The continental shelf in the southwestern Atlantic Ocean is among the six richest marine regions for biodiversity in the Southern Hemisphere, and its subtropical region is one of the fastest-warming hotspots. Thus, climate change could profoundly affect future species distributions. We investigated future climate-induced changes in fish larvae and harvested fish taxa in the subtropical southwestern Atlantic Ocean using a community-based modelling technique (gradient forest). This approach integrates information on multiple species rather than treating each species individually, as is typical in many species distribution approaches. We addressed two primary questions: how might climate change affect fish larval communities, and will communities of harvested fish (juveniles and adults) taxa respond similarly. Using two climate change scenarios (moderate RCP 4.5 and ‘business as usual’ 8.5), we found that fish larvae and harvested taxa are influenced differently by environmental variables, with differences in both the level and shape of the response to environmental drivers. Chlorophyll a and sea surface temperature were the most important predictors for fish larvae communities, while depth and sea surface salinity best predicted the harvested community. However, both communities are expected to move southwards in response to climate change, with greater changes in community composition predicted in the southern portion of the study area for both fish larvae and harvested taxa. To our knowledge, this is the first study to investigate the effect of future climate change on a suite of taxa of fish larvae and adults. We also suggest that modelling the integrated response of a suite of species to environmental predictors using community-modelling approaches such as gradient forest could provide robust projections and novel insights into community changes.

西南大西洋的大陆架是南半球生物多样性最丰富的六个海洋地区之一，其亚热带地区是增速最快的热点之一。因此，气候变化可能会深刻影响未来的物种分布。我们使用基于社区的建模技术（梯度森林）调查了亚热带西南大西洋海洋中未来气候引起的鱼幼虫和收获鱼类群的变化。这种方法整合了多个物种的信息，而不是像许多物种分布方法中那样对每个物种单独进行处理。我们解决了两个主要问题：气候变化将如何影响鱼类幼虫群落，收获的鱼类群落（幼鱼和成年）分类群也会做出类似反应。使用两种气候变化情景（适度的RCP 4.5和“照常营业” 8.5），我们发现鱼幼虫和收获的分类单元受环境变量的影响不同，对环境驱动因素的响应的水平和形式也有所不同。叶绿素a和海面温度是鱼类幼虫群落最重要的预测指标，而深度和海面盐度最能预测收获的群落。但是，预计这两个社区都将向南移动以应对气候变化，预计在研究区域的南部，鱼幼虫和收获的分类单元的社区组成将发生更大的变化。据我们所知，这是第一个研究未来气候变化对一组鱼幼虫和成虫的影响的研究。我们还建议，使用诸如渐变森林之类的社区建模方法对一组物种对环境预测因子的综合响应进行建模，可以提供可靠的预测和对社区变化的新颖见解。