Correlates of the change in Ring Ouzel Turdus torquatus abundance in Scotland from 1988-91 to 1999.

Capsule The change was correlated with environmental, habitat and management variables. Aims To identify factors correlated with a population decline and range contraction between 1988-91 and 1999. Methods Regression models are used to describe the variation in Ring Ouzel abundance in 1988-91, and the change in Ring Ouzel abundance between 1988-91 and 1999. Results The abundance of Ring Ouzels in 1988-91 was negatively correlated with improved pasture, but positively with heather/smooth grass mosaic and Nardus/Molinia grassland. There was also a positive quadratic correlation with rock cover. A decline in Ring Ouzels between 1988-91 and 1999 was most likely on tetrads with shallow gradients, and outside of a 350-750 m altitudinal range. The change in ouzel numbers was also negatively correlated with both the initial area of conifer forest and heather/smooth grass mosaic. The rate of loss was greatest in those areas with a low Ring Ouzel abundance in 1988-91, although this effect was non-significant once habitat was accounted for. Conclusions The abundance of Ring Ouzels was linked to heather-grass mosaics, while declines were greater in tetrads lacking topographical features known to be selected by nesting ouzels. The negative relationship between forestry and population change suggests that large-scale afforestation of the uplands is a cause of concern. [ABSTRACT FROM AUTHOR]