**About the Editors**

## **Sophia Rhizopoulou**

Sophia Rhizopoulou is a Professor Emeritus at the Department of Biology, National and Kapodistrian University of Athens, Greece. She graduated in and completed her Ph.D. at the National and Kapodistrian University of Athens. She has been a visiting scientist at Kuwait University (Kuwait), Bielefeld University (Germany), and Sabanci University (Turkey), and a Consultation Board Member of IDOSI as well as the Vice President of the Scientific Sectoral Council in agricultural production, food, agrobiotechnology, and aquaculture, National Council for Research and Innovation, etc. Her research interests include ecophysiology, environmental science, plant biology and physiology, Mediterranean plants, biodiversity, historical botany, roots and water uptake, biomimetics (...art imitates nature...), etc. See more at http://publicationslist.org/sophia.rhizopoulou.

## **Maria Karatassiou**

Maria Karatassiou is an Associate Professor at the Laboratory of Rangeland Ecology, Aristotle University of Thessaloniki, Greece. Her research interests include plant ecology and evolution, plant–water relations, plant crown architecture and hydraulic architecture, the survival of forage species under biotic or abiotic stress, etc. See more at https://www.researchgate.net/profile/ Maria-Karatassiou.

## **Efi Levizou**

Efi Levizou is an Associate Professor at the Department of Agriculture Crop Production and Rural Environment, University of Thessaly, Greece. She graduated in and completed her Ph.D. at the University of Patras, Greece. Her research interests include alternative crop nutrition systems in the context of a circular economy in addition to their effects on plants' functional characteristics and the effects of abiotic stress factors on plant physiology and production. See more at http://agr.uth.gr/ en/wp-content/uploads/2022/06/Levizou-short-CV-english June-22.pdf.

## **Preface to "Mediterranean Plants"**

Plants grown and exposed to Mediterranean climatic conditions are a source of information of a natural heritage. Mediterranean plants have been presented as vehicles for expressing historical knowledge and environmental attributes; scientific reports have given us important insights into plant growth, structure, and function. With the current environmental and social threats, mainly posed by expanding touristic and anthropogenic activities, the importance of Mediterranean plants will once again be appreciated. In this book, the function, structure, diversity, biogeography, conservation, seasonality, and interactions of Mediterranean plants with the abiotic and biotic environment are highlighted.

> **Sophia Rhizopoulou, Maria Karatassiou, and Efi Levizou** *Editors*

## *Article* **Seasonal Changes in the Plant Growth-Inhibitory Effects of Rosemary Leaves on Lettuce Seedlings**

**Kwame Sarpong Appiah 1,2,\*, Richard Ansong Omari 3,4, Siaw Onwona-Agyeman 5, Christiana Adukwei Amoatey 2, John Ofosu-Anim 6, Abderrazak Smaoui 7, Abdelkarim Ben Arfa 8, Yoko Suzuki 9, Yosei Oikawa 1, Shin Okazaki 1, Keisuke Katsura 1, Hiroko Isoda 10, Kiyokazu Kawada 10,\* and Yoshiharu Fujii <sup>1</sup>**

	- abderrazak.smaoui@gmail.com

**Abstract:** Plant biodiversity has been studied to explore allelopathic species for the sustainable management of weeds to reduce the reliance on synthetic herbicides. Rosemary (*Rosmarinus officinalis* L., syn *Salvia rosmarinus* Spenn.), was found to have plant growth-inhibitory effects, and carnosic acid was reported as an allelochemical in the plant. In this study, the effects of seasonal variation (2011–2012) on the carnosic acid concentration and phytotoxicity of rosemary leaves from two locations in Tunisia (Fahs and Matmata) were investigated. The carnosic acid concentration in rosemary leaves was determined by HPLC, and lettuce (*Lactuca sativa* L.) was used as the receptor plant in the phytotoxicity bioassay. The highest carnosic acid concentration was found in rosemary samples collected in June 2011, which also had the highest inhibitory activity. Furthermore, a significant inverse correlation (*r* = −0.529; *p* < 0.01) was found between the inhibitory activity on lettuce hypocotyl and the carnosic acid concentration in rosemary leaves. Both temperature and elevation had a significant positive correlation with carnosic acid concentration, while rainfall showed a negative correlation. The results showed that the inhibitory effects of rosemary leaf samples collected in summer was highest due to their high carnosic acid concentration. The phytotoxicity of rosemary needs to be studied over time to determine if it varies by season under field conditions.

**Keywords:** Mediterranean climate; elongation; allelochemicals; specific activity; phytotoxicity

## **1. Introduction**

Interference from weeds can have a significant impact on the growth and development of field crops, resulting in substantial crop production losses [1]. The use of synthetic herbicides to minimize crop loss due to weed infestation has become the predominant

**Citation:** Appiah, K.S.; Omari, R.A.; Onwona-Agyeman, S.; Amoatey, C.A.; Ofosu-Anim, J.; Smaoui, A.; Arfa, A.B.; Suzuki, Y.; Oikawa, Y.; Okazaki, S.; et al. Seasonal Changes in the Plant Growth-Inhibitory Effects of Rosemary Leaves on Lettuce Seedlings. *Plants* **2022**, *11*, 673. https://doi.org/10.3390/ plants11050673

Academic Editors: Sofia Rhizopoulou, Maria Karatassiou and Efi Levizou

Received: 30 January 2022 Accepted: 24 February 2022 Published: 1 March 2022

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

**Copyright:** © 2022 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 (https:// creativecommons.org/licenses/by/ 4.0/).

weed control strategy. However, the global increase in herbicide-resistant/tolerant weeds has triggered the need to diversify the existing weed control practices [2,3]. Subsequently, there has been a growing interest in the utilization of natural products in the management of weeds. Secondary metabolites produced by plants have no direct role in the basic processes of plant growth and development. After being released into the environment, some of these bioactive molecules (allelochemicals) influence the growth and development of other surrounding species, a phenomenon known as allelopathy [4–6]. These compounds can improve a plant's ability to compete in its local environment [7–9]. Allelochemicals interfere with various physiological processes of plants, including respiration, photosynthesis, and hormone balance, to affect the germination and growth of surrounding plants [10,11].

The phytotoxic effects of plant species have been explored to diversify existing weed management strategies for sustainable agriculture [12,13]. Plant extracts have been utilized to control pests [14], and isolated allelochemicals have the potential to be used in weed control or herbicide formulation [15]. However, allelopathy is a complex phenomenon since the production and release of plant secondary metabolites can be altered by environmental conditions. Seasonal changes in biotic and abiotic variables, such as pathogen presence [16], temperature [17], precipitation [18], and nutrient availability [19], can have a significant impact on the production and release of allelochemicals, which can contribute to seasonal fluctuations in plant phytotoxicity. Furthermore, soil bacteria can break down allelochemicals into less hazardous molecules or transform them into more toxic compounds [20,21]. Although the concentration of allelochemicals in plant tissues (flowers, leaves, stems, bark, and roots) might change during the growing season [22,23], most studies on the potential phytotoxicity of plants focus on a particular evaluation period during the season. However, understanding the potential phytotoxicity of plant species and gaining insight into the ecological interactions of plants with their environment necessitates the investigation of seasonal fluctuation [24].

*Rosmarinus officinalis* L. (Lamiaceae), also known as rosemary, is an evergreen shrub that grows wild in the Mediterranean region. A recent phylogenetic analysis merged the genus *Rosmarinus* with the genus *Salvia*. *Rosmarinus officinalis* is now known as *Salvia rosmarinus* [25–27]. Rosemary is an aromatic plant with needle-like leaves. The plant is now cultivated worldwide and has several reported therapeutic properties, including antidepressant [28], antiproliferative [29], and antidiabetic [30] activities. Diterpenes (such as carnosol and carnosic acid) and rosmarinic acid, both with strong antioxidant activity, have also been found in rosemary extracts [31–33]. The leaf extract of rosemary was reported to be potentially phytotoxic, and carnosic acid was reported as an allelochemical in the leaves of the plants [34]. Rosemary leaves also contain volatiles such as 1,8-cineole, which showed inhibitory effects on lettuce growth [35]. Carnosic acid has only been identified in a few plant species, all of which belong to the Lamiaceae family [36–38]. Richheimer et al. [39] reported carnosic acid concentrations in rosemary leaves between 1.7% and 3.9%. Subsequently, rosemary cultivars such as Daregal, VAU3, 4 English, Farinole, and Severn Seas were developed with higher levels of carnosic acid (4–10% on a weight basis of air-dried leaves) [40]. In addition, the concentration of carnosic acid in rosemary can also be modulated by growing conditions and the influence of genetic background. Climatic and environmental stress both affect the production of carnosic acid in rosemary [38], which further increases the importance of phytotoxic evaluation of the plant under Mediterranean climatic conditions. Although there are studies on the seasonal variation of carnosic acid concentration in rosemary, the seasonal variation in the biological activities of the plant has mainly focused on antioxidant activity [37,41].

Consequently, there is no available report on the relationship between carnosic acid concentration and the inhibitory activity of rosemary leaves. This study, therefore, aimed to investigate (i) how carnosic acid concentration in rosemary leaves changes with the season, (ii) which environmental factors play a role in this change, and (iii) whether this seasonal dependence of carnosic acid concentration is related to the inhibitory effect of leaves on lettuce seedling growth.

## **2. Materials and Methods**

*2.1. Collection of Plant Samples*

Plant samples were collected from the northern (Fahs) and southern (Matmata) parts of Tunisia (Figure 1). Matmata has an annual mean temperature of 20.6 ◦C, while Fahs has an annual mean temperature of 18.0 ◦C. The annual mean precipitation at Matmata and Fahs are 204 and 451 mm·year−<sup>1</sup> respectively. Fahs belongs to the Mediterranean or steppe climate zone, while Matmata belongs to the desert climate zone with a drier climate [42]. The vegetations of the collection sites are affected by the Mediterranean climate, which has less precipitation in the summer. Matmata is drier than Fahs throughout the year. The monthly mean temperature and monthly mean precipitation at the sampling locations over the sampling period are shown in Figure 2.

The meteorological data for the sampling locations were assessed using WorldClim 2.1 [43]. These collecting sites were chosen because they feature rosemary-dominated vegetation, allowing rosemary plants from various climate zones to be compared. Sampling was done four times a year by randomly selecting five sites from each of the two areas of Fahs and Matmata. A total of 40 rosemary plant samples were collected from individual rosemary plants from June, September, and November of 2011, as well as February 2012. Sampling was done while avoiding spring when nutrients are used for flower growth rather than leaves. The sampling locations at Matmata were 535–620 m above sea level, whereas those at Fahs were 300–430 m above sea level. The elevation was recorded using a GPS (Colorado 300, Garmin, Olathe, KS, USA). The samples used in this study were only those collected in the growing season, each from a single rosemary plant (Table 1).

**Figure 1.** Map of sampling areas in Tunisia.

**Figure 2.** The monthly mean precipitation (bar graph: gray bar is Fahs; white bar is Matmata) and monthly mean temperature (line graph: the solid line is Fahs; the dotted line is Matmata) at the two sampling locations.



UT-ARENA: the University of Tsukuba Alliance for Research on the Mediterranean and North Africa.

Each rosemary sample was given a unique ID (UT-ARENA management number) and stored at the Alliance for Research on the Mediterranean and North Africa's herbarium at the University of Tsukuba in Japan. Rosemary leaves were collected from the tops of the individuals that were the most exposed to the sun. The collected samples were air-dried in a well-ventilated room and then placed in a light-shielding bottle for storage in a cool and dark place.

## *2.2. Extraction Procedure*

The crude extracts were obtained from the air-dried rosemary leaf samples. In brief, 200 mg of air-dried rosemary leaves of each sample were accurately measured and placed into a 50 mL falcon tube containing 20 mL of solvent (80% ethanol). The leaf–solvent mixture was sonicated for 30 min at room temperature, filtered through filter paper No.1 (Advantec Toyo Roshi Kaisha, Tokyo, Japan), and centrifuged using Hitachi himac CR22N (6000 rpm, 10 min); then, the supernatants were collected. The residue was re-extracted using the same procedure as above, and the supernatants were combined and used as the working solutions.

## *2.3. Chemicals and Reagents*

Carnosic acid used in this study was purchased from Tokyo Chemical Industry (TCI, Tokyo, Japan). Formic acid and acetonitrile for analytical chromatography were purchased from Fluka, Sigma-Aldrich (Steinheim, Germany) and Fisher Scientific (Madrid, Spain), respectively. A Milli-Q system from Millipore (Bedford, MA, USA) was used to purify the water used in all the analyses.

## *2.4. High-Performance Liquid Chromatography (HPLC) Analysis*

A total of 50 mg of ground rosemary samples (leaves) was accurately weighed, put into a 50 mL falcon tube, and extracted, as described in the extraction procedure. An aliquot of the extract after centrifugation was filtered through a 0.2 μm syringe filter before the injection of 10 μL in LC-20AD liquid chromatography (Shimadzu, Japan) for the HPLC analysis. An Inertsil ODS 2 column (250 × 4.6 mm, 5 μm particles, GL Sciences Inc, Tokyo, Japan) was used. Mobile phases A and B were water with 0.1% formic acid and acetonitrile, respectively. The column temperature was kept at 30 ◦C, and the flow rate of the mobile phase was set at 0.5 mL·min<sup>−</sup>1. The following multistep gradient with different proportions of mobile phase B was applied: 0 min, 20% B; 10 min, 40% B; 15 min, 90% B maintained for 5 min. The initial conditions were maintained for 5 min. The analysis was monitored using an SPD-M20A detector at 210 nm. The quantification was done by comparing the peak areas of the targeted carnosic acid with the abundance of the compound in the corresponding standard used in the calibration curve. All chemical analyses were done in triplicate.

## *2.5. Phytotoxic Activity Bioassay*

The radicle and hypocotyl elongation of *Lactuca sativa* (Great Lake 366, Takii Co., Kyoto, Japan) was evaluated in the phytotoxic activity bioassay using ethanol crude extracts of each of the 40 samples of rosemary leaves. In the phytotoxic activity bioassay, 40 samples of ethanol crude extracts of rosemary leaves were tested on the radicle and hypocotyl elongation of *Lactuca sativa*. The concentration range of the rosemary crude extracts (0.5, 1.0, 3.0, 5.0, and 10 mg DW·mL<sup>−</sup>1) was adapted from a previous study [34]. In a 27 mm diameter glass Petri dish, a filter paper (27 mm, Toyo Roshi Kaisha, Ltd., Tokyo, Japan) was inserted. A total of 0.7 mL of test solution was added to the filter paper and dried completely in vacuo. Five lettuce seedlings (pre-germinated for 20 h) were placed on the filter paper after adding 0.7 mL of 0.05% dimethyl sulfoxide (DMSO) and incubated (CN-25C, Mitsubishi Elec., Tokyo, Japan) for 52 h at 22 ◦C in dark conditions. The control treatments were set up with no crude extract but only 0.05% DMSO. Three replications were set for each

treatment. The radicle and hypocotyl lengths were measured after the incubation period, and elongation percentages were calculated using the following equation:

$$E = A/B \times 100\tag{1}$$

where *E* is the elongation percentage, *A* is the average length of radicle/hypocotyl in the treatment, and *B* is the average length of radicle/hypocotyl in the control.

## *2.6. Statistical Analysis*

The IBM statistics tool SPSS (SPSS Inc., Chicago, IL, USA, version 21) was used to analyze the data. Data were subjected to a two-way analysis of variance (ANOVA) to determine the significant differences among the samples collected in different months and locations. The sampling months and locations were considered as the independent factors in the analysis. Mean differences among the treatments were compared using the Tukey test at *p* < 0.05. Pearson's correlation analysis was conducted to establish significant relationships among the measured parameters.

## **3. Results**

## *3.1. Variations in Carnosic Acid Concentration in Rosemary Leaves during the Growing Season*

The concentration of carnosic acid in the leaves of rosemary samples collected from the two different locations (Fahs and Matmata) in Tunisia was studied over a growing season using reversed-phase high-performance liquid chromatography (RP-HPLC) (Figure 3). The equation for the calibration curve for carnosic acid was *y* = 84051*x* + 240721, *R*<sup>2</sup> = 0.9994. The limit of detection (LOD) and limit of quantification (LOQ) were determined at signalto-noise (S/N) ratios of 3 and 10, respectively. The LOD and LOQ were 0.0150 mg·g−<sup>1</sup> and 0.0455 mg·g−1, respectively. This study focused primarily on carnosic acid, as it was previously found to be the major allelochemical responsible for the plant growth-inhibitory effect of rosemary leaves [34]. The results of this study showed that the accumulation of carnosic acid in rosemary leaves depended on the time of sampling.

**Figure 3.** Chromatograph of an ethanol extract from rosemary leaves (**a**) and synthetic carnosic acid (**b**).

The carnosic acid concentration in leaves of rosemary samples collected in Tunisia during the study period varied widely between 2.9 and 28.4 mg·g−<sup>1</sup> dry weight (Figure 4). The results showed that the highest average carnosic acid concentration (15.1 mg·g−<sup>1</sup> dry weight) was measured in June (early summer), while the lowest concentration was measured in February (8.3 mg·g−<sup>1</sup> dry weight) (Figure 5). It was observed that the concen-

tration of carnosic acid in the leaves of rosemary was higher in the samples from Matmata than in those from Fahs at all sampling times.

**Figure 4.** Distribution of carnosic acid concentration in the crude extracts of rosemary leaf samples collected from Tunisia (June 2011–February 2012). Values are the means of three replicates ± SD. CA: carnosic acid (expressed on a dry weight basis).

## *3.2. Influence of Precipitation, Elevation, and Temperature on Carnosic Acid Concentration in Rosemary Leaves*

The two sampling locations had different annual precipitation, temperature, and elevation. Matmata has a hot climate, while Fahs has a moderately hot climate. To determine which environmental factors might be related to the observed seasonal variation in carnosic acid concentration, a Pearson correlation analysis was performed based on carnosic acid concentration and environmental factors (temperature, precipitation, and altitude) during sampling (Table 2). Carnosic acid concentration showed a significant positive correlation with temperature (*r* = 0.30; *p* < 0.05) and altitude (*r* = 0.33; *p* < 0.05). However, there was a significant inverse relationship between carnosic acid concentration and precipitation at the sampling locations (Table 2). The results show that temperature and precipitation variations influence the concentration of carnosic acid in rosemary leaves during the season.


**Table 2.** Pearson correlation analysis for carnosic acid concentration, precipitation, elevation, and temperature.

CA: carnosic acid. Significance level: \* *p* < 0.05, \*\* *p* < 0.01.

## *3.3. Effects of Seasonality on the Plant Growth-Inhibitory Potential of Rosemary Leaves*

The concentration of carnosic acid in rosemary leaves showed seasonal variation and a significant relationship with precipitation and temperature. The study also investigated whether the seasonal variation in carnosic acid concentration could influence the phytotoxic activity of rosemary during the sampling season. The phytotoxic activity assay was tested on lettuce elongation. The inhibitory effect of rosemary leaf ethanol crude extracts on lettuce radicle and hypocotyl elongation was dose-dependent. The ranges of inhibition of lettuce radicle and hypocotyl elongation were 18.3–123% and 15.6–100% (percentage of control), respectively (Table S1). Lettuce hypocotyl elongation was more sensitive to rosemary crude extract than the radicle.

The concentration of rosemary leaf extracts required for 50% growth inhibition (EC50 or specific activity) of lettuce elongation was determined for all the collected rosemary samples. The inhibitory effect (expressed as EC50) on lettuce growth ranged from 2.1–8.6 mg DW·mL−<sup>1</sup> and from 0.7–7.2 mg DW·mL−<sup>1</sup> for radicle and hypocotyl, respectively (Table S1). The observed phytotoxicity of rosemary leaves on lettuce length growth showed seasonal variations. Samples collected in September and November had the lowest EC50 values (strong inhibition) for lettuce hypocotyl elongation (Figure 6a).

**Figure 6.** Effect of the sampling period and location on the growth-inhibitory activity of rosemary leaves on lettuce (**a**) hypocotyl and (**b**) radicle elongation. Different letters (a, b, location for each month; A–C, sampling month during the season) above the error bars show treatments with significant differences throughout the season (*p* < 0.05). Values are means ± SD (*n* = 5).

High EC50 values (low inhibition) for lettuce hypocotyl were measured in February at both locations, which coincided with the lowest carnosic acid concentration in rosemary leaves. The average specific activity of samples collected in November, September, June, and February on lettuce radicle elongation was 4.8, 5.0, 5.6, and 6.2 mg DW·mL<sup>−</sup>1, respectively (Figure 6b). Except for samples collected in September, there was no significant difference in inhibitory activity between sampling locations during the season (Figure 6). The effect of sampling location, sampling period, and their interaction on lettuce carnosic acid concentration and growth elongation are shown in Table 3. Except for the effect of sampling location on hypocotyl and radicle growth, all other effects and interactions were significant. However, the seasonal variation in phytotoxicity and concentration of carnosic acid in rosemary leaves should be evaluated over 1 year in a Mediterranean climate to fully understand this relationship.

**Table 3.** Summary of the analysis of variance (ANOVA) for carnosic acid concentration, growth elongations, sampling location, and period.


\* Significant at the 0.05 level of probability. \*\* Significant at the 0.01 level of probability. *p* > 0.05: not significant. CA: Carnosic acid. Growth is expressed as a percentage of the control. MS: means of squares.

## *3.4. Correlation between Carnosic Acid Concentration and Phytotoxicity of Rosemary Leaves*

To determine the relationship between carnosic acid concentration and phytotoxicity of rosemary leaf extracts, a Pearson correlation analysis was performed based on the results of carnosic acid concentration and EC50 (specific activity) of the leaves. The resulting graph represents a natural dose-response curve for carnosic acid in rosemary leaves. The correlation study showed a significant inverse relationship (*p* < 0.01; *r* = −0.529) between carnosic acid concentration and inhibitory effects (expressed as EC50 or specific activity) of rosemary leaves for hypocotyl elongation (Figure 7a). This result shows that the contribution of carnosic acid to the inhibitory effect of rosemary leaves on lettuce hypocotyl elongation is high, but low on radicle elongation. The results indicate that rosemary leaves with a high concentration of carnosic acid have great phytotoxic potential, which can be further explored. However, the degree of correlation between carnosic acid concentration and phytotoxicity of rosemary leaves indicates that other compounds may also contribute to the phytotoxicity of rosemary leaves.

**Figure 7.** Relationship between carnosic acid concentration and phytotoxicity (expressed as EC50) of the leaf extract of rosemary on lettuce (**a**) hypocotyl and (**b**) radicle elongations.

## **4. Discussion**

The RP-HPLC analysis of rosemary leaves collected from the two locations in Tunisia showed that carnosic acid concentration varied throughout the season (as shown in Figure 4). Other studies reported similar variations in carnosic acid concentration in rosemary samples from different geographical zones [44,45]. The average carnosic acid concentration was highest in early summer for both sampling locations in this study. In line with the results of this study, Hidalgo et al. [45] also reported an increase in carnosic acid concentration in rosemary leaves in summer (46.2 mg·g−<sup>1</sup> in July 1996), while the lowest values were observed in February of the same year. In Brazil, the reported carnosic acid concentration in rosemary leaves was highest in leaf samples collected in summer [46]. In contrast, Luis and Johnson [37] observed a decrease in the carnosic acid concentration of about 50% during the summer months characterized by high temperatures. The discrepancy in the concentration of carnosic acid in rosemary leaves could be due to the influence of growing conditions and other factors. The influence of environmental factors on the variation of carnosic acid concentration in rosemary leaves was reported previously [46,47]. The seasonal variations in carnosic acid concentration observed in this study may indicate that the synthesis of the compound is influenced by changes in certain climatic factors.

The results also showed a relationship between environmental conditions at the time of sampling and carnosic acid concentration in rosemary leaves. Temperature, precipitation, and elevation of sampling locations showed significant correlations with carnosic acid concentration in rosemary leaves. Similar to the results of this study, Hidalgo et al. [45] reported increasing carnosic acid concentration in rosemary leaves with increasing temperature. Lemos et al. [46] also reported the highest carnosic acid concentration in the month with the highest temperature. In contrast, Munne-Bosch et al. [48] reported a negative linear relationship between carnosic acid concentration and temperature. However, an increased amount of carnosic acid was detected during the summer with high rainfall and temperature in Brazil [46]. Borras et al. [44] reported that the observed variations in the altitude of sampling locations had significant effects on the concentration of plant metabolites (including carnosic acid) in rosemary leaves. Compared to other native Mediterranean plants, rosemary can withstand prolonged drought by avoiding damage to its photosynthetic organs [47]. Seasonal variation is associated with certain changes in soil moisture and temperature, which may lead to variations in the biosynthetic pathways of primary and secondary metabolites [17,18]. Carnosic acid was found mainly in June 2011, followed by September 2011 and November 2011. The biosynthetic pathway of terpenes could explain this observation. Terpenes are synthesized in the cytosol and plant plastids [49]. The pathway leads to the formation of sesquiterpenoids in the cytoplasm and the formation of diterpenes and tetraterpenes in the plastid. However, these processes are associated with the capture of sunlight and a photoprotective function in cell membranes [49]. Thus, according to the biosynthetic mechanisms, rosemary leaves harvested in June 2011 increased the synthesis of terpenes (including carnosic acid) in plastids at the high temperatures (26.6 ◦C). Moreover, carnosic acid is one of the most important antioxidants in rosemary leaves, and its concentration increases under stress conditions [46]. It should be considered that the production of carnosic acid in rosemary depends on the genetic background, plant part, and growing conditions [50], which could also explain part of the discrepancy between the results reported in different studies.

Although carnosic acid was reported as the principal allelochemical in rosemary leaves [34], other compounds found in the plant, such as ferulic, caffeic, gallic, chlorogenic, and rosmarinic acids, have been linked to phytotoxicity [51]. The antioxidative mechanism of carnosic acid in plants has been reported [52]; however, there has been no reported study on its mode of action as a plant growth inhibitor. Since other compounds contribute to the inhibitory effects of rosemary leaves, the physiological actions of some of these compounds are discussed. According to Araniti et al. [53], rosmarinic acid inhibited the main reactive oxygen species (ROS)-scavenging enzymes, resulting in high ROS levels that cause alterations in mitochondrial ultrastructure and function, leading to cell death in *Arabidopsis*

seedlings. Rudrappa et al. [54] asserted that gallic acid elevated the level of ROS in the roots of *Arabidopsis*. The activated ROS caused the root architecture of susceptible plants to be disrupted by impairing the microtubule assembly. According to dos Santos et al. [55], ferulic acid may be channelled into the phenylpropanoid pathway, where it may increase the quantity of lignin monomer in the cell wall, hardening the cell wall and inhibiting root growth. Similarly, caffeic acid channelled into the phenylpropanoid pathway increased lignin monomers that solidify the cell wall and inhibit root growth [56]. 1,8-Cineole, a significant essential oil in rosemary, decreased root growth in other plants by impeding DNA synthesis in the apical meristem of *Brassica campestris* roots [57]. Monoterpenes, which are abundant in rosemary, inhibited chlorophyll content, as well as the biosynthesis of several phenolic compounds [58].

The rosemary leaves sampled in this study showed variations in carnosic acid concentration, suggesting that the growth-inhibitory effect of the leaves may change over the season. The study further confirmed that the phytotoxicity of rosemary leaves changed during the sampling period. The changes in carnosic acid concentration and the expression of biological activities during different seasons have been reported in other studies [46,59,60]. Although the antimicrobial activity of rosemary leaves changed during the growing season [61], seasonal changes in the phytotoxicity of rosemary have not been reported. The concentration of carnosic acid in rosemary leaves showed a significant correlation (*r* = −0.529; *p* < 0.01) with growth inhibition at the hypocotyl of lettuce. Our results agree with other studies that showed that allelochemicals and growth inhibition are related in allelopathic species. Ben-Hammaouda et al. [62] reported that the phytotoxicity of sorghum hybrids had a positive correlation (*r* = 0.66) with the total concentration of phenolic compounds. Similarly, Reberg-Horton et al. [63] reported that the inhibitory effect of aqueous extracts of *Secale cereale* tissue correlated with the amount of DIMBOA extracted from the harvested tissue. In another study, the concentration of phenolic acids together with DIBOA and DIMBOA explained about 90% of the variation in growth inhibition observed in annual ryegrass [64]. Although a significant relationship was found between carnosic acid concentration and growth inhibition, the contribution of other compounds to rosemary leaf phytotoxicity should not be ignored.

## **5. Conclusions**

The concentration of carnosic acid in rosemary leaves and the inhibitory effect of ethanolic extracts of rosemary leaves were both influenced by seasonal variations. The carnosic acid concentration in rosemary leaves peaked in early summer at both sampling locations in Tunisia and then gradually decreased until winter. Rosemary leaf phytotoxicity (expressed as EC50) followed a similar pattern throughout the season and showed a significant (inverse) relationship with carnosic acid content. It is important to evaluate the seasonal variation in the inhibitory activity of rosemary leaves to avoid over-or underestimating the phytotoxicity of the plant. The efficacy of rosemary as a potential weed control agent needs further investigation under field conditions.

**Supplementary Materials:** The following supporting information can be downloaded at https: //www.mdpi.com/article/10.3390/plants11050673/s1: Table S1. The effects of rosemary leaf samples from different seasons and at different locations on lettuce growth elongation; Table S2. Moran's I statistic.

**Author Contributions:** Conceptualization, K.S.A. and Y.F.; methodology, K.S.A., R.A.O., A.S., A.B.A., J.O.-A., K.K. (Kiyokazu Kawada), and Y.F.; formal analysis, K.S.A., R.A.O., C.A.A., S.O.-A., and Y.F.; investigation, K.S.A.; resources, A.S., A.B.A., Y.S., H.I., and K.K. (Kiyokazu Kawada); data curation, K.S.A., Y.O., K.K. (Keisuke Katsura), C.A.A., Y.S., S.O., and R.A.O.; writing—original draft preparation, K.S.A.; writing—review and editing, K.S.A., R.A.O., K.K. (Kiyokazu Kawada), H.I., S.O.-A., J.O.-A., A.S.; A.B.A.; Y.S., S.O., Y.O., C.A.A., K.K. (Keisuke Katsura), and Y.F.; validation, K.K. (Kiyokazu Kawada), K.S.A., and Y.F.; visualization, K.S.A. and R.A.O.; supervision, Y.F.; project administration, K.K. (Kiyokazu Kawada), H.I., and Y.F.; funding acquisition, Y.F., K.K. (Kiyokazu Kawada), and H.I. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was supported by a grant-in-aid for Research on Agriculture and Food Science (25029AB) from the Ministry of Agriculture, Forestry, and Fisheries of Japan. JST CREST Grant Number JPMJCR17O2, Japan, JSPS KAKENHI Grant Number 26304024, and the Technology Research Partnership for Sustainable Development (SATREPS) of JST/JICA (JPMJSA1506) also supported this work.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data included in the article with Supplementary Materials.

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

## **References**


## *Article* **Pre-Germination Treatments at Operational Scale for Six Tree Species from the Sclerophyll Forest of Central Chile**

**Eduardo Cartes-Rodríguez 1, Carolina Álvarez-Maldini 2,\*, Manuel Acevedo 1, Marta González-Ortega 1, Alejandro Urbina-Parra <sup>1</sup> and Pedro León-Lobos 3,4**


**Abstract:** Sclerophyll forest in Mediterranean central Chile has been subjected to severe degradation due to anthropic disturbances and climate change and is in need of restoration. Since direct seeding is usually unsuccessful, we need to research seed propagation to produce plants for restoration. Our objective was to assess pre-germination treatments for six native woody species (*Acacia caven*, *Lithraea caustica*, *Quillaja Saponaria*, *Porlieria chilensis*, *Kageneckia angustifolia*, and *Ceratonia chilensis*) of the sclerophyll forest, considering its operational applicability and consequences for nursery plant production. Treatments were selected according to previous studies, and operational applicability in nurseries. Germination and level of seeds water imbibition were assessed. Results indicate that time for seed water imbibition is critical for germination in *A. caven*, *P. chilensis* and *K. angustifolia*, with an average germination of 90.2 ± 2.0%, 85.0 ± 4.7%, and 47.4 ± 2.3%, respectively. Gibberellin did not improve germination compared to water soaking in *Q. Saponaria*, *K. angustifolia* and *P. chilensis*. In addition, physical scarification is a suitable treatment for *L. caustica* and *C. chilensis*, instead of chemical scarification, avoiding handling toxic and corrosive compounds in nurseries. We recommend assessing seed water imbibition rates as a key factor for proper germination processes.

**Keywords:** Mediterranean; nursery production; seeds; water imbibition

## **1. Introduction**

There are five Mediterranean-type climate regions in the world that represent less than 5% of global surface but and are catalogued as biodiversity hotspots because they harbor 20% of endemic vascular plants of the planet [1,2]. These regions are characterized by higher levels of endemism of the vascular flora, which are adapted to seasonal water deficit and high summer temperatures [3]. One of these regions is found in central Chile, between 30◦ S and 36◦ S approximately; this represents a transition between the Atacama Desert and mixed temperate forest of south Chile, with shrub formations and sclerophyll forests that dominate the landscape [1,4,5].

The Mediterranean-type climate region of Chile has been subjected to strong anthropic pressures [6] due to changes in soil use, which has reduced the coverage, structure, and composition of this ecosystem, and this has been worsened by the effect of forest and soil degradation and bushfires [7–10]. These events have increased in severity during the last decade, as Chile has been facing a mega-drought since 2010 [11–13]. In recognition of the central role of forests for carbon storage in vegetation and soil [14–16], the United Nations (UN) has declared the 2021–2030 the "Decade of Ecosystem Restoration" as an

**Citation:** Cartes-Rodríguez, E.; Álvarez-Maldini, C.; Acevedo, M.; González-Ortega, M.; Urbina-Parra, A.; León-Lobos, P. Pre-Germination Treatments at Operational Scale for Six Tree Species from the Sclerophyll Forest of Central Chile. *Plants* **2022**, *11*, 608. https://doi.org/10.3390/ plants11050608

Academic Editors: Sofia Rhizopoulou, Maria Karatassiou and Efi Levizou

Received: 30 January 2022 Accepted: 16 February 2022 Published: 24 February 2022

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

**Copyright:** © 2022 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 (https:// creativecommons.org/licenses/by/ 4.0/).

important tool to mitigate the increasing loss of biodiversity and rise in CO2 emissions. Altamirano et al. [17] reveal the urgent need to restore forest ecosystems at a global scale, which represents the best option for the achievement of goals described by the UN. This is even more relevant for sclerophyll forests because it requires greater restoration efforts considering its actual degradation state [18].

Chile and 114 other countries have subscribed to a series of restoration commitments [19]. In the context of the Paris Agreement and the update of the Determined Contribution at National Level in 2020, Chile is committed to afforest 70,000 hectares of native species for the formation of a permanent forest cover and to restoration processes of 1,000,000 hectares of landscape [20]. However, Bannister et al. [21] identified three bottlenecks for successful forest restoration in Chile: (1) a lack of national plan for forest landscape restoration; (2), poor quality and low supply of native plants species, which was thoroughly described by Acevedo et al. [22]; and (3) poor results in the establishment phase. Likewise, León-Lobos et al. [23] describe a fourth bottleneck focused on the low availability and seed quality of native species, which impairs plant production and the achievement of restoration goals. In addition, Mediterranean species from central Chile present several issues that negatively affect seed supply such as low density and species diversity in seed banks; besides low germination after sampling from soil [24–26], this has proven to be even more severe in landscapes dominated by *Quillaja Saponaria* Molina (Quillajaceae) and *Lithraea caustica* (Mol.) Hook. and Arn. (Anacardiaceae) [25]. Besides supply problems, seeds from the sclerophyll forest are exposed to inadequate environmental conditions during seed germination and establishment. Summer season in Mediterranean climate imposes important levels of stress, due to higher temperatures and the absence of relevant precipitations [27,28]. In the context of central Chile there is little experimental evidence that native species produce seed without dormancy at the moment of seed dispersal [26]. Evidence indicates that physical dormancy imposed by a hard and impermeable seedcoat is present in native species such as *Acacia caven* (Molina) Molina (Fabaceae) and *L. caustica* [29], which can be prompted to germinate with mechanic scarification or acid application. Likewise, the development of fleshy fruits in species such as *L. caustica* and *Porlieria chilensis* can induce dormancy due to the presence of chemical inhibitors [30,31], and in natural environments germination can be triggered by the passage through the intestinal tract of frugivore species such as native fox *Pseudalopex culpaeus* (Mol.) [32]. In these cases, the removal of the pericarp, acids or gibberellin treatment can be applied at operational levels.

Despite that seed anatomy and the environmental conditions can shed light regarding dormancy in seeds of species from the sclerophyll forest, there is a lack of information regarding specific dormancy and the identification of suitable pre-germination treatments that maximize its germination [23,33,34]. Although higher germination is a pre-requisite to unlock the following bottlenecks for restoration [21], such treatments should consider their applicability by local nurseries, that in most cases lack the proper training to, for example, handle toxic chemicals for acid stratification such as sulfuric acid [22,35].

The objective of this research is to assess pre-germination treatments for six native woody species (*A. caven*, *L. caustica*, *Q. Saponaria*, *P. chilensis* I.M. Johnst., *Kageneckia angustifolia* D. Don (Rosaceae), and *Ceratonia chilensis* (Molina) Stuntz emend. Burkart (Fabaceae) of the sclerophyll forest from central Chile, considering their operational applicability and consequences for nursery plant production.

## **2. Results**

## *2.1. Seed Characterization*

From the estimation of the number of seeds per kilogram, it is observed that bigger seeds belong to *A. caven* with 7,554 ± 216 seeds kg−1, followed by *L. caustica* and *C. chilensis* with 23,066 ± 660 seeds kg−<sup>1</sup> and 24,359 ± 543 seeds kg<sup>−</sup>1, respectively, and finally *P. chilensis* with 37,169 ± 2379 seeds kg−<sup>1</sup> and *Q. saponaria* with 112,413 ± 1,396 seeds kg−1; seeds with the lower weight were observed in *K. angustifolia* with 136,935 ± 3855 seeds kg<sup>−</sup>1.

A significant effect was observed in the time of water imbibition on the moisture content of seeds for all species (all *p* < 0.0012). The seeds initial moisture content was on average 14.9 ± 7.6%, which corresponded to seed water content immediately after storage and before soaking. After soaking in water for 24 h, significant increase was observed in this variable in all evaluated species (all *p* < 0.0058), while only for *A. caven*, *P. chilensis* and *K. angustifolia* a significant increase in moisture content was observed after 48 h of soaking (all *p* < 0.0152) (Table 1).

**Table 1.** Seed moisture content (*w*/*w* dry basis, g g−1) initial and after soaking in water for 24 and 48 h of sclerophyll species of central Chile (mean ± s.d.; *n* = 3). Letters indicate significant differences in moisture content for each species, between the different measurement times (Tukey, *p* < 0.05).


## *2.2. Nurseries Survey*

Surveyed nurseries represent the 41.3% (*n* = 6) of nation-wide plant production for the selected species in this study during the 2017 to 2019 seasons, where chemical and scarification treatments present higher restrictions at an operational scale. Restrictions are mainly related to technical capabilities, due to lack of knowledge about how to apply the treatments at an operational scale, and to a lesser degree due to restrictions in infrastructure and equipment (Table 2).

**Table 2.** Restrictions identified in surveyed nurseries for application of pre-germination treatments at operation scale in evaluated species. (1): Gibberellic acid; (2) Hot water; (3) Sulfuric acid.


\* Gibberellic acid concentrations were not declared by the nurseries surveyed.

Among the most used pre-germination treatments in nurseries are soaking in water at room temperature and chemical scarification with sulfuric acid (Table 3) for the selected species. Soaking in water ranged between two and 72 h for most nurseries and species, while exposure to sulfuric acid varies between 30 and 180 min depending on the protocol of each nursery and species. Surveyed nurseries do not apply plant hormones as pregermination treatments, stratification, or mechanical scarification, for the assessed species.


**Table 3.** Pre-germination treatments applied by nurseries at operational scale, for the species evaluated in the sclerophyll forest. (1): Hot water; (2): Sulfuric acid.

## *2.3. Germination Experiment*

For all evaluated species, a significant effect was observed in the interaction between pre-germination and measurement time (MED × TRAT) (all *p* < 0.0430).

In *Q. saponaria* average germination at the end of the experiment reached 90.2 ± 2.0%; 30 days after sowing, no significant differences between pre-germination treatments were observed. After day 16 no significant increments in germination were observed (all *p* > 0.1745), with an average of 86.1 ± 1.2% without significant differences between pregermination treatments (Figure 1a).

In *L. caustica*, at the end of the germination experiment (63 days after sowing), average germination reached at 45.4 ± 5.2%, and no significant differences between pre-germination treatments were observed (all *p* = 1). Chemical scarification with sulfuric acid induced a reduction in germination times of 33 and 23 days in comparison with control treatment and physical scarification, respectively. With chemical scarification, germination did not show significant increases after day 18 of sowing (38.1 ± 2.4%) (all *p* > 0.2327), while in control and physical scarification treatments, germination remained constant at day 51 (40.6 ± 6.3%) and 46 (38.4 ± 4.3%), respectively (Figure 1b).

In *A. caven* pre-germination treatments showed significant differences in germination (all *p* < 0.0001). At the end of the evaluation period (day 53 since sowing), seed soaking in water for 48 h germination was 70.1 ± 6.8%, which was significantly higher than with physical and chemical scarification that reached an average germination of 24.3 ± 6.5%, which did not show significant differences (*p* = 1) (Figure 1c). At day 25 after sowing, we did not observe significant increments in germination in the control treatment that reached an average of 61.1 ± 4.2% (all *p* > 0.9071).

Similarly, final germination of *P chilensis* seeds showed significant differences between treatments (all *p* < 0.0001). Physical scarification caused low germination (0.3 ± 0.2%) compared to control and the application of gibberellic acid. Thus, at the end of the germination period (63 days since sowing), no significant differences were observed between control treatment and gibberellic acid application at 200 mg L−<sup>1</sup> (*p* = 1), with an average germination of 47.4 ± 2.3%. However, soaking in gibberellic acid decreased the germination period by 11 days compared to control treatment (Figure 1d).

In *K. angustifolia* no significant differences in germination were observed at the end of the experiment (30 days after sowing) between pre-germination treatments (all *p* > 0.8581), an average germination of 85.0 ± 4.7% was reached. At day 16 after sowing germination had not increased significatively in any treatments (*p* > 0.3008), with an average germination of 80.5 ± 6.0% (Figure 1e).

**Figure 1.** Germination (%) observed in *Quillaja Saponaria* (**a**), *Lithraea caustica* (**b**), *Acacia caven* (**c**), *Porlieria chilensis* (**d**), *Kageneckia angustifolia* (**e**), and *Ceratonia chilensis* (**f**) according to different pre-germination treatments. Arrows indicate the day after sowing where no further increment in germination was observed. Symbols indicate mean + s.d.

At the end of the evaluation period in *C. chilensis*, 53 days after sowing, pre-germination treatments showed a significant effect in germination (all *p* < 0.0001). Chemical and physical scarification induced a significantly higher germination (83.7 ± 11.0%) compared to the control (16.1 ± 3.0%). No significant differences were observed between chemical and physical scarification (*p* = 0.5807) (Figure 1f). In the case of chemical scarification, 11 days after sowing germination did not increase significantly (all *p* > 0.690), reaching 70.4 ± 9.4%. For physical scarification a similar pattern was observed at day 30 after sowing (all *p* > 0.3874), and 85.6 ± 2.6% of germination was observed. In the control a stable germination level of 11.1 ± 2.4% was observed at day nine after sowing (all *p* > 0.8457).

## **3. Discussion**

## *3.1. Seed Characterization*

A low water content after 48 h of imbibition was observed in *L. caustica*, *P. chilensis*, and *C. chilensis*, in relation to the other assessed species (Table 1), which can be attributed to physical dormancy, previously reported in *L. caustica* and *C. chilensis* [25]. In *P. chilensis*, despite that our results show higher water imbibition than the results reported by Cabello et al. [36], there is agreement that water imbibition reaches a peak after 48 h of soaking in water, indicating that seedcoat impermeability is not an impairment for germination in this species. Similar results regarding water imbibition were observed in *A. caven*, with increased water content after 48 h of soaking in water, which does not agree with results from previous research [25,37,38]. Specifically, Funes and Venier [38] indicate that low water imbibition and low germination in non-scarified seeds (0%) vs. scarified seeds (96.6 ± 1.6%) is evidence of physical dormancy imposed by an impermeable seed coat, which could act as protection against humidity and temperature fluctuations [39]. However, in the research of Funes and Venier [38], imbibition was evaluated only up to 24 h, while our results indicate that maximum imbibition was reached 48 h after soaking (Table 1). In addition, germination was assessed for five days whereas, according to our results, maximum germination is reached after 25 days (Figure 1c), indicating that while seed coat impermeability could be present it can be overcome with longer water imbibition and germination periods. Thus, longer imbibition times in *A. caven* could be linked to prevention of germination in sites with unpredictable or sporadic rainfall [40], a characteristic of its habitat in central Chile.

According to Hartmann and Kester [41], seed germination occurs with water content from 40% to 60%, which would indicate that the only seeds that did not reach the needed water content for germination are the ones from *C. chilensis*, and this could explain the low germination levels observed in the control treatment (Table 2, Figure 1f). In addition, desiccation tolerance of the species can also affect germination, which would be directly linked to the initial water content of seeds (Table 1). However, the seed storage behavior of species included in this study had not been experimentally evaluated. Most species from dryland environments, such as central Chile are likely to have desiccation tolerance seeds [42]. According to the Seed Information Database [43] seeds of these species should be desiccation tolerant (Orthodox).

## *3.2. Germination Experiment*

High germination in *Q. saponaria* seed is in agreement with previous reports [44,45]. The absence of significant differences between pre-germination treatments (Figure 1a) indicate that *Q. saponaria* seeds do not present dormancy, which contradicts Figueroa and Jaksic [25], who indicate that the presence of an undetermined dormancy in *Q. saponaria* seeds and a physiological dormancy as proposed by Baskin and Baskin [46] according to Donoso and Cabello [47] results. However, no information regarding seed storage conditions was displayed by authors that mentioned an undetermined dormancy in this species, which could also affect germination. In addition, soaking in water for 24 h should be enough to achieve high water content in seeds (Table 1) and promote a high germination (90.2 ± 2.0%), which is a smaller imbibition time than the 72 h reported by Benedetti et al. [48]. According to nursery survey results, 83% of nurseries a proper treatment for seed germination.

In *L. caustica*, Donoso and Cabello [47] reported a germination capacity of 59% and recommended to treat seeds with sulfuric acid for at least three hours. Similar results were reported by [49] after remotion of seeds epicarp. Our results indicate that chemical scarification with sulfuric acid accelerates the germination process; however, it does not increase germination capacity in comparison to physical scarification or water soaking (Figure 1b), treatments that become in a suitable alternative to chemical scarification. However, these last alternatives extend the germination process to 46 and 51 days, respectively. These periods should be taken into consideration by nursery managers for sowing planification activities. According to our survey, at least 80% of nurseries apply a proper pre-germination treatment in this species.

In *A. caven* our results do not agree with previous reports [37,50,51], where seeds exposure to sulfuric acid before sowing promotes high germination (between 70% to 90%) as result of physical dormancy breaking of seeds [25]. However, germination observed after physical and chemical scarification treatments suggest deterioration of the seeds, questioning the existence of physical dormancy (Figure 1c). According to nursery surveys, we observed that at least 80% of nurseries applied chemical or physical scarification (Table 3) as a substitute for chemical scarification; these treatments could be simplified by water soaking for longer periods of time, such as 48 to 72 h.

According to Cabello et al. [36], *P. chilensis* seeds do not have physical dormancy related to impermeable seed coat, which agrees with our results regarding an increased water imbibition until 48 h since sowing. Despite that Loayza et al. [52] indicated that germination in response to physical scarification depends on seed provenance, our results showed that scarification with hot water caused a decrease in germination compared to control treatment (0.3 ± 0.2% vs. 47.4 ± 2.3%, respectively) (Figure 1d), suggesting that *P. chilensis* seed coat it is not impermeable to water and imbibition in hot water should have a detrimental effect on seed viability of this species (Figure 1d). Instead, Cabello et al. [36] reported that *P. chilensis* seeds present endogenous physiological dormancy, which could be broken with 60 days of cols stratification or soaking in 400 mg L−<sup>1</sup> of gibberellic acid, reaching a germination capacity of 78.2%, higher than results obtained in this research. Although our gibberellic acid treatment was 200 mg L−<sup>1</sup> and no differences were observed in germination compared to control, gibberellic acid decreased germination times by 11 days relative to control treatment, indicating that physiological dormancy could be present in this species. Despite that gibberellic acid application seems the most appropriate pre-treatment for this species, 100% of nurseries that produce *P. chilensis* only apply soaking in water as pre-germination and no specific treatment to break physiological dormancy is considered (Table 3).

Results observed in *K. angustifolia* regarding germination capacity agree with Takayashiki et al. [53], although the authors indicate a soaking in water for four days as pre-treatment, while according to our results 48 h in water is enough to achieve a water content of 105.4 ± 8.9% and to promote a germination of 85.0 ± 6.0%. These last results are consistent with several authors [54–56] and the surveyed nurseries that achieved germination between 70% to 80% without pre-germination treatments in direct sowing, indicating that *K. angustifolia* seeds are not dormant.

Scarification has been reported by several authors [57–60], as a method to break physical dormancy in *Prosopis* species seeds. In fact, it had been reported that passage through a digestive tract of frugivores and cattle induce germination by promoting seed coat rupture [61,62]. However, Vilela and Ravetta [60] indicated that chemical scarification for 15 min reduced germination *C. chilensis*, while physical scarification (dipped in boiling water until water reached room temperature) induced higher germination. In *Prosopis ferox*, a similar species, Ortega et al. [59] obtained higher germination after physical scarification (93.0 ± 0.03%) and chemical scarification (91.0 ± 0.02%) compare with hydrochloric acid (14.0 ± 0.02%). Our results agree with the statement that *C. chilensis* germination is promoted by chemical (with sulfuric acid) or physical scarification (Figure 1f). Physical and chemical scarification did not caused differences in germination, but affected the time when germination reached a stable value (30 vs. 9 days since sowing, respectively) (Figure 1f). Although 66% of surveyed nurseries apply a proper scarification treatment (physical or chemical, 33% each), the time needed to complete the germination process is a factor that should be considered at a large operational scale.

## *3.3. Operational Applicability of Pre-Germination Treatments*

Among the main problems mentioned by nurseries, we highlight the lack of technical capabilities (Table 3); there is a knowledge breach regarding to preparation, application, and manipulation of some chemical products for the implementation of treatments. This lack of information could be amended through instances of training and technological transference instances. This agrees with the diagnosis reported by León-Lobos et al. [23] where lack of knowledge regarding the dormancy of breaking in seeds of several native species is identified as a bottleneck for the fulfilment of Chile restoration commitments.

In regard to chemical scarification, its operational implementation (Table 3) is limited in some nurseries by the manipulation of corrosive chemical such as sulfuric acid, and managers indicate concern regarding the risk to staff safety and chemical residue disposal [35]. On the other hand, physical scarification application is limited due to technical restrictions and lack of infrastructure (Table 3), in particular the need for equipment to process large numbers of seeds at operation scale, technology that is not widely distributed in Chilean nurseries.

## **4. Materials and Methods**

## *4.1. Species Selection and Locations for Seed Collection*

Six tree species from the sclerophyll forest were selected between Valparaíso and Biobío regions, three were selected according to dominance and three according to the degree of ecological vulnerability (for a full description of the species see Table S1). Vulnerability was referred to the conservation state according to the Classification Regulation of the Species from the Environmental Ministry of Chile [63]. Regarding the dominance criteria *Q. saponaria*, *L. caustica* and *A. caven* were selected, while *P. chilensis*, (vulnerable), *K. angustifolia* (near threatened) and *C. chilensis* (vulnerable) were selected according to conservation criteria.

Seed collection was performed between January and March of 2020 in the populations indicated in Figure 2. Seeds were sampled from at least 10 trees for each species with a minimal distance of 15 m between each tree. Once collected, seeds were transferred to the Centro Tecnológico de la Planta Forestal from the Instituto Forestal (36◦50.9 S; 73◦7.9 W), Biobío region, Chile, for cleaning and storage at 4 ◦C until mid-May of 2020.

**Figure 2.** Distribution of sclerophyll forest (in grey) in Mediterranean central Chile and seed collection locations for *Acacia caven*, *Quillaja saponaria*, *Lithraea caustica*, *Porlieria chilensis*, *Kageneckia angustifolia* and *Ceratonia chilensis*.

## *4.2. Seed Characterization*

As part of seed characterization prior to pre-germination treatment application, the number of seeds per kilogram, the initial water content of seeds, and the increase in water seed imbibition was recorded, seed weight was recorded in three samples of 100 seeds (replicates) for each species to estimate the number of seeds per kilogram. Then, each sample was divided in two sub-samples of 50 seeds each, which were used to determine initial moisture content (*w*/*w*, dry basis), and after 24 and 48 h of soaking in distilled water. For this, one seed sub-sample of each species was weighted before soaking and after 24 and 48 h of soaking, while the other sub-sample was oven dried in a forced ventilation oven (Binder, model FD115, Tuttlingen, Germany) at 105 ◦C until constant weight. Weight of fresh and dried seeds were recorded on a 0.001 g precision scale (Quimis, Q DH-203, São Paulo, Brazil).

## *4.3. Pre-Germination Treatments*

Three pre-germination treatments were evaluated for each species, which were selected according to available information from previous experiments [36,37,44,47,50,53,64–70] and information obtained from a survey performed to six nurseries to consider pre-germination treatments that were feasible to apply at operational scale at that were commonly applied in nurseries (Table S2). Selected nurseries for the survey produced a larger amount of the native species selected for this research between 2017 and 2019 nationally [71].

The survey identified the pre-germination treatments operationally applied for each evaluated species, and restrictions for proper treatment application linked to technical capabilities, infrastructure, or equipment. In addition, knowledge gaps regarding the benefits of the application of pre-germination treatments was assessed.

For *L. caustica*, *A. caven*, and *C. chilensis* seeds with reported physical dormancy [25] two scarification treatments were applied: (1) chemical scarification, consisted of exposure of seeds to sulfuric acid (PQM Fermont, Monterrey, Mexico) at a concentration of 97.3% for 90 min, then seeds were rinsed with distilled water and soaked in water at room temperature for 48 h; (2) physical scarification, consisted of exposure of seeds to water at 80 ◦C, seeds were cooled until room temperature and soaked in water for 48 h. For *Q. Saponaria* and *K. angustifolia*, pre-germination treatments consisted of the use of gibberellin to break physiological dormancy, seeds were soaked in gibberellin at 200 mg L−<sup>1</sup> and 600 mg L−<sup>1</sup> (GA3, Giberplus, Anasac, Santiago, Chile) for 48 h. In the case of *P. chilensis*, treatments were the use of gibberellin at 200 mg L−<sup>1</sup> and physical scarification as previously described. For all species, seed soaking in distilled water for 48 h corresponded to control treatment.

## *4.4. Germination Experiment*

The germination experiment was performed in the Centro Tecnológico de la Planta Forestal of the Instituto Forestal, in greenhouse conditions (36 m2). Maximum air temperature was limited to 25 ◦C through a forced ventilation system. A photoperiod of 12 day and 12 dark was established with six halide lamps of 400 Watts each (Philips Master HPI-Plus, Brussels, Belgium). To characterize environmental conditions during germination, air temperature (◦C) and relative humidity (%) was measured with an Atmos 14 sensor (METER Group Inc., Pullman, WA, USA) and for substrate temperature (◦C) a RT-1 (METER Group Inc.) sensor was used. Environmental data were recorded every 30 min with a ZL6 datalogger (METER Group Inc.). During the germination experiment an average air temperature of 19.3 ± 4.4 ◦C was observed, with a daily thermal oscillation of 10.2 ± 4.5 ◦C. Minimum and maximum daily substrate temperatures were 15.1 ± 3.0 ◦C and 22.4 ± 2.7 ◦C, respectively. Air relative humidity ranged between 35% and 84%, with an average of 63 ± 9%.

Sowing for six species was performed in June of 2020 in expanded polystyrene trays of 0.13 L (15 cm depth) and 84 cavities (336 cavities m<sup>−</sup>2). Three seeds were sowed in each cavity at 1 cm depth approximately, three trays (replicates) were sowed for each germination treatment. Composted *Pinus radiata* bark was used as substrate with particles smaller than 10 mm (pH = 5.3 ± 0.01; organic matter = 76.1 ± 2.8%; total nitrogen = 0.9 ± 0.1%; C/N relation = 48.5 ± 6.4; N-NO3 = 233.6 ± 37.3 mg kg<sup>−</sup>1; N-NH4 = 662.6 ± 56.3 mg kg<sup>−</sup>1; water retention = 0.45 m<sup>3</sup> m−3). Irrigation was performed with watering cans once a day, maintaining high humidity in the surface of the substrate.

Germination was measured three times per week (Monday, Wednesday and Friday), the number of germinated seeds was recorded out of the total sowed seeds (84 cavities × 3 seed cavities<sup>−</sup>1) for each of the three trays (replicates), species and pregermination treatment. Since the sowing, germination was monitored for 30 days in *Q. Saponaria* and *K. angustifolia*, 53 days in *A. caven* and *C. chilensis*, and 63 days in *L. caustica* and *P. chilensis*. Occurrence of germination was considered when the epicotyl emerged from the substrate surface.

## *4.5. Data Analysis*

The average values of number of seeds per kilogram was calculated for each species from the weight of 100 seeds. The analysis of relative water content of seeds was performed through a repeated measurement analysis of one way for each species, considering the time for seed water imbibition time as factor (MED). Environmental data collection allowed the estimation of average temperature, minimum and maximum daily temperatures in air and substrate, and average relative humidity during the germination period.

For the germination, the experimental design corresponded to a completely randomized design with three pre-germination treatments (TRAT) for each species and with three replicates for each treatment. For each species, a germination analysis was performed through a repeated measures analysis of two ways for measurement time (MED) and pre-germination treatment (TRAT), assessing the main effects and interactions.

Repeated measure analysis was performed through a generalized mixed model using PROC GLIMMIX (SAS Institute, Cary, NC, USA) with selection of distribution and structure of the variance-covariance residual considering the Akaike Information Criteria (normal distribution and unstructured variance and covariance matrix for every analysis). Multiple comparison tests were performed for significant effects according to Tukey.

The time during germination with no further significant differences in the proportion of germinated seeds was evaluated performing comparison tests considering significant effects in the variance model (MED × TRAT).

Graphs development for data visualization were designed with SigmaPlot 10.0 software (Systat Software Inc., Chicago, IL, USA).

## **5. Conclusions**

Although several nurseries apply the proper pre-germination treatment in some species such as *Q. saponaria*, *P. chilensis* and *K. angustifolia*, there is room for improvement in applied treatments in the rest of the species. For example, in *A. caven* the water imbibition time should be considered to improve germination before the application of other pregermination treatments. In addition, in *L. caustica* and *C. chilensis* chemical scarification could be replaced by physical scarification to avoid the issues linked to manipulation of chemicals.

The evaluation of the rate of water imbibition is needed for the evaluated species before the implementation of required pre-germination treatments at an operational scale to avoid seed germination decay.

Physical scarification with hot water (80 ◦C) is a proper alternative to chemical scarification. However, higher germination times should be considered in sowing calendarization at a large operational scale.

Extension and technical transference instances could be helpful to reduce the breach in knowledge indicated by national nurseries. This could help to achieve an optimal implementation of pre-germination treatments and to avoid the application of incorrect treatments such as in *A. caven*, *P. chilensis*, and *C. chilensis*, which could generate high losses in seed supply. These actions tackle the bottleneck related to the lack of a proper seed supply to achieve reforestation commitments in Chile.

**Supplementary Materials:** The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/plants11050608/s1, Table S1: Selected species detailed description. Table S2: Pre-germination treatments applied by surveyed nurseries y reported by previous research for each of the evaluated species. N.R: Not reported.

**Author Contributions:** Conceptualization and methodology: E.C.-R., M.A. and M.G.-O. Data collection and germination measurements: A.U.-P. Data analysis: E.C.-R. Writing of the manuscript: E.C.-R., C.Á.-M., M.A., M.G.-O. and P.L.-L. Manuscript review and editing: E.C.-R., C.Á.-M. and P.L.-L. Translation: C.Á.-M. All authors have read and agreed to the published version of the manuscript.

**Funding:** Research was funded Corporación Nacional Forestal (CONAF), through the "Fondo de Investigación del Bosque Nativo" research grant, number 025/2019.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The datasets generated for this study are available on request to corresponding author.

**Acknowledgments:** Authors would like to thank the Chilean Agricultural Ministry for supporting this research.

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

## **References**


## *Article*
