Fig. 2. Choice of prey as a function of visual occlusion (experiment I). Each data point is the proportion (mean ± s.e.m.; N=6 birds; 9 trials per bird per day over 2 days) of choice of prey side (`correct choice'). The prey (live or a dead fish, 12–15 cm in length) was presented behind a transparent Perspex door (open squares) on days 1, 2, 3 and 9, or behind a visually opaque fabric door (solid squares) on days 4 to 8. Targets were 1.4 m from the Y-junction. Broken line indicates the critical value for significant choice (0.78).