We investigated which environmental variables of floodplain lakes act as potential drivers of fish assemblages and how they explain the variation in taxonomic and functional diversity of the community. We evaluated the taxonomic richness, functional dispersion (the distribution of species abundances with different traits) and redundancy (how similar the species are) of fish communities from six floodplain lakes of the upper Paraná River floodplain, evaluating whether they respond to the same set of predictor variables. We predict that the variation in taxonomic richness will be explained by limnological variables that express the characteristics of the water and the functional variation by variables that express the physical structure of the habitat. We sampled limnological and habitat structural variables and fish communities of each floodplain lake in a span of 14 years. Functional diversity was evaluated from six functional traits. The three diversity indices were used as response variables in Generalized Linear Mixed Models (GLMMs), while environmental variables were used as predictor variables. Taxonomic richness was best explained by total phosphorus, water transparency, depth and water level, while functional dispersion was explained by water level. Functional redundancy was explained by the water transparency variable. Our results show that taxonomic and functional diversity metrics may have environmental predictors in common, but the taxonomic richness is more predictable depending on the environmental gradient than the functional diversity. Besides, the floodplain lakes showed high functional redundancy. This may suggest that in functionally redundant communities, the main drivers of different diversity metrics do not differ.

中文翻译：

---

功能冗余的社区在不同多样性指标的主要环境驱动因素上没有表现出差异

我们调查了漫滩湖泊的哪些环境变量是鱼类组合的潜在驱动因素，以及它们如何解释群落分类和功能多样性的变化。我们评估了巴拉那河上游泛滥平原六个泛滥平原湖泊鱼类群落的分类丰富度、功能分散（具有不同特征的物种丰度的分布）和冗余度（物种的相似程度），评估它们是否对同一组预测变量。我们预测分类丰富度的变化将由表达水特征的湖沼学变量和表达栖息地物理结构的变量的功能变化来解释。我们在 14 年的时间里对每个洪泛区湖泊的湖沼学和栖息地结构变量以及鱼类群落进行了采样。从六个功能性状评估功能多样性。三个多样性指数被用作广义线性混合模型（GLMMs）中的响应变量，而环境变量被用作预测变量。总磷、水透明度、深度和水位最好解释分类丰富度，而水位解释功能分散。功能冗余由水透明度变量解释。我们的结果表明，分类学和功能多样性指标可能具有共同的环境预测因子，但与功能多样性相比，根据环境梯度，分类丰富度更容易预测。除了，泛滥平原湖泊显示出高度的功能冗余。这可能表明，在功能冗余的社区中，不同多样性指标的主要驱动因素没有差异。