The effects of forestry on golden eagles on the island of Mull, western Scotland

1 The afforestation of previously open habitats continues to involve conservation organizations in assessing effects on important species. We investigated the effects of commercial afforestation on golden eagles Aquila chrysaetos on the island of Mull, western Scotland, using long-term data on eagle reproductive success and occupancy on 30 home ranges, largely during 1981–99. 2 We modelled home range parameters in a geographical information system (GIS) that gave geographical location and predicted range use as a percentage of the total use. Resolution was to 50 × 50-m (equivalent) pixels, each with a predicted value of percentage use. Forest cover was created as a separate GIS layer set to be temporally dynamic and to reflect the stage at which commercial plantations made open ground unsuitable for golden eagles by canopy closure (12 years). 3 The layers for forest cover and range use were overlapped in the GIS to produce year-on-year estimates of the extent of open canopy forest (trees < 12 years), closed canopy forest (trees ≥ 12 years), semi-natural woodland, and open ground within each golden eagle range. 4 Based on their history of productivity, golden eagle ranges were classified using cluster analysis as either productive or unproductive. These two groups did not differ significantly in range size, mean elevation, variation in elevation, terrain ruggedness or mean cover of closed canopy forest. Nor was productivity related to these measures on ranges unaffected by commercial forestry. 5 Two golden eagle ranges were apparently abandoned by breeding eagles as a result of afforestation, but these losses were balanced by the formation of new ranges elsewhere. However, on ranges where forests were planted, standardized values of golden eagle productivity fell significantly after canopy closure. 6 Temporal trends in eagle productivity on ranges where forest had been planted differed significantly from ranges where no forest was planted. The productivity of forested ranges declined markedly in the mid-1990s when forest cover exceeded 10–15% of the areas probably used by range-holding golden eagles. 7 In a general linear model, using ranges with commercial forestry, productivity after canopy closure was positively associated with productivity before closure. Productivity after canopy closure was unrelated to range size, and only weakly related to the change in forest cover (P = 0·09, 13 ranges). Changes in eagle productivity due to increased forest cover were thus too variable on individual ranges to be predicted with confidence. 8 This study demonstrates that commercial forestry can adversely affect the productivity of golden eagles but the exact scale of effect is difficult to predict as even small plantations can have an adverse influence. Vacant neighbouring ranges may also influence the response of golden eagles to increasing forest cover. We caution against using set criteria of the extent of forest cover to predict whether a range will be abandoned.