Chen G.G., Gu X., Liu Y., Wang W.Q., Wang M. and Maiorano L.. 2021. Diversity and Distributions, 28(2):331-345.

Aim Due to differences in community functional traits and composition, the response of communities to environmental changes can be variable, and thus an understanding of the mechanisms underlying differential functional group responses is necessary. We investigated a mangrove soil mollusc community before (in 2007) and after (in 2008 and 2009) an extreme cold event (ECE) to determine the effects of the ECE on co-occurrence patterns, spatiotemporal turnover and ecological processes in different functional groups (i.e. in-faunal shelled deposit feeders [ISDFs] and hard-bodied, mobile scavenger/predators [MS/Ps]). Location Southeast coasts of China. Method We used network analysis and null models to test the co-occurrence patterns of the soil mollusc community. We used the analysis of distance and time decay to characterize the spatial turnover and temporal succession of the mollusc community based on presence/absence, relative abundance and biomass data. We performed random forest (RF), variation partitioning analysis (VPA) and Mantel tests to detect the effects of local environmental filtering and dispersal limitations on the soil mollusc community. Results We found a decrease in species richness and biomass and an increase in abundance after an ECE, reflecting a significant change in mangrove soil mollusc communities. The network modularization, divergent succession and geographical differentiation of the ISDFs increased from 2008 to 2009, but environmental filtering and regional dispersal limitations decreased. The temporal trends of the network modules, geographical differentiation and dispersal limitations of MS/Ps followed a hump-shaped pattern from 2008 to 2009, whereas the temporal trends of environmental filtering and temporal divergence exhibited an invertedhump-shape pattern. The ECE changed the main environmental drivers of ISDFs and MS/Ps from water salinity to meteorology and vegetation factors. Main conclusions Mangrove soil mollusc communities in different functional groups invoked different spatiotemporal turnover models and ecological processes to shape unique co-occurrence patterns in response to an ECE. We highlight the value of dividing different functional groups in developing predictive frameworks for a better understanding of community functional responses to global change.

Figure 1. The species richness (a, b, c), abundance (d, e, f) and biomass (g, h, i) of the mollusc communities before (2007) and after (2008 and 2009) an extreme cold event (ECE). The “n” indicates the number of samples. p < 0.05 was the significance level in the one-way analysis of variance (ANOVA). “ISDF” indicates the in-faunal shelled deposit feeders, “MS/P” reflects the hard-bodied, mobile scavenger/predators and “All” means the total mollusc community. “*” indicates p < 0.05 and that it passed α_o = 0.05 (<0.017) in the Bonferroni test, “ns” indicates p > 0.05 and that it did not pass the level of α_o = 0.05 (>0.017) in the Bonferroni test.