Resolving the context-dependency of local heterogeneity-diversity relationships across rocky reefs worldwide

Environmental heterogeneity is widely thought to promote biodiversity, yet its variable effects limit its predictive power. This variability can be better understood by considering key mediating factors: different forms of local heterogeneity, organismal groups and their associated traits, and the broader environmental context (e.g., latitude, substrate type). To address these factors, we analysed 144 studies (24,412 data points and 2,815 effects) from rocky reefs worldwide. Heterogeneity was defined as spatial variability in an area, and heterogeneity metrics were grouped into facets such as three-dimensional structure (e.g., substrate rugosity), complexity (e.g., fractal dimension), and feature variability (e.g., mussel size classes). All facets promoted biodiversity, but effects were context-dependent, with three-dimensional structures having the strongest impact, likely due to increased niche provision and substrate area. Responses also varied across organismal groups, with small-bo..., , , # Resolving the context-dependency of local heterogeneity - diversity relationships across rocky reefs worldwide

[https://doi.org/10.5061/dryad.jh9w0vtm5](https://doi.org/10.5061/dryad.jh9w0vtm5)

This repository provides the data and code used to produce the results presented in Sola & Griffin (2024): *Resolving the context-dependency of local heterogeneity-diversity relationships: a framework and key moderators from rocky reefs worldwide.*

 Following a metanalytical approach, heterogeneity effects in rocky reef literature were analysed to assess the effects of heterogeneity facet, organismal group, environmental context and ecological processes on the strength and shape of HDRs. This approach allowed to develop an operational framework of heterogeneity metrics linked to multiple forms of heterogeneity at the local scale, or ‘heterogeneity facets’ (e.g., Heidrich et al. 2020).

## Description of the data and file structure

Here, you can find the code used to compute the effect sizes...,