Evaluating seasonal change in the live bivalve assemblage of Mumbai and the impact of pollution

Abstract

Live bivalve communities are sensitive indicators of shallow marine ecosystem health, as they are influenced by pollution and seasonal changes. Yet studies simultaneously evaluating both stressors on shallow marine communities remain limited, particularly in tropical monsoon-influenced systems. This study evaluates seasonal changes in live bivalve assemblages and the impact of pollution on them across six selected shallow marine sites (two open coastal and four creek) in Mumbai. These sites span a gradient of pollution. Sampling was conducted during the pre-monsoon (February) and post-monsoon (November) seasons in the year of 2024 at multiple depths. Pollution data was obtained from the Maharashtra Pollution Control Board (MPCB) 's annual water quality and pollution report.

Bivalve community composition shifted significantly between seasons (PERMANOVA R² = 0.061, p = 0.001), with post-monsoon assemblages showing higher taxonomic diversity indices when all sites are pooled together. At the site level, the same pattern was observed, but it was not significant. Habitat type (creek vs open coastal) explained marginally more community variation than season alone (R² = 0.079), with both factors together explaining 14% of total variation. Despite significant taxonomic turnover, functional trait composition did not change significantly across seasons, indicating functional redundancy- only pollution-tolerant functional types mostly persist year-round. There was no significant correlation between taxonomic diversity and water quality index across sites. To understand what drives changes in the bivalve community in polluted Mumbai waters, multiple Generalized Linear Models (Poisson family, log link) were run, with water quality parameters as predictors of species richness and functional richness. BOD and DO are the best predictors of pre-monsoon species richness, with statistical significance. For the dual-stress scenario, season was also added as a predictor, and the BOD+COD+DO+season model was the best. Pollution parameters and season couldn’t explain a large percentage of the variation in the community, suggesting the need to add other environmental parameters to the analysis, potentially reflecting other sources of stress or influence. Two non-native, pollution-tolerant taxa - Mytella strigata and Mytilopsis sp. - dominated the rank- abundance distribution, raising concern about pollution-facilitated biological invasion.