|| 要旨トップ | 本企画の概要 |||日本生態学会第70回全国大会 (2023年3月、仙台) 講演要旨
シンポジウム S02-1 （Presentation in Symposium）
The framework of modern coexistence theory (MCT) has emerged as a powerful approach to understanding the nature of competitive interactions and community assembly. It provides an intuitive, simple-to-measure method for quantifying the nebulous concept of niche differences and the myriad roles of environmental variation in competitive outcomes. Thus, the framework has been widely adopted by empiricists eager to measure these mechanisms in their own systems, resulting in many high-profile studies over the past decade. In many of these cases, however, statistical rigor of coexistence studies is abandoned in favor of presenting simple, clear results. I will present an exhaustive literature review and a co-occurrence analysis concerning some overlooked statistical pitfalls in empirical coexistence studies and the consequences thereof. These include easily-addressed omissions of model testing and error propagation as well as more existential issues of the framework's overall predictive utility and often-restrictive assumptions. I conclude by identifying ways in which empiricists can best circumvent these statistical pitfalls while also advocating for a more predictive and interdisciplinary empirical coexistence programme which can usefully inform management and other pressing environmental issues.