Soil communities and decomposition in focus of the metabolic theory and the ecological stoichiometry

Ecosystem functioning is maintained by the interplay of a multitude of species. However, in a time of global change there are high rates of species extinctions which can drastically reduce the functionality of natural ecosystems. Consequently, a mechanistic understanding of the relationship between biodiversity and ecosystem functioning becomes increasingly important. Diversity effects are mediated by complex interactions and impact multiple levels of biological organization. Changes in consumer-resource interactions at the individual level, for example, affect population densities, spreading across communities and ultimately scaling up to entire food webs. Hence, research at all of these levels is required to establish a comprehensive understanding how ecosystem functioning is maintained. The metabolic theory of ecology and the theory of ecological stoichiometry both explore and describe various patterns ranging from species interactions to entire ecosystems. However, to date both theories focus on different perspectives: the conceptual idea behind the metabolic theory is allometry - non-linear scaling relationships of biological rates and processes with body mass and temperature. Ecological stoichiometry instead focuses on imbalances in elemental contents between units, such as consumers and their resources. However, the allometric and stoichiometric principles of the two theories are not mutually exclusive. Moreover, they should complement each other to form a more powerful, unified theory. This new theory would aid our understanding of the complexity of interactions amongst species within food webs. Surprisingly, to date these concepts have rarely been used in a combined framework. In this thesis, I aimed to make an important step towards such a unified perspective of metabolic theory and ecological stoichiometry by studying the interplay of body mass, temperature and stoichiometry. The ecosystem function I focused on is decomposition, which directly influences nutrient cycling and thus is crucial for the ecosystem´s productivity. I concentrated on invertebrate species and communities of forest floors and studied possible direct and indirect effects on the process of litter decomposition. To disentangle the interplay of body mass and resource stoichiometry across several levels of organization, I conducted experiments and analyzed field data of natural environments. “A mechanistic understanding of how interactions between temperature and litter stoichiometry are driving decomposition rates is currently lacking” (Ott et al. 2012). In Chapter 2, I “filled this void by quantifying decomposer consumption rates” in a laboratory experiment (Ott et al. 2012). To realize this, I applied for the first time the concept of functional responses that consists of the parameters handling time and attack rate. In systematic variations of body masses of a woodlouse species, environmental temperature and the resource quality, I “found that attack rates increased and handling times decreased (1) with body masses and (2) temperature” (Ott et al. 2012). Notably, “these relationships interacted with litter quality” (Ott et al. 2012). Small woodlice possibly showed avoidance behavior of poor resource, whereas the consumption rates of large woodlice increased on the poor resource. This contrast suggests that larger woodlice have to compensate a higher metabolic demand with decreasing resource quality in relation to smaller woodlice. The combination of variables associated with “metabolic theory and ecological stoichiometry provided significant mechanistic insights into how warming and varying litter quality may modify consumption rates” of differently sized decomposers (Ott et al. 2012). I investigated this interdependency of factors in different trophic levels of a forest floor community in Chapter 3. In a microcosm study, I manipulated horizontal (within a trophic level) and vertical (across trophic levels) diversity to examine multi-trophic diversity effects on the decomposition. While litter mass loss in general increased with total diversity (i.e., combined decomposer and predator richness), I found that this total diversity effect was driven by horizontal diversity. Moreover, effects of vertical diversity were surprisingly neutral to positive for ecosystem functioning, even though intra-guild predation likely could have released the decomposer prey from top-down pressure. I argue that the interplay between interference competition among decomposers and low top-down pressure by predators should be responsible for these results. Overall, I found interwoven effects of horizontal and vertical diversity on litter decomposition in forest ecosystems. As a possible stimulus for future research, my study provides an example how to systematically disentangle horizontal and vertical diversity effects on ecosystem functioning. In Chapters 2 and 3, I combined effects of allometry and stoichiometry to consumer-resource interactions and multi-trophic mechanisms of diversity on decomposition in a small manipulated community. In Chapter 4, I extend the level of complexity to populations in soil food webs. “Metabolic theory predicts variance in biomass density within communities in dependence of population average body masses, whereas the ecological stoichiometry” considers resource stoichiometry to cause variation in density across communities via nutritional limitations on the consumers (Ott et al. 2014b). I integrated these two theories into one novel framework to analyze biomass densities of “populations of soil invertebrates across 48 forest sites” (Ott et al. 2014b). Using linear mixed effects models, I investigated “how the scaling of biomass densities with population-averaged body masses systematically interacts with stoichiometric variables” (Ott et al. 2014b). The integrated model with allometric and stoichiometric predictors proved superior to the allometric null model. Moreover, the integrated model explained deviations from predicted allometric scaling while accounting for phylogenetic groups as co-predictor in the random structure of the model. In Chapter 5, I applied the model concept developed in Chapter 4 to twelve phylogenetic groups of the same dataset separately. I “investigated how the populations´ biomass densities of temperate forest soil communities depend on 1) the stoichiometry of the basal litter according to the ecological stoichiometry concepts and 2) the population average body mass as predicted by the metabolic theory” (Ott et al. 2014a). “Following various ecological stoichiometry hypotheses, I tested for effects of the carbon-to-element ratios of 10 elements” (Ott et al. 2014a). “Additionally, I included the abiotic litter characteristics habitat size, litter diversity and pH, as well as forest type as an indicator for human management” (Ott et al. 2014a). For ten out of the twelve phylogenetic groups “the biomass densities scaled significantly not only with population-averaged body masses but also with stoichiometric and abiotic co-variables” (Ott et al. 2014a). Out of 14 predictors, “the four most frequent co-variables were 1) forest type, 2) the carbon-to-phosphorus ratio, 3) the carbon-to-sodium ratio, and the carbon-to-nitrogen ratio” (Ott et al. 2014a). While these results confirmed some element-specific hypothesis, I revealed that scaling relationships from taxa between functional groups (meso- and macrofauna) and trophic groups (decomposers and predators) were best predicted by the integrated model approach. In this comprehensive analysis, I demonstrated “how the elemental stoichiometry of the litter as the basal resource constrains population densities across multiple trophic levels of soil communities” (Ott et al. 2014a). Moreover, I confirmed the predictive power of the integrated model approach. In Chapter 6, I extended the laboratorial (Chapters 2 and 3) and analytical (Chapters 4 and 5) approaches to situations under natural conditions on forest plots. I conducted a litter-bag study and examined several factors affecting litter decomposition. I examined: 1) litter quality using leaf litter of two tree species (maple and beech) that differed in litter stoichiometry; 2) the absence and presence of meso - and macrofauna compared to microorganisms alone; 3) land use via a gradient of intensive forest management; 4) exclusion of living roots compared to an associated control on each plot; 5) species richness of the soil communities. I found significant interactive effects that yielded highest litter mass loss on maple litter, in presence of meso - and macrofauna and in most intensively managed forests. Summarizing, my study reports striking insights into how major components of the decomposition process interact with each other in managed forest ecosystems. Furthermore, the interactive effects of litter quality and with body mass (presence of meso-macrofauna) corroborate the finding from the previous chapters. In a nutshell, I provide promising novel experimental solutions for measurements of interaction strengths of decomposers (Chapter 2) and for disentangling effects of multi-trophic diversity (Chapter 3). Moreover, the integrative model framework with interdependent allometric and stoichiometric variables successfully predicted population biomass densities superior to the allometric null model (Chapter 4 and 5). Due to its flexibility, this framework has much potential to be broadly applicable. This thesis represents a highly promising step towards unifying metabolic theory and ecological stoichiometry. My results emphasize that a combination of body mass, temperature and stoichiometry yields superior predictive power on species interactions and population densities, which scale up to food webs. Ultimately, I demonstrate that a combination of metabolic theory and ecological stoichiometry provides a promising basis for future research on biodiversity and ecosystem functioning