Fish Crow Predation on Eggs of the White Ibis at Battery Island, North Carolina

FISH CROW PREDATION ON EGGS OF THE WHITE IBIS AT BATTERY ISLAND, NORTH CAROLINA

MARK A. SHIELDS• AND JAMESF. PARNELL Departmentof BiologicalSciences, University of North Carolina, Wilmington,North Carolina28403 USA

ABSTRACT.--Westudied predation by Fish Crows (Corvusossifragus) on eggsof the White Ibis (Eudocimusalbus) during the 1983 and 1984nesting seasonsat BatteryIsland, southeastern North Carolina. Crow predation accountedfor the lossof 32% (n = 223) of ibis eggsin 1983 and 44% (n = 538) in 1984.Crows usually took all eggsin a clutch. An estimated6 pairs of Fish Crows nested on the island each year. We believe these individuals were responsible for mostegg loss.The predation rate of ibis clutcheswas highest in plots nearestcrow nests and lowest in two plots that containedobservation blinds. Resultsof experimentsusing simulatedibis nestssuggested that crowswere wary of the blinds. Predationdeclined with nest age,apparently due to increasednest attentivenessby adult ibisesduring the last week of incubation.The overall predation rate in 1984 was significantly higher than in 1983. Greaternest densities and lesssynchronous breeding by ibisesin 1984may have contributed to the higher predationrate. Ibis productivitywas estimatedat 1.22-1.30fledglings per pair in 1983and 1.05-1.12 in 1984.This level of reproductionappeared sufficient for maintenance of the population.Thus, egg predationby Fish Crows during our study did not appearto be a seriousthreat to the productivityof this White Ibis population.Received 24 June1985, accepted22 January1986.

CROWS(Corvus spp.) are well known as pred- The White Ibis (Eudocimusalbus) reaches its ators of birds' eggs.Upon finding an unguard- northern limit of regular breeding in coastal ed nest, a crow typically flies off with an egg, North Carolina (A.O.U. 1983). White Ibis eats or caches it, and then returns to the same breedingbiology was studiedby Allen-Grimes area to stealanother egg (Tinbergen et al. 1967, (1982)at BatteryIsland, site of the largestWhite Croze 1970, Montevecchi 1976). Because crows Ibis colony in North Carolina. She found nest- often return to sites of previous prey capture, ing successto be significantly lower than in their predation successincreases with increas- Florida populations (Rudegeair 1975, Kushlan ing prey density (Tinbergen et al. 1967, Gfr- 1977) and attributed this, in part, to high egg ansson et al. 1975, Montevecchi 1977). Thus, loss.Allen-Grimes frequently saw Fish Crows predation on eggs of colonial nesters,particu- (C. ossifragus),which alsonested on the island, larly wading birds (Ciconiiformes)that do not carryingoff eggs.She suggestedthat crow pre- exhibit group-mobbingantipredator behavior, dation may have been a major reasonfor low may be severe(Krebs 1978).Crow predation on nesting successof this ibis population. We ini- eggsof wading birds is documentedthorough- tiated our study specifically to examine Fish ly (e.g. Bent 1926, Baker 1940, Meanley 1955, Crow predation on ibis eggsat Battery Island. Dusi and Dusi 1968,Rudegeair 1975, Burger and Our objectiveswere to determine the extent Hahn 1977, Maxwell and Kale 1977, Tremblay of Fish Crow predation on White Ibis eggs, to and Ellison 1979, Allen-Grimes 1982, Frederick examine temporal and spatial patterns of pre- 1985). We can find no detailed studies, how- dation, and to evaluatethe impact of predation ever, of the extentof egg predationduring more on ibis productivity. than one breeding season,factors affecting the rate of predation, or the impact of this egg loss STUDY AREA AND METHODS on wading bird populations. We conductedthe studyfrom April to August1983 and 1984 at Battery Island (33ø54'N,78ø01'W), a Na- • Present address:Wyoming Cooperative Fishery tional Audubon Societysanctuary located in the Cape and Wildlife ResearchUnit, Box3166, University Sta- Fear River estuary !km southeastof Southport, tion, Laramie, Wyoming 82071 USA. Brunswick Co., North Carolina. This island has sup-

531 The Auk 103:531-539. July 1986 532 S•ELr•sAND PARNELL [Auk,Vol. 103 ported a mixed-speciesheronry since at least 1938 We usually visited nestsevery 3-4 days to record (Brimley 1938).Recently it has held the largesther- contents.To minimize colony disturbanceand ther- onry in North Carolina (Parnell and Soots1979, Par- mal stressto eggs and young, visits were made as nell and McCrimmon 1984) and the northernmost quicklyas possible (<30 min/plot) during 0700-1200. large (>100 pairs) breeding colony of White Ibises If eggswere missing,we searchedthe area immedi- in North America (Allen-Grimes 1982, Shields and ately surroundingthe nest for the presenceof egg- Parnell 1983). shells. In addition, we regularly collected all egg- Comprisingan area of about40 ha, BatteryIsland shells and shell fragments within each plot. We is mostlysalt marsh dominated by Spartinaalternifiora. recordeda predation lossif an egg was punctured in Two woodeduplands, North (1 ha) and South(7 ha) the manner characteristicof crows (Rearden 1951) or colonies,provide nesting habitat for the large wad- if the disappearanceof an eggcould not be attributed ing bird assemblage.White Ibises nest only in the to the egg hatchingor falling from the nest (Burger South Colony. and Hahn 1977, Montevecchi 1977, Gottfried and The South Colony site, createdby depositionof Thompson1978, Miller and Burger 1978). dredgedmaterial (Funderburg 1960), is a grass/forb- We computeda daily predationrate (DPR) by di- covered dome fringed by a maritime thicket. The viding the numberof clutchespredated (partial clutch thicket vegetationis composedof red cedar(Juniperus losswas small) by the amount of nest exposure(in virginiana),yaupon (Ilex vomitoria),Hercules'-club nest-days)(Mayfield 1961, 1975). The data were (Zanthoxylumclava-herculis), wax myrtle (Myrica ceri- groupedby plot into 1-weekperiods, with the first fera),groundsel-tree (Baccharis halimifolia), marsh eld- week beginning on the date of our first visit, to ex- er (Ira frutescens),and severalother woody species. aminespatial and temporalpatterns in DPR eachyear. Trees and shrubsare also scatteredin clumps across We comparedDPRs between two groupsof nestsus- the dome. Woody vegetationon the dome is domi- ing the variance estimator and statistical test de- nated by red cedar, with lessernumbers of yaupon, scribedby Hensler and Nichols (1981). When more wax myrtle, Hercules'-club,cherry (Prunussp.), red than two groupswere compared,we employed the mulberry (Morusrubra), and buckthorn (Bumelialy- Bonferroni multiple comparison method (Miller cioides). 1966). In 1983 the South Colony containedan estimated We estimatedibis nesting successusing the May- 3,737 White Ibis nestsand 637 nestsof eight other field methodto evaluatethe impactof egg predation wading bird species(Casmerodius albus; Bubulcus •bis; on ibis productivity. We calculatednest successsep- Egrettathula; E. caerulea;E. tricolor;Butorides striatus; arately for the egg and nestling stages.We defined Black-crowned Night-Heron, Nycticorax nycticorax; the egg stageas the period from the day the first egg and Plegadisfalcinellus). The SouthColony held 4,849 in a clutch was laid to the day before hatching of the White Ibis nests and 852 nests of the other eight first egg. The nestling stageextended from the day speciesin 1984.At least 2 pairs of Fish Crows nested the first egg hatched to 10 days after that date, at in the South Colony in 1983. We located 6 nestsdur- which time nestling ibisesbegan to leave their nests ing extensivesearches in 1984. Other potential egg upon our approachand no longer could be associated predators in the South Colony included Black- with specificnests (cf. Custeret al. 1983). crowned Night-Herons, Boat-tailed Grackles (Quis- If the first clutchis destroyed,many wading birds caiusmajor), Norway rats (Rattus norvegicus), and snakes will lay a replacementclutch, often using the origi- (Elapheobsoleta quadrivittata and E.g. guttata).No large nal nest (Jenni 1969, Milstein et al. 1970, Maxwell mammals inhabited the island. and Kale 1977).Failure to take renestinginto account We determined the fates of White Ibis nests in ten will result in an underestimate of nesting success 15 x 15 m plots selectedrandomly from a pre-estab- (Custer and Pitelka 1977). We estimated the extent of lished grid. We used the same plots in both years; renesting by ibises to reduce this bias. Becauseun- White Ibisesdid not nest in one plot (R15) in 1984, attendedWhite Ibis nestsare quickly (often within however, and another (J28)was chosento replaceit. one day) dismantled by neighboring birds (Rude- White Ibis nests comprisedover 95% of all wading geair 1975,Shields pers. obs.), we considereda clutch bird nestsin the plots used each year. We attempted to be a replacementif the nest was not dismantled to individually mark all ibis nests and eggs in each between the time of orginal egg lossand subsequent plot. The highest nests in most plots could not be laying (Schreiber1979). reachedeasily without endangeringlower nestsand thereforewere excludedfrom study.Unmarked nests RESULTS AND DISCUSSION accountedfor about 15%of all ibis nestsin the study plots. We marked nestswith inconspicuousplastic We individually marked 694 eggs in 262 tags wired beneath the nest bowl. Eggs were num- clutchesin 1983and 1,213eggs in 493 clutches bered on both ends using waterproofIndia ink. We in 1984. We marked 98% of the nests during did not mark empty nests. the first week of incubation. July1986] CrowPredation onIbis Eggs 533

TABLE1. Fatesof White Ibis eggsin sampleclutches our visits. Based on the 6 crow nests found in at Battery Island. the South Colony in 1984,we assumedthat the

1983 1984 residentpopulation was composed of 12adults. The 2 additional crows may have been nest Fate n % n % helpers(McNair 1985) or nonresidentcrows. Hatched 410 59.0 596 49.1 We frequentlysaw 2-3 crowsperched together Predated 223 32.1 538 44.3 near two known crow nests, but rarely saw Survived incubation, but did not hatch 43 6.2 59 4.9 crows flying between the South Colony and Abandoned 8 1.2 4 0.3 the mainland (a distance of I km) or between Unknown a 8 1.2 0 0.0 the South and North colonies (a distance of 0.5 Dump eggb 2 0.3 8 0.7 km). Fish Crow pairs tolerate helpers at nest Other c 0 0.0 8 0.7 sites and food caches,but vigorously defend Total 694 100.0 1,213 100.0 these areas against intrusion by other Fish ' The outcome of 3 clutches in 1983 could not be Crows (McNair 1985). We therefore believe that determined. the major egg predatorswere resident pairs, bEgg laid in another ibis's nest. All dumped eggs someperhaps with helpers,and that territorial were laid well after hosts' clutches were completed; behavior limited the number of nonresidents no dumpedeggs hatched due to inadequateincuba- tion. foraging on the island. cIncludes 6 eggscracked or dented, 1 egg dropped Spatialpatterns of predation.--Theoverall DPR during handling, and I runt egg in 1984. of ibis clutches varied considerably among study plots in each year (Table 2). However, Extentand sources of predation.--Predationwas trends in relative predation intensity among the mostimportant source of egg mortality and plots were quite similar in both years (Spear- accountedfor the loss of 223 (32.1%) eggs in man's rank correlation, r = 0.78, n = 9, P < 1983and 538 (44.3%)eggs in 1984(Table 1). We 0.01) despitethe higher DPRs in 1984 (Table believe that Fish Crows were responsiblefor 2). For example, plots K23, M21, and FI6 had most,if not all, egg predation.Other potential the lowest DPRs in both years, while plots KI5 avian predators were common in the colony, and N24 had the highest. This similarity be- but during > 150 h of observationlogged over tween yearssuggests that the intensity of pre- three nesting seasons(1982-1984) we saw only dation was related to some characteristic(s) of Fish Crows preying on ibis eggs.We observed the plots that remained relatively stable over only one snakeeach year. The effectof Norway both years. rats is unknown. Nesting habitat did not appear to be related Crowstook eggsfrom I00 (38.2%)clutches in to predation intensity as the three plots with 1983and 244 (49.5%)clutches in 1984.All eggs the lowest DPRs in both yearshad quite differ- were removed from 89 (89.0%) clutches in 1983 ent vegetative structures:M21 consistedof a and 231 (94.7%) in 1984. Lossof all eggs in a low (<2 m), densegrowth of yaupon;K23 was clt:tch is typical of crow predation (Rearden made up of a 3.4-m-high red cedar and a 2.4- 1951).Fish Crow predationaccounted for 97.3% m-high Hercules'-club;and FI6 was composed of all clutch losses in 1983 and 99.1% in 1984. of a tall (6.5 m) red cedar with an understory Estimatedpredation loss of ibis eggsin 1981 of yaupon and buckthorn.Plots KI5 and N24 (calculated from Allen-Grimes 1982) was 5% consisted of tall (>5 m) cedars similar to FI6, lower than in 1983 and 17% lower than in 1984. yet these two plots had the highest DPRs in Predation rates at Battery Island were compa- both years (Table 2). The degree of nest con- rable to thoseat some other wading bird colo- cealmentdid not differ significantlybetween nies where Fish Crowsalso nested (e.g. Burger predated and nonpredated clutches (Shields and Hahn 1977, Miller and Burger 1978), but 1985). were more than double the rate of crow pre- The location of a plot within the colony, dation at another White Ibis colony in which however, was related to predation intensity. In crows did not nest (Frederick 1985). general,overall DPR declined with increasing Resident crows appeared to be the primary distancefrom a Fish Crow nest (Fig. I). Despite predators.We observeda maximum of 14 adult weekly variation, the DPRs in KI5, N24, and crows in the South Colony each year during H11 (the three plots nearestcrow nestsin 1984) 534 SHIELDSAND PARNELL [Auk, Vol. 103

TABLE2. Daily predationrate (DPR) of White Ibis clutchesby studyplot and year. n = amountof nest exposure(in nest-days)during the egg stage.

1983 1984

Plot DPR + SD n DPR + SD n

K23 0.0000 ñ 0.0698 205 0.0268 ñ 0.0062 672 M21 0.0043 ñ 0.0030 466 0.0104 ñ 0.0039 674 F16 0.0057 ñ 0.0033 529 0.0262 ñ 0.0055 839 Hll 0.0144ñ 0.0059 415 0.0375 ñ 0.0080 560 I28 0.0173 ñ 0.0070 347 0.0281 ñ 0.0064 676 R15 0.0234 ñ 0.0087 299 ------El3 0.0247ñ 0.0065 567 0.0318 ñ 0.0074 566 K27 0.0322 ñ 0.0112 248 0.0271 ñ 0.0109 221 K15 0.0403 ñ 0.0073 720 0.0715 ñ 0.0085 909 N24 0.0478 ñ 0.0125 293 0.0539 ñ 0.0096 557 J28 ------0.0326 ñ 0.0064 765 Overall 0.0217 ñ 0.0023 4,089 0.0358 ñ 0.0023 6,439

were consistentlyhigher than the rates in the highest overall DPR in 1984, while El3 was one other sevenplots (Fig. 2). A similar pattern was of the three plots with consistentlyhigh DPRs observed in 1983, with K15, N24, and E13 con- throughout the 1983 breeding season.We be- sistently having the highest DPRs (Fig. 2). lieve that nesting by crows in the same terri- Corvids often obtain much of their food from tories,but not necessarilythe same nest sites, within their territories (Goodwin 1976), and in successiveyears (see Butler et al. 1984) ac- several authors (Jonesand Hungerford 1972, countedfor the similarity of predation patterns Loman and G6ransson1978, Erikstad et al. 1982) among plots in 1983 and 1984. similarly reported decreasingrates of nest pre- The DPRs in plots K23 and M21 were lower dation as distance from corvid nests increased. in 1984 than might be expected based on the Such a spatial pattern of predation may be ex- proximity of these plots to crow nests(Fig. 1). pected becauseit would be less energetically These plots also had the lowest DPRs in 1983 expensive for crows to forage as close as pos- (Table 2), when only two clutcheswere pre- sible to their nests(Loman and G6ransson1978). dated in both plots combined. The presenceof Although we did not obtain complete data a small cloth-covered observation blind in each on Fish Crow nest locations in 1983, we believe plot may have contributed to the lower than that crowsoccupied the sameterritories in 1983 expectedDPRs. Preliminary studiesconducted and 1984. Northwestern Crows (C. caurinus) in May 1982 indicated that Fish Crows were maintain one or more feeding stations within wary of the blind in M21. We placed 10 simu- their territories and use the samesites year af- lated ibis nests, each containing two chicken ter year (Verbeek 1982). We found a Fish Crow eggscolored to resembleWhite Ibis eggs,with- feeding stationat the study site about 5 m from in this plot. The nestssurvived three dayswith K15 (which contained a crow nest in 1984) in only one loss,while 12 similar nestsplaced in the 1982-1984 breeding seasons.This suggests areas of the thicket where blinds were absent that a FishCrow territory alsoencompassed plot were predated within 24 h. We observedcrows K15 in 1982 and 1983. In 1984 we found a crow from the blind in M21 for 9 h on the first day nest about 20 m from the site of the one known of the experiment. Crows frequently flew over crow nest in 1983. In 1983 we observed a re- the plot and perched in nearby trees but made cently fledged crow from an undiscoverednest no attempt to land near the nests,all of which perched in the same cedar in which the nest were 5-10 m from the blind and in full view closest to N24 was located in 1984. We fre- of the crows. Montevecchi (1976) also reported quently saw several adult crows perched in this that Fish Crowswere wary of a blind, although tree during the 1983breeding season. Two crow he felt that crows were wary becauseof his nests were located within 30 m of Hll and presencein the blind. The lossof only one nest within 50 m of El3 in 1984. Hll had the third in three days in our experiment suggeststhat July1986] CrowPredation on Ibis Eggs 535

ILl 0.08 '1 I• ß K15 0.06 14•,• 1983

0.06 Z ,,,0.04 ß N24 <• 19 u \ 1 O 11 < 0.04 2 1 Hll ß E13 J28 ß K27 ß O ß ß ß O 12 19 26 3 10 17 24 31 7 K23 F16 i28 >. 0.02 < 0.16 25**

o M21 • 0.12 1984

20 40 60 80 • 0.08 **** DISTANCE TO NEAREST CROW NEST (M) Fig. 1. Daily predation rate of White Ibis clutches in relation to the distancefrom the study plots to the nearestFish Crow nest in 1984. Open circlesdenote plots that contained observation blinds. APRIL MAY JUNE Fig. 2. Weekly variation in the daily predation the crows may have learned to associatethe rate of White Ibis clutches.Open circles denote the blind with human presenceand therefore ex- 3 study plots nearest Fish Crow nests;solid circles hibited cautionwhether or not we were pres- denotethe other 7 plots.The numberof clutchespre- ent. dated each week in both groups of plots is shown. Temporalpatterns of predation.--Throughout Significancelevels for comparisonsof the DPRs be- the breedingseason DPR remainedhigher in •een the two groupsof plotsby week are indicated the three plots closestto crow neststhan in the by asterisks:• 0.05 < P < 0.10, ** P < 0.05, *** P < other sevenplots (Fig. 2). DPR increasedfrom 0.005, **** P < 0.0001. the firstto secondweeks in both groupsof plots in both years.Egg predation might be expected remainedrelatively high through the first 4-5 to be lower in the first week of study than in weeks in both years (Fig. 2). The DPRs in the the secondwhile the crowsdevelop a search- otherplots, however, declined sharply after the ing image for new prey items (Croze 1970). secondweek, althoughpredation in theseplots However, we do not believe the development in 1984was still fairly high through the third of an egg-specificsearching image by Fish week (Fig. 2). The declinesin DPRs in the two Crowscan completely explain the dramaticpeak groupsof plots cameduring or just before the in DPR during the secondweek in 1984 (Fig. week of peakhatching of clutchesin both years 2). Weatherduring this periodwas cloudy and (Shields 1985). White Ibisesappeared to flush windy. Our observationsduring the two years less readily and to return to their nests more of study indicated that ibises tended to flush quickly late in incubation and when small from their nestsmore readily and with greater young were present.For example,early in in- frequency,even in the absenceof human dis- cubation all adults flushed from their nests as turbance,during windy conditionsthan when we approachedto within 10-15 m of a plot. the air was calm. Such behavior would result Towardhatching and when small (< 10-day-old) in increasedexposure of eggsto predatorsand young were in nests, adults usually did not may account for the high rate of predation. flush until we were within 5 m, and often sev- Windy conditionsfacilitated American Crow (C. eral adultsremained perched in the topsof trees brachyrhynchos)predation on Double-crested in which we were working (see also Skutch Cormorant(Phalacrocorax auritus) eggs, by caus- 1962, Ellison and Cleary 1978). This increased ing cormorantsto readily leave their nests(El- nest attentivenessprobably accounts for the lison and Cleary 1978). rapidly declining DPRsnear the period of peak The DPR in the three plots nearestcrow nests hatching. 536 SHIELDSAND PARNELL [Auk,Vol. 103

To examine quantitatively the relationship 50 •'2 betweenage of nestand predation,we divided • 61 1983 the egg stageinto three 7-day age groups and calculatedthe DPR for each group. We estimated the date of initiation of each clutch by u• 3o back-dating,allowing two days for each egg laid (Rudegeair 1975, Kushlan 1977). In 1983 dutches in the 15-21-day-old group had a • 10 '• lower overall DPR (0.0120 + 0.0030) than ,, clutches in either the 1-7-day-old group 0 5 12' 19 26 3 10 17 2'4 3'1 (0.0258 + 0.0044) or the 8-14-day-old group 0 52 (0.0270 + 0.0042), although the differences Z 30 85 among groupswere not significantstatistically (P > 0.05). In 1984, 15-21-day-old clutches mO20 4t I:i:]•I:• 2• had a significantly(P < 0.05) lower overall DPR • 10 (0.0096 + 0.0022) than 1-7-day-old clutches (0.0506 + 0.0047) and 8-14-day-old clutches (0.0434 + 0.0042). This supports the observa- 4 11' 18 25 2 9 16 23 30 tion that, as hatching nears, adult ibises be- APRIL MAY come more attentive and thereby reduce egg Fig. 3. Frequen• distributionof White Ibis clutch losses to Fish Crows. initiations by week. Fi•t nests and renests in the 3 Breedingextended later into the seasonin the study plots nearest Fish Crow nestsare denoted by three plots nearestcrow nestsin both yearsbe- hatched and cross-•tched ba•, respectively; first causehigh initial levels of predationresulted nestsand renestsin the other 7 plots are denoted by in few early clutchessurviving to hatch and in openand doRedba•, respectively.The number above the laying of many replacementclutches. In each bar is the number of clutches initiated. The dates addition, most late-arriving ibisesnested with- of initiation of 3 clutches in 1984 could not be deter- in these three plots in 1983 (Fig. 3), perhaps mined. Asterisks denote the dates of our first visits attractedby the availabilityof abandonednest to the plots. Mean ß SD age of clutchespresent on our first visit was 2.0 ß 0.8 days (n = 88) in 1983 and sites or the activities of tenesters, or both. Thus, 2.6 ß 1.3 days (n = 64) in 1984. many clutchesin the more vulnerable early stagesof incubationwere availablethroughout the first 5-6 weeks of each season,and preda- maximum nest densities in all but one of our tion ratesremained high. Lower levelsof pre- study plots were greater in 1984 than in 1983. dation in the other seven plots allowed more Becausepredation successof crows increases early clutchesto hatch,and fewer replacement with increasingprey density (Tinbergen et al. clutcheswere laid. Thus, young clutcheswere 1967, Croze 1970, G6ransson et al. 1975, Mon- available for a shorter period, and predation tevecchi1977), the greaterdensity of ibis nests ratesin theseplots declined earlier in both years in 1984may have accountedfor the higher pre- (Fig. 2). dation rate. Comparisonof predationbetween years.--A1- Synchronyof nestingis often citedas an an- though trends in relative predation intensity tipredator adaptation of colonial nesting in among plots were similar in both years, the birds (Darling 1938, Patterson 1965, Nisbet overall DPR in 1984 was significantly(z = 4.33, 1975). Differences in ibis breeding synchrony P < 0.0001)higher than in 1983 (Table 2). The betweenyears also may have contributedto the difference may have been due to changesin difference in overall DPR. Nesting in all plots Fish Crow or White Ibis nest densities between was less synchronous in 1984 (Fig. 3). The years.Although the total numberof FishCrow availability of many young clutchesthrough- nests in the South Colony in 1983 was un- out much of the breeding seasonmay have in- known, our observationsof crow activity did creasedpredation above the rate in 1983, when not indicate an increase in crow numbers in overall nesting was more synchronous. 1984. More than 1,100 more ibis nests were Impactof predationon ibisproductivity.--Nest- counted in 1984 than in 1983, however, and ing-successdata for White Ibisesat Battery Is- July1986] CrowPredation onIbis Eggs 537

TA•SLE3. Nesting successof White Ibises in study would have been produced per pair of White plots at Battery Island. Nest successestimated us- Ibises in 1983 and 1.05-1.12 in 1984. ing Mayfield (1961, 1975) method. Productivityof ibisesin our study plots was

1983 1984 comparableto that reportedfor the other White Ibis colonies(e.g. Kushlan 1977). Becausehu- Nest success--eggstage (A) 0.6302 0.4645 man disturbancemay facilitate Fish Crow pre- Hatching success(B) a 0.8801 0.8801 Nest success--nestlingstage (C) 0.9630 0.9214 dation (Bent 1926, Schreiber and Risebrough Nestling success(D) b 0.9220 0.9361 1972, Montevecchi 1977, Shields 1985) and be- Egg success(A x B x C x D) • 0.4925 0.3526 cause most of the South Colony was undis- Mean clutch size (E) 2.64 2.44 turbed by our activities, we believe that the Number of 10-day-oldyoung/nest (A x B x C x D x E) 1.30 0.86 nesting successof the colony as a whole was Number of clutches/pair (F) 1.11 1.44 higher than indicatedby our sample.Produc- Number of 10-day-oldyoung/pair tion of young at a level above that needed for (A x B x C x D x E x F) 1.44 1.24 replacementmay explain the steadyincrease in aThe probability of an egg hatching given that the the breedingpopulation of White Ibisesat Bat- nest survivesthe 21-day egg stage. tery Island since the early 1960's(Parnell and bThe probability of a young surviving to 10 days Soots1979) and alsomay accountfor the recent of age given that the nest survives the 10-day nestling stage. northward expansionof nesting by ibises in ½The probabilityof an egg producinga 10-day-old coastal North Carolina (Shields and Parnell young. 1983).Thus, at presentcrow and ibis densities, egg predationdoes not appearto be a serious threat to the productivityof the BatteryIsland land in 1983and 1984are given in Table 3. All White Ibis population. clutch losses were used in the calculations, but becauseFish Crow predation accountedfor over 97%of clutch lossesin both years,success rates ACKNOWLEDGMENTS for the egg stagereflect primarily the effectsof Dargan Frierson, Jr., Paul E. Hosier, and Donald predation. The probability of an egg surviving A. McCrimmon, Jr., provided valuable suggestions to hatch (A x B, Table 3) was 55.5% in 1983 and on methodologyand data analysis,and also made 40.9%in 1984.Survival of nestlingsto 10 days helpful commentson an earlier versionof the manu- of age(C x D, Table 3) wasmuch higher (88.8% script.Many people aided us in the field. We partic- in 1983and 86.3%in 1984).Thus, egg loss,due ularly thank Robin D. Bjork, Peter B. Colwell, W. mainly to predation by Fish Crows, was the Walker Golder, and Jennifer E. Slack for their help on numerous occasions. The comments of Thomas primary causeof nest failure. Wray II and two anonymousreviewers greatly im- A mean of 1.30 10-day-oldyoung was raised proved the manuscript.Major funding for this study per nest in 1983 and 0.86 in 1984 (Table 3). wasprovided by the SanctuaryDepartment, National Takingrenesting into account,we estimatedthe Audubon Society;the New Hope and Forsyth Au- number of 10-day-old ibises raised per pair at dubon societies;and the University of North Caro- 1.44 in 1983 and 1.24 in 1984. Because White lina at Wilmington through its Department of Bio- Ibises fledge at 40-50 days of age (Kushlan logical Sciencesand Biology Trust Fund. 1977), fledging successat Battery Island was undoubtedly lower due to nestling mortality LITERATURE CITED after the ageof 10 days.Most nestlinglosses in the coastalFlorida coloniesstudied by Kushlan ALLEN-GRIMES,A.W. 1982. Breedingbiology of the (1977) occurredin the first 20 daysof age; the White Ibis (Eudocimusalbus) at Battery Island, mortality rate of young betweenthe agesof 20 North Carolina. Unpublished M.S. thesis, Wil- mington, Univ. North Carolina. and 40 days was only 10%. Similar trends in AMERICAN ORNiTHOLOGiSTS' UNION. 1983. Check-list nestling mortality rates have been noted in of North American birds, 6th ed. Washington, other wading bird coloniesnot subjectto pre- D.C., Amer. Ornithol. Union. dation by large mammals (e.g. Wolford and BAKER,R. H. 1940. Crow predation on heron nest- Boag1971, Miller and Burger 1978).Assuming ing colonies.Wilson Bull. 52: 124-125. a 10-15% mortality rate of young between 10 BENT,A.C. 1926. Life histories of North American days of age and fledging, 1.22-1.30 fledglings marsh birds. U.S. Natl. Mus. Bull. 135. 538 SHIELDSAND PARNELL [Auk, Vol. 103

]3RIMLEY,H.H. 1938. The BatteryIsland rookerynear KUSHLAN,J. A. 1977. Population energeticsof the SouthportßN.C. Chat 2: 41-43. American White Ibis. Auk 94: 114-122. ]3URGER,J., & D.C. HAHN. 1977. Crow predation on LOMAN,J., & G. G•SRANSSON.1978. Egg shell dumps Black-crownedNight Heron eggs. Wilson Bull. and crow Corvuscornix predation on simulated 89: 350-351. birds' nests. Oikos 30: 461-466. ]3UTLER,R. W., ]N[. A.M. VERBEEK,& H. RICHARDSON. MAXWELLßG. R., II, & H. W. KALEII. 1977. Breeding 1984. The breeding biology of the Northwest- biologyof five speciesof heronsin coastalFlor- ern Crow. Wilson Bull. 96: 408-418. ida. Auk 94: 689-700. CROZE,H. 1970. Searchingimage in Carrion Crows. MAYFIELD,H. F. 1961. Nesting successcalculated BerlinßPaul Parey. from exposure.Wilson Bull. 73: 255-261. CUSTER,T. W., G. L. HENSLER,& T. E. KAISER. 1983. 1975. Suggestionsfor calculatingnest suc- Clutch sizeßreproductive successßand organo- cess. Wilson Bull. 87: 456-466. chlorine contaminants in Atlantic coast Black- MCNAIR,D. B. 1985. An auxiliary with a mated pair crowned Night-Herons. Auk 100: 699-710. and food-caching behavior in the Fish Crow. ß & F. A. PITELKA. 1977. Demographic fea- Wilson Bull. 97: 123-125. tures of a Lapland Longspur population near MEANLEY,]3. 1955. A nestingstudy of the Little Blue Barrow, Alaska. Auk 94: 505-525. Heron in eastern Arkansas. Wilson Bull. 67: 84- DARLING,F. F. 1938. Bird flocks and the breeding 99. cycle. LondonßCambridge Univ. Press. MILLER,L. M., & J. BURGER.1978. Factorsaffecting DusI, J. C., & R. T. DusI. 1968. Ecologicalfactors nestingsuccess of the GlossyIbis. Auk 95: 353- contributing to nesting failure in a heron col- 361. ony. Wilson Bull. 80: 456-466. MILLERß R. G. 1966. Simultaneous statistical infer- ELLISON,L. N., & L. CLEARY. 1978. Effects of human ences. New Yorkß McGraw-Hill. disturbance on breeding Double-crested Cor- MILSTEIN, P. LE S., I. PRESTT,& A. A. ]3ELL. 1970. The morants. Auk 95: 510-517. breedingcycle of the Grey Heron. Ardea 58: 171- ERIKSTAD,K. E., R. BLOM, & S. MYRBERGET. 1982. Ter- 257. ritorial Hooded Crows as predatorson Willow MONTEVECCHI,W. A. 1976. Egg size and the egg Ptarmigan nests.J. Wildl. Mgmt. 46: 109-114. predatorybehaviour of crows.Behaviour 57: 304- FREDERICK,P. C. 1985. Mating strategiesof White 320. Ibis (Eudocimusalbus). Unpublished Ph.D. disser- 1977. Predation in a salt marsh Laughing tation, Chapel Hillß Univ. North Carolina. Gull colony. Auk 94: 583-585. FUNDERBURG,J. ]3. 1960. The breeding birds of Bat- tery Island, North Carolina.Chat 24: 19-20ß36. NISBET,I. C. T. 1975. Selectiveeffects of predation GOODWINßD. 1976. Crows of the world. IthacaßNew in a tern colony. Condor 77: 221-226. York, Cornell Univ. Press. PARNELLßJ. F., & D. A. McCRIMMON, JR. 1984. 1983 G•SRANSSON,G., J. KARLSSON,S. G. NILSSONß& S. supplement to atlas of colonial waterbirds of ULFSTRAND. 1975. Predation on birds' nests in North Carolina estuaries. Univ. North Carolina Sea Grant Publ. UNC-SG-84-07. relationto antipredatoraggression and nestden- sity: an experimentalstudy. Oikos 26: 117-120. --, & R. F. SooTs, JR. 1979. Atlas of colonial waterbirds of North Carolina estuaries. Univ. GOTTFRIEDßB. M., & C. F. THOMPSON.1978. Experi- North Carolina Sea Grant Publ. UNC-SG-78-10. mental analysisof nest predation in an old-field habitat. Auk 95: 304-312. PATTERSONßI.J. 1965. Timing and spacingof broods HENSLER,G. L., & J. D. NICHOLS.1981. The Mayfield in the Black-headed Gull Larus ridbundus. Ibis 107: method of estimating nesting success:a model, 433-459. estimators and simulation results. Wilson Bull. REARDEN,J.E. 1951. Identification of waterfowl nest 93: 42-53. predators.J. Wildl. Mgmt. 15: 386-395. JENNI,D. A. 1969. A study of the ecologyof four RUDEGEAIR,T. J., JR. 1975. The reproductivebehav- speciesof heronsduring the breedingseason at ior and ecologyof the White Ibis (Eudocimusal- Lake Aliceß Alachua County, Florida. Ecol. bus). Unpublished Ph.D. dissertationßGaines- Monogr. 39: 245-270. ville, Univ. Florida. JONES,R. E., & K. E. HUNGERFORD.1972. Evaluation SCHREIBERßR. W. 1979. Reproductive performance of nestingcover as protectionfrom magpiepre- of the eastern Brown Pelican, Pelecanus occiden- dation. J. Wildl. Mgmt. 36: 727-732. talis. Contrib. Sci. Nat. Hist. Mus. Los Angeles KREBS,J. R. 1978. Colonial nesting in birdsßwith Co. 317: 1-43. specialreference to the Ciconiiformes.Pp. 299- ß & R. W. RISEBROUGH. 1972. Studies of the 314 in Wading birds (A. Sprunt VI, I. C. Ogden, Brown Pelican. Wilson Bull. 84:119-135. and S. Winckler, Eds.). Natl. Audubon Soc. Res. SHIELDS,M. A. 1985. An analysisof Fish Crow pre- Rept. No. 7. dation on eggs of the White Ibis at Battery Is- July1986] CrowPredation onIbis Eggs 539

landßNorth Carolina. Unpublished M.S. thesisß TREMBLAY,J., & L. N. ELLISON. 1979. Effects of hu- WilmingtonßUniv. North Carolina. man disturbanceon breeding of Black-crowned ß& J. F. PARNELL.1983. Expansionof White Night Herons. Auk 96: 364-369. Ibis nestingin North Carolina.Chat 47: 101-103. VERBEEK,N. A.M. 1982. Egg predationby North- SKUTCH,A. F. 1962. The constancyof incubation. western Crows: its association with human and Wilson Bull. 74: 115-152. Bald Eagleactivity. Auk 99: 347-352. TINBERGEN, N., M. ][MPEKOVEN,& D. FRANCK. 1967. WOLFORD,J. W., & D. A. BOAG. 1971. Distribution An experiment on spacing-out as a defence and biology of Black-crownedNight Herons in againstpredation. Behaviour28: 307-321. Alberta. Can. Field-Natur. 85: 13-19.

ERRATUM

The American Robin (Turdusmigratorius) should be deleted from Table 1 of "The function of singing in female Black-headedGrosbeaks (Pheucticus melanocephalus): family-group maintenance" by Gary Ritchison (1983, Auk 100: 105-116).