Community assembly in harsh environments: the prevalence of ecological drift in the heath vegetation of South America

Abstract

Metacommunity dynamics is marked by a gradient ranging from pure ecological determinism to pure stochasticity. This gradient encompasses compositional turnover that is governed by ecological drift, selection and dispersal. Here we estimate the influences of selection, dispersal limitation acting in concert with drift, drift acting alone and homogenizing dispersal on the structure of tropical restinga heath vegetation growing under stressful conditions in north‐eastern South America. We hypothesize that if abiotic heterogeneity is strong enough, it could select distinct sets of colonizing species from neighbor ecosystems, with stress sensitive species occupying refuges created by abiotic heterogeneity and stress tolerators dominating the more exposed areas. In this case selection would occur at both biogeographical and local scales. Under its prevalence, we expect selection to have a major signature in the woody plant community structure in a phylogenetic null model. Alternatively, if abiotic heterogeneity is not strong enough to impose significant selection, the environment would be homogeneously stressful for the majority of species, and would be dominated species selected at biogeographical scale only. Under its prevalence, we expect drift to have a major signature in the phylogenetic null model. We used an analytical framework based on phylogenetic and community structure null models to assess the relative importance of ecological processes. We also aimed to characterize the ecosystem features that impose selection and dispersal limitation. We found that 95.1% of turnover in composition is attributable to drift, 2.4% to homogenizing dispersal, 2.1% to selection, and just 0.40% to dispersal limitation, thus confirming our Neutrality hypothesis. As expected, both soil and topographic variables influenced metacommunity structure. However, contrary to our prediction, light availability and vegetation structure were also important. The predominance of coarse spatial patterns correlated to topographic and soil properties suggests that coarse differences in wind exposure and associated vegetation and soil factors represent the main selective forces acting on the studied vegetation. The dominance of drift in the assembly of restinga heath vegetation is likely to result from homogeneously stressful environmental conditions and also from the ongoing colonization process that is taking place in the restinga by immigrants from species‐rich neighboring ecosystems.

Common species distribution and environmental determinants in South American coastal plains

“…For example, restingas in the windy, hot, and drought‐prone northeastern coast have been shown to share several species with the Caatinga dry forests, Cerrado savannas, and semi‐deciduous Atlantic Forests (Silva et al. ). The influence of neighboring major vegetation types imply an important role for dispersal in the explanation of restinga heath vegetation floristic variation.…”

“…, Silva et al. ), and one community null model analysis indicated that stochastic processes were dominant in the assembly of a restinga community (Silva et al. ).…”

“…), and one community null model analysis indicated that stochastic processes were dominant in the assembly of a restinga community (Silva et al. ). These evidences add up to our results and support the view of local restinga heath communities as unlikely to form cohesive networks of interacting species (Leaper et al.…”

“…, Silva et al. ). We used a large set of climatic and soil variables in restinga heath vegetation along ~9000 km of eastern South American coastal plains.…”

Weak whole‐plant trait coordination in a seasonally dry South American stressful environment

“…Lack of root traits for this functional axis could also explain part of the trait independence seen in our data, as leaf, stem, and root traits are jointly the basis for species strategies (Fortunel et al., ). However, homogeneity of soil nutrients distribution in the study area (Silva et al., ) makes strong root trait coordination unlikely.…”