Many services and functions of ecosystems are the achievements of the local food web. Each single species contributes to the system’s emergent properties. The highest complexity, however, culminates in food webs with high diversity of predatory generalist species, feeding on many prey species at the same time, including other predators. As a consequence, the net effect of losing a predator species can be positive, negative or neutral, depending on the food-web context and the species-specific properties. These species ‘identity effects’ render the consequences of random species loss on ecosystem functioning and ecosystem services unpredictable. The estimation of these effects with general models is the primary objective of this thesis. To achieve this aim, I apply allometric theory to estimate numerous properties of the species, including metabolism, population density, prey range and feeding intensity by the easy-to-assess parameter ‘body mass’. Especially, two assumptions are applied in the experiments and models: Firstly,
allometric feeding rates predict predators to prefer prey of a certain body mass, with declining feeding rates towards larger and smaller prey; Secondly, allometric mass–abundance scaling predicts small species to be more abundant than large species. I show, that by providing this general framework, allometric theory integrates all levels of complexity, from population level over predator-prey systems towards food webs and ecosystems. It thus provides answers to four research questions: (Q1) How can body mass explain the observed feeding rates of a predator on a prey? (Q2) Do these allometric feeding rates predict the effect of a predator in the context of a community? (Q3) How do these predictions scale with increasing predator diversity? (Q4) Can allometrically-defined predator species explain patterns at the ecosystem level?
As an experimental model system, I chose the food web of the litter layer in deciduous forests, which comprises body-mass structured communities of generalist predators. The process of litter decomposition by detritivores, which is important for nutrient recycling and CO2-release, is the focal ecosystem function.
In Chapter 2, I parameterized a mechanistic model for allometric feeding rates on basis of a comprehensive experimental data-set of feeding experiments. In microcosm experiments, I combined multiple differently-sized, generalist predator species by applying a novel allometric design, which balances the bias by individual density and biomass density.
In Chapter 3, I decomposed the nested effects of three predators, which were combined in a full factorial treatment design (the centipede Lithobius forficatus, the spider Pardosa lugubris, the predatory mite Hypoaspis miles), on the lower trophic level springtail population and on microbial biomass. Here, intraguild predation of the large centipedes dampened the suppressive effect of the small mites on springtails. In Chapter 4, these interactions could be predicted by simulating bodymass dependent feeding of the three species.
In Chapter 5, I tested the predictability of net interaction strength of five generalist predators on seven detritivore species. Predator body mass and target species population density proved to be the best explanatory parameters. Surprisingly, the predictability in complex communities was higher than in simple predator monocultures. This was due to the occurrence of body-mass driven intraguild predation in the complex communities.
Chapter 6 presents a mesocosm study where I manipulated population body-mass structure of the top predator (the shore crab, Carcinus maenas) using an allometric design in a subtidal habitat as a simulation of climate change consequences. The community responded to gradually decreasing body mass of the top predator with increasing total biomass and individual body masses. Furthermore, this triggered a gradual trophic cascade.
In Chapter 7, I applied the validated model of allometric feeding rates in a mechanistic simulation of the relationship between predator diversity and ecosystem function. With increasing diversity of predators the biomass stock of the basal trophic level was suppressed although the total predator biomass was low due to high respiration rates and intraguild feeding. This contradicts the expectations of community-level trophic cascades.
In summary, the projects provided the following answers to the research questions (Q1–4):
(A1) The body masses of predator and prey predict the quantitative feeding rates.
(A2) The context dependent positive and negative, weak and strong effects of a predator species on the lower trophic level are resolved by body-mass dependent feeding rates.
(A3) In increasingly complex communities the prediction of interaction strengths becomes simpler.
(A4) A mechanistic model of predator diversity predicts negative effects of diversity on the trophic level below. | English |