LONG-TERM DYNAMICS OF HAZEL GROUSE POPULATIONS IN SOURCE-DOMINATED AND SINK-DOMINATED PRISTINE TAIGA LANDSCAPES

Hazel grouse Bonasa bonasin were censused dong transects every autumn for 30 yr in a source-dominated and a sink-dominated landscape in a pr istine taiga forest near the Ural Mountains of European Russia. The so urce-dominated landscape contained somewhat less than 4.7 times the pr oportion of source habitat than the sink-dominated landscape. Populati on density was twice as high in the source-dominated landscape (P = 0. 002), but variability was not different (P > 0.10). In the source-domi nated landscape there was a residual carry-over effect from par to yea r, as indicated by a significant autocorrelation in density with a l-y r time lag. However, in the sink-dominated landscape there was no sign ificant autocorrelation with a I-yr time lag. Population trends were s tationary over the 30-yr period in both landscapes. Power analyses sug gested that autocorrelation with I-yr time lag, i.e. correlation betwe en density in year x and x+l, was better to distinguish source from si nk habitats than differences in density or variability.