Food availability is thought to strongly influence the leaving phase of natal dispersal in animals with condition-dependent dispersal. We conducted a food supplementation experiment to determine the influence of extra food on the onset of dispersal and early dispersal movement patterns of 42 radio-tagged northern goshawks (Accipiter gentilis atricapillus) from 28 broods in north-central New Mexico during 1992 and 1993. We randomly assigned half of the broods each year as treatments and the other half as controls. Treatment broods were given supplemental food from hatching (late April) until mid-October. Control broods received the same visitation rate but no food. Birds were located approximately every 2 days from fledging until mid-October in 1992 and 1993. Timing of fledging and independence were not affected by the treatment. However, extra food significantly influenced post-fledging movements of juvenile goshawks. During the late fledgling-dependency period (>65 days of age until independence) control birds were located in the natal area (<2 km from nest tree) more frequently than supplemented birds. This pattern reversed after independence (approximately 82 days of age) when supplemented birds were located more frequently in the natal area than controls. After independence the control birds were never located in the natal area and by the end of September in both years the controls had all left the study area (study area boundaries were approximately 25 km from nest tree). Supplemented birds were never located outside of the study area for the duration of the experiment. We conclude that the control birds dispersed out of the study area and the supplemented birds remained. Since the experimentally fed juveniles remained near a known food source and the controls did not, our study demonstrates that food availability influences at least the first 4 months of post-fledging movement patterns in this population. These results also suggest individuals base dispersal decisions on knowledge of their environment at a local scale, which can influence juvenile recruitment.

In the Materials and methods section, under Food supplementation experiment, 3rd paragraph, 3rd sentence: Table 1 in Ward and Kennedy 1996 should not have been linked in the online version.In Fig. 2: the number over the second bar from the left should be 10 and not 7.The online version of the original article can be found at http:// dx.

No abstract

In 1992 and 1993, 28 Northern Goshawk (Accipiter gentills) broods in northcentral New Mexico were used in a supplemental-feeding experiment to determine if there was a causal relationship between food availability and survival of young goshawks. We randomly assigned 28 nests as treatments or controls. Every two days (from hatching [late April] until juvenile dispersal [mid-October]) we visited control nest stands and provided food at treatment nest stands. We measured morphological characteristics and attached tarsalmounted transmitters with mortality switches to 42 nestling Northern Goshawks when they were 21 days old. In both 1992 and 1993, we relocated birds via telemetry every other day until mid-August. We located birds at least once a week from mid-August until mid-October in 1992, and until late-November 1993. There were no significant differences in nestling size. Treatment birds had a significantly higher survival rate during the nestling period in 1993, but not in 1992. Because most control nestlings died from predation, we attribute higher nestling survival to increased time spent in nest stands by adult females, whose presence probably deterred predators.