The Condor 96:331-340 0 The CooperOmithological Society 1994 THE KLEPTOPARASITIC INTERACTIONS BETWEEN GREAT FRIGATEBIRDS AND MASKED BOOBIES ON HENDERSON ISLAND, SOUTH PACIFIC ’ J. A. VICKERY2 Universityof East Anglia, Schoolof Biological Sciences,Norwich, NR4 7TJ, U.K. M. DE L. BROOKE Department of Zoology, Cambridge University,Downing Street, Cambridge,CB2 3EJ, U.K. Abstract. Kleptoparasitic activities by Great Frigatebirds (Fregata minor) on Masked Boobies (Sula dactylatru) were quantified on Henderson Island in the South Pacific. The frequency of kleptoparasitic attacks increased towards dusk, the time when most Masked Boobiesreturned to the roost site. Masked Boobiesreturning to the roost low (i 30 m) over the water were significantlymore likely to be chasedthan thosereturning at a higher altitude. Single boobies were also significantly more likely to be attacked than boobies returning as one of a group of birds. However, group size and height were positively correlated. There was no effect of group size on the likelihood of attack. Sixteen percent of chases were successfuland the successrate was not influenced by either the height or the distance of the target from the roost. The duration of the chase was, however, influenced by the position of the target: chases on distant or high targets lasted significantly longer than chaseson targetsthat were low or close to the roost. In addition successfulchases were significantly longer than unsuccessfulones. We discussthese results in relation to the chase tactics of frigatebirds, the avoidance tactics of boobies and the energetic costs and benefits of klep- toparasitism. Approximately 40% of daily energy expenditure of some individual Great Frigatebirds may be securedthrough kleptoparasitism. However, on average, frigatebirds may be meeting under five percent of their daily energy demands by this feeding method. Key words: Boobies;chase duration; energeticxfrigatebirds; kleptoparasitism; Pa& Ocean. INTRODUCTION and has been well studied in the latter three (e.g., Kleptoparasitism, the stealing by one animal of Dunn 1973, Anderson 1976, Verbeek 1977, Bur- food which has already been caught by another, ger 198 1, Fumess and Hislop 1981, Birt and occurswidely throughout the animal kingdom in Cairns 1987). However, despite the fact that frig- a range of taxonomic groups including insects, atebirds are considered among the most spe- fish, reptiles, birds and mammals (Thompson cialized kleptoparasites (Fumess 1987), few de- 1986). Although kleptoparasitismmay reducethe tailed studies of the interactions between costs of foraging, by using the time and energy frigatebirds and their hosts exist (Fumess 1987, investment of others, it involves a high degree Osomo et al. 1992). of specialization and is likely to be profitable only The aims of this study were to carry out a under certain ecologicalconditions, for example, detailed study of the kleptoparasitic behavior of where numerous potential hosts carry large the Great Frigatebird (Fregata minor) on one of quantities of food and behave in a predictable its main target species,the Masked Booby (Sula way (Brockmann and Barnard 1979). dactylatru) and to evaluate the importance of this Kleptoparasitism is an important feeding feeding method in satisfying the energy needs of method in four seabird families; Fregatidae (frig- the frigatebirds. Data are presented on the daily atebirds), Chionididae (sheathbills), Ster- activity pattern of Great Frigatebirds and Masked corariidae (skuas) and Laridae (gulls and terns) Boobies over inshore waters around Henderson Island, South Pacific;the nature of the flight paths adopted by Masked Boobies as they return to I Received1 September 1993. Accepted 10 January their roost site; the location, duration and out- 1994. come of kleptoparasitic attacks;and the number, * Present address: Division of Biological Sciences, size, and species of fish that may be obtained University of Edinburgh, Ashworth Laboratories, Ring’s Buildings, West Mains Road, Edinburgh EH9 from a successfulattack. This information is then 3JT. used, in conjunction with published figures for [3311 332 J. A. VICKERY AND M. DE L. BROOKE metabolic costsof flight and energy values of fish, came into view. Any time during which no in- to assessthe proportion of the daily energy re- dividuals were present was recorded. quirements of Great Frigatebirds that may be obtained from kleptoparasitism. FLIGHT PATHS OF MASKED BOOBIES RETURNING TO THE ROOST SITE STUDY AREA AND METHODS Detailed observations were made of the flight The study was carried from April to June 199 1, paths of Masked Boobies returning to the roost on Henderson Island (24”22’S, 128”2O’W), one site during ten 2-hr watches at dusk (16:00 to of the Pitcairn Islands in the South Pacific. Hen- 18:00 hr). Individual Masked Boobies were derson is a raised coral atoll with limestone cliffs sighted up to 600 m offshore and followed until reaching 34 m above sealevel, a fringing reef 50- they had either landed at the roost site or were 100 m wide and three beaches;one on each of lost from view out to sea. For each individual the north, west and east coasts(Fig. 1). The East bird the following information was recorded: (i) Beach was approximately 2 km long with a 20- whether the booby was single or one of a group 30 m wide strip of vegetation, primarily Pan- and the size of the group, (ii) the height at which danus tectorius and Argusia argentea, between the booby crossedthe reef (estimated to the near- the beach and the base of the cliffs. During the est 5 m in relation to the height of the cliff) and study period, 70 to 80 Masked Boobies roosted (iii) whether it was chasedand, if so, whether the at the base of the cliffs on East Beach. The roost chasewas successful(a chasewas consideredsuc- site was located in a natural clearing on the sea- cessfulif the booby was forced to regurgitateand ward edge of the beach vegetation. All obser- the frigatebird obtained the food). The same in- vations of frigatebirds and boobies were made formation was recorded for the small number of from a position on the edgeof the vegetation 150 Red-footed Boobies (Sula sula) that were also m south of the roost. seen crossingEast Beach to roost further inland. A maximum of 126 Great Frigatebirds was recordedover East Beachduring the study period KLEPTOPARASITIC BEHAVIOR OF GREAT FRIGATEBIRDS (pers. observ.). Frigatebirds on Henderson nest- ed and roosted in the forest of the interior pla- Individual frigatebirds were followed during ten teau. Following laying in June, the number of 2-hr watches at dusk (16:00 to 18:00 hr). For frigatebirds continued low, probably under 100 each kleptoparasitic attack the following infor- pairs. mation was recorded: (i) the time at the start of the chase (taken as the moment the frigatebird DAILY ACTIVITY PATTERN OF GREAT switched from gliding to direct flapping flight FRIGATEBIRDS AND MASKED BOOBIES towards the booby, (ii) the distance or initial lo- The numbers of Masked Boobies returning to cation of the target booby from the roost site roost and the activity of frigatebirds were re- (identified as being in one of three zones: over corded during five watches from 06:OOto 18:OO the beach, between the reef and the beach or hr and five additional 2-hr watches from 16:00- beyond the reel), (iii) the height of the target 18:00 hr (local time = GMT less 8.5 hr). Indi- above the sea or beach, (iv) whether the target vidual frigatebirds were followed using 10 x 40 was single or one of a group, (v) the duration of binoculars and the total time spent perching, the chase, (vi) the outcome of the chase and, if gliding (slow, often circling flight with few or no successful,(vii) the number of fish the frigatebird wing beats) and chasing (rapid, flapping flight) obtained. was recorded during each 2-hr observation pe- Information on sneciesand mass of fish that riod. Individual birds were selected at random were carried by boobies to the roost was obtained and followed constantly until they were lost from from regurgitateswhich usually comprised easily view as they flew either out to sea, around the separated,barely digested fish. Masked Boobies cliffs to the north or south of the beach, or over- will readily regurgitate food if disturbed by peo- head and into the interior of the island. The time ple. Fish were obtained from this sourceand also at which a bird was lost from view was recorded as a result of chaseswhere the frigatebird either and observations were switched to another ran- missedor dropped the regurgitatedfood. The fish domly selectedindividual. This switch was made were identified to family and weighedto the near- immediately or as soon as another individual est 0.5 g using a O-100 g Pesola spring balance. KLEPTOPARASITIC BEHAVIOR OF GREAT FRIGATEBIRDS 333 Pacific Ocean Christmas Island Equator . _____------___-------__--_--_________--------__________ Tahiti a ~_-------__-_--~--~~~--------------------- Oenos+Henderson__________-__________________ Pitcairns Duois . Easter Island 1 New Zealand Henderson Island I- 1Km *Masked Soobv Roost Site FIGURE 1. The study site. RESULTS inshore waters for approximately half of the time (mean = 51.0%, 95% Confidence Intervals [CI] DAILY ACTIVITY PATTERN OF MASKED = 39.1-62.8%, 30 2-hr periods; arcsinetrans- BOOBIES AND GREAT FRIGATEBIRDS n = formations). When visible over theseinshore wa- Throughout the day (06:OO to 18:00 hr) one or ters frigatebirds spent 16.9% (95% CI = 1 l.O- more frigatebirds were visible in the air over 23.9%) of their time perched, 82.0% (95% CI = 334 J. A. VICKERY AND M. DE L. BROOKE u .9 time spent chasing by frigatebirds ki a q time spent perching by frigatebirds 5 c] time spent gliding by frigatebirds _? (a) 0 number of returning boobies 6-8 8-10 10-12 12-14 14-16 16-18 Time of day loo- 80 - 60 - 40 - 1600 1630 1700 1730 1800 Time of day FIGURE 2.