Funded applicants: François Buscot, Nico Eisenhauer, Anna Heintz-Buschart, Kirsten Küsel, Carl-Eric Wegner

Co-applicants with project responsibility: Simone Cesarz, Antonic Chatzinotas, Carlos Guerra, Johannes Sikorski

International cooperation partner: Ika Djukic

Further project group: Guillaume Pantoine, Tesfaye Wubet

The project is performed in cooperation with international partners. Ika Djukic (Swiss Federal Institute for Forest, Snow and Landscape Research WSL, Zürcherstrasse 111, 8903 Birmensdorf; - Switzerland; ika.djukic@umweltbundesamt.at) coordinates the global efforts of TeaComposition, the soil samples of which will be analysed in this project. Ika Djukic is not financially involved beyond the provision of the teabags to the cooperating scientists. There is no financial involvement in the proposed analyses for either partner.

Next generation sequencing will be performed at a DFG Sequencing Centre, the NGS Competence Center Tübingen (NCCT), Germany.

Human activities are altering the biodiversity of ecosystems around the globe, which has provoked concern regarding potential consequences for ecosystem functions. The notion that biodiversity could be an important determinant of ecosystem functions stems from observations of natural communities (Elton 1958) as well as from theoretical models (Tilman et al. 1997; Yachi and Loreau 1999) and has been around for decades. Hundreds of biodiversity-ecosystem function (BEF) experiments have revealed that biodiversity is essential for the provisioning of various ecosystem functions and services (Cardinale et al. 2012; Naeem et al. 2012). Given the limitations of these experiments in realistically representing the non-random biodiversity loss (Haddad et al. 2008) determined by the simultaneous interplay of abiotic and biotic filters in time and space (Götzenberger et al. 2011; Wardle 2016), future BEF research has to answer the question: How does non-random biodiversity loss affect ecosystem functioning in real ecosystems? Answering this question is particularly important in order to apply BEF results to, e.g., agriculture and biodiversity conservation. However, this is also a very challenging question as biodiversity change and species turnover may be hard to predict due to multiple co-occurring and interacting global change drivers and their context-dependent effects on different organisms and interactions (Schmid and Hector 2004). Environmental change and heterogeneity determine both the biodiversity and the functioning of ecosystems (Fig. 1A). In addition, changes in biodiversity can have significant effects on ecosystem functioning, which can be masked by strong environmental gradients (Fig. 1B; Isbell et al. 2013).

Future approaches to study BEF relationships should thus move from studying the relationships between random species loss and ecosystem functioning to the complex interplay between anthropogenic drivers, non-random biodiversity changes (including gains, losses, and changes in evenness/dominance and composition/traits), ecosystem functioning, and ecosystem services at multiple spatial and temporal scales. Given these complications, evidence for significant BEF relationships across environmental gradients is scarce (e.g., Grace et al. 2016). However, recently developed international networks of collaborating researchers now allow for the study of BEF relationships across environmental gradients (e.g., Djukic et al. 2018).

Currently, most BEF studies focus on aboveground diversity, although the ecosystem functions are driven by belowground biological processes. In particular, soils play pivotal roles in important terrestrial ecosystem functions including nutrient cycling, sustaining plant growth, water purification, litter decomposition, and carbon storage (Bardgett and van der Putten 2014, Wall et al. 2015). Microbial communities in the soil drive these critical functions (Quince et al. 2008), which is why they are often referred to as the ‘functional backbones’ of terrestrial ecosystems (van der Heijden et al. 2008). Soil microbial communities are among the most diverse in the world, with key taxa comprising bacteria, archaea, fungi and protists. The high levels of soil microbial diversity have led to the assumption that many taxa are functionally redundant. However, at least for bacteria, it appears that rare and common taxa play fundamentally different roles in ecosystem functioning (Rivett and Bell 2018, Delgado-Baquerizo et al. 2016). Soil microbial communities drive plant diversity and productivity (van der Heijden et al. 2008). However, globally, belowground alpha diversity is not always correlated with above-ground diversity (Cameron et al. 2019). Soil microbial diversity is also only poorly correlated with plant diversity, especially for bacteria and fungi, and therefore focusing on plant diversity is not helpful to predict responses of soil microbiomes, e.g., to global change factors (Prober et al. 2014) or biodiversity loss (Fanin et al. 2019). As a consequence, while plant diversity influences soil microbial biomass (Chen et al. 2019) and plant community composition may improve predictions of microbial communities, the variation in microbial communities explained by plant community composition at the global scale is low (Leff et al. 2018). Recent research has revealed strong spatial variability in the distribution of soil microbial diversity (Ramirez et al. 2017, Delgado-Baquerizo et al. 2018) and functional capacity (Bahram et al. 2018), and has revealed major drivers of bacterial alpha-diversity across biomes (Delgado-Baquerizo and Eldridge 2019). However, data on soil biodiversity still is very limited (Cameron et al. 2018) and has rarely been linked to ecosystem functions. Therefore, a more comprehensive understanding of BEF relationships for soil microbiomes is urgently needed.

Although microcosm experiments suggest positive microbial BEF relationships (e.g., Tiunov and Scheu 2005, Jousset et al. 2011), just a few studies address the potential links between microbial diversity and ecosystem functions in natural ecosystems and across broad environmental contexts (Soliveres et al. 2016). A key challenge is to quantify the functional roles of microbial taxa to understand how ecosystem properties change at spatial and temporal scales. A number of different methods for the assessment of soil microbial communities and their diversity currently co-exist, including the most commonly employed meta-barcoding (amplicon sequencing) estimates of relative abundances of highly resolved taxonomic groups and microbial diversity. Increasingly, studies assess microbial traits and gene content by metagenomics and, more rarely, expressed functions by metatranscriptomics and metaproteomics. Due to their different aims and different resolutions, these techniques can lead to diverging conclusions, but integration of both taxonomic and trait data is expected to improve predictions of ecosystem functions (Fry et al. 2018). Traditionally, overall respiration rates and specific in vitro enzyme activities have often been measured. While most studies employ a combination of some of the mentioned techniques and relate the results to drivers or response variables, systematic comparisons are rare (Fierer et al. 2012). While data collected in independent studies can be combined to improve predictions (Johnston and Sibly 2018), sequencing-based profiles are hardly ever collected in the same studies as ecosystem function data. Hence, data on soil biodiversity are not as coherent and exhaustive as the ones on aboveground biodiversity (Cameron et al. 2018). It is therefore necessary to employ comprehensive methods that deliver the most predictive read-outs for ecosystem function and response to changes.

Among the important functions fulfilled by soil microbial communities that are strongly influenced by soil conditions, climate, and human activities (e.g., Tedersoo et al. 2014, Leff et al. 2015) is the breakdown of organic matter. Litter decomposition and the cycling of the respective nutrients influences plant productivity, and therefore aboveground biodiversity and carbon storage within the soil (Hättenschwiler et al. 2005). The microbial actors and their enzymes are well described under controlled conditions (Schneider et al. 2012). Noteworthy, organisms from all domains of life interact in this process. Copiotrophic, fast growing bacteria degrade easily accessible substrates. Oligotrophic, slow growing bacteria feed on more recalcitrant substrates. In particular the more recalcitrant fractions of litter are more efficiently decomposed by saprotrophic fungi (van der Wal et al. 2013). Mycorrhizal fungi also play a significant role in litter decomposition, as they on the one hand have stimulatory effects, and on the other hand contribute to stabilization of litter-derived carbon (Verbruggen et al. 2015). Microbial grazers and predators also shape the microbial community and the turnover rates (Geisen et al. 2015, Xiong et al. 2017).

We lack a systemic understanding of how this cycling potential is influenced by local (e.g. along soil profile) and interregional variations (climate, soil type, stoichiometry) via their impacts on soil biodiversity. For the longest time, global litter decomposition has been considered to be mainly controlled by climate (temperature and moisture) and litter quality (i.e. the chemical composition; Tenney and Waksman 1929). While climatic drivers are often observed (Sherman and Steinberger 2012), further important controllers may exist (Keiser et al. 2014, Bradford et al. 2015). Soil microbial communities are prime candidates for these controllers (Couˆteaux et al. 1995, Hättenschwiler et al. 2005, Manzoni et al. 2008, Allison et al. 2013). For example, N-fertilization is consistently observed to lower litter decomposition (Knorr et al. 2005). A possible reason for lower decomposition rates may be lower microbial biomass, as observed in N-fertilized grasslands (Treseder 2008, Riggs and Hobbie 2016, Zhang et al. 2018). Despite high functional redundancy, certain functions can be reduced or lost when microbial diversity declines (Singh et al. 2014). For carbohydrate active enzymes in particular, shifts in genetic diversity have been observed, e.g. in response to precipitation (Diamond et al. 2019). Precipitation is a major driver of both litter decomposition (Djukic et al. 2018) and bacterial functional diversity (Bahram et al. 2018), suggesting a significant relationship between functional diversity and decomposition as an ecosystem function, but it is unknown whether this will be true across biomes. While factors like climate, soil conditions, and fertilization directly influence the soil microbiomes (e.g., Tedersoo et al. 2014, Ramirez et al. 2017, Leff et al. 2015), indirect effects through changes in litter quality are also observed (Meier and Bowman 2008). It is therefore important to assess decomposition of standardized materials (Keuskamp et al. 2013) to explore general relationships between microbial diversity and decomposition.

With the proposed project, we aim to perform an integrative, global assessment of the relationship between microbial diversity and litter decomposition in the soil, using a set of existing monitoring platforms operating along environmental gradients (Fig. 1B). This project builds on the successful establishment of a “network of networks” on litter decomposition (Djukic et al. 2018). In an unprecedented initiative, experimental and observational ecological networks have been united using a common methodology to assess litter decomposition (Keuskamp et al. 2013). In preparation of this proposal, this network conducted a concerted and highly standardized soil sampling campaign for the analysis of soil microbial communities. The analysis of microbial taxa and functional diversity using a standardized set of next generation sequencing (NGS)-based tools in relation to functional measurements of decomposition processes should enable us to determine how to combine the available technologies most effectively in BEF research as well as in mechanistic microbiological studies. These analyses necessitate substantial investment in the form of NGS, being applied for within this proposal.

Given the strong societal relevance of global soil biodiversity and function (Bardgett and van der Putten 2014), high levels of integration between scientific disciplines as well as spatial and temporal scales are required. The applicant iDiv members cover a uniquely broad expertise ranging from the analysis of microbial communities by NGS to soil biodiversity and function research that has little parallel worldwide. Researchers at iDiv have started compiling a comprehensive dataset on the global distribution of soil biodiversity (www.idiv.de/sworm), which will be hosted in EDAPHOBASE (https://portal.edaphobase.org) as a first trial for a global soil biodiversity database hosted by iDiv (such as outlined in Ramirez et al. 2015). These data demonstrate for the first time at a global level that hotspots and coldspots of aboveground (mammals, birds, and amphibians, and vascular plants) and soil biodiversity (soil macrofauna, and fungi) do not always coincide (Cameron et al. 2019).

Moreover, some of the PIs have been active participants in the TeaComposition network (Djukic et al. 2018), and all PIs are now leading an initiative to study microbial diversity-function relationships across the sites worldwide. In this study, decomposition measurements have been taken by our international partners together with a collection of soil. From the 395 projected field sites (Fig. 2, 790 samples), 258 (516 samples) were already pre-processed for soil functional analysis in our labs (soil aggregate stability, soil respiration, and microbial biomass). 402 samples have been processed to extract DNA in our labs and the remaining samples are expected to arrive until the end of June 2019.

In summary, this project will foster integration between the above-mentioned approaches to reach a comprehensive scale-independent understanding of environmental drivers and anthropogenic effects on the structural and functional diversity of microbial communities and the subsequent consequences for ecosystem functioning. Therefore, we will produce extensive datasets on the main soil microbial taxa (bacteria, archaea, fungi, and protists) across a global network of sites in relation to climate and land use. Functional relationships will be deduced by multivariate analyses (e.g., structural equation modelling; Grace et al. 2016) of data on microorganisms, physicochemical parameters, and ecosystem functions (decomposition, including litter mass loss and soil microbial respiration). Building on the TeaComposition network will enable us to create unique synergies between an existing, vibrant global network of collaborators working on litter decomposition and cutting-edge next generation sequencing of the soil biodiversity that drives decomposition processes. By doing this, our ultimate goal is to optimize an approach that can be used as standard for future global soil biodiversity and function analyses in relation to different environmental (change) drivers.

Reich PB, Tilman D, Isbell F, Mueller K, Hobbie SE, Flynn DFB and Eisenhauer N (2012) Impacts of biodiversity loss escalate through time as redundancy fades. Science 336: 589-592.

This project proposal is to request funding of NGS analyses, as detailed in section 2 of this proposal. Sufficient funding for all other costs are available to the applicants within the framework of iDiv. Therefore, no funding is requested for any of the following points:

Funding for Staff; Direct Project Costs; Equipment up to 10 000 €, Software and Consumables; Travel Expenses; Visiting Researchers; Expenses for Laboratory Animals; Project-related publication expenses; Instrumentation; Equipment exceeding 10 000 €; Major Instrumentation exceeding 50 000 €.