Heredity 75 (1995) 390—397 Received 27 February 1995 Population structure and gene flow in Stomion: a species swarm of flightless beetles of the Galapagos Islands TERRIE L. FINSTON* & STEWART B. PECK Department of Biology, Carleton University, Ottawa, Ontario, Canada, K1Y5P4 Stomion is a swarm of 13 flightless tenebrionid beetle species endemic to the Galapagos Islands. Their distribution is patchy and largely restricted to the littoral and arid zones of the archipelago. Each taxon is found on one or a few geographically close islands. Thirty-five populations representing nine of the taxa were examined at eight polymorphic enzyme loci using cellulose acetate electrophoresis to measure patterns of gene flow and investigate models of dispersal in a relatively young species group exhibiting spatially isolated populations. Genetic subdivision is high, particularly among populations of taxa which inhabit more than one island, revealing restricted gene flow and confirming the high potential for reproductive isolation among subpopulations. The mean FST across taxa was 0.30. The genetic differentia- tion occurring between spatially isolated populations may explain the exuberant speciation of the genus in the Galapagos. Tests of gene flow models give support to the stepping-stone model of dispersal. Because of their lack of flight wings, interisland dispersal of Stomion probably occurred by oceanic drift as pleuston or on floating debris, with individuals colonizing nearby islands more frequently than ones at a greater distance. Keywords:allozymes,beetles, Galapagos Islands, gene flow, population genetics. Introduction of gene flow and infer levels of migration. Two models of gene flow are commonly used. The rare Theextent of isolation and gene flow among spa- alleles model of Slatkin (1985) uses the average fre- tially separated populations determines the potential quency of alleles found only in single subpopulations for genetic differentiation, reproductive isolation (private alleles) to infer the amount of gene and speciation. The resulting population differentia- exchange between those subpopulations. Wright's F- tion may result from heterogeneous selection on dis- statistics (1978) offer an additional method of junct populations or the stochastic effects of finite measuring genetic differentiation among subpopula- population size and genetic drift. Thus, an analysis tions whereby the fixation indices of subpopulations of population structure and gene flow is an essential with respect to the total population are calculated. first step towards understanding the relative effects Generally there is a predictable relationship of selection and drift in producing the observed between individual vagility and population subdivi- population differentiation. In addition, patterns of sion; sedentary organisms tend to show higher geographical allelic variation in members of a values of FST whereas more vagile organisms show species complex may provide insight into past evolu- lower values (Zera, 1981; Liebherr, 1986). Data tionary events affecting the evolution of the group. from several recent studies have provided estimates The migration of individuals may be measured of gene flow through the examination of population using both direct and indirect methods. Mark and genetic structure in beetles which have localized dis- recapture studies can be used to measure individual tributions as a result of host-plant specificity vagility and provide estimates of the number of (McCauley & Eanes, 1987) or habitat restrictions migrants between populations. Alternately, popula- (King, 1987; Crouau-Roy, 1989). The present study tion genetic structure can be used to estimate levels supplements a field mark and recapture study (Fin- ston, 1993) which suggested that individual vagility * Correspondence:Department of Zoology, University of was limited in three genera of flightless tenebrionid Guelph, Guelph, Ontario, Canada, N1G 2W1. beetles among quadrats at a single site and provides 390 1995The Genetical Society of Great Britain. POPULATION STRUCTURE AND GENE FLOW IN STOM/ON 391 a test of gene flow levels in beetles restricted to In the present study, cellulose acetate electro- island habitats. phoresis was employed to examine the population Island and island-like communities provide structure of the genus Stomion. Beetles from 35 sites models for studying the genetic effects of isolation were examined at eight variable enzyme loci with an and population subdivision, and their roles in micro- aim towards understanding patterns of gene flow evolutionary genetic change. The Galapagos Islands and models of dispersal in a young insect species have been available for colonization for approxi- swarm showing spatial subdivision. mately 3—4 million years (Hall, 1983; Hickman & Lipps, 1985), and like other isolated archipelagos they are rich in endemic faunas. In particular, three Materials and methods genera of flightless tenebrionid beetles comprise the greatest endemic beetle assemblage in the islands Electrophoresis (Peck & Kukalova-Peck, 1990). Of these, the radia- Samples of beetles were made at 35 sites on 19 tion of the endemic genus Stomion Waterhouse is different islands. Individuals were identified to believed to have occurred following colonization by species using the key of Van Dyke (1953). Eight of a single ancestral species, which has since become the 13 described taxa were available for analysis. In extinct (Van Dyke, 1953). The limited dispersal addition, one population, DARW, represented a ability and patchy distribution of this species swarm taxon which could not be placed in the existing key. of flightless tenebrionid beetles provides an ideal Large islands were sampled at several sites whereas model for characterizing population subdivision. smaller islands were generally represented by only a G.R. Waterhouse erected the genus Stomion in single collection site. The sites and the sample sizes 1845 to encompass three new species from collec- used for electrophoretic analysis are shown in Table tions made by Charles Darwin in 1835. The most 1 and site locations are shown in Fig. 1. Live beetles recent comprehensive taxonomic treatment of the from each of the 35 sites, representing nine taxa, genus (Van Dyke, 1953) recognized eight species were frozen on dry ice and shipped to the laboratory and two subspecies. Three additional subspecies in Ottawa where they were stored in a —80°Cultra- were subsequently described (Von Kaszab, 1970; freezer until further analysis. Variation at enzyme Franz, 1985). The genus is widespread in the archi- loci was analysed using cellulose acetate electro- pelago but each taxon is found on only one or a few phoresis. This technique permitted individuals to be islands which are geographically close (Van Dyke, examined for variation at multiple enzyme loci. 1953; Finston, 1993). Members of the genus are Cephalic tissue of dissected individuals, rather than found primarily in the low elevation littoral and arid thoracic or abdominal tissue, was analysed, because zones of the islands where the substrate is generally the latter often contained food particles or eggs and sand or cinder. caused interference with allozyme banding patterns. 920 910 900 Pinta 19, 20 12 1°15' 018 Genovesa 920 Marchena 00 Santiago 2 34 Santa Cruz o30-32 SantaFe 23 — 35 San Cristobal jO Fig. 1 Sampling sites for 35 populations 0 50 9,10 0.11 6-8 of Stomion in the Galapagos Islands. km Floreana Espanola Map numbers as in Table 1. The Genetical Society of Great Britain, Heredity, 75, 390—397. 392 T. L. FINSTON & S. B. PECK Table1Sitesand sample sizes (N) used in the electrophoretic examination of Stomion Island Site N Taxon Map no. Baltra BALT 44 S. linelli 1 Bartolomé BART 44 S. linelli 2 Daphne DAPH 49 S. linelli 4 Eden EDEN 15 S. linelli S Rabida RABI 12 S. linelli 21 Santa Cruz GRAN 44 S. linelli 26 SHOR 15 S. linelli 29 Seymour SEYM 13 S. linelli 34 Española BGAR 37 S. g. punctlennis 6 MANZ 50 S. g. punctipennis 7 PTSU 44 S. g. punctipennis 8 Floreana FLBB 9 S. g. galapagoensis 9 FLCU 38 S. g. galapagoensis 10 Gardner GARD 36 S. g. galapagoensis 11 Genovesa DARW 25 'Genovesa' 12 Caamaño CAAM 42 S. laeviagtum 3 Isabela ARIM 44 S. laevigatum 13 TAGU 43 S. laevigatum 16 Santa Cruz BTOR 44 S. laevigatum 25 Santiago SGPE 44 S. laevigatum 33 Isabela WVIL 16 S. longulum 17 AZUL 44 S. longulum 14 NVIL 44 S. longulum 15 Santa Cruz BARR 44 S. longulum 24 HALF 4 S. longulum 27 PENS 44 S. longulum 28 Marchena MARC 38 S. rugosum 18 Pinta FORE 18 S. rugosum 19 IBET 36 S. rugosum 20 San Cristóbal AERO 30 S. helopoides 22 LIDO 33 S. helopoides 23 Santa Fe SFEB 44 S. obesum 30 SFEP 9 S. obesum 31 SFTR 16 S. obesum 32 Tortuga TORT 30 S. obesum 35 Map numbers refer to Fig. 1. One reference standard (an individual of S. linelli 1989; Hebert & Beaton, 1993) to identify informa- Blair from Isla Daphne) was included in each row of tive loci. Eight variable loci were found in the pilot 12 individuals for comparison of electromorph study: hexokinase (Hk-2), mannose-6-phosphate iso- mobilities. In addition, on some gels, four indivi- merase (Mpi), phosphoglucomutase (Pgrn-1), pepti- duals from a site were carried over to the next set of dase-A (Pep, utilizing phe-pro), glutamate gels to provide a second source of comparison of oxaloacetate transaminase, supernatant and mito- electromorphs. chondrial forms (Got-s, Got-rn), 6-phosphogluconate Electrophoresis was carried out using standard dehydrogenase (6Pgdh) and phosphoglucose isomer- methods (Murphy, et a!., 1990; Hebert & Beaton, ase (Pgi).Wherepossible, 44 individuals from each 1993). Two buffer systems were employed: Tris-gly- site were analysed for these eight loci. The protein cine (pH 8.5) was used for all enzyme systems products of these presumptive gene loci were visual- except 6Pgdh, for which Tris-citrate (pH 7.3) was ized using staining protocols outlined in Hebert & employed. Beaton (1993) and Murphy et al. (1990). The result- Initially, 174 animals from six different sites were ing bands were scored with respect to their relative analysed for 20 commonly resolvable loci (Hsiao, mobilities: the slowest migrating electromorph was The Genetical Society of Great Britain, Heredity, 75, 390—397.