Aquatic Heteroptera in New Guinea 327 Assembling New Guinea: 40 million years of island arc accretion as indicated by the distributions of aquatic Heteroptera (Insecta)

Dan A Polhemus1 and John T Polhemus2 1Department of Entomology, MRC 105, National Museum of Natural History, Smithsonian Institution, Washington, D% C%, USA 20560 2Colorado Entomological Museum, 3115 S% York St%, Englewood, CO, USA 80110

Key words: New Guinea, island arcs, geology, biogeography, aquatic Heteroptera

Abstract tory have revealed that at least two major epi- sodes of collision and terrane accretion have The island of New Guinea consists of the northern margin occurred in the northern half of the island of the Australian continental craton that has collided over (Pigram and Davies, 1987; Davies, 1990; Smith, the past 40 million years with a series of migrating island arcs Each of these arcs has had a separate tectonic history 1990; Pigram and Symonds, 1991), and each of and carried a correspondingly different biota Recent faunal these has in turn left a biological signature The surveys and phylogenetic analyses, coupled with evolving current challenge for regional biogeographers tectonic knowledge, are now allowing us to identify ele- lies in deciphering these signatures, and deter- ments within the New Guinea aquatic Heteroptera biota that may be correlated with particular accreted arc systems mining, if possible, which plant and animal It appears that many Asian-derived groups arrived via a groups aggregated to the present faunal mixture Papuan arc that collided obliquely with the northern margin via which arcs of Australia between the Late Eocene and Early Oligocene, In a previous paper, Polhemus (1996) re- while other distinctively Melanesian groups evolved on an viewed the history of biogeographic concepts isolated Solomons arc that was initiated in the Oligocene and accreted terranes to northeastern New Guinea from the relating to New Guinea, in particular the devel- Miocene into the Pliocene This gradually improving under- opment of the ‘Inner arc’ versus ‘Outer arc’ para- standing of the island’s interrelated tectonic and biotic his- digm that has dominated most modern analyses tory is permitting a better comprehension of the complex of the island’s biogeography The underlying patterns of faunal fusion and disjunction currently present geological concepts upon which this hypothesis within the Melanesian region was based were shown to be correct in the broad sense, but flawed in many other aspects Introduction In particular, the concept of a single ‘Outer arc’ extending from New Guinea to Tonga was A long standing goal of zoogeographers and his- shown to be simplistic, since the arcs grouped torical geologists has been to integrate biologi- under this concept represented at least four cal information and Earth history models No- separate systems with dissimilar ages, origins, where is this better pursued than in the young and subduction polarities (those readers unfa- and tectonically complex region surrounding miliar with the nomenclature, geological struc- New Guinea The island represents the northern ture, and mobilistic nature of island arcs are also margin of the Australian continental craton, referred to the paper cited above, which con- which has been uplifted through a series of col- tains a review of these subjects) In this paper lisions with southward-migrating island arcs that we seek to correlate particular groups of aquatic have formed along the boundary between the and semi-aquatic true bugs (referred to hereafter Australian and Pacific tectonic plates (Hamilton, as aquatic Heteroptera) with particular arc sys- 1979, 1988; Kroenke, 1984) Recent advances in tems and terranes, assessing their probable time our understanding of the island’s geological his- of arrival in New Guinea, and their subsequent

Biogeography and Geological Evolution of SE Asia, pp 327-340 Edited by Robert Hall and Jeremy D Holloway © 1998 Backhuys Publishers, Leiden, The Netherlands 328 D% A% Polhemus and J% T% Polhemus speciation or dispersal within the island east of New Guinea appear to have been formed In the analysis below, brief synopses of cur- during alternating episodes of volcanism along rent geological hypotheses for the assembly of two distinct zones of crustal weakness in the New Guinea are presented, noting also areas of Solomons and northern New Guinea According current uncertainty One difficulty in providing to Kroenke’s (1984) model, each time that sub- such a discussion lies in the confusing number duction along one of these zones became inacti- of different names that have been given by vari- vated, it would reactivate along the other The ous authors to the Tertiary arc systems that sequence below has been postulated (terminol- formed north of New Guinea In the present pa- ogy for trenches and arc systems follows per we follow the terminology of Kroenke Kroenke, 1984) (1984), recognizing both his Papuan and Northern New Guinea subduction episodes Solomons arcs The former is equivalent to the and corresponding trench systems: Salumei System of Davies (1990), while the lat- 1 Early Eocene to Early Oligocene: Aure- ter corresponds to the Melanesian arc of Smith Moresby-Pocklington trenches (1990) and Weiland and Cloos (1996) 2 Early Miocene to Late Miocene: Wewak- The known distributions of aquatic Heter- Trobriand trenches (= the New Guinea trench of optera are compared to these geological mod- Davies, 1990) els, and the degree of congruence discussed Solomons subduction episodes and corre- The analysis of such patterns is far from com- sponding trench systems: plete Several of the most potentially informative 1 Early Oligocene to Early Miocene: North Solo- groups, such as Rhagovelia (Veliidae) and mon trench Ptilomera (Gerridae) have not yet been taxo- 2 Late Miocene to Holocene: South Solomon nomically revised, much less cladistically trench analyzed, and as a result this paper relies heavily These alternating episodes of subduction ac- on currently unpublished data (i%e , undescribed tivity in eastern New Guinea are correlated with species in the collections of the authors) The the following tectonic events (for a CD ROM analysis should thus be seen as a preliminary video of the hypothesized plate and arc motions evaluation of biogeographic hypotheses based consult Yan and Kroenke, 1993): on work in progress, rather than a validation of Cretaceous – Rifting and basin formation oc- such hypotheses; the latter can only be based on curred along the passive northern Australian phylogenetic analyses of the insect groups in- margin This rifting, which took place in several volved Many such analyses are in the process of phases, isolated small slivers of continental crust being completed, however, and in the near fu- outboard of marginal basins These slivers have ture it should be possible to provide more rigor- been referred to as the ‘Inner Melanesian arc’ by ous tests of the hypotheses presented below many previous biogeographers While there may have been some arc related activity and back-arc spreading in the marginal basins dur- Geological setting ing the final phases of this process, it was for the most part a rifting event, similar to the process The northern half of New Guinea is a composite seen in current day East Africa of island arcs that have accreted to the northern Middle Eocene – Eastward subduction below margin of Australia at various times over the last the Pacific plate along the Aure-Moresby- 40 million years Kroenke (1984), in a synthesis Pocklington trenches, lying well north and east of covering the eastern half of the island, recog- the rifted Australian margin, produced a south- nized three such arc systems, the Papuan, ward migrating Papuan arc in an oceanic setting Trobriand, and Solomons arcs, that had already Early Oligocene – The eastward subduction collided with New Guinea, and a fourth, the Bis- below the Papuan arc ceased as the arc collided marck arc, that was in the process of doing so with the Australian margin, causing overthrust- In the west, Hamilton (1979, 1988) discussed the ing of arc terranes onto the Australian craton incipient collision of a fifth system, the Banda New westward subduction was initiated below arc, that is approaching New Guinea from the the Australian plate in the Solomons zone along southwest the North Solomon trench, forming the eastward The Banda arc system in the west is formed migrating Solomons arc in an isolated oceanic along the boundary of the Australian and Eura- setting Western extensions of this arc, linked by sian plates above the well defined Sunda sub- transforms, appear to have extended to the area duction zone By contrast, the arcs lying to the north of Irian Jaya Aquatic Heteroptera in New Guinea 329

Fig1 Map of New Guinea showing geological provinces discussed in the text Unstippled areas = Australian craton and Vogelkop craton fragment; black areas = Papuan arc terranes, accreted in the Eocene and Oligocene; dark stippled areas = Solomons arc terranes, accreted in the Miocene; light stippled areas = Banda arc terranes and associated uplifts, currently in the process of accretion; very finely shaded areas = mountainous regions at elevations above 2000 metres Areas in the Bird's Neck are of uncertain geological affinity; although assigned to the Australian craton in this figure, they are also likely to contain accreted terranes from other sources

Early Miocene – The Ontong Java plateau ward over the Solomon Sea In the west, the jammed the North Solomons trench, ending this accreted terranes of northern Vogelkop were episode of subduction in the Solomons zone dismembered by left lateral faulting, while the South or west-dipping subduction was reacti- Banda arc began the first stages of collision from vated in the New Guinea zone along the Wewak the southwest and Trobriand trenches (Hall, 1997), producing All the arc systems associated with these epi- onshore volcanism in New Guinea, forming the sodes of subduction discussed above, with the Trobriand arc to the east, and gradually consum- exceptions of the recent New Britain arc and the ing the sea floor separating the Solomons arc eastward migrating Banda arc, appear to have islands from northern New Guinea contributed accreted terranes, and thus accreted Middle Miocene – Terranes associated with biota, to the modern island of New Guinea the Solomons arc system collided obliquely with (Fig 1), although the sequence of accretion is northern New Guinea from west to east (Davies still poorly understood in the western half of the et al , 1996; Hall, 1997), ending subduction in island New Guinea as we see it today thus con- the New Guinea zone for a second time This sists of a complex fusion of geological fragments collision continued into the Pliocene, with the derived from the sources summarised below last unit to be sutured consisting of the Adelbert- Finisterre terrane Subduction then reactivated in the Solomons zone along the South Solomons The Australian craton and Vogelkop trench, but in an eastward direction opposite to the previous polarity of the North Solomons The southern half of New Guinea consists of the trench In the west, the Vogelkop peninsula was Australian continental craton, which was a pas- sutured to the main body of New Guinea sive margin until the Cretaceous, when renewed Holocene – Complex fracturing of plates rifting began along its eastern side, followed in northeast of New Guinea created many small the Eocene by the series of arc collisions dis- arcs and subduction zones, including the New cussed above Although the margin itself lay in a Britain arc, which began to advance southeast- roughly east-west orientation, a basement high 330 D% A% Polhemus and J% T% Polhemus extended north in the area of the current Papua lided In addition, another cratonic fragment, New Guinea-Irian Jaya border as far as the Bor- now present in the Jimi Terrane of Papua New der Mountains west of the Sepik basin, and re- Guinea (i%e%, the Bismarck-Kubor block), rifted mained subaerial during the Mesozoic when from the Australian margin and functioned as a most of the remaining platform was submerged separate island during the Late Cretaceous be- (Davies, 1990) Evidence for this high can still be fore being reintegrated into New Guinea at or seen in current drainage patterns, which flow near the time of the collision with the Papuan away from it to both the east and west This arc (Pigram and Davies, 1987; Davies, 1990); this structure has had an important and previously Bismarck-Kubor block appears to have re- underestimated influence on aquatic zoogeogra- mained subaerial from the Cretaceous onward phy in the region, since it represented a sub- (Davies, 1990) Other subaerial blocks of conti- aerial salient of Australia that projected north- nental crust may also have been rifted from ward into the tropics and subtropics when the northern Australia and followed similar histories rest of the continent was still in cooler climatic as small islands north of the main craton during zones It also seems to have acted like a wedge, the Late Cretaceous, but if so, then all of them splitting the accreted terranes of the Oligocene were trapped and crushed between the Papuan Papuan arc arc and the Australian margin in the Eocene The Vogelkop peninsula for most of its history (Rogerson and Hilyard, 1990) was a detached piece of the Australian continen- tal craton (the Kemum terrane) that separated in the Mesozoic (Hamilton, 1979; Pigram and The Papuan arc Davies, 1987) It was probably never far away from the northwestern margin of the main Aus- The first arc collision event in the assembly of tralian land mass, however, since its post-Creta- New Guinea occurred when the Papuan arc col- ceous tectonic history has been substantially lided with the northern margins of Vogelkop similar to that of the main body of New Guinea; and Australia This collision was oblique from it has received several waves of arc terrane lami- west to east, thus the initial contact appears to nations onto the northern and eastern margins have been possibly latest Cretaceous or Early of its continental core, and in generally the same Eocene in Irian Jaya (Davies et al , 1996), ex- historical sequence The Vogelkop, after a con- tending through the Late Eocene and Early siderable period of isolation, was finally fused to Oligocene in Papua New Guinea (Kroenke, the remainder of New Guinea in the Miocene 1984; Pigram and Davies, 1987; Davies, 1990, The suture zone is marked by the Wandamen Pigram and Symonds, 1991) The collision em- peninsula and adjoining Mangguar Terrane of placed the overthrust central Irian Jaya ophiolite Pigram and Davies (1987) and metamorphic belts (the Rouffaer terrane of Pigram and Davies, 1987), along with correlative April Ultramafics, Marum Complex, and Papuan The ‘Inner Melanesian arc’ Ophiolite Belt in Papua New Guinea (the Sepik, Marum, and Bowutu terranes of Pigram and The ‘Inner Melanesian arc’ (a term that is mis- Davies, 1987), and apparently the Tamrau leading and probably ought to be dropped, ex- Mountains of Vogelkop This collision also re- cept for its long history of usage; see Polhemus, sulted in the depression of the Australian conti- 1996) was not an arc at all, but a set of thin strips nental margin south of the overthrust terranes, of continental crust rifted from the eastern mar- which were thus separated from the subaerial gin of the Australian continent during the devel- craton by a water gap through most of the opment of the Tasman and Coral Sea basins in Miocene (Pigram and Symonds, 1991) The lime- the Late Cretaceous and Early Tertiary The New stones that formed in this shallow sea would Guinea sector of this ‘arc’ is represented by later end up on the summit of the New Guinea Pigram and Davies (1987) as a long, narrow sali- central ranges after a second episode of orogeny ent of Australian continental craton lying north driven by the collision of the Solomons arc in of the Coral Sea basin (the Eastern and Papuan the Late Miocene (Nash et al , 1993) plateaus) These crustal slivers north of the Coral Waigeo is of the correct age and composition Sea were apparently never emergent, but did to be part of the Oligocene episode of terrane constitute a topographic barricade against collision (Charlton et al , 1991), and has had which the southward migrating arc systems that clear historical associations with greater New formed the Papuan peninsula eventually col- Guinea based on its biota, but its high degree of Aquatic Heteroptera in New Guinea 331 species level endemism also implies an ex- Solomons arc system via a transform, and tended period of isolation by water gaps The Pigram and Davies (1987) similarly suggest that presence of thick and extensively exposed lime- it was previously part of a system that included stone strata also indicates that much of the is- the Bismarcks and Solomons, and formed in an land was submerged during the Miocene isolated oceanic setting to the northeast of New (Charlton et al , 1991) It seems possible that Guinea Given its late arrival in comparison to Waigeo was a part of the Papuan arc that was other terranes linked to the Solomons arc sys- accreted to northern Vogelkop in the area west tem, it is possible that this terrane occupied an of the Tamrau Mountains during the Oligocene, isolated position on a transform connected sec- then sheared off and carried further westward tor of the arc lying between the Solomons to the along the Sorong Fault zone to its present posi- east and the other Miocene terranes to the west tion during the Pliocene In the far west, all of the Miocene accreted terranes of the Vogelkop have been highly frag- mented by left-lateral shearing For instance, The Solomons arc Pigram and Davies (1987) hypothesized that the Arfak Mountains, Yapen, and parts of Biak are A large number of terranes unrelated to those of all portions of the same original terrane that was the Papuan arc accreted to northern New ripped apart by faulting In central Irian Jaya, the Guinea during the Miocene and into the Miocene arc collision also caused a massive de- Pliocene These were formed along a separate formation of the Australian continental craton arc system that lay north of the Papuan arc, pos- south of the overthrust ophiolitic and metamor- sibly a transform connected western extension phic terranes of the Irian Jaya Mobile Belt that of Kroenke’s (1984) Solomons arc subduction had been emplaced in the Late Eocene or Early system They are hypothesized to include, from Oligocene (see previous discussion), pushing west to east, the Arfak Mountains, Biak, Yapen, the Miocene limestones that had formed south the Van Rees, Foja, Cyclops, Torricelli, Prince of this belt to elevations of over 5000 metres, Alexander, Adelbert, Finisterre and Saruwaged and creating a large, overthrust anticlinal struc- Mountains, and possibly the Papuan peninsula ture, the Mapenduma anticline (Nash et al , As with the terranes of the Papuan arc that pre- 1993) This anticline, which appears to be struc- ceded them, these terranes converged obliquely turally related to the Muller anticline of west from west to east, with collision beginning in the central Papua New Guinea, was then subse- Late Miocene in Irian Jaya, but continuing until quently eroded in an asymmetrical fashion from the Pliocene in Papua New Guinea (Cooper and the south, due to orographically induced rain- Taylor, 1987) fall, producing the huge escarpments that form Kroenke (1984) considered the Papuan pe- the southern flank of the present Irian Jaya cen- ninsula to have been accreted to New Guinea in tral ranges (Weiland and Cloos, 1996) the Late Eocene or Early Oligocene, at the same The tectonic history of western New Guinea time that the ophiolite belts of the central ranges in the Early to Middle Miocene is still poorly un- were emplaced, and this interpretation is also derstood, particularly in regard to the accretion implied in the maps presented by Weiland and of terranes in the northern coastal ranges of Irian Cloos (1996) Pigram and Davies (1987), by con- Jaya and the formation of the Weyland Moun- trast, considered the Papuan peninsula to con- tains in the Bird’s Neck region (Milsom, 1991) sist of a complex terrane composed of multiple Our assignment of Irian Jaya’s Miocene terranes arc fragments (much like the present day Philip- to a westerly extension of the Solomons arc sys- pines) that was assembled offshore in the tem is thus tentative, pending more detailed Oligocene, then accreted to the remnant subma- geological and biogeographic analyses rine continental sliver of the ‘Inner Melanesian arc’ lying north of the Coral Sea basin in the early to middle Miocene, and finally sutured to Other arc systems the remainder of New Guinea Nearly all tectonic reconstructions agree that In addition to the systems discussed above, two the Adelbert-Finisterre terrane accreted to New other arcs, the New Britain arc and the Banda Guinea in the Late Miocene or Early Pliocene, arc, are advancing toward New Guinea, and will and was the last of the Miocene terranes to ar- collide with the island over the next ten million rive Kroenke (1984) provides a structural and years if current plate motions continue Al- tectonic analysis linking this terrane to the though these incipient collisions have not yet 332 D% A% Polhemus and J% T% Polhemus

Table 1 Currently known distributions of aquatic Heteroptera genera occurring on New Guinea in relation to arc system terranes and other geographic areas discussed in the text (see Fig1) Codes used in table: AC = Australian craton; VK = Vogelkop craton fragment; IA = ‘Inner Melanesian arc’ terranes (ie, Jimi terrane); PA = Papuan arc terranes; SA = Solomons arc terranes; NB = New Britain arc and Bismarck Islands; SL = Solomon Islands; BA = Banda arc

GENUS AREA AC VK IA PA SA NB SL BA

BELOSTOMATIDAE Appasus XX X Lethocerus XX CORIXIDAE Cnethocymatia X Micronecta XX Sigara X GELASTOCORIDAE Nerthra XXXXXXXX GERRIDAE Andersenella X Calyptobates X Ciliometra XX Halobates XX XXXX Iobates XX X Limnogonus XXXXXXXX Limnometra XXXXXXXX Metrobatoides X Metrobatopsis XXX Neogerris XX Pseudohalobates XX Ptilomera XXXXXXXX Rhagdotarsus XX XX X Rheumatometroides XX XXXX Stenobates XX X Stygiobates XX Tenagogonus XX XX X Thetibates XXX HEBRIDAE Hebrus XXXXXXXX HERMATOBATIDAE Hermatobates XX XXXX HYDROMETRIDAE Hydrometra XXXXXXXX LEPTOPODIDAE Valleriola XX X MESOVELIIDAE Mesovelia XX XXXX Phrynovelia X NAUCORIDAE Aphelocheirus XXX Aptinocoris XX Cavocoris XX Idiocarus XXXXX Nesocricos XX Quadricoris X Sagocoris X XXX Tanycricos XX Warisia X NEPIDAE Cercotmetus XX X X Laccotrephes X Ranatra XX X X NOTONECTIDAE Anisops XXXXXXXX Enithares XXXXXXXX OCHTERIDAE Ochterus XXXXXXXX PLEIDAE Paraplea XX Aquatic Heteroptera in New Guinea 333

Table 1 Continued

GENUS AREA AC VK IA PA SA NB SL BA

SALDIDAE Pentacora XXX Saldula XXX VELIIDAE Aegilipsicola X Halovelia XX XXXX Halovelioides XX Microvelia XXXXXXXX Neusterensifer XX X Rhagovelia XXXXXXXX Strongylovelia XX XXXX Tanyvelia XX Tarsovelia XXXXX Veliohebra X Xenobates XX

emplaced terranes into the main body of the is- period between the Mesozoic and the Miocene, land, they have provided dispersal corridors for the biological evidence from aquatic Heter- biota, and produced buckling of the continental optera clearly indicates that it could not have shelf limestones in the west (Polhemus, 1996), been very far away during this period Although creating young uplifts such as the Fakfak penin- the Vogelkop fauna shows strong local ende- sula These unaccreted arc systems are not mism at the species level, it shares most of the treated in further detail in the current paper; same regionally endemic genera found on the readers wishing additional background are in- main body of New Guinea (Table 1), while hav- stead referred to Hamilton (1979, 1988) and ing very few endemic genera of its own One Polhemus (1996), who provide discussions of exception occurs in the gerrid genus Stygiobates, the incipient collision of the Banda arc with which is currently known only from Vogelkop western New Guinea, and the resulting biogeo- and Morotai (Polhemus and Polhemus, 1993) graphic consequences

Inner Melanesian arc distributions Biological responses The evidence for distributional patterns in The general distribution patterns displayed by aquatic Heteroptera potentially influenced by plants and animals on island arc systems were terranes of an ‘Inner Melanesian arc’ has been reviewed by Polhemus (1996) In the section reviewed in detail by Polhemus (1995, 1996), below, we attempt to identify aquatic Heter- and is not repeated at length here The currently optera distributions within New Guinea that ap- accepted geological hypothesis of marginal ba- pear to have been influenced by particular epi- sin formation isolating continental slivers along sodes of terrane accretion discussed in the sec- the Australian continental margin in the late Cre- tion above These distributions are further sum- taceous to early Paleocene is consistent with the marized in Table 1 biological evidence Polhemus (1995, 1996) ar- gued on the basis of aquatic Heteroptera distri- butions that such slivers must have been to Australian craton and Vogelkop distributions some extent emergent above sea level, carrying a distinctive and in some cases relict (i%e%, New Even though the geological evidence indicates Caledonia gymnosperms) biotic assemblage that Vogelkop was a separate island lying west eastward with them into the Pacific at large of the main Australian craton for a considerable Groups of aquatic Heteroptera conforming to 334 D% A% Polhemus and J% T% Polhemus

Fig2 Currently known distributions of Ptilomera water strider species occuring on New Guinea and associated arcs Most of the nineteen species recognized from this region on the basis of recent character analysis are still undescribed This fast water genus is absent in Australia and the North Moluccas, and current sampling also indicates that it does not occur on Biak, Salawati, or the Fakfak peninsula The highly localized distributions of species in the northern and central ranges indicate that speciation in this group may have been influenced by the accretion of terranes associated with the Papuan and Solomons arc systems (see text) such patterns typically show di- or tricentric dis- Papuan arc distributions tributions involving widely separated geo- graphical areas along the western margin of the The accretion of the Papuan arc in the late Pacific plate, including the southern Philippines Eocene and early Oligocene was a complex Three groups are notable in this context: event whose biological signature as reflected by Rhagovelia (Veliidae), novacaledonica spe- aquatic Heteroptera distributions is still being cies group – members of this group occur in the worked out Groups influenced by terranes from uplands of New Caledonia, New Guinea, and this arc might be expected to show localized Mindanao (Polhemus, 1995) species level endemism in the central New Rhagovelia (Veliidae), caesius species group Guinea ophiolite and metamorphic belts and in – members of this group occur in New Guinea the mountains of the northern Vogelkop penin- and the southern Philippines (Polhemus, 1995) sula, and to have outgroup relationships with Naucoridae, subfamily Cheirochelinae, tribe taxa occurring to the west, in the Malay archi- Sagocorini – creeping water bugs in this tribe pelago and mainland SE Asia Several groups are found in New Guinea and the Philippines appear to fall in this category: (Polhemus and Polhemus, 1987; Polhemus, Microveliinae (Veliidae) – the genus Aegili- 1996) psicola is currently known only from the north- In none of the above instances are members ern slopes of the central New Guinea highlands of these groups present in Australia, the In addition, another currently undescribed Moluccas, or Sulawesi Such patterns probably microveliine genus, referred to in manuscript as date from the Late Cretaceous, and are visible ‘Papuavelia’, is known only from the central now only at the tribal or intrageneric species Irian Jaya ophiolite and metamorphic belts group level Since the terranes these species oc- Ptilomera (Gerridae) – several undescribed cupied formed the initial cores of what would species in this genus appear on the basis of cur- eventually become larger islands, such as New rent collections to be localized endemics with Guinea and Mindanao, their biological signa- distributions centred on the Rouffaer and Sepik tures have been somewhat obscured by subse- terranes of Pigram and Davies (1987) The situa- quent geological and biological events tion is not clear-cut, however, since other en- Aquatic Heteroptera in New Guinea 335 demic species have ranges centred around chemistries, and occupy catchments character- terranes associated with the Solomons arc, ized by high gradients, frequent waterfalls, and which accreted in the Miocene, indicating that poor integration of drainage; in numerous in- members of this genus may have been distrib- stances these streams disappear into caves in the uted across multiple arc systems (as in fact they karst terrain, then reappear great distances away are today; see Polhemus, 1996 and Fig 2) as resurgences in the form of springs, often Nesocricos (Naucoridae) – members of this bursting as high waterfalls from sheer cliffs genus are entirely confined to the central moun- Such habitats, though scenically spectacular, are tains of New Guinea, and certain species appear not particularly easy for aquatic Heteroptera to to have ranges correlating with terranes defined colonize, since they present many significant by Pigram and Davies (1987) For instance, a topographic barriers In addition, during dry pe- currently undescribed species is apparently con- riods much of the water retreats to underground fined to the central Irian Jaya metamorphic and conduits, and those permanent surface water- ophiolite belts that comprise the Rouffaer courses that do exist are often discontinuous terrane of the above authors Stream basins in the ophiolite and metamor- Tanycricos (Naucoridae) – as with Nesocricos, phic terrains, by contrast, show very different members of this genus are endemic to New characteristics They tend to have relatively pH Guinea, confined to the central mountains of the neutral waters flowing in well sorted beds with island, and unknown from any of the north high substrate heterogeneity, and occupying coastal terranes that accreted in the Miocene well integrated catchments with moderate gradi- This suggests that they or their ancestors also ents This allows the development of a far richer arrived via the Papuan arc aquatic Heteroptera biota than streams in lime- The metamorphic and ophiolite belts of cen- stone at similar elevations The dichotomy is tral New Guinea are large, relatively cohesive particularly well illustrated by a transect over the geological units, and this is reflected in their central ranges just west of Puncak Jaya At aquatic Heteroptera biotas In Irian Jaya, for in- Tembagapura, on the south slope at an eleva- stance, the mid-montane aquatic Heteroptera tion of 2000 metres, the mountain streams are assemblage typical of these exposures is distinc- high gradient, originate in limestone catch- tive but broadly distributed, occurring through- ments, and support no aquatic Heteroptera At out the upper basin of the Mamberamo River Bilogai, on the north slope at 2000 metres, the The local dispersal of this biota has no doubt mountain streams are moderate gradient, origi- been facilitated by the long river valleys that nate in metamorphic catchments, and contain a trend east to west along strike through the diverse aquatic Heteroptera biota This leads to Derewo Fault Zone an interesting possibility – that, based on sam- Interestingly, the Heteroptera biota of the ples of aquatic insects, inferences about the pe- ophiolite and metamorphic belts emplaced in trology of a catchment can be made a priori, the Oligocene seems to be considerably richer even when the geological information is not than that of the highland limestone belt lying precisely known immediately to the south, which was elevated Not only are the aquatic Heteroptera biotas of later, in the Late Miocene and Pliocene This is ophiolitic and metamorphic terrains richer than partly due to elevation; the limestones fre- those of other areas, they are also regionally dis- quently lie at elevations ranging from 3000 to tinctive This is hypothesized to result from the 5000 metres, and the jagged crest they form rep- fact that, in addition to providing suitable physi- resents an insurmountable cold water barrier to cal habitats, the ophiolites are in many instances aquatic Heteroptera, breached only by the markers for accreted terranes that have travelled Baliem Valley to the east and the Paniai basin to considerable distances from their original point the west Both of these gaps in the high crest of formation, bringing isolated suites of aquatic drain to the south, and, as shown by the distri- insect species with them (Polhemus, 1996) bution of crayfishes (Holthuis, 1982), have Even after an arc terrane has been incorporated clearly acquired their aquatic biotas from that into a larger island such as New Guinea, its asso- direction; as such, they have offered little oppor- ciated aquatic insects often appear to retain a tunity for north to south faunal interchange high fidelity to the accreted block, probably be- In addition, the limestones produce stream cause such accretions are frequently surrounded basin characteristics that many aquatic species by topographic basins containing lakes and find unsuitable Streams in these limestone re- slow water streams that are barriers to the dis- gions usually have marly beds, basic water persal of upland species, or because the 336 D% A% Polhemus and J% T% Polhemus

Fig3 Distribution of the near-shore marine water strider genus Thetibates on New Guinea and surrounding arcs accreted ophiolitic terranes are often closely jux- terranes can be made at this time: taposed with adjacent limestone terranes also Ptilomera (Gerridae) – this genus occurs from uplifted in the course of the accretion (as in the India through Indochina and westward through central mountain chain discussed above), which the Malay archipelago to New Guinea, New Brit- once again serve as effective barriers to disper- ain, New Ireland, and Bougainville; it is absent sal of many aquatic species in Australia, the Moluccas and Philippines (ex- cept Palawan) Such a distribution implies that the group is of Asian affinities, and arrived in the Solomons arc distributions New Guinea area from the west Patterns from the central New Guinea ophiolites suggest that There is abundant evidence to indicate that the this genus was distributed on the Papuan arc accreted terranes of Miocene age in the northern prior to its accretion, and the degree of localized coastal ranges of New Guinea and the Papuan speciation in the northern coastal ranges peninsula harbour distinctive suites of endemic strongly suggests that species occupied the aquatic Heteroptera species that may be linked Solomons arc as well A distribution map of spe- to their former history as isolated land masses cies endemism within New Guinea, including The situation is made complex, however, by the undescribed species, was presented by large number of terranes, and the resulting large Polhemus (1996); although accurate in its broad number of endemic species Several groups details, this figure was overly simplistic Ongo- have the potential to be particularly informative ing character analysis has shown that many ad- in this regard, but since two of the most critical ditional species are present, and that terrane fi- are in the process of taxonomic revision, only a delity among them is even more pronounced limited analysis of patterns within the Miocene than was suggested in that paper (Fig 2) Aquatic Heteroptera in New Guinea 337

Fig4 Distribution of the water strider genus Metrobatopsis on New Guinea and surrounding arcs

Rhagovelia (Veliidae) – a large number of process of being deciphered, the biological evi- species in this group show patterns of localized dence from water bugs does strongly indicate speciation centred around accreted terranes of that the northern and southern sections of the Miocene age within New Guinea, suggesting Papuan peninsula have functioned as separate that their distributions have been strongly arc units, although it does not provide direct sup- influenced, but only a handful of species have port for whether the final amalgamated unit was been described (Lansbury, 1993) The genus assembled on or offshore The following genera promises to provide excellent resolution of have endemic species pairs distributed on the terrane linked distributions, but this must await northern and southern sides of the peninsula, completion of a taxonomic and phylogenetic re- and thus conform to the predictions of the vision currently underway by the authors terrane assembly model: Ptilomera (Gerridae; Metrobatini (Gerridae) – two genera in this see Hungerford and Matsuda, 1965); Ciliometra tribe, Andersenella and Metrobatoides, are con- (Gerridae; see Polhemus and Polhemus, 1993; fined to Miocene terranes in northern and east- Lansbury, 1996); Rhagovelia (Veliidae); Aptino- ern New Guinea A phylogenetic analysis cur- coris (Naucoridae); Sagocoris (Naucoridae) rently in progress indicates that both genera are In all cases these species pairs have their clos- most closely related to groups occurring in New est putative relatives in the main body of New Guinea, Australia, and the Solomons, supporting Guinea Although these patterns indicate that the hypothesis of a Melanesian derivation for the northern and southern sections of the penin- some of the Miocene arc biota sula functioned as separate biological units, they Although the biological patterns present do not strongly support the idea that this com- among aquatic Heteroptera occurring on the plex was assembled at any great distance from Miocene terranes are still for the most part in the the main land mass of New Guinea In addition, 338 D% A% Polhemus and J% T% Polhemus

Fig5 Distribution of the riffle bug Rhagovelia biroi on New Guinea and surrounding arcs

the evidence does not permit rejection of an al- Thetibates (Gerridae) – this marine genus ternative hypothesis to the terrane assembly (Fig 3) is known only from the Solomons, New model, namely, that the current north to south Britain, and the coastal zones of the Adelbert- disjunction may be the result of vicariance due Finisterre terrane on New Guinea (Polhemus to the formation of the Owen Stanley Range in and Polhemus, 1996) situ Metrobatopsis (Gerridae) – members of this Another interesting set of patterns is present freshwater genus (Fig 4) occur on the Solomons, in regard to the Adelbert-Finisterre terrane If New Britain, Mussau, the Adelbert-Finisterre one accepts the geological hypothesis of terrane; future collecting will undoubtedly re- Kroenke (1984) regarding the formation of the veal them on New Ireland as well Several spe- Solomon Islands and the eventual accretion of a cies are also distributed along the north coast of western extension of the Solomons arc into New Guinea westward to Yapen and Biak northeast New Guinea, and assumes that the (Polhemus and Polhemus, 1993) This could in- Adelbert-Finisterre terrane was an isolated sec- dicate that some of the Miocene terranes and tor of this arc, then taxa associated with this sys- islands in this latter area represent westward ex- tem should have distributions encompassing the tensions of the Miocene Solomons arc, but the Solomons, the Bismarcks, and the Adelbert, pattern could also be the result of dispersal fol- Finisterre, and Saruwaged ranges of mainland lowing terrane accretion New Guinea, and should show outgroup rela- Rhagovelia (Veliidae) – R biroi (Fig 5) occurs tionships including taxa in Melanesia and Aus- on New Ireland and New Britain, plus the tralia Three separate lineages of aquatic Heter- Adelbert-Finisterre terrane of New Guinea optera are known to display such patterns: (Lansbury, 1993); extensive surveys have not Aquatic Heteroptera in New Guinea 339 found it elsewhere on the latter island tions that took place during the assembly of These patterns suggest that the distributions modern New Guinea of these groups were linked to the tectonic his- tory of the Solomons arc, and support Kroenke’s (1984) hypothesis that the Adelbert-Finisterre Acknowledgements terrane is an accreted portion of this arc The authors wish to thank Dr Warren Hamilton, of the U S Geological Survey, Denver; Dr Loren Conclusions W Kroenke of the University of Hawaii, Manoa; and Dr Kevin T M Johnson of the Bishop Mu- Based on current geological hypotheses and seum, Honolulu, all of whom generously took known species distributions, it has been possi- time out of their busy schedules to educate us in ble to hypothesize linkages between the distri- the complexities of island arc geology In addi- butions of certain groups of aquatic Heteroptera tion, Kevin Hill of La Trobe University, provided present on New Guinea and particular island arc a useful review of the geological discussion and terranes that accreted to the northern margin of suggested many areas for improvement This re- the island during the Tertiary search was sponsored in part by a series of The Vogelkop peninsula possesses a biota grants from the National Geographic Society, typical of that seen on New Guinea as a whole, Washington, D C (1806-77, 2698-83, 3053-85, indicating that it and the main body of New 4537-91), and by grant BSR-9020442 from the Guinea have been in close proximity since at National Science Foundation, Washington, D C least the beginning of the Tertiary The central ophiolite belt of New Guinea pos- sesses a number of distinctive genera not seen in References the north coast ranges; these may have been derived from the Papuan arc which accreted in Charlton, T R, Hall, R and Partoyo, E 1991 The geology the Oligocene and tectonic evolution of Waigeo Island, NE Indonesia The terranes of the north coastal ranges, de- Journal of Southeast Asian Sciences 6: 289–297 rived from a Miocene arc, show pronounced lo- Cooper, P and Taylor, B 1987 Seismotectonics of New Guinea: a model for arc reversal following arc-continent cal endemism, with many species restricted to collision Tectonics 6: 53–67 single terranes This indicates that speciation Davies, H L 1990 Structure and evolution of the border may have occurred while these terranes were region of Papua New Guinea In Petroleum Exploration separate islands in an arc system north of New in Papua New Guinea: Proceedings of the First PNG Pe- Guinea troleum Convention, Port Moresby, 12–14th February 1990 pp 245-270 Edited by G J Carman and Z Car- The northern and southern portions of the man PNG Chamber of Mines and Petroleum, Papua New Papuan peninsula also show marked regional Guinea endemism, and have basically behaved as two Davies, H L, Winn, R D, and KenGemar, P 1996 Evolu- separate islands from a biogeographic stand- tion of the Papuan Basin – a view from the orogen In point This is compatible with the hypothesis of Petroleum Exploration, Development and Production in Papua New Guinea: Proceedings of the 3rd PNG Petro- offshore assembly followed by subsequent ac- leum Convention, Port Moresby, 9th–11th September cretion to the main body of New Guinea 1996 pp 53-62 Edited by P G Buchanan PNG Cham- The Adelbert-Finisterre terrane has a distinct ber of Mines and Petroleum, Papua New Guinea biota which frequently shows sister group rela- Hall, R 1997 Cenozoic tectonics of SE Asia and Australasia tionships with the Solomons, rather than with In Petroleum Systems of SE Asia and Australasia pp 47- 62 Edited by J V C Howes and R A Noble Indonesian the remainder of New Guinea This accords well Petroleum Association, Jakarta with the hypothesis that it formed as part of the Hamilton, W B 1979 Tectonics of the Indonesian region Solomons arc system U S Geol Survey Professional Paper 1078 U S Govern- Such hypotheses, although interesting, carry ment Printing Office, Washington, D C 345 pp Hamilton, W B 1988 Plate tectonics and island arcs Geo- little weight until they are subsequently tested logical Society of America Bulletin 100: 1503–1527 by cladistic analyses of the groups involved Holthuis, L B 1982 Freshwater Crustacea Decapoda of Analyses of this type are now underway, and New Guinea In Biogeography and Ecology of New promise to provide a more detailed picture of Guinea Edited by J L Gressitt Dr W Junk, The Hague the relationships between the species assem- Hungerford, H B and Matsuda, R 1965 The genus Ptilomera Amyot and Serville (Gerridae: Hemiptera) blages occurring on the various terranes In the University of Kansas Science Bulletin 45: 397–515 near future, we may thus have a much better Kroenke, L W 1984 Cenozoic development of the South- comprehension of the biogeographic interac- west Pacific United Nations Economic and Social Com- 340 D% A% Polhemus and J% T% Polhemus

mission for Asia and the Pacific, Committee for Co-ordi- Polhemus, D A and Polhemus, J T 1987 A new genus of nation of Joint Prospecting for Mineral Resources in Naucoridae (Hemiptera) from the Philippines, with com- South Pacific Offshore Areas, Technical Bulletin 6: 122 pp ments on zoogeography Pan-Pacific Entomologist 63: Lansbury, I 1993 Rhagovelia of Papua New Guinea, Solo- 265-269 mon Islands, and Australia (Hemiptera-Veliidae) Polhemus, J T and Polhemus, D A 1993 The Trepoba- Tijdschrift voor Entomologie 136: 23–54 tinae (Heteroptera: Gerridae) of New Guinea and sur- Lansbury, I 1996 Two new species of Ciliometra Polhemus rounding regions, with a review of the World Fauna Part (Hem, Gerridae) from Papua New Guinea Entomolo- 1 Tribe Metrobatini Entomologica Scandinavica 24: 241–284 gist’s Monthly Magazine 132: 55–60 Polhemus, J T and Polhemus, D A 1996 The Milsom, J 1991 Gravity measurements and terrane tecton- Trepobatinae (Heteroptera: Gerridae) of New Guinea ics in the New Guinea region Journal of Southeast Asian and surrounding regions, with a review of the World Earth Sciences 6: 319–328 Fauna Part 4 The marine tribe Stenobatini Nash, C R, Artmont, G, Gillan, M L, Lennie, D, O’Connor, Entomologica Scandinavica 27: 279–346 G, and Parris, K R 1993 Structure of the Irian Jaya mo- Rogerson, R J and Hilyard, D B 1990 Scrapland: a suspect bile belt, Irian Jaya, Indonesia Tectonics 12: 519–535 composite terrane in Papua New Guinea In Petroleum Pigram, C J and Davies, H L 1987 Terranes and the accre- Exploration in Papua New Guinea: Proceedings of the tion history of the New Guinea orogen Bureau of Min- First PNG Petroleum Convention, Port Moresby, 12–14th eral Resources, Journal of Australian Geology and Geo- February 1990 pp 271-282 Edited by G J Carman and physics 10: 193–211 Z Carman PNG Chamber of Mines and Petroleum, Pigram, C J and Symonds, P A 1991 A review of the tim- Papua New Guinea ing of the major tectonic events in the New Guinea Smith, R I 1990 Tertiary plate tectonic setting and evolu- orogen Journal of Southeast Asian Earth Sciences 6: tion of Papua New Guinea In Petroleum Exploration in 307–318 Papua New Guinea: Proceedings of the First PNG Petro- Polhemus, D A 1995 Two new species of Rhagovelia from leum Convention, Port Moresby, 12–14th February 1990 the Philippines, with a discussion of zoogeographic rela- pp 229-244 Edited by G J Carman and Z Carman PNG tionships between the Philippines and New Guinea Chamber of Mines and Petroleum, Papua New Guinea Journal of the New York Entomological Society 103: 55- Weiland, R J and Cloos, M 1996 Pliocene-Pleistocene 58 asymmetric unroofing of the Irian fold belt, Irian Jaya, Polhemus, D A 1996 Island arcs, and their influence on Indonesia: Apatite fission-track thermochronology Geo- Indo-Pacific biogeography In The Origin and Evolution logical Society of America Bulletin 108: 1438–1449 of Pacific Island Biotas, New Guinea to Eastern Poly- Yan, C Y and Kroenke, L W 1993 A plate tectonic recon- nesia: Patterns and Processes pp 51-66 Edited by A struction of the SW Pacific 0-100 Ma In Proceedings of Keast and S E Miller SPB Academic Publishing bv, the Ocean Drilling Program, Scientific Results 130: 697- Amsterdam 709 Edited by T Berger, L W Kroenke, L Mayer et al