To what extent will a change in biodiversity result in a change in the ecosystem services provided to society in natural ecosystems? This question remains unresolved and debated. The answer to this question will, in part, depend on how much the bulk of ecosystem services depends on a few dominant species orwhether many rare species contribute significantly. Indeed, rare species comprise the vast majority of species richness and are more likely to be lost. Two approaches attempt to address this question: scaling up results from experiments and attempting to infer causality from observational studies in natural ecosystems. Predictions based on results from small-scale experiments versus conclusions drawn from observational studies in natural ecosystems reach different answers. Using either approach to draw conclusions about the role of richness and particular species, including dominants, in natural systems relies on assumptions. We review the assumptions– and how they influence the conclusions drawn from each study type and drive several key gaps in our knowledge. We aim to bridge findings from experimental and observational studies, both which have enormous value, and identify ways to improve understanding of how natural systems and their diversity of species provide services currently and under future conditions.
Results/Conclusions
Attempting to bridge findings from experimental and observational studies reveals several gaps in our understanding, which will be discussed. For one, under what conditions do rare (or less dominant) species contribute substantially to ecosystem services? And, how often or under what circumstances the dominant species, contributing to the bulk of services, will change under different and novel conditions. We propose future directions to help understand and predict when these potential changes occur, the factors driving them (e.g., environmental, management, or community structure variables), and their consequences for service provisioning at a larger scale. We provide two routes for future research in this area: one based on biological mechanisms and the other on statistical techniques for causal inference in observational data that have not been widely adopted in ecology. We provide examples for how these approaches can help take advantage of newly available and existing data sources with greater spatial and temporal coverage but also with the messy challenges associated with using observational data. Merging the perspectives and research frontiers discussed in this session will help identify how and when shifts in species abundances and diversity will impact ecosystem functions and services in changing environments.