J. Field Ormthol., 65(1):109 114

AERIAL FOOD TRANSFER AS A DEMAND BEHAVIOR IN THE MARSH

CARMELO FERN•,NDEZ 1 Estaci& Bwldg•cade Dor7ana CSIC Avda, M" Luisa Pabelldndel Perd, 41013 Sewlla, Spare

PAZ AZKONA Soc. L•arra Tafalla 34-4 -• 37003 Parnplona,Spain Abstract.--Sixtypercent of the foodsupplied by male Marsh Harriers (Circusaeruginosus) was transferredto the femalein the air when shewas presentin the territory. The frequency of aerial foodpassing increased as the breedingcycle advanced, from 26% during incubation to 77% in the late nestlingstages, and also increasedwith increasingbrood size. This may be explainedby a greaterfemale demand behavior as the foodrequirements of the brood grow. Thus aerial foodtransfer may be interpretedas a femaledemand behavior, which forcesthe male to increasehunting effort. The useof the aerial foodtransfer as a demand behavior among raptors should be limited to maneuverablespecies with reversedsexual dimorphismand female dominance.

EL TRASPASO DE CEBAS EN VUELO COMO CONDUCTA DE DEMANDA DE ALIMENTO EN CIRCUS AERUGINOSUS Sinopsis.--E160% de las cebasaportadas por los machosdel Aguilucholagunero (Circus aeruginosus),cuando las hembrasestaban presentes en el firea de nidificaci6n,fueron trans- œeridasen vuelo (n = 283). La proporci0nde cebasen vuelo aument6conforme avanzaba el cicloreproductivo, de 26% durante la incubaci6na 77% en la filtima fase de estanciade los polluelosen el nido, y con el tamafiode las polladas.Esto puedeset explicadopor una intensificaci0n en la conducta de demanda de alimento asociada a un incremento en las necesidadesalimenticias de los polluelos. Asl, el traspaso de cebas en vuelo puede ser interpretadocomo un comportamientode demandade alimento por parte de las hembras, que obliga a los machosa incrementarel esfuerzode caza. En las rapaces,el traspasode cebasen vuelodebe estar restringido alas especiesmils figiles, con dimorfismo sexual inverso y dominanciapor parte de las hembras. Male Marsh Harriers Circusaeruginosus transfer food items to females in flight during the incubationand nestlingphases (Cramp and Simmons 1980,Johannesson 1975). Hypothesesregarding the significanceof aerial foodpassing during the breedingperiod include: (1) maintenanceof pair bonds(Lack 1940, Watson 1977), (2) avoidanceof food lossby missing small nests(Newton 1986), (3) preventionof chick aggressiontowards small adult males (Newton 1978, 1986), and (4) reduction of predation on nests(Watson 1977) or at terrestrial transfer sites(Simmons 1991). In this studywe evaluatethe relativeimportance of threefood transfer methods(to the nest, on the ground and aerial) usedby Marsh Harriers during the breedingseason. We analyze variationsin the use of these

• Currentaddress.' Soc. Ugarra, Tafalla 34-4', 31003Parnplona,

109 1 10] C. Ferndnclezand P. Azkona J.Field Ornithol. Winter 1994 transfermethods during the breedingseason and in relation to broodsize. We suggestthat aerial food transfer during breeding may be a function of female demandbehavior, which informs the male of the hunger state of the youngand stimulatesmale hunting effort.

STUDY AREA AND METHODS The studywas conductedin spring 1990 on two nature reservewetland areas (El Juncal and Dos Reinos) in the Upper Ebro Valley (Navarra, Northern Spain). Ten monogamouslybreeding pairs were observed,two in E1 Juncal and eight in Dos Reinos (Fernfindez 1991). Observations were made by two observersequipped with binoculars (8 x 35) and telescopes(20-45 x 60) from fixed positionsabout 500 m from the nests. The breeding period was divided into five stages:incubation and four periodsof 11 d each during the nestlingstage. A total of 631 h of obser- vationswas made: 128 h during incubation, and 131, 124, 140 and 108 h, respectively,in each four nestlingstage. Aging of broodswas achieved by measuringbody developmentof the chicksin the nest (Altenburg et al. 1987). Time spent by either the male or the female on the nest and in the nestingarea (on the wetlands)was recorded.Time spentout of the nesting area was consideredas a measureof hunting time (Altenburg et al. 1982). For each food item delivered, a record was kept of: (1) sex of the adult supplyingprey; (2) presenceof female in the nestingarea; (3) type of delivery:to the nest(directly to the chickson the nest),on the ground(to the femaleout of the nest)or aerial transfer(to the femalein flight); and (4) successof aerial passes,noting the dropping and lossof food items. For statistical estimatesonly prey items supplied by the male when the female was present were considered.Statistical differences in the proportionof three typesof food transfer during the five breedingstages were tested by means of chi-squared test in 3 x 4 contingencytables (Sokaland Rohlf 1969). Similarly, we comparedthe frequenciesof three methodsof foodtransfer in relationwith the sizebroods (comparing pairs with 1, 2, 3 or 4 chicks) using chi-squared test. Also, we calculated Pearson'scorrelation coefficients (r) betweenthe proportionof foodtrans- fers by malesin flight, on the groundor in the nest and the five age intervals of the chicks (Sokal and Rohlf 1969).

RESULTS A total of 467 prey deliverieswas observed(0.74 food item/h) of which 372 (79.7%) were supplied by the male. In 283 (76.1%) of these 372 casesthe female was presentin the nestingarea. The femalessupplied directlyto the nest85 (18.2%) fooditems, probably hunted by themselves. The females almost never left the nest during incubation or the first stageafter hatching(1-11 d), whereasat the end of nestlingrearing (34- 44 d) femalesonly spent30.1% of their time in the nestingarea (Table 1). Males spent 43.8% of their time around the nest during incubation but only 25.5% at the beginning(1-11 d) and 12.6% at the end (34-44 d) Vol.65, No. 1 FoodTransfer zrz Marsh Harrzer [11 1

TABLE1. Time spentin nestingarea and numberof prey items suppliedby male and female Marsh Harriers during breedingperiod.

Nestling phase Incubation 1-11 d 12-22 d 23-33 d 34-44 d

% time spentby females in nestingarea 100 100 93.5 57.2 30.1 % time spent by males in nestingarea 43.8 25.5 21.7 12.8 12.6 Total food items suppliedby hour 0.43 0.53 0.81 0.97 0.88 % suppliedby females -- -- 8.0 29.1 41.6 % suppliedby males 100 100 92.0 70.9 58.4 of the nestlingstage. As a result,the numberof fooditems brought directly to the nestby the male when the femalewas absentwas higherin the latter stagesof breeding. The proportionof foodtransfers by malesin flight,on the ground,and in the nestwhen femaleswere present(Table 2) variedover the breeding period(X 2 = 81.53, df = 8, P < 0.001). Aerial transfersincreased as the breedingcycle advanced (r = 0.892,df = 4, P < 0.05). Conversely,ground transfers,very frequentduring incubation,decreased notably over the nestlingperiod (r = -0.980, df = 4, P < 0.001). Numbersof fooditems broughtdirectly by malesto the nestwhen femaleswere presentin the territorywere always very low (<23%), increasingslightly as the chicks developed(r = 0.977,df = 4, P < 0.001) (Fig. 1). On 12 occasionsfemales flew to receivefood but were refusedby the male (6.6%, n = 181), who deliveredfood on the ground(5.5%) or to the nest (1.1%). We noted a relationshipbetween brood size and type of food pass (Table 3). Aerial transferswere morefrequent in pairswith largerbroods (x2 = 17.95, df = 6, P < 0.01), though also this could be due to the greaterskills or experienceof the mostsuccessful breeding pairs.

TABLE2. Frequencyand variationin the modesof foodpass from male to femaleMarsh Harriers during breeding period.

Incuba- Nestlingphase tion 1-11 d 12-22 d 23-33 d 34-44 d Total

Total food items suppliedby males 55 70 94 97 56 372 Food supplied by males when femalespresent 55 70 79 53 26 283 Transfers in nest 0 6 9 8 6 29 Transferson ground 41 24 14 6 0 85 Aerial transfers 14 40 56 39 20 169 Food itemsdropped (lost) 0 2 (1) 3 (0) 1 (1) 0 6 (2) 112] C. Ferndndez and P. Azkona J. Field Ornithol. Winter 1994

IO0

80-

60- ß AERIAL 1. eJe ß ON GROUND ß IN NEST 40-

20-

I I I I 1.11 12.22 23.33 34'44 DAYS INCUBATION , NESTLING PHASES '

FIGURE 1. Variation in the proportion of three types of food pass (in the nest, on the groundand aerial) from male Marsh Harriers during breedingperiod when females were presentin the nestingarea.

Food items were dropped in six (3.6%) of the 169 aerial transfers attempted. Two of these items were lost, with adults unable to find them either in the water or amongcattails (Typha spp.). This representsa loss of 1.2% of aerial delivered food items.

DISCUSSION Various authorshave suggestedthat aerial food passingin raptors is a meansof pair bonding(Lack 1940), or a systemto avoidfood lossby missing small nests (Newton 1986) or nests situated in areas of dense vegetation(Watson 1977). Another hypothesisis that, by transferring foodaway from the nest,the small adult male avoidsaggression from his large (particularlyfemales) offspring (Newton 1978, 1986). Finally, in the caseof harriers,there are two additionalhypotheses that may explain aerial transfers:(1) aerial foodpassing is a methodto avoidnest detection and nest predationin a genuswith particularly conspicuousmales and nestson the ground (Watson 1977), and (2) the safe substrates(such as trees or cliffs) usedby other raptors to transfer food are not availableto harriers(Simmons 1991). Thesehypotheses, however, hardly explain the changesin frequencyof food passingthat take place in Marsh Harriers during breeding and with brood size. During our research we observedthat there were females who con- trolledthe modeof foodpassing, because the femaleswere rarely refused Vol.65, No. I FoodTransfer in MarshHarrier [ 1 13

TABLE 3. Food passingpatterns from male to female Marsh Harriers dependingon brood size.

Type of food pass Brood size n In nest On ground Aerial Total 1 chick 2 5 (17.9%) 6 (21.4%) 17 (60.7%) 28 2 chicks 3 14 (17.1%) 24 (29.3%) 44 (53.7%) 82 3 chicks 1 1 (5.9%) 4 (23.5%) 12 (70.6%) 17 4 chicks 4 9 (8.9%) 10 (9.9%) 82 (81.2%) 101 Total 10 29 44 155 228

by males when they flew to receivefood (cf. Simmonset al. 1987). We suggestthat, besidesother functions,aerial foodtransfer during breeding has a componentof demand behavior that servesto inform the male of the hunger state of the female and the young. Thus, the progressive increase in aerial transfers would be due to an increase in the brood's foodrequirements, which stimulatesthe femaleto go out to collectfood from the male before he has time to deliver it to the nest or leave it on the ground.The fact that aerial passeswere more commonin pairs with larger brood sizesand greater food requirementsseems to supportthis view. As a demandbehavior the aerial foodtransfer may be similar and complementaryto dismissalscreams and aggressiondescribed in other of prey (Balfour 1957; Haverschmidt 1953; Newton 1978, 1979, 1986; Schnell 1958; Sherrod 1983; Watson 1977) and could be related to reversedsexual dimorphism and female dominance(Smith 1980). The main benefitof the female forcingthe male into aerial foodpasses would be to increasemale foraging rate (Cade 1960, Schnell 1958). The resultsof Altenburget al. (1982) as well as thosefrom this studysuggest that monogamousmale Marsh Harriers spendmore time foragingas the breedingcycle advances. Increased male foragingtime couldresult in two ways: (1) time savedin foodtransfer, as suggestedby Simmons(1991), and (2) demandbehavior of femalesstimulating the maleswhen females come out to receive prey items. Undoubtedlythe advantagesof aerial transfersare partially offsetby the risk of food loss (Watson 1977), but as we have seen, lossesin the Marsh Harrier, as well as in other harriers (Jimanez and Jaksic 1988, Simmons1991), are insignificant.Losses are probably compensatedby extra hunting time and the consequentincrease in food items supplied by the male. In other raptors,which are not able to capturebirds in flight, or whosestaple prey is heavy (e.g., large mammals) the risk of foodloss would be much greater and not compensatedby increasedfood searching time. Then one would expectthe use of aerial food passesas a demand behavioramong birds of prey that are aerially maneuverable,that hunt handy prey and that havereversed sexual dimorphism, as occurin genera Circus,Falco and Accipiter (Cade 1960,Jim•nez andJaksic 1988, Newton 1986, Simmons 1991, Watson 1977). 114] C. Ferndndezand P. Azkona J.Field Ornithol. Winter 1994

ACKNOWLEDGMENTS

This study was supportedby the Environmental Serviceand an FPI-grant from the Service of Universities and Research of the Navarran Government in collaboration with the Dofiana BiologicalStation (CSIC). D. , J. A. Donfizar, F. Hiraldo, S. K. Sherrod and R. E. Simmonscontributed with their commentstoward improving the original manuscript. N. C. B. Bowles translated the paper into English.

LITERATURE CITED

ALTENBURG,W., J. BRUINENBERG,P. WILDSHUT,AND M. ZIJLSTRA. 1987. Colonization of a new area by the Marsh harrier. Ardea 75:213-220. , S. DAAN, J. STARKENBURG,AND m. ZIJLSTRA. 1982. Polygamy in the marsh harrier, (Circusaeruginosus): individual variation in hunting performanceand number of mates. Behaviour 79:272-312. BALFOUR,E. 1957. Observationson the breedingbiology of the in Orkney. Bird Notes 27:177-183. CADE, T.J. 1960. Ecology of Peregrine and Gyrfalcon populationsin AlaskaßUniv. California Publ. Zool. 63:151-290. CRAMP, S., AND m. E. SIMMONS. 1980. The birds of the western Palearctic, Vol. II. Oxford Univ. Press.Oxford, . 695 pp. FERNANDEZ,C. 1991. Censo,fenologia y 6xito reproductordel Aguilucholagunero (Circus aeruginosus)en Navarra. Munibe 41:89-93. HAVERSCHMIDT,F. 1953. Observationson the Marsh harrier with particular reference to clutch-sizeand nestingsuccess. Brit. Birds 46:258-259. JIMgNEZ, J. E., AND F. M. JAKSIC. 1988. Ecology and behaviour of southernSouth American Cinereous Harriers Circus cinereus. Revista Chilena de Historia Natural 61: 199-208. JOHANNESSON,H. 1975. Activitiesof breedingMarsh Harriers (Circusaeruginosus). Var FagelvSrld 34:197-206. LACIq,D. 1940. Courtship feeding in birds. Auk 57:169-178. NEWTON,I. 1978. Feeding and developmentof Sparrowhawk (Accipiternisus) nestlings. J. Zool. Lond. 184:465-487. 1979. Populationecology of raptors.T. and A.D. Poyser.Calton, United King- do•.399 pp. ß 1986. The Sparrowhawk.T. and A.D. Poyser.Calton, United Kingdomß396 pp. SCHNELL,J.H. 1958. Nestingbehavior and foodhabits of Goshawksin the Sierra Nevada of California. Condor 68:113-151. SHERROD,S.K. 1983. Behavior of fledgling Peregrines.The PeregrineFund, Inc. Ithaca, New York. 202 pp. SIMMONS,R. E. 1991. The efficiencyand evolutionof aerial food passingin harriers. Gabar 6:51-56. --, P. E. BARNARD,AND P. C. SMITH. 1987. Reproductivebehaviour of Circuscyaneus in North America and : a comparison.Ornis Scand. 18:33-41. SMITH, S.M. 1980. Demand behavior:a new interpretationof courtshipfeeding. Condor 82:291-295. SOKAL,R. R., AND F. J. ROHLF. 1969. Biometry. Freeman and Co., San Francisco, California. 832 pp. WATSON,D. 1977. The Hen Harrier. T. and A.D. Poyser.Berkhamsted, United Kingdom. 307 pp. Received:22 Feb. 1993; accepted5 May 1993.