We analyzed migration counts of American Kestrels (Falco sparverius) at 20 autumn migration hawkwatch sites throughout North America to estimate population trends. Prior to trend analysis, we converted counts to indexes adjusted for effort and patterns of passage. In eastern North America, autumn counts showed a pattern of significant declines along the Atlantic Coast and eastern Appalachian Mountains, stable numbers in the eastern Great Lakes region, and a significant increase in the western Great Lakes region from 1974 to 2004. From 1994 to 2004, significant declines occurred at most hawkwatch sites in eastern North America, with nonsignificant declines recorded in the western Great Lakes, northeastern Quebec, and Florida. In western North America, three counts from the Intermountain and Rocky Mountain regions that spanned the mid-1980s to 2005 showed mixed long-term trends. From 1995 to 2005, counts decreased significantly at three and nonsignificantly at two of six western hawkwatch sites active throughout the period. These results suggest population declines across much of interior western North America and the Pacific Northwest; however, kestrel counts remained comparatively stable since the mid-1980s in the southern Rocky Mountains. Migration counts along the Gulf of Mexico were variable from 1995 to 2005, with no strong indication of population changes in this region. The overall pattern of migration counts suggests that kestrel populations have undergone a long-term decline in northeastern North America and more recent declines in the midwestern and western regions of the continent.

Declines in American Kestrel (Falco sparverius) populations are widely reported, and Breeding Bird Survey (BBS) data suggest that the North American population declined significantly from 1984 to 2007. Potential causes include the spread of West Nile virus (WNV), increases in populations of Cooper's Hawks (Accipiter cooperii), and loss of suitable habitat. We examined trends in the numbers of both migratory and resident kestrel populations that use nest boxes in eight study areas in Florida, Georgia, Virginia and Maryland, New Jersey, Massachusetts, Pennsylvania, Saskatchewan, and the Yukon Territory, 1984–2007. All eight populations underwent significant declines; the mean annual decline in nest-box occupancy rate was 3.0% and ranged from 0.6% in Pennsylvania to 4.7% in New Jersey. Except for the most recent nest-box program, established in 1995 and declining since 2002, all nest-box populations began to experience declines before WNV arrived in North America in 1999. To test whether changes in kestrel population densities generally are associated with the opposite trend in Cooper's Hawks, we examined the 42 BBS physiographic regions for which trends for both species were available. No significant correlations were detected for the period 1966–2007, or for 1980–2007, more closely concurrent with our nest-box data. Christmas Bird Count data from 1959 through 1988 also failed to demonstrate a significant correlation. Finally, the habitat within our study areas still appears suitable, and the remaining kestrels appear healthy and have high reproductive success. Thus, the principal cause of the decline probably lies elsewhere, perhaps on the wintering grounds or along migration routes. Further, for both migratory and resident populations, the decline in nest-box occupancy may reflect regional declines, which would reduce the number of individuals available for replacing breeding birds that have died or dispersed.

We monitored American Kestrels (Falco sparverius) nesting in boxes from 1986–2006 in southwestern Idaho. The number of boxes available to kestrels each year varied from 34 to 126. Approximately 20% of boxes were on highway signs along Interstate 84, 20% were on trees in rural residential properties, and 60% were on wooden poles in rangeland, irrigated agricultural, and exurban areas. Occupancy rates increased over the 21-year period, despite increasing human populations, housing developments, and vehicular traffic near the boxes. The percent of boxes occupied each year averaged 48 ± 14, and the percent of nesting attempts that raised young to 22 d old averaged 64 ± 10. Clutch size averaged 4.8 and exhibited no significant trend over time. Nesting success and productivity also showed no trends over time. Mean number of young fledged by kestrel pairs each year ranged from 1.3 to 3.5 and averaged 2.6. Brood sizes at successful boxes averaged 4.0. Nests in boxes on trees near farmsteads and in boxes on interstate highway signs were less successful than nests in boxes on poles in less-inhabited areas, suggesting that signs on interstate highways may not be optimal nesting habitat for kestrels. Kestrel nest-initiation and hatching dates became earlier during the study period, and mean winter temperatures increased significantly during the study period, suggesting a possible effect of climate change.

We conducted a nest-box study from 1989–93 in north-central Florida, in an effort to increase the population of the threatened Southeastern American Kestrel (Falco sparverius paulus) and to evaluate the effect of providing artificial nest sites on the local population. A total of 388 nest boxes were erected in the experimental area. Over 75% of these nest boxes were placed >250 m away from roadside transects we established. This enabled us to measure changes in population density without the potential bias associated with the concentration of kestrels at specific nest-box sites. Annual censuses were conducted along 20 16-km transects to quantify population density in the control (no nest boxes) and treatment zones. Breeding attempts in nest boxes increased during each year of the study. Kestrel densities along the experimental transects increased from 5.0 birds/100 km² to levels equivalent to the control zone within 3 yr of nest box placement (32.3 vs. 34.9 birds/100 km²). Kestrels preferred nest boxes in sandhill habitats over those in hammock habitats, and nesting attempts in sandhills had greater hatching rates, more fledglings produced, and greater overall nesting success. Compared with sites with nest boxes that were not used by kestrels, sites with selected nest boxes were characterized by having more longleaf pine (Pinus palustris), cabbage palm (Sabal palmetto), and ground cover <10 cm in height, and less broadleaf deciduous trees, oaks (Quercus spp.), roadside berms, hayfields, water, and gravel.

American Kestrels (Falco sparverius) nesting near Hawk Mountain Sanctuary in eastern Pennsylvania exhibited widespread West Nile virus (WNV) seroprevalence in 2004. Here we examine the dynamics of WNV in kestrels nesting near Hawk Mountain Sanctuary in 2004–06. We tested kestrel saliva for viral RNA using oral swabs analyzed by RT-PCR, and all swabs were negative for WNV. We used plaque-reduction neutralization testing (PRNT) to measure WNV antibodies in serum. In 2004, we tested 22 adult kestrels and most (95%) were positive for WNV. In 2005, we tested blood samples from six adults and seven nestlings, and WNV seroprevalence was 33% in adults and 14% in nestlings. In 2006, we sampled seven adults and eight nestlings, and 29% of the adults and no nestlings were positive. WNV was widespread in birds, mosquitoes, and humans in Pennsylvania in 2003, and subsequently has declined in these populations. The prevalence of WNV antibodies in the kestrels was compared to surveillance data for mosquitoes, dead birds, and humans in Pennsylvania. WNV antibody prevalence in kestrels was significantly correlated with prevalence of WNV infection in dead birds from the previous year (correlation coefficient 1.000; P  =  0.009), but not with prevalence in mosquitoes or humans from the previous year or in dead birds, humans, and mosquitoes from the same year. This suggested that the antibody levels observed in the kestrel population we studied resulted from infections that occurred during the previous year. Our data indicated that WNV prevalence has declined in this population of kestrels over time.

Previous studies have examined vegetative cover and land use immediately surrounding American Kestrel (Falco sparverius) nest sites. However, the relationship of landscape-level habitat structure to nest-site selection has received little attention. Between 1995 and 2007, we erected nest boxes in pastures and meadows in northwestern New Jersey. The number of breeding pairs in boxes ranged from 2 in 1995 to 59 in 2002. We used a Geographic Information System to model habitats used by kestrels (open areas dominated by herbaceous vegetation) and delineated patches of contiguous suitable habitat within the study area and statewide. Nest boxes available in large (>1000 ha) patches of suitable habitat were occupied by kestrels at rates significantly higher than those in medium (250–1000 ha) or small (<250 ha) patches. Large patches exhibited relatively high occupancy rates every year, medium patches were most likely to be occupied during years when kestrels were abundant in the study area, and small patches were used infrequently, regardless of how many breeding pairs were present in the study area. That kestrels disproportionately select nest sites in contiguous patches sufficiently large to include many breeding territories has important implications for conservation strategies. We currently are using information on patch size to implement a statewide nest-box program for kestrels in New Jersey.

During 1997 and 1998, we studied American Kestrels (Falco sparverius) using nest boxes at Kesterson Reservoir (Merced County, central California), a site with elevated soil selenium concentrations. We radiotagged male kestrels to determine their foraging ranges. Nest-box use was lower than expected; eleven nest boxes at Kesterson were used during the two years (16% and 20% occupancy) and five offsite nest boxes were used (60 and 40% occupancy). Kestrel pairs in 60% and 67% of onsite nest boxes that were used successfully fledged chicks in 1997 and 1998, respectively; pairs in offsite nest boxes had 100% (1997) and 50% (1998) success. Concentrations of Se in blood and eggs of kestrels at Kesterson were similar to concentrations in offsite kestrels. We analyzed 31 pellets and we used previously measured selenium concentrations in prey items from Kesterson to estimate dietary Se exposure of 8.1 µg Se/g for kestrels at Kesterson. We estimated that radiotagged male kestrels foraged onsite 58% of the time in 1997 and 53% of the time in 1998 and that they foraged from 0 to 3.4 km (maximum foraging distances observed: 0.5–3.4 km) from the nest box during foraging bouts.

Stable-isotope analyses of feathers of migrant birds can help identify the locations where these birds bred or molted (i.e., breeding or molt origins). We examined stable hydrogen isotope (δD) values in feathers of American Kestrels (Falco sparverius) wintering at sites east (South Carolina) and west (Tennessee) of the Appalachian Mountains to assess their breeding origins. We used the dataset provided by Lott, and Smith (2006, Auk 123:822–835) to create a feather isotope basemap that considered error propagation associated with uncertainty in feather δD values and used the predicted mean growing-season precipitation δD of Bowen et al. (2005, Oecologia 143:337–348). We also measured feather δ¹⁸O on a subset of wintering kestrels and found that the general linear relationship found between feather δD and δ¹⁸O measurements deteriorates for feather δD more positive than −20‰. We suggest that δ¹⁸O measurements can be used to screen raptor feather isotope datasets to identify possible outliers that may unduly influence the assignment of birds to origin using feather δD basemaps. Our analyses indicated that the wintering population consisted of 13.1% ± 5.5% (SD) residents for the Tennessee site and 9.3% ± 3.5% (SD) residents for South Carolina. Migrants originated from the north-central portion of the U.S. breeding range (east of the Rocky Mountains) and southeastern Canada. This was consistent with leapfrog migration in this species.

We studied reproductive parameters, agonistic interactions, foraging behavior, prey deliveries, and diet of American Kestrels (Falco sparverius) using nest boxes in semiarid forests in central Argentina. The breeding season (from egg-laying to fledging) began in late October and ended in late December. Average clutch size was 4.3 ± 0.3 (SE) eggs (N  =  6 nests), hatching success was 73%, and nest success was 66%. Incubation was primarily by females, whereas males mostly hunted and defended the nest area. During early and late nestling stages, both sexes showed aggressive behavior against intruders. The primary hunting method we observed kestrels using near the nest box was perch-hunting (99%), with a peak in hunting activity during midday for females and during morning for males. Both sexes visited the nest most frequently during the late nestling stage (males: 1.04 ± 0.47 prey/hr; females: 1.22 ± 0.35 prey/hr), likely because of high energy demands of the nestlings for growth during that time. The majority of the nestling diet was arthropods (71.4% of diet as determined by observations; 71% by pellets; 89% by prey remains). Reproductive characteristics of this species in the semiarid forest of central Argentina, including reproductive rate, role partitioning between sexes, foraging behavior, and diet, are similar to those previously recorded for some populations in North America.

The American Kestrel (Falco sparverius) is a common North American falcon. As a result of a number of characteristics, including small size, annual reproduction, ability to be housed in large numbers, and ease of breeding in captivity, these falcons have been used as wildlife models for over four decades. Kestrels are best known for their use in the demonstration that 1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane (DDT), and 1,1-dichloro-2,2-bis(p-chlorophenyl)ethylene (DDE) caused eggshell thinning and reproductive failure in birds, and their use in the amelioration of captive propagation programs for the Peregrine Falcon (F. peregrinus). Since that time, kestrels have continued to play a prominent role in wildlife research, particularly in the field of toxicology, where they remain the only raptor species to be bred in sufficient numbers in captivity for statistically significant research. Kestrels have also been used for studies of avian biology, physiology, nutrition, and behavior. In recent years, concerns have been raised about factors that might affect the future use of captive kestrels as wildlife models, namely the declines in wild kestrel populations and the possible deleterious effects of long-term captive breeding, which could have implications for the comparability of wild and captive populations.