524 Short Communicationsand Commentaries [Auk, Vol. 114

LITERATURE CITED and E Gill, Eds.). Academy of Natural Sciences, Philadelphia, and American Ornithologists' BENNETT,A. J. 1996. Pacificand Americangolden- Union, Washington,D.C. plovers breeding in southwesternAlaska. JOHNSON,O. W., ANDP.M. JOHNSON.1983. - Northwestern Naturalist 77:49-51. molt-age relationshipsin "over-summering" [SLEIB,M. E. P. 1979. Migratory shorebirdpopula- and migratoryLesser Golden-Plovers. Condor tions on the Copper River Delta and eastern 85:406-419. Prince William Sound, Alaska. Studies in Avian JOHNSON,O. W., M. L. MORTON,P. L. BRUNER,AND P. Biology2:125-129. M. JOHNSON.1989. Fat cyclicity,predicted mi- IVERSON,G. C., $. E. WARNOCK,R. W. BUTLER,M. A. gratoryflight ranges, and features of wintering BISHOP,AND N. WARNOCK.1996. Springmigra- behavior in Pacific Golden-Plovers. Condor 91: tion of WesternSandpipers along the Pacific 156-177. Coast of North America: A telemetry study. WARNOCK,N., AND $. WARNOCK. 1993. Attachment Condor 98:10-21. of radio-transmittersto sandpipers:Review and JOHNSON,O. W., AND P. G. CONNORS.1996. Ameri- methods.Wader Study Group Bulletin70:28-30. can Golden-Plover (Pluvialis dominica) and Pa- cific Golden-Plover(Pluvialis fulva). In The Received2 December1996, accepted 25 February1997. of North America, nos. 201 and 202 (A. Poole AssociateEditor: J. S. Marks

TheAuk 114(3):524-530, 1997

The Effectsof FluctuatingFood Availability on Breeding Arctic Terns (Sternaparadisaea)

D. SUDDABY 1 AND N. RATCLIFFE 2'3 •RoyalSociety for theProtection ofBirds, The Baelans, Fetlar, Shetland ZE2 9DJ,United Kingdom; and 2RoyalSociety for theProtection of Birds,The Lodge, Sandy, BedfordshireSG19 2DL, UnitedKingdom

Reproductionis a costlyprocess that involvesan treme food shortage,condition and survival of investmentof effort that can decreaseparental con- adultsmay be adverselyaffected (Hamer et al. 1991). ditionand survival(Partridge 1989, Stearns 1992). In Foodsupply is expectedto affectbreeding parame- long-lived iteroparousspecies such as seabirds,re- terswithin a specificrange of prey availability,out- productivecosts produce a tradeoff in energyallo- side of which it will have little effect (Cairns 1987, cationbetween current and future breeding attempts Phillips et al. 1996). (Clutton-Brock1984), and thebreeding effort should The abundanceof sandeels(Ammodytes marinus) be adjusted to maximize lifetime reproductivesuc- was severely reduced in Shetland waters between cess.Food availability can have substantial effects on 1985 and 1990 due to successiveyears of poor re- the relative costsof reproductionand thus maybe an cruitmentof the group-0cohort (i.e. fish spawnedin important factorin determiningreproductive strat- the currentyear) and a subsequentdecline in spawn- egiesin seabirds(Boekelheide and Ainley 1989,Sy- ing-stockbiomass (Wright and Bailey 1993). Seabirds deman et al. 1991,Pons and Migot 1995). respondedby increasingtheir foraging effort (Uttley Cairns(1987) hypothesized that during periodsof 1992,Hamer et al. 1993,Monaghan et al. 1994) and reducedfood supply,seabirds should increase for- exploiting alternativeprey (Martin 1989, agingeffort in orderto bufferthe effects of foodscar- Hamer et aL 1991).Despite these changes in behav- city on breedingperformance. Further reductions in ior, the breedingsuccess of surface-feedingspecies food may result in parentsbeing unable to increase was adverselyaffected (Monaghan et al. 1989;Ham- foragingeffort without incurringexcessive costs, re- er et al. 199l, 1993;Phillips et al. 1996). suiting in a declinein reproductivesuccess (Mon- Arctic Terns (Sternaparadisaea) suffered six con- aghanet al. 1989,1992). Finally, during periods of ex- secutiveyears of almost completebreeding failure on Shetland (Walsh et al. 1990) due to adults aban- 3Address correspondenceto this author. E-mail: doning clutchesand chicksstarving shortly after [email protected] hatching(Monaghan et al. 1989, 1992).The number July1997] ShortCommunications and Commentaries 525 of breedingArctic Terns at Shetlanddeclined by 55% Arctic Ternswere monitored annually at 15 to 22 between 1981 and 1989 due to lack of recruitment coloniesthroughout Shetland from 1990to 1994.Col- and possiblyto increasedrates of nonbreeding,em- onieswere visitedevery 7 to 10 days,from mid-May igration, and mortality (Avery et al. 1993). Arctic until early Augustin eachyear. Randomly sampled Ternsare morevulnerable to reductionsin food sup- nests were selectedin each colony and their clutch ply than most other seabirdsat Shetlandowing to size was determined. The largest egg in a clutch their low degreeof behavioralplasticity (Monaghan (hereafterthe "A" egg, becausethe largest egg in a et al. 1992).This resultsfrom smallbody size (Craik clutchalways hatched first) wasweighed to the near- and Becker1992), specialization toward surface feed- est 0.1 g using a Pesolabalance, and maximum ing at short rangesfrom the colony (Pearson1968), length (L) and breadth (B) were measured to the and lack of alternativeprey species(Uttley et al. nearest0.1 mm using Verniercalipers. Internal egg 1989). volume (V) wasestimated from the equationof Davis In 1991,a dramaticincrease in the availabilityof (1975): sandeelsat Shetlandresulted from largenumbers of V = 0.00048 LB2 . (1) group-0 fish being swept north from Orkney by cur- rents (Wright and Bailey1993). Recruitment of these Egg density(D) wasestimated by dividing eggmass fish into the Shetland stock resulted in an abundance by eggvolume. The laying date of the A eggin each of group-1 fish in 1992and an increasedspawning- clutchwas predicted using the relationshipbetween stock biomass in subsequentyears (Anonymous egg densityand daysto hatching: 1995). This increase in sandeel abundance was as- days to hatching = 111.9 D - 102.2. (2) sociatedwith improved breeding performancefor other speciesof surface-feedingseabirds, including The relationshipwas highly significant(r • = 0.62, F Black-leggedKittiwakes (Rissatridactyla; Hamer et = 425.1,df = 1 and 256, P < 0.0001)and predicted al. 1993), Parasitic Jaegers(Stercorarius parasiticus; the hatching date to within 3 days for 95% of the Phillips et al. 1996), and Great Skuas(Catharacta skua; eggs.The laying date was calculatedby subtracting Ratcliffe 1993). Despite the extremesensitivity of the incubationperiod of 21 days(Cramp 1985)from Arctic Ternsto changesin food availability,no pub- the hatchingdate, and the data were expressedin lished informationexists on their breedingperfor- days after 30 April. mance sincethe recoveryof the sandeelpopulation We usedanalysis of variance(ANOVA) to analyze at Shetland.Here, we examinethe responsesof Arc- data on laying datesand egg volume and G-teststo tic Terns to variations in sandeel abundance at Shet- analyzedata on clutchsize. We usedlogistic regres- land from 1990 to 1994. sion to test whetherclutch size was independentof Methods.--Theavailability of prey to Arctic Terns laying date,with the proportionof one-eggclutches was inferred from the SOAFD(Scottish Office Agri- being the dependentvariable. The effectsof year, culture and FisheriesDepartment) estimatesof san- clutchsize, and laying date on A-egg volume were deel abundance in Shetland waters (Anonymous testedwith analysisof covariance(ANCOVA). 1995).The dataprovided a meaningfulindex of prey To estimate adult body condition, we captured availabilityto ternsbecause the locationsof the san- ternson nestsusing walk-in trapsduring the last 10 deel fishinggrounds and tern colonieswere closely days of incubationfrom 1990 to 1993. We measured juxtaposed (Monaghan 1992), and the sea-surface the maximum flattened wing chord (+ 1 mm) and availability of sandeelsis directly related to their body mass (ñ 1 g) for all birds. Chick growth was abundance(Wright and Bailey 1993). Estimatesof estimatedby measuringwing length (+- 1 mm) and sandeelabundance were made from Virtual Popu- body mass (+ 0.1 g) of chicks encounteredin the lation Analysis(VPA) of catch-at-agedata from re- study colonies. The age of chicks with unknown search trawls and commercialcatches at fishing hatching dates was estimatedby modeling wing groundsaround Shetland(Cook and Reeves1993). length against age for chickswith known hatching Abundancewas estimated separately for the first (up dates.Wing length (WL) increasedwith age (a) ac- to 30 June) and secondhalves of the year (after 30 cording to a logistic equation (r2 = 0.96) that pre- June).The biomassof group-0and group-1sandeels dictedthe age of 95% of chicksto within one day: (in metrictons) in eachhalf of theyear was estimated 177.19 W/• - (3) by multiplying the estimatedgroup abundanceby 1 •- e 255 0.23(a) the average mass of an individual (Anonymous 1995). VPA suffers from inaccuraciesdue to the dif- The mean growth rate of chicksin each year was ficulty in samplingcohorts randomly and the as- analyzedduring the linear stageof development(4 sumptions concerningnatural mortality rates and to 14 days old, Nisbet et al. 1995). Only one mea- terminal mortality (Cook and Reeves 1993). How- surement from each individual was included in the ever, it is the best estimate of trends in sandeel abun- analysis.The growth rate was analyzedusing AN- danceavailable for Shetlandwaters, and it provides COVA, with year definedas a categoricalvariable a coarseindependent index of tern food supply. and chickage as the covariate.Mean growth incre- 526 ShortCommunications and Commentaries [Auk, Vol. 114

TABLE1. Annual variation in biomass (thousands There was a strongtendency for the proportion of of metric tons)of group-1 sandeelsduring the first one-eggclutches to increasewith later laying dates half of the year and group-0 and group-1 sandeels (logisticregression, G 2 = 23.5, df = 1, P < 0.001). during the secondhalf of the year. Data are from This could result in annual variations of clutch size Anonymous(1995). beingan artifactof differencesin layingdate in those Group-1 Group-1 Group-0 years.However, the effectof year was still significant Year (lst half) (2nd half) (2nd half) independentof laying date,with 1990having a sig- 1987 24.8 14.1 3.0 nificantlyhigher percentage of one-eggclutches than 1988 2.0 1.1 9.3 other years (logisticregression, G 2 = 83.2, df = 4, P 1990 6.8 1.4 7.5 < 0.001). This suggeststhat one-eggclutches were 1991 3.2 1.9 112.9 more likely in years of low food availability than 1992 97.0 58.3 13.4 wouldbe expectedfrom thelaying date of the clutch. 1993 11.5 6.9 50.1 The volume of the A egg varied significantlybe- 1994 43.0 25.8 6.2 tween years (ANOVA, F = 42.8, df = 4 and 1,345, P < 0.001;Table 2), being significantlylower in 1990 than in all otheryears and lowerin 1991than in 1993. ments(g per day) were calculatedas the slopeof the The effectof year remainedsignificant independent regressionbetween age and masswithin eachyear. of laying date (ANCOVA;laying date,F = 22.6,df = Productivity estimateswere basedon peak counts 1 and 1,345, P < 0.001; year, F = 2.46, df = 5 and of fledgedchicks for the whole colonyand the num- 1,345, P < 0.05). There was no effect of clutch size on ber of Arctic Ternpairs at eachcolony as estimated A-egg volume independentof laying date (ANCO- from flush counts(Walsh et al. 1995). Productivity VA; laying date, F = 35.14, df = I and 1,345, P < wasexpressed as the overallproportion of pairs that 0.001; clutch size, F = 2.83, df = 2 and 1,345, P > fledgedchicks; data were analyzedusing logistic re- 0.05). gression. Body massof incubatingadults at Shetlandfrom Results.--Theabundance of group-0 and group-1 1990to 1993averaged 112.9 -+ SE of 0.63 g (n = 166). sandeelswas very low in 1990 (Table 1), as it had The stageof incubation(number of daysuntil hatch- been since1987 (Anonymous1995). Sandeel recruit- ing) had no effect on the incubationmass of adults ment was extremelyhigh in 1991,low in 1992,mod- (F = 0.29, df = 1 and 166, P > 0.5). Becausewing erate in 1993, and very poor in 1994. The abundance of group-1 sandeelsin the first half of the year was length(used as an indexof bodysize) was positively related to mass (r 2 = 0.02, F = 4.13, df = 1 and 166, a functionof group-0 recruitmentin the previous year (Table 1). P < 0.05) we used ANCOVA to examine annual vari- Laying date of Arctic Terns on Shetland differed ationsin massindependent of body size.Adult mass- significantlyamong years (ANOVA, F = 405.2, df = es were 15% lower in 1990 than in other years (F = 4 and 1,345,P < 0.0001;Table 2), being latestin 1990, 26.9, df = 3 and 166, P < 0.0001;Table 2) but did not intermediatein 1991,and uniformly early in subse- vary significantlyfrom 1991 to 1993. quentyears. Clutch size differed significantly among We did not obtain growth data in 1990 becauseall years (G2 = 499.7, df = 8, P < 0.001; Table 2). The of the chicksstarved before the age of 4 days.Chick percentageof one-eggclutches was very high (52%) growth varied significantlyfrom 1991to 1994.Year in 1990 comparedwith other yearsbut declinedin had a significanteffect on the slopeof the regression subsequentyears. The averageclutch size increased linesbetween age and chickmass (ANCOVA, age x from 1990until 1993and declinedslightly in 1994. year; F = 8.50, df = 4 and 1,088,P < 0.001;Table 2).

TABLE2. Annual variationin breedingparameters of Arctic Ternson Shetland,1990 to 1994.Values are œ +- SE (n in parentheses)with the exceptionof productivity(see footnote).

1990 1991 1992 1993 1994

Laying datea 39.8 -+ 0.3 (392) 32.9 -+ 0.3 (315) 27.4 _+0.3 (357) 23.7 -+ 0.4 (157) 26.8 +- 0.4 (129) Clutch size 1.5 _+0.03 (392) 1.9 -+ 0.03 (315) 2.2 _+0.03 (357) 2.5 +- 0.04 (157) 2.3 +- 0.05 (129) A-egg volume (ml) 16.2 +_0.1 (392) 16.9 +- 0.1 (315) 17.1 _+0.1 (357) 17.3 -+ 0.1 (157) 17.2 -+ 0.1 (129) Chick growth (g/day) No data 7.8 -+ 0.1 (380) 7.8 +_0.1 (330) 6.9 +- 0.2 (313) 7.2 +- 0.3 (73) Productivityb 0.00 (588) 0.70 (594) 0.72 (1,215) 0.38 (1,074) 0.21 (707) Adult mass(g) 97.8 -+ 2.3 (14) 113.9 _+0.7 (78) 115.9 +- 1.0 (34) 113.7 +- 1.1 (40) No data Days after 30 April. Numberof chicksraised per pair. July1997] ShortCommunications andCommentaries 527

0.9 ÷ ally high during the secondhalf of 1991 owing to 0.8 ÷ large numbersof fish being swept into Shetlandwa- ters by currentsfrom northeasternOrkney. This re- suited in a further increase in abundance and a wid- er distributionof sandeels(Wright and Bailey1993). Recruitmentwas poor in 1992 and 1994 and mod- .--. 0.5. erate in 1993 (Anonymous 1995), resulting in 0.4. group-1 fish being scarcein 1993 and moderately abundantin 1994.These data clearlyindicate that the food supply of Arctic Ternsfluctuated considerably • 02• during our studyperiod.

0.1 Studiesof larids haveshown that food supply is an important determinantof laying date (Safinaet al. 0.0 1988), clutch size (Sydemanet al. 1991, Pons and 20 40 60 80 100 120 Migot 1995),and egg volume (Hiom et al. 1991,Bol- Biomassof sandeels (thousandsoftons) ton et al. 1992, Oro et al. 1995). Egg productionis costlyfor Arctic Terns,with eachegg representing FIc. 1. Relationshipbetween biomass of sandeels 16% of adult body mass(Cramp 1985).The group-1 and productivity of Arctic Terns.Data on sandeel fishthat normally are fed to femaleArctic Terns were biomasstaken from Anonymous(1995). Data on tern very scarcefrom 1988to 1990,and a largeproportion productivityfrom 1987 to 1989were from Walshet of the diet in 1988 was made up of post-larval al. (1990). group-0 fish (Monaghanet al. 1992) that have very low calorificvalue (Hislop et al. 1991).The qualityof courtshipfeeding can affect egg laying (Nisbet 1973, Growth rates in 1991 and 1992 were 0.7 g per day 1977),and the lack of group-1sandeels from 1988to higher than in 1993 and 1994. 1990 was associatedwith decreasedegg production. Productivityvaried significantly among years (lo- The increasedavailability of group-1sandeels from gisticregression, G 2 = 1,186.4,df = 4, P < 0.001).No 1991 to 1994 coincided with earlier laying dates, chicksfledged from any of thestudy colonies in 1990, largerclutch sizes, and largerA eggsin ArcticTerns. but productivityimproved dramatically to approxi- When food supply fluctuates unpredictably mately0.7 chicksper pair in 1991and 1992.Produc- throughoutthe breedingseason, females should lay tivity decreasedsignificantly to 0.38per pair in 1993 as many eggs as prelaying nutrient reservesallow and further decreasedto 0.21chicks per pair in 1994. and opt for a brood-reductionstrategy later in the There was a significantrelationship between tern seasonshould conditionsnot improve (Bolton et al. productivity and the combinedbiomass of group-0 1992).Because Arctic Ternsfeed mainly on group-1 and group-1 sandeelsbetween 1987 and 1994 (data fish during prelaying, but provision chickswith for 1987 to 1989 from Walsh et al. 1990) that con- group-0fish that appearafter the laying period, the formedto a logisticcurve (F = 50.9,df = 5, P < 0.005; availabilityof sandeelsearly in the seasonprobably Fig. 1): is not a good indicator of their abundancelater on 0.73 (Wright and Bailey 1993).It is probablethat clutch P - 1 + e4-2s(ø 0s•)' (4) size in Arctic Terns is maximized within the limita- tions of nutrient reserves and that regulation of where P is tern productivity and M is sandeelbio- breedingeffort occurslater in the seasonby brood mass (thousandsof tons). Variation in sandeel bio- reductionor clutchdesertion (Monaghan et al. 1992). mass explained 96% of the variance in Arctic Tern The very low body massesof Arctic Terns during productivity.Productivity was uniformly low when 1987,1988, and 1990 (Avery et al. 1992;Monaghan et sandeel biomass was below 25,0000 tons and in- al. 1989, 1992;this study) indicatethat somaticre- creasedlinearly to an asymptoteof 0.73 chicksper servesmay indeedhave limited eggproduction dur- pair when sandeelabundance exceeded 70,000 tons ing those years. (Fig. 1). Monaghanet al. (1992)proposed a thresholdlevel Discussion.--From 1988 to 1990, the distribution of of adult bodymass in ArcticTerns below which terns sandeelsaround Shetland was patchy and their avail- shouldabandon breeding attempts in order to con- ability was very low (Wright and Bailey 1993). In serveresidual reproductivevalue. The fact that Arc- 1991,acoustic transects detected an influx of group-1 tic Terns continued to incubate below the "thresh- and older sandeels into Shetland waters that resulted old" body massin 1990(Avery et al. 1992)suggests in increasedabundance early in that year (Wright that birds incurred seriousreproductive coststhat and Bailey 1993).These sandeels were not sampled year.The declinein the tern populationsize on Shet- by the VPA surveysand so are not representedin Ta- land hasbeen greater than would be predictedfrom ble 1. The recruitmentof group-0fish was exception- low productivityalone (Avery et al. 1993),suggest- 528 ShortCommunications and Commentaries [Auk, Vol. 114 ing that the low masseswere associatedwith in- al. 1996, Phillips et al. 1996). Further researchis re- creasedmortality. Adult body mass increasedin quired to understandhow the interactionbetween 1991 and has remained at a similar level in subse- sandeelabundance and avian predation affectstern quentyears, suggesting that the increasedavailabil- productivity. ity of sandeelshas allowed Arctic Ternsto maintain Cairns (1987) suggested that relationships be- an improvedbody condition.The body massesob- tween breeding performanceand food supply served between 1991 and 1993 were higher than shouldbe nonlinear,being responsiveonly within a thosereported for WestScotland (Craik and Becker narrow rangeof prey availabilityand being relative- 1992) and Orkney (Sim et al. 1993) and similar to ly constantabove and below certain thresholdval- thosefor the CoquetIsland (Monaghanet al. 1989). ues.The relationshipbetween Arctic Tern productiv- These masses were well above the threshold levels ity and the combined biomass of group-0 and for clutch desertion proposedby Monaghan et al. group-1sandeels conformed to a logisticcurve (Fig. (1992), and no clutches were deserted between 1991 1), which provides some support for Cairns' (1987) and 1994. model. Similar logistic relationshipsbetween pro- Variationsin growth ratesof chickshave been as- ductivity and 'food supply have been demonstrated sociatedwith changesin sandeelavailability in many for ParasiticJaegers at Shetland(Phillips et al. 1996). studiesof Shetlandseabirds (Hamer et al. 1991,Dan- Acknowledgments.--Wethank all of the RSPBand chin1992, Uttley et al. 1994,Phillips et al. 1996).Sim- ScottishNatural Heritagewardens in Shetlandfor ilarly, growth ratesof Arctic Tern chickswere asso- their assistancein collectingdata on reservesduring ciated with the abundanceof group-0 and group-1 this project.We are gratefulto ICESfor allowingac- sandeels.During 1991and 1992,when sandeel avail- cess data on sandeel abundance and catch-at-age abilitywas high, growth rates at Shetlandwere high- data. We also thank Mark Avery, Pete Ellis, Eric er than those recorded in 1983 and 1984 (Ewins 1985) Meek, Adrian del Nevo, and JaneSears for logistic and also were higher than those documentedfor support,project planning, and commentson earlier Arctic Ternsnesting north of the Arctic Circle (Lem- drafts. metyinen 1972, Klaassenet al. 1989). In 1993 and 1994, growth rates at Shetlandwere lower, coincid- LITERATURE CITED ing with reductionsin sandeelabundance. Monagh- an et al. (1992)suggested that group-0sandeels were ANKER-NILSSEN,T. 1995. Long termbreeding disas- used primarily to feed chicks,but during the years ter of Puffins at Rost;causes and consequences. of their study all of the tern chicksdied beforethe Page7 in Threatsto seabirds(M. L. Tasker,Ed.). ageof 5 days.Older chickscan ingest larger group-1 Proceedingsof the 5th International Seabird sandeels(Ewins 1985)that havea higher total energy Group Conference,The Seabird Group, Peter- content(Hislop et al. 1991).Thus, the availabilityof borough,United Kingdom. both group-0 and group-1 fish can affect chick ANONYMOUS.1995. Report of the working group on growth. the assessmentof Norway pout and sandeel. Arctic Ternsfailed to raise any chicksto fledging ICES CM 1995/Assessment5. Copenhagen, agein 1990,a resultsimilar to that from 1985to 1989 Denmark. (Walsh et al. 1990). Prolongedbreeding failures as- AVERY,M. I., D. BURGESS,J. N. DYMOND, M. MELLOR, sociatedwith poor food supply alsohave been doc- AND P.M. ELLIS. 1993. The status of Arctic Terns umented in Atlantic Puffins (Fraterculaarctica) on Sterna paradisaeain Orkney and Shetland in Rost in Norway (Barrett et al. 1987,Anker-Nilssen 1989. Seabird 15:17-23. 1995)and in Black-LeggedKittiwakes in the Bering AVERY, M. I., D. SUDDABY,P.M. ELLIS,AND I. M. W. Sea (Springeret al. 1986).The group-0 fish are es- S•M. 1992. Exceptionallylow body-weightsof peciallyimportant for small Arctic Tern chicks(0 to Arctic Terns Sternaparadisaea on Shetland.Ibis 4 daysold) that wouldhave difficulty ingesting larg- 134:87-88. er prey. The low availability of group-0 fish in the BARRETT,R. t., t. ANKER-NILSSEN,E RIKARDSEN,K. late 1980sand 1990was responsiblefor masspost- VALDE,N. Rov, AND W. VADER. 1987. The food, hatchingstarvation of tern chicksat most Shetland growthand fledgingsuccess of NorwegianPuf- colonies(Monaghan et al. 1992). fin chicks Fratercula arctica in 1980-1983. Ornis Predationby gulls(Larus spp.) and GreatSkuas oc- Scandinavica 18:73-83. curred at many Shetlandtern colonies(Ewins 1985, BOEKELHEIDER. J., ANDD. G. AINLEY.1989. Age, re- Monaghanet al. 1989,pers. obs.), and this alsocould source availability and breeding effort in be indirectly related to sandeelavailability. Decreas- Brandt's Cormorant. Auk 106:389-401. esin marinefood suppliescan result in obligatepis- BOLTON,M., D. HOUSTON, AND P. MONAGHAN. 1992. civoreshaving to spendmore time foraging at theex- Nutritionalconstraints on egg formationin the penseof nest defense,whereas polyphagous gulls Lesser Black-backed Gull: An experimental and skuas could switch from a diet of fish to seabird study. Journalof Ecology61:521-532. chicks(Hamer et al. 1991,Spear 1993,Bukacinska et BUKACINSKA,M., D. BUKACINSKA,AND A. L. SPANNS. July 1997] ShortCommunications and Commentaries 529

1996. Attendance and diet in relation to breed- changingprey availability. Study 36: 181- ing successin Herring Gulls (Larusargentatus) 191. Auk 113:300-309. MONAGHAN, P. 1992. Seabirds and sandeels: the CAIRNS, D. K. 1987. Seabirds as indicators of marine conflictbetween exploitation and conservation food supplies.Biological Oceanography 5:261- in the northernNorth Sea.Biodiversity and Con- 271. servation 1:98-111. CLUTTON-BROCK,t. H. 1984. Reproductive effort MONAGHAN, P., J. D. UTTLEY,AND M.D. BURNS. 1992. and terminal investmentin iteroparousanimals. Effectsof changesin food availabilityon repro- American Naturalist 123:212-229. ductive effort in Arctic Terns. Ardea 80:71-81. COOK, R. g., AND S. A. REEVES.1993. Assessment of MONAGHAN,P., J. D. UTTLEY,M.D. BURNS,C. THAINE North Seaindustrial fish stocks with incomplete AND J. BLACKWOOD.1989. The relationship be- catch-at-agedata. ICES Journal of Marine Sci- tween food supply, reproductive effort and ence 50:425-434. breeding successin Arctic Terns Sternaparadi- CRAIK,J. C. A., AND P. H. BECKER.1992. Temporal saea.Journal of Animal Ecology58:261-274. and spatial variationsin the body weights of MONAGHAN, P., P. WALTON, S. WANLESS,J. D. UT- Common and Arctic Terns. Seabird 14:43-47. TLEY,AND M.D. BURNS.1994. Effects of prey CRAMP,S. (Ed.). 1985. Arctic Tern. Pages87-100 in abundanceon the foragingbehaviour, diving ef- Handbook of the birds of Europe,the Middle ficiencyand time allocationof breeding Guille- East and North Africa. The birds of the western mots Uria aalge.Ibis 136:214-222. Palearctic,vol. 4. Oxford UniversityPress, Ox- NISBET,I. C. t. 1973. Courtshipfeeding, egg size and ford. breeding successin Common Terns.Nature 241: DANCHIN,E. 1992. Food shortageas a factor in the 141-142. 1988 Kittiwake Rissatridactyla breeding failure NISBET,I. C. t. 1977. Courtshipfeeding and clutch in Shetland. Ardea 80:93-98. sizein CommonTerns Sterna hirundo. Pages 101- DAVIS,J. W. ]• 1975. Age, eggsize and breedingsuc- 109 in Evolutionaryecology (B. Stonehouseand cess in the Herring Gull Larus argentatus.Ibis C. Perrins, Eds.). Macmillan Press, London. 117:460-473. NISBET,I. C. t., j. A. SPENDFLOW,AND J. S. HATFIELD. EWINS,P. J. 1985. Growth,diet and mortalityof Arc- 1995. Variations in growth of Roseate Terns chicks. Condor 97:335-344. tic Tern Sternaparadisaea chicks in Shetland.Sea- bird 8:59-68. ORO, D., L. JOVER,AND X. RUIZ. 1995. The effect of HAMER, K. C., R. W. FURNESS,AND R. W. G. CALDOW. a trawl moratorium on somebreeding parame- ters of Audouin's Gull Larus audouinii in the Ebro 1991. The effectsof changesin foodavailability Delta, N.M. Spain.Pages 36-37 in Threatsto sea- on the breeding ecologyof Great SkuasCathar- birds (M. L. Tasker,Ed.). Proceedingsof the 5th actaskua in Shetland.Journal of Zoology (Lon- International Seabird Group Conference,The don) 223:175-188. SeabirdGroup, Peterborough, United Kingdom. HAMER, K. C., P. MONAGHAN,J. D. UTTLEY,P. WAL- PARTRIDGE,L. 1989. Lifetime reproductivesuccess TON, AND g.D. BURNS. 1993. The influence of and life historyevolution. Pages 421-440 in Life- food supply on the breedingecology of Kitti- time reproductionin birds (I. Newton, Ed.). Ac- wakesRissa tridactyla in Shetland.Ibis 135:255- ademic Press, London. 263. PEARSON,t. H. 1968. The feedingbiology of seabird HIOM, L., M. BOLTON, P. MONAGHAN, AND D. WOR- speciesbreeding on the Farne Islands, North- RAIL. 1991.Experimentalevidence for foodlim- umberland. Journalof Animal Ecology37:521- itation of egg productionin gulls. Ornis Scan- 552. dinavica 22:94-97. PHILLIPS, R. A., R. W. G. CALDOW, AND R. W. FUR- HISLOP, J. R. G., M.P. HARRIS, AND J. G. g. SMITH. NESS.1996. The influenceof food availabilityon 1991. Variation in the calorific value and total the breeding effort and reproductivesuccess of energy contentof the lessersandeel and other Arctic Skuas Stercorariusparasiticus. Ibis 138: fish preyedon by seabirds.Journal of Zoology 410-419. (London) 224:501-517. PONS,J. M., ANDP. MIGOT. 1995. Life-historystrat- KLAASSEN,M., C. BECH, D. MASMAN, AND G. SLAGS- egyof the HerringGull: Changesin survivaland VOLD. 1989. Growth and energeticsof Arctic fecundityin a populationsubjected to various Tern chicks(Sterna paradisaea). Auk 106:240-248. feeding conditions.Journal of Animal Ecology LEMMETYINEN,R. 1972. Growth and mortality in the 64:592-599. chicksof Arctic Terns in the Kongs•ord area, RATCLIFFE,N. 1993. The age-specificbreeding per- Spitzbergenin 1970. Ornis Fennica49:45-53. formance of Great Skuas Catharacta skua on Shet- MARTIN, A. R. 1989. The diet of Atlantic Puffin Fra- land. Ph.D. thesis, University of Glasgow,Glas- tercula arctica and Northern Gannet Sula bassana gow, Scotland. chicksat a Shetlandcolony during a period of SAFINA,C., J. BURGER,M. GOCHFELD,AND R. H. 530 ShortCommunications and Commentaries [Auk, Vol. 114

WAGNER.1988. Evidencefor prey limitation of UTTLEY, J. D., P. MONAGHAN, AND S. WHITE. 1989. Common and Roseatetern reproduction.Con- Differential effects of reduced sandeel availabil- dor 90:852-859. ity on two sympatricallybreeding speciesof SIM, I., D. SUDDABY,AND M. I. AVERY. 1993. Body tern. Ornis Scandinavica 20:273-277. weights of incubating Arctic Terns Sterna par- UTTLEY, J. D., P. WALTON, P. MONAGHAN, AND G. adisaeaon Orkney and Shetlandin 1990 and AUSTIN. 1994. The effects of food abundance on 1991. Seabird 15:24-29. breedingperformance and adult time budgetsof SPEAR,L. B. 1993. Dynamicsand effectsof Western GuillemotsUria aalge.Ibis 136:205-213. Gulls feeding on a colony of Guillemots and WALSH, P.M., M. I. AVERY, AND M. HEUBECK. 1990. Brandt'sCormorants. Journal of AnimalEcology Seabirdnumbers and breedingsuccess in 1989. 62:399-414. Report No. 1071, Nature ConservancyCouncil, SPRINGER,A.M., D. G. ROSENEAU,D. S. LLOYD ,C. P. Peterborough,United Kingdom. McRo¾, AND E. C. MUR•'H¾. 1986. Seabird re- WALSH, P.M., D. J. HALLEY, M.P. HARRIS, A. DEL sponsesto fluctuating prey availability in the NEVO, I. M. W. SIM, AND M. L. TASKER. 1995. easternBering Sea. Marine EcologyProgress Se- ries 32:1-12. Seabirdmonitoring handbook for Britain and STEARNS,S.C. 1992. The evolution of life histories. Ireland. Joint Nature ConservationCommittee, OxfordUniversity Press, Oxford. Royal Societyfor the Protectionof Birds,Insti- SYDEMAN,W. J., J. E PENNIMAN, t. M. PENNIMAN, P. tute of Terrestrial Ecology, and The Seabird PYLE,AND D. G. AINLEY. 1991. Breedingperfor- Group,Peterborough, United Kingdom. mance in the WesternGull: Effects of parental WRIGHT,P. J., AND M. C. BAILEY.1993. Biologyof age,timing of breedingand yearin relationfood sandeelsin the vicinity of seabird coloniesat availability.Journal of Animal Ecology60:135- Shetland.Fisheries Research Report No. 15/93. 149. SOAFDMarine Laboratory,Aberdeen, Scotland. UTTLEY,J. D. 1992. Foodsupply and allocationof pa- rental effort in Arctic TernsSterna paradisaea. Ar- Received13 May 1996,accepted 27 February1997. dea 80:83-92. AssociateEditor: L. J. Petit