Illustrating the importance of meta‐analysing variances alongside means in ecology and evolution

Abstract

Meta‐analysis is increasingly used in biology to both quantitatively summarize available evidence for specific questions and generate new hypotheses. Although this powerful tool has mostly been deployed to study mean effects, there is untapped potential to study effects on (trait) variance. Here, we use a recently published data set as a case study to demonstrate how meta‐analysis of variance can be used to provide insights into biological processes. This data set included 704 effect sizes from 89 studies, covering 56 animal species, and was originally used to test developmental stress effects on a range of traits. We found that developmental stress not only negatively affects mean trait values, but also increases trait variance, mostly in reproduction, showcasing how meta‐analysis of variance can reveal previously overlooked effects. Furthermore, we show how meta‐analysis of variance can be used as a tool to help meta‐analysts make informed methodological decisions, even when the primary focus is on mean effects. We provide all data and comprehensive R scripts with detailed explanations to make it easier for researchers to conduct this type of analysis. We encourage meta‐analysts in all disciplines to move beyond the world of means and start unravelling secrets of the world of variance.

Poor nutritional condition promotes high‐risk behaviours: a systematic review and meta‐analysis

“…This contrasts with the expectation that poor condition may increase phenotypic variability (e.g. by exposing cryptic genetic variation), but agrees with a recent meta‐analysis showing that developmental stress does not seem to influence variation in behavioural traits across species (Sánchez‐Tójar et al ., 2020). Heterogeneity was generally lower in lnCVR models relative to lnRR ones, which is likely because variance effect sizes are generally associated with larger sampling variances (Sánchez‐Tójar et al ., 2020).…”

“…by exposing cryptic genetic variation), but agrees with a recent meta‐analysis showing that developmental stress does not seem to influence variation in behavioural traits across species (Sánchez‐Tójar et al ., 2020). Heterogeneity was generally lower in lnCVR models relative to lnRR ones, which is likely because variance effect sizes are generally associated with larger sampling variances (Sánchez‐Tójar et al ., 2020). Variance meta‐analyses are expected to be more data hungry, although this is unlikely to be the cause of the overall weak lnCVR effect found in our study given the large data set used.…”

“…There has, however, been growing interest in individual‐level variation in recent years (Westneat, Wright, & Dingemanse, 2015), and new tools to meta‐analyse variances alongside means are revealing that meta‐variance effects may be both prevalent and often overlooked (Nakagawa et al ., 2015). While a recent meta‐analysis of variance has shown that diet restriction can increase variation in longevity (Senior et al ., 2017), another has shown little evidence of environmental stress (including diet restriction) effects on phenotypic behavioural variance (Sánchez‐Tójar et al ., 2020). Furthermore, case studies have shown increased within‐individual behavioural variation in high‐condition animals, via an increased capacity to express behavioural plasticity (Royauté & Dochtermann, 2017; Royauté et al ., 2019).…”

Scale‐dependent effects of niche specialisation: The disconnect between individual and species ranges

“…This kind of data bears great potential for disentangling the niche‐related mechanisms underlying distribution patterns at different levels of biological organisation (e.g. Sánchez‐Tójar et al, 2020).…”

Herbivore‐induced defenses are not under phylogenetic constraints in the genusQuercus(oak): Phylogenetic patterns of growth, defense, and storage

“…To better understand responses across species, we used a nonparametric sign test (Siegel & Castellan, 1981) to analyze overall trends. Trait variances often display trends that give insights into ecological and evolutionary processes that are not always visible when analyzing mean effects alone (Sánchez‐Tόjar et al., 2020).…”