Salt marshes supply vital ecosystem services (ES), providing material goods and recreation space, regulating natural hazards, and supporting diverse wildlife. However, increases in the utilisation of one ES can lead to reductions or ‘trade-offs’ in others. Because salt marshes are commonly used for grazing livestock, it is important to understand how this grazing impacts the saltmarsh ecosystem, and the consequences for ES supply. This thesis (i) uses a global meta-analysis to investigate the effects of livestock grazing on saltmarsh properties, and finds multiple significant changes to soil, vegetation and fauna properties. The meta-analysis reveals that the response of soil carbon is context dependent – there is no effect in Europe but a reduction in the Americas. (ii) Extensive surveys of soil carbon in grazed and ungrazed US marshes, controlling for key covariates, confirm that grazing trades-off against carbon storage in US marshes. These observational surveys, together with 18-month experimental exclusion of horses from a salt marsh in Georgia, show that grazing also disrupts the plant community in US marshes, but has little effect on resident invertebrates. (iii) Focussing on bees in salt marshes, a three-year study in south Wales, UK shows that grazing trades-off against bee habitat by reducing the flower cover of two key food plants, and that increases in plant diversity with grazing do not compensate for this negative effect. (iv) Spatial analyses of seven saltmarsh ES supplied by an estuary complex in south Wales show that marshes are not achieving their potential as a bee habitat here, due to the predominance of grazing. These analyses also show that the provision of ES by salt marshes is spatially heterogeneous, dependent on management, size and location. As a whole, this thesis adds to the understanding of grazer impacts and ES trade-offs, and supplies crucial data to support evidence-based management of salt marshes.
Livestock grazing alters multiple ecosystem properties and services in salt marshes: a meta-analysis
