SHORT COMMUNICATIONS 187

T/w Condor 102:1 X7-1 9” 0 The CooperOmthological Soc~ty 2000

TAXONOMIC AFFINITIES OF AUDUBONS’ SHEARWATER FROM EUROPA ISLAND ’

MAITHIEU LE CORRE~ CEBC-CNRS Villiers en Bois. 79360 Beauvoir/Niort.

Abstract. The western holds five Island, a remote islet of the southern subspecies of the pantropical Audubon’s Shearwater Channel (Le Corre and Jouventin 1997) although no (PufJinus Iherminieri), but none was known to breed breeding colonies were known in the Mozambique in the . Here, I describe a newly Channel, or along the eastern of discovered population of Audubon’s Shearwaters on (Feare 1984, Langrand 1995). Europa Island, southern Mozambique Channel. Com- Recent studies on geographic variation of parison of geographic variation of morphometric char- of the western Indian Ocean showed that two pelagic acters showed that the birds of this population belong species from Europa Island were distinct from all other to the P. 1. bailloni subspecies, previously thought to populations of the Indian Ocean (Le Corre and Jou- be endemic to the Mascarenes Islands. They were dis- ventin 1999, Le Corre, in press). The oceanic environ- tinct from the three subspecies of the Comoro, Alda- ment of the Mozambique Channel, particularly the bra, and group, suggesting that few suc- high sea-surface temperature (Piton 1989) may consti- cessful exchanges of mdividuais occu; between north tute an ecological barrier for pelagic seabirds. If such of the Mozambiaue Channel and Eurooa Island. This an ecological barrier exists in the Mozambique Chan- biogeographic pattern is similar to that of two phylo- nel, it should have strong effects on highly sedentary genetically independent pelagic species from Europa species like Audubon’s Shearwater. Island. This suggests that a common cause related to The aim of my study was to determine the taxonom- geographic isolation and oceanic conditions in the ic affinities and biogeographic relationships of this Mozambique Channel may explain the apparent iso- newly discovered population of Audubon’s Shearwa- lation of these three populations from Europa ters. Biometrics and color patterns of the birds from Island. Europa Island are described and compared with those of the other subspecies of the Indian Ocean. Based on Key words: Audubons’ Shearwater, endemism, Eu- these comparisons, I discuss the hypothesis of biogeo- ropa Island, geographic variation, Indian Ocean, Puf- graphic isolation of seabirds of the southern Mozam- finus lherminieri. bique Channel. I also describe the breeding habitat and breeding season, estimate population size, and identify Although pelagic seabirds forage over large oceanic possible threats induced by human-related factors to surfaces, many species show high natal philopatry and this population. breeding-site fidelity (Weimerskirch et al. 1985), which may lead to geographic isolation between col- METHODS onies. The oceanic environment also influences move- Europa (22”2O’S, 40”21 ‘E) is a 28 km2 coralline island ments of pelagic seabirds at sea (Ashmole 1971), of the southern Mozambique Channel. Its remoteness which may have important consequences on exchanges and absence of fresh water preserved it from extensive between colonies. human exploitation, and the island still holds large sea- Petrels and shearwaters have evolved numerous en- bird breeding colonies (Le Corre and Jouventin 1997). demic forms, particularly in tropical waters (Warham During my field work, observations and censuses of 1990, 1996). In the tropical Indian Ocean, six species shearwaters were made at dusk from vantage points on of petrels and shearwaters are known to breed; two are the coastline, when the birds were returning to the is- endemic species from Reunion Island (Barau’s Petrel land. The colonies were located by crepuscular and Pterodroma baraui and Mascarene Black Petrel Pseu- nocturnal observations and listening for displaying dobulweria aterrima). A third species, the Audubon’s adults, together with diurnal searching for burrows. Shearwater (Puf/inus Iherminieri), is pantropical but The birds examined were caught by hand at the en- has evolved five distinct subspecies in the area (Jouan- trance of their burrows or mist-netted when flying in and Mougin 1979, Louette and Herremans 1985, above their colonies. Some of them (4 out of 23) were Shirihai and Christie 1996). In 1994, a small popula- lured with playback vocalizations of P. 1. bailloni. All tion of Audubon’s Shearwaters was found at Europa captured birds were measured, photographed, ringed, and released. Their coloration pattern was recorded (plumage and bare parts). Measurements taken (in ’ Received 22 September 1998. Accepted 12 July mm) were flattened wing length, tarsus length, culmen 1999. length, bill depth at gonys, maximal unguis, and Z Current address: SEOR, Museum d’Histoire Na- weight (g). I also checked for incubation patch. I was turelle. Rue Poivre 97400 Saint Denis. Reunion Island. not able to sex the birds in the field. Bretagnolle and e-mail: lecorre@univ-reunionfr Attit (1996) found no sex differences in biometrics of 188 SHORT COMMUNICATIONS

TABLE 1. Variation of plumage pattern and coloration of bare parts of Puffinus Zherminieri subspecies of the western Indian Ocean and of P. atrodorsalis.

Under tail Under wing Locality (Taxa) Upper parts coverts coverts Tarsus

Europa Island (bailloni ?) brownish black mostly white mostly white black and blue Reunion Island (bailloni) brownish black mostly white mostly white black and blue ? (P. atrodorsalis) black mostly white mostly white black and blue Island (colstoni) brownish black brown mostly white pink Seychelles (nicolae) brownish black brown mostly white pink Comoro (temptator) brownish black brown brown black and blue Arabian Sea (persicus) brownish black brown brown ?

adult P. 1. bailloni at Reunion Island. The birds from and in September and October 1995 (I did not visit the Europa were compared to birds from the other subspe- island in July and August). I did not observe them in cies of the Indian Ocean: P. 1. bailloni from Reunion December and January despite intensive field work in Island (unpubl. data), P. 1. nicolue from the Seychelles the appropriate places. Therefore, this species probably (Jouanin 1970), P. 1. colstoni from Aldabra (Shirihai breeds during the winter at Europa Island, like two and Christie 1996), P. 1. persicus from the Arabian other pelagic species from this island (Sooty Terns Sea, and P. 1. temptutor from the Comoro Islands Sterna jiiscata, and Red-footed Boobies Sula sula) as- (Louette and Herremans 1985), and the newly de- sociated with surface-feeding tunas (unpubl. data). The scribed P. atrodorsalis (Shirihai et al. 1995). Measure- places where all displaying birds were observed and ments of adult P. 1. bailloni from Reunion Island were heard were in an area of about 1 km*, covered with made on skins and live specimens, but there were no shrubs (Pemphis acidula and Psiadia altissima). The significant differences in measurements between the ground was composed of coralline limestone rocks, two groups, therefore these data were pooled. Statistics regularly interrupted with sandy holes. Several bur- were computed with LOGITHEQ and STAT-ITCE rows were discovered in some of these holes with birds Biometrics are expressed as mean 2 SD. entering just after landing, but I was able to check the content of only one of these burrows, in September RESULTS 1995. Although I did not find any egg or chick, the COLORATION PATTERN AND BIOMETRICS presence of feces, down, feathers, and typical odor of Twenty-three birds from Europa Island were caught shearwater strongly suggested that the birds had at- from April 1995 to May 1997. All birds had similar tempted to breed. plumage pattern and bare part coloration (Table 1). Out I never observed more than 25 birds per evening of 19 birds for which the incubation patch was ex- and per breeding place (maximum 23 different birds amined, 7 had partially (5) or completely (2) bare in- counted on 20 April 1995). As there are only 2 or 3 cubation patch, suggesting that they were adults, breeding places on the island, population size must be breeding or about to breed. Measurements did not dif- very small, probably less than 100 individuals. fer according to whether birds had an incubation patch or not but the birds which had an incubation patch DISCUSSION were heavier (272 t 12 g and 234 5 11 g, respec- TAXONOMIC AFFINITIES tively, Mann-Whitney U-test, U = 0, P < 0.001). As there was no difference in biometrics between birds The systematics of the “little shearwater” group, in- with and without an incubation patch, I considered that cluding P. assimilis, P. lherminieri and P. atrodorsalis the birds captured at Europa were all adults or sub- are very complex and not fully agreed upon (Jouanin adults, some of them were probably breeding or about and Mougin 1979, Shirihai et al. 1995, Bretagnolle and to breed (heaviest birds with bare incubation patch), AttiC 1996). However, most subspecies of the two gen- and others were merely prospecting for a nesting site erally accepted species (P. assimilis and P. Zhermini- (lightest birds with no incubation patch). eri) can be separated both by size and plumage pattern Birds from Europa had a different plumage pattern (Jouanin and Mougin 1979). The birds from Europa than those from Aldabra, the Seychelles, the , do not belong to any subspecies of P. assimilis because or the Persian Gulf (Table 1). They were similar to of their larger size (all the P. assimilis subspecies are adults from Reunion Island in plumage pattern (Table smaller, see data in appendix in Shirihai et al. 1995) l), but tended to have a larger bill depth and a longer plumage pattern, and ecology (P. assimilis breeds in wing length (Table 2). They also had a longer wing, subtropical or temperate rather than in tropical waters). larger bill depth, and larger unguis than juveniles from They also differ from all the Atlantic and Pacific sub- Reunion Island (Table 2). species of P. lherminieri by their white undertail co- verts (brown in the other subspecies) and black and BREEDING SEASON, HABITAT, AND POPULATION blue legs (pink in the other subspecies). In the western SIZE Indian Ocean, comparisons of measurements and The birds were observed displaying above the breed- plumage pattern show that they are distinct from all ing places from March to June 1995, 1996, and 1997, known subspecies except P. 1. bailloni from which SHORT COMMUNICATIONS 189 they differ only by slight differences in bill depth and wing length (Tables 1 and 2). I will not discuss the affinities between the birds of Europa with P. atrodorsnlis because the validity of this species is not fully agreed upon (Bretagnolle and AttiC 1996, Shirihai et al., unpubl. data). I suggest that the birds from Europa constitute an isolated population of P. 1. bailloni, which was previously thought to be en- demic to the (Jouanin 1987, Barre et al. 1996). The slight differences in measurements may indicate that there is little successful exchange between the two populations but that this apparent iso- lation is recent. No study attempted to identify the dis- persal of fledglings and nonbreeding adults in this spe- cies, so that we have no data of the level of isolation between adjacent colonies. However, the present num- ber of known subspecies worldwide (13) and espe- cially in the western Indian Ocean (5) suggests that the different populations are highly isolated from each oth- er. Further studies using both molecular techniques and ringing data are necessary to address this question.

BIOGEOGRAPHIC IMPLICATIONS It has been argued elsewhere that the geographic iso- lation of Europa Island and high sea-surface tempera- tures in the center of the Mozambique Channel may have contributed to the isolation of two seabird spe- cies, the White-tailed Tropicbird Phaethon lepturus and the Red-footed Booby Sula sulu (Le Corre and Jouventin 1999, Le Corre, in press). Warm waters are known to be poor feeding areas for pelagic seabirds (Ashmole 197 1). Thus, the central Mozambique Channel waters, which are on average 1°C warmer than those of the Indian Ocean at same latitudes (Piton 1989), may constitute an ecological barrier for seabirds. The data on Audubon’s Shear- water suggest that they are different from those breed- ing to the north of Madagascar (P. 1. temptator, P. 1. nicolae, P. 1. persicus, and P. 1. colstoni). This is con- sistent with the hypothesis that there are few successful exchanges between seabird colonies north of the Moz- ambique Channel and those in the south (Europa Is- land). Interestingly, White-tailed Tropicbirds, Red- footed Boobies, and Audubon’s Shearwaters have the same pattern of geographic variation in the western Indian Ocean, although they are phylogenetically in- dependent, suggesting a common response to environ- mental constraints.

POPULATION SIZE AND THREATS The number of birds observed every night when re- turning to the island suggests that the population size may be less than 100 birds. Such a small population is very vulnerable and at risk of local extinction. Black rats (Rattus ruttus) were introduced to Europa at least 100 years ago (Le Corre and Jouventin 1997) and they are now widespread all over the island, where they prey upon all ground nesting seabird species (pers. ob- serv.) and may be responsible for the small population size of Audubon’s Shearwaters at Europa Island. Barn Owls (Tyto alba) are also abundant at Europa Island (Le Corre and Jouventin 1997). I have no evidence of predation of Audubon’s Shearwater by Barn Owls at Europa, but, in the Seychelles, they prey upon this species (Diamond and Feare, in Lever 1987). 190 SHORT COMMUNICATIONS

The existence of this small population of P. 1. bail- fins Puj$nus Iherminieri de 1’OcCan Indien occi- loni, suggest that this taxon has a larger range than dental. Oiseau et RF0 40:303-306. previously thought. Knowing the size of the breeding JOUANIN, C. 1987. Notes on the nesting of Procellari- population of Audubon’s Shearwater at Europa and the formes in RCunion, p. 359-363. In A. W. Dia- potential threats, a conservation program should be un- mond [ED.], Studies of Mascarene Island birds. dertaken to preserve this small and isolated population. Cambridge Univ. Press, Cambridge. JOUANIN, C., AND MOUGIN, J. L. 1979. Order Procel- This work was supported by the Direction Rtgion- lariformes, p. 48-121. In E. Mayr and G. W. Cot- ale de 1’Environnement (DIREN-Minis&e de trell IEDS.~. Check-list of birds of the world. Vol. 1’Environnement) of R&union Island and by the Of- 1. M&. C%mp. Zool., Cambridge. fice National de la Chasse (ONC), to whom I am LANGRAND, 0. 1995. Guide des Oiseaux de Madagas- particularly grateful. The stays on the island were au- car. Delachaux et NiestlC, Lausanne. thorized by the Prefect of Reunion Island, and orga- LE CORRE, M. (In press). Plumage polymorphism of nized with the help of the Direction RCgionale de Red-footed Boobies (S&a sula) in the western In- dian Ocean: an indicator of biogeographic isola- M&Co France. Transports to Europa were allowed tion. J. Zool. 249. with the collaboration of the Force Armte de la Zone LE CORRE, M., AND l? JOUVENTIN. 1997. Ecological sig- Sud de I’OcCan Indien (FAZSOI). I am indebted to nificance and conservation priorities on Europa Is- Sonia Ribes, curator of the Natural History Museum land (western Indian Ocean), with special refer- of Reunion Island, and Pierre Jouventin, director of ence to seabirds. Terre et Vie 52:205-220. the Centre d’Etudes Biologiques de ChizC (CNRS), LE CORRE, M., AND P JOUVENTIN. 1999. Geographic who gave me the opportunity to conduct research on variation in the White-tailed Tropicbird (Phaethon Europa. I also thank J. Bried, E Lagarde, H. Lormte, lepturus) with the description of a new subspecies and j. M. Probst for field assistance, C. Jouanin, J. endemic to Europa Island (western Indian Ocean). Bried. W. D. Koenig. and two anonymous reviewers Ibis 141:233-239. for their useful comments on an early draft, and V. LEVER, C. 1987. Naturalized birds of the world. Long- Bretagnolle for providing recorded vocalizations of mann Scientific and Technical, Harlow, England. P. 1. bailloni from Reunion Island. LOUETTE, M., AND M. HERREMANS. 1985. A new race of Audubon’s Shearwater Pujfinus lherminieri at LITERATURE CITED Moheli, Comoro Islands. Bull. Brit. Ornithol. Club 105:42-49. ASHMOLE, N. P 1971. Sea bird ecology and the marine MOORS, l? J., AND I. A. E. ATKINSON. 1984. Predation environment, p. 223-286. In D. S. Farner, J. R. on seabirds by introduced animals and factors af- King, and K. C. Parkes [ED%], Avian biology.-. Ac- fecting its severity, p. 667-690. In l? J. Moors ademic Press, New York. [ED.], Status and conservation of the world’s sea- ATKINSON. I. A. E. 1985. The soread of commensal birds. Int. Council Bird Preserv. Tech. Publ. 2. species of Rattus to oceanic’islands and their ef- PITON, J. 1989. Ouelaues aspects nouveaux sur la cir- fects on island avifauna. Int. Council Bird Pres- culation sup&ficielle dins le Canal du Mozam- erv., Tech. Publ. 3:35-81. bique (o&an Indien). Documentation Scientifique BARR& N., A. Barau, and C. Jouanin. 1996. Les Orstom 14:289-299. oiseaux de La Reunion. Edition du Pacifique, Par- SHIRIHAI, II., AND D. A. CHRISTIE. 1996. A new taxon is. of small shearwater from the Indian Ocean. Bull. BRETAGNOLLE,V., AND C. AT@. 1996. Coloration and Brit. Ornithol. Club 116: 180-I 86. biometrics of fledgling Audubon’s Shearwater SHIRIHAI, H., I. SINCLAIR, AND l? R. COLSTON. 1995. A Pufjinus lherminieri from RCunion Island, Indian new species of Pufinus shearwater from the west- Ocean. Bull. Brit. Ornithol. Club 116:194-197. ern Indian Ocean. Bull. Brit. Ornithol. Club 115: FEARE, C. J. 1984. Seabirds status and conservation in 75-87. the tropical Indian Ocean. Int. Council Bird Pres- WARHAM, J. 1990. The petrels: their ecology and erv. Tech. Publ. 21457-471. breeding systems. Academic Press, New York. IMBER, M. J. 1978. The effects of rats on breeding WARHAM, J. 1996. The behavior, population biology success of petrels, p. 67-72. In P R. Dingwall, I. and physiology of the petrels. Academic Press, A. E. Atkinson, and C. Hay [EDS.], The ecology New York. and control of rodents in New Zealand nature re- WEIMERSKIRCH, H., J. L. MOUGIN, J. C. STAHL, AND J. serves. Dept. Land and Survey Information Se- C. VAN BEVEREN. 1985. Banding recovery and ries, Auckland, New Zealand. dispersion of seabirds breeding in the French Aus- JOUANIN, C. 1970. Note taxonomique sur les Petits Puf- tral and Antarctic territories. Emu 85:22-23.