Quantifying the effects of Entry Level Stewardship (ELS) on biodiversity at the farm scale: the Hillesden Experiment

The Hillesden experiment has proved to be a valuable, practical test of the effectiveness of the Entry Level Stewardship scheme for a range of taxa at the farm-scale. The experiment was a large scale, randomised block experiment with three treatments applied to replicated land areas of c. 50-60 ha: i) Cross compliance (CC, conventional intensive arable management subject to the EU rules of cross compliance), ii) Entry Level Stewardship (ELS, a small suite of options selected to replicate a ‘standard’ agreement comprising 1% area out of production), iii) Entry Level Stewardship Extra (ELS X, a more extensive selection of options targeted at farmland wildlife of conservation interest comprising 5% area out of production). The results show that over a five year period (2006-2011) the effects on (i) habitat quality and food resources, and (ii) the abundance, diversity and population dynamics of some key farmland taxa were significantly enhanced by both targeted local management and enhancements at the landscape scale. However there were clear effects of scale on the responses of different taxa (see Table 1.1). Mobile species like bumblebees, moths, birds and small mammals were more subject to the effects of wider landscape context and showed a net response to the whole farm manipulation. The response of other taxa e.g. plants and some invertebrate groups, were more influenced by local environmental effects and plot level manipulations. In general resource quality was higher in the more ‘targeted’ ELSX options. However there was high variability in results reflecting species specific responses to habitats at both local and landscape scales. Critical to the success of any ELS agreement is the quality of plot establishment and maintenance. Often if a sown option is not doing well after the first year, it will be unlikely to improve and should be abandoned. Both ground conditions at establishment and subsequent weather conditions contributed to the high plot variability observed. We suggest that greater emphasis should be placed on creating a much higher proportion of field margin habitat that gives a higher probability of providing a range of critical both winter and summer resources for farmland wildlife. It is not possible to provide these resources within a single margin type, so a diversity of margins is required at the farm scale.