LAUSR

Winter quantitative studies on the Rook Corvus frugilegus, Hooded Crow Corvus cornix and Jackdaw Corvus monedula were conducted independently by two observers in the season 1999/2000 in fragments of urban green areas in Bydgoszcz and Toruń. Each observer visited all plots six times at different dates and did two subsequent counts the same day, noted as the early and the late one, respectively. The number of birds from all single counts done in the same plot by one observer at a stable time of a day was averaged and called a control. Comparisons of controls obtained from the same plot by the same observer were called the “effect of repeatability of counts” while comparisons of controls between the observers from the same area were defined as the “observer’s effect”. No statistically significant differences were found in estimates of the similarity of the bird (Renkonen’s coefficient) domination structure between the “observer’s effect” and the “effect of comparability of counts” (Mann-Whitney U-test: U = 12, n = 12, P = 0.50). Statistically significant differences between totals for each species stated during single counts by the same observer within the same day were found in both study plots and for both observers (χ2-test: df = 1, P < 0.05). Differences in bird numbers stated in subsequent controls in the same site were statistically significant only in the case of the plot BYD (χ2 = 25.93, df = 3, P = 0.00001). These differences occurred for both „effect of repeatability of counts” and „observer’s effect” (χ2-test: df = 1, P < 0.05). The impediment to unambiguous estimation of the total numbers of the studied corvids in small fragments of urban green areas mainly resulted from nomadic movements of these birds over large areas and their non-directional movements caused by high human penetration.