89? 729743 Original Article COCOS ISLAND BARK BEETLES (SCOLYTINAE) L. R. KIRKENDALL and B. H. JORDAL Biological Journal of the Linnean Society, 2006, 89, 729–743. With 5 figures The bark and ambrosia beetles (Curculionidae, Scolytinae) of Cocos Island, Costa Rica and the role of mating systems in island zoogeography LAWRENCE R. KIRKENDALL1* and BJARTE H. JORDAL2 1 Department of Biology, University of Bergen, Allegaten 41, N-5007 Bergen, Norway Museum of Natural History and Archaeology, University of Science and Technology, N-7491 Trondheim, Norway 2 Received 11 November 2004; accepted for publication 15 January 2006 Cocos Island is a small oceanic island midway between Costa Rica and the Galápagos Archipelago; about 2 Myr in age, it is the only tropical oceanic island in the eastern Pacific with tropical wet forest. We identified several hundred bark beetle specimens collected during recent expeditions by INBio, the National Biodiversity Institute of Costa Rica, and re-examined all specimens from earlier collections. We report 19 species in ten genera, seven or eight of which are endemic, making scolytines the largest group of beetles known from the island. We describe as new Pycnarthrum pseudoinsulare, Xyleborinus cocoensis, and Xyleborus sparsegranulosus, resurrect Xyleborus bispinatus as separate from X. ferrugineus, and report six other species as new to Cocos Island. Three-quarters of the scolytines reproduce by brother–sister mating, and we argue that inbreeders are superior island colonists because they are less affected than are outbreeders by problems of mate location and inbreeding depression. The fauna and flora of Cocos Island arrived by dispersal and human transport. We examine natural colonization patterns for the fauna, using the distributions of the relatives of island endemics: most colonization came from the Americas, but the closest relatives to some endemics are found on Caribbean or Galápagos islands. © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743. ADDITIONAL KEYWORDS: Baker’s rule – colonization – inbreeding – island biogeography. INTRODUCTION Since the voyage of the Beagle, investigations of island biotas have led to key insights into major evolutionary processes behind the origin of organismal diversity. Research on the Hawaiian, Canary, and Galápagos Islands in particular has provided a wealth of information on modes of speciation, relationships between ecology and speciation, and the time scales over which speciation processes occur (Juan et al., 2000; Gillespie & Roderick, 2002). Islands that are isolated over extended geological time scales ultimately foster divergent phenotypes or ecotypes, often sufficiently different from their continental relatives to be named *Corresponding author. E-mail: Lawrence.Kirkendall@bio.uib.no as evolutionarily separate entities. However, detailed analyses of island diversity and endemism are critically dependent upon thorough sampling and sound taxonomic analyses of focal taxa. Small islands are therefore excellent model systems, because these can be sampled more thoroughly for hyperdiverse taxa such as arthropods, resulting in more complete taxonomic lists. Cocos Island (Isla del Coco) is a small island (46.6 km2) belonging to Costa Rica, located midway between that country and the Galápagos Archipelago, about 500 km west of Costa Rica and 630 km northeast of the Galápagos. The island was formed by recent (2 Mya) volcanic activity; its highest peak, Cerro Iglesias, is 575 m (Castillo et al., 1988). Because of its small size, young age and isolation, this tropical island is particularly suited to the study of colonization and establishment of populations; among evolu- © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 729 730 L. R. KIRKENDALL and B. H. JORDAL tionary biologists, it is perhaps best known for studies of ecological release or niche expansion, in which single species experiencing little or no interspecific competition evolve an extraordinary degree of ecological and behavioural plasticity [as in a theridiosomatid spider (Eberhard, 1989) and the endemic ‘Darwin’s finch’ (Werner & Sherry, 1987)]. Of particular importance is the fact that Cocos Island, with its 7-m rainfall annually, is the only tropical oceanic island in the eastern Pacific that has tropical wet (as opposed to dry) forest, a habitat which is similar in form and structure (though not in species composition) to those of most of the closest potential mainland source areas in Central and northern South America (Hogue & Miller, 1981). As is typical for such a small isolated island, the flora and fauna are depauperate relative to their continental source communities, and taxonomically imbalanced (‘disharmonious’), with the island lacking many abundant mainland taxa (Fournier, 1966; Hogue & Miller, 1981; Brown, 1990). Among terrestrial vertebrates, for example, there are only four resident bird species, two reptiles, one freshwater fish, and no amphibians or native mammals (World Conservation Monitoring Centre, 1997). Fewer than 400 insect, arachnid, and myriapod species have been recorded from the island (Hogue & Miller, 1981; Bright, 1982; Cwikla & Freytag, 1982; Johnson, 1982; Adams, 1983; Nickle, 1983; Bellamy, 1986; Brown, 1990; Brown, Donahue & Miller, 1991; Woodley, 1991; Davis, 1994; Jordal, 1998a; Brown & Miller, 1999; Anderson & Lanteri, 2000). Wood-boring, leaf-mining, and epiphytedwelling species are among the more species-rich insect groups on the island, supporting a key role for rafting on plant material in the colonization process (Hogue & Miller, 1981). Comparative studies of island faunas have demonstrated a close relationship between mating behaviour and the bark beetle (Curculionidae, Scolytinae) communities of islands: species with regular brother–sister inbreeding are overrepresented on small islands compared with neighbouring mainland forests; tropical islands have a higher proportion of inbreeders than do temperate islands; outbreeding species are more likely than are inbreeders to speciate on small islands (Kirkendall, 1993; Jordal, Beaver & Kirkendall, 2001). Faunistic details for particular islands and archipelagos are needed in order to explore such patterns further and to understand more fully the historical and ecological conditions necessary for successful colonization (Bright & Peck, 1998). In this context, we here update the faunal composition of scolytine beetles for Cocos Island, roughly doubling the number of species known from the island. (Scolytines we refer to collectively as ‘bark beetles’, but this term is also used in a more restrictive, eco- logical sense to refer to those scolytines breeding in inner bark.) Four papers have dealt with the bark beetle fauna of Cocos Island (Schedl, 1974; Bright, 1982; Wood, 1982; Wood, 1986a), bringing the number of species previously known from the island to 10. We describe three species new to science, report another seven species as new to the island, and correct previous misidentifications, resulting in a compilation of 19 species of Scolytinae – making bark beetles the largest group of beetles, tied with ants as the largest family or subfamily of arthropods currently known from Cocos Island (Solomon & Mikheyev, 2005). We conclude with a brief discussion on general zoogeographical trends that can be deduced from patterns of endemism for the terrestrial fauna of Cocos Island. MATERIAL AND METHODS Cocos Island collections examined and type material are deposited in the National Institute for Biodiversity, Costa Rica (INBio) unless noted otherwise. Other collections examined included those from the Museum of Comparative Zoology, Harvard University (MCZ); Natural History Museum, London (NHM); California Academy of Science (CAS); Natural History Museum of Los Angeles County (LACM); Stephen L. Wood, Brigham Young University, property of the US National Museum (SLWC); private collection of Lawrence R. Kirkendall, University of Bergen, Norway (LRKC). Material of the X. ferrugineus complex was borrowed from the Zoological Museum of the University of Copenhagen (ZMUC), the Royal Museum for Central Africa (RMCA, in Belgium), Institute Royal des Sciences Naturelles de Belgique (IRSNB), and the Karl Schedl collection in the Natural History Museum of Vienna (NHMW). Species determinations of problematic taxa were made by direct comparison with type material wherever possible. Identification of scolytines often requires meticulous observation of surface sculpture details (e.g. Wood, 1982) which are difficult or impossible to see with normal light sources (such as fibre optic lights); we used one or two small flourescent lamps with 7-W bulbs as sources of flat light for this purpose. Terminology follows Wood (Wood, 1986b) and measurements were in accordance with methods described by Jordal (1998a). Note that all measurements were based on perpendicular views of the dorsal aspect of a specimen. Species are listed alphabetically by genus, with the order of genera following the classification by Wood (1982). Full synonymies and other taxonomic information for previously described species can be found in Wood & Bright (1992) and are not repeated here. Complete label data are given only for type material of new species; label data were translated © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 COCOS ISLAND BARK BEETLES (SCOLYTINAE) 731 from Spanish for the INBio collections, including most place names. Unless otherwise stated, information on overall species distribution and host plant usage were obtained from Wood & Bright (1992), and ecological data (such as type of breeding material used) from Wood (1982). Classification as inbreeding or outbreeding follows Kirkendall (1993; see also Hamilton, 1967). Thus, all Xyleborini, Coccotrypes and Hypothenemus, as well as the small Araptus complex discussed below, were classified as inbreeding by sibling mating because: (1) in all cases where males are known, they are morphologically adapted to extreme inbreeding, having reduced eyes and nonfunctional flight wings, (2) usually only one or a few males occur per brood, and (3) females reproduce alone (males are found with breeding females in most outbreeding species, Kirkendall, 1983). ANNOTATED LIST OF SPECIES (1) PHLOEOTRIBUS PACIFICUS BRIGHT New records: Wafer Bay (close to sea level), January 1994 (2), February 1994 (3), and March 1994 (10), Malaise traps, Yolanda Camacho (INBio). Comments: The type material was taken at a blacklight trap at Wafer Bay in March 1978, and at light at Chatham Bay in April 1979. According to Bright (1982), the species is most similar to P. biguttatus Blandford; the latter is known from Panama, Columbia and Venezuela, the host for which is given as Brosimum. All species in this genus are outbreeding and breed in inner bark. (2) PYCNARTHRUM BROSIMI WOOD Comments: This record is based on eight specimens (SLWC) collected from Brosimum by G.C. Stevens in April 1980 (Wood, 1982), and the identification of these specimens was verified by L.R.K., by comparison with type material in SLWC. This species has previously been collected only in Venezuela and Columbia. All species in this genus are outbreeding and breed in inner bark. (3) PYCNARTHRUM PSEUDOINSULARE SP. NOV. (FIG. 1) Type material: Holotype female(?): COSTA RICA, Puntarenas, Isla del Coco, Wafer Bay, 25 March 1978, C. Hogue & S. Miller (Steele exp.) (LACM). Etymology: The name is based on the similarity to P. insulare Blair, endemic to the Galápagos Islands. Figure 1. Comparison of Pycnarthrum pseudoinsulare sp. nov. (left side) and Pycnarthrum insulare Blair (right side). A, frons. B, dorsum (on the right is the scutellum, enlarged). C, posterior face of left protibia. Description of female (male unknown): Length 2.0 mm, 2.1 times longer than wide; colour yellowish brown. Frons (Fig. 1A) weakly convex above, flattened below; surface granulate, reticulate, subshining; granules associated with base of long hairlike setae which are evenly distributed from epistoma to well above upper level of eyes. Eyes separated above by more than twice their width. Antennal club 1.8 times longer than wide, with two weakly procurved, oblique sutures; funicle six-segmented. Pronotum 0.93 times as long as wide; sides subparallel to anterior third, then rather broadly rounded in front (Fig. 1B); surface reticulate, subshining, punctures shallow, separated by a distance equal to their diameter; vestiture consisting of bristles and hairlike setae. Elytra 1.45 times longer than wide, 1.8 times longer than pronotum, sides straight and parallel on basal © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 732 L. R. KIRKENDALL and B. H. JORDAL two-thirds, broadly rounded behind (Fig. 1B); scutellum (Fig. 1B) with tiny hairlike setae, base of elytra extending to and partly embracing anterior part of scutellum; striae 1 weakly impressed, other not impressed, punctures separated by a distance equal to their diameter; interstriae three times as wide as striae, punctures minute, obscure. Declivity steep, broadly flattened, striae 1 more deeply impressed than on disc. Vestiture consisting of scalelike interstrial setae in three irregular rows, bristles in median rows stouter and twice as long as those in lateral rows; tiny hairlike setae arising from strial punctures. Procoxae separated by 0.4 times the width of one procoxa; protibiae armed by a lateral row of irregularly sized unsocketed teeth, distal teeth one and two larger than the others (Fig. 1C). Lateral, apical edge of metatibiae truncately arched, with four small, socketed lateral teeth on distal fifth. Sex judged to be female by comparison with males and females of P. insulare. Comments: The holotype was compared with two syntypes of P. insulare (CAS) and eight other males and three other females of the latter species (MCZ). The new species is distinguished from the otherwise similar P. insulare by the different female frons, scutellum, protibiae and general outline (Fig. 1). P. pseudoinsulare is stouter (2.1 vs. 2.3 times longer than wide), has more widely spaced eyes, has a hairy scutellum that is embraced anteriorly by the elytral bases, and has protibiae which lack an additional tooth inside the lateral margin (between lateral teeth 2 and 3). In addition, the pronotal punctures are distinctly smaller in P. pseudoinsulare. The specimen identified by D.E. Bright as P. insulare (Bright, 1982) is our designated holotype of P. pseudoinsulare. The species described by Blair is consequently removed from the list of species on Cocos Island. (4) SCOLYTODES PACIFICUS JORDAL Comments: Previous records were included with the description of the species (Jordal, 1998a). This outbreeding species colonizes petioles of fallen Cecropia pittieri leaves. It is most closely related to S. impressus Wood, known from Cartago in central Costa Rica (Wood, 1982) and from the Osa Peninsula on the Pacific coast where it was taken from Xylopia sp. (Annonaceae) (Jordal, 1998b). (5) COCCOTRYPES CYPERI (BEESON) New record: Isla del Coco, April 1980, G.C. Stevens (2) (SLWC). Comments: Recorded for the first time on Cocos Island. In Costa Rica, this species is common at lower elevations, and collected from trunks, branches, petioles, and seeds of a wide variety of woody plants (Wood, Stevens & Lezama, 1991). According to Wood (1982; p. 736), this inbreeding species was introduced recently to the western hemisphere (where it is widespread) from the Indo-Malayan area. The earliest American record is from San José, Costa Rica (6 May 1915). (6) AMBROSIODMUS PAUCUS WOOD Comments: This species was described from a single female specimen collected in April 1980 from ‘Isla del Coco’ by G.C. Stevens (SLWC) (Wood, 1986a). No subsequent collections exist. It is related closely to A. devexulus (Wood), a Caribbean species, and there are several other similar species from Mexico and Central America. Ambrosiodmus is predominantly Asian and was separated only recently from Xyleborus (see Wood, 1986b); none of the neotropical species are collected frequently in Central America. As for all Xyleborini, Ambrosiodmus species are inbreeding ambrosia beetles. (7) XYLEBORINUS COCOENSIS SP. NOV. (FIGS 2, 3) Material: Holotype female: COSTA RICA, Puntarenas, Isla del Coco, Cerro Iglesias, 600 m, October 1994, Intersección, #3352, J.F. Quesada (INBio). Paratype females: same data as holotype (13); Bahia Chatan, October 1994, yellow pan traps, #3350, J.F. Quesada (83); Bahia Wafer, February 1994, Y. Camacho, Malaise, #2601 (2); 26–28 February 1980, T.K. Werner & T.W. Sherry, Malaise trap (1) (INBio); 17–21 March 1980, T.K. Werner & T.W. Sherry, Malaise trap (5) (INBio). Other material: ‘treefall, top C. Iglesias’, 28 February 1980, T.W. Sherry & T.K. Warner, Xyleborinus intersetosus (Bldf.), D.E. Bright 1980 (this individual was used for the SEM photos in Fig. 2). Etymology: The name is derived from the type locality, Isla del Coco. Description of female (male unknown): Length 1.78– 1.94 mm, 2.5–2.7 times longer than wide; mature colour dark brown, holotype yellow. Frons broadly convex; surface rugosely reticulate, a few coarse, shallow punctures; vestiture consisting of a few hairlike setae. Funicle five-segmented, antennal club obliquely truncate. Pronotum (Fig. 2A) 1.1–1.2 times longer than wide; sides straight and subparallel, rather broadly rounded in front; summit at middle; anterior slope coarsely asperate; posterior half reticulate, punctures fine, © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 COCOS ISLAND BARK BEETLES (SCOLYTINAE) 733 Figure 2. SEM photographs of three Xyleborinus species showing whole body and declivity. A, B, X. cocoensis sp. nov. C, D, X. intersetosus. E, F, X. gracilis. (The lighter area on the pronotum in ‘C’ is an artifact.) closely separated. Vestiture consisting of hairlike setae, confined mostly to marginal areas. Elytra (Fig. 2A) 1.5–1.6 times longer than wide, 1.4 times longer than pronotum; sides straight and paral- lel on basal two-thirds, then weakly tapered to the broadly rounded apex; disc confined to slightly more than basal half; scutellum conical and does not fit into the sutural notch; striae not impressed, punctures © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 734 L. R. KIRKENDALL and B. H. JORDAL tles of X. intersetosus are more widely spaced (by their length vs. by half their length), stouter, and in more complete rows. The declivital apex is more acutely rounded in X. intersetosus (Fig. 2C). (8) XYLEBORINUS GRACILIS (EICHHOFF) (FIGS 2, 3) New records: Chatham Bay, 5–9 February 1993, hand sampling, F. Quesada (2); Wafer Bay, Malaise trap, February 1994, Y. Camacho (1); Chatham Bay, October 1994, yellow pan trap, J.F. Quesada (4); Cerro Iglesias, 600 m, flight intercept trap, October 1994, J.F. Quesada (4). Figure 3. Posterior face of left protibia for three xyleborine species: A, Xyleborinus intersetosus; B, X. cocoensis sp. nov.; C, X. gracilis. Comments: First record for Cocos Island. The seven specimens collected from Cocos Island deviate slightly from mainland Costa Rican material in that the major denticle on the third interstriae on the declivity is not as dramatically produced. On the mainland, this ambrosia beetle has been taken from a variety of branches, lianas and tree saplings. (9) XYLEBORINUS INTERSETOSUS (BLANDFORD) (FIGS 2, 3) minute, separated within a row by two to four times their diameter; interstriae smooth, shining, five times wider than striae, punctures slightly smaller than those of striae. Declivity (Fig. 2B) moderately steep, broadly convex; striae 1 and 2 slightly impressed; interstriae 1 with tubercles absent or obscurely indicated, intestriae 2 smooth and with no tubercles, interstriae 3 with one minute tubercle near base and a larger one close to apex which is as large as the largest on the posterolateral margin. Vestiture consisting of rows of small hairlike strial setae, and rows of slightly confused interstrial bristles, each bristle twice as long as the distance between bristles within a row. Procoxae contiguous; protibiae (Fig. 3B) broad, with eight lateral, socketed teeth on distal half. Comments: Specimens were compared with the holotype of X. intersetosus Blandford (NHM) and with numerous specimens of that species from Cocos Island and the Costa Rican mainland deposited in the INBio and LRK collections. The new species is readily distinguished from the closely allied X. intersetosus by the numerous teeth on the protibiae (Fig. 3B), the latter having only four socketed teeth on distal third (Fig. 3A). Furthermore, X. intersetosus is smaller (max. 1.6 mm vs. max. 1.9 mm), and has rows of subequal tubercles on interstriae 1, 2 and 3 on the declivity, none of which are as large as are those on ventrolateral angles of the declivity. Interstrial bris- New Records: Wafer Bay, Malaise trap, March 1994, Y. Camacho (4). Comments: The single specimen of X. intersetosus previously reported from Cocos Island (Bright, 1982) belongs to the new species described above. Hence, the four specimens reported here constitute the first true record of this polyphagous ambrosia beetle from this island. 10. XYLEBORUS AFFINIS EICHHOFF New Records: Wafer Bay, Malaise trap, Feb (4), March (5), April (31), May (11), June (8) 1994 (total 59), Y. Camacho; Chatham Bay, 5–9 February 1993, J.F. Quesada, Malaise trap (75), by hand (10); Chatham Bay, yellow pan trap, October 1994, J.F. Quesada (11); Cerro Iglesias, 600 m, flight intercept trap, October 1994, J.F. Quesada (3); El Guarumal, trail WaferChatham, 30 m, 29 December 1997, at light, E. Ulate and C. Flores (3); Iglesias waterfall at Bahia Iglesias, hand sampling, 21 December 1997, E. Ulate and C. Flores (1); El Gallinero coast, trail to Cerro Iglesias, 260 m, Malaise trap 28 December 1997–28 February 1998, E. Ulate and C. Flores (1). Comments: Previous records are from blacklight, white light, and hand sampling. No host records have been recorded from the island, but this cosmotropical inbreeding ambrosia beetle is extremely polyphagous. © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 COCOS ISLAND BARK BEETLES (SCOLYTINAE) 735 Figure 4. Comparisons of elytral interstriae and of lateral profiles for Xyleborus bispinatus and X. ferrugineus (digital photos). A, dorsal view of elytra of X. bispinatus. B, dorsal view of elytra of X. ferrugineus. C, lateral view of X. bispinatus (above) and X. ferrugineus (below). 11. XYLEBORUS BISPINATUS EICHHOFF (REINSTATED SPECIES) (FIG. 4) New Records: Wafer Bay, Malaise trap, June 1994, Y. Camacho (3); Chatham Bay, October 1994, J.F. Quesada (9); Chatham Bay, yellow pan trap, October 1994, J.F. Quesada (6); Cerro Iglesias, 600 m altitude, flight intercept trap, October 1994, J.F. Quesada (9). Comments: We believe that neotropical specimens currently being identified by bark beetle specialists as X. ferrugineus actually comprise two species, Xyleborus ferrugineus s.s. and a second, larger species. These two forms are broadly sympatric in Central and northern South America, and were clearly distinguished by Blandford (1905) for Central America. The smaller taxon with the pair of large denticles in the middle of the declivity (Fig. 4B, C) is X. ferrugineus (Fabricius) (lectotype examined by Kirkendall). The larger species agrees with the Brazilian syntypes of X. bispinatus Eichhoff, which is currently a synonym of X. ferrugineus (Schedl, 1960b); X. bispinatus is the senior-most of several synonyms which we believe apply to this second species. [The taxonomy of North American ‘ferrugineus’ is also unclear, and the subject of current research; two forms are found, one of which is X. impressus Eichhoff (Rabaglia, 2005), the other of which seems to be consistantly different from X. ferrugineus and X. bispinatus and which should probably be treated as X. fuscatus Eichhoff (L.R. Kirkendall, unpubl.)]. Xyleborus bispinatus is larger on average than is sympatric X. ferrugineus (seen clearly when series of the two are compared directly) and usually slightly darker; it has the major pair of denticles on the declivity closer to the base of the declivity (Fig. 4A, C). It has regularly spaced hairlike setae in the discal interstriae (the interspaces between the rows of punctures) of the elytra (Fig. 4A); these setae are few or absent in X. ferrugineus, especially near the base of the elytra (Fig. 4B). X. ferrugineus has a weak, broad impunctate vertical median carina on the frons. The frons of X. bispinatus is usually flattened medially (with at © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 736 L. R. KIRKENDALL and B. H. JORDAL most a very short broad carina just above the epistoma); in the place of the carina is an impunctate median area in the shape of an inverted triangle, the upper portion of which occasionally appears slightly impressed. Specimens of X. bispinatus have been examined from countries as far north as Belize and as far south as Ecuador, Bolivia and Brazil, but an exhaustive search of museum collections has not yet been carried out, so its exact distribution is not yet known. Lectotype designation: We hereby designate one of Eichhoff ’s seven Brazilian syntypes as the lectotype for X. bispinatus. The original, round locality label reads ‘S. Cathr./Deyr.’, which we interpret as being collected by Henri Deyrolle from the province of Santa Catarina in southern Brazil (see collection data for species 99 in Schedl, 1960a). The handwritten round label has been mounted on a printed purple museum label reading ‘Coll. R. I. Sc. N. B./Brazil’, to which is also glued an original ‘Coll. Chapuis’ label. Also on the pin is a ‘Type’ label, and a mounted, handwritten label in Eichhoff ’s hand, ‘Xyleborus bispinatus m’ (‘meus’, mine). Four other specimens have similar mounted round labels but with the locality given as ‘Bresil’. All seven specimens have printed ‘Type’ labels. All have labels saying they were determined by Eichhoff; four (including the lectotype) have determination labels written by Eichhoff. We designate the other six specimens as paralectotypes, and have labelled the seven in accordance with this treatment. 12. XYLEBORUS FERRUGINEUS (F.) New Records: Chatham Bay, Río Genio, Malaise trap, 5–9 February 1993, J.F. Quesada (17); Wafer Bay, Malaise trap, June 1994, Y. Camacho (3); Cerro Iglesias, 600 m altitude, flight intercept trap, October 1994, J.F. Quesada (1); Casa Wafer, 15–31 December 1997, hand collecting, Ulate and Flores (1); El Guarumal, WaferChatham trail, 30 m, at light, 29 December 1997, Ulate and Flores (10). Comments: See comments under X. bispinatus. Previously recorded Cocos Island specimens of X. ferrugineus were taken from rotting wood and at blacklight. One host has been recorded, Ochroma pyrimidalis (balsa) (Schedl, 1974), but these ambrosia beetles are quite polyphagous. 13. XYLEBORUS PARALLELOCOLLIS EGGERS (FIG. 5) New Records: Chatham Bay, 5–9 February 1993, by hand, J.F. Quesada (1); Chatham Bay, October 1994, J.F. Quesada, yellow pan trap (2), unspecified (1). Figure 5. Outline of declivity. A, Xyleborus sparsegranulosus sp. nov. B, X. parallelocollis. Comments: These specimens are identical to the specimen collected in a blacklight trap by the Steele expedition (determined by D.E. Bright as X. parallelocollis Eggers). L.R.K. studied the holotype of X. parallelocollis, and material which was compared directly with the holotype was compared in turn with numerous specimens in SLWC, UCR and INBio collections which were collected from both the Pacific and Atlantic sides of Costa Rica and from Venezuela, and with a series of specimens from Ecuador (LRKC). These specimens are clearly conspecific. Published records of X. parallelocollis are from Costa Rica, Colombia and © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 COCOS ISLAND BARK BEETLES (SCOLYTINAE) French Guiana (Wood & Bright, 1992). (See also comments under species 15, below.) 14. XYLEBORUS PERFORANS WOLLASTON New Records: Chatham Bay, February 1993, Malaise trap (18) and by hand (4), J.F. Quesada; Chatham Bay, October 1994, yellow pan trap, J.F. Quesada (7); Wafer Bay, Malaise trap, Feb (2), March (6), April (2), and June (2) 1994, Y. Camacho; Cerro Iglesias, 600 m, flight intercept trap, Oct. 1994, J.F. Quesada (5); Iglesias waterfall, Iglesias Bay, by the trail, 21 December 1997, Ulate and Flores (26); El Guarumal, WaferChatham trail, 30 m, at light, Ulate and Flores (12); El Gallinero coast, trail to Cerro Iglesias, 260 m, Malaise trap 28 December 1997–28 February 1998, E. Ulate and C. Flores (1). Comments: The earlier collection (two individuals) by Sherry and Werner (X. volvulus (F.) according to Bright, 1982) was at light. No host records are known, but the species is quite polyphagous. Occasionally the east Asian X. perforans Wollaston has been considered a synonym of the neotropical X. volvulus (e.g. Bright, 1982), but the two remain as doubtfully separate taxa which are very difficult to distinguish by morphological characters (Wood & Bright, 1992). Mitochondrial DNA sequences of cytochrome oxidase I from two Cocos Island specimens were identical; to our surprise, this Cocos haplotype was very similar to haplotypes from three Papua New Guinean populations of X. perforans but very divergent from sequences from Costa Rican populations of X. volvulus (unpubl. data). For the present, we must consider the Cocos Island population, then, to be X. perforans rather than X. volvulus (in agreement with morphology: Schedl, 1974; contra Bright, 1982) and we consequently remove X. volvulus from the island list. (It is very likely, of course, that future molecular taxonomic research will reveal the presence of X. volvulus on the island as well, given the abundance of that species throughout the neotropics.) 15. XYLEBORUS SPARSEGRANULOSUS SP. NOV. (FIG. 5) Material: Holotype female: COSTA RICA, Puntarenas, Isla del Coco, Bahia Chatan, intersección, October 1994, #3352, J.F. Quesada [INBio]. Paratypes: same data as holotype, 8 females [INBio]; Bahia Chatan, 1–100 m, October 1994, #3350, J.F. Quesada [INBio] (1 female); Bahia Wafer, 1 m, October 1994, #3314, J.F. Quesada [INBio] (1 female). Etymology: The new species has relatively few declivital granules compared with X. parallelocollis and related species. 737 Description of female (male unknown): Length 3.1– 3.2 mm, 2.5–2.6 times longer than wide, colour brown. Frons broadly convex, with a broad, median, weakly elevated carina; surface reticulate, with rather coarse shallowly impressed punctures. Vestiture consisting of a few hairlike setae, mostly close to epistoma. Antennal club obliquely truncated; funicle five-segmented. Pronotum 1.15 times longer than wide, subquadrate; summit at middle; asperate on anterior slope, posterior disc smooth, tiny asperities extending laterally behind summit, punctures minute, sparse and obscure; vestiture consisting of a few hairlike setae in asperate area. Elytra 1.6 times longer than wide, 1.45 times longer than pronotum; lateral margins subparallel to posterior third, narrowly rounded behind; striae not impressed, punctures small, distinctly impressed, separated by their diameter; interstriae twice as wide as striae, smooth, with confused minute punctures separated by a distance equal to their diameter. Declivity (Fig. 5A) broadly flattened, transversely impressed on lower third, striae impressed, punctures shallow, separated by less than their diameter; on average, three to four large granules on interstriae 1–3 above transverse impression, the largest granule more than three times as large as the smallest, the larger granules separated within a row by three to four times their diameter. Vestiture consisting of a few scattered interstrial bristles on disc, and much shorter interstrial bristles on declivity, each declivital bristle equal in length to one third the distance between rows, and minute interspersed interstrial and strial hairlike setae. Procoxae contiguous, protibiae with eight lateral socketed teeth on distal half; metatibiae with nine to 11 lateral socketed teeth on distal two thirds, lateral edge evenly arched to apex. Comments: Although species delimitation of inbreeding beetles (such as those of Xyleborus) can be difficult from a limited sample, X. sparsegranulosus is distinct from all similar species. It is readily identifiable by the pattern of declivital granules (Fig. 5). In addition, it is distinguished from X. parallelocollis and X. semipunctatus by its larger size, and from sympatric X. parallelocollis also by the more rounded anterior margin of the pronotum. It can be distinguished from X. discretus and X. semipunctatus by the more gradually sloping declivity, and from X. concentus by the obtuse (vs. acute) apex of the declivity. Examination of many X. parallelocollis from Costa Rica, Ecuador and Venezuela, as well as holotypes of X. semipunctatus, X. concentus, and X. usticus (a junior synonym of discretus Eggers) led us to conclude that the Cocos Island population deserves specific status. © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 738 L. R. KIRKENDALL and B. H. JORDAL 16. HYPOTHENEMUS ERUDITUS WESTWOOD New Records: Wafer Bay, Malaise trap, February 1994 (1), Wafer Bay, Malaise trap, March 1994 (1), Wafer Bay, Malaise trap, May 1994 (1), Cerro Iglesias, 600 m, flight intercept trap, Oct. 1994, J.F. Quesada (1); Wafer Bay, 17–22 April 1975, by hand, C.L. Hogue (2). Comments: Recorded for the first time on Cocos Island. The genus had not been collected previously from the island but was to be expected there, since many species of Hypothenemus are good colonists and are easily spread by human activity (see e.g. Wood, 1982). For example, ten of 24 scolytines on the Galápagos Islands are species of Hypothenemus. This tiny (1.0–1.3 mm) species is argueably the most abundant and widespread scolytine; it has been collected from the Americas, all of Africa, southern Europe, the warmer parts of Asia and Australia, and many tropical and subtropical islands around the world. It is ecologically flexible, having been collected from several hundred plant species and from a wide variety of plant tissues. On mainland Costa Rica, the species has been collected by Malaise traps at elevations ranging from sea level to 3000 m altitude (L.R. Kirkendall, unpubl. data). Not infrequently, H. eruditus is the only species, or one of only a few species, of bark beetle collected on isolated small islands (e.g. Bright & Peck, 1998) and in dry, hot microclimates such as sun-exposed bark or petioles (e.g. Jordal & Kirkendall, 1998). All Hypothenemus are inbreeders. The taxonomy of H. eruditus is very uncertain, as is the case for many inbreeding scolytines, and 70 synonyms of this taxon are listed in Wood and Bright’s (1992) catalogue. There is considerable morphological variability among specimens assigned to this species, particularly in characters such as the sculpture of the frons, body colour, and vestiture. 17. HYPOTHENEMUS SERIATUS (EICHHOFF) New Records: Wafer Bay, March 1994, Malaise trap (1). Comments: Recorded for the first time in Cocos Island. This and previously determined Cocos Island specimens (Bright, 1982) were compared with the lectotype of H. seriatus and with specimens of H. seriatus from Costa Rica, and with type and nontype material of H. obscurus and H. crudiae, two species with which H. seriatus is confused frequently. This widespread, polyphagous species breeds in twigs and branches, seeds, seedpods, and other plant material (Wood, 1982). Its primary distribution is from eastern North America south through Mexico and Central America to Brazil, but it is also found on tropical and subtropical islands throughout the world, including the Galápagos (Bright & Peck, 1998). 18. ARAPTUS SP. New Records: Wafer Bay, 24 March 1978, blacklight trap, C. Hogue and S. Miller, Steele expedition (1); Bahia Chatan, 5–9 February 1993, F. Quesada (2); Cerro Iglesias, 600 m altitude, flight intercept trap, October 1994, J.F. Quesada (3); banks of the Río Genio, 10 m altitude, 17–19 December 1997, flight intercept trap, Ulate and Flores (5: 2 are incomplete, 3 were extracted for DNA and then mounted); Los Llanos, 24 December 1997, beating sheet (‘sombrereta’), Ulate and Flores (1). Comments: The specimen previously determined as ‘Pityophthorus sp.’ (Bright, 1982), collected from the Steele expedition blacklight trap in March 1978, is an Araptus, and is clearly a member of the distinctive A. costaricensis complex of inbreeding bark beetles. A. costaricensis, A. laevigatus (Eggers), and at least five undescribed species occur in Costa Rica and Panama. The Cocos Island species can be distinguished from other species of the complex by characteristics of the frons and by body size and shape. Species of this Araptus complex are apparently all inbreeding by sibmating, and five are known to breed primarily in unusual plant material, such as fallen woody pods and fallen woody leafstalks. (The taxonomy and ecology of this complex will be published separately by L.R.K.). 19. CORTHYLUS SP. New record: Hill between Chatham and Cape Atrevida, flight intercept trap, 600 m altitude, October 1994, J.F. Quesada (1). Comments: The single, damaged male of this Corthylus species could not be identified from the literature because keys to this genus rely heavily on characters of the female frons. Neither could it be identified by comparison with the material at hand or in the collection of the University of Costa Rica. We compared it with paratypes or authenticated male specimens of a variety of Costa Rican species of similar form and size, including C. collaris Blandford, C. comatus Blandford, C. retusus Wood, C. splendens Wood, C. spinifer Schwarz, and C. villus Bright, as well as with males of several unidentified, probably new species from Costa Rica (in the INBio collections). The male specimen has no unusual features. It is 2.2 mm long, black, with an antenna similar to that of C. comatus, and with a coarsely faceted eye; the declivital interstriae 1 are raised and flattened (but interstriae 3 are not raised). All Corthylus are outbreeding ambrosia beetles. © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 COCOS ISLAND BARK BEETLES (SCOLYTINAE) BEHAVIOURAL ECOLOGY OF COCOS ISLAND SCOLYTINAE Although botanists have long recognized the potential significance of selfing for successful colonization of islands (‘Baker’s Rule’: Baker, 1955; Brown & Burdon, 1987; McMullen, 1987; Anderson et al., 2001; Rambuda & Johnson, 2004), it seems that zoologists have largely ignored the possible importance of inbreeding in this context. Bark beetles are not the only terrestrial arthropods with clades in which both outbreeding and incestuous species are found – regular close inbreeding occurs in, for example, certain mites, thrips, aphids, solitary bees and wasps, ants and social spiders (Thornhill, 1993; Wrensch & Ebbert, 1993). However, they are apparently the only group in which an association between zoogeography and close inbreeding has been documented (Kirkendall, 1993; Jordal et al., 2001). The proportion of the scolytine fauna that inbreeds by repeated sib-mating increases with decreasing latitude, and is highest on tropical islands. In agreement with these patterns, species which regularly inbreed by brother–sister matings represent nearly three quarters (74%, Table 1) of the known scolytine fauna of Cocos Island. This is about average for small tropical islands, compared with 40–50% for mainland (e.g. Costa Rican lowland forests) and large island tropical faunas (Bright & Peck, 1998; Kirkendall, 1993; Jordal et al., 2001: fig. 7.1). 739 The over-representation of inbreeders on islands such as Cocos suggests that they have a higher success rate than do outbreeders for establishing viable populations after colonization. Females of inbreeding species mate with a brother before dispersal from the natal breeding material; mate finding in recently established, sparse inbreeding populations is not the problem that it would be for outbreeders. Furthermore, presumably inbreeding colonists do not suffer the sudden decrease in absolute fitness due to inbreeding depression that is expected to afflict small initial populations of outbreeders. Eleven of the 19 species reported here are ambrosia beetles (Table 1), a typical proportion for small tropical islands (Kirkendall, 1993, fig. 7.2D). Unlike most phloem-feeding bark beetles, ambrosia beetles are host plant generalists; they feed on mats of fungal hyphae growing in their tunnels instead of on host tissues directly, and these fungal symbionts are host generalists. The ability to utilize a broad range of plant taxa should increase their likelihood of colonization success relative to that of host specialists. COCOS ISLAND ZOOGEOGRAPHY Only seven or eight (depending on the Corthylus species) of the 19 species reported here are endemic to Cocos Island (Table 1), compared with the estimate of 52% endemic for the total arthropod community Table 1. Systematic placement (Wood, 1982), distribution and mating systems of Scolytinae found on Cocos Island; the Xyleborini and Corthylus are ambrosia beetles Tribe Species Distribution Mating system Phloeotribini Ctenophorini Phloeotribus pacificus Bright Pycnarthrum brosimi Wood Pycnarthrum pseudoinsulare sp. nov. Scolytodes pacificus Jordal Coccotrypes cyperi (Beeson) Ambrosiodmus paucus Wood Xyleborinus cocoensis sp. nov. Xyleborinus gracilis (Eichhoff) Xyleborinus intersetosus (Blandford) Xyleborus affinis Eichhoff Xyleborus bispinatus Eichhoff Xyleborus ferrugineus (F.) Xyleborus parallelocollis Eggers Xyleborus perforans Wollaston Xyleborus sparsegranulosus sp. nov. Hypothenemus eruditus Westwood Hypothenemus seriatus (Eichhoff) Araptus sp. Corthylus sp. Endemic Costa Rica-Colombia Endemic Endemic Pantropical Endemic Endemic Tropical America Tropical America Pantropical Tropical America Pantropical Tropical America Pantropical Endemic Pantropical Pantropical Endemic ? Outbreeding Outbreeding Outbreeding Outbreeding Inbreeding Inbreeding Inbreeding Inbreeding Inbreeding Inbreeding Inbreeding Inbreeding Inbreeding Inbreeding Inbreeding Inbreeding Inbreeding Inbreeding Outbreeding Dryocoetini Xyleborini Cryphalini Corthylini © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 740 L. R. KIRKENDALL and B. H. JORDAL (Hogue & Miller, 1981). However, three (possibly four) of the five outbreeding species are endemic, which is more in line with the general pattern for arthropods. The contrast with inbreeding species (only four of 14 endemic) is in accordance with a pattern seen for other tropical islands (Jordal et al., 2001) including the intensively collected Galápagos archipelago (Bright & Peck, 1998). These data, then, add support to our previous conclusions that, even when correcting for ecological differences between inbreeders and outbreeders, (1) inbreeders are clearly superior colonists, and (2) successful colonization of small islands by outbreeders is more likely to result in species-level morphological divergence than is that by inbreeders, resulting in higher levels of endemism in the former (Jordal et al., 2001). There are three patterns of colonization for the putative sister relationships we could find for insects, a scorpion, and a bird endemic to Cocos Island (Table 2). Primary colonization routes have been from the Pacific side of Central and northern South America (13 cases), the Caribbean (five), and the Galápagos (two). Five of the seven endemic scolytine species are related to species found primarily in Central or northern South America, which suggests that the main route of overseas dispersal for these beetles was westwards from the Americas, as surmised for the Galápagos insect fauna as a whole (Peck, 1994). Species not endemic to Cocos Island, but with narrow geographical distributions, further support this trend, as demonstrated by P. brosimi and X. parallelocollis. The fact that the neotropical Xyleborinus intersetosus and Xyleborus parallelocollis appear to be sister species to Cocos Island endemics strongly suggests that there have been repeated successful colonizations from the American mainland by these inbreeding lineages, and that the island was initially colonized without the aid of human transport. Anthropogenic spread, however, is the only plausible explanation for the immigration of X. perforans and the tineid moth Erechthias flavistriata (Davis, 1994) from the West Pacific, since these species are otherwise not known from the Americas. Several lepidopteran groups and one ambrosia beetle seem to have a Caribbean islands–Cocos Island vicariant distribution (Table 2), which could be explained by dispersal by an extinct or uncollected Central or South American ancestor rather than dispersal directly from any of the Caribbean islands. Lepyrotica acantha is found on both the Galápagos and Cocos Island, a distribution pattern shared with Pycnarthrum insulare/P. pseudoinsolare and with Darwin’s finches (Sato et al., 2001). Only phylogeographical studies can reveal the details of the colonization direction, but at least for Darwin’s finches, Sato and co-workers argue that Cocos Island was colonized from the Galápagos. Cocos Island originated c. 500 km south-west of its current position and hence was initially much closer to the Galápagos Islands than it is today (Castillo et al., 1988). CONCLUDING REMARKS The scolytine fauna of Cocos Island is notable for its species richness relative to other arthropod groups, when compared with mainland faunas. La Selva Biological Station on the Caribbean side of Costa Rica is the site of a long-term, intensive survey of selected arthropod groups (e.g. Colwell & Longino, 2005). In this typical mainland wet tropical forest, compared with bark beetles there are twice as many species of ants, two thirds as many buprestids, one and a half times as many cerambycids and probably an order of magnitude more weevils (other than scolytines) (L.R. Kirkendall and J.T. Longino, unpubl. data; H.A. Hespenheide, pers. comm.; F. Hovore, pers. comm.). On Cocos Island, bark beetles and ants are equal in terms of numbers of species, while bark beetles outnumber buprestids by 19 : 1, cerambycids by about 2 : 1, and other weevils by about 2 : 1 (Hogue & Miller, 1981; E. Ulate, INBio, upubl. data). Are these unusual proportions simply an example of the stochasticity of long-distance colonization? We argue that scolytines should be extraordinarily successful colonizers because of a combination of factors. First, as discussed above, many scolytines benefit (in this context) from being inbreeders. Second, scolytines are an appropriate size for long-distance aerial transport (e.g. Hespenheide, 1977). Third, biologists have long known that insects and spiders living in temporary habitats (such as dead woody tissues) are good colonizers because of their adaptations to a lifestyle dependent upon long-distance dispersal (e.g. Peck, 1994). Fourth, unlike most other wood-boring beetles, both adults and immatures tunnel in dead wood, and adults are capable of tightly blocking tunnel entrances with frass or their own bodies (which could protect tunnel inhabitants from flooding), increasing the likelihood of surviving a long oceanic voyage in dead floating vegetation. Understanding the evolution of the Cocos Island fauna depends on taxonomic research. Unfortunately, despite the large number of insect specimens collected and prepared by INBio personnel in the past decade, the comprehensive summary of Hogue & Miller (1981) has not yet been updated; in fact, our paper is the first review of Cocos Island representatives of any arthropod family or subfamily which includes the INBio collections, and apparently only the second publication of newly discovered insect species from those collections (the other being Anderson & Lanteri, 2000). We hope that our paper will inspire others to follow suit: we feel that Cocos Island is especially valuable for studies of © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 Order Endemic taxon Sister taxon/taxa Distribution (sister taxon/taxa) Reference Neuroptera Orthoptera Orthoptera Lepidoptera Lepidoptera Lepidoptera Lepidoptera Lepidoptera Diptera Diptera Diptera Diptera Diptera Coleoptera Coleoptera Coleoptera Coleoptera Coleoptera Coleoptera Coleoptera Coleoptera Scorpiones Passerine birds Chrysoperla externa cocosensis Parascopioricus binoditergus Cocconotus bellicosus Oxydia hoguei Coelostathma insularis Protodarcia cocosensis Lepyrotica acantha* Tortyra hoguella Paracanace hoguei Merosargus insularis Limonia hoguei Helius brunneus Orimaga flavescens Halecia cocosae Phloeotribus pacificus Scolytodes pacificus Pycnarthrum pseudoinsulare sp. nov. Ambrosiodmus paucus Xyleborinus cocoensis sp. nov. Xyleborus sparsegranulosus sp. nov. Araptus sp. Opisthacanthus valerioi Pheuticus ludovicianus C. e. externa P. cordillericus Not given O. lalanneorum C. binotata Protodarcia spp. (4) Lepyrotica spp. (6) Tortyra spp. P. aicea and P. leban M. penai L. willistoniana H. micracanthus O. omissinervis H. auropunctata P. biguttatus S. impressus P. insulare X. devexulus X. intersetosus X. parallelocollis A. costaricensis complex (c. 6 spp.) Opisthacanthus spp. (4) P. aureoventris Central America, SE US, Antilles Columbia Columbia Guadeloupe Mexico Caribbean islands Caribbean islands Mexico to South America Greater Antilles Equador Lesser Antilles & Costa Rica Brazil Bolivia Nicaragua to Peru Panama-Columbia/Venezuela Costa Rica Galápagos Caribbean islands Tropical America Tropical America Costa Rica-Panama Northern S. America, Peru Galápagos Adams (1983) Nickle (1983) Nickle (1983) Brown et al. (1991) Brown & Miller (1999) Davis (1994) Davis (1994) Heppner (1980) Mathis & Wirth (1978) Woodley (1991) Byers (1981) Byers (1981) Byers (1981) Bellamy (1986) Bright (1982) Jordal (1998a) This paper Wood (1986a) This paper This paper This paper Lourenço (1980) Sato et al. (2001) SE, south-eastern. COCOS ISLAND BARK BEETLES (SCOLYTINAE) © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 Table 2. Summary of well-documented sister relationships in a bird, a scorpion and various insects endemic to Coco Island, including the scolytine species described in this paper; the distribution of putative closest relatives are given along with reference to primary sources 741 742 L. R. KIRKENDALL and B. H. JORDAL neotropical biogeography, given that it is the only eastern Pacific oceanic island with wet tropical forest. ACKNOWLEDGEMENTS Many thanks to Scott Miller (at that time, LACM) and to the staff at INBio (especially to Elaine Ulate) for making specimens and collection information available for this study. Our job was simplified by those who responded promptly to our requests for difficultto-obtain literature: Robert Anderson, Charles Bellamy, George W. Byers, Don Bright, John Brown, Jorge Cortés, Don Davis, William Eberhard, Caroline Godoy, Robert Haack, and Scott Miller. We are grateful for prompt replies to loan requests by Ole Martin (ZMUC), Marc de Meyer (RMCA) and Didier Drugmond (IRSNB), and for specimens of ‘X. ferrugineus’ from Robert Rabaglia and Roger Beaver. We thank Arguitxu de la Riva Caballero for preparing the Spanish abstract. Finally, Nicole Berti’s assistance during a visit by L.R.K. to the Muséum National d’Histoire Naturelle (Paris) was much appreciated. REFERENCES Adams PA. 1983. A new subspecies of Chrysoperla externa (Hagen) from Cocos Island, Costa Rica (Neuroptera: Chrysopidae). Bulletin of the Southern California Academy of Sciences 82: 42–45. Anderson GJ, Bernardello G, Stuessy TF, Crawford DJ. 2001. Breeding system and pollination of selected plants endemic to Juan Fernandez Islands. American Journal of Botany 88: 220–233. Anderson RS, Lanteri AA. 2000. New genera and species of weevils from the Galapagos Islands, Ecuador, and Cocos Island, Costa Rica (Coleoptera; Curculionidae; Entiminae; Entimini). American Museum Novitates 3299: 1–15. Baker HG. 1955. Self-compatibility and establishment after ‘long-distance’ dispersal. Evolution 9: 347–349. Bellamy CL. 1986. A new species of Halecia from Cocos Island, Costa Rica, with a review of the neotropical genera of the tribe Chalcophorini. Coleopterists Bulletin 40: 381– 387. Blandford WFH. 1905. Fam. Scolytidae. Biologia CentraliAmericana 4: 281–298, Tab. 9. Bright DE. 1982. Scolytidae (Coleoptera) from the Cocos Islands, Costa Rica, with a description of one new species. Coleopterists Bulletin 36: 127–130. Bright DE, Peck SB. 1998. Scolytidae from the Galápagos Islands, Ecuador, with descriptions of four new species, new distribution records, and a key to species (Coleoptera: Scolytidae). Koleopterologische Rundschau 68: 233–252. Brown JW. 1990. Sphingidae (Lepidoptera) of Isla del Coco, Costa Rica, with remarks on the macrolepidoptera fauna. Brenesia 33: 81–84. Brown AHD, Burdon JJ. 1987. Mating systems and colonizing success in plants. In: Gray AJ, Crawley MJ, Edwards PJ, eds. Colonization, succession and stability. Oxford: Blackwell Scientific Publications, 115–131. Brown JW, Donahue JP, Miller SE. 1991. Two new species of geometrid moths (Lepidoptera: Geometridae: Ennominae) from Cocos Island, Costa Rica. Natural History Museum of Los Angeles County, Contributions in Science 423: 11–18. Brown JW, Miller SE. 1999. A new species of Coelostathma Clemens (Lepidoptera: Tortricidae) from Cocos Island, Costa Rica, with comments on the phylogenetic significance of abdominal dorsal pits in Sparganothini. Proceedings of the Entomological Society of Washington 101: 701–707. Byers GW. 1981. The crane flies (Diptera: Tipulidae) of Cocos Island, Costa Rica with descriptions of new species. Natural History Museum of Los Angeles County, Contributions in Science 335: 1–8. Castillo P, Batiza R, Vanko D, Malavassi E, Barquero J, Fernandez E. 1988. Anomalously young volcanoes on old hot-spot traces. I. Geology and petrology of Cocos Island. Geological Society of America Bulletin 100: 1400–1414. Colwell R, Longino JT. 2005. Project ALAS, Arthropods of La Selva. http://viceroy.eeb.uconn.edu/alas/alas.html (last viewed 4 October 2006). Cwikla PS, Freytag PH. 1982. Three new leafhoppers (Homoptera: Cicadellidae) from Cocos Island. Proceedings of the Entomological Society of Washington 84: 632–635. Davis DB. 1994. Neotropical Tineidae, V: The Tineidae of Cocos Island, Costa Rica (Lepidoptera: Tineoidea). Proceedings of the Entomological Society of Washington 96: 735–748. Eberhard WG. 1989. Niche expansion in the spider Wendilgarda galapagensis (Araneae, Theridosomatidae) on Cocos Island. Revista de Biologia Tropical 37: 163–168. Fournier LA. 1966. Botany of Cocos Island, Costa Rica. In: Bowman RI, ed. The Galápagos. Berkeley: University of California Press, 183–186. Gillespie RG, Roderick GK. 2002. Arthropods on islands: colonization, speciation, and conservation. Annual Review of Entomology 47: 595–632. Hamilton WD. 1967. Extraordinary sex ratios. Science 156: 477–488. Hepner JB. 1980. (1981). A new Tortyra from Cocos Island, Costa Rica (Lepidoptera: Choreutidae). Journal of Research on the Lepidoptera 19: 196–198. Hespenheide HA. 1977. Dispersion of and the size composition of the aerial insect fauna. Ecological Entomology 2: 139– 141. Hogue CL, Miller SE. 1981. Entomofauna of Cocos Island, Costa Rica. Atoll Research Bulletin 250: 1–29. Johnson C. 1982. Ptiliidae from the Galapagos and Cocos Islands. Brenesia 19/20: 189–199. Jordal BH. 1998a. A review of Scolytodes Ferrari (Coleoptera: Scolytidae) associated with Cecropia (Cecropiaceae) in the northern Neotropics. Journal of Natural History 32: 31–84. Jordal BH. 1998b. New species and new records of Scolytodes Ferrari (Coleoptera: Scolytidae) from Costa Rica and Panama. Revista de Biologia Tropical 46: 407–420. Jordal BH, Beaver RA, Kirkendall LR. 2001. Breaking taboos in the tropics: inbreeding promotes colonization by wood-boring beetles. Global Ecology and Biogeography 10: 345–358. © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743 COCOS ISLAND BARK BEETLES (SCOLYTINAE) Jordal B, Kirkendall LR. 1998. Ecological relationships of a guild of tropical beetles breeding in Cecropia petioles in Costa Rica. Journal of Tropical Ecology 14: 153–176. Juan C, Emerson BC, Oromi P, Hewitt GM. 2000. Colonization and diversification: towards a phylogeographic synthesis for the Canary Islands. Trends in Ecology and Evolution 15: 104–109. Kirkendall LR. 1983. The evolution of mating systems in bark and ambrosia beetles (Coleoptera: Scolytidae and Platypodidae). Zoological Journal of the Linnean Society 77: 293–352. Kirkendall LR. 1993. Ecology and evolution of biased sex ratios in bark and ambrosia beetles (Scolytidae). In: Wrensch DL, Ebbert MA, eds. Evolution and diversity of sex ratio: insects and mites. New York: Chapman & Hall, 235–345. Lourenço WR. 1980. Two new Opisthacanthus species for the neotropical region: Opisthacanthus valerioi, n.sp. of the Island of Coco, Costa Rica and Opisthacanthus heurtaultae, n.sp. of French Guiana (Scorpiones: Scorpionidae). Revista Nordestina de Biologia 3: 179–194. Mathis WN, Wirth WW. 1978. A new genus near Canaceoides Cresson, three new species and notes on their classification (Diptera: Canacidae). Proceedings of the Entomological Society of Washington 80: 524–537. McMullen CK. 1987. Breeding systems of selected Galapagos Islands angiosperms. American Journal of Botany 74: 1694– 1705. Nickle DA. 1983. A new species of pseudophylline katydid from Cocos Island, Costa Rica (Orthoptera: Tettigoniidae). Entomological News 94: 1–6. Peck SB. 1994. Aerial dispersal of insects between and to islands in the Galápagos Archipelago. Annals of the Entomological Society of America 87: 218–224. Rabaglia RJ. 2005. The validity of Xyleborus impressus Eichhoff (Coleoptera: Curculionidae: Scolytinae) as distinct from Xyleborus ferrugineus (Fabricius). Coleopterists Bulletin 59: 261–266. Rambuda TD, Johnson SD. 2004. Breeding systems of invasive alien plants in South Africa: does Baker’s rule apply? Diversity and Distributions 10: 409–416. Sato A, Tichy H, O’hUigin C, Grant PR, Grant BR, Klein J. 2001. On the origin of Darwin’s finches. Molecular Biology and Evolution 18: 299–311. 743 Schedl KE. 1960a. Chapuis Platypodidae, eine Revision mit Ergänzungen. 184. Beitrag zur morphologie und systematik der Scolytoidea. Institut Royal Des Sciences Naturelles de Belgique, Memoires 62: 1–68. Schedl KE. 1960b. Synonymies of bark beetles (Scolytidae) IV. Coleopterists Bulletin 14: 5–12. Schedl KE. 1974. Scolytoidea from the Galapagos and Cocos Islands. Studies of the Neotropical Fauna 9: 47–53. Solomon SE, Mikheyev AS. 2005. The ant (Hymenoptera: Formicidae) fauna of Cocos Island, Costa Rica. Florida Entomologist 88: 415–423. Thornhill NW. 1993. The natural history of inbreeding and outbreeding: theoretical and empirical perspectives. Chicago: University of Chicago Press. Werner TK, Sherry TW. 1987. Behavioral feeding specialization in Phinaroloxias inornata, the ‘Darwin’s Finch’ of Cocos Island, Costa Rica. Proceedings of the National Academy of Sciences, United States of America 84: 5506–5510. Wood SL. 1982. The bark and ambrosia beetles of North and Central America (Coleoptera: Scolytidae), a taxonomic monograph. Great Basin Naturalist Memoirs 6: 1–1359. Wood SL. 1986a. New synonymy and new species of American bark beetles (Coleoptera, Scolytidae), part XI. Great Basin Naturalist 46: 265–273. Wood SL. 1986b. A reclassification of the genera of Scolytidae (Coleoptera). Great Basin Naturalist Memoirs 10: 1–126. Wood SL, Bright DE. 1992. A Catalog of Scolytidae and Platypodidae (Coleoptera), Part 2: Taxonomic index. Great Basin Naturalist Memoirs 13: 1–1553. Wood SL, Stevens GC, Lezama HJ. 1991. Scolytidae (Coleoptera) de Costa Rica II. Clave para la subfamilia Scolytinae, tribus: Scolytini, Ctenophorini, Micracini, Ipini, Dryocoetini, Xyleborini y Cryphalini. Revista de Biología Tropy 39: 279–306. Woodley NE. 1991. Stratiomyidae of Cocos Island, Costa Rica (Diptera). Proceedings of the Entomological Society of Washington 93: 457–462. World Conservation Monitoring Centre. 1997. World Heritage Site: Isla del Coco marine and terrestrial conservation area. http://www.wcmc.org.uk/protected_areas/data/wh/ coco.html. (Last updated: 5 January 2001). Wrensch DL, Ebbert MA, eds. 1993. Evolution and diversity of sex ratio in insects and mites. New York: Chapman & Hall. © 2006 The Linnean Society of London, Biological Journal of the Linnean Society, 2006, 89, 729–743