Factors affecting the nitrogen status of soils and ground flora in Beech woodlands

In Autumn 2001, a UK wide survey was conducted to investigate the nitrogen status of soils and vegetation under 20 beech woodlands on brown earth soils. Predictor variables (including nitrogen pollution levels, ecosystem biological demand and soil–water flux) were used to explain measurements of nitrate at depth (55–60 cm). Indicators of nitrogen pollution levels included estimates from the national mapping database, the distance to the edge of each wood, foliar nitrogen levels in ground flora and measurements of soil and litter nitrogen. Biological demand was represented by ground cover, soil carbon and tree age, diameter at breast height (DBH) and density. Indices used to reflect soil water flux included precipitation, precipitation excess and particle size distribution. Data were also collected to explain the character of the ground flora at the sites as expressed by Ellenberg values of nitrogen, light and reaction. Results showed that the distribution of ground cover was primarily related to canopy gap fraction—this also governed species composition along with soil pH and nitrogen inputs (as represented by site proximity to edge of the wood). However, ground flora composition itself was a poor predictor of nitrate levels at depth as were soil and litter nitrogen levels. The latter were controlled by soil moisture, carbon content and pH. Nitrate levels in the sub soils were highly skewed; 80% of the measurements were less than 40 mg l^-1, yet the mean concentration was 37 mg l^-1 due to extreme values at three sites with extraordinary local conditions. A multiple linear regression, including bramble (Rubus) foliar nitrogen levels, DBH, precipitation, tree density and ground vegetation cover, explained 85% of the variation in sub soil nitrate levels when these three extreme site were excluded. An acceptable regression which requires minimal field work is also presented.