An experimental study examining the anti-predator behaviour of Sabine s gulls (Xema sabini) during breeding

Transcription

1 Springer-VerlagTokyohttp:// of EthologyJ EtholLifeSciences /s J Ethol (2005) 23: Japan Ethological Society and Springer-Verlag Tokyo 2005 DOI /s ARTICLE Iain J. Stenhouse H. Grant Gilchrist William A. Montevecchi An experimental study examining the anti-predator behaviour of Sabine s gulls (Xema sabini) during breeding Received: January 19, 2004 / Accepted: October 10, 2004 / Published online: January 15, 2005 Abstract Anti-predatory behaviour is widespread among a broad range of animal taxa, including birds. Nest defence is not without risk, however, and parent birds face a trade-off between the survival of their offspring and the risk of injury or mortality to themselves. This study focused on the antipredator behaviour of the Sabine s gull (Xema sabini), a ground-nesting, Arctic breeder. Specifically, we quantified the gulls behavioural response towards natural predators, a human intruder, experimental predator decoys, and a nonpredatory decoy. Neither the distance at which nesting Sabine s gulls first reacted to natural predators, nor the relative intensity of their response, differed with incubation stage or predator type. However, response behaviour was highly variable among pairs. In response to decoys, Sabine s gulls responded strongly towards predatory species, compared with a non-predatory species. The distance at which they first swooped at a human intruder was also variable, and there was no seasonal trend. Sabine s gulls were often joined in nest defence by conspecifics, Arctic terns, and shorebirds nesting nearby, although the number of conspecifics involved in attacks was not related to the proximity of neighbouring nests. Key words Anti-predator behaviour Nest defence Predation Sabine s gull Xema sabini Introduction Predation is a powerful selective pressure on most species, particularly during reproduction (Lack 1954; Krebs 1973). I.J. Stenhouse (*) W.A. Montevecchi Cognitive & Behavioural Ecology Programme, Department of Psychology, Memorial University of Newfoundland, St. John s, NF, Canada, A1B 3X9 Tel ; Fax istenhouse@audubon.org H.G. Gilchrist Canadian Wildlife Service, National Wildlife Research Centre, Carleton University, Ottawa, ON, Canada, K1A 0H3 The extent of predation experienced by a species is influenced by the abundance of predators, and if the predators are generalists, by the availability of alternative prey (Bêty et al. 2002). Vulnerable species may forego breeding in years of intense predation pressure (see Spaans et al. 1998), while others attempt to limit the extent of predation via life history (Martin 1995) or behavioural adaptation (Kruuk 1964). In defending offspring, parents increase the chances of survival of their young, and, in turn, increase their own reproductive fitness (Clutton-Brock 1991). There are costs associated with anti-predator behaviour, however, including direct mortality or injury caused by a predator (Myers 1978), and a reduction in the time and energy available for other essential activities (Walters 1982). Consequently, the level of defence should vary with the predator type, the value of offspring, the relative risk to adults themselves, and possibly individual age and experience (Kruuk 1964; Curio 1975; Andersson et al. 1980; Patterson et al. 1980). Anti-predator behaviour is widespread in a broad range of animal taxa, including birds (Curio 1976; Redondo 1989). Among birds, predators are often the main cause of nest failure (Clark and Wilson 1981), and nest defence can be critically important in reproductive success, particularly in ground-nesting birds, where the nest is easily accessible to predators (Gochfeld 1984). Sabine s gulls, Xema sabini, nest on the ground in the Arctic. Their nest dispersion is variable, from solitary to loosely colonial, and they exhibit conspicuous aggressive and defensive behavioural displays towards predators (Day et al. 2001). Their distraction display is particularly notable, being more characteristic of shorebirds (Scolopacidae) and some jaegers (Stercorariidae) than other gulls (Laridae; Brown et al. 1967; Day et al. 2001), perhaps suggesting that they have evolved in an environment where predation pressures are intense. The nests of Sabine s gulls are vulnerable to numerous mammalian and avian predators, including Arctic fox, Alopex lagopus; red fox, Vulpes vulpes; glaucous gull, Larus hyperboreus; herring gull, Larus argentatus; parasitic jaeger, Stercorarius parasiticus; and common raven, Corvus corax (Day et al. 2001). Peregrine falcons, Falco peregrinus, have been observed to attack isolated adults in

2 104 flight during breeding (Parmelee et al. 1967; Day et al. 2001). Although few details of their reproductive ecology are known, Sabine s gulls are highly variable in their contributions to other types of parental investment and care, with considerable variation observed between individuals within pairs (Stenhouse et al. 2004). This study focused on the antipredator behaviour of breeding Sabine s gulls. Specifically, through observation and experimental manipulation, we examined whether Sabine s gulls (1) distinguished between predators and a non-predatory species in their defensive response, (2) responded differently to an avian versus a mammalian predator, (3) exhibited variation among pairs in their responses to predators, and (4) altered their response as the season progressed. This latter prediction is based on the idea that the theoretical value of a clutch/ brood increases with time (Barash 1975; Weatherhead 1979; Andersson et al. 1980). Methods Field work was conducted from late May to early August of in the East Bay Migratory Bird Sanctuary (64 01 N, W), Southampton Island, Nunavut, in the eastern Canadian Arctic. The East Bay Sanctuary encompasses an area of approximately 1,200 km 2. Despite being located within the Low Arctic region, it is generally High Arctic in character, influenced in its ecological and physical characteristics by the deep, cold waters of the Foxe Channel. Land-fast sea ice generally remains in East Bay well into July, and daily minimum temperatures are usually close to freezing throughout summer (Jun Aug). The study area is a km area of low-lying coastal wetland tundra, which includes numerous brackish and freshwater ponds. The region supports a diverse avian community (Abraham and Ankney 1986). During the years of this study, pairs of Sabine s gulls, pairs of herring gulls, and 2 pairs of parasitic jaegers nested within the study area. One peregrine falcon nest was located at the nearest cliff face, approximately 15 km east of the study area. Common ravens were occasionally observed flying over the study area, but were never observed to land. At this site, Sabine s gulls generally nest solitarily, with an average inter-nest distance of approximately 100 m (Stenhouse et al. 2001). Nest phenology was calculated from known laying dates of first eggs, or from the hatch date of the first egg, based on a 21-day incubation period (Stenhouse et al. 2001). Nest phenology was categorized as early incubation (days 1 7), mid-incubation (days 8 14), late incubation (days 15 21) or hatch (days 22 24). During interactions involving other birds (group defence), the identification of the species involved was recorded. Response to predators The response of breeding Sabine s gulls to predators was recorded from portable canvas blinds, placed ~100 m from nests, during 3-h behavioural watches at the nests of three pairs in 2000 (69 h) and eight different pairs in 2001 (162 h). Observations concentrated on the behaviour of the focal nesting pair in response to the presence of predatory species. The time the pair spent in the interaction was recorded, and the distance (to the nearest 5 m) of the predator from the nest at the first reaction of the nesting birds was estimated visually. The response of Sabine s gulls was scored on a scale of increasing intensity and risk: 0 = no response, 1 = flies out, 2 = chases predator, 3 = chases and calls, 4 = swoops/dives, 5 = strikes predator. Although some interactions were missed, as Sabine s gulls appeared to react to predators that could not always be seen in advance from the blind, we believe this happened rarely enough that it did not bias our results. Response to predator decoys In 2001, decoys of two known nest predators (a mounted red fox and a carved herring gull) and a novel nonpredatory control species (a plastic hunting decoy painted to resemble a female harlequin duck, Histrionicus histrionicus) were used to examine the type and relative intensity of parental response. Red foxes are native to Southampton Island, as are Arctic foxes. Observations were carried out at eight nests, at which the decoys were presented in a random order. Decoys, covered by a sheet of burlap attached to a line, were placed 25 m from nests in a random direction, so that they were clearly visible from nests. The observer then retreated to approximately 50 m from the nest. Once the incubating bird had been back on the nest for 10 min, the decoy was exposed by tugging on the line and removing the burlap, and the behaviour of the nesting pair and neighbouring birds was recorded for 15 min. Nest phenology was controlled in this experiment, as all trials were conducted during late incubation. An attack was defined as any swooping or diving behaviour directed at the decoy. For each trial, we recorded the following: decoy type, distance to nearest neighbouring Sabine s gull nest, time to first attack (latency), total time spent in attack (duration), number of birds involved in attack, and the intensity of attack. An intensity index focused on the behavioural response of the nesting pair ranged from none to intense (Table 1). This experiment examined whether the type of decoy affected the behav- Table 1. Intensity index, response type and behaviour of Sabine s gulls in response to predator decoys close to the nest Index Behaviour Response intensity 0 No response from either parent None 1 Incubating bird calls from nest, no response Low from mate 2 One bird circles over decoy Mild 3 Both birds circle over decoy Mild 4 One bird attacks decoy Intense 5 Both birds attack decoy Intense

3 105 ioural response of Sabine s gulls, by comparing their reactions to known predators (fox, gull) and a non-predatory species (duck). In addition, whether Sabine s gulls responded differently to an avian versus a mammalian predator was also tested. The presentation of a stationary decoy does not mimic the approach of a natural predator, particularly for avian species. The reactions of Sabine s gulls, however, appeared to be unaffected by the sudden appearance and the static nature of the decoys, which elicited realistic responses similar to their reactions toward natural predators (above). Response to human intruders The nest defence behaviour of Sabine s gulls in response to a human intruder was also recorded. Observations were made during the incubation period at 12 nests. However, four of these nests were depredated between trials, with the result that all 12 nests were visited at least twice, and 8 were visited four times. Nests were approached during calm, dry weather by a single observer at a slow, steady walk, in a direct line of sight of the incubating bird. The distances at which the behaviours of the incubating bird changed were measured to the nearest metre with a rangefinder (Yardage Pro Sport Laser Rangefinder, Model , Bushnell, KS, USA; accuracy ± 1 m), which required the observer to pause for a few seconds to take the reading. The recorded behaviours included: flew up, called, swooped, made contact, defecated, and performed distraction display. The number of Sabine s gulls involved in the attack, as well as the number and species of other birds involved, were also recorded. The intruder remained standing over the nest for 2 min and continued to record any changes in behaviour. Nest phenology was calculated and categorized as above. Statistical analysis Statistical significance was recognized at P 0.05, and values are generally reported as mean ± SD unless otherwise stated. Associations between two variables were measured using the Pearson product-moment correlation. Differences among sample means were examined using analyses of variance (ANOVA). Where multiple observations were recorded at each nest, predictor variables were nested within gull pairs for analysis in a nested ANOVA design. Associations among categories were explored with the Bonferroni post hoc test, and reported as the difference, standard error (SE) and probability. All statistical tests were performed using DataDesk statistical software v. 6 (Data Description, Ithaca, NY, USA). Results Response to predators For both years combined, a total of 105 interactions with natural predators were observed; in 2000, responses were recorded for only two predators (n = 29, Table 2), herring gulls (93%) and parasitic jaegers (7%). In 2001, responses were recorded for five different predators (n = 76; Table 2); herring gulls (70%), parasitic jaegers (23%), Arctic foxes (3%), glaucous gulls (3%), and a peregrine falcon (1%). The distance at which nesting Sabine s gulls first reacted to the most commonly encountered predators (parasitic jaegers and herring gulls) did not differ with incubation stage (F 10,99 = 1.8, P = 0.08), or predator species (F 7,99 = 1.9, P = 0.09), and was close to significance among pairs (F 10,99 = 1.9, P = 0.05). Similarly, the relative intensity of their response did not differ with incubation stage (F 10,99 = 1.6, P = 0.11) or predator species (F 7,99 = 1.4, P = 0.20), but did differ significantly among pairs (F 10,99 = 2.7, P 0.001). There were no differences, however, in the length of time they spent in behavioural interactions with these two predators (incubation stage F 9,93 = 0.8, P = 0.61; predator species F 6,93 = 1.3, P = 0.24; among pairs F 10,93 = 1.1, P = 0.37). Predation pressure was not high in either 2000 or Of the three pairs under observation in 2000, none lost eggs, and of the eight pairs under observation in 2001, only one lost eggs due to predation. This depredated pair showed the lowest mean intensity of response towards predators (1.8 ± 1.4; range for other 10 pairs = ). The likelihood that other Sabine s gulls joined in defensive attacks was not related to the proximity of the nearest conspecific nest (F 8,104 = 1.4, P = 0.20). During instances of communal defence, however, the following species participated: Sabine s gulls (30% of total observed interactions); Arctic terns (13%); ruddy turnstones, Arenaria interpres (2%); and white-rumped sandpipers, Calidris fuscicollis (1%). Table 2. The number of observations, mean ± SD distance from nest, mean ± SD response score, and the percentage of interactions involving communal defence for natural predators observed in 2000 and 2001 Year Predator Number (n) Distance from nest (m) Response score Communal defence (%) 2000 Herring gull ± ± Parasitic jaeger ± ± Herring gull ± ± Parasitic jaeger ± ± Glaucous gull ± ± Peregrine falcon Arctic fox ± ±

4 106 Table 3. The mean (±SD) latency time, attack duration, and response score, and the percentage of interactions involving communal defence, for each predator decoy type Decoy type Number (n) Latency (s) Duration (s) Response score Communal defence (%) Duck ± 361a 30 ± 45a 1.4 ± 1.3a 0a Gull ± 179b 352 ± 174b 4.1 ± 0.3b 4a Fox ± 125b 397 ± 292b 4.5 ± 0.5b 4a Letters indicate significant differences or lack thereof between variables Response to predator decoys The mean time to initial response (F 2,21 = 12.1, P < 0.001), duration of attack (F 2,23 = 8.2, P = 0.002), and attack intensity (F 2,23 = 33.2, P < 0.001) all differed among decoy types. In all cases, response was greater for the predators (gull, fox), when compared with the non-predatory species (duck; Table 3). There were no significant differences, however, in response to the gull or fox decoys in terms of the time it took Sabine s gulls to respond (diff = 0.87 ± SE, P = 1.00), or attack duration (diff = ± 99.0 SE, P = 0.96) or intensity (diff = 0.37 ± 0.42 SE, P = 0.76). The number of birds involved in attacks did not differ among decoy types (F 2,23 = 3.2, P = 0.06), increase the total number of swoops at a predator (r 2 = 0.18), or appear to have any association with the proximity of the nearest conspecific nest (r 2 = 0.01). Two of 24 trials (8%) involved communal attacks, neither of which was directed at the non-predatory decoy. Sabine s gulls were involved in both of these communal attacks, and Arctic terns were involved in one. Response to human intruders There were no significant differences in the mean distance at which Sabine s gulls first flew (F 12,39 = 0.41, P = 0.94), called (F 11,34 = 0.68, P = 0.73), or struck (F 7,13 = 3.5, P = 0.17) a human intruder in relation to nest phenology. There was, however, a significant difference in the distance at which they first swooped at a human intruder between incubation stages (F 14,30 = 3.6, P = 0.03) and among pairs (F 14,30 = 6.3, P = 0.007). The overall mean distance at first swoop was lower in mid-incubation than in early or late incubation, and highly variable among pairs throughout. There was no significant difference in the mean number of birds involved in attacks in relation to incubation stage (F 15,39 = 2.3, P = 0.06), although there was a significant difference among nests (F 8,39 = 7.1, P < 0.001), with some pairs gaining conspecifics in their attacks more often than others. The number of Sabine s gulls involved in attacks, however, did not have a strong association with the proximity of the nearest conspecific nest (r = -0.48). One pair displayed both the greatest mean distance to first swooping at a human intruder (81.5 ± 14.9 m), and on average more birds were involved in attacks at that nest (4.5 ± 2.4 birds). Overall, additional Sabine s gulls (i.e. not members of the residential pair) participated in 47% of interactions, Arctic terns in 15%, and herring gulls in 5% (n = 40). Discussion Sabine s gulls reacted intensely to Arctic foxes, glaucous gulls and peregrine falcons at considerable distances ( m), although these predators were rarely encountered in both 2000 and 2001 (Table 2). Herring gulls were the most commonly encountered predator and, consequently, were attacked most often. Behavioural interactions with parasitic jaegers, however, appeared to occur more often than their nesting density in the study area (0.4 nests/km 2 ) might predict, suggesting that they may represent a greater threat to nests than herring gulls. Sabine s gulls responded to humans approaching the nest at a relatively short distance (~90 m) and with a comparatively weaker intensity than displayed to natural predators (Table 2). Unlike many other bird species (Biermann and Robertson 1981; Shields 1984; Redondo and Carranza 1989; Brunton 1990; Mallory et al. 1998; Galeotti et al. 2000), there was no trend to suggest that the level of anti-predatory behaviour displayed by Sabine s gulls changed over the course of the incubation period. Although pairs varied in the distance at which they first swooped at a human intruder in relation to incubation stage, these results may not be biologically significant because pairs did not consistently increase or decrease. Overall, these results suggest that neither habituation (a waning of response), nor positive reinforcement (an intensifying of response) were involved (see Knight et al. 1987). They also suggest that the value of the current breeding episode did not increase with offspring age, or with a diminishing renesting potential (see Montgomerie and Weatherhead 1988). For species breeding at high latitudes, where the reproductive window is extremely short, it is likely that renesting potential is zero throughout the season, and that the current breeding attempt is of maximum value from the outset. As parent birds face a trade-off between the survival of their offspring and the risk of injury or mortality to themselves, we assumed that the level of nest defence would reflect (1) the potential threat posed by a predator to the eggs and/or young, and (2) the potential threat posed by the predator to the parents themselves (Andersson et al. 1980). However, the experimental results suggest either that the herring gull and the fox posed a similar threat to Sabine s gull eggs/young, or perhaps that the fox was a greater threat towards adults, which dampened the intensity of the response.

5 107 Although parent birds would be expected to defend their offspring more vigorously from a predator that poses a low risk to themselves, they may put themselves at greater risk from other predators in the process. The fresh remains of three adult Sabine s gulls found within the study area suggest that adults are at some risk of predation during breeding. All three cases appeared to result from peregrine falcon attacks. Although peregrine falcons occur rarely in the study area, and are unlikely nest predators, they are known to take both young and adult Sabine s gulls in flight (Day et al. 2001). Thus, Sabine s gulls involved in aerial chases and/or focused on attacking other predators might put themselves at greater risk of attack by a peregrine falcon. In response to predators, both natural and decoys, most cases of communal defence involved conspecifics. Sabine s gulls contributing to communal defence included neighbouring nesters and/or failed or non-breeders in the area. Arctic terns were also involved when these species nested close together. On a few occasions, shorebirds nesting near Sabine s gulls also contributed to communal defence, particularly ruddy turnstones. Interestingly, red phalaropes, Phalaropus fulicaria, were never observed to be involved in communal defence, despite nesting at a relatively high density (2.9 nests/km 2 ) in the same area as Sabine s gulls (Smith 2003). Herring gulls were rarely involved in attacks on human intruders, and only at the few sites where they nested within 250 m of Sabine s gulls. This was likely due to the fact that a human intruder can be seen at more than 1 km in the flat landscape of East Bay. Therefore, we conclude that herring gulls were defending against a perceived threat towards their own nests, rather than making a deliberate contribution to communal defence with other species. Generally, Sabine s gull pairs showed considerable variation in their response to predators and human intruders, with some pairs consistently attacking sooner and with greater intensity than others. This may be related to their age and/or previous experience (Montgomerie and Weatherhead 1988), but we have insufficient information on individuals at this time to investigate these issues. In addition, some nesting pairs gained the help of conspecifics more often than others, although this was not due to the proximity of neighbouring nests. Perhaps those pairs with the most obvious aggressive attacks drew the greatest response from conspecifics. This appears to be particularly true for those pairs that vocalized loudly during their attack (personal observations), and we suspect that vocally active pairs may be more likely to attract the attention of conspecifics, even at considerable distance (see Curio et al. 1978). Annual fluctuation in predation pressure appears to be the current, dominant influence on the breeding productivity of Sabine s gulls and varies considerably among years (Stenhouse 2003). Under these circumstances, pairs that defend their nests most aggressively in years of high predation should contribute the most to subsequent generations. Based on the extent of variation in anti-predatory behaviour observed among pairs in this study, some are clearly more aggressive than others and may be better at defending their nests from both mammalian and avian predators. Further examination of common behavioural features of highly successful Sabine s gull pairs, particularly in years of high predation, may provide insights into the long-term costs and benefits of anti-predator behaviour for this species. Acknowledgements We are grateful to the community of Coral Harbour for permission to work in East Bay, Southampton Island, Nunavut. We are deeply indebted to Rachel Bryant for her dedication and assistance in the field, and to all members of the East Bay field crews for their assistance, encouragement and companionship over the years of this study. We thank Greg Robertson for statistical advice, and two anonymous referees for their insightful comments on an earlier draft. Research was supported by the Northern Conservation Division of the Canadian Wildlife Service (HGG), NSERC (via a Discovery Grant to WAM), Memorial University of Newfoundland (via a graduate fellowship and the Hatcher Scholarship to IJS), and The Seabird Group (via a research grant to IJS). References Abraham K, Ankney CD (1986) Summer birds of East Bay, Southampton Island, Northwest Territories. Can Field Nat 100: Andersson M, Wiklund CG, Rundgren H (1980) Parental defense of offspring: a model and an example. Anim Behav 28: Barash DP (1975) Evolutionary aspects of parental behavior: distraction behavior of the alpine accentor. Wilson Bull 87: Bêty J, Gauthier G, Korpimäki E, Giroux JF (2002) Shared predators and indirect trophic interactions: lemming cycles and arctic-nesting geese. J Anim Ecol 71:88 98 Biermann G, Robertson RJ (1981) Residential reproductive value and parental investment. Anim Behav 31: Brown RGB, Blurton-Jones NG, Hussel DJT (1967) The breeding behaviour of Sabine s gulls (Xema sabini). Behaviour 28: Brunton DH (1990) The effect of nesting stage, sex, and type of predator on parental defense by killdeer (Charadrius vociferous): testing models of avian defense. Behav Ecol Sociobiol 26: Clark AB, Wilson DS (1981) Avian breeding adaptations: hatching asynchrony, brood reduction, and nest failure. Q Rev Biol 56: Clutton-Brock TH (1991) The evolution of parental care. Princeton University Press, Princeton Curio E (1975) The functional organization of antipredator behaviour in the pied flycatcher: a study of avian visual perception. Anim Behav 23:1 115 Curio E (1976) The ethology of predation. Springer, Berlin Heidelberg New York Curio E, Ernst V, Vieth W (1978) The adaptive significance of avian mobbing III. Z Tierpsychol 48: Day RH, Stenhouse IJ, Gilchrist HG (2001) Sabine s gull (Xema sabini). In: Poole A, Gill F (eds) The birds of North America, vol 593. The Birds of North America, Philadelphia Galeotti P, Tavecchia G, Bonetti A (2000) Parental defence in longeared owls (Asio otus): effects of breeding stage, parent sex and human persecution. J Avian Biol 31: Gochfeld M (1984) Antipredator behavior: aggressive and distraction displays of shorebirds. In: Burger J, Olla BL (eds) Shorebirds: breeding behavior and populations. Plenum, New York, pp Knight RL, Grout DJ, Temple SA (1987) Nest-defense behavior of the American crow in urban and rural areas. Condor 89: Krebs JR (1973) Behavioral aspects of predation. In: Bateson PPG, Klopfer PH (eds) Perspectives in ethology. Plenum, New York, pp Kruuk H (1964) Predators and anti-predator behavior of the blackheaded gull (Larus ridibundus L.). Behaviour 11(Suppl):1 129 Lack D (1954) The natural regulation of animal numbers. Oxford University Press, Oxford Mallory ML, McNicol DK, Walton RA, Wayland M (1998) Risk-taking by incubating common goldeneyes and hooded mergansers. Condor 100: Martin TE (1995) Avian life-history evolution in relation to nest sites, nest predation, and food. Ecol Monogr 65:

6 108 Montgomerie RD, Weatherhead PJ (1988) Risks and rewards of nest defence by parent birds. Q Rev Biol 63: Myers JP (1978) One deleterious effect of mobbing in the southern lapwing (Vanellus chilensis). Auk 95: Parmelee DF, Stephens HA, Schmidt RH (1967) Birds of southeastern Victoria Island and adjacent small islands. Natl Mus Can Bull 222:1 229 Patterson TL, Petrinovich L, James DK (1980) Reproductive value and appropriateness of response to predators by white-crowned Sparrows. Behav Ecol Sociobiol 7: Redondo T (1989) Avian nest defence: theoretical models and evidence. Behaviour 111: Redondo T, Carranza J (1989) Offspring reproductive value and nest defense in the magpie (Pica pica). Behav Ecol Sociobiol 25: Shields WM (1984) Barn swallow mobbing: self-defence, collateral kin defence, group defence or parental care? Anim Behav 32: Smith PA (2003) Factors affecting nest site selection and reproductive success of tundra nesting shorebirds. MSc Thesis, University of British Columbia, Victoria Spaans B, Blijleven HJ, Popov IU, Rykhlikova ME, Ebbinge BS (1998) Dark-bellied brent geese (Branta bernicla) forego breeding when Arctic foxes (Alopex lagopus) are present during nest initiation. Ardea 86:11 20 Stenhouse IJ (2003) The reproductive behaviour and ecology of Sabine s gulls (Xema sabini) in the eastern Canadian Arctic. PhD Thesis, Memorial University of Newfoundland, St. John s Stenhouse IJ, Gilchrist HG, Montevecchi WA (2001) Reproductive biology of Sabine s gull in the eastern Canadian Arctic. Condor 103: Stenhouse IJ, Gilchrist HG, Montevecchi WA (2004) Reproductive investment and parental roles in Sabine s gulls (Xema sabini). J Ethol 22:85 89 Walters JR (1982) Parental behavior in lapwings (Charadriidae) and its relationship with clutch sizes and mating systems. Evolution 36: Weatherhead PJ (1979) Do savannah sparrows commit the Concorde fallacy? Behav Ecol Sociobiol 5: