**Soilless Culture, Growing Media and Horticultural Plants**

Editors

**Nazim S. Gruda Brian Eugene Jackson**

Basel • Beijing • Wuhan • Barcelona • Belgrade • Novi Sad • Cluj • Manchester

*Editors* Nazim S. Gruda University of Bonn Bonn, Germany

Brian Eugene Jackson NC State University Raleigh, NC, USA

*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland

This is a reprint of articles from the Special Issue published online in the open access journal *Agronomy* (ISSN 2073-4395) (available at: https://www.mdpi.com/journal/agronomy/special issues/Soilless Media Horticultural).

For citation purposes, cite each article independently as indicated on the article page online and as indicated below:

Lastname, A.A.; Lastname, B.B. Article Title. *Journal Name* **Year**, *Volume Number*, Page Range.

**ISBN 978-3-0365-8760-8 (Hbk) ISBN 978-3-0365-8761-5 (PDF) doi.org/10.3390/books978-3-0365-8761-5**

Cover image courtesy of Nazim S. Gruda

© 2023 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license. The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons Attribution-NonCommercial-NoDerivs (CC BY-NC-ND) license.

## **Contents**


Different Growth Stages Reprinted from: *Agronomy* **2020**, *10*, 139, doi:10.3390/agronomy10010139 .............. **131**



## **About the Editors**

## **Nazim S. Gruda**

Nazim S. Gruda, Professor of Horticulture at the University of Bonn in Germany, vice-chair of the ISHS-Division Vegetables, Roots, and Tubers, and chair of one ISHS-Working Group. Professor Gruda is an internationally recognized leading authority in soilless culture, growing media, and controlled environment cultivation. He has been working as a successful researcher, research advisor, project expert and evaluator, member of steering committees and editorial boards, editor, and scholar for more than three decades. He obtained his Doctorate at the Technical University of Munich and Habilitation at the Humboldt University of Berlin, both in Germany. His research focused on the scientific understanding and application of innovative and sustainable horticultural food production. He has achieved an impressive track record of over 300 publications. He has co-edited four Acta Horticulturae for the ISHS. Dr. Gruda is one of the authors and the editor of the successful book "Advances in Horticultural Soilless Culture", from Burleigh Dodds Science Publishing Limited, Cambridge, UK, in 2021. According to the evaluation carried out by Stanford University over the last three years, Prof. Gruda is in the top 2% of the most influential scientists in the world ("World's Top 2% Scientists"). In recognition of his excellent research, Professor Gruda was awarded the "Dr Heinrich-Baur-Prize" 2003 by the Technical University of Munich, Germany, the "National Scientific Prize" 2018 by the Academy of Science of Albania, and was elected as "Distinguished Scientist" 2020 by the Academy of Science of China. In addition, Professor Gruda is a full member, a foreign correspondent member, and an honorary member of three European Academies of Sciences.

### **Brian Eugene Jackson**

Brian Eugene Jackson is a Professor and Director of the Horticultural Substrates Laboratory in the Department of Horticultural Science at North Carolina State University. Brian earned both his graduate degrees (M.S and Ph.D.) in soilless substrates and has continued as a soilless substrate scientist for the past 15 years at NC State University. Brian's major research focus and international expertise are in the engineering and development of alternative substrates in soilless growing systems, primarily wood fiber, and other forest and agricultural-based biomass materials. He has advised or co-advised 5 Ph.D. and 14 M.S. students and has served on the committees of 18 others nationally and internationally. Brian has developed a global reputation as one of the leading voices, educators, researchers, and advocates for the soilless substrate industry. Brian has traveled to over 50 countries to research, collaborate, or consult on soilless substrate-related problems and innovations in the past decade. At home, his career efforts and successes have been recognized by his receipt of the 2020 Outstanding Undergraduate Educator Career Award and the 2021 Outstanding International Horticulturalist Career Award from the American Society of Horticultural Science.

## *Editorial* **Advances in Soilless Culture and Growing Media in Today's Horticulture—An Editorial**

**Nazim S. Gruda**

Division of Horticultural Sciences, Institute of Crop Science and Resource Conservation, University of Bonn, Auf dem Hügel 6, 53121 Bonn, Germany; ngruda@uni-bonn.de

**Abstract:** The soilless culture system is a promising, intensive, and sustainable approach with various advantages for plant production. The Special Issue "Soilless Culture, Growing Media, and Horticultural Plants" includes 22 original papers and 1 review written by 84 authors from 15 countries. The purpose of this Special Issue was to publish high-quality research articles that address the recent developments in the cultivation of horticultural plants in soilless culture systems and solid growing media. The published articles investigated new developments in simplified and advanced systems; the interaction between soilless and environmental factors with their effects on plant growth and photosynthesis, and the accumulation of secondary metabolites; the analyses of nutrient solution and hydraulic properties of substrates and mixtures; and the microbe–plant growing media interactions. Climate change and environmental and ecological issues will determine and drive the development of soilless culture systems and the choice of growing media constituents in the near future. Bioresources and renewable raw materials have great potential for use as growing medium constituents.

#### **1. Introduction**

Decreasing arable land, rising urbanisation, water scarcity, and climate change have placed pressure on agricultural producers [1]. The soilless culture system (SCS) is a promising approach with different advantages for plant production. As an intensive and sustainable cultivation method, SCSs have rapidly expanded worldwide, particularly in areas close to cities or with a shortage of water supply, poor soil quality, and problems with soil-borne diseases and salinity. These systems produce pot ornamentals, seedlings, and transplants and increase plant metabolites in fruits, vegetables, and medicinal and aromatic plants. Production technology affects plant growth, yield, and overall plant quality, which, in turn, improves the cumulative benefits of plants [2–4].

Horticultural crops, such as vegetables, floral crops, ornamentals, and fruits, have become essential components of aesthetics and nutrition in our daily life. Currently, SCSs have received significant interest and are used for the intensive production of vegetables, floral crops, ornamentals, green roofs, and rain gardens [2,3,5]. Furthermore, because of their lightweight and sustainable resource efficiency, soilless systems are especially suitable for urban areas, including green infrastructure projects and vertical farming [6]. The increased worldwide production of crops in controlled environmental systems has been further accelerated by the increased interest in growing small/soft fruit crops, greens, herbs, and cannabis in soilless container systems. In addition, they are used to increase metabolites in medicinal and aromatic plants and to introduce new crops [2,3]. As a result, the demand for SCS and growing media continues to increase worldwide, as does the need for novel research to address problems and continue creating opportunities for this industry [4].

The purpose of this Special Issue was to publish high-quality research articles that address the recent developments in the cultivation of horticultural plants using SCSs with or without solid growing media. It aims to provide contributions from various currently relevant topics in horticultural sciences, physiology, root medium properties,

1

**Citation:** Gruda, N.S. Advances in Soilless Culture and Growing Media in Today's Horticulture—An Editorial. *Agronomy* **2022**, *12*, 2773. https://doi.org/10.3390/ agronomy12112773

Received: 2 November 2022 Accepted: 4 November 2022 Published: 7 November 2022

**Copyright:** © 2022 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

plant propagation, plant nutrition and chemistry, substrate hydrology and physics, compost and waste management, engineering, and all other research fields familiar with soilless culture and growing media.

The Special Issue "Soilless Culture, Growing Media, and Horticultural Plants" includes 22 original papers and 1 review written by 84 authors from 15 countries. Considering this is just the tip of the iceberg, the remaining papers were rejected during the published review process, showing the great interest in this Special Issue from the scientific community. Writing an editorial after several years allowed us to analyse the papers' importance. Following citations from the publication date until the end of October 2022, papers from this Special Issue were cited 363 times, with an average of 16.5 times per paper, which is relatively high. The review article [1] received the highest number of citations (126) among all the published papers. It should be mentioned that this article received the second-best paper award on the tenth anniversary of the journal, while the Web of Science-Clarivate lists it as a highly cited paper (1% of all papers included in the database). The article from Dou et al. [7] received the highest number of citations among the research papers, with 41 citations.

#### **2. Soilless Culture Systems**

Soilless culture is a modern cultivation technology applied mainly in greenhouses, which has developed rapidly during the last 30–40 years [5]. Most SCS plants are grown in high-tech greenhouse structures with fully automatic climate control features [2,3]. This Special Issue focused on new developments in simplified, advanced, and complex SCSs. For instance, Bentrary et al. [8] and Michelon et al. [9] investigated the feasibility of a low-tech SCS for cultivating *Pelargonium zonale* and *Lactuca sativa*, respectively. As a result, the yield for lettuce cultivation in tropical areas was improved by +35% and +72% in Brazil and Myanmar, respectively, and the water-use efficiency (WUE) was 7.7 and 2.7 times higher in Brazil and Myanmar, respectively, compared to traditional on-soil cultivation [9]. The soilless system typology can also significantly affect the rooted cutting growth, commercial features, and WUE. For example, adopting an open-cycle drip system significantly improved all commercial crop characteristics of geranium (*Pelargonium zonale*) compared to a substrate and a nutrient film technique system. The water consumption of this treatment system was higher than that of the other systems. However, it induced the highest fresh weight and, therefore, the highest WUE [8].

Given its flexibility in manipulating the nutrient status and efficient utilisation of nutrient components, SCSs could be used as an efficient tool for producing high-value vegetables and herbs and crucial root vegetables in temperate and tropical zones, such as sweetpotatoes (*Ipomoea batatas*) [10].

In recent years, research on soilless culture has mainly focused on the automation of nutrient and water supply, particularly in closed systems [5]. A closed-loop SCS is an environment-friendly cultivation method. However, variations in nutrients can lead to instability in nutrient management. Ahn and Son [11] analysed nutrient variation in a closed-loop SCS based on a theoretical model and found fluctuations around the target value. However, the total nutrient concentration did not continuously deviate from the target value in the conventional method and showed a tendency to increase. Therefore, the authors concluded that these characteristics of the alternative method could help minimise nutrient and water emissions from the cultivation system.

Theoretical and experimental analyses of nutrient solutions, variations in electrical conductivity, fertiliser selection, and nutrient solution replenishment methods have been discussed in the papers published in this Special Issue. The fertiliser used in the SCS should contain balanced elements without any precipitates [12]. For sweet pepper yields, the commercial fertiliser 5N-4.8P-21.6K was responsible for the highest yield of both cultivars, 'Bentley' and 'Orangella'. Fertilisers and cultivars did not affect the shape index. For eggplant, the shoot fresh weight was greater for 'Angela' than for 'Jaylo' at 5N-4.8P-21.6K and 7N-3.9P-4.1K. Furthermore, both eggplant cultivars were affected by yellowing fruits

for all the fertiliser treatments after two months, probably due to the accumulation of nutrients in the closed hydroponic system [12].

#### **3. The Interaction of SCS with Environmental Greenhouse Factors**

The effects of the interaction of soilless culture and different environmental greenhouse factors, such as supplemental lighting intensity, UV radiation, and CO2 enrichment, on biomass accumulation, gas exchange properties, and plant quality are addressed in this Special Issue. For instance, Llewellyn et al. [13] found that increasing levels of supplemental light had only minor effects on vegetative growth (young plants) and the size and quality of harvested flowers (mature plants). However, cut gerbera (*Gerbera jamesonii*) plants grown under higher light intensity produced 10.3 and 7.0 more total and marketable flowers per plant than the lowest light intensity and matured faster [13].

One other factor is the CO2 concentration in the air. According to Li et al. [14]), the accumulation of cucumber biomass can be significantly increased by elevated CO2 concentrations and high N supply. In addition, a high N supply can further improve photosynthesis. The authors concluded that if we had a greater understanding of the mechanisms that control mineral concentration changes in cucumber plants in response to elevated CO2, mineral fertilisation could be optimised to improve the growth of plants under elevated CO2 conditions. Thus, sustainable vegetable production with higher C and N use efficiency and lower CO2 emissions and fertiliser input could be achieved [14].

#### **4. SCS and Produce Quality**

Using SCSs to control nutrients, the temperature in the root area, and managing environmental and agronomic factors can improve product quality [1,15]. This Special Issue investigated the effects on plant photosynthesis and growth, the accumulation of secondary metabolites, and seasonal antioxidant changes. For instance, Neocleous et al. [16] indicated that lower solar irradiance, ultraviolet radiation, and temperature in Mediterranean greenhouses could be insufficient to stimulate phytochemical production during late autumn and winter in peppers. Thus, plant light interception must be more actively managed. Furthermore, Ellenberger et al. [17] investigated how stress affects the content of secondary metabolites in leaf bell papers. Therefore, high UV stress should be considered a tool for enriching plant leaves with valuable secondary metabolites.

The absence of ultraviolet (UV) radiation and low photosynthetic photon flux density (PPFD) in a controlled environment reduced the phenolic compounds in herbs. Dou et al. [7] investigated green and purple basil to characterise the optimal UV-B radiation dose and PPFD for enhancing the synthesis of phenolic compounds in basil plants (*Ocimum basilicum*). Plants were grown at two PPFDs, 160 and 224 <sup>μ</sup>mol·m−2·s−1, and treated with five UV-B radiation doses. In purple basil plants, the concentrations of phenolics and flavonoids increased after 2 h·d−<sup>1</sup> of UV-B treatment. Among all treatments, 1 h·d−<sup>1</sup> for 2 d of UV-B radiation under a PPFD of 224 <sup>μ</sup>mol·m−2·s−<sup>1</sup> was the optimal condition for green basil production in a controlled environment [7].

Interestingly, Giménez et al. [18] found that compost in growing media boosted the product's final quality, with a higher total phenolic content and antioxidant capacity in the leaves of baby leaf lettuce in a floating system, particularly during the second cut.

#### **5. Growing Media and the Diversity of Inorganic and Organic Substrates**

In SCSs, solid inorganic or organic substrates are used for plant cultivation, usually in containers. Therefore, studies submitted to this Special Issue have investigated the physicochemical and hydraulic properties of organic and mineral substrates and mixtures and the substrate volumetric water content to improve water-use efficiency in growing media. Furthermore, the chemical properties and the microbe–plant growing media interactions were investigated.

According to Gohardoust et al. [19], an essential first step towards developing advanced soilless culture management strategies is the comprehensive characterisation of

the growing media's hydraulic and physicochemical properties. These parameters can be applied to the engineering of growing media by mixing organic and inorganic constituents at different ratios to meet specific plant physiological demands. Furthermore, these results could also be used to visualise three-dimensional numerical computer codes to simulate water and nutrient dynamics in containerised growth modules.

Moreover, Currey et al. [20] found that the growth of basil, dill, parsley, and sage can be affected by the water supply, with no signs of stress or visual damage resulting from the reduced volumetric water content of the substrate. Therefore, restricting irrigation and substrate volumetric water content is an effective non-chemical growth control method for containerised culinary herbs.

Bacterial enhancement has a significant potential to modulate plant performance in horticultural systems. However, the effectiveness of bacterial amendment regarding plant performance depends on the bacterial source and its interaction with the growth medium. Therefore, an appropriate selection of the plant growth medium composition is critical for the efficacy of bacterial amendments and optimal plant performance in a plant factory with artificial lighting [21].

#### **6. Peat Alternatives in Growing Media Mixtures**

Peat is the most commonly used substrate constituent in horticulture. However, the use of peat in horticulture has been strongly criticised because of environmental and climate change concerns [1–3]. Therefore, new peat additives and/or peat alternative growing media, such as biochar, green compost, olive oil-processing waste composts, and vermicompost, were investigated in the Special Issue. In addition, the raw materials used as growing media constituents should be free from phytotoxic compounds [22] and should demonstrate good chemical properties, such as a suitable pH [23,24] and the content of certain elements and/or salt content [18,25].

Composts from different raw materials, such as vineyard waste, tomato waste, leek waste, and olive mill cake, can be alternatives to peat in producing baby leafy vegetables in a floating system. The use of 25% compost as a component of the growing media in the production of baby leafy vegetables in a floating system not only favours crop yield and product quality, but also suppresses *Pythium irregulare* [18].

Tüzel et al. [26] found that compost obtained from two-phase and three-phase olive mill solid wastes and olive oil wastewater sludge that can be used in a ratio of 25% in mixtures with peat was appropriate for most of the measured tomato seedling properties.

Moreover, biochar has been proposed as a soil amendment and a growing medium component that positively affects plant growth and yield [24]. Chrysargyris et al. [25] investigated four types of commercial-grade biochar from wood-based materials used in mixtures with peat for cabbage seedling production. Biochar material had a high K content and a pH ≥ 8.64, which increased the growing media's pH. In addition, the leachate pH of all biochar mixes was higher than that of the control [27]. Potassium, phosphorous, copper accumulation and magnesium deficiency in cabbage leaves were related to the presence of biochar. Therefore, wooden biochar from beech, spruce, and pine species and fertilised biochar from fruit trees and hedges is promising for cabbage seedling production [25].

While recent studies on biochar mentioned the importance of the feedstock used, Prasad et al. [24] stated for the first time the need for information on particle size because the fractions from the same biochar can have different levels of total extractable nutrients and pH levels. Particle size could have a profound effect on the nutrient availability of Ca and Mg. This could lead to nutrient imbalances during the cultivation of plants on substrate mixtures. In addition to nutrient ratios, a suitable pH level for a given species should be achieved [24].

Mixes with 80% biochar and vermicompost had lower container capacities than the control. Nevertheless, plants in the BC mixes had similar growth indices and total dry weights concerning those in 100% commercial substrate. Therefore, BC mixed with vermiculite has the potential to replace commercial peat-based substrates for container-grown plants [27].

Yu et al. [28] conducted a greenhouse experiment to evaluate the potential of replacing mixed hardwood biochar with sugarcane bagasse. Both tomato and basil plants grown in biochar-incorporated mixes had a similar or higher growth index, leaf greenness, and yield than bark-based commercial substrates. The authors concluded that hardwood and sugarcane bagasse biochar could replace 50% and 70% of bark-based substrates for tomato and basil plants without adverse growth effects [28].

#### **7. Concluding Remarks and Future Trends**

The articles published in this Special Issue stated that climate change and environmental and ecological issues would soon determine and drive the development of soilless cultural systems and the choice of growing media constituents. It is clear that while much has been achieved in this Special Issue, many challenges remain. Understanding the optimisation of root-zone conditions [29] and clarifying the mechanism of interaction between roots and surroundings will contribute to a better understanding of SCS. Advances in soilless culture will be supported by findings from other scientific fields that will contribute to the further development of soilless cultures. In addition, bioresources and renewable raw materials have great potential for use as growing media constituents. We expect these publications to promote further discussion about these two exciting topics.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The author declares no conflict of interest.

#### **References**


## **Increasing Sustainability of Growing Media Constituents and Stand-Alone Substrates in Soilless Culture Systems**

## **Nazim S. Gruda**

Department of Horticultural Science, INRES-Institute of Crop Science and Resource Conservation, University of Bonn, 53121 Bonn, Germany; ngruda@uni-bonn.de

Received: 2 May 2019; Accepted: 5 June 2019; Published: 9 June 2019

**Abstract:** Decreasing arable land, rising urbanization, water scarcity, and climate change exert pressure on agricultural producers. Moving from soil to soilless culture systems can improve water use efficiency, especially in closed-loop systems with a recirculating water/nutrient solution that recaptures the drain water for reuse. However, the question of alternative materials to peat and rockwool, as horticultural substrates, has become increasingly important, due to the despoiling of ecologically important peat bog areas and a pervasive waste problem. In this paper, we provide a comprehensive critical review of current developments in soilless culture, growing media, and future options of using different materials other than peat and rockwool. Apart from growing media properties and their performance from the point of view of plant production, economic and environmental factors are also important. Climate change, CO2 emissions, and other ecological issues will determine and drive the development of soilless culture systems and the choice of growing media in the near future. Bioresources, e.g., treated and untreated waste, as well as renewable raw materials, have great potential to be used as growing media constituents and stand-alone substrates. A waste management strategy aimed at reducing, reusing, and recycling should be further and stronger applied in soilless culture systems. We concluded that the growing media of the future must be available, affordable, and sustainable and meet both quality and environmental requirements from growers and society, respectively.

**Keywords:** biochar; compost; climate change; hydroponics; growing medium; life cycle analysis; organic bioresources; peat alternatives; renewable raw materials; rockwool; waste; wood fibers

## **1. Introduction**

According to the United Nations, the current world population of 7.79 billion people will increase to 9.77 billion people by 2050 [1], while the arable land per capita continues to be reduced. This development is following the same pattern in all countries, although the rate varies between countries. For instance, in North America there were 1.06 ha, and in the European Union 0.32 ha, per person available in the year 1961, while in 2015 only 0.55 ha and 0.21 ha per person, respectively. This is nearly to 2× and more than 1.5× reduction for North America and the European Union, respectively (Figure 1) [2].

**Figure 1.** The arable land per person has been continuously reduced in the recent past. Arable land in hectares per person from 1961–2015 for North America, the European Union, and worldwide, according to World Bank [2].

In addition, worldwide urbanization is increasing rapidly. In 2008, the global urban population overtook the rural population for the first time in history. Today, over 50% of the world's population lives in cities; by 2030, this number is projected to increase to 70% [3].

Future climate change scenarios predict more frequent occurrence of extreme conditions, such as drought years and the uneven distribution of precipitation during the year [4]. The possible increase in water shortage and extreme weather events may cause lower yields and higher yield fluctuations [5]. These disadvantages will be predominately in warmer regions worldwide. Therefore, besides securing sufficient water, it will become increasingly important to improve the use efficiency of this resource [6–8]. Water, as a valuable resource, can be used more efficiently in protected vegetable production, which is considered less dependent on weather conditions than open field production, because micro-climates can be manipulated [6,7].

Decreasing arable land, rising urbanization, water scarcity, and climate change exert pressure upon agricultural producers. One of the most promising approaches to tackle this challenge is termed "sustainable intensification", which tries to combine increased production without damaging its supporting ecosystem. Examples for this approach are protected, soilless culture systems (SCS) [9]. "Soilless culture" is defined as the cultivation of plants in systems without soil in situ [10]. The percentage of SCS to the total commercial horticultural protected cultivation area varies from country to country. For instance, in the Netherlands and Almeria, Spain, soilless culture represents the main cultivation system used [11]. In Europe, Canada, and in the large horticultural industry complexes in the US, 95% of greenhouse tomatoes are produced in SCS [12,13].

Growing media, "substrates" or "plant substrates" provide a root environment that is initially free of plant pathogens and properties that ensure an adequate aeration, water, and nutrient supply. In the horticultural industry, generally, mixtures of growing media constituents and additives are used. Organic or inorganic materials can be used as constituents, while additives include fertilizers, liming materials, and bio-control or wetting agents [14–16].

Blok and Urrestarazu [17] estimated an area of more than 10,000 ha cultivated in rockwool slabs worldwide, including 6000 ha greenhouse area in Europe, mainly in Northern Europe. Rockwool has a low volume weight, is inert, and has a buffering capacity, limited to the quantity of nutrients and water held within the pore space of the medium [18]. To feed the plant with water and fertilizer a complete nutrient solution is supplied through the irrigation system (Figure 2).

**Figure 2.** Tomato production in soilless culture with rockwool as a growing medium: (**a**) Transplants in rockwool cubes, shortly before the start of greenhouse cultivation; (**b**) tomato plants in rockwool slabs (photos: Gruda, private collection).

However, it is important to note that the disposal problem for mineral wool has led to criticism of its current usage. Some authors, such as Bussell and Mckennie [19], showed some options to reuse rockwool, but when analyzing the life cycle assessment of horticultural growing media, Quantis [20] reported that mineral wool has the highest negative impacts on human health. In addition, freight costs are relatively high.

Besides rockwool, other inorganic substrates, such as perlite, volcanic rock, tuff, expanded clay granules, vermiculite, zeolite, pumice, sand, and synthetic materials could be used directly or in combination with other materials as a growing medium.

Of all organic materials, peat is the most used substrate constituent in horticulture [7]. The leading peat-production countries are Finland, Ireland, Germany, Sweden, Belarus, Canada, and Russia, which account for 80% of the world's production. Commercial applications include lawn and garden soil amendments, potting soils, and turf maintenance on golf courses [21]. The extensive use of peat as a basic and main component of substrates is due to relatively low costs in these areas, its excellent chemical, biological, and physical properties with low nutrient content, low pH, a unique combination of high water-holding capacity by high air space and drainage characteristics, light weight, and freedom from pests and diseases [14,16,21]. The unique microporous properties of *Sphagnum* peat and its resistance to degradation are matched by few other growing medium constituents [22].

However, peat is a limited resource with a great demand, and the extraction of peat bogs causes negative impacts on environment. Peatlands are areas with a layer of dead plant materials (peat) at the surface. The water-saturated and oxygen-free conditions prevent peat from fully decomposing. Peatlands are a habitat with special ecological value with the most important long-term carbon sinks and one of the most effective eco-systems in the terrestrial biosphere, providing different environmental services, such as biodiversity, carbon (C) storage, regulation of the local water quality, and local hydrology conditions, including flood protection [23–25]. Covering only about 3% of Earth's land area, they may store 21% [26] to 33% [27] of the total world's terrestrial organic carbon. In the long-term, peatlands are the largest stores of organic carbon of all terrestrial ecosystems [28]. However, when peat bogs are drained or destroyed, i.e., used in agriculture, forestry, and/or horticulture, they no longer act as carbon sinks. Degraded peatlands contribute disproportionally to global greenhouse gas emissions, with approximately 25% of all CO2 emissions from the land use sector [29]. Annual emissions equivalent of 15 million tons of carbon are estimated [23,24,30,31]. In addition, the renewal process of peatlands takes a very long time, and in arid areas peat is imported, with an impact both in environmental and economic terms. Therefore, Quantis [20] indicates that peat has the highest impact on "climate change" and "resources" of all commonly-used substrate materials.

Recently, the energy use and carbon emissions in horticultural production systems have moved into the public spotlight. Thus, retailers increased the pressure and are now requiring not only traceable healthy and safe horticultural products, but also "clean and green" produce with a low carbon footprint. On the other hand, due to limited natural resources and waste recycling issues, environmentally acceptable solutions are needed for materials used as growing media constituents.

The objective of this paper is to critically review and expand the knowledge of impacts of soilless culture and growing media on the environment, targeting an improvement of sustainability of all horticultural systems. First, an overview on the pros and cons of soilless culture and growing media use is provided. Second, different important economic and environmental factors are analyzed. Moreover, different organic materials are explored with the objective to recognize successful alternatives for peat and rockwool.

#### **2. Results and Discussion**

#### *2.1. Soilless Culture and Growing Media: Pros and Cons*

Soilless culture systems are commonly integrated in controlled environment agriculture, i.e., heated greenhouses, that in turn are associated with environmental concerns and the production of high amounts of greenhouse gases (GHGs). Indeed, major studies conducted showed that from an environmental point of view, plants cultivated directly in soil in tunnels or greenhouses without using auxiliary systems perform better than those with heating in SCS [32–34]. However, even if the heated protected cultivation systems present a good opportunity to move from soil to SCS, we do not have to associate SCS only with heated greenhouses. The specific features along the entire production system in these structures include the large amount of energy consumption for heating during the cold season, artificial lighting, the greenhouse structure itself [35], the use of fertilizer and growing media [7], postharvest transport, and packaging [36]. The equipment of SCS contribute to some degree to an increase of the energy needed together with growing media used in these systems. But, on the other hand, SCS contributes to a reduction of many problems associated with traditional cultivation on soil in situ, such as soil-borne diseases and pests, and to an exact control of water and fertilizer supplies. As a consequence, higher yields at a reasonable production cost and high product quality can be attained in these systems [13,37]. Recently, the greenhouses production is increasingly carried out with machines as an "unmanned working model" in some soilless systems [38].

Moreover, high precision in modulating nutrient solution composition, the exact dosage and controlled exposure, make SCS a good instrument to predict the product supply and enhance the organoleptic plant parameters and bioactive quality components. Moderate salinity and/or nutritional stress and the biofortification of vegetables with beneficial micronutrients to human health, such as iodine, iron, molybdenum, selenium, silicon, and zinc are well known methods that have been successfully used to enhance the health-promoting phytochemicals in vegetables [13,39–42].

Therefore, in general, growing plants in soilless media is a sustainable production manner. This is due to the inherent space, nutrient, and water use efficiencies of this production method; all of which are higher than soil-grown crops [9]. At present, life cycle analysis (LCA) is used for the classification of growing media constituents, based on their environmental impact and sustainability, environmental protection, and the application of "green technologies" for their production [7,16]. Mugnozza et al. [43] determined, using LCA, that soilless cultivation reduced the environmental impact by more than double, due to lower levels of fertilizers and pesticides emitted to the environment, compared to soil cultivation. The total GHG emissions from a tomato rockwool culture averaged 853 g (exp. 1) and 999 g CO2 equivalent (exp. 2), and from a soil-based production averaged 1303 g (exp. 1) and 1509 g CO2 equivalent (exp. 2), respectively. In addition, 16S ribosomal ribonucleic acid gene abundance in soil samples was 10-fold higher than in rockwool samples [44].

Every year, the majority of freshwater, approx. 87%, is used worldwide for agricultural production [45]. The lack of freshwater resources is an acute issue for arid and semiarid areas in Africa, the Middle East, Southern Europe, and South America that may not only threaten economic development, but also lead to drastic environmental and social problems. One of the major advantages of using SCS is water economization. For instance, lettuce nutrient film technique (NFT) production in South-East Spain requires 62% less water than soil cultivation [46]. In this context, sometimes a comparison between local and imported products is discussed. Stoessel et al. [47] studied a wide range of vegetables, including tomatoes, and concluded that, from a carbon footprint viewpoint, it is often better to import vegetables produced in warm Southern countries during periods when Northern production requires heating. However, surprisingly, sometimes LCA studies, e.g., for tomato production in different Mediterranean countries, have been carried out without considering the impacts of freshwater use [48–51]. Webb et al. [52] also did not address the impacts of freshwater use in their comparison of locally produced tomatoes in the UK and imported tomatoes from Spain [53].

Tomato is the most important vegetable crop in the world [54] and the most cultivated in SCS. When comparing water consumption and water use efficiency (WUE), defined as the obtained yield per unit of irrigation water, vast improvements in WUE are made, with varying degrees, when moving from traditional, soil-based production to protected SCS cultivation methods (Figure 3). For instance, for one kilogram of tomatoes produced in the field, on average about 200 ± 100 L of water are needed. Using drip irrigation, this amount is reduced to about 60 L/kg [55,56]. Moving from soil to SCS can further improve WUE.

**Figure 3.** Applying new techniques and new irrigation systems can significantly improve water use efficiency, here calculated as L/kg tomatoes, using different growing systems. Soilless culture system (SCS).

The SCS could be either open-loop or closed-loop cultivation systems. The latter, which involves re-using any drainage solution, can substantially reduce potential pollution of water resources by nitrates and phosphates, while contributing to an appreciable reduction in water and fertilizer consumption [10], even if an ion accumulation (Na<sup>+</sup> and Cl−) is a challenge in these systems [57]. Comparing data from a commercial tomato farm in Italy and referring to one summer growing season, the savings from a closed irrigation system were 25%, 40%, 24%, and 11% in water, nitrogen (N), phosphorus (P), and potassium (K), respectively [58]. In an open system, where drainage water is not captured and recycled, 10–20% water and fertilizers can be saved, while production and quality can be improved [59]. However, in closed-loop or recirculating water systems that recapture the drainage water for reuse [13], the average use is between 14 and 20 L/kg, i.e., reduced by a factor of up

to 5–10 [51,55,60] (Figure 3). By combining a modern irrigation system with modern environmental management, such as the use of closed/semi-closed greenhouses [8] with the regaining and reusing of condensed evaporated water [55], the use of light selective shading and evaporative cooling systems [60,61] make more water savings possible. To come back to our example regarding tomatoes, according to van Kooten et al. [55], it is possible to achieve WUE of 1.5 L water per kg tomato. Under practical conditions, the levels of WUE are rather higher than this. However, these values are possible to achieve and every reduction in water consumption is a step in the right direction. Moreover, under the expected climate change scenarios and water limitation for agriculture, desalinated seawater coupled with hydroponic systems could be a valuable strategy to sustain a high productive agriculture [46].

WUE has a direct economic and environmental effect [8]. Apart from WUE, growing crops with high water requirements in water-scarce areas has important implications. Payen et al. [51] compared the production of tomatoes in Morocco with a production in France. They found that, although the water use efficiency was similar, Moroccan tomato freshwater deprivation was almost four times higher, with 28.0 L H2O eq kg<sup>−</sup><sup>1</sup> for Moroccan tomatoes and 7.5 L H2O eq kg<sup>−</sup><sup>1</sup> for French tomatoes. This was explained by the high-water stress index of the cultivation area. Therefore, the authors concluded that, from a water perspective, sourcing vegetables from water-scarce countries is questionable [51].

Because of their light weight and sustainability in terms of resource efficiency, soilless systems are especially suitable for urban areas as well as hobby gardening. These systems allow for an exact dosage and application of nutrients [3,62]. Nowadays, "vertical farming systems" in tower shapes have started to be applied. This system provides 10x more plants per unit area, a 50% reduction in the harvest period, water and fertilizer savings, clean production, and all year-round production [38].

The major disadvantages of soilless cultures are the high investment and energy costs that are required for the initial installation, as well as the increased technical skills that are needed. Other advantages and disadvantages by using certain organic materials as growing media constituents and stand-alone substrates are analyzed below.

#### *2.2. Organic Materials Other than Peat Used as Growing Media Constituents*

Different organic materials may play an important role in decreasing the C footprint of the horticultural industry by fully or partly replacing peat-based substrates. Compost, coir, bark, and wood fiber are some organic materials that are already being used in a commercial way as an alternative to peat [23]. In addition, some inorganic materials, such as vermiculite, perlite, clay granules, lava, and pumice are used instead of rockwool or in mixture with peat and other combinations, while new organic materials, such as *Sphagnum* moss, waste and digestates, biochars, and hydrochars are still in their test phase. Below, some of these organic materials and bioresources are briefly described.

#### 2.2.1. Compost, as a Bioresource and Growing Media Constituent

Compost is a general term, describing all organic matter that has undergone thermophilic, aerobic decomposition. It represents a bioresource and a sustainable use case for a potential waste material [9,63]. Several materials are used as growth media after adequate composting. Abad et al. [64] created a database with 105 materials suitable for use as growing media for ornamental potted plant production in Spain. The authors differentiate between urban, sea, agricultural, forest, animal, industrial, and food waste. The disposal needs for waste materials is already an environmental problem, and their recycling in the form of potting media provides a suitable solution. However, some of these materials cannot be used directly. They either contain pathogens, are not stable, or have high water [65] or nutrients content. In some cases, the legal basis needs also be clarified.

Table 1 presents several waste materials used for compost production, which, afterwards, alone or in mixture with other materials, can be used as plant substrates. These include urban and municipal solid wastes, animal manure, grape marc, olive mill, and other food processing wastes; bark, sewage sludge, paper waste, greenhouse waste, pruning waste, spent mushroom compost, and green waste. Different nursery, ornamental, and vegetable plants can grow into these substrates (Table 1). Materials

such as bark, wood, several shells or hulls, and coconut coir possess good physical properties after composting. However, being relatively resistant to decomposition, these materials should be subjected to long and well-controlled composting, which may be shortened using N and N-rich organic matter, such as animal manures [66]. According to Raviv [66], high temperatures may cause ashing of these materials, which leads to reduced porosity and increased bulk density. Therefore, temperatures above 65 ◦C are not desirable.


**Table 1.** Waste materials used for compost, which, in turn, is used as a plant substrate on its own or in a mixture with other materials.

Superscripted reference numbers (e.g., 1, 2, 3, 4, 5) link feedstock waste and growing media with the corresponding literature, applicable only within rows, not columns.

Some value-added benefits have to be highlighted here. These are based on specific properties, such as the potential to suppress some diseases and the capacity to control some plant pathogens. Biofertilization and biostimulation could be mentioned as well. However, composts are variable with respect to physical, chemical, and biological properties. Volume weight, air space, water retention, pH, and available plant nutrient elements can vary greatly from batch to batch as well as with the degree of microbiological degradation and primary organic material used. Even within the different green composts there are differences concerning the quality of the compost. For instance, only the use of selected raw material from greenhouse vegetables, nursery shrubs, and green wastes, i.e., plant trimmings, prunings, and crop residues, could contribute to the production of high-quality compost [87]. The selected green compost was found to be a valuable growing medium for peat substitution, while

the green compost derived from mixed raw material negatively influenced *Pelargonium* plant nutrition and photosynthesis, thus significantly reducing plant biomass accumulation and quality. Raw material selection increases the production costs of compost. Therefore, according to Massa et al. [87], efforts should involve the adaptation of new technologies for tracking raw materials and supporting sustainable circular chains for compost production at a local level. In addition, strict quality control procedures are essential in preparing composts for use in growing media [22].

Composts produced from so-called green materials, such as prunings, shredded branches, plant debris, and waste from gardens and nurseries, are widely used as components of growing media in the Netherlands, the United Kingdom, Italy, and Germany, primarily in media for the hobby market [22]. However, they can be used as a component of a growing medium up to 50%, but not as stand-alone substrates [88]. The limiting factor regarding the use of composted green waste is its high electrical conductivity (EC) and potassium (K) concentration. There can also be a problem of plant pathogens, human pathogens, and weed contamination if the composting process is not properly conducted, i.e., if the temperature time exposure is not sufficient [14]. Moreover, compost has a low (5–10%) carbon efficiency, which is reflected in material mass and volume reduction and a relatively high pH.

The use of waste as composting material with a further use as growing media and/or growing media constituents is of a dual benefit. For instance, the removal and disposal of large volumes of plant biomasses of *Posidonia*, a marine phanerogam endemic of the Mediterranean Sea, represent, on one hand, a high cost for local administrations [79]. On the other hand, posidonia-based compost, produced from posidonia residues, may have a considerable potential as a peat substitute in horticultural substrates. Several studies evidenced its use for production of tomatoes [78], lettuce transplants [79], melon and tomato seedlings [80], pot basil [81], and pot sea fennel [82].

The same is true for mushroom substrates. Over three million tons of spent mushroom substrates are produced in Europe every year as a by-product of the cultivation of *Agaricus bisporus* [89]. Due to its physical properties and nutrient content, spent mushroom substrate has great potential to be employed as a growing medium in horticulture. However, spent mushroom substrate should be first matured and stabilized through a composting system [89] before being used, e.g., for vegetable production (Figure 4).

(**a**) (**b**)

**Figure 4.** Spent mushroom substrate used as growing media in simple soilless culture systems (SCS) in Shandong province in China. (**a**) Spent mushroom substrate. Mushroom production is usually placed in the North part of the greenhouse. (**b**) Tomato production in simple SCS in the South part of the greenhouse. Here, the spent mushroom substrate is utilized as a growing medium (Photos: Gruda, private collection).

Compost, when mixed into growing media, is a source of fiber, i.e., a rooting medium, as well as an important source of nitrogen (N), phosphorus (P), and potassium (K). Therefore, the substrate mixtures containing compost required adjusted fertigation due to nutrients supplied by the compost [90]. In addition, the degree of infection with powdery mildew and aphids was strongly positively correlated with the N status of the crop, pointing at the risks of high N supply for the crop [90].

As an alternative to conventional composting, the action of worms and their gut microorganisms can be used to break down organic waste materials to produce vermicompost. Particle-size distribution and fertility were superior in the vermicompost-based media than in the conventional compost-based media. The compost-based media showed an approx. 2.2× higher coarseness index than the vermicompost medium that possessed more fine particles as compost, due to the effect of earthworms [91]. Earthwoms increase the quantity of small particles by ingesting, mixing, grinding, and then egesting organic material [92]. In addition, the nutrient level was higher and the heavy metal concentration was lower in vermicompost [91,93,94]. Moreover, the supplement of additives could counteract some negative aspects of composting processes, such as emissions of GHGs and odorous molecules.

Due to the large range of raw materials used, composting durations and conditions leads to different compost qualities are produced. Concerning the reproducibility, this is a weakness. However, on the other hand, the diversity of final materials may be treated as a force. The use for plant growth and the properties of materials should meet plant biological requirements.

#### 2.2.2. Coir, a Growing Media Constituent and Stand-Alone Substrate

Coir is the material that forms the middle layers or mesocarp of coconut fruits (*Cocos nucifera* L.). Coir pith, coir fibers, and coir chips are some of the most abundant plant-derived organic waste materials in many tropical and subtropical countries, notable as a rapidly renewable resource. The use of coir as and in growing media has vastly increased since 2004, particularly in Europe but also in the western United States [22].

Similar to peat, coir is used in mixtures for the potting industry as it is a lightweight material and has good air and water holding characteristics. Since coir contains more lignin and less cellulose than peat, it is more resistant to microbial breakdown and usually shrinks less; it is also more hydrophilic and easier to re-wet after drying than peat moss and tends to retain its basic structure when wet or dry [18,95,96].

Leaching of nitrogen is marginally higher and the total water-holding capacity is lower than in peat when comparing materials of a similar particle size, and sometimes natural higher total soluble salts, sodium, and chloride levels are found in coir, depending on their origin [96–98].

However, coir pith has the highest impact on "ecosystem quality", which is often due to land occupation during the coconut harvesting stage [20]. Therefore, efforts have been undertaken to investigate and develop growing media from locally sourced materials, such as, for instance, bark or other wood-based materials, co-products from a forest harvest, or wood processing industries [99–102].

#### 2.2.3. Bark and Wood-Based Materials as Bioresources, Growing Media Constituents, and Stand-Alone Substrates

Bark is a major component of growing media, particularly in areas where peat is scarce or expensive [22], due to transportation cost. It is a lightweight material with a bulk density of 0.1–0.3 g cm−<sup>3</sup> [63]. Similar to coir, bark can be produced in different particle sizes, which makes adjusting the air and water-holding capacities possible by varying the percentage of fine particles [103].

As with coir, pine bark is not produced specifically for use in growing media and tends to have variable physical, chemical, and biological properties [24]. Bark is usually used as a composted or aged material, in order to avoid potential problems with phytotoxicity, since the presence of phenolic compounds, terpenes, and tannins are typical in the chemical composition [30]. High manganese content, especially at low pH could also be a source of potential phytotoxicity [104]. In addition, N deficiency is a common issue, depending on the origin of the material used and the processing method. Recent studies showed that hydrothermal treatments were effective regarding phytotoxicity removal from industrial bark. After this treatment, bark maintains a very high air content that can be a plus in aeration improvement when added to commercial peat-based substrates [31].

Wood fiber, wood chips, and sawdust are renewable resources from the woodworking industry. All these products are characterized by low water retention and good air content. Depending of the initial material, they could sometimes contain phytotoxins that may affect the plant growth at the beginning of cultivation. In this case, a pretreatment with substrate washing would be recommended [105]. Particle-size distribution determines further physical properties, e.g., water retention and water-holding capacity [99,100,106]. A very good correlation was detected between the high percentage of particles <1 mm and max. water holding capacity, and therefore plant growth [101,107].

Wood fibers are further used to optimize the physical properties of other material components, e.g., reducing bulk density, increasing air space, improving re-wetting capacity [24,107,108] and/or as an organic mulch to reduce soil temperature fluctuations, and soil water evaporation and suppress weeds [109,110].

## 2.2.4. Biochar and Hydrothermal Carbonization Products as Bioresources and Growing Media Constituents

Different investigations have been carried out to search for methods that transform agricultural, industrial, and municipal wastes into materials that can be used in growing media. The benefit of diverting wastes from landfills and providing large quantities of organic growing media in the future is particularly important for arid and semiarid regions of the globe [22,23].

Biochar and hydrothermal carbonization (HTC) might play a more important role as constituents of growing media. Whereas biochar is manufactured by heating organic matter in an anoxic situation (pyrolysis), the HTC process requires only moderate temperatures [31] and pressures and is usually used for materials with high water content, e.g., plants. Both processes, pyrolysis and HTC, show great potential for the production of sustainable CO2-neutral energy from biomass, because plants capture the sun energy and convert carbon dioxide from the atmosphere into carbohydrates via photosynthesis [23].

Biochar and HTC char have physical and chemical properties that are variable, depending on the raw material used and the carbonization technique. Usually, the electrical conductivity (EC) and pH values are similarly low in peat and HTC and are slightly increased in biochar [25].

Biochars contain various amounts of different micronutrients in addition to P and K. These nutrients are usually slowly available to plants much like slow release fertilizers, rather than being immediately available [65]. However, there are some problems that need to addressed. For instance, biochar usually contains about 1% nitrogen (N). A high N-immobilization occurs in hydrochar as well. This, and the presence of some phytotoxic substances, were the factors that lead to reduced growth of potted basil, even in mixtures of only 30% by volume [111]. After composting, N-immobilization was reduced and phytotoxic substances degraded within a few weeks [111]. However, as mentioned before, low carbon efficiency, high volume reduction, and time needed for composting make this process not particularly economically attractive. Therefore, apart from feedstock choice, carbonization processes seem to be important for future research.

#### 2.2.5. Other Organic Materials as Bioresources and Growing Media Constituents

Apart from materials analyzed above, several more novel materials and bioresources are used at a small scale and/or have the potential to be used as growing media constituents. These include untransformed waste stream materials, which are affordable and available in certain areas. Waste materials can include, e.g., rice hulls [112–114], almond shell waste [115–117], hazelnut husks [118–120], and paper waste [121]. The main disadvantage of using these materials in commercial soilless media is that they are not produced specifically for horticultural applications; they can therefore be highly inconsistent. As such, they are almost always used in conjunction with more traditional materials [24].

Furthermore, peat moss (*Sphagnum*) from paludiculture has recently been used as a sustainable high-quality alternative to fossil white peat, i.e., as a raw material for plant substrates. *Sphagnum* farming refers to the cultivation of *Sphagnum* mosses to produce *Sphagnum* biomass sustainably [122]. Moreover, *Sphagnum* farming is a feasible large-scale, climate-friendly, and sustainable land use option for abandoned cutover bogs and degraded bog grassland [123]. It reduces human pressure on the remaining natural peatlands in surroundings areas [122].

In areas where forestry activity is minimal, but arable farmland is abundant, the development of soilless growing media from crops normally used as biofuels has been investigated [24]. *Miscanthus* is one such fast-growing crop. *Miscanthus* is a renewable raw material and a low-input crop that can be locally produced, providing ecosystem services, such as CO2 mitigation and biodiversity [124]. Moreover, switchgrass (*Panicum virgatum* L.) [125,126], giant reed (*Arundo donax* L.) [127], reed canary grass (*Phalaris arundinacea* L.) straw [128], and willow (*Salix* spp.) [126] have been used in plant production alone or in mixtures with other materials.

#### *2.3. Growing Medium Choice*

The question as to which is the best growing medium does not have a single answer. This will depend on the location, the availability and cost of potential growing medium constituents, and the crop production system envisaged.

The materials for growing media have to fulfil different requirements: First, they should be available consistently from batch to batch and economically feasible, i.e., the materials and the production process should not be very expensive. Second, the physical, chemical, and biological properties of the growing medium should meet the biological plant requirement. However, there is no universal substrate or mixture that is valid for all plant species and in all situations of cultivation [11,14,23]. Gruda et al. [14], Barrett et al. [24], Savvas, and Gruda [16] also speak for the performance of growing media. Here, they included not only substrate properties, but also the ability to perform well in real growing conditions.

Third, the material used for production and growing media itself should be sustainable and environmentally friendly. Carbon footprint analyses show that the largest share of emissions from heated greenhouse farms results from energy consumption, followed by substrate, packing, and containers used [129]. The biodiversity concern and climate change emphasize the significance of peat bogs as carbon sinks. Generally, avoiding or reducing the use of peat as a growing media constituent, can substantially reduce the carbon footprint in horticulture [23,130]. Apart from extraction, processing, manufacturing, and transportation are important business factors to distinguish between materials from specific sources [131]. Therefore, the authors suggested a list of eight criteria that reflect current, and potentially future, social and environmental issues in relation to the use of growing media. These include the energy and water used in previously mentioned business factors, the social compliance, ensuring minimum labor standards, continuity of supply, habitat and biodiversity, pollution, renewability, and resource use efficiency. In order to guarantee a continued growth and sustainable development of soilless cultivation, it is important to identify effective and environmentally sustainable materials for growing media [24].

Selecting growing media is not an easy task because environmental issues and technical and financial implications must be considered [14,20]. The geographical location, the selection of plant cultivation and production types, the substrate cost and performance, as well as other societal concerns, govern which growing media has to be selected. In addition, the evidence indicates that growers and gardeners tend to favor the types of growing media they are accustomed to and know how to manage. Hence, inertia is also a barrier to change [132]. In the following, we identified two perspectives and functions that we found important to consider: Production systems and transportation distances.

#### 2.3.1. Production Systems

#### 2.3.1.1. Nursery Production

Peat-based growing media are mainly used for production of seedlings and transplants for vegetables and ornamental plants. Nowadays, efforts in the substrate industry are made toward peat reduction in the entirety of the components, used for growing media. Even 10% wood fiber mixed in pure black peat would significantly reduce the carbon footprint for lamb's lettuce, grown in 4 cm press pots [133]. Higher percentages of wood fiber can result in additional emission reductions. For instance, Gruda and Schnitzler [107] reported that, from a performance point of view, the optimal percentage of wood fiber for the prevention of considerable degradation of press pots was approximately 30% in volume. Similarly, biochars can be favorably used as an amendment to peat-based substrates for the development of sustainable greenhouse production [134]. The authors evaluated the effects of additional biochars at a rate of 15% (*v*/*v*) to a peat-based substrate and found that the biochar addition increased the C, decreased the N availability in fertigated peat-based growing media, and mitigated CO2, CH4, and N2O emissions. To increase microbial activity, compost at a rate of 4% (*v*/*v*) was added. This reaction is similar to results reported for agricultural soils by an additional biochar application.

On the other hand, using the large definition of a plant nursery that includes the production of plants for gardens, agriculture, forestry and conservation biology, bark, and wood fiber substrates are the standards in nursery production. This sustainable way of production will remain steady in the near future.

#### 2.3.1.2. Greenhouse Vegetable Production

Growing media have been used traditionally, mostly for plant propagation, bedding, and pot plant production, but this range of use has expanded to include the total production of many food crops, especially high-value crops grown under protection in greenhouses [14]. For instance, stand-alone substrates, such as rockwool and perlite are used for the commercial soilless production of vegetables [15,16].

The use of polythene-wrapped rockwool, originally produced as insulation in the construction industry, aided by its lightweight and ease of handling, has become the dominant soilless culture system for greenhouse vegetables worldwide and especially in Europe [10]. The advantages of rockwool are substrate uniformity, ease of handling, and ease plant production steering.

Materials which can be pressed in slabs, such as coir, can be successfully used instead of rockwool. The water-buffering capacity is lower in coir dust than in rockwool and peat, and the level of air space varies considerably depend on the origin of the material [97]. Hence, mixing different particle sizes and ratios together or adding other materials is recommended to meet crop-specific moisture and aeration requirements in order to use coir products as stand-alone substrates. For instance, adding perlite to coir improved the physical and hydraulic characteristics of the media, such as total porosity and wettability, by manipulating the porosity and capillarity [135]. However, while coir products can make excellent growing media, the long transportation distance makes this alternative less attractive for many areas, such as Northern Europe and North America (see Section 2.3.2. for more information).

White spruce and fir bark alone or mixed with low-grade peat showed high potential for greenhouse tomato production and represented an environmentally sound alternative to rockwool [136]. Moreover, pine bark can be successfully used as a stand-alone substrate for the cultivation of vegetables, such as bell pepper, cucumbers, and muskmelons [137–139]. An economic analysis determined that pine bark was nearly one-eighth the cost of perlite and could be reused for several consecutive crops, resulting in reduced production costs and greater profits. However, bark could become a limited resource due to the changing timber industry and the fact that it is an effective energy source [140], increasingly used as fuel.

Wood chips and fibers are also gaining traction as an alternative to rockwool for slab culture [141]. Depardieu et al. [142], stated that sawdust- and bark-based materials can be used for strawberry soilless culture production, as long as an initial basic fertilization is applied to avoid the initial tie up. Additional N fertilization from the beginning of plant cultivation is recommended to overcome N immobilization in wood fiber substrates [143].

Recently, Kraska et al. [124] found that cucumbers and tomatoes grown on different stand-alone Miscanthus substrates, such as shreds, chips, and fibers, obtained comparable cumulative yields to rockwool. Generally, by using rockwool alternative substrates, the plant cultivation technology has to be adapted to the growing medium's properties [7].

#### 2.3.1.3. Greenhouse Ornamental Production

The standard substrate component used for the production of greenhouse ornamentals is peat moss. Several stand-alone substrates, such as perlite and volcanic lava are used to produce cut ornamentals. If SCS, such as ebb-and-flow bunches or floors, are applied, pot ornamentals could also be cultivated in alternative peat substrates. Other materials, such as bark, wood fibers etc., can be used up to 100% to produce plants. Since nutrient solution is used to supply the plants, the substrate function is vital to keep and support the plants.

However, depending on the crops and technologies used, the portion of usage of growing media constituents other than peat in pot ornamentals varies between 20–50%. Apart from porosity that is much higher in growing media, an important difference between soil and substrate culture is the limited volume of plant roots in a container. This provides a reduced root system for a comparable and sometimes much higher developed aerial part. According to Savvas and Gruda [16], the particle size of the growing media used and the container geometry have to be properly selected to balance water availability and aeration in the root zone. In addition, an adaptation in cultivation methods, mainly in irrigation systems, is required. Furthermore, investing in SCS demands excellent water quality, drainage water collection systems, and an increase in laborers' skills. A soilless crop is much more sensitive to mistakes as there is hardly any buffer [59].

Bark is used as stand-alone substrate in the production of orchids and as a growing media constituent in pot ornamentals, whereas wood fiber substrates are becoming more and more popular in ornamental plant production. Wood chips and sawdust are usually used in the proportion of 20−30% (volume basis) in mixtures with other substrate components. A reduction in particle size, an increase in volume weight, and an increase in the irrigation frequency is recommended [99,100,106]. Furthermore, clay is added, to increase the water holding capacity and nutrient buffer ability of potting mixes.

Álvarez et al. [144] showed that it is possible to grow container plants of geranium (*Pelargonium peltatum* (L.) L'Hér. ex Aiton) and petunia (*Petunia x hybrida* hort. ex E. Vilm.) using a peat-based substrate mixed with biochar and/or vermicompost. Plants in these substrates showed a similar or enhanced physiological response to those grown under control using a commercial peat-based substrate. When compost is used, perlite may be utilized as a growing medium constituent to increase the drainage and air content of the growing media mix.

Several studies reported that biochar in potting media results in the same ornamental plant growth as in peat-based standard substrates [65,145,146]. According to Kern et al. [25], char materials must not necessarily remain on the level of a minor ingredient, but have the potential to be used as major constituents. Furthermore, since they are characterized by a high porosity and a high water-holding capacity, these materials may also be usable as a substitute for constituents, which are already established in the growing media market, but which have a limited supply [25,147,148]. For instance, rice hull-derived biochar would be a practically applicable amendment to improve the properties of growing media, in terms of an increased cation exchange capacity and water content [149]. The typically high porosity and surface area of biochars promote the retention of water and the sorption of nutrients [25].

Non-decomposed *Sphagnum* has been used with great success in the cultivation of orchids as well as together with peat substrates for the cultivation of *Tagetes patula* L. [150]. These results were confirmed by investigations with *Pelargonium* and *Petunia* [151]. Adding *Sphagnum* fibers to peat increased water retention and hydraulic conductivity, but either reduced or had no impact on air-filled porosity. Moreover, the quality of brown peat can be improved by adding a minimum of 30% *Sphagnum* fibers to sieved peat. Therefore, Jobin et al. [151] stated that *Sphagnum* biomass production will most likely continue to develop, offering the growing mix industry an alternative material with a low carbon footprint and a better use of peatlands.

However, the chosen substrate has to be stable enough and possess a good bulk density within the entire cultivation period and after the sale to the end-consumer. For bed, balcony, bowl, and hanging basket plants, the irrigation management of the end-consumer is a challenge. Since the end-consumers are usually inexperienced, mistakes occur. Any incorrectness is frustrating and associated with product rejection. End-consumers think that they do not possess the "green fingers" and this in turn creates a great loss for horticulture, not only from the profit side.

#### 2.3.2. Transportation Distances

The second perspective is a function of growing media use from distances from sources of primary raw materials to growers. Due to transportation ways, the cost of a growing medium is also a function of location. For instance, in peat-rich regions, such as Northern Europe and Canada, where the transportation distances are relatively short, peat may still be an economical option. Similar to peat, coconut coir is produced in specific locations (mainly South-East Asia) and, if not used locally, has to be transported to growers in other parts of the world, with unavoidable costs [9,23]. This is the reason why regional substrates, such as volcanic lava and pumice are and will certainly remain important in the South of Europe in the future. However, location is not only important from an economical point of view, but also from a sustainability perspective, due to the high CO2 footprint. Therefore, compost, together with biochar and hydrochar, has good chances, since usually they are locally produced. Materials, whether sourced from industrial, agricultural, or municipal waste are being investigated as soilless substrate components [24]. A particular trend has been the use of renewable raw materials locally sourced, natural in occurrence and fast-growing, in particular in industrialized countries [16,30].

#### *2.4. Disposal Concerns and Waste Management*

The disposal issue is one of the biggest concerns of using soilless culture and growing media. The question is, what can be done with several fertilizer leachates and water waste during the cultivation period as well as the growing media after its end-of-life?

The generally accepted waste management hierarchy includes the three Rs: Reduce, reuse, and recycle [152]. Reducing the amount of growing medium per plant contributes to reducing CO2 emissions in the production chain of plants [7].

In the seedling and transplant industry there has recently been a trend among producers towards more cells per tray, which decreases the need for growing medium and increases the number of seedlings or transplants produced per unit area [153]. However, the reduction of growing media amount is not always a viable option, due to a direct influence on yield and product quality parameters [9,13]. For instance, Gruda and Schnitzler [153] reported that a reduction of the pot size decreased the quality of the lettuce seedlings. However, no differences were found in the lettuce yield after transplanting to the field and this is of much importance. Certainly, culture methods, such as irrigation and a good root development of seedlings in wood fiber substrates, have been responsible for these results [153].

On the other hand, using SCS means using a reliabe and precise dosage of both fertilizer and water, and this is one of the advantages of using closed systems, at least theoretically. However, in practice, soilless culture vegetables are usually over-fertilized, and an excessive synthetic N fertilizer is applied to ensure that no nutrient deficiency occurs. Indeed, as Truffault et al. [154] reported, over-fertilized tomatoes provided an accumulation of N in leaves and stems. However, yield, leaf photosynthetic activity, and plant architecture were not significantly improved. In addition, the quality of tomato fruits decreased in terms of their sugar:acid ratio and dramatically decreased in the pericarp, whereas the locular gel composition remained similar [154]. Therefore, the reduction of fertilizer used, first and foremost the N fertilizer, is the first appropriate and sustainable step that should be undertaken. The impacts are not only related to the use of fertilizers itself but also to the amount of energy, materials, and transport processes involved in the production of fertilizers [155] and manufacturing facilities. As Gruda et al. [7,8] reported, the fertilizer reduction is directly linked with a reduction of N-emissions (N2O, NH3, and NOx) that, in turn, have an enormous effect on GHGs.

One way to address the runoff nutrient wastewater pollution in open-loop hydroponic systems is the reuse of drained nutrient solutions to a second greenhouse crop. This system is called the "cascade cropping system" [156,157]. Muñoz et al. [157] reported that the N leachate from a soilless tomato system decreased by more than 60% when the nutrient solution was used in a tomato soil system. Moreover, intense and year-round crop production, high N-fertilizer application, suitable temperatures, and frequent irrigation make the greenhouse system an ideal environment for high N-emissions that are considered to be extremely damaging to the ozone layer [7]. The adoption of the cascade crop system reduced the environmental impact by 21%, but increased the eutrophication category by 10% because of the yield reduction [157]. Similarly, cherry tomatoes may be grown with an exhausted nutrient solution that is flushed out from a culture of a salt-sensitive tomato cultivar in semi-closed soilless systems [156]. Several other studies stated that nutrient solution discharged from hydroponic culturing systems can be reused for the production of several vegetables in indoor or outdoor conditions, such as Chinese cabbage [158], melon, and cucumber [159]. These results are in agreement with the growth promotion of poinsettias (*Euphorbia pulcherrima* Willd. ex Klotzsch) after reusing the waste nutrient solution from rose hydroponic cultures [160].

Growing media can be reused as well. Reuse is the best approach in terms of its environmental impact and the results of LCA [9]. For instance, multiple cucumber cycles can be produced on the same growing media in soilless or substrate culture systems, whereas a reuse of substrates in containers systems is generally not common. However, reusing could be associated with distributions of pathogen infections and the possible deterioration of substrate properties. Therefore, in accordance with the Directive EU2018/851 of the European Parliament and of The Council, "waste management in the European Union should be improved and transformed into sustainable material management, with a view to protecting, preserving, and improving the quality of the environment, protecting human health, ensuring prudent, efficient, and rational utilization of natural resources, promoting the principles of the circular economy ... " [161]. The directive further regulates how to reuse and prepare for reuse and recycling, in line with the waste hierarchy. With regards to growing media, the reuse of substrates may induce a higher compaction with increased volume weight (bulk density) and reduction of porosity, due to shrinkage [9,162], with a limited air and low water buffer capacity [101] accompanied by failures and a bottleneck situation of nutrients [163]. On the other hand, the gradual accumulation of nutrients in organic substrates during growing season may have adverse effects on plant development [148], and these effects are further increased by a substrate reuse. Xing et al. [164] identified a total of 358 differentially abundant proteins, including 11 mineral ion binding and transport related proteins, such as a calmodulin-like protein and a nitrate transporter 3.2 under peat-vermiculite and coir tomato cultivation. Xing et al. [164] suggested that these indicators could contribute to a better control of SCS and a waste reduction.

The investigations of crop response to the cultivation in reused growing media compared to virgin substrates show contradictory results: (a) Reduction of crop yield and/or produce quality in reused media, (b) minimal differences between virgin and reused substrates, or (c) even better results in reused materials [165]. Similar to virgin growing media, the reused materials have to possess good physical, chemical, and biological properties. Therefore, generally, some remediation steps are recommended to amend the substrate properties before reusing [9].

First, growing media should be free from any infection with pests and diseases, otherwise a disinfection process has to be undertaken. For instance, cleaning and disinfecting perlite with hot water at a temperature of 96 ◦C before reuse produced a better marketable tomato yield in comparison to a virgin one, due to the collective effect of salt reduction, medium disinfection, and the optimum level of nutrients [166]. Second, the nutrient level of growing media should be analyzed and eventually adjusted according to crop demands. This step is very important when a nutrient solution is not used in the second crop. Third, physical properties have to be amended by breaking up and sifting growing media as well as by removing older roots [165].

Further, organic substrates with high microbial activities, such as compost, are often added to used peat substrates, because of their suppressive properties against soilborne diseases, such as *Pythium*. In addition, an artificial inoculation with selected microorganisms or the introduction of microbial antagonists, preliminarily isolated from suppressive soils and/or used soilless media, could be used to increase the suppressive properties against root rot diseases [165,167]

Recycling is the final approach in the waste management hierarchy. To recycle something means that it will be transformed again into raw material, which can be shaped into a new item [152] for second or multiple life uses. Until recently, growing media were always the last step of the value chain, and usually it was all about how to dispose of them without further negative impact on the environment and climate. Composting offers a good option to drastically reduce this impact, as shown in Section 2.2.1. Organic substrates can be used immediately or after their composting as soil amendments. This method is highly evaluated in arid and semi-arid areas, increasing not only organic matter in soil but also improving water holding capacity. In addition, composted materials can be used to cultivate less-demanding crops, such as forest tree saplings [9]. Moreover, Kraska et al. [124] opted for a cascade way of recycling and found a subsequent use of *Miscanthus*-based growing medium for combustion feasible, after the production of cucumbers and tomatoes on different stand-alone *Miscanthus* substrates. As mentioned before, *Miscanthus* is a renewable raw material and a low-input crop that can be locally produced.

#### *2.5. Other Factors Having an Impact on Sustainability*

In temperate regions, controlled environment systems are characterized by large amounts of energy consumption for heating during the cold season. Large energy consumption is the greatest environmental concern [7,8]. As Eigenbrod and Gruda [3] stated, the motto for future plant production should not be "local at any price," but "as sustainable as possible." Therefore, Gruda et al. [7,8] recommend the implementation of so-called next generation culture methods: Better insulation thanks to double cladding and triple screens, following biological and nature-oriented culture techniques, dehumidifying the blown-in air, and, if necessary, humidifying (rewetting) and "harvesting" greenhouse existing heat amounts. In addition, the use of alternative energy sources can fundamentally increase and improve the sustainability of protected cultivation systems and nursery production. Replacing or recycling rockwool and plastic items are other important factors [7,8].

Plastic containers, pots, bags, and trays have been the predominant containers in greenhouse and nursery production. However, most plastics are derived from petroleum—a nonrenewable resource [168]. Therefore, different examples of alternative containers made from plantable and compostable materials, such as bamboo, coconut or wood pulp fiber, rice hulls, and recycled paper have been developed. The use of these containers will furthermore contribute to sustainable systems along with suitable growing media.

Moreover, the lifetime of structure materials, e.g., plastic covers and auxiliary equipment, e.g., drippers, should be further extended and manufactured out of biodegradable material to reduce waste. Better management of the nutrient supply as well as the reduction of fertilizer use is required [7].

Another way to reduce the amount of peat (not only for SCS), used as soil improvements for acidophilic plants, is the breeding of new varieties that have neutral requirements related to pH in the root zone. In addition, the use of plant biostimulants, such as humic substances, protein hydrolysates, seaweed extracts, and beneficial microorganisms, such as mycorrhizal fungi and nitrogen fixation bacteria [37,167,169], can contribute to improve effectiveness and interaction in the root zone of plants into growing media.

#### **3. Conclusions**

In conclusion, soilless culture is one of the best techniques to overcome local water shortages, while also producing high quality produce, even in areas with poor soil structure and problematic conditions. Reduce, reuse, and recycle issues should be more frequently applied in SCS. The application

of these systems is likely to increase close to existing cities as well as in mega-cities worldwide in the near future.

In this paper, we reviewed different organic materials and bioresources used or intended to be used as growing media constituents in the future. All of these have their respective advantages and disadvantages. Different areas in the world, with different conditions and requirements, require different crops, different distances to sources of primary raw materials used as growing media components, and different technologies used to produce plants.

However, factors such as climate change, CO2 emissions, and other ecological issues will determine and drive the adoption and influence of growing media in the near future. Materials that are easily available, financially feasible, environmentally friendly, and that can provide a high-quality growing medium will become replacements for rockwool and peat in the future.

Further research on the innovative approaches in SCS and materials used as growing media components is required.

**Acknowledgments:** Special thanks to Michael Maher, Dublin, Ireland for language revision and valuable recommendations.

**Conflicts of Interest:** I declare no conflict of interest.

## **References**


© 2019 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*

## **Pre-Harvest UV-B Radiation and Photosynthetic Photon Flux Density Interactively A**ff**ect Plant Photosynthesis, Growth, and Secondary Metabolites Accumulation in Basil (***Ocimum basilicum***) Plants**

## **Haijie Dou 1, Genhua Niu 2,\* and Mengmeng Gu <sup>3</sup>**


Received: 19 July 2019; Accepted: 5 August 2019; Published: 7 August 2019

**Abstract:** Phenolic compounds in basil (*Ocimum basilicum*) plants grown under a controlled environment are reduced due to the absence of ultraviolet (UV) radiation and low photosynthetic photon flux density (PPFD). To characterize the optimal UV-B radiation dose and PPFD for enhancing the synthesis of phenolic compounds in basil plants without yield reduction, green and purple basil plants grown at two PPFDs, 160 and 224 <sup>μ</sup>mol·m−2·s<sup>−</sup>1, were treated with five UV-B radiation doses including control, 1 h·d−<sup>1</sup> for 2 days, 2 h·d−<sup>1</sup> for 2 days, 1 h·d−<sup>1</sup> for 5 days, and 2 h·d−<sup>1</sup> for 5 days. Supplemental UV-B radiation suppressed plant growth and resulted in reduced plant yield, while high PPFD increased plant yield. Shoot fresh weight in green and purple basil plants was 12%–51% and 6%–44% lower, respectively, after UV-B treatments compared to control. Concentrations of anthocyanin, phenolics, and flavonoids in green basil leaves increased under all UV-B treatments by 9%–18%, 28%–126%, and 80%–169%, respectively, and the increase was greater under low PPFD compared to high PPFD. In purple basil plants, concentrations of phenolics and flavonoids increased after 2 h·d−<sup>1</sup> UV-B treatments. Among all treatments, 1 h·d−<sup>1</sup> for 2 days UV-B radiation under PPFD of 224 <sup>μ</sup>mol·m−2·s−<sup>1</sup> was the optimal condition for green basil production under a controlled environment.

**Keywords:** UVR8; PPFD; dose-dependent; photosynthesis; chlorophyll fluorescence; phenolic compounds

## **1. Introduction**

Decreasing arable land, rising urbanization, water scarcity, and climate change exert pressure on agricultural producers [1]. Conventional food production is severely limited by seasonality, unpredictable weather, pests/diseases, and resources such as land and water. Indoor controlled environment agriculture (CEA) systems, which can be built anywhere, have the potential to be a suitable alternative to open field and greenhouse production [2]. However, crops cultivated in indoor CEA systems using artificial lighting are not exposed to ultraviolet radiation. Ultraviolet (UV) radiation is an important environmental signal that initiates plant responses in photosynthetic function, cell division, plant growth, and development [3,4]. In previous studies, UV-B radiation was mainly considered as a stress factor to plants, focusing on the effects of increasing solar UV-B radiation reaching Earth's surface due to stratospheric ozone depletion [5,6]. Recent studies have highlighted supplemental UV-B radiation as a eustress (i.e., positive stress), and reported that low to moderate

UV-B radiation induces a range of favorable processes in plants, such as synthesis of UV-absorbing compounds (anthocyanin, phenolic acids, and flavonoids) and antioxidants (carotenoids, ascorbate, and glucosinolate) [7–9]. These bioactive compounds represent an important source of antioxidant molecules in human diet reducing the risk of cardiovascular diseases, chronic diseases, and specific forms of cancer [10,11].

Manipulation of secondary metabolites in horticultural crops through supplemental UV-B radiation have demonstrated at least two UV-B signaling pathways, which is determined by UV-B radiation dose [11,12]. Under low UV-B radiation dose, the UV-B specific photoreceptor, UV RESISTANCE LOCUS 8 (*UVR8*), initiates an *UVR8*-dependent pathway [13]. Specifically, *UVR8* stimulates gene expression such as CONSTITUTIVELY PHOTOMORPHOGENIC 1 (*COP1*), ENLONGATED HYPOCOTYL 5 (*HY5*), and HY5 HOMOLOG (*HYH*), which play key roles in the synthesis of phenolic compounds, as well as growth retardation such as the inhibition of hypocotyl elongation [14,15]. Under high UV-B radiation dose, UV-B light acts as a damaging agent inducing formation of reactive oxygen species (ROS), causing damage to plant cells, DNA, proteins, and photosynthesis apparatus and, subsequently, negatively affect plant growth and induces synthesis of antioxidants [16,17].

In addition to being dose-dependent, plant responses to supplemental UV radiation also varied among species and cultivars [18]. For example, anthocyanin concentration of red leaf lettuce (*Lactuca sativa*, 'Red Cross') increased by 11% after 12-days UV-A radiation at 18 <sup>μ</sup>mol·m−2·s−<sup>1</sup> for 16 h·d−<sup>1</sup> prior to harvest (controlled environment, PPFD of 300 <sup>μ</sup>mol·m−2·s<sup>−</sup>1) [19]. Synthesis of anthocyanin and other polyphenols in another red leaf lettuce cultivar ('Red Fire', controlled environment, PPFD of 150 <sup>μ</sup>mol·m−2·s−1) significantly increased after 3-days UV-B radiation at a much lower dose, 1.5 <sup>μ</sup>mol·m−2·s−<sup>1</sup> for 16 h·d−<sup>1</sup> prior to harvest [4]. Furthermore, glucosinolate concentration in 7-day-old broccoli (*Brassica oleracea*) sprouts (controlled environment, PPFD not mentioned) was enhanced by 19% after 1-day UV-B radiation at 7.0 <sup>μ</sup>mol·m−2·s−<sup>1</sup> for 2 h·d<sup>−</sup>1, compared to 63% enhancement at 10.3 <sup>μ</sup>mol·m−2·s−<sup>1</sup> for 2 h·d−<sup>1</sup> [9].

Basil (*Ocimum basilicum*) plants have been considered a source of valuable healthy substances due to their unique flavor and relatively high content of phenolic compounds [20,21]. To improve the yield of high-quality basil, more growers are turning to controlled environment production, which has been proven to be a suitable alternative to open field and greenhouse basil production, due to its high environmental controllability and improved resource utilization efficiency (arable land and clean water) [2,22]. However, crops cultivated in controlled environment systems using artificial lighting are not exposed to UV-B radiation, bearing a direct impact on basil flavor and visual appearance [10]. Meanwhile, considering energy saving, the photosynthetic photon flux density (PPFD) in controlled environment systems is much lower compared to sunlight intensity in open field, resulting in further reduction of secondary plant metabolites [21]. Therefore, there is an increasing interest in the use of supplemental UV-B radiation to enhance the synthesis of health-beneficial phenolic compounds to produce premium quality basil products under controlled environment [3,23,24].

Although some studies investigated the effects of supplemental UV-B radiation on phytochemical accumulation of basil plants, most studies were conducted in the open field or greenhouse using photo-selective film covers, and results varied largely in both biomass production and phenolic contents [25–27]. Meanwhile, most studies only focused on the effects of UV-B radiation on secondary metabolites accumulation, not considering yield reduction caused by UV-B radiation [25,28]. Furthermore, considering the significantly low PPFD used in controlled environment systems, little information is known about the interactive effects between supplemental UV-B radiation and PPFD. Collectively, to identify the optimal combination of UV-B radiation dose and PPFD that enhance concentrations of phenolic compounds without significant yield reduction, further investigation is warranted to characterize the physiological, morphological, and biochemical responses in basil plants to supplemental UV-B radiation and different PPFDs under a controlled environment.

Accordingly, in the present study, we exposed two basil cultivars to five pre-harvest supplemental UV-B radiation doses in order to characterize plant responses to supplemental UV-B radiation under two PPFDs in a controlled environment system. Photosynthetic photon flux density of 224 <sup>μ</sup>mol·m−2·s−<sup>1</sup> for basil plants was selected according to our previous study [21], and a low PPFD of 160 <sup>μ</sup>mol·m−2·s−<sup>1</sup> was selected to test if UV-B radiation can compensate for the reduced accumulation of phenolic compounds in basil plants grown under low PPFD.

#### **2. Materials and Methods**

## *2.1. Plant Materials and Culture*

Experiments were conducted in a walk-in growth room in Texas AgriLife Research Center at El Paso, TX, USA, from 8 August to 15 September 2017 on green basil 'Improved Genovese Compact' and from 5 September to 19 October 2017 on purple basil 'Red Rubin' (Johnny's Selected Seeds, Winslow, ME, USA), respectively. For both experiments, one basil seed per cell was sown in 72 square cell trays (length 3.86 cm; height 5.72 cm; volume 59 cm3) with Metro-Mix® 360 (peat moss 41%, vermiculite 34%, pine bark 25%, Sun Gro® Horticulture, Bellevue, WA, USA). All trays were put under mist in a greenhouse for germination. Temperature under the mist was maintained at 32.7 ◦C/22.2 ◦C day/night. Seedlings were moved out from the mist after the emergence of cotyledons and grown in a greenhouse for two weeks. Temperature and relative humidity in the greenhouse were maintained at 29.1 ◦C/21.6 ◦C and 48%/66% day/night, respectively. When one pair of true leaves fully expanded, basil seedlings were transplanted into square pots (length 9.52 cm, height 8.26 cm, and volume 574 cm3) filled with the Metro-Mix® 360, and uniform plants were selected and moved to the walk-in growth room for different treatments.

After transplanting, multi-layer cultivating shelves were used with mechanical mini fans (LS1225A-X, AC Infinity, City of Industry, CA, USA) circulating air to achieve uniform temperatures across treatments. Plant canopy temperature in each treatment was maintained at 23.9 ◦C/21.2 ◦C day/night. All plants were manually sub-irrigated with a nutrient solution containing 1.88 g·L−<sup>1</sup> (277.5 ppm N) 15N-2.2P-12.5K (Peters 15-5-15 Ca-Mg Special, The Scotts Company, Marysville, OH, USA) as needed. The nutrient solution was mixed and stored in a 100-gallon tank with a lid, and the electrical conductivity (EC) and pH were adjusted to 2.0 dS·m−<sup>1</sup> and 6.0, respectively, using an EC/pH meter (Model B-173, Horiba, Ltd., Japan).

### *2.2. Supplemental Ultraviolet B (UV-B) Radiation and Photosynthetic Photon Flux Density (PPFD) Treatments*

Uniform green and purple basil plants were grown under two PPFDs of 160 and 224 <sup>μ</sup>mol·m−2·s−<sup>1</sup> with a 16-h photoperiod provided by cool white fluorescent lamps (Philips Lighting, Somerset, NJ, USA). Two or five days prior to harvest (basil plant height reaching about 25 cm), UV-B lamps were switched on and basil plants were treated with one of the five UV-B radiation doses including no supplemental UV-B radiation (control), 1 h·d−<sup>1</sup> for 2 days (1H2D), 2 h·d−<sup>1</sup> for 2 days (2H2D), 1 h·d−<sup>1</sup> for 5 days (1H5D), or 2 h·d−<sup>1</sup> for 5 days (2H5D) with UV-B light intensity at 16.0 <sup>μ</sup>mol·m−2·s−<sup>1</sup> (equal to 18.7 kJ·m−2·h<sup>−</sup>1). There were a total of 10 treatments created by the combination of two PPFDs and five UV-B radiation doses, and 12 plants per treatment. Supplemental UV-B radiation treatments were applied from 8:00 in the morning and provided by Philips TL 40W/12 and 20W/12 UV-B broadband lamps (wavelength: 270–400 nm, maximum emission wavelength at 315 nm, Svetila.com d.o.o., Domzale, Slovenia, EU). The cool white fluorescent lamps at PPFD of 160 and 224 <sup>μ</sup>mol·m−2·s−<sup>1</sup> radiated low intensity of UV radiation, which was 2.2 and 2.5 <sup>μ</sup>mol·m−2·s<sup>−</sup>1, respectively. The UV-B light intensity (including UV radiation provided by broadband UV-B lamps and cool white fluorescent lamps) and PPFD in each treatment were measured at 15 cm underneath the lamps at 9 spots using a MU-200 UV radiation meter (Apogee Instruments, Logan, UT, USA) and PS-100 spectroradiometer (Apogee Instruments, Logan, UT, USA), respectively, before placing the plants. To minimize the disproportionate light distribution within each treatment, all plants were systematically rearranged every 3 days.

#### *2.3. Measurements*

#### 2.3.1. Growth Parameters

Growth parameters of basil plants such as plant height, width, the number of internodes, leaf area, and yield including shoot fresh weight (FW) and dry weight (DW) were recorded at harvest (on 15 September and 19 October 2017 for green and purple basil plants, respectively). Plant width was calculated as the average of the widest point and its perpendicular width of plant canopy. A leaf area meter (LI-3100, LI-COR, Lincoln, NE, USA) was used to measure the leaf area. Shoot DW was determined after shoot tissues were dried at 80 ◦C in an oven (Grieve, Round Lake, IL, USA) for 3 days. Specific leaf area (leaf area per unit leaf dry weight) was calculated as an indicator of leaf thickness.

#### 2.3.2. Gas-Exchange Rate, Relative Chlorophyll Concentration, and Chlorophyll Fluorescence

A portable gas exchange analyzer (CIRAS-3, PP Systems International, Amesbury, MA, USA) was used to measure the gas exchange rate, including net photosynthetic rate (Pn), transpiration rate (E), and stomatal conductance (Gs) of basil leaves at harvest. A PLC3 leaf cuvette with light-emitting diode (LED) light unit (white light, in which the proportions of red, blue, and green light were 38%, 25%, and 37%, respectively) was used. The PPFD, temperature, relative air humidity, and CO2 concentration inside the leaf cuvette were set at 800 <sup>μ</sup>mol·m−2·s−1, 25 ◦C, 50%, and 390 <sup>μ</sup>mol·mol−1, respectively. The third pair of leaves from the top was used for measuring and measurements were taken until the Pn reached a steady state.

Soil plant analysis development (SPAD) index of basil leaves was recorded on the third pair of leaves from the top at harvest to quantify the relative chlorophyll concentration of basil leaves using a chlorophyll meter SPAD-502 (Konica-Minolta cooperation, Ltd., Osaka, Japan). Three measurements were taken for each leaf and the average was recorded for data analysis.

Chlorophyll fluorescence parameters of basil plants were measured at harvest using a pocket Plant Efficiency Analyzer chlorophyll fluorimeter (PEA, Hansatech Instruments Ltd., Norfolk, UK). The third pair of leaves from the top were dark adapted for at least 30 min prior to the measurement. Minimal fluorescence values (F0) and maximal fluorescence values (Fm) in the dark-adapted state were measured, and maximum quantum use efficiency of photosystem II (PSII) in the dark-adapted state was calculated as Fv/Fm = (Fm − F0)/Fm. Performance index (PI ABS, where "ABS" specifies that the reaction centers' density is expressed per absorption), dissipation of energy per cross section (DI0/CS), trapped energy flux per cross section (TR0/CS), and electron transport flux per cross section (ET0/CS) parameters were calculated using the PEA Plus software (V1.10, Hansatech Instruments Ltd., Norfolk, UK).

#### 2.3.3. Secondary Plant Metabolites

Five basil plants were randomly selected for the measurement of concentrations of anthocyanin, phenolics, and flavonoids, and antioxidant capacity of basil leaves at harvest. Fresh basil leaves were collected in a cooler and immediately stored in a deep freezer (IU1786A, Thermo Fisher Scientific, Marietta, OH, USA) at −80 ◦C until phytochemical evaluation.

Extraction. Approximately 2 g fresh basil leaves were ground in liquid nitrogen and extracted with 15 mL 1% acidified methanol at 4 ◦C in dark. After overnight extraction, the mixture was centrifuged (Sorvall RC 6 Plus Centrifuge, Thermo Fisher Scientific, Madison, WI, USA) at 13,200 rpm (26,669× *g*) for 15 min, and the supernatant was collected for phytochemical evaluation [29].

Anthocyanin evaluation. Absorbance of the extract was measured at 530 nm using a spectrophotometer (Genesys 10S ultraviolet/Vis, Thermo Fisher Scientific, Madison, WI, USA), and anthocyanin concentration was expressed as mg cyanidin-3-glucoside equivalent per 100 g FW of basil leaves using a molar extinction coefficient of 29,600 [30].

Phenolics evaluation. A modified Folin-Ciocalteu reagent method [29] was used to determine the phenolics concentration of basil leaves: 100 μL extraction sample was added to a mixture of 750 μL 1/10 dilution Folin–Ciocalteau reagent and 150 μL distilled water. After 6 min reaction, 600 μL 7.5% Na2CO3 was added and the mixture was incubated at 45 ◦C in a water bath for 10 min before the absorbance was measured at 725 nm using a microplate reader (EL×800, BioTek, Winooski, VT, USA). Results were shown as mg of gallic acid equivalent per g FW of basil leaves.

Flavonoids evaluation. Flavonoid concentration of basil leaves was determined [21] as the following: 20 μL extraction sample was added to a mixture of 85 μL distilled water and 5 μL 5% NaNO2. After 6 min reaction, a 10 μL of 10% AlCl3·6H2O was added to the mixture. After another 5 min reaction, 35 μL of 1M NaOH and 20 μL distilled water were added to the mixture and the absorbance was measured at 520 nm using the aforementioned microplate reader. Results were shown as mg of (+)-catechin hydrate equivalent per g FW of basil leaves.

Antioxidant capacity evaluation. A 2,2'-azino-bis (3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) method [31] was used to determine the antioxidant capacity of basil leaves: 150 μL extracted sample was added to 2.85 mL of ABTS<sup>+</sup> solution and incubate at room temperature for 10 min. The absorbance of mixed solution was measured at 734 nm using the aforementioned spectrophotometer. Results were shown as mg of Trolox equivalent antioxidant capacity per 100 g FW of basil leaves.

### *2.4. Statistical Analyses*

Experiments were arranged in a two factors factorial design. Five plants per treatment were randomly selected for measurement. After verifying the significance of the two main factors (UV-B and PPFD) and their interaction (PPFD × UV-B), a one-way analysis of variance among 10 treatments was conducted for green and purple basil plants, respectively, according to Student's *t* method (*p* < 0.05). Some data were pooled from two PPFDs because effect of PPFD was not statistically significant. Pairwise correlations method (*p* < 0.05) was used to test correlations between parameters. All statistical analyses were performed using JMP software (Version 13, SAS Institute Inc., Cary, NC, USA).

#### **3. Results**

#### *3.1. Gas Exchange Rate, Relative Chlorophyll Concentration, and Chlorophyll Fluorescence*

Supplemental UV-B radiation suppressed plant photosynthesis, in which Pn, E, and Gs in both basil cultivars were lower compared to plants grown under control, while PPFD showed no effects (Table 1). In green and purple basil leaves, Pn, E, and Gs was 68%/70%, 55%/68%, and 65%/76% lower under treatment 2H5D compared to plants grown under control, respectively. Relative chlorophyll concentration of green and purple basil plants was 9%–15% and 6%–8% lower under supplemental UV-B radiation compared to plants grown under control, respectively, while PPFD showed no effect on green basil plants but increased relative chlorophyll concentration in purple basil plants (Figure 1).

Supplemental UV-B radiation inhibited plant chlorophyll fluorescence parameters in green basil plants, including Fv/Fm and PI ABS. However, in purple basil plants, Fv/Fm showed no differences between control and 1H2D treatment, and PI ABS was only lower under the highest UV-B radiation dose, 2H5D treatment (Figure 2A,B). Similarly, TR0/CS and ET0/CS in green basil plants were lower after UV-B radiation, while they were not affected by UV-B radiation in purple basil plants (Figure 2D,E). On the contrary, DI0/CS in purple basil plants was significantly higher under treatments 1H5D and 2H5D, while in green basil plants no treatment effect was observed (Figure 2C). Chlorophyll fluorescence parameters in basil plants were not affected by PPFD.


**Table 1.** Net photosynthetic rate (Pn), transpiration rate (E), and stomatal conductance (Gs) of green basil 'Improved Genovese Compact' and purple basil 'Red Rubin' plants under five supplemental UV-B radiation treatments, including no supplemental UV-B radiation (control), 1 h·d−<sup>1</sup> for 2 days (1H2D), 2 h·d−<sup>1</sup> for 2 days (2H2D), 1 h·d−<sup>1</sup> for 5 days (1H5D), and 2 h·d−<sup>1</sup> for 5 days (2H5D).

Data were pooled from two photosynthetic photon flux density (PPFD) treatments. <sup>z</sup> Means followed by the same lower/upper case letters are not significantly different for green/purple basil plants, according to Student's *t* mean comparison (*p* < 0.05).

**Figure 1.** Relative chlorophyll concentration (soil plant analysis development (SPAD) index) of green basil 'Improved Genovese Compact' and purple basil 'Red Rubin' plants at different treatments. There were 10 treatments created by the combination of two photosynthetic photon flux density (PPFD) of 160 and 224 <sup>μ</sup>mol·m−2·s−<sup>1</sup> and five ultraviolet B (UV-B) radiation treatments, including no supplemental UV-B radiation (control), 1 h·d−<sup>1</sup> for 2 days (1H2D), 2 h·d−<sup>1</sup> for 2 days (2H2D), 1 h·d−<sup>1</sup> for 5 days (1H5D), and 2 h·d−<sup>1</sup> for 5 days (2H5D). Means followed by the same lower/upper case letters are not significantly different for green/purple basil plants, according to Student's *t* mean comparison (*p* < 0.05). Bars represent standard errors.

### *3.2. Growth Parameters and Crop Yield*

Supplemental UV-B radiation inhibited plant growth in both basil cultivars and performed as lower plant height, width, and leaf area, and the detriment increased with increasing UV-B radiation doses (Table 2). Specifically, under high PPFD (224 <sup>μ</sup>mol·m−2·s<sup>−</sup>1), plant height of both basil cultivars was the highest under treatments control and 1H2D, followed by treatments 2H2D and 1H5D, and the lowest under treatment 2H5D. Leaf area of green/purple basil plants was 14%/17%, 28%/30%, 28%/34%, and 44%/44% lower, respectively, under treatments 1H2D, 2H2D, 1H5D, and 2H5D compared to control. Specific leaf area (leaf area per unit leaf dry weight) was calculated and used as an indicator of leaf thickness. In the present study, specific leaf area of both basil cultivars was lower under supplemental UV-B radiation, indicating increased leaf thickness after supplemental UV-B radiation (Table 2). Under

higher UV-B radiation doses such as 1H5D and 2H5D treatments, basil plants also showed leaf bronze, chlorosis, waxy appearance, and premature leaf defoliation (Figure 3).

**Figure 2.** Chlorophyll fluorescence parameters, including maximal photochemical efficiency of Photosystem II (Fv/Fm) (**A**), performance index (PI ABS, where "ABS" specifies that the reaction centers' density is expressed per absorption) (**B**), dissipation of energy per cross section (DI0/CS) (**C**), trapped energy per cross section (TR0/CS) (**D**), and electron transport flux per cross section (ET0/CS) (**E**) of green basil 'Improved Genovese Compact' and purple basil 'Red Rubin' plants under different supplemental UV-B radiation treatments including control, 1H2D, 2H2D, 1H5D, 2H5D. Data were pooled from two photosynthetic photon flux density (PPFD) treatments. Means followed by the same lower/upper case letters are not significantly different for green/purple basil plants, according to Student's *t* mean comparison (*p* < 0.05). Bars represent standard errors.


**Table 2.** Plant height, width, leaf area, and specific leaf area of green basil 'Improved Genovese Compact' and purple basil 'Red Rubin' plants under different treatments. There were 10 treatments created by the combination of two photosynthetic photon flux density (PPFD) levels of 160 and 224 <sup>μ</sup>mol·m−2·s−<sup>1</sup> and five UV-B irradiation treatments including control, 1H2D, 2H2D, 1H5D, 2H5D.

<sup>z</sup> Means followed by the same lower/upper case letters are not significantly different for green/purple basil plants, according to Student's *t* mean comparison (*p* < 0.05). Asterisks (\*) indicate significant differences (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001). NS indicates non-significant differences (\* *p* < 0.05).


**Figure 3.** Green basil 'Improved Genovese Compact' and purple basil 'Red Rubin' plants under different treatments at harvest. There were 10 treatments created by the combination of two photosynthetic photon flux density (PPFD) levels of 160 and 224 <sup>μ</sup>mol·m−2·s−<sup>1</sup> and five UV-B radiation treatments including control, 1H2D, 2H2D, 1H5D, 2H5D.

Shoot FW and DW of green and purple basil plants were generally lower in plants grown under supplemental UV-B treatments, and interactive effects (UV-B × PPFD) were observed on shoot FW (*p* = 0.01) and shoot DW (*p* = 0.02) in purple basil plants, while only interactions in shoot DW were observed in green basil plants (*<sup>p</sup>* <sup>=</sup> 0.03). Specifically, under low PPFD (160 <sup>μ</sup>mol·m−2·s<sup>−</sup>1), treatment 1H2D showed no effects on shoot FW in green basil plants. So did the 1H2D and 1H5D treatments in purple basil plants, while under high PPFD (224 <sup>μ</sup>mol·m−2·s<sup>−</sup>1), shoot FW in both cultivars was lower under UV-B treatments compared to control (Figure 4A,B).

**Figure 4.** Shoot fresh weight and shoot dry weight of green basil 'Improved Genovese Compact' plants (**A**), and purple basil 'Red Rubin' plants (**B**) under different treatments. There were 10 treatments created by the combination of two photosynthetic photon flux density (PPFD) of 160 and 224 <sup>μ</sup>mol·m−2·s−<sup>1</sup> and five UV-B radiation treatments including control, 1H2D, 2H2D, 1H5D, 2H5D. Means followed by the same lower/upper case letters are not significantly different for green/purple plants, according to Student's *t* mean comparison (*p* < 0.05). Bars represent standard errors.

Plant height, leaf area, leaf thickness, shoot FW, and shoot DW in both basil cultivars were higher under high PPFD (Table 2, Figure 4A,B). Without supplemental UV-B treatments, plant height, leaf area, leaf thickness, shoot FW, and shoot DW in green/purple basil plants were 16%/12%, 24%/21%, 15%/9%, 44%/34%, and 59%/35% higher under high PPFD (224 <sup>μ</sup>mol·m−2·s−1) compared to plants grown under low PPFD (160 <sup>μ</sup>mol·m−2·s<sup>−</sup>1), respectively.

#### *3.3. Secondary Plant Metabolites Accumulation and Antioxidant Capacity*

Concentrations of phenolic compounds in green basil plants, including anthocyanin, phenolics, and flavonoids were 9%–23%, 28%–126%, and 80%–169% greater, respectively, after UV-B radiation compared to control (Table 3). Concentrations of anthocyanin and flavonoids in green basil plants were not affected by PPFD, while phenolics concentration was greater under high PPFD (224 <sup>μ</sup>mol·m−2·s<sup>−</sup>1). In purple basil plants, only 2 h·d−<sup>1</sup> UV-B treatments (2H2D and 2H5D) enriched concentrations of phenolics and flavonoids, while UV-B treatments showed no effects on anthocyanin concentration (Table 3). Specifically, under 2H2D and 2H5D treatments, concentrations of phenolics and flavonoids in

purple basil plants were 29%–63% and 37%–79% greater, respectively. Concentrations of anthocyanin and phenolics in purple basil plants were greater under high PPFD (224 <sup>μ</sup>mol·m−2·s<sup>−</sup>1), while flavonoid concentration was not affected by PPFD (Table 3).

**Table 3.** Anthocyanin concentration (conc.), phenolics conc., and flavonoids conc. of green basil 'Improved Genovese Compact' and purple basil 'Red Rubin' plants under different treatments. There were 10 treatments created by the combination of two photosynthetic photon flux density (PPFD) of 160 and 224 <sup>μ</sup>mol·m−2·s−<sup>1</sup> and five UV-B radiation treatments including control, 1H2D, 2H2D, 1H5D, 2H5D.


<sup>z</sup> Means followed by the same lower/upper case letters are not significantly different for green/purple basil plants, according to Student's *t* mean comparison (*p* < 0.05). Asterisks (\*) indicate significant differences (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001). NS indicates non-significant differences (\* *p* < 0.05).

The total amounts of phytochemicals per plant (i.e., anthocyanin, phenolics, and flavonoids) were calculated by multiplying the phytochemical concentrations by leaf FW per plant (Table 4). Under low PPFD (160 <sup>μ</sup>mol·m−2·s−1), total amount of anthocyanin in green basil plants was 23% lower under treatment 2H5D compared to control, while total amounts of phenolics and flavonoids were 49%–79%% greater (Table 4). Under high PPFD (224 <sup>μ</sup>mol·m−2·s<sup>−</sup>1), total amounts of anthocyanin and phenolics in green basil plants were 15%–39% lower under supplemental UV-B treatments compared to control, while total amount of flavonoids was 43%–44% higher under treatments 1H2D and 1H5D compared to control (Table 4). In purple basil plants, all supplemental UV-B radiation treatments showed negative or no effects on the total amount of phenolic compounds regardless of PPFD (Table 4).


**Table 4.** Total amount of anthocyanin, phenolics, and flavonoids per plant of green basil 'Improved Genovese Compact' and purple basil 'Red Rubin' plants under different treatments. There were 10 treatments created by the combination of two photosynthetic photon flux density (PPFD) of 160 and <sup>224</sup> <sup>μ</sup>mol·m−2·s−<sup>1</sup> and five UV-B radiation treatments including control, 1H2D, 2H2D, 1H5D, 2H5D.

<sup>z</sup> Means followed by the same lower/upper case letters are not significantly different for green/purple basil plants, according to Student's *t* mean comparison (*p* < 0.05).

Antioxidant capacity in basil plants were not affected by PPFDs. Antioxidant capacity in green basil plants was higher under all supplemental UV-B radiation treatments, while it was only higher under 2 h·d−<sup>1</sup> UV-B treatments (2H2D and 2H5D) in purple basil plants (Figure 5A). Correlation between antioxidant capacity and UV-B radiation doses was analyzed in three terms according to different UV-B radiation patterns, all UV-B treatments (Figure 5A), 1 h·d−<sup>1</sup> UV-B treatments (1H2D and 1H5D, Figure 5B), and 2 h·d−<sup>1</sup> UV-B treatments (2H2D and 2H5D, Figure 5C). Antioxidant capacity in green basil plants were all positively related to UV-B radiation doses regardless of radiation patterns, while antioxidant capacity in purple basil plants showed no correlation with 1 h·d−<sup>1</sup> UV-B radiation treatments (1H2D and 1H5D, *p* = 0.1994).

Correlation between antioxidant capacity with concentrations of phenolic compounds was analyzed in basil plants. In green basil plants, concentrations of anthocyanin, phenolics, and flavonoids were all positively related to antioxidant capacity (Figure 6A). In purple basil plants, concentrations of phenolics and flavonoids were positively related to antioxidant capacity, while anthocyanin concentration showed no relationship (*p* = 0.8812) (Figure 6B).

**Figure 5.** Correlation between antioxidant capacity of green basil 'Improved Genovese Compact' and purple basil 'Red Rubin' plants with UV-B radiation doses. Correlation test was conducted in three terms according to different UV-B radiation patterns, five supplemental UV-B radiation treatments including control, 1H2D, 2H2D, 1H5D, 2H5D (**A**), control and 1 h·d−<sup>1</sup> UV-B radiation treatments (**B**), and control and 2 h·d−<sup>1</sup> UV-B radiation treatments (**C**). Data were pooled from two photosynthetic photon flux density (PPFD) treatments. Means followed by the same lower/upper case letters are not significantly different for green/purple basil plants, according to Student's *t* mean comparison (*p* < 0.05). Bars represent standard errors. Dashed lines show the regression between antioxidant capacity with supplemental UV-B radiation dose, according to the pairwise correlation method.

**Figure 6.** Correlation between antioxidant capacity and concentrations of anthocyanin, phenolics, and flavonoids in green basil plants (**A**), and purple basil plants (**B**). Dashed lines show the regression between concentrations of phenolic compounds with antioxidant capacity according to Pairwise Correlation method.

#### **4. Discussion**

#### *4.1. Impacts of UV-B and PPFD on Photosynthesis, Relative Chlorophyll Concentration, and Chlorophyll Fluorescence*

Photosynthesis is one of the most sensitive metabolic processes in plants responding to environmental condition changes, such as supplemental UV-B radiation and PPFD. In the present study, Pn in basil leaves was lower after UV-B radiation, which was mainly caused by the direct damage of PSII components and led to reduced photosynthetic capacity, subsequently decreased Gs [32–34]. Meanwhile, relative chlorophyll content in basil leaves was also lower after UV-B radiation, either through degradation or inhibition of enzymes involved in the chlorophyll biosynthetic pathways [34]. However, compared to depressed photosynthesis and reduced chlorophyll content by supplemental UV-B radiation in our study, a meta-analysis of field studies (more than 450 reports from 62 papers) reported unaffected photosynthesis and chlorophyll content after supplemental UV-B radiation [35]. Differences between our study (controlled environment with artificial lighting) from previous field studies (sunlight) probably resulted from significantly low PPFDs and relatively high UV-B proportion used in our study. Firstly, in controlled environment systems, due to the high cost of powering artificial lighting, lower PPFDs are normally used compared to that of sunlight intensity in an open field. Subsequently, lower PPFDs resulted in depressed photochemical protection system of plants, such as decreased photosynthetic capacity, decreased leaf thickness, and reduced concentrations of UV-absorbing agents [21], which aggravated the negative effects caused by UV-B radiation. Secondly, the damage caused by UV-radiation increases with decreasing UV wavelength, since short UV wavelength has more energy than long UV wavelength [36]. The UV component of sunlight consists of 95% UV-A and 5% UV-B, of which the small portion UV-B radiation shows stronger mutagenic and carcinogenic effects compared to UV-A radiation [36,37]. For example, a less prominent and less long-lasting activation of p53 gene ("guardian of the genome") after UV-A radiation compared to UV-B was observed, suggesting stronger effects of UV-B radiation than UV-A [36]. In the present study, the UV radiation provided by broadband UV-B lamps was mainly UV-B radiation with relatively low UV-A radiation, contributing to aggravated negative effects on plant photosynthesis compared to previous field studies, of which mainly consists of UV-A radiation.

Chlorophyll fluorescence parameters provide precise and objective information with regard to photochemical efficiency and non-photochemical de-excitation involved in the conversion of light energy under different conditions [28,38]. The less reduced Fv/Fm, PI ABS, TR0/CS, and ET0/CS after UV-B radiation in purple basil plants than green basil plants clearly indicate that purple basil plants are more tolerant to UV-B radiation, resulted from its improved capacity to process excess UV-B energy through PSII [39]. Meanwhile, the uninfluenced DI0/CS under UV-B treatments in green basil plants suggests its inability to dissipate absorbed UV-B energy in the form of harmless heat, even under the smallest UV-B radiation dose, 16.0 <sup>μ</sup>mol·m−2·s−<sup>1</sup> at 1 h·d−<sup>1</sup> for 2 days, while purple basil plants coped with excess UV-B energy by increasing heat dissipation. Mosadegh et al. (2018) also reported that the DI0/CS of green basil plants was not affected after 2-weeks UV-B radiation at 68 and 102 kJ·m−2·d−1, confirming that green basil plants failed to dissipate UV-B energy as harmless heat [28]. Differences in chlorophyll fluorescence parameters between green and purple basil plants may be due to the relatively higher concentrations of UV-protective antioxidants in purple basil plants such as anthocyanins, phenolics, and flavonoids, which are known to provide plants with strong protection from excess UV-B energy [40].

In our previous study, the gas exchange rate in green basil plants was positively correlated with PPFD [21], while it was not affected in the present study. This may be due to the large variation of Pn, E, and Gs caused by UV-B radiation at each PPFD. In green basil plants, Pn ranged from 3.7 to 12.6 <sup>μ</sup>mol·m−2·s−<sup>1</sup> at low PPFD (160 <sup>μ</sup>mol·m−2·s<sup>−</sup>1), and ranged from 4.8 to 13.8 <sup>μ</sup>mol·m−2·s−<sup>1</sup> at high PPFD (224 <sup>μ</sup>mol·m−2·s−1). Also, it was observed that the Pn in purple basil plants was much lower compared to the Pn in green basil plants. One hypothesis is that the differences between two cultivars is due to the lower quantum efficiency of photosynthetically active radiation (PAR) in purple basil plants compared to green basil plants. In purple basil plants, the relatively high concentration of anthocyanins and flavonoids absorbs more PAR light, which decreases the absorption of PAR light by chloroplasts and subsequently decreases the photochemistry energy transferred to reaction centers, resulting in decreased Pn in purple basil plants compared to green basil plants [41].

#### *4.2. Impacts of UV-B and PPFD on Growth and Yield*

Plant leaf expansion is invariably inhibited by supplemental UV-B radiation and other leaf morphogenesis changes such as reduced leaf area, increased leaf thickness, and accumulation of leaf surface waxes are also observed across a range of plant species [14,42,43]. Internode length is also a very sensitive growth parameter that responds to UV-B radiation [44]. Kaiserli (2018) reported that most cell-wall elongation genes induced by BRI1-EMS-SUPPRESSOR 1 (BES1) are negatively regulated by UV-B radiation [45]. Meanwhile, the biosynthesis and signaling of plant growth hormone auxin, a key regulator of stem elongation, was also suppressed in arabidopsis (*Arabidopsis thaliana*) and coriander (*Coriandrum sativum*) plants after UV-B radiation, thereby reducing plant stem elongation and promoting a compact phenotype [46]. In the present study, similar results such as reduced leaf area, increased leaf thickness, accumulation of leaf surface waxes, and reduced leaf internode length were observed, which are plant acclimation responses to supplemental UV-B radiation. In addition to protecting plants from receiving excess UV-B energy, these acclimation responses also provide plants with improved tolerance to other adverse environmental conditions, such as heat stress and mechanical handling during postharvest [6,47,48].

Reduced gas exchange rate and leaf expansion, and inhibition of stem elongation of basil plants under supplemental UV-B radiation resulted in a reduction in plant size and yield. The greater yield reduction by the UV-B radiation under high PPFD than low PPFD may be due to its taller plants, which shortened the distance between basil plants and UV-B light tube, resulting in increased UV-B radiation intensity sustained by basil plants, and subsequently severer yield reduction.

#### *4.3. Impacts of UV-B and PPFD on Phytochemical Accumulation and Antioxidant Capacity*

Across a range of plant species, phenolic compounds, especially flavonoids, act as efficient UV-screening agents to reduce excess UV light received by photosynthetic tissues to protect plants from possible harm [40,49]. Enhanced accumulation of phenolic compounds by supplemental UV-B radiation has been supported by a large body of experimental evidence [50,51], which was confirmed in this study. Ghasemzadeh et al. (2016) reported that total phenolic and flavonoid content in green basil plants increased by 16% and 85%, respectively, after a 13 kJ·m−2·h−<sup>1</sup> post-harvest UV-B radiation for 4–10 h, but anthocyanin content was not measured [52]. It was also reported that upon supplemental UV-B radiation, the gene expression of phenylalanine ammonia lyase (PAL) and chalcone synthase (CHS), two key molecular markers for phenolic compounds biosynthesis increased significantly [46,53]. Noticeably, in the present study, the enhancement of flavonoids and phenolics by UV-B radiation was much greater than anthocyanin. Consistently, antioxidant capacity was significantly correlated with concentrations of phenolics and flavonoids in both basil cultivars, while marginally or not correlated to anthocyanin concentration. This might be due to the higher ROS-scavenging capacity of phenolics and flavonoids than anthocyanins, resulting in more sensitive reactions of phenolics and flavonoids to UV-B radiation [54]. Csepregi et al. (2017) also reported such differential regulation of different phenolic compounds by UV-B radiation, in which quercetins with additional hydroxyl group on ring-B increased up to 10 folds while kaempferol increased 3–4 fold, due to their different ROS-scavenging capacity [55].

Enhancement of phenolic compounds after UV-B radiation was greater in basil plants grown under low PPFD compared to those grown under high PPFD, indicating basil plants are more sensitive to UV-B radiation under low PPFD. In a similar way, Behn et al. (2010) reported that under low PPFD (550 <sup>μ</sup>mol·m−2·s<sup>−</sup>1), essential oil quality in peppermint plants was improved in terms of an enhanced menthone to menthol conversion after UV-B radiation, while not affected by UV-B treatment under high PPFD (1150 <sup>μ</sup>mol·m−2·s−1) [56]. As mentioned, this may be due to a depressed photochemical and biochemical protection system of plants grown under low PPFD, such as lower leaf thickness and reduced concentrations of UV-absorbing agents [21]. As we hypothesized, concentrations of phenolic compounds in basil plants grown under low PPFD with UV-B radiation was significantly higher compared to those of plants grown under high PPFD without UV-B radiation, suggesting that UV-B radiation could be used as a tool to compensate for reduced accumulation of phenolic compounds in basil plants grown under controlled environment.

Similar to plant responses on chlorophyll fluorescence, different responses in phytochemical accumulation between green and purple basil plants were also observed. Specifically, purple basil plants showed fewer biochemical changes than green basil plants after UV-B radiation, which performed as unaffected anthocyanin concentration and less induction of phenolics and flavonoids. Our hypothesis is that the relatively high concentrations of phenolic compounds in purple basil plants act as potent UV-screening agents as well as free-radical scavengers to protect purple basil plants from excess UV-B light. Under high PPFD without UV-B treatment, concentrations of anthocyanin, phenolics, and flavonoids and antioxidant capacity in purple basil leaves were 3.33, 1.47, 1.93, 3.72 times those in green basil leaves, respectively. This hypothesis was confirmed by Tattini et al. (2014), in which he reported that purple basil 'Red Rubin' showed lower metabolic cost of photoprotective mechanisms than green basil 'Tigullio' when being moved from 30% to 100% sunlight condition [57].

## *4.4. Impacts of UV-B Radiation Doses and Radiation Patterns on Phytochemical Accumulation and Antioxidant Capacity*

With the radiation doses and different radiation patterns used in the present study, green basil plants were more dose-dependent, while purple basil plants were both dose-dependent and radiation pattern-dependent. Antioxidant capacity in green basil plants was significantly correlated with the UV-B radiation dose for both 1 h·d−<sup>1</sup> and 2 h·d−<sup>1</sup> UV-B radiation patterns, while antioxidant capacity in purple basil plants was not affected by 1 h·d−<sup>1</sup> UV-B radiation treatments. With the similar UV-B radiation dose (1H5D and 2H2D treatments), after 1 h·d−<sup>1</sup> UV-B radiation treatments, the recovery time until next day treatment (23 h) allowed purple basil plants' signaling and metabolic adaptation to (at least partially) reset to pre-stress level, without increasing phenolic compounds accumulation, while after 2 h·d−<sup>1</sup> UV-B radiation (recovery time of 22 h until next treatment), purple basil plants failed to recover from UV-B radiation stress and resulted in an overall increase of phenolic compounds to cope with excess UV-B energy. This indicated that radiation patterns play an important role in regulating purple basil responses to UV-B radiation, while radiation dose is the determining factor in regulating green basil biochemical responses. Mosadegh et al. (2018) also reported that with the same UV-B radiation dose of 102 kJ·m−2, phenolics concentration of green basil 'Genovese' was the same level regardless of UV-B radiation pattern, continuous 1-d UV-B radiation or discontinuous 6-d UV-B radiation [28]. However, at lower UV-B radiation doses of 8.5, 34, and 68 kJ·m−2, when 'Genovese' green basil plants were treated with the same UV-B radiation dose, continuous 1-d UV-B radiation resulted in significant higher phenolics concentration compared to plants treated with discontinuous 6-d UV-B radiation [28]. Thus, plant responses to UV-B radiation in green basil plants may also depend on radiation patterns, which are affected by the total UV-B radiation dose.

#### *4.5. Implications of Study Findings*

Different plant responses to UV-B radiation are observed in studies conducted in the open field with sunlight than in a controlled environment with artificial lighting, due to different PPFDs and components of UV radiation [13,35,58,59]. The novel finding of the present study is that plants grown under a controlled environment with lower PPFDs are more sensitive to UV-B radiation. Therefore, for future studies under a controlled environment, a lower UV-B radiation dose should be applied to reduce its negative effects on plant photosynthesis, growth, or yield. Furthermore, we see differential

responses in green and purple basil plants to UV-B radiation doses and radiation patterns. Therefore, to better understand plant responses to supplemental UV-B radiation, more plant species/cultivars, lower radiation doses, and different radiation patterns need to be investigated in future studies.

Plant acclimation responses to supplemental UV-B radiation lead to plant cross-protection against other environmental stresses, through photochemical, morphological, and biochemical mechanisms [60]. For example, *UVR8* was recently shown to be involved in regulating thermomorphogenesis, shade-avoidance responses, and plant immunity, underlining the importance of signaling crosstalk among UV-B radiation, hormone, and defense pathways [47,61]. As a result, supplemental UV-B radiation could be used as a tool to improve plant tolerance to other adverse environmental conditions, and interactions between supplemental UV-B radiation and other key environmental factors still need to be studied.

## **5. Conclusions**

Results of the present study suggest that a short period of pre-harvest supplemental UV-B radiation could significantly improve phytochemical concentrations in basil plants, and plant responses to UV-B radiation vary among plant cultivars, radiation doses, and radiation patterns. Meanwhile, effects of UV-B radiation on basil plants interacted with PPFDs used in the cultivation system, and high PPFD improved plant tolerance to UV-B radiation. Also, supplemental UV-B radiation could compensate for the reduced accumulation of phenolic compounds in basil plants grown under low PPFD. Therefore, combining plant growth performance, yield, and accumulation of health-promoting phenolic compounds, a pre-harvest UV-B radiation of 1 h·d−<sup>1</sup> for 2 days under a PPFD of 224 <sup>μ</sup>mol·m−2·s−<sup>1</sup> was recommended for green basil 'Improved Genovese Compact' production under a controlled environment. However, supplemental UV-B radiation doses used in this study decreased the total amount of phenolic compounds in purple basil plants due to yield reduction, and UV-B radiation is not recommended for purple basil 'Red Rubin' production under a controlled environment.

**Author Contributions:** Conceptualization, H.D., G.N. and M.G.; data analysis, H.D.; methodology, H.D., G.N. and M.G.; writing—original draft preparation, H.D.; writing—review and editing, G.N., M.G.

**Funding:** This research was partially funded by the USDA National Institute of Food and Agriculture Hatch project TEX090450, Texas A&M AgriLife Research, and Texas A&M AgriLife Extension Service.

**Acknowledgments:** The authors appreciate the assistance from Youping Sun, Zhanyang Xu, Christina Perez, Triston Hooks, and the student workers at Texas A&M AgriLife Research Center at El Paso, TX.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*

## **Comparative Study of Three Low-Tech Soilless Systems for the Cultivation of Geranium (***Pelargonium zonale***): A Commercial Quality Assessment**

## **Luca Brentari 1, Nicola Michelon 2,\*, Giorgio Gianquinto 2, Francesco Orsini 2, Federico Zamboni <sup>3</sup> and Duilio Porro <sup>1</sup>**


Received: 25 August 2020; Accepted: 18 September 2020; Published: 20 September 2020

**Abstract:** The study evaluated the feasibility of simplified hydroponics for the growth of rooted cuttings of geranium (*Pelargonium zonale*) for commercial purposes in local farms in Northern Italy. Tested systems included a control where soilless system on substrate (peat) (T-1), usually adopted by local farmers, was compared against an open-cycle drip system on substrate (peat) (T-2), and a Nutrient Film Technique system (T-3). For commercial features, assessed parameters included flowering degree (flowering timing, numbers of inflorescences plant−1, and number of flowers inflorescence−1), numbers of leaves plant−1, number of branches plant−1, final height of plant, and the aesthetic-commercial assessment index. Assessed parameters also included fresh and dry weight, SPAD Index, the water consumption, and the water use efficiency (WUE). The soilless systems typology significantly affected rooted cuttings growth, commercial features, and WUE. The adoption of an open-cycle drip system (T-2) resulted in a significant improvement of all the crop commercial characteristics as compared with other treatments, making plants more attractive for the market. The water consumption was higher in T-2 as compared with T-1 and T-3, but it allowed for the highest fresh weight, and therefore also the highest WUE. The results indicate that the typology of soilless system significantly enhances the commercial characteristics of geranium.

**Keywords:** *Pelargonium zonale*; low-tech soilless cultivation system; commercial quality

## **1. Introduction**

The global gardening pots market size was valued at USD 1.7 billion in 2018. A growing interest in gardening is expected to remain a favorable factor for industry growth [1]. In Italy, the cultivation of cut flowers and potted ornamentals in both greenhouses and open field accounts for a relevant share of the market. In 2017, out of the 2.5 billion euros associated with the national floricultural and ornamental crop sector, about 1.15 billion euros are associated with flower production and potted plants. The Italian cut flowers and potted ornamental sector accounts for 27,000 companies, 100,000 workers, and almost 29,000 ha of farmland. When considering only figures for the ornamental seedling production, 2000 farms for a total area of 1500 ha are also found in Italy [2].

Ornamental plants are typically characterized by a fast growth rate and a large consumption of both nutrients and water, which should be of elevated quality given the limited salt tolerance

of these plant species [3]. Furthermore, farmers generally tend to overwater these crops, with the consequence that ornamental plants generally present low water use efficiency (WUE) values. Accordingly, despite the existing variability among ornamental species in terms of water and fertilizer requirements [4], the sector generally accounts for high environmental impact due to the losses of both water and fertilizers [3]. In this scenario, the increasing awareness of environmental pollution caused by agriculture, the scarcity of resources such as water, the need to reduce the production costs, and the growing demand for healthy foods are forcing operators to move towards more sustainable cropping techniques. In greenhouse cultivation, the adoption of soilless culture, coupled with techniques such as fertigation, drip irrigation, integrated plant protection, and climate control, can provide a high-quality product with efficient use of resources, e.g., water, while also increasing the potential yield [5–7] as well as decreasing nutrient losses [8]. Soilless culture can be defined as "any method of growing plants without the use of soil as a rooting medium, in which the inorganic nutrients absorbed by the roots are supplied via the irrigation water" [9]. The soilless systems are classified according to the presence and type of substrate, to the irrigation system, and to the nutrient solution (NS) management, namely the reuse or not of the leaching fraction [9], which results, respectively, in the so-called "closed" or "open" loop systems. In open-loop systems, an excessive amount of NS (120–150% of actual water requirements) is supplied to avoid salt accumulation in the substrate, and the leaching fraction is not reused and commonly released into the environment. In closed soilless systems, on the other hand, water supply is generally higher (150–170% of daily water requirements), but the leaching fraction is reused after being disinfected [10]. Accordingly, water and fertilizer saving (which has both environmental and economic benefits), are the main advantages of closed systems. However, closed systems require more complicated NS management that ultimately results in higher equipment and management costs. Primarily, the higher risks of pest outbreak (mainly root diseases) require the disinfection of the leached fraction. Moreover, controlling nutrients and non-essential ions in the recirculating NS becomes more difficult, especially in the case of saline waters or high concentration of non-essential or scarcely absorbed ions (e.g., Na<sup>+</sup> and Cl−) [10]. Overall, it is acknowledged that closed systems show a better WUE, at the expenses of possible yield decay in response to salt build-up in the root zone as compared to open systems [5,11–16]. Classification of soilless systems may also be done according to those that feature the presence of a solid inorganic or organic medium, which offers support to the plants and systems without substrate (water based soilless systems), where the bare roots of plants lie directly in the NS [9]. Different features characterize the two groups of systems. Soilless systems on substrate are surely the most popular systems for cut flowers and pot ornamentals [17]. Water-based soilless systems are, on the other hand, associated with reduced environmental impact and costs related to the substrate disposal. However, in water-based soilless systems, the resilience to stresses (e.g., drought) is affected by the absence of a buffer offered by the substrate, and a considerably higher risk of outbreak of root-borne diseases may also be experienced [18].

In soilless systems, fertilization is performed administering a NS containing macro- and micro-nutrients, generally through different types of irrigation systems (drip irrigation, sub-irrigation, or overhead system). Such fertigation can be continuous or discontinuous.

Based on these assumptions, the current research comparatively assessed three low-tech soilless systems for the cultivation of geranium (*Pelargonium zonale*), targeting the identification of the system that would allow for optimal commercial production and improvement of WUE.

#### **2. Materials and Methods**

#### *2.1. Location*

The experiment was conducted in a greenhouse covered with polyethylene within a commercial farm located in Vigolo Vattaro, Province of Trento, Northern Italy, 46◦00 N, 11◦19 E, at an altitude of 725 m a.s.l. Plants were grown under natural light conditions. The local climate, according to Köppen's classification, is Cfb type [19], which is a mesothermic climate, with the absence of a dry season and cool summer with temperature during the hottest month falling below 22 ◦C. The experiment was conducted from 25 March 2017 to 2 June 2017.

*2.2. Treatments and Experimental Design*

Three low-tech soilless systems were compared:


In each of the three treatments, all 90 rooted cuttings were arranged in rows with 10 plants each, with 15 cm between rooted cuttings and 42 cm between rows, resulting in a planting density of about 16 plants m<sup>−</sup>2, following common commercial practices. Three replicate plots for each treatment (rows), composed of ten rooted cuttings each (*n* = 30), were arranged in a randomized complete block design.

**Figure 1.** T1, farm system with substrate. Schematic representation of the growing system used.

**Figure 2.** T-2, open-cycle drip system with substrate. Schematic representation of the growing system used.

**Figure 3.** T-3, Nutrient Film Technique system. Schematic representation of the growing system used.

#### *2.3. Plant Material and Crop Management*

At the beginning of the trial, rooted cuttings were selected to have uniform plant material (4 cm height and 3 leaves) among the 90 individual plants used for the experimentation. T-1 (control) was managed following traditional practices from local farmers. It was irrigated only with water once every 2 days from the 1st to the 7th week, and once a day from 8th to 10th week, by hand. T-1 was fertilized only three times (discontinuous fertigation, on 4 April, 22 April, and 12 May), with a granular fertilizer (Manna Lin A, Mannafert V., Bolzano, Italy). Granular fertilizer was solubilized in water to have a 2.08 g·L−<sup>1</sup> concentration for a total amount of 50.50 g applied. Manna Lin A is composed of 7% N-NO3, 13% N-NH4, 5% P2O5, 10% K2O, 2% MgO, 0.025% B, 0.005% Cu, 0.06% Fe, 0.025% Mn, 0.0025% Mo, and 0.02% Zn. The microelements were supplied as chelates.

Unlike T-1, in T-2 and T-3, a continuous fertigation was adopted, using the same NS. The composition of macronutrients of full strength NS was: 10.00 mM NO3 <sup>−</sup>, 1.00 mM NH4 <sup>+</sup>, 2.00 mM H2PO4 <sup>−</sup>, 5.01 mM K+, 4.00 mM Ca2+, 1.50 mM Mg2<sup>+</sup>, and 3.53 mM SO4 <sup>2</sup>−. A mixed fertilizer for micronutrients was used, with the following full strength NS: 20.00 μM Fe3+, 0.63 μM Cu2+, 4.29 μM Zn2<sup>+</sup>, 13.88 μM B3<sup>+</sup>, 19.66 μM Mn2<sup>+</sup>, and 0.42 μM Mo6<sup>+</sup>. For all fertigation treatments, NS was prepared using fresh water (pH = 8.00, EC = 359 <sup>μ</sup>S·cm−<sup>1</sup> at 20 ◦C). The final EC of full strength NS ranged between 1829 and 1963 <sup>μ</sup>S·cm−<sup>1</sup> and pH ranged between 5.5 and 6.2. During the first week, in T-2 and T-3, a lower strength NS for macronutrients was used (T-2 top-fertilized by watering can) (EC = 1021 <sup>μ</sup>S·cm−1, pH = 5.5, 5.4 mM NO3 <sup>−</sup>, 0.50 mM NH4 <sup>+</sup>, 1.0 mM H2PO4 <sup>−</sup>, 2.5 mM K+, 2.0 mM Ca2<sup>+</sup>, 0.97 mM Mg2<sup>+</sup>, and 0.75 mM SO4 <sup>2</sup>−) to allow the roots to adapt to the new growing environment before using the full strength NS. The EC of leaching fraction was measured every week in both T-2 and T-3 treatment.

T-2 fertigation scheduling took into account the leaching fraction measurement, having drainage around 30% per day as a target. It changed during the crop cycle and ranged from 1 irrigation every 3–5 days at the beginning to 2 irrigations per day at the end of the trial. The NS volume provided for all pots ranged from 3.6 L during the 2nd week to 78.9 L during the 10th week, corresponding to the flowering stage. In T-3, the NS was continuously supplied from sunrise to sunset, by submerged adjustable flow pump with a measured flow rate for every plastic trough of 1.83 L·min<sup>−</sup>1.

Inside the greenhouse, temperature and relative humidity were monitored every 15 min by GEMINI data logger Tinytag Plus 2. The greenhouse temperature ranged between 12 and 33 ◦C, and day/night humidity from 30% to 85%, respectively.

#### *2.4. Sampling and Analysis*

In the first week, EC, pH, the drained volume of T-2, volume of leftover NS in the 210 L reservoir tank of T-3, its EC and pH were daily measured after the sunset. During the trial, on 22 April and on 13 May, 100 L of fresh NS each were added to the T-2 and T-3 reservoir tanks. EC, pH, and total volume were measured again after the additions. From the 2nd to 10th week, all these parameters were measured weekly. EC was measured by Adwa AD31 Waterproof EC/TDS Tester and pH was measured by Artiglass IP67 pocket pH Tester. All testers were weekly calibrated.

Progressive and final plant heights, determined as the distance from the surface of the medium to the top of the plant, for all 90 rooted cuttings were measured. To evaluate the flowering timing and its quality, the starting date of appearance of inflorescences and their numbers per plant, together with dates of beginning and full flowering, were recorded in ten plants per replicate. Furthermore, in three plants per replicate, weekly counts of the number of fully-grown leaves was performed, as well as counts of the number of flowers of the first inflorescence and number of branches.

The estimation of leaf chlorophyll concentration was performed at the end of the trial through a non-destructive measurement with SPAD-502 (Konica-Minolta, Tokyo, Japan). Measures were taken on the leaf nearby the oldest inflorescence from 10 plants per replicate. The Minolta SPAD meter (Soil Plant Analysis Development) used indirectly measures chlorophyll content in a non-destructive manner. SPAD values were determined by measuring the ratio of light transmitted through the leaf at a red wavelength (650 nm) and an infrared wavelength (940 nm).

At the end of the experiment, all 90 plants were divided into leaves (leaves with petioles), trunks, roots, and inflorescences. Plant organs were weighed for the fresh and dry weight determinations after drying in a ventilated oven at 105 ◦C for 48 h.

Furthermore, at the end of the experiments, an aesthetic-commercial assessment of 10 plants per replicate was performed. For each rooted cutting, three parameters (vegetative growth, foliage compactness, and general aspect) were evaluated by assigning a score from 1 to 5, with the score 3 being the threshold value for marketability. Whenever at least one of the three parameters received a score below 3, the rooted cutting was evaluated as not marketable. The aesthetic-commercial assessment was performed by the local farmer in a randomized way without being aware of the specific treatments.

#### *2.5. Statistical Analysis*

For phenological data regarding the flowering degree (appearance of inflorescences, flowering start, and full flowering) and fully-grown leaves, no statistical analysis was applied, but only kinetic behaviors in relation to the treatment were shown. Analysis of variance (ANOVA) was used to determine the effect of the growing system used on the number of inflorescences plant−1, number of flowers inflorescence<sup>−</sup>1, final height of the plant, number of branches plant−1, SPAD index, water contents of organs, aesthetic commercial assessments, fresh weight, total water consumption, and water use efficiency. All data were statistically processed using Systat software package (Systat Software 9.0, San Jose, CA, USA).

#### **3. Results**

#### *3.1. Climate and Nutrient Solution Monitoring during the Experiment*

During the experiment, inside the greenhouse, a data logger was used to measure temperature and humidity every fifteen minutes. Maximum air temperature ranged between 14.7 and 39.0 ◦C, with an average of 32.2 ◦C. Minimum air temperature ranged between 8.2 and 20.7 ◦C, with an average of 12.1 ◦C. The average daily temperature was 18.7 ◦C. The daily maximum relative humidity ranged between a minimum of 50.0% and a maximum of 93.3%, with an average of 83.9%. The daily minimum relative humidity ranged between a minimum of 16.5% and a maximum of 85.6%, with an average of 31.7%. The average daily humidity was 65.3%. During the experiment, the mean value of daily global radiation, outside the greenhouse, was 15.87 MJ·m−2·day−<sup>1</sup> in April and 20.25 MJ·m−2·day−<sup>1</sup> in May.

A NS was applied only in T-2 and T-3 treatments, with periodical control of both EC and pH. During the first week, in which a lower strength NS (EC = 1021 <sup>μ</sup>S·cm−<sup>1</sup> and pH = 5.5) was used, EC of the leaching fraction ranged between 1026 and 1040 <sup>μ</sup>S·cm−1, while the pH ranged between 5.5 and 5.8. From the second week, when a full strength NS (EC ranged 1829–1963 <sup>μ</sup>S·cm−<sup>1</sup> and pH ranged 5.5–6.2) was used, EC of leaching fraction ranged between 2171 and 3923 <sup>μ</sup>S·cm<sup>−</sup>1, while the pH ranged between 5.8 and 6.5.

#### *3.2. Date of the Appearance of Inflorescences*

The starting date of appearance of inflorescences was not affected by the soilless systems (16–17 days after transplanting) (Figure 4a). Concurrently, T-2 and T-3 showed a more extended period (3–4 days) to conclude this phase as compared to T-1 (Figure 4a).

**Figure 4.** Effect of growing systems on *Pelargonium zonale*: (**a**) plants with inflorescences just visible; (**b**) plants at flowering start phase; (**c**) plants at full flowering phase; and (**d**) leaf number. T-1, farm system with substrate; T-2, open-cycle drip system with substrate; T-3, Nutrient Film Technique; DAT, Days After Transplanting.

#### *3.3. Date of Flowering Start*

The three soilless systems affected the date when the flowering started. As compared with T-1, flowering started six and four days earlier, respectively, in T-3 and T-2 (Figure 4b). Flowering was concluded between 64 and 66 days in all treatments, independently from the growing system.

#### *3.4. Date of Full Flowering*

There were no differences between T-1 and T-2 treatments in terms of date of starting of the full flowering phase (58 days after transplanting) and flowering duration (14 days) (Figure 4c). Conversely, full flowering was anticipated by about six days and lasted five days longer under the T-3 treatment (Figure 4c).

#### *3.5. Biometrical Parameters*

All biometrical parameters were significantly affected by treatments (Table 1), with highest values always associated with T-2 and lowest values found in plants grown under T-3. The plants' height was also affected by treatment (Table 1 and Figure 5): at the final assessment, T-2 had higher values than T-1, which in turn was significantly higher than T-3.

**Table 1.** Mean *Pelargonium zonale* biometrical responses to growing systems. Within-columns mean values followed by different letters are significantly different by Tukey test.


With significance (\*\*\*) for *p* ≤ 0.001. T-1, farm system with substrate; T-2, open-cycle drip system with substrate; T-3, Nutrient Film Technique.

**Figure 5.** Height of *Pelargonium zonale* plant during growing period in response to the growing system used. Mean values ± standard error. T-1, farm system with substrate; T-2, open-cycle drip system with substrate; T-3, Nutrient Film Technique.

#### *3.6. Number of Leaves*

The three soilless systems affected the number of fully-grown leaves, which was the highest in T-2 (Figure 4d). In general, T-1 and T-2 treatments resulted in a different crop kinetic behavior as compared to T-3.

#### *3.7. Leaf Chlorophyll*

Leaf greenness of plants (SPAD values) was affected by treatment. Higher SPAD values were detected in rooted cuttings grown in T-3 treatment as compared with plants grown in T-1 and T-2 (Table 2).

**Table 2.** SPAD values of *Pelargonium zonale* in response to the growing system used. Within-columns mean values followed by different letters are significantly different by Tukey test.


With significance at *p* ≤ 0.001 (\*\*\*). T-1, farm system with substrate; T-2, open-cycle drip system with substrate; T-3, Nutrient Film Technique.

#### *3.8. Fresh and Dry Weight*

The three soilless systems affected both fresh (Figure 6a) and dry (Figure 6b) weight, which were the highest in T-2 and the lowest in T-3 and T-1. Leaves, flowers, and branches fresh weights had similar behavior of total biomass, presenting highest values in T-2 as compared with T-1 and T-3, while the fresh weight of roots was not affected by treatment (Figure 6a). Among dry weights (Figure 6b), different behaviors were found across organs. Plants grown under T-2 presented the highest leaf and flower dry biomass as compared to T-1 and T-3. On the other hand, higher dry biomass of both branches and roots was associated with T-1 and T-2 as compared with T-3.

**Figure 6.** Effect of growing system used on *Pelargonium zonale*. (**a**) Plan fresh weight (g plant<sup>−</sup>1) in relation to growing system and relative partitioning into different organs. Means values ± standard error for total biomass. (**b**) Plan dry weight (g plant<sup>−</sup>1) in relation to growing system and relative partitioning into different organs. Mean values ± standard error for total biomass. T-1, farm system with substrate; T-2, open-cycle drip system with substrate; T-3, Nutrient Film Technique.

Treatment affected the water contents of different organs (Table 3): in particular, T-3 showed higher values than T-1, except for flowers where the difference was not significant. T-1 always presented the lowest levels, while T-2 had an intermediate behavior, with high values for both flowers and branches and low value for roots. It is interesting to note that roots of T-3 plants showed the highest values of water contents.

**Table 3.** Water contents (**%**) of *Pelargonium zonale* organs in response to the growing system used. Within-columns mean values followed by different letters are significantly different by Tukey test.


n.s., not significant; \* significance for *p* ≤ 0.050 and *p* ≥ 0.010; \*\*\* significance for *p* < 0.001. T-1, farm system with substrate; T-2, open-cycle drip system with substrate; T-3, Nutrient Film Technique.

#### *3.9. Aesthetic-Commercial Assessment*

T-3 always had the lowest values for all investigated parameters. T-2 showed the best scores in all the parameters evaluated, except for the vegetative growth, in which T-1 had the highest score as the absolute value, even if statistical analysis did not detect any significant difference as compared to T-2 (Table 4).

**Table 4.** Aesthetic-commercial assessment of *Pelargonium zonale* in response to the growing system used. Within-columns mean values followed by different letters are significantly different by Tukey test.


With significance for *p* < 0.001 (\*\*\*). MV is the arithmetic mean among vegetative growth, foliage compactness, and general aspect values. T-1, farm system with substrate; T-2, open-cycle drip system with substrate; T-3, Nutrient Film Technique.

#### *3.10. Water Consumption and WUE*

The comparison of the three treatments revealed significant differences as regards the biomass produced, the water consumption (leaching fraction included), and the related WUE values (Table 5). In particular, T-2 differed from the other two for its most considerable vegetative development. Total water consumption revealed significantly different values among the three treatments: T-2 showed the highest values, T-3 the intermediate ones, and T-1 the lowest ones. The calculated values of WUE, therefore, showed the highest values in T-2, and the lowest ones in T-3.

**Table 5.** Total biomass (fresh weight), total water consumption and WUE of *Pelargonium zonale* in response to the growing system used. Within-columns mean values followed by different letters are significantly different by Tukey test.


With significance for *p* < 0.001 (\*\*\*). T-1, farm system with substrate; T-2, open-cycle drip system with substrate; T-3, Nutrient Film Technique; FW, Fresh Weight; TWC, Total Water Consumption; WUE, Water Use Efficiency.

#### **4. Discussion**

The adoption of different soilless cultivation systems significantly affected the growth (including flowering, fresh weight, and dry weight) and the commercial characteristics (number of inflorescences per plant, number of flower per inflorescence, number of branches per plant, and number of leaves) of geranium grown in a greenhouse, in Northern Italy. As also reported by Rouphael and Colla [20], the optimal concentration of fertilizer solutions for greenhouse crops may be affected by irrigation method, because it influences the accumulation of nutrients in the growing medium, which in turn affects the nutrient uptake by plants. For example, Cardarelli [21] reported that, when averaged over NS concentration, the number of geranium flowers per plant was significantly (27%) higher with sub-irrigation than with drip-irrigation.

The growth of rooted cuttings of geranium continues until full bloom, when they are ready for sale. Given the characteristics of the market in the area where the trial was conducted, where farmers generally supply local retailers, a gradual flowering could help producers. However, T-1 and T-2 showed no differences (Figure 4), displaying both a 14-day flowering window (from 23 May to 5 June). T-3, on the other hand, showed a much more scalar flowering (20 days, from 17 May to 5 June) and earlier than T-1 and T-2 (Figure 4). Despite this, T-3 did not develop adequate commercial characteristics for the market. Furthermore, some of the rooted cuttings of T-3 treatments highlighted a delayed growth demonstrating stress conditions, which may have caused the observed flowering pattern. In fact, according to Riga [22], stress conditions in geranium can influence the flowering timing, anticipating the opening of the flowers.

The application of different cultivation systems, to which three different fertigation managements are associated, significantly affected all the plant commercial features. T-2 showed the highest number of inflorescences (13.67 inflorescences plant<sup>−</sup>1), followed by T-1 and T-3, where 10.63 and 6.00 inflorescences plant−<sup>1</sup> were observed, respectively (Table 1). T-2 showed significant differences from both T-1 and T-3, and T-1 from T-3. T-2 also showed the highest number of flowers per inflorescence (143.11 flowers inflorescence−1), followed by T-1 and T-3, where values of 96.33 and 64.67 flowers inflorescence−<sup>1</sup> were observed, respectively. No significant differences were observed by comparing T-1 and T-3 (Table 1). Regarding the vegetative behavior (number of fully grown leaves at the end of the experiment, number of branches plant−1, and final height of plants), as well as considered commercial characteristics, T-2 always showed the highest values, demonstrating a better efficiency of this treatment, as also confirmed by fresh and dry weight results (Figure 6a,b). Overall, T-1 and T-2 developed adequate leaf mass for marketing, whereas T-3 showed insufficient development. Cardarelli [21], reported that the net assimilation of CO2 of geranium was significantly affected by the irrigation systems with the highest values recorded with the drip-irrigation. The mean value of the number of fully-grown leaves (Figure 4d), at the end of the experiment, was 12.10 in T-2, followed by 8.33 in T-1 and 4.33 in T-3. Concerning the number of branches plant−1, T-2 developed 8.33 branches plant<sup>−</sup>1, followed by T-1 and T-3, where 5.33 and 2.78 branches plant−<sup>1</sup> were observed, respectively (Table 1). T-2 showed significant differences from both T-1 and T-3, and T-1 from T-3. The final height of plants was 15.13 cm in T-2, followed by T-1 with a value of 11.66 cm and T-3 with a value of 9.41 cm. T-2 showed significant differences from both T-1 and T-3, and T-1 from T-3. Moreover, regarding the growth trend, as reported in Figure 5, it is possible to see that, only two weeks after transplanting, treatment significantly modified the rate of growth until the final assessment. During the first two weeks, plants had a similar trend due to low temperatures registered, which strongly depressed growth.

Regarding the SPAD values, T-1 had the lowest values (47.16) and differed from both T-2 and T-3, with values of 54.88 and 65.63, respectively, which in turn were significantly different (Table 2) [23]. These behaviors reflected management of fertilization: in fact, when only three fertilizer supplies were provided (T-1), the lowest values were recorded, while, for other treatments (T-2 and T-3), in which the concentration of nutrient (nitrogen in particular) was constantly kept, SPAD values were always high. T-3 presumably had too elevated SPAD values, confirmed by the worst performances, while T-2

reached good SPAD levels, suggested to have more equilibrated leaf greenness [23]. Previously, in geranium, SPAD values were linearly correlated with total chlorophyll in fresh tissue. For example, in geranium "Ringo Deep Scarlet", there the following correlation was observe: SPAD = 14.96 + 37.30 \*chlorophyll content (mg·g−<sup>1</sup> of dried tissue), r<sup>2</sup> <sup>=</sup> 0.95 and *<sup>p</sup>* <sup>&</sup>lt; 0.001 [24]. In this experiment, the high SPAD values of T-2 and T-3 are attributable to the high nutrient concentration of NS provided. This is confirmed by EC values of leaching fraction, always showing higher values compared to EC of NS applied. EC values of leaching fraction fluctuated between 2171 and 3923 <sup>μ</sup>S·cm<sup>−</sup>1. The fact that the percentage of leaching fraction has always been sufficient (around 30%) [25] and that no particularly high temperatures were experienced during the experiment may have resulted in the use of a too concentrated NS, thus suggesting that the use of a less concentrated NS should be recommended. This was also confirmed by the fact that, during the first week, when a lower concentration of the NS was used, the EC of the leaching fraction did not increase as compared to the EC of NS supplied.

The visual assessment (aesthetic-commercial assessment) confirmed that the rooted cuttings of T-2 reached the best score, except for vegetative growth, and developed the best characteristics for the market (Table 4). Only one rooted cutting of T-2 treatment scored 2, and therefore was considered unmarketable because of an excessive asymmetry of the shape of the canopy. In T-3 treatments, 50% of the plants reached a MV score below 3, mainly since they showed a reduced growth. All of these plants also showed roots darkening. In particular, seven plants scored below 3 in one of the three parameters, six plants in two parameters, and two plants in all parameters. The roots darkening and the stunted growth could be due to the reuse of the non-sterilized leaching fraction, favoring the spread of root pathogens to the whole system. Indeed, spreading of root-borne diseases may occur, thus sterilization of the solution must be provided to avoid pathogens outbreak [9].

The application of different cultivation systems, each featuring a different fertigation management, also affected both water consumption and WUE (Table 5). Water consumption was highest in T-2 (10.11 L·plant<sup>−</sup>1), followed by T-3 and T-1 with 8.68 and 7.50 L·plant<sup>−</sup>1, respectively. T-2 showed significant differences from both T-1 and T-3, and T-1 from T-3. Despite these results, T-2 showed the highest value of WUE (21.85 g·L<sup>−</sup>1), followed by T-1 and T-3 with 17.63 and 13.07 g·L<sup>−</sup>1, respectively. No significant differences were observed by comparing T-2 and T-1 (Table 5). It should be considered, however, that, when converting the T-2 treatment into a closed system, the values could significantly improve [26–28]. In this case, the consumption of the NS could be lower (thanks to recycling of the drained solution), if compared with the other two treatments [28].

However, it is important to underline that in this scenario (closed system) the changing relationships between the nutrients in the drained solution need to be carefully considered, since they constitute an aspect that could influence the development of the rooted cuttings. Closed systems show a better water use efficiency, despite a slightly lower yield due to salt build-up in the root zone as a consequence of degradation of NS quality compared to the open systems [5,11–16]. In some cases, according to Savvas et al. [9], "switching over to closed cultivation systems does not seem to restrict crop yield or product quality". Given that in the context considered the farms often integrate their income with other crops, the drained fraction can also be used in open-air crops [29].

#### **5. Conclusions**

The experiment shows that the adoption of simplified soilless technology may allow enhancing the commercial characteristics of geranium, making it more attractive for the market and ultimately improving water and nutrient management. In particular, the adoption of a cultivation system with continuous fertigation on the substrate (peat) with drip irrigation can enable to obtain more attractive plants for the market. Moreover, this strategy improves water use efficiency, which could also be further improved with the adoption of a closed system. Modernization in the cultivation system and fertigation management may help to improve the commercial features of geranium even without using high technologies currently still not economically sustainable for most of the often family-run farms operating in the cut flowers and pot ornamentals sector in Trento province.

**Author Contributions:** Conceptualization, L.B. and F.Z.; methodology, L.B.; validation, L.B., F.Z., D.P., N.M. and F.O.; formal analysis, L.B. and D.P.; investigation, L.B. and F.Z.; resources, L.B. and F.Z.; data curation, L.B., F.Z., and D.P.; writing—original draft preparation, L.B.; writing—review and editing, D.P., N.M., F.O., and G.G.; visualization, L.B., D.P., N.M., F.O., and G.G.; supervision, D.P., N.M., F.O., and G.G.; project administration, L.B.; and funding acquisition, L.B. and F.Z. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** We thank the Zamboni Stefano farm located in Vigolo Vattaro (Northern Italy) for providing the infrastructure and equipment to carry out the experiments. We also thank Maria Vender for her contribution to the revision of the English version of the article.

**Conflicts of Interest:** The authors declare no conflict of interest.

## **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article*

## **Strategies for Improved Yield and Water Use <sup>E</sup>**ffi**ciency of Lettuce (***Lactuca sativa* **L.) through Simplified Soilless Cultivation under Semi-Arid Climate**

## **Nicola Michelon 1, Giuseppina Pennisi 1, Nang Ohn Myint 2, Giacomo Dall'Olio 1, Lucrecia Pacheco Batista 3, Adeodato Ari Cavalcante Salviano 4, Nazim S. Gruda 5, Francesco Orsini 1,\* and Giorgio Gianquinto <sup>1</sup>**


Received: 14 August 2020; Accepted: 9 September 2020; Published: 12 September 2020

**Abstract:** Simplified soilless cultivation (SSC) systems have globally spread as growing solutions for low fertility soil regions, low availability of water irrigation, small areas and polluted environments. In the present study, four independent experiments were conducted for assessing the applicability of SSC in the northeast of Brazil (NE-Brazil) and the central dry zone of Myanmar (CDZ-Myanmar). In the first two experiments, the potentiality for lettuce crop production and water use efficiency (WUE) in an SSC system compared to traditional on-soil cultivation was addressed. Then, the definition of how main crop features (cultivar, nutrient solution concentration, system orientation and crop position) within the SSC system affect productivity was evidenced. The adoption of SSC improved yield (+35% and +72%, in NE-Brazil and CDZ-Myanmar) and WUE (7.7 and 2.7 times higher, in NE-Brazil and CDZ-Myanmar) as compared to traditional on-soil cultivation. In NE-Brazil, an eastern orientation of the system enabled achievement of higher yield for some selected lettuce cultivars. Furthermore, in both the considered contexts, a lower concentration of the nutrient solution (1.2 vs. 1.8 dS m−1) and an upper plant position within the SSC system enabled achievement of higher yield and WUE. The experiments validate the applicability of SSC technologies for lettuce cultivation in tropical areas.

**Keywords:** urban agriculture; simplified soilless culture; hydroponics; conventional agriculture

## **1. Introduction**

The detrimental effects of climate change are resulting in dramatic environmental, economic, and social consequences across the world [1]. Current projections show an overall increase in temperatures, with rainfall being irregularly distributed and characterized by heavy downpours [2–5]. Erratic climate can negatively affect natural resources availability (e.g., water and agricultural land), as well as posing severe risks on both ecosystems and human health [2,6]. Many developing

countries, also located in tropical areas, are vulnerable to climate change due to their dependence on rain-fed agriculture, widespread poverty, and limited access to innovative technologies and improved agricultural practices [7]. An evident interdependence between climate change, economic vulnerability and migrations exists [8]. Accordingly, climate change is also resulting in a growing rate of migration toward urban and periurban areas of large cities. However, adaptation mechanisms are not yet in place, or are not strong enough, to mitigate the economic vulnerability of the most impoverished strata of the population [8]. Particularly in the tropical areas of Latin America and South-East Asia, health concerns are related to different forms of malnutrition frequently associated with a lack of micronutrients and vitamins in the population diet and low dietary diversification [9].

In Latin America, the Piaui State, located in the north-east area of Brazil (NE-Brazil), is one of the areas most affected by climate change due to its natural resources scarcity and extreme climatic conditions (i.e., semi-dry zone with a rainy season from December to May) [8]. Furthermore, after years of deforestation for agricultural purposes, the soil has a low amount of organic matter which negatively affects agricultural production [10]. Similar to Piaui State, the central dry zone in Myanmar (CDZ-Myanmar) is considered one of the most food-insecure regions of south-east Asia [11]. The climate of CDZ-Myanmar is characterized by a dry season without precipitation from November to March, which compromises and minimizes the agricultural choices of the farmers. Accordingly, climate and water scarcity are considered among the most significant problems of this area [12] and are expected to worsen in the future due to climate change [13].

In both NE-Brazil and CDZ-Myanmar, the introduction of innovative agricultural technologies which allow vegetable production even in urban and periurban areas, while fostering water-saving techniques to improve crop water use efficiency (WUE), is a crucial priority. According to Gianquinto et al. [14], it may be advisable to adopt simplified soilless cultivation (SSC) systems, which are independent of soil fertility and soil-borne diseases, do not require large spaces and intensive work labor and are characterized by high water and nutrient use efficiency thanks to the use of recirculating systems for the nutrient solution [14,15]. SSC systems are adapted from the concept of commercial hydroponics by integrating the advantages of easy construction and maintenance, while also reducing the initial economic investment or input requirements [16]. Different system designs exist for SSC, which mainly differ in the construction material, substrate used for plant growth and management of the nutrient solution [17].

The aim of this study is to assess the viability of an SSC system for the production of lettuce compared to traditional on-soil cultivation techniques in both NE-Brazil and CDZ-Myanmar, considering yield, water use efficiency and the overall physiological plant response. The assumption is that the adoption of SSC can increase yield and reduce water consumption of lettuce as compared to traditional on-soil grown plants also in very highly challenging contexts where soil quality is poor, climate is unfavorable and access to land by many people living in urban and periurban areas of large cities is limited. Moreover, the study integrates figures from different crop features and management strategies, such as crop positioning, garden orientation and cultivar traits to elaborate specific recommendations on the optimal management of the SSC systems proposed.

#### **2. Materials and Methods**

#### *2.1. Location*

**North-East of Brazil (NE-Brazil):** The experiments were carried out at the Horticulture Demonstration and Research Centre located at Fazenda Nova Esperança (5◦01 S and 42◦46 W, 87 m a.s.l.), owned by the Foundation Pe. Antonio Dante Civiero, located on the outskirts of the city of Teresina, capital of Piaui. According to Köppen's classification, the local climate is Aw type, with a dry summer and a rainy season between January and May.

**Central Dry Zone of Myanmar (CDZ-Myanmar):** The experiments were conducted at the Soil and Water Research Station of Yezin Agriculture University (19◦83 N and 96◦27 E, 122 m a.s.l.), located at the university campus in the periurban fringes of the capital NayPyiTaw. According to Köppen's classification, the local climate is Aw type, with a dry summer and a rainy season between June and October.

All experiments were carried out in the open field during the dry season, although the simplified soilless systems were equipped with a shading net (see description in Section 2.3).

## *2.2. Experimental Design*

Four independent experiments were performed with commercial varieties of lettuce. In all experiments, lettuce (*Lactuca sativa* L.) was sown manually in 105 cells plastic seedling trays, and seedlings were transplanted 21 days after sowing (DAS).


## *2.3. Simplified Soilless Cultivation System*

The SSC system used was the so-called Bottles system (Figure 1), developed and tested in the northeast of Brazil since 2005 [18]. It is composed of a wooden/bamboo frame and a gravity-flow system, where nutrient solution drains from a tank of 310 L volume placed above the system at 2 m height. Hydraulic pipes with an emitter flow rate of 2 L h−<sup>1</sup> direct the flux into the declined garden with a slope of 24%, which is composed by connecting plastic drinking bottles that host both substrate (rice husk in all experiments) and plants. The excess nutrient solution is then directed through a drainage pipe system to another tank placed below. A 50% shading net was placed above the system, to reduce light intensity. In NE-Brazil, the system used for the experiments was 6 m long and 3 m wide (18 m2) and accounted for 20 lines of 2 L plastic bottles (8 bottles line−1). Each bottle had two holes for hosting plants. Therefore, at full regime, the system could accommodate 320 plants. In CDZ-Myanmar, the system was tailored to the local context to meet the vegetables production needs of individual

households. Accordingly, the system size was reduced (5 m long and 2 m wide, resulting in a garden surface of 10 m2), and a smaller tank for the nutrient solution (100 L) was adopted. Each module hosted 240 plants. When also considering the surrounding paths allowing for garden access (about half a meter on each side and in internal paths), the net planting density was of 26 plants m−<sup>2</sup> in both NE-Brazil and CDZ-Myanmar.

In NE-Brazil, a nutrient solution (NS1.6) previously adopted for local SSC cultivation was used [18,19]. The NS1.6 was prepared with locally available simple mineral salts and soluble fertilizers and was characterized by an electrical conductivity (EC) of 1.6 dS m−<sup>1</sup> and a pH of 6.5. In CDZ-Myanmar, for both experiments, the NS was prepared by using locally available NPK fertilizer (15-15-15). During experiment 2, the adopted nutrient solution presented an EC of 1.2 dS m−<sup>1</sup> and a pH of 7.7, while in experiment 4 the nutrient solution was prepared at two concentrations, respectively, 0.6 g L−<sup>1</sup> in NS1.2 (EC = 1.2 dS m<sup>−</sup>1, pH = 7.3) and 0.8 g L−<sup>1</sup> in NS1.8 (EC = 1.8 dS m<sup>−</sup>1, pH = 7.5).

Details on macronutrient and micronutrient concentrations of nutrient solutions are reported in Tables 1 and 2.

**Figure 1.** (**a**) Schematic drawing of the growing system used with measurements (in meters) adopted. The system includes a top (A) and a drainage (B) tank, as well as a fresh nutrient solution reservoir (C). The system is fitted with a gravity flow drip-irrigation system (D) that deliver the nutrient solution to 20 lines of recycled plastic bottles (E). Excess nutrient solution is then drained to a re-collection pipe (F) which is connected (G) to the drainage tank (B). UP = Upper position; LP = Lower position. Images of the systems in the cities of (**b**) Teresina (Piaui, Brazil) and (**c**) NayPyiTaw (Myanmar).


**Table 1.** Macronutrient concentrations in water and nutrient solutions (NS) adopted in the experiments in NE-Brazil and in CDZ-Myanmar.

nd = not determined.

**Table 2.** Micronutrient concentrations in water and nutrient solutions (NS) adopted in the experiments in NE-Brazil and in CDZ-Myanmar.


nd = not determined.

### *2.4. Traditional on-Soil Cultivation*

The soil of the two regions had a loamy sand texture with similar hydrological soil parameters (wilting point and field capacity at 6% *v:v* and 13% *v:v*, respectively). The physical and chemical characteristics of the soil in the two locations are described in Table 3. In both NE-Brazil and CDZ-Myanmar, the soil was overturned and dug with a hoe prior to cultivation. Soil fertilization provided a supply of 1.5 kg m−<sup>2</sup> of cattle manure and 3.75 g m−<sup>2</sup> of N, P, and K (mineral fertilizer 10-10-10 and Nitrophoska 15-15-15 in NE-Brazil and CDZ-Myanmar, respectively). Fertilizer was manually applied three days before transplanting. No additional fertilizer was applied during the crop cycles. Due to low soil pH in NE-Brazil, 0.15 kg m−<sup>2</sup> of dolomitic limestone was added into the soil. The plots were raised by 20 cm and a trapezoid shape was developed, ensuring a base 1.2 m wide and a top 1.0 m wide. Finally, each plot was adjusted with a rake. Between the experimental plots, a space of approximately 0.7 m was left to facilitate maintenance, data collection and harvesting process. In both countries, plant spacing was 0.25 m between rows and 0.3 m within rows, resulting in a planting density of 13.3 plant m<sup>−</sup>2, according to the habits of the local farmers. The elemental unit consisted of a plot of 10 m2 (133 plants) or 5.4 m2 (72 plants) in NE-Brazil and CDZ-Myanmar, respectively.


**Table 3.** Chemical characterization of the soils in the two locations.

<sup>1</sup> OM = Organic Matter; <sup>2</sup> EC = Electrical Conductivity; <sup>3</sup> CEC = Cation Exchange Capacity, nd = not determined.

#### *2.5. Irrigation Management*

In the SSC system, nutrient solution flux started early in the morning (at 7:00 am) and continued until dusk (6:00 pm). Three times per day (at 7:00 am, 11:00 am, and 3:00 pm), the drained nutrient solution was moved back to the upper tank. The daily nutrient solution consumption was calculated by the difference between the nutrient solution volume between the upper tank (at the beginning of the day) and the bottom tank (at the end of the day). The nutrient solution in the system was refreshed every day by adding new nutrient solution to a set level

When plants were grown on the soil-based system, the irrigation management was different in NE-Brazil and CDZ-Myanmar experiments. In NE-Brazil, irrigation management was carried out based on the traditional local habit of the farmers by using manual irrigation. Water was distributed across experimental plots through manual labor, and a 12 L watering bucket was used. The amount of water distributed in a plot was based on farmers' experience. In CDZ-Myanmar, the irrigation management of soil-based treatments was based on crop evapotranspiration (ETc), restoring 100% of crop ETc by means of a drip irrigation system

ETc was calculated by using the following equation (Equation (1))

$$\rm ET\_c = ET\_0 \times K\_c \tag{1}$$

where ETc (mm day−1) is the calculated crop evapotranspiration, ET0 (mm day−1) is the reference evapotranspiration, and Kc is the FAO crop coefficient for lettuce [20].

For the estimation of the reference evapotranspiration (ET0), the Hargreaves-Samani (HS) equation (Equation (2)) was used,

$$\rm{ET}\_0 = 0.0023 \times (T\_{\rm{mean}} + 17.8) \times (T\_{\rm{max}} - T\_{\rm{min}})^{0.5} \times R\_\text{a} \tag{2}$$

where ET0 (mm day−1) is the reference evapotranspiration rate, Tmean, Tmax and Tmin are the mean, maximum and minimum temperature (◦C) of the day, respectively, and Ra (W m−<sup>2</sup> day−1) is the extraterrestrial solar radiation [20].

The meteorological data for the determination of the reference evapotranspiration were daily downloaded from the website of the Agro-Meteorological Department of Yezin Agriculture University (http://www.yau.edu.mm/), located inside the university campus, excluding extraterrestrial radiation Ra that was calculated according to Duffie and Beckman [21].

The amount of water used for each irrigation was calculated based on plant water balance considering soil properties, root depth, and climate data (including rainfall, if any). Daily ETc was estimated considering the FAO crop coefficient for lettuce crop growth stages. Lettuce cycles were divided into three growth stages, and the Kc used was 0.7, 1.0 and 0.95, respectively. The time of irrigation was determined when readily available soil water (50% available soil water) was depleted.

Sixteen mm diameter drip pipes were used. Drippers had a flow rate of approximately 1.3 L h<sup>−</sup>1, and each plant was supplied with a single dripper. A flow rate test and calculation of distribution uniformity (DU) were carried out before transplanting. The DU was calculated following the indications from Baum et al. [22]. Irrigation management (time and rate) was manually performed.

#### *2.6. Plant Measurements*

At harvest, plants were weighed to determine the fresh weight (g plant−1). Yield (kg m−2) was assessed by excluding external leaves which appeared damaged or wilted. Leaf number was also counted. Water use efficiency (WUE) was determined as the ratio between fresh weight and the volume of water used and was expressed as g FW L−<sup>1</sup> H2O. In experiment 4, leaf stomatal conductance was also measured using a handheld photosynthesis measurement system model CI-340 (Camas, WA, USA), equipped with 6.25 cm2 cuvette. Measurements were made at 27 DAT on the upper surface of the canopy on three leaves per each plant from 10:00 to 14:00 taking approximately one hour to

complete each replication. All plants were measured on a single day. In the system, EC and pH were constantly monitored using a Combo pH/EC/TDS/Temp tester Model HI98130 (HANNA®, Villafranca Padovana (PD), Italy). In experiments 2 and 4, the nutrient solution temperature was also monitored twice a week.

#### *2.7. Statistical Analysis*

Data were collected on 12 plants from the central part of each plot. Data from experiments 1 and 2 were analyzed using one-way ANOVA. Data from experiments 3 and 4 were analyzed by using twoand three-way ANOVA, respectively. Means were separated using the Tukey HSD test at *p* ≤ 0.05. Before the analysis, all data were checked for normality and homogeneity of variance. Averages and standard errors (SE) were calculated. Statistical analysis was carried out using R statistical software (version 3.3.2, package "emmeans" and "car").

#### **3. Results**

## *3.1. Climate during the Experiments.*

## **NE-Brazil**

During experiment 1, maximum air temperature ranged between 31.4 and 34.7 ◦C with an average of 33.0 ◦C. Minimum temperature ranged between 17.8 and 22.4 ◦C with an average of 19.8◦C. The daily relative humidity (RH) ranged between a minimum of 57% and a maximum of 97% (Table 4). Furthermore, 20.3 mm of effective rainfall occurred. During experiment 3, maximum air temperature ranged between 31.7 and 35.4 ◦C with an average maximum temperature of 34.0 ◦C. Minimum temperature ranged between 17.8 and 22.4 ◦C with an average minimum temperature of 19.8 ◦C. The maximum relative humidity (RH) was 97% and the minimum RH was 55% (Table 4). The growing degree days (GDD) from transplanting to harvest ranged from 710 ◦C (experiment 3) to 920 ◦C (experiment 1).


**Table 4.** Main climatic features during the experiments.

<sup>1</sup> RH = Relative Humidity; <sup>2</sup> DLI = average daily light integrals; <sup>3</sup> GDD = growing degree days, calculated based on a crop base temperature of 4 ◦C [23]; \* on soil-based system (40 days cropping cycle); \*\* on simplified soilless system (31 days crop cycle).

### **CDZ-Myanmar**

During the experiment 2, maximum air temperature ranged between 25.0 and 34.0 ◦C with an average of 31.5◦C. Minimum temperature ranged between 14.4 and 20.4 ◦C, with an average of 16.7 ◦C. The daily relative humidity (RH) ranged between a minimum of 48% and a maximum of 73% (Table 4). During experiment 4, maximum air temperature ranged between 30.7 and 38.6◦C with an average maximum temperature of 35.6 ◦C. Minimum temperature ranged between 16.0 and 23.0◦C with an average minimum temperature of 19.5 ◦C. The maximum relative humidity (RH) was 59% and the

minimum RH was 30%. No rainfall occurred during the experiments. The growing degree days (GDD) from transplanting to harvest ranged from 662 ◦C (experiment 2) to 731 ◦C (experiment 4) (Table 4).

#### *3.2. Experiment 1—NE-Brazil*

Lettuce yield was higher (+35%) in the SSC system, with a mean value of 2.3 kg m<sup>−</sup>2, as compared to 1.7 kg m−<sup>2</sup> achieved on soil (Figure 2a). This was mainly due to larger size of the leaves (data not shown), while leaf number was higher in plants grown on soil (Figure 2b). The increased yield was obtained with a daily water use (L m−<sup>2</sup> d<sup>−</sup>1) approximately four times lower in the SSC system, as compared to conventional on-soil cultivation (1.8 vs. 7.5 L m−<sup>2</sup> d−1) (data not shown). As a consequence, WUE in SSC system was 7.7 times higher, as compared to the conventional on-soil system, with mean values of 43.7 and 5.6 g L−<sup>1</sup> H2O, respectively (Figure 2c).

#### *3.3. Experiment 2—CDZ-Myanmar*

During experiment 2 the average minimum temperature of the nutrient solution was 19.6 ± 1.64 ◦C while the average maximum temperature was 29.2 ± 1.69 ◦C. The average pH was 7.7, ranging from 7.4 to 8.1. The average EC was 1.28, ranging from 1.12 to 1.46 dS m<sup>−</sup>1.

Yield (kg m−2) was increased by 72% (3.1 vs. 1.8 kg m−2) in SSC in comparison to the soil treatment (Figure 2d) and the leaf number was significantly higher in soil-grown lettuce compared to soilless-grown plants (Figure 2e). Daily water use (L m−<sup>2</sup> d<sup>−</sup>1) was approximately two times lower in the SSC system (2.66 L m−<sup>2</sup> d<sup>−</sup>1) as compared to conventional on soil production (4.07 L m−<sup>2</sup> d<sup>−</sup>1). WUE in the SSC system was found to be 2.7 times higher than that obtained with conventional cultivation, with average values of 37.1 and 13.7 g L−<sup>1</sup> H2O, respectively (Figure 2f).

**NE-Brazil** 

**Figure 2.** Results from experiments 1 (Brazil, top row) and 2 (Myanmar, bottom row). Lettuce yields (**a**,**d**), leaf number (**b**,**e**), and water use efficiency (WUE, **c**,**f**). Vertical bars represent standard errors. Significant differences at *p* ≤ 0.01 (\*\*), *p* ≤ 0.001 (\*\*\*).

## *3.4. Experiment 3—NE-Brazil*

Considering the system orientation, significant differences for yield were found only in Veronica and Banchu cultivars, for which the west-oriented system showed a reduction in yield of 10 and 44%, respectively, as compared to the east-oriented one. In contrast, yields of cv Isabela and cv Mimosa were not affected by the SSC system orientation (Figure 3). Daily water use was about 1.8 L m−<sup>2</sup> d<sup>−</sup>1, as for experiment 1.

**Figure 3.** Results from experiments 3 (NE-Brazil). Yield response to the simplified soilless system orientation (east, grey columns; west, white columns) in four lettuce cultivars (Isabela, Veronica, Banchu, and Mimosa). Vertical bars represent standard errors. Significant differences at *p* ≤ 0.01 (\*\*), ns = not significant differences.

## *3.5. Experiment 4—CDZ-Myanmar*

The average minimum temperature of the nutrient solution was 20.7 ± 1.1 ◦C while the average maximum temperature was 39.5 <sup>±</sup> 0.87 ◦C. Daily water use was 2.50 L m−<sup>2</sup> d−<sup>1</sup> for NS1.2 and 2.23 L m−<sup>2</sup> d−<sup>1</sup> for NS1.8. The average pH was 7.3 and 7.5 for NS1.2 and NS1.8, respectively. pH ranged from 6.4–8.7 for the former, and 6.6–8.9 for the latter. Average EC was 1.25 and 1.83 dS m−<sup>1</sup> for NS1.2 and NS1.8, respectively, ranging from 1.14–1.48 dS m<sup>−</sup><sup>1</sup> for solution NS1.2 and 1.59–2.06 dS m−<sup>1</sup> for solution NS1.8 (data not shown). Results of analysis of variance in Table 5 show that the EC of the nutrient solution (EC), lettuce cultivar (Cv), and plants position (P) significantly affected plant morphological and productive parameters, as well as WUE and the crop physiological response.

**Table 5.** Results from the ANOVA on experiment 4 (CDZ-Myanmar). Effect of EC of the nutrient solution (EC), cultivar (Cv), and position within the garden (P) on lettuce yield, leaf number, and water use efficiency (WUE). Significant differences at *p* ≤ 0.05 (\*), *p* ≤ 0.01 (\*\*) and *p* ≤ 0.001 (\*\*\*), ns = not significant differences.


Yield, stomatal conductance and WUE were affected by EC, Cv, and P—wherein a significant interaction between the three factors was noted—while leaf number was only affected by EC and Cv, with a significant interaction between the two factors (Table 5). Yield of plants placed in the lower position (LP) was not affected by Cv and EC, while for both cultivars the plants in the upper position (UP) yielded more when NS1.2 was used (Table 6). The yield of plants belonging to cv Thai and grown by using NS1.2 was four times higher, as compared to yield of Thai lettuce supplied with NS1.8 and placed in the same position within the system (Table 6). The increased yield was mainly due to

leaf number, as Thai plant grown adopting NS1.2 showed the highest number of leaves (12.9 leaves plant<sup>−</sup>1) while no differences were observed between the other treatments (data not shown). Stomatal conductance was highest (212 mmol m−<sup>2</sup> s<sup>−</sup>1) in Thai lettuce grown on the upper part of the system by using NS1.2 (Table 6). For plants grown in the lower part of the system (LP), stomatal conductance was only affected by CV, and was higher in cv Thai for both considered EC (Table 6). Leaf temperature was only affected by the position (P) in the system (data not shown), and was lowest in plants grown on the top of the system (28.8 ◦C compared to 29.8 ◦C measured in plants grown at the bottom of the system). In cv Thai, WUE was highest in plants fed with NS1.2 and grown in the upper position (UP) of the system, while for cultivar Rapido 344 the only statistically significant difference was evidenced between plants grown on the upper position and fed with NS1.2 and plants grown in the lower position of the system and fed with NS1.8 (Table 6).

**Table 6.** Results from experiment 4 (CDZ-Myanmar). Effects of factorial combination of EC of the nutrient solution (EC, 1.2 vs. 1.8 dS m<sup>−</sup>1), cultivar (Cv, Thai vs EW) and position (P, upper position, UP vs lower position, LP) within the garden on lettuce yield, stomatal conductance (gs)and water use efficiency (WUE). Different letters indicate significant differences at *p* ≤ 0.05.


#### **4. Discussion**

The application of different cropping systems significantly affected yield, physiological response and water use efficiency of lettuce grown in both NE-Brazil and CDZ-Myanmar.

Water availability is one of the major constraints for agricultural development and food production. The first and second experiments aimed to determinate whether SSC lettuce production is a suitable and sustainable alternative to conventional on-soil production in both locations. Barbosa et al. [23], when comparing commercial (high-tech) hydroponic greenhouses against on-soil lettuce production, found that hydroponics could increase yield by 11-folds, thanks to improved nutrition and environmental control. According to our results obtained in both experiments, the use of a simplified (low tech) soilless system allowed increase in the yield of lettuce but to a lesser extent (+35% in NE-Brazil and +72% in CDZ-Myanmar, (Figure 2a,d)). Yield increase can be the result of higher planting density (26 vs. 13 plants m<sup>−</sup>2, on SSC and on-soil cultivation respectively), fast plant growth and precocity of production (31 vs 40 DAT according to experiment 1) and the improved environmental conditions maintained within the SSC system, including plant nutrition, uniform and constant irrigation, as well as the shading cover integrated in the SSC system. According to Zhao et al., the adoption of a shading net as a cover for lettuce production in the summer season in Kansas led to a slightly lower daily maximum air temperature relative to the open field, with an average reduction of ≈0.4 ◦C [24]. Moreover, Zhao et al. reported that the shading net has a significant impact on soil temperature and leaf temperature [24]. Indeed, in comparison with open field conditions, when shading net is adopted, a considerable reduction of leaf surface temperature, by 1.5 to 2.5 ◦C, was observed [24], thus affecting the plants' capacity to absorb water and nutrients [25]. In the SSC system, the higher fertigation frequency probably affected production capacity. Silber et al. experimented on the effect of fertigation frequency on yield, water and nutrient uptake of lettuce [26], finding that high fertigation frequency (from 2 to 10 events a day) induced a significant increase (13–15%) in lettuce fresh weight (FW) [26].

Furthermore, SSC systems are also considered water-saving technologies, thanks to the capability to deliver water directly to the plant root [27,28]. Despite limited soil exploration by the shallow rooting system of lettuce, in NE-Brazil, when a conventional growing system was adopted, irrigation water was applied by means of a bucket or can on the entire soil surface and consequently a significant amount of water is lost through evaporation and percolation into the sandy soil. Increase in the use of low-flow and more targeted irrigation techniques, such as the adoption of a drip irrigation system, could lower the overall water use of conventional farming [23]. As a matter of fact, drip irrigation was used as a control treatment in the CDZ-Myanmar experiments. Accordingly, the adoption of an SSC system enabled a reduction of water use by 76% and 59% in NE-Brazil and in CDZ-Myanmar, respectively, as compared to on-soil production. The observed water savings are consistent with previous literature, e.g., when a SSC system was adopted in Colombia, water use was reduced by 90% as compared to the traditional on-soil cropping system [29].

A consequence of higher yield and lower water use was an increased WUE in the SSC systems. In NE-Brazil and CDZ-Myanmar, WUE was, respectively, 7.7 and 2.7-fold higher, as compared to conventional on-soil production (Figure 2c,f). Similarly, WUE for lettuce in hydroponics was previously found in the range of 2.9 g of dry mass per L−<sup>1</sup> H2O [30], or 41 g of fresh mass per L<sup>−</sup><sup>1</sup> H2O [31]. Lettuce grown in high-tech hydroponic conditions showed a reduction in water use by 13-fold, as compared with traditional on-soil cultivation [23]. Under the expected climate change scenarios and water limitation for agriculture, SSC systems could be a valuable strategy to sustain highly productive agriculture where the adoption of high-technology systems is not affordable [3,5,28,32].

In the third experiment, the adaptability of four lettuce cultivars to two different garden orientations was addressed to have a deeper understanding of the SSC system management. It emerged that the response of plant growth to the garden orientation was cultivar dependent (Figure 3). Accordingly, Veronica and Banchu achieved a higher yield with the eastern garden exposure (Figure 3), which received a lower amount of solar radiation in the afternoon mitigating the high air temperatures. Wheeler et al. [33] indicated 23 ◦C as the optimal daily temperature for growing lettuce, a condition that is far below the mean temperatures observed in NE-Brazil during the experiment. Heat stress could also result in a greater osmotic stress caused by the nutrient solution, resulting in lower water uptake and reduced plant growth [27]. Moreover, due to elevated temperatures in the root zone environment, hypoxia may occur, inhibiting root respiration, mineral uptake and water movement into the roots [34].

In the fourth experiment, the growth response of plants grown in different positions within the system was addressed as a function of both plant cultivar and the nutrient solution concentration (Table 5). Accordingly, the highest yield was associated with Thai cultivar supplied with NS1.2 (Table 6). Possibly, under the local climate, plants preferred a nutrient solution with lower EC, and the Thai cultivar better responded to reduced osmotic stress. Furthermore, yield was the highest when Thai cultivar was grown in the upper part of the system. (Table 6). As reported by Gianquinto et al. [14], this aspect could be due to the increased temperature reached by the nutrient solution during the flow between the top and the bottom tank. It would suggest that by the time the nutrient solution reached the lower section of the SSC, it was significantly warmed up by irradiance in the plastic bottles, although this statement should further be confirmed by determination of nutrient solution and substrate temperatures in the different positions. Thompson et al. [35] showed that a 24 ◦C root temperature in hydroponic systems is the ideal temperature whereby lettuce growth can be maximized, even with elevated air temperature. Different studies also reported that high nutrient solution temperature depresses water and nutrient uptake through reduced oxygen availability, also affecting physiological processes such as root browning and active transport in membranes [36]. Moreover, it was also observed that high solution temperature might decrease nutrient concentration (particularly of N, K and Ca) in the root, which may ultimately decrease crop growth [25]. It should be further studied, however, whether this may be associated with increased nutrient solution temperature as water flows through the system, or with selective absorption of specific nutrients from those plants that receive the

nutrient solution first. In this regard, it could also be considered to add additional hydraulic pipes with emitters in the middle of the system.

### **5. Conclusions**

The study addressed the application of SSC technologies [37–39] for lettuce production in tropical wet and dry climates. Elevated potentialities, in terms of both yield increase and improved water use efficiency in comparison with traditional on-soil cultivation technologies, were evidenced in both locations. Furthermore, the study explored alternative crop management strategies evidencing differences in cultivar adaptability and potential productivity. For instance, garden orientation was shown to affect crop productivity on a cultivar-dependent basis. Finally, under the elevated temperatures that are locally experienced, it is advisable to reduce the concentration of the nutrient solution (with EC of 1.2 dS m−<sup>1</sup> providing better results than EC of 1.8 dS m<sup>−</sup>1). Interestingly, the yield was also improved when plants were located in the upper positions of the garden. Government and local support services could influence the future of soilless farming, as subsidies could be used to offset the relatively high initial cost of SSC infrastructure. We conclude that a simplified soilless system could become one of the efficient strategies for contributing to sustainably feeding the world's growing population, especially in challenging areas such as the north east of Brazil and the central dry zone of Myanmar.

**Author Contributions:** Conceptualization, N.M., A.A.C.S., F.O., G.G.; methodology, N.M.; validation, N.M., G.G., G.P., F.O.; formal analysis, N.M., G.P., G.G.; investigation, N.M., L.P.B., G.D., N.O.M.; resources, F.O., G.P.; data curation, N.M., G.P.; writing—original draft preparation, N.M.; writing—review and editing, G.P., G.G., F.O., N.S.G.; visualization, N.M., G.P., G.G.; supervision, F.O., G.P.; project administration, N.M.; funding acquisition, F.O., G.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** We thank Padre Humberto Pietrogrande, former President of Pe. Antonio Dante Civiero Foundation, Nang Hseng Hom, Rector of Yezin agriculture University of Myanmar and Terre des Hommes Italia INGO for providing the area, infracture and equipment to carry out the experiments.

**Conflicts of Interest:** The authors declare no conflict of interest.

## **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **E**ff**ect of Nutrient Solution Concentration on the Growth of Hydroponic Sweetpotato**

## **Masaru Sakamoto \* and Takahiro Suzuki**

Faculty of Biology-Oriented Science and Technology, Kindai University, Wakayama 649-6493, Japan; tksuzuki@waka.kindai.ac.jp

**\*** Correspondence: sakamoto@waka.kindai.ac.jp; Tel.: +81-0736773888

Received: 24 September 2020; Accepted: 3 November 2020; Published: 4 November 2020

**Abstract:** Nutrient solution concentration (NSC) is a critical factor affecting plant growth in hydroponics. Here, we investigated the effects of hydroponic NSC on the growth and yield of sweetpotato (*Ipomoea batatas* (L.) Lam.) plants. First, sweetpotato cuttings were cultivated hydroponically in three different NSCs with low, medium, or high electrical conductivity (EC; 0.8, 1.4, and 2.6 dS m<sup>−</sup>1, respectively). Shoot growth and storage root yield increased at 143 days after plantation (DAP), depending on the NSC. Next, we examined the effect of NSC changes at half of the cultivation period on the growth and yield, using high and low NSC conditions. In plants transferred from high to low EC (HL plants), the number of attached leaves increased toward the end of the first half of the cultivation period (73 DAP), compared with plants transferred from low to high EC (LH plants). Additionally, the number of attached leaves decreased in HL plants from 73 DAP to the end of the cultivation period (155 DAP), whereas this value increased in LH plants. These changes occurred due to a high leaf abscission ratio in HL plants. The storage root yield showed no significant difference between HL and LH plants. Our results suggest that the regulation of hydroponic NSC during the cultivation period affects the growth characteristics of sweetpotato.

**Keywords:** nutrient solution concentration; hydroponics; sweetpotato; storage root; leaf abscission

#### **1. Introduction**

Sweetpotato (*Ipomoea batatas*) is an important root vegetable cultivated in temperate and tropical zones, especially in Asia and Africa [1]. Storage roots of sweetpotato contain relatively high amounts of carbohydrates that support the demand for food in developing countries [2,3]. In recent years, sweetpotato has been also evaluated as a candidate for biofuel production [4,5]. Sweetpotato could potentially be used as an alternative to corn-based ethanol production to reduce fertilizer, water, and pesticide inputs and to utilize its ability to fix relatively large amounts of solar energy into starch in storage roots [6,7]. Several efficient methods of extracting biofuels and their residues (hydrogen, ethanol, and methane) from sweetpotato have been reported to date [5,8–11]. Because the demand for sweetpotato is gradually increasing worldwide [12], it is necessary to establish an efficient and cheap cultivation method with low fertilizer requirement.

Fertilizers are widely used in agriculture to increase crop production. In sweetpotato, soil amendment using manure and inorganic fertilizers has a significant impact on plant growth and storage root development [13]. Among chemical fertilizers, nitrogen (N), phosphorus (P), and potassium (K) are the major elements required for supporting shoot and root growth in sweetpotato [14,15]. Under N deficient conditions, the application of N fertilizers significantly increases the storage root weight [16–18]. The relative proportion of N and K fertilizers applied also affects the storage root yield, photosynthesis product distribution, and leaf enzyme activities in field-grown sweetpotato [19–21]. Administration of an adequate quantity of K fertilizer (K2 was supplied to 300 kg ha−1) has shown to increase the ratio of storage root yield relative to the total yield [22]. Furthermore, N application rate influences the lateral root development at the early growth stage, with 50 kg ha−<sup>1</sup> application being the most developed [23]. These early root developments are thought to the initiation of storage root formation [24].

In hydroponics, fertilizers are supplied as ions in the nutrient solution [25]. Several formulations of essential macro- and micronutrients have been developed to enhance nutrient uptake and plant growth [26]. Because the nutrient solution is the only source of mineral nutrients in hydroponically-grown plants, extremely low concentrations of nutrients generally leads to growth inhibition [27]. On the other hand, extremely high nutrient solution concentration (NSC) causes osmotic stress, ionic toxicity, and growth restriction [27]. Several studies have demonstrated that NSC influences the growth and components of spinach, tomato, cucumber, salvia, bean, artichoke, wasabi, and lettuce plants [28–37]. In a hydroponic NSC with high electrical conductivity (EC), the growth of tomato plants was restricted, whereas the level of sugars and lycopene in tomato fruits, and consequently fruit quality, were enhanced [30]. In strawberry, flower bud initiation was promoted by treatment with low NSC [38–40].

Sweetpotato plants fail to develop storage roots under continuous waterlogging conditions [41–43]. Therefore, several studies have established hydroponic methods of sweetpotato cultivation to avoid soaking the hypertrophic sites of roots in water [41–49]. Substrates that ensure proper root aeration, such as rockwool, vermiculite, and sand, have been used for the hydroponic cultivation of sweetpotato [43–45]. Additionally, rockwool slab-based hydroponic systems have been demonstrated to produce thickened sweetpotato storage roots between the hydroponic solution surface and rockwool slabs [46]. Similarly, the nutrient film technique and modified deep flow technique have been shown to induce storage root formation at an area where roots are not continuously immersed in the hydroponic solution [41,47–49]. Although some hydroponic methods for sweetpotato have been developed to date, studies on sweetpotato hydroponic NSC are limited. Here, using previously developed vermiculite-based hydroponic methods [43], we investigated the effect of NSC on the growth and yield of sweetpotato.

#### **2. Materials and Methods**

#### *2.1. Experimental Conditions*

Sweetpotato (*Ipomoea batatas*) cultivar "Narutokintoki" was used in this study. The hydroponics system for sweetpotato was prepared as described previously [43]. Briefly, this system consisted of vermiculite-filled vinyl pots (4.5 L) placed in nutrient solution-filled containers (59 cm × 39 cm × 18 cm). Storage roots developed in vinyl pots, and fibrous roots of sweetpotato plants extended from the upper vinyl pots into the containers placed below. Plants could absorb the nutrient solution from the vermiculite, as the bottom of each pot was in direct contact with the water absorption sheet that extended into the nutrient solution below. Therefore, vermiculite in vinyl pots remained saturated with the nutrient solution throughout the cultivation period. The surface of pots and bottom containers was covered with insulation sheets to maximize the utilization of sunlight by reflection for photosynthesis. The nutrient solution was prepared by mixing OAT house 1: OAT house 2 (OAT Agrio Co., Ltd., Tokyo, Japan) at a ratio of 3:2, and the NSC was adjusted to obtain the target EC (described below). The mixed nutrient solution contains N, 260 mg L<sup>−</sup>1, P2O5, 120 mg L−1; K2O, 405 mg L−1; CaO, 230 mg L−1; MgO, 60 mg L−1; MnO, 1.5 mg L−1; B2O3, 1.5 mg L−1; Fe, 2.7 mg L−1; Cu, 0.03 mg L−1; Zn, 0.09 mg L−<sup>1</sup> and Mo, 0.03 mg L−1. Reduction in the water level in the tank due to evaporation was compensated by adding more water to the maximum level. Therefore, the EC of the NSC in each tank gradually decreased over time (Supplementary Figure S1). The nutrient solution was renewed approximately once every 30 days.

Two separate experiments were conducted (experiment 1 in 2018, and experiment 2 in 2019) to examine the effect of NSC on the growth and yield of sweetpotato (Figure 1). In experiment 1, three NSCs, each with low (0.8 dS m−1), medium (1.4 dS m−1), or high (2.6 dS m−1) EC, were used throughout the cultivation period, and the effect of each NSC on plant growth was observed. The initial pH of high, medium, and low NSCs were 6.14, 6.46, and 6.72, respectively. In experiment 2, effects of changes in NSC on plant growth were examined using nutrient solutions with low EC (0.8 dS m−1) and high EC (2.6 dS m−1). Four treatments were conducted in experiment 2: (1) LL, plants were grown in low EC nutrient solution throughout the cultivation period; (2) LH, plants were grown in low EC nutrient solution until the end of the first half of the cultivation period, and then transferred to high EC nutrient solution and maintained until the end of the cultivation period; (3) HL, plants grown in high EC nutrient solution were transferred to low EC nutrient solution at the end of the first half of the cultivation period and maintained thereafter; (4) HH, plants were grown in high EC nutrient solution throughout the cultivation period.

**Figure 1.** Experimental design. EC: electrical conductivity.

In experiment 1, sweetpotato stem cuttings were planted in vermiculite-filled vinyl pots and grown for 18 days by drenching in nutrient solution with medium EC (1.4 dS m<sup>−</sup>1). The experiment was started by transferring the pots to the hydroponic system with different NSCs (four pots per container). Plants were then cultivated for 143 days (from 29 May to 19 October in 2018) at the open experimental field of Kindai University (Faculty of Biology-Oriented Science Technology, Wakayama, Japan). The average temperature of Wakayama city in 2018 were 19.7 ◦C, 23.2 ◦C, 28.8 ◦C, 29.1 ◦C, 24.3 ◦C, and 19.5 ◦C in May, June, July, August, September, and October, respectively, according to the website of Japan Meteorological Agency [50]. The average relative humidity was 69%, 77%, 74%, 69%, 78%, and 67% in May, June, July, August, September, and October, respectively [50]. The experimental field is about 17 km away from the meteorological station in Wakayama City, and 90 m higher than the station. The nutrient solution was renewed on 26 June, 31 July, and 30 August. In experiment 2, stem cuttings were planted in pots and grown for 25 days under the same growth conditions as those used for experiment 1. Pots were then transferred to the hydroponic system and cultivated for 155 days (from 18 May to 20 October in 2019) at

the open experimental field of Kindai University. The average temperature of Wakayama city in 2019 was 20.1 ◦C, 23.5 ◦C, 26.3 ◦C, 28.5 ◦C, 26.4 ◦C, and 20.7 ◦C in May, June, July, August, September, and October, respectively [51]. The average relative humidity was 60%, 72%, 79%, 76%, 71%, and 73% in May, June, July, August, September, and October, respectively [51]. The nutrient solution was renewed on 22 June, 12 July, 3 August, 31 August, and 28 September. The EC of the nutrient solution was changed for HL and LH plants on 3 August 2019.

#### *2.2. Measurement of Plant Growth and Yield*

In experiments 1 and 2, shoot and storage root fresh weight (FW) and storage root number were measured at 143 and 155 days after planting (DAP), respectively. Enlarged roots weighing more than 20 g were included as storage roots. In experiment 2, the number of attached leaves and the maximum length of the stem were measured at 3, 73, and 155 DAP. Total leaf number, abscised leaf ratio, stem number, and stem diameter were measured at 155 DAP. The number of total leaves was counted by adding the number of attached leaves and leaf petioles (without leaves). The abscised leaf ratio was calculated by dividing the number of total leaves with the number of attached leaves. In experiments 1 and 2, all measurements were recorded as the average of eight plants.

#### *2.3. Measurement of Chlorophyll Contents*

Relative chlorophyll contents were measured via a nondestructive assay using the Soil and Plant Analyzer Development (SPAD) chlorophyll meter (SPAD-502; Konica Minolta, Tokyo, Japan). Measurements were conducted at 3, 73, and 153 DAP using the second young fully expanded leaf of each plant.

#### *2.4. Data Analysis*

Data were analyzed using the JMP statistical package (SAS Institute, Cary, NC, USA). Significant differences among treatments were determined by one-way analysis of variance, followed by the Tukey–Kramer honest significant difference test for pairwise comparisons at *p* < 0.05.

#### **3. Results**

#### *3.1. E*ff*ect of NSC on the Growth of Hydroponic Sweetpotato (Experiment One)*

Experiment one was conducted to examine the effect of NSC on the growth of sweetpotato in a hydroponic system. At 143 DAP, the shoot FW was the highest in the nutrient solution with high EC, followed by medium EC, and low EC (Figure 2A). The storage root FW showed the same trend as that described above (Figure 2B). Additionally, the number of storage roots showed no significant difference among the three treatments (Figure 2C).

**Figure 2.** Effects of nutrient solution concentration on shoot fresh weight (**A**), storage root fresh weight (**B**), and number of storage roots (**C**) of sweetpotato at 143 days after plantation in experiment 1. Vertical bars represent the means ± SE (*n* = 8). Different letters indicate significant differences among the treatments at *p* < 0.05 by Tukey–Kramer's test.

#### *3.2. E*ff*ect of Variation in NSC on the Growth of Hydroponic Sweetpotato (Experiment Two)*

Next, we examined whether changes in NSC affect the growth of hydroponic sweetpotato. Plant shoot growth was measured at three time points: 3 DAP, 73 DAP (4 days before changing the NSC), and 155 DAP (harvest day). The leaf chlorophyll content increased from 3 to 73 DAP in all plants, reaching similar levels in all treatments (Figure 3A). No significant differences were detected among treatments at each time point, although the leaf chlorophyll contents of HH and LH plants at 155 DAP tended to be higher than that of HL and LL plants (Figure 3A). The number of attached leaves was higher in HH and HL plants than in LH and LL plants at 73 DAP (Figure 3B). Compared with 73 DAP, the number of attached leaves at 155 DAP was approximately 1.51-fold change in HH plants, 0.58-fold change in HL plants, 2.84-fold change in LH plants, and 1.09-fold change in LL plants (Figure 3B). Reduction in the number of attached leaves during cultivation suggests the induction of leaf abscission. This coincides with the pictures of shoots of HL plants at 155 DAP showing that leaves were rarely attached to the petiole at the bottom and middle sections of the stem (Figure 4). To examine leaf abscission, we counted the number of total leaves, including previously abscised leaves, at 155 DAP. HH plants showed the highest number of total leaves, followed LH, HL, and LL plants (Table 1). The abscised leaf ratio was the highest in HL plants, followed by LL, HH, and LH plants (Table 1).


**Table 1.** Effects of nutrient solution concentration on number of total leaves and abscised leaf ratio of sweetpotato at 155 days after plantation in experiment 2. Different letters indicate significant differences among the treatments at *p* < 0.05 by Tukey–Kramer's test.

LL: plants were grown in low EC nutrient solution throughout the cultivation period; LH: plants were grown in low EC nutrient solution until the end of the first half of the cultivation period, and then transferred to high EC nutrient solution and maintained until the end of the cultivation period; HL: plants grown in high EC nutrient solution were transferred to low EC nutrient solution at the end of the first half of the cultivation period and maintained thereafter; HH: plants were grown in high EC nutrient solution throughout the cultivation period.

At 73 DAP, the maximum shoot length was higher in HH and HL plants than in LH and LL plants (Figure 3C). At 155 DAP, shoot length was the highest in HH plants and lowest in LL plants, while HL and LH plants showed similar intermediate shoot lengths (Figure 3C). The number of stems was significantly higher in HH plants compared with plants in the other treatments (Figure 5A). Stem diameter was the highest in HH plants, followed by HL and LH plants, and the lowest in LL plants (Figure 5B).

**Figure 3.** Effects of nutrient solution concentration on number of attached leaves (**A**), maximum shoot length (**B**), and the Soil and Plant Analyzer Development (SPAD) (**C**) of sweetpotato in experiment 2. These parameters were examined after 3, 73, and 155 days after plantation. Vertical bars represent the means ± SE (*n* = 8). Different letters indicate significant differences among the treatments at *p* < 0.05 by Tukey–Kramer's test.

**Figure 4.** Effects of nutrient solution concentration on the shoot morphology of sweetpotato at 155 days after plantation in experiment 2. Scale bars = 47 cm.

**Figure 5.** Effects of nutrient solution concentration on number of stems (**A**) and stem diameter (**B**) of sweetpotato at 155 days after plantation in experiment 2. Vertical bars represent the means ± SE (*n* = 8). Different letters indicate significant differences among the treatments at *p* < 0.05 by Tukey–Kramer's test.

The biomass of shoots and storage roots was measured at 155 DAP. Shoot FW was the highest in HH plants, followed by LH, HL, and LL plants (Figure 6A). Storage root FW was the highest in HH plants, followed by HL and LH plants, and the lowest in LL plants (Figure 6B). The number of storage roots showed no significant difference among treatments (Figure 6C). Storage roots developed within vinyl pots

in all treatments. Storage roots were round in shape, with a short length and partially undeveloped parts (Figure 7), consistent with previous observations [43]. These morphological characteristics of storage roots exhibited no variation among the different treatments (Figure 7).

**Figure 6.** Effect of nutrient solution concentration on shoot fresh weight (**A**), storage root fresh weight (**B**), and number of storage roots (**C**) of sweetpotato at 155 days after plantation in experiment 2. Vertical bars represent the means ± SE (*n* = 8). Different letters indicate significant differences among the treatments at *p* < 0.05 by Tukey–Kramer's test.

**Figure 7.** Effect of nutrient solution concentration on the storage root morphology of sweetpotato at 155 days after plantation in experiment 2. Scale bars = 10 cm.

#### **4. Discussion**

In hydroponics, the optimal NSC varies among plant species, with EC ranging from 1.5 to 2.5 dS m−<sup>1</sup> [52,53]. Several studies have shown that high NSCs reduce the growth and photosynthetic parameters of hydroponically-grown plants [30,32,35,54–56]. High NSC-dependent growth restrictions are observed at EC > 1.8 dS m−<sup>1</sup> in peace lily and at EC > 2.8 dS m−<sup>1</sup> in peppermint and lettuce [37,54,55]. By contrast, hydroponically-grown bush snap beans can tolerate EC up to 3.6 dS m−<sup>1</sup> [33]. In the current

study, the growth of shoots and storage roots of hydroponic sweetpotatoes increased in an NSC-dependent manner up to an EC of 2.6 dS m−<sup>1</sup> (Figure 2). Given that continuous growth of sweetpotato plants in a nutrient solution with an EC of 2.6 dS m−<sup>1</sup> did not influence the leaf chlorophyll content (Figure 3A), this NSC appears to be more favorable for plant growth and storage root development rather than an osmotic stress condition that would deter growth and photosynthetic activity.

Plants sense the nutrient dose and alter the biomass partitioning accordingly [57]. In sweetpotato, the dose of N fertilizer alters the biomass partitioning of storage roots and shoots [58,59]. Increasing the N fertilizer dose from 0 to 1.2 g N per plant increases the biomass partitioning to storage roots [58]. However, at a higher N dose, the storage root biomass decreases, whereas the shoot biomass increases [58]. In experiment one, the growth of shoots and storage roots were enhanced as the NSC increased to an EC of 2.6 dS m−<sup>1</sup> (Figure 2A,B). Considering the reports that hydroponic plants have different growth characteristics compared with soil grown plants [60,61], the responsiveness of hydroponically-grown sweetpotato to the nutrient dose might be different from that of soil grown sweetpotato plants. In hydroponic potatoes, shoot growth was enhanced when NSC was increased up to EC 2.4 dS m−1, whereas tuber biomass was not affected by the EC of the nutrient solution [56]. Therefore, NSC-dependent partitioning of biomass may differ among plant species in hydroponics. The number of storage roots tend to be higher in experiment two (Figure 6C) compared to experiment one (Figure 2C). This may be caused by the different cultivation periods between experiments. Because two experiments were conducted only one time, the data may be influenced by the environmental condition.

The nutritional requirements of plants vary with the developmental stage [62]. Several studies have shown that changes in NSC influence plant growth characteristics [29,63–66]. In strawberry, restriction of N application at an early developmental stage increased the fruit biomass by enhancing reproductive growth [63]. In hydroponic tomato, increasing the NSC during fruit development reduces the fruit size and increases the sugar content [66]. Nutrient solution formulations have been developed for various growth stages in hydroponic tomato [67]. In sweetpotato, the timing of N fertilizer application influences plant growth and storage root yield [68–71]. Split application of N fertilizer could increase the storage root yield of sweetpotato by improving the efficiency of N uptake [69,70]. The timing of N fertilizer application is also important for increasing the marketable sweetpotato yield [71]. In experiment two, the storage root FW showed no significant difference between HL and LH plants (Figure 6B). This suggests that the timing of high-dose N application is not important for storage root development in hydroponically-grown sweetpotato. On the other hand, the shoot biomass and total leaf number were higher in LH plants compared with HL plants (Figure 6A, Table 1). These results implicate that higher dose of nutrient application at the storage root hypertrophic stage (the second half of growth stage) may enhance the development of shoots as well as storage roots. It should also take into account for the high abscised leaf ratio in HL plants (Table 1) because the abscised leaves, which did not contribute to shoot biomass, were partly responsible for the low shoot FW of HL plants. In general, at the late stage of sweetpotato cultivation, the storage root growth is enhanced, whereas shoot growth is retarded [59]. Thus, nutrient limiting condition at the hypertrophic stage of storage roots in HL plants may represent the field-grown sweetpotato characteristics in the shoot and root development. Because the amount of photosynthetic products translocated to storage roots partly depends on the shoot biomass, modifying the timing of NSC changes might improve the storage root yield.

Leaf abscission occurs during the senescence process and is induced by various stress responses [72]. Before the onset of leaf cell separation, the abscission zone encounters the repression of auxin biosynthesis and enhancement of ethylene production and sensitivity, resulting in the activation of cell wall degradation enzymes [73–75]. In experiment two, leaf abscission was induced at the late stage of cultivation in all NSC treatments (Table 1). This growth stage-dependent leaf abscission in sweetpotato has also been observed in open field conditions [76–78], suggesting a consistent senescence related phenomenon. N or P limitation is

known to induce leaf abscission by enhancing ethylene production and sensitivity [79]. Sweetpotato leaves also abscise at the late growth stage under low N or P condition [80]. In HL plants, plant shoot biomass (the number of attached leaves and maximum shoot length) increased during the first half of the cultivation period in the nutrient solution with high EC (Figure 3B,C); however, these shoots grew in relatively poor nutrient conditions during the second half of the cultivation period. These nutrient poor conditions might trigger the high ratio of leaf abscission associated with N or P deficiency. Leaf senescence is accompanied by the breakdown of chlorophyll [81]. HL and LL plants showed a higher abscised leaf ratio and lower relative chlorophyll content (Figure 3A, Table 1), suggesting accelerated leaf abscissions by the progression of senescence. N deficiency also causes oxidative stress to the leaf [82]. Given that oxidative stress could trigger leaf abscission [83–86], it is possible that HL plants exhibit leaf abscission during the second half of the cultivation period due to oxidative stress triggered by N deficiency. On the other hand, LH plants were relatively nutrient-rich condition at the late cultivation period. Therefore, leaf senescence and abscission were thought to be suppressed by relatively rich-N supplement.

## **5. Conclusions**

Compared to traditional soil culture systems, sweetpotato hydroponics saves absorbent material (soil) and can be used anywhere exposed to sunlight. In addition, hydroponics can efficiently utilize nutrient components as supplied components are not dispersed to the soil. In fact, almost all nutrients were absorbed in plants grown on EC 0.8 and 1.4 in this study (Supplementary Figure S1). Here, we presented NSC-dependent storage root yield in hydroponic sweetpotato (experiment one). Although the timing of high and low NSC did not have a significant impact on the storage root yield, shoot growth was apparently increased by high NSC (experiment two). A more precise adjustment of the NSC may increase the yield of storage roots relative to the fertilizer input. Thus, given its flexibility in manipulating the nutrient status, hydroponics could be used as an efficient tool for sweetpotato production.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4395/10/11/1708/s1, Figure S1: Time-course changes of nutrient solution EC in experiment 1. The nutrient solution was renewed on June 26.EC was measured two containers of each experimental plot.

**Author Contributions:** Conceptualization, M.S. and T.S.; formal analysis, M.S.; investigation, M.S.; writing—original draft preparation, M.S.; writing—review and editing, M.S. and T.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was partly funded by the Japan Society for the Promotion of Science (JSPS), KAKENHI Grant Number 20K12247 to T.S. and 20K06329 to M.S.

**Acknowledgments:** The authors thank Kokoro Kubota, Takashi Tobe, Kaito Nakano, Daisuke Fukada, Hyuki Kishida, and Rihito Yamabe for supporting the field experiments.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**






**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*

## **Theoretical and Experimental Analysis of Nutrient Variations in Electrical Conductivity-Based Closed-Loop Soilless Culture Systems by Nutrient Replenishment Method**

## **Tae In Ahn and Jung Eek Son \***

Department of Plant Science and Research Institute of Agriculture and Life Sciences, Seoul National University, Seoul 08826, Korea; genisleaf@outlook.com

**\*** Correspondence: sjeenv@snu.ac.kr; Tel.: +82-2-880-4564

Received: 11 September 2019; Accepted: 16 October 2019; Published: 17 October 2019

**Abstract:** In closed-loop soilless culture systems, variation in nutrients can lead to instability in the nutrient management and forced discharge of nutrients and water. Total nutrients absorbed by plants are replenished in an electrical conductivity-based closed-loop system, and fluctuation in electrical conductivity within a certain range around the initial value can be expected. However, this is not always observed in systems using conventional nutrient-replenishment methods. The objectives of this study were to analyze nutrient variation in a closed-loop soilless culture system based on a theoretical model and derive an alternative nutrient-replenishment method. The performance of the derived alternative method was compared with a conventional nutrient-replenishment method through simulation analysis. A demonstration experiment using sweet peppers was then conducted to confirm whether the theoretical analysis results can be reproduced through actual cultivation. The average amounts of injected nutrients during the experimental period of four months in the conventional and alternative methods were 2257 and 1054 g, respectively. There was no significant difference in the yield of sweet peppers between the two methods. The substrate electrical conductivity in the alternative method was maintained at 2.7 dS·m−<sup>1</sup> <sup>±</sup> 0.5 within the target electrical conductivity value, while that in the conventional method gradually increased to 5.0 dS·m−<sup>1</sup> <sup>±</sup> 1.2. In a simulation study, results similar to the demonstration experiment were predicted. Total nutrient concentrations in the alternative method showed fluctuations around the target value but did not continuously deviate from the target value, while those in the conventional method showed a tendency to increase. As a whole, these characteristics of the alternative method can help in minimizing nutrients and water emissions from the cultivation system.

**Keywords:** growing medium; nutrient uptake; nutrient variation; simulation model; sweet pepper

### **1. Introduction**

Closed-loop nutrient-management techniques are essential for sustainable soilless cultures with resource savings [1]. Nutrients in soilless culture systems are managed primarily with an open-loop nutrient supply [2,3]. Open-loop soilless culture systems are easier to implement, but resource losses are inevitable. Moreover, due to the intensive use of fertilizers, the threat posed to aquatic environments by repeated discharging a certain ratio of drainage is serious enough to warrant regulation by national governments [3–6]. Since a closed-loop soilless culture system reuses its drainage, the resulting variation in nutrient concentration can significantly affect the plant growth as the reuse period becomes longer [5,7–9]. It is therefore difficult to intuitively explain or interpret nutrient-variation management techniques, unlike open-loop systems. In order to appropriately apply those techniques, theoretical models are required and the problems should be precisely defined [10].

In both closed-loop and open-loop soilless culture systems, the electrical conductivity (EC) of the nutrient solution in the mixing tank is adjusted to a target value before the solution is applied to the plant [9,11,12]. However, unlike an open-loop system, the mixing ratio of tap water to stock solution in a closed-loop system is adjusted by considering the change in nutrient concentration due to the inflow of drainage [12]. Alternatively, in simple systems, a premixed standard nutrient solution of a certain EC is supplied based on the difference between initial and current water levels in the circulation tank, which simultaneously performs drainage collection and nutrient-solution feeding [13,14]. In an EC-based closed-loop soilless culture system, supply of stock solution or standard nutrient solution and tap water is intended to replenish nutrients and water consumed in the system [12]. For a single system in which the plants are grown directly in a nutrient solution container, the nutrients and water consumed due to absorption of plants in the system can be estimated almost exactly [15]. However, errors may occur in systems in which the root zone and nutrient supply are separate from drainage collection. Both elements are widely used in commercial farming conditions.

Considering the functional objective of nutrient and water replenishment in a closed-loop soilless culture system, relatively stable fluctuations within a certain range around the initial EC value should be observed. However, EC changes far exceeding the initial values in the system have generally been observed [13,14,16]. In addition, the effects of these fluctuations are linked to forced discharge of recirculated nutrient solution outside of the system [13,14]. The problems associated with variations in nutrient concentration or EC observed in soilless culture systems are presumed to be inevitable due to the nutrient uptake concentration affected by the environment [5,14]. The experimental results are interpreted depending on the responses of the system according to the treatment application [9,14,16–18], and these have proven difficult to interpret in an integrated way. As a result, technical approaches to managing nutrient variation and the design of experiments are limited. To block nutrient emissions from a soilless culture system, nutrient reuse practices must be standardized, which requires a precise problem definition based on variations of nutrient concentrations or EC.

The objectives of this study were to analyze the cause of EC variation in closed-loop soilless culture systems based on a theoretical model, to derive an alternative nutrient-replenishment method for managing nutrient fluctuation, and to evaluate the performance through theoretical and experimental analyses.

#### **2. Materials and Methods**

#### *2.1. Soilless Culture System Model*

The model used in this study simulated nutrient changes in a soilless culture system with an automated nutrient-mixing system (Figure 1). The basic structures of the soilless culture system and plant growth models were constructed by referring to the nutrient transport model in a substrate condition [6,19–21]. The measured data of incident radiation intensity in the greenhouse from 10 September 2011 to 9 March 2012 were used as an input variable of transpiration and irrigation control in the simulation. Some units of parameters and variables were converted from the references for simulating the minute-based time scale of the automated soilless culture system. Values and description of the parameters used in the simulation were summarized in Table 1.

**Figure 1.** Schematic description of a closed-loop soilless culture system in a simulated condition. Solid lines indicate water and nutrient flow, and dotted lines indicate data flow for nutrient solution mixing and irrigation control. CM and AM mean conventional and alternative nutrient-replenishment methods, respectively.


**Table 1.** Parameters used for the simulations of soilless culture system.

## *2.2. Water and Nutrient Transport in a Substrate*

According to standard practices for automated irrigation of a soilless culture system, the mixing process for a nutrient solution is initiated in the mixing tank, and the nutrient solution is supplied to the substrate after mixing. The target nutrients for the simulation were selected as macronutrient cations (K<sup>+</sup>, Ca2<sup>+</sup>, and Mg2<sup>+</sup>).

$$\frac{dV\_n}{dt} = Q\_{n-1} - Q\_n - T\_n \tag{1}$$

The volume of water in a substrate layer (*Vn*, L) was calculated depending on the flow rate of the water from the former (*Qn*−1, L min<sup>−</sup>1) and to the next layer (*Qn*, L min<sup>−</sup>1) and the evapotranspiration rate (*Tn*, L min<sup>−</sup>1). The flow rate of the water to the first substrate layer (*V*1) was the irrigation flow rate (*Q*0). *Qn* is the difference between the flow rate of the water from the former layer and the evapotranspiration rate (*Qn*−<sup>1</sup> − *Tn*) or the difference between the irrigation rate and the evapotranspiration rate in the first substrate layer (*Q*<sup>0</sup> − *T*1) [23]. The field capacity (*F*, dimensionless) and difficult available water (*WDAW*, dimensionless), respectively restrict *Qn* and *Tn*. The flow for *Qn* occurs only when *Vn* > *Ssub*,*nF*, and *Tn* flows only when *Vn* > *Ssub*,*nWDAW*. *Ssub*,*<sup>n</sup>* (L) is volume of the substrate layer *n*.

The flow of nutrients in the medium is generated by the flow rate of water.

$$V\_n \frac{d\mathbf{C}\_n^l}{dt} = Q\_{n-1}\mathbf{C}\_{n-1}^l - Q\_n\mathbf{C}\_n^l - P\_{RSA}\mathbf{I}\_n^l \tag{2}$$

*C* is the molar concentration of nutrient (mM), superscript *I* is the type of ions (K+, Ca2+, Mg2+), *J I <sup>n</sup>* is the uptake rate of nutrients (mmol m−<sup>2</sup> min−1), and *PRSA* is the specific root surface area (m2), which is described as the root length density and specific root surface area.

#### *2.3. Plant Variables and Growth Parameters*

In this simulation, evapotranspiration and nutrient uptake rates were applied as plant variables in the substrate. In general, the plant parameters relate to changes in evapotranspiration and nutrient uptake rates with plant growth. The relationship between solar radiation and evapotranspiration is adjusted by the leaf area index [24]. The parameters related to the nutrient uptake rate are derived from the characteristics of the plant ion transporters and are modeled as increasing with growth of the root surface area [25].

#### 2.3.1. Leaf Area Index

The Boltzmann sigmoid equation was used to apply changes in the leaf area index to the evapotranspiration rate:

$$P\_{LAI,t} = \frac{a\_{LAI}}{\left[1 + e^{\frac{x\_0 - t}{b\_{LAI}}}\right]} \tag{3}$$

where *aLAI*, *bLAI*, and *x*<sup>0</sup> are constants, and *t* is time.

#### 2.3.2. Evapotranspiration

The evapotranspiration rate was modeled using the simplified Penman-Monteith equation by Baille et al. (1994) [24].

$$T\_n = a\_T \left[ 1 - e^{-k\_{LAl}P\_{LAl}} \right] \frac{R}{A} + b\_T \tag{4}$$

*Tn* (L min<sup>−</sup>1, numbers were converted from kg min−1) was calculated depending on the radiation for a minute (*R*, MJ m−<sup>2</sup> min−1), the latent heat of vaporization (λ, MJ kg−1), the light extinction coefficient (*kLAI*), and the leaf area index (*PLAI*). *aT* (dimensionless) and *bT* (kg m−<sup>2</sup> min−1) are regression parameters.

#### 2.3.3. Root Length Density and Specific Root Surface Area

Root length density was used to calculate the specific root surface area and modeled using a logistic function of time [20,26]:

$$P\_{lcn,t} = \frac{KLD\_{max}}{1 + K\_1e^{-k\_1t}}\tag{5}$$

$$P\_{RSA,t} = 2\pi r\_0 P\_{len,t} \tag{6}$$

where *RLDmax* (m m−3) is the maximal root length density, and *K*<sup>1</sup> and *k*<sup>1</sup> are coefficients. *r*<sup>0</sup> is the mean root radius (m). Root length density was set to start at the top layer of the substrate and be sequentially assigned to the subsequent layer as the value increased. The allocation of root length density for each layer was calculated by dividing *RLDmax* by the total number of layers.

#### 2.3.4. Nutrient Uptake

The nutrient uptake rate of the plant in the substrate was simulated as a function of Michaelis–Menten:

$$J\_n^I = \frac{J\_{\max}^I \left(\mathbb{C}\_n^I - \mathbb{C}\_{\min}^I\right)}{K\_m^I + \left(\mathbb{C}\_n^I - \mathbb{C}\_{\min}^I\right)}\tag{7}$$

where *J I max* (mmol m−<sup>2</sup> min−1) is the maximum absorption rate of nutrient *I*, *K<sup>I</sup> <sup>m</sup>* (mM) is the Michaelis-Menten constant, and *CI min* (mM) is the minimal concentration at which *J I <sup>n</sup>* = 0.

#### *2.4. Mixing of Nutrient Solutions*

The conventional mixing process for stock solution, tap water, and drainage under the automated closed-loop soilless culture system is performed in the mixing tank [11,12]. When the system receives an irrigation command, the entire volume of collected drainage is diluted with tap water within the range of the irrigation volume, and the stock solution is added to the target EC.

However, because drainage is included in the automated mixing process in closed-loop soilless culture systems, the Equation needs to solve for target EC with mixing stock solution, drainage, and water [12]. The nutrient solution mixing process occurs intermittently according to the irrigation interval, and the basic Equation for conventional nutrient replenishment can be summarized based on the dilution Equation:

$$V\_T \mathbf{C} \mathbf{r} = V\_D \mathbf{C} \mathbf{p} + V\_W \mathbf{C} \mathbf{w} + V\_S \mathbf{C}\_S \tag{8}$$

$$V\_w = V\_T - V\_D - V\_S \tag{9}$$

$$V\_S = \frac{\mathbb{C}\_T V\_T - \mathbb{C}\_W V\_T + \mathbb{C}\_W V\_D - \mathbb{C}\_D V\_D}{\mathbb{C}\_S - \mathbb{C}\_W} \tag{10}$$

where *VT* (L) is the target irrigation volume per event, *CT* (mEq L−1) is the target total equivalent concentration, *VD* is the drainage volume, *CD* (mEq L−1) is the total equivalent concentration in drainage, *VW* (L) is the amount of tap water input to the mixing tank, *VS* (L) is the amount of stock solution input to the mixing tank, *Cw* (mEq L<sup>−</sup>1) is the total equivalent concentration in tap water, and *CS* (mEq L<sup>−</sup>1) is the total equivalent concentration of the stock solution. Equation (8) can be summarized as Equation (10) by substituting Equation (9) for *VW*. Equation (10) is calculating the amount stock solution input based on the total equivalent concentration. In this simulation, we assumed the total equivalent concentration as EC based on the linear relationship between EC and the total equivalent concentration of nutrient solution presented by Savvas and Manos (1999) [27].

The amount of stock solution input to the mixing tank was calculated through this process, and when the irrigation control command was generated during the simulation, the mixing process began based on the volume of drainage stored in the drainage tank at that moment. If the calculated value of the Equation (10) was less than zero, dilution using tap water could not be adjusted to the target concentration within the range of irrigation amount. In this case, the amount of tap water

required for diluting the drainage to target total equivalent concentration (*CT*) was calculated, and then the ratio between the drainage and calculated tap water was multiplied by *VT*. When the doses of *VD*, *VW*, and *VS* were determined through the abovementioned calculation, a flow rate was generated until the corresponding amount was transferred to *VM* according to *Qdrg*, *Qwtr*, and *Qstk*, respectively. In the simulation, irrigation was controlled by a radiation integral method, which is conventionally used in automated irrigation control [28]. 140 mL of nutrient solution per plant in the mixing tank were supplied whenever the accumulated radiation reached 100 J m<sup>−</sup>2.

## *2.5. Experimental Analysis*

### 2.5.1. Cultivation Conditions

Three sweet pepper (*Capsicum annuum* L. "Derby") plants were grown in a rockwool slab, and seven slabs were used per row. Four cultivation lines were installed in a Venlo-type greenhouse at the experimental farm of Seoul National University (Suwon, Korea, Lat. 37.3◦ N, long. 127.0◦ E). Each line was an independent closed-loop soilless culture system with a mixing tank, drainage tank, and stock solutions. The stock solution was prepared based on the PBG nutrient solution of the Netherlands. In the greenhouse, daytime temperature was maintained at 25–35 ◦C and nighttime temperature at 17–22 ◦C. The solar radiation-based irrigation control was applied; when the cumulative radiation measured by a pyranometer (SP-110-L10, Apogee Instruments, Logan, Utah, USA) reached 100 J cm<sup>−</sup>2, 150 mL of the nutrient solution was supplied to each plant. However, the irrigation amounts were adjusted according to meteorological conditions to maintain a drainage ratio of approximately 30%. The composition of nutrient solution was 14.17 mM of NO3 <sup>−</sup>, 1.14 mM of H2PO−, 5.92 mM of K+, 4.43 mM of Ca2<sup>+</sup>, 1.59 mM of Mg2<sup>+</sup>, and 1.6 mM of SO4 2– as macro-elements; and 0.019 mM of Fe2<sup>+</sup>, 0.01 mM of Zn2<sup>+</sup>, 0.002 mM of Cu2+, 0.01 mM of Mn2<sup>+</sup>, and 0.0005 mM of MoO4 2– as micro-elements. After an irrigation event, the drainage solution was returned to the drainage tank (11.7 L). The EC and pH of tap water were 0.17 dS·m−<sup>1</sup> and 7.11, respectively, and contained 0.21 mM of Na+, 0.29 mM of Cl−, 0.04 mM of K<sup>+</sup>, 0.36 mM of Ca2<sup>+</sup>, 0.11 mM of Mg2<sup>+</sup>, 0.10 mM of SO4 <sup>2</sup>−, 0.39 mM of NO3 −, and 0.0 mM of PO4 3–.

#### 2.5.2. Measurement of Fruit Yield and Analyses of Nutrient Content in Leaves and Substrate

The total yield and average fruit weight during the experimental period were measured. The proportion of blossom-end rot (BER) fruits on a sweet pepper plant was measured. At the end of the experiment, 18 leaves (including petiole) from the middle to the top nodes of a sweet pepper were randomly collected from each treatment. Leaves were washed in tap water and dried for 48 h at 70 ◦C in an oven. The dried leaves were ground, and 0.5 g of each ground sample was digested using concentrated nitric acid. Next, 1 mL of concentrated perchloric acid was added to maintain a set solution temperature of 180 ◦C, and the digestion process was accelerated on a hot plate at 90 ◦C for approximately one h, until a clear-colored solution was obtained. After digestion, the tube was cooled, filled with 25 mL deionized water, and the total contents of K<sup>+</sup>, Ca2<sup>+</sup>, and Mg2<sup>+</sup> in the leaves were determined with an inductively coupled plasma-optical emission spectrometer (ICP-730ES, Varian, Mulgrave, Australia). To determine the nutrient concentrations in the rockwool substrate, samples of nutrient solution in the rockwool slabs were extracted using a syringe. The collection points of the nutrient solution in the rockwool slab were randomly selected to ensure representative samples of the overall concentration in the rockwool slabs. Five 10 mL samples of a rockwool slab nutrient solution were collected for each extraction, for a final volume of 50 mL sample. Four 50 mL samples per treatment were collected every week. SAS (version 9.2, SAS Institute, Cary, NC, USA) was used for statistical analysis.

#### 2.5.3. Nutrient-Replenishment Method

A conventional nutrient-replenishment method (CM) and an alternative nutrient-replenishment method (AM) derived from the theoretical analysis in this study were performed in the mixing tank with two applied nutrient solution mixing modules. In the CM, as explained in Section 2.4, when the system received an irrigation command, the entire drainage volume was diluted with tap water within the range of irrigation volume, and the stock solution was added to match the fixed target EC [12]. In the case when the calculated volume of the diluted drainage exceeds the irrigation volume, the injection ratio of drainage and water was multiplied by the irrigation volume, and the converted drainage and water volumes were injected into the mixing tank without injection of the stock solution. In the AM, the additional volume of the stock solution was determined by the equation derived from the simulation study at every irrigation event (Equation (14)).

#### 2.5.4. Nutrient Solution Mixing Module and Data Collection

The ECs of the nutrient solutions in the mixing tank and drainage tank were measured by EC sensors (SCF-01A, DIK, Chuncheon, Korea). Light intensity in the greenhouse was measured with a pyranometer (SP-110, Apogee, Logan, UT, USA) and used for input data for solar radiation-based irrigation control. Data were measured every 10 s from 15 October to 31 December 2014. Mean values for every hour were used. A datalogger (CR1000, Campbell Scientific, Logan, UT, USA) was used to measure and control the drainage and nutrient mixing process. Water levels of the stock solution tanks and the drainage tanks were monitored by ultrasonic sensors (UHA-300, Unics, Daegu, Korea) and used to estimate the stored volume changes of stock and drainage solutions. ECs in the substrates were measured at intervals of two to five days using a multimeter (Multi 3420 SET C, WTW, Weilheim, Germany).

#### **3. Results and Discussion**

#### *3.1. Theoretical Analysis: Reconsideration of Problem and Derivation of Possible Solution*

The total concentration of nutrients in the system using CM for nutrient replenishment gradually increased with diurnal level fluctuations, and after approximately 60 days, the total concentration showed repeated fluctuations within a certain range (Figure 2a). The changes with an increasing tendency in total nutrient concentrations relative to initial values have been typically reported in most EC-based closed-loop, semi-closed-loop, and open-loop soilless culture systems [8,13,14,16,29]. Theoretically, the concentration of nutrient solutions in the substrates can be explained by the difference between the concentration of irrigated solution and the concentration of nutrient uptake when the boundary area is limited to a substrate [5]. This can simply explain the nutrient variations in open-loop soilless culture systems. In closed-loop soilless culture systems, on the other hand, the concentration of irrigated solution is also affected by the drained solution, but most of studies on nutrient variations in closed-loop systems have been carried out with a premise that nutrient variations are the result of the changing dynamics of uptake concentrations [5,14,16,30–32].

The total amount of nutrients in the system using CM also increased with time (Figure 2b). In a closed-loop system, the changes in the total amount of nutrients can be interpreted more straightforwardly. The increasing tendency in the total amount of nutrients indicates the accumulation of surplus nutrients supplies. However, most of the previous studies did not attempt to interpret the fluctuations from the perspective of total amount of nutrients. Thus, our theoretical analyses reconsider problems for the nutrient concentration changes in the closed-loop soilless culture system; the nutrient fluctuation with increasing tendency is mainly caused by the accumulated difference between nutrient uptake and replenishment.

**Figure 2.** Changes in total ion concentration in the substrate (**a**) and total ions in the system (**b**) in the closed-loop soilless culture system using conventional (CM) and alternative (Equation (12) applied) nutrient-replenishment methods.

In a simple cultivation system sharing root-zone and nutrient solutions in a single container, measurement of changes in nutrient concentration, and water volume in the container corresponds to the actual amount of consumed nutrients in the system [15]. On the other hand, a typical soilless culture system consists of subsystems, including mixing tank, drainage tank, and substrates [6,12]. However, in the conventional nutrient solution mixing method of closed-loop soilless culture, the amount of nutrients replenishment has been mainly determined as a function of EC and volume of irrigation water and drainage [12]. Thus, to remove the errors between the actual nutrient consumption by plants and nutrients supplies in the closed-loop soilless culture system, the determination of the replenishment amount should consider the system-wide nutrients and water. We summarized equations for the estimation of nutrient consumption in the typical soilless culture system as Equation (11) and for the determination of nutrient replenishment as Equation (12)

$$V\_{\rm init} \mathbb{C}\_{\rm init} = V\_{dry} \mathbb{C}\_{dr\_X} + V\_{\rm sub} \mathbb{C}\_{\rm Sub} + V\_{\rm mix} \mathbb{C}\_{\rm mix} + V\_{\rm II} \mathbb{C}\_{\rm II} \tag{11}$$

$$V\_{stk} = \frac{\mathbb{C}\_{init} V\_{init} - \mathbb{C}\_{drg} V\_{drg} - \mathbb{C}\_{mix} V\_{mix} - \mathbb{C}\_{sub} V\_{sub}}{\mathbb{C}\_{stk}} \tag{12}$$

where *Vinit* is the initial volume of water in the system; *VU* is the amount of water absorbed by the plant; *Cinit* is the initial total concentration of the system; *CU* is the average total nutrient uptake concentration; *Vdrg*, *Vsub*, *Vmix*, and *Vstk* are the volumes of water stored in the drainage tank, substrate, mixing tank, and the input volume of stock solution, respectively; and *Cdrg*, *Csub*, *Cmix*, and *Cstk* are the total nutrient concentrations in the drainage tank, substrate, mixing tank, and the stock solution concentration, respectively.

In the calculation using Equation (12) for nutrient replenishment by stock solution, the total nutrient concentration showed repeated fluctuations near the initial concentration (Figure 2a). The amount of total nutrients in the system also stayed near the initial value without any apparent increasing or decreasing tendency (Figure 2b).

Precise measurements for the variables in Equation (12) in a real cultivation system have technical limitations. In particular, the amounts of total nutrients *Csub* and *Vsub* in the substrate are difficult to estimate. In a soilless culture system, the field capacity (*F*) of a substrate corresponds to the parameters of the system, and the volume of water cannot exceed the volume of the substrate multiplied by the field capacity. The EC of the drainage (*Cdrg*) can be indicative of a change in concentration of substrate. Considering this, we can modify Equation (12) as follows for an alternative nutrient-replenishment method (AM):

$$V\_{stk} = \frac{\mathbb{C}\_{init} V\_{init} - \mathbb{C}\_{dry} V\_{dry} - \mathbb{C}\_{mix} V\_{mix} - \mathbb{C}\_{dry} F}{\mathbb{C}\_{stk}} \tag{13}$$

When the EC of the drainage (*Cdrg*) and the field capacity (*F*) are substituted for *Csub* and *Vsub*, respectively, errors may occur. However, in this case, total ion concentration fluctuated around the initial concentration (Figure 3).

**Figure 3.** Changes in total ion concentration in the substrate according to the nutrient-replenishment method (**a**) and mean and standard deviation of total ion concentration in the substrate according to the nutrient-replenishment method (**b**). CM is the conventional nutrient-replenishment method and AM is alternative nutrient-replenishment method (Equation (13) applied).

In the existing problem definition, the EC variation in the closed-loop soilless culture system was derived from the dynamic change in nutrient uptake concentration [14,16,30–32]; thus, there were restrictions on active control and interpretation. However, a series of analysis steps leading to Equation (13) makes it possible to convert EC control in the closed-loop soilless culture system to the problem of proper gain search through arbitrary adjustment of system parameters. That is, in Equation (13), all but *Cdrg* can be viewed as parameters and the process of calculating the difference between *CinitVinit* and the product of the parameters and *Cdrg* is performed in every mixing process.

### *3.2. Experimental Analysis: Demonstration Experiment for the Theoretical Analysis*

The AM showed stable changes in the EC control of substrate and drainage against the CM (Figure 4). While the EC of substrate and drainage in the AM was maintained near the initial value of the system, an increasing tendency in stored drainage volume in the drainage tank was not observed (Figure 5). The average level of stored drainage level in the CM was higher than in the AM, and the range of variation was relatively wider (Figure 5).

**Figure 4.** Comparison of electrical conductivity (mean ± SD) in the rockwool substrate (**a**) and the drainage (**b**) of the closed-loop soilless culture system during the experimental period between conventional (CM) and alternative (AM) nutrient-replenishment methods.

**Figure 5.** Changes in stored drainage volume in the drainage tank (**a**) and box-plot comparison between conventional (CM) and alternative (AM) nutrient-replenishment methods (**b**) during the experimental period.

The mixing ratio of drainage, water, and stock solution in the conventional nutrient solution mixing process depends on the target EC for the irrigation solution. However, this aspect could generate significant fluctuations in the stored volume of drainage. No increasing or decreasing trend in EC or stored drainage volume can be inferred over the entire experimental period in the closed-loop system, meaning that total nutrient input to the system adequately followed total nutrient uptake by the plant. In the CM, the EC of the rockwool substrate was relatively higher, and gradual increase was observed. The EC in the substrate can eventually be reflected in the EC of the drainage. A high EC value in a closed-loop soilless culture system where concentration control of the recycled nutrient solution is carried out can lead to an increase in the volume of stored drainage solution and subsequently to discharge of drainage when it exceeds system capacity [13,14]. This can be a factor in system instability. The EC changes in the rockwool substrate of the AM applied system indicate a normal effect of the proportional gain adjustment, as in the theoretical analysis in this study.

The cumulative amount of nutrients supplied to the system using the AM increased at a low rate in comparison with the CM, and the final amount of supplied nutrients was also lower than that of the CM; 1054 g for AM and 2257 g for CM, respectively (Figure 6). The AM appeared to work normally, and a reduction in fertilizer input compared with the CM was also observed. In addition, measurement of cumulative amount of nutrients in a state with no overall increases in EC and stored

drainage volume were not observed indicates that the system can detect the total nutrient requirement of a plant. This measure could be used as an as index for plant nutritional status, one that is not provided in the CM.

**Figure 6.** Accumulated amounts of fertilizers injected into the soilless culture systems with conventional (CM) and alternative (AM) nutrient-replenishment methods.

In the case of stock solution input volume change, it was confirmed that the input amount of the AM was relatively evenly distributed during the cultivation period (Figure 7b). On the other hand, in the case of the CM, a concentrated period of nutrient solution injection occurred, and relatively long periods during which the input of stock solution was blocked were observed (Figure 7a). The irregular feeding rate of the stock solution could be an adverse factor in nutrient-balance control when nutrient correction in the system is performed by input of stock or standard nutrient solution [13,14,32].

**Figure 7.** Changes in volume of injected stock solution with conventional (CM, **a**) and alternative (AM, **b**) nutrient-replenishment methods.

In the CM, overall tendencies of increasing Ca2<sup>+</sup> and Mg2<sup>+</sup> and decreasing K<sup>+</sup> were observed (Figure 8). In the AM, Ca2<sup>+</sup> and Mg2<sup>+</sup> concentrations were stable at a level relatively close to the initial value, but K<sup>+</sup> values showed a rapid decline and then fluctuated at a low concentration (Figure 8a–c). For CM, variations in nutrient concentrations similar to those reported in previous studies were observed [9,14,16]. Previous research on closed-loop soilless culture systems has determined that nutrient variations are a result of dynamic changes in nutrient uptake concentrations, and following those changes is challenging. [5,14,30,33]. However, Figure 8 indicates that a more deterministic change

occurred in the system when nutrient replenishment was synchronized with total nutrient uptake through the AM system.

**Figure 8.** Changes in nutrient concentrations (mean <sup>±</sup> SD) of K<sup>+</sup> (**a**), Mg2<sup>+</sup> (**b**), and Ca2<sup>+</sup> (**c**) and changes in cumulative standard deviation of nutrient concentrations of K<sup>+</sup> (**e**), Mg2<sup>+</sup> (**f**), and Ca2<sup>+</sup> (**g**) in the rockwool substrates using the conventional (CM) and alternative (AM) nutrient-replenishment methods, respectively.

The cumulative standard deviations of the AM were maintained at a lower level than those of the CM during the entire experimental period, and gradually decreasing tendencies were observed in K<sup>+</sup> and Mg2<sup>+</sup> for the AM (Figure 8e–g). This means that the changes in nutrient concentration in the AM applied system were maintained close to the average concentration values during the experimental period compared with the CM. Considering the nutrient variations of the AM system itself, there may be a limit to defining it as steady state in the strict sense. However, in the actual cultivation conditions in this experiment, input of nutrients and water into the root zone by irrigation occurs intermittently, and the variation in the section where no input occurs cannot be controlled until the next input event. Furthermore, the frequency of changes of such input can affect system fluctuations [34–36], and the AM applied system is also under this influence. Considering these constraints and the CM changes, it can be assumed that the AM entered an average steady state that fluctuated within a certain range. The nutrient concentration control in the soilless culture system can therefore be seen as shifting the fluctuation range of the average steady state to the target range through a compositional change in the stock nutrient solution.

However, because the K<sup>+</sup> concentration of the AM was maintained at a very low level in this study, the impacts on sweet pepper productivity need to be considered [37]. Total sweet pepper yields during the experiment were compared (Figure 9). The average total yield was 827.5 g per plant (standard deviation [SD] ±106.5) in the CM and 838.8 g per plant (±109.8) in the AM, and statistically significant differences were not observed (*t*-test, *P* > 0.05; *n* = 10 per treatment). The average fruit weights were 133.7 g (±35.2) and 137.8 g (±38.6) for the CM and AM, respectively, but no significant effect was observed.

**Figure 9.** Comparisons (mean ± SD) of total yield (**a**) and fruit weight (**b**) of sweet pepper during the experimental period between conventional (CM) and alternative (AM) nutrient-replenishment methods (*t*-test). NS: Not significant (*P* > 0.05); *n* = 10 per treatment.

In the case of blossom-end rot, the mean value was low in the AM but not by a significant difference (Figure 10). This is considered to be due to the difference in concentration of the root zone when considering the characteristics of sweet pepper responses to root zone nutrient concentration [38].

**Figure 10.** Comparison of blossom-end rot (mean ± SD) of sweet pepper between conventional (CM) and alternative (AM) nutrient-replenishment methods (*t*-test). NS: Not significant.

When comparing the changes in the nutrient ratio in the substrate during the experiment, the AM showed a tendency to accumulate calcium (Figure 11), but it was not in the range of physiological limitations of Steiner's standard [39]. Leaf analysis confirmed that absorption selectivity is maintained by achieving the ratio range of standard nutrient solutions, unlike the ratio of nutrients in the substrate nutrient solution (Figure 11). In the AM, the concentration of K<sup>+</sup> was maintained at a low level, but the supply interval of the stock solution was relatively uniformly distributed, resulting in a periodic supply. That could correspond to the prevention effect of nutrient deficiency through the constant feeding rate of nutrients even at lower concentrations [40].

**Figure 11.** Nutrient balance changes in the rockwool substrates and dried leaves using the conventional (CM) and alternative (AM) nutrient-replenishment methods.

Previous studies and techniques for the EC-based closed-loop soilless culture systems interpreted the nutrient variations mainly focused on the discrepancies between supplied nutrient concentrations and uptake concentrations. Due to the dynamic features in the uptake concentrations and seemingly complex changes of each nutrient in the substrate, this has been a limiting factor in the systematic approach and the development of appropriate technologies so far. Therefore, most of the studies have been carried out through relative comparison by controlled experiments. However, there was no proper theoretical platform for nutrients variation in the closed-loop soilless culture system, so the stability of the cultivation has been verified by changing the terminal factors such as the irrigation, composition of the nutrient solution, and reuse period [9,14,16–18,41,42]. Our study redefined the problem of nutrient variation control in the EC-based closed-loop soilless culture system in the whole system perspective through the theoretical analysis and deduced the proper solution. The experimental results showed theoretically-predicted behaviors in the EC variation control. In addition, the ion concentrations showed convergent changes, which are providing a basis for future studies for technical advancement.

#### **4. Conclusions**

The effects of synchronized total nutrient supply on total nutrient uptake by the alternative nutrient-replenishment method (AM) were confirmed and compared with those of the conventional nutrient-replenishment method (CM) in the soilless culture system for sweet pepper cultivation. In the AM, electrical conductivity (EC) was maintained close to the initial value, and the use of fertilizers was reduced by about 45% without significant yield losses compared with the CM. This could mean that a closed-loop soilless culture system, showing complicated nutrient variations, can be stably controlled. Through this study, the problem of EC variation in closed-loop soilless cultures was theoretically analyzed. In addition, more advanced and sustainable control techniques could be applied based on the problem definition provided by this study and repeated experiments for other crops are required to ensure the on-site feasibility.

**Author Contributions:** Formal analysis, T.I.A.; investigation, T.I.A. and J.E.S.; methodology, T.I.A. and J.E.S.; supervision, J.E.S.; writing—original draft, T.I.A.; writing—review & editing, J.E.S.

**Funding:** This work was supported by the Korea Institute of Planning and Evaluation for Technology in Food, Agriculture, Forestry and Fisheries (IPET) through the Agriculture, Food and Rural Affairs Research Center Support Program funded by the Ministry of Agriculture, Food and Rural Affairs (MAFRA; 717001-07-1-HD240).

**Conflicts of Interest:** The authors declare no conflict of interest.

## **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Selection of Fertilizer and Cultivar of Sweet Pepper and Eggplant for Hydroponic Production**

## **Hardeep Singh 1,\*, Bruce L. Dunn 1, Mark Payton <sup>2</sup> and Lynn Brandenberger <sup>1</sup>**


Received: 6 July 2019; Accepted: 5 August 2019; Published: 7 August 2019

**Abstract:** Dutch bucket hydroponic trials were conducted with the aim to evaluate the effects of different hydroponic fertilizers (5N-4.8P-21.6K, 5N-5.2P-21.6K, and 7N-3.9P-4.1K) on growth, fruit production, and the fruit quality (fruit shape index) parameters of two cultivars of sweet pepper (*Capsicum annuum* L.) and on two cultivars of eggplant (*Solanum melongena* L.). For sweet pepper yield, the 5N-4.8P-21.6K fertilizer was responsible for the greatest yield for both cultivars. For sweet pepper fresh and dry shoot weight interaction, the 'Orangella' cultivar had greater growth in 5N-4.8P-21.6K and 5N-5.2P-21.6K fertilizers, whereas there was no difference among cultivars in 7N-3.9P-4.1K. Shape index was not affected by fertilizers or cultivars. For the eggplant yield, there was no main effect nor interaction between fertilizers and cultivars for fruit yield, while the interaction between fertilizers and cultivars was significant for shoot fresh weight production. Shoot fresh weight was greater for 'Angela' than 'Jaylo' in 5N-4.8P-21.6K and 7N-3.9P-4.1K. Furthermore, both eggplant cultivars were affected with yellowing of fruits in all fertilizer treatments after 2 months, which was probably due to the accumulation of nutrients in the closed hydroponic system. Therefore, hydroponic producers could select 5N-4.8P-21.6K and 5N-5.2P-21.6K fertilizers for the cultivation of the 'Orangella' cultivar of sweet pepper based on yield. It is important to evaluate more fertilizers and cultivars for eggplant hydroponic cultivation.

**Keywords:** soilless culture; water soluble fertilizers; vegetables; *Capsicum annuum* L.; *Solanum melongena* L.; nutrients; shape index

## **1. Introduction**

Problems such as soil salinity, lack of fertile soil, and soil-borne diseases are causes of hindrance for vegetable production in soil. Therefore, to overcome these problems, soilless culture was developed [1]. Involving growing plants without soil, soilless culture is considered a sustainable method for the cultivation of various greenhouse vegetable crops such as tomatoes (*Solanum lycopersicum* L.), cucumbers (*Cucumis sativus* L.), peppers (*Capsicum* L.), lettuce (*Lactuca sativa* L.), Swiss chard (*Beta vulgaris* L.), and eggplant (*Solanum melongena* L.). It is considered good for increasing agriculture sustainability as well as improving environmental health [2]. Soilless culture has various classification systems and methods such as hydroponics, aeroponics, gravel culture, and rockwool culture [3–5]. Dutch bucket system was introduced in the early 1980s by Dutch and Belgian growers and is defined as a container-type hydroponics system filled with substrates to provide support to the plant and nutrient solution supplied by drippers to each container [6].

The most important factor affecting crop yield and quality in hydroponics is nutrient solution [7]. The fertilizer used in hydroponic production should have balanced amounts of essential elements and should not form any precipitates during its use [8]. In most studies, nutrient solutions such as Copper's, Hoagland and Arnon's and Yamazaki's solution, which require self-preparation, have been

evaluated in the hydroponic production of various crops. The self-preparation of nutrient solution for hydroponic production is effective for large-scale growers, whereas small scale growers face difficulties in managing nutrient concentration [9]. Therefore, commercially prepared, also known as one or two bag approach fertilizers, are gaining popularity. According to Mattson and Peters [9], a single bag fertilizer performed well for the production of peppers, cucumbers, and tomatoes at the University of Arizona Controlled Agriculture Center greenhouse. One of the reasons for the importance of a suitable fertilizer selection in hydroponics is that under field conditions, plants can influence nutrient availability by releasing root exudates or exploring new soil regions by growing their roots, while in hydroponics, it is not possible for plant roots to expand because of the confined area for root growth and the low buffering capacity of roots [10]. Furthermore, the accumulation of nutrients into plant structures may occur if nutrients are supplied in excess, posing health risks when plant products are consumed [11]. In addition, if a food product high in nitrate content is ingested, it is transformed into nitrite and subsequently nitrite, and in combination with amines, may form some carcinogenic compounds [12].

Soilless culture not only offers the possibility of growing crops with considerable savings of water and fertilizers, it is also considered as an easy and rapid method for screening cultivars of different crops for production, drought tolerance, and for physiological disorders [13]. Moreover, cultivar selection for hydroponics is not comparable to cultivar selection for field production. The data derived from field experiments for cultivar selection cannot be directly applied for hydroponic production due to the great difference in growth conditions between the two systems [14]. Some studies have evaluated the performance of sweet pepper cultivars for different objectives. Twelve sweet pepper cultivars were evaluated using a hydroponic system and it was concluded that 'Special' and 'Cupra' for red, 'Boogie', 'Fellini', and 'President' for orange, and 'Fiesta' and 'Derby' for yellow color had greater yields compared to other cultivars [15]. Mineral nutrition has the greatest impact on some physical and quality characteristics of sweet pepper, which include soluble solids, pH, fruit shape index, firmness, and pulp thickness [9]. It has also been also suggested that fruit weight and fruit shape index are two important characteristics of sweet peppers determining consumer preference and acceptability [16].

Various cultivars are available in the market for each crop, but for hydroponic cultivation, it is also necessary that the cultivar have a high economic value due to high input costs [5,8]. The yellow and orange colored cultivars of sweet peppers have a higher economic value than green colored cultivars. Therefore, 'Orangella' and 'Bentley' are orange and yellow colored cultivars of sweet pepper, respectively. Among the eggplant cultivars, 'Angela' and 'Jaylo' have been reported to have higher economic values due to their greater fruit size and white stripped fruits, respectively. Due to their high economic value, these cultivars has been tested with different objectives in hydroponic production. 'Bentley' has been tested for susceptibility to *Fusarium* spp. and other water-borne diseases in hydroponic cultivation [17]. Nevertheless, scientific literature evaluating these sweet pepper and eggplant cultivars using different commercially available hydroponic fertilizers is still lacking. Therefore, the objectives of our study were to evaluate the effect of three different commercial hydroponic fertilizers on growth, fruit production, and fruit quality (fruit shape index) parameters of different cultivars of sweet pepper and eggplant in the Dutch bucket system.

#### **2. Materials and Methods**

#### *2.1. Plant Materials and Growth Conditions*

Seeds of sweet peppers 'Bentley' and 'Orangella' and eggplants 'Angela' and 'Jaylo' were obtained from Johnny's Selected Seeds (Winslow, ME, USA) and sown on 12 February 2016. The seeds were sown in 1.5 cm<sup>3</sup> rockwool starter cubes with a sheet of 98 cubes (Grodan, Milton, ON, Canada) and transplanted into a Dutch bucket system on 20 March 2016 at the Department of Horticulture and Landscape Architecture Research Greenhouses in Stillwater, OK, USA. The average daily temperature, measured using a data logger (T & D Corporation, Nagano, Japan), was 27.2 ◦C. Light was measured

using the same sensor and the daily light integral (DLI) was calculated from this data by multiplying 7992.48 lux by 0.0185 (standard conversion factor for sunlight to convert lux to PPFD), then multiplying 172.9 μmol m <sup>−</sup><sup>2</sup> s−<sup>1</sup> by 0.0864 (standard conversion based on the total number of seconds in a day divided by 1 million) to obtain a DLI average of 12.8 mol m−<sup>2</sup> d−<sup>1</sup> for sweet pepper production [8]. For eggplant production, the average lux for the growth period was 8701.23 lux, therefore, DLI was equal to 13.90 mol m−<sup>2</sup> d−1. No nutrition was provided during nursery production. Seeds for the second replication were sown on 15 February 2017 and transplanted into the system on 23 March 2017. A single plant was transplanted into each bucket. The Dutch buckets were placed 50 cm apart and the rows were 100 cm apart and arranged on the opposite side of the irrigation and drainage pipes. Water was provided to each plant by a drip emitter, which supplied 3.75 L of water per h. Buckets were filled with expanded clay pebbles (Mother Earth Hydroton, National Garden Wholesale Sunlight Supply, Vancouver, WA, USA). The water that drained away was recirculated from a 150 L capacity storage tank using an electric pump.

## *2.2. Fertilizers*

Both crops were fertilized by 5N-4.8P-21.6K (Jack's, J.R. Peters, Allentown, PA, USA), 5N-5.2P-21.6K (Peters, J.R. Peters, Allentown, PA, USA), and 7N-3.9P-4.1K (Dyna Gro, Richmond, CA, USA). The fertilizers used in this experiment had different elemental compositions (Table 1). Fertilizers 5N-4.8P-21.6K and 5N-5.2P-21.6K did not contain calcium (Ca) in their formulation, therefore, it was recommended by the manufacturer to add calcium nitrate (CaNO3) (Haifa North America, Inc., Altamonte Spring, FL, USA) to supply Ca and a fraction of nitrogen (N). Fertilizer 7N-3.9P-4.1K contained all the recommended dosages of nutrients in one formulation. Tap water with an electrical conductivity (EC) of 0.5 dS m−<sup>1</sup> and a pH of 7.8 was used to prepare the nutrient solution.


**Table 1.** Nutrient concentrations (in ppm) of hydroponic fertilizers when 3.69 kg were dissolved in 3785.4 L of water (as suggested by manufacturer).

#### *2.3. EC, pH, and Data Collection*

Sweet pepper fruits were harvested when 80% color (yellow or orange) development occurred and eggplants were harvested when they reached full size (i.e., weighing 250–400 g). Harvesting was carried out once or twice a week depending on number and maturity stage of fruits. The EC of all the nutrient solutions was maintained at 2.5–3.5 dS m<sup>−</sup>1. If EC was higher than the recommended limit, then water was added and if EC was lower, then some fertilizer was added. The pH was maintained at 5.5–6.5 for eggplants and 5.5–6 for peppers. The commercially available product pH down (General Hydroponics, Santa Rosa, CA, USA) were used to adjust pH. This product was reported to be best among different organic and inorganic products used for pH maintenance in hydroponics [18]. The pH and EC of the solution was checked every alternate day.

At each harvest, data were collected on fruit weight and fruit shape index (for sweet pepper). Shape index was defined by the equatorial to longitudinal length ratio and calculated by dividing the maximum height (H) of fruit to the maximum width (W) of fruit (H/W) [19]. The height and width of each fruit were measured from randomly selected fruit. Nutrient analysis was conducted for the leaves of sweet peppers and eggplants. The nutrient analysis data for sweet pepper are not presented because there were no nutritional disorders in sweet pepper and nutrient concentration were within recommended limits. At the end of the trial, data were collected on fresh shoot weight, dry shoot, and root weight (shoots and roots dried for 2 days at 56 ◦C). Nutrient analysis of leaf samples was analyzed by the Soil, Water and Forage Analytical Laboratory at Oklahoma State University, using a nutrient analyzer (TruSpec Elemental Analyzer; LECO Corp, St. Joseph, MI, USA).

## *2.4. Experimental Setup and Data Analyses*

The experimental design was a split plot design with two replications over time. The factors were fertilizer (main plots, three levels) and cultivars (sub plots, two levels for each crop). The experimental unit for the fertilizer was 18 plants, while the experimental unit for the cultivar was nine plants of each crop. Therefore, for each fertilizer treatment, there were nine replicas of each cultivar. Tests of significance were performed at the 0.05, 0.001, and 0.0001 levels. Least significance difference (LSD) method was used for comparing differences between treatment means. Data analysis was generated using SAS/STAT software (version 9.4) [20].

### **3. Results**

## *3.1. Sweet Pepper*

Interactions between fertilizer and sweet pepper cultivars occurred for shoot fresh and dry weight, and average fruit weight (Table 2). Shoot fresh weight and dry weight were significantly greater for 'Orangella' as compared to 'Bentley' when fertilized with 5N-4.8P-21.6K and 5N-5.2P-21.6K (Figures 1 and 2). There was no significant difference between shoot fresh and dry weight between sweet pepper cultivars when fertilized with 7N-3.9P-4.1K (Figures 1 and 2). Average fruit weight was significantly greater for 'Orangella' as compared to 'Bentley' when fertilized with 5N-4.8P-21.6K, whereas there was no significant difference between two cultivars when fertilized with 5N-5.2P-21.6K and 7N-3.9P-4.1K (Figure 3). Average fruit weight ranged from 122–172 g.


**Table 2.** Interaction and main effect for sweet pepper ('Bentley' and 'Orangella'), eggplant ('Angela' and 'Jaylo'), and hydroponic fertilizers (5N-4.8P-21.6K, 5N-5.2P-21.6K, and 7N-3.9P-4.1K).

<sup>z</sup> NS, \*, \*\*, \*\*\* indicates non-significant or significant at *p* ≤ 0.05, *p* ≤ 0.001, or *p* ≤ 0.0001, respectively.

For fruit yield and root weight, there was a significant fertilizer effect, while there was no fertilizer or cultivar effect for shape index (Table 2). The fruit yield of sweet pepper was significantly greater in 5N-4.8P-21.6K and 5N-5.2P-21.6K as compared to 7N-3.9P-4.1K (Table 3). The root weight of sweet pepper was significantly greater in 5N-4.8P-21.6K as compared to 5N-5.2P-21.6K and 7N-3.9P-4.1K (Table 3).

**Figure 1.** Interaction between sweet pepper cultivars ('Bentley' and 'Orangella') and hydroponic fertilizers (5N-4.8P-21.6K, 5N-5.2P-21.6K, and 7N-3.9P-4.1K) for shoot fresh weight (g) per plant (n = 9). Data are presented as means ± SEM. Means with same lowercase letter are not significantly different by LSD (*p* ≤ 0.05) between cultivars within fertilizers. Means with same uppercase letter are not significantly different by LSD (*p* ≤ 0.05) among fertilizers within each cultivar.

**Figure 2.** Interaction between sweet pepper cultivars ('Bentley' and 'Orangella') and hydroponic fertilizers (5N-4.8P-21.6K, 5N-5.2P-21.6K, and 7N-3.9P-4.1K) for shoot dry weight (g) per plant (n = 9). Data are presented as means ± SEM. Means with same lowercase letter are not significantly different by LSD (*p* ≤ 0.05) between cultivars within fertilizers. Means with same uppercase letter are not significantly different by LSD (*p* ≤ 0.05) among fertilizers within each cultivar.

**Figure 3.** Interaction between sweet pepper cultivars ('Bentley' and 'Orangella') and hydroponic fertilizers (5N-4.8P-21.6K, 5N-5.2P-21.6K, and 7N-3.9P-4.1K) for average fruit weight (g) per plant (n = 9). Data are presented as means ± SEM. Means with same lowercase letter are not significantly different by LSD (*p* ≤ 0.05) between cultivars within fertilizers. Means with same uppercase letter are not significantly different by LSD (*p* ≤ 0.05) among fertilizers within each cultivar.

**Table 3.** Main effect of hydroponic fertilizers (5N-4.8P-21.6K, 5N-5.2P-21.6K, 7N-3.9P-4.1K) pooled across cultivars for per plant sweet pepper fruit yield and root weight. (n = 18).


<sup>z</sup> Means within a column followed by same lowercase letter are not significantly different by LSD (*p* ≤ 0.05). Data are presented as means ± SEM.

#### *3.2. Eggplant*

Interactions between fertilizer and eggplant cultivars occurred for shoot fresh weight. The shoot fresh weight was significantly greater for 'Angela' as compared to 'Jaylo' when fertilized with 5N-4.8P-21.6K and 7N-3.9P-4.1K. There was no significant difference between the shoot fresh weights of eggplant cultivars when fertilized with 5N-5.2P-21.6K (Figure 4).

**Figure 4.** Interaction between eggplant cultivars ('Angela' and 'Jaylo') and hydroponic fertilizers (5N-4.8P-21.6K, 5N-5.2P-21.6K, and 7N-3.9P-4.1K) for shoot fresh weight (g) (n = 9). Data are presented as means ± SEM. Means with the same lowercase letter are not significantly different by LSD (*p* ≤ 0.05) between cultivars within fertilizers. Means with same uppercase letter are not significantly different by LSD (*p* ≤ 0.05) among fertilizers within each cultivar.

A fertilizer effect was found on the shoot dry weight, while a cultivar main effect was only found for the shoot dry weight of eggplant (Table 2). There was no significant difference for yield and root weight among different fertilizer treatments (Table 2). The shoot dry weight of eggplant was significantly greater in 5N-4.8P-21.6K as compared to 5N-5.2P-21.6K and 7N-3.9P-4.1K when pooled across cultivars (Table 4). The shoot dry weight of 'Angela' was significantly greater than 'Jaylo' when pooled across fertilizers (Table 4).

**Table 4.** Main effect of hydroponic fertilizers (5N-4.8P-21.6K, 5N-5.2P-21.6K, 7N-3.9P-4.1K) pooled across cultivars (n = 18) and eggplant cultivars ('Angela' and 'Jaylo') pooled across fertilizers for shoot dry weight. (n = 9).


<sup>z</sup> Means within a column followed by same lowercase letter are not significantly different by LSD (*p* ≤ 0.05). Data are presented as means ± SEM.

Eggplant fruits developed an abnormal color after 2 months of production in both years. The fruits of the 'Jaylo' cultivar turned brownish-purple in color, while the 'Angela' cultivar fruits developed a yellow color. Foliar analysis found that the concentration of all nutrients was above the recommended upper limit except N (Table 5). A pairwise comparison was performed between the recommended foliar nutrient concentration by Flores et al. [21] and foliar nutrient concentration of plants grown in different fertilizers was observed.


**Table 5.** Average foliar nutrient concentration for eggplant in comparison to recommended nutrient level by Flores [21] (n = 18).

<sup>z</sup> Means within a row followed by same letter are not significantly different by paired t test (*p* ≤ 0.05). y, \*, \*\*, \*\*\* indicates non-significant or significant at *p* ≤ 0.05, *p* ≤ 0.001, or *p* ≤ 0.0001, respectively. Data are presented as means ± SEM.

#### **4. Discussion**

Fruit weight and fruit shape index are two important characteristics of sweet peppers, determining the fruit quality [22]. For sweet pepper, fruits weighing less than 100 g are considered to be unmarketable [22]; in the current trial, the sweet pepper average fruit weight ranged from 122–172 g (Figure 3). Rubio et al. [22] also looked for the response of Ca and K on the yield and fruit quality of sweet pepper and found that adequate management of Ca and K fertilization could help improve yield and fruit quality (fruit shape index) of sweet pepper in hydroponics. The findings from the current experiment for sweet pepper fruit yield support the results from Rubio et al. [22], as high yielding fertilizers 5N-4.8P-21.6K and 5N-5.2P-21.6K were high in Ca and K as compared to 7N-3.9P-4.1K, whereas there was no effect on fruit quality (fruit shape index). Fertilizer 5N-4.8P-21.6K has been recommended for hydroponic production of tomatoes, cucumbers, and peppers and was found to be similar in nutrient content with the hydroponic recipe prepared by the University of Arizona, which provided remarkable results [9].

Another study evaluated the effect of nutrition and irrigation on sweet pepper production in hydroponics and concluded that in a closed system, the fertilization of nitrogen (N) 240, phosphorus (P) 60, (K) 300, magnesium (Mg) 50, ferrous (Fe) 6, manganese (Mn) 3, boron (B) 1.6, zinc (Zn) 2, (Ca) 90, copper (Cu) 0.8 and molybdenum (Mo) 0.12 (mg L<sup>−</sup>1) was appropriate for sweet pepper production [23]. Therefore, there is a possibility of a further increase in fruit yield for current sweet pepper cultivars because all the nutrient levels of the current fertilizers were lower than the levels recommended by Gul et al. [23] (Table 1). Adding potassium peroxide at a rate of 1 g L−<sup>1</sup> has also been reported to result in a 20% increase in sweet pepper yield in hydroponics [24].

For hydroponic eggplant production, we did not find any recommendations of specific fertilizers in the literature other than self-preparation of Hoagland's solution [25]. However, since the manufacturer recommended that the fertilizers tested in the current trial were suitable for fruiting vegetable crops, they were tested for eggplants. Both the form and quantity of N play important roles in hydroponic as well as field vegetable production. The nitrate form of N should dominate in the nutrient solution, while the ammoniacal form should be lower [25]. In the current study, fertilizers 5N-4.8P-21.6K and 5N-5.2P-21.6K had a total N in nitrate form while 7N-3.9P-4.1K had 2.6% as ammoniacal form and 4.4% as nitrate form. In terms of the quantity of N, the recommendation of total N for hydroponic production of eggplants was 120–170 ppm, which was satisfied by the fertilizers used in our study [26]. There was limited literature providing information regarding micronutrient requirements of eggplant in soilless culture. It has been reported that eggplants need 15, 10, 5, 0.75 and 0.5 μM of Fe, Mn, Zn, Cu, and Mo, respectively [26].

The yellowing of eggplant leaves and fruits was initially suspected to be caused by a deficiency of some nutrients. Eggplant is susceptible to boron deficiency and young fully developed leaves turn yellow at the distal end [27]. However, foliar analysis of eggplant revealed that the concentration of all the nutrients was above the recommended limit except N (Table 5). Therefore, the yellowing in plants was more likely due to the toxicity of nutrients. A possible reason explaining this nutrient toxicity in hydroponic eggplant production is the use of expanded clay balls as a stand-alone substrate. Some substrates may have a higher cation-exchange capacity, thereby leading to the localization of some nutrients in root zones and to the toxicity of nutrients. Pine bark has been suggested to be the best stand-alone substrate for fruit vegetable production [28]. Another reason explaining nutrient toxicity could be the higher accumulation of macro and micronutrients in closed hydroponic systems reported in some studies [29]. Therefore, the selection of an adequate stand-alone substrate is important for hydroponic vegetable production to avoid yield loss due to nutrient toxicity [30]. Moreover, there is need for an appropriate method to monitor nutrient concentration in solution during growing cycles.

Many studies have reported different EC ranges for the hydroponic cultivation of eggplants. The response of eggplant to salinity in a recirculating hydroponics system was studied by Savvas et al. [26], who found that high salinity significantly affected osmotic potential due to reduced water uptake leading to less water being directed towards fruit development and they recommended an EC of 1.5 dS m−1. Moazed et al. [31] and Mahjoor et al. [32] recommend an EC of 2.5 dS m−1. According to the foliar nutrient concentrations, by maintaining the EC in the recommended range (2.5–3.5 dS m−1), plants were not able to maintain nutrient concentration in required limits as the concentration of all the nutrients except N was higher than the recommended range. Therefore, some researchers have reported that the EC is not a good indicator for estimating the nutrient concentration of solution, as EC indicates total dissolved ion concentrations only and cannot be used directly to determine individual ion concentrations. Thus, controlling nutrients based on EC in hydroponics may lead to excess or deficiency of some nutrients [33]. Periodic tissue sampling is reported to be the best way to evaluate if the nutrients provided are adequate for the growth stage and growing conditions [9]. Furthermore, some other non-destructive precision agriculture tools, such as mobile phone plant nitrogen applications, can be used to monitor nutrient concentrations in greenhouse production [34].

#### **5. Conclusions**

From the results of the present experiment, 5N-4.8P-21.6K and 5N-5.2P-21.6K can be recommended for sweet pepper production in hydroponics because fruit yield was not significantly different between these fertilizers, whereas it was significantly greater than with 7N-3.9P-4.1K. For cultivar evaluation, 'Orangella' produced significantly greater shoot fresh and dry weight in 5N-4.8P-21.6K and 5N-5.2P-21.6K. Nevertheless, vegetable producers are more interested in fruit yield and quality. The average fruit weight of 'Orangella' was significantly lower than 'Bentley' when grown in 5N-4.8P-21.6K. Moreover, some other factors needed to be evaluated to recommend these cultivar for hydroponic production because some studies reported 'Bentley' to be susceptible to *Fusarium* and to water borne disease [17]. Two months data for eggplants showed that there was no effect of cultivar or fertilizer on eggplant yield, while the main effects of cultivar and fertilizer were observed for shoot dry weight, with 'Angela' producing significantly greater results than 'Jaylo' and 5N-4.8P-21.6K producing significantly greater results among the three fertilizers. An interaction among fertilizers and eggplant cultivars was observed for eggplant shoot fresh weight, with 'Angela' producing significantly greater weight in 5N-4.8P-21.6K and 7N-3.9P-4.1K. Based on the results of the current study, it is not possible to recommend either fertilizer or cultivar for the hydroponic production of eggplant, as after 2 months, almost all the fruits were non-marketable due to yellowing. Therefore, future studies are needed to investigate the physiology behind the yellowing of eggplant fruits, and to identify a better indicator of nutrient concentration than EC. Different recycling rates of nutrient solutions and alternatives for stand-alone substrates for eggplant hydroponic production should be evaluated in future studies because this will also affect nutrient accumulation into plant parts.

**Author Contributions:** Conceptualization, B.L.D.; H.S.; methodology, H.S.; formal analysis, M.P.; investigation, H.S.; writing—original draft preparation, H.S.; writing—review and editing, B.L.D.; H.S.; and L.B.; supervision, B.L.D.

**Funding:** This research was funded by ODAFF Specialty Block grant #G00000475.

**Conflicts of Interest:** We declare no conflict of interest.

## **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*

## **Increasing Levels of Supplemental LED Light Enhances the Rate Flower Development of Greenhouse-grown Cut Gerbera but does not A**ff**ect Flower Size and Quality**

## **David Llewellyn, Katherine Schiestel and Youbin Zheng \***

School of Environmental Sciences, University of Guelph, 50 Stone Road East, Guelph, ON N1G 2W1, Canada; dllewell@uoguelph.ca (D.L.); katie.schiestel@gmail.com (K.S.)

**\*** Correspondence: yzheng@uoguelph.ca

Received: 13 August 2020; Accepted: 3 September 2020; Published: 4 September 2020

**Abstract:** To investigate the influence of supplemental lighting intensity on the production (i.e., rate of flower development, flower quality, and yield) of cut gerbera during Canada's supplemental lighting season (November to March), trials were carried out at a research greenhouse. Five supplemental light emitting diode (LED) light intensity (LI) treatments provided canopy-level photosynthetic photon flux densities (PPFD) ranging from 41 to 180 μmol m−<sup>2</sup> s<sup>−</sup>1. With a 12-h photoperiod, the treatments provided 1.76 to 7.72 mol m−<sup>2</sup> d−<sup>1</sup> of supplemental light. Two cultivars of cut gerbera (*Gerbera jamesonii* H. Bolus ex Hook.f) were used to evaluate vegetative growth and flower production. Plugs of 'Ultima' were assessed for vegetative growth and rate of flower development. There were minor LI treatment effects on number of leaves and chlorophyll content index and flowers from plants under the highest versus the lowest LI matured 10% faster. Reproductively mature 'Panama' plants were assessed for flower yield and quality. 'Panama' flowers from the highest LI treatment had shorter stems than the three lowest LI treatments, and flowers from the middle LI treatment had larger diameter than the other treatments. Flowers from the lowest LI treatment had lower fresh mass than the three highest LI treatments. There were linear relationships between LI and numbers of flowers harvested, with the highest LI treatment producing 10.3 and 7.0 more total and marketable flowers per plant than the lowest LI treatment. In general, increasing levels of supplemental light had only minor effects on vegetative growth (young plants) and size and quality of harvested flowers (mature plants), but flowers from plants grown under higher LIs were more numerous and matured faster.

**Keywords:** flower bud development; flower number; flower quality; *Gerbera jamesonii*; growth; DLI

## **1. Introduction**

In greenhouses at higher latitude regions, such as northern USA and Canada, it is often considered necessary for growers of year-round commodities (e.g., cut flowers) to use supplemental lighting to meet the crops' economic minimum lighting requirements during the darker months, due to low natural light conditions and short daylengths. While many economic (e.g., capital cost of fixtures and electricity prices) and practical (e.g., fixture positioning and capacity of electrical supply infrastructure) elements are considered when outfitting a greenhouse with supplemental lighting systems, the response of the crop(s) to additional lighting is a key factor which can only be evaluated through careful production trials.

The photosynthetic responses of plants to increasing levels of photosynthetically active radiation (PAR), generally described in terms of photosynthetic photon flux density (PPFD, μmol m−<sup>2</sup> s−1), have been well established for many plant species and environments. When considering supplemental

lighting in greenhouse production scenarios, crops are generally subjected to light intensities (LI) that are on the linear portion of the photosynthetic light response curve (i.e., far lower than LIs needed to saturate the photosynthetic machinery). By extension, the yield responses of many greenhouse crops are commonly generalized as being directly proportional to the levels of light provided, with every 1% increase in lighting resulting in concomitant 0.5% to 1% increases in production [1]. This relationship has borne out for some economically relevant production indices in various floriculture commodities, such as cut gerbera [2,3], potted begonias [4], and cut roses [5,6].

For optimization of commercial greenhouse production scenarios which utilize supplemental lighting, it is necessary to determine the impact supplemental daily light integral (DLI), supplied within practical constraints of intensity and photoperiod, has on economically relevant production indices for target commodities. This may be especially relevant for cut flower production, such as cut gerbera, where the harvestable product represents a relatively minor component of total plant biomass production. Therefore, yield may be less directly related to rates of photosynthesis or carbon assimilation, while other crop x lighting interactions, such as photomorphogenic effects and flower bud development, may become more relevant with increasing levels of supplemental DLI [7]. We are unaware of any other references which have directly investigated the effect of DLI on vegetative growth, days between transplant and first visible flower bud, rate of flower development from visible bud to harvest, or fresh harvest metrics of cut gerbera flowers.

Auito [2] investigated cut gerbera production under a range of supplemental PPFD and photoperiods provided by high pressure sodium (HPS) lights. Their results showed that 12-h photoperiod maximized flower production for a given supplemental DLI. Conversely, Pettersen and Gislerød [8] found that a 20-h photoperiod had a higher production of cut gerbera than a 10-h photoperiod. However, the trials were done in a growth chamber, with a fixed PPFD, thus confounding the effects of photoperiod and DLI (i.e., the 20-h photoperiod had twice the DLI as the 10-h photoperiod), making it difficult to extrapolate their results to greenhouse environments. In a parallel study, Auito [2] also found linear or near linear relationships between supplemental light intensity and cut gerbera production in a greenhouse, using PPFD levels ranging from 75 to 300 μmol m−<sup>2</sup> s−<sup>1</sup> with a 12-h photoperiod (i.e., DLIs of 3.2 to 13.0 mol m−<sup>2</sup> d−1). However, natural lighting was only reported as seasonal mean values for outdoor DLI throughout the 6-month trial period (<sup>≈</sup> 7.4 mol m−<sup>2</sup> d−1, November to April) with an (estimated) greenhouse transmission value of 50%. Therefore, it is not possible to draw conclusions based on the absolute light levels (i.e., natural + supplemental) within the treatment plots. Approximate values for natural DLI at crop level in this study would probably have averaged between 3 and 4 mol m−<sup>2</sup> d−<sup>1</sup> (based on 50% of 7.4 mol m−<sup>2</sup> d−1), which is similar to the winter lighting conditions in the research greenhouse facility used in the present study [9]. Spanomitsios et al. [3] found a positive linear relationship between mean daily solar radiation and rate of cut gerbera flower production. In this study, the slope of the relationship between light and production (slope = 0.47) indicated that ≈ 0.5% increase in flower production could be expected for every 1% increase in total light. However, it takes approximately four weeks for a cut gerbera flower to mature from visible bud to harvestable stage. Therefore, the reported relationship between daily net radiation and flower yield would have been more realistically portrayed if harvest data had been related to the average DLI for the four weeks prior to each harvest. Further, it is not possible to infer PPFD or DLI at crop level in this study as it is not clear how or where light data were collected or how the data were processed. Mustapi´c-Karli´c et al. [10] found a positive influence of supplemental lighting on flower yield of two cut gerbera cultivars. They compared treatments of natural lighting with natural + supplemental HPS lighting providing <sup>≈</sup> 3 mol m−<sup>2</sup> d−<sup>1</sup> of additional PAR (i.e., PPFD of <sup>≈</sup> 70 <sup>μ</sup>mol m−<sup>2</sup> s−<sup>1</sup> with a 12-h photoperiod). However, the DLI at crop level is unknown because the natural light levels at crop level were not reported. Similarly, Gagnon and Dansereau [11] found increases in potted gerbera productivity and reductions in time to flowering with increasing levels of supplemental HPS lighting (ranging from 1.7 to 5.2 mol m−<sup>2</sup> d<sup>−</sup>1). However, the authors also did not report natural light levels, making it impossible to draw conclusions about the absolute influence of the lighting

treatments on production. While these trials clearly indicate positive relationships between increasing levels of supplemental lighting and production of cut gerbera, insufficient information on canopy-level lighting conditions make it difficult for readers to critically evaluate the total amount of PAR received by the plants in these trials [7].

With respect to the quality of supplemental light, research has shown that at similar PPFD, supplemental PAR from light-emitting diode (LED) technologies have resulted in similar crop production metrics as traditional HPS in greenhouse commodities, such as leafy vegetables [12], fruiting vegetables [13–15], ornamentals [16–18], and cut flowers [19]. While the capital costs of LED technologies are still considerably higher than HPS, LEDs have many advantages over HPS. LEDs can provide narrow wavebands of light specifically targeted at the maximum absorption bands of photosynthetic machinery. LEDs are touted to have greater than twice the lifetime as HPS and also have the potential to achieve higher efficacies (i.e., conversion of electricity into PAR). Moreover, LEDs are naturally dimmable, providing the capacity to adjust intensity according to natural lighting conditions, as well as on-demand customization of spectral recipes, providing greater plasticity for photoperiod and photomorphological control within a single fixture [20,21]. Accordingly, leading researchers and industry professionals consider it only a matter of time before LED technologies replace HPS as the benchmark technology for supplemental lighting in greenhouse applications [22].

The objectives of this study were to evaluate the relationships between increasing levels of supplemental lighting from LEDs during the darker months in Canada on the growth, flower development, yield, and quality of greenhouse grown cut gerbera.

#### **2. Materials and Methods**

#### *2.1. Location, Trial Bench, and Greenhouse*

The study took place at the University of Guelph in Guelph, ON, Canada, (43.55 ◦ N, 80.25 ◦ W) beginning on 9 November 2015 and ending on 25 February 2016 (i.e., 107 d). The study was set up within a single 7.2 × 7.2-m glass-clad research greenhouse compartment, containing four 4.57 × 1.07 m benches, with 0.91 m spacing between them. The long sides of the benches were positioned in an east-west direction (i.e., parallel with the track of the sun).

#### *2.2. Lighting Treatments and Plant Distribution*

There were five PPFD treatments, two pots of plants (i.e., two subsamples) for each of two cultivars under each PPFD treatment on each bench, as well as four replicates (i.e., benches) within the greenhouse compartment.

There were four LED fixtures (Pro 325; LumiGrow, Novato, CA, USA) per bench, located 30 and 100 cm (measured on-center of each fixture's LED array) from both ends of each bench. The lights were centered along the long axis of the bench and fixed with the LED arrays 140 cm above pot level. Each fixture was affixed with shrouds arranged parallel with the long sides of the benches made of white vinyl siding (Cedar Creek D4D; Abtco, Milton, ON, Canada) to reduce stray lighting from adjacent benches. The fixtures were set with an area-averaged photon flux ratio of blue (B, 400 to 500 nm) to red (R, 600 to 700 nm) of B22:R78. Fixture positioning and mapping light distribution patterns were done at night using a radiometrically-calibrated spectrometer (USB2000+; Ocean Optics, Dunedin, FL, USA) coupled to a 400-μm diameter UV-VIS optical fiber with a CC-3 cosine corrector (Ocean Optics, Dunedin, FL, USA). Light distribution (intensity and quality) was measured at pot level on a 2 × 12 rectangular grid (i.e., 24 specific locations), centered on the geometric center of the bench, with 30 cm separating adjacent measurement locations. For the trial, individual cut gerbera pots were centered on each of these bench locations and remained there for the duration of the trial. In this configuration, the supplemental light treatment at pot level of each plant was kept at a constant, known value. This design resulted in five unique supplemental PPFD treatment levels on each bench (labeled T1 to T5).

Two cut gerbera (*Gerbera jamesonii* H. Bolus ex Hook.f) cultivars, 'Panama' and 'Ultima', were used for this trial. 'Panama' plants were sourced from an active production environment (≈ 5 months of active flower production) from a local grower (Bayview Flowers Ltd., Lincoln, ON, Canada). Flower stems longer than 2.5 cm were removed from 'Panama' plants at the beginning of the study. 'Ultima' plants came from the supplier, Florist Holland B.V. (De Kwakel, The Netherlands), as 'Jiffy 4' plugs.

On 8 October 2015, the plugs were transplanted into round 19 cm diameter × 19 cm tall pots; filled with coarse coir mix typically used by and obtained from a local cut gerbera grower. 'Ultima' plants began the trial in the vegetative stage, with no visible flower buds. Equal numbers of plants from each cultivar were positioned on the benches such that the cultivars were arranged in an alternating fashion. This arrangement resulted in two plants of each cultivar per treatment per bench, plus two border plants on the ends of each bench. The planting density was <sup>≈</sup> 7 plants m<sup>−</sup>2, which was consistent with local commercial cut gerbera greenhouses. Although the location of each plant was fixed, the plants were rotated one-quarter turn weekly to reduce pot-location effects.

## *2.3. Environmental Management*

The greenhouse environment parameters were set at similar levels to those used by local cut gerbera producers. Supplemental LED lighting was turned on daily 12 h before dusk and turned off at dusk, resulting in a constant 12-h photoperiod. Day and night temperature setpoints were 21 and 14 ◦C, respectively. Relative humidity (RH) was maintained at 70% using an aerial fogger system located at gutter level. Temperature and humidity dataloggers (HOBO U12-013; Onset Computer Corporation, Bourne, MA, USA) were located at canopy level in the center of each bench. PAR sensors (SQ-110; Apogee Instruments Inc., Logan, UT, USA) were located 1.75 m above the center of each bench (i.e., just above the top of the LED fixtures) and connected to the HOBO dataloggers. Temperature, RH, and PPFD were logged every 120 s throughout the study. Previous light uniformity data, collected by simultaneously logging the natural PPFD at fixture-level and bench-level (supplemental light fixtures present but left off) during a prior supplemental lighting season (i.e., November to March), indicated strong correlations in DLI measured between bench- and fixture-level locations on each bench. Coefficients relating natural DLI at fixture-level to bench-level derived from these data (not shown), were applied to the fixture-level PPFD data collected during the present trial to determine natural DLIs at canopy level on each bench.

#### *2.4. Irrigation Management*

Plants were drip irrigated using 20N-3.5P-16.6K All Purpose water soluble fertilizer (250 ppm N, pH 5.5; Plant Products Co. Ltd., Brampton, ON, Canada) with temporary substitutions of well water (pH and EC of 7.9 and 1000 μS cm<sup>−</sup>1, respectively), when necessary, to maintain an approximate root zone pH of 5.5 and EC of 2500 μS cm<sup>−</sup>1. Pulse irrigation occurred every second day, at 0915 and 1315 HR for 180 s each time. This irrigation protocol was aimed at producing approximately 10% to 25% leachate. Hand-watering was used as needed to supplement the drip irrigation.

#### *2.5. Plant Growth, Leaf Chlorophyll Content Index, Flower Quality, and Yield Metrics*

The number of leaves and chlorophyll content index (CCI) were measured approximately monthly on each 'Ultima' plant using a chlorophyll meter (CCM-200 Plus; Opti-Sciences, Hudson, NH, USA). CCI measurements were taken (three measurements per leaf with the average CCI value recorded), near the leaf margin (i.e., avoiding larger venation) of the youngest fully-expanded leaf of each plant. 'Ultima' plants were also checked twice weekly for the development of flower buds. Once each stem was ≥ 1 cm long, it was tagged with a unique identifier and the respective date was recorded as the date of appearance. This provided the days from transplant to first visible flower bud (i.e., stems ≥ 1 cm), as well as insight into the rate of flower development (i.e., the time between visible flower bud appearance and harvest).

Flowers on 'Panama' plants were harvested twice weekly. Flowers were deemed harvestable once they developed one complete ring of matured anthers. Fresh mass, flower diameter (measured petal tip to petal tip on the widest part of the flower), and stem length (measured from heel to the base of the flower) were measured on each harvested flower. Flower quality was also classified subjectively as either marketable or unmarketable according to the severity of malformations and pest damage.

#### *2.6. Statistical Analysis*

The experiment was a block design with 5 treatments and 4 concurrent replications. All data sets were analyzed using JMP® (version 13; SAS Institute Inc., Cary, NC, USA, 1989–2017). Least squares analysis was used for light treatment uniformity; vegetative growth, rate of appearance of visible flower buds, and flower development metrics in 'Ultima'; and accumulated total and marketable flowers harvested per plant in 'Panama'. Flower yield metrics in 'Panama' were analyzed using the Mixed-Models add-in, which accounts for the different numbers of flower stems harvested from each plant. Data were evaluated using a significance level of *p* ≤ 0.05 using Tukey's honestly significant difference (HSD) test. Days between the appearance of flower buds and harvest on 'Ultima' and accumulated total and marketable flowers harvested per plant on 'Panama' underwent regression analysis (*p* ≤ 0.05), using total DLI (i.e., natural + supplemental) as the independent variable.

#### **3. Results**

Weekly average canopy-level natural DLI for the 17-week trial ranged from <sup>≈</sup> 1 to 6 mol m−<sup>2</sup> d−<sup>1</sup> with an overall average of 3.6 mol m−<sup>2</sup> d−<sup>1</sup> (Figure 1), which was consistent with previous years' light characterizations within the same experimental greenhouse (data not shown). Daytime (i.e., daily timeframe when supplemental lighting was on) and nighttime (i.e., daily timeframe when supplemental lighting was off) temperatures were (mean±SD) 20.4±2.0 °C and16.6 ± 1.24 °C, respectively.

**Figure 1.** Weekly natural daily light integral (DLI) at canopy level (average ± SE, n = 7). The overall average natural DLI, during the 17-week trial, was 3.6 mol m−<sup>2</sup> d<sup>−</sup>1.

The supplemental PPFD treatments ranged from 40.7 to 179 μmol m−<sup>2</sup> s<sup>−</sup>1, corresponding to 1.8 to 7.7 mol m−<sup>2</sup> d−<sup>1</sup> of daily supplemental PAR with a 12-h photoperiod (Table 1).


**Table 1.** Canopy-level supplemental photosynthetic photon flux density (PPFD) of the five supplemental light-emitting diode (LED) treatments and their associated supplemental and total daily light integrals (DLI).

<sup>z</sup> There were no block or bench position effects on supplemental PPFD within each treatment, so data are pooled means for each treatment ± SE (n = 16). Values in the same column followed by the same letter are not different at *p* < 0.05, using Tukey's honestly significant difference (HSD). <sup>y</sup> DLI from supplemental LEDs were calculated using mean PPFD from each treatment and 12-h photoperiod. <sup>x</sup> Total DLI is the sum of supplemental DLI from LED treatments and experiment-wise mean natural DLI of 3.6 mol m−<sup>2</sup> d<sup>−</sup>1.

'Ultima' plants chosen for each treatment had uniform CCI and number of leaves at the start of the trial (9 November 2015). After one month of treatment (8 December 2015), plants in T5 had ≈ 4 more leaves than plants in T4. After two months of growth under the supplemental light treatments (6 January 2016), plants in T4 had higher CCI values than T1, T2, and T3, and plants in T5 had ≈ 6 more leaves than plants in T2 (Table 2).


**Table 2.** Chlorophyll content index (CCI) of the youngest fully-expanded leaf, and number of leaves per plant, measured at ≈ 4-week intervals post-transplant of 'Ultima' plants.

<sup>z</sup> There were no block effects within each treatment at each measurement date, so data are pooled averages for each treatment ± SE (n = 8). Values in the same column with the same measuring day followed by the same letter are not different at *p* < 0.05, using Tukey's HSD.

Flowers in T5 matured (i.e., time between appearance of flower buds and harvest) ≈ 3.6 d faster than plants in T1, which represents ≈ 10% reduction in flower development time (Table 3).

There were only minor treatment effects in fresh flower harvest metrics on 'Panama' flowers (Table 4). Flowers grown in T5 had marginally shorter stems than flowers grown in T1, T2, and T3. Flowers grown in T3 were marginally larger and flowers grown in T1 were smaller than the other treatments (with < 0.2 cm difference in diameter). Flowers grown in T3 also had higher fresh mass than flowers grown in T1 and T2.


**Table 3.** Days between appearance of flower buds (i.e., stems ≥ 1 cm) and harvest for all 'Ultima' flowers harvested during the trial, for different total daily light integral (DLI) treatments.

<sup>z</sup> There were no block effects, so data are pooled averages for each treatment ± SE (n <sup>=</sup> 8). Values in the same column followed by the same letter are not different at *p* < 0.05, using Tukey's HSD.

**Table 4.** Stem length, flower diameter, and fresh mass of 'Panama' flowers harvested throughout the trial, for different total daily light integral (DLI) treatments.


<sup>z</sup> There were no block effects, so data are pooled means for each treatment ± SE (n <sup>=</sup> 8). Values in the same column followed by the same letter are not different at *p* < 0.05, using Tukey's HSD.

Regressing 'Panama' flower harvest numbers against total DLI indicated that every 1% increase in DLI increased cumulative flower yield by ≈ 1.5% (Figure 2). The trend was similar in terms of marketable flowers, where a 1% increase in DLI resulted in a concomitant ≈ 1% increase in the number of marketable flowers produced per plant.

**Figure 2.** Cumulative total and marketable flowers harvested per plant, for 'Panama', in response to total daily light integral (DLI). Each point represents the treatment mean ± SE (n = 8); however, the equations are linear regressions of all of the harvest data on a per-plant basis.

#### **4. Discussion**

The range of supplemental PPFD levels used in this study raised the total canopy-level DLI to levels that approximately match the DLI range deemed necessary to produce minimum acceptable quality (6 mol m−<sup>2</sup> d<sup>−</sup>1) to high quality (12 mol m−<sup>2</sup> d<sup>−</sup>1) gerbera [23]. Vegetative growth and flower development indices were investigated using transplanted plugs of the 'Ultima' cultivar, while mature plants of the 'Panama' cultivar were used to assess the size, quality, and numbers of flowers produced.

There were no commercially-relevant LI treatment differences (or trends) in number of leaves or CCI of 'Ultima' plants. While there were also no LI treatment effects on the days from transplant to first visible flower (data not shown), flowers in T5 matured ≈ 10% faster than flowers in T1. Linear regression of the treatment means for days between appearance of first visible flower bud and harvest in 'Ultima' (in Table 3) against DLI indicates that each additional mol m−<sup>2</sup> d−<sup>1</sup> of DLI (e.g., <sup>≈</sup><sup>23</sup> <sup>μ</sup>mol m−<sup>2</sup> <sup>s</sup>−<sup>1</sup> of supplemental PAR over a 12-h photoperiod) shortened the time between flower bud appearance and harvest by 0.53 d. For example, adding <sup>≈</sup> 90 <sup>μ</sup>mol m−<sup>2</sup> s−<sup>1</sup> of supplemental PAR with a 12-h photoperiod could shorten the flower production time by 2 d, during the darker months.

There were only minor (i.e., probably not commercially relevant) LI treatment effects on stem length, flower diameter, and fresh mass of marketable 'Panama' flowers. However, there were LI treatment effects on the total and marketable numbers of 'Panama' flowers harvested per plant, with plants in T5 producing ≈ 40% more flowers than plants in T1. Subjecting the cumulative flower production metrics to linear regression analysis showed that DLI could be used to predict the cumulative flowers produced per plant (Figure 2). Similarly, Bredmose [5,6] found linear relationships between supplemental light (HPS) intensity and numbers of flowers produced by mature plants of two rose cultivars, within the range of 0 to 174 μmol m−<sup>2</sup> s−1. Auito's [2] investigation on the effects of supplemental light intensity and photoperiod on cut gerbera production is the most comprehensive to date. However, insufficient information was provided about the natural lighting environment under which the crops were grown; making it difficult to assess the actual lighting conditions (e.g., total DLI) in these trials. Despite this drawback, the author concluded that cut gerbera plants utilize supplemental light for flower production most efficiently at shorter photoperiods (i.e., 12 h), which is in line with local production practices. Auito [2] noted some cultivar-specific responses to increased supplemental PAR, although total flowers per plant and total dry mass generally increased linearly with increasing supplemental DLI (between 3.2 and 13.0 mol m−<sup>2</sup> <sup>d</sup>−<sup>−</sup>1, with a 12-h photoperiod)

In the present study, it was shown that doubling the total DLI from 6 to 12 mol m−<sup>2</sup> d−<sup>1</sup> by providing an additional 6 mol m−<sup>2</sup> d−<sup>1</sup> of supplemental PAR from LEDs could increase the number of flowers produced by nine flowers per plant (over 107 d). At typical commercial plant densities of 7 m<sup>−</sup>2, this would result in monthly increases in flower production of <sup>≈</sup> 18 more flowers/m2. In practical terms, if a grower provided 100 μmol m−<sup>2</sup> s−<sup>1</sup> of supplemental PAR, with a 12-h photoperiod, they could potentially increase the total number of flowers produced per plant during the darker months by <sup>≈</sup> 30%. To further contextualize in terms of energy cost, the efficacy factor of 1.29 <sup>μ</sup>mol J−<sup>1</sup> for the LumiGrow Pro 325 fixtures used in this study [24] can be used to estimate that <sup>≈</sup> 1.3 kWh m−<sup>2</sup> d−<sup>1</sup> would be needed to add 6 mol m−<sup>2</sup> <sup>d</sup>−<sup>1</sup> of supplemental PAR from LEDs, which would be <sup>≈</sup> 2 kWh per additional flower produced, in the above scenario. However, the efficacy of some horticultural LED fixtures has more than doubled versus the fixtures used in this study [25], which would reduce the energy input per flower to less than 1 kWh for modern LED fixtures.

Future research should include broadening the range of commodities investigated under supplemental LED lighting intensity regimens, as well as investigating applications of targeted spectrum treatments (especially at night, where applicable) for manipulating crop morphology. A promising example of spectrum-mediated change in morphology are the increases in stem extension rates without some of the negative "shade avoidance" effects of high far red (700–800 nm) treatments by using low fluence rates of monochromatic blue light, applied at nighttime [26].

## **5. Conclusions**

This investigation examined the influence of different levels of supplemental PAR, supplied by red and blue LEDs, on the production of cut gerbera during the darker months at higher latitudes. While there were few commercially-relevant LI treatment effects in the vegetative growth and harvested flower quality indices, higher light was shown to proportionally increase the rate of flower development and cumulative numbers of flowers produced. These relationships can be used by growers to assess the economic viability of using supplemental LED lighting to produce cut gerbera within their own production environments.

**Author Contributions:** Conceptualization, D.L. and Y.Z.; methodology, D.L., K.S. and Y.Z.; validation, D.L. and Y.Z.; formal analysis, D.L.; investigation, K.S. and D.L.; writing—original draft, D.L. and Y.Z.; writing—review & editing, D.L. and Y.Z.; visualization, D.L.; supervision, Y.Z.; funding acquisition, Y.Z. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the International Cut Flower Growers Association and the Joseph H. Hill Memorial Foundation, Inc.

**Acknowledgments:** LumiGrow, Inc. supplied the LED lighting fixtures. Plant materials were donated by Van Geest Bros Ltd (Ontario, Canada) and Bayview Flowers Ltd. (Ontario, Canada). None of these contributors were involved in the conduction of the study or submission of this article for publication.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

## **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Interactive E**ff**ects of the CO2 Enrichment and Nitrogen Supply on the Biomass Accumulation, Gas Exchange Properties, and Mineral Elements Concentrations in Cucumber Plants at Di**ff**erent Growth Stages**

**Xun Li 1, Jinlong Dong 1, Nazim S. Gruda 2, Wenying Chu 1,3,\* and Zengqiang Duan <sup>1</sup>**


Received: 24 December 2019; Accepted: 15 January 2020; Published: 17 January 2020

**Abstract:** The concentration changes of mineral elements in plants at different CO2 concentrations ([CO2]) and nitrogen (N) supplies and the mechanisms which control such changes are not clear. Hydroponic trials on cucumber plants with three [CO2] (400, 625, and 1200 μmol mol<sup>−</sup>1) and five N supply levels (2, 4, 7, 14, and 21 mmol L−1) were conducted. When plants were in high N supply, the increase in total biomass by elevated [CO2] was 51.7% and 70.1% at the seedling and initial fruiting stages, respectively. An increase in net photosynthetic rate (Pn) by more than 60%, a decrease in stomatal conductance (Gs) by 21.2–27.7%, and a decrease in transpiration rate (Tr) by 22.9–31.9% under elevated [CO2] were also observed. High N supplies could further improve the Pn and offset the decrease of Gs and Tr by elevated [CO2]. According to the mineral concentrations and the correlation results, we concluded the main factors affecting these changes. The dilution effect was the main factor driving the reduction of all mineral elements, whereas Tr also had a great impact on the decrease of [N], [K], [Ca], and [Mg] except [P]. In addition, the demand changes of N, Ca, and Mg influenced the corresponding element concentrations in cucumber plants.

**Keywords:** dry weight; N levels; elevated CO2; open-top chamber; nutrient transportation; photosynthesis; transpiration; dilution effect

## **1. Introduction**

In the cold of winter, most greenhouses in China are not opened or ventilated in order to keep a warmer air temperature for vegetable growth. Therefore, carbon dioxide (CO2) is often depleted rapidly in these closed greenhouses, and this lack becomes one of the biggest adverse factors that depress the photosynthesis and growth of vegetables [1]. CO2 enrichment has been widely implemented in greenhouses in Europe, North America, and Japan since the 1950s, and was introduced and used in greenhouses in China in the late 1980s [2–4]. CO2 enrichment has been found to have a dramatic effect on faster growth, greater biomass, and higher yield [5,6] due to the increased photosynthesis and carbohydrate accumulation, particularly in C3 plants [7,8]. Nevertheless, the longer and further researches reported some drawbacks of CO2 enrichment. These drawbacks included the decline of mineral concentrations in plant tissues [7,8], worsened taste due to the increased cellulose content [9,10], as well as the photosynthetic acclimation and the weak sustainability of its fertilization effects on

yield improvement [11–13]. The main reason was the increased photosynthates and accumulated carbohydrates in plant tissues under elevated CO2 concentrations ([CO2]), which decreased the nitrogen to carbon ratio (N/C) and caused the imbalance between source and sink [7,13,14]. To deal with this problem, many researchers recommended higher N fertilization to minimize the reduction in N/C associated with high [CO2] conditions [15–17]. However, when the dry matter accumulation outpaces N uptake, enriched CO2 will still reduce N concentrations ([N]) in plants even if N uptake is enhanced by higher N supply [18].

A reduction in mineral concentrations has been frequently demonstrated in crops grown under elevated [CO2] conditions [7–9,18]. Among these mineral elements, the possible mechanisms of the reduction of [N] in plants under high [CO2] conditions have been extensively studied [9,10,19,20]. Four hypotheses are well-documented: (1) Dilution effect due to accumulation of non-structural carbohydrates [10,12]; (2) reduced mass flow and transpiration due to decreased stomatal conductance [21,22]; (3) decreased Rubisco protein concentrations and N demands due to increased plant N use efficiency [10,23,24]; (4) inhibited photorespiration-dependent nitrate assimilation under high [CO2] [25,26]. However, the effects of elevated [CO2] on other mineral elements in plants have received far less attention. In plants, N, phosphorus (P), and sulfur (S) are mainly bounded to C to form organic molecules, whereas potassium (K), calcium (Ca), and magnesium (Mg) tend to remain in ionic forms or be chelated with enzymes in plant tissues. Considering their different uptake pathways, existence forms, and physiological functions, concentration changes of different mineral elements in plants at different [CO2] and N supply levels are possible with different mechanisms potentially contributing to each.

Cucumber (*Cucumis sativus* L.) is globally one of the most important vegetables that prefers to be cultivated in greenhouses with CO2 enrichment [27]. While the growth, photosynthesis [28], nitrogen metabolism [29], yield [1], fruit quality [30], root morphology [31], root exudate [32], and water use efficiency [33] of cucumber grown in elevated [CO2] conditions have been studied, the effects of different [CO2] and N supply levels on mineral element concentrations in cucumber and the key factor leading to these changes have received far less attention. Moreover, the available information about the optimum N supply under different elevated [CO2] is extremely limited. A better understanding of the concentration changes of mineral elements in cucumber plant responding to elevated [CO2] and N supplies is necessary for optimizing [CO2] and N fertilization in order to obtain high quality greenhouse products with higher C and N use efficiency [34,35] and to deal with future climate change scenarios with less CO2 emission and fertilizer input [36,37].

In previous studies, we have found the optimum N supply was 7 and 14 mmol L−<sup>1</sup> for the seeding and mature plants of cucumber, respectively, and the saturated and semi-saturated [CO2] of cucumber plants under natural solar radiation (about 600 mol m−<sup>2</sup> s<sup>−</sup>1) was 1200 and 625 μmol mol<sup>−</sup>1, respectively [28,30–32]. Therefore, in this study, we carried out the hydroponic trials on cucumber plants with three [CO2] (400, 625, and 1200 μmol mol<sup>−</sup>1) combined with five N supply levels (2, 4, 7, 14, and 21 mmol L−1) during seedling and initial flowering stages. Then, we investigated the effects of [CO2], N supply levels, and growth stages on cucumber growth, gas exchange, and macro-nutrient elements concentrations in different tissues. Pearson correlation coefficient is commonly used to quantify the degree of linear relationship between two factors and has been used to evaluate the relationships between root exudates and root morphological traits in our previous work [31]. In this study, we also used the Pearson correlation analysis to evaluate what were the key factors affecting the concentration changes of each mineral element in cucumber plants under different [CO2] and N supply conditions.

#### **2. Materials and Methods**

#### *2.1. Plant Culture and Growth Conditions*

Three open-top chambers (OTCs) (2.3 m length × 0.8 m width × 1.4 m height) made of poly (methyl methacrylate) were established in the glasshouse at Institute of Soil Science, Chinese Academy of Sciences, Nanjing, P.R. China (32.0596◦ N, 118.8050◦ E). The OTCs were transparent for minimizing the shading effect and received solar radiation with natural day length. The OTCs also have a pair of opposite side doors, which can be opened like wings for inside operation and cooling. The experiments were carried out as a split-plot design where [CO2] was the main treatment, and N supplies were considered as the sub-plot treatment. Five N treatments were set at 2 (N1), 4 (N2), 7 (N3), 14 (N4), and 21 (N5) mmol L−1, which were repeated six times in each chamber, and the thirty pots in each chamber were rotated within and among chambers every two weeks to minimize chamber effects. The [CO2] in three identical OTCs was set at 400 (ambient: C1), 625 (elevated: C2), and 1200 (super-elevated: C3) μmol mol−<sup>1</sup> respectively and was reset to the corresponding treatment condition following plant rotation. The [CO2] in OTCs was controlled and monitored continuously through an infrared gas analyzer (Ultramat 6, Siemens, Munich, Germany) started on the day after transplanting (DAT). The [CO2] in OTCs was elevated from 0800 to 1700 h every sunny and cloudy day. The temperature and relative humidity within the OTCs were recorded by a L95-83 data logger (Hangzhou loggertech Co., Ltd., Hangzhou, China) every 15 min. The chambers were opened for cooling and CO2 was not supplied when the temperature inside was above 35 ◦C. The accumulated CO2 treating time was 338 h within the whole experiment period of 62 days. The average temperature in OTCs was 22.9–23.5 ◦C, and the average humidity was 61.2–63.2%, respectively.

Cucumber (*Cucumis sativus* L.) seeds of 'Jinyou 38 (Tianjin Lvfeng Co., Ltd., Tianjin, China) were germinated on moist filter paper in constant-temperature incubator at 28 ◦C and relative humidity of 70% for 48 h, and then seeds with radicles were sown into trays containing peat-vermiculite (2:1, *v*/*v*) substrate. When the third true leaf emerged, healthy seedlings were selected and transplanted to 5 L polyvinyl chloride polymer (PVC) pots with two plants per pot. Each pot was filled with 4 L modified Yamazaki nutrient solutions for cucumbers [38] with five nitrogen levels. To keep the same P, K, Ca, and Mg concentrations ([P], [K], [Ca], and [Mg]) in nutrient solutions with five N levels, anions or cations were balanced with SO4 <sup>2</sup>−, NO3 <sup>−</sup>, or NH4 <sup>+</sup> respectively (Table 1). All the nutrient solutions contained the same concentration of micro-nutrients composed of (mg L−1): Na2Fe-EDTA (29.27), H3BO3 (2.86), MnSO4·4H2O (2.03), ZnSO4·7H2O (0.22), CuSO4·5H2O (0.08), and (NH4)6Mo7O24·4H2O (0.02). The pH of the nutrient solution was adjusted to 6.5 with dilute NaOH. All pots were aerated intermittently for 30 min in every hour and renewed every four days.


**Table 1.** Components of macro-nutrient solutions in different N levels.

#### *2.2. Sampling and Measurements*

#### 2.2.1. Gas-Exchange Rate Measurements

The gas exchange properties of cucumber plants, including net photosynthetic rate (Pn), transpiration rate (Tr), and stomatal conductance (Gs) were measured using a portable photosynthesis system (Li-6400, Li-Cor Inc., Lincoln, OR, USA) with a standard leaf chamber (2 cm × 3 cm) (6400-02B) with a LED light source. The photosynthetic photon flux density, temperature, relative air humidity, and the air flow rate inside the leaf chamber were set at 1500 mol m−<sup>2</sup> s<sup>−</sup>1, 25 ◦C, 50%, and 500 μmol s<sup>−</sup>1, respectively. The [CO2] of the flow-in air was set the same as that in the corresponding OTC where the plant was grown. The measurements were conducted on 13, 38, and 60 DAT and six replicates of the third leaves from the top of cucumber plants in each treatment were used for measuring.

#### 2.2.2. Plant Harvest and Biomass Determination

One plant in each pot was harvested at the seedling stage (T1) when the seedling had five to six true leaves (18 DAT) and the other was harvested at the initial fruiting stage (T2) when small fruits of 5–8 cm of length formed (62 DAT). After harvest, plants were separated to root, stem, and leaf samples and washed with tap water followed by distilled water. Dry weight (DW) of each tissue was determined by drying the fresh tissues at 105 ◦C for 30 min and then at 75 ◦C to a constant weight in an electro-thermostatic blast oven.

#### 2.2.3. Mineral Element Concentration Determination

The dry samples were ground to pass through a 0.5-mm screen. Next, 0.2 g dry samples were soaked in 5 mL concentrated H2SO4 for 24 h then digested at 180 ◦C for 5 h, followed by intermittent addition of 0.5 mL H2O2 for 2 or 3 times. The extracted solution was diluted to 500 mL with deionized water and the [N] was analyzed using a discrete auto-analyzer (Smartchem200, Alliance, France) [28]. Another portion of 0.2 g dry samples was digested with 5 mL HNO3-HClO4 (85:15 *v*/*v*) at 190 ◦C, and [P], [K], [Ca], and [Mg] were determined by an inductively coupled plasma atomic emission spectrometer (IRIS Advantage, Thermo Elemental, Franklin, MA, USA) [39].

#### *2.3. Statistical Analysis*

Statistical analysis was performed using SPSS software (Version 22.0; IBM Corp., Armonk, NY, USA). All data were shown as mean ± standard error. The means of DW, gas exchange properties, and mineral concentrations with six replicates in each treatment were compared using Duncan's multiple range test at a significance level of *p* = 0.05 in one-way analysis of variance (ANOVA). The effects of N supply, [CO2], growth stage, and their interaction on DW, gas exchange properties, and mineral concentrations were quantified using a general linear model. Correlation and significance tests between each mineral concentration in different tissues of cucumber and [CO2], N supply, transpiration rate were calculated using the Pearson correlation coefficient with two-tailed test. All figures were generated by OriginPro (Version 8.0; OriginLab Corp., Northampton, MA, USA).

#### **3. Results**

#### *3.1. Dry Weight and Root to Shoot Ratio*

The effects of [CO2] levels, N levels, growth stages, and their interactions on root, stem, leaf, and total DW as well as root to shoot ratios (R/S) of cucumber plants are shown in Table 2. The growth stage significantly affected the DW of roots, stems, leaves, and total biomass of cucumber plants. The average DW of total plants was increased from 1.01 g plant−<sup>1</sup> at T1 stage to 6.14 g plant−<sup>1</sup> at T2 stage. As the aerial parts of cucumbers grew faster than root, the R/S significantly decreased from 0.107 to 0.093 during a growth period of 44 days.

**Table 2.** The dry weight of roots, stems, leaves, whole plants and root/shoot ratios of cucumber grown under different [CO2] and N levels at the seedling and initialfruitingstages(*<sup>n</sup>*=6).



**Table 2.** *Cont.*

Growth stage: T1, seedling stage (18 DAT); T2, initial fruiting stage (62 DAT); 2 CO2 level: C1, C2, and C3: 400, 625, and 1200 μmol mol−1; 3 N level: N1, N2, N3, N4, and N5: 2, 4, 7, 14, and 21 mmol L−1; 4 Means within rows at the same stage not followed by the same lower case letters are significantly different among different N levels in the same CO2 level, and not followed by the same upper case letters are significantly different among different CO2 levels in the same N level, according to Duncan's test at *p* < 0.05; 5 C: [CO2] level; N: N level; T: growth stage; 6 Asterisks (\*) indicate significant differences (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001). NS indicates non-significant differences (*<sup>p</sup>* ≥ 0.05).

1

N supply levels also significantly affected the biomass accumulation of cucumber plants, especially at T2 stage. At T1 stage, the DW of stems, leaves as well as total biomass of cucumber plants was significantly increased with the increase of N supply, especially at higher [CO2] treatments (C3), whereas the DW of roots was not influenced by N levels at all [CO2] levels. Specifically, the increase of total biomass from N1 to N5 was 28.0%, 37.0%, and 68.5% in treatments C1, C2, and C3, respectively. At T2 stage, the positive effects of N supplies were more obvious than that at T1 stage, and the DW of all parts of cucumber plants were much greater in N5 treatment than those in N1 treatment. The total biomass was increased by 1.57, 2.85, and 4.70 folds from N1 to N5 in treatments C1, C2, and C3, respectively. Since the improvement of DW in the aerial parts was more noticeable than that in roots, the R/S at each [CO2] treatment was significantly decreased with the increase of N supply at both growth stages.

With respect of [CO2] levels, the increase of DW by super-elevated [CO2] (C3) were more dramatic in moderate (N3) and high N supplies (N4 and N5), whereas there was little increase or even a decrease in low N supplies (N1 and N2). Generally, the increase of total biomass from C1 to C3 was 29.4%, 58.4%, and 51.7% at T1 stage, and was 68.9%, 69.2%, and 70.1% at T2 stage, in N3, N4, and N5 treatments, respectively. The R/S was not significantly affected by [CO2] levels, irrespective of the N supply and growth stage.

The interactions of [CO2] × N and [CO2] × N × growth stage had significant effects on the DW of stems, leaves, and total cucumber plants but it was not significant on root DW. The interactions of [CO2] × growth stage and N × growth stage had significant effects on the DW of all parts of and total cucumber plants. The interaction of neither two nor three of these factors had significant effects on R/S.

#### *3.2. Gas Exchange*

Generally, [CO2] levels, N levels, growth stages and their interactions all had significant effects on the Pn of cucumber plants (Figure 1). On 13 DAT, super-elevated [CO2] (C3) significantly increased the Pn in N3 and N4 treatments. On 38 DAT, the Pn under super-elevated [CO2] was the highest in all N levels. On 60 DAT, the Pn under super-elevated [CO2] was also the highest in all N levels except for N2. Compared with ambient [CO2] (C1), the increase of the Pn by super-elevated [CO2] (C3) was 60.1%, 115.5%, and 77.7% in N3, N4, and N5 treatment, respectively. However, elevated [CO2] (C2) did not significantly increase the Pn compared with C1 in all N levels at three growth stages. The Pn was usually increased with the N supply increasing at the same [CO2] and growth stage. The increase of the Pn from N1 to N5 in C3 treatment was 2.35, 0.89, and 2.24 folds on 13, 38, and 60 DAT, respectively. As the plant grew, the Pn was gradually increased in high N supplies (N4 and N5) under super-elevated [CO2]. Whereas in low N supplies (N1 and N2), the Pn reached its highest value earlier on 38 DAT.

The Gs of cucumber plants was significantly influenced by [CO2] levels, N levels, growth stages, and the interactions of [CO2] × growth stage and N × growth stage (Figure 1). The Gs was much lower in N1 treatment on 13 DAT and in N1 and N2 treatments on 60 DAT than that in other higher N levels in the same [CO2] level. The Gs was also depressed by higher [CO2] (C2 and C3) in N5 treatment on 13 DAT and in N1, N2, and N5 treatments on 60 DAT. In other N treatments and growth stages, there was only a decreasing trend not a significant decrease in the Gs by higher [CO2] treatments. Averaged across all N treatments, the decrease of Gs from C1 to C3 was 27.4%, 27.7%, and 21.2%, on 13, 38, and 60 DAT, respectively.

The changes of Tr of cucumbers grown under different N and [CO2] levels at three growth stages were similar to those of Gs (Figure 1). [CO2] levels, N levels, growth stages, and the interactions of [CO2] × growth stage and N × growth stage had significant effects on the Tr of cucumber plants. The Tr was inhibited by super-elevated [CO2] (C3) in N2, N4, and N5 treatments on 13 DAT, in N4 on 38 DAT, and in N1, N2, and N3 on 60 DAT. In other treatments, a decreasing trend in C3 was usually observed compared with C1. The average decrease of Tr from C1 to C3 among all N treatments was 31.9%, 26.4%, and 22.3%, on 13, 38, and 60 DAT, respectively. Increasing N supply improved the Tr in C2 and C3 treatments on 60 DAT, and the increase from N1 to N5 was 72.6% and 130.2% in C2 and C3 treatment, respectively. The Tr differences among the three growth stages were not significant.

**Figure 1.** Net photosynthesis rate, stomatal conductance, and transpiration rate of cucumbers grown under different N and [CO2] levels at three growth stages (*n* = 6). Bars represent standard errors. CO2 levels: C1, C2, and C3: 400, 625, and 1200 μmol mol<sup>−</sup>1. N levels: N1, N2, N3, N4, and N5: 2, 4, 7, 14, and 21 mmol L<sup>−</sup>1. Means not followed by the same lower case letters are significantly different among different N levels in the same CO2 level and growth stage, and not followed by the same upper case letters are significantly different among different CO2 levels in the same N level and growth stage, and not followed by the same Greek letters are significantly different among different growth stages in the same CO2 and N level, according to Duncan's test at *p* < 0.05. In the internal table, C: [CO2] level; N: N level; T: growth stage. Asterisks (\*) indicate significant differences (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001); indicates non-significant differences (*p* ≥ 0.05).

#### *3.3. Mineral Nutrient Concentration*

As shown in Figure 2, [CO2] levels, N levels, growth stages, and the interaction of [CO2] × N had significant effects on the [N] in all three parts of cucumber plants, and the interaction of [CO2] × growth stage, N × growth stage, and [CO2] × N × growth stage also had significant effects on the [N] in leaves of cucumbers. [N] in roots, stems, and leaves of cucumbers was significantly increased with the increasing N supply. At T1 stage (18 DAT), the average increase of [N] from N1 to N5 among all [CO2] treatments was 71.4%, 85.9%, and 61.2% in roots, stems, and leaves respectively, whereas the value was 50.0%, 71.7%, and 27.7% at T2 stage (62 DAT). [N] in leaves was significantly decreased by super-elevated [CO2] (C3) compared with ambient [CO2] (C1) in all N treatments at both growth stages except for N1 and N2 at T2 stage. Averaged across all N treatments, the decrease of [N] in leaves from C1 to C3 was 6.1% and 9.3%, at T1 and T2 stage, respectively. However, [CO2] levels did not affect the [N] in roots and stems in most treatments. As the plants grew, the [N] in stems in N2, N3, and N4 treatments in all [CO2] levels, in leaves in N1, N4, and N5 treatments in higher [CO2] levels (C2 and C3) were decreased.

**Figure 2.** Nitrogen concentrations in different tissues of cucumber plants grown under different N and [CO2] levels at two growth stages (*n* = 6). Bars represent standard errors. CO2 levels: C1, C2, and C3: 400, 625, and 1200 μmol mol<sup>−</sup>1. N levels: N1, N2, N3, N4, and N5: 2, 4, 7, 14, and 21 mmol L−1. Means not followed by the same lower case letters are significantly different among different N levels in the same CO2 level and growth stage, and not followed by the same upper case letters are significantly different among different CO2 levels in the same N level and growth stage, according to Duncan's test at *p* < 0.05. Means with asterisks are significantly different between two growth stages in the same CO2 and N level (\**p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001). In the internal table, C: [CO2] level; N: N level; T: growth stage. Asterisks (\*) indicate significant differences (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001); indicates non-significant differences (*p* ≥ 0.05).

The [P] in different tissues of cucumber plants grown under different N and [CO2] levels at two growth stages was shown in Figure 3. [CO2] levels, N levels, growth stages, and the interaction of N × growth stage had significant effects on the [P] in all three parts of cucumber plants. At T1 stage, [P] in stems was gradually decreased with the N supply increased, and the average decrease from N1 to N5 among all [CO2] treatments was 17.3%. At T2 stage, [P] in roots and stems was significantly decreased as the N supply increasing, and the corresponding average decrease from N1 to N5 was 44.2% and 38.3%, respectively. At T1 stage, [P] in moderate (N3) and high (N4 and N5) N supplies were usually lower in C3 than that in C1 treatment, specifically in roots of the N3 and N4 treatments, stems of the N4 and N5 treatments, and leaves of N3 and N5 treatments. At T2 stage, the differences of [P] between C1 and C3 were not significant except for N1 and N4 in stems and N2 and N4 in leaves.

**Figure 3.** Phosphorus concentrations in different tissues of cucumber plants grown under different N and [CO2] levels at two growth stages (*n* = 6). Bars represent standard errors. CO2 levels: C1, C2, and C3: 400, 625, and 1200 μmol mol<sup>−</sup>1. N levels: N1, N2, N3, N4, and N5: 2, 4, 7, 14, and 21 mmol L<sup>−</sup>1. Means not followed by the same lower case letters are significantly different among different N levels in the same CO2 level and growth stage, and not followed by the same upper case letters are significantly different among different CO2 levels in the same N level and growth stage, according to Duncan's test at *p* < 0.05. Means with asterisks are significantly different between two growth stages in the same CO2 and N level (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001). In the internal table, C: [CO2] level; N: N level; T: growth stage. Asterisks (\*) indicate significant differences (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001); indicates non-significant differences (*p* ≥ 0.05).

Figure 4 showed the [K] in different tissues of cucumber plants grown under different treatments. Similar to the [P], [CO2] levels, N levels, growth stages, and the interaction of N × growth stage had significant effects on the [K] in all three parts of cucumber plants. At T1 stage, [K] in roots was gradually decreased as the N level elevating, and the average decrease from N1 to N5 among all [CO2] treatments was 40.5%, whereas they were gradually increased in leaves as the N level elevated, and the average increase from N1 to N5 was 13.8%. Compared with ambient [CO2] (C1), [K] was significantly decreased by super-elevated [CO2] (C3) in N3 and N4 in roots, N3, N4, and N5 in stems and N2, N4, and N5 in leaves. At T2 stage, a decrease of [K] by elevated [CO2] was observed in N4 and N5 in roots and N4 in leaves. In terms of growth stages, [K] was decreased from T1 stage to T2 stage in stems of cucumbers in all treatments.

**Figure 4.** Potassium concentrations in different tissues of cucumber plants grown under different N and [CO2] levels at two growth stages (*n* = 6). Bars represent standard errors. CO2 levels: C1, C2, and C3: 400, 625, and 1200 μmol mol<sup>−</sup>1. N levels: N1, N2, N3, N4, and N5: 2, 4, 7, 14, and 21 mmol L<sup>−</sup>1. Means not followed by the same lower case letters are significantly different among different N levels in the same CO2 level and growth stage, and not followed by the same upper case letters are significantly different among different CO2 levels in the same N level and growth stage, according to Duncan's test at *p* < 0.05. Means with asterisks are significantly different between two growth stages in the same CO2 and N level (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001). In the internal table, C: [CO2] level; N: N level; T: growth stage. Asterisks (\*) indicate significant differences (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001); indicates non-significant differences (*p* ≥ 0.05).

[CO2] levels, N levels, growth stages, and the interaction of [CO2] × growth stages had significant effects on the [Ca] in all three parts of cucumber plants (Figure 5). [Ca] was always the lowest in highest N treatment (N5) at two growth stages in all three parts of cucumbers except for that in roots at T1 stage. The average decrease from N1 to N5 among all [CO2] treatments was 16.9% and 15.6% in stems and leaves respectively at T1 stage, and was 32.4%, 33.9%, and 15.3% in roots, stems, and leaves respectively at T2 stage. [Ca] was also decreased by elevated [CO2] in all three tissues regardless of N supply at T1 stage. The average decrease from C1 to C3 among all N treatments was 16.4%, 14.3%, and 10.4% in roots, stems, and leaves, respectively. Growing caused a significant decrease of [Ca] in 10 of 15 treatments in root and 11 of 15 treatments in leaves.

**Figure 5.** Calcium concentrations in different tissues of cucumber plants grown under different N and [CO2] levels at two growth stages (*n* = 6). Bars represent standard errors. CO2 levels: C1, C2, and C3: 400, 625, and 1200 μmol mol<sup>−</sup>1. N levels: N1, N2, N3, N4, and N5: 2, 4, 7, 14, and 21 mmol L<sup>−</sup>1. Means not followed by the same lower case letters are significantly different among different N levels in the same CO2 level and growth stage, and not followed by the same upper case letters are significantly different among different CO2 levels in the same N level and growth stage, according to Duncan's test at *p* < 0.05. Means with asterisks are significantly different between two growth stages in the same CO2 and N level (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001). In the internal table, C: [CO2] level; N: N level; T: growth stage. Asterisks (\*) indicate significant differences (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001); indicates non-significant differences (*p* ≥ 0.05).

The [Mg] in different tissues of cucumber plants growing under different N and [CO2] levels at two growth stages was shown in Figure 6. Generally, [CO2] levels, N levels, and growth stages had

significant effects on the [Mg] in all three parts of cucumber plants. [Mg] was almost the highest in lowest N treatments (N1) and the lowest in highest N treatment (N5) in stems and leaves at T1 stage and in roots and leaves at T2 stage. The corresponding average decrease from N1 to N5 among all [CO2] treatments was 19.1%, 27.1%, 24.5%, and 24.8%. At T1 stage, [Mg] was significantly decreased by higher [CO2] (C2 and C3) compared with ambient [CO2] (C1) at all N levels in roots except for N1, and in leaves except for N2. The average decrease among all N treatments was 26.4%, 0.2% (not significant), and 6.6% in roots, stems, and leaves respectively from C1 to C2, and the decrease from C1 to C3 was 16.9%, 10.0%, and 7.8%. Similar to [Ca], growing also caused a significant decrease of [Mg] in 10 of 15 treatments in root and 11 of 15 treatments in leaves.

**Figure 6.** Magnesium concentrations in different tissues of cucumber plants grown under different N and [CO2] levels at two growth stages (*n* = 6). Bars represent standard errors. CO2 levels: C1, C2, and C3: 400, 625, and 1200 μmol mol<sup>−</sup>1. N levels: N1, N2, N3, N4, and N5: 2, 4, 7, 14, and 21 mmol L<sup>−</sup>1. Means not followed by the same lower case letters are significantly different among different N levels in the same CO2 level and growth stage, and not followed by the same upper case letters are significantly different among different CO2 levels in the same N level and growth stage, according to Duncan's test at *p* < 0.05. Means with asterisks are significantly different between two growth stages in the same CO2 and N level (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001). In the internal table, C: [CO2] level; N: N level; T: growth stage. Asterisks (\*) indicate significant differences (\* *p* < 0.05; \*\* *p* < 0.01; \*\*\* *p* < 0.001); indicates non-significant differences (*p* ≥ 0.05).

## *3.4. Correlations between [CO2], N Supply, Transpiration Rate, and Mineral Nutrient Concentration*

To further evaluate the relationship between mineral nutrient concentration, [CO2], N supply levels, and Tr, the Pearson correlation coefficient was calculated (Figure 7). Tr was negatively correlated to [CO2] at both growth stages and was positively correlated to N supply levels at T2 stage. [N] in all parts of cucumber plants was significantly positively correlated to N supply levels and Tr at both growth stages except for the insignificant positive correlation between [N] in root and Tr at T1 stage. [N] was negatively correlated to [CO2] only in leaves at both growth stages. [P] in all parts of cucumber plants was significantly negatively correlated to N supply except for levels at T2 stage. [P] was only negatively correlated to [CO2] in stems at T1 stage and in leaves at T2 stage, and was not significantly correlated to Tr. [K] in all parts of cucumber plants was significantly negatively correlated to [CO2] at T1 stage and in leaves at T2 stage. [K] in all parts of cucumber plants was also significantly positively correlated to Tr at both growth stages except for that in roots at T1 stage. [K] in leaves was also significantly positively correlated to N levels, whereas it was negatively correlated to N levels in root at T1 stage. [Ca] in all parts of cucumber plants was significantly negatively correlated to [CO2] at T1 stage, whereas this correlation was not significant at T2 stage. [Ca] was also significantly negatively correlated to N levels in all parts of cucumber plants at both growth stages except for that in roots at T1 stage. [Ca] in all parts of plants at T1 stage and in leaves at T2 stage also had a strong positive correlation to Tr. [Mg] in all parts of cucumber plants at T1 stage and in stems at T2 stage was significantly negatively correlated to [CO2]. [Mg] in stems and leaves at T1 stage and in roots and leaves at T2 stage was also significantly negatively correlated to N levels. [Mg] in aerial parts of cucumber was all significantly positively correlated to Tr at both growth stages.

**Figure 7.** Pearson correlations between [CO2], N supply, transpiration rate and mineral nutrient concentration of cucumber grown under different N and [CO2] levels at two growth stages. Tr: transpiration rate.

#### **4. Discussion**

#### *4.1. Impacts of [CO2] and N Supply on the Growth of Cucumbers*

The beneficial effects of CO2 enrichment on stimulating plant growth and biomass accumulation of crops have been extensively reported [2,5,6,40]. An increase in the DW ranging from 13.5% to 34.4% has been reported for cucumbers, when [CO2] was elevated to 500–760 μmol mol−<sup>1</sup> [1,31,33,41,42], and ranged from 43.9% to 128.0% when [CO2] was elevated to 1000–1200 μmol mol−<sup>1</sup> [28–31,43]. In the present work, in moderate N (N3), the increase in the DW was 30.5% and not significant when [CO2] was elevated to 625 μmol mol<sup>−</sup>1, and 29.4% and 68.9% when [CO2] was elevated to 1200 μmol mol<sup>−</sup>1, at T1 and T2 stage, respectively (Table 2). This increased DW caused by more fixation of CO2 and accumulation of biomass under elevated [CO2] combined with moderate and high N supplies [5,6]. Our results showed that the increase in Pn was greater than 60% by super-elevated [CO2] compared with ambient [CO2] in moderate and high N supplies (Figure 1), which were in close conformity with this explanation.

Moreover, the present work also clearly demonstrated that the DW accumulation by CO2 enrichment depended on the N supply, in which there was significant increase in DW in moderate and high N supplies but no change or decrease in low N supplies (Table 2). This is consistent with the previous findings that limited N will inhibit the synthesis of photosynthetic related proteins, lower the photosynthetic capacity, and reduce the photo-assimilate accumulation [12,15,44,45]. The significant decrease of [N] in leaves at T1 stage, as well as no change or decrease in Pn by elevated [CO2] observed in low N supply treatments in this work, also gave a clear indication that the stimulation of Pn at high [CO2] was only partial or counteracted when N supply was limited (Figures 1 and 2). It is worth mentioning that the improvement in Pn by CO2 enrichment maintained at 60 DAT even when the [N] in leaves was also deceased in high N supplies (Figures 1 and 2). A possible reason is that the [N] in leaves treated with high N supplies was still enough for guaranteeing the RuBP regeneration and Rubisco activity to match the increased C-fixation [10,46]. These results also confirmed that increasing N supply could alleviate or prevent the photosynthetic acclimation under elevated [CO2] condition and ensure the sustainability of the [CO2] enrichment fertilization effects on crop growth [15–17,33].

Decreased Gs and Tr are the most obvious and universal changes observed in C3 plants including cucumbers grown in elevated [CO2] condition [41,47–49]. Elevated [CO2] causes partial stomatal closure and decreases the Gs by 8–44% for C3 plants, consequently with a reduction of Tr by 20–40% [2,50]. In the present work, the decrease in Gs and Tr of cucumbers in super-elevated [CO2] compared with the ambient [CO2] was 12.2–27.7% and 22.9–31.9%, respectively (Figure 1), which was in good agreement with previous reports. Additionally, we found a greater reduction in Gs and Tr in low N supply treatments (Figure 1). Low N causes reductions in Rubisco concentration and activity hereby forces the reduction in Gs and Tr in order to maintain a constant ratio of internal leaf [CO2] to that of outside air [50]. The improvement in Gs and Tr under higher N treatments might also result from the synthesis of more photosynthates and increased cell wall rigidity [51].

[CO2] and N Supplies also affect the biomass allocation of the plants. In the present work, the R/S was significantly decreased with N supplies increasing, but was not affected by [CO2] levels (Table 2). A similar phenomenon was observed in other reports and our previous works [23,28,30,31,52]. Incorporated with current views in the literature, two possible reasons were usually proposed. Firstly, the accumulation of nitrate in shoots under high N supply down-regulates the growth of roots relative to shoots, resulting in lower R/S [53]. Secondly, plants always allocate more biomass to the apparatus in nutrient-limiting conditions, so more photosynthates will be invested in roots for exploring and acquiring more nutrients in N-deficient conditions [54].

#### *4.2. Key Factors A*ff*ecting the Mineral Nutrient Concentrations*

The changes of [N] under different [CO2] and N supply conditions have been extensively studied. The results that [N] was decreased as [CO2] increasing and increased as N supply increasing have been

frequently observed for crops and vegetables [8,9,19]. The reasons for the increase in [N] with increasing N supply are transparent, whereas the reasons for the decrease in [N] with increasing [CO2] are more complicated. The most common reason is that the [N] is diluted by more accumulated carbohydrates in high [CO2] conditions [9,10,12]. However, tissues respond differently to elevated [CO2] conditions, and it has been reviewed that the average decrease in [N] in leaves was 16%, which was larger than that in stems (9%) and roots (9%) [7,9]. In the present work, only the [N] in leaves was significantly decreased by 6.1–9.3% in super-elevated [CO2] treatment, whereas [N] in stems and roots were not influenced by [CO2] levels (Figures 2 and 7). Therefore, the dilution effect is not the only cause of [N] decrease in elevated [CO2]. Previous elevated [CO2] studies have found that the NO3 − assimilation was enhanced in roots by increased photosynthate translocation to roots, whereas it was inhibited in leaves caused by the competition for reductants between the carbon fixation and NO3 − reduction [17,25,26]. Hence, the dilution effect on [N] could be counteracted by the enhanced NO3 − assimilation in roots while aggravated by the inhibited NO3 − assimilation in leaves. Additionally, the decrease in Tr caused by the closed stomata could result in a reduced flow of NO3 − from roots to leaves, which also leads to more decrease in [N] in leaves than that in roots [17,19,21,22,55]. In the present work, a significant positive correlation between [N] and Tr, especially in leaves, also gave convincing evidence for this reason (Figure 7).

In this study, [P] in moderate and high N supplies was decreased in higher [CO2] levels at T1 stage, and was significantly decreased in roots and stems with the increasing N supply (Figure 3). The similar results were also found in tomatoes [17], beans [56], and soybeans [57]. This decrease of [P] in elevated [CO2] especially associated with higher N supplies has been considered as the result of the dilution effect [10,17]. When the larger biomass was accumulated in higher N supplies and elevated [CO2] conditions, [P] in plants will decease if the P supply was not changed [20]. Besides, we also found there was little correlation between [P] and Tr (Figure 7). This lack of response of [P] to Tr may be explained by the free transportation pathway of phosphate in the xylem [58,59]. When phosphate is delivered from root to shoot in xylem sap, it can also be redistributed between different tissues according to their own demands, and excess phosphate will be stored in the vacuoles to maintain the cellular phosphate homeostasis, which is less affected by Tr.

In moderate and high N supplies, a decrease of [K] by elevated [CO2] was found in the whole plant at T1 stage, and in roots and leaves at T2 stage (Figure 4). There was also a significant negative correlation between [K] and [CO2] level in the whole plant at T1 stage, and in leaves at T2 stage (Figure 7). These results are consistent with the average decrease of 10% [K] in plants by elevated [CO2] in previous reviews [7,22]. The dilution effect and reducing Tr are considered as two key factors driving this decrease [17,22,56,60,61]. The significant positive correlation between [K] and Tr at both growth stages also confirmed the importance of Tr on the transportation of K in cucumber plants (Figure 7). Interestingly, at T1 stage, [K] in leaves was significantly increased with the increasing N supply, whereas [K] in roots was decreased. A possible reason is that K<sup>+</sup> is always transported accompanied with NO3 <sup>−</sup> from roots to shoots to maintain the balance between K<sup>+</sup> and NO3 − in xylem sap [62,63]. So, transportation of more NO3 − in high N supply will cause a synchronous increase of K<sup>+</sup> in xylem sap, and result in an increase in [K] in leaves and a decrease in [K] in roots in higher N treatments.

Although an average decrease of 8% in leaf [Ca] has been reviewed [7,22], the change of [Ca] in different tissues and species under elevated [CO2] were different [60,61,64]. On one hand, being different from N, P, and K, the largest [Ca] are always found in cell walls in plants, where Ca2<sup>+</sup> are stably fixed not only by electrostatic interactions with carboxylic groups of pectin, but also by coordination linkage with hydroxylic groups of polysaccharides [65]. So, when plant growth is improved under elevated [CO2], the demands of Ca<sup>2</sup><sup>+</sup> are also increased, which could partially offset the dilution effect [64]. On the other hand, Ca in xylem sap is mainly in the form of ions or chelate, so its transportation from roots to the aerial parts largely depends on the Tr [22]. However, due to the inhibited transportation of Ca in the phloem, Ca could hardly move and be reused from old tissues to young tissues [65]. Therefore, the accumulation of Ca could happen in old tissues and counterbalance the negativity of the dilution effect and reduced Tr on [Ca]. In the present work, an average decrease of [Ca] by 16.4%, 14.3%, and 10.4% in roots, stems, and leaves was observed under [CO2] enrichment at T1 stage, respectively (Figure 5). The significant negative correlation between [Ca] and [CO2], and the positive correlation between [Ca] and Tr, implied that the dilution effect and reducing Tr were two key factors driving [Ca] decrease in young plants (Figure 7) [7,63]. At T2 stage, elevated [CO2] had little effects on [Ca], and this might be due to the accumulation of Ca in older tissues that offsets the dilution effect and reduced Tr (Figure 5). Since the leaves we analyzed did not include the old leaves that had fallen, the [Ca] in leaves was still influenced by Tr, and the positive correlation between [Ca] and Tr was observed only in leaves. Besides, the decrease of [Ca] with the increasing N supply may also be the result of the dilution effect by the increased biomass accumulation in higher N levels [18].

A decrease in [Mg] in elevated [CO2] conditions has been frequently reported, and the average decrease value was 10% in leaves [7,18,60,64]. McGrath and Lobell found a 20% decrease in [Mg] but only a 10% decrease of other mineral elements in leaves of wheat [22]. Based on the chlorophyll concentration analysis and mass flow experiment, they calculated that the dilution effect accounted for a 10% reduction of [Mg], reduced Tr accounted for 3–10%, and reduced chlorophyll content accounted for 1–5% (represents the reduced demands). In the present work, we found an average decrease in [Mg] by 16.9%, 10.0%, and 7.8% in roots, stems, and leaves in elevated [CO2] conditions, respectively, and an average decrease of [Mg] by 19.1–27.1% under higher N supply (Figure 5). The significant negative correlation between [Mg] and [CO2] as well as [Mg] and N supply indicated the dilution effect had a strong impact on [Mg] (Figure 7). Meanwhile, the significant positive correlation between [Mg] in the aerial parts and Tr implied that reduced Tr also had detrimental effects on [Mg] (Figure 7).

#### **5. Conclusions**

According to our results, the cucumber biomass accumulation could be significantly increased by elevated [CO2] accompanied by high N supplies. High N supplies could further improve the Pn and offset the decrease of the Gs and Tr by elevated [CO2]. Thus, increasing N supply could alleviate or prevent the photosynthetic acclimation under elevated [CO2] conditions. Based on the mineral nutrient concentrations in different [CO2] and N supply treatments and the correlation analysis, we proposed the key factors affecting the concentration changes of each mineral element. The dilution effect was the main factor that reduced all mineral elements, whereas Tr had a large impact on the decrease of [N], [K], [Ca], and [Mg] except [P]. The decreased demands of N and Mg and the increased demands of Ca also influenced the concentrations of the corresponding elements in cucumber plants. However, this study was just a qualitative analysis. A quantitative analysis of the effect of each factor on the concentration changes is urgently needed. When we have better understanding of the mechanisms controlling the mineral concentration changes in cucumber plant responding to elevated [CO2], we could optimize the mineral fertilization in order to improve the growth of cucumber plant under elevated [CO2] conditions. Thus, a sustainable vegetable production with higher C and N use efficiency and less CO2 emission and fertilizer input will be achieved.

**Author Contributions:** Conceptualization, X.L., J.D., W.C., and Z.D.; methodology, X.L., J.D., and W.C.; investigation, X.L., J.D., and W.C.; data curation, X.L. and W.C.; writing—original draft preparation, X.L. and W.C.; writing—review and editing, J.D. and N.S.G.; funding acquisition, W.C., X.L., and Z.D. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Natural Science Foundation of China (41877103), the Strategic Priority Research Program of the Chinese Academy of Science (XDA23020401), and the Frontier Project of Knowledge Innovation Program of Institute of Soil Science, Chinese Academy of Sciences (ISSASIP1635).

**Acknowledgments:** The authors are grateful to Hua Gong for assistance with the ICP analyses and to two anonymous reviewers for their constructive comments and suggestions.

**Conflicts of Interest:** The authors declare no conflict of interest.

## **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Antioxidant Seasonal Changes in Soilless Greenhouse Sweet Peppers**

## **Damianos Neocleous <sup>1</sup> and Georgios Nikolaou 2,\***


Received: 10 October 2019; Accepted: 6 November 2019; Published: 8 November 2019

**Abstract:** This study was commissioned to study the effect of the growing season on the antioxidant components of greenhouse sweet pepper crops, which is of scientific interest because of their possible beneficial health effects. The total antioxidant activity (estimated by ferric reducing antioxidant power-FRAP assay) major antioxidants (ascorbic acid, phenolics and carotenoids) and taste fruit quality characteristics (soluble solids, titratable acidity, dry matter and sugars) were recorded in soilless-grown sweet pepper cultivars of red, orange, yellow and green color at four harvesting season months, i.e., February (winter), May (spring), July (summer) and October (autumn). The results showed seasonal variations in antioxidant components and activity of pepper fruits. In most cases measured parameters showed higher values in spring (May) and summer (July) compared with winter (February) and autumn (October) growing seasons. This study indicates that during late autumn and winter, lower levels of solar irradiance, ultraviolet radiation and temperature in Mediterranean greenhouses can be insufficient to stimulate phytochemicals production in peppers; thus, plant–light interception must be more actively managed.

**Keywords:** *Capsicum annuum* L.; colored sweet peppers; antioxidant activity; phenolics; ascorbic acid; carotenoids; solar and ultraviolet radiation; soilless culture

## **1. Introduction**

Peppers are among the most consumed vegetables worldwide. In the United States consumption of fresh bell peppers (*Capsicum annuum* L.) increased up to 20% the last decade and averaged 11.2 pounds per person in 2018 [1]. Bell peppers have become one of the most important cultivated fruiting vegetable in Mediterranean greenhouses. Particularly, solanaceous crops (tomato, pepper and eggplant) constitute about 60% of greenhouse-cultivated areas, which are often cultivated in soilless culture to enhance yield, product quality, water use efficiency and sustainability [2,3]. To date, peat-based substrates are most widely used in fruiting vegetable production systems, although rockwool is still the dominant soilless culture system in Europe [4]. Furthermore, alternative ecofriendly substrates (e.g., biowaste materials) with a lower carbon footprint coupled with soilless culture systems may be considered as a useful tool in sustainable greenhouse horticulture [5].

Recently, Mediterranean sweet peppers gained a growing interest to produce brightly colored (e.g., red, yellow, orange) fruits throughout the year based on the favorable climatic conditions (high radiation and mild autumn and winter temperatures) of the region [6], modern soilless culture technologies and the good marketability of the product. In this context, EU [7] has established specific marketing standards for sweet peppers. However, many consumers have additional quality requirements, which can go beyond legislation and standards [8]. Thus, the interest in foods of plant origin as a source of phytochemicals, increases throughout the years [9]. Specifically, the consumption of natural antioxidant compounds coming from fruits and vegetables such as phenolics, ascorbic acid and carotenoids have been associated with prevention of chronic diseases (e.g., cardiovascular disease and different forms of cancer) due to their ability to neutralize free radicals in the human body [10,11]. Peppers are among vegetable crops that are considered as naturally abundant in plant phytochemicals and their composition is of major importance for the beneficial health effects of the product [12,13]. As already mentioned peppers are widely grown in Mediterranean greenhouses, however, the lack of information regarding seasonal influences on the accumulation of dietary antioxidants in colored pepper fruits represents a drawback (e.g., in diets based on antioxidant intake). It is relevant that in other crops such as tomatoes and lettuce, seasonality can affect their antioxidant composition [14,15].

It is well known that genetic and environmental factors may directly affect the antioxidant composition of plant parts [16,17]. Particularly, under Mediterranean climatic conditions, increased solar irradiation and mild temperatures during winter months can affect plant antioxidant content and eventually fruit quality [18,19]. Although different species may have different responses [20], there is a general notion that solar ultraviolet radiation (UV total radiation 280–400 nm) has relevant biological effects on agroecosystems and induce the accumulation of phenolic compounds in the plant [14,19,20]. However, global UV fraction (i.e., ratio of the UV to global solar radiation) is highly dependent on variations in the concentration of clouds, water vapor and aerosols in the atmosphere and may vary from 2.0% to 9.5% [21,22]. Such environmental factors impact the quality of greenhouse vegetables. For example, a positive correlation between light levels and levels of secondary metabolites such as ascorbic acid in sweet peppers has been reported [8]. In accordance, the increase levels of Ultraviolet B (UVB) radiation (280–315 nm) enhanced several defense compounds such as carotenoids and flavonoids in bell peppers [23]. Ultraviolet A (UVA) radiation (315–400 nm) also enhanced the amounts of secondary metabolites, soluble carbohydrates, free amino acids and proteins in greenhouse peppers [24]. Consequently, it is assumed that changes in global solar and UV radiation and air temperature levels at different harvesting seasons, may affect the antioxidant components of greenhouse soilless-grown peppers, which is of scientific interest because of their possible beneficial effects on human health.

To document this response, this study was designed to evaluate seasonal effects in total antioxidant activity (estimated by ferric reducing antioxidant power-FRAP assay) and important antioxidant compounds (i.e., ascorbic acid, phenolics, carotenoids) so as in other fruit quality characteristics (soluble solids, titratable acidity, pH, dry matter, reducing sugars) in soilless sweet peppers of red, orange, yellow, and green color, in Mediterranean greenhouses.

## **2. Materials and Methods**

#### *2.1. Plant Material and Agronomic Features*

Data was collected from the same plants of a year round sweet pepper crop, giving harvests in 2017, at greenhouse facilities of Agricultural Research Institute of Cyprus (34◦94 N, 33◦19 E, altitude 40 m). Three colored pepper (*Capsicum annuum* L.) cultivars (Agroglobal, Hungary), namely red (*cv. Castello*), orange, (*cv. Donat*), yellow (*cv. Solero*) and one green local variety (*cv. Glikes*) were grown on rockwool substrates (Grodan Company, Denmark; slabs dimensions 100 cm × 20 cm × 7.5 cm). Substrates were placed into Polygal-gutters (Mapal Plastics, Israel) 12-m long, which were supported by metal frames in 12 single rows. Each experimental unit consisted of one Polygal-gutter planted with 24 plants. Three replications (one Polygal-gutter per replication) for each cultivar were randomly arranged in three blocks. The plants were vertically supported ('V' system) giving a planting density of 2.0 plants m−<sup>2</sup> [25]. Colored fruits (red, orange, yellow) ripened on the vine following the mature green stage. The crop was transplanted on rockwool slabs 3 months prior harvesting, which started in February and terminated in October 2017. Harvested fruits of each experimental unit were weighted and counted to determine fresh yield and average weight of the fruit. Total marketable fruit yield was the combined total of Extra Class and Class I according to EU marketing standards [7]. For quality

analysis, fruits were sampled at four harvesting times, particularly in February (winter), May (spring), July (summer) and October (autumn).

The irrigation schedule was controlled by Fertimix hydroponic head unit (Galgon, Kfar Blum, Israel) and adjusted to light conditions [26]. The start of irrigation was depended from light sums according to the growth stage (1500–2800 kJ/m2) targeting a leaching fraction of about 20%. Drip emitters delivered the nutrient solution directly to the root zone of pepper plants. Electrical conductivity (EC) and pH values were monitored in both irrigation and drainage water. The target EC levels of the irrigation nutrient solution were adjusted in response to radiation differences (± 0.3 dS/m; higher EC at low radiation and lower EC at high radiation). The hydroponic fertigation head prepared a nutrient solution (NS) for growing soilless peppers in Mediterranean greenhouses with NS composition originating from the literature [2]. The following NS was delivered to the plants at the vegetative stage: 5.4 mM K+, 4.65 mM Ca2+, 1.6 mM Mg2<sup>+</sup>, 1.2 mM NH4 <sup>+</sup>, 13.7 mM NO3 <sup>−</sup>, 1.2 mM H2PO4 −, 1.85 mM SO4 <sup>2</sup>−, 15 μM Fe as Fe-EDDHA, 10 μM Mn, 5 μM Zn, 0.8 μM Cu, 30 μM B, and 0.5 μM Mo. Corresponding EC and pH values were 2.20 dS/m and 5.6, respectively. At the reproductive stage the plants were fed with the following nutrient solution: 5.8 mM K+, 4.5 mM Ca2+, 1.40 mM Mg2+, 0.6 mM NH4 <sup>+</sup>, 13.0 mM NO3 <sup>−</sup>, 1.2 mM H2PO4 <sup>−</sup>, 1.75 mM SO4 <sup>2</sup>−, 15 μM Fe as Fe-EDDHA, 10 μM Mn, 5 μM Zn, 0.8 μM Cu, 30 μM B, and 0.5 μM Mo. The EC and pH values of this NS were 2.10 dS/m and 5.6, respectively.

#### *2.2. Greenhouse Facilities and Climatic Data*

The experiment was conducted in a North–South oriented greenhouse with a total ground area of 216 m2, with cutter height 3.50 m, ridge height 5.26 m, spans width 6 m and total length 18 m. The gable end and side walls were covered with double-walled polycarbonate and the roof was covered with a common polyethylene film (88% light global transmission, 55% light diffused transmission and 88% thermal efficiency). In each greenhouse span there was a single continuous rooftop window for natural ventilation. In addition, evaporative cooling was performed by a fan-pad system consisted of four fans, two at each span and a wetted pad. The greenhouse floor was completely covered by a white, water permeable polypropylene sheet.

External climatic parameters measured were air relative humidity (RHo, %) and temperature (To, ◦C) (Sensor type PT 100; Galcon, Kfar Blum, Israel) and net solar radiation (Gh, kJ/m2) with a pyranometer at 3 m above the greenhouse (Sensor type TIR-4P; Bio Instruments Company, Chisinau, Moldova). The same types of sensors were used for monitoring relative humidity and air temperature within the greenhouse. All measurements were recorded every 30 s on a data logger system (Galileo controller; Galcon, Kfar Blum, Israel) and a ten-minute average was estimated. Vapor-pressure deficit (VPD) was estimated based on greenhouse air temperature and relative humidity. The mean daily value of ultraviolet radiation over a month was calculated based on global solar radiation, following Equation (1). According to this formula [27], the hourly and daily values of both radian fluxes are highly correlated with a general linear relationship of the following form providing coefficients of determination of R2 always greater than 0.91 for hourly and 0.88 for daily fittings in the case of Cyprus.

$$\mathbf{Gw} \mathbf{v} = \mathbf{a} \mathbf{Gh} \tag{1}$$

where Guv is the solar global ultraviolet radiation (kJ/m2); Gh is the solar global radiation (kJ/m2); a is the slope corresponding to measurements.

#### *2.3. Fruit Quality Measurements*

The fruit quality characteristics (i.e., ascorbic acid, sugars, total soluble solids, pH, titratable acidity and dry matter) were determined at commercial maturity stage (Figure 1) in randomly selected samples excluding outliers, from each experimental unit. The edible part of the fruit was homogenized and soluble solids (◦Brix; Atago PR-1, Tokyo, Japan), pH (Mettler Toledo, Switzerland), titratable acidity (titration with sodium hydroxide solution to pH 8.2, % citric acid), and fruit dry matter (g/100 g FW, drying at 70 ◦C) were recorded. The ration between total soluble solids and titratable acidity was calculated (TSS/TA). The content of fruits in ascorbic acid (mg AA/100 g FW) and reducing sugars (mg Glucose+Fructose/g FW) were determined by a Merck RQflex reflectometer. Briefly, ascorbic acid reduces yellow molybdophosphoric acid to phosphormolybdenum blue that is determined reflectometrically as reducing sugars after enzymatic conversion with glucose-6-phosphate dehydrogenase and diaphorase according to the company protocols (Merck, Darmstadt, Germany). For the determination of the total phenolic content and antioxidant activity, subsamples of fruits were kept-frozen at −30 ◦C until the date of analysis. Quantitative determination of phenolic substances was performed in fruits samples (10 g) homogenized with 25 mL acidified acetone (acetone: water: acetic acid 70:29.5:0.5, v:v:v), following the Folin–Ciocalteu procedure [28]. The absorbance of the reaction mixtures (0.25 mL extract, 2.5 mL FolinCiocalteu's reagent (previously diluted 1:10 with deionized water) and 2 mL 7.5% Na2CO3) after 5 min at 50 ◦C was measured at 760 nm (UV-Vis spectrophotometer Helios Zita, Thermo Fisher Scientific, USA). The results were expressed in gallic acid equivalents (mg GAE) per g of fresh weight, using a calibration curve (GAE/g FW) [29]. For the determination of the antioxidant capacity of pepper fruits by the ferric reducing antioxidant power method (FRAP; [30,31], sample extracts (100 μL) were mixed with 3 mL FRAP reagent (1:1:10 mixture of 20 mM FeCl3, 10 mM TPTZ and 0.3 M acetate buffer at pH 3.6) and after 4 min at 37 ◦C the absorbance at 593 nm was recorded. Ascorbic acid (AA) was used as standard and the results were expressed per g of fresh weight (μmol AA/g FW) as previously described [29]. Chlorophyll content was determined in green fruit samples blended with 80% acetone measuring the absorbance of the supernatant at 648 and 664 nm [29]. Total carotenoids content in colored fruits extracts (hexane: acetone: ethanol 50:25:25, v:v:v) was determined at 450 nm following concentration calculations as previously reported [32]. The results were expressed as mg β-carotene per g of FW.

**Figure 1.** Fruit maturity at the time of harvest in (**a**) red; (**b**) orange; (**c**) yellow; (**d**) green sweet pepper (*Capsicum annuum* L.) cultivars grown in greenhouse soilless culture.

## *2.4. Statistical Analysis*

Experimental layout in the greenhouse consisted of three replicates for each cultivar arranged in a randomized complete block design. SAS software system (ver. 9.2, Cary, NC, USA) was used for analysis of variance (ANOVA) for all traits studied and means were separated using DMRT at 5% level of significance. Pearson correlation coefficients between antioxidant variables studied were calculated.

## **3. Results**

## *3.1. Greenhouse Microclimate and External Climatic Data*

The monthly mean values, of outdoor climate data (i.e., air temperature and relative humidity) and inside greenhouse microclimate; global solar radiation and calculated ultraviolet radiation are presented in Table 1. The monthly variability of both radiant fluxes, Gh and Guv, is shown in Figure 2.

**Figure 2.** (**a**) Monthly means of global solar and ultraviolet radiation (kJ/m2; the bars are in relation but not proportional to the data they encode); (**b**) Sun orientation during the experiment; Straight-line embedded in the graph (**a**) represents minimum radiation requirements for cultivation of thermophilic vegetable species in Mediterranean greenhouses [33].

The mean estimated values kJ/m<sup>2</sup> (±standard deviation) of Gh and Guv were respectively 1378 (799.99) and 42 (24.08) in winter (D-J-F), 2087 (1216.19) and 74 (43.48) in spring (M-A-M), 2451 (1222.93) and 77 (38.76) in summer (J-J-A) and 1788 (976.99) and 69 (39.42) in autumn (S-O-N). Seasonal variations of Guv value followed seasonal variations of Gh. Particularly, higher Guv values observed during summer and lower values at winter; as affected by yearly length of a day and the solar zenith angle. However, from Table 1 we can observe that, despite the decrease of Gh from August to September by 15%, Guv values increased by 17%. The line in Figure 2a represents minimum radiation requirements for cultivation of thermophilic vegetable species in Mediterranean greenhouses during N-D-J according to the literature [33].


Monthly mean values (±standard deviation) of outdoor climate data and inside greenhouse microclimate for daylight hours.

**Table**

**1.**

greenhouse air temperature (◦C); RHi, inside greenhouse air relative humidity (%); VPD, inside greenhouse air vapor pressure deficit (kPa).

#### *3.2. Antioxidants and Other Fruit Quality and Yield Parameters*

Season and cultivar were in most cases significant sources of variation (Table 2). Because of some interactions observed between season and cultivar, data were graphically presented within each cultivar (Figure 3). The antioxidant activity (FRAP values; μmol AA/g FW) of the pepper cultivars tested showed higher values in spring (May) and summer (July) compared with winter (February) and autumn (October) (Figure 3). An increase was also observed for total phenolics (GAE/g FW) during May compared with February in orange and yellow cultivars, however, in red and green cultivars the increase was not significant (*p* < 0.05; Figure 3). Accordingly, ascorbic acid content (mg AA/100 g FW) showed higher value in May and July and lower in February and October in all cases (Figure 3). Similarly, sugars (mg Glucose + Fructose/g FW) were accumulated at higher levels during May and July compared with the other two months in red, orange and yellow pepper fruits, whereas in green fruits the values observed were not differentiated with harvest time (Figure 3). Changes in total soluble solids ( ◦Brix) with harvesting time-followed alterations of the sugar content as may be expected. However, in some cases (orange cultivar) differences were not consistent (Figure 3). The titratable acidity (% citric acid) was higher during July compared with February for red, orange and yellow cultivars, whereas no variation was observed among harvest times for the green cultivar. Yet importantly, carotenoids content (mg β-carotene/g FW) at harvesting times May and July was enhanced in colored pepper fruits in relation with the other two months depending on the cultivar (Figure 3). On the contrary, total chlorophyll (a + b) content in the green cultivar remained unaffected by the growing season (Figure 3), so as the dry matter content in most of the cases. Similarly, the estimated ratio total soluble solids to titratable acidity (TSS/TA) was not differentiated among harvest months in the cultivars tested. Overall, total marketable fruit yield (kg/m2) and mean fruit weight (g/fruit) was greater in colored peppers in relation to the green cultivar (Figure 4). On the contrary, more fruits per m2 were produced by the green than the rest of the colored cultivars (Figure 4). Last but not least, FRAP values were highly correlated (*p* < 0.001) with phenolics (*r* = 0.81) and ascorbic acid (*r* = 0.84), whereas pigment phytochemicals had a lower influence to the reducing potential. In addition, phenolics were highly correlated with ascorbic acid (*r* = 0.77) so as both with reducing sugars (*r* = 0.60 and *r* = 0.83, respectively).


**Table 2.** Analysis of variance table and levels of significance (*p* < 0.05, *p* < 0.01, *p* < 0.001).

Antioxidant activity (FRAP), phenolics (Ph), ascorbic acid (AA), sugars (Sug), total soluble solids (TSS), titratable acidity (TA), carotenoids (Car), total chlorophyll (Chl) and dry matter (DM).

q

**&DURWHQRLGVPJȕFDURWHQHJ):**

**Figure 4.** Total marketable fruit yield (kg/m2), average fruit weight (g/fruit) and fruit number per m2, in red, orange, yellow and green sweet pepper cultivars from February 2017 until October 2017. Different lower-case letters above the bars indicate significant differences between mean values at *p* < 0.05 according to Duncan's test. Error bars indicate ± standard errors of the mean.

#### **4. Discussion**

Pepper plants were grown in a plastic greenhouse under soilless conditions giving harvests from February until October 2017, to study seasonal variations in fruit antioxidants and antioxidant activity in cultivars of red, orange, yellow and green color. Experimental results on other crops such as spinach and tomatoes demonstrated that antioxidant activity and phytochemicals including phenolics and ascorbic acid were greatly affected by the growing season [14]. In this study, total antioxidant activity and major antioxidant components including, phenolics, ascorbic acid, and carotenoids were higher in pepper fruits during harvesting on May and July in relation to the other two harvesting periods (i.e., February and October) depending on the cultivar (Figure 3). This increase was associated with an increase (avg. 30%) of solar and ultraviolet (UV) radiation and elevated temperature conditions inside the greenhouse from the autumn–winter to spring–summer period (Table 1). Ultraviolet radiation (i.e., Guv values) followed seasonal variations of global solar radiation (i.e., Gh values) as affected by the yearly length of a day and the solar zenith angle Figure 2; [27]. In this context, literature suggests that light intensity is closely related with the biosynthesis of bioactive (i.e., biologically active) compounds of secondary metabolism in plants such as phenolics, because light increases the activities of key enzymes in the phenolic synthesis such as phenylalanine ammonia lyase (PAL) [14]. It is also known that the synthesis of secondary metabolites in plants is involved in the defense mechanism against several stresses such as UV (280–400 nm) radiation [11]. In accordance, other authors [9] reported phenolics accumulation in different fruits and vegetables in response to UVB (280–315 nm) exposure due to increase expression of phenylpropanoid pathway genes. For example, UV-treated sweet pepper plants contained higher amounts of bioactive compounds such as phenolics so as soluble carbohydrates and photosynthetic pigments at earlier reports [24]. In addition, biosynthesis of phenolic constituents with well-known antioxidant properties in *Capsicum annuum* and other *Capsicum* species including flavonoids, quercetin and luteolin [34], were connected to UVB radiation [11]. These irradiance effects on metabolic functions may also be used to explain phenolics changes with growing season in the current study. In absolute values, even though phenolic and ascorbic acid accumulation is greatly affected by pre- and post-harvest factors [16], the values observed in this study were in the same range with those reported in other experiments for hydroponic sweet peppers grown in a Mediterranean type climate [35].

Moreover, light and temperature in the optimal range stimulate photosynthesis, which leads to the accumulation of reducing sugars and soluble solids in the fruits [36]. Indeed, the increase of ascorbic acid in colored pepper fruits of *C. annuum* L. during May and July was accompanied by an increase in reducing sugars and soluble solids, which supports previous findings for other *Capsicum* species, correlating ascorbic acid biosynthesis with light intensity [8] and reducing sugars (ascorbic acid is synthesized from D-glucose) [37]. Overall, ascorbic acid in fruits of red, orange and yellow colors varied in a range of 80–110 mg/100 g FW, which seems to fall close to that reported previously for greenhouse-grown colored sweet peppers in Spain [38]. Noteworthy, the values observed during spring–summer months were higher than 90 mg/100 g FW and the values during autumn-winter months were lower than 90 mg/100 g FW (Figure 3). Taking into consideration that 90 mg/day is the threshold of ascorbic acid recommended daily allowances for adult men set by the Food and Nutrition Board of the Institute of Medicine in the United States as cited in [37], this study clearly shows that the challenge to eliminate nutritional variations all year round of the selected crops is fundamental. Accordingly, growing season affected carotenoids formations in colored pepper fruits, with higher values observed during spring and summer at elevated light and temperature conditions. Earlier studies have clearly demonstrated that greater exposure to sunlight and higher temperature enhances carotenoid biosynthesis (isoprenoid pathway) in fruits [32]. Thus, the physiological mechanism implicated in the differences between seasons in the current is presumably based on biosynthesis of secondary metabolites from carbon skeletons derived from photosynthetic process [36]. Thus, it is reasonable to conclude that pepper fruit biochemistry was upregulated in response to prevailing environmental conditions (light and temperature) as previously suggested [8]. Considering the total antioxidant activity, the higher activity in spring and summer months was in accordance with the elevated concentrations of phenolics, ascorbic acid and carotenoids in pepper fruits, which confirms the close relationship between these antioxidant components with antioxidant activity [35,39] and their synergistic effect [40]. In general, antioxidant activity reflects the cumulative antioxidant function of a food product [13] and may serve as a tool in epidemiological studies [12]. Particularly, peppers had the second highest total antioxidant capacity among 34 vegetables as reported previously [13]. Summarizing, these results let us suggest that in Mediterranean greenhouses during late autumn and winter light conditions, they need to be more carefully managed (Figure 2) to stimulate brightly colored peppers with higher content of phytochemicals. On the other hand, there is a growing interest among vegetable producers to better control pest and diseases using UV-absorbing films as greenhouse material [24,41], however, UV exclusion may lead to lower concentrations of secondary metabolites in plants and deterioration of product nutritional quality [8]. Therefore, it can be hypothesized that

much stricter selection of the greenhouse covers UV blocking or transmitting properties in conjugation with the cultivated crop and production practices (e.g., crop orientation, harvesting time, planting density), would be beneficial to the growers to reduce pesticide use without a negative effect on phytochemical composition of selected crops in Mediterranean greenhouses. Improvement of product quality in soilless cultivations by manipulating nutrient solution composition has also been stated in several cases [42]. Moreover, the use of artificial light sources (e.g., UV light-emitting diodes) in the greenhouse could not be ruled out, however at the moment is of low usability in horticulture due to operating costs and law restrictions [43]. Furthermore, the data set of this study indicated that although variations in total soluble solids and titratable acidity of pepper fruits may exist at different times of the year, the sensory TSS/acid ratio remained unaffected with time. This may suggest that taste quality of peppers would probably not greatly vary among harvest months in any of the cultivars tested, which is of importance for the market value of the product but may not always coincide with the micronutritional quality of the fruits [44]. Yield results also revealed that there is always a need to validate the results of the earlier studies with the new high yielding cultivars, modern growing systems and prevailing environmental conditions. Indicatively, red cultivar showed greater yield and average fruit weight, followed by orange and yellow, with the lowest values observed in the green one. Total marketable fruit yield and mean fruit weight varied from 8.4 to 9.1 (kg/m2) and 162 to 171 (g/fruit), respectively, for colored fruited peppers. In this content, greenhouse pepper production in Spain yields about 7 kg/m2 yearly, whereas colored peppers grown in Florida yielded 6.9 to 11.3 kg/m2 in a harvesting period from October to March with an average fruit weight from 161 to 212 g/fruit [45].

#### **5. Conclusions**

This study clearly shows the challenge to eliminate fruit antioxidant phytochemical variations in yearly grown greenhouse colored pepper crops. It was clearly shown that the total antioxidant activity and major antioxidant components including phenolics, ascorbic acid, and carotenoids tend to accumulate in higher amounts in sweet pepper fruits at harvesting times with higher solar and ultraviolet radiation and elevated temperature (i.e., spring and summer). Collectively, these results indicate that in Mediterranean greenhouses during late autumn and winter, light conditions can be insufficient to stimulate brightly colored peppers with elevated content of antioxidants, thus the antioxidant activity. This further suggests that a proper selection of greenhouse type and cover material in response to plant–light interception in conjugation with the selected crop and cultivation system may be a prerequisite to optimize environmental conditions for plant growth and elevated antioxidant phytochemicals in yearly grown sweet colored peppers in Mediterranean greenhouses.

**Author Contributions:** Formal analysis, D.N. and G.N.; Investigation, D.N.; Methodology, D.N.; Writing— review & editing, D.N. and G.N.

**Funding:** This work was supported by the Agricultural Research Institute of Cyprus and authors did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*
