Decline effects are rare in ecology: Comment

This is a Preprint and has not been peer reviewed. The published version of this Preprint is available: This is version 4 of this Preprint.

Add a Comment

You must log in to post a comment.


There are no comments or no comments have been made public for this article.


Download Preprint

Supplementary Files

Yefeng Yang, Malgorzata Lagisz, Shinichi Nakagawa


Recently, Costello and Fox (2022) tested, with a large dataset, the hypothesis of whether there is a widespread decline effect in the discipline of ecology. In other words, the magnitude of the reported ecological effect sizes declines over time (Leimu and Koricheva 2004). Contrary to early results from much smaller datasets (Jennions and Møller 2002, Barto and Rillig 2012), Costello and Fox (2022), using 466 ecological meta-analyses with > 100,000 effect sizes, concluded that there was no systematic decline effect across the field of ecology – only ~5% of ecological meta-analyses showed statistical evidence of a decline effect. This conclusion seems to be “good news” and has important field-wide implications. For example, the temporal stability of the cumulative evidence can alleviate the concerns about policy-making for conservation and environmental management (Koricheva and Kulinskaya 2019).



Ecology and Evolutionary Biology, Life Sciences, Physical Sciences and Mathematics, Statistical Methodology, Statistical Models, Statistics and Probability


decline effect, meta-analysis, meta-research, publication bias, time-lag bias


Published: 2022-06-06 14:34

Last Updated: 2022-06-07 09:57

Older Versions

CC-BY Attribution-NonCommercial-ShareAlike 4.0 International