Prolonged Parental Care in Royal Terns and Other Birds

PROLONGED PARENTAL CARE IN ROYAL TERNS AND OTHER BIRDS

N. PHILIP ASH•OLE ^ND HU•BERTO TOV^R S.

DURING the last thirty years ornithologistshave accumulateda vast body of quantitativeinformation on many aspectsof the breedingbiology of a large variety of bird species,but adequateinformation on the weeks followingdeparture of the young birds from the nest is available for only a handful of species.As this period includesthe critical time when the young becomeindependent of their parents--and one when they suffer heavy mortality--the seriousnessof this gap in present knowledgeis ob- vious. The gap exists becauseof the difficulties of obtaining data after young birds have left the nest. For most speciesintensive efforts are necessaryto overcomethese difficulties. Our interest in the postfledging period was stimulated by a chanceobservation that led us to considerthe selectivepressures determining the length of the period of dependenceof young birds.

OBSERVATIONS

We watchedgroups of Royal Terns (Thalasseusmaximus) and Elegant Terns (Thalasseuselegans) at Laguna Grande, Departamento de Ica, Peru, on severaloccasions between early December 1966 and early Janu- ary 1967. The Elegant Tern winters in considerablenumbers on the Peru- vian coast, but the Royal Tern is seen less often. However, a group of Royal Terns is regularly presentduring the northern winter, and odd birds stay during the northern summerat Laguna Grande, which at 14ø S may be the southernmostwintering station for the specieson the west coast of SouthAmerica (cf. Koepcke,1964). At the time of our visitson 5, 6, and 10 December 1966 more than 1,000 Elegant Terns and at least 100 Royal Terns were present at Laguna Grande; on 11 January 1967 about 300 Royal Terns were present. Most of the Royal Terns were adults in the middle of the annual molt of their remiges,but perhapsone tenth were first winter birds (i.e., about 6 monthsold), distinguishablein flight by their worn, and thereforeblack- ish, primaries, and when perchedby the darker scapularsand inner sec- ondaries. At least four of the Royal Terns had bands, and of these one adult and one first winter bird were shot 10 December 1966 to determine their placeof origin. The adult, a female,proved to have been bandedas a juvenile by Lovett Williams near Pascagoula,Mississippi in July 1964. The young bird, also a female, was banded as a juvenile by H. T. Davis on Wainwright Island, Pamlico Sound,North Carolina, 4 July 1966, and

90 The Auk, 115: 90-100. January, 1968 1968Jan. ] AS•0LEA•ND TOVAR S.,Prolonged Parental Care 91 was thus probably a little lessthan 6 monthsold when collected. In each wing the innermosttwo juvenal primariesand the innermostfive second- aries (numbers 20 through 24) had been replaced by new feathers. No remigeswere in the processof growth, and "discontinuities"(Ashmole, 1963b) were noticeablebetween the old and new feathers; probably the molt had started in the northernhemisphere and then stoppedduring the southwardmigration. (In someyoung birds noneof the juvenal primaries had yet been replaced.) In the tail the two innermostpairs of juvenal rectriceshad been replacedand the vanesof the new third pair were just breaking their sheaths. Somegrowing body featherswere presenton the head, mantle, and dorsal surfaceof the wings. On the first day we watchedthis group of Royal Terns, we severaltimes heard individualsutter a seriesof loud calls in flight, and noted that these birds were always carrying fish. Twice we saw the fish given to anoth'er bird, but could not determinethe age of the recipient. As behavior of this kind is usedby adult Thalasseusterns during courtshipas well as when feedingyoung, it was clearly important to determinewhether we were wit- nessingpremature or residual courtshipbehavior or prolongedparental care. On 10 December we returned to the area and watched from 0545 until 0930, using a telescopeas well as binoculars,and three times saw adults give fish to young birds. Th'e first observationwas of an adult fish- ing, with a young bird flying below and behind it and persistentlyusing the squeaky begging call. Eventually the adult caught a fish and then flew somedistance away from the main group of birds, followedby the beggingjuvenile that then settled on the water. The adult flew down and gave it the fish without itself settling; we could not see whether the fish was actually passedfrom bill to bill or droppedto the youngbird from just above its head. The young bird continued begging for at least 10 minutes,partly in flight below a fishingadult (perhapsthe one that had fed it), but later also on the ground, when the begging was apparently directed towardsa flying bird or birds. A little later an adult with a fish flew in past th'e restinggroup, calling persistently.It was soonjoined by a young bird, which followedit for a minute or more as it flew away from the main group. The adult then droppedthe fish--evidently deliberatelyfor it made no effort to retrieve it as thesebirds normally do--and the young bird picked it up from the water surface. Later this juvenile was seen hunting by itself, though it was not seen to catch anything. The third observationwas of a calling adult that flew in with a fish, was joined by a young bird, and was also pursuedfor a short time by another adult. The adult with the fish and the young bird then flew out to sea, the young bird settled, and the adult hoveredand gave it the fish. The young bird in this casewas probably 92 ASH•OLEAND TOVAR S., ProlongedParental Care [ Auk [ Vol. 85 a banded individual. The sameday we watched an adult fly in with a fish and circle four times around the various groupsof perched birds, calling about every 2 secondsover a period of some 5 minutes, but it only suc- ceededin attracting the attention of a Band-tailed Gull (Larus belcheri) which pursuedit vigorously.Eventually the adult tern swallowedthe fish itself. Adults were again seen feedingyoung (by H. T. S.) at Laguna Grande on 11 January 1967. Subsequentlysmall groups of Royal Terns were watched (by H. T. S.) at Lagunilla, a bay a few miles north of Laguna Grande,on 13 March, 31 March, 7 April, and 21 April 1967, but juveniles were not heard beggingfor food, nor were adults seenfeeding them. On 13 March juveniles were seen successfullyfishing for themselves. We have describedour observationsin somedetail becausethey indicate that adult Royal Terns sometimescontinue to feed their young until they are about 7 monthsold and have been able to fly for more than 5 months. Althoughwe have no proof that the adults feed only their own young,the persistentcalling by both adults and young strongly suggeststhat individual recognitionis involved. Adult feedingof youngafter fledginghas previouslybeen reportedin at least two other membersof the genusThalasseus. After observingElegant Terns in San Diego Bay, California during the summerand fall seasonsof 1953, 1954, and 1955, Monroe (1956) says "Iramatures continuebegging for food, and have beennoted doing so into late November. Beggingis accomplishedexactly as in the Royal Tern. The immaturebird, upon ar- rival of an adult with food, lowers its head, thereby erecting the crest feathers,chases the adult with head low and slightlyuptilted, and peeps continuouslyuntil fed." As only small numbersof Elegant Terns were presentin San Diego Bay later than October (the area being visited by the speciesmainly in the late summerand early fall) the fact that Monroe did not recordbegging later may merely reflect the absenceof birds. How- ever we did not observeany beggingamong the wintering Elegant Terns in Peru in Decemberor any birds arriving at the resting area carrying fish. In contrastthe feedingbehavior by Royal Terns was conspicuous,despite the smaller numbersof this species. Adult Elegant Terns feedingyoung in California were also observedby Pyle and Small (1951) in August 1951. Among other membersof the genus,we are aware of relevantdata only for the SandwichTern (Thalas- seussandvicensis); Tinbergen (1953: 229) watchedadults feedingchicks during migrationin August,and David Lack (pers. comm.) saw adults of this speciesfeeding young in Norfolk, England,at least20 milesfrom their breedingcolony. Similar behavior certainly occursin a number of other ternsand gulls,but seemsnot to have beenstudied systematically. Palmer 1968Jan. ] AS•MOLEAND TOVAR S.,Prolonged Parental Care 93

(1941: 103) in his study of the Common Tern (Sterna hirundo) records that "As somejuvenals are still being fed by their parents during migration, the beggingnote is often heard in traveling flocks. Not all juvenals do this, for someflocks, reported by variouswriters, were apparentlycom- posedentirely of youngbirds." Ralph W. Schreiberinformed us (orally) that he regularly observedyoung Herring Gulls (Larus argen.tatus)being fed by adults--under circumstancesstrongly suggestingthat they were the parents--4)nrubbish dumps in Maine from Augustto October; someof the gulls involvedwere banded in coloniesseveral hundred miles away. Simi- larly David Lack (pers. comm.) has observedyoung Herring Gulls re- peatedlybegging their (presumed)parents for food in Cornwall,England at the end of October. Thus evidencesuggests that parentalcare persists after departurefrom the breedingarea in a number of larids. However the feeding of young Royal Terns, apparently by their own parents, more than 5 months after fledging and after a trans-equatorialmigration of at least 3,000 miles, seemsto have no recordedparallel in other species. The closestanalogy may be with' the geesein which family parties persistthrough the south- ward migration,the followingwinter, and even the return migrationto the breedinggrounds (see Hochbaum,1955: 93).

DISCUSSION

In view of the high mortality sufferedby young birds of a wide variety of speciesshortly after they learn to fly, which certainly resultslargely from their relative inefficiencyin obtainingfood, it is perhapssurprising that more speciesdo not continuefeeding their fledgedyoung over an ex- tendedperiod. The difficulty of maintainingcontact between parents and their young,especially in speciesthat showa postbreedingdispersal from th'enesting area, probably constitutesa seriousbarrier to the evolutionof prolongedpostfledging parental care in many species.Royal Terns share with other membersof the genusThalasseus a number of ecologicalchar- acteristicsthat may have been important preadaptations. First as the young wander from the nest site at an early age, this speciesdoubtless evolved,like someother larids, mechanismsfor individual recognitionbe- tween parents and young chicks (Davies and Carrick, 1962), probably largely by variation among individual calls; suchrecognition could readily persistafter fledging. Secondthe birds tend to roost at night and to rest duringthe day, both on migrationand in winter quarters,in compactflocks closeto where they obtain their food; this must greatly facilitate the maintenanceof contactamong the membersof a family. Third they catchprey of a substantial size which can be easily and quickly transferred to the young. 94 As•Mo•;E ANDTOVAR S., ProlongedParental Care [ Auk I_ Vol. 85

Similar reasoningis usedby Nelson (1966b) to explainthe fact that the youngof tropicalboobies (Sula spp.) are fed by their parentsnear the nest site for a longtime after they can fly (and apparentlysometimes until they are nearly a year old--Verner, 1961), while youngGannets (Sula bassana), which go to sea even before they can fly, are not fed after they leave the breedingcolony. Nelsonbelieves that it would be impracticablefor adult Gannetsto maintain contactwith their own young once the young leave the immediatevicinity of the colony. Amongresident species, on the other hand,the maintenanceof contactbetween parents and offspringshould not presentan insuperableobstacle to the evolutionof prolongedparental care, but feedingyoung for more than a few weeksafter fledgingappears to be uncommonamong birds. This impliesthat in most speciesthe youngsoon becomesufficiently expert at findingfood for themselves,and that any se- lectiveadvantage gained by parentscontinuing to feed their young is out- weighedby the disadvantageousdiversion of energyresources of the adults from other important functions,especially the raisingof secondbroods or the replacementof feathers(cf. Pitelka, 1958). On the otherhand in those speciesthat needa high degreeof skill for h•nting successand whoseyoung require a considerabletime to perfect their skill, prolongedparental care is essentialand may prevent any attempt to raise a secondbrood if the first is successful. The raising of a large brood implies that the parents can easily obtain large quantitiesof food in the breedingseason. Thus in speciesthat lay large clutcheseven inexperienced juveniles may generallyhave little difficulty in obtainingthe muchsmaller quantities of food they need to maintain themselvesin the time just after fledging,so one might predict that prolongedparental care would not often occurin specieswith large clutch sizes. On the other hand low clutch size, and especially the minimum clutch of one, implies scarcityof food or strongcompetition for it (Ash- mole, 1963a,Amadon, 1964), so prolongedparental care shouldbe especially commonamong birds that have small clutchesand also use special- ized feedingmethods which can be perfectedonly by a prolongedlearning process. The most critical situation will be in those speciesthat capture relatively large prey at infrequent intervals, becausein these birds even a few successivemissed opportunities by an inexperiencedbird could have fatal results. Although we have not attempted a comprehensivereview, the literature providessome support for the predictionthat thosebirds whose methods of obtainingfood require great skill shouldhave a considerableperiod of post- fledgingparental feeding of the young,and shouldoften lay smallclutches and be single-brooded. Perhaps the most spectacularexample is that of the Crowned Eagle Jan.1968 ] AS•r•O•EAND TOVAR S.,Prolonged Parental Care 95

(Stephanoaetuscoronatus), studied by Brown(1966), which feeds mainly onagile prey such as small antelopes, hyrax, and monkeys. In thisspecies theyoung eagle is fed,almost always at thenest, for aslong as 11.5 months after the time of fledging,until it is about 15 monthsold. Howeverit sometimeskills for itself long beforeit becomesindependent, apparently whenit has not beenfed by the parentsfor severaldays. As mightbe expected,Crowned Eagles breed only in alternateyears, as the femaleat leastis involvedin breedingactivities for a total of 17-18 months.This speciesprovides an excellentexample of a bird eatinglarge, relatively scarceprey, in whicheven the adultsmay sometimeshave difficulty in makingkills often enough, as suggested by failureto breedin a yearwhen the normalprey wereless abundant than usual. The observationsof Fowler and Cope (1964) on the Harpy Eagle (Harpiaharpyja) in BritishGuiana show that the breeding biology of this speciesis similarto thatof theCrowned Eagle. A juvenile10 monthsold wasstill being fed by an adultand showed no signsof huntingfor itself, althoughreplacement of the juvenalrectrices and primaries had already startedat theage of about8 months.In mostmembers of theAccipitridae theperiod of parentalcare after the young leave the nestis muchshorter, but it amountsto severalmonths in someother large species(see Brown, 1952-53; 1955). Th'eFalconidae normally raise only one broodper year and the parents feedthe young for severalweeks after fledging (see Cade, 1960). The same is true for someowls. The Tawny Owl (Strix aluco) which eats mainly smallrodents (Southern, 1954) is single-brooded,and in Britain the clutch is smalland the youngare dependenton their parentsfor foodin the 21/.2 to 3 monthsafter fledging;nevertheless many young die of starvationafter the adultsstop feeding them (Southernet al., 1954). During the depen- dentperiod the younglearn the way aboutthe territoryof their parents, but havenot beenseen practicing hunting; the behaviorof the youngdur- ing the last few weeksbefore they becomeindependent is so difficultto observethat there are few relevant observations.This examplethus sup- portsthe hypothesisoutlined above only if it can be assumedthat the youngbirds are acquiring skills or knowledgein thepostfledging dependent periodthat will makethem more efficient when they haveto hunt for themselves.Southern's (1954) discussionof the situationin other owls showsthat somespecies (Barn Owl, Tyto alba,and Short-earedOwl, Asio Jlammeus),which must be regardedas equallyskilled in hunting,contrast with the Tawny Owl in havingrelatively short periods of parentalcare and raisingseveral broods in seasonswhen prey is very abundant. Amongfish-eating birds the kingfishersare clearlyrelevant to this discussion.Although we havenot seena detailedaccount of the postfledging 96 ASHMOLE^•DTOV^R S.,Prolonged Parental Care [ Vol.Auk 85 period of any species,the descriptionby Bralliar (in Bent, 1940: 117) of how young Belted Kingfishers(Megaceryle alcyon) are taught to fish by their parentssuggests the way in which membersof this group achievethe transition to independence.According to Brailiar th'e learning period in this specieslasts only a week or so, but it presumablyvaries betweenspe- cies accordingto the abundanceof prey at the critical time. In the Green Kingfisher (Chloroceryleamericana) Skutch (in Bent, 1940: 144) de- scribeshow a youngbird that had beenout of the nest for 29 days begged for a fish from its father, but was chasedaway. Young of the Kingfisher (Alcedoatthis) apparentlybecome independent fairly soonafter fledging. This speciesis one of the few birds in which successivebroods sometimes overlapsubstantially, so that the male may be simultaneouslyinvolved in feedingyoung at one nestand incubatingat another (Brown, 1934; Swan- berg, 1952). This fact and also the high clutch size (generally 6-7) suggestthat, like someowls when rodentsare very abundant,both adults and young of this speciesof kingfishercan obtain food easily during the breeding season,in spiteof their specializedmethod of feeding. The frigate-birds(Fregatidae) are anothergroup in which the acquisition of food requiresgreat skill, but their food is often scarce. In the As- censionIsland Frigate-bird (Fregata aquila) the chicksbegin to fly during their 6th or 7th month, but remain at least partly dependenton their parentsfor a further3 or 4 months. Stonehouseand Stonehouse(1963), who studied the species,suggest that "the long period of partial dependence may help th'eyoung birds to attain proficiencyin accurateflying, a neces- sary skill in a specieswhich obtainsmuch of its food either in the air, or by skimming over the sea or ground and taking food without settling." Nelson (1966a), studying the Great Frigate-bird (Fregata minor) on Tower Island in the Galfipagos,found that the young birds were dependent on their parents for 6 months after fledging, and that a successful breeding cycle took more than a year; apparently these birds, like the Crowned Eagle• can breed successfullyonly in alternate years. Nelson found that in spite of the prolongedpostfledging feeding several young died of starvation about when they were becomingindependent. This ex- ample demonstratesparticularly well the difficulty of the transitionto independencein specieswhose food is sometimesscarce and whosefeeding methodsare highly skilled. Members of several of the groups discussedabove have been observed performing"play" activitiesof a kind that must aid in the acquisitionof hunting skill. Stonehouseand Stonehouse(1963) report juvenile Ascen- sion Island Frigate-birdsin small groupsover the coast "flying together, catchingfeathers and strandsof sea-weedfrom each other in mid-air, and attacking incoming boobies." Similarly Gibson-Hill (1947) says of Jan.1968 ] A$1tlVIOLEAI•'DTOVAR S.,Prolonged Parental Care 97

Fregata minor "They seem,in fact, to be much addictedto picking things up, and small groupsof birds of various ages often play follow-my-leader over a rock-pool,each in turn swoopingdown over the water and pausing just aboveits surface. They also take a great delight in pulling up pieces of dead creeper,and in trying to break prominentbranches off low trees." Swanberg(1952) recordsa young Kingfisher probably 50 days old "several timesdiving for and fetchinga twig, 1.5 cm long, floatingalong in the water." Among the birds of prey there are a number of recordsof com- parable behavior. L. Tinbergen (quoted by Thorpe, 1963: 362) watched young Kestrels (Falco tinnunculus) playfully hunting a variety of inani- mate objects, although they were already able to hunt successfullyfor themselves,while Munro (1954) observeda Prairie Falcon (Falco mexi- canus) of unspecifiedage repeatedlydropping or tossing,and then catch- ing, piecesof dried cow manure. After submittingthis article we had an opportunityto observeyoung Inca Terns (Larosternainca) practicinghunting around Isla Chincha Sur, Peru. For severalhours on 15, 16, and 17 August 1967 we watched about 20 juvenileswhich had recently learned to fly (certainly in the previous month) and which probably averagedabout 2 monthsold. From time to time one of thesebirds would leave the rock on which it was perched,circle low over the water and then "plunge to surface"and pick up an inani- mate object in the water; the birds also sometimesused "contact dipping." (In "plungingto surface"the bird dropsto the surfacebut doesnot sub- mergecompletely, while in "contactdipping" forward flight doesnot stop and usually only the bill touchesthe water: see Ashmole and Ashmole, 1967.) The objectspicked up were mainly piecesof algae betweenabout 1 and 10 cm long. After obtainingone of these,and sometimesmandibu- lating it, the bird wouldfly a shortdistance (between about 1 and 50 m), drop it, and then frequentlyretrieve it. This procedurewas repeatedper- sistently,in one caseat least ten times with the sameobject. This kind of behaviorin one juvenile often attractedothers, which either competedfor the sameobject or picked up different ones. Occasionallya bird showed anothertype of behavior,flying low and fast and executingrapid twists and turns without touchingthe surface. Unlike the plungingand dipping movements,which were generallyor always triggeredby the sight of an objectin the water, theseaerobatics were apparently not dependenton any specificexternal stimulus, but occurredspontaneously. The huntingpractice occurred at all timesof day--perhapsespecially in the early morning--but it was very intermittent; there were often periods of as muchas half an hour when all the juvenileswere resting,followed by a shortburst of activity. To give an idea of the intensitythis activity could 98 Asn•ro]zsAtqD Tovaa S., ProlongedParental Care [ Auk [ Vol. 85 reach,in the 16 minutesbetween the time we noticed that one of the juve- nileswas recognizableby a gap in one wing and the time we lost it round a cornerof the island,this individualmade 8 separatesorties from its base on an offshorerock, and in the courseof thesemade 40 to 50 contactswith the water. The great majority of th'esecontacts were plungesto surfaceto retrieve objects,but a few were followedby bathing behavior. Adult Inca Terns plungeto surfacein a similar way when they start bathing as well as while hunting,but we have never seenthem pickingup inedibleobjects. During our observationswe often saw the juveniles beg for fish from adults,and feedingwas severaltimes observed. So far as we couldtell the juvenilesat this age do not go out to seaand fish, but spenda considerable proportionof their time and energy developingtheir hunting skills. Plung- ing to surfaceand contact dipping are the two normal feeding methodsof the species,and both must be executedwith great speedand skill if they are to be successful.The value of practicing them before independenceis self-evident. We hope this brief discussionmay serve to emphasizethe need for more information about the critical period when young birds are learning to catch prey for themselves.In particular it would be of great interest to have for a numberof speciesquantitative data on the proportionof prey suppliedby the parents,and the proportioncaught by th'eyoung bird at different stagesduring the postfiedgingperiod. Anotherpossible approach is exemplifiedby observationsN. P. A. made on AscensionIsland in 1958, whereAscension Island Frigate-birdsfrequently preyed on small chicksof SootyTerns (Sterna ]uscata). The frigate-birdswould patrol over the tern colonies,swooping at intervals to try to pick up exposedchicks without alighting. Casual observationssuggested that adults were more often suc- cessfulat this than white-headedjuveniles, so a tern colony was watched for th'ewhole of two days,22 and 28 February 1958. During the first day adultsobtained 25 chicksin 89 attempts,while juvenilesobtained 6 chicks in 44 attempts. On the secondday adults took 32 chicksin 129 attempts, but juvenilesmade only 9 classifiableattempts, of which 2 were successful. These data suggestthat the adults were more efficient than juvenilesin the use of this huntingtechnique, but the differenceis nonsignificant.It was not possibleto collectfurther data on this species,but we hope to in- vestigatethe acquisitionof hunting skill in terns by making similar observations on them in the future.

ACKNOWLEDGMENTS

Field work on Peruvian sea birds was supported by NSF grant GB-3983 to N. P. A., while H. Laurence Achilles also gave vital assistance. R6mulo Jordan, David Lack, and Charles G. Sibley kindly read and criticized the manuscript. Jan.1968 ] AS•tMO•,EANDTOVAR S.,Prolonged Parental Care 99

SUMMARY During December 1966 and January 1967 adult Royal Terns wintering on the coastof Peru were seento feed juvenileshatched about six month's previouslyin North America. Feedingof youngby parentsafter departure from the breeding grounds has been recorded in other membersof the genusThalasseus and in other Laridae, but the age of the young and the distance traveled make this case unusual. Some instancesof prolonged parental care in other groupsare briefly reviewed; its occurrenceappears to be correlatedwith the use of highly skilled feeding methods and the exploitationof scarcefoods, and thus with low dutch size. Speciesshowing it do not normally have secondbroods. Juvenilesof speciesexploiting ac- tive prey animalsdifficult to catch often show"play" activitiesrelated to the acquisition of hunting skill. More information is needed about the critical period of transition to independencein young birds.

LITERATLIRE CITED

AM^•)ON,D. 1964. The evolution of low reproductive rates in birds. Evolution, 18: lo5-110. AS•t•OLE, N.P. 1963a. The regulation of numbers of tropical oceanic birds. Ibis, 103b: 458-473. AS•-•rOLE,N.P. 1963b. Molt and breeding in populations of the Sooty Tern Sterna /uscata. Postilia (Yale Univ.), no. 76: 1-18. ASltMOLE, N. P., AI,IDM. J. ASI-LMOLE.1967. Comparative feeding ecology of sea birds of a tropical oceanicisland. Peabody Mus. Nat. Hist., Yale Univ., Bull. 24. BE•T, A.C. 1940. Life historiesof North American cuckoos,goatsuckers, humming- birds and their allies. Ll. S. Nat. Mus., Bull. 176. BRow•r, L.H. 1952-53. On the biology of the large birds of prey of the Embu district, Kenya Colony. Ibis, 94: 577-620; 95: 74-114. BRow•r, L.H. 1955. Supplementary notes on the biology of the large birds of prey of Embu district, Kenya Colony. Ibis, 97: 38-64, 183-221. BROW•, L.H. 1966. Observationson some Kenya eagles. Ibis, 108: 531-572. BROW•, R.L. 1934. Some breeding-habits of Kingfishers. Brit. Birds, 28: 83-84. C^D•, T. J. 1960. Ecology of the peregrine and gyrfalcon populations in Alaska. Univ. Calif. Publ. Zo/31.,63(3): 151-290. D^vms, S. J. J. F., A•D R. C^RRICX. 1962. On the ability of CrestedTerns, Sterna bergii, to recognizetheir own chicks. Austral. J. Zool., 10: 171-177. FowleR, J. M., am) J. B. CovE. 1964. Notes on the Harpy Eagle in British Guiana. Auk, 81: 257-273. GIBSo•-Hr•., C. A. 1947. Notes on the birds of Christmas Island. Raffles Mus., Singapore,Bull., 18: 87-165. HOCHB^tr•r, H. A. 1955. Travels and traditions of waterfowl. Minneapolis, Univ. Minnesota Press. KOEVCXE,M. 1964. Las Aves del Departamento de Lima. Lima, published by the author. Mo•Ro•, B. L., JR. 1956. Observationsof Elegant Terns at San Diego, California. Wilson Bull., 68: 239-244. M•O•rRO,D.A. 1954. Prairie Falcon "playing." Auk, 71: 333-334. 100 ASaMOLE^•) Tov^}•S., ProlongedParental Care 1_[ Vol. Auk 85

NELSO•r,J.B. 1966a. The Man-o'-War Bird. Natural History, 75(5): 32-39. NELSO•r,J.B. 1966b. The breeding biology of the Gannet SuIa bassanaon the Bass Rock, Scotland. Ibis, 1011: 584-626. PALM•}•, R. S. 1941. A behavior study of the Common Tern (Sterna hirundo hirundo L.). Proc. Boston Soc. Nat. Hist., 4•.(1). P•T•LXA, F.A. 1958. Timing of molt in Steller Jays of the Queen Charlotte Islands, British Columbia. Condor, 60: 38-49. PYL•, R. L., ^•rv A. SMALL. 1951. Southern Pacific Coast Region. Audubon Field Notes, $: 307-309. SovT•R•r, H.N. 1954. Tawny Owls and their prey. Ibis, 06: 384-410. SOySmERe,H. N., R. V^VCnA•r, ^•rv R. C. Mv•}•. 1954. The behaviour of young Tawny Owls after fledging. Bird Study, l: 101-110. S•o•E•ovs•, B., A•V S. S•ro•n•ovs•. 1963. The Frigate Bird Fregata aquila of AscensionIsland. Ibis, 103b: 409-422. SWA•m•% P.O. 1952. Observationson feeding,brooding and bathing habits in a pair of Kingfishers (Alcedo arthis) [in Swedish, English summary]. V•tr. F•gelv•/rld, 11: 49-66. Tao}•, W.H. 1963. Learning and instinct in animals. 2nd ed. London, Methuen. T•mE}•r, N. 1953. The Herring Gull's world. London, Collins. ¾•}•r•, J. 1961. Nesting activities of the Red-footed Booby in British Honduras. Auk, 711: 573-594. Departmento/ Biologyand PeabodyMuseum o• Natural History, Yale University, New Haven, Conn., and Instituto del Mar del Per•, Callao, Lima, Peru.