Coastal grasslands, such as salt marshes and sand dunes, provide many important ecosystem services including ‘supporting services’ (soil formation, primary productivity and nutrient cycling), ‘provisioning services’ (fresh water supply, food and fibre products, bio-chemical or genetic resources), ‘regulating services’ (equable climate, pollution control, flood prevention, invertebrate pollination and pest regulation) and ‘cultural services’ (recreation, education and aesthetic appreciation). Historically, salt marsh and sand dune grasslands were commonly used as agricultural livestock grazing land. Currently, some of these coastal grasslands are ‘conservation grazed’ (i.e. extensively grazed to maximise plant diversity and to provide a suitable habitat for over-wintering bird species), others have been ‘abandoned’ (i.e. large herbivores removed) due to the removal of agricultural subsidies or remain historically ‘un-grazed’. Grazing management of coastal grasslands influences biological and physical habitat characteristics, ecosystem function, biodiversity and ecosystem service delivery. Understanding the impact of grazing is therefore vital to enable future robust management recommendations. Biodiversity is often used as an indicator of ecosystem health and ecosystem service provision with conservation priorities allocated accordingly. It is therefore essential to critically assess just how important biodiversity is to the provision of ecosystem services within a wide range of habitats. The review chapter draws together evidence for this argument from salt marsh and sand dune habitats with the conclusion that functional diversity and composition are more important than biodiversity per se (Chapter 2). The experimental chapters of this thesis deal with the impact of grazing upon temperate salt marsh and sand dune grassland biodiversity and ecosystem service provision. ‘Grazed’ (cattle grazed < 8 cm) and historically ‘un-grazed’ upper salt marsh plots were compared. ‘Fully grazed’ (ponies 0.2 ha-1, cattle 0.05 ha-1 and rabbits 45 ha-1), ‘rabbit grazed’ and ‘un-grazed’ (for 8 years) fixed sand dune grassland plots were also evaluated. Firstly, how grazing management affected ecosystem service provision of sand dune grassland was examined, by measuring a wide range of biophysical variables as proxies for ecosystem services (Chapter 3). ‘Supporting’ and ‘regulating’ services were provided predominantly by the un-grazed, ‘provisioning’ and ‘cultural’ services by the extensively grazed grassland. Secondly, the impact of short sward cattle grazing on the abundance, composition and diversity of the ground dwelling invertebrate community of an upper salt marsh was assessed using pitfall traps (Chapter 4). The findings showed that both cattle grazed and un-grazed saltmarsh habitat should be maintained to maximise invertebrate abundance and diversity and provide suitable habitat for coastal specialists. Thirdly, greenhouse gas emissions from grazed and un-grazed salt marsh were measured monthly for one year. Additionally, below-ground gas sampling tubes were used to measure soil methane concentrations (Chapter 5). Carbon dioxide efflux was greater from the un-grazed marsh soil but ‘hotspots’ of methane efflux were only found on the grazed marsh. Finally, the influence of grazing on the soil microbial community of both salt marsh and sand dune grasslands was measured by microbial biomass (fatty acid phospholipids: PLFAs), bacterial growth rate (Leucine incorporation) and respiration rates (Chapter 6). Microbial biomass, PLFA markers and bacterial growth rate were all influenced by grazing management. In summary, this work concludes that grazing management clearly affects biological and physical habitat characteristics, biodiversity, ecosystem function and ecosystem service delivery (Chapter 7). Management of coastal grasslands evidently involves trade-offs between biodiversity conservation and multiple ecosystem service provision.