Breeding skylarks Alauda arvensis on Environmentally Sensitive Area arablereversion grass in southern England: survey-based and experimental determination of density

1. In the South Downs and South Wessex Downs (southern England), designated an Environmentally Sensitive Area (ESA) in 1987 and 1993, respectively, fa rmers may undertake to conserve existing areas of downland turf(chalk grass land rich in plants and invertebrates) and to create new areas of grassland out of previously cultivated land (arable reversion), in exchange for annu al payments. These areas are subject to strict management prescriptions. 2. Singing skylarks were surveyed on 40 km(2) of downland turf, arable reve rsion, non-ESA grass and arable land between 1994 and 1996 (April-July) in the two ESAs. The emphasis was on two types of arable reversion: permanent grassland sown with agricultural grass species, and chalk downland sown wit h traditional chalk grass species. 3. The surveys showed that singing skylark density was greater on chalk dow nland arable reversion than on permanent grassland reversion, in swards abo ve 15 cm in height and in fields without boundary scrub. In general, perman ent grassland reversion did not support more skylarks than winter wheat or grass not entered in the ESA schemes; highest densities were in undersown s pring barley. Singing skylark density decreased after mowing in May-June in the South Wessex Downs ESA. 4. In 1995 and 1996, a fully randomized experiment was carried out on 12 fi elds (each c. 5 ha) at Plumpton, Sussex, to verify the survey findings. Six fields were grazed to maintain a sward below 10 cm in height (short-grazed ), and the others grazed to maintain a award between 15 and 25 cm (long-gra zed) during the skylark breeding season (April-July). 5. On average, singing skylarks were six times as abundant, and non-singing ones twice as abundant, in experimental long-grazed fields as in short-gra zed fields. The total number and number of taxa of chick-food invertebrates were about twice as large on average, and the number of grass seed heads r ecorded in July was two to 15 times as great, in long-grazed as in short-gr azed fields. 6. Grazing-exclusion cages within a randomized block experiment at New Erri ngham, Sussex, were sampled for invertebrates in mid-May 1995 and 1996. Ung razed cages (sward up to 40 cm tall) supported over five times as many inve rtebrates and over three times as many invertebrate taxa as grazed areas (s ward <2 cm tall). 7. These results contributed to revised management prescriptions for the So uth Downs (1997) and South Wessex Downs (1998) ESAs. Prescriptions to encou rage taller swards and to prevent mowing during the nesting season in the S outh Wessex Downs, and removal of the 10-cm limit on sward height, allowanc e of fertilizer on some arable reversion grassland, and an arable tier enco uraging undersown spring cereals and overwinter stubbles in the South Downs , will benefit grassland birds.