2009 SOUTHEASTERN NATURALIST 8(1):93–106 Activity of Post-fl edging Peregrine Falcons in Different Rearing and Habitat Conditions Matthew R. Dzialak1,*, Kristina M. Carter2, Michael J. Lacki3, David F. Westneat4, and Katie Anderson5 Abstract - To assist with recovery of Falco peregrinus (Peregrine Falcon) in the southeastern United States, we compared pre-dispersal activity budgets between falcons reintroduced at sites chosen for their contrasting habitats (agriculture vs forest). We also compared behavior of our hacked birds with nearby wild-produced juveniles. We classified pre-dispersal behavior into nine activities depicting fl ight and non-fl ight. We logged 901 hr of observation and found that wild-produced falcons spent more time in low fl ight, soaring, and mock combat during a 4-wk postfl edging period (mean ± 95% CI) than hacked birds. Peregrine Falcons hacked in mixed agricultural habitat spent more time soaring and perching alertly than those hacked in forest habitat; falcons in forest habitat perched inactively with higher frequency. Dispersal time (mean ± SD) differed among groups (F2,31 = 11.4, P < 0.001). Falcons hacked in forest habitat spent 15.2 ± 12.2 days on the post-fl edging areas before dispersing, whereas those hacked in agricultural habitat spent 31.0 ± 3.3 days and wild-produced birds spent 35.9 ± 10.1 days. It appeared that transitional habitat supporting available prey and the presence of adults during the post-fl edging period were important in the expression of key behavior repertoires including hunting, defense, and social fl ight activity. Our results suggest that further recovery of the Peregrine Falcon in the southeastern United States would be poorly served by additional hacking, particularly in forest habitat. Rather, managers should continue to monitor and encourage productivity in existing occupied habitat; eventually offspring from occupied habitat may occupy adjacent habitats. Introduction The transition from fledging to independence is an important lifehistory stage for many bird species. In raptors, the post-fledging period provides stimuli that contribute to development of behavior such as social interaction among conspecifics, prey recognition, pursuit, and defense (Newton 1979). Understanding the factors that affect post-fledging behavior in raptors can inform their management, particularly for species of conservation concern that might be subject to manipulations in rearing conditions (e.g., hacking [housing juveniles in enclosures at the release site from several days to up to four weeks before release] or cross-fostering) during recovery efforts. Attention to the events that lead to formation 1Hayden-Wing Associates LLC, 2308 South 8th Street, Laramie, WY 82070. 2Kristina Carter Photography, Aurora, IL 60504. 3Department of Forestry, University of Kentucky, Lexington, KY 40546. 4Department of Biological Sciences, University of Kentucky, Lexington, KY 40546. 5Department of Biology, Northern Michigan University, Marquette, MI 49855. *Corresponding author - matt@haydenwing.com. 94 Southeastern Naturalist Vol. 8, No. 1 of fully functional adults could make the difference between success and failure in raptor management for recovery. Falco peregrinus Turistall (Peregrine Falcon) has been the focus of several conservation-oriented studies that examined behavior. A comprehensive description of the transition to independence in hacked and wild-produced Peregrines Falcons has been completed (Sherrod 1983). Many facets of behavior developed similarly in these two groups despite the absence of adult infl uence on hacked birds, which suggests that instinct and sibling interaction among hacked falcons were key factors in this similarity. This finding had important conservation implications for development and refinement of recovery efforts because it characterized the resilience of Peregrine Falcons to alternate rearing strategies. Behavior in the Peregrine Falcon has since become well known, particularly as it relates to managing eyries (Carlier and Gallo 1995, Hovis et al. 1985, Palmer et al. 2003). Knowledge gaps remain, however, including information on factors affecting behavior in areas where this falcon is still considered to be at risk of extirpation. The Peregrine Falcon has recovered throughout much of North America. In the Arctic, it was removed from the endangered species list in 1994, and the population in the lower 48 United States was removed from the endangered species list in 1999. Recovery has progressed more slowly in the southeastern United States (Southeast: US Fish and Wildlife Service [FWS], Region 4). Up to 70 eyries may have occurred historically in the Southeast, but at present, the region supports approximately 18 breeding pairs—the fewest among regions designated by FWS (Dzialak et al. 2005a). Given the extent of apparently suitable habitat, several states in this region still list this falcon as endangered. We hacked peregrines in Kentucky as part of an effort to augment Peregrine Falcon recovery in the Southeast. Typically, hacked raptors remain in the vicinity of the hacking station for several weeks after release and then disperse after a post-fl edging period. Our objectives were to compare pre-dispersal activity time budgets among Peregrine Falcons hacked at sites specifically chosen for their contrasting habitat configurations and to compare time budgets and time to dispersal of hacked groups with nearby wild-produced juveniles to develop insight into relative infl uences of rearing and habitat on Peregrine Falcons in the Southeast. Methods We hacked 33 Peregrine Falcons at Red River Gorge Geologic Area in Daniel Boone National Forest (DBNF) and at Tom Dorman State Nature Preserve (TDSNP), KY. Cliffs at these sites are 60–80 m in height. These sites were chosen, in part, for their contrasting habitat features. DBNF, located in the Cumberland Plateau physiographic region, provided non-forest corridors within a forest matrix. TDSNP, about 75 km to the west in the Bluegrass region, provided forest corridors within a largely non-forest matrix (Fig. 1; Dzialak et al. 2005b). We summarized several key differences in habitat 2009 M.R. Dzialak, K.M. Carter, M.J. Lacki, D.F. Westneat, and K. Anderson 95 between these regions (Dzialak et al. 2005a; Table 1). We released 28 falcons at DBNF during 2001–2003 at two hacking stations 1.5 km apart, and Figure 1. (A) The Red River Gorge Geologic Area, Daniel Boone National Forest, KY. Numerous massive sandstone outcrops within a largely forested matrix characterize the region; photo by J.L. Larkin (B). Aerial views of largely nonforest corridorlike landscape features increasing in spatial extent from Cumberland Plateau into Bluegrass physiographic regions near Daniel Boone National Forest; photo by M.R. Dzialak (C) and (D). Aerial views of the largely forested Kentucky River drainage near Tom Dorman State Nature Preserve within a mixed agricultural landscape; photos by M.R. Dzialak. Figure adapted from Dzialak et al. (2005b). Table 1. Characteristics of cliffs in the Bluegrass and Cumberland Plateau physiographic regions, KY. Nine Cumberland Plateau cliffs were located in Red River Gorge, with two used as hack sites. One Bluegrass cliff was in Dorman Preserve. Table modified from Dzialak et al. (2005a). Data presented as mean ± SD. Variable (units) Bluegrass (n = 3) Cumberland Plateau (n = 26) Height (m) 50 ± 20 34 ± 21 Estimated horizontal extent (m) 1269 ± 1300 204 ± 131 Elevation (m) 208 ± 34 323 ± 50 Nearest agriculture >2 ha (m) 244 ± 77 2089 ± 1455 Nearest developmentA (m) 825 ± 101 13,190 ± 4920 Nearest open water (m) 19 ± 5 1192 ± 2795 Land use: forest (%)B 50 ± 1 92 ± 4 Land use: agriculture (%)C 47 ± 1 6 ± 2 Land use: development (%)D 1 ± 0 0 ± 0 Land use: total non-forested (%)E 50 ± 1 8 ± 4 ARestricted to incorporated city, town, borough, or locality. BPercentage of landscape (20,096 ha) primarily forest within 8-km radius of the cliff. CPercentage of landscape (20,096 ha) primarily agriculture within 8-km radius of the cliff. DPercentage of landscape (20,096 ha) primarily developed/residential within 8-km radius of the cliff. EPercentage of landscape (20,096 ha) non-forest within 8-km radius of the cliff. 96 Southeastern Naturalist Vol. 8, No. 1 five at TDSNP in 2003 following methods of Sherrod et al. (1982). Data on 22 of 28 falcons hacked in DBNF were available for this study in addition to data on the five falcons hacked from TDSNP. We fitted each bird with a tarsal-mounted RI-2CM transmitter (Holohil Systems, Ltd., ON, Canada) and gave it a unique identification marking on the dorsal humeral region using nontoxic paint. Prerelease veterinary evaluation and health certification of all falcons was coordinated through the University of Minnesota College of Veterinary Medicine Raptor Center. All methods used were approved by the University of Kentucky Animal Care and Use Committee (protocol # 00504A2002). We observed eight wild-produced Peregrine Falcons in 2002–2003 at two breeding locations along a 60-km segment of the Ohio River in northern Kentucky where adults used nest boxes affixed to smokestacks of coal-fired power plants. Four breeding pairs occupied the smokestacks along a 175-km segment of the Ohio River in the Bluegrass region. We used a geographic information system and Mid-American Remote Sensing Center data (Murray State University, Murray, KY) to characterize four habitat classes within 8 km of Ohio River eyries and hacking stations; we arcsine-transformed the data and used the Fisher exact test (Zar 1996) for analysis. We observed Peregrines Falcons from 18 June–14 August at a distance of 200–400 m using 10 x 30 binoculars and a 20 x 60 mm spotting scope (Fujinon®, Wayne, NJ). Observation began immediately following fl edging and ended with dispersal. We visited a site for two consecutive days before switching to a different one. The experimental unit was a fl edgling; we observed each fl edgling for a 1-hr block. When a fl edgling moved beyond survey range for ≥15 min, we selected a different fl edgling for observation and began a new 1-hr block. Observation sessions were 10-hr blocks, with start times staggered to encompass daylight hours (0630–2100). We classified behavior into nine types (Table 2; Cade 1960, Harris and Clement 1975, Herbert and Herbert 1965, Sherrod 1983). We converted data into activity budgets calculated as percent frequency of time spent in a behavior by a fl edgling divided by total time in view per day. We averaged these data among fl edglings over 1-wk intervals and evaluated a 4-wk post-fl edging period (Sherrod 1983, Varland et al. 1991). The dataset was unbalanced across groups (hereafter, groups refers to wild-produced, DBNF, and TDSNP), rendering a two-factor (group and week) repeated measures ANOVA unreliable. We present the data graphically and discuss results in terms of central tendency and associated precision (i.e., mean ± 95% CI). Non-overlapping 95% CI is a conservative estimate of significance (i.e., likely has a lower type I error rate and higher type II error rate at α = 0.05; Schenker and Gentleman 2001). There are techniques to correct for bias in error rates (type I error-averaged confidence intervals; Goldstein and Healy 1995) but, given the small sample size and probable deviations from underlying assumptions of equal variance and normality, we opted to present the more conservative results. We also compared the proportion of Peregrine Falcons achieving 2009 M.R. Dzialak, K.M. Carter, M.J. Lacki, D.F. Westneat, and K. Anderson 97 independence among sites using the Fisher exact test (Zar 1996). Data on the proportion of wild-produced young that achieved independence during 2001–2003 were from Vorisek and Carter (2004). Beginning at fl edging, we tabulated the number of days elapsed before each peregrine dispersed and used a single-factor ANOVA to compare dispersal time among groups. Differences for the Fisher exact test and ANOVA were considered significant at P ≤ 0.05. All analyses were performed using SAS® software (SAS Institute, Cary, NC). Results We logged 901 hr of observation. It appeared that the activity types we examined adequately characterized behavior among sites (Table 2); we observed no behavior unique to a particular site. Habitat composition at TDSNP and two Ohio River eyries was similar in terms of percent forest (0.48, 0.44, and 0.47, respectively), agriculture (0.48, 0.48, and 0.46, respectively), open water (0.02, 0.06, and 0.06, respectively) and development (0.02, 0.02, and 0.01, respectively). The proportion of forest, agriculture, open water, and development at DBNF was 0.94, 0.05, <0.01, and 0.01, respectively. Habitat attributes were different among sites (Fisher exact test; df = 3, P < 0.001), refl ecting the contrast between the heavily forested DBNF and the mixed forest-agricultural TDSNP and Ohio River eyries. Peregrine Falcons achieved independence similarly among sites: 5 of 5 at TDSNP, 17 of 22 at DBNF and 10 of 26 at the Ohio River sites (Fisher exact test; df = 2, P = 0.30). Dispersal time differed among groups, with DBNF falcons leaving the post-fl edging area at 15.2 ± 12.2 days, TDSNP at 31.0 ± 3.3 days, and wild-produced falcons at 35.9 ± 10.1 days (F2,31 = 11.4, P < 0.001). Time-activity budgets differed among groups and with week-sincefl edging based on mean ± 95% CI criteria (Figs. 2, 3, and 4). Table 2. Nine predispersal activities identified for hacked and wild-produced Falco peregrinus (Peregrine Falcons) in Kentucky, June to August, 2001–2003. Behavior Description Non-fl ight activity Alert perching Active visual surveying, head-bobbing, or movement while perched. Inactive perching Sunning, resting, lying down, or other stationary perching. Movement Flapping, walking, hopping, stationary play with inanimate objects, preening, feaking, or scratching. Flight activity Simple fl ight Perch to perch fl ights, short fl ight with high wing-stroke frequency. Low fl ight Low altitude fl ight restricted to areas adjacent to hacking stations or eyries, moderate wing-stroke frequency. Soaring Prolonged high-altitude fl ight, infrequent or no wing-stroke. Pursuit Activity Mock combat Social aerial activity including stooping, chasing, vocal protest, or display. Hawking Effort to capture fl ying invertebrates. Direct pursuit Effort to capture vertebrate prey. 98 Southeastern Naturalist Vol. 8, No. 1 Figure 2. Predispersal activity budgets for non-fl ight behavior including inactive perching (A), alert perching (B), and movement (C) among hacked (Daniel Boone National Forest and Tom Dorman State Nature Preserve) and wild-produced Falco peregrinus (Peregrine Falcon) in Kentucky, June to August, 2001–2003. Sample sizes shown in Figure 2A apply to 2B and 2C. 2009 M.R. Dzialak, K.M. Carter, M.J. Lacki, D.F. Westneat, and K. Anderson 99 Non-fl ight activity Week effects within groups. Wild-produced Peregrine Falcons spent considerable time perching inactively during weeks one and two (27–38%), but reduced inactive perching in weeks three and four (5–6%) (Fig. 2). Falcons at TDSNP also perched inactively with high frequency in week one (31%), but then reduced their inactive perching through time (9–11%). In contrast, DBNF falcons perched inactively throughout the study (18–36%). All groups generally maintained a consistent frequency of alert perching; the only notable effect was more frequent alert perching at DBNF in week one (52%) compared to week three (27%). Movement activity was most frequent in all groups in week one (6–7%) compared to weeks two to four (0.4–4%). Figure 3. Predispersal activity budgets for fl ight behavior including simple fl ight (A), low fl ight (B), and soaring (C) among hacked (Daniel Boone National Forest and Tom Dorman State Nature Preserve) and wild-produced Falco peregrinus (Peregrine Falcon) in Kentucky, June to August, 2001–2003. Sample sizes shown in Figure 3A apply to 3B and 3C. 100 Southeastern Naturalist Vol. 8, No. 1 Comparison of wild-produced and hacked falcons. Hacked Peregrine Falcons at DBNF perched inactively more frequently than wild-produced birds (≈30% vs ≈5%) during weeks three to four (Fig. 2). Similarly, hacked falcons generally perched alertly more frequently than wild-produced falcons; differences existed between TDSNP (67–79%) and wild-produced (33–51%) Peregrine Falcons in weeks two and four. Movement activity generally was more frequent in hacked falcons; differences existed between wild-produced and DBNF birds in week four (1% vs 3%). Comparison of hacked falcons by habitat. DBNF falcons perched inactively more frequently than ones at TDSNP during all weeks, but differences existed only in week three (31% vs 9% inactive perching). Peregrine Falcons at TDSNP perched alertly more frequently than at DBNF, but differences Figure 4. Predispersal activity budgets for pursuit behavior including mock combat (A), direct pursuit (B), and hawking (C) among hacked (Daniel Boone National Forest and Tom Dorman State Nature Preserve) and wild-produced Falco peregrinus (Peregrine Falcon) in Kentucky, June to August, 2001–2003. Sample sizes shown in Figure 4A apply to 4B and 4C. 2009 M.R. Dzialak, K.M. Carter, M.J. Lacki, D.F. Westneat, and K. Anderson 101 were only significant in weeks two to four (≈70% vs ≈30%). Movement activity was similar between hacked groups (Fig. 2). Flight activity Week effects within groups. Simple fl ight was consistently infrequent throughout the study (Fig. 3). Low fl ight peaked in frequency in week three and was lowest in week one, but this difference was significant only in wildproduced falcons (0.5% vs 13%). Frequency of soaring exhibited the same trend as low fl ight by peaking in all groups in week three; the most apparent week effect was less frequent soaring in all groups in week one (≈0.5%) compared to week three (≈6%). Comparison of wild-produced and hacked falcons. We found no difference between wild-produced and hacked Peregrine Falcons in simple fl ight (Fig 3). Wild-produced falcons exhibited low fl ight more frequently than hacked Peregrines during weeks two and four (≈8% vs. ≈1%). Wildproduced birds generally spent more time soaring than hacked ones, but differences were apparent only between wild-produced and DBNF falcons in week two (14% vs 2%) and between wild-produced and both hacked groups in week four (7% vs ≈1%). Comparison of hacked falcons by habitat. We found no difference between hacked groups in frequency of simple fl ight or low fl ight (Fig. 3). TDSNP birds soared more frequently than those at DBNF, but differences were significant only in weeks three (8% vs 2%) and four (2% vs 0.1%). Pursuit activity Week effects within groups. No week effect existed for direct pursuit, mock combat or hawking activities; however, wild-produced falcons engaged in mock combat, more frequently in weeks two and four (≈9%) compared to week one (≈2%) (Fig. 4). Comparison of wild-produced and hacked falcons. Direct pursuit was observed infrequently in hacked birds, but was seen with comparatively high frequency in wild-produced ones in weeks two to four (Fig. 4). Hawking was rarely observed during the study. Mock combat was more frequent in wildproduced falcons in week four (9% vs. ≈0.4%). Comparison of hacked falcons by habitat. We found no differences between hacked groups in direct pursuit, hawking, or mock combat activities (Fig. 4). Discussion Changes in the frequency of activities that we monitored were consistent with what is known about the post-fledging period in raptors, especially for young Peregrine Falcons during development from fledging to dispersal (i.e., Balbontin and Ferrer 2005, Newton 1979, Sherrod 1983). For example, Peregrine Falcons generally exhibited perching activities, movement, and simple flight with relatively high frequency soon after fledging. Concurrent with a decrease in these activities with age 102 Southeastern Naturalist Vol. 8, No. 1 (week effect) was an increase in frequency of more complex flight behavior such as mock combat and soaring. Wild-produced falcons spent more time in mock combat, pursuit, soaring, and low fl ight activities than hacked birds, although these differences were not always apparent because of low sample sizes. Wild falcons also remained on the post-fl edging area longer than hacked birds. More conspicuous social behavior and a longer post-fl edging period in wild-produced versus hacked Peregrine Falcons has been reported previously (Sherrod 1983). The activities we identified as more frequent among wild-produced falcons often occurred as part of adult-juvenile interaction, including aerial food transfers, family hunting events, mimicking adult defensive behavior, and adult-juvenile antagonism (Sherrod 1983). We have corroborated many previous findings, particularly those on the infl uence of adults in conspecific interaction and defense (Sherrod 1983). Research on hacked fl edgling Haliaeetus leucocephalus Linnaeus (Bald Eagle) that had been captive-reared or removed from active nests at about seven weeks of age (wild-reared) indicated that wild-reared eagles generally dispersed sooner than captive-reared ones and also sooner than eagles fl edged at wild nests (Meyers and Miller 1992). However, none of the Bald Eagles was reared in a setting with adults as was the case with wild Peregrine Falcons in this study. We found differences in pre-dispersal activity between Peregrine Falcons hacked in two different habitats. Falcons at TDSNP (mixed forest-agriculture) spent more time in alert perching and soaring behavior and less time perching inactively than falcons at the forested DBNF. Timing of dispersal also differed considerably between hacked groups, with those at TDSNP remaining on the post-fl edging area for an average of 31 days compared to only 15 days for birds at DBNF. It is important to note the similarity between TDSNP and wild-produced falcons, which while reared differently, shared a similar habitat. Did habitat differences cause behavioral differences between falcons at DBNF and TDSNP? Inferring cause and effect would require replicated landscapes and perhaps manipulation of habitat, which we could not investigate. Also, several factors associated with variation in hacking protocol can confound these observations including health condition, sex, release age, and release cohort size. We controlled for such factors as much as possible. All Peregrines received pre-release veterinary evaluation and appeared healthy at release. Some limited variation occurred in other factors, but given the conservation objective of restoration, it was in our interest to follow hacking protocol (Dzialak et al. 2006). For example, all falcons were 43–54 days old at release, with males generally released slightly younger than females. Consistency in rearing conditions (hacking protocol) was generally maintained between sites, so we suggest that it is worthwhile to consider how habitat might play a role in the behavioral differences between hacked groups (sensu Bélisle et al. 2001, Evans et al. 2003, Lima and Zollner 1996). The value of exploring this is in developing further hypotheses on the infl uence of habitat 2009 M.R. Dzialak, K.M. Carter, M.J. Lacki, D.F. Westneat, and K. Anderson 103 on raptor development and in offering insight into factors affecting the success of Peregrine Falcon hacking and recovery in the Southeast. Habitat affects the falcon’s activity by infl uencing prey availability, the presence of competitors, the risk of predation, mating opportunities, and microclimate. From previous work, we knew that preferred prey were significantly more abundant at TDSNP compared to DBNF (Carter et al. 2003, Dzialak et al. 2005a). More abundant prey at TDSNP appeared to result in a higher frequency of alert perching, and also appeared to trigger pursuit. Once in fl ight, pursuit often gave way to soaring. In contrast, falcons at DBNF had to move several km from the hack site to encounter edge habitat and associated prey species (Dzialak et al. 2005b). Most falcon prey in the immediate vicinity of DBNF hack sites were inconspicuous forest birds, so visual stimulus associated with prey abundance was limited; therefore, falcons at DBNF tended to rest (perched inactive). The importance of prey resources has been observed previously, particularly in terms of breeding success or movement behavior in raptors (Barclay and Cade 1983, Hickey 1942, Holroyd and Banasch 1990, Kopimärki and Norrdahl 1991). It is important to have abundant and available prey in close proximity to hacking stations. Yet, the importance of prey abundance and availability in the development of behavior critical to hunting, defense, and perhaps conspecific interaction has received little attention because the effects of prey resources at hack sites have focused on physical requirements rather than behavioral development (Holroyd and Banasch 1990, Sherrod et al. 1982). An alternative hypothesis is that differences in activity between hacked groups refl ected differences between sites in predation (by owls) or competition. Previous work suggested that risk of predation by raptors (i.e., Bubo virginianus Gremlin [Great Horned Owl]) was greater at TDSNP (Dzialak et al. 2005a). However, no mortality occurred at TDSNP; in contrast, five deaths occurred at DBNF, all from unknown causes (Dzialak et al. 2007). Competition for food may also have been greater at DBNF. Species observed at DBNF but not at TDSNP that scavenged on dead quail (Coturnix sp.) we provided as food for falcons, included Urocyon cinereoargenteus Schreber (Gray Fox), Crotalis horridus Linnaeus (Timber Rattlesnake), and Lynx rufus Schreber (Bobcat). Reduced alertness and activity (observed at DBNF), however, is not supported in the literature as an adaptive response to predation or competition in raptors. A post-fl edging period of 23–30 days has been observed for hacked Peregrine Falcons (Fyfe 1988, Powell et al. 2002). Hypotheses on resource competition, physical condition, and density-dependent social interaction have been offered in investigations of dispersal timing in raptors (Balbontin and Ferrer 2005, Belthoff and Dufty 1998, Willey and van Riper 2000). We found no evidence that hacked groups in the two locales differed in physical condition or access to resources, and our data do not support the notion of density-dependent social confl ict (i.e., more falcons were released at DBNF, but there was less activity there). Also, week of dispersal was highly variable 104 Southeastern Naturalist Vol. 8, No. 1 at DBNF; some individuals dispersed in the first week of fl edging, while others remained on the post-fl edging area for 39 days and defended territories there in subsequent years. Peregrine Falcons in other regions occupy forest habitat (Corser et al. 1999), but an association between forest and a short post-fl edging period has not been reported, and so it seems questionable that simple preference for non-forest habitat explains the short and highly variable post-fl edging period at DBNF. Rather, individual experiences among these falcons in release conditions, successful food acquisition, or conspecific interaction post-release may better account for this variability (S. Sherrod, G.M. Sutton Avian Research Center, Bartlesville, OK, pers. comm.). What factors limit Peregrine Falcon recovery in the Southeast? An absence of resident reproductive adults following extirpation, limited foraging habitat near cliffs, or frequent recreational use of cliff habitat might contribute to an aversion to these areas among potential colonizers. To these considerations should be added the possibility that a diminished social repertoire in the absence of abundant prey can infl uence choices of habitat over the long term. To our knowledge, no additional Peregrine Falcon releases are scheduled in the Southeast in the foreseeable future. This shift away from falcon introductions in the Southeast is appropriate as further recovery there may be poorly served by additional hacking in forest habitat. Perhaps the best option biologically and financially would be to monitor and encourage productivity at existing strongholds of Peregrine Falcon activity such as western North Carolina or coastal regions of Virginia and Maryland. Some of these offspring may eventually adapt to contemporary selective pressures in the Southeast and begin to occupy inland portions of the Appalachian Plateaus (Mengel 1939) and Mississippi River bottomland forests (Bellrose 1938, Spofford 1943). Acknowledgments Funding and other support for this project was provided by the Kentucky Department of Fish and Wildlife Resources; the University of Kentucky, College of Agriculture; and the US Department of Agriculture, Daniel Boone National Forest. We thank S. Sherrod for his comments and suggestions, which improved the manuscript. This research (KAES 03-09-065) is associated with a project of the Kentucky Agricultural Experiment Station and is published with the approval of the Director. Literature Cited Balbontín, J., and M. Ferrer. 2005. Factors affecting the length of the post-fl edging period in the Bonelli’s Eagle Hieraaetus fasciatus. Ardea 93:189–198. Barclay, J.H., and T.J. Cade. 1983. Restoration of the Peregrine Falcon in the eastern United States. Bird Conservation 1:3–39. Bélisle, M., A. Desrochers, and M-J. Fortin. 2001. Infl uence of forest cover on the movements of forest birds: A homing experiment. Ecology 82:1893–1904. Bellrose, F. Jr. 1938. Duck Hawks nesting in western Tennessee. Wilson Bulletin 50:139. 2009 M.R. Dzialak, K.M. Carter, M.J. Lacki, D.F. Westneat, and K. Anderson 105 Belthoff, J.R,. and A.M. Dufty, Jr. 1998. Corticosterone, body condition, and locomotor activity: A model for dispersal in Screech-owls. Animal Behavior 55:405–415. Cade, T.J. 1960. Ecology of the Peregrine and Gyrfalcon populations in Alaska. University of California Publications in Zoology 63:151–290. Carlier, P., and A. Gallo. 1995. What motivates the food-bringing behaviour of the Peregrine Falcon throughout breeding? Behavioral Processes 33:247–256. Carter, K.M., M.J. Lacki, M.R. Dzialak, L.S. Burford, and R.O. Bethany. 2003. Food habits of Peregrine Falcons in Kentucky. Journal of Raptor Research 37:351–356. Corser, J.D., M. Amaral, C.J. Martin, and C.C. Rimmer. 1999. Recovery of a cliffnesting Peregrine Falcon, Falco peregrinus, population in northern New York and New England, 1984–1996. Canadian Field-Naturalist 113:472–480. Dzialak, M.R., M.J. Lacki, and K.M. Carter. 2005a. Characterization of potential release sites for Peregrine Falcon reintroduction. Natural Areas Journal 25:188– 196. Dzialak, M.R., M.J. Lacki, K.M. Carter, K. Huie, and J.J. Cox. 2006. An assessment of raptor hacking protocol during a reintroduction. Wildlife Society Bulletin 34:542–547. Dzialak, M.R., M.J. Lacki, and S. Vorisek. 2007. Survival, mortality, and morbidity among Peregrine Falcons reintroduced in Kentucky. Journal of Raptor Research 41:61–65. Dzialak, M.R., M.J. Lacki, J.L. Larkin, K.M. Carter, and S. Vorisek. 2005b. Corridors affect dispersal initiation in reintroduced Peregrine Falcons. Animal Conservation 8:421–430. Evans, K.L., R.B. Bradbury, and J.D. Wilson. 2003. Selection of hedgerows by swallows, Hirundo rustica, foraging on farmland: The infl uence of local habitat and weather. Bird Study 50:8–14. Fyfe, R.W. 1988. The Canadian Peregrine Falcon recovery program, 1967–1985. Pp. 599–610, In T.J. Cade, J.H. Enderson, C.G. Thelander, and C.M. White (Eds.). Peregrine Falcon Populations: Their Management and Recovery. Peregrine Fund Inc., Boise, ID. 949 pp. Goldstein, H., and M.J.R. Healy. 1995. The graphical presentation of a collection of means. Journal of the Royal Statistical Society A. 158:175–177. Harris, J.T., and D.M. Clement. 1975. Greenland Peregrines at their eyries: A behavioral study of the Peregrine Falcon. Meddelelser Om Grønland 205:1–28. Herbert, R.A., and K.G.S. Herbert. 1965. Behavior of Peregrine Falcons in the New York City region. Auk 82:62–94. Hickey, J.J. 1942. Eastern population of the Duck Hawk. Auk 59:176–204. Holroyd, G.L., and U. Banasch. 1990. The reintroduction of the Peregrine Falcon, Falco peregrinus anatum, into southern Canada. Canadian Field-Naturalist 104:203–208. Hovis, J., T.D. Snowman, V.L. Cox, R. Ray, and K.L. Bildstein. 1985. Nesting behavior of Peregrine Falcons in west Greenland during the nestling period. Journal of Raptor Research 19:15–19. Kopimärki, E., and K. Norrdahl. 1991. Numerical and functional responses of Kestrels, Short-eared Owls, and Long-eared Owls to vole densities. Ecology 72:814–826. Lima, S.L., and P.A. Zollner. 1996. Towards a behavioral ecology of ecological landscapes. Trends in Ecology and Evolution 11:131–135. 106 Southeastern Naturalist Vol. 8, No. 1 Mengel, R.M. 1939. Summer records from Cumberland National Forest. Kentucky Warbler 15:45–47. Meyers, J.M., and D.L. Miller. 1992. Post-release activity of captive- and wildreared Bald Eagles. Journal of Wildlife Management 56:744–749. Newton, I. 1979. Population Ecology of Raptors. T. and A.D. Poyser, Berkhamsted, UK. 399 pp. Palmer, A.G., D.L. Nordmeyer, and D.D. Roby. 2003. Effects of jet aircraft overfl ights on parental care of Peregrine Falcons. Wildlife Society Bulletin 31:499–509. Powell, L.A., D.J. Calvert, I.M. Barry, and L. Washburn. 2002. Post-fl edging survival and dispersal of Peregrine Falcons during a restoration project. Journal of Raptor Research 36:176–182. Schenker, N., and J.F. Gentleman. 2001. On judging the significance of differences by examining the overlap between confidence intervals. American Statistician 55:182–186. Sherrod, S.K. 1983. Behavior of Fledgling Peregrines. Peregrine Fund, Ithaca, NY. 202 pp. Sherrod, S.K., W.R. Heinrich, W.A. Burnham, J.H. Barclay, and T.J. Cade. 1982. Hacking: A Method for Releasing Peregrine Falcons and Other Birds of Prey. Peregrine Fund, Ithaca, NY. 62 pp. Spofford, W.R. 1943. Peregrines in a west Tennessee swamp. Migrant 14:25–27. Varland, D.E., E.E. Klass, and T.M. Loughin. 1991. Development of foraging behavior in the American Kestrel. Journal of Raptor Research 25:9–17. Vorisek, S., and K.M. Carter. 2004. Interim report submitted to the US Department of Interior Fish and Wildlife Service: Peregrine Falcon monitoring. Wildlife Diversity Program, Kentucky Department of Fish and Wildlife Resources, Agreement No. 401814G017, Frankfort, KY. Willey, D.W., and C. van Riper III. 2000. First-year movements by juvenile Mexican Spotted Owls in the canyonlands of Utah. Journal of Raptor Research 34:1–7. Zar, J.H. 1996. Biostatistical Analysis. Prentice Hall, Upper Saddle River, NJ. 662 pp.