January1995] ShortCommunications and Commentaries 257 steroid measurementsto the assessmentof gonadal faecal steroid analysis.J. Zool., Lond. 224:649- activity in free-living Kakapo. The survival of this 668. endangered speciesdepends on egg laying by female COCKP,EM, J. F. 1989. Reproductivephysiology and Kakapo. Thus, conservation efforts are focussedon the managementof the Kakapo.Ecology Division the females. The shortageof female samplesin our Report No. 18. Lower Hutt, Ecology Division, study meansthat further samplecollection programs Department of Scientificand Industrial Research, are neededto realize the potential of this new method Lower Hutt, New Zealand. to provide information on the reproductivecondition COCKREM,J. F., AND J. R. ROUNCE. 1994. Faecal mea- of female Kakapo. In a wider context, fecal steroid surements of estradiol and testosterone allow the measurementscan now be appliedto otherfree-living non-invasive estimation of plasma steroid levels for which handling and blood sampling is dif- in the chicken. Br. Poult. Sci. 35:433-443. ficult or inappropriate. FOLLETT,B. K. 1984. Birds. Pages283-350 in Mar- Acknowledgments.--Wethank T. Greene, B. D. Lloyd, shall's physiology of reproduction,4th ed. Vol. R. G. Powlesland,C. Smuts-Kennedy,and other work- 1, Reproductivecycles of vertebrates(G. E. Lam- ers on Little Barrier Island for the collectionof drop- ming, Ed.). Churchill Livingstone,Edinburgh. pings.B. D. Lloyd, D. V. Merton, and R. G. Powlesland KOFUJI,H., M. KANDA,AND T. OISHI. 1993. Breeding generously provided information about Kakapo. S. cyclesand fecal gonadal steroidsin the Brown Ishii and C. M. Bishopkindly made available details Dipper Cincluspallasii. Gen. Comp. Endocrinol. of their methodsfor fecal-samplepreparation. Tl•e 91:216-223. work was funded by the Threatened SpeciesTrust POWLESLAND,R. G., B. D. LLOYD, H. A. BEST,AND D. through the Department of Conservationas part of V. MERTON. 1992. Breeding biology of the Ka- the Kakapo RecoveryProgramme, and D. J. Melior kapo Strigopshabroptilus on Stewart Island, New provided accessto laboratoryfacilities. Zealand. Ibis 134:361-373. WILKINSON,L. 1988a. SYGRAPH. Systat,Evanston, LITERATURE CITED Illinois. WILKINSON,L. 1988b. SYSTAT. Systat,Evanston, Il- BERCOVITZ,A. B., J. COLLINS,P. PRICE,AND D. TUTTLE. linois. 1982. Noninvasive assessment of seasonal hor- WINGFIELD,J. C., G. F. BALL,A.M. DUFTY,JR., R. E. mone profiles in captive Bald Eagles(Haliaeetus HEGNER, AND g. RAMENOFSKY. 1987. Testoster- leucocephalus).Zoo Biol. 1:111-117. one and aggressionin birds. Am. Sci. 75:602-608. BISHOP,C. M., AND M. R. HALL. 1991. Non-invasive monitoring of avian reproduction by simplified Received15 September1993, accepted 12 June1994.

TheAuk 112(1):257-260, 1995

Polygyny in the Asian Openbill (Anastomusoscitans)

TANMAY DATTA AND B.C. PAL Departmentof Zoology,North BengalUniversity 734430, Dist. Darjeeling, West Bengal, India

Although monogamy is the predominant mating We describeseveral cases of polygyny in a breeding systemin birds, there are few strictly monogamous population of a , the Asian Openbill (Anastomus species(see review by Ford 1983). Other than oscitans).We also report information concerningthe the widespread occurrenceof polygyny in various formation of suchmating groupsand the successex- species,occasional polygyny also has been reported perienced by them in comparison to monogamous in normally monogamousspecies (Armstrong 1955, pairs. Verner and Willson1969, Logan and Rulli 1981,Marks Methods.--Our study was conductedat the Raiganj et al. 1989). Although Brown (1987) did not include Wildlife Sanctuary(25ø36'N, 38ø10'E),West Bengal, any membersof the Ciconiiformesin his list of 222 India, where AsianOpenbills are found breedingfrom speciesof communallybreeding birds, occasionalpo- July through December,as are five other waterbird lygyny also has been found in this group (Lancaster species. 1970, Cramp 1977, Fujioka 1986, McKilligan and We studied this population of openbills from 1987 McConnell 1989).However, storksalways have been to 1991. Nests with more than two adult birds were found to breed monogamously(Ali and Ripley 1968, consideredas a possible caseof polygyny, and we Cramp 1977, Coulter et al. 1989). kept closewatch on thosenests. Data from thosenests 258 Short Communicationsand Commentaries [Auk, Vol. 112 were taken into accountonly when the presenceof rushed to that spot and got involved in the distur- more than two adult birds was confirmed by direct bance.In these cases,fighting continued for 5 to 18 observation for at least three consecutivedays. min (2 = 588 + SD of 296.74 s), involving 6 to 13 Membersof 37 suchgroups were colormarked with birds (2 = 9.29 _+2.29). The duration of fighting was nontoxic textile dyes using a sprayer from a high directly proportional (r = 0.946) to number of birds vantage position above the nest-rim level. For each involved. At the end of fighting, three or four birds nest we maintainedan identity card with the records were found to rest peacefullyin that nesting spot. of the position of the nest in the sanctuaryand the The new group sometimesincluded one or both of marking patternsor identifiable physicaltraits of the the original pair, but this was not always the case. members. Sexes were ascertained from behavior. Po- Within a very shortperiod, the birdsat the nestmade sition during copulationwas treated as a good indi- joint up-down displays(Kahl 1972) in very closesuc- cator of sex, as "reverse mounting" is very rare in cession.Thereafter, they behaved as a unit in all re- this species.Above all, we think our continuousand spects,exhibiting pair-forming displays,as done by careful observationswere effective in making certain monogamouspairs. Such intraspecificattacks were that we did not introduce the confounding effect of very commonin this colony and, in most cases,they reverse mounting. ended with only two birds in the nest. Nestswere inspectedeach day beforehatching and We were able to surveythe developmentand suc- thereafter on alternate days. Observationwas made cessof clutchesat 11 polygamousnests. Polygamous from a high vantageposition or from the groundwith clutchesshowed greater variation in layingorder than minimum interference to birds in the area. To com- those in monogamousnests, where eggs were con- pare successrates, four monogamouspairs were se- sistentlylaid on alternatedays. In polygamousnests, lectedas controlsfor eachpolygamous nest. Control mostof the eggswere laid on alternatedays; however, nestswere selectedfrom the closevicinity of the po- in six instanceseggs were added to the clutch on the lygamousnest; the physicaland breedingstatuses of sameday. In another nest, there was a three-day in- all control birds were similar. Observations were made terval between eggs.Clutch size of polygamousnests with the naked eye or with 20 x binoculars. Mea- (œ= 4.45 _+ 1.44, n = 11) was significantly higher (t surementsof eggswere taken with a vernier caliper, = 3.76, df = 53, P < 0.001) than that for the control and massesof young were taken with a Pesolaspring monogamousnests (2 = 3.41 _+0.54, n = 44). balance. Intraclutch variation of egg dimensiondid not fol- Results.--Duringthis five-year study, we found 51 low definite order in polygamousclutches, in contrast monogamousgroups of 2,115 active nests.In 2 cases, to the situation in monogamousclutches. In monog- four birds used a single nest, and the other 49 nests amousclutches, larger eggsalways were laid in the were occupiedby trios. middle of the clutch.In three- or four-egg clutches, In this colony, openbills usually started nesting the secondegg was largest,while in five-eggclutches from the first week of July. The polygamousnesting the third was largest.In polygamousclutches, no such attemptswere not seenuntil August,when almostall order was maintained and, in some cases, the first or of the nesting trees were invaded by monogamous last egg of the clutch was the largest.Intraclutch dif- pairs. Nesting by monogamouspairs also continued ferencesbetween the largest and smallestegg were until the lastweek of September.After the third week significantlygreater in polygamousnests (t = 3.76, df of July, most of the new settlershad no option other = 53, P < 0.001). In 44 control monogamousclutches, than to choosea preoccupiednesting tree for nesting, the range of egg volumes was never more than 30% which often resulted in intraspecificfighting. Natu- (œ= 14.8%)of the smallestegg volume. However, in rally polygamousgroups always nested in a preoc- 8 of 11 polygamousclutches, this differencewas more cupied nesting tree with many monogamousnests. than 35% 07 = 30.6%). Polygamousnests were not confinedto any particular Hatching successand fledging successwere con- zone of the colony, but were scatteredthroughout. siderablyhigher in polygamousnests, while lossdue Copulatory behavior confirmed the sexesof the to intraspecificattack was higher in monogamousnests members of 21 trios in which two females mated with (Table 1). The massof young birds at fledging was a common male. In 15 other cases,behavioral (other significantly higher (t = 3.2, df = 49, P < 0.005) in than copulation)and physicalcharacters of the mem- polygamousnests. bersalso suggested that only one male wasinvolved. All members of the polygamous nests shared in However, for the remaining 15 cases,sexes of the mostof the parentalactivities, from nest building to membersof thosemating groupswere not identified. food provisioningof young. We sawno obvioussigns We witnessedthe formation of suchmating groups of parental discrimination in these polygamous on seven occasions.It was a direct consequenceof groups. As a result, the whole clutch or brood re- intraspecificattack. In each casethe unattached bird ceivedmore care than did thosein monogamousnests. madean attackon an alreadymated monogamous pair During the egg stage,more than one parent attended with a four- to seven-day-oldnest. The resultingfight the polygamousnest for more than 85% of daytime attractedother unattachedbirds of the colony,which hours (316.5 h of 360 h observed), while this occurred January 1995] ShortCommunications and Commentaries 259

TABLE1. Reproductivesuccess in polygamousand control monogamousnests (n in parentheses).

Hatching Fledging Lossdue to Massat success success intraspecific fledging (%) (%) attack (%) (g; œ+ SD) Polygamous 77.6 (11) 89.5 (11) 5.0 (40) 1,279.8 + 78.7 (21) Monogamous 66.0 (44) 81.8 (44) 15.9(44) 1,208.2+ 76.0 (30) only 40% of the time during the day (375.1 h of 926 successwere higher than found for monogamousnests. h observed)at monogamousnests. Before hatching, Although productivity (number of young) per female the total time when birds were not incubatinga nest of a polygynousnest looksto be slightly lower than during daytimehours was significantlyhigher (t = that of a monogamousfemale, quality (asreflected in 11.7,df = 53, P < 0.001)in controlmonogamous nests fledging mass)of the young of polygynousnests was (• = 121.8 + 11.2 min) than in polygamousnests (• greater than that of monogamousnests. Moreover, = 93.5 + 7.3 min). Thesefactors probably make the polygynousnesting reducedthe probability of total polygamousnest more resistant to intraspecificattack. reproductivefailure due to intraspecificattack. Such None of the polygamousnests observed were de- failure was very commonat the monogamousnests, stroyedcompletely by intraspecificattack, but 2 of 44 particularly in the later part of the breeding season. control monogamousclutches were destroyedcom- In the situationwe studied,the effectof territory pletely for this reason. quality (Verner 1964) or female-biasedsex ratio (Em- Discussion.--Inthis sanctuary, polygamous nests len and Oring 1977) seemsto be lesssignificant; po- were found in all the years studied. All membersof lygynous nestswere distributed throughout the sanc- such groups have been found to take active part in tuary and replacementof nesting malesby new ones breeding, and the data suggestthat more than one during intraspecificfighting or establishmentof mo- femalelaid eggsin a given nest.The reasonsfor these nogamous pairs up to the last week of September conclusionsinclude: (1) in each of these groups, a suggeststhe abundanceof unattachedmales. Proba- single male was found to copulatewith the other two bly, scarcityof suitable nesting sites was the causal membersof the group;(2) largerclutch sizes seem to factorof intraspecificattack by unattachedbirds on be the result of laying by more than one female;(3) nes,ted pairs. Joint nesting by female Asian Openbills the laying sequencewas abnormal,particularly with seemed to be an effective device to withstand the the laying of two eggsin one day; (4) the abnormal menaceof intraspecificattack with no major loss in intraclutch variation of egg dimensions relative to productivity. laying sequenceseems to be due to asynchronous Acknowledgments.--Thisstudy was supportedby a laying by more than one female;and (5) the greater grant from the University GrantsCommission (UGC), intraclutchdifferences between smallestand largest New Delhi. The divisional forest officer of West Di- egg volumes may be due to laying by femalesof dif- najpur Division kindly allowed us to work in this ferent physicalstatus. sanctuary.We thank the beat officer and other staff From direct observation, it was evident that these of RaiganjWildlife Sanctuaryfor providingvarious mating groupsformed as a consequenceof fighting kinds of help in the field. to establishsole dominance.Probably none of the femaleswas dominant to the point that it could dis- place or chaseoff other females. The choice to share LITERATURE CITED a nest could be advantageousbecause continuation of fighting is likely not only to be expensiveener- ALl, S., AND S. D. RIPLEY. 1968. Handbook of the getically but alsocould attract other unattachedbirds, birds of India and Pakistan, vol. 1. Oxford Univ. which may jeopardize the former attacker'schance of Press,Bombay. winning. However, if two or three of these females ARMSTRONG,E.A. 1955. The Wren. Collins, London. jointly stand againstintruders, the result could sig- BROWN,J.L. 1987. Helping and communalbreeding nificantly increasethe fitnessof the defendingbirds. in birds. Princeton Univ. Press, Princeton, New Mating polygynously is disadvantageousto fe- Jersey. males becauseof heavier pressureand COULTER,M. C., S. BALZANO,R. E. JOHNSON,C. E. KING, poorer food condition (Verner 1964, Orians 1969), or AND P. W. SHANNON(EDs.). 1989. Conservation due to lack of male assistance(Martin 1974, Smith et and captivemanagement of .Proceedings al. 1982). In JapaneseWagtails (Motacillagrandis), po- of an International Workshop at the New York lygynousnests had a higher frequencyof unhatched Zoological Society's Wildlife Survival Center, eggs (Nakamura 1985). However, in the caseof the Stork Interest Group. Saint Catherine's Island, openbills, polygynousnests were lessvulnerable to Georgia, USA. intraspecificattack, and both hatching and fledging CRAMP,S. (CHIEFED.). 1977. Handbook of the birds 260 ShortCommunications and Commentaries [Auk, Vol. 112

of Europe,the Middle Eastand North Africa, vol. dencesuggesting a caseof bigyny in the Inter- 1. Oxford Univ. Press, Oxford. mediate Egret Egretta intermedia.Aust. Bird EM•,EN,S. T., ANDL. W. OraNG. 1977. Ecology,sexual Watcher 13:98-99. selectionand the evolution of mating systems. NAKAMUP•,S. 1985. Clutch size and breeding suc- Science 198:215-223. cessof the JapaneseWagtail Motacilla grandis, with FORD,N.L. 1983. Variation in mate fidelity in mo- a specialreference to its habitatand mating sys- nogamousbirds. curr. Ornithol. 1:329-356. tem. J. Yamashina Inst. Ornithol. 17:84-104. FUIIOKA,M. 1986. Two casesof bigyny in the Cattle ORI,*a•S,G. H. 1969. On the evolutionsof mating Egret Bubulcusibis. Ibis 128:419-422. systemsin birds and mammals.Am. Nat. 103:589- KAHL,M.P. 1972. Comparativeethology of the Ci- 603. coniidae,part 5. The Openbill Storks( An- SMITH, J. N.M., Y. YoM-Tov., ,•a•D R. MOSES. 1982. astomus).J. Ornithol. 113:121-137. Polygyny, male parental care and sex ratio in L,*a•C•TER,D.A. 1970. Breedingbehaviour of the Song Sparrows:An experimentalstudy. Auk 99: Cattle Egretin Columbia.Living Bird 9:167-194. 555-564. LOGAN,C. A., ANDM. RULU. 1981. Bigamyin a male VERNER,J. 1964. Evolutionof polygamyin the Long- Mockingbird.Auk 98:385-386. billed Marsh Wren. Evolution 18:252-261. Mmucs, J. S., J. H. DO•,EMUS,,*a•D R. J. C,*a,•NINGS.1989. VERNœR,J., AND M. F. WI•,LSON.1969. Mating sys- Polygyny in the Northern Saw-whetOwl. Auk tems, sexual dimorphism and the role of male 106:732-734. North Americanpasserine birds in the nesting M•RT•N, S. G. 1974. Adaptationsfor polygynous cycle.Ornithol. Monogr. 9. breeding in the Bobolink, Dolichonyxoryzivorus. Am. Zool. 14:109-119. Received30 August1993, accepted20 November1993. McKILLIGAN, N. G., ,•a•D P. McCoNNELL. 1989. Evi-

The Auk 112(1):260-262, 1995

Water and Energy Limitations on Flight Range

MARCEL KLAASSEN • Max-Planck-Institutfiir Verhaltensphysiologie,D-82346 Andechs, Germany Carmi et al. (1992) proposeda computermodel that I adjustedthe model by Carmi et al. to accountfor furthersour knowledgeof whether energy or water this continuouschange in body mass due to water are the greaterphysiological limitation to bird flight loss(in their energy-budgetpart, Carmi et al. used range.With the trans-Saharamigrating Willow War- Pennycuick's[1989] flight-range model, which al- bler (Phylloscopustrochilus) as an example, the model ready accountsfor massloss as a result of use of en- is a major theoreticaleffort to identify several im- ergy reservesunderway). In the alternative"extend- portant parameters on which migratory research ed" model, where the energy and water budgetsof should be focused in the future. However, there is a the flying bird are closelylinked, the model'sparam- shortcomingin the model, as it calculatesthe flying eter values change continuouslyduring flight per- bird's energy and water budgetlargely independent formance.To approximatethese continueschanges, of eachother. Only the calculatedenergy-consump- all parameterswere evaluated and actualized at 15- tion rate in the energy budget part of the model is rain flight intervals.In thiscommentary, I will discuss passedto the water-budget component for the cal- some of the major discrepanciesthat occur between culationof respiratorywater lossand metabolicwater the original model of Carmi et al. (1992) and the ex- production.However, the usual negativewater bal- tended model outlined above. ance,as pointed out by Carmi et al. (1992), will result For this comparison,in concordancewith Carmi et in a decreaseof body massthat should be accompa- al.'spaper, I simulatedthe flight of a smallPalaearctic nied by changesin optimal flight speedand costsof passerine,the Willow Warbler (with a wingspan of flight.Thus, water and energy budgets influence each 0.17 m and an initial body massof 10 g, including 5 other mutually. g water [of which a 30%loss is tolerated]and 3 g fat), during its annual migration over the Sahara desert. When not statedotherwise, a flight altitude of 2,000 • Presentaddress: Netherlands Institute of Ecology, m, an air temperature of 14 øC, and an oxygen-ex- Centrefor Limnology,Rijksstraatweg 6, NL-3631AC traction coefficient of 0.039 are assumed, all in accord Nieuwersluis, The Netherlands. with the parametersettings used by Carmi et al. (1992).