Stocking density shapes phytoplankton community stability in rice-crayfish integrated systems.

The rice-crayfish integrated system (RCIS) has expanded rapidly across China, yet its ecological impacts on phytoplankton remain inadequately understood. To fill this gap, we conducted a 40-d field experiment with five crayfish (Procambarus clarkii) stocking densities (0, 1, 2, 3, and 4 ind./m2; CK, T1-T4) in replicated paddy plots. Water samples were collected weekly, and phytoplankton were identified and enumerated under light microscopy. Community structure, diversity indices (Margalef, Pielou, and Shannon-Wiener), co-occurrence networks, neutral community model (NCM), and Spearman correlations were analyzed. One-way ANOVA with Tukey tests was used for treatment comparisons, and network stability was assessed by robustness and vulnerability. Our results showed that crayfish stocking significantly reduced species richness and total biomass relative to CK (P < 0.05) and shifted dominant taxa. Diversity indices declined to their lowest values at T2 and then recovered: Margalef decreased from 2.67 ± 0.05 in CK to 1.40 ± 0.10 in T2 (P < 0.001), Pielou from 0.88 ± 0.01 to 0.80 ± 0.02 (P < 0.05), and Shannon-Wiener from 2.91 ± 1.03 to 2.11 ± 0.10 (P < 0.001), corresponding to reductions of 47.57%, 9.09%, and 27.49%, respectively. Co-occurrence networks were simplest and least stable at T2, with the highest vulnerability, whereas T3 had the lowest vulnerability among stocked treatments. NCM fitting showed higher stochasticity in stocked treatments (Rsqr = 0.683-0.806) and lower migration rates (Nm = 3-4) than CK (Nm = 11). Spearman correlations (significance at P < 0.05) identified dissolved oxygen, water temperature, and total nitrogen as key drivers in CK, whereas stocked treatments showed more complex nutrient-phytoplankton relationships. From a phytoplankton-based perspective, a stocking density of 3 ind./m2 appears ecologically favorable, but this recommendation requires integrated testing with production, water quality, and multi-trophic indicators.