Population Estimates for Northern Juvenile Peregrine Falcons With Implications for Harvest Levels in North America

The peregrine falcon (Falco peregrinus) is one of the most intensively studied raptor species in the world (Olsen and Olsen 1989; Ratcliffe 1993; Rosenfield et al. 1995; Quinn and Kokorev 2000; Ganusevich et al. 2004; Anctil et al. 2013). In North America, wildlife managers focused interest on this species due to well-established population declines associated with contamination from organochlorine pesticides (Enderson et al. 1968; Peakall and Kiff 1979; Peakall et al. 1983; Risebrough and Peakall 1988; Ellis et al. 1989; Court et al. 1990; Franke et al. 2010). Researchers first summarized wide-scale population declines in North America at the 1965 Madison Peregrine Falcon Conference (Hickey 1969); these declines were found to be extensive in the United States (virtually total extirpation in the east and severe declines in the west) and throughout southern parts of Canada east of the Rocky Mountains. Of the three subspecies that exist in the North America, the American peregrine falcon (F. p. anatum) and Arctic peregrine (F. p. tundrius) were listed as “endangered” in the United States pursuant to the Endangered Species Act (ESA 1973, as amended) in 1970; the Peale's peregrine (F. p. pealei) was never listed in the United States (Figure 1). In Canada, the Committee on the Status of Endangered Wildlife in Canada listed the American peregrine as endangered, the Arctic peregrine as threatened, and the Peale's peregrine as a species of special concern pursuant to the Species at Risk Act (SARA 2002, as amended) in 1978.

Starting in 1970, researchers in Canada and the United States established 5-y surveys to monitor changes in the size of the breeding population and assess productivity where populations remained. On the basis of these surveys, the low point in the population is thought by researchers to have occurred around 1975 (Fyfe et al. 1976; Enderson et al. 1995). This was supported by migration data from Assateague Island (located off the coasts of Maryland and Virginia), which indicated that the population low occurred between 1970 and 1978 (Ward et al. 1988).

Wildlife biologists think that the northern portion (boreal, subarctic, and arctic) of the peregrine falcon population is large, and reportedly accounts for approximately 80% of the total population in North America (Kiff 1988). However, the spatial extent over which northern surveys can realistically occur constrains data collection at broad geographic spatial scales (Cade 2013). Nevertheless, researchers have attempted to estimate the size of the northern population; Cade (1960) proposed an Alaskan population of 1,000 pairs based primarily on surveys of major rivers, and Fyfe (1969), extrapolating from estimated densities in known survey areas of small spatial extent, estimated around 7,500 pairs in Canada north of 55°N.

Northern-breeding American and Arctic peregrines are highly migratory (Yates et al. 1988; Fuller et al. 1998; Schmutz et al. 1991), and although fall migration occurs over a broad geographic range (Fuller et al. 1998), Yates et al. (1988) indicated that “separate and distinct autumn migratory populations pass through the east and Gulf coasts” of the United States. By 1979, efforts to band nestlings existed throughout North America, including Greenland (Mattox 2003). In addition, researchers mounted an ongoing and substantial effort to trap and band migrating peregrine falcons at known concentration points on the east coast (Assateague Island off of the coasts of Maryland and Virginia, and Cape May, New Jersey) and in the Gulf of Mexico (North and South Padre islands, Texas) in the United States (Seegar et al. 2003). These efforts allowed for an estimation of the size of the migratory population using the Lincoln–Peterson (LP) mark–recapture method (Lincoln 1930). Using this method, Sheppard (1983) published a report for the U.S. Fish and Wildlife Service (USFWS) based on limited data from 1976 to 1981, and suggested that approximately 20,000 young peregrines were produced yearly. Using the same method, Cade et al. (1988) extended the time frame to include banding data from 1982 to 1985, and similarly concluded that 10,000–20,000 young might be produced annually. Other estimates focused on enumerating the number of breeding territories or breeding pairs; for example, Kiff (1988) estimated the size of the entire North American population during the pre-DDT era to be 7,000–10,000 territories with 80–90% occupied in any one year. Cade (2003) provided a slightly higher estimate for the pre-DDT era of 10,600–12,000 breeding pairs, of which 8,600–9,000 were thought to populate regions north of 55°N. In an analysis conducted to determine a harvest quota for falconry, the USFWS (2008) indicated that the upper population estimate for all of North America was 10,368 breeding pairs.

In the United States, the Arctic peregrine and American peregrine were removed from the list of endangered species in 1994 (USFWS 1998) and in 1999 (USFWS 1999), respectively. In Canada, the Arctic peregrine and American peregrine were evaluated together (rather than as separate subspecies) and designated by the Committee on the Status of Endangered Wildlife in Canada as species of special concern in 2007, and were legally listed as such on Schedule 1 of the Species at Risk Act in 2012. However, there remains great uncertainty among wildlife managers with regard to the size of the northern migratory population. Estimates range from a low of 3,005 pairs for all of North America (Green et al. 2006) to as high as 340,000 “individuals” in Canada and the United States (Rich et al. 2004). Estimates such as these are based on extrapolation of small-scale surveys of breeding pairs and expert opinion, and were, in part, used by the USFWS (2008) to set harvest limits of hatch-year (HY; the cohort of individuals known to have hatched during the calendar year in which they were banded) peregrine falcons in the United States. The USFWS completed an environmental assessment for take of passage (i.e., HY fall migrants) peregrine falcons in the United States in 2008 (Millsap 2008). For the purposes of managing take, and to allocate harvest levels for passage and nestling/postfledging peregrines, the environmental assessment considered three management populations: 1) “northern,” consisting of American and Arctic peregrines originating north of latitude 54°N; 2) “western,” consisting of all American peregrine falcons originating from natal sites west of longitude 100°N and south of latitude 54°N and all Peale's peregrines; and 3) “eastern,” consisting of all American peregrines originating from natal sites east of longitude 100°W and south of latitude 54°N.

United States regulations currently permit an annual take of up to 36 HY migrant peregrine falcons from September 20 through October 20 from anywhere in the United States east of longitude100°W. However, any reassessment of harvest of HY migrant peregrines in North America requires that wildlife managers have a better understanding of population size (USFWS 2008). To my knowledge there has been no attempt to apply the LP method to the problem of estimating the size of the northern population since Cade et al. (1988). The primary objectives of this study were to estimate the size of the HY population of northern migratory peregrine falcons in North America (including Greenland) using the LP method applied to banding records from 1970 through 2010, and provide an estimate of sustainable harvest for this population.

The LP estimator is a simple mark–recapture method for estimating population size that involves a single marking and a single recapture episode. Field biologists capture and mark (M) a sample of animals in a population. They then release the marked individuals, and allow the marked animals to mix throughout the population. Field biologists then capture a second sample of animals, C, containing R previously marked individuals. An estimate of the size of the population at the time of the original marking, N̂, can be calculated as

The LP estimator assumes that N (i.e., the number of individuals in the population) is constant and the population is closed (i.e., no births or deaths, immigration, or emigration), and that all individuals in the first sample have the same probability of being captured. In addition, the method assumes catchability remains unaffected by marking (i.e., no trap-happy or trap-shy individuals), and that individuals do not lose marks between the two sampling periods. The first assumption can be relaxed when there is natural morality between the two samples, and it applies to both the marked and unmarked cohort equally; in this event, the estimate then applies to the population size at the time that the first sample was released (Seber 1982).

The number of M and C samples can lead to poor estimates, particularly when M = 0, which leads to an infinite estimate of N (Robson and Regier 1964). To overcome this, Chapman (1951) proposed the following approximately unbiased LP estimator;

where N̂ is the estimated size of the population, M is the number of individuals marked on occasion 1, C is the total number of individuals captured on occasion 2, and R is the number of individuals in C that were previously marked. Estimates are said to be nearly unbiased where R ≥ 7 (Krebs 1999). However, the resulting population estimate is difficult to evaluate without an associated measure of dispersion, such as standard deviation, which is calculated as follows for equation (2):

Data were provided by the Canadian Bird Banding Office (BBO 2012) and the Danish Zoological Museum (DZM 2014) in Copenhagen, Denmark, for all peregrine falcons encountered (i.e., any report of a banded bird, alive or dead) in North America from 1970 through 2010.

I defined the northern population as all migratory peregrines originating north of latitude 60°N in Alaska, north of latitude 54°N in Canada, and in Greenland. I used latitude 60°N for Alaska in order to exclude nestlings originating in the range generally considered to be occupied by the nonmigratory subspecies F. p. pealei (the Aleutians and the coastal panhandle). For Canada, I followed the USFWS (2008) and used 54°N as the northern cutoff. All of Greenland is north of latitude 60°N (Figure 1), and the peregrines that breed there are highly migratory (Restani and Mattox 2000).

Yates et al. (1988) concluded breeding populations that originated in the western and eastern portions of Arctic and sub-Arctic tend to remain separate during outward migration. Similarly, based on latitudinal and longitudinal origins, USFWS (2008) defined three management populations: northern for falcons originating north of 54°N, and western and eastern for those originating west and east of 100°W, respectively. To characterize the degree to which northern HY migrants disperse longitudinally, I calculated relative recapture rates in Texas and in New Jersey, Maryland, and Virginia of HY falcons that originated in Alaska, Canada north of 54°N (i.e., Yukon, Northwest Territories, Nunavut, northern British Columbia, northern Alberta, northern Saskatchewan, and northern Manitoba, northern Quebec, northern Labrador, northern Ontario), and Greenland as follows:

where R is the number of recaptures and M is the number of locally marked nestlings.

I included counts of only those nestlings marked in Alaska north of latitude 60°N, those marked in Canada north of latitude 54°N, and those marked in Greenland. In addition, I partitioned the data to characterize the northwestern and northeastern populations (Yates et al. 1988; USFWS 2008). All American states and Canadian provinces and territories falling within the Pacific and Central flyways were considered to be western, while those jurisdictions falling within the Mississippi and Atlantic flyways as well as Greenland were considered to be eastern (Figure 1). I further excluded those years in which fewer than 100 nestlings were marked to account for any effect of low banding effort.

I restricted the captured sample (C) to all “how obtained” encounters coded “89” (banded bird trapped and released during banding operations in different 10-min block than the 10-min block where originally banded) in Texas, and in New Jersey, Maryland, and Virginia, then further restricted the sample to exclude any year in which fewer than 100 HY falcons were captured. I combined occasion 2 captures from New Jersey, Maryland, and Virginia because substantial trapping effort has occurred in the region and because of their geographic proximity to one another (Figure 1). Although Kiff (1988) indicated that the northern population accounted for as much as 80% of the total population in North America, I down-adjusted the captured sample to account for the proportion of C that was not northern by assuming that the origins of falcons in C were proportional to the origins of birds in R.

I tallied only live recaptures of HY falcons that were included in the sample of M (i.e., occasion 1) and C (i.e., occasion 2). I accounted for recaptures of marked HY peregrines from jurisdictions excluded on the basis of geographic location (not northern) or sample size in the unmarked sample (C). For example, I excluded all live recaptures of birds marked as nestlings in New Jersey, Maryland, and Virginia, as well as HY birds previously captured and banded in New Jersey, Maryland, and Virginia, from R but included them in C.

I used equation (2) to calculate an estimate of the population size at the time of the original marking by combining banding data across all years for nestlings marked (M) in all northern jurisdictions, C and R (i.e., pooled sample, Table 1). For comparison, I then restricted these data to include only those jurisdictions in the northwest and northeast for years in which M and C were both greater than 100 (i.e., restricted sample; Table 1). I calculated an estimate using only the restricted sample to account for any influence of years in which banding effort was relatively low. In this analysis, I annualized LP estimates by dividing the output of the above formula by the number of years over which data were pooled. I used equation (3) to estimate dispersion (SD) for estimates calculated using the pooled and restricted samples.

To quantify juvenile mortality from the time that marks were originally applied to the time usually associated with capture during migration, I converted annual survival estimated for first-year peregrine falcons (0.54) to daily survival (Craig et al. 2004) as follows:

where S_D is equal to the daily survival rate and S_A is equal to the annual survival rate. I used known breeding phenology from the long-term study at Rankin Inlet, Nunavut (A. Franke, unpublished data), to estimate the number of days from the time the marks were applied (occasion 1; typically ∼25 d of age) to occasion 2 during migration (typically ∼90 d of age).

To provide estimates of harvest levels for HY peregrines, I used the management recommendation from Millsap and Allen (2006). The authors indicated that passage peregrine falcons could withstand a maximum sustainable yield rate of 0.16 (i.e., the rate of harvest that did not produce a decline in the number of breeding adults). However, USFWS (2008) indicated that actual harvest rates for HY peregrine falcons should not exceed one-half of the estimated maximum sustainable yield up to a maximum of 5% of estimated minimum annual production.

There were 5,073 nestlings marked in Alaska north of latitude 60°N, 3,203 in Canada north of latitude 54°N, and 2,454 in Greenland—a total of 10,730 nestlings marked (M) from 1970 through 2010. Over the same period, 6,839 HY falcons were captured and subsequently released in the states of New Jersey, Maryland, and Virginia, and 5,169 in Texas for a total of 12,088 HY falcons captured (C). However, after down-adjusting C to account for nonnorthern falcons, C_corrrected totaled 7,973 falcons in the pooled sample, and 3,117 and 1,177 HY falcons in each of the restricted samples for the west and east, respectively. A total of 129 (1.2%) falcons recaptured (R) originated from northern regions (Table 2). After excluding years in which M and C were both fewer than 100, I tallied M, C, and R in the west and east separately. In the west, I retained 19 of 41 y , which resulted in marking 5,764 nestlings (M). In the east, I retained only 11 of 41 y, which resulted in marking 1,665 nestlings. Recaptures totaled 58 in the west and 33 in the east (Table 2).

Recapture rates of HY falcons calculated for each of the two main (east and west) trapping locations were strongly influenced by natal origin (Figure 2). For example, for every 1,000 nestlings marked in Alaska, 9.3 were recaptured in the Texas and only 0.4 were recaptured in the New Jersey–Maryland—Virginia region. Similarly, for every 1,000 nestling falcons marked in Greenland, 16.7 were recaptured in the east, and only 2.04 were recaptured in the west. Recapture rates of northern falcons in Texas declined from west to east, and were best explained by a linear function (R² = 0.76, Figure 2). Recapture rates in New Jersey, Maryland, and Virginia of northern HY falcons decreased from east to west, but were best explained by a quadratic function (R² = 0.80, Figure 2).

Without accounting for annual fluctuations or long-term trend, the LP population estimate annualized across all 41 y (i.e., pooled sample) was 16,055 ± 2,984 birds (all values mean ± SD). However, after excluding years in which M and C both comprised fewer than 100 individuals (i.e., restricted sample), the LP population estimate in the west was 16,035 ± 2,040 falcons. Similarly, using only the restricted sample, the LP population estimate in the east was 5,245 ± 500 falcons.

Mortality for the 65-d duration between typical banding age of nestlings (∼25 d) to arrival of HY falcons (∼90 d) in Texas and New Jersey–Maryland–Virginia was estimated to be 10.4%. Using the LP estimate from the pooled sample, and accounting for mortality, the estimated harvest level at 0.05 maximum sustainable yield was 586 − SD and 853 + SD HY falcons. Harvest estimates based on the LP estimates calculated from the restricted samples ± SD in the west and east were 627 and 818 HY falcons, and 213 and 257 HY falcons respectively (Table 3), or 840 and 1,075 HY falcons, when summed.

The annualized LP population estimate was approximately 21,000 HY falcons when I combined estimates for the east and west. Jaffre et al. (2015) reported mean productivity for three breeding populations in the Canadian Arctic (Melville Peninsular, 1.9 ± 0.9; Baffin Island, 1.2 ± 0.9; and Rankin Inlet, 0.9 ± 0.4), and Ritchie and Shook (2011) reported mean productivity for east-central Alaska (1.6 young/occupied site). Using average productivity from these studies combined (1.4 young/occupied site) results in an estimate of the northern breeding population of more than 15,000 pairs or 30,000 breeding adults. Newton (1988) suggested that the number of nonbreeding peregrine falcons in a growing population in Scotland would stabilize at a minimum of three birds per breeding pair where environmental factors were not limiting. Assuming that the number of nonbreeding adults was equal to the number of breeding adults, the estimated total annualized adult breeding-aged population was in excess of 60,000 falcons, and the total combined population (individuals of any age class) at the end of a breeding season was in the order of 80,000 falcons by the year 2000. However, in the absence of a long-term, systematic trapping program on the Gulf Coast on the Mississippi Flyway, my estimate of the northern population cannot account for an unknown portion of the population migrating through the Mississippi Flyway. Indeed, Russell (2005) reported that large numbers (11,000–66,000) of peregrines were observed on oil platforms in the Gulf of Mexico, including the region at the terminus of the Mississippi Flyway.

Most previous attempts to estimate the size of the northern peregrine population are based on extrapolation of counts of breeding pairs from “intensive surveys in limited areas” (i.e., small spatial extent) and “subjective impressions of habitat capabilities” (White et al. 2013 for explaination of challenges). For example, to estimate the total population, Fyfe (1969) used known nesting densities in well-surveyed, small-scale studies and (based on presumed habitat quality) extrapolated densities to areas not surveyed. Based on 252,061 km² of optimum nesting habitat and another 331,945 km² of “more limited nesting habitat in northern Canada” Fyfe (1969) estimated a northern population of 7,500 pairs. Although Fyfe's (1969) estimate likely represented the best population estimate to date, I suggest that it was very likely too low. I acknowledge that the annualized population estimate reported here does not in any way account for change in population size through time. Cade (2013) reported that the North American population has undergone rapid recovery, making the annualized estimate reported here an integrated measure of the population low point (1970–1975) and a presumably much larger size by 2010.

Yates et al. (1988) concluded breeding populations originating in the western and eastern portions of Arctic and sub-Arctic tend to separate longitudinally during outward migration, and reflects a similar migratory pattern for peregrines breeding in northern Eurasia (Dixon et al. 2012). In addition, the analysis of recapture rates reported here supports the finding that western and eastern populations tend to remain separate, and are best treated as two distinct populations. Although rates of recapture in Texas for falcons originating from Yukon, Northwest Territories, and Nunavut are lower than those of falcons originating in Alaska, they are considerably higher than those for falcons originating in Greenland and northern Quebec. I suggest that the lower recapture rates of falcons originating from northern Canada (compared to Alaska birds in Texas, or Greenland birds in the east) probably result from a tendency to migrate directly south across the Gulf of Mexico to Mexico, or through Florida, and therefore, these birds do not mix completely with either the western (Central Flyway) or eastern (Atlantic Flyway) population prior to reaching the trapping stations. I suggest that HY falcons originating from Yukon eastward to the western shores of the Hudson Bay and the Melville Peninsula mix throughout the Mississippi Flyway to a much greater extent than they do with birds originating east of Hudson Bay. Similarly, falcons that originate from northern Quebec and Greenland appear to remain concentrated within the Atlantic Flyway.

In Canada, the legal status of the peregrine falcon has been changed from threatened to special concern (COSEWIC 2007), and in the United States, the species has been removed from the list of threatened and endangered species (USFWS 1999) resulting in legal harvest of wild-caught migratory peregrine falcons, which were used regularly for the practice of falconry (Ward and Berry 1972) from 1938 until 1970. Using the best data available to estimate the size of the northern peregrine falcon population, I conclude that the size of the northern migratory peregrine population in North America is larger than previous estimates with the exception of the estimate offered by Rich et al. (2004), and suggest that Alternative 8 (USFWS 2008), which would allow harvest of up to 5% (the management guideline set the by the USFWS) of first-year peregrine falcons from all management populations through any combination of resident and migrant harvest, is a safe, sustainable long-term approach for managing falconry harvest of peregrine falcons. Furthermore, using the USFWS harvest guideline, and the annualized estimate of HY falcons reported here (after mortality), it appears that the combined annual harvest limit in Canada, the United States, and Mexico could be conservatively set at 840 HY falcons without negative impact to the breeding population. However, in the event that harvest levels of this magnitude are implemented, I recommend that managers of wildlife agencies in Canada and the United States responsible for management of peregrine falcons implement ongoing monitoring programs capable of estimating trends in rates of occupancy and productivity. Ideally, I recommend that monitoring programs be accompanied by studies that investigate potential causes that underlie variability in occupancy and productivity.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding authors for the article.

Data S1. Banding data (1970–2010) used for the analysis of population estimates for northern juvenile peregrine falcons (Falco peregrinus) with implications for harvest levels in North America.

Found at DOI: http://dx.doi.org/10.3996/062015-JFWM-050.S1 (5359 TXT).

Data S2. Recovery data (1970–2010) used for the analysis of population estimates for northern juvenile peregrine falcons (Falco peregrinus) with implications for harvest levels in North America.

Found at DOI: http://dx.doi.org/10.3996/062015-JFWM-050.S2 (1056 KB TXT).

Reference S1. Green M, Swem T, Morin M, Mesta R, Klee M, Hollar K, Hazlewood R, Delphey P, Currie R, Amaral M. 2006. Monitoring results for breeding American peregrine falcons (Falco peregrinus anatum), 2003. U.S. Fish and Wildlife Service. Report BTP-R1005-2006.

Found at DOI: http://dx.doi.org/10.3996/062015-JFWM-050.S3; also available at http://nctc.fws.gov/resources/knowledge-resources/pdf/peregrine_breeding.pdf (489 KB PDF).

Reference S2. Rich TD, Beardmore CJ, Berlanga H, Blancher PJ, Bradstreet MSW, Butcher GS, Demarest DW, Dunn EH, Hunter WC, Iñigo-Elias EE, et al. 2004. Partners in Flight North American landbird conservation plan. Cornell Lab of Ornithology. Ithaca, New York.

Found at DOI: http://dx.doi.org/10.3996/062015-JFWM-050.S4 (3347 KB PDF); also available at http://www.partnersinflight.org/cont_plan/PIF2_Part1WEB.pdf (3.3 mB PDF) and http://www.partnersinflight.org/cont_plan/PIF3_Part2WEB.pdf (1.2 MB PDF).

Reference S3. Russell RW. 2005. Interactions between migrating birds and offshore oil and gas platforms in the northern Gulf of Mexico: final report. U.S. Department of the Interior, Minerals Management Service, Gulf of Mexico OCS Region, New Orleans, Louisiana. OCS Study MMS 2005-009.

Found at DOI: http://dx.doi.org/10.3996/062015-JFWM-050.S5; also available at www.data.boem.gov/PI/PDFImages/ESPIS/2/2955.pdf (5755 KB PDF).

Reference S4. [SARA] Species at Risk Act S.C. 2002, c. 29.

Found at DOI: http://dx.doi.org/10.3996/062015-JFWM-050.S6; also available athttp://laws-ois.justice.gc.ca/eng/acts/s-15.3/ (725 KB PDF).

This project was supported by the Government of Nunavut and ArcticNet Network of Centres of Excellence.

I thank the Canadian Bird Banding Office and the Danish Zoological Museum for providing banding data for peregrine falcons encountered in North America and Greenland from 1970 through 2010. Use of these data conforms to Environment Canada Bird Banding Office and the U.S. Geological Survey Bird Banding Laboratory protocol for publication of banding data. This analysis would not have been possible without the banding efforts of many volunteers and biologists in Greenland, Canada, and the United States; I acknowledge their dedication and countless hours of work banding nestlings in the north and trapping passage falcons on migration. I am indebted to B. Millsap, C. Francis, S. Ward, D. Haukos, C. Dykstra, M. Yates, M. Fuller, C. Farmer, and three anonymous reviewers who all provided insightful comments and original ideas that led to a considerably improved manuscript. I am particularly grateful to L. Oliphant for his contribution to initial development of this manuscript, and for the many hours spent discussing peregrine falcons in general, and the size of the northern peregrine population in particular.

Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.