Effects of Human Disturbance on Breeding Least and Crested Auklets at St. Lawrence Island, Alaska

EFFECTS OF HUMAN DISTURBANCE ON BREEDING LEAST AND CRESTED AUKLETS AT ST. LAWRENCE ISLAND, ALASKA

JOHN F. PIATT, BAY D. ROBERTS,WAYNE W. LIDSTER, JOHN L. WELLS,AND SCOTTA. HATCH AlaskaFish and WildlifeResearch Center, U.S. Fish and WildlifeService, 1011 East Tudor Road, Anchorage,Alaska 99503 USA

A13STRACT.--Westudied breeding success,chick growth, and diets of Least (Aethia pusilla) and Crested (A. cristatella)auklets on St. Lawrence Island, Alaska, in summer 1987. Least Auklets had higher breeding successon control plots (50-66%) than on disturbedplots (36%). Crested Auklets had a breeding successof 42% on disturbed plots. Predation by microfine rodents and weather accountedfor most natural chick mortality. Least Auklet chicks grew at a maximum rate of 4.9 g/day, and CrestedAuklet chicksat 12.8 g/day. LeastAuklet chicks were fed mostly copepods(Neocalanus plumchrus), whereas CrestedAuklet chicks were fed Thysanoessaeuphausiids. Received 13 December1988, accepted 30 November1989.

LEASTAuklets (Aethia pusilla) and Crested bance. Because fewer Crested Auklets were Auklets (A. cristatella)are the most abundant found, they were studied at only one (high) planktivorousseabirds in the North Pacific,and level of disturbance. We also measured chick breeding populationscoexist at many colonies growth and collected food samples to deter- in the Bering Sea(Sowls et al. 1978).Both species mine the compositionof auklet diets. lay their eggsin natural crevicesformed in the rubble of talus slopes.There have been few de- STUDY AREA AND METHODS tailed studiesof auklet breeding biology, probably becauseauklet colonies are remote, and The study was conductedfrom 28 May to 2 Sep- tember 1987at KongkokBay (63ø24'N, 171ø49'W) near breedingsites are difficult to observe.We stud- the southwesterncape of St. LawrenceIsland (Fig. 1). ied Least and Crested auklets on St. Lawrence The avifauna, climate, and seabird breeding habitats Island in summer 1987 to learn more about their of St. Lawrence Island have been describe•, elsewhere basicbreeding biology and to evaluatemethods (Fay and Cade 1959,B6dard 1969a, Sealy 1975). used to measurebreeding success. We establishedsixteen 200-m2 auklet study plots After several years of study on St. Lawrence on the southeast-facing(inland) talus slopeof Owalit Island, B6dard(1969a, b) describedthe breeding Mountain (Fig. 1). At least 20,000 auklets (2:1 Least/ habitat and diets of auklets, and Sealy (1968) Crested)breed on thoseslopes (Searing 1977,Piatt et reported on chick growth and development al. 1990).All 16 plotswere inspected 3-4 timesover (Sealy1973, 1981), factors that influencebreed- 13 daysduring the early egg-layingperiod to locate eggs,check the statusof previouslylocated eggs, and ing chronology(Sealy 1975,1984) and survival mark breeding sites. Some eggs were located after of eggs and chicks (Sealy 1982). Elsewhere, minor excavation,but mostwere locatedby a line-of- Knudtson and Byrd (1982) estimatedLeast and sight searchwith strong flashlights.All inspections Crestedauklet breeding successat Buldir Island were conducted during mid-day (1400-1800) when in the western Aleutians, and Roby and Brink attendanceby adults on the surface of talus slopes (1986) studied Least Auklet breeding biology was minimal (Piatt et al. 1990). LeastAuklet breeding during two summersat St. GeorgeIsland in the siteson one setof eight "Intensive"plots were checked Pribilofs. every day weather permitted through the hatching All previous investigators have noted the period (27 July-13 August),every 2-4 daysup to early negativeeffect of human disturbanceon auklet fledging (25 August), and every day possiblethere- after until 2 September.We recorded weights and breedingsuccess, but this factorhas never been wing-chord lengths of LeastAuklet chickson Inten- quantified. We attempted to measurethe effect sive plots only. To reducedisturbance, Least Auklet of human disturbance on auklet breeding suc- breedingsites in the other setof eight "Control" plots cessand to identify natural sourcesof egg and on Owalit Mountain were inspectedonly four times chick mortality. Selected Least Auklet study after the initial searchfor eggs (but seebelow). Least plots were subjectedto three levels of distur- Auklet chickswere not handled on Control plots. The

342 The Auk 107: 342-350. April 1990 April1990] BreedingBiology ofAuklets 343 final check of breeding sites on all plots (including thosebelow) was especiallythorough, and excavation was sometimes used to determine the status of chicks. On 23-25 July, in another set of eight control plots on the north-facingtalus slope of KongkokBasin (Fig. 1), we searchedLeast Auklet breeding sitesfor eggs. These plots were then checkedtwice before chicks fledged to determine the statusof chicks. Observer disturbancein Kongkokwas minimal comparedwith Owalit, where observerswere on the slopesdaily to checkplots and weigh chicks.Least Auklet densities were higher at Kongkok,and total numberswere far greater(ca. 1 million birds;Searing 1977,Platt et al. Kongkok 1990) than at Owalit. Becausewe located fewer Crest- Bay ed Auklet breedingsites, we studiedthem on all 16 Owalit plots with the same level of intensity as the "Intensive" LeastAuklet study plots. CrestedAuklet chicks were weighed and measuredin all 16 study plots. Although we tried to avoid disturbing Least Aukletsin their breedingcrevices, our work on Crest- ed Auklets increased disturbance to Least Auklets on the eight "Control" Owalit plots. It is difficult to monitor the breeding statusof Least and Crested auklets because neither builds a nest, and eggsor chicksare never in plain view. Detailedwrit- Fig. 1. Locationsof auklet study plots on Owalit ten instructionswith referenceto orangespray-paint- Mountain and in Kongkok Basin. ed marks and numbers on rocks were used to locate breeding sites again. When eggs and chicks disappeared between site-checks,their fate was difficult to by native huntersin Gambellin May and from birds assess.Many reappearedon subsequentchecks, but shot near Kongkok Bay for diet studiesin June and those that never were locatedagain may have been July. To obtain samplesof chickmeals, we mist-netted taken by predators,moved permanently, or died of 56 Least Auklets and 5 Crested Auklets on Owalit other causes. Mountain between 7 August and 1 September,and Estimatesof hatchingsuccess were biasedbecause we retrieved the contentsof food regurgitated from eggs found at a late stage of incubation were more gular pouches.Food samplesfrom 18 Least and 49 likely to hatch than those found earlier (Mayfield Crestedauklets were obtainedfrom gular pouchesof 1975, Johnson 1979). To correct for that effect, we adultsshot near KongkokBay on 30 August.Collected calculatedhatching success(Mayfield 1975) as fol- meals were preserved in 5% formalin and later sub- lows: (1) we divided the total number of egg-days sampledand sortedto identify prey items. Weights observed by the total number of eggs lost, which of the various taxa identified were extrapolatedfrom yielded the egg-survivalrate per day; (2) we multi- numbersoccurring in subsamples(Springer and Ro- plied that rateby itself n times,where n is the average seneau 1985). number of incubation days, which yielded the egg- survival rate per season;and (3) we multiplied that rate by the proportion of eggssurviving to hatching RESULTS age (i.e. eggsthat actually hatched). For describingand contrastingchick growth rates, Breedingchronology.--We first observedLeast we followed Sealy (1973) and Gaston(1985) to deter- and CrestedAuklets near shore on 15 May, and mine the "instantaneous(maximum) growth rate" at at colonies near Gambell on 18 May. We in- the inflection point of the growth curve. We deter- spected breeding habitat at Owalit and Kong- mined maximumgrowth ratesfrom our own and pub- kok throughoutJune, and first eggswere found lisheddata by fitting the steepestpossible tangent to on 27 June at Owalit (4 Least and 1 Crested). compositegrowth curves(Gaston 1985). We calculat- Becausewe were rarely sure we had found an ed median chick weights for comparativepurposes as the midweight between hatching and asymptotic egg on the first day of incubation, only a few records were used to estimate auklet incubation weights (Gaston 1985). We also calculated t•0_90,the time (days)it took for chicksto grow from 10 to 90% periods(Table 1). Combinedwith hatching data of asymptoticweight (Ricklefs 1967). (Table 1, Fig. 2), we suggestthat median laying Adult body weightswere obtainedfrom birds shot occurred around 1-2 July. Only 4 of 369 eggs 344 Px^vrœt ^r. [Auk,Vol. 107

T^13Lœ1. Breedingchronology of Leastand Crestedauklets on St. LawrenceIsland in 1987.

Least Auklet Crested Auklet n œ_+ SE Range n i + SE Range Incubation period (days) 31 30.1 _+ 0.51 25-39 20 33.8 + 0.63 29-40 Chick-rearingperiod (days) 34 29.3 + 0.37 25-33 6 33.2 + 0.05 27-36 Median laying date -- 1 Julya -- 2 Julya Median hatching 76 30 July 75 4 August Median fledging 38 28 August -- 6 September' Extrapolatedfrom median hatchingdates using mean incubationand chick-rearingperiods.

(1.1%)were suspectedto havebeen re-lays (Ta- or abandonmentof eggson the Intensive study ble 2). plots. A higher proportion of eggsdisappeared Hatching was highly synchronous(Fig. 2). for unknown reasonson the lessdisturbed plots. Approximately80% of Leastand Crestedauklet Most CrestedAuklet egg losseswere from non- eggshatched over 7- and 10-day periods,re- hatching or abandonment. A smaller propor- spectively.Fledging of LeastAuklets was also tion was lost to vole predation (asevidenced by highly synchronous(ca. 80% over 7 days).The teeth marks and punctures of the shell, Sealy longer incubation and fledging periods of 1982). Estimatesof hatching successusing the Crested Auklets produced median fledging Mayfield correctionwere ca. 5-10% lower than around6 September.Fledging began a few days estimates made from raw data. before our departureon 2 September. The losses of Least Auklet chicks from unex- Breedingsuccess.--Least Auklet hatching suc- plained disappearances,mortality, and preda- cesswas higher at Kongkok than at the Inten- tion were higher on heavily disturbed(Inten- sive study plots on Owalit Mountain (X2 = 5.06, sive) plots than on the moderately disturbed df = 1, P < 0.05; other comparisonswere not (Control) plots on Owalit and the lightly dis- significant,Table 2). The biggestdifference be- turbed plots at Kongkok. This produced pro- tween siteswas the high rate of nonhatching gressivelyhigher fledging successfrom Inten-

T^•3Lœ2. Breedingsuccess of Leastand Crestedauklets on St. LawrenceIsland in 1987.Values in parentheses are percentagesof hatched eggs.

Least Auklet Crested Auklet Intensive Control Kongkok Intensive n % n % n % n %

Visits 28 7 3 28 Total eggsfound 109 100.0 101 100.0 56 100.0 103 100.0 Egg lossesdue to: Non-hatch/abandoned 24 22.0 7 6.9 3 5.4 14 13.6 Unknown 4 3.7 13 7.9 4 7.1 4 3.9 Breaking 2 1.8 4 4.0 1 1.8 5 4.9 Vole predation 3 2.8 1 1.0 0 0.0 2 1.9 Eggsre-laid a 2 -- 1 Hatching success(A) 76 69.7 76 75.3 48 85.7 78 75.7 (Mayfield estimatelB]) 66.4 69.8 80.5 65.6 Chick losses due to: Disappearance 16 (21.1) 16 (21.1) 6 (12.5) 18 (23.1) Death 12 (15.8) 5 (6.8) 3 (6.3) 6 (7.7) Vole predation 7 (9.2) 1 (1.3) 0 (0.0) 2 (1.9) Fledging success(C) 41 (53.9) 54 (71.1) 39 (81.3) 50 (64.1) BreedingSuccess (A x C) 37.6 53.5 69.7 48.5 (Mayfield estimate,B x C) 35.8 49.6 65.5 42.0 Not included in subsequentcalculations. April 1990] BreedingBiology of Auklets 345

HATCHIN' *G .•'•"•"• •CRESTED - ' FLEDGING/' ' /' LEA8Ti..• (n.76) LEASTi..•

ti ti ISAPPEARED .... I ....,, .... ,, .... ,, .... I JULY AUGUST SEPT. i '!!' CHICK r'lLeastAuklet Fig. 2. Breedingphenology of Leastand Crested • 20.""111!,l'l'lL&2 ss E s B CrestedAuklet auklets. 0 ILl

U. sive to Control plots (X2 = 5.32, df = 1, P < 0.05), o and from Control to Kongkok plots (X2 = 3.91, 0-3 4-7 8-11 12-15 16-19 20-23 24+ df = 1, P < 0.05; Table 2). However, chick losses CHICK AGE did not vary consistentlywith levels of disturbance.For example, similar proportions of chicks Fig. 3. Sourcesand frequencyof aukletchick mor- disappearedon Control and Intensive plots(but tality with age (days) of chicks. see below), and similar proportions of chicks were found dead on Control and Kongkok plots. CrestedAuklet chick mortality rates were sim- (Fig. 3). The proportion of chick lossesascribed ilar to those of Least Auklets on Owalit, with to "disappearance"increased with age up to 22 an overall fledging successthat was interme- days.Some of thosechicks may have wandered diate between Least Auklet success on Intensive from their breedingsites and starved,been eat- and Control plots. en, or survived undetected. Most chicks were Vole predationwas highest on Intensivestudy locatedagain on repeatedinspections. The few plots (Table 2). Voles did not usually kill chicks that couldnot be locatedafter survivingto >23 outright, but removed flesh from the lower back days old were assumedto have fledged in our or flank regions. Death or disappearanceusu- calculationsof breeding success. ally followed vole attacksby a few days.Thus, Overall breeding successof LeastAuklets var- many of the chicks that "disappeared"on the ied significantlyamong study plots (X2 = 16.1, Control plots were probably killed by voles,but df = 2, P < 0.001), and reflected the different vanishedbetween inspections. Presumably they hatchingand fledgingrates among plots (above, were removed by an adult or scavenger,buried, and Table2). Breedingsuccess was significantly or washed away. Vole predation was most in- tense on chicks 5-20 days old (Fig. 3). Chicks were protectedby attending adults in the first few days of life (Fig. 4), and were able to fend for themselvesat a later age (>20 days). Most chick mortality occurredwithin a week of hatching (Fig. 3). Overall, 38% (Least Auklet) and 30% (Crested Auklet) of all chicks that died or disappeared were handled by us for mea- surements.Many dead chicks were found sat- urated with water, and it is likely that rain and hypothermia contributed to mortality. More than twice as much rain fell between 2 and 17 August (5.1 cm) than in the previous months of June and July (2.1 cm), and much of this fell CHICK AGE (POST-HATCHING) over two periods (2-3 and 9-12 August) im- Fig. 4. Frequency of attendanceof adult auklets mediately following chick hatching. with chicksof varying age (days).Sample sizes in- No chicksolder than 22 dayswere found dead dicatednext to eachdata point. 346 I•IATTET At. [Auk,Vol. 107

LEAST AUKLETS CRESTED AUKLETS

5 110' ' '115.... 20.... 215.... 310' '

Z 4O-

AGE (DAYS POST-HATCHING) AGE (DAYS POST-HATCHING) Fig. 5. Increasein body massand wing lengths + SE,and samplesizes) of aukletchicks with age.

higheron Kongkokplots than on Controlplots than 22 days post-hatchwas similar (98%) to (X2 = 3.9, df = 1, P < 0.05), which in turn ex- adults in mid-July (Table 3). Crested Auklet ceededsuccess on Intensive plots (X2 = 5.32, df chick wing lengths increased after an initial = 1, P < 0.05). CrestedAuklets had a higher 5-daylag to a maximumof 85%of adultlengths. breeding successthan Least Auklets on Inten- Diets.--Least Auklet chick meals sampled siveplots, but the differencewas not significant between 7 August and 1 Septemberwere dom- (X2 = 2.58,df = 1, P > 0.05).The Mayfieldcor- inated by calanoid copepods,particularly Neo- rection reducedestimates of auklet breeding calanusplurnchrus (Table 5). A variety of other success 2-6% below estimates from raw data. prey were taken in much smaller quantities. Chickgrowth.--The mean massof Least Auk- These included Parathernistoamphipods, Thy- let chicksincreased steadily for ca.21 days post- sanoessaeuphausiids, and Pandalusshrimp zoeae. hatch, reachedan asymptoteat 27 days,and MostCrested Auklets were collected on oneday declined thereafter (Fig. 5). The mean massof (30 August),when mealsconsisted largely of chicksolder than 22 days,or at fledging,was Thysanoessaeuphausiids. Copepods and am- close to that of adults in late June (Table 3). phipods contributedonly slightly to the diets Maximumchick growth rates,midweights, and of Crested Auklets. developmenttimes appeared to be "normal" for St. Lawrence Island (Table 4), but statistical DISCUSSION evaluation based on variance of individual growth rateswas not possiblebecause only Chronology.--Breedingof aukletson St. Law- compositedata were available from previous renceIsland followed the samechronology as studies.After an initial lag as primariesdevel- in most previous years (1964-1967, 1976, 1981; oped,wing lengthsincreased steadily to 90%of B&dard1969b, Sealy 1968, Searing 1977, Rose- adult wing lengthsat 31 dayspost-hatch. neauet al. 1985),which is 7-10 dayslater than CrestedAuklet chickmass increased steadily at St. Matthew Island (60ø27'N;Springer et al. up to 23 daysand reachedan asymptoteat 29 1985a,b), 2-3 weekslater than at St. George days post-hatch.Maximum growth rates and Islandin the Pribilofs(56ø55'N; Roby and Brink midweightswere slightly higher than in pre- 1986), and 3-4 weeks later than at Buldir Island viousyears (Table 4). No fledgingweights were in the western Aleutians (52ø21'N; Knudtson obtained,but the averagemass of chicksolder and Byrd 1982).The differencesare probably April 1990] BreedingBiology of Auklets 347

TABI•E3. Massand wing-chordlength of adult aukletsand chicksfrom St. LawrenceIsland, 1987.

Least Auklet Crested Auklet n œñ SE Range n œñ SE Range Adult mass(g) May 18-20 18 283 ñ 5.6 250-315 June 6 18 86.6 _+ 1.5 77-101 6 272 ñ 6.0 256-298 June 30 8 82.0 _+ 1.8 76-90 34 260 ñ 2.4 227-285 July 18 22 265 ñ 3.2 228-294 Chick mass(g) Asymptote 6 90.8 ñ 3.1 81-100 3 269 ñ 6.1 258-279 Fledgling 12 82.2 ñ 2.8 72-99 22+ days 46 85.0 ñ 1.3 64-105 38 261 ñ 2.8 221-290 Wing length (ram) Adult 18 97.6 ñ 0.5 95-102 20 143 ñ 0.9 137-153 Fledgling 12 87.9 ñ 1.1 83-95 1 122(max)

related to the timing of prey availability in each not checkedfrequently enough to pinpoint lay- region (Sealy 1968,Birkhead and Harris 1985), ing and hatchingdates (Sealy 1984; V. Byrdpers. although cold temperaturesand snow may also comm.). Chick-rearing periods were similar to delay breeding at northern colonies in some thosereported at St. Lawrence Island and else- years (Sealy 1975). where (Sealy 1968,1984; Roby and Brink 1986). Least and Crested auklets are synchronized Breedingsuccess.--All previous investigators in laying eggs,but hatchingis out of phaseby havesuggested that their estimatesof hatching ca.5 days,and fledgingby ca.9 days.Like Sealy successfor auklet eggs were biased upwards (1968, 1981), we found that Least Auklets in- becauseof early egg lossesto predation or other cubatedeggs for ca.31 daysand CrestedAuklets causes(Sealy 1968, Searing 1977, Knudtson and for 34 days. Incubation times may vary consid- Byrd 1982, Roby and Brink 1986). We concur erably depending on the degree of disturbance and believe that hatchingsuccess was probably to incubating birds (Sealy 1968, 1984), but in- overestimatedby 5-10%, and breeding success cubationperiods reported for LeastAuklets (35- by 2-6%. 36 days) and Crested Auklets (40-41 days) at Our measure of egg-hatching successwas Buldir Island (Knudtson and Byrd 1982) were biaseddownwards by human disturbancethat probably inflated becausebreeding siteswere led to high ratesof nonhatchingand abandon-

TABLE4. Aukletgrowth at St.Lawrence Island (SLI) and other colonies in the BeringSea (St. Matthew Island [SMI] and the Pribilofs [PRI]).

Area Year MGR (g/day)a Mid-weightb t•0_90(days) c Sources Least Auklet SLI 1987 4.9 51.4 20 This study SLI 1981 4.7 -- -- Roseneau et al. 1985 SLI 1976 4.4 50.5 19 Searing1977 SLI 1966-1967 5.3 49.7 18 Sealy 1968 PRI 1981-1982 5.7 53.9 17 Roby and Brink 1986 SMI 1982 4.9 -- -- Springeret al. 1985a SMI 1983 3.9 -- -- Springeret al. 1985b Crested Auklet SLI 1987 12.8 147.8 17 This study SLI 1976 11.1 146.0 20 Searing 1977 SLI 1966-1967 12.5 144.9 22 Sealy 1968

ßMGR is maximum growth rate. bMid-weight is the averageweight of chicksmid-way through growth to asymptoticweight. ßt•e•o is the time (days)for growth between10 and 90%of asymptoticweight. 348 PIATTET AL. [Auk, Vol. 107

TABLE5. Compositionof Least (n = 74) and Crested (n = 54) auklet chick meals collectedat St. Lawrence Island between 7 August and 1 September,1987.

Least Auklet Crested Auklet

Prey species n % No. % Wt. n % No. % Wt. Neocalanus cristatus 725 1.4 2.4 280 3.2 0.6 Neocalanusplurnchrus 44,417 82.9 87.0 909 10.4 1.2 Calanus rnarshallae 5,646 10.5 5.5 285 3.3 0.2 Calanoid copepod 690 1.3 1.0 20 0.2 0.1 Metridia pacifica 40 0.1 0.1 0 0 0 Eucalanusbungii 286 1.5 0.3 0 0 0 Euchaetaelongata 2 0.1 0.1 0 0 0 Apherusaglacialis 1 0.1 0.1 0 0 0 Anisogamrnaruspugettensis 48 0.1 0.1 0 0 0 Parathernisto libellula 135 0.3 1.0 9 0.1 0.1 Parathernistopacifica 571 1.1 0.7 24 0.3 0.1 Parathernistospp. 28 0.1 0.1 27 0.3 0.1 Thysanoessaspp. adult 36 0.1 0.8 7,048 80.3 97.8 Thysanoessaspp. furcilia 413 0.8 0.1 165 1.9 0.1 Pandalusspp. zoea 525 1.0 1.0 13 0.1 0.1 Paguridaespp. zoea 23 0.1 0.1 0 0 0 Lirnacina helicina 3 0.1 0.1 0 0 0 Squid 1 0.1 0.1 0 0 0 Fish 1 0.1 0.1 0 0 0

ment. Knudtson and Byrd (1982) and Searing methodology and levels of disturbance,how- (1977) also reported high rates (29% and 24%, ever, it is impossibleto saywhether thesevalues respectively)of nonhatching and abandonment reflect differences in the biology of Least Auk- for LeastAuklets. Similarly, CrestedAuklet egg lets among areasor years. The sameis true for lossesfrom nonhatchingand abandonmentob- Crested Auklets, where breeding successwas served by Searing (50%), Knudtson and Byrd lower on disturbed plots at St. Lawrence Island (22%), and us suggestthat human disturbance (42%) than at Buldir Island (51%). interferes with incubation activities and re- Other factors may also bias estimates of duces hatching success.High egg lossesfrom breeding success.For example, optimal breed- nonhatchingand abandonmenthave alsobeen ing sitesfor auklets (especiallyCrested), are in reported for other crevice- or burrow-nesting deep, inaccessiblecrevices, where the effectsof alcids. These include the ParakeetAuklet (Cy- human disturbance,predation, and weather are clorrhynchuspsittacula; 20%, Sealy and B•dard lesspronounced than near the surface(B•dard 1973), Cassin'sAuklet (Ptychoramphusaleuticus; 1969a,Sealy 1968).Also, auklets occurin much 23-30%, Vermeer and Lemon 1986), Xantus' higher abundanceand density in Kongkok Ba- Murrelet (Synthliboramphushypoleucus; 14%, sin than on Owalit Mountain, and breeding suc- Murray et al. 1983), and Ancient Murrelet (S. cessmay be correlatedpositively with bird den- antiquus;Gaston et al. 1988). Auklet fledging sity (Birkhead 1985). If we assumethat Least successwas also reduced by human disturbance Auklet breeding successat Kongkok was most as evidencedby the high rates of chick death representativefor the population in our study and disappearanceon disturbedplots. Similar- area in 1987, and divide the 10 unaccounted for ly, mostchick losses of both specieson St. George egg and chick lossesproportionately between and Buldir islands resulted from death or dis- all observed outcomes(i.e. 7 chicks fledged), appearance (16-19%, Roby and Brink 1986, then "typical" breeding successmight approach Knudtson and Byrd 1982). 82%.Similarly, CrestedAuklet breeding success Least Auklet breedingsuccess (number of might have been much higher than measured. chicks fledged per egg laid) on the least dis- Effectsof predationand weather.--Red-backed turbed Kongkok plots (66%)was higher than at Voles (Clethrionomysrutilus) were abundant and Buldir Island (51%) and a little lower than at St. widespreadon breeding slopes,and frequently George Island (67-73%). Given differences in preyed on auklet eggsand chicks.Sealy (1968, April 1990] BreedingBiology of Auklets 349

1982)and Searing(1977) observedsimilar rates Nettleship, and Alan Springer improved this manu- of predation(1-11%) by voles.Predation on sea- scriptwith their helpful commentsand discussions. birds by microtine rodents is unusual (Sealy This work was supportedjointly by the Minerals 1982),but it has also been reported (Sealy and Management Service (Intra-agency Agreement No. 14-12-0001-30391), the Alaska Fish and Wildlife Re- B•dard 1973, Murray et al. 1983, Gastonet al. search Centerß the Alaska Maritime National Wildlife 1988). We could not account for the losses of Refuge, and Mr. Theodore Crossof New York. aukleteggs and chicks from predationby Arctic foxes(Alopex lagopus), but aukletsare not pre- LITERATURE CITED dominant prey of foxes that den near seabird colonieson St. Lawrence Island (Fay and Ste- B•DARD,J. 1969a. The nestingof the Crested,Least, phenson 1989). and Parakeet auklets on St. Lawrence Island, We confirmed (Roby and Brink 1986) that Alaska. Condor 71: 386-398. adult attendance of Least and Crested auklet 1969b. Feeding of the Least,Crested, and chicksdecreased rapidly during the first 10days Parakeet auklets around St. Lawrence Islandß Alaska. Can. J. Zool. 47: 1025-1050. after hatching. As chicks do not becomeho- BIRKHEAD,T. R. 1985. Coloniality and socialbehav- meothermicuntil they are ca.4-5 daysold (Sea- ior in the Atlantic Alcidae. Pp. 355-383 in The ly 1968),unattended chicks are particularly vul- AtlanticAlcidae (D. N. Nettleshipand T. R. Birk- nerable to cold, wet weather in the first week head, Eds.). Orlando, Florida, Academic Press. of life. This accounts for the increased chick --, & M.P. HARRIS.1985. Ecologicaladaptations mortalitythat followed two prolongedperiods for breedingin the Atlantic Alcidae.Pp. 205-231 of rain in August. in The Atlantic Alcidae (D. N. Nettleship and T. Chickdiets and growth.--Auklets had no dif- R. Birkhead, Eds.). Orlando, Floridaß Academic ficulty feedingand rearingchicks in 1987.Me- Press. FAY,F. H., & T. J.CADE. 1959. An ecologicalanalysis dian masses,maximum growth rates,develop- of the avifauna of St. Lawrence Island, Alaska. ment times, and diets were similar to those found Univ. of Calif. Publ. Zool. 63: 73-150. in other years and locations.The composition ß•: R. O. STEPHENSON.1989. Annualß seasonal, of dietswas similar to that reportedby B•dard and habitat-relatedvariation in feeding habitsof (1969b)for CrestedAuklets, but was markedly the arctic fox (Alopexlagopus) on St. Lawrence different for Least Auklets. Whereas we, and Island, Bering Sea. Can. J. Zool. 67: 1986-1994. others (Searing 1977, Springer and Roseneau GASTON,A. J. 1985. Development of the young in 1985),found that LeastAuklets consumedpre- the Atlantic Alcidae. Pp. 319-354 in The Atlantic dominantlyNeocalanus plumchrus, and generally Alcidae (D. N. Nettleship and T. R. Birkhead, much lower proportionsof Calanusmarshallae Eds.). Orlando, Florida, Academic Press. and N. cristatus,B•dard reportedthat C. finmar- -, I. L. JONES,& D. G. NOBLE.1988. Monitoring Ancient Murrelet breeding populations.Colon. chicus(now C. marshallaein the Pacific) was the Waterbirds 11: 58-66. dominantcalenoid copepod in auklet dietsand JOHNSONßD. H. 1979. Estimating nest success:the reported no N. plumchrus.However, it seems Mayfield methodand an alternative.Auk 96: 651- likely that B•dard identifiedN. plumchrusas C. 661. finmarchicus(B•dard 1969b:1036). The remain- KNUDTSON,E. R., & G. V. BYRD. 1982. Breedingbi- der of prey identified by B•dard were similar ology of CrestedßLeast, and Whiskeredauklets to thosereported by later investigators. in Buldir Islandß Alaska. Condor 84: 197-202. MAYFIELD,H. 1975. Suggestionsfor calculatingnest ACKNOWLEDGMENTS success. Wilson Bull. 87: 456-466. MURRAY, K. G., K. WINNETT-MURRAY,Z. A. EPPLEY,G. We thank the Sivuqaq and SavoongaNative Cor- L. HUNT,& D. B. SCHWARTZ.1983. Breedingbi- porations for permission to work on St. Lawrence ologyof the Xantus'Murrelet. Condor85: 12-21. Island.Lane Iyakitan (Sukilpaq)provided invaluable PLATT,J. F., B. D. ROBERTS,& S. A. HATCH. 1990. At- assistanceat our camp and in locating auklets and tendancepatterns and populationsof Least and study plots. We appreciatethe help provided on St. Crested auklets on St. Lawrence Island. Condor Lawrence Island by Darlene Apangalook, Gerome 92: 97-106. Apatiki, Chris Haney, JoelHubbardß Lewis Iyakitan, RICKLEFS,R.E. 1967. A graphicalmethod of fitting GerrardKanooka, and Bruce Rasmussen. Alan Spring- equationsto growth curves.Ecology 48: 978-983. er, Martha Springer,and Amy Stoneidentified the ROBY,D. D., & K. L. BRINK. 1986. Breedingbiology contents of auklet chick meals. Alan H. Brush, Vern of Least Auklets on the Pribilof Islandsß Alaska. Byrd, Tony Gaston,Chris Haney, Ed MurphyßDavid Condor 88: 336-346. 350 PLATTET AL. [Auk,Vol. 107

ROSENEAU,D. G., A.M. SPRINGER,E. C. MURPHY, & M. cliff-nestingseabirds at Kongkok Bay, St. Law- I. SPRINGER.1985. Population and trophic stud- renceIsland, Alaskaduring 1976.U.S. Dep. Com- ies of seabirdsin the northern Bering and eastern merce, NOAA, Outer Continental Shelf Environ- Chukchi Seas,1981. U.S. Dep. Commerce,NOAA, mental AssessmentProgram, Annu. Rep. 5: 263- Outer Continental Shelf Environmental Assess- 412. ment Program,Final Rep. 30: 1-58. SOWLS,A. L., S. A. HATCH, & C. J. LENSINK. 1978. SEALY,S.G. 1968. A comparativestudy of breeding Catalog of Alsakan seabird colonies.U.S. Dep. ecologyand timing in plankton-feedingalcids Interior, Fish Wildl. Serv. FWS/OBS-78/78. (Cyclorrynchusand Aethia spp.) on St. Lawrence SPRINGER,A.M., E. C. MURPHY, D. G. ROSENEAU,& M. Island, Alaska. M.S. thesis, Vancouver, Univ. I. SPRINGER.1985a. Populationstatus,reproduc- British Columbia. tive ecology, and trophic relationships of sea- --. 1973. Breedingbiology of the Horned Puffin birds in northwesternAlaska. U.S. Dep. of Com- on St. LawrenceIslandß Bering Sea,with zoogeo- merce, NOAA, Outer Continental Shelf graphicalnotes on the North Pacificpuffins. Pa- EnvironmentalAssessment Program, Final Rep. cific Sci. 27: 99-119. 30: 127-242. 1975. Influence of snow on egg-laying in ßD. G. ROSENEAU,E. C. MURPHY, & M. I. SPRING- auklets. Auk 92: 528-538. ER. 1985b. Populationsand trophic studies of 1981. Variation in fledging weight of Least seabirds in the Northern Bering and eastern Auklets Aethiapusilla. Ibis 123: 230-233. Chukchi Seas,1983. U.S. Dep. Commerce,NOAA, 1982. Volesas a sourceof egg and nestling Outer Continental Shelf Environmental Assess- lossamong nesting auklets.Murrelet 63: 9-14. ment Program,Final Rep. 30: 59-126. 1984. Interruptions extend incubation by ß&--. 1985. Copepod-basedfood webs: Ancient Murrelets, Crested Auklets, and Least aukletsand oceanographyin the BeringSea. Mar. Auklets. Murrelet 65: 53-56. Ecol. Prog. Set. 21: 229-237. --, & J. B•DARD.1973. Breedingbiology of the VERMEER,K., & g. LEMON. 1986. Nesting habitsand Parakeet Auklet (Cyclorrhynchuspsittacula) on St. habitats of Ancient Murrelets and Cassin's Auk- Lawrence Islandß Alaska. Astarte 6: 58-68. lets in the Queen Charlotte Islands, British Co- SEARING,G.F. 1977. Someaspects of the ecologyof lumbia. Murrelet 67: 34-44.