POPULATION ECOLOGY OF THE LONG-BILLED (NUMENIUS AMERICANUS) IN WESTERN IDAHO

ROLAND L. REDMOND AND DONALD A. JENNI Departmentof Zoology,University of Montana,Missoula, Montana 59812 USA

ABSTRACT.--IIlwestern Idaho, a breeding population of Long-billed (Numenius americanus)remained relatively stable over a 7-yr period.Productivity was monitored closely from 1977through 1979. Clutch mortality rates did not differ significantlyamong years, and meanclutch survival for a 32-daynesting period (4 daysegg laying and 28 daysincubation) was40% for all 3 yr combined.Females laid justone clutch(usually of 4 eggs)each season, and meanclutch size was significantlysmaller in 1979than in 1977and 1978.Estimated fledgingsuccess ranged from 0.40in 1977to 0.17in 1978and wasconsistently greater for femalesthat nestedearly eachseason. Mean annualadult survivalwas estimatedat 85% basedon resightingsof color-markedindividuals. Limited data for subadultsurvival pre- cludedcomplete demographic analysis. Nevertheless, given the estimatesof productivity and adult survivalpooled for 3 yr, survivalof subadultsfrom fledginguntil first breeding as3-year-olds needed to beonly 58% to maintainthe stablepopulation size observed through 1983. Received10 June1985, accepted 7 April 1986.

THE population ecology of shorebirds (Cha- land et al. 1982), and Common Redshank (Yates radrii) is not well known, probably because 1982). Larger speciestend to be better studied many speciesare wary, breed in remote areas, during the nonbreeding season,and survival or both. Most thoroughly studied are estimatesbased on annual returns to wintering that nest at temperate latitudes such as the Eur- grounds exist for Bar-tailed (Limosa asianOystercatcher (Haematopus ostralegus; Goss- lapponica)and Eurasian Curlews (Numeniusar- Custard et al. 1982, Safriel et al. 1984), Common quata)(Evans and Pienkowski 1984). Ringed Plover (Charadriushiaticula; Laven 1940, We studied breeding density, clutch size, Bub 1962, Pienkowski 1984a), and Common nesting mortality, reproductive success,and Redshank (Tringatotanus; Grosskopf 1959, 1964; adult survival in a population of Long-billed Yates 1982). Many aspectsof breeding biology Curlews (Numenius americanus). These demo- are known for other shorebirds, especially in graphic data are integrated by life-table anal- Europe and Scandinavia (see Cramp and Sim- yses,and the longer-term reproductive perfor- mons 1982, Evans and Pienkowski 1984). But mance of the population is evaluated. even among calidridine (Scolopaci- dae), a subfamily with diverse social adapta- STUDY AREA AND METHODS tions and mating systems(Pitelka et al. 1974, Studyarea.--The study area was an upland strip of Myers 1981), detailed demographicparameters shortgrassrangeland (•21,600 ha) in western Idaho have been reported only for the Dunlin (Cali- (Redmond and Jenni 1982). We concentrated on a dris alpina; Soikkeli 1967, 1970a, b) and Tem- portion of the area (• 1,600 ha) that supportedhigh minck's Stint (C. temminckii;Hild•n 1978). Trin- breeding densities of Long-billed Curlews. This gine sandpipersexhibit a wider range of body smaller plot was grazed intensively by large bands sizes and tend to breed at lower latitudes than of sheepfrom late March through early May each do calidridines (Johnsgard1981). Hence, they year;other portions of the studyarea were grazedby should be more accessiblefor study, and also cattle from November to June. Methods.--Thenesting cycle was divided into pre- more accessiblefor comparative demographic laying, laying, incubation,and prefledgingperiods. analyses.Long-term reproductive performance A standardclutch of 4 eggswas laid in 4-7 days,and and population dynamics,however, are known incubation began after the last egg was laid. Time for only a few smaller speciessuch as the Spot- from laying to hatchingof the last egg was consid- ted ( macularia;Oring et al. ered the incubation period; it ranged from 28 to 30 1983), Common Sandpiper (A. hypoleucos;Hol- days(•? = 28.4 days).

755 The Auk 103: 755-767. October 1986 756 REDMONDAND JENNI [Auk, Vol. 103

Population densitieswere estimatedeach year by fornicus;Ring-billed Gulls, L. delawarensis;and Frank- a modified Finnish line-transect method (Redmond lin's Gulls, L. pipixcan)occasionally passed over the et al. 1981),and nestswere locatedduring egg laying study area during April and May. Only ravens were or incubation according to methods described by mobbed by adult curlews. Redmond (1986). When found, each nest was marked Grazing livestock also destroyed curlew nests. with a numbered wooden stake (2.5 x 5 x 40 cm) Characteristically,one or more eggsdisappeared from placed 10 m upslope from the nest cup. To deter a nest coincidentwith the presenceof sheepor cattle mammalianpredators from following our scenttrails, in the vicinity. Sometimes a crushed egg remained paradichlorobenzene(PDB) crystalswere sprinkled in the cup or an egg was dislodgeda short distance. at a 1-m radiusaround each nest cup. On subsequent Gophersnakes (Pituophis melanoleucus) were the only visits, a nest was always approached directly from reptiles capableof preying on curlew eggsor chicks. the marking stake;when departing, a line of PDB Thesesnakes were encounteredregularly during late crystalswas spreadacross this path. Becausethe crys- spring, but individuals were always relatively small tals were small and very volatile, they evaporated and probably unable to swallow a curlew egg whole. quickly and probably had only a short-term effect. The existence of individual gopher snakes large Whereasthe efficacyof this method for discouraging enough to ingest whole curlew eggs was document- mammalian predatorshas never been tested,Lehner ed in Nebraska (Tremaine 1975), but we doubt that et al. (1976) showed that similar chemicals deterred such predation was frequent on this study area. coyotes(Canis latrans) and dogs(Canis familiaris) from Daily mortality rates were calculatedfor all clutch- food rewards. Similarly, Knight (1983) found PDB es, for individual eggs within successfulnests, and and napthaleneto be useful in repelling skunksfrom for chicks before fledging according to Mayfield's specificareas. (1975) method as modified by Willis (1981). When Nests found during incubation usually were the day of nest destructionwas not known, it was checkedevery 2-5 daysuntil the eggsfirst starredor assumed to have occurred at 40% of the interval since pipped, and then every day until hatching was com- the last visit (Johnson 1979). Variance estimates of pleted. Nests located during egg laying were not thesemortality ratesfollow Johnson(1979), and rates checked for the first 2-3 weeks of incubation. Clutch were compared by G-testsof independence (Willis size was estimatedby the maximum number of eggs 1981). Capture and marking techniques were de- found in a given nest. scribedby Redmond and Jenni (1982), and Redmond The diversityof potentialpredators on and around (1984) provided details about radio-tracking meth- the studyarea made it difficult to interpret every act ods. of predation and to identify responsiblespecies. Resightingsof natally philopatric curlews (indi- Mammalian carnivores that hunted regularly over viduals color-bandedas chicksin the study area and portionsof the studyarea included coyotes, fetal dogs, resightedthere in later years) were made opportun- feral cats(Felis domesticus), badgers (Taxidea taxus), and istically along survey routes from 1979 through 1981. long-tailedweasels (Mustela frenata). Red foxes(Vulpes By themselves,these resightings represent minimum fulva), striped skunks (Mephitis mephitis),and rac- survival rates for subadults.Because of a significant coons(Procyon lotor) occurred on adjacentagricultur- male bias in natal philopatty (Redmond and Jenni al lands but were never seen hunting in upland hab- 1982), however, subadult survival for males was es- itat. timated as one half the number of color-marked chicks Nest losswas attributed to a badger if the nest cup known to fledge and later resighted. Estimatesof an- was damaged and buried under several centimeters nual adult survival utilize resighting data from 1978 of soil. Large, crushedshell fragmentsoften lay be- through 1981 and are basedon the maximum-likeli- neath this soil or in the immediate vicinity of the hood method of Cormack (1964). To obtain resight- destroyednest. Canids were presumedto be respon- ings of color-marked adults, their previous nesting sible for an empty, undamagednest cup whenever territoriesand adjacentterritories were searchedreg- the previous visit had revealed an intact clutch. In ularly during the prelaying and incubationperiods. several casesalso attributed to canids, the nest cup was damaged but not buried, and large shell frag- RESULTS ments lay nearby. Avian predators usually did not destroy entire Breedingdensity.--Because of high variances, clutches;instead, they openedand consumedpor- there were no significant differences among tions of 1 or 2 eggs.Partially destroyedclutches were abandoned by adult Long-billed Curlews, and re- annual density estimates of territorial male maining eggs were subsequently consumed. Black- Long-billed Curlews on the 1,600-hastudy plot billed Magpies (Pica pica) were the most abundant (Fig. 1). Thus, for the duration of the study, we avian egg predator. Common Ravens (Corvuscorax) assumedthat the breeding population was rel- and severalgull species(California Gulls, Laruscali- atively stable in size, even though a gradual October1986] Long-billedCurlew Population Ecology 757

TABLE1. Clutch size of Long-billed Curlews in western Idaho.

Frequencyof clutch size a o Year 3 4 5 Mean + SE b

1977 I 24 0 3.96 ñ 0.039 1978 1 36 1 4.00 ñ 0.037 1979 8 32 0 3.80 ñ 0.063 1977-1979 10 92 1 3.91 ñ 0.000

a Clutch size could not be determined for 16 addi- do d, d3 tional nestsbecause they were either destroyeddur- ing laying or found during hatching. YEAR bANOVA amongyears: F = 4.56,2,100 df, P = 0.01. Fig. l. Annual densityestimates (mean + SD) of maleLong-billed Curlews attempting to breedon the study site from 1977 to 1983.Estimates are basedon tended, abandoned, or wild) were the most fre- Finnish line-transect method of J•irvinen and V•iis•i- quently seen.Coyotes were seen infrequently, nen (1975)as modifiedby Redmondet al. (1981).Ac- and red foxes were sighted occasionallyin ag- tual data are available in Redmond (1984). ricultural fringe areaswithin 3-5 km of curlew nesting habitat. Other causesof clutch losswere grazing live- decline in males attempting to breed probably stock,other , and investigatordisturbance occurred between 1978 and 1981 (see Discus- (Table 2). When large bands of sheep passed sion). through an area, chanceseemed to play a role Clutch size.--Female Long-billed Curlews in the actual destruction of Long-billed Curlew commonlylaid a single clutchof 4 eggsin 1977 nests. On two occasions in 1977, bands of more and 1978 (Table 1). In 1979, however, the fre- than 500 ewesand lambstrampled pastthe same quencyof 3-eggclutches increased dramatical- two neighboring nests en route to a water ly, and mean clutch size was significantlyless trough. Neither clutch was damaged,but both than in the previous 2 yr (GT-2 multiple com- times one of the wooden marking stakes was parison test, P < 0.05). There was no evidence snappedoff and crushed.Soon after each in- of renesting by any color-banded or radio- cident, the adult curlews returned and re- marked female in any year. sumed incubation. Similar, but less dramatic, Eggand clutchmortality.--Mean clutch mor- interactionsresulted in severalbroken eggsand tality rate for all years was 0.029 nestslost per 4 nest desertions in 1978 and 1979. day (Table 2). Daily clutch mortality rateswere Adult Long-billed Curlews could readily de- remarkablyconstant in 1977 and 1978,and in- fend their nestsagainst magpies and other avi- creasedinsignificantly in 1979 (G = 0.02, 2 df, an egg predators. We therefore suspect that P = 0.90). these birds probably gained accessto curlew Mammalian carnivores were the most impor- clutchesonly as a result of nest desertion or tant predators of Long-billed Curlew eggs and abnormal inattendanceby the adults. Two un- clutches(Table 2). The relative intensity of pre- usual casesof egg damageand nest abandon- dation by canidsand badgersdiffered signifi- ment were attributed to birds (1 each in 1977 cantly among years (G = 6.10, 2 df, P < 0.05). and 1979) and may have been causedby other During 1977 badgerswere particularly destruc- Long-billed Curlews.All eggsin both clutches tive of curlew eggs in an area of high Town- were pecked, but their contentswere not eaten. send ground squirrel (Spermophilustownsendii) The holes, approximately 15-30 mm in diame- density, where 67% of all nests were found. ter, were considerablysmaller than those made During 1978 and 1979, when only 15% of all by magpies.Both of thesenests were within 40 nests were found in this area, canids were the m of other active curlew nests. dominant egg predator (Table 2). Individual egg lossesor hatching failures It is difficult to identify any specificcanid as were recordedeach year in neststhat produced the most important egg predator. Dogs (unat- one or more chicks (Table 3). This mortality 758 REDMONDAND JENNI [Auk, Vol. 103

TABLE2. Lossesof Long-billedCurlew clutches.

1977 1978 1979 Total Percent

Clutches 30 40 49 119 Losses to: Canids 3 8 10 21 17.7 Badgers 7 3 3 13 10.9 Birds 2 2 4 8 6.7 Livestock 0 3 2 5 4.2 Trapping 0 0 2' 2 1.7 Unknown 0 0 1 1 0.8 Total 12 16 22 50 42.0 Nest-days 449 592 705 1,749 Mortality/day 0.0277 0.0280 0.0294' 0.0285 $E 0.00761 0.00666 0.00625 0.00391 aTwo clutcheswere abandoned becauseof trapping efforts in 1979; mortality rate assumesthese would have hatchedsuccessfully. was a result of addled eggs (n = 7), parental and 2 of 62 in 1979). The 3 nonviable hatch- abandonmentof late, asynchronouslyhatching lings in 1977 were from 3 different 4-egg eggs(n = 5), and grazing livestock(n = 6). Pat- clutches,and each weighed 20-25% less than terns of single egg lossesor hatching failures their siblings.All had poorly healed umbilical in otherwise successfulnests did not differ sig- regionsand incompleteyolk sacretention, with nificantly among years (Gaai= 6.00, 4 df, P = small pieces of dried shell or membranesad- 0.20). hering to the yolk sac.There was nothing pe- Hatchingsuccess.--Probabilities of clutch sur- culiar about the appearance of nonviable vival for a 32-daynesting period were remark- hatchlingsin 1978or 1979,and they may have ably constanteach year (Table 4). Egg survival perished from inadequatebrooding. in successfullyhatching nests showed more Long-billed Curlew chicks have cryptic variability among years (Table 4), probably be- plumageand behavior,which makesthem very cause of the absenceof livestock damage in difficult to locateafter they leave the nest. Con- 1977; nevertheless, these differences were not sequently,we relied heavily on the fatesof ra- significant (G = 4.47, 2 df, P > 0.80). dio-marked individuals to determine chick Chickmortality.--Each year individual chicks mortality rates. The averageage when chicks that had not survived more than a few hours were radio-markeddecreased each year from were found in or near a nest cup (n = 3 of 79 15.8 (1977), to 14.0 (1978), to 6.0 days (1979). color-marked chicks in 1977, 1 of 69 in 1978, We found no mortality of radio-marked chicks

TABLE3. Egg lossesand hatching failuresfrom successfulLong-billed Curlew nests.

1977 1978 1979 1977-1979 Eggsin successfulnests 68 95 91 254 Number of eggs Addled 1 3 3 7 Abandoned 3 1 1 5 Damaged' 0 4 2 6 Egg-days 1,258 1,702 1,743 4,703 Lossratea/day + $E ( x 103) 0.00 2.42 1.18 1.31 +0.000 +1.174 +0.811 +0.521 Percentagesurviving 32 daysb 100.0 92.5 96.3 95.9 Percentageaddled or abandoned 5.88 4.40 4.49 4.84 aCaused by grazing livestockonly. It is assumedthe 4 other eggsdamaged in 1979 becauseof trapping efforts would have hatched. bFour days for egg laying and 28 daysfor incubation. October1986] Long-billedCurlew Population Ecology 759

TABLE4. Estimatesof hatching and fledging successper breeding adult.

Mean Nesting Hatchingsuccess Fledging clutchNo. of young Year classa Clutches Eggsb successc size fledged/adult 1977 Early 0.408 0.972 0.565 4.00 0.45 Late 0.408 0.906 0.437 3.91 0.32 All 0.408 0.941 0.495 3.96 0.38 1978 Early 0.403 0.878 0.293 4.00 0.21 Late 0.403 0.890 0.100 a 4.00 0.07 All 0.403 0.884 0.229 e 4.00 0.16 1979 Early 0.385 0.856 0.485 3.77 0.30 Late 0.385 0.978 0.216 3.86 0.16 All 0.385 0.920 0.318 3.80 0.21 1977-1979 Early 0.397 0.896 0.436 3.92 0.30 Late 0.397 0.927 0.262 e 3.92 0.19 All 0.397 0.913 0.350 e 3.91 0.25 aEarly clutcheswere completedbefore the mediandate of clutchcompletion each year; late clutcheswere completedon or after the median date. bHatchability of eggsin successfulclutches (i.e. thosein which at least1 egg hatched). cBased on chick survival estimatesfrom age 6 daysto fledging (Table 7) times 75% in 1977 and 1978 and 65% in 1979, except as noted below (see text). a Fractionof color-marked,late-hatching nestlings that were resightedpostfledging in 1978(2/20). eBased on adjustedchick loss (n = 1.87rather than 8) amonglate-nesting females in 1978;see Table 6 and text.

in the 0-5-day-old age classduring 1977 or 1978, fledging. One chick was eaten by a long-tailed but our samplesizes were only 3 and 4 for each weasel, and the fates of 2 others were un- year. In 1979,when our samplewas much larg- known. In 1978, only 2 of 13 radio-marked er, mortality in this very young age classwas chicks(15%) survived to fledge, and both were 57% (12/21). Six conspicuouslyfeeble chicks killed soon thereafter. Raptorswere responsi- perished within a few hours of hatching. Two ble for at least 73% (8/11) of all prefledging others died on their fourth and fifth day, re- lossesin 1978, and no mortality was attributed spectively.Neither gained weight nor showed to mammalian carnivores. The fate of 1 indi- signsof feeding;we inferred that both starved. vidual was unknown, however, and it may have One chick died mysteriouslyon its secondday, been captured by a mammal and carried un- and the radio signal from another disappeared derground.In 1979,predation was again light, when the chick was 5 days old. Finally, 2 sib- lings may have succumbedto heat stresson their fourth day. They hatched late in the sea- TABLE5. Mortality of radio-marked Long-billed son (18 June) within 100 m of a well-traveled Curlew chicksaged 0-45 days. road. The female parent disappearedjust after hatching, leaving the male to provide all pa- 1977 1978 1979 Total Percent rental care. The family crossedthe road on 20 Chicks 12 13 26 51 June and entered a sparse,crested wheatgrass Losses to: (Agropyroncristatum) planting where grasshop- Raptors 0 8 1 9 17.7 pers were abundant and shade was scarce.Ex- Mammals 1 0 2 3 5.9 Aspergillosis 0 2 0 2 3.9 cessivetraffic along this road on 21-22 June Other factors a 0 0 12 12 23.5 (both clear, hot days)may have interfered with Unknown b 2 1 2 5 9.8 the male's ability to provide needed shade to Total 3 11 17 31 60.8 the chicks. Predation on radio-marked chicks varied aSee text for fatesof these12 chicksaged 0-5 days in 1979. dramatically among years (Table 5). In 1977, bRadio signalswere lost, and chicksalmost certain- 75% (9/12) of these chicks survived through ly were depredated. 760 REDMONDAND JENNI [Auk,Vol. 103

TABLE6. Survival of radio-markedLong-billed Curlew chicksfrom age 6 daysto fledging.

Nesting No. of No. Chick- Mortality rate/day Days to Percent Year classa chicks lost days + SE pb fledgec survival 1977 Early 5 1 116 0.009+ 0.0086 0.60 32.0 75.3 Late 7 2 123 0.017 + 0.0114 32.0 58.3 All 12 3 239 0.013 + 0.0072 32.0 66.0 1978 Early 5 3 130 0.024+ 0.0132 0.001 39.5 39.1 Late 8 8 42 0.211 + 0.0606 40.0 0.01 All 13 11 172 0.066 + 0.0187 39.8 6.5 1979 Early 6 1 120 0.009+ 0.0083 0.15 34.3 74.6 Late 8 4 122 0.034 + 0.0161 32.0 33.2 All 14 5 242 0.021 + 0.0091 33.2 48.9 1977-1979 Early 16 5 366 0.014+ 0.0061 0.01 35.3 60.8 Late 23 14 287 0.051 + 0.0127 34.7 16.4 All 39 19 653 0.030 + 0.0066 35.0 34.4 aEarly clutcheswere completedbefore the median date eachyear; late clutcheswere completedon or after median dates. bProbability that daily mortality rates do not differ between early and late clutches;obtained by G-tests (Willis 1981). cMeans for each period in each year. with just 4% (1/26) of the samplebeing lost to The estimate for 1979 was lower because small- raptors and 8% (2/26) to mammals. er eggs were laid that year (Redmond 1986), The only otherknown causeof mortalityto and in other there is an in- radio-marked chicks was aspergillosis,a pul- verse relationship between egg size and chick monary mycotic infection from which 15% (2/ mortality (Lundberg and V•iis•inen1979). 13) perishedin 1978(necropsies and diagnoses The variable intensity of predationon radio- by Dr. B. W. O'Gara, Montana Cooperative marked chicksaged 6 daysto fledging suggests Wildlife ResearchUnit). Aspergillosishas been variation among years in either predation pres- reportedfor a wide rangeof free-living,avian sure or effectivenessof parental care, or both. species,particularly waterfowl and gamebirds Becauselate-hatching chicks tend to receive re- (O'Mearaand Witter 1971),but never for Long- duced parental care as a result of early depar- billed Curlews. ture by females(Redmond 1984),we compared Fledgingsuccess.--Survival of very young daily mortality ratesfrom age 6 days to fledg- chicks(aged 0-5 days)probably depends more ing between chicksthat hatched early (before on their learningto feedeffectively and receiv- median date) and late (on or after the median) ing occasionalthermoregulatory assistance from eachyear. In 1977 and 1979 theserates did not parents than on avoiding predation. In 1977 differ significantly,but in 1978 late-hatching and 1978 few chicks were monitored for this chicks experienced significantly greater mor- period, and no mortality was detected.From tality than those that hatched early (Table 6). the larger sample in 1979, we estimated 14.1% Resightings of color-marked juveniles that of chickslost per day (12 lost over 85 chick- were marked as nestlingsprovide a minimum days) or 40.2% survival to age 6 days. These estimate of fledging success.The fractions re- figureswere almostcertainly biased by stress sighted in 1977 (16/56) and 1979 (1/47) were to very young chicksassociated with their car- much lower than the estimates of survival from rying a radio transmitter and being disturbed age 6 days to fledging for radio-markedchicks daily. Such stressmight have increased mor- (Table 6). In 1978, however, we resighted as tality as much as 50-100%,which would result juveniles 17%(5/30) of the color-markedchicks in survival values of 60-80% from hatching to that hatched early and 10%(2/20) of those that age 6 days. Our estimatesof fledging success hatched late. This 10% minimum survival of (Table 4) were based on 75% of chicks surviv- late-hatchingchicks was considerablygreater ing to 6 days in 1977 and 1978, and 65% in 1979. than the 0.01% estimated from the sample of October 1986] Long-billedCurlew Population Ecology 761

TABLE7. Resightingsof natally philopatricLong-billed Curlews. a

No. Year resighted Sex Hatch year resighted 1978 1979 1980 1981 Males 1977 5

1978 4

1979 0 Females 1977 1 1978 0 1979 ! aEach horizontal line representsthe resightingof one Long-billedCurlew that was color-markedas a chick;a dashedline indicatesnonbreeding status for that year. radio-markedchicks. To adjustfor the possibil- was related to human disturbance activities. ity that in 1978 mortality was greater among Carcassesof 9 adults were found during May radio-markedchicks than among color-marked and June (8 in 1977, 1 in 1979). Three were shot chicks (see Discussion), we assigned a 10% with small-caliber firearms, but the other 6 were fledging successrate to the late-nestingclass too decomposedand damaged to ascertainthe (Table 4). Basedon this adjustment,estimated causeof death. We believe that they also were fledging successin 1978 was 23% for all fe- shot becauseall were found along roadsacces- males. sible to the public and during late May, when Becausefemale Long-billed Curlews are mo- many humans used the upland for recreation nogamousand lay just one clutch each season, and when breeding adults were mostvigorous the number of young fledged per breeding in mobbing intruders (Redmond 1984). adult equalshalf the number fledged per nest. Survivorship.--Becauseof a significant male Estimatesof young fledged per female varied bias in natal philopatry (Redmond and Jenni considerablyamong years (Table 4). More im- 1982),and the tendencyfor malesnot to return portantly, early-nesting pairs consistently and attempt to breed until they are 3 or more fledged more young than did late-nestingpairs years of age (Table 7), subadultsurvival could (from 0.13 to 0.14 more; see Table 4). be estimatedonly for malesfrom the 1977 and Juvenileand adult mortality.--Only 2 radio- 1978 cohorts.For these years, 29 and 16 color- marked chicks were known to perish after markedchicks were sightedpostfledging. If half fledging. Both were taken by raptors in 1978 of these were actually males, then subsequent within a week of fledging. All other radio- resightingsof 21% (3/14) and 50% (4/8) for the marked chicks that were monitored past fledg- respective cohorts provide a minimum esti- ing (8 each in 1977 and 1979) survived. mate of survival for the 33-month period be- There was no evidence of nonhuman pre~ tween fledging and the first attempt to breed. dationon adult Long-billedCurlews at any time These estimates were unreliable because both during this study. The fact that many clutches resighting components(numbers returning to were lost to predators,often in twilight or com- breed and numbers resighted postfledging) plete darkness,suggests that incubating adults representminimums, and therefore actual sur- can effectively detect and escapemammalian vival could be either greater or less than the carnivores.One unsucceessfulattack by a Prai- estimates.Given the relatively stable popula- rie Falcon (Falco mexicanus)was observed in tion size (Fig. 1) and the results of a demo- early April 1979. graphicanalysis (see below), both figuresprob- All adult mortalityobserved during thisstudy ably underestimatedsubadult survival. 762 REDMONDAND JENNI [Auk,Vol. 103

TABLE8. Resighting data and annual survival esti- 13.3 matesa for adult Long-billed Curlews of both sexes breeding during 1977-1981.

Estimated

seenNo. probabilityof No. re- for Disap- No. sight- last pear- Year marked ed time ing Surviving + SD • 8.0- -.7 1977 9 -- 1 1978 17 b 8 9 0.11 0.89 ñ 0.10 1979 17 16 12 0.36 0.64 ñ 0.10 1980 4 c 11 6 0.36 0.84 ñ 0.16 0 1981 6 a 19 ------a.

a Based on maximum-likelihood method of Cot- bJ mack (1964). Z - 3.4 bSixteen adults captured and color-banded + 1 al- bino male (seen from 1978 to 1980). 2.0- c All 4 individuals were color-banded as chicks in 1977 and first sighted as adults in 1980. d Another 6 individuals color-banded as chicks in 1978-1979 and first sighted as adults in 1981. 0 0.5 1.0 1.5 2.0 • YOUNGFLEDGED PER • PER YEAR

Fig. 2. Net reproductiverate (•) as a function of Estimates of annual adult survival based on female fledging successand su•ival of subadults(Sz) resightingsof color-markedindividuals (Table from fledging to first breeding at 3 yr (2 yr for line 8) represent minimum values. We believe that A) and of adults thereafter (S•). See text for details. survival was substantiallyhigher than 64% in 1979 (seeDiscussion), and we consideraverage adult survivalto be approximately85% per year. For an adult just starting to breed, such a figure tality per year, R0 equals 13.3 (Fig. 2: line A). predicts a mean further expectedlife span of At the other extreme, if all eggs and young 6.15 yr (Brownie et al. 1978:39). Thus, depend- fledge but survival from fledging to first breed- ing on age at first breeding, which appearsto ing at 3 yr drops to 36%, and minimum adult be 2-3 yr for femalesand 3-4 yr for males,the survival thereafter is set at the observed mean averagelongevity of Long-billed Curlews may of 79% (Table 8), then Ro drops to 3.4 (Fig. 2: be 8-10 yr. line B). Demographicanalysis.--To evaluate the repro- By definition, R0 equals 1.0 in a stable pap- ductiveperformance of this population,annual ulatian. For this population of Long-billed fledging successrates were comparedwith a Curlews to be stable, with annual survival of range of valuesfor a stablepopulation that as- 79% for adults (mean observed for all years) sumemaximum possible (100%) and minimum and 36% for subadultsfrom fledging to 3 yr, observed(36%) survival from fledging until first females must fledge on average 0.59 female breeding. These values were obtained by plot- young per year (where line B crossesR0 = 1). ting net reproductiverates (R0, calculated after This did not occurin any year. Estimatedfledg- Ricklefs 1973: 408) as a function of fledging ing successper female in 1978 (0.17) was close successper female for different rates of sub- to the intercept of line A with R0 = 1, confirm- adult survival (Fig. 2). Adult mortality was as- ing that this was a year of poor productivity. sumed to be age-constant(Deevey 1947, Lack Dependingon actualvalues of subadultsurviv- 1954), and the maximum possible number of al and age at first breeding, the 1979 estimate female young fledged per female equals2.0 (or of averagefledging success (0.21 femaleyoung/ the number of females in the largest mean female) was probably marginal. In 1977, and clutch size, 4.0/2, assuminga primary sexratio especially among early-nesting females, esti- of 1:1). If all eggs and young survive to breed matedfledging successreached or exceeded0.38 at age 2 yr, and thereafterexperience 15% mar- femaleyoung. Suchproductivity may have been October1986] Long-billedCurlew Population Ecology 763

enough to offset the relatively poor successin We believe that clutch size actually was re- the following years.Mean fledging successper duced in 1979, especiallyamong early-nesting female pooled for all years was 0.25 female females, and that this was an adaptive re- young, which, assuminga stablepopulation size sponse.Unusually dense vegetation covered and 85% adult survival, predicts survival from mostbreeding territories early that season,and fledging to first breeding at 3 yr to be 58% (Fig. curlews flew considerable distances to find food 2: line C). (Redmond 1986). Clutch size and egg quality are associatedwith levels of protein reserves stored in the flight musclesof gulls (Houston DISCUSSION et al. 1983) and other birds (Jones and Ward 1976,Fogden and Fogden1979). If female Long- Breedingdensity.--The modified Finnish line- billed Curlews also rely on flight musclesfor transect method can be used to estimate den- protein storage, then strenuous use of these sities of territorial male Long-billed Curlews muscles before laying might limit protein re- only. Femalesare less detectable before hatch- servesand thereby affect reproductive effort. ing of their eggs,and consequentlytheir den- We suggestthat reduced clutch size was an al- sity cannot be estimatedreliably (Redmond et ternative to delayed laying and reduced egg al. 1981). Nevertheless, spot-mapping during size for some females in 1979 (see Redmond prelaying and incubation in 1978 and 1979 con- 1986). We also predict that intraspecificvaria- firmed that at least85% of malesacquired mates, tion in clutch size is more likely to be detected, and there was no evidence of a skewed, breed- at least among shorebirds,during seasonsof ing sex ratio (Redmond et al. 1981). marked environmental contrastthat may stress The gradual decline in male density from female reproductive physiology. 1978to 1981was likely real in spite of the large Reproductivesuccess.--The hatching success variances. During August 1981, a range fire of shorebirds is extremely variable among burned 142 ha of the study plot. Transectcounts species,and even within speciesit tends to vary during April 1982 were substantially greater, amongyears (see reviews by Goss-Custard1981, and estimatedbreeding density increasedal- Evans and Pienkowski 1984). Evidence for a most 30% from the previous year (Fig. 1). Ter- general decline in hatching successwith lati- ritorial males were especially common in the tude (as proposedby Ricklefs 1969) is mount- burned area during the spring of 1982, proba- ing for shorebirds(Pienkowski 1984a).Consis- bly becauseof reduced vegetative cover. Bicak tent with such a trend, the mean clutch et al. (1982) found a significant negative rela- mortality rate of Long-billed Curlews in west- tionship between vegetative cover and num- ern Idaho (0.0285/day) was significantlyhigher bers of breeding Long-billed Curlews on this than that of Whimbrels (Numeniusphaeopus) study area. nesting on the Canadian subarctic tundra Clutch size.--Significant annual variation in [0.0150/day,calculated from Skeel's(1983) data clutch size has never been reported among for all three habitats combined; G = 4.34, P < scolopacidshorebirds. At north temperate and 0.05]. arcticlatitudes nearly all shorebirds(Charadrii) Reports of clutch lossesbetween 20 and 50% lay clutchesof 4 eggs, and especiallywithin are common for shorebirds that nest at tem- Scolopacidae,intraspecific variation is slight perate latitudes. These figures, however, are (Maclean 1972).Yet in 1979,female Long-billed biasedin many studies(exceptions are Sordahl Curlews laid significantly more 3-egg clutches 1980, Page et al. 1983, Pienkowski 1984a,Lank than in other years. Becauseclutch size was es- et al. 1985)and underestimateactual mortality. timatedby the maximumnumber of eggsfound This is becausenests were found throughout in a nest, it is conceivable that many of these incubation, yet rates were not adjustedfor un- nestsin 1979 initially contained 4 eggs,but that known losses(Mayfield 1975, Johnson 1979). If 1 egg was lost from each nest before its being this bias is ignored for Long-billed Curlews, found. If this were so, one would expect single the percentageof clutcheslost each year was egg lossesin 1979 to continue after nestswere 40-45%, which, although still high, is more found and to be significantly greater than in comparableto data from theseother shorebirds other years. Single egg losseswere greatest, (see Cramp and Simmons 1982). however, in 1978 (Table 3). Survival of shorebird chicks is difficult to de- 764 REDMONDAND JENNI [Auk, Vol. 103

termine because of their precocial develop- amonglate-hatching than early-hatchingnests ment and nidifugoushabit. Estimatesof fledg- in 1978and 1979.This probably reflectsthe de- ing successare availablefor only a few species cline in numbersof grazing livestockthat oc- (Safriel 1975, Page et al. 1983, Pienkowski curred throughout May each season. 1984b).Chick mortality between hatching and Abnormally heavy spring rainfall in 1978 learning to feed is inferred to be high for many producedunusually lush vegetativecover (ver- species(Holmes 1966, Soikkeli 1967, Jehl 1973, tical density and biomass;Redmond 1986).The Hild•n 1978, Oring et al. 1983). Among Com- tall, thick vegetation probably reduced small mon Ringed Plovers breeding in northeastern mammal vulnerability to raptor predation, and England, 53% of all chick mortality occurred hindered the movements of Long-billed Cur- during the first week after hatching, and it was lew chicks,especially as they grew larger. If attributable to predation and starvation (Pien- chicks then favored more open areas, they kowski 1984b).Mortality of young Long-billed might have becomemore vulnerable to avian Curlews (age 0-5 days) was most often the re- predators. We recorded no losses of radio- suit of inadequate parental care (resulting in marked chicks to raptors before 14 June 1978, overheating,chilling, or predation),starvation, two days after most adults abandoned their or some physical inviability. broods (Redmond 1984). This emphasizes the In 1977, 5% (3/57) of nestlings handled importance of parental care to chick survival showed signs of incomplete yolk sacretention and suggeststhat the abnormally early depar- and adherence of eggshell fragments. All 3 ture of adults in 1978 facilitated a major shift chicks died within hours of hatching. There in diet by local raptors to include Long-billed was no evidence of similar inviability from 118 Curlew chicks. The causeof this adult depar- nestlings handled in 1978 and 1979. Incom- ture remainsunclear, but the tall, densevege- plete yolk sac retention and a poorly healed tation may have interfered with their ability to umbilicus are signs of either inadequateor ex- tend chicksor forage efficiently,or both. cessive warming of eggs during incubation Survivorshipand demographic analysis.--An im- (Rol'nik 1968). Adherence of chick embryosto portant assumption underlying estimates of portions of the eggshell occurs under condi- adult survival is that all marked individuals tions of insufficient humidity (Rol'nik 1968). alive in future years are equally likely to be The drought conditions in 1977 (Redmond resighted (Cormack 1964). Becauseno sex bias 1986) may have reduced humidities in Long- wasdetected in annualresightings of malesand billed Curlew nests, which in turn may have females (Redmond and Jenni 1982), we as- increased egg water loss (Ar and Rahn 1980, sumed approximately equal survival rates for Walsberg 1980). The responseof Long-billed both sexes.But females that were captured and Curlew embryosto dehydration is not known, subsequentlylost their clutch were less likely but yolk sacadherence to the outer membranes to be resighted than females that nested suc- and eggshell proper may have been a conse- cessfullyin their year of capture(Redmond and quence of excessiveembryonic water loss. Jenni 1982). The same was not true of males. Late-nestingpairs of Long-billedCurlews had Resightingsof females captured in 1978 were lower reproductive success(in terms of num- particularly low in 1979 (6/12; Redmond and ber of young fledged per adult) than early- Jenni 1982: table 1), but not because of poor nestingpairs (Table 4). This was due primarily hatching success.We suspectthat femaleswere to differencesin fledging successbetween the discouragedfrom nesting in this area oncethey two groups.Fledging success was greateramong encountered the hostile vegetative conditions chicks that hatched before the median date in presentduring March and April of 1979.Thus, all three years. In 1978 this difference was par- if female breeding dispersalrather than adult ticularly dramatic becauseof the intense pre- mortality produced the lower resightings in dation of radio-marked chicks by raptors late 1979, then the 64% survival estimate for that in the season.Although likely to exist, differ- year was too low. Given the estimatesof 84% encesin clutchsurvival between early- and late- and 89% for two separateyears, and the like- nesting pairs could not be evaluatedbecause lihood that some breeding dispersal was un- laying datesof many destroyednests were not detected in 1979, we consider 85% to be a rea- known. Egg survival, however, was higher sonable estimate of average annual adult October1986] Long-billedCurlew Population Ecology 765

survival. Comparativedata, however, suggest has important ecologicaland managementim- that even this figure might be too low for such plications for Long-billed Curlews or any a largeshorebird (Goss-Custard 1981, Evans and specieswith geographicallyisolated popula- Pienkowski 1984). tions and sex differencesin natal philopatty The life-table analysisalso assumes that both and dispersal(see Greenwood 1980). It means clutch size and adult mortality are age-con- that adequateproductivity, subadult survival, stant. The former is reasonablefor a species and adult survival in one population may not with a relatively fixed clutch size; the latter is be sufficientto ensure its growth or stability. widely accepted(Deevey 1947,Lack 1954,Rick- Clearly, the productivity of other populations lefs 1973) but difficult to evaluate (see Botkin that supply female recruits,as well as subadult and Miller 1974). Becauseannual resightings survival of these recruits, also must be consid- estimate minimum adult survival, calculations ered. In other words, the fates of geographi- of net reproductive rate (R0)based on such es- cally isolated populations may be intercon- timatesare necessarilyconservative. nected by strong, sex-biasednatal dispersal. The critical data missing from this analysis To conclude that this breeding population are annual survival ratesfor juveniles and sub- was stable, one must be relatively certain that adults. For this population to be stable in size local recruitment from marginal, surrounding with femalesproducing an averageof 0.25 fe- areaswas not responsiblefor the stable densi- male fledglings per year (the mean observed ties observedon the study plot. A method was for all years), survival from fledging to first devisedto monitor numbersof breedingadults breeding 33 months later must be 58%. This over about 70% of the entire study area, and no requiressurvival of about 84% of juveniles and major declineswere detectedbetween 1977 and subadultseach year. Mortality is likely to be 1983 (Redmond et al. unpubl. data). higher than this, however,at least for juvenile Productivity on the 1,600-ha study plot var- birds undertaking their first migration. Year- ied considerably among years. In 1977, a lings were never sighted on this breeding droughtyear, fledgling productionwas great- ground,and reportsof Long-billedCurlews re- est and probably sufficient to offset the poor maining on their winter range throughoutthe productivity associatedwith abnormally wet year (McCaskie 1970, Jurek 1974) suggestthat conditions the next year. But the abnormal yearlings, and perhaps some 2-year-olds,may rainfall in the springof 1978had an additional not attempta northwardmigration. In this case, effect on Long-billed Curlew reproduction in subadultsurvival approaching 95% on the win- 1979,mediated by the abundant, standingdead tering grounds would be sufficient to offset vegetationpresent during prelaying. Thus, the mortality of up to 36%of juvenilesduring their population faced unusual environmental con- first migration and still provide for 58% return- ditions each year we monitored its productiv- ing to attempt breeding as 3-year-olds. These ity. If, in years with more normal rainfall and figuresare indeedplausible considering that in vegetativecover (see Redmond 1986), fledgling Britain 50% of Eurasian Curlews banded as production varies between the values estimat- chicks survived to 1 year of age when the ed for 1979 and 1977, and if our estimates of specieswas legal quarry (Bainbridgeand Min- survival after fledging are realistic, then this ton 1978). Other limited data indicate that population should remain stable within limits 2-year-oldEurasian Curlews survivebetter than set by habitat quality and density-dependent adults, perhaps at a rate well over 90% (Evans mortality factors(Lack 1966). and Pienkowski 1984). The fact that population surveysin 1980 and Even if 58% representsa reasonableestimate 1981 did not reflect the poor fledgling produc- of average subadult survival for Long-billed tion in 1978 is no doubt a function of the Curlews,a further complicationremains. Males species'longevity. Populationsof longer-lived were more likely to return and breed in their organisms,because they are made up of indi- natal area than were females (Table 7), sug- viduals that vary widely in age, are less vul- gestingthat femalesdisperse before first breed- nerable to short-term fluctuations. Yet this nu- ing (Redmond and Jenni 1982). The relative merical inertia makes population declines isolation of this breeding population suggests difficult to detect until they are well underway that femalesmay disperselong distances.This unlessage structurecan be determined. 766 REDMONDAND JENNI [Auk, Vol. 103

ACKNOWLEDGMENTS FOGDEN, M.P. L., & P.M. FOGDEN. 1979. The role of fat and protein reservesin the annual cycle We thank Kevin Barth, Pedro Beraun, Tom Bicak, of the Grey-backed Camaroptera in Uganda Jim Eastman,Rodolpho Heredia, Leland Hersh, John (Aves:Sylvidae). J. Zool. London 189:233-258. Parker, Alan Sands, Michael Shantz, and Chuck GOss-CUSTARD,J. D. 1981. Role of winter food sup- Weichler for their invaluable help with fieldwork. plies in the populationecology of commonBrit- Drs. D. L. Kilgore, Jr., R. McKelvey, A. L. Sheldon, ish wading birds. Verh. Ornithol. Ges. Bayern P. L. Wright, L. Zwarts,and an anonymousreviewer 23: 125-146. provided helpful commentson earlier versionsof the , S. E. A. DURELL,H. P. SITTERS,& R. SWINFERN. manuscript.DeWayne Williams preparedthe figures. 1982. Age-structureand survival of a wintering This researchwas funded by the U.S. Department of population of Oystercatchers. Study 29: 83- Interior, Bureauof Land Management(BLM) through 98. contracts YA-512-CT7-54 to DAJ and ID-010-CT1-0015 GREENWOOD,P.J. 1980. Mating systems,philopatry, to RLR. We are particularly grateful for the cooper- and dispersal in birds and mammals. Anim. Be- ation and encouragementof the following personnel hav. 28: 1140-1162. at the BLM, Boise District Office: A1 Bammann, Mike GROSSKOPF,G. 1959. Zur Biologiedes Rotschenkels Kochert, Mike Rath, Bud Sherrets,A1 Tripp, and es- (Tringat. totanus),II. J. Ornithol. 100: 210-236. pecially Alan Sands.This paper representsa portion 1964. Sterblichkeit und durchschnittsalter of a dissertationpresented at the University of Mon- einiger K/istenv6gel.J. Ornithol. 105: 427-449. tana by RLR. HILD•N, O. 1978. Population dynamics of Tem- LITERATURE CITED minck's Stint temrninckii. Oikos 30: 17-28. HOLLAND, P. K., J. E. ROBSON,& D. W. YALDEN. 1982. AR, A., & H. RAHN. 1980. Water in the avian egg: The breedingbiology of the CommonSandpiper overallbudget of incubation.Amer. Zool. 20: 373- Actitishypoleucos in the Peak District. Bird Study 384. 29: 99-110. BAINBRIDGE,I. B., & C. D. T. MINTON. 1978. The mi- HOLMES,R. T. 1966. Breeding ecology and annual gration and mortality of the Curlew in Britain cycle adaptationsof the Red-backedSandpiper and Ireland. Bird Study 25: 39-50. (Calidrisalpina) in northern Alaska. Condor 63: BICAK, T. K., R. L. REDMOND,& D. A. JENNI. 1982. 3-46. Effectsof grazing on Long-billed Curlew (Nu- HOUSTON, D.C., P. J. JONES,& R. M. SIBLEY. 1983. rneniusamericanus) breeding behavior and ecol- The effectof female body conditionon egg lay- ogy in southwestIdaho. Pp. 74-85 in Wildlife- ing in LesserBlack-backed Gulls Larusfuscus. J. livestockrelationships symposium: proceedings Zool. London 200: 509-520. 10 (J. M. Peek and P. D. Dalke, Eds.). Moscow, J•RVlNEN,O., & R. A. V•IS•NEN. 1975. Estimating Idaho, Univ. Idaho, Forest,Wildl. and RangeExp. relative densitiesof breeding birds by the line- Sta. transect method. Oikos 26: 316-322. BOTKIN,D. B., & R. S. MILLER.1974. Mortality rates JEHL,J. R., JR. 1973. Breedingbiology and system- and survival of birds. Amer. Natur. 108: 181-192. atic relationshipsof the Stilt Sandpiper. Wilson BROWNIE, C., D. g. ANDERSON, K. P. BURNHAM, & D. Bull. 85: 115-147. S. ROBSON. 1978. Statistical inference from band JOI-INSGARD,P. 1981. The plovers, sandpipers,and recoverydata--a handbook.U.S. FishWildl. Serv. snipes of the world. Lincoln, Univ. Nebraska Res. Publ. 131. Press. BUB,H. 1962. Planberingungenam Sandregenpfei- JOHNSON,D.H. 1979. Estimating nest success:the fer (Charadriushiaticula). J. OrnithoL 103: 243- Mayfieldmethod and an alternative.Auk 96:651- 249. 661. CORMACK, g. g. 1964. Estimates of survival from JONES,P. J., & P. WARD. 1976. The level of reserve the sighting of marked . Biometrika 51: protein as the proximate factor controlling the 429-438. timing of breeding and clutch size in the Red- CRAMP,S., & K. E. L. SIMMONS(Eds.). 1982. The birds billed Quelea Queleaquelea. Ibis 118: 547-574. of the western Palearctic, vol. III. Oxford, Oxford JUREK,R. M. 1974. California shorebird study: ac- Univ. Press. celeratedresearch program for shoreand upland DEEVEY,E. S., JR. 1947. Life tables for natural pop- migratory gamebirds. Final Rept., California ulations of animals. Quart. Rev. Biol. 22: 283- Dept. Fish & Game, Sacramento,California. 314. KNIGHT,J. E. 1983. Skunks. Pp. 81-86 in Prevention EVANS,P. R., & M. W. PIENKOWSKI.1984. Population and control of wildlife damage(R. M. Timm, Ed.). dynamicsof shorebirds.Pp. 83-123 in Shore- Lincoln, Nebraska,Great Plains Agr. Council and birds: breeding behavior and populations (J. Nebraska Coop. Ext. Serv. Burger and B. L. Olla, Eds.).New York, Plenum LACK, D. 1954. The natural control of num- Press. bers. Oxford, Clarendon Press. October1986] Long-billedCurlew Population Ecology 767

1966. Population studiesof birds. Oxford, 1986. Egg size and laying date of Long-billed Clarendon Press. Cur!ewE (Numeniusamericanus): implications for LANK, D. B., L. W. ORING, & 5. J. MAXSON. 1985. female reproductivetactics. Oikos 46: 330-338. Mate and nutrient limitation of egg laying in a , T. K. BICAK,& D. A. JENNI. 1981. An eval- polyandrousshorebird. Ecology 66: 1513-1524. uation of breedingseason census techniques for LIVENßI-•. 1940. Beitr•ge zur Biologiedes Sandre- Long-bil!edCurlews (Numeniusamericanus). Stud. genpfeifers (Charadriushiaticula L.). J. Ornithol. Avian Bio!. 6: 197-201. 88: 183-287. , & D. A. JENN1.1982. Natal philopatry and LEHNER, P. N., R. KRUMM, & A. T. CR1NGAN. 1976. breeding area fide!ity of Long-bi!!ed Curlews Testsfor olfactoryrepellants for coyotesand dogs. (Numeniusamericanus): patterns and evo!utionary J. Wildl. Mgmt. 40: 145-150. consequences.Behav. Ecol. Sociobiol.10: 277-279. LUNDBERG,C., & R. A. VXISXNEN. 1979. Selective RICKLEES,R. E. 1969. An ana!ysisof nesting mortal- correlationof egg size with chick mortality in ity in birds. Smithsonian Contrib. Zool. 9: 1-48. the Black-backed Gull. Condor 81: 146-156. 1973. Fecundityßmorta!ity, and avian de- MACLEAN, G. L. 1972. Clutch size and evolution in mography. Pp. 366-435 in Breeding biology of the Charadrii. Auk 89: 299-324. birds (D. S. Farner, Ed.). Washington, D.C.ßNatl. MAYFIELD,I-•. F. 1975. Suggestionsfor calculating Icad. Sci. nest success. Wilson Bull. 87: 456-466. ROL'NIK,V. V. 1968. Bird embryo!ogy. Jerusa!em, McCASKIS, G. 1970. Shorebird and waterbird use of Israel Program for Sci. Transl. the Salton Sea. California Fish Game 56: 87-95. $AFRIEL,U. N. 1975. On the significanceof clutch MYERSßJ.P. 1981. Cross-seasonal interactions in the size in nidifugous birds. Ecology56: 703-708. evolution of sandpipersocial systems.Behav. ß M.P. HARRIS, M. DE L. BROOKE, •r C. K. Ecol. Sociobiol. 8: 195-202. BRITTON. 1984. Surviva! of breeding Oyster- O'MEARA,D.C., •r J. F. WITTER.1971. Aspergillosis. catchersHaematopus ostralegus. J. Anim. Ecol. 53: Pp. 153-162in Infectiousand parasiticdiseases 867-877. of wild birds (J. W. Davis, R. C. Anderson, L. $KEEL,M. A. 1983. Nesting success,density, philo- Karstad, and D. O. Trainer, Eds.). Amesß Iowa patry, and nest-siteselection of the Whimbre! State Univ. Press. (Numeniusphaeopus) in different habitats.Can. J. ORING, L. W., D. B. LANK, & 5. J. MAXSON. 1983. Zool. 61: 218-225. Populationstudies of the polyandrousSpotted S¸IKKELI,M. 1967. Breedingcycle and population Sandpiper.Auk 100: 272-285. dynamics in the Dunlin (Calidris alpina). Ann. PAGE,G. W., L. E. $TENZEL,D. W. W1NKLER,•r C. W. Zoo!. Fennica 4: 158-198. SWAr•TH. 1983. Spacing out at Mono Lake: 1970a. Dispersalof the Dunlin in relation breedingsuccess, nest density, and predationin to sitesof birth and breeding. Ornis Fennica47: the Snowy Plover. Auk 100: 13-24. 1-9. PIENKOWSKI,M. W. 1984a. Breeding biology and 1970b. Morta!ity and reproductive rates in populationdynamics of RingedPlovers Charadri- a Finnish popu!ation of Dunlin Calidrisalpina. us hiaticulain Britain and Greenland:nest pre- Ornis Fennica 47: 149-158. dation as a possiblefactor limiting distribution S¸r•OAHI,,T.A. 1980. Antipredatorbehavior and pa- and timing of breeding.J. Zool. London202: 83- rental care in the American Avocet and Black- 114. necked Stilt (Aves: Recurvirostridae).Unpub- 1984b. Behaviorof young Ringed Plovers lishedPh.D. dissertationßLogan, Utah StateUniv. Charadriushiaticula and its relationshipto growth TREMAINE, M. N. 1975. Life at an avocet nest. Au- and survival to reproductiveage. Ibis 126: 133- dubon 77: 71. 155. WALSBERG,G. E. 1980. The gaseousmicroclimate of PITELKA,F. A., R. T. HOLMES,•r 5. F. MACLEANßJR. the avian nest during incubation. Amer. Zool. 1974. Ecologyand evolution of socialorgani- 20: 363-372. zation in arctic sandpipers.Amer. Zool. 14: 185- WILL•S,E. O. 1981. Precautionsin ca!culatingnest- 204. ing success.Ibis 123: 204-207. REDMONDßR. L. 1984. The behaviora!ecology of YATES,B. J. 1982. Population biology of breeding Long-billedCurlews breeding in westernIdaho. RedshanksTringa totanus.Unpublished Ph.D. Unpublished Ph.D. dissertation,Missoula, Univ. dissertationßLiverpool, United KingdomßLiver- Montana. pool Polytechnic.