Abstract：
Environmental filtering, spatial factors and species interactions are fundamental ecological mechanisms for community organisation, yet the role of such interactions across different environmental and spatial settings remains mostly unknown. In this study, we investigated fish community organisation scenarios and seasonal species-to-species associations potentially reflecting biotic associations along the Qiupu River (China). Based on a latent variable approach and a tree-based method, we compared the relative contribution of the abiotic environment, spatial covariates and potential species associations for variation in the community structure, and assessed whether different assembly scenarios were modulated by concomitant changes in the interaction network structure of fish communities across seasons. We found that potential species associations might have been underestimated in community-based assessments of stream fish. Omnivore species, since they have more associations with other species, were found to be key components sustaining fish interaction networks across different stream orders. Hence, we suggest that species interactions, such as predation and competition, likely played a key role in community structure. For instance, indices accounting for network structure, such as connectance and nestedness, were strongly correlated with the unexplained residuals from our latent variable approach, thereby re-emphasising that biotic signals, potentially reflecting species interactions, may be of primary importance in determining stream fish communities across seasons. Overall, our findings indicate that interaction network structures are a powerful tool to reflect the contribution of potential species associations to community assembly. From an applied perspective, this study should encourage freshwater ecologists to empirically capture and manage biotic constraints in stream ecosystems across different geographical and environmental settings, especially in the context of the ever-increasing impacts of human-induced local extinction debts and species invasions.
摘要：
环境过滤,空间因素和物种相互作用是社区组织的基本生态机制，然而，这种相互作用在不同环境和空间环境中的作用仍然未知。在这项研究中,我们调查了可能反映秋浦河沿岸生物关联的鱼类群落组织情景和季节性物种间关联(中国).基于潜在变量方法和基于树的方法，我们比较了非生物环境的相对贡献，群落结构变化的空间协变量和潜在物种关联，并评估了不同的组装方案是否受到季节鱼类群落相互作用网络结构的伴随变化的调节。我们发现，在基于社区的溪流鱼类评估中，潜在的物种关联可能被低估了。杂种，因为它们与其他物种有更多的联系，被发现是维持不同流顺序的鱼类互动网络的关键组成部分。因此，我们建议物种相互作用，比如捕食和竞争，可能在社区结构中发挥了关键作用。例如,考虑网络结构的指标，例如连通性和嵌套，与我们潜在变量方法中无法解释的残差密切相关，从而再次强调生物信号，可能反映物种相互作用，在确定跨季节的溪流鱼类群落方面可能是最重要的。总的来说,我们的发现表明，相互作用网络结构是反映潜在物种关联对群落组装贡献的有力工具。从应用的角度来看，这项研究应该鼓励淡水生态学家在不同的地理和环境背景下，以经验方式捕获和管理河流生态系统中的生物约束，特别是在人类引起的当地灭绝债务和物种入侵的影响越来越大的背景下。
