d05

Individual niche specialisation in the light of evidence synthesis

The number of niche-related publications is ever growing and evidence synthesis, and specifically meta-analysis, offers a powerful tool for quantitatively integrating findings across publications. By integrating heterogeneous datasets, meta-analyses allow two important insights that cannot be gained from single studies: (i) An assessment of the overall evidence and its generality and (ii) an evaluation of higher-order context-dependencies and/or hypotheses. Meta-analyses have recently been complemented by influence synthesis. Influence synthesis builds on bibliographic methods and on modern artificial intelligence algorithms such as deep learning. It is particularly suitable for reviewing established scientific concepts such as “ecological niche”. In the first phase of the NC³, we have achieved three main goals: First, we conducted multiple meta-analyses, including two on environmental and developmental effects on risk-taking behaviour and fitness, respectively, and one on transgenerational effects. Second, we developed methods for variance decomposition and for meta-analysing variances. These tools will be essential in the second phase of the NC³. Third, we started an influence synthesis using a topic-modelling analysis to evaluate temporal changes and development of the concept of “ecological niche”. In the second phase of the NC³ we plan to: (a) Conduct specific meta-analyses on topics that are actively studied within the NC³ and on additional topics of great relevance to individual niche specialisation; (b) Conduct an influence synthesis of the concept of “niche construction” using topic-modelling analyses; (c) Support data management and curation within the NC³ by organizing training workshops, promoting modern open science developments and providing practical advice on data organisation and curation. The meta-analyses represent the largest part of the project and will focus on the following topics: population density effects on risk-taking behaviour, behavioural niche specialisation and its dependency on the species’ social system, individual variation in reaction norms, and effects of ontogenic dominance on behavioural variation.