Clustering Coefficient Supports Phytoplankton Network Stability During Short-Term Hydrological Pulses in a Lake...

Long-term environmental disturbance studies have shown that the network stability is primarily affected by specific network complexity parameters, particularly the average degree and modularity. Whether these conditions also apply to short-term environmental disturbances remains unknown. Using short-term hydrological pulses in Poyang Lake as a model, this study integrated field surveys with co-occurrence analysis to investigate phytoplankton dynamics and network structure across distinct hydrological periods (low- and high-water levels). Our results revealed that elevated water levels triggered compositional homogenization and simplified co-occurrence network structure. This structural simplification was indicated by simultaneous weakened network complexity (including reduced network size, average degree, and average clustering coefficient) and diminished stability (less keystone taxa, lower robustness, and higher vulnerability) during the high-water level period. Contrary to long-term disturbances linked to network modularity or average degree, the clustering coefficient was identified as primary stabilizer under short-term hydrological pulses. In Poyang Lake, the high clustering coefficient was strongly positive associated with phosphorus availability (total phosphorus: TP and phosphate: PO₄^3-). These findings demonstrate a timescale-dependent mechanism of ecological network stability, where stabilizing short-term responses relies on local connectivity shaped by system specific resource drivers (e.g. phosphorus), thereby providing a new perspective for predicting and managing aquatic ecosystem stability under climate change.