Direct and indirect estimates of Peregrine Falcon population size in Northern Eurasia

ABSTRACT.-We used two different methods to estimate the density of nesting Peregrine Falcons (Falco peregrinus) across different parts of northern Eurasia. In the "territory-density" method, we extrapolated our density estimate of 406 km^sup 2^ per territory (95% CI = 295 to 650 km^sup 2^ per territory) in a high-density area, the Pyasina basin on the Taymyr Peninsula, to other similar areas across the range defined by published estimates. To estimate numbers in low-density areas, we used published data that suggested that Peregrine Falcon territories occur every 1,000 km^sup 2^. Based on the nesting association between Peregrine Falcons and Redbreasted Geese (Branta ruficollis), we used a second, post hoc method to provide a comparative estimate where the ranges of the two species overlap. This model was based primarily on the population ecology of the Red-breasted Goose and included parameters such as the proportion of the goose population nesting with peregrines, the proportion of peregrine pairs associated with geese, goose population size, and three other variables. Some of these variables were already known, whereas others had been estimated as part of another study. The territory-density and nesting-association methods led to estimates of 1,586 (95% CI = 991 to 2,179) and 2,417 (95% CI = 1,306 to 3,528) falcon territories, respectively, across the common range of Peregrine Falcons and Red-breasted Geese; the first method suggested a population of 3,652 falcon territories (95% CI = 2,282 to 5,018) across the entire range F p. calidus. Although both approaches entailed several major assumptions, together they provide the only quantitative estimate of this remote population of Peregrine Falcons. Received 26 February 1999, acce ted 1 October 1999.

THE REMOTENESS of the breeding grounds of Peregrine Falcons (Falco peregrinus) in northern Eurasia means that little is known about their numbers in this region. Even in North America, where Peregrine Falcons are relatively well studied, the continental estimate of 7,000 to 10,000 breeding pairs before the pesticide-induced decline in numbers was extrapolated from suitable information from only a few areas (Kiff 1988). Kolosov (1983) suggested that the Peregrine Falcon population east of the Ural Mountains contained several thousand pairs, but no firm estimate of numbers relates specifically to the range of E p. calidus, which is the most widespread subspecies in the region.

Much of the literature on Peregrine Falcon densities in northern Eurasia relates to a time when the population was severely depressed by pesticides (e.g. Ivanov et al. 1983, Kokorev 1995), but numbers have increased since then (Quinn et al. 2000), as they have in many other parts of the world (Enderson et al.1995). A current population estimate is needed to help assess the influence that trapping for falconry has on the population as it migrates through Eurasia (Eastham et al. 2000).

Here, we estimate the size of the E p. calidus population by extrapolation using density estimates from our study area in Taymyr and by classifying regions across the entire range into those likely to hold low or high densities. We also use a second method to verify the estimate across part of the range by applying a model that incorporates population parameters linked to the nesting association of Red-breasted Geese (Branta ruficollis) with Peregrine Falcons and other birds of prey. The geese are thought to depend heavily on Peregrine Falcons to maintain an area free of terrestrial predators, especially arctic foxes (Alopex lagopus), around goose nests (e.g. Dementiev and Gladkov 1951, Kretschmar and Leonovich 1967). Although many nesting associations have been described (e.g. Bogliani et al. 1992, Ueta 1994, Larsen and Grundetjern 1997), Red-breasted Geese are possibly unique among bird species in the extent to which this association occurs. Of the 23 Peregrine Falcon nests found by Kretschmar and Leonovich (1967), 19 had Red-breasted Geese nesting next to them. Kostin and Mooiji's (1995) data agree with these findings, as do ours. We suggest that the nesting-association model can provide a useful way to verify the scale of the estimate provided by the first method. Furthermore, it describes an unusual scenario in which the population of one species theoretically can be estimated using parameters based entirely on the population of another species. We do not necessarily suggest that it should become a standard way to monitor the falcon population. In our case, it was possible to do so on a post hoc basis because the parameter estimates were already available or could be easily derived. Some of the parameters can be estimated annually with ease, for example the population size of Red-breasted Geese. Others parameter estimates are more difficult to obtain, however, so the nesting-association method may not always be a cost-effective alternative to the standard territory-density method.

STUDY AREA AND METHODS

The breeding range of F. y. calidus stretches eastward from the White Sea near the border of Fennoscandia and Russia (35degE) to the Lena Delta (130degE) and lies above 62degN (White and Boyce 1988, Stepanyan 1995; Fig. 1), but the exact locations of the boundaries are uncertain. Peregrine Falcons in northern Fennoscandia generally are believed to belong to the nominate race, although they may belong to F p. calidus (C. J. Henny pers. comm.). The breeding range of the Red-breasted Goose occurs above 65N and includes the Yamal (70degE), Gydan (80degE), and Taymyr (100degE) peninsulas (Krivenko 1983, Hunter and Black 1996) and covers approximately 40% of the entire range of F. p. calidus. Seventy percent of the Red-breasted Goose population is thought to occur on Taymyr (Kostin 1985), which stretches from 71 to 78degN and 80 to 114degE and is the most northerly part of the Eurasian continent. The data presented in this paper were collected along 823 km of river in the Pura and Pyasina basins in the tundra zone of western Taymyr (Fig. 2) each breeding season from 1995 to 1998.