Environment and Rural Affairs Monitoring & Modelling Programme - ERAMMP Year 1 Report 21: GMEP outstanding analysis part 2 - Revisiting trends in topsoil carbon from CS2007 to GMEP 2013-2016

New analysis was carried out to explore the reported loss of topsoil-C between 2007 and 2016 in the ‘Habitat’ category in the final GMEP report. This ‘Habitat’ category is defined as all habitats except woodlands, arable and improved grassland. The GMEP survey squares were selected using Countryside Survey protocols stratified according to Land Classes. The final GMEP survey sample from 2012-2016 consists of 7% previously surveyed Countryside Survey squares. Further analysis was needed to explore, and account for, unintended shifts in environmental variables which could have contributed to the reported topsoil carbon decline. The results indicate: 1. The reported change in the ‘Habitat’ category is driven by trends in upland habitats (median elevation of 400m). 2. In upland habitats, soil carbon is positively associated with dwarf shrub cover (particularly ericoid e.g. heather cover), Sphagnum, presence of peat, elevation and moisture conditions. 3. The coverage of dwarf shrubs was lower in GMEP than in Countryside Survey 2007, mostly due to lower cover of ericoids i.e. heather. This is consistent with decreasing soil carbon in upland habitats. Other variables (i.e. potential drivers) did not differ between surveys, or direction of change was inconsistent with reported C trends. 4. Re-analysis of Countryside Survey data (1978-2007) provides evidence that shifts over time from dwarf shrub to grass-dominated habitats are associated with a decline in topsoil carbon. 5. Overall, this suggest a potential role of ongoing vegetation change in upland habitats (i.e. conversion of dwarf shrub to grass-dominated) contributing to topsoil carbon loss. Further work is needed to: • Confirm recent vegetation change in upland habitats using independent data e.g. satellite data; • Explore relationships between specific plant species and topsoil carbon in Countryside Survey where we have a high number of true repeat samples; This work highlights the importance of the findings of the next ERAMMP survey, which will be more powerful than the combined CS-GMEP approach reported here.