Locale-Specific Sugar Packet Opening by Lesser Antillean Bullfinches In

J. Field Ornithol. 73(1):82–85, 2002

Locale-speciﬁc sugar packet opening by Lesser Antillean Bullﬁnches in Barbados

Simon M. Reader,1,3,4 Daniel Nover,2 and Louis Lefebvre2

1 Bellairs Institute of McGill University, Barbados, West Indies 2 McGill University, Montre´al, Canada Received 8 January 2001; accepted 11 April 2001

ABSTRACT. Lesser Antillean Bullfinches Loxigilla noctis were observed opening packets of sugar on the grounds of a hotel on the west coast of Barbados. We presented closed sugar packets at 40 sites in 10 areas on the west coast, near the site where opening was originally observed, to determine whether the foraging behavior was locale- specific. Bullfinches only touched, opened, and fed from sugar packets at the original opening site, suggesting that the behavior pattern was specific to that location.

SINOPSIS. Individuos de Loxigilla noctis abren paquetes de azućar en una a´rea localizada de Bar- bados Se observaron individuos de Loxigilla noctis abriendo paquetes de azućar en un hotel de la costa oeste de Barbados. Para determinar si este patroń de forrajeo era especı´fico de esta zona, colocamos paquetes de azućar en 40 sitios repartidos entre 10 localidades de la costa cercanas al lugar donde observamos la conducta. Las aves de la zona del hotel fueron las uńicas que abrieron los paquetes y se alimentaron del azućar, lo que sugiere que el patroń de conducta es especı´fico de la zona. Key words: behavioral flexibility, foraging techniques, innovation, opportunism, social learning

Lesser Antillean Bullfinches are omnivorous, (grid reference 5547 5922), in the parish of St. feeding on fruits, seeds, insects, and routinely James, Barbados. A bullfinch fed from a small visiting scraps of food left by humans (Pinchon hole in the flat side of a single-serving packet 1964; Evans 1990; Dolman et al. 1996). Bull- of white sugar (5.5 ϫ 3.7 cm), removing the finches will readily feed on dissolved sugar and packet into a nearby bush after feeding for ca. fruit juice (Pinchon 1964; Webster and Lefebv- 60 s. It was not possible to determine the sex re 2000), including leftover cocktails and syrup of the individual involved, since male bullfinch- bottles, and have also been observed feeding es on Barbados have a similar plumage to fe- from open sugar bowls (S.M. Reader, pers. males, in contrast to the rest of the eastern Ca- obs.). In 2000, bullfinches on Barbados were ribbean (Bond 1985; Evans 1990). To deter- observed opening an unusual food source, mine whether bullfinches were responsible for sealed paper packets of sugar. We describe here packet opening, a single packet (obtained from our observations of the behavior pattern, and the hotel) was placed approximately 1 m from experiments to determine whether the behavior two observers. A bullfinch approached the pattern is locale-specific. Field observations of packet within 30 s, and began pecking at the opening behavior in birds are of particular in- flat packet surface. After ca. 20 s, the finch had terest because they add ecological validity to the made a hole and began feeding on the sugar, many captive tests that utilize opening of an occasionally widening the hole by pushing in apparatus. the beak. The packet was removed for exami- The feeding behavior was first observed in nation, and a second packet placed 0.3 m from the mid afternoon (14:45–15:45) on 16 May the observers. The packet was moved by a bull- 2000 in the grounds of the Colony Club hotel finch to a distance of 1 m after a period of 5– 10 min, where the finch repeatedly turned the 3 Current address: Department of Biology, McGill packet over, but did not make pecking attempts University, 1205 avenue Docteur Penfield, Montreál, and eventually abandoned the packet. A second Que´bec, H3A 1B1, Canada. bullfinch, distinguishable from the previously 4 Corresponding author. Email: Ͻsimon.reader@ observed birds by virtue of its smaller size, ap- mcgill.caϾ proached and opened the packet by pecking at

82 Vol. 73, No. 1 Bullfinch Locale-specific Foraging Behavior 83 the flat surface. A larger bullfinch subsequently morning, the grounds of the Colony Club ho- fed from this hole. Examination of the opened tel, the north of the Coral Reef Club hotel (grid packets revealed a hole in each of ca. 5 ϫ 2 reference 5551 5913), the south of the Coral mm, of similar appearance to holes made by Reef Club (5547 5910), the west of Bellairs bullfinches in ripe fruit such as banana. Research Institute (5555 5905), and to the In total, four observations were made of bull- north of St. James’ Church (5570 5888). In the finches feeding from sugar packets on 16 May, afternoon of the following day, we tested at the with at least two different individuals observed following sites: the south of St. James’ Church opening the packets. One observation was (5571 5881), the east of Bellairs (5567 5906), made of an individual feeding from a packet the east of the Coral Reef Club (5555 5911), opened by another. To our knowledge, this is the Colony Club, and the west edge of the the first recorded observation of this behavior Balmford estate to the north of the Colony pattern. Lefebvre et al. (1997, 1998) collected Club (5542 5932). Grid references are the stan- over 1800 reports of foraging innovation in dard metric system for plotting precise locations birds from the published literature, and the in Barbados and many other nations (including only remotely similar cases we found in this Canada, Great Britain, and Australia). Grid ref- database were Passeriformes and Piciformes ex- erences are typically split into three parts: the ploiting hummingbird nectar feeders (Taylor 100 km square in which they lie, the easting (a 1972; Fisk 1973; Leck 1974; Fisher 1975; Mar- measure of longitude, read from the horizontal tin 1977; Stokes and Stokes 1984, 1985; Beach map edge), and the northing (a measure of lat- 1988). On 24 May 2000 further observations itude). Barbados is covered by one 100 km map of packet opening were made at the Colony square, so the first part of the grid reference is Club hotel. Initially, brown colored packets of not required. The furthest study area was 500 raw sugar were placed at the site, but after 35 m from the Colony Club sites. min of observation only one finch had ap- At each of these 10 areas, two observers with proached the packets and made a failed attempt synchronized stopwatches conducted tests at to open them. We then placed the white pack- four sites simultaneously, so that if opening was ets of white sugar used by the hotel on the observed at more than one site we could deter- ground. Bullfinches were observed picking up, mine how many birds were involved. Conduct- opening, and carrying the packets away. Bana- ing the four tests consecutively in the same area naquits (Coereba flaveola) picked up the packets would risk attracting the same bird to each site and fed on sugar which had fallen out of an in turn, producing a false picture of the open- opened packet, and a Zenaida Dove (Zenaida ing activity in each area. It was unlikely that aurita) pecked at the same packet that a finch the same finches were observed in different had recently contacted, as well as feeding on study areas, since finches show very high site sugar it knocked out of an already opened fidelity (Webster and Lefebvre 2001), and on packet. each day the study areas were at least 50 m To determine whether the sugar packet open- apart and visually isolated from one another. All ing behavior was locale-specific or part of the the test sites were open to the public, but were usual bullfinch behavioral repertoire, we con- chosen to be as free from disturbance and hu- ducted a simple field experiment. We placed man traffic as possible. packets of sugar at a total of 40 sites, in 10 In the experiments, finches only contacted different areas in St. James, Barbados, and and opened sugar packets at the Colony Club watched for any bird activity. We obtained sug- sites. In the morning session, rapid opening was ar packets of a similar design and color to those observed at one site at the Colony Club. A used at the Colony Club hotel from Barbados finch picked up a packet after 148 s, carried it Hotel Foods, St. Michael. We placed five pack- ca. 3 m, and opened and fed from the packet ets in a row ca. 35 cm long at a variety of sites 156 s after the beginning of the trial. After 295 where finches had been recently observed. We s a second finch chased away the opener and observed each site for 20 min, from at least 5 fed from the hole, and several other displace- m. We conducted half the observations in the ment incidents were observed. At another Col- morning and half in the afternoon to control ony Club site a finch touched, but did not for time of day. The study areas were, in the open, a packet. In the afternoon testing ses- J. Field Ornithol. 84 S. M. Reader et al. Winter 2002 sions, finches were observed contacting the bullfinches may be socially learned by individ- packets at one site at the Colony Club study uals taking advantage of the innovation of other area, but no opening events were observed birds. In Blue Tits (Parus caeruleus) a novel for- within 20 min (though a finch did open a pack- aging behavior, nectar robbing from the flowers et 56 s after the end of one trial). At all other of the Crown Imperial, was restricted to two study areas, in both the morning and the after- areas, even though flowers were available in noon, finches were observed near the sugar nearby areas (Thompson et al. 1996). Thomp- packets, passing within 5 cm, 15 cm and 75 son et al. (1996) argued that the fact that flow- cm at three sites, for example, but on no oc- ering coincides with nesting, when Blue Tits casion did a finch approach, touch, or open the defend territories, might restrict the spread of packets. Despite the small number of opening nectar robbing to within a mating pair. Terri- events observed, our finding that birds only toriality may also restrict the spread of the novel touched and opened packets at the Colony opening behavior in bullfinches. In our study, Club site suggests that the behavior is highly other bird species fed from the packets opened locale-specific. A survey of hotels and restau- by bullfinches, which might suggest that social rants in the local Holetown area determined learning could facilitate spread of the use of a that the Colony Club was the only open-air novel food source between species. establishment serving sugar in packets. Two ACKNOWLEDGMENTS similar field experiments presented opened dishes of dissolved sugar (Webster and Lefebvre We thank the staff of the Colony Club hotel for per- mission to work on their grounds. This work was fund- 2000) and closed boxes of seed (Webster and ed by a Commander C. Bellairs postdoctoral fellowship Lefebvre 2001) in two areas (Bellairs and St. to SMR and an NSERC grant to LL. James’ Church) where we did not observe packet opening. Bullfinches readily approached both LITERATURE CITED tasks, and opened the box at three of the 10 sites tested, suggesting that between-site differ- BEACH, P. E. 1988. Northern Oriole at hummingbird ences in the willingness to approach novel ob- feeder. Delmarva Ornithologist 21: 2. jects or to avoid humans are unlikely explana- BOND, J. 1985. Birds of the West Indies, 5th ed. Collins, London. tions for our results. DOLMAN, C. S., J. TEMPLETON, AND L. LEFEBVRE. 1996. The fact that birds did not open, or even Mode of foraging competition is related to tutor contact, sugar packets in other sites suggests preference in Zenaida aurita. Journal of Compara- that birds at the Colony Club site have learned tive Psychology 110: 45–54. EVANS, P. G. H. 1990. Birds of the eastern Caribbean. both to recognize the packets as a source of Macmillan, London. food and to open them. This localized behav- FISHER, G. F. 1975. Additional records of unusual nectar ioral variation may be the result of social learn- feeding. Wilson Bulletin 87: 421. ing from conspecifics, with birds perhaps learn- FISHER, J., AND R. A. HINDE. 1949. The opening of ing as a result of feeding from previously milk bottles by birds. British Birds 42: 347–357. FISK, L. H. 1973. Notes on nectar feeding by orioles. opened packets or by observing other bullfinch- Auk 90: 208–209. es opening packets. Alternatively, individuals GALEF, B. G., JR. 1992. The question of animal culture. may have learned through a process of ‘‘natural Human Nature 3: 157–178. shaping’’ (Galef 1992) at the Colony Club site, LECK, C. 1974. Further observations of nectar feeding perhaps feeding from abandoned human- by orioles. Auk 91: 162. LEFEBVRE, L. 1995. The opening of milk-bottles by opened sugar packets before learning how to birds: evidence for accelerating learning rates, but open them for themselves. In Black-capped against the wave-of-advance model of cultural Chickadees (Poecile atricapilla), the rate of transmission. Behavioural Processes 34: 43–53. learning to open tubs of food or milk-bottles ———, A. GAXIOLA,S.DAWSON,S.TIMMERMANS,L. ROSZA, AND P. K ABAI. 1998. Feeding innovations was increased in birds which encountered open and forebrain size in Australasian birds. Behaviour food sources (Sherry and Galef 1984, 1990), 135: 1077–1097. and this is thought to be one of the processes ———, P. WHITTLE,E.LASCARIS, AND A. FINKELSTEIN. responsible for the spread across mainland Brit- 1997. Feeding innovations and forebrain size in ain of milk-bottle opening in tits (Fisher and birds. Animal Behaviour 53: 549–560. MARTIN, D. J. 1977. Nectar feeding in the Black-capped Hinde 1949; Lefebvre 1995). This raises the Chickadee. Murrelet 58: 18. possibility that the opening of sugar packets by PINCHON, R. 1964. Faune des Antilles Françaises: les Vol. 73, No. 1 Bullfinch Locale-specific Foraging Behavior 85

oiseaux. Muse´um d’Histoire Naturelle, Fort-de- TAYLOR, P. M. 1972. Hovering behavior by House France, Martinique. Finches. Condor 74: 219–221. SHERRY,D.F.,AND B. G. GALEF,JR. 1984. Cultural THOMPSON, C. F., G. F. RAY, AND R. L. PRESTON. 1996. transmission without imitation: milk bottle open- Nectar robbing in Blue Tits Parus caeruleus: failure ing by birds. Animal Behaviour 32: 937–938. of a novel feeding trait to spread. Ibis 138: 552– ———, and———. 1990. Social learning without im- 553. itation: more about milk bottle opening by birds. WEBSTER,S.J.,AND L. LEFEBVRE. 2000. Neophobia by Animal Behaviour 40: 987–989. the Lesser-Antillean Bullﬁnch, a foraging generalist, STOKES, D., AND L. STOKES. 1984. The behavior watcher’s notebook: separating sexes. Birdwatcher’s Digest and the Bananaquit, a nectar specialist. Wilson Bul- 6: 75–79. letin 112: 424–427. ———, and L. STOKES. 1985. The behavior watcher’s ———, and ———. 2001. Problem solving and neo- notebook: bird song (part I). Birdwatcher’s Digest phobia in a Columbiforme–Passeriforme assem- 7: 84–89. blage in Barbados. Animal Behaviour 62: 23–32.