Estuarine and coastal environments provide a wide range of societal goods and services that need to be strategically managed to ensure sustainable use of resources. Ecosystem service (ES) assessments are transitioning from individual ES to multi-service assessments that consider interactions and grouping of ES (i.e. bundles). This thesis investigates the use of ecological mechanism (i.e. the links between ecological processes, functions, and ES) to derive insights into the associations between multiple ES provided by marine bivalves and the implications for management. In data-scarce marine environments, conventional spatial methods for assessing ES interactions and bundles are not always feasible. However, advancements can be made by focussing on ecological mechanisms, as for bivalves their ecological role has been extensively studied. The complex links between processes, functions, and ES were identified for bivalves, which were used to derive four ES bundles based on shared underpinning mechanisms. This study provides detailed descriptions of the specific (set of) ecological mechanisms per bundle, thereby gaining insights in how the ES are formed, the interactions between ES (synergies and trade-offs) and the environmental stressors they are prone to. Quantification of ES remains challenging, and measuring them in functionally similar species in different habitats is rarely considered, although ecological studies indicate differences are likely. I measured in situ bivalve contribution to water quality regulation for two functionally similar bivalve species (infaunal suspension-feeders) that dominate different estuarine habitats (subtidal and intertidal). Benthic chambers were used to measure solute fluxes (oxygen and nitrogen) and bed clearance rates as proxies for ES. Empirical findings indicated higher hourly productivity, nitrogen recycling, and water column filtration in the intertidal. However, when converted to daily ES estimates these patterns did not persist, emphasising the unequal contribution to functions and ES by functionally similar species and the non-linear scaling between them. Many ecological processes and functions are density dependent, hence spatial information on the distribution and density of bivalves is an important prerequisite to estimate ES. Probability of occurrence and density were predicted using Species Distribution Models (SDMs) for two estuarine bivalve species. Fine scale (100 m resolution) predictions showed different spatial patterns depending on habitat association. Species with a narrow distribution displayed good congruence between occurrence and density predictions, whereas species with a wide distribution demonstrated that a high probability of occurrence does not always equate to high density. Simultaneously considering both occurrence and density will improve decision-making and identify areas of greatest ecological value to the species of interest. Ecological mechanisms can be further applied to advance spatial predictions of multiple ES in marine environments, through process-based models that incorporate ecological principles to derive ES potential and assess natural variability in ES. Four ES (food provision, water quality regulation, nitrogen removal, and sediment stabilisation) were spatially predicted for two bivalve species. Spatial patterns in ES related to the species habitat association, with varying quantities and were driven by environmental conditions. Hotspot analyses on combined ES maps per species identified high-density areas for the provision of multiple ES simultaneously. These models provide a versatile tool to inform current management practices and can be further applied to test management scenarios. Collectively the results from this thesis demonstrated the utility and effectiveness of ecological mechanisms to determine ES interactions and bundles, and the spatial heterogeneity observed across estuaries. It also highlighted the interconnectedness of marine ES and the implications for their management.