Aim

Soil microorganisms are essential for the functioning of terrestrial ecosystems. Although soil microbial communities and functions may be linked to tree species composition and diversity, there has been no comprehensive study of how general these potential relationships are, or if they are context-dependent. Here, we examine tree diversity–soil microbial biomass and respiration relationships across environmental gradients using a global network of tree diversity experiments.

Location

Global

Time Period

2013

Major Taxa Studied

Soil microorganisms

Methods

Soil samples collected from eleven tree diversity experiments in four biomes were used to measure microbial respiration, biomass, and respiratory quotient using the substrate-induced respiration method. All samples were measured using the same analytical device, method, and procedure to reduce measurement bias. We used linear mixed-effects models and PCA to examine the effects of tree diversity (taxonomic and phylogenetic), environmental conditions, and interactions on soil microbial properties.

Results

Abiotic drivers, mainly soil water content, but also soil carbon and soil pH, significantly increased soil microbial biomass and respiration. Optimal soil water content reduced the importance of other abiotic drivers. Tree diversity alone had no effect on the soil microbial properties, but interactions with phylogenetic diversity indicated that diversity effects are context-dependent and stronger in drier soils. Similar results were found for soil carbon and soil pH.

Main conclusions

Our results point to the importance of abiotic variables and especially soil water content for maintaining high levels of soil microbial functions and modulating the effects of other environmental drivers. Planting tree species with diverse water-use strategies and structurally complex canopies and high leaf area may crucial for maintaining high soil microbial biomass and respiration. Since higher phylogenetic distance alleviated unfavorable soil water conditions, reforestation efforts accounting for traits improving soil water content or choosing more phylogenetically distant species may assist in increasing soil microbial functions.

Soil samples were taken in 2013 from eleven tree diversity experiments that are part of the global network TreeDivNet (Verheyen et al., 2016; http://www.treedivnet.ugent.be/). Experiments are independent of each other with different experimental designs and plot configurations. Experiments are distributed across four continents (Asia, Europe, North, and South America) and four different biomes (boreal, temperate, tropical, subtropical, Olson et al., 2001), and differ in age, with the youngest experiments running for three years and the oldest for fourteen years as of 2013 (i.e., the year of the sampling campaign). In total, 106 tree species were included in this study. Experiments had a mean ± SD number of diversity levels of 3.7 ± 1.0, with diversity levels ranging from monocultures to 18 tree species in sub/tropical regions. All experiments had an experimental gradient in tree species richness, with the exception of one (BIOTREE-FD) which manipulated functional diversity at a constant level of tree species richness.

Soil sampling

Soil samples were taken from a depth of 0 – 10 cm, excluding the litter layer using a soil corer. Depending on the size of the experimental plot, different numbers of subsamples were taken per plot to create one composite sample. For plots <100 m², three subsamples were taken per plot, while ten subsamples were taken for plots >100 m2. These subsamples were taken to capture the spatial heterogeneity of the plot and to represent as many different combinations of tree species as possible. Soil samples were always taken in the center of surrounding trees. Immediately after sampling, soil samples were stored at 5°C until sieving at 2 mm and then were stored at -20°C until and during shipping to minimize changes in microbial activity, biomass, and composition. Alternatively, if shipping at -20°C was not possible, samples defrosted during shipping and were measured shortly after arrival. Altogether, 1010 plots were sampled across the eleven tree diversity experiments.

Measurement of soil microbial properties

Before the start of microbial measurements, samples were kept at +20°C for five days to unfreeze and to adapt the soil microbial community to a constant and standardized temperature. Three different soil microbial community properties were assessed using an automated O₂ micro-compensation system (Scheu, 1992). First, basal respiration (µl O₂ h^-1 g^-1 dry soil) was measured as the mean oxygen consumption per hour without the addition of any substrate. The mean oxygen consumption was measured for hours 15 to 20. Basal respiration reflects the active part of the soil microbial community at the time of sampling. Second, microbial biomass carbon was measured by substrate-induced respiration, i.e., the respiratory response of microorganisms to glucose and water addition. To saturate catabolic microbial enzymes, 8 mg glucose g^-1 soil dry weight was added as an aqueous solution to the soil samples. The lowest substrate-induced respiration of three contiguous hours within the first 10 h was taken as the maximum initial respiratory response (MIRR) – a period when microbial growth has not started. Microbial biomass (µg C g^-1 dry soil) was calculated as 38 × MIRR (µl O₂ h^-1 g^-1 dry soil) following Beck et al., (1997). By providing water and glucose, the maximum potential of the living microbial biomass is activated that is able to use glucose, whereas for basal respiration only a fraction of the entire community is active. Third, the microbial-specific respiratory quotient (µl O₂ mg^-1 Cmic h^-1) was calculated as the ratio of basal respiration and soil microbial biomass. The specific respiratory quotient is a measure of soil microbial carbon-use efficiency. Carbon-use efficiency is high when microbial biomass can be built up without high investment in basal respiration, which is indicated by a lower specific respiratory quotient. All measurements were conducted at +20°C in an air-conditioned laboratory using the same analytical devices (RMS Schuller, Darmstadt, Germany).

Diversity metrics

In addition to tree species richness, we aimed at testing a tree diversity metric that captures the functional diversity of each experimental forest plot. However, no comparable trait measurements were available from all experiments. We chose not to use data from trait databases such as TRY because some tropical tree species were absent from these databases and because traits demonstrate substantial plasticity to their abiotic and biotic environment (i.e., intraspecific trait variation). Instead, we used phylogenetic diversity as a proxy for multi-trait functional diversity (Tucker et al., 2018). Phylogenetic diversity indices have been shown to be powerful predictors of biodiversity-ecosystem functioning relationships (e.g., Craven et al., 2018) and are suggested to work when key functional traits are not available (Paquette et al., 2015). We used the molecular phylogeny from previous studies (Pietsch et al., 2014; Zanne et al., 2013) as a backbone to build a phylogeny of all species within the tree diversity experiments, conservatively binding species into the backbone using dating information from congeners in the tree. We used the comparative.comm function in the R package pez to calculate a set of phylogenetic diversity indices, specifically MPD (mean phylogenetic diversity), MNTD (mean nearest taxonomic distance), and the standardized version of both to account for correlation with species richness (Pearse et al., 2015). MNTD was found to correlate less (using Pearson correlation) with log species richness and, therefore, was used in all following analyses. Taxonomic names of tree species were standardized using the website http://tnrs.iplantcollaborative.org/index.html.

Soil characteristics

We included a set of explanatory variables to describe the experimental sites, which were shown to have an effect on soil microbial properties and reflect the designs and local conditions of the different experiments (Fig. S1). Gravimetric soil water content was measured as % H2O from fresh soil weight by drying the whole sample at 75°C for three days. Soil pH and soil C (%) were measured at the block level to obtain information about soil quality characteristics of each experiment. Therefore, equal proportions of dry soil were weighed from each sample to form a composite sample. The whole sample was ground, and a fraction of 10 g was used for pH measurements by adding 0.01 m CaCl2. Soil C concentrations were analyzed by using the ground soil with an elemental analyzer (Vario EL Cube, Elementar). We further extracted clay (%), sand (%), and silt (%) content from the SoilGRIDS database (Hengl et al., 2014).

Environmental conditions

For each experimental site, we extracted mean annual temperature (MAT), the seasonal variability of temperature (Season Temp.), annual precipitation (MAP), and the seasonal variability of precipitation (Season Prec.) from the WorldClim database (http://www.worldclim.org/current) with 2.5 arc-minutes resolution. Potential evapotranspiration (PET) and the aridity index (MAP/PET) were extracted from CGIAR-CSI (https://srtm.csi.cgiar.org/). In addition, we obtained the age of the experiment (years) and tree density (trees m^-²) from publications associated with each experiment and the TreeDivNet website (http://www.treedivnet.ugent.be/). Biomes were assigned based on Olson et al., (2001).