Biomass Production and Quality of Twelve Fast-Growing Tree Taxa in Short Rotation under Mediterranean Climate

Abstract

Sustainable production of lignocellulosic biomass for energy use can contribute to climate change mitigation. This work aims to compare the biomass production of twelve fast-growing woody taxa belonging to the Eucalyptus, Casuarina, Populus and Paulownia genera, the quality of their biomass for energy use and its valorizing through transformation into pellets, as well as the effect of the crop on the soil. Over the course of two rotations, plant growth and biomass production were assessed. The yield of aboveground dry biomass ranged from 9 to 61 Mg ha⁻¹ year⁻¹ (equivalent to 137–867 GJ ha⁻¹ year⁻¹). The highest yields were obtained for Eucalyptus clones (51–61 Mg ha⁻¹ year⁻¹). The N-fixing species Casuarina equisetifolia and two Populus × euramericana clones (‘Adige’ and ‘AF2’) also achieved high yields (28–33 Mg ha⁻¹ year⁻¹), though significantly smaller than those of the eucalypts. Due to its low wood density, Paulownia fortunei was not very productive in terms of biomass (18 Mg ha⁻¹ year⁻¹), despite its good growth in diameter and height. However, some management practices, such as not removing nutrient-rich and poor-quality biomass fractions for energy use (leaves and thin branches) from the harvested crop, as well as the use of N-fixing species and by taking into account all nutrient inputs and outputs, can ensure the sustainability of the cultivation systems and improve degraded soils.

1. Introduction

Biomass, as a renewable and clean energy source, has aroused growing interest during the first decades of the 21st century, emphasized by the objectives of reducing CO₂ emissions in the short and medium term set out in international commitments [1,2,3,4]. Furthermore, the inevitable depletion of fossil fuels, along with their negative impact on the environment, has seen an increase in all types of renewables to the point that they accounted for 19.3% of the total global energy production in 2015 [5]. In the same year, global energy production from biomass accounted for 68.4% of all renewable sources [6]. While the primary use of biomass, mainly lignocellulosic biomass, in developing countries remains that of cooking, developed countries are significantly increasing its use in the production of thermal and electric energy due to the value-added products derived from it (pellets, chips, biogas, etc.) [7,8,9], apart from other industrial uses (wood pulp, wood-based panels, etc.) [10].

Ensuring that the growth in demand for lignocellulosic biomass can be met without adversely affecting agroforestry systems requires a regular, manageable and renewable supply of fast-growing tree species [11,12,13]. This supply is not only compatible with the cultivation of food crops but also contributes to the preservation of the environment and biodiversity [14,15] and rural development [16,17], as well as providing an alternative use for abandoned degraded land. Only in the European Union, 147 million hectares are degraded for one reason or another (soil erosion, low organic matter content, excess inorganic nitrogen fertilization). Many are in danger of abandonment, but many are suitable for establishing short-rotation tree plantations [18,19]. It is worth remembering that one of the first conclusions of the recent study commissioned by the Committee on Agriculture of the European Parliament, entitled “The challenge of land abandonment after 2020 and options for mitigating measures” (https://bit.ly/39ElcFJ, accessed on 20 March 2023), is that about 30% (56 million hectares) of the valuable agricultural area of the European Union run a moderate and high risk of being abandoned (5 million hectares in the short term, before 2030).

Under the Mediterranean climate, woody tree species have been cultivated over the last few decades. Therefore, studies have been carried out on various taxa, such as short rotation forestry crops, mainly belonging to the genera Populus and Eucalyptus but also Salix, Paulownia, Robinia, Ulmus, etc. [16,20,21,22,23,24,25]. However, most of them consider taxa (provenances/clones) of a single species, and very few compare taxa of different species growing at the same site [17,26,27]. Traditionally, plantations of these fast-growing species have been principally intended for wood production or paper pulp and, under optimal conditions, can produce up to 25 t ha⁻¹ year⁻¹ dry matter (poplars, willows) and 40 t ha⁻¹ year⁻¹ (Eucalyptus and Paulownia) [16,17,21,26,28]. However, under sub-optimal conditions exacerbated by climate change, environmental and cultivation factors (soil fertility, drought, frosts, irrigation, etc.) differentially affect each taxon [17,29]. In this respect, N-fixing species such as Casuarina sp. [28] can help to reduce the energy and economic costs of supplying nitrogenous inorganic fertilizers [30]. Thus, it is not clear that currently cultivated taxa are optimal in terms of both their yield and physical–chemical characteristics regarding energy use, and it is worth studying them on degraded soils [9,31,32].

The low energy density of biomass is the main drawback to its efficient energy use caused by the high moisture content, which reduces its heating value, and the low density, which entails handling large volumes, complicating transport and storage [33]. Pelletization presents clear advantages over other biomass fuels that have not been densified since it reduces transportation costs and increases energy conversion efficiency thanks to the low moisture content (less than 10%) [34,35,36]. Debarked softwood, mainly from pine, is one of the most used for the manufacture of pellets due to its wide worldwide distribution, as well as its low ash content and high heating value, higher than that of hardwood species and of non-woody species. Still, many other species are used with similar or slightly lower values [37].

Hence, this paper presents a study into the suitability of twelve tree taxa for biomass production in southwest Europe under a Mediterranean climate with mild winters. Eucalyptus, a genus widely used in forest plantations in the study area, along with other genera, namely Populus, Casuarina and Paulownia, for which fewer data are available in these climatic conditions, were evaluated. The objectives of this study were to assess: (i) plant growth and biomass production of twelve selected taxa belonging to Eucalyptus, Populus, Casuarina and Paulownia genera; (ii) the suitability of their biomass for energy use, including its transformation into commercial-quality pellets; and (iii) the changes of the upper soil layer after a cropping period of eight years.

2. Materials and Methods

2.1. Study Area and Experimental Design

The study site was located in the province of Huelva (Spain) UTM, zone 29S, X: 685,016, Y: 4,119,147, 10 m asl., on a permeable alkaline soil (pH = 8.3) with low salinity (EC = 135 μS cm⁻¹), 1.56 kg dm⁻³ bulk density and 18% stoniness. It was a nutrient-poor soil except for Ca and Mg, with low organic matter content and free of active limestone or heavy metal contamination (1st year in

Table 1. Physical–chemical properties of the upper soil layers at the beginning and the end of the study period for all taxa as a whole (mean (SE)). See also Table S1. Asterisks (*) denote significant differences between measurement dates within a soil layer (p < 0.05).

Variable	Soil Layer (Depth Range)
	0–15 cm		15–30 cm
	1st Year	8th Year	1st Year	8th Year
pH (H2O, 1:2.5)	8.32 (0.05)	7.98 (0.04) *	8.34 (0.05)	8.05 (0.05) *
Organic Matter (%)	1.22 (0.02)	1.63 (0.06)	0.70 (0.05)	1.17 (0.06)
C/N ratio	4.08 (0.08)	6.81 (0.40)	3.02 (0.04)	6.14 (0.40)
N (%)	0.17 (0.01)	0.15 (0.01) *	0.14 (0.01)	0.11 (0.01) *
Available P (mg kg−1) (1)	8.59 (0.12)	5.40 (0.28) *	7.85 (0.34)	4.50 (0.29) *
Available K (meq/100 g)	0.17 (0.01)	0.16 (0.02)	0.12 (0.01)	0.12 (0.01)
Available Mg (meq/100 g)	1.97 (0.05)	1.51 (0.04) *	1.81 (0.06)	1.54 (0.04) *

Differences between measurement dates for organic matter and C/N ratio were not significant (p = 0.086 and 0.059, respectively) but marked an increasing trend. The 4th year has not been shown in the table since it did not differ significantly from the 1st or 8th year and presented intermediate values between both extreme dates. ⁽¹⁾ The total P averaged 140.2, 88.3, 132.2 and 72.8 mg kg⁻¹, respectively.

, Tables S1 and S2; the “S” indicates information in the Supplementary Material section). The water table was less than 2 m deep during the wet season but more than 3 m during the summer. The area had been cultivated with woody legumes eight years prior to plantation establishment. The climate is Mediterranean, marked by mild winters due to oceanic influence and hot and dry summers, with an average annual temperature of 17.4 °C and an average annual precipitation of 558 mm. More details of the climatic variables during the study period are shown in Table S3.

Without going too deeply into details, the main characteristics taken into account for selecting the taxa for this study were rapid growth rate and high biomass yield; adaptability to the edaphoclimatic conditions of the plantation area; regrowth vigor from the stump after tree felling; and tree structure (trunk and branching) suitable for harvest mechanization. After a previous study, 12 taxa were selected (

Table 2. Woody plant taxa (species/clones) used in the present study and some examples of studies that refer to the plant growth and biomass production of the selected species.

Species	Hybrid	Seedling or ‘Clone Name’	Cited
Casuarina cunninghamiana		Seedlings	[38,39]
Casuarina equisetifolia		Seedlings	[28,40]
Eucalyptus camaldulensis		‘ENCE’	[41,42,43]
Eucalyptus × trabutii	E. botryoides × E. camaldulensis	‘Biopoplar’	[44,45,46]
Eucalyptus × urograndis	E. urophylla × E. grandis	‘ENCE’	[16,47,48]
Eucalyptus globulus		‘ENCE’	[29,44,46,49]
Paulownia fortunei		‘UHU’	[17,21,50]
Populus × euramericana	P. deltoides × P. nigra	‘Adige’	[17]
Populus × euramericana	P. deltoides × P. nigra	‘I-214’	[17,20,26,51,52,53]
Populus × euramericana	P. deltoides × P. nigra	‘AF2’	[17,53,54,55]
Populus × euramericana	P. deltoides × P. nigra	‘Oudenberg’	[26,56]
Populus × interamericana	P. trichocarpa × P. deltoides	‘Raspalje’	[17,57,58,59]

Species: Casuarina cunninghamiana Miq.; Casuarina equisetifolia L.; Eucalyptus camaldulensis Dehnh.; Eucalyptus × trabutii Vilm. ex Trab.; Eucalyptus globulus Labill.; Paulownia fortunei (Seem.) Hemsl.; Populus × euramericana (Dode) Guinier.; Populus × interamericana van Broekhuizen.; Eucalyptus urophylla S.T. Blake; Eucalyptus grandis W. Hill ex Maiden; Eucalyptus botryoides Sm.; Populus nigra L.; Populus deltoides W. Bartram ex Marshall; Populus trichocarpa Torr. and A. Gray ex Hook.

). Some have already been tested in geographical areas similar to this study. Others have been less frequently used in short rotation crops but, due to their characteristics, were considered likely to be a productive source of biomass in a Mediterranean climate [16,17,21,26,28,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59]. Because of their evergreen condition, Eucalyptus and Casuarina can benefit from the mild winter in the area [16,28]. Eucalyptus globulus is the most cultivated species in the area, therefore, well adapted. However, Eucalyptus camaldulensis and Eucalyptus trabutii tolerate drought and clay soils better than the former. Eucalyptus × urograndis presents very high growth rates but needs high soil moisture content [16]. Casuarina sp. are N-fixing species and contribute to soil nitrogen fertilization [60]. Poplar plantations are widespread in the Mediterranean area [27], mainly with I-214, an old-known clone usually used as a reference in many studies [52]. The other selected poplar clones show advantages over I-214 under certain growth conditions [26,53,55].

The experimental unit was composed of three parallel lines of five plants each, with a separation of 2 m between lines and 1 m between plants in the line (5000 plants per hectare) (Figure S1). This high plantation density was selected because some field studies indicated that high planting densities increase the biomass production of the system and shorten the harvest rotation period while decreasing the percentage of bark on the trunks [22,26,27,51,52,53,54,55,56,61,62,63,64]. A randomized complete block design was established, with four blocks (four experimental units per taxon, one experimental unit per block). This field trial configuration was replicated twice (5 plants × 3 lines × 4 experimental units × 2 replicates = 120 plants per taxon). An additional line of plants was planted around each replicate, to avoid the border effect, but not inside or between experimental units (Figure S1). So, each replica occupied an area of approximately 1800 m², to which another 800 m² were added. These correspond to the plants grown in similar conditions that were used for the allometric equations described below.

Planting took place in mid-spring 2011, according to the availability of the plants (Figure S2a). Plants of 25 to 40 cm in height coming from rooted shoot cuttings were used in the case of Eucalyptus; 15–25 cm tall plants coming from root cuttings in the case of Paulownia; hardwood cuttings of 20–25 cm in length for Populus; and 20–30 cm tall seedlings for Casuarina. Before planting, a deep plowing at 50 cm followed by a shallow tillage (harrowing) and pre-emergent herbicide (Oxifluorfen 24%, 2.5 L ha⁻¹) were applied. Fertilizers were supplied by fertigation: 75, 62.3, 21.8 and 16.0 kg ha⁻¹ year⁻¹ of N, K, P and S, respectively; 3 kg ha⁻¹ year⁻¹ of Fe–Zn–Mn (4.5%–1.5%–0.5%) fertilizer chelated by EDDHA; 75 g ha⁻¹ year⁻¹ of Cu (14%) chelated by EDTA; and 75 and 5 g ha⁻¹ year⁻¹ of H₃BO₃ and (NH₄)₂MoO₄, respectively. Support irrigation was provided every year throughout the summer (from May to September) by means of drippers along the lines, providing between 245 and 550 mm per year, according to the amount of precipitation recorded up to the time of the start, to total 800 to 900 mm annually (rainfall + irrigation). The irrigation water contained 4.6, 0.11 and 0.14 mg L⁻¹, respectively, of N, P and K. Weeding was carried out mechanically three months after planting. Still, it was not necessary to repeat it thereafter due to the high planting density and the rapid growth of the new sprouts after harvesting. It was unnecessary to apply phytosanitary treatments, although some poplar clones (‘I-214’ and ‘Raspalje’) suffered slight attacks from Melampsora rust and aphids.

2.2. Plant Growth and Biomass Estimation

Throughout the three-year duration of each of the first two rotation periods, measurements of the stem diameter (D, at 10 cm above the ground surface to a precision of 0.1 mm in two perpendicular directions), plant height (H) and the number of main stems (Nst) were taken, along with mortality rate and phytosanitary status. The plants usually had a single main stem during the first rotation (Figure S2c–e), but several re-sprouted shoots (2 to 5) from the stump after felling (2nd rotation) (Figure S2f–h). These measurements were carried out at the end of each season (4 times per year) on 24 plants per taxon each time (3 plants/taxon × 4 experimental units × 2 replicates = 24 plants/taxon). In order to avoid interference in growth by neighboring taxa, only 3 plants of each experimental unit (2nd, 3rd and 4th plants) were measured, corresponding to the central line and avoiding the two plants at the ends of said line (1st and 5th plants). At the end of each rotation, all the trees were cut down, and all the aboveground biomass (leaves, trunks and branches) was removed from the field trial. Some of these felled trees were used to elaborate the allometric equations described below. A third rotation in 2020 was planned, but it was not measured due to the COVID-19 pandemic, and consequently, the plantation was left unattended. For this reason, only data from two rotations are shown.

The biomass production assessment was carried out every year at the end of winter (February to March), before spring bud break, based on the data of D and H. For this, allometric equations relating D and/or H to shoot dry mass were previously developed [65,66,67]. Sixteen different equations were tested (including linear, power, exponential, polynomial, logarithmic and quadratic) with combinations of D, H, and D² as predictor variables, similar to those performed by other authors [16,17,23,25,40,56,68]. Most of them resulted in a high degree of prediction and fitted well to the measurement data (Tables S4 and S5, [69,70]). However, as the measurement of tree height in high-density stands is less accurate than the diameter of the main stem, for practical simplicity, the potential model was chosen in which only the diameter (D) is used as an independent variable, Dry mass = aD^b (

Table 3. Relationship between stem diameter at 10 cm above the ground surface (D, cm) and aboveground dry mass (AGDM, kg). AGDM = a D^b. D ranged from 5 to 160 mm. Significance level, p < 0.001 for all cases. n = sample size.

Taxon	a	b	R2	n
Casuarina cunninghamiana	0.031	2.586	0.985	12
Casuarina equisetifolia	0.026	2.695	0.992	12
Eucalyptus camaldulensis ‘ENCE’	0.027	2.670	0.919	12
Eucalyptus × trabutii ‘Biopoplar’	0.025	2.704	0.981	13
Eucalyptus × urograndis ‘ENCE’	0.043	2.495	0.969	15
Eucalyptus globulus ‘ENCE’	0.031	2.597	0.985	20
Paulownia fortunei ‘UHU’	0.006	3.143	0.971	14
Populus × euramericana ‘Adige’	0.033	2.704	0.979	18
Populus × euramericana ‘AF2’	0.024	2.721	0.971	22
Populus × euramericana ‘I-214’	0.044	2.448	0.971	15
Populus × euramericana ‘Oudenberg’	0.036	2.719	0.986	18
Populus × interamericana ‘Raspalje’	0.039	2.637	0.980	15

). At least 12 plants per taxon, with a stem diameter (D) ranging from 5 to 160 mm, were cut 10 cm from the ground, oven-dried at 80 °C to constant weight and weighted. They included the full range of sizes obtained in this study. Some plants harvested at the end of the 1st and 2nd rotations were used for this. In addition, other younger plants with a smaller diameter were grown next to the experimental plot under the same growth conditions (Figure S1). Any attempt to apply these models to areas different from the study should first be trialled. Each felled tree was divided into six different fractions: one for leaves and five diameter-based classes for the aboveground woody parts, that is, branches and stems (d < 25 mm, 25 mm < d < 50 mm, 50 mm < d < 75 mm, 75 mm < d < 100 mm, d > 100 mm; d being the diameter of the branch or stem). Each fraction was separately weighed. Samples of each class were taken and oven-dried at 80 °C to constant weight in order to determine their moisture content and dry weight. Likewise, due to its effect on biomass quality (i.e., on ash and nutrients content, etc.), the proportion of bark and wood for each diameter class was determined on 5 cm thick slices per taxon and diametric class.

2.3. Physical–Chemical Characterization of Plant and Soil Material

The soil sampling took place in March 2011, 2014, and 2018 (1st, 4th and 8th year, respectively) under each of the taxa, but only for one of the two replicates of the study. Two soil subsamples per experimental unit and soil depth (0–15 and 15–30 cm depth) were taken with a soil core at an equidistant distance (0.5 m) from the planting lines (2 subsamples × 4 experimental units = 8 subsamples per taxon and depth). Previously, at each sampling point, the litterfall (Figure S2b) contained in a soil surface of 0.25 m² was sampled and subsequently oven-dried at 80 °C, weighed and stored (15–20 °C) for further analysis. All the subsamples belonging to each taxon and depth were pooled. The soil samples were air-dried at 15–20 °C and sieved (2 mm mesh). Regarding the aboveground plant biomass, eight trees per taxon (one tree per experimental unit, considering the two replicates) were randomly sampled at the end of the 2nd rotation, pooled, oven-dried at 80 °C, ground and stored at 15–20 °C for further physical-chemical characterization. All samples (litterfall, soil, plant biomass) were processed by standardized methods (Table S2). Although the physicochemical properties of biomass allow for evaluating its suitability for energy use, densification into pellets would increase its commercial value. For this reason, pellets were manufactured following the methodology described by Alesso et al. [17]. Briefly, it was ground (Woodstock 3PH, Smartec^®, Villar San Costanzo, Italy), sieved to a 0.2–5.0 mm particle size and densified with a pelleting press (PLT-400, Smartec^®, Villar San Costanzo, Italy) at 95–105 °C using a raw material with 14% moisture content. The die channels were six millimeters in diameter and 24 mm in length for Populus and Paulownia but 20 mm for Eucalyptus and Casuarina. After that, their physical, mechanical and chemical properties were assessed according to the standard ISO 17225-2:2021 [71].

2.4. Data Analysis

All statistical analysis was carried out in the Statistical Package for Social Sciences (SPSS^® 20.0, IBM^®). The block effect and the replicate were not statistically significant for any parameter, so they were excluded from the models in order to simplify them. The growth (D, H, Nst) and biomass yield were evaluated in the same plants during the 1st and 2nd rotations, so repeated measures ANOVA was applied [72] by testing the assumptions of normality and sphericity in advance. Hence, the plant (within-Experimental unit, Eu) was considered a random effect, and the taxon (T) and annual growth (Year) as fixed effects. However, each rotation was analyzed separately, as the trees were felled at the end of the first rotation, and new shoots emerged. The full model included the interactions between the main factors:

y_ijkl = Eu_i + Y_j + T_k + (Eu × Y)_ij + (Eu × T)_ik + (Y × T)_jk + (Eu × Y × T)_ijk + ε_ijkl

Soil characteristics were analyzed by a three-way ANOVA by considering taxon (T), soil layer (Sl: 0–15 cm and 15–30 cm) and sampling date (Year: 2011, 2014 and 2018) as fixed effects. Normality and homoscedasticity were tested in advance. It was not appropriate to apply repeated measures ANOVA because the sampling points were not the same on all three dates.

y_ijkl = μ +T_i + Sl_j + Y_k + (T × Sl)_ij + (T × Y)_ik + (Sl × Y)_jk + (T × Sl × Y)_ijk + ε_ijkl

A two-way ANOVA was performed to determine the physical-chemical properties of biomass and litterfall, the taxon (T), and the biomass fraction (BF), or the year (Y), being the fixed effects.

y_ijk = μ + T_i + Y_j (or BF_j)+ ε_ijk

Data transformation was not necessary. Significant differences were established at p ≤ 0.05. Tukey’s honestly significant difference (HSD) test was used to differentiate within homogeneous groups.

3. Results

3.1. Growth and Biomass Partitioning

Significant differences in height (H), stem diameter (D) and biomass yield (Figure 1 and Figure 2), both among taxa (p ≤ 0.001) and years (p ≤ 0.001), were obtained but not among experimental units or for the main effect interactions (p ≥ 0.075). The plant height at the end of each rotation cycle averaged 10–12 m for Eucalyptus and Populus, 9–11 m for Paulownia and 8–9 m for Casuarina. At the end of the first rotation, the diameter of the single main stem averaged between 6.9 cm for Populus × interamericana ‘Raspalje’ and 14.6 cm for Eucalyptus × trabutii (Figure 2, Table S6), while in the second rotation, the new sprouts grown from the stump gave rise to 2.5–3.3 main stems on average, according to the taxon, whose mean diameters varied from 4.9 to 10.4 cm (Table S6). The largest plants at the end of the rotations exceed 40, 30 and 20 kg of total aboveground dry mass for Eucalyptus, Casuarina and Populus, and Paulownia, respectively. For the set of taxa studied, the four eucalyptus taxa stood out for their biomass production. At the same time, among the remainders, Casuarina equisetifolia and poplar clones ‘Adige’ and ‘AF2’ differed from the least productive (Figure 1). It should be noted that the biomass production during the first year after planting was consistently lower than that of the other years of the first and second rotations.

Regarding the distribution of the different fractions of biomass in the plants (Figure 3), we can highlight that overall, the proportion of the largest diameter class (d > 100 mm) was lower in the second harvest (end of the second rotation); at the time of the first harvest (1st rotation) the fraction of leaves amounted to 11% for Paulownia, 14% for Populus, 16% for Eucalyptus and 25% for Casuarina, while in the second harvest, they were respectively 12%, 15%, 23% and 34%; similarly, the fraction of leaves + thin branches (d < 50 mm) in the first rotation amounted to 42, 55, 40 and 55% for Paulownia, Populus, Eucalyptus, and Casuarina, respectively, and 41, 62, 57 and 66%, respectively, in the second rotation. The proportion of wood and bark in the stems and branches varied depending on their diameter (

Table 4. Stems and branches bark proportion (mean (SE)) for each taxon and diametric class on a dry mass basis. Sample size = 3–4 slices per taxon and diametric class. Different letters within a column mean significant differences among taxa (p ≤ 0.05), but in the last row, differences among diametric classes are shown (p ≤ 0.001).

Taxon		Diametric Class (% Bark)
	d < 25 mm	d = 25–50 mm	d = 50–75 mm	d = 75–100 mm	d > 100 mm
Casuarina cunninghamiana	30.0 (2.3) bc	21.4 (1.9) b	19.9 (1.6) b	14.3 (1.3) ab	12.5 (1.2) b
Casuarina equisetifolia	32.3 (3.9) bc	19.1 (3.4) ab	17.3 (3.1) b	13.2 (2.6) ab	12.3 (1.7) ab
Eucalyptus camaldulensis	25.5 (2.5) ab	17.3 (1.7) ab	16.2 (1.3) ab	14.3 (1.5) ab	12.1 (1.4) ab
Eucalyptus × trabutii	31.3 (3.8) bc	17.2 (1.9) ab	16.9 (1.4) ab	15.2 (1.3) b	12.2 (1.3) ab
Eucalyptus × urograndis	24.7 (1.7) a	18.4 (2.0) ab	19.3 (2.3) b	13.8 (1.8) ab	8.6 (1.6) ab
Eucalyptus globulus	24.0 (1.8) a	19.3 (1.9) ab	16.1 (1.2) ab	15.4 (1.3) b	12.7 (1.2) b
Paulownia fortunei	23.1 (2.0) a	16.8 (2.6) ab	13.4 (1.1) ab	10.3 (1.4) a	7.7 (1.5) a
Populus ‘Adige’	33.1 (1.5) c	17.4 (1.8) ab	13.6 (1.6) ab	10.9 (1.4) a	10.6 (1.3) ab
Populus ‘AF2’	33.4 (2.3) c	18.3 (2.3) ab	15.3 (1.8) ab	10.2 (1.5) a	10.1 (1.5) ab
Populus ‘I-214’	26.0 (4.6) ab	19.0 (1.5) ab	13.1 (1.6) ab	12.7 (1.3) ab	10.5 (1.3) ab
Populus ‘Oudenberg’	33.9 (3.3) c	16.3 (1.4) a	11.2 (1.5) a	11.5 (1.2) ab	10.6 (1.2) ab
Populus ‘Raspalje’	31.4 (2.7) bc	16.6 (1.7) ab	14.6 (2.3) ab	11.5 (1.7) ab	10.5 (1.6) ab
Total	29.2 (1.1) d	18.1 (0.9) c	15.6 (0.8) bc	12.8 (0.8) ab	11.0 (0.9) a

), always resulting in less than 15% when the stems or branches exceeded 75 cm in diameter.

Except for the three least productive taxa of this trial (Populus × interamericana ‘Raspalje,’ Populus × euramericana ‘Oudenberg’ and Casuarina cunninghamiana), at the end of the second rotation, the biomass yield increased between 5.8% and 18.4% compared to that obtained at the end of the first rotation. Thus, despite the significantly smaller diameter of stems in the second rotation than in the first, the higher number of stems compensated for biomass production. Considering the two harvests (first and second rotations together), the mean annual increment in aboveground dry biomass averaged 19–33 Mg ha⁻¹ year⁻¹ for Casuarina sp., 51–61 Mg ha⁻¹ year⁻¹ for Eucalyptus clones, 18 Mg ha⁻¹ year⁻¹ for Paulownia fortunei, and 9–28 Mg ha⁻¹ year⁻¹ for hybrid clones of Populus.

The litterfall accumulated on the ground surface at harvest time averaged 6.92 (0.36) Mg ha⁻¹ of dry matter, without significant differences among taxa (p = 0.333).

3.2. Physical–Chemical Characterization of Plant and Soil Material

Most of the properties of the two uppermost layers of the soil (0–15 and 15–30 cm) did not differ significantly from the beginning (1st year) to the end (8th year) of the trial (0.051 < p < 0.770), although the general trend resulted in a decrease in mineral nutrient content and an increase in organic matter and C/N ratio (Table 1 and Table S1). The only exceptions were the significant decreases in pH (p = 0.001), N (p = 0.014), P (p < 0.001) and Mg (p < 0.001). In a quick guesstimate for one of the reference nutrients such as N, which initially (1st year) started with an average of 0.17% (0–15 cm) and 0.14% (15–30 cm), and decreased to 0.15% and 0.11%, respectively, seven years later (8th year), we can deduce the following:

• Taking into account the stoniness (18%) and bulk density (1.56 kg dm⁻³) of the soil, the amount of total N per hectare in the 0–30 cm layer at the beginning of the trial averaged: (0.0017 + 0.0014 kg_{of N}/kg_{of soil}) × (0.15 m × 10 000 m²) × (1.0 − 0.18) × 1.56 = 5948 kg of N;
• The amount of total N per hectare in the 0–30 cm soil layer seven years later (8th year) averaged: (0.0015 + 0.0011 kg_{of N}/kg_{of soil}) × (0.15 m × 10 000 m²) × (1.0 − 0.18) × 1.56 = 4989 kg of N;
• Therefore, after seven years of cultivation, the N content of the soil decreased by an average of 5948 − 4989 = 959 kg of N. That is, 137 kg ha⁻¹ year⁻¹ of N, but in a range from 27.4 kg ha⁻¹ year⁻¹ (Casuarina equisetifolia) to 274.1 kg ha⁻¹ year⁻¹ (E. × trabutii).

Many of the soil fertility parameters (organic matter and mineral nutrient contents) were higher in the upper layer of the soil (0–15 cm) than in the lower one (15–30 cm). Still, the evolution during the 8 years analyzed did not differ significantly (i.e., the interaction Year × Soil layer was not significant, p ≥ 0.300). Likewise, no significant differences were obtained in soil properties due to the cultivation of the different taxa during the first two rotations (p ≥ 0.150). In this case, only the two taxa that presented the most extreme values of soil N content in the 8th year were differentiated: Casuarina equisetifolia (0.15% for the soil layer 0–30 cm altogether) differed from E. × trabutii (0.11%), p = 0.043.

The mineral composition of the litter sampled at the end of the rotation did not differ among taxa (p = 0.280), the contents being (mean (SE)): 1.15 (0.10), 0.13 (0.02), 0.21 (0.02), 1.41 (0.06), 0.32 (0.02), 0.03 (0.01), 0.14 (0.01), 46.3 (1.1), 5.69 (0.31)% of N, P, K, Ca, Mg, S, Cl, C and H, respectively; and 316.9 (28.6), 63.3 (4.7), 154.3 (14.8) and 5.8 (0.6) mg kg⁻¹ of Fe, Mn, Zn and B, respectively.

Regarding the chemical composition of the harvested biomass and the energetic and physical-mechanical properties of the manufactured pellets, the following can be noted:

• No significant differences among experimental units or between the two replicates of the test were found, so global mean values as a whole are shown;
• The mineral composition of the different biomass fractions differed significantly (0.001 ≤ p ≤ 0.016), with the only exception being the H content (p = 0.492). The leaves and bark presented the highest ash and mineral percentages. On the opposite side was the thick wood, with a tendency to higher values if the bark was preserved, but without differing significantly from the debarked one (

  Table 5. Mineral composition (mean (SE)) and ash content of each biomass fraction, expressed on a dry mass basis, for all the taxa and trials as a whole. Different letters within a row mean significant differences among biomass fractions (p ≤ 0.016). Significance level for H: p = 0.492.

  Variable		Biomass Fraction
  	Wood with Bark	Debarked Wood	Bark	Thin Branches	Leaves
  C (%)	48.1 (0.3) ab	49.9 (0.5) a	48.2 (0.4) ab	47.3 (0.4) b	47.1 (0.4) b
  H (%)	6.81 (0.27) a	6.32 (0.39) a	6.02 (0.35) a	7.02 (0.34) a	6.79 (0.34) a
  N (%)	0.38 (0.04) a	0.35 (0.06) a	1.48 (0.06) c	0.66 (0.57) b	1.69 (0.05) d
  P (%)	0.06 (0.01) a	0.03 (0.01) a	0.11 (0.01) b	0.10 (0.01) b	0.15 (0.01) c
  K (%)	0.35 (0.04) a	0.22 (0.03) a	0.51 (0.04) b	0.54 (0.04) b	0.92 (0.05) c
  Ca (%)	0.47 (0.07) a	0.35 (0.10) a	1.31 (0.10) b	1.16 (0.09) b	2.34 (0.10) c
  Mg (%)	0.14 (0.01) a	0.13 (0.02) a	0.18 (0.02) a	0.14 (0.02) a	0.30 (0.02) b
  S (%)	0.03 (0.02) a	0.02 (0.02) a	0.07 (0.03) a	0.06 (0.03) a	0.21 (0.03) b
  Cl (%)	0.08 (0.03) a	0.05 (0.04) a	0.26 (0.04) b	0.04 (0.04) a	0.31 (0.04) b
  Fe (mg kg−1)	50.4 (5.2) a	56.6 (6.8) a	80.2 (6.8) ab	103.9 (6.3) b	531.4 (7.5) c
  Mn (mg kg−1)	30.9 (3.1) ab	27.6 (4.12) a	75.2 (4.1) c	35.7 (3.8) ab	45.6 (4.5) b
  Zn (mg kg−1)	33.5 (4.1) ab	23.6 (5.4) a	58.3 (5.4) c	47.9 (4.9) bc	99.6 (5.9) d
  B (mg kg−1)	8.3 (1.0) a	8.9 (1.3) a	16.6 (1.3) b	13.1 (10.4) ab	49.7 (1.4) c
  Ash (%)	1.52 (0.05) b	1.07 (0.06) a	6.59 (0.06) d	2.65 (0.06) c	5.28 (0.08) e

  Wood: main stems and thick branches with d ≥ 5 cm; Thin branches: d < 5 cm.

  ), possibly due to the low percentage of bark in this fraction (11.0%–15.6% on average).

• The differences among taxa were not significant for most of the mineral elements analyzed (0.058 ≤ p ≤ 0.921), except for N (p < 0.001), Fe (p < 0.001) and C (p = 0.017). The Taxon × Biomass fraction interaction was not significant in any case (0.130 ≤ p ≤ 0.996).

  Table 6. Physical, mechanical, chemical and energetic properties (mean (SE)) of the studied taxa, expressed on a dry mass basis, for aboveground woody biomass (d ≥ 5 cm with bark). Those taxa that did not differ for any parameter measured have been grouped in a single column in order to reduce the table size and clarify it. Different letters within a row mean significant differences among taxa (p ≤ 0.017).

  	Populus ‘Adige’ ‘AF2’	Populus ‘I-214’, ‘Oudenberg’ ‘Raspalje’	Casuarina sp.	Paulownia fortunei	Eucalyptus globulus	Eucalyptus × urograndis	Eucalyptus camaldulensis E. × trabutii
  C (%)	48.7 (0.6) ab	49.5 (0.7) b	48.9 (0.5) ab	49.2 (0.5) b	48.0 (0.6) ab	48.4 (0.5) ab	47.7 (0.7) a
  N (%)	0.47 (0.05) bc	0.34 (0.06) ab	0.52 (0.05) c	0.30 (0.04) a	0.31 (0.05) a	0.25 (0.05) a	0.27 (0.06) a
  S (%)	0.04 (0.01)	0.03 (0.01)	0.03 (0.01)	0.03 (0.01)	0.02 (0.01)	0.03 (0.01)	0.04 (0.01)
  Cl (%)	0.04 (0.01)	0.04 (0.01)	0.10 (0.02)	0.03 (0.01)	0.10 (0.02)	0.10 (0.02)	0.10 (0.02)
  Fe (mg kg−1)	57.0 (8.0) b	44.5 (9.2) ab	61.9 (10.2) ab	23.1 (9.7) a	52.0 (15.1) b	55.7 (14.4) b	64.9 (13.6) b
  Ash (%)	1.20 (0.07) b	1.00 (0.07) b	2.15 (0.10) c	0.76 (0.06) a	1.30 (0.12) b	1.17 (0.11) b	2.23 (0.13) c
  HHV (MJ kg−1)	19.1 (0.3)	19.3 (0.2)	19.2 (0.3)	19.5 (0.3)	19.2 (0.3)	19.5 (0.4)	18.9 (0.3)
  LHV (MJ kg−1)	17.8 (0.3)	18.0 (0.2)	17.8 (0.3)	18.1 (0.3)	17.7 (0.3)	18.0 (0.3)	17.8 (0.4)
  Bdp (kg m−3) (1)	659 (14)	667 (14)	674 (15)	632 (11)	648 (10)	690 (12)	682 (10)
  MDp (%) (1)	97.6 (1.2)	97.5 (0.9)	96.8 (1.0)	97.7 (0.8)	97.0 (0.8)	96.6 (1.2)	96.5 (1.1)
  Moisturep (%) (1)	6.4 (0.4)	6.7 (0.5)	6.3 (0.6)	6.1 (0.4)	6.8 (0.4)	6.9 (0.5)	6.6 (0.6)
  LHVp (MJ kg−1) (2)	16.5 (0.3)	16.6 (0.3)	16.5 (0.2)	16.9 (0.3)	16.5 (0.3)	16.6 (0.3)	16.5 (0.2)
  Wd (kg dm−3)	0.42 (0.05) b	0.36 (0.02) b	0.65 (0.03) d	0.27 (0.03) a	0.54 (0.04) c	0.55 (0.03) c	0.61 (0.03) cd
  LHVv (MJ dm−3)	7.5 (0.2) b	6.5 (0.1) b	11.6 (0.2) d	4.9 (0.1) a	9.6 (0.2) c	9.9 (0.2) c	10.8 (0.3) cd

  Casuarina sp., C. equisetifolia and C. cunninghamiana are included. HHV (higher heating value) and LHV (lower heating value) on a dry mass basis. Bd_p—bulk density of pellets. MD_p—mechanical durability of pellets. Moisture_p—pellets moisture. Wd—wood density. LHV_v—lower heating value of the dry biomass on a volume basis (p = 0.037). Significance level (p-values) for thick woody biomass of the taxon effect: C, p = 0.017; N, p < 0.001; S, p = 0.921; Cl, p = 0.098; Fe, p < 0.001; Ash, p < 0.001; Wd, p < 0.001. ⁽¹⁾ Assessed according to ISO 17225-2:2021. ⁽²⁾ On a wet basis (with the moisture content after manufacturing). Without significant differences among taxa (p ≥ 0.251), the manufactured pellets had a diameter of 6.01 (0.02) mm, a length of 21.5 (4.6) mm, and an ash melting temperature of ≥ 1230 °C.

  shows the physical-mechanical and chemical properties of the wood (main stems and thick branches) for the different taxa. Paulownia fortunei wood stood out for its low percentage of minerals and ash but had the lowest wood density of all the taxa studied. However, Casuarina sp., Eucalyptus camaldulensis and Eucalyptus × trabutii, on the one hand, had the handicap of high mineral concentrations but, on the other hand, they had a high-density wood;
• As a rough estimate, during the studied period, and for all the taxa and the two rotations as a whole, the following can be noted:

  ○

  According to Figure 3, on average, the leaves, thin branches and thick wood represented 16.6, 36.3 and 47.1%, respectively, of the aboveground dry mass, which contained the N concentrations shown in Table 5. So, the average amount of N removed from the field plot with the harvested biomass was 227.6 kg ha⁻¹ year⁻¹ (from 63.6 kg ha⁻¹ year⁻¹ for Populus ‘Raspalje’ to 308.9 kg ha⁻¹ year⁻¹ for Eucalyptus × trabutii), of which an average of 40.1% (91.3 kg ha⁻¹ year⁻¹) corresponded to leaves;

  ○

  The litterfall contained an average of 11.4 kg ha⁻¹ year⁻¹ of N;

  ○

  The amount of N supplied by the fertilizer was 75 kg ha⁻¹ year⁻¹, which must be added to 18.5 kg ha⁻¹ year⁻¹ supplied by irrigation water (400 mm annually on average);

  ○

  The removed 227.6 kg ha⁻¹ year⁻¹ of N from the biomass is quite close to the N supplied by fertilization and irrigation water (93.5 kg ha⁻¹ year⁻¹); plus, the N removed from the soil litterfall set (137–11.4 = 125.6 kg ha⁻¹ year⁻¹); which amounts to 219.1 kg ha⁻¹ year⁻¹ of N on average;

  ○

  Other inputs and outputs of N have not been considered in this rough approximation (leaching, emission of N oxides, N supplied by rainwater, rock decomposition, other deeper layers of soil, etc.).

4. Discussion

The rate of diameter growth was highest in spring for all taxa due to the coincidence of favorable temperatures with well-water availability [73]. However, autumn and winter growth rates varied according to species, with significant growth recorded for the evergreens (Eucalyptus and Casuarina) and considerably less or almost zero growth in the case of the deciduous trees such as Populus and Paulownia. So, as a general rule, the evergreens, Eucalyptus and Casuarina were observed to increase their diameter throughout the year [16,29]. By contrast, the deciduous trees, Populus and Paulownia, achieved close to zero growth during the cold season [17,29] as they began to slow down metabolic activity, bud set and leaf fall. This means that, in climate areas with mild winters, evergreen species can take advantage of deciduous ones, taking advantage of sunny days with favorable temperatures for their growth. Similarly, high summer temperatures decreased growth due to their effect on net photosynthesis rates and transpiration [73], which was not as high as in spring, despite having irrigation water.

In terms of genera, the highest diameter growth rates were achieved by Eucalyptus, such that the four taxa in this genus exceeded 14 cm in diameter at the end of the first rotation period (3rd year after planting), but with no significant differences among them [74,75]. However, between the two Casuarina species, Casuarina equisetifolia stood out above Casuarina cunninghamiana. These growths were somewhat greater than those obtained by other authors with different Casuarina species in other soil and climate conditions than in the present study [28,76]. In the case of Populus clones, the good growth response of ‘Adige’ and ‘AF2’ over other clones has been shown in other studies [17,26,54,77].

In general, the lowest annual yields in biomass corresponded to the first year after planting. This may be due to the post-transplant effect, whereby plants do not achieve their maximum growth rates until they have reached a certain root system size. For this reason, the first year after the felling, the growth was vigorous since it started from an already installed root system [16,17]. The maximum mean annual biomass at the end of the rotation period was recorded by the Eucalyptus genus (51–61 Mg ha⁻¹ year⁻¹ of aboveground dry biomass). Behind the four Eucalyptus taxa, Casuarina equisetifolia stood out by producing up to 33 Mg ha⁻¹ year⁻¹ of dry biomass. These high yields could be due in part to the fact that both genera are evergreen, and thus they can maintain continuous growth throughout practically the whole year, where mild winters allow, as is the case in the study area. In addition, this leads us to conclude that the use of Casurina equisetifolia in short-rotation forest plantations for energy production could be appropriate in areas such as that in the current study, especially those with degraded soils. Despite smaller yields than other species (such as Eucalyptus), the ability of Casuarina sp. to fix atmospheric nitrogen [78,79] makes them highly suitable for cultivation in such kinds of soils without resorting to excessive fertilization and, at the same time, contributing to their restoration [19,28,29,80] and to energy saving [16,17,30]. In this study, it was estimated that up to 173 kg ha⁻¹ year⁻¹ of N have been fixed by Casuarina equisetifolia, taking into account the N contained in its harvested biomass (282.5 kg ha⁻¹ year⁻¹), the supplied by fertilization and irrigation (93.5 kg ha⁻¹ year⁻¹) and the removal from the whole soil litterfall (16 kg ha⁻¹ year⁻¹). In the case of the Populus clones, ‘Adige’ and ‘AF2’ stood out by giving biomass yields (about 28 Mg ha⁻¹ year⁻¹) not dissimilar to Casuarina equisetifolia. The biomass yield of Paulownia fortunei, such as Populus, a deciduous species, was low (18 Mg ha⁻¹ year⁻¹), despite having achieved a growth of D and H similar to the more productive Populus clones (‘Adige’ and ‘AF2’) and Casuarina equisetifolia. This poor performance of Paulownia fortunei can be accounted for by the low density of the wood (0.27 kg m⁻³), which restricts the conversion of volume into dry weight. The low density is a factor that should be taken into account when this species is considered for energy use since it could substantially increase transport and storage costs [80] if it is not densified into pellets. For example, its LHV on a dry mass basis (18.1 MJ kg⁻¹) is in the same range as that of the other taxa studied (17.7–18.0 MJ kg⁻¹), whereas if it is expressed on a volume basis (4.9 MJ dm⁻³), it differs significantly from all other taxa (6.5–11.6 MJ dm⁻³). In terms of energy, the LHV of the harvested woody biomass (excluding leaves) was equivalent to 725–867 GJ ha⁻¹ year⁻¹ (or 201–241 MWh ha⁻¹ year⁻¹) for Eucalyptus, 238–413 GJ ha⁻¹ year⁻¹ (or 66–115 MWh ha⁻¹ year⁻¹) for Casuarina, 288 GJ ha⁻¹ year⁻¹ (80 MWh ha⁻¹ year⁻¹) for Paulownia and 137–428 GJ ha⁻¹ year⁻¹ (38–119 MWh ha⁻¹ year⁻¹) for Populus.

The obtained biomass yields, surprisingly high for Eucalyptus, are in the highest range reported for these plant genera when grown under favorable conditions [16,17,22,24,25,26,28,52,53,81,82,83,84]. This is most likely due to the favorable climatic and edaphic conditions of the field plots, characterized by a Mediterranean climate with mild winters, good availability of nutrients and water, and well-drained fertile soil. It should be noted that in this study, data of the total aboveground dry biomass are mainly shown, which includes leaves, branches and the trunk with bark. In any case, even knowing the production potential of these taxa, caution must be exercised when extrapolating the data to an industrial plantation since the two field plots in this study were very homogeneous, small in area, and have not suffered mortality or damage by biotic or abiotic agents. However, commercial plantations usually extend over larger areas, making them more heterogeneous and more likely to suffer damage from unfavorable agents. There are multiple environmental and plantation management factors that affect growth and biomass production, such as availability of water and mineral nutrients, temperature and precipitation, soil physicochemical properties, plant material (genotype), crop density, etc. [21,26,27,29,52,63,84], making it difficult to compare studies and attribute differences to only one or a few factors. Comparing some results reported for the Iberian Peninsula and other areas under the Mediterranean climate with those of the present study, in plantations supplied with irrigation and fertilization, it can be observed that the annual biomass production is reduced by approximately 10%–55% in areas with lower annual mean temperature (7–14 °C) than that of this study (17 °C) [17,21,26,27,46,49,53,55,64,85]. However, for deciduous species such as those of the genus Populus, growth is more related to the maximum temperatures of the warm season than to the mean annual temperature, as well as to the rainfall regime and the dry summer period [52], since during the cold season they are in vegetative rest, without leaves. Suppose deficient management is added to the effect of temperature, which does not provide the plants with all the water (at least >500 mm of annual precipitation or availability of groundwater) or the necessary nutrients to achieve their maximum growth potential. In that case, the latter will be reduced proportionally to such an extent that yields can be below 5 Mg ha⁻¹ year⁻¹ in some instances [22,24,29,52,54,56,57,58,59,84,85].

As a result of global warming caused by climate change, the Mediterranean region’s temperature is expected to rise by 0.9–5.6 °C by the end of the 21st century [86]. This seems to favor higher latitudes or altitudes in terms of crop production. Likewise, it could expand the potential area of evergreen species. For example, allowing frost-sensitive species such as eucalyptuses to reach higher altitude sites. However, this increase in temperature is accompanied by a reduction in rainfall of 4%–22%, as well as an increase in extreme weather events, such as heat waves and prolonged droughts, which will negatively affect the plantations and increase dependence on irrigation to achieve high production. It is predicted that many areas from current oceanic or humid subtropical climates might change to a Mediterranean climate in the coming years, and areas from current Mediterranean climates to semi-arid climates [86]. A climate change scenario that implies an increase of 2 °C in temperature and a decrease of 30–40 mm in annual precipitation could mean an 18% reduction in the productivity of these crops [25]. For all these reasons, further research should be carried out on the selection and improvement of cultivated genotypes for their water use efficiency and drought resistance [17,26,52,59,73,87], as well as on improving cultivation techniques to make them more efficient. At the same time, we try to mitigate climate change through the use of renewable energy sources, such as biomass.

The high growth rates of some of the taxa studied gave rise to a trend, although not yet significant, of increasing organic matter and the C/N ratio in the upper layers of the soil, as well as in the litter layer, in increment ranges similar to other studies with fast-growing species [16,17,19,28]. This may be due to a large amount of litterfall every year, mainly leaves, which degrade rapidly in the shady and humid environment of this crop. As the period studied has been short, and the processes of soil evolution are usually slow, it is not surprising that significant differences have not yet been obtained with respect to the original state of the soil that previously supported a fast-growing legume species for 12 years. However, for some of the nutrients analyzed, such as N, P and Mg, significant decreases were detected with respect to the initial contents in the soil. This opposite trend between the increase in organic matter and the decrease in N possibly explains the increase in the C/N ratio. This means that, in this study, the natural loss of nutrients (leaching, denitrification) and the removal of nutrients with the harvested biomass was greater than that supplied by fertilization, and, therefore, the plants took them from the soil according to their growth needs. For example, on average, for all the taxa as a whole, in this study, the decrease in N of the whole soil litterfall was 125.7 kg ha⁻¹ year⁻¹. Therefore, in order to reduce harvest-related nutrient exports, a more conservative approach involving the removal of only larger diameter stems and branches (d > 2.5 cm) might be recommended to retain more of the accumulated nutrient capital on-site [28]. Similarly, on average, if the leaves of the harvested plants had been left on the cultivation soil, this would only imply leaving 16.6% of the harvested dry biomass in the field but 40.1% of the nitrogen (91.3 kg ha⁻¹ year⁻¹), reducing the fertilization costs and the environmental impact, since only 34.4 kg ha⁻¹ year⁻¹ of N would have been removed from the soil (125.7 − 91.3 = 34.4). This does not apply to deciduous species since they do not have leaves at harvest time. Furthermore, if, in addition to the leaves, the other lower quality biomass fractions for energy use were also left in the cultivation land, that is, the thinnest branches (d < 2.5 cm) and the bark of thick branches and stems, the economic and energy impact would be reduced to a greater extent [16]. The high mineral nutrient and ash content of the bark and the high proportion of bark on the thinnest branches (23%–34%) support this recommendation [71,88,89]. Finally, considering the use of N-fixing species, as is the case with the genus Casuarina discussed above, alone or mixed with other more productive species would reduce nitrogen fertilization requirements [17,24,28].

Regarding the biomass quality for energy use, on the one hand, and according to the ISO 17225-4 standard [88], wood chips made from thick woody biomass are of good commercial quality, mainly for hybrid poplars, Paulownia fortunei, Eucalyptus globulus and Eucalyptus × urograndis (ash content < 1.5%), but also for Casurina sp., Eucalyptus camaldulensis and Eucalyptus × trabutii (ash content < 3.0%). On the other hand, adding the thinnest branches and leaves to the wood chips would decrease their commercial quality by increasing the ash content, and for the same reason, the debarking of the thick wood would improve it.

Furthermore, according to standard ISO 17225-2 [71], due to their physical–mechanical, energetic and chemical characteristics (N and S contents, heating values, bulk density, mechanical durability, ash content, moisture, etc.), the manufactured pellets corresponded to commercial quality categories ENplus^®-A2 (Paulownia fortunei, hybrid clones of Populus) and ENplus^®-B (Eucalyptus globulus, Eucalyptus × urograndis), all of them valid for residential use. However, due to the high ash content, Casurina sp., Eucalyptus camaldulensis and Eucalyptus × trabutii must be classified as industrial, commercial category I3. None of the manufactured pellets contained heavy metals or halogenated organic compounds, more than typical virgin material values or typical values of the site of origin. Chlorine content was slightly higher than usual for these species [89,90]. Still, it is mainly justified by the proximity to the coast of the field trial, not because it is an intrinsic property of these studied taxa. The resulting pellets possessed the characteristics of hardwood species and were of better quality than herbaceous species pellets. However, they did not reach the maximum standardized quality of pellets derived from conifer debarked wood [16,17,19,79,90,91,92,93,94]. All this is closely related to the chemical composition of the biomass, mainly the lignin and extractive contents. These are two factors that affect the bonding quality of the biomass particles during pelletization [95,96,97,98].

Finally, the most immediate use could be in large thermal boilers or cogeneration plants after transformation into chips, especially Casuarina sp., Eucalyptus camaldulensis and Eucalyptus × trabutii because they are evergreen species and because of the high density and ash content of the wood. However, to give added value to the biomass, for small or medium-sized thermal boilers, Eucalyptus globulus, Eucalyptus × urograndis and Populus wood chips are more recommended. On the other hand, in the case of Paulownia fortunei, due to its low density and ash content, its transformation into chips is less recommended, as it can be better valued by transforming it into pellets. In addition, for all the taxa studied, the transformation into pellets of the thick wood of branches and trunks (d > 5 cm) would endow them with added commercial value, thanks to the high biomass production of these species.

Under the Mediterranean climate, woody species have been cultivated for several decades, so agroforestry owners have experience with these crops, mainly in the case of poplars and eucalyptuses [26,27,87]. Traditionally these crops were used to obtain wood and paper pulp, but in the last decade also for biomass for energy use. The commercialization of the product is assured due to the presence of biomass thermoelectric plants. For example, in the region where this study was carried out (Andalusia, Spain), there are 17 power plants with a biomass electricity generation capacity of more than 270 MW, which demand more than 2.5 million tons of agroforestry biomass annually [99], so any place in the region is less than 100 km away from a biomass power plant. In the rest of Spain, another 15 plants with similar characteristics are currently in operation. Moreover, to obtain additional added value through the transformation into pellets, it is worth saying that the pellet market has been expanding in recent years. World pellet production in 2019 was estimated at 40.5 million tons, an increase of 125% compared to seven years ago [100,101]. The consumption of pellets in 2020 in the EU exceeded 19 million tons, an increase of 5% compared to the previous year [102]. Regarding Spain, about 100 pellet production factories, evenly distributed throughout the national territory, produce 1 million tons per year (666% more than 12 years ago), but with a production capacity of up to 2.5 million tons per year.

It would be interesting to analyze in the future if the wood of the studied taxa could be mixed with debarked pine wood to produce commercial pellets of the highest quality for residential use [101], as well as delve into the effect on the soil due to the cultivation process in order not to degrade it and continue advancing in genetic improvement programs that achieve highly productive taxa by increasing their energy properties and nutrients and water use efficiency.

5. Conclusions

• The biomass yield of short-rotation forest tree species (hybrid clones of Populus and Eucalyptus, monospecific clones of Eucalyptus and Paulownia fortunei, and Casuarina species) averaged 9–61 Mg ha⁻¹ year⁻¹ under a Mediterranean climate with mild winters, good availability of nutrients and water, and well-drained fertile soil;
• In terms of energy, the LHV of the woody biomass was equivalent to 38–241 MWh ha⁻¹ year⁻¹ (equivalent to the replacement of about 3400–22 000 L ha⁻¹ year⁻¹ of diesel);
• Plantation managers must take into account nutrient inputs and outputs in order to ensure the sustainability of the system and to prevent any loss in soil fertility and productivity;
• The biomass quality varied among the taxa studied, but they all have enough commercial quality to add value through their transformation into chips or pellets;
• It is recommended that the lowest quality biomass fractions for energy use (leaves and thinnest branches < 25 mm diameter) should not be removed from the cultivation land due to their contribution to the nutrient cycle;
• The use of N-fixing species, such as Casuarina sp., should be considered in future plantations, mainly in degraded soils.