melipona (melikerria) (hymenoptera: apidae:...

1

The Panama microplate, island studies and relictual species of Melipona

(Melikerria) (Hymenoptera: Apidae: Meliponini)

David W. Roubik

Smithsonian Tropical Research Institute

Ancon, Balboa, Republic of Panama

[email protected]

João Maria Franco de Camargo

University of São Paulo

Ribeirão Preto, SP, Brazil

(deceased)

2

Abstract. The endemic stingless honey-making bee Melipona (Melikerria) insularis

sp. nov. on Coiba and Rancheria Islands in Pacific Panama is described, along with the

proposed sister species, M. ambigua sp. nov. from NE Colombia. The Coiba Island

group and Panama mainland were surveyed, yielding one meliponine endemic (M.

insularis sp. nov.) and six meliponine genera and species. The poor Coiba fauna of

amphibians and birds corresponds to the poor social bee fauna and suggests habitat

barriers generally precluded re-colonization from the mainland during glacial periods.

Many animals became extinct, yet some remain as relicts. M. insularis sp. nov. was

isolated on accreted terranes of Coiba rain forest in the Panama microplate. Morphology

suggests M. insularis sp. nov. is not a direct descendant of the San Blas-E. Panama

endemic Melikerria, M. triplaridis. A phylogenetic hypothesis corroborates disjunct

distributions. Rain forest endemics such as Peltogyne purpurea (Fabaceae) and

Ptilotrigona occidentalis (Apidae, Meliponini) also occur as relictual, disjunct

populations in Central and South America. These may have been isolated before

accelerated biotic exchange began 2.4 Ma. Our work supports the geological findings of

both a volcanic arc and the San Blas massif providing a substantial bridge for Melikerria

from Colombia and Panama in Eocene to Miocene times. We suggest there have been

taxon cycles permitting re-colonization during glaciations, whereby colonies of M.

insularis sp. nov. were able to re-colonize Rancheria, a 250 ha island, 2 km from Coiba.

However, rafting colonies nesting in trees, carried on vegetation mats, may have

produced founding populations of Melipona in Central America and on oceanic islands

such as Coiba.

3

Introduction

Because their species have strong dispersal limitations (Rasmussen & Cameron, 2010),

meliponine bees are of particular interest for the study of relationships between islands

and continents. Coiba Island is largely volcanic, Cretaceous, and after an origin in the

Galapagos hotspot was pushed by the Cocos plate. The Eocene Coiba Island emerged far

from present day South America but now lies 23 km from Panamanian mainland

(Castroviejo, 1997). Here we describe two new species of Neotropical honey making

bee, genus Melipona, one from Coiba Island and nearby Rancheria Island, and one from

northern Colombia. Our studies on the Melipona compressipes (Fabricius) group

(subgenus Melikerria) provide a comparative element of the biogeographically complex

Caribbean region (Liebherr, 1988; Ricklefs & Bermingham, 2001; Morrone, 2006),

including one of the oldest extant Neotropical stingless bee genera (Rasmussen &

Cameron, 2010), also used by Maya people since antiquity for beekeeping and honey

(Villanueva et al., 2005). Workers of subgenus Melikerria are the size of the western

honey bee Apis mellifera but maintain much smaller colonies; all but one species build

nests only in tree hollows. The distribution of Melikerria (26˚S to 22˚N) (Camargo &

Pedro, 2007; Ramírez et al., 2010) including both the Amazon rim and basin also

implies antiquity that should be useful in reconstructions of biotic evolution and

diversification in tropical America.

When organizing the taxonomy of Melipona, Moure (1992) created subgenus

Melikerria based on two apomorphies, following Schwarz (1932: 265): “The inner half

4

of the apex of the mandibles subdivided by strong emargination into two distinct

contiguous teeth [the median tooth with sinuosity], and “The antero-lateral angles of the

mesonotum with a rust-red patch of hair …”. Besides these features, and corroborating

the monophyly of the subgenus, the males possess scythe-shaped genitalic gonostyles

(JMC, pers. obs.). No new Melikerria has been described since Schwarz (1932) and all

eight species of the subgenus known until now (Camargo & Pedro, 2007) are

continental. Their distribution patterns are well defined and based on substantial field

work: Melipona beecheii Bennett, 1831 from Mexico to Costa Rica (records for Cuba

and Jamaica derive, probably, from introductions by indigenous peoples from

continental areas [see Michener, 1982: 37] and a Panama record is likely erroneous [see

Schwarz, 1934: 8 and Michener, 1954: 164], DWR, pers. obs.); M. triplaridis Cockerell,

1925 in eastern Panama; M. salti Schwarz, 1932, northern Colombia – Magdalena; M.

compressipes (Fabricius, 1804), from Venezuela, Guyana, Suriname, French Guiana,

Brazil – north of the Amazon and Negro Rivers, and Amazonian Colombia. Such a

distribution is classified ‘northern Amazon’ (‘NAm’, of stingless bee distribution,

Camargo & Pedro, 2003); M. interrupta Latreille, 1811, from Guyana, Suriname, French

Guiana and Brazil, north of the Amazon River, east of the Branco River to the state of

Amapá (overlapping with the eastern distribution of M. compressipes); M. grandis

Guérin, 1834, southwest Amazonia (the ‘SWAm’ meliponine biogeographic component,

Camargo & Pedro, 2003); M. fasciculata Smith, 1854, eastern Amazonia – Pará,

Maranhão, Piauí, Tocantins, Mato Grosso; M. quinquefasciata Lepeletier, 1836, from

northeastern Brazil to Argentina – Missiones, Paraguay and Bolivia (including an

undescribed species, see Camargo & Pedro, 2007: 343). Of interest is the absence of this

5

subgenus along the Pacific mainland forest, from Ecuador to the northern Chocó,

Colombia. We discuss the evidence, below, that this lineage colonized Coiba Island in

the Miocene, went extinct on the mainland, but founded an ancestral population of the

only Melikerria in Central America and Mexico, the ‘Maya honey bee’, M. beecheii.

Materials and methods

Coiba Island, in the Gulf of Chiriquí, is 23 km minimum distance from the Pacific coast

of Panama, near Punta Gorda and the town of Pixvae (DWR, pers. obs.). Its area is

50,314 ha, between 7º10΄4″ - 7º53΄27″N and 81º32΄25″ –81º56΄15″W, with a maximum

elevation 416 m. Annual rainfall is over 3000 mm and Coiba Island flora has major

Darién affinities (Velayos et al., 1997). Rancheria island, 21.7 km from the closest point

to the mainland, with an area of 250 ha and 130 m at its highest point, is the second

island where one of our two new bee species was found. Coiba is the largest island, and

a group of eight lesser islands and 30 islets surround it. It was emergent (above sea

level) during the late Eocene (Cardiel et al., 1997). Coiba Island and the area we refer to

as the ‘Panama microplate’ was transported with the Cocos Plate to the subduction zone

of Panama and Costa Rica, and then incorporated in the continental border (Meschede &

Barckhausen, 2000; Lissinna et al., 2001; Trenkamp et al., 2002; Coates et al., 2004;

Kerr & Tarney, 2005). Coiba Island originated as one of the exotic terranes, like the

Azuero and Soná highlands flanking it on the mainland, and Osa and Nicoya Peninsulas

in Costa Rica. During multiple glaciations that began over 3.0 Ma in the Pliocene,

terrestrial connections to Coiba Island, lasting up to approximately 90,000 years, were

6

possible when sea levels lowered by as much as 130 m. The sea floor is presently at a

maximum 100 m depth between Coiba and the mainland, both to the north and to the

east.

Field collections were made for this study between 1979 and 2011, when the first

author surveyed Coiba Island and nearby mainland biota 12 times and spent 50 field

days there (Appendix S2). Playa Rosario, Playa Hermosa, Playa Venado, Playa Damas,

La Central, La Salina and two points on western Coiba near its highest hills were used to

assess the island forest biota, as we made trails inland to its hilly forest. Rancheria Island

was studied intensively (nine visits, 18 days), Montuosa, Jicaron, Cébaco, the Outer

Channel, and Santa Catalina Islands were surveyed for one or two days. Only the island

of Cébaco is largely degraded or secondary forest. Collections on flowering plants, or

using bee attractants (honey-water sprayed on vegetation; volatiles effective for both

Meliponini and Euglossini, i.e. vanillin, methyl salicylate, skatole) were made during the

morning and early afternoon. On the mainland, north and east of Coiba Island, the

principal forest remnant resembling Coiba and Rancheria Islands was located in a 5 km2

area at Punta Gorda, also near mainland closest to Coiba and Rancheria Islands. The

peninsular Veraguas Province region has no other extensive lowland forest of closed

canopy and is degraded. Two additional forest patches were surveyed between Pixvae

and Soná. Collections in Colombia made by collaborators and beekeepers provided

additional material (DWR collection, STRI).

Taxonomic methods include the terminology and measurements of Camargo &

Pedro (2009). Abdominal terga and sterna are indicated as TII–TVIII and SII–SVIII

respectively; antennal flagellomeres are fl.1, fl.2, etc., from the base to apex of the

7

flagellum; the first abdominal tergum, incorporated in the thorax, is referred to as the

propodeum, and the ‘propodeal disc’ as the metapostnotum; legs and respective

podomeres are followed by Roman numerals according to the segment, e.g. tibia III

refers to the hind tibia. Alveolocellar distance is measured from the alveolus to the

anterior margin of the lateral ocellus of the same side; interocellar distance between the

lateral ocelli; clypeocellar distance from the apex of the clypeus to the anterior margin of

the median ocellus; length of the head from the apex of the clypeus to the vertex aligned

on the same focal distance; length of the forewing from the apex of the costal sclerite to

the tip of the wing, except when indicated. Measurements were taken with an ocular

reticule and are given in millimeters. Photos were taken with digital Pentax Camera

K10D coupled onto with stereomicroscopy LEICA MZ APO.

We used limited molecular data provided by the Barcode of Life study

(University of Guelph, Ontario, Canada), because the method supplies information

concerning cryptic species, relative species ages and, because we lacked male specimens

of the new species, an added means of testing species hypotheses. The mtDNA gene

fragment CO1-5P was analyzed from specimens collected by the first author. These

were combined in a distance matrix with sequences deposited in GenBank (Ramírez et

al., 2010), after alignment using the ‘Clustal X algorithm’, for CO1 in Melikerria.

Sequence data were used to produce values for a similarity matrix, using BioEdit

(http://www.mbio.ncsu.edu/BioEdit/bioedit.html).

Results

8

Melipona (Melikerria) insularis sp. nov.

(Figs. 1a,e,i, 2b; Appendix S1)

Trigona interrupta; Cheesman, 1929: 149

Melipona compressipes (distinct form of); Camargo, Moure & Roubik, 1988: 156

Diagnosis. Pilosity on head and thorax entirely whitish, with exception of yellow-

ferruginous tufts on the anterior corners of mesoscutum (Fig. 1i); paraocular stripes

expanded below (Fig. 1a), 1.3x scape diameter; yellowish bands on distal margins of

metasoma TII-V, complete, with slight median indentation on TII and TIII and brief

interruption on TIV and TV, width greater than or equal to scape diameter (Fig. 1e),

slighter narrower and more broadly interrupted on TVI.

Holotype, worker.

Dimensions. Total length approximately 11.0 mm; forewing length, 8.32 mm (including

tegula, 9.78 mm); maximum head width, 4.22 mm; TIII width, 4.18 mm.

Integument color. Predominantly black; anterior surface of scape, yellowish; posterior

surface of scape, pedicel and posterior surface of flagellum dark ferruginous; anterior

flagellar surface ferruginous. Mandible, except apex and base, and all tarsomeres, fusco-

ferruginous; basitarsus III and distal part of tibia dark ferruginous. Wing membranes,

ferrugineous. Tegula and wing veins, dark ferruginous, like tarsomeres. Head, pronotum

and metasomal terga, with distinct yellowish markings; on face, paraocular stripe

occupying inferior 2/3 of eye, pointed above and gradually widened below –lower 1/3

ca. 1.3x wider than diameter of scape; clypeus with complete median-longitudinal band

(reaching epistomal suture, above, to distal margin of clypeus; some paratypes with band

9

faint superiorly ), little wider than half the scape diameter, with subtriangular mark on

each side, on the edges, largely separated from the median band (in one paratype barely

a narrow marginal stripe, greatly muted, touching the median band at the edge

markings); supraclypeal mark an isosceles triangle with very widened base. Labrum

with yellowish band, diffuse edges along its base. A yellowish stripe, very narrow, on

each side of pronotum. Bands on distal margin of TII-V, wide as scape diameter,

complete on TII-III, having median indentation, and briefly interrupted on TIV-V; on

TVI the band slightly narrower, reduced at edges and having slightly wider median

interruption.

Pilosity. Whitish, pale silver, with ferruginous tufts on anterior corners of mesoscutum,

ferruginous hairs on tarsomeres and internal face of basitarsi, and ventro-distal

trochanter II tuft; black hairs on metasomal apex. Metasomal terga with very narrow

fimbriae of small whitish hairs; TIII with simple hairs ca. 0.10 mm length; TIV-V with

simple and branched hairs, 0.12 and 0.15 mm length, respectively; TVI without fimbria.

TIII disc area with simple hairs, semi-erect, to 0.20 mm length; TIV-V having

decumbent small hairs, ca. 0.10 mm length; TVI-VII with erect black hairs, interspersed

with whitish plumose hairs, increasing in length and density, to 0.53 length on TVII

margin, curving inward and mostly branched. SVII with black setae, longer at sides,

with fulvous tips, narrower and denser on disc.

Integument. Finely dull-reticulate on head and thorax, slightly smoother and shiny on

gena; scutellum with space between piligerous punctation, smooth and shining.

Propodeum and metapostnotum dull-reticulate. Abdominal terga with fine, dull

10

micropuncturing, more shiny on TVI-VII. Tibia III (see Methods)very smooth and

shining, basitarsus slightly less so.

Shape and dimensions (Appendix S1). Head 1.27x wider than long and 2.56x wider

than clypeocellar distance. Compound eyes, 2.3x longer than wide, converging slightly

below. Malar area short, ca. 0.58x diameter of fl.3 (see Methods). Clypeus 0.58x shorter

than its maximum width, 0.68x clypeocellar distance. Mandible length 1.16x

clypeocellar distance. Labrum normal, convex, with slight medial longitudinal sulcus.

Scape 1.06x alveolocellar distance, approximately cylindrical and slightly wider than

third fl.3. Fl.1 1.43x longer than wide. Interocelar distance 1.03x greater than

ocelloorbital and 2.0x median ocellar diameter. Scutellum 0.48x shorter than wide, distal

margin in slightly pointed semi-circle. Tíbia III 0.82x width of head and 2.45x longer

than wide, posterio-distal edge projected tooth like. Basitarsus III 1.75x longer than

wide, posterio-distal edge slightly acute angled. Forewing 2.7x longer than wide and

1.97x longer than head width.

Male. Unknown.

Type specimens. Holotype, worker, “PANAMA: Veraguas Prov. Coiba Island, Playa

Hermosa, 1.6 Km W, 17 Dec 2007, D. Roubik”, deposited at USNM. Paratypes, 102

workers: 3 with the same collection label as holotype; 3 from “PANAMA, Veraguas

Prov. Coiba Island, 21 Oct 1979 (63 specimens); 1, ibidem, idem, 22 Oct. 1979, nos. 6,

and 3, ibidem, idem, 26 Oct. 1979, no. 124, and remaining collections, 2007-2010,

deposited at RPSP, STRI, USNM, and University of Panama, Fairchild Entomological

Collection.

Etymology. From Latin, insularis, isolated, island dwelling.

11

Geographic distribution. Coiba Island, Rancheria Island, Veraguas Province, Panamá.

Nest. Unknown.

Remarks. Melipona (Melikerria) insularis sp. nov. is distinguished clearly from M.

(Melikerria) triplaridis Cockerell, 1925 – a bee endemic to eastern Panama, from the

canal area to the Darién– by slightly smaller size (head width 4.18-4.27 mm, in M.

triplaridis 4.27-4.46 mm), by larger yellow head markings, in particular the inferior

paraocular stripe wider (Fig. 1d), by narrower and paler yellow metasomal bands (ca.

0.24 mm, in M. triplaridis 0.37-0.38 mm, Fig. 1h) and by the entirely whitish thoracic

pilosity (except anterio-lateral margins of the mesoscutum), lighter than that of M. salti

(Fig. 1j); M. triplaridis has yellowish ivory to fulvous pilosity (mesoscutal margins,

axillae and scutellum; Fig. 1l). The ‘Barcode of Life’ results give a 98.31% similarity in

the mtDNA sequence between M. triplaridis and M. insularis, based upon 657 base

pairs. Another figure, 97.983% similarity, was obtained using BioEdit

(http://www.mbio.ncsu.edu/BioEdit/bioedit.html) with the Clustal X algorithm for p-

values (see also Molecular Results, below). In addition, M. insularis sp. nov. differs

from M. ambigua sp. nov. of northern Colombia by the thoracic pilosity, which in the

latter is yellowish ivory, slightly lighter than in M. triplaridis, and by the yellowish

clypeal markings– in M. ambigua sp. nov. the median clypeal band is attached to the

side markings, making an ‘anchor’ shape; in M. insularis sp. nov. the arms of the anchor

are interrupted below. Plumose pilosity of TVI in M. insularis sp. nov. is largely

glabrous in the discal area and black simple hairs are more abundant and robust; on the

other hand M. ambigua has denser plumose and simple whitish pilosity and few longer

black hairs interspersed.

12

Evidence for a taxon cycle is given by a population of M. insularis on Rancheria

Island. During glaciations colonies of the large bee were likely able to colonize this 250

ha island, 2 km from Coiba (see Discussion), and its superior flight and foraging ranges

may confer survival ability.

Melipona (Melikerria) ambigua sp. nov.

(Figs. 1b,f,k, 2b; Appendix S1)

Diagnosis. Pilosity, on head and thorax, except ferruginous tuft at anterior corners of

mesoscutum, entirely yellowish ivory, slightly more yellow at mesoscutal margins,

axillae and scutellum (Fig. 1k). Paraocular bands expanded slightly below, ca. 1.1-1.2x

scape diameter; clypeal marking anchor-shaped (Fig. 1b). Yellow bands on distal margin

of TII-VI, complete, with slight median constriction, more evident on TIV-VI, as wide

as greatest diameter of scape (Fig. 1f); TVI with complete band, as anterior segments.

Holotype, worker.

Dimensions. Total length, approximately 11.0 mm; forewing length, 8.70 mm

(including tegula, 10.2 mm); maximum head width, 4.41 mm; TIII, width 4.56 mm.

Integument color. Predominantly black, like M. insularis sp. nov., with the same

variation, only legs with lighter coloration. Markings yellowish on head, pronotum and

metasomal terga, pronounced; paraocular stripe occupying inferior 3/7 of compound

eye, pointed above and gradually widening below – at the inferior extreme ca. 1.2x

flagellar diameter; clypeus having anchor shaped marking – the median longitudinal

stripe little wider than scape diameter, above not reaching the epistomal suture, below

13

touching lateral markings with narrowed stripes; supraclypeal marking an isosceles

triangle, broad at base. Labrum with wide yellow band at base. Pronotal markings

extensive, almost united at midpoint. Yellowish bands on distal margin of TII-VI, as

wide as scape diameter, like those of M. insularis sp. nov., band of TVI more complete.

Pilosity. Ivory yellow, except ferruginous tufts on anterior edges of mesoscutum,

ferruginous setae on tarsomeres and internal surface of all basitarsi, a tuft on the ventro-

distal trochanter II, and the black setae of metasomal apex, like M. insularis sp. nov.

Mesoscutal margins and scutellum with slightly darker yellowish pilosity. Tergite III

with fine, simple pilosity, semi-erect, longest ca. 0.20 mm, TIV-V with small decumbent

simple hairs, reaching 0.10 mm length,; TVI with more erect hairs, longer and

interspersed with plumose hairs, on the lateral margins, and some black setae; TVII

clothed with long black branched hairs (to 0.53 mm), interspersed with abundant

plumose ivory yellow hairs. Fimbriae on TIII-V borders exactly like M. insularis sp.

nov. Sternite VII with black hairs.

Integument. Like M. insularis sp. nov.

Shape and porportions (Appendix S1). Head ca. 1.27x wider than long and 2.70x

wider than clypeocellar distance. Compound eyes 2.41x longer than wide and slightly

convergent below. Malar area 0.39x diameter of fl.3. Clypeus length 0.63x of its

maximum width and 0.77x clypeocellar distance. Mandible 1.16x clypeocellar distance.

Labrum normal, convex, with soft median sulcus. Scape 1.12x clypeocellar distance,

approximately cylindrical and as wide as fl.3 diameter. Fl.1 1.6x longer than wide.

Interocellar distance 1.24x larger than ocelloorbital distance and 2.44x diameter of

median ocellus. Scutellum 0.53x longer than wide, distal margin a slightly pointed semi-

14

circle. Tíbia III 0.77x head width and 2.30x longer than wide, posterio-distal edge

slightly pointed tooth like. Basitarsus III ca. 1.60x longer than wide, posterio-distal edge

slightly closed. Forewing 2.82x longer than wide and 1.97x longer than head width.

Male. Unknown.

Type specimens. Holotype, worker, from “COLOMBIA, Bolivar Dpto. Cartagena. 15

km SW, September. 1980 – nest, D. Roubik”, at USNM. 24 paratypes, workers: 17 with

the same label as the holotype, at RPSP, STRI, USNM, National University of Colombia

at Bogota (LABUN), 7 additional workers, from “COLOMBIA, Bolivar Cart. Jardin

Bot. Piñeres, XII-17-79, 150 m, G. Parra coll.”.

Etymology. From Latin, ambiguus, uncertain, doubtful.

Geographic distribution. Known only from type locality.

Nest. Tree cavity, details unknown.

Remarks. Differing from M. (Melikerria) salti Schwarz, 1932, an endemic of the Santa

Marta region, Magdalena, Colombia, primarily in the tergal abdominal bands. In M.

salti, the band of TII is expanded at the lateral margins and slightly constricted in the

middle, bands on TIII-IV are more interrupted medially, and are absent, or present as

lateral markings, on TVI-VII (Fig. 1g); in M. ambigua sp. nov. bands are complete (see

Diagnosis). For comparisons with other species see “Remarks” for M. insularis sp. nov.

Molecular results

Seven species of Melikerria with CO1 sequences were analyzed, those accessioned

from GenBank (Ramírez et al., 2010) were of 491 bp, and those from Barcode of Life

15

(DWR collection, STRI) were 657 bp. Two or three individuals were used for some

species, and their similarities to the others were averaged. A mean similarity of 96.5%

(SD 2.18%) exists among the seven taxa (Table 1). The Mexican and Central American

species M. beecheii was substantially less similar to its relatives. If M. beecheii is

excluded, average similarity among the other bees is 97.81%, SD 0.63%. Even

considering this lower SD, the remaining six species show similar degrees of

divergence.

Discussion

Morphological and ecological perspectives. Recognition of new Melikerria, one

endemic to Coiba and Rancheria islands, referred to in earlier literature by Cheesman

(1929: 149, as T. interrupta) and Camargo et al. (1988) as “a distinct form of

compressipes”, and the other endemic Melikerria of northern Colombia, brings new

elements to the evolutionary history and biogeography of Melipona, and lends credence

to previous assertions on the biogeographic history and origin of Panamanian biota (Fig.

2a). We rejected the hypothesis of post-Pliocene dispersal by M. insularis sp. nov. to

Coiba and Rancheria islands in part because we did not find it in mainland forests, and

in part because the other island meliponines are widespread (Appendix S2), which we

interpret to signify recent, Holocene colonization. Melipona visits many common forest

edge flowers such as Cassia, Mimosa, Solanum and melastomes, and M. insularis sp.

nov. actively seeks and collects vanillin crystals (DWR, unpublished data) and recruits

well to honey water baits. This species, of large size, is particularly likely to be detected.

16

Its discovery on Rancheria Island —a single worker vibrating a flower of Cassia for

pollen— was surprising. There has been no mainland population to colonize this small

island, and it must periodically go extinct there (Wilson, 1961), which leads us to

propose a taxon cycle; glaciation allowed colonies of M. insularis sp. nov. to re-colonize

Rancheria from Coiba.

Colonies of Meliponini must be founded by groups of workers flying to and from

nest sites, and considerable time is required to prepare the new nesting site, prior to

colony reproduction, including nest occupation by a mated queen. Pollen and nectar are

harvested from flowers and stored in the incipient nest, and honey transferred by the

mother colony. Mother and daughter nests are united, sometimes over months or even

years (Michener, 1974: 171; Roubik, 2006; Camargo, 2008). As a result, even small

rivers and certainly oceanic gaps prevent meliponine colony dispersal. There is an

intriguing specialization common both to Apis and meliponines, in that workers direct

the queen to her new nesting site (see Seeley, 2010). Tree-nesting species may

nonetheless disperse in floating trees or in trees on vegetation rafts (and honey bee

swarms disperse many kilometers over open ocean, Roubik & Boreham, 1990). For

marine islands, however, we favor the hypothesis of dispersal via terrestrial connections

(Camargo et al., 1988; Roubik et al., 1997). How, then, did Melikerria arrive on Coiba

Island; how and when did its ancestors spread among southern South America, Central

America and Mexico? A Melipona on Coiba Island suggests that at some period it has

been part of continuous mainland forest community, of sufficient duration to be used by

founding swarms of meliponine bees, or that bee colonies numbering only a few to

hundreds (see Nogueira-Neto, 2002; Lynch & Lande, 1997; Kerr & Vencovski, 1982;

17

Alves et al., 2010) were able to raft there on vegetation mats. Moreover, Melipona is

preadapted to isolation in natural habitat by their tendency to replace the colony’s queen

and curb diploid male production (Nogueira-Neto, 2002; Alves et al., 2010). Despite

substantial investigation (Appendix S2), M. variegatipes was the only extant island

endemic known for the Neotropics. Found on Guadeloupe and Dominica, it had no

options for arrival there except from South America during glaciations (Camargo et al.,

1988; Rasmussen & Cameron, 2010). We suggest that large body size and flight range,

and exceptional communication ability (Michener, 1974; Roubik, 1989; Nieh & Roubik,

1995) increases survival on forested islands, forestalling extinction that we believe has

been frequent among meliponines.

Based on morphology, the proposed general area cladogram for Neotropical

meliponine bees (Camargo & Moure, 1996; Camargo, 1996; Camargo & Pedro, 2003)

indicates three clades, and elucidates Neotropical biogeography in a time frame relevant

to the present discussion (see also Amorim & Pires, 1996; Sigrist & Carvalho, 2009).

One meliponine clade includes taxa that range from Pacific Ecuador and the Chocó to

Panama, then to Mexico (called the ‘Chocó-Ca’ biogeographic component). Two

biogeographically proximal clades are of the southwestern Amazon (SWAm) and

northern Colombia and Panama to Mexico, the NAm biogeographic component

(Camargo & Pedro, 2003). Camargo (1996) has evaluated Amazonian meliponine

biogeography and finds congruence in three historically distinct areas, one that produced

the southeastern South America and Atlantic Brazilian taxa, and two others, which were

connected, in two distinct periods (Amorim & Pires, 1996; Camargo, 2008), to Central

America. The logical implication is that Panama was the conduit of two major

18

continental stingless bee dispersal events. Furthermore, Melikerria lacked a

corresponding component of the Chocó-Ca. That component is quite possibly

represented by the relictual Coiba species that we describe, now extinct on the mainland,

and descendants from northward dispersal, now called M. beecheii.

The proposed sister species of M. insularis sp. nov., M. ambigua sp. nov. allows us

to make some inferences. The CO1 data indicated an average difference of 2.6% (SD

0.6%) among Melikerria. Variation within M. beecheii has been reported at 1.2%

sequence divergence, but is analyzed from only one gene fragment (Quezada-Euán et

al., 2007). The mtDNA similarities among taxa lead to no particular argument, save one:

M. beecheii is a divergent taxon with a longer biogeographic divergence than found

between its consubgeners, and this matches the conclusion from much more extensive

data (Rasmussen & Cameron, 2010; Ramírez et al., 2010a). Its within-species

differentiation may be considered relatively high (see Magnacca et al., 2010). Detailed

morphological examination of Melikerria provides an enhanced perspective. No

comprehensive phylogeny of Melipona has been presented, although roughly half the

taxa were included in a molecular phylogeny (Ramírez et al., 2010a). Morphological

attributes suggest M. insularis sp. nov., M. ambigua sp. nov. and M. salti constitute a

subgroup or monophyletic branch. The three species share: 1) paraocular markings

gradually widened below (Figs. 1a,b,c); 2) fimbria on TIII very straight and of only

simple hairs (Figs. 1e,f,g), and 3) fimbria on TIV-V straight and of simple and branched

hairs. Melipona insularis sp. nov. and M. ambigua sp. nov. have wide and complete

metasomal bands, like M. triplaridis and unlike M. salti. The remaining species of

Melikerria have different combinations of these and other characters. We may thus

19

propose a phylogenetic hypothesis (((M. insularis sp. nov. – M. ambigua sp. nov.) M.

salti) M. triplaridis). In this context, M. triplaridis, a species of eastern continental

Panama, is basal, perhaps sister to the other three species, but not a direct relative of M.

insularis sp. nov. Ancestral populations of Melikerria are implied as candidate direct

ancestors of M. beecheii.

A Bayesian molecular clock estimate for Melikerria dates the group as mid-

Miocene, and the Mexican and Central American endemic, M. beecheii, possibly

diverged from congeners by late Miocene (Ramírez et al., 2010a). However, the branch

support for estimated divergence time, based on four genes, is weak. The distinctive

Mexican and Central America M. beecheii is apparently a sister to other extant

Melikerria, thus its ancestral population was already in place well before the Panama

isthmus arose. Because Melipona, determined by analysis of nine nuclear genes

(Rasmussen & Cameron, 2010), is sister to almost all extant Neotropical stingless bee

genera, an Eocene dispersal to Mexico from South America cannot be discounted, yet

Melikerria is unlikely to be the oldest lineage among several Melipona endemic to

Mexico (Ayala, 1999). Nonetheless, ancestral Melikerria may have been the only

meliponine both present and able to disperse to and maintain a population on Coiba

Island in the Miocene, when Coiba was more isolated from the mainland than at present.

Geological events in the isthmian region. Interesting details on terranes comprising

southeastern Costa Rica, Panama and the Colombian Chocó (Tremkamp et al., 2002;

Coates et al., 2004; Kerr & Tarney, 2005, Montes et al., in press, see Fig. 2a) provide a

possible explanation for disjunct distributions of closely related taxa or single species. In

the late Eocene and Miocene, Central America was a peninsula (not a series of islands)

20

that extended 300 km farther east than the Canal area of Panama. South America was

separated from Central America by approximately 150 km of deep ocean (Montes et al.,

in press). In addition, combined exotic terranes from an ancient volcanic arc (a

microplate), which included the Nicoya and Osa peninsulas in Costa Rica, the Soná

highlands and Azuero peninsula in the Gulf of Chiriquí, Coiba Island, and perhaps

extending south to the Sapo and Baudó Massif to Gorgona Island (Gansser et al., 1979),

could have been emergent from the mid Miocene (12.8-9.5 Ma) until the late Miocene

(7-6.3 Ma) (Duque-Caro, 1990). That block of exotic terranes and land now subducted

may have been available for colonization during a limited period in the mid to late

Miocene, and also separated from newer terranes at the Caribbean margin. Thus we

believe Melikerria may have crossed the South America-Central America gap in

Miocene times, occupying both the San Blas massif and the Pacific side of the forming

isthmus. An effective interruption in contact between Panama and South America then

lasted until the Plio-Pleistocene at 3.7-3.1 Ma.

Comparative biogeography. Chronostragraphic studies of sufficient detail to allow

correctly dated fossil evidence of ancient dispersal events between North and South

America are few, but it is now clear that large land animals dispersed from Central

America to Peru before closing of the isthmus of Panama (Campbell et al., 2010; Kirby

et al., 2008; Retellack & Kirby, 2006). Cody et al., (2010) point out that no geological

evidence proves that a land connection dated earlier than approximately 3 Ma. However,

recent advances in geological sediment characterization indicate a relatively narrow gap

between continents during as early as Eocene times (Montes et al., in press).

Furthermore, beginning about 2.5 Ma, glaciation events were relatively longer and sea

21

levels were reduced by over 100 m. At 2.4 Ma there is a telltale rush of species crossing

the isthmus (Coates, 1997; Webb, 1997). However, according to molecular divergence

time estimates, the earliest freshwater fish dispersal occurred 7 Ma (Bermingham &

Martin, 1998). Cody et al. (2010) summarize dozens of molecular dating and phylogeny

studies, finding that all mammals crossed the isthmian area ≤ 10 Ma, and plants ≤ 50

Ma, but only 24% of those plants and 39% of the animals did so ≤ 3 Ma. Birds clearly

dispersed more often north than south (Weir et al., 2009), and their forest lineages

increased dramatically in dispersal rate after the Panamanian isthmus was complete.

Thus, there has been a steady flow of biota across the isthmian region since the Eocene.

That largely oceanic region was a very leaky barrier.

Xeric conditions have been related to low species diversity, local extinction, and

failure by amphibians to recolonize Coiba Island, where eight bird families went extinct

(Castroviejo, 1997). Only four plants species may be endemic on Coiba Island (Velayos,

et al., 1997) but the putative endemics are not trees and perhaps less easily documented,

thus potentially extant elsewhere. In contrast, a valued timber species endemic to rain

forests is found on Coiba Island, Peltogyne purpurea “purpleheart” (Fig. 2d), and its

distributional pattern is fragmented (Silva, 1976; Salinas & Cárdenas, 2006), in areas

that correspond strikingly well to those of the species we describe of Melikerria.

Disjunct distributions among Neotropical meliponines occur both at the species level

(Ptilotrigona occidentalis, Meliponini, Fig. 2c, Appendix S2) and at genus level.

Proplebeia (as an amber fossil of Miocene age) is found in Chiapas, Mexico and

Dominica/Hispaniola (Camargo et al., 2000). Meliwillea is an apparent relict, confined

to cloud forest of Costa Rica and Panama (Roubik et al., 1997) and putatively sister to a

22

widespread genus, Scaptotrigona, with endemic species throughout its range. Either

extinction or more complex historical biogeography (Ricklefs & Bermingham, 2001;

Camargo & Pedro, 2003; Morrone, 2006) may explain such patterns. It is likely that wet

conditions prevailed on the higher elevations of Coiba Island since the Pliocene, while

xeric or non-forest conditions prevented colonization between island wet forest biota

and continental lowlands.

Conclusions. A mixture of northern and southern biogeographic elements

composes the flora of Panama, which on the mainland is primarily N. American

(Graham, 1990; MacFadden, 2009), northwest of the limit of endemic Darién Melipona

triplaridis at the lowland division that contains the Panama Canal. In contrast, the Coiba

Island flora is predominantly like the Darién and South America, rather than like that of

Caribbean wet forests (Velayos et al., 1997), and contains our new Panamanian

Melikerria.

While acknowledging a Laurasian presence of stingless bees in North America,

Mexico and Greater Antilles, an older Gondwanan origin and vicariant dispersal with

the rifting of South America from Africa places meliponines in South America in the

mid Cretaceous (Camargo et al., 1988; Ramirez et al., 2010a; Rasmussen & Cameron,

2010). The island-endemic, M. (Melipona) variegatipes of Lesser Antilles, arrived from

South America across land (Camargo et al., 1988), but general dispersal of meliponines

farther northward was precluded until the Neogene Panama connection. Dispersal

avenues in the Caribbean, to Central America or Mexico, may exist for non-colonial

female euglossine and other bees (Ramírez et al., 2010b) but are unlikely for colony-

dispersing meliponines. We see a pattern for multiple meliponine lineages, and their

23

vicariance, as the result of an early Neogene filter bridge between North and South

America.

A final note on the conservation of a large, endemic stingless bee on the largest

island on the Pacific coast of Central American is appropriate (see Kier et al., 2009).

The Africanized honey bee colonized western Panama in 1985 (Roubik, 2002) and its

foragers are now common on flowers, many visited by M. insularis on Coiba, and by the

honey bees on most of the other islands (DWR, pers. obs.). The two comparably large

honey-making bees, from the New and Old World tropics, have coexisted on Coiba for

over 25 years. At present, Coiba is protected wildland and no beekeeping ventures are

allowed. An escalation in the honey bee colonization process or diminution of the flora

(Roubik & Villanueva, 2009) could lead to untimely demise of an ancient bee species.

Acknowledgments

We thank Drs. Silvia Menezes Pedro and Claus Rasmussen for their help, anonymous

reivewers for comments, marine and field assistants in Panama for aid, and the

Panamanian authorities of ANAM and Futuroforestal (Cébaco Island) for support.

References

Alves, D. A., Imperatriz-Fonseca, V. L., Francoy, T. M., Santos-Filho, P. S., Billen, J.,

& Wenseleers, T. (2010). Successful maintenance of a stingless bee population

despite a severe genetic bottleneck. Conservation Genetics doi: 10.1007/s10592-

010-0171-z.

24

Amorim, D. S. & Pires, M. R. S. (1996) Neotropical biogeography and a method for

maximum biodiversity estimation. Biodiversity in Brazil, a first approach. (ed. by

C.E.M. Bicudo & N.A. Menezes, pp. 183–219. CNPq, São Paulo.

Ayala-B., R. (1999). Revisión de las abejas sin aguijón de México (Hymenoptera:

Apidae: Meliponini). Folia Entomológica Mexicana, 106, 1–126.

Bermingham, E. & Martin, A. P. (1998) Comparative mtDNA phylogeography of

neotropical freshwater fishes: testing shared history to infer the evolutionary

landscape of lower Central America. Molecular Ecology, 7, 499–517.

Camargo, J. M. F. (1996) Meliponini Neotropicais (Apinae, Apidae, Hymenoptera):

biogeografia histórica. Anais do Encontro sobre Abelhas de Ribeirão Preto, SP,

Ribeirão Preto, USP, 2, 107–121.

Camargo, J. M. F. (2008) Biogeografia histórica dos Meliponini (Hymenoptera, Apidae,

Apinae) da região Neotropical. Abejas sin Aguijón y Valorización Sensorial de su

Miel (ed. by P. Vit), pp. 13–26. APIBA-DIGECEX, Universidad de Los Andes,

Mérida.

Camargo, J. M. F. & Moure, J. S. (1996) Meliponini Neotropicais: o gênero Geotrigona

Moure, 1943 (Apinae, Apidae, Hymenoptera), com especial referência à filogenia e

biogeografia. Arquivos de Zoologia, 33, 95–161.

Camargo, J. M. F. & Pedro, S. R. M. (2003) Meliponini Neotropicais: o gênero

Partamona Schwarz, 1939 (Hymenoptera, Apidae, Apinae) – bionomia e

biogeografia. Revista Brasileira de Entomologia, 47, 311–372.

25

Camargo, J. M. F. & Pedro, S. R. M. (2004) Meliponini Neotropicais: o gênero

Ptilotrigona (Hymenoptera, Apidae, Apinae). Revista Brasileira de Entomologia,

48, 353–377.

Camargo, J. M. F. & Pedro, S. R. M. (2007) Meliponini. Catalogue of Bees

(Hymenoptera, Apoidea) in the Neotropical Region. (ed. by J. S. Moure, D. Urban

& G. A. R. Melo), pp. 272–578. Sociedade Brasileira de Entomologia, Curitiba.

Camargo, J. M. F. & Pedro, S. R. M. (2009) Neotropical Meliponini: the genus

Celetrigona Moure (Hymenoptera: Apidae, Apinae). Zootaxa, 2155: 37-54.

Camargo, J. M. F., Grimaldi, D. A. & Pedro, S. R. M. (2000) The extinct fauna of

stingless bees (Hymenoptera: Apidae: Meliponini) in Dominican Amber: two new

species and redescription of the male of Proplebeia dominicana (Wille and

Chandler). American Museum Novitates, 3293, 24 pp.

Camargo, J. M. F.; J. S. Moure & Roubik, D. W. (1988) Melipona yucatanica new

species (Hymenoptera: Apidae: Meliponinae); stingless bee dispersal across the

Caribbean Arc and post-Eocene vicariance. The Pan-Pacific Entomologist, 64, 147–

157.

Campbell, K. E., Prothero, D. R., Romero-Pittman, L., Hertel, F. & Rivera, N. (2010)

Amazonian magnetostratigraphy: dating the first pulse of the great American faunal

interchange. Journal of South American Earth Sciences 29, 619–626.

Cardiel, J. M., Castroviejo S., & Velayos, M. (1997) El Parque Nacional de Coiba: El

Medio Fisico Flora y Fauna del Parque Nacional de Coiba (Panamá), Inventario

Preliminar (ed. by S. Castroviejo), pp. 11–30. AECI, Madrid.

26

Castroviejo, S. (1997) Introducción y consideraciones biogeográficas. Flora y Fauna del

Parque Nacional de Coiba (Panamá), Inventario Preliminar. (ed. by S.

Castroviejo), pp. 1–8. AECI, Madrid.

Cheesman, L. E. (1929) Hymenoptera collected on the “St. George” Expedition in

Central America and the W. Indies. Transactions of the Entomological Society of

London, 77, 141–154.

Coates, A. G., Collins, L. S., Aubry, M. P., & Berggren, W. A. (2004) The Geology of

the Darién, Panama, and the late Miocene-Pliocene collision of the Panama arc with

northwestern South America. Geological Society of America Bulletin

November/December, 1327–1344.

Cody, S., Richardson, J. E., Rull, V., Ellis, C. & Pennington, R. T. (2010) The Great

American Biotic Interchange revisited. Ecography 33, 326-332.

Duque-Caro, H. (1990) Neogene stratigraphy, paleoceanography and paleobiogeography

in northwest South America and evolution of the Panama seaway.

Palaeogeography, Palaeoclimatology, Palaeocology, 7, 203–234.

Gansser, A., Dietrich, V. J. & Cameron, W. E. (1979) Palaeogene komatiites from

Gorgona Island. Nature, 278, 545–546

Graham, A. (1990) New angiosperm records from the Caribbean Tertiary. American

Journal of Botany, 77, 897–910.

Kier, G., Kreft, H., Ming Lee, T., Jetz, W., Ibish, P. L, Nowicki, C., Mutke, J. &

Barthlott, W. (2009) A global assessment of endemism and species richness across

island and mainland regions. Proceedings of the National Academy of Science

(USA), doi/10.1073/pnas.08103-06106.

27

Kerr, W. E., & Vencovski, R. (1982) Melhoramento genético em abelhas. Efeito do

número de colonias sobre o melhoramento. Revista Brasileira de Genética, 5, 279–

285.

Kerr, A. C. & Tarney, J. (2005) Tectonic evolution of the Caribbean and northwestern

South America: The case for accretion of two late Cretaceous oceanic plateaus.

Geology, 33, 269–272.

Kirby, M. X., Jones, D. S. & McFadden, B. J. (2008) Lower Miocene stratigraphy along

the Panama Canal and its bearing on the Central American peninsula. PLoS ONE 3,

e2791. doi:10.1371/journal.pone.0002791.

Liebherr, K. K. editor. (1988) Zoogeography of Caribbean insects. Cornell Univ. Press,

Ithaca, New York.

Lissinna, B. K., Hoernle, P, van den Bogaard, P. & Werner, R. (2001) New evidences

for ocean Island accretion in eastern Panamá. Abstract in EU6 XI. Symposium

OS09. Geochemistry.

Lynch, M. & Lande, R. (1997) The critical effective size for a genetically secure

population. Animal Conservation 1, 70–72.

MacFadden, B. J. (2009) Three-toed browsing horse Anchitherium (Equidae) from the

Miocene of Panama. Journal of Paleontology 83, 489–492.

Magnacca, K. N. & Brown, M. J. F. (2010) Mitochondrial heteroplasmy and DNA

barcoding in Hawiian Hylaeus (Neoprosopis) bees (Hymenoptera: Colletidae).

BMC Evolutionary Biology, 10:174. Doi: 10.1186/1471-2148-10-174.

28

Meschede, M. & Barckhausen, U. (2000) Plate Tectonic Evolution of the Cocos – Nazca

Spreading Center. In: E. A. Silver; G. Kinura & T. H. Shipley (Eds.) Proc. ODP,

Sci. Results, 170, pp. 1–10. College Station, TX (Ocean Drilling Program).

Michener, C. D. (1954) Bees of Panamá. Bulletin of the American Museum of Natural

History, 104, 1–175.

Michener, C. D. (1974) The social behavior of the bees: a comparative study. Harvard

University Press, Cambridge, Massachusetts.

Michener, C. D. (1982) A new interpretation of fossil social bees from the Dominican

Republic. Sociobiology, 7, 37-45.

Montes, C., Cardona, A., MacFadden, R., Morón, S. E., Silva, C. A., Restrepo-Moreno,

S., Ramírez, D. A., Wilson, J., Farris, D., Bayona, G. A., Jaramillo, C. A., Valencia,

V., Bryan, J. & Flores, J. A. (2011). Evidence for middle Eocene and younger

Emergence in Central Panama: implications for Isthmus closure. Geological Society

of America Bulletin (in press).

Morrone, J. J. (2006) Biogeographic areas and transition zones of Latin America and the

Caribbean Islands based on panbiogeographic and cladistic analysis of the

entomofauna. Annual Review of Entomology, 51, 467–494.

Moure, J. S. (1992). Melikerria e Eomelipona, dois subgêneros novos em Melipona

Illiger, 1806. (Hymenoptera – Apidae. Anais do Encontro Brasileiro de Biologia de

Abelhas e Outros Insetos Sociais. Homenagem aos 70 Anos de Warwick Estevam

Kerr. ( ed. By C. C. Cruz Landim & J. Chaud Netto, pp. 45–49. Naturalia, special

tome, UNESP, Rio Claro, Brazil.

29

Nieh, J. C. & Roubik, D. W. (1995) A stingless bee, Melipona panamica, indicates food

location without using a scent trail. Behavioral Ecology and Sociobiology, 37, 63–

70.

Nogueira-Neto, P. (2002). Inbreeding and building up small populations of stingless

bees. Revista Brasileira de Zoologia, 19, 1181–1214.

Quezada-Euán, J. J. G., Paxton, R. J., Palmer, K. A., Itzá, W. J. M., Tay, W. T., & Oldroyd, B.

P. (2007) Morphological and molecular characters reveal differentiation in a Neotropical

social bee, Melipona beecheii (Apidae: Meliponini) Apidologie, 38, 1–12.

Ramírez, S. R., Nieh, J. C., Quental, T. B., Roubik, D. W., Imperatriz-Fonseca, V. &

Pierce, N. E. (2010a) A molecular phylogeny of the stingless bee genus Melipona

(Hymenoptera: Apidae). Molecular Phylogenetics and Evolution, 56, 519–525.

Ramírez, S. R, Roubik, D. W., Skov, C., & Pierce, N. (2010b) Phylogeny,

diversification patterns and historical biogeography of euglossine orchid bees

(Hymenoptera: Apidae). Biological Journal of the Linnean Society, 100, 552–572.

Rasmussen, C. & Cameron, S. A. (2010) Global stingless bee phylogeny support ancient

divergence, vicariance, and long distance dispersal. Biological Journal of the

Linnean Society 99, 206–232.

Retellack, G. J. & Kirby, M. X. (2007). Middle Miocene global change and

paleogeography of Panama. Palaios, 22, 667–679.

Ricklefs, R. E. & Bermingham, E. (2001) Nonequilibrium diversity dynamics of the

Lesser Antillean Avifauna. Science, 294, 1522–1524.

Roubik, D. W. (1989) Ecology and natural history of tropical bees. Cambridge

University Press, New York.

30

Roubik, D. W. (1992) Stingless bees (Apidae: Meliponinae): a guide to Panamanian and

Mesoamerican species and their nests. Insects in Panama and Mesoamerica:

Selected Studies. (ed. by D. Quintero & A. Aiello), pp. 495–524. Oxford University

Press, Oxford.

Roubik, D. W. (1993) Direct costs of forest reproduction, bee-cycling and the efficiency

of pollination modes. Journal of Bioscience, 18, 537–552.

Roubik, D. W. (2002) African honey bees augment neotropical coffee yield. Pollinating

bees: the conservation link between agriculture and nature. (ed. by P. G. Kevan &

V. L. Imperatriz-Fonseca), pp. 255–266. Ministry of Environment, Brasilia, Brazil.

Roubik, D. W. (2006) Stingless bee nesting biology. Apidologie, 37, 124–143.

Roubik, D. W., & Boreham, M. M. (1990) Learning to live with Africanized honeybees.

Interciencia, 15, 146–153.

Roubik, D. W., & Villanueva, R. (2009) Invasive Africanized honey bee impact on

native solitary bees: a pollen resource and trap nest analysis. Biological Journal of

the Linnean Society, 98, 152–160.

Roubik, D. W., Lobo, J. A., & Camargo, J. M. F. (1997) New stingless bee genus

endemic to Central American cloud forests: phylogenetic and biogeographic

implications (Hymenoptera: Apidae: Meliponini). Systematic Entomology, 22, 67–

80.

Salinas, M. R., & Cárdenas, D. L. (2006). Peltogyne purpurea Pittier, NazarenoLibro

Rojo de Plantas de Colombia, Especies maderables amenazadas, I parte (versión

preliminar. (ed. By L. D. Cárdenas & N. R. Salinas), pp. 144-146.

http:www.sinchi.org.co/herbariov/documentos/LibroRojoMaderables,pdf).

31

Schwarz, H. F. (1932) The genus Melipona: the type genus of Meliponidae or stingless

bees. Bulletin of the American Museum of Natural History 63: 231–460.

Schwarz, H. F. (1934) The Social Bees (Meliponidae) of Barro Colorado Island, Canal

Zone. American Museum Novitates, 731, 1–23.

Seeley, T. D. (2010) Honeybee democracy. Princeton University Press, Princeton.

Sigrist , M. S. & Carvalho, C. J. B. De (2009) Historical relationships among areas of

endemism in the tropical South America using Brooks Parsimony Analysis (BPA).

Biota Neotropica, 9, 1–22.

Silva, M. F. (1976) Revisão taxonômica do gênero Peltogyne Vogel (Leguminosae –

Caesalpinioidea). Acta Amazonica, 6, suppl. 61 pp.

Trenkamp, R., Kellog, J. N., Freymueller, J. T., & Mora, H. P. (2002) Wide plate margin

deformation, southern Central America and Northwestern South America, CASA

GPS observations. Journal of South American Earth Sciences, 15, 157–171.

Velayos, M., Correa, M., Galdames, C., Castroviejo, S., & Arauz, B. (1997) Primera

aproximación al catálogo de las plantas vasculares de la isla de Coiba (Panamá).

Flora y Fauna del Parque Nacional de Coiba (Panamá), Inventario Preliminar.

(ed. by S. Castroviejo), pp. 245–328. AECI, Madrid.

Villanueva-G. R., Roubik, D. W., & Colli-Ucán, W. (2005) Extinction of Melipona

beecheii and traditional beekeeping in the Yucatán Peninsula. Bee World, 86, 35–

41.

Webb, S. D. (1997) The great American faunal interchange. Central America, a natural

and cultural history (ed. by A. G. Coates), pp. 97–122. Yale University Press, New

Haven.

32

Weir, J. T., Bermingham, E., & Schluter, D. (2009) The great American biotic

interchange in birds. Proceedings of the National Academy of Science (USA), 106,

21737–21742.

Wilson, E. O. (1961) The nature of the taxon cycle in the Melanesian ant fauna.

American Naturalist, 95, 169-193.

33

Table 1. COl mitochondrial DNA similarity matrix for Melipona subgenus Melikerria,

‘*’ = Average from multiple collection localities (Melikerria): beecheii beec comp gran insul inter quin compressipes 93.83 grandis* 93.30 98.51 insularis 92.67 97.56 97.35 interrupta* 93.40 98.78 98.17 97.56 quinquefasciata 92.87 97.96 97.35 96.33 97.55 triplaridis 93.48 98.37 98.17 97.96 98.37 97.15

34

FIGURES:

Figures 1a-l, workers. Melipona (Melikerria) insularis sp. nov., holotype: a; paratype: e,

i. M. (Melikerria) ambigua sp. nov.., holotype: b, f, k. M. (Melikerria) salti, paratype: c,

g, j. M. (Melikerria) triplaridis, from Panama, Colon Prov., Sta. Rita Ridge: d, h, l.

Scales = 1 mm.

Figures 2a-e. a.Panamá microplate. Dark gray area indicates probably continuous exotic

terranes during the Middle to Upper Miocene time (12.0 to 6.3 Ma.). Dashed area

indicates remainders (geological relicts) of exotic terranes accreted on the continent

during mid Miocene. White triangles indicate the San Blás – Dárien Massif in mid

Miocene. Based on data of Duque-Caro (1990), Coates et al. (2004), Kerr & Tarney

(2005), Tremkamp et al. (2002), and others. b. Geographical distribution of Melipona

(Melikerria) insularis sp. nov., M. (Melikerria) ambigua sp. nov., and related species. c.

Geographical distribution of Ptilotrigona occidentalis Schulz, 1904 in South America

(Chocó region) and the relictual population in Osa Peninsula. From Camargo & Pedro

(2004). d. Relictual distribution pattern of Peltogyne purpurea Pittier. From Silva (1976)

and Salinas & Cárdenas (2006).

37

(Appendix S1). Measurements (mm) of holotype Melipona (Melikerria) insularis sp.

n. and M. (Melikerria) ambigua sp. nov.

Measurements

M. insularis sp. nov. M. ambigua sp. nov.

Total body length 11.00 11.00

Width of head 4.22 4.41

Length of head

(from clypeal apex to vertex) 3.31 3.46

Length of compound eye 2.62 2.75

Width of compound eye 1.14 1.14

Upper interorbital distance 2.51 2.56

Maximum interorbital distance 2.68 2.77

Lower interorbital distance 2.37 2.43

Diameter of median ocellus 0.33 0.32

Interocellar distance 0.66 0.78

Ocellorbital distance 0.64 0.63

Interalveolar distance 0.49 0.51

Alveolorbital distance 0.61 0.65

Alveolocellar distance 1.35 1.35

Alveolar diameter 0.40 0.42

Length of clypeus 1.12 1.25

Maximum width of clypeus 1.94 1.99

Intertentorial distance,

38

width of clypeus 1.14 1.16

Clypeocellar distance 1.65 1.63

Length of malar space 0.11 0.09

Length of scape 1.44 1.52

Diameter of scape 0.21 0.23

Diameter of third flagellomere 0.19 0.23

Length of pedicel + flagellus 3.06 3.23

Length of first flagellomere 0.30 0.32

Length of second flagellomere 0.27 0.32

Length of third flagellomere 0.27 0.32

Length of mandible 1.84 1.90

Length of forewing

(to apex of costal sclerite) 8.32 8.70

Length of forewing (including tegula) 9.78 10.20

Width of forewing 3.08 3.08

Length of pterostigma 0.95 0.95

Width of pterostigma 0.17 1.17

Length of marginal cell 2.76 2.66

Width of marginal cell 0.46 0.49

Length of first abscissa of M 1.51 1.69

Length of first abscissa of Cu 1.61 1.69

Length of Rs+M plus

2nd abscissa of M 1.08 0.96

39

Hamuli number 12 13

Length of mesoscutum 2.95 3.19

Width of mesoscutum 2.88 2.93

Width of scutellum 2.06 1.98

Length of scutellum 1.00 1.06

Length of tibia III 3.45 3.42

Width of tibia III 1.41 1.48

Length of basitarsus III 1.59 1.60

Width of basitarsus III 0.91 0.99

Width of tergum III 4.18 4.56

40

(Appendix S2). Our hypothesis is that current bee fauna on Coiba, the largest island off

the Pacific coast of Central America, contains ancient endemic species, now relictual,

and more recent immigrants from the Holocene.

Panama is very biogeographically complex. In the stingless bees (Meliponini)

the discussion by Camargo et al. (1988) outlines the basic condition that various groups

were already present in Central America and Mexico in the mid Miocene. Biogeogrpahic

relationships that concern glacial periods and reduces sea levels are applicable to Gorgona

Island of Pacific Colombia (Fig. 2a), the extreme end of the exotic terranes that we argue

may have provided a dry land link between Colombia and Central America in the Miocene.

Gorgona is 26 km2 in area (9 km long, 2.5 km wide). Its closest point to the continent is 30

km in distance from Ponta Reyes and ocean depth is no greater than 50 m. It has at least

nine species of Meliponini in the genera Melipona, Partamona, Plebeia, Tetragonisca,

Trigonisca, Trigona, Nogueirapis, Nannotrigona and Scaptotrigona (Cheesman, 1929,

www.uky.edu/cgi-‐bin/cgiwrap/mjfharo/spgorgona.cgi and museum specimens), including

two that make subterranean nests Trigona fulviventris Guérin, 1835 and Nogueirapis

mirandula (Cockerell, 1917).

On Coiba island, our recent studies confirm earlier reports of Cheesman (1929) with

one important exception—there is no Trigona species on the island and the earlier

report was likely a mistake, probably referring to Frieseomelitta paupera, which occurs

from Costa Rica and Panama to northern Colombia, Venezuela and Trinidad, and in the

Coiba Island group, found also at least on Rancheria, Canal de Afuera, and Santa Catalina.

DWR collections on Coiba also include Scaptotrigona sp. (unpublished, by H. F. Schwarz,

manuscript name), as on Rancheria Island, and in the Gulf of Panamá (= T. pectoralis

41

panamensis in Cheesman, o.c.), on Taboga Island and in Veraguas Province, on the

mainland, with a distribution from Pacific Panama to northern Colombia; Trigonisca sp.

(unpublished, J. S. Moure) of Coiba and Rancheria Islands, and Taboga Island, corresponds

to a continental population (Chepo and Balboa —Panama Prov., Portobelo— Colon and San

Blas Prov., DWR & JMC, pers. obs.). On Coiba Island, Nannotrigona perilampoides (Cresson,

1878) also occurs, and is found nesting in some of the former prison buildings [also on

Taboga Island, Bay of Panama, as N. testaceicornis (Lep.) in Cheesman, 1929]. In addition,

Cébaco Island (DWR pers. obs.) has many Trigona fulviventris and Partamona peckolti

Friese, 1901, and Plebeia sp., N. perilampoides, Tetragonisca angustula Latreille, 1911,

Scaptotrigona sp. nov., and very likely Trigonisca sp. nov. Trigona and Partamona, present

nowhere in the Coiba Island group, nest in the ground or exposed, or on epiphytes (Roubik,

2006). They are unable to colonize across oceanic gaps.

Notably, species richness on Coiba island, despite large size, is far less than the 20-‐47

meliponine species found in forest areas of equal size on Panamanian mainland (Roubik,

1992, 1993). The island Cébaco lies 7 km from the mainland and is separated by relatively

shallow water of 10 m depth. It has the richest island meliponine fauna of Pacific Panama,

but included only seven meliponines. Because only Geotrigona, Nogueirapis and some

Trigona nest exclusively in the ground (Roubik, 2006), we suggest extinction and dispersal

limitation have reduced higher taxa of island stingless bees that are separated by oceanic cepths

>50 m and not close to the mainland, among them Tetragonisca, Tetragona, Scaura,

Nogueirapis, Plebeia, Trigona, Partamona, Cephalotrigona, Paratrigona, Oxytrigona,

Geotrigona and Lestrimelitta (Roubik, 1992). Extinction has far outpaced re-colonization in

general. However, the first Pliocene glaciations and sea levels lowered by 100 m, which

42

coincided with a general breach of the isthmian barrier (2.4 Ma), could certainly have allowed

various taxa to re-colonize Coiba, including M. insularis sp. nov.

melipona (melikerria) (hymenoptera: apidae:...

Documents