Welsh Journals

The effect of agricultural improvement on some upland breeding birds in Wales A. C. Knight and K. B. Shepherd In recent years there has been considerable concern amongst conservationists about the effects of modern agricultural practices on upland bird communities (Dixon 1984; RSPB 1984). The present study was designed to investigate the effects of moorland improvement on breeding Meadow Pipits (Anthus pratensis) and Skylarks (Alauda arvensis). It was also designed to develop the work carried out in the same area by E. Kahrom and J. M. Edington in 1982 (Kahrom & Edington 1983). It involved three main investigations: (1) a comparison of breeding bird populations on unimproved and agriculturally improved land; (2) an attempt to explain the distribution of Meadow Pipits and Skylarks, by correlation with vegetation types; (3) an investigation of the areas used by Meadow Pipits whilst rearing young. Study Area The work was carried out at Pwllpeiran Experimental Husbandry Farm near Aberystwyth. The Farm, established in 1955, is one of a chain of centres operated by the Agricultural Development and Advisory Service and set up to investigate various problems of upland sheep farming and to explore the scope for integration between agriculture and forestry (MAFF 1979). Much of the farm is rough moorland grazing, but there are several sections within which land has been improved to swards of short-cropped grass (Pwllpeiran 1974). Improvement can only be carried out on land accessible to machinery, so its intensity varies between farms. At Pwllpeiran, the sections designated for improvement contain inaccessible wet or steeply sloping areas. Within the study area, 43% of the improved section has been transformed into improved pasture. In this respect, Pwllpeiran is somewhat atypical. Improvement is usually more intensive, with little provision of unimproved patches. The unimproved patches, however, are grazed more heavily because of the higher stocking levels in the improved sections (2-0 ewes/hectare on improved, 1.3 ewes/hectare on unimproved). A 330 hectare study area was chosen, comprising two adjacent 165 hectare sections of improved and unimproved land, and lying at an elevation of 490-570 metres. The land improvement within the study area was carried out during 1972-76. Vegetation types are shown in Fig. 1 and described in Table 1. Methods (a) Breeding bird survey. The survey followed the technique devised by the British Trust for Ornithology for the Common Birds Census, derived from the Enemar technique (BTO 1983). Particular attention was given to the simultaneous location of different birds. The two 165 hectare sections of improved and unimproved land were divided into 9 plots of suitable size such that one could be thoroughly worked per man per morning (Fig. 1). Census work was carried out between soon after dawn and late morning on 24 days between 18 April and 4 June. The same amount of time per hectare was spent on each of the 9 plots; a total of 58 man-hours was spent on each of the two 165 hectare sections. On each visit, a plot was traversed in a systematic manner and activities of potential significance to territory location (e.g. singing, fighting) were plotted on daily maps. The maps used were large-scale (1 :3400) ones drawn up from a detailed vegetation map available from the Farm, the Ordnance Survey maps and additional features (fences, gates, posts and peat-erosion features) obvious to the observer in the field. Because of the abundance of Meadow Pipits, coupled with the known tendency of the species to behave unobtrusively (Seel & Walton 1979), any accurate determination of individual territory boundaries throughout the whole study area was unrealistic. To overcome this problem, two 15 hectare sample plots were marked out, in positions judged to be representative of their respective sections in terms of elevation and vegetation type (Fig. 1). Within the sample plots, a concentrated effort was made to determine Meadow Pipit