Southern Fulmars Molt Their Primary Feathers While Incubating ’

SHORT COMMUNICATIONS 563

The Condor 100563-566 0 The Cooper Ornithological Society 1998

SOUTHERN FULMARS MOLT THEIR PRIMARY FEATHERS WHILE INCUBATING ’

CHRISTOPHEBARBRAUD AND OLIVIER CHASTEL C. E. B. C./C. N. R. S., 79360 Villiers en Bois, France, e-mail: petrel@cebc.cnrs,fr

Abstract. Molt-breeding overlap and the extent of or soon after the egg hatches (Hunter 1984). Warham molt of primary feathers were studied in the Southern (1962) also observed some Cape Petrels (Duption ca- Fulmar (Fulmarus glacialoides) in relation to breeding pense) losing their primaries in late incubation and status, sex, breeding experience, and age. Primary molt during chick rearing, and more recently Monteiro and was observed among breeders and nonbreeders during Fumess (1996) found Cory’s Shearwaters (Culonectris the incubation period. Nonbreeders renew more feath- diomedea) molting their primaries during chick rear- ers than breeders. In breeding individuals, males and ing. females renew the same number of feathers. No dif- Here, we report primary molt-breeding overlap in ference in primary molt score was found between in- Southern Fulmars (Fulmarus glacialoides) during inexperienced and experienced breeders. However, the cubation. Southern Fulmars breed during the brief aus- extent of primary molt was significantly related to age tral summer in Antarctica and islands of the southern among breeders; older birds had less new feathers than ocean. Laying occurs in early December, egg hatching younger birds. These results suggest that primary molt- lasts half of January, and departure occurs about mid- breeding overlap in Southern Fulmars is a consequence March to April (PrCvost 1953). Birds breed annually of high food availability during the short summer and and body molt begins during incubation in early Jan- a reallocation of energy into molt during incubation. uary (Bierman and Voous 1950). These latter authors Possible reasons for differences in molt extent owing noted that outwards primaries molt from late Decem- to age and breeding status are discussed. ber to early April, but gave no information concerning Key words: breeding, Fulmarus glacialoides, molt, the status and age of birds inspected. Southern Fulmar. We document simultaneous breeding and primary molt in this fulmarine petrel, but also examine the wing molt in relation to the breeding status and age of Temporal overlap between molt and breeding is generally uncommon in bird species because of the great the birds using a population of birds of known age and energy demands of each process (Payne 1972, Ricklefs breeding status. 1974). However, overlap of breeding and molt has been found in several bird families, especially in the METHODS tropics (Payne 1969, Foster 1974). In seabirds, molt- The study was carried out at Pointe GCologie, Terre breeding overlap is observed in Arctic/Antarctic spe- AdClie (66”2O’S, 14O”Ol ‘E) in January 1994. At Pointe cies where resources are abundant for only a brief pe- Gtologie, the entire population of Southern Fulmars riod (Carrick and Dunnet 1954, Fumess 1988). Most (20 to 50 breeding pairs each year) has been banded of these studies indicate that body molt begins during since 1963, and thereafter captured every year (Wei- incubation and that molt of the primary feathers in merskirch 1990a). Thus, most birds were of known age successful breeders does not commence until the and breeding experience. Birds were checked during young are near fledging, whereas in unsuccessful the last part of the incubation period (l-10 January). breeders primary molt may begin within a week of egg Molt scores were collected from primary feathers on or chick loss. the right wing only as there is no apparent difference Some authors suggest that the selective advantage of delayed primary molt in breeding birds is to retain between both wings (Weimerskirch 1991). The scoring maximum wing efficiency while feeding young (Beck method used was that of Ashmole (1962) in which 1970). In most seabirds of the order Procellariiformes, feather growth for each primary is estimated on a scale the primary molt period and the breeding period are of 0 (old) to 5 (new). Scores for each primary are then separated (Streseman and Streseman 1966). For these summed to obtain an aggregate for the wing. We species the molt-breeding overlap is narrow and birds checked both breeders and nonbreeders. On average, usually do not shed their primaries until their chick is 34% of experienced breeders skip breeding each year close to flying (Warham 1996), the only exception be- in Adelie Land (Chastel, unpubl. data). In this study, ing the two species of Giant-Petrel (Mucronectes gi- nonbreeders refer to birds known to have bred in a ganteus and M. hali) in which wing molt starts before previous season. The extent of primary molt was studied in relation to breeding status and age. All statistical analyses were conducted using SYSTAT (1992). A sig- I Received 17 November 1997. Accepted 16 March nificance level of P < 0.05 was used for all tests. Val- 1998. ues are presented as mean + SD. 564 SHORT COMMUNICATIONS

TABLE 1. The mean primary scores for Southern Fulmar breeders, nonbreeders,unsuccessful and successful breedersin January 1994 at Pointe Gtologie (sample sizes in parentheses).

Breedinr status Primary scores (ia P Nonbreeders 8.89 -c 3.41 (9) Breeders 3.37 ? 1.80 (49) 54.5

a Mann-Whitney U-test

RESULTS mar-us glacialis) primary molt starts after departure We inspected 58 birds and all were found in active from the breeding colony (late August) after successful primary molt in early January.There was a significant nesting, earlier i? failed (Cramp and Simmons 1977, difference between breeders and nonbreedersin pri- Brown 1988). However. Cat-rick and Dunnet (1954), mary scores (Table 1). Nonbreedershad significantly reported few birds molting their primaries while rear- more new primary feathers emerging or growing than ing a nestling. Molt-breeding overlap can occur only breeders (G-test, G, = 22.9, P < 0.001). There was if energy and nutrients are available and adequatefor no significant difference between unsuccessfulbreed- both activities to occur simultaneously.Foster (1974) ers and successfulbreeders in primary scores (Table and King and Murphy (1985) proposedtwo hypotheses 1). Considering breeding experience, first time breed- to account for this overlap: a dependenceon nutrient ers did not differ significantly from experiencedbreed- reserves and a reduction or reallocation in energy or ers in their primary molt scores,although there was a nutrient expenditure, or both. The absenceof weight tendency for inexperienced birds to exhibit lower changeat the beginning of shifts during incubationand scores (2.80 t 0.84, n = 5; 3.43 t 1.87, n = 44; the short time spent on foraging trips suggestthat food respectively; F,,,, = 0.5). In breeding birds, the pri- availability is high and/or food is readily accessibleto mary score was significantly correlated with the age Southern Fulmars (Weimerskirch 1990b). Moreover, of the birds, older birds having lower scores (Fig. 1). Weimerskirch suggestedthat during the incubationpe- There was no significant difference in primary scores riod the birds could have spent longer shifts on the between males (3.53 2 1.79, n = 17) and females nest, consequentlylosing more weight and consuming (3.44 ? 1.83, n = 27; F,,,, = 0.03). more energy. This result and ours suggest that Southern Fulmars might reduce or reallocate the en- DISCUSSION ergy or nutrient expenditure into primary molt during There is a clear primary molt-breeding overlap in the incubation. Southern Fulmar at Pointe Geologic, as primary molt The stage of molt was more advancedin nonbreed- was observed during the last part of the incubation ers than breeders,probably reflecting the trade-off be- period. Although for Procellariiformes the incubation tween present reproductive effort and the timing of period is less energy demandingthan the brooding pe- molt (Weimerskirch 1991). One oossible exnlanation riod (Ricklefs 1983), our result contrastswith previous is that nonbreedersstart molting primary feathers ear- studieson small fulmarine petrels indicating a delaved lier than breeders. Becausethey do not invest energy molt of the primary feathers until the yo&g is near into reproduction,nonbreeders could divert the energy fledging (Carrick and Dunnet 1954, Brown 1966, Beck not used for maintenanceinto the molting process.On 1970). In the closely related Northern Fulmar (Ful- the other hand, breedersand nonbreederscould initiate molting similarly, but breederscould delay their molt or molt at a slower rate than nonbreedersdue to their high energy demand for reproduction. However, we did not study molting during the entire breeding cycle and cannot verify theseshypotheses. The stage of molt was identical between unsuccessful breeders and successfulbreeders during the study period. This result contrastswith previous studies on albatrosses indicating that failed breeders replaced more primaries than successfulbreeders and where primary molt and breeding do not overlap (Furness1988, Weimerskirch 1991, Prince et al. 1993). When birds were inspected, unsuccessful breeders were birds 0 1 , / I / which had lost their egg. Becauseprimary molt usually 10 20 30 40 begins within a week after egg loss (Warham 1990), Age (years) the time interval was probably too short to detect differences in primary molt between unsuccessfuland FIGURE 1. Primary scoresin breeding male and fe- successfulbreeders at this stage of breeding. male SouthernFulmars in relation to age (r = -0.48, Males and females had similar molt scores,probably n = 31, P < 0.01). indicating identical molting strategiesfor both sexes. SHORT COMMUNICATIONS 565

In Southern Fulmars, males and females do not differ ritories directed by I? Jouventin. We thank Peter Prince significantly in weight losses during incubation and for suggestions and comments on an earlier version of chick brooding, probably indicating similar energy ex- this manuscript. penditure for both sexes (Weimerskirch 1990b). It ap- pears that sex-specific differences in molting strategies LITERATURE CITED occur in species where males and females differ significantly in different aspects of their morphology, be- ASHMOLE, N. F! 1962. The Black Noddy Anous ren- havior, and life history (Payne 1972, Weimerskirch uirostris on Ascension Island. Part 1. General bi- 1991), which could partly explain the absence of dif- ology. Ibis 103:235-273. ference found in Southern Fulmars. BECK, J. R. 1970. Breeding seasons and moult in Studies conducted on albatrosses have shown that some smaller Antarctic petrels, p. 542-550. In M. inexperienced breeders delayed primary molt com- W. Holdgate [ed.], Antarctic ecology. Vol. 1. Ac- pared to experienced birds (Furness 1988, Weimer- ademic Press, London. skirch 1991). Although inexperienced Southern Ful- BIERMAN, W. H., AND K. H. Voous. 1950. Birds ob- mars tended to have lower scores than experienced served and collected during the whaling expedi- birds, the difference was not significant. This might tions of the ‘Willem Barendsz’ in the Antarctic, have been caused by the small sample size for inex- 1946-1947 and 1947-1948. Brill, Leiden, Neth- perienced birds. erlands. The stage of molt was clearly related to age of birds BROWN, D. A. 1966. Breeding biology of the Snow among breeders. Older birds molted less primary Petrel Pago&ma nivea (Forster). ANARE Sci. feathers than younger birds. Conversely, in albatrosses Rep. Ser. B.(l) Zool. Pub. 89:1-63. where primary molt and breeding do not overlap, first BROWN, R. G. B. 1988. The wing-moult of fulmars time breeders-renewed less prim&es than experienced and shearwaters (Procellnriidae) in Canadian and older birds (Weimerskirch 1991). One oossible Arctic waters. Can. Field-Nat. 102:203-208. cause might be that among breeding fulmars, young CARRICK, R., AND G. M. DUNNET. 1954. Breeding of birds started breeding later than old birds. Indeed, the fulmar Fulmarus glacialis. Ibis 96:356-370. among Procellariiformes experienced and old birds CHASTEL, 0. 1995. Influence of reproductive success usually lay earlier than inexperienced and young birds on breeding frequency in four southern petrels. (Warham 1990). In Southern Fulmars, Weimerskirch Ibis 137:360-363. (1990a) found that inexperienced birds hatched their CRAMP, S. 1977. Handbook of the birds of Europe, eggs later than experienced birds. Consequently, young the Middle East and North Africa: the birds of the birds might have to start molting earlier than old birds Western Palearctic. Vol. I. Oxford Univ. Press, to complete molt before the end of summer. Oxford. But why did not Southern Fulmars delay molting FOSTER,M. S. 1974. A model to explain molt-breed- after breeding as apparently do other small fulmarine ing overlap and clutch size in some tropical birds. species? One possible cause might be that Southern Evolution 28:182-190. Fulmars breed during a short summer season during FURNESS,R. W. 1988. Influences of status and recent which resource availability is high. Similarly, Hunter breeding experience on the moult strategy of the (1984) suggested that overlap between molt and breed- Yellow-nosed Albatross Diomedea chlororhyn- ing in Giant Petrels is a consequence of very abundant chos. J. Zool. 215:719-727. and easily available summer food supplies (mainly seal HUNTER, S. 1984. Moult of the Giant Petrels Mucro- carrion from pupping fur seals). Indeed, the fact that nectes halli and M. giganteus at South Georgia. both breeding and wing molt, which are high energy Ibis 126:119-132. demanding processes, occur simultaneously could KING, J. R., AND M. E. MURPHY. 1985. Periods of partly explain why Southern Fulmars lose weight at nutritional stress in the annual cycle of endo- the highest daily rates known in Procellariiformes therms: fact or fiction. Am. Zool. 25:955-964. (Weimerskirch 1990b). Moreover, as fledging occurs MONTEIRO, L. R., AND R. W. FURNESS. 1996. Molt of in late March to early April when food availability Cory’s Shearwater during the breeding season. decreases in Antarctic waters, molting might be too Condor 98:216-221. energy demanding at this time of the year. Molting and PAYNE, R. B. 1969. Overlap of breeding and molting restoring body reserves after the breeding season might schedules in a collection of African birds. Condor compete for energy, and this could have a negative 71:140-145. influence on winter survival or future reproduction. Fi- nally, as suggested for other petrel species (Hunter PAYNE, R. B. 1972. Mechanism and control of molt, 1984, Chaste1 1995). if Southern Fulmars delaved molt p. 103-155. In D. S Farmer and J. R. King [eds.], after breeding, timing also may be critical to b&h com- Avian biology. Vol. 11. Academic Press, New plete molt and restore body condition before the next York. breeding attempt. PROVOST,J. 1953. Notes sur la reproduction du Fulmar Antarctique Fulmarus glacialoides. Alauda 21: This study was supported by the administration of 157-164. Terres Australes et Antarctiques FranGaises and the by PRINCE, P A., S. RODWELL, M. JONES,AND F?ROTHERY. the Institut FranGais pour la Recherche et la Technol- 1993. Moult in Black-browed and Grey-headed ogie Polaires which provided the logistic and financial Albatrosses Diomedeu melanophrys and D. chry- support. This study is a part of the programs on the sostoma. Ibis 135:121-131. ecology of birds and mammals of French southern ter- RICKLEFS, R. E. 1974. Energetics of reproduction in 566 SHORT COMMUNICATIONS

birds, p. 152-297. In R. A. Paynter [ed.], Avian WARHAM,J. 1996. The behaviour,population biology energetics.Publ. Nuttall Ornithol. Club 1.5. and physiology of the petrels. Academic Press, RICKLEFS, R. E. 1983. Some considerationson the re- London. productive energetics of pelagic seabirds. Stud. WEIMERSKIRCH,H. 1990a. The influence of age and Avian Biol. 8:84-94. experience on breeding performance of the Ant- STRESEMAN,E., AND V. STRESEMAN.1966. Die Mauser arctic Fulmar Fulmarus glacialoides. J. Anim. der Vogel. J. Ornithol. 107:l-445. Ecol. 59:867-875. SYSTAT 1992. SYSTAT for windows: statistics.5th WEIMERSKIRCH,H. 1990b. Weight loss of Antarctic ed. SYSTAT, Inc., Evanston, IL. Fulmars Fulmarus glacialoides during incubation WARHAM,J. 1962. The biology of the Giant Petrel and chick brooding. Ibis 131:68-77. Macronectes giganteus. Auk 79: 139-160. WEIMERSKIRCH,H. 1991. Sex-specific differences in WARHAM, J. 1990. The petrels: their ecology and molt strategy in relation to breeding in the Wan- breeding systems.Academic Press, London. dering Albatross. Condor 93:731-737.

The Condor 100566-568 0 The CooperOmthological Society 1998

REJECTION OF ARTIFICIAL PARASITE EGGS BY GRAY KINGBIRDS IN THE BAHAMAS’

MICHAELE. BALTZ Division of Biological Sciences, University of Missouri, Columbia, MO 6521 I, e-mail: c623359@showme,missouri,edu DIRK E. BURHANS North Central Research Station, 1-26 Agriculture Building, University of Missouri, Columbia, MO 65211, e-mail: [email protected]

Abstract. We added artificial Shiny Cowbird (Mol- over 90 years, enough time to have evolved egg ejec- othrus bonariensis) eggs to Gray Kingbird (Tyrannus tion behavior (Rothstein 1975b). However, Shiny dominicensis) nests in the Bahamas, where Shiny Cowbirds have only recently arrived in the Bahamas Cowbirds were first recorded in 1993. Gray Kingbirds (Baltz 1995). BecauseShiny Cowbirds have not been ejected 85% of artificial eggs within 48 hr of addition. in prolonged contact with potential host speciesin the Based upon the short time of contact between the two Bahamas,egg rejection behavior should not have had species, we suggestthat egg ejection by Gray King- time to evolve in responseto interspecific parasitism. birds in the Bahamasis retention of ejection behavior We predicted that the Gray Kingbird (Tyrannus dom- from ancestralpopulations. inicensis), a known rejector of parasiteeggs in Puerto Rico and St. Lucia (Cruz et al. 1985, Post et al. 1990), Key words: brood parasitism, egg ejection, Gray would not reject Shiny Cowbird eggs in the Bahamas. Kingbird, Molothrus bonariensis,Shiny Cowbird, Ty- We tested this prediction by documentingthe response rannus dominicensis. of Gray Kingbirds in the Bahamasto artificial cowbird eggs added to their nests. The Shiny Cowbird (Molothrus bonariensis) is a gen- eralist brood parasitethat has invaded the West Indies METHODS from South America during this century, recently ar- The experiment was conductedfrom 3-14 July 1996, riving in North America (Post and Wiley 1977b, Post in the vicinity of StaniardCreek, North Andros Island, et al. 1993, Baltz 1995). The Shiny Cowbird parasit- Bahamas. We generally followed Rothstein’s (1975a) izes several host species in the West Indies region method of experimentally parasitizing nests. Artificial (Cruz et al. 1995), and some of these have been doc- eggs were shaped by hand from Sculpey Modeling umented rejecting experimentally added cowbird eggs Clay (Polyform Products Inc., Elk Grove Village, Il- in Puerto I&co and St. Lucia (Cruz et al. 1985, Post et linois), baked to harden, and painted with waterproof al. 1990). In many areas in the West Indies, Shiny acrylic and enamel paints to resemblereal Shiny Cow- Cowbirds have been in contact with host species for bird eggs: white ground color with “burnt umber” maculation. Real Shiny Cowbird eggs average 20.3 X 16.7 mm (Post and Wiley 1977a) and the artificial eggs I Received 1.5 December 1997. Accepted 16 April averaged 20.8 X 16.4 mm (n = 24). Single artificial 1998. Shiny Cowbird eggs were added to 20 Gray Kingbird