Effects of tree identity dominate over tree diversity on the soil microbial community structure

Soil Biology & Biochemistry 81 (2015) 219e227

Contents lists available at ScienceDirect

Soil Biology & Biochemistry journal homepage: www.elsevier.com/locate/soilbio

Effects of tree identity dominate over tree diversity on the soil microbial community structure Andrea Scheibe a, Christina Steffens b, 1, Jasmin Seven c, Andreas Jacob d, Dietrich Hertel d, Christoph Leuschner d, Gerd Gleixner a, * €ll-Straße 10, P.O.B. 100164, 07701 Jena, Germany Max Planck Institute for Biogeochemistry, Hans-Kno €ttingen, Soil Science of Temperate and Boreal Ecosystems, Büsgen Institute, Büsgenweg 2, 37077 Go €ttingen, Germany Georg-August-University Go c €ttingen, Department of Forest Botany and Tree Physiology, Büsgen Institute, Büsgenweg 2, 37077 Go €ttingen, Germany Georg-August-University Go d €ttingen, Plant Ecology and Ecosystem Research, Albrecht-von-Haller Institute for Plant Sciences, Untere Karspüle 2, 37073 Georg-August-University Go €ttingen, Germany Go a

b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 10 June 2014 Received in revised form 17 November 2014 Accepted 19 November 2014 Available online 2 December 2014

This study investigated the possible effects of tree species diversity and identity on the soil microbial community in a species-rich temperate broad-leaved forest. For the first time, we separated the effects of tree identity and tree species diversity on the link between above and belowground communities in a near-natural forest. We established 100 tree clusters consisting of each three tree individuals represented by beech (Fagus sylvatica L.), ash (Fraxinus excelsior L.), hornbeam (Carpinus betulus L.), maple (Acer pseudoplatanus L.), or lime (Tilia spec.) at two different sites in the Hainich National Park (Thuringia, Germany). The tree clusters included one, two or three species forming a diversity gradient. We investigated the microbial community structure, using phospholipid fatty acid (PLFA) profiles, in mineral soil samples (0e10 cm) collected in the centre of each cluster. The lowest total PLFA amounts were found in the pure beech clusters (79.0 ± 23.5 nmol g
1 soil dw), the highest PLFA amounts existed in the pure ash clusters (287.3 ± 211.3 nmol g
1 soil dw). Using principle components analyses (PCA) and redundancy analyses (RDA), we found only for the variables ‘relative proportion of beech trees’ and ‘living lime fine root tips associated with ectomycorrhiza’ a significant effect on the PLFA composition. The microbial community structure was mainly determined by abiotic environmental parameters such as soil pH or clay content. The different species richness levels in the clusters did not significantly differ in their total PLFA amounts and their PLFA composition. We observed a tendency that the PLFA profiles of the microbial communities in more tree species-rich clusters were less influenced by individual PLFAs (more homogenous) than those from species-poor clusters. We concluded that tree species identity and site conditions were more important factors determining the soil microbial community structure than tree species diversity per se. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Tree species identity Tree species diversity Soil microbial community structure Variation partitioning analysis Phospholipid fatty acids

1. Introduction In recent years, the interest in investigating the linkages between above- and belowground biodiversity has steadily increased (Wardle, 2006; Cardinale et al., 2011). However, we still poorly

* Corresponding author. Tel.: þ49 (0) 3641 576172; fax: þ49 (0)3641 5770. E-mail address: [email protected] (G. Gleixner). 1 €t Hamburg, Institute of Soil Science/Center for Earth Present address: Universita System Research and Sustainability, Allende-Platz 2, D-20146 Hamburg, Germany. http://dx.doi.org/10.1016/j.soilbio.2014.11.020 0038-0717/© 2014 Elsevier Ltd. All rights reserved.

understand, whether the soil microbial communities of mixed forests are influenced to a larger extent by tree species identity or by tree species diversity (Scherer-Lorenzen et al., 2005; Prescott and Grayston, 2013). The microbial communities are influenced by the microhabitat conditions in the soil (Ranjard and Richaume, 2001; Neumann et al., 2013). Individual tree species, with their litter inputs or root activities, are able to directly or indirectly influence microbial communities through changes of the abiotic environmental variables, e.g. such as the soil pH (Eviner and Chapin, 2003). In the

220

A. Scheibe et al. / Soil Biology & Biochemistry 81 (2015) 219e227

upper soil horizons, the microbial communities depend on the decomposition of dead plant material (leaf or root litter) and/or the €ttenschwiler et al., 2005; Bais availability of rhizodepositions (Ha et al., 2006; Kramer and Gleixner, 2006; Herman et al., 2012). Tree species differ in their litter quality (e.g. C/N) and may also produce different root exudates, thereby influencing the nutrient availability (Jacob et al., 2009; Vesterdal et al., 2012; Mitchell et al., 2012a; Fender et al., 2013). Generally, the variability in litter quality and often also in nutrient availability in mixed stands due to patches of different tree species can significantly influence decomposition processes. In these stands, decay rates of individual litter types in litter-mixtures are non-additive in most cases (Wardle et al., 1997; Gartner and Cardon, 2004; H€ attenschwiler et al., 2005). Therefore, results from pure stands cannot simply be transferred to mixed stands. However, to which extent the microbial communities depend on the different environmental parame€ et al., 2010; ters (biotic and abiotic), is still under discussion (Merila Mitchell et al., 2012b). In sown grassland mixtures, higher plant diversity was found to positively influence the microbial biomass (Zak et al., 2003; Eisenhauer et al., 2010; Lange et al., 2014). These positive influences can in some cases even cause a higher microbial biomass with higher plant diversity than in the weighted average of the respective monocultures (overyielding effect), which is explained by plant complementary resource use (Eisenhauer et al., 2012; Guenay et al., 2013; Lange et al., 2014). However, from studies in grasslands, it was also suggested that soil microbial communities depend more on the presence/absence or abundance of individual plant species (“singular hypothesis”), which influence the functioning of the microbial community through specific traits, than on plant diversity per se (Porazinska et al., 2003; Eisenhauer et al., 2010). So far, it is poorly investigated, if plant species identity is also a dominant factor for the microbial community composition in forest ecosystems, as it is in synthetic grasslands. The linkages between above- and belowground biodiversity were investigated in a wide range of grassland and forest systems (Bardgett and Wardle, 2010). In a near-natural forest, Thoms et al. (2010) recently observed an increase in the microbial biomass and a shift in the microbial community structure with increasing tree diversity. However, it was not possible to clearly differentiate if the observed changes in the microbial community were caused by higher tree species diversity or by decreasing abundance of one important tree species, European beech. We designed a comparative field study as a follow-up of the study of Thoms et al. (2010) aimed to answering the question, whether tree species diversity or rather tree species identity had a more important effect on microbial community structure. We further wanted to elucidate, which environmental (biotic and/or abiotic) variables are the most important drivers of the microbial community composition. We applied a tree cluster approach to investigate the effect of tree species identity and diversity on the soil microbial community structure in a near-natural deciduous forest (Hainich National Park, Germany). We selected 100 tree clusters consisting of each three mature trees of variable species identity (Leuschner et al., 2009). The trees belonged to the six most common tree species in this forest (Fagus sylvatica L., Fraxinus excelsior L., Carpinus betulus L., Acer pseudoplatanus L. and Tilia cordata Mill. or T. platyphyllos Scop.), which are known to differ in litter quality (Jacob et al., 2009, 2010a). The clusters were divided into three species richness levels (SR) with clusters consisting of one (SR 1), two (SR 2) or three (SR 3) tree species. The soil microbial community structure was investigated using phospholipid fatty acid (PLFA) profiles. We hypothesized that the microbial biomass increases with increasing a) tree species diversity and b) litter quality.

Furthermore we hypothesized that c) the microbial community structure changes with increasing tree species diversity and d) the microbial community in the upper mineral soil horizon (0e10 cm) is mainly determined by biotic factors (e.g. tree species composition or litter C/N ratio). 2. Material and methods 2.1. Sampling site The study was performed in the Hainich National Park (Thuringia, Germany). The National Park protects the largest continuous temperate mixed broad-leaved forest in central Germany (ca. 7500 ha). Due to historical forest utilization regimes, a small-scale mosaic of forest patches developed in which a maximum tree €lder et al., species diversity is as high as 14 tree species per ha (Mo 2006; Leuschner et al., 2009). The forest at the study sites is an oldgrowth forest, which has existed for over 200 years (Mund, 2004). Large parts of the Hainich forest were used as military training area since 1960, until the southern part of the Hainich became a National Park in 1997. Over these last 50 years, the forest was able to grow nearly undisturbed and is now very close to the original deciduous forests of Central Europe (Mund, 2004; Gleixner et al., 2009). The stem density (>7 cm dbh ha
1) ranged from 291 ± 69 to 581 ± 101 in the study area (Jacob et al., 2010b). The mean annual temperature is 7.7  C and the mean annual precipitation is 590 mm (period 1973e2004, station Weberstedt/Hainich; 51060 N, 10 310 E, 270 m a.s.l., Deutscher Wetterdienst 2005). The forest grows on a Luvisol developed on loess over Triassic limestone (Guckland et al., 2009). 2.2. Experimental setup For this study, 100 tree clusters were randomly selected in the study area. The clusters consisted of three individual trees standing in a triangle. Each tree was represented by one of the six main tree species of the National Park: beech (Fagus sylvatica L.), ash (Fraxinus excelsior L.), hornbeam (Carpinus betulus L.), maple (Acer pseudoplatanus L.) and lime (Tilia cordata Mill. and/or T. platyphyllos Scop.), whereby we considered the two Tilia species collectively in the further data analyses (Leuschner et al., 2009). All five tree genera not only differ in their leaf litter quality (see Table A1 in the Supplementary Material), but also in their mycorrhizal symbiont association type (ecto-vs. arbuscular mycorrhizal). Trees of beech and hornbeam are associated with ectomycorrhizal symbionts, whereas trees of ash and maple are associated with arbuscular mycorrhizal symbionts. Lime trees belong to the few tree genera, which can be associated by both ecto- and arbuscular mycorrhizal fungi (Pigott, 1991; Timonen and Kauppinen, 2008). We established three species richness levels (SR) in this field study. The SR 1 clusters consisted of only one tree species, SR 2 clusters of two (2:1 or 1:2) and SR 3 clusters of three different tree species. We randomly chose 25 variants of tree combinations with four replicates of each combination. Two replicates were localized at the Thiemsburg (n ¼ 50) and two at the Lindig site (n ¼ 50). The cluster trees were mature trees with a mean breast height diameter of 0.43e0.46 m and of 80e160 years of age, which formed a closed canopy (Jacob et al., 2013, 2014). No significant differences between the three SR level of the clusters were found for mineral soil (0e10 cm) parameters, such as pH, C/N ratio or clay content (see Table A2 in the Supplementary Material). The soil texture of the mineral soil varied between a silty loam to silty clayey loam characterized by a high (~74%) silt content and low (<5%) sand content (Guckland et al., 2009).

A. Scheibe et al. / Soil Biology & Biochemistry 81 (2015) 219e227

2.3. Sample collection, preparation and analysis for microbial community structure To investigate the microbial community structure soil samples were collected in May 2008. A mineral soil sample of 0e10 cm (Ø 6.4 cm) was taken in the center of each cluster. The litter and O horizon were directly removed in the field. The soil samples were kept on ice, transferred to the lab and kept frozen (
20  C) until phospholipid fatty acid (PLFA) extraction. Soil samples were defrosted at 4  C overnight and sieved (Ø 2 mm). Animals, seeds, roots and plant material were carefully removed. To determine the water content, a subsample of 5 g was dried at 105  C for 24 h. For PLFA extraction, ~20 g (dw) of mineral soil was used, following a modified method after Bligh and Dyer (1959), Zelles and Bai (1993) and Kramer and Gleixner (2006). The whole lipid fraction was extracted with a mixture of methanol, chloroform and phosphate buffer (2:1:0.8 v/v/v). Phospholipids were isolated using a silica column (Mega BE-SI, Agilent Technologies, Waldbronn, Germany). Fatty acid methyl esters (FAMEs) were isolated after a mild alkaline hydrolysis, a methylation using a methanolic KOH solution and a NH2 column (Chromabond NH2, Machery-Nagel GmbH & Co.KG, Düren, Germany). Individual FAMEs were quantified with a gas chromatograph (GC) using a flame ionization detector (GC-FID, HP 6890 Series, Agilent Technologies, Waldbronn, Germany). The nonadecanoic methylester (C19:0) was used as an internal standard for the GC measurements. The FAMEs were separated using a HP-Ultra2 (50 m  0.32 I.D., 0.25 mm film thickness, Agilent Technologies, Waldbronn, Germany) column and He as carrier gas with a flow rate of 2 ml min
1. Samples were measured three times with 1 ml of sample injection. For separation the temperature program started at 140  C for 1 min and heated up to 270  C (for 6 min) with an increase of 2  C min
1. With a rate of 30  C min
1, the final temperature of 320  C was reached, which was kept for 3 min. The identification of individual FAMEs was done using a gas chromatograph (GC 5890 series II, Agilent Technologies, Waldbronn, Germany) with a GCQ ion trap mass spectrometer (Thermo-Fischer, Bremen, Germany) and an identical column and temperature program as described above. Individual FAMEs were identified by their mass spectra in comparison to spectra of standard mixtures of saturated and unsaturated fatty acids (Supelco, Bellefonte, USA) and by using an internal mass spectral data set (Thoms et al., 2010). In total, we were able to identify up to 56 different PLFAs per sample. PLFAs can be used to investigate the actual living soil microbial community (White et al., 1996; Bossio et al., 1998; Kaur et al., 2005; Frostegård et al., 2011). Specific PLFAs can be applied as markers for different microbial groups (Vestal and White, 1989; Zelles, 1999). According to the literature, we considered the PLFAs 17:0cy, 19:0cy, 16:1u7 and 18:1u7 as biomarkers for Gram-negative bacteria, whereas Gram-positive bacteria are often represented by the PLFAs 15:0i/a and 17:0i/a (O'Leary and Wilkinson, 1988; White et al., 1996; Zelles, 1999). As a biomarker for fungi, the 18:2u6,9 PLFA is used (Frostegård and Bååth, 1996; Olsson, 1999; Bååth and Anderson, 2003). The PLFA 16:1u5 is widely applied as a marker for arbuscular mycorrhizal (AM) fungi, although it can also occur in high amounts in Gram-negative bacteria (Nordby et al., 1981; Nichols et al., 1986; White et al., 1996; Olsson, 1999). 2.4. Data analysis The program SPSS (PAWS Statistics 18) was used for statistical analyses. The data were tested for normality using the ShapiroeWilk test. We applied a log-transformation of the data to

221

achieve a normal distribution. To investigate significant differences across the different SR 1 clusters (between the different tree species) and the three species richness levels we applied a oneway ANOVA followed by Tukey's HSD post hoc test. To compare the clusters with different mycorrhiza types (ecto-vs. arbuscular mycorrhiza) and the two study sites (Thiemsburg vs. Lindig) we applied a t-test. With Spearman's rank correlation coefficients (rho), we investigated the relationship between total PLFA amounts and litter or physical soil parameters across the different clusters. For multivariate statistics, we utilized Canoco 5.0 software (Microcomputer Power, Ithaca, NY, USA). For all multivariate analyses, we used the relative proportions (mol%; log transformed) of the 22 most dominant individual PLFAs in the analysis. To describe the similarity or dissimilarity in the PLFA composition between the different tree clusters we used the principle components analyses (PCA). With a partial redundancy analyses (RDA), we were able to summarize this part of the variability in the PLFA composition, which can be explained by a linear combination of environmental variables, after removing the effects that could be explained by a covariable ‘site’ (the two different sites Thiemsburg and Lindig). In these analyses, statistical significant environmental variables were determined using a stepwise forward selection after a Monte Carlo permutation test. For the partial RDA we included different biotic and abiotic variables. As biotic environmental factors we included the following variables directly linked to the tree species: the amount of humus (L or L þ Of) in the forest floor; for each tree species the relative proportion of their leaf litter in a cluster, living and dead fine root mass and the relative proportion of root tips infected with ectomycorrhiza (in case of beech, hornbeam and lime). As abiotic environmental factors, we included parameters of the mineral soil (0e10 cm): bulk density, soil moisture, soil pH, clay content, C/N ratio, concentrations of Corg (organic carbon) and Nt (total nitrogen), the salt-exchangeable concentrations of various cations (Naþ, Kþ, Ca2þ, Mg2þ, Fe2þ, Mn2þ) and cation exchange capacity (CEC; for further methodological details see A1 in the Supplementary Material). For RDA the individual assignment to abiotic or biotic factors was not relevant. In the ordination diagrams, each PLFA and environmental variable points in the direction of the steepest increase starting from the origin. We conducted a variation partitioning analysis using Canoco 5.0 software to explore, which parts of the variation in the soil microbial community structure were explicable by biotic and/or abiotic environmental parameters (unique and simple effects; Legendre, 2008). We used the same environmental parameters as included in the partial RDA and assigned them into two groups: biotic and abiotic (plus site) environmental parameters (as described above). To test for simple effects (groups as explanatory variables, but never as covariables) we conducted a variation partitioning analysis with stepwise forward selection of the group members after a Monte Carlo permutation test. The same group members resulting from the stepwise forward selection were used in a second variation partitioning analysis to test for conditional (unique) effects. This analysis includes three partial RDAs. In the first two partial RDAs, the amount of variability explained by one group was calculated using the other group as covariable (abiotic group as explanatory variable and biotic group as covariable and vice versa). In the third RDA, the amount of variability was calculated, which was explicable together by both groups. More details about the variation partitioning method in Canoco can be found in  ter Braak and Smilauer (2012). The significance level was set to P  0.05.

Total PLFA amounts [nmol PLFA/g soil (dw)]

222

A. Scheibe et al. / Soil Biology & Biochemistry 81 (2015) 219e227

600 b

500

400 b 300 ab ab

200

100

a

0 Beech

Hornbeam

Lime

Maple

Ash

SR1 clusters Fig. 1. Total amounts of PLFAs (mean values ± s.d.) detected in the species richness level 1 (SR 1) clusters, which consisted of only one tree species (n ¼ 4). Different lower case letters indicate significant differences between the cluster types (P < 0.05, oneway ANOVA followed by Tukey's HSD post hoc test). Tree species are arranged by increasing litter quality (decreasing C/N ratio).

a decreasing abundance of beech trees (P ¼ 0.057; Fig. 3a). The opposite was found for clusters containing maple or ash trees (Figs. 3d and e). The clusters containing hornbeam showed the highest total PLFA amounts in SR 2 clusters, whereas limecontaining clusters showed the lowest total PLFA amounts (Figs. 3b and c). Overall, in the five types of monospecific SR 1 clusters, the total PLFA amounts showed a significant positive relation to Nt and a significant negative one to C/N ratio and Mn concentration of the litter produced by the cluster-building tree species (see Table A3 in the Supplementary Materials). A significant positive correlation was found between the total PLFA amounts and the P concentrations of the litter mixtures in the SR 1 clusters. The total PLFA amounts were also significantly positively correlated with the Ca, Mg and Fe concentrations for both the individual litter types and the litter mixtures in the clusters. The Nt, Corg, pH, clay content and CEC of the mineral soil, were significantly positively correlated with total PLFA amounts in the SR 1 clusters and for all clusters. A significantly negative correlation was found between total PLFA amounts and the C/N ratio and bulk density of the mineral soil, the amount of humus (L or L þ Of), and the relative proportion of beech litter in the total leaf litter mass of a cluster. 3.2. Microbial community structure

3. Results 3.1. Total microbial biomass The total PLFA amounts over all clusters ranged from 28.1 to 593.2 nmol PLFAs g
1 soil (dw). In the species richness level (SR) 1 clusters the lowest PLFA amounts were detected in beech clusters, followed by hornbeam and lime clusters (Fig. 1). Maple and ash SR 1 clusters yielded significantly higher PLFA amounts in comparison to the SR 1 beech clusters. The total PLFA amounts did not significantly differ between the three species richness levels (SR 1e3) in the tree clusters (P ¼ 0.342, with n ¼ 20 for SR 1 and n ¼ 40 for SR 2 and SR 3; Fig. 2). Moreover, the three species richness levels for the individual tree species were not significantly different from each other when all clusters with presence of a given target species were compared among the five species (Fig. 3aee). By trend, the total PLFA amounts increased with

Total amounts of PLFA [nmol/g soil (dw)]

350 300 250 200 150 100 50 0 SR1

SR2

SR3

Species richness levels Fig. 2. Calculated total amounts of PLFAs (mean values ± s.d.) for the different species richness levels (SR; P ¼ 0.342, one-way ANOVA with SR 1: n ¼ 20, SR 2 and 3: n ¼ 40).

3.2.1. Effect of tree identity and diversity on the microbial community For the PCA and RDA, we used the 22 most dominant PLFAs, which represented on average 87.3 (±0.9) % of the total PLFA amounts. The first two principle components (PC) explained approximately 70% of the variation in the PLFA composition, with 54.6% explained by the first and 16.7% explained by the second PC (Fig. 4). Along the first PC, the PLFAs 14:0i, 15:0a and 17:0a showed strong positive loadings in comparison to the PLFAs 15:0i, 16:0 and 18:0 with strong negative loadings. The loadings of the individual SR 1 clusters along the first PC were significantly different in their PLFA composition between the pure beech tree clusters and the pure maple, ash and lime clusters (P < 0.01, F4,15 ¼ 9.6, n ¼ 4). Except for the pure beech clusters, the loadings of all monospecific cluster types overlapped in accordance with the leaf litter quality parameters (see Table A1 in the Supplementary Materials). We also found a clear separation (P < 0.001, n ¼ 12) of the tree species clusters with ectomycorrhiza (beech and/or hornbeam) from the tree species clusters with arbuscular mycorrhiza (ash and/or maple) along the first PC. Along the second PC, the PLFAs 16:1u7, 18:2u6,9 and 16:1u5 showed strong negative loadings in comparison to the PLFAs 17:0i and 10Me17:0 with strong positive loadings. The SR 1 clusters or clusters with the different mycorrhiza types (ecto- or arbuscular mycorrhiza) did not significantly differ along the second PC. No significant differences were found between the species richness levels of the clusters, neither for the first nor for the second (and third) PC. We observed a slight tendency that with increasing species richness level the vector length of the individual PLFAs decreased and that the resulting mean vectors were smaller (Fig. A1), which suggests that the total PLFA composition was the less influenced by individual PLFAs. 3.2.2. Effects of environmental factors on the microbial community The first two axes of the RDA biplot explained approximately 55% of the variability in the PLFA composition between the clusters (Fig. 5). Soil pH (pseudo-F ¼ 69.4, P < 0.01) and clay content (pseudo-F ¼ 17.1, P < 0.01) in the upper mineral soil (0e10 cm) were the most significant environmental variables after a Monte Carlo

A. Scheibe et al. / Soil Biology & Biochemistry 81 (2015) 219e227

500

a)

Beech

b)

223

Hornbeam

400 300 200

Total PLFA amounts [nmol / g soil (dw)]

100 0 500

SR1

SR2

SR3

c)

SR1 Lime

SR2

d)

SR3 Maple

400 300 200 100 0 SR1 500

SR2

SR3

e)

SR1 Ash

SR2

SR3

Species richness levels

400 300 200 100 0 SR1

SR2

SR3

Species richness levels Fig. 3. Determined total amounts of PLFAs (mean values ± s.d.) at the different species richness levels for the individual tree species: a) beech, b) hornbeam, c) lime, d) maple and e) ash. With increasing species richness levels (one, two or three tree species per cluster) the number of trees of a species in the tree clusters decreases with all three trees belonging to the target species in SR 1 (n ¼ 4), two or one trees in SR 2 (n ¼ 16) and only one tree in SR 3 (n ¼ 24) (P > 0.05, one-way ANOVA).

permutation test with forward selection. The salt-exchangeable Mg2þ concentration (pseudo-F ¼ 5.0, P < 0.05) and the C/N ratio (pseudo-F ¼ 4.2, P < 0.05) of the mineral soil (0e10 cm) also significantly influenced the PLFA composition in the clusters. The relative proportion of beech trees (pseudo-F ¼ 3.5, P < 0.05) and the relative proportion of living lime root tips with ectomycorrhiza colonization (pseudo-F ¼ 3.5, P < 0.05) in the mineral soil (0e20 cm) were the only investigated biotic environmental factors with a significant influence. The PLFAs 15:0a and 17:0a showed a strong positive correlation to the clay content of the clusters, whereas in the opposite direction, the PLFA 18:2u6,9 was highly correlated with the soil C/N ratio (Fig. 5). We found a strong negative correlation between the PLFA 19:0cy and soil pH, in contrast to the PLFA 18:1u7, which was positively correlated. The PLFAs 15:0i, 16:0, 18:0 and 18:1u9 were highly associated with the relative proportion of beech trees in a cluster. In accordance to the PCA, we found a significant separation of the pure beech cluster (SR 1) loadings from all four other tree species (P < 0.05, F4,15 ¼ 8.6, n ¼ 4) and significant differences between the loadings of the tree clusters with different mycorrhiza types (ecto- or arbuscular mycorrhiza; P < 0.001, n ¼ 11) along the first RDA axis, but no significant differences along the second axis.

We also found no significant differences between the different species richness levels along the first and second RDA axes, but a similar tendency of a shorter mean vector length with increasing species richness level, indicating a decreasing effect of specific environmental factors (biotic or abiotic) on the microbial community composition. The applied variation partitioning analyses revealed that 38.5% of the total explained variability in the PLFA compositions, regarding the conditional effects, referred to abiotic factors, whereas only 4.6% of the total explained variability was attributable to biotic factors. We found that the two groups represented a highly significant contribution to explanatory power both in their simple and in their conditional effects (P ¼ 0.002 for both groups). Most of the explained variability was referred to an overlap between the two groups, which means that abiotic and biotic environmental factors jointly described 56.9% of the total explained variability. 4. Discussion 4.1. Effects of tree species diversity on the microbial community In synthetic grasslands, plant species diversity seemed to be an important factor stimulating the microbial biomass and influencing

224

A. Scheibe et al. / Soil Biology & Biochemistry 81 (2015) 219e227

1

1 17:0i

10Me17:0

16:0i

17:0i

19:0cy

10Me17:0

16:0i 10Me16:0

17:1

17:0a 15:0a

14:0i 18:0

10Me18:0

18:1ω5 16:0

18:1ω9

clay**

19:0cy

17:0cy

RDA Axis 2 (7.1 %)

PCA Axis 2 (16.7 %)

15:0i

17:0a 15:0a [Mg]*

15:0i

17:0cy

Beech*

18:1ω5

18:1ω9 18:0 16:0

14:0i

C/N* 18:2ω6,9

16:1ω5

pH** 16:1ω11

16:1ω5 16:1ω7

10Me16:0

10Me18:0 18:1ω7

18:1ω7 18:2ω6,9

17:1

Myc_L* 16:1ω7 16:1

16:1ω11 16:1

-1

-1 -1

1

PCA Axis 1 (54.6 %) PLFAs Maple Beech

Ash Hornbeam

-1

-1

RDA Axis 1 (47.4 %) Lime

Fig. 4. Biplot of a principle components analysis (PCA) showing the loadings of the individual phospholipid fatty acids (PLFAs; blue lines) and the cluster points (mean values ± s.d., n ¼ 4) of the species richness level 1 for the five tree species along the first two principle component axes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

the microbial community structure (Scherber et al., 2010; Reich et al., 2012). In contrast to these investigations and in agreement with an investigation in boreal forests (Lucas-Borja et al., 2012) we found no tree diversity effects on the microbial community (total PLFA amounts and PLFA composition) in the mineral soil of the studied deciduous forest (Figs. 2 and 3). In sown grasslands it was found that higher plant diversity can positively influences the microbial biomass, especially due to longterm effects of plant species facilitating decomposers or arbuscular mycorrhiza fungi (Eisenhauer et al., 2012; Guenay et al., 2013). However, we found no positive influencing effects of the forest tree diversity on the soil microbial biomass. The total PLFA amounts in the mixed clusters (SR 2 and SR 3) were overall significantly (P < 0.001; n ¼ 80) lower than theoretical amounts calculated from the total PLFA amounts in the SR 1 clusters (see A2 and Figs. A2 and A3 in the Supplementary Material). The measured total PLFA amounts in the mixed clusters represented approximately 83% of the theoretically calculated amounts. However, our results are in line with studies that investigated the above-ground biomass production (leaves, fruits, and wood) and the fine roots (<2 mm diameter) in the Hainich forest (Meinen et al., 2009; Jacob et al., 2010a, 2013, 2014). With the rates of leaf litter and root litter supply not being systematically different between monospecific and mixed plots in the Hainich forest, the microbial communities receive similar amounts of organic material as substrates. Due to the strong dependence of the microbial communities on the substrate supply, a significant increase in the microbial communities with increasing tree diversity is unlikely in this forest. We also could not find significant differences in the PLFA composition between the species richness levels of the clusters. By tendency, the average PLFA composition was more evenly

PLFAs Environmental Variables Fig. 5. Biplot of a partial redundancy analysis (RDA) representing the loadings of the individual PLFAs and the six significant environmental variables determined after a Monte Carlo permutation test with forward selection. Significance is marked with asterisk (**P < 0.01, *P < 0.05). Acronyms: Beech ¼ relative proportion of beech trees [%] at a cluster; clay ¼ clay content [%] of the mineral soil (0e10 cm); [Mg] ¼ exchangeable Mg2þ concentration [mmolc kg
1 (dw)] in the mineral soil (0e10 cm); pH ¼ pH after KCl extraction in the mineral soil (0e10 cm); C/N ¼ C/N of the mineral soil (0e10 cm); Myc_L ¼ % of living lime root tips with ectomycorrhiza colonization in the mineral soil (0e20 cm).

distributed with increasing tree species richness, i.e. the PLFA composition was less influenced by individual PLFAs and specific environmental factors. This was indicated in the PCA and RDA by a (non-significant) decrease of the mean vector length with tree species richness (Fig. A1). One possible interpretation might be that with higher tree diversity the microbial community is less influenced by specific interactions of specific phyla/classes like AMF formation. This could be interpreted as “selection effect” or as support of the “insurance hypothesis”. However, future studies in near-natural forests with complex structures in combination with research in synthetic tree stands are necessary to confirm these assumptions (Leuschner et al., 2009). To conclude, in contrast to our first and third hypotheses tree species diversity does not significantly influence the amount and composition of the soil microbial community. 4.2. Tree identity effects on the microbial community In agreement with our second hypothesis, we suggest that higher abundances of soil microorganisms may partly depend on the tree species present, which likely is caused by a higher nutrient availability under trees with higher leaf litter quality (e.g. with decreasing C/N ratio). The five monospecific cluster types of SR 1 differed among each other significantly in their litter quality (see Table A1 in the Supplementary Material). In accordance with the literature, the litter quality of the five tree species was lowest in beech, followed by hornbeam, and was highest in maple, ash and lime (Jacob et al., 2010a; Langenbruch et al., 2012; Vesterdal et al., 2012). The PLFA

A. Scheibe et al. / Soil Biology & Biochemistry 81 (2015) 219e227

amounts increased with Nt of the leaf litter and also with the P concentration in the litter mixtures in the clusters, which suggests a marked tree species effect on the microbial communities (see Table A3 in the Supplementary Material). According to the low soil P concentrations and the small P fluxes with leaf litterfall (Talkner et al., 2009), P seems to be a limiting element or of low availability for the tree species in the Hainich forest. Here, the interaction between the tree species and their associated mycorrhizal partners might be important. As mycorrhizal fungi differ in their efficiency to acquire P, tree species have different advantages in periods of nutrient (P) shortage (Leake et al., 2004; van der Heijden et al., 2008; Courty et al., 2010). In the Hainich forest, it was already observed that the various tree species significantly differ in their associated mycorrhizal community (Lang et al., 2011). In our PCA analysis, we further found that specific PLFA mixtures existed under the different tree species, which clearly supports tree species-specific effects on the microbial community structure (Fig. 4). This also provides evidence that the “singular hypothesis” is also valid in forests. Long-term plantemicrobe interactions could promote the development of specific decomposer communities in adaptation to the microhabitat structure and nutrient availability (Vivanco and Austin, 2008). The cluster loadings for the different SR 1 cluster types in the PCA (Fig. 4) and RDA showed a clear separation between beech and the other four tree species. As fungal biomarker, the PLFA 18:2u6,9 indicated higher abundances of fungi (or EM mycelium) below beech trees in comparison to the other tree species. These higher abundances are probably related to a stronger participation of fungi in the decomposition of the more recalcitrant beech leaf litter. We also found significant differences in the PLFA composition of clusters with ectomycorrhizal tree species (beech and/or hornbeam) and arbuscular mycorrhizal tree species (ash and/or maple) indicating differences in microbial community composition. However and in accordance to Thoms and Gleixner (2013), the PLFA 16:1u5 often used as a biomarker for arbuscular mycorrhizal (AM) fungi was highly correlated to the abundance of lime trees, even though the roots of this genus show predominantly ectomycorrhizal colonization (Lang et al., 2011). Also in the PCA and RDA, the PLFA 16:1u5 was only weakly correlated with clusters containing arbuscular mycorrhizal tree species (data not shown). It is known that Gram-negative bacteria can also contain high amounts of the PLFA 16:1u5 (Nichols et al., 1986). A high proportion of Gramnegative bacteria below lime trees, as displayed by a high correlation of the signature PLFAs 16:1u7 and 18:1u7 with lime (Fig. 4), might have strongly influenced the abundance of PLFA 16:1u5 (Olsson, 1999; Ngosong et al., 2012). Therefore, we assume that the PLFA 16:1u5 in our investigation was mainly produced by a specific bacterial community adapted to the mycorrhizosphere of lime trees rather than indicating the abundance of AM fungi. This suggests that for identifying AM fungi in samples from species-rich forests, the neutral-lipid fatty acid (NLFA) 16:1u5 would be a better choice than the PLFA 16:1u5 (Herman et al., 2012; Ngosong et al., 2012). In their study, Thoms et al. (2010) were unable to clearly separate between a tree species diversity effect and the influence of variable abundance of beech, when investigating the microbial communities in plots with either 1, 3, or 5 tree species in the Hainich forest. Our study with a large number of small plots was designed as a follow-up of this investigation and allowed to separate between these two effects. We show that the abundance of a specific tree species (e.g. beech see Fig. A4 in the Supplementary Material) in a cluster is more important for the microbial community structure as reflected in the PLFA amounts and composition than tree diversity per se. This corroborates the assumption of Thoms et al. (2010, 2013) that their observed trend of increasing PLFA amounts with increasing tree diversity and associated shifts in

225

PLFA composition (especially in early-summer samples) was caused by decreasing beech abundance rather than by increasing tree diversity. However, the results of Thoms et al. (2010) also suggest that the observed species-specific effects were partly related to specific edaphic conditions in the plots, which demonstrate the necessity to disentangle the influence of biotic and abiotic environmental factors on the PLFA composition in forest soils. 4.3. Biotic versus abiotic environmental parameters influencing the PLFAs In contradiction to our fourth hypothesis, we did not find that the microbial community in the upper soil horizon was mainly determined by direct effects of biotic factors. Instead, we found that edaphic parameters directly explained most of the variability in the PLFA compositions among the different clusters. In accordance to the literature, the RDA revealed that the PLFA composition was mainly driven by abiotic environmental parameters such as soil pH and clay content of the mineral soil (Fig. 5; Hackl et al., 2005; Thoms et al., 2010). In general, after soil pH the clay content is an important determinant of soil nutrient stocks due to its effect on CEC and the exchangeable pools of base cations and its association with the organic carbon content in the soil (Guckland et al., 2009; Langenbruch et al., 2012). Therefore, clay content directly influences the microbial community in the soil. The variation partitioning analyses demonstrated that ~39% of the total explained variability in the soil microbial community structure was related to variations of abiotic environmental factors. Biotic factors also impacted the PLFA composition, like the proportion of beech and the infection with ectomycorrhiza, however this effect was very small (~5% of the total explained variability). This indicates that soil-related parameters have much stronger direct effects on the formation of the microbial community structure than direct effects of plant-related parameters. In contrast to our results, Mitchell et al. (2012b) observed that variation in plant community composition explained nearly the same or even higher amounts of the variability in the microbial community as did variation in soil chemistry when comparing contrasting moorland and deciduous woodland ecosystems in Northern Scotland. Nevertheless, Mitchell et al. (2012b) and our study agree in that changes in the soil microbial communities in certain parts are related to changes in both abiotic and biotic variables, although the part directly related to changes in plant attributes was in our study very small. However, with the variation partitioning analyses we also found a strong overlap between influencing abiotic and biotic variables that jointly describe ~57% of the total explained variability. This part of explained variability cannot be assigned to plant- or soil-related parameters alone, but indicates a strong interrelationship between influencing plant and soil properties (Meril€ a et al., 2010; Mitchell et al., 2012b). The large overlap of ~57% might stand for the indirect effects of the tree species on the soil (Eviner and Chapin, 2003; Langenbruch et al., 2012). As for the litter quality parameters, we detected significant differences between the soil properties of the different monospecific SR 1 clusters (Table A3). The most important factor influencing the PLFA composition in the clusters was the soil pH (Fig. 5). In this context, it is likely that beech trees are able to cause specific changes in the microbial community structure due to the nature of their root exudates, which decreased the soil pH (Cesarz et al., 2013; Fender et al., 2013). In our investigation, the effect of pH was mainly expressed in the ratio of the PLFAs 19:0cy and 18:1u7, which are partially used to calculate stress ratios (Bååth and Anderson, 2003; Kaur et al., 2005) (see Fig. A5 in the Supplementary Material). However, based on our investigations it's not possible to separate effects of pH induced physiological

226

A. Scheibe et al. / Soil Biology & Biochemistry 81 (2015) 219e227

stress and possible effects on the membrane composition from shifts in the community structure. The pure beech clusters were further characterized by the highest soil C/N ratios and the lowest soil pH of all five tree species. The relatively high soil C/N ratios in the pure beech clusters are probably resulting from the high C/N ratio of beech leaf litter, which exceeded the ratio of all other species in our study. Leaf litter can act as constant long-term C and nutrient source in soils and strongly impact on the microbial community as this was already found for the bacterial community composition in the rhizosphere of ash and beech (Pfeiffer et al., 2013). Nevertheless, this implies that the different tree species likely are capable of influencing the nutrient status and the microhabitat structure in the soil in specific ways (Binkley and Giardina, 1998). Largest tree effects are to be expected in the top soil through species-specific differences in litter quality and/or root exudates (Guckland et al., 2009; Jacob et al., 2009; Langenbruch et al., 2012; Cesarz et al., 2013; Pfeiffer et al., 2013). To summarize, variation in abiotic environmental factors directly explained most of the variability in the PLFA composition in the Hainich forest, but these factors were themselves largely under the influence of the different tree species and their specific traits; the only exception is soil clay content. In other words, tree speciesrelated biotic factors do not have a large direct influence on microbial community composition, but they influence PLFA variability mainly indirectly through their effects on soil pH, soil C/N ratio, nutrient availability and other edaphic properties. Disentangling these indirect interrelationships will increase our understanding of the interactions between the plant and soil microbial communities. 5. Conclusion Total PLFA amounts and the structure of the microbial communities in the top soil covaried with the tree species composition of the forest stands and the specific functional traits of the tree species (notably the C/N ratio of leaf litter). Tree identity effects clearly dominated over tree diversity effects on the composition of the microbial community. However, abiotic environmental factors such as soil pH had the largest effect on the microbial community structure, although these parameters (except clay content) are in large parts indirectly influenced by tree species identity. Before the results of our case study can be generalized, further investigations are required applying the tree cluster approach in other forest types, including conifer tree species as well as mixtures of conifer and broad-leaved forest stands. Acknowledgement The authors would like to thank the German Research Foundation (DFG) for financial support of the Graduiertenkolleg (GRK) 1086. We are grateful to the National Park administration for granting permission to conduct the research in the Hainich National Park. Thanks to Carolin Thoms for the introduction into the PLFA extraction method. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.soilbio.2014.11.020. References Bååth, E., Anderson, T.-H., 2003. Comparison of soil fungal/bacterial ratios in a pH gradient using physiological and PLFA-based techniques. Soil Biology & Biochemistry 35, 955e963.

Bais, H.P., Weir, T.L., Perry, L.G., Gilroy, S., Vivanco, J.M., 2006. The role of root exudates in rhizosphere interactions with plants and other organisms. Annual Review of Plant Biology 57, 233e266. Bardgett, R.D., Wardle, D.A., 2010. Aboveground-belowground Linkages: Biotic Interactions, Ecosystem Processes, and Global Change. Oxford University Press, Oxford. Binkley, D., Giardina, C., 1998. Why do tree species affect soils? The warp and woof of tree-soil interactions. Biogeochemistry 42, 89e106. Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid extraction and purification. Canadian Journal of Biochemistry and Physiology 37, 911e917. Bossio, D.A., Scow, K.M., Gunapala, N., Graham, K.J., 1998. Determinants of soil microbial communities: effects of agricultural management, season, and soil type on phospholipid fatty acid profiles. Microbial Ecology 36, 1e12. Cardinale, B.J., Matulich, K.L., Hooper, D.U., Byrnes, J.E., Duffy, E., Gamfeldt, L., Balvanera, P., O'Connor, M.I., Gonzalez, A., 2011. The functional role of producer diversity in ecosystems. American Journal of Botany 98, 572e592. Cesarz, S., Fender, A.-C., Beyer, F., Valtanen, K., Pfeiffer, B., Gansert, D., Hertel, D., Polle, A., Daniel, R., Leuschner, C., Scheu, S., 2013. Roots from beech (Fagus sylvatica L.) and ash (Fraxinus excelsior L.) differentially affect soil microorganisms and carbon dynamics. Soil Biology & Biochemistry 61, 23e32. e, M., Diedhiou, A.G., Frey-Klett, P., Le Tacon, F., Rineau, F., Courty, P.-E., Bue Turpault, M.-P., Uroz, S., Garbaye, J., 2010. The role of ectomycorrhizal communities in forest ecosystem processes: new perspectives and emerging concepts. Soil Biology & Biochemistry 42, 679e698. Eisenhauer, N., Reich, P.B., Scheu, S., 2012. Increasing plant diversity effects on productivity with time due to delayed soil biota effects on plants. Basic and Applied Ecology 13, 571e578. Eisenhauer, N., Beßler, H., Engels, C., Gleixner, G., Habekost, M., Milcu, A., Partsch, S., Sabais, A.C.W., Scherber, C., Steinbeiss, S., Weigelt, A., Weisser, W.W., Scheu, S., 2010. Plant diversity effects on soil microorganisms support the singular hypothesis. Ecology 91, 485e496. Eviner, V.T., Chapin, F.S., 2003. Functional matrix: a conceptual framework for predicting multiple plant effects on ecosystem processes. Annual Review of Ecology, Evolution, and Systematics 34, 455e485. Fender, A.-C., Gansert, D., Jungkunst, H.F., Fiedler, S., Beyer, F., Schützenmeister, K., Thiele, B., Valtanen, K., Polle, A., Leuschner, C., 2013. Root-induced tree species effects on the source/sink strength for greenhouse gases (CH4, N2O and CO2) of a temperate deciduous forest soil. Soil Biology & Biochemistry 57, 587e597. Frostegård, Å., Tunlid, A., Bååth, E., 2011. Use and misuse of PLFA measurements in soils. Soil Biology & Biochemistry 43, 1621e1625. Frostegård, Å., Bååth, E., 1996. The use of phospholipid fatty acid analysis to estimate bacterial and fungal biomass in soil. Biology and Fertility of Soils 22, 59e65. Gartner, T.B., Cardon, Z.G., 2004. Decomposition dynamics in mixed-species leaf litter. Oikos 104, 230e246. Gleixner, G., Tefs, C., Jordan, A., Hammer, M., Wirth, C., Nueske, A., Telz, A., Schmidt, U.E., Glatzel, S., 2009. Soil carbon accumulation in old-growth forests. In: Wirth, C., Gleixner, G., Heimann, M. (Eds.), Old-growth Forests: Function, Fate and Value. Springer, New York, pp. 231e266. Guckland, A., Jacob, M., Flessa, H., Thomas, F.M., Leuschner, C., 2009. Acidity, nutrient stocks, and organic-matter content in soils of a temperate deciduous forest with different abundance of European beech (Fagus sylvatica L.). Journal of Plant Nutrition and Soil Science e Zeitschrift für Pflanzenern€ ahrung und Bodenkunde 172, 500e511. Guenay, Y., Ebeling, A., Steinauer, K., Weisser, W.W., Eisenhauer, N., 2013. Transgressive overyielding of soil microbial biomass in a grassland plant diversity gradient. Soil Biology & Biochemistry 60, 122e124. Hackl, E., Pfeffer, M., Donat, C., Bachmann, G., Zechmeister-Boltenstern, S., 2005. Composition of the microbial communities in the mineral soil under different types of natural forest. Soil Biology & Biochemistry 37, 661e671. €ttenschwiler, S., Tiunov, A.V., Scheu, S., 2005. Biodiversity and litter decompoHa sition in terrestrial ecosystems. Annual Review of Ecology, Evolution, and Systematics 36, 191e218. Herman, D.J., Firestone, M.K., Nuccio, E., Hodge, A., 2012. Interactions between an arbuscular mycorrhizal fungus and a soil microbial community mediating litter decomposition. FEMS Microbiology Ecology 80, 236e247. Jacob, A., Hertel, D., Leuschner, C., 2014. Diversity and species identity effects on fine root productivity and turnover in a species-rich temperate broad-leaved forest. Functional Plant Biology. http://dx.doi.org/10.1071/FP13195. Jacob, A., Hertel, D., Leuschner, C., 2013. On the significance of belowground overyielding in temperate mixed forests: separating species identity and species diversity effects. Oikos 122, 463e473. Jacob, M., Viedenz, K., Polle, A., Thomas, F.M., 2010a. Leaf litter decomposition in temperate deciduous forest stands with a decreasing fraction of beech (Fagus sylvatica). Oecologia 164, 1083e1094. Jacob, M., Leuschner, C., Thomas, F.M., 2010b. Productivity of temperate broadleaved forest stands differing in tree species diversity. Annals of Forest Science 67. Jacob, M., Weland, N., Platner, C., Schaefer, M., Leuschner, C., Thomas, F.M., 2009. Nutrient release from decomposing leaf litter of temperate deciduous forest trees along a gradient of increasing tree species diversity. Soil Biology & Biochemistry 41, 2122e2130. Kaur, A., Chaudhary, A., Kaur, A., Choudhary, R., Kaushik, R., 2005. Phospholipid fatty acid e a bioindicator of environment monitoring and assessment in soil ecosystem. Current Science 89, 1103e1112.

A. Scheibe et al. / Soil Biology & Biochemistry 81 (2015) 219e227 Kramer, C., Gleixner, G., 2006. Variable use of plant- and soil-derived carbon by microorganisms in agricultural soils. Soil Biology & Biochemistry 38, 3267e3278. Lang, C., Seven, J., Polle, A., 2011. Host preferences and differential contributions of deciduous tree species shape mycorrhizal species richness in a mixed Central European forest. Mycorrhiza 21, 297e308. Lange, M., Habekost, M., Eisenhauer, N., Roscher, C., Bessler, H., Engels, C., Oelmann, Y., Scheu, S., Wilcke, W., Schulze, E.-D., Gleixner, G., 2014. Biotic and abiotic properties mediating plant diversity effects on soil microbial communities in an experimental grassland. Plos One 9. http://dx.doi.org/10.1371/ journal.pone.0096182. Langenbruch, C., Helfrich, M., Flessa, H., 2012. Effects of beech (Fagus sylvatica), ash (Fraxinus excelsior) and lime (Tilia spec.) on soil chemical properties in a mixed deciduous forest. Plant and Soil 352, 389e403. Leake, J.R., Johnson, D., Donnelly, D.P., Muckle, G.E., Boddy, L., Read, D.J., 2004. Networks of power and influence: the role of mycorrhizal mycelium in controlling plant communities and agroecosystem functioning. Canadian Journal of Botany e Revue Canadienne De Botanique 82, 1016e1045. Legendre, P., 2008. Studying beta diversity: ecological variation partitioning by multiple regression and canonical analysis. Journal of Plant Ecology 1, 3e8. Leuschner, C., Jungkunst, H.F., Fleck, S., 2009. Functional role of forest diversity: pros and cons of synthetic stands and across-site comparisons in established forests. Basic and Applied Ecology 10, 1e9. rez, D.C., Serrano, F.R.L., Andre s, M., Bastida, F., 2012. AltitudeLucas-Borja, M.E., Pe related factors but not Pinus community exert a dominant role over chemical and microbiological properties of a Mediterranean humid soil. European Journal of Soil Science 63, 541e549. Meinen, C., Hertel, D., Leuschner, C., 2009. Biomass and morphology of fine roots in temperate broad-leaved forests differing in tree species diversity: is there evidence of below-ground overyielding? Oecologia 161, 99e111. €, P., Malmivaara-L€ Merila ams€ a, M., Spetz, P., Stark, S., Vierikko, K., Derome, J., Fritze, H., 2010. Soil organic matter quality as a link between microbial community structure and vegetation composition along a successional gradient in a boreal forest. Applied Soil Ecology 46, 259e267. Mitchell, R.J., Keith, A.M., Potts, J.M., Ross, J., Reid, E., Dawson, L.A., 2012a. Overstory and understory vegetation interact to alter soil community composition and activity. Plant and Soil 352, 65e84. Mitchell, R.J., Hester, A.J., Campbell, C.D., Chapman, S.J., Cameron, C.M., Hewison, R.L., Potts, J.M., 2012b. Explaining the variation in the soil microbial community: do vegetation composition and soil chemistry explain the same or different parts of the microbial variation? Plant and Soil 351, 355e362. €lder, A., Bernhardt-Ro €mermann, M., Schmidt, W., 2006. Forest ecosystem Mo research in Hainich National Park (Thuringia): first results on flora and vege€kologie Online 3, tation stands with contrasting tree species diversity. Waldo 83e99. Mund, M., 2004. Carbon Pools of European Beech Forests (Fagus sylvatica) under € ttingen. Different Silvicultural Management (PhD dissertation). Universit€ at Go Neumann, D., Heuer, A., Hemkemeyer, M., Martens, R., Tebbe, C.C., 2013. Response of microbial communities to long-term fertilization depends on their microhabitat. FEMS Microbiology Ecology 86, 71e84. Ngosong, C., Gabriel, E., Ruess, L., 2012. Use of the signature fatty acid 16:1u5 as a tool to determine the distribution of arbuscular mycorrhizal fungi in soil. Journal of Lipids. http://dx.doi.org/10.1155/2012/236807. Nichols, P., Stulp, B.K., Jones, J.G., White, D.C., 1986. Comparison of fatty-acid content and DNA homology of the filamentous gliding bacteria Vitreoscilla, Flexibacter, Filibacter. Archives of Microbiology 146, 1e6. Nordby, H.E., Nemec, S., Nagy, S., 1981. Fatty acids and sterols associated with citrus root mycorrhizae. Journal of Agricultural and Food Chemistry 29, 396e401. O'Leary, W.M., Wilkinson, S.G., 1988. Gram-positive bacteria. In: Ratledge, C., Wilkinson, S.G. (Eds.), Microbial Lipids, vol. 1. Academic Press, London, England, UK, San Diego, California, USA, pp. 117e202. Olsson, P.A., 1999. Signature fatty acids provide tools for determination of the distribution and interactions of mycorrhizal fungi in soil. FEMS Microbiology Ecology 29, 303e310. Pigott, C.D., 1991. Tilia cordata Miller. Journal of Ecology 79, 1147e1207. Pfeiffer, B., Fender, A.-C., Lasota, S., Hertel, D., Jungkunst, H.F., Daniel, R., 2013. Leaf litter is the main driver for changes in bacterial community structures in the rhizosphere of ash and beech. Applied Soil Ecology 72, 150e160.

227

Porazinska, D.L., Bardgett, R.D., Blaauw, M.B., Hunt, H.W., Parsons, A.N., Seastedt, T.R., Wall, D.H., 2003. Relationships at the aboveground-belowground interface: plants, soil biota, and soil processes. Ecological Monographs 73, 377e395. Prescott, C.E., Grayston, S.J., 2013. Tree species influence on microbial communities in litter and soil: current knowledge and research needs. Forest Ecology and Management 309, 19e27. Ranjard, L., Richaume, A.S., 2001. Quantitative and qualitative microscale distribution of bacteria in soil. Research in Microbiology 152, 707e716. Reich, P.B., Tilman, D., Isbell, F., Mueller, K., Hobbie, S.E., Flynn, D.F.B., Eisenhauer, N., 2012. Impacts of biodiversity loss escalate through time as redundancy fades. Science 336, 589e592. Scherber, C., Eisenhauer, N., Weisser, W.W., Schmid, B., Voigt, W., Fischer, M., Schulze, E.-D., Roscher, C., Weigelt, A., Allan, E., Beßler, H., Bonkowski, M., Buchmann, N., Buscot, F., Clement, L.W., Ebeling, A., Engels, C., Halle, S., € nig, S., Kowalski, E., Kummer, V., Kuu, A., Kertscher, I., Klein, A.-M., Koller, R., Ko Lange, M., Lauterbach, D., Middelhoff, C., Migunova, V.D., Milcu, A., Müller, R., Partsch, S., Petermann, J.S., Renker, C., Rottstock, T., Sabais, A., Scheu, S., Schumacher, J., Temperton, V.M., Tscharntke, T., 2010. Bottom-up effects of plant diversity on multitrophic interactions in a biodiversity experiment. Nature 468, 553e556. €rner, C., Schulze, E.-D., 2005. The functional significance of Scherer-Lorenzen, M., Ko €rner, C., Schulze, E.-D. forest diversity: a synthesis. In: Scherer-Lorenzen, M., Ko (Eds.), Forest Diversity and Function: Temperate and Boreal Systems. SpringerVerlag, Berlin, Berlin, pp. 377e389. Talkner, U., Jansen, M., Beese, F.O., 2009. Soil phosphorus status and turnover in central-European beech forest ecosystems with differing tree species diversity. European Journal of Soil Science 60, 338e346.  ter Braak, C.J.F., Smilauer, P., 2012. Canoco Reference Manual and User's Guide: Software for Ordination (Version 5.0). Microcomputer Power, Ithaca, New York, USA. Thoms, C., Gleixner, G., 2013. Seasonal differences in tree species' influence on soil microbial communities. Soil Biology & Biochemistry 66, 239e248. Thoms, C., Gattinger, A., Jacob, M., Thomas, F.M., Gleixner, G., 2010. Direct and indirect effects of tree diversity drive soil microbial diversity in temperate deciduous forest. Soil Biology & Biochemistry 42, 1558e1565. Timonen, S., Kauppinen, P., 2008. Mycorrhizal colonisation patterns of Tilia trees in street, nursery and forest habitats in southern Finland. Urban Forestry & Urban Greening 7, 265e276. van der Heijden, M.G.A., Bardgett, R.D., van Straalen, N.M., 2008. The unseen majority: soil microbes as drivers of plant diversity and productivity in terrestrial ecosystems. Ecology Letters 11, 296e310. Vestal, J.R., White, D.C., 1989. Lipid analysis in microbial ecology e quantitative approaches to the study of microbial communities. Bioscience 39, 535e541. Vesterdal, L., Elberling, B., Christiansen, J.R., Callesen, I., Schmidt, I.K., 2012. Soil respiration and rates of soil carbon turnover differ among six common European tree species. Forest Ecology and Management 264, 185e196. Vivanco, L., Austin, A.T., 2008. Tree species identity alters forest litter decomposition through long-term plant and soil interactions in Patagonia, Argentina. Journal of Ecology 96, 727e736. Wardle, D.A., 2006. The influence of biotic interactions on soil biodiversity. Ecology Letters 9, 870e886. Wardle, D.A., Bonner, K.I., Nicholson, K.S., 1997. Biodiversity and plant litter: experimental evidence which does not support the view that enhanced species richness improves ecosystem function. Oikos 79, 247e258. White, D.C., Stair, J.O., Ringelberg, D.B., 1996. Quantitative comparisons of in situ microbial biodiversity by signature biomarker analysis. Journal of Industrial Microbiology 17, 185e196. Zak, D.R., Holmes, W.E., White, D.C., Peacock, A.D., Tilman, D., 2003. Plant diversity, soil microbial communities, and ecosystem function: are there any links? Ecology 84, 2042e2050. Zelles, L., 1999. Fatty acid patterns of phospholipids and lipopolysaccharides in the characterisation of microbial communities in soil: a review. Biology and Fertility of Soils 29, 111e129. Zelles, L., Bai, Q.Y., 1993. Fractionation of fatty acids derived from soil lipids by solidphase extraction and their quantitative analysis by GC-MS. Soil Biology & Biochemistry 25, 495e507.