Quantitative fatty acid signature analysis: a new tool to determine tropic interactions in soil food webs

点击翻译按钮获取中文摘要

Soil food webs are important drivers for ecological key functions in soil systems such as carbon and nutrient cycling. However, food web models generelly lack quantitative data from empirical studies to describe these energy flows. This is mainly due to the shortage in high-throughput methods to assess the dietary composition of the soil fauna. The aim of this project is to implement Quantitative Fatty Acid Signature Analysis (QFASA) to study belowground feeding relationships. In a precursor project, the classical QFASA model, designed for marine ecosystems and vertebrate predators, was applied to Collembola as important soil decomposers. This work resulted in essential model prerequisites, such as a comprehensive lipid library for basal resources in the green and brown food chain with 229 entries. Moreover, first sets of Calibration Coefficients (CCs) for common Collembola consumers were gained and were tested on a variety of single and mixed diets comprising basal food web resources (i.e. bacteria, fungi, algae, plants). However, model application revealed difficulties not encountered in marine ecosystems, such as highly variable CCs in soil invertebrates and smaller footprints of fatty acid (FA) signatures in the consumers at the food web base. In a first model adaptation a better QFASA performance was achieved by considering the fat content of the diet and adjusting the CCs via linear or exponential equations.The goal of the continuation application is a substantial model development, both empirically (CCs and FA subsets) and mathematically (iterative analyses). This includes experiments with formulated diets using FAs with highly variable CCs and offered over a broad range of fat contents. Thereby, the metabolic control, i.e. the variation of CCs by consumers, is better integrated into the model. The mathematical model application will be systematically elaborated with the focus on: i) CC robustness and distance measurements, ii) best-fit FA and library subsets, and iii) constrains by compositional data. This will result in CCs with significantly improved functionality. Additionally, a model extension from low trophic levels (Collembola) to top predators (mites) is performed. Finally, the best-fit model will be applied to field populations of Collembola and mites to assign their diets in situ. In sum, the proposed work will enable “shifting systems”, i.e. translating the application of QFASA from marine to soil food webs.