Data from: Precipitation mediates the effect of human disturbance on the Brazilian Caatinga vegetation

Seasonally dry tropical forests (SDTFs) are one of the most threatened forests worldwide. These species-rich forests not only cope with several acute (e.g. forest loss) and chronic (e.g. overgrazing and firewood extraction) human disturbances, but also with climate change (e.g. longer and more severe droughts); yet, the isolated and combined effects of climate and acute and chronic human disturbances on SDTF vegetation are poorly known.
Given the environmental filter imposed by drought in SDTFs, the composition and structure of vegetation is expected to be strongly associated with annual precipitation, and thus the effects of human disturbances on vegetation may also depend on precipitation (i.e. interacting effect).
We tested these hypotheses in the Brazilian Caatinga – a SDTF threatened by climate change and human disturbances. We evaluated the isolated and combined (both additive and multiplicative) effect of precipitation, a chronic disturbance index and acute disturbance (landscape forest cover) on the diversity, stem density, evenness, taxonomic composition and aboveground biomass of adult trees and shrubs across 19 0.1-ha plots distributed along a disturbance and precipitation gradients.
We recorded 5541 stems from 129 species. Precipitation showed a stronger (positive) effect on species diversity than acute and chronic disturbances and, as expected, the effect of disturbance depended on precipitation (interacting effect): i.e. species diversity (especially the number of rare species) was negatively related to forest loss but positively related to chronic disturbance in wetter sites, whereas in drier sites, species diversity was weakly related to forest cover, but strongly and negatively related to chronic disturbance. Contrary to species diversity, community evenness, stem density and biomass were weakly related to all predictors.
Synthesis. Precipitation appears to be a strong environmental filter determining the distribution of water-demanding plant species. Chronic disturbance in wetter (high-productive) forests may favour species diversity by increasing ecosystem heterogeneity (intermediate disturbance hypothesis). Yet, the biodiversity costs of chronic disturbance are higher in drier (low-productive) forests; i.e. there is a co-limitation imposed by drought and disturbance in drier forests. Overall, our findings indicate that rapid climatic changes in the region will probably have strong negative effects on this SDTF.

季节性干旱热带林（Seasonally Dry Tropical Forests, SDTFs）是全球受威胁程度最高的森林类型之一。这类物种丰富的森林不仅需承受多种急性（如森林丧失）与慢性（如过度放牧、薪柴采伐）人为干扰，还面临气候变化带来的压力（如持续时间更长、强度更高的干旱事件）；然而，气候因子、急性与慢性人为干扰对SDTF植被的单独及联合效应，目前仍缺乏深入研究。
鉴于SDTF生境中干旱所施加的环境过滤（environmental filter）作用，植被的组成与结构预计与年降水量密切相关，因此人为干扰对植被的效应可能也会随降水水平发生变化（即存在交互效应）。
我们在受气候变化与人为干扰威胁的巴西卡廷加（Caatinga）——一类典型SDTF——中验证了上述假说。我们在沿干扰与降水梯度布设的19块0.1公顷样地内，评估了降水、慢性干扰指数与急性干扰（景观森林覆盖度）的单独及联合（加性与乘性）效应，对成年乔灌木的物种多样性、茎干密度、群落均匀度、分类组成与地上生物量的影响。
本次调查共记录到129个物种的5541株茎干。降水对物种多样性的正向效应强于急、慢性人为干扰；且如预期所示，人为干扰的效应依赖于降水水平（即存在交互效应）：具体而言，湿润生境中的物种多样性（尤其是稀有种数量）与森林丧失呈负相关，与慢性干扰呈正相关；而在干旱生境中，物种多样性与森林覆盖度关联较弱，但与慢性干扰呈强负相关。与物种多样性的响应模式不同，群落均匀度、茎干密度与地上生物量与所有预测变量的关联均较弱。
综合分析结果表明，降水是决定需水植物物种分布的关键环境过滤因子。在湿润（高生产力）森林中，慢性干扰可通过提升生态系统异质性（中度干扰假说，Intermediate Disturbance Hypothesis）促进物种多样性。然而，慢性干扰带来的生物多样性代价在干旱（低生产力）森林中更为显著——即干旱生境中存在干旱与干扰的共同限制作用。总体而言，本研究结果显示，该区域快速的气候变化可能会对这类SDTF产生强烈的负面影响。