IN many high northern regions, optimum conditions for nesting exist for only a short time. Consequently most bird species have restricted breeding seasons and start nesting at about the same time each year (see Immelmann 1971, 1973). The timing of restrictive breeding involves two major problems: the phylogenetic aspect, which includes the de- velopment of species-specific and population-specific breeding seasons through natural selection (the "ultimate" factors), and the current physiological aspect, which involves the precise adjustment of the in- dividual to the phylogenetically determined reproductive season (the "proximate" factors). Each spring Least Auklets (Aethia pusilia), Crested Auklets (A. cristatella), and Parakeet Auklets (Cyclorrhynchus psittacula) return about mid-May to their nesting slopes on Sevuokuk Mountain, St. Law- rence Island, Alaska. Although the nesting slopes at this time are still covered with snow, with yearly variations in extent and depth, the auk- lets begin to visit and settle on these slopes each day. Here courtship activities take place while their previous pair bonds are renewed and strengthened; meanwhile the snow melts and the nesting crevices among the boulders immediately below become available to the birds. During studies of breeding ecology of these auklets (Sealy 1968, Sealy and Bgdard 1973) it became evident that snow on the nesting slopes in- fluenced the timing of egg-laying and productivity within the auklet population, particularly in 1967 when the snow melted early. Bgdard (1969a) demonstrated that food supply is an important ultimate factor controlling the general timing of these auklets' breeding on St. Lawrence Island. The present paper examines the role of snow cover and the sudden emergence of the nest sites as proximate factors. METHODS I spent the 1966 and 1967 nesting seasons (mid-May to mid-September) on Sevuokuk Mountain (Fig. 1) on the northwest cape of St. Lawrence Island, Bering Sea, studying the breeding ecology of auklets. I arrived in May before the auklets' return in 1966 and shortly after their return in 1967. The nesting and marine environ- ments of these auklets on and near St. Lawrence Island have been described else- where (Fay and Cade 1959; Sealy 1968, 1973; Bdard 1969a, 1969b; Sealy et al. 1971). Snow cover and its disappearance from the nesting slopes were recorded photograph- ically (Fig. 2) and estimated by the ratio of bare ground to snow-covered ground (Fig. 3) in the same eight randomly selected plots (five on the west slope and three July 1975] Auklet tgg-laying 529 530 SPEICER G. SEALY' [Auk, Vol. 92 I00= & 0 0 966 -  1967 < date auklets  first seen on nesting slopes 80- _o o o 20- I i 25 31 5 10 15 May Juno Fig. 3. Pattern of snowmelt on Sevuokuk Mountain in 1966 and 1967. Each point represents the mean value of estimated snow-free nesting habitat from each of 8 plots. Value of 0 indicates snow- and ice-free nesting crevices. on the northeast slope of Sevuokuk Mountain) in Least and Crested Auklet nesting habitat. In 1966-67 using "noose carpets" (see Berger and Hamerstrom 1962) I live- trapped, banded, and color-marked 17 Least Auklets and 6 Crested Auklets sitting on the snow in one of the plots on the northeast slope of Sevuokuk Mountain. Reproductive development of birds sitting on bare rocks (rock-sitters) and on the snow (snow-sitters) was studied in 121 specimens collected between 4 June and 11 July 1967, with about 10 individuals being taken from each habitat each week. The ovaries of all females (n = 56) were removed, the diameter of the largest follicle was measured to the nearest 0.1 mm, and the presence of a postovulatory follicle or shelled egg in the oviduct was recorded. All gonads were fixed for 24 h in Bouin's fixative and stored in 70% ethanol. Egg-laying dates of each species were determined by dissection of adult females and by visiting nest sites daily until the single eggs appeared. These visits were made in the late morning or early afternoon when the nest sites were empty and therefore the disturbance did not affect the timing of egg-laying. In 1967 I filled six Least Auklet nests I located in 1966 with snow for up to 6 days after the ice and snow had melted out of the other nest crevices, to prevent the pairs from entering them and laying their eggs. Egg-laying in seven other nests in the same portion of the slope was permitted to proceed normally. Fig. 4. Crested and Least AukleB sitting on the snow along the brow of the northeast slope of Sevuokuk Mountain, 25 June 1967. Note the nonrandom dis- persion of birds on the snow. RESULTS Snow and its disappearance Jrom the nesting slopes.--The extent and persistence of snow encountered by the returning and settling auklets on Sevuokuk Mountain differed considerably in 1966 and 1967 (Figs. 2 and 3). In 1966, a late season, snow covered about 95% of the nesting habitat when the first auklets settled on 24 May. In 1967, an early season, it covered only about 60% when the first Least Auklets settled on 20 May. The mean temperature at Gambell (3 km west of Sevuokuk Mountain) was -1.0 ø C from 21 May to 15 June 1966 and 7.0 ø C from 26 May to 15 June 1967. Thus most of the snow, except that along the brow of the mountain (Fig. 2C), melted earlier in 1967. Despite the earlier overall melting of the snow on the lower reaches of the slope in 1967, the total snow cover when the first eggs were laid in both years was about the same. The snow melted at different rates on different parts of the slope depending upon the exposure and elevation, leaving some portions covered with snow (Figs. 2C and 4) while other parts were bare. In both years, snow persisted along the brow of the northeast slope of Sevuokuk Mountain until about 12 July despite the difference in the general phenology of snow melt those years. Fig. 4 shows the snow on the brow of the mountain covering much auklet nesting habitat. ,q!' , ..'  ß ..-..'-  _ Fig. 5. Least and Crested Auklets settled on the roc (the rock-sitters) and snow (the snow-sittem) on the northeast slope of Sevuokuk Mountain, 25 June 1967. Note fidelie of birds to snow-covered ro&s despite adjacent snow-free rock. Timing o! egg-laying.--In 1966 egg-laying in nests studied on the lower reaches of the slope spanned the period 24 June to 7 July. Those auklets faithful to nest sites covered by snow along the brow have access to snow-free nesting habitat a few days later and therefore are able to lay their eggs at that time. Thus egg-laying was essentially synchronized within the entire population in 1966. In 1967 egg-laying spanned 13 25 June for those nests studied on the lower snow-free portions of the slope that were completely free of snow by 8 June (Fig. 3). In 1967 the snow persisting on the brow prevented birds there from reaching their nest sites. On 22 June Least and Crested Auklets began to lay their eggs on this snow. These eggs were sub- sequently destroyed either by rolling down the slope and breaking on the exposed rocks below or were taken by Herring Gulls (Larus argentatus). No attempt was made by the birds to incubate eggs that were laid on the snow. These individuals did not renest. During the pre-egg stage, the auklet population (particularly Crested and Least Auklets) is divided into two groups, those settling and court- ing on the rocks and actually entering nest sites and those courting on the snow and unable to enter the nest sites (Fig. 5). The dispersion of auklets settling on the snow-covered slope was not random (Fig. 4). These birds, unclassified as to experienced or inexperienced breeders, were apparently familiar with the nesting slope, and their acute sense of TABLE 1 BANDING AND SUBSEQUENT OBSERVATIONS AND/OR RECAPTURES OF LEAST AND CRESTED AUXZETS N 1966 ANn 1967  Species Date banded Date observed Activity Least Auklet 15 June 1966 5 July 1966 Incubating egg " " 15 June 1966 9 July 1966 Incubating egg " " 19 June 1966 3 August 1966 Carrying food to chick " " 19 June 1966 10 August 1966 Carrying food to chick Crested Auklet 15 June 1966 4 August 1966 Carrying food to chick " " 12 June 1967 15 July 1967 Incubating egg ' ' 12 June 1967 3 August 1967 Incubating egg " " 12 June 1967 21 August 1967 Carrying food to chick  Each adult was live-trapped on top of the snow in a pbt at about the 20-m contour on the northeast slope of Sevuokuk Mountain and seen later at the same spot. location was borne out by the banded and color-marked birds captured on the snow that subsequently occupied the nest sites below (Table 1). The auklets exhibit mate retention and nest site tenacity (Sealy 1968). General timing of egg-laying in auklets on Sevuokuk Mountain in 1966 and 1967 is presented in Fig. 6. Egg-laying started 10 days earlier in 1967 in all species. The mean laying date in seven Least Auklet nests in a sector of nesting habitat with early snow melt was 17 June 1967. The six nests in the same sector of the slope that were plugged with snow for 6 additional days had a mean laying date of 28 June. The interval be- tween absence of snow in places where it melted early to the mean egg- laying date was 20 days in 1966 (n = 36) and 10 days in 1967 (n: 28). Where the snow cover was prolonged this interval averaged 4 days (n = 3) in the Parakeet Auklet, 3.1 days (n = 4) in the Crested Auklet, and 3.6 days (n: 9) in the Least Auklet. Gonada! development.--Gonadal development in male snow-sitting and rock-sitting Least Auklets was examined elsewhere (Sealy 1968). There was no relationship between testicular development, as indicated by the weight of the left testes, and whether or not the birds were able to enter snow-free nest sites. Follicular maturation in adult female Least Auklets from rock and snow areas is summarized in Fig. 7; the Fig. 6. Crested Auklet 1967 Least Auklet Parakeet Auklet  June July Egg-laying in the Parakeet, Crested, and Least Auklets on Sevuokuk Moun- tain, 1966 and 1967. Shaded blocks show egg-laying dates obtained by direct obser- vations in nests, open blocks dates obtained by dissection of fully shelled eggs from the oviduct. more rapid (earlier) maturation of ova of females taken on sites that became free of snow earlier is evident. No atretic follicles were found in any of the females collected. DISCUSSION The effects of prolonged snow cover in delaying egg-laying or even preventing nesting in certain years has been amply documented. Barry (1962) demonstrated the effects of late seasons on Brant (Branta ber- nicla) where fewer young are produced in some seasons than in others; in fact no young are produced in some years. Although Brant return to their breeding grounds on the same date each year, they must wait for the snow to melt and uncover nesting habitat. If the snow persists, energy reserves are depleted and follicular astresia occurs, which Barry I postovulatory ß o o  follicles ß ß : 25 2 June July Fig. 7. Follicular maturation of snow-sitting and rock-sitting Least Auklets as indicated by the diameter of the largest follicle. The solid line and circles indicate rock-sitters; the dashed line and open circles indicate snow-sitters. The horizontal lines indicate mean and vertical lines the extremes; numbers beneath extremes are the sample sizes. argues is an effort to retain nourishment as their food is unavailable under the snow. If the snow persists too long, the entire clutch is re- sorbed. Besides providing nourishment, the follicular atresia prevents extremely late nesting, which would jeopardize the entire population by allowing too little time to complete nesting activities and molt and re- gain the power of flight before freezeup. Lack (1933) found that Arctic Terns (Sterna paradisaea) in northern Norway laid their eggs on different dates in different places as the nest- ing spots successively became available after the snow melted. He sug- gested that the delayed availability of nest sites coupled with the in- ability of the gonads to remain in breeding condition may combine to make reproduction impossible in a particular year. Holmes (1966) noted that the timing of arrival of Dunlins (Calidris alpina) and most other arctic nesting shorebirds differs each season and is correlated with year-to-year differences in climatic events that either prolong or accelerate snowmelt, thereby influencing the availability of snow-free tundra. Once the Dunlins return, vigorous activity ensues with territories being established and defended on the patches of tundra as they become available. The situation in the auklets differs from those described above. In auklets the final production of eggs and egg-laying is at least partially controlled by the proximate environmental stimulus of a snow4ree nest site. Rock-sitters presented with this stimulus sooner than snow-sitters will lay earlier. The auklets appear unable to halt their follicular mat- uration in response to prolonged unavailability of nesting habitat, and lay their eggs on the snow. In geese and shorebirds, the food supplies and the nest sites lie beneath the snow. The longer snowmelt is prolonged, the longer the birds must do without food. Hence the follicles cannot develop or, if development has progressed, atresia results. The auklets make regular trips to sea to feed and therefore have the energy necessary for egg production readily available. Manniche (1911) attributed the nonbreeding of aquatic birds in a late season in Greenland to the lack of food because ice remained in the inlet all summer. Arctic Terns laid eggs on the snow, which suggests that their gonads developed as usual, and that they probably obtained food by fly- ing to open water. Parakeet Auklets return in spring before Crested and Least Auklets, which lay their eggs earlier (see Fig. 6). Most of the 2000 breeding pairs of Parakeet Auklets (Bdard 1969b) on Sevuokuk Mountain use crevices among splintered parent rock (Sealy and Bdard 1973), which is largely distributed along the brow of the mountain where the greatest accumula- tion of snow occurs (Fig. 2D) and melts last. Parakeet Auklets, by nest- ing later, are able to use these nest sites that are unavailable 7-10 days earlier when the Crested and Least Auklets have begun egg-laying. In 1967 the Parakeet Auklets nested successfully along the brow of the mountain after the Crested and Least Auklets had laid their eggs on the snow. ACKNOWLEDGMENTS The work on auklet breeding ecology, conducted from the University of British Columbia, was funded by National Research Council of Canada grants to M.D. F. Udvardy. Jean Bdard aided in the fieldwork and has offered much constructive criticism throughout all phases of the study. I wish to thank many residents of Gainbell for help in the field. SUMMARY Egg-laying in relation to snowmelt in a population of Crested, Least, and Parakeet Auklets on Sevuokuk Mountain, St. Lawrence Island, Alaska, is examined. In 1966 the snow melted late and egg-laying was synchronized throughout the population. In 1967 individuals faithful to nesting habitat from which the snow melted early laid their eggs earlier than individuals faithful to portions of the slope where snowmelt was prolonged. Crested and Least Auklets faithful to nesting habitat along the brow of the mountain where the snow remained until mid-July laid their eggs on the snow, which caused nesting failure in about 5% of the population. Snow cover as a proximate factor influencing timing of egg-laying on an individual basis is discussed. LITERATIJRE CITED BARRY T. W. 1962. Effect of late seasons on Atlantic Brant reproduction. J. Wildl. Mgmt. 26: 19-26. BDARD J. 1969a. Feeding of the Least, Crested, and Parakeet Auklets around St. Lawrence Island, Alaska. Canadian J. Zool. 47: 1025-1050. BDARD, J. 1969b. The nesting of Least, Crested, and Parakeet Auklets on St. Law- rence Island, Alaska. Condor 71: 386-398. BERGER D. G., AND F. HA2VIERSTRO2VI. 1962. Protecting a trapping station from rap- tor predation. J. WiIdl. Mgmt. 26: 203-206. FAY, F. H., AND T. J. CADE. 1959. An ecological analysis of the avifauna of St. Lawrence Island, Alaska. Univ. California Publ. Zool. 63: '73-153. HOLMES, R. T. 1966. Breeding ecology and annual cycle adaptations of the Red- backed Sandpiper (Calidris alpina) in northern Alaska. Condor 68: 3-46. IM3iELMANN, K. 1971. Ecological aspects of periodic reproduction. pp. 341-389 in Avian biology, vol. 1 (D. S. Farner and J. R. King, Eds.). New York, Academic Press. IELANN, K. 1973. Role of the environment in reproduction as a source of "pre- dictive" information. Pp. 121-147 in Breeding biology of birds (D. S. Farner, Ed.). Washington, D.C., Natl. Acad. Sci. LACK, D. 1933. Nesting conditions as a factor controlling breeding time in birds. Proc. Zool. Soc. London 1933: 231-237. MANNIC{E, A. L. V. 1911. Danmarck Ekspeditionen til Gron]ands Nordostkyst 1906-1908. Bind v. nr. 1. SEALY, S. G. 1968. A comparative study of breeding ecology and timing in plank- ton-feeding aIcids (Cyclorrhynchus and Aethia spp.) on St. Lawrence Island, Alaska. Unpublished M. Sc. thesis, Vancouver, Univ. British Columbia. SEALY, S. G. 1973. Breeding biology of the Horned Puffin on St. Lawrence Island, Bering Sea, with zoogeographical notes on the North Pacific puffins. Pacific Sci. 27: 99-119. SEALy, S. G., AND J. BDARD. 1973. Breeding biology of the Parakeet Auklet (Cyclorrhynchus psittacula) on St. Lawrence Island, Alaska. Astarte 6: 59-68. SEALY, S. G., F. H. FAy, J. BDARD, AND M.D. F. UDVARDY. 1971. New records and zoogeographical notes on the birds of St. Lawrence Island, Bering Sea. Condor 73: 322-336. Department o/ Zoology, University o/ Manitoba, Winnipeg, Manitoba, Canada. Accepted 2 July 1974.