effects of fluctuating food availability on breeding Arctic terns (Sterna paradisaea), The

Auk, The, Jul 1997

D. SUDDABYl AND N. RATCLIFFE23

1 Royal Society for the Protection of Birds, The Baelans, Fetlar, Shetland ZE2 9DJ, United Kingdom; and 2 Royal Society for the Protection of Birds, The Lodge, Sandy, Bedfordshire SG19 2DL, United Kingdom

Reproduction is a costly process that involves an investment of effort that can decrease parental condition and survival (Partridge 1989, Stearns 1992). In long-lived iteroparous species such as seabirds, reproductive costs produce a tradeoff in energy allocation between current and future breeding attempts (Clutton-Brock 1984), and the breeding effort should be adjusted to maximize lifetime reproductive success. Food availability can have substantial effects on the relative costs of reproduction and thus may be an important factor in determining reproductive strategies in seabirds (Boekelheide and Ainley 1989, Sydeman et al. 1991, Pons and Migot 1995).

Cairns (1987) hypothesized that during periods of reduced food supply, seabirds should increase foraging effort in order to buffer the effects of food scarcity on breeding performance. Further reductions in food may result in parents being unable to increase foraging effort without incurring excessive costs, resulting in a decline in reproductive success (Monaghan et al. 1989, 1992). Finally, during periods of extreme food shortage, condition and survival of adults may be adversely affected (Hamer et al. 1991). Food supply is expected to affect breeding parameters within a specific range of prey availability, outside of which it will have little effect (Cairns 1987, Phillips et al. 1996).

The abundance of sandeels (Ammodytes marinus) was severely reduced in Shetland waters between 1985 and 1990 due to successive years of poor recruitment of the group-0 cohort (i.e. fish spawned in the current year) and a subsequent decline in spawning-stock biomass (Wright and Bailey 1993). Seabirds responded by increasing their foraging effort (Uttley 1992, Hamer et al. 1993, Monaghan et al. 1994) and exploiting alternative prey species (Martin 1989, Hamer et al. 1991). Despite these changes in behavior, the breeding success of surface-feeding species was adversely affected (Monaghan et al. 1989; Hamer et al. 1991, 1993; Phillips et al. 1996).

Arctic Terns (Sterna paradisaea) suffered six consecutive years of almost complete breeding failure on Shetland (Walsh et al. 1990) due to adults abandoning clutches and chicks starving shortly after hatching (Monaghan et al. 1989, 1992). The number of breeding Arctic Terns at Shetland declined by 55% between 1981 and 1989 due to lack of recruitment and possibly to increased rates of nonbreeding, emigration, and mortality (Avery et al. 1993). Arctic Terns are more vulnerable to reductions in food supply than most other seabirds at Shetland owing to their low degree of behavioral plasticity (Monaghan et al. 1992). This results from small body size (Craik and Becker 1992), specialization toward surface feeding at short ranges from the colony (Pearson 1968), and lack of alternative prey species (Uttley et al. 1989).

In 1991, a dramatic increase in the availability of sandeels at Shetland resulted from large numbers of group-0 fish being swept north from Orkney by currents (Wright and Bailey 1993). Recruitment of these fish into the Shetland stock resulted in an abundance of group-1 fish in 1992 and an increased spawningstock biomass in subsequent years (Anonymous 1995). This increase in sandeel abundance was associated with improved breeding performance for other species of surface-feeding seabirds, including Black-legged Kittiwakes (Rissa tridactyla; Hamer et al. 1993), Parasitic Jaegers (Stercorarius parasiticus; Phillips et al. 1996), and Great Skuas (Catharacta skua; Ratcliffe 1993). Despite the extreme sensitivity of Arctic Terns to changes in food availability, no published information exists on their breeding performance since the recovery of the sandeel population at Shetland. Here, we examine the responses of Arctic Terns to variations in sandeel abundance at Shetland from 1990 to 1994.

Methods.-The availability of prey to Arctic Terns was inferred from the SOAFD (Scottish Office Agriculture and Fisheries Department) estimates of sandeel abundance in Shetland waters (Anonymous 1995). The data provided a meaningful index of prey availability to terns because the locations of the sandeel fishing grounds and tern colonies were closely juxtaposed (Monaghan 1992), and the sea-surface availability of sandeels is directly related to their abundance (Wright and Bailey 1993). Estimates of sandeel abundance were made from Virtual Population Analysis (VPA) of catch-at-age data from research trawls and commercial catches at fishing grounds around Shetland (Cook and Reeves 1993). Abundance was estimated separately for the first (up to 30 June) and second halves of the year (after 30 June). The biomass of group-0 and group-1 sandeels (in metric tons) in each half of the year was estimated by multiplying the estimated group abundance by the average mass of an individual (Anonymous 1995). VPA suffers from inaccuracies due to the difficulty in sampling cohorts randomly and the assumptions concerning natural mortality rates and terminal mortality (Cook and Reeves 1993). However, it is the best estimate of trends in sandeel abundance available for Shetland waters, and it provides a coarse independent index of tern food supply.