Degradation of freshwater ecosystems requires efficient tools for assessing the ecological status of freshwater biota and identifying potential cause(s) for their biological degradation. While diatoms and macroinvertebrates are widely used in stream bioassessment, the potential utility of microbial communities has not been fully harnessed. Using data from 113 Finnish streams, we assessed the performance of aquatic leaf-associated fungal decomposers, relative to benthic macroinvertebrates and diatoms, in modelling-based bioassessment. We built multi-taxon niche -type predictive models for fungal assemblages by using genus-based and sequence-based identification levels. We then compared the models’ precision and accuracy in the prediction of reference conditions (number of native taxa) to corresponding models for macroinvertebrates and diatoms. Genus-based fungal model nearly equalled the accuracy and precision of our best model (macroinvertebrates), whereas the sequence-based model was less accurate and tended to overestimate the number of taxa. However, when the models were applied to streams disturbed by anthropogenic stressors (nutrient enrichment, sedimentation and acidification), alone or in combination, the sequence-based fungal assemblages were more sensitive than other taxonomic groups, especially when multiple stressors were present. Microbial leaf decomposition rates were elevated in sediment-stressed streams whereas decomposition attributable to leaf-shredding macroinvertebrates was accelerated by nutrients and decelerated by sedimentation. Comparison of leaf decomposition results to model output suggested that leaf decomposition rates do not detect effectively the presence of multiple simultaneous disturbances. The rapid development of global microbial database may soon enable species-level identification of leaf-associated fungi, facilitating a more precise and accurate modelling of reference conditions in streams using fungal communities. This development, combined with the sensitivity of aquatic fungi in detecting the presence of multiple human disturbances, makes leaf-associated fungal assemblages an indispensable addition in a stream ecologist’s toolbox.

淡水生态系统的退化需要有效的工具来评估淡水生物群系的生态状况并确定其生物降解的潜在原因。虽然硅藻和大型无脊椎动物广泛用于河流生物评估中，但尚未充分利用微生物群落的潜在用途。使用来自113个芬兰河流的数据，我们在基于模型的生物评估中评估了与底栖大型无脊椎动物和硅藻相关的水生叶相关真菌分解剂的性能。通过使用基于属和基于序列的识别水平，我们为真菌组合建立了多分类群小生境类型的预测模型。然后，我们将模型在参考条件（天然分类单元数）的预测中的准确性和准确性与大型无脊椎动物和硅藻的相应模型进行了比较。基于属的真菌模型的准确度和精确度几乎等于我们最好的模型（无脊椎动物）的准确度和精确度，而基于序列的模型的准确度则较低，并且往往高估了分类单元的数量。但是，当将模型单独或组合应用于人为压力源（营养富集，沉淀和酸化）干扰的河流时，基于序列的真菌组合比其他分类组更敏感，尤其是在存在多个压力源的情况下。沉积物胁迫流中微生物叶片的分解速率增加，而叶片切碎的大型无脊椎动物的分解由于营养而加速，而由于沉积而降低。将叶片分解结果与模型输出进行比较表明，叶片分解速率无法有效检测到多个同时发生的干扰。全球微生物数据库的快速发展可能很快使物种相关的与叶相关的真菌的鉴定成为可能，从而有助于使用真菌群落对流中参考条件进行更精确的建模。这种发展与水生真菌在检测多种人为干扰中的敏感性相结合，使与叶相关的真菌组合成为河流生态学家工具箱中必不可少的组成部分。有助于使用真菌群落对流中参考条件进行更精确的建模。这种发展与水生真菌在检测多种人为干扰中的敏感性相结合，使与叶相关的真菌组合成为河流生态学家工具箱中必不可少的组成部分。有助于使用真菌群落对流中参考条件进行更精确的建模。这种发展与水生真菌在检测多种人为干扰中的敏感性相结合，使与叶相关的真菌组合成为河流生态学家工具箱中必不可少的组成部分。