Land-use change is reshaping ecological assemblages worldwide. As a result of expanding human-dominated land uses and the intensification of land, many local ecological communities have shown stark reductions in abundance and richness, and significant turnover in species identities. However, less is known about how these pressures affect the functional diversity of communities, even though functional diversity can be more important than species richness for ecosystem function. We ask how changes in land use and intensity impact the values and ranges of functional traits in plant communities, assessing both the position (community weighted mean) and spread (functional dispersion) of communities in functional trait space. We do this by using a mixed-effects modelling framework to combine the PREDICTS (Projecting Responses of Ecological Diversity In Changing Terrestrial Systems) database—a collation of biodiversity data from the published literature—and the Global Spectrum of Plant Form and Function – a recent framework that distils the complexity of plant functional traits into two core axes.
The functional composition of plant communities shows strong and highly significant shifts in response to land use. Mature secondary vegetation appears functionally distinct from primary vegetation, with markedly lower values of the axis corresponding to size and woodiness, and higher values of the axis corresponding to leaf economics (i.e., less conservative leaves). Therefore, although recent studies have shown high biomass resilience of some secondary forest sites, functional trait composition appears to show more severe recovery debt, which could have impacts on other facets of biodiversity. Land-use intensity also impacted functional composition; in plantation forests, low-intensity use (e.g., low-input organic farming) maintained plant communities having a similar functional composition to primary vegetation in terms of the first trait axis (size and woodiness) but not the second axis (plantations had lower values of the second axis, corresponding to more conservative leaves). By contrast with functional composition, the functional dispersion of communities—the range of traits displayed by individuals in a community—was not strongly influenced by land use. We combine our statistical models with global data on land use to infer the geographic pattern of shifts in functional composition caused by land use and related drivers.