The Land Fauna of Asce~wio~~Island: New Data from Caves and Lava Flows, and a Reconstruction of the Prehistoric Ecosystem N. Philip Ashmole; Myrtle J. Ashmole Jour~ulofBiogeogruphy, Vol. 24, No. 5 (Sep., 1997),549-589. Stable URL: ht~p://links.~stor.org/sici?sici=O3O5-0270%2#199709%2924%3A5%3C549%3ATLFOA1%3E2.0.CO%3B2-K Journal of Biogeography is currencly pubIished by BIackweII Science, Inc.. Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, availabIe ac http://www.~s~or.org/aboutl~eims.h~ml.JST0R1s Terms and Conditions of Use provides, in parc, chat unIess you have obtained prior permission, you may not downIoad an entire issue of a journa1 or muItipIe copies of arcicIes, and you may use contenc in the JSTOR archive only for your personal, non-commercial use. PIease contacc he pubIisher regarding any furher use of chis work. Publisher contacc information may be obcained at ht~p://www.js~or.org/joui~als/bIacksci-inc.hcm1. Each copy of any parc of a JSTOR transmission musc concain the same copyrighc nocice that appears on the screen or printed page of such transmission. JSTOR is an independent not-for-profic organization dedicated to creating and preserving a digita1 archive of scholarly journals. For more information regarding JSTOR, p1eae concact J stor-info @Jscor.org. http://www.js~or.org/ Sat Apr 10 1853:13 2004 The land fauna of Ascension Island: new data from caves and lava flows, and a reconstruction of the prehistoric ecosystem N. PR~LIPASHMOLE and MYRTLEJ. ASHMOLEDivkion ojBidogical Sciencex, Univerxi~yof Edinburgh, Edznburgh EH9 3Jz U,K. Abstract. Tarestria1 arthropods were investigated in 1990 populat~ons and perhaps reinforced at intervals by and 1995 on Ascension, a young and extremely isolated additional groups of colonists The more extensive and volcanic island in the equatorial Atlantic. Three new genera largely cryptogamic vegetation on the central peak had a (of Araneae, Oribatida and Psocoptera) were discovered poor fauna probab[y composedmainly of micro-arthropods. and the number of apparently endemic species was increased A[ong coasts, on islets and in the extensive seabird colonies from fourteen to about twenty-nine. Some 311 species of there were additional arthropod species and also a flightless land animals (a few now extinct) are known to have rail and a night heron (both now extinct). established themselves on the idand; In a few groups there Invertebrate stocks that colonized Ascension underwent a may be significant numbers of unrecorded spies. About variety of evoluttonary changes including phyletic evolution ninety-five endemic and non-endemic species are considered leading to endemic status, adaptation to subterranean life to be native (including two marine turtlesj twelve seabirds (Araneae, Pseudoscorpiones, Collembola and Psocoptera), and two extinct landbirds); seventy-eight are of doubtful character release (phorid Dipterajj and probably splitting of status; and at least 138 were probably introduced by humans. lineages (speciation) with~nthe island (Isopoda, Collembola Natural colonists arrived mainly by air (drifting with the and gryllid Orthoptera). The relatively high diversity of wind during migratory movements) but some evidently Pseudosmrpiones (five species in five genera) and their 100% travelled on floating objects or attached to birds or other apparent endemicity is notable. animak Natural colonization was almost entirely from The md~genousfauna of Ascension provides a view of Afr~ca. an early stage in the processes of mlonization, adaptive Before the arrival of humans, Ascension had an early evolution and radiation which-over much longer sumssiona[ ecosystem. The fauna in the lava and cinder periods-give rise to the richer and more distinctive faunas deserts of the lowlands-both on the surface and in of older oceanic islands such as Ascension's nearest subterranean cracks and caves-wz dominated by neighbour, St Helena, where a few invertebrate c[ades have taxonomically varied scavengers and mainly arachnid undergone repetitive speciation and some adaptive predators. The scattered angiosperms here and in the radiation. foothills supported some host-spec~fic herbivores with associated predators; they were also exploited (especially Key words. Ascension Island, St Helena, zoogeography, after exceptional rains) by a number of Orthopteraj insect m~gration, co[onization, idand evolution, lava Hemiptera and Lepidoptera derived from migratory African habitats, cave biology, arthropods. climate have alltred the spape and nature of some other fuule INTRODUCTION info fhis, I Ieave if to the learned to dkpu~eof.' 'I hove heretofore asked the quutkn concerning Mauritius henm and duhs, thatt seeing those could neither fly nor Peter Mundy in June 1656 (Temple & Anstey, 1936) xwyme, beeing cloven fouled and withoutl wingx on an ild far from any olher land, and none to bee seence elce where, The quotation from Peter Mundy highlights the particular haw lhey shld come ~h~ther?Soe now ag0in.e concerning lhe biological significance of isolatd oceanic islands. We can Ascenfian birds a[Isoe, thatt cmnectherjiy nor swymme. The now generally assume that the ancestors of all their ildbeecng aboutt 300 Ieaws from the coast of Guinnea indigenous, terrestrial biota arrived over water, The and 160 leagues from the ilatzd of St Matheo, the nearut composition of the fauna and flora provides evidence on the land to it, the pestion i~,how they shold bee generaled, pattern of transoceanic movement, and the characteristics whithey created theye from lhe beginning, or hall the earlh of island species throw light on evolutionary processes. pro&celh them of itx owne accord, m mice, serpetmx, jies, Although Ascension cannot compete with older oceanlc wormes, etts. hects, or whithey the wture of the earth and islands and archipelagos in terms of diversity or taxonomic I997 Blackwell Science Lcd 550 N, Phi/;p Ashmole and Myrtle J. Ashmole FIG, 1, Map of Ascension IsIand, South AtIantic Ocean, showing Iocations of coIkctmg sites Numbered grid Iines (indicated at 5km intervals) are those of the Universal Transverse Mercator Grid* zone 28; they should be prefixed with the letters ES. Contour inkrval distinctiveness in its endemic fauna, it can provide a 200km closer to Africa than it is now. The island is snapshot of a relatively early stage in the process of arrival geologically active and the last on-shore eruption may have and diversification of land animals on what wa-in the been within the last millennium (Rosenbaum} 1992). relatively recent past-a pile of almost sterile steaming rock Ascension Island now has an area of c. 97 km2}but eustatic emerging from the tropical Atlantic almost midway between changes in sea level resulting from Pleistocene glaciations the African and South American continents. will have led to substantial fluctuat~onsin thls throughout Ascension is an extremely isolated and relatively young its existence. The lsland has rugged volcanic terrain (Figs, oceanic island, Lying at 7-57's 14O22'W (Fig. I), it is 2-31. The relatively low and dry western part is dominated 1300 km from St Helena, 2232 km from Recife on the coast by scoria cones and basaltic lava flows, mantled in many of Brazil and 1504 km SSW of Cape Palmas (Liberia). To places with hepyroclastic deposits, The higher eastern end the east, the coast of Gabon is just over 2600 km away and is moister with many domes of trachyte lava, culminating that of Angola and northern Namibia-another potential in Green Mountain, which has an altitude of 859 m. Just source for the biota-about 3000 km. north of the eastern end of the main island lies The island is situated about 300 km northwest of the Boatswainbird Island, with an area of about 5 ha; there are plume of the Ascension ho&pot (Brozena & White, 1990), also a number of rocky stacks around the mast (Fig. I), some 100km west of the Mid-Atlantic Ridge and These sites, which are free of alien mammals, are now of immediately south of the Ascension Fracture Zone. The sea crucial importane for breeding seabirds {Ashmole, floor around the island is 54million years old (Brozena, Ashmole & Simmons, 1994). 1986) and the island cannot be older than this. The oldest Ascens~onhas a tropical but oceanic climate, with little dated subaerial rocks are estimated at 1.5 million years seasonal change. Monthly average maximum temperatures (Harris, Bell & Atkins 1982; Harris, 1986). Ascension is at sea level vary only between about 27 and 3I0C}with moving away from Africa at a rate of some 36 mm per year the warmest weather in March-April and molest around (Brozena} 19861 in conformity with spreading at the Mid- September (Mathmon, 1990; note that figures quoted in Atlantic Ridge, but cannot ever have been more than about Ashmole et a!.} 1994 from Dufey, 1964 appear to be in 0 Blackwell Science Lrd 1997,jaurnaf of Bfogeafrophy, 24, 549-5a9 FIG. 2, Aerial photograph of Ascension Island, 1989. Crown apyright error). The island Iles in the path of the South-East Trade southward displacements of the Intertropical Convergence Winds, and southeast and easterly winds blow for more Zone, which affected the tropical Atlantic at these times than half the time in every month (Duffey, 1964; Scullion, (Philander, 1986). Ascension is also affected by southern 1990). The trade wind inversion at a height of c, 1000-l500 m hemisphere 'easterly waves', disturbances apparently acts as a strong cap inhibiting vertical cloud development originating in centra1 quatoria1 Africa and moving and thus ensures generaIIy Iow ramfa11 (RiehI, 1979). westwards across the tropical Atlantic (Hall, 1989). Nevertheless, as the winds rise over the eastern end of the Ascension Island was discovered by a Portuguese island they bring cool and misty conditions to the ridge of navigator in 1501; we use the word 'prehistoric' to refer to Green Mountain, and a mean annual rainfall (at an altitude events prior to that date.