Molecular Ecology (1998) 7, 519–531 Speciation and phylogeography of Hawaiian terrestrial arthropods G. K. RODERICK and R. G. GILLESPIE Center for Conservation Research and Training, 3050 Maile Way, Gilmore 409, University of Hawai’i, Honolulu, Hawaii 96822, USA Abstract The Hawaiian archipelago is arguably the world’s finest natural laboratory for the study of evolution and patterns of speciation. Arthropods comprise over 75% of the endemic biota of the Hawaiian Islands and a large proportion belongs to species radiations. We classify patterns of speciation within Hawaiian arthropod lineages into three categories: (i) single representatives of a lineage throughout the islands; (ii) species radiations with either (a) single endemic species on different volcanoes or islands, or (b) multiple species on each volcano or island; and (iii) single widespread species within a radiation of species that exhibits local endemism. A common pattern of phylogeography is that of repeated colonization of new island groups, such that lineages progress down the island chain, with the most ancestral groups (populations or species) on the oldest islands. While great dispersal ability and its subsequent loss are features of many of these taxa, there are a number of mechanisms that underlie diversification. These mechanisms may be genetic, including repeated founder events, hybridization, and sexual selection, or ecological, including shifts in habitat and/or host affiliation. The majority of studies reviewed suggest that natural selection is a primary force of change during the initial diversification of taxa. Keywords: biodiversity, conservation, island biology Introduction morphology, ecology and behaviour. These radiations are associated with high frequencies of endemism: Remarkable suites of endemic and often rare species are greater than 81% in birds and an estimated 99% in terres- characteristic of isolated land masses. This is particularly trial molluscs and arthropods (Eldredge & Miller 1995, true of the Hawaiian Islands, where the biota is well 1997; Miller & Eldredge 1996). Second, the tremendous known for its extraordinarily high levels of endemism, topographical range and consequent environmental with species frequently confined to minute distributional diversity (wet and dry habitats; bogs, shrublands and ranges. At the same time, anthropogenic disturbance, forests; elevations from tropical coastlands to alpine either direct or indirect, is causing rapid species decline. deserts) provide a huge ecological spectrum for species No other area in the USA has such a high proportion of differentiation. Third, the series of islands provides a endemic species nor suffers the impact from as many new replicated system for examining within- and between- species annually as the Hawaiian Islands (Dobson et al. island patterns of species formation across similar 1997). As a consequence, the archipelago represents a ranges of environmental extremes. Finally, because the microcosm for studies in evolution and conservation. high islands are arranged by age from Kaua’i (oldest) to There are several geographical features that render the Hawai’i (youngest), the archipelago allows examination Hawaiian archipelago an unparalleled scientific labora- of species formation within an identifiable chronological tory for studying processes of evolution. First, the framework (Carson & Clague 1995) (Fig. 1). The initial extreme isolation of the Hawaiian Islands has allowed stages of population subdivision can be found on the repeated and explosive diversification of species from a youngest island, with the dynamic volcanic state of the single ancestor, often accompanied by radical shifts in island providing shifting barriers to gene flow (Carson 1990a). Later stages in species formation can generally Correspondence: R. G. Gillespie. Tel.: +01-(808)-956–8884; Fax: be found on the older islands. Upon this palette, the +01-(808)-956–9608; E-mail: [email protected] primary factors responsible for generating the high © 1998 Blackwell Science Ltd 520 G. K. RODERICK AND R. G. GILLESPIE Fig. 1 Major land masses of the Hawaiian Islands. Dark circles mark the highest (> 900 m) volcanoes. Elevations and ages of the volcanoes are given (from Carson & Clague 1995). diversity of endemic species of arthropods are genetic species radiations). This phenomenon was first described factors associated with small population sizes, and eco- in the seminal papers of Carson (1970, 1974, 1987b) who logical factors arising from habitat and/or host shifts. used chromosomal sequences to infer phylogenetic pat- Here, we examine patterns of species formation and terns of the Hawaiian Drosophila. In addition, species (or phylogeography within lineages of Hawaiian arthropods. populations) tend to have the narrowest ranges on the Arthropods make up over 75% of the Hawaiian endemic older islands (Gillespie et al. 1997). Species on the biota (Eldredge & Miller 1997) and include some of the youngest island (Hawai’i) tend to be widely distributed world’s best-known species radiations. We first document over the island and in some groups these species are phylogeographic patterns within lineages and then shared with Maui, the next-youngest island. A number of discuss mechanisms that have contributed to these different within-island patterns may be superimposed on patterns in different taxa. In this discussion we use these general patterns. Here we discuss three categories currently recognized species designations and do not of patterns found in the native Hawaiian taxa: (i) single attempt to impose our own assessments of species bound- representatives of a lineage throughout the islands; (ii) aries. However, it should be noted that species designa- species radiations in which there are either (a) single tions are particularly problematic in Hawai’i, as endemic species on different volcanoes or islands, or (b) hybridization may be a recurrent theme, and there is multiple species on each volcano or island; and (iii) sin- increasing evidence for the existence of a number of para- gle widespread species within a radiation of species that phyletic ‘species’ (see Discussion below). We consider exhibits local endemism. phylogeography in its broadest context, ‘the study of principles and processes governing the geographical dis- 1. Single species in a lineage tributions of genealogical lineages including those at the intraspecific level’ (Avise 1994). While this definition may A number of taxa in the Hawaiian Islands are repre- encompass a greater evolutionary spectrum than others in sented by single species. Carson (1987a) has compiled the volume, it is the relevant focus for many of the impor- genus:species ratios for selected Hawaiian taxa, and has tant patterns of evolution in Hawai’i. suggested genetic mechanisms that could provide for the development of speciose vs. nonspeciose lineages. These single species are generally endemic to the Patterns archipelago, and are often morphologically unique. The The phylogeographic pattern that predominates in most fact that they have failed to proliferate has often been Hawaiian taxa, both species and populations, is a step- attributed to high dispersal abilities. For example, two like progression down the island chain from the oldest to species of dragonflies, Anax strenuus (Aeshnidae) and the youngest islands (see Funk & Wagner (1995) for a Nesogonia blackburni (Libellulidae), are both single rep- review of the progression rule as applied to Hawaiian resentatives of Hawaiian lineages (Howarth & Mull © 1998 Blackwell Science Ltd, Molecular Ecology, 7, 519–531 PHYLOGEOGRAPHY OF HAWAIIAN ARTHROPODS 521 1992). The lack of diversification in these taxa is consis- that colonization is relatively recent, or that possible tent with their tremendous dispersal abilities. In partic- habitats into which they could diversify have already ular, the sister species of A. strenuus is A. junius, a been ‘filled’. species that is widespread throughout the Pacific. Among Lepidoptera, the two endemic Hawaiian butter- Single representatives among relict taxa flies, Vanessa tameamea (Nymphalidae) and Udara black- burni (Lycaenidae) are the sole representatives of their An alternative explanation for the nonspeciose nature of lineages in the Hawaiian Islands. Again, the non- certain taxa is that they may be relict lineages. Certain speciose nature of these taxa has been attributed to the taxa are characterized by a suite of apomorphies which, in fact that, at least for Vanessa tameamea, they belong to some cases, has even caused them to be placed in mono- groups that are noted for migrations or strong flight typic genera. For example, the spider Doryonychus raptor (Howarth & Mull 1992). Among Hemiptera, the genus (Tetragnathidae) has abandoned the web-building Hyalopeplus (Miridae) has not radiated ecologically or behaviour characteristic of the family, and developed differentiated by island. The single endemic Hawaiian long claws at the ends of the tarsi of leg pairs I and II species, H. pellucidus, belongs to a genus that is (Gillespie 1991b, 1992). The claws are employed in a widespread in the Pacific. This pattern contrasts to that unique behaviour whereby insects are impaled directly of related mirids in endemic genera that have speciated from the air. D. raptor has all the features characteristic of extensively (Asquith 1997). Finally, the koa bug, the genus Tetragnatha, but is phylogenetically basal when Coleotichus blackburniae, is a lone representative of its compared to