Mixed-Species Plantation Effects on Soil Biological and Chemical Quality and Tree Growth of A Former Agricultural Land

: Tree planting on abandoned agricultural land could both restore the soil quality and increase the productivity of economically valuable woody species. Here, we assess the impact of mixed-species tree plantations on soil quality at a site in Central Italy where tree intercropping systems were established 20 years ago on a former agricultural land. These intercropping systems include two species of economic interest, Populus alba and Juglans regia , and one of three different nurse trees, i.e., Alnus cordata , Elaeagnus umbellata , both of which are N-ﬁxing species, and Corylus avellana . We measured tree growth and compared how soil organic matter, soil extracellular enzymes, and nematodes of different feeding groups varied among the intercropping systems and relative to a conventional agricultural ﬁeld. Our results indicate that tree plantation led to an increase in soil carbon and nitrogen, and enhanced enzyme activities, compared with the agricultural land. The proportion of nematode feeding groups was heterogeneous, but predators were absent from the agricultural soil. Multivariate analysis of soil properties, enzymatic activity, nematodes, and tree growth point to the importance of the presence N-ﬁxing species, as the presence of A. cordata was linked to higher soil quality, and E. umbellata to growth of the associated valuable woody species. Our ﬁndings indicate that intercropping tree species provide a tool for both restoring fertility and improving soil quality.


Introduction
Cropland abandonment is an important process in many regions of the world and one of the dominant processes of land use change in Europe, where abandoned agricultural lands are estimated to cover between 20,277 and 211,814 km 2 by 2040 [1]. Policies and management of these areas have spurred debate regarding the loss of (traditional) agricultural landscapes and potential impacts on biodiversity and ecosystem services [2].
we expect different effects of native and non-native N-fixing nurse trees, as found by Vilà et al. (2011) [27].

Study Site
The experimental area is located near Brusciana, Tuscany, Central Italy (43 • 40 29 N, 10 • 55 21 E), at 35 m a.s.l. The location has a mean annual precipitation of 850 mm and a mean annual temperature of 15 • C. The soils of the site developed on recent (Holocene) fluvial deposits and are Fluventic Haplustepts coarse-loamy, mixed, thermic of the U.S. Soil Taxonomy, according to the 1:250,000 soil map by Regione Toscana (http://sit.lamma.rete.toscana.it/websuoli/, accessed on 5 March 2021). In 1996, an experimental tree plantation was established for the production of wood on agricultural land. This plantation was polycyclic, i.e., different crop trees combining fast growing and rotation periods were planted together: (i) broadleaved crop trees with medium-long rotation (generally 20-30 years); (ii) crop trees with medium-short rotation (generally 10-15 years), namely poplar clones; (iii) crop trees with short rotation (generally 3-7 years). Trees were planted at a fixed distance to reach commercially valuable size [28]. In this trial, two target broadleaved tree species widely used in Italy for timber production, poplar (Populus alba L., Salicaceae) and walnut (Juglans regia L., Juglandaceae), were planted together, using a triangular layout with a distance of 8 m (179 trees ha −1 ), and intercropped with N-fixing species, such as Elaeagnus umbellata and Alnus cordata and with another nurse shrub such as Corylus avellana, using a rectangular layout of 3.5 × 4 m 2 (715 trees ha −1 ) (Figure 1). No thinning was performed in the plantation. Planting plots (≈0.4 ha each) were compared using a randomized blocks design with three replicates:

1.
Poplar and walnut (plots PJ), planted in a high-density mixture.

3.
Poplar and walnut intercropped with autumn olive (Elaeagnus umbellata Thunb., Elaeagnaceae) (plots PJE), an alien N-fixing shrub used in tree farming plantations for its high ability to fix nitrogen in the soil.

4.
Poplar and walnut intercropped with Italian alder (Alnus cordata (Loisel.) Duby, Betulaceae) (plots PJA), an N-fixing tree species common in Southern Italian forests and widely used in tree farming plantations.
Soil samples were also collected from an adjacent active agricultural field (AL).

Sampling
The following sampling procedure was carried out in each field replicate (in both afforested and agricultural soil). In November 2018, in the centroid of each of the 0.4 ha planting plots, we designed a regular 12 × 12 m 2 grid. At the central point of each grid and at four microsites 6 m away (in the direction of the cardinal points), we collected mineral soil with a core sampler (diameter 5 cm; length 10 cm) to a depth of 10 cm, resulting in an overall sample size for soil analyses of 75 (i.e., 5 soil cores × 3 field replicates × 5 treatment plots). Within each tree planting plot, four dominant trees of P. alba and four trees of J. regia were sampled using a 5.1 mm diameter Haglof increment borer, collecting two cores per tree at an angle of 120 • from each other.
It should be taken into account that sampling error may have been comparatively large due to small sample size [29].
All the P. alba and J. regia presented in the plots were alive and of comparable diameter and height (tree height and mean diameter at breast height, respectively, 43 ± 6 cm and 22 ± 3 m P. alba; 17 ± 3 cm and 12 ± 2 m, J. regia).

Physical and Chemical Soil Analysis
Once in the laboratory, the soil samples were air-dried and then sieved to remove rock fragments (>2 mm). Soil pH was measured potentiometrically using a 2.5:1 (v/w) 0.01 M CaCl 2 solution to soil ratio. Particle size distribution was determined by the hydrometer method on one combined sample per plot (i.e., three replicates per stand type), obtained by bulking equal aliquots from each plot. Soil bulk density was measured by dividing the weight of the undisturbed core samples, after drying at 105 • C in an oven to constant weight, by the volume of the steel cylinder of the sampler.
Soil organic matter (SOM) was determined gravimetrically by incineration in a muffle furnace at 375 • C for 16 h [30]. Total carbon and nitrogen (TC and TN) were determined by dry combustion on an elemental analyser (Elementar Analysensysteme GmbH, Vario EL III from Elementar, Langenselbold, Germany) [31], while total inorganic carbon (TIC) was measured volumetrically using a Scheibler calcimeter [31]. Total organic carbon (TOC) was calculated as the difference between TC and TIC. All chemical variables were determined in each plot on 5 field cores and 3 laboratory replicates.
The AI3 index, as validated by Puglisi et al. (2006) [37], is calculated by a linear combination of three different enzymes as follows: This index is used to assess the quality of soils affected by various forms of degradation such as irrigation with saline water, contamination by heavy metals, organic fertilizer, erosion, and contamination by industrial and/or urban waste, as well as forest soils [32].
High values indicate a low soil biological quality [37], while the index usually assumes negative values in high-quality soils.

Nematodes
Soil nematodes were extracted from 50 g fresh soil aliquots using Baermann funnels over 3 days. The aliquots were obtained by mixing together 10 g of soil from each field replicate, in order to obtain a pool. Each aliquot was then divided in two replicates of 25 g each. Nematodes collected in water were removed daily and stored at 4 • C prior to identification. Within 5 days of extraction, the nematodes were counted and identified using an optical microscope (Leica DMLS, Wetzlar, Germany; 100× magnification). Based on morphology, they were allocated to the following feeding groups: bacterivores, fungivores, herbivores, omnivores, and predators [38]. A few individuals that could not be assigned with confidence to any of the above were classified as unknown.

Tree-Ring Measurement and Basal Area Increment of Individual Trees at the Species Level
The core wood samples were air-dried and then polished with a series of successively finer sand-paper grits until rings were clearly visible. Tree rings were measured to the nearest 0.01 mm using a binocular stereoscope and a LINTAB measuring device (Rinntech, Heidelberg, Germany). Cross-dating of the tree-ring series was checked using the program COFECHA [39].
Basal area increment (BAI) of both target species, J. regia and P. alba, was calculated as the average accumulated increment of the tree rings following the equation: where BAI at year t is equivalent to the annual ring area, r t is the stem radius at the end of the annual increment, and r t−1 is the stem radius at the beginning of the annual increment. BAI series were used as they minimize the effect of tree size and age on annual growth trends while retaining the high-and low-frequency signals contained in the tree ring series (see Battipaglia et al. (2017) [11] for details), unlike ring width series and with the exception of the first years (3-5 years) of increasing juvenile growth rates.

Statistical Analyses
One-way analysis of variance (ANOVA) was used to test the effects of the different stands on the response variables, followed by Tukey's post-hoc test (α = 0.05). The model residuals were checked for normality and homogeneity of variance. In order to investigate patterns of covariance between soil biological data (i.e., enzyme activities and proportion of nematode functional groups) and tree basal area, we employed a multivariate statistic known as Two-Block Partial Least Squares (2B-PLS). This statistical approach can provide reliable inference when dealing with matrices that have a comparatively low sample size as well as highly correlated variables [40,41]. This technique has recently been applied in several ecological contexts [32,42,43]. Patterns of covariance between the two matrixes can be represented by a scatterplot for the first axis of the 2B-PLS, where the x-axis and the y-axis represent the two multivariate matrixes, respectively. Patterns of positive or inverse correlation can be asserted both within and between matrixes. All analyses were done in R 3.6.1 [44], using packages 'plsdepot' and 'ggplot2'. Results are reported as mean ± standard error.

Soil Properties
Soil texture was mostly sandy loam. The contribution of the clay fraction was rather homogeneous in the different stands, ranging between 15% and 20% of the total mineral mass (Table 1). However, there were some differences in terms of sand and silt fractions. In particular, PJA stands had the lowest sand content and AL had the highest silt content. In spite of these differences in particle size distributions, soil bulk density was virtually the same throughout the investigated area. Soil pH was neutral and did not differ significantly between the stands, while it was slightly higher in AL. Table 1. Soil particle-size distribution, bulk density, and pH of the investigated tree stands and the arable land. Site labels: PJ = white poplar and common walnut; PJC = PJ intercropped with common hazel; PJA = PJ intercropped with Italian alder; PJE = PJ intercropped with autumn olive; AL = agricultural land. Values are mean and standard deviation (n = 5 for soil particle-size distribution and 15 for bulk density and pH). Superscript letters indicate significant differences within means in the column according to one-way ANOVA at p ≤ 0.05 and Tukey's test.

Stand
Overall, soil TC, TN, and SOM contents were higher in the tree stands than in AL ( Table 2). Among the stands, PJE had the smallest TC and TOC values, and PJA the highest TC, TN, and TOC. Soils supporting the N-fixing species (PJA and PJE) were the poorest in TIC, while the arable soil (AL) was the richest. Table 2. Soil chemical properties of the investigated tree stands and the arable land (see Table 1 for site labels). Values are mean and standard error (n = 15). Superscript letters indicate significant differences within means in the column according to one-way ANOVA at p ≤ 0.05 and Tukey's test.
The activity of the enzymes β-glucosidase, urease, and acid phosphomonoesterase was about twice as high in the tree stands as in AL (Figure 3).
Overall, enzymatic activities were twice as high in the tree plantation plots as in the agricultural field. As a result, the AI3 index calculated from the three enzyme activities was highest in AL, indicating lower soil biological quality (Figure 3). This suggests that tree plantations improved soil quality.

Nematodes
Herbivores were the nematode functional group with the highest abundance overall, with the maximum relative proportion in PJ (0.59) and PJE (0.66), followed by bacterivores, with the greatest relative abundance in AL (0.49) and PJC (0.45) (Figure 4). The highest proportion of fungivores was found in AL (0.10) and PJ (0.8), while omnivores were more abundant in PJC and PJE (0.05). Generally, predators were the least represented group, ranging from none in AL to 0.04 in PJC. The proportion of unknown was low across land uses (min. 0.01 in AL, PJ, PJE and max 0.03 in PJC). Overall, the agricultural field tended to have higher relative abundance of bacterivores and fungivores compared with the tree plantations, but the differences were not statistically significant.  Table 1 for site labels).

Basal Area Increment (BAI)
For both target tree species J. regia and P. alba (Figure 5a,b), BAI was highest in the intercropping with E. umbellata (PJE). PJC and PJA stands had similar BAI (possibly with an outlier in PJA determining this similarity) and distinct from PJE and PJ. Finally, the PJ stand was also distinct from PJE (despite an outlier). Different letters indicate significant differences (p ≤ 0.05, Tukey's post-hoc test following one-way ANOVA). See Table 1 for site labels.

Two-Block Partial Least Squares Analysis
The 2B-PLS ( Figure 6) separated the PJE plot from the other plots. Considering the large degree of variability within plots (as shown by the error bars), on the y-axis (i.e., basal area increment) PJA, PJC, and PJ were largely overlapping. On the x-axis (i.e., biological data), instead, PJA was marginally distinct from PJC and PJ. The correlation coefficients within and between blocks indicated that PJA (and to a lesser extent PJC and PJ) had lower cumulative basal area compared with the other stands; this was inversely correlated with the enzyme activities of β-glucosidase, acid phosphomonoesterase, and urease as well as the proportion of unknown and bacterivore nematodes. On the other hand, PJE was different, showing the highest cumulative basal area, which was linked to the highest AI3 values and positively correlated with the proportion of predatory and herbivore nematodes. Overall, considering that higher AI3 values are linked to lower soil quality, a difference between PJA and PJE was found: PJA showed a higher soil quality but lower tree growth, while the opposite was true for PJE. Enzyme activities and functional abundance of nematodes were consistent with this pattern. FDAH and the proportion of omnivore and fungivores nematodes had a low correlation coefficient within Block 1, indicating that these variables contribute poorly to the observed pattern of covariance.

Soil Properties
Soil pH may respond rapidly to afforestation because of the change in the base cation cycle and the addition of plant residues [45], which also lead to higher accumulation of organic matter in soil [46]. Therefore, the slight but significantly lower pH we measured in the tree plantation plots compared with the agricultural land was plausibly due to higher plant uptake of base cations [45] and higher organic residue deposition and decomposition, which imply a net input of protons to the soil [47]. N-fixing trees promote soil acidification in mixed stands because of both greater nitrate leaching and release of strongly acidic soil organic matter (SOM) [48]. In our study, however, soil pH clustered around neutrality.
The tree plantation soils showed significant differences from AL in terms of all the other investigated soil abiotic properties, i.e., TC, TOC, TN, and SOM. Previous research highlighted that changes in land use influence soil fertility and quality indicators because of subsequent alterations in abiotic and biotic factors along with SOM stabilization [49]. Land use change alters C and N inputs and dynamics [50], which subsequently regulate microbial processes and soil physical and chemical features [51]. Afforestation of arable lands has been shown to enhance soil C stocks, albeit only in the long term [52]. In fact, whereas tillage destroys the soil structure, enhancing organic matter mineralization and CO 2 emissions [53], afforestation implies a continuous input of above-ground and root litter [46]. In this study, we found differences between the tree stands and the arable field in terms of TOC and TN, as well as SOM content, despite the relatively short time since the tree plantations' establishment. The differences between the different tree mixtures were smaller. The PJA stands, with alders, showed the highest soil TC, TOC, and TN contents, consistently with Chiti et al. (2007) [54] and Chodak and Niklinska (2010) [55]. Alder is symbiotic with N-fixing actinomycetes and promotes soil N enrichment, leading to nutrient-rich litter and higher soil fertility [56]. Symbiotic N fixation may increase not only soil TN but also TOC [57], as a result of both the slower decomposition rate of humified organic matter and greater C inputs [58]. Innangi et al. (2017) [43] showed that Italian alder leaf litter has low lignin and high ethanol extractable contents, which promote a fast litter decomposition rate and, consequently, a high quantity of soil TOC that can build up as microbial C [59]. The stands without N-fixing species, i.e., PJ and PJC, gave contrasting results in terms of soil C org /TN ratio; we expected this ratio to be higher here than in the other stands, but this was true only for PJC.

Enzymatic Activities and Soil Quality
All enzymatic activities had lower values in AL compared with tree stands, except the acid phosphomooesterase in PJC (Figures 2 and 3). The enhanced enzymatic activities could be attributed to the increase in soil TOC and TN stocks due to plant residue inputs with afforestation [60], as this provides more substrate for microbial assimilation [61].
Overall, soil microbial activity was not significantly different under N-fixing and non-N-fixing trees, except for FDAH ( Figure 2) and acid phosphomonoesterase (Figure 3). Similarly, previous studies found no difference in microbial community activity between the non-N-fixing Pseudotsuga menziesii (Mirb.) Franco and N-fixing Alnus rubra Bong. trees in the forest of northwestern North America [62], or in microbial biomass C and N under Acacia mangium Willd. compared with Eucalyptus grandis W. Hill in a 20-month-old mixed-species planting in Brazil [63]. Nonetheless, in the latter case, there was significantly more dehydrogenase enzyme activity under A. mangium than E. grandis, suggesting some differences in the microbial community activity underneath these N-fixing and non-Nfixing tree species. Our results on soil enzymatic activity are in agreement with Bini at al. (2013) [63] since FDAH, as well as dehydrogenase, has usually been considered a good indicator of microbial biomass in soils [42].
While FDAH was higher in the presence of nitrogen-fixing species, acid phosphomonoesterase increased with C. avellana. Acid phosphomonoesterase is involved in P cycling as it catalyzes the hydrolysis of organic P esters to inorganic P [64]. Since the synthesis of phosphomonoesterase may be suppressed by the presence of inorganic P, high phosphomonoesterase activity in PJC could indicate insufficient P supply [55].
The activity of several soil enzymes is considered a reliable indicator of soil quality under different tillage practices and soil management [65]. This is supported by our findings. The AI3 index, based on enzymatic activities, showed that establishment of mixed-tree intercropping systems increases the quality of the soil compared with the agricultural field (Figure 3). Although the mixtures in our study have a low number of species, this result is consistent with the finding by Lange et al. (2015) [66] that tree species diversity can increase SOM and biochemical properties, presumably via higher root inputs and other yet unidentified mechanisms.

Nematodes
Microfauna grazing on microorganisms, such as nematodes, affects the growth and metabolic activities of microorganisms and alters the microbial community, thus regulating rates of decomposition [67] and nutrient mineralization [68]. Our study showed a heterogeneous pattern among the different forest associations, particularly regarding whether herbivores and bacterivores dominated the community. Yeates and Bongers (1999) [69] showed a dominance of bacterivores for both Populus and Alder soils (0.43 and 0.80, respectively), while herbivores had lower proportions (0.27 Populus; 0.07 Alder). In our study, the only intercropping system in line with these results is PJC, where the proportion of bacterivores was slightly higher than that of herbivores. By contrast, in PJE herbivores were more than three times as abundant as bacterivores. Armendàriz and Arpin (1996) [70] found a greater relative abundance of bacterial feeder nematodes during the first stages of colonization of the soil community succession, while an increase in plant-feeders was observed later. However, the stands in our study were established in the same year and had generally similar soil properties. The greater abundance of herbivores in PJE could be due to the higher tree biomass, the presence of the alien N-fixer E. umbellata, and/or due to unmeasured soil or biological properties. Regardless of the causes, the pattern suggests a greater energy flow from living roots to the soil food web in PJE than in PJC. Laboratory experiments and field studies have demonstrated that nematodes that feed on bacteria and fungi play important roles in influencing the turnover of soil microbial biomass and, thus, the availability of plant nutrients [15]. Differences in relative abundance of bacterial-feeding and fungal-feeding nematodes reflects differences in decomposition pathways or channels [71].
Plant-feeding nematodes have been reported to be more sensitive to changes in plant community composition than other groups of nematodes, presumably due to their selective feeding on certain host plants [67,69]. In fact, a global analysis of nematode communities found that edaphic factors (such as sand content and pH) were stronger predictors of communities dominated by bacterivores, whereas vegetation cover was predictive of herbivore-dominated communities [19]. The dominance of bacterivore nematodes in the agricultural soil in our study is in line with results of Yeates and Bongers (1999) [69], who found that bacterivores dominated with a relative proportion of 0.43 in a ploughed corn/soybean rotation cultivated field. Bacterivore abundance generally increases with the incorporation of cover crop residues, often attributed to the increase in bacterial biomass after cover crop additions [72]. However, in the study of Yeates and Bongers (1999) [69], the relative proportion of fungivore nematodes was much higher (0.27) than in the agricultural soil of our study (0.10). The greater relative quantity of other groups could be explained by the fact that fungi and fungivore nematodes are more likely to regulate decomposition of surface residues in no-till soils, whereas bacteria and bacterivorous nematodes regulate decay rates in incorporated residues in conventional-till soils [67], as in our case. Lastly, we found few predatory nematodes overall, but they were notably absent only from the agricultural plots, suggesting a more mature food web in the tree plantations.

Basal Area Increment
Tree growth is influenced by a wide variety of factors. Although we sampled trees in the centre of each plot, our results could be influenced by the limited extension of the plots and by edge effects, in which conditions near the boundary of two habitats differ from the interior. This can affect tree community composition, structure, and possible tree mortality [73,74]. However, mixed plantations, which are often more productive than monocultures [75], are characterized by pluri-stratified canopies that may increase light availability for the species in the dominant layer [76]. Elaeagnus umbellata is a shrub; hence, it does not compete for light with larger, economically valuable tree species, while at the same time enriching soil in nitrogen as it is an N-fixing species. As reported by Forrester et al. (2006) [76], regarding the availability of N-fixing species at stand level, once N has been transferred to the non-N-fixing species, this can result in higher CO 2 assimilation rates [77]. However, our study showed that PJA had lower basal area values for both target species (Figure 5a,b) compared with PJE. Battipaglia et al. (2017) [11] reported a significant increase in basal area in Quercus robur L. growing together with the native N-fixing species A. cordata. Mixing valuable tree crops such as J. nigra L. or J. regia with N-fixing E. umbellata can improve tree growth performance of the former species as a result of increased soil N mineralization rates [78]. Panshke et al. (1989) [79] estimated N mineralization rates in a 19-year-old intercropping of black walnut (J. nigra) with N-fixing autumn olive (E. umbellata.) or black alder (A. glutinosa) and in pure walnut plantings at two locations in Illinois, USA. Walnut size was highly correlated with soil N mineralization, particularly soil NO 3 -N production in a plot. Total N and TOC contents were smallest in Elaeagnus interplantings at both sites, in concordance with our results, but NO 3 -N concentration was higher in the Elaeagnus plot than in the Alnus plot. Khamzina et al. (2009) [78] also showed that the increase in plant-available soil N was significantly higher in Elaeagnus angustifolia L. plots than in Populus euphratica Oliv. and Ulmus pumila L. plots. All this suggests that Elaeagnus favours the formation of N-NO 3 , which, in turn, could positively affect the growth of target species.

Overall Trends
The plots with N-fixing species were markedly different, as highlighted by the 2B-PLS ( Figure 6). While PJA was distinguished by its higher enzymatic activity values and higher soil quality as indicated by AI3 index, PJE was distinct because of its lower soil quality but higher BAI values. A key difference between the two N-fixing species lies in the fact that E. umbellata is an exotic species. Large differences in microbial community have been reported in ecosystems invaded by exotic N-fixers [80]. Several alien plant species have been shown to decrease local plant species diversity, increase ecosystem productivity, and alter the rate of nutrient cycling [81]. As a strong impact on nutrient cycling affects plant performance (e.g., plant resource allocation, plant competitive ability, plant resistance to herbivores, etc.) and, hence, community structure, Vilà et al. (2011) [27] assumed alien N-fixing plants to have greater community impacts than alien non-N-fixing species. Ge et al. (2018) [82] found that adding root extracts of the alien species Alternanthera philoxeroides (Mart.) Griseb. to the soil significantly decreased the soil microbial community activity and the relative abundance of microorganisms. Liang et al. (2016) [83] showed that allelochemicals released by invasive species might not only affect seed germination, root growth, and biomass of native plant species, but also impact soil enzymatic activities, soil microbial community, and soil functions, thus both directly and indirectly affecting the growth and development of native plants. However, further analyses are needed to clarify the reasons for the different results we obtained in the presence of a native and non-native N-fixing species.

Conclusions
Our results suggest that the afforestation of arable soils is a valid strategy for improving soil quality. Soil C, N, and enzymatic activities were all greater in the tree plantations than in the agricultural soil. As expected, the presence of N-fixing species had an impact on both the above-ground and below-ground compartments. As shown by the multivariate analysis, the non-native N-fixing Elaeagnus umbellata improved the growth of both Populus alba and Juglans regia, the target species, and supported a relatively high abundance of herbivorous nematodes. The native N-fixing species Alnus cordata, on the other hand, was associated with higher soil quality than E. umbellata. These findings suggest potential trade-offs to consider when selecting which tree species to use in mixture plantations, as some promote the growth of the target tree species while others promote soil chemical and biological quality. Nevertheless, our results support the use of mixed-tree plantations in the afforestation of former agricultural land both to restore soil quality and ensure an economic yield. We suggest that increasing tree growth and reducing the rotation length of the target species in the mixed plantation may be sufficient to offset the increased costs associated with planting and managing a mixed-species plantation.
Author Contributions: T.D. conceptualized the research design, was involved in the soil sampling and analyses, and led the writing of the manuscript. G.C. was involved in the soil sampling and analyses. G.G. was involved in the soil sampling and analyses. M.I. was involved in the soil sampling and analyses, and performed the multivariate statistical analyses. G.M. was involved in the soil sampling and analyses. F.N. was involved in the dendroecological sampling and analyses. F.P. was involved in the dendroecological sampling, contributed to field design and identification, and undertook the formal identification of the plant material used in the study. W.S.A. was involved in nematode analyses. G.B. was involved in the dendroecological sampling and analyses, and contributed to field design and identification. A.F. conceptualized the research design, and was involved in the soil sampling and analyses. All authors contributed to writing and revisions of the text. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.