Geology of the Solomon and New Hebrides Islands, as Part of the Melanesian Re-entrant, Southwest Pacific! P. J. COLE MAN2
THE SOLOMON ISLANDSand the N ew Hebrides Company) ; A. A. Day (University of Sydney) ; Archipelago are examples of fractured island J. G. Speight and D. H. Blake (Commonwealth "arcs," autochthonous geological systems, in Scientific and Industrial Research Organisation, which patterns of straight-line fractures and Canberra); and Y. Miezitis (Bureau of Mineral vertical and horizontal movement of blocks are Resources, Canberra). Certain geological and the dominant structural feature . The two groups geophysical information obtained by the recent have similar but yet distinctive geologies. They United Nations Special Project (Director: H. have been studied systematically only over the Winkler) and by expeditions from the Unive r last 20 years, so that many of their features are sity of W isconsin and the Hawaii Institute of still not understood, but it is clear that the re Geophysics (Director: G. P.W oollard) has sults so far obtained amount to a significant been incorporated. contribution to the understanding of fractured For the New Hebrides the main sources in island arcs and, by extrapolation, to the under clude, similarly, the work of the Geological standing of the genesis and development of the Survey (D. 1. J. Mallick, Senior Geologist; A. southwestern Pacific. H. G. Mitchell and A. J. W arden , former The aim of this paper is to present a synoptic Senior Geologist); G. P. Robinson (British account of the essential geology of the Solomon Petroleum Company); K. Liggett (New Zea and New Hebrides groups. That is, it is the land Geological Survey) ; also publ ished work rock types and successions, their ages and struc by Mawson (19 05) , Obelliane ( 1958), de la ture, which are stressed. But certain geophysical Rue ( 1937) , and Sagatzky (1959) . evidence is also incorporated, particularly where I am particularly indebted to the people of the geological implications are clear, even the Geological Surveys, not simply for their in though some of this evidence is discussed in formation and encouragement, but also for help detail in the accompanying paper in this issue in the field, often given at the cost of incon ( Furumoto et al., 1970) . There is some repe venience and personal discomfort. tition, but this is prefe rable to fostering a sense of dichotomy between the geology, sensu stricto, and the geophysics-especially in this REGIO NAL FRAMEWORK area where there is usually a nice correspondence The Solomon Islands and the New Hebrides between geological and geophysical data. Archipelago are prominent elements within the Apart from the writer's own field and labora Melanesian Re-entrant. This is one of the more tory studies, the account which follows derives forceful aspects of Carey's Tethyan Tors ion from a variety of major sources. For the Solo System (Carey, 1963) . On any recent bathy mons they include published and unpublished metric map of the southwest Pacific it shows results from Geological Survey personnel, past as a great angle jutting into the Pacific Basin. and present, in particular : J. D . Bell, J. C. The apex is the area of the Fiji-Tonga groups ; Grover (former Director) , B. D . Hackman, the northern boundary is at first neatly defined P. Pudsey-Dawson, and R.B. Thompson (Di by New Ireland and the Solomons, and then by rector); also studies by R. L. Stanton (Univer a markedly en echelon arrangement of bathy sity of New England), a long-term worker in metric highs and lows which form the northern the area; F. K. Rickwood (British Petroleum edge ( Fairbridge, 1961) of the Fiji Plateau (badly named on many maps as the N orth Fiji 1 Manuscript received Decem ber 2, 1969. 2 Geology Department, Un iversity of Western Aus or Pandora Basin) ; the eastern limb is sharp ly tralia, Perth. defined by the line of the Tonga and Kermadec 289 290 PACIFIC SCIENCE , Vol. 24, July 1970
islands, which connect with the north island of echelon disposition (Bougainville-Choiseul New Zealand. The New Hebrides form, as it Isabel-Malaita) . A minor deep reflects this ar were, a minor re-entrant which at their southern rangement and separates the Solomon Block end swings to the northeast by way of the from the Ontong Java Platform, which supports Hunter Ridge, connecting with Fiji and so con Ontong Java and other prominent atolls. This taining the Fiji Plateau (Fig. 1) . platform is a tectonically puzzling feature; it The Melanesian Re-entrant and much of the may be a remnant of the Darwin Rise (Menard, area within it is made of elongated crustal 1964). New Georgia Sound, a severely rhombic blocks which find topographic expression as depression 2,000 meters deep, separates the island chains. New Caledonia, with the Loyalty north and south flanks. The latter (BougainviIle Islands, is an advanced example, so that a large New Georgia-Russell-Guadalcanal-San Cristo part of the block is now a land mass; the Tonga bal) shows some offsetting of the islands but is Kermadec Block is youthful, the islands being not so regularly en echelon as the north flank. relatively small and widely separated ; the Solo Over its middle portion, it appears to be linked mons and New Hebrides come between and are by NE-trending horst and graben with the good examples of so-called fractured island arcs. Louisiade-Rennell Ridge (Krause, 1966). The Within the Re-entrant there are rhornbo east part of the flank, however, is delineated by chasms (the Coral Sea is a large one, New the South Solomon Trench (about 7,000 me Georgia Sound in the Solomons is small but ters) , as is Bougainville, at the other end, by a well defined) , sphenochasms (the Tasman part of the Planet Deep (9 ,000 meters) (Fig. 1) . Sphenochasm, between southeast Australia and The axis of the Solomon Block appears sig the Lord Howe Rise, is the most obvious), and moidal, but this is probably accidental and re orotaths (the Kermadec-Tonga Ridge is named flects the relative dominance of one or the other as such by Carey) . These are Carey's terms for of two lineament sets with preferred orienta features which accompany crustal sundering and tions, NW-SE and W-E, respectively. These spreading (Carey, 1958). sets, with a third subsidiary one trending NE , Emphatic straight-line lineaments are com define the outlines and relative positions of the mon ; curved or arcuate lineaments are less so, larger islands. To the west, the Block ends with and some of those shown even on modern maps the Lihir group of small volcanic islands, which may represent the subjective, rounding-off touch are separated by a minor deep from New Ire of the cartographer. Others are probably real land . This connects with northern coastal New for example, those of the Hebrides-Hunter Guinea by way of the "swirl" of the Bismarck Trench and the markedly rotational, curved fea Archipelago (New Britain, New Ireland). At tures of the eastern Fiji Plateau. Horsts and its eastern end, the Block is terminated trans graben dominate the structure, even of particu versely by the northern extremity of the New ular islands, so that in general the region exem Hebrides Block. This "cut-off" relationship is plifies the taphrogenic structural style. Tensional shared by the South Solomon Trough, terminat effects are-and appear to have been-the rule. ing against the Torres Trench, and is reflected The rim of the Re-entrant and the New on the northeastern flank by the confused ba Hebrides is tectonically unstable and seismically thymetry in the area east of Ulawa, with its active; the New Hebrides are an area of intense pocket depressions (Ulawa Deep) and linear seismicity. The Solomons, New Hebrides, and deeps (Cape Johnson Trough) . other parts of the Re-entrant are areas of mark The New Hebrides lie within the Re-entrant. edly anomalous positive gravity (Furumoto et Although not so markedly as the Solomons, the aI., 1970 ; Malahoff and Woollard, 1969) . New Hebrides Archipelago is also in part a The Solomon Islands, which help define the double chain, closed at the north by the Santa northern limb of the Re-entrant, are in part a Cruz group and at the south by a tail of vol double chain closed at the northwestern end by canic islands (Epi-Efate-Erromango-Tanna Bougainville and at the southeastern end by San Aneityum) (Fig. 2). The axis of the New Cristobal. The northern flank of the group is Hebrides Block is slightly sigmoidal, marked by made up of large islands with a distinct en a line of active and recent volcanoes. As with 0 o $ 180 g2 ~ ~ ..,.,o ~ en o 0" ~ 8 \ o ::J ~
AUSTRALIA
300
\ 140~ -\- F IG. 1. Generalized bathymetric and locality map of the Mel anesian Re-entrant. Bathymetry based on data from N orth American Aviation/ Autonetics; Th e Times Atlas of the World; and H . W . Menard and colleagues. Contours : heavy line, 1,000 meters; open bacbured line, 3,000 meters. Trenches and IV linear deeps : enclosed, close bactnoed line. \0.... 292 PACIFIC SCIENCE, Vol. 24, July 1970
- , LE GE ND • ~ Scbm e r-Ine c o nt o u r , . ,5 0 0 rn. ~ Act ive vo lc a n o e s @ O.RSTO.M. Seismolcgic el st at ion • Supe rf;OaI < 50 Depfh of l oci .. Sh allow <100 in kilometers {v l o-t e r-rn e die-l-e ) 100 <200 ... De ep )200
Magnitude of f• • v ., <5 U.S.C.G.S . shock \-. '7 '" )5 U.S.C.G.S.
16·
.. o 18
o 20
D
FIG. 2. Locality map of the New Hebrides Archipelago with seismicity for 1967 (after Benoit in Mallick, 1969) . the Solomons, it has deeps on the "reverse," covering the Fiji Plateau. To the north, the New that is, continental, side (Torres Trench to the Hebrides Block ends with a deep-part of the north and N ew Hebrides Trench to the south), Johnson Trough-Vitiaz Trench complex. At the which are separated by a shallower midd le por south, the trench swings to the east, becoming tion. This has E-W trending horsts and graben the Hunter Deep and is not crossed by any con which tail off into the New Hebrides Basin and nection with the nearby N ew Caledonia-Norfolk in part connect with the southerly extension of Ridge. As a block, the New Hebrides is rela the Rennell Ridge. The eastern flank passes tively isolated. more or less smoothly into the deeper water The high seismicity of the Solomon and New CJ ",o ~ 5'5 10'S 170'f 17s"E o..... Ulo \ . 0" " g " \ ::l 'J> b 0...
1 Z ~ :r: 3-..., <,' 0.: (,." q;~ [] I pO'''~INGTON (') TIl £ NCH o r ~ ~ ~ . ~ tIl ~I"l'.~,,~ l'-~1",J ~O ~'t-"" ~ Ol.lP :s: !f)\I)S CHOISE'lJL 0 A MAI..A,ITA sA IJf" Gl ul. -:)t- NEw btd _ " . ...' N qJ-1" 8 0IJGAJNVII..I.f ..r" N E w G .!a Ao ~ \"'~o2>Jo~J- t.~b~,~m '?'t'- GU.4D.... LcANAL S AJJ C~ 1 5T O eAL. (,\ ?' I'IU (l DIO S t-~ :> Z -=--=- = :: -~~' -:. ~;;~: .~¥~~i~~~ _'C~4~?~¥~~i~~~~~~~: ~~::C-~~~~7~:~~ ~C ~~ -:~:-~~?~ -=-~
~~;~~~~~f~~~~~~¥CCC " -- - _~~= =~ -~='~----__~~~~ .__
------~ _-::-_ -=- . - - - ~7 --~-~ - - ~ - -==~~~~~~;;i:~::-=£~ a ;>"<
~~-':f .,;.: :. I\E L ",T IVE MAG NITUDES OF DEEP SHOGKS INOteATED IlY G1ZE, 11 >nOk", 8'1' SQUARES ~.~' ~1. . h <58"Ok" 8'( DI AMONDS :'-.. ··:·1;jY ····'/·' ·1·
~ONG/ T"P INA~ SECTION ';,z.J'!}~'~ :;7 "";/'" -- 700 I ~., .
FIG . 3. Seismicity in the Solomon Islands (after Grover, 1% 9) . tv \0 \.>.l 294 PACIFIC SCIENCE, Vol. 24, July 1970
Hebrides blocks is a noteworthy regional fea a steep easterly dip for the seismic zone, al ture (Grover, 1969) (Fig. 3).3 Over the Solo though not as steep as the Solomons example. mon Block most of the earthquake shocks are If the Fiji Plateau is regarded as "continent," of shallow focus, less than 60 km, and tend to then the seismic zone, the trenches, and the dis be distributed along the southern flank of the tribution of volcanic centers become more nearly Block, adjacent to the northeastern side of the "normal." Solomon Trough and the east Planet Deep. As with the Solomons, but less spectacular in Intermediate shocks (70 to 300 krn) are con detail, the New Hebrides area also shows abnor centrated at each end, in the Planet Deep and mally high positive gravity anomalies. These do Guadalcanal-San Cristobal areas; a few inter not seem to be related directly to the seismic mediate and deep shocks have been recorded pattern. below New Georgia Sound, northeast of New Earlier published estimates of crustal thick Georgia Island . In profile, the seismic zone is ness in the Solomons and New Hebrides areas either near-vertical or steeply inclined toward were based on slender seismic evidence and pre the Pacific-an unusual case. The longitudinal liminary gravimetric results (Officer, 1955 ; profile is asymmetric, the deepest hypocenters, Coleman and Day, 1965) . The value given was up to 500 km, being confined to the western end. 15 km or less. Later seismic and gravity evi Allowing for a wide margin of error in fixing dence suggests that this is a minimum value and the hypocenters, certainly exceeding 25 km for that the actual thickness varies between 20 and the deeper shocks, it does seem that there is 30 km over the combined length (over 2,000 something of a gap in their distribution between km) of the two groups . A tentative estimate 200 and 400 km, and that centers deeper than (unpublished) by the Bureau of Min eral Re 500 km are extremely rare. Miyamura (1968) sources, Canberra, for the arc of the Bismarck considers that a gap in the distribution of inter Archipel ago is more than 30 km. The structural mediate-depth hypocenters is a feature of the and gravity picture is consistent with consider seismicity of the circum-Pacific regions. able but varying uplift of crustal and subcrustal The significance of the seismic pattern with blocks, so that variation in estimates of crustal regard to local regional structure is not readily thickness could be expected. apparent. At least superficially, the areas of deep . foci relate to the deeps at each end of the southern flank of the Solomon Block, but what GEOLOGY OF T HE SOLOMONISLAND S this means is uncertain. Similarly, there is a The major geological features of the Solo very approximate link of intermediate centers mon Islands have been described by Coleman, with some of the areas of highest positive grav Grover, Stanton, and Thompson (1965), and ity anomaly, but again it is not clear why this Coleman (1965, 1966a). These papers include should be so (Fig. 4) . reference to others which provide details, espe The New Hebrides Block is an area of even cially of the geology of individual islands. The higher seismicity and is, indeed , one of the reader is also referred to the revised regional most active in the world. Most shallow shocks geological map of the British Solomons (British originate below a belt along the western edge Solomon Islands Geological Survey, 1969) . The of the Block. Distribution is fairly even, al summary given here includes observations ob though there is some concentration of foci at tained over the last few years. the Santa Cruz end, where the South Solomon Th e Solomon Islands are an autochthonous Trough abuts against the Torres T rench (see geological system; that is, the rocks present in Fig. 3) . Inte rmediate and deep hypocenters are the area originated within it and are not de more common and more evenly distributed than rived from some nearby separate entity. On the in the Solomon Islands region. They indicate basis of rock type and structure the group can be divided into three provinces: the Volcanic, 3 Th e most recent seismic picture for the New Central, and Pacific provinces (Coleman, 1965) Guinea-Solomons region is given by D enham (Journ al of Geophysical Research, vol. 74, no. 17, pp . ( Figs. 5 and 6) . 4290-4299 ) . The Volcanic Province includes nearly all of CJ (l) 0 , f>' 0- C1Q BOUGUER GRAVITY ANOMALY MAP OF '<: THE SOLOMONS AREA 0 CONTOUR IN TERVAL : 50 M IL LIGA LS ...... rJJ 0 0- S 0 ::l til I>' ::l p... Z (l) ~ :r: (l) ....er ,0' 0.: I (l) til ~:,. I n 0 t""' ttl ;:: > Z
I ,;>0
\
\t. ~ Ill. 0.
FIG. 4. Regional Bouguer gravi ty anomalies in the Solomon Sea area (after Rose et a!., 1968). IV \0 VI 296 PACIFI C SCIENCE, Vol. 24, July 1970
FIG. 5. Geological pro vinces 111 the Solomon Islands, with major faults and distri bution of volcanoes and ultrabasic bodies.
the active and recent volcanoes. It is boomerang volcano) and, with Savo, the western tip of shaped, the western arm trending NW- SE, the Guadalcanal. eastern and shorter arm, roughly W-E. Th e Three sets of faults are found in the Volcanic apex is centered on the N ew Georgia group, Province. In the west, a NW-SE set is dom which is an assemblage of cones linked by their inant, reflected by the elongation of Bougain own extrusives and by fringing and barrier reefs. ville and the disposition of its volcanoes, more' According to Stanton and Bell (1969) , the or less along the axis of that island. Th e eastern lavas in this most typical part of the Province arm of the Province reflects the dominance of range from highly mafic picrite basalts, contain a W-E trending set, parallel to the southeastern ing over 50 percent olivine, to hornblende an side of New Georgia Sound. These sets meet desites. Olivine basalts and pyroclastics are most and are equally impor tant in the New Georgia common, along with basaltic andesites. Stanton group; this was probab ly a controlling factor in and Bell consider that petrogenetically, "the the origin and development of New Georgia . suite appears to have affinities with both mid As well, a third set is conspicuous here. This ocean and island-arc associations, the latter to set trends NE, is probably tensional, and is much lesser degree ." marked by the linear disposit ion of cones. The Eastern New Georgi a Island has a noteworthy fault pattern shown in the Province is a common occurrence of a coarse-grained pyroxene diori te, one throughout the Solomons (Fig. 7). a stock intrusive into the lavas and possibly of The Volcanic Province, in terms of the con Quaternary age (W right, 1969) . There are rela tinuing development of the Solomons, may be tively few active volcanic centers but a great regarded as a youthful extension, a young ver many (over 35) well-preserved cones. Most of sion, of the Central Province. The Central Prov the curren t activity is on Bougainville, where ince has a Mesozoic basal complex (Lower there is an overlap of the Volcanic and Centra l Cretaceous being the older age limit) made up provinces. The lavas and pyroclastics in this part of chloritic "greenstones" and amphibolitic of the Province, as described by Blake and schists, and less altered basic lavas. It also in Miezitis ( 1967), are dominantly hornblende cludes lightly metamorphosed, so-called pelagic andesites, although augite andesites are not un sediments or deep-water oozes and, possibly, common. Basalts are rare. These authors regard small bodies of gabbro ic and granitoid rocks the volcanic suite as being fairly typical of the ( the age of some of these is uncertain ) . On orogenic calc-alkaline suite. The presence of Choiseul, at least, the albite-epido te-amphibolite older diorite, granodiorite, and similar granitoid rocks were metamorphosed in the Lower Eocene rocks in this intruded part of the Central Prov (Richards et al., 1966). ince may have affected the petrography of the This basal complex is intruded and covered younger intrusives. by Upper Eocene- Oligocene intrusives, lavas, The eastern arm includes two active sub and volcanic agglomeratic and pyroclastic layers; marine volcanoes, Mborokua Island, the Russell basalts and basaltic pillow lavas predominate, . group (the northern remnant of a collapsed although andesites are also common. The lavas C) Ibc FLO RIDA G~U P 0- ~,~ WN 'U c=J \JC/ .....AVO.". ~ " ~ ·LAVAS ( SOME VOLCANI CLASTI CS ) QUATERNARY ~ 0 ~ " RE'EF LIM ESTONE ...... I llO NOOI I...... ". ~. I....u.cAHIC. "EONONT . SER' ENTINCUS .. C/l ~IAVO SEDIM ENTS ( UNCONSOLIOATm) 0 RUSSEL L VOLCANICS ( LAVAS ON "l , GEORGiA ) ~ 0- Pl10CE NE S .. WACKES , CALCAREN ITES ('S\f.;,S),J I 0 .: . (. ::J NoVUVU ~~ G ~.... NIC... en M _U Mi OCENE wACKES, CALCARENITES []]]]]]]Ill ~ MIOCE NE ::J c, IEARLY «roc LARGELY REE FA L !!!!Ill ( NOT MALAllA ) Z v- U. CRET. TO Ib CALCrW TlrES. OOZ ES ( MA.l..AITA. ULAwA) 1---1 D..IGOCE NE ::E! ~ ' IU[I" UlTR ABASICS, 'ERPENT!NlTES( OLlG, 1) ::r: ~ Ib LAVA S ~ MAJNLY ANOESITI C ....0- EARLY TERT IARY ILAVAS- MAINLY BASALTIC 1=,=',:;,1 ~ (U. CRET. ON Ib MALAITA I srecss _ DIORlTIC . GABBROIC 01iJ en ~ I \ METAM ORP HICS n UND IF FERENT IATED BASE MENT ~ 0 r- '1010;0110" w l'h C rt '"~ e, " cU. . .. t'l'1 \ 0 1_ Vol cOIl OfJ • ~ :> BOUGA1NVlLLE - SOUTH Z
aOUGAINVllLE ~ .. K" Q "b ~CHO I SEUL
Q ~ . ?o - ...~ NEW GECOG" ~ . ~ - .. ~ "'-'-'~~. . .. : ~..... : ~ o " ow . GlCfUICAI ~ 'fJ ' VANGUNU .. W. ¥ • • A.[NOO VA SAN CIUSTO~ ~T" .A' UIn' ""'''''0 " ~ ~ • TlTlPAlIll t: ~ ,tS' - ",' 1112' FIG. 6. Geological sketch map of the Solomon Islands. tv \0 -...J 298 PACIFIC SCIEN CE, Vol. 24, July 1970
.., ...... "0 .., I J I J '"1
BRITISH SOLOMON ISLANDS
!oH O"TLAHD ~P . STRUCTURAL TRENDS o -.
FOLD AXES
FAULTS / - '!
PROBABLE MA JOR FAU LT
MA JOR REGIONA L FAULT _
WITH IN FERRED EXTEN SIONS
I I I I I 1 ..0 .., '" ~ '" ...... FIG. 7. Stru ctural trends in the Solomon Islands.
may be up to 1,000 meters thick. They are over played the dominant role. On Guad a1canal, the lain by a pile of reefal sediments, pyroclastics, suggestion is that durin g the lower Tertiary the rare lavas, and sediments derived from all of NE and W sets were dominant, the former these and also the rocks of the basal complex. having the greatest strike-slip, dextral, move The quartz content is extremely low, so there ment. In the midd le Tertiary the latter took are no quartzose sediments. The sediments range over, the movement now being sinistral. More in age from early Miocene through Recent. recently, the set trending NW has had a major They were deposited in subsiding fault-bounded shap ing influence, with dextra l strike-slip move troughs, one such trough in eastern Guada1canal ment. The north coast of western Choiseul with accumulating over 5,000 meters of sediment. its off-setting of fault blocks shows this particu The structure of the Central Province is of larly well. horst and graben type, the blocks and troughs This intense fracturing has produced a sys shaped mainly by the interaction between two tematic result only on Guada1canal which has a major sets of faults, one set trending N W (3 10 spine of block-faulted ridges up to 2,600 meters to 330 degrees), the other W (260 to 280 de high, asymmetric in cross section with a steep grees). A third, tensional set of fractures is fall-away to the southern coast. Much of this also important and trends approximately NE coast, as also the coasts of Choiseul and San (40 to 60 degrees) . Cristobal, appears to border complementary, Most faults appear to be normal, with north foundered off-shore blocks. erly dips and with strong strike-slip components Although it is probably true that the Central (Hackman, 1969). Some of these are antithetic Province has been more or less continuously mo to larger faults which are possibly the funda bile since the late Cretaceous, nevertheless there mental shaping fractures, but no simple basic were times of accentuated uplift activity and de- . pattern as yet emerges. The intensity and com formation by faulting. These were, approximate plexity of the fracturing is further complicated ly: late Paleocene, Oligocene, Upper Miocene, by the likely fact that at different times during and Quaternary. Recent work, indeed, suggests the Cainozoic one or other of the fault sets has that much of the uplift which has produced Geology of Solomons and New Hebrides- Co LEMAN 299
Guadalcanal's block-mounta in spine (highest ciated with radiolarian chert bands, there may poin t nearly 2,600 meters) took place in the be up to 5 percent of a complex marine man Quaternary (B. D. Hackman, personal com ganese mineral. munication) . There is at least one relatively shallow-water Th e boundary between the Central and Pa benthonic fauna, of Upp er Eocene age. The cific provinces is marked by a fracture system Upper Miocene and younger sediments show which appears sigmoidal but is actually made up an increasing amount of terrigenous (volcanic) of straight faults, alternating in trend; the two materia l. There are no reefal sediments or lithic trends are again the common ones, that is, terrigenous arenites. Malaita did not experience roughly 310 to 330 degrees and 260 to 280 subaerial erosion before the Pliocene. degrees, respectively. These faults are especially The whole sequence is faulted and (unusual well developed on Santa Isabel ( the Korigole for the Solomons) folded with moderate inten Thrust : Stanton, 1961) and on Choiseul (Cole sity into a series of rather irregular folds trend man, 1962) (again, Fig. 7). ing NW (310 to 330 degrees) and disposed Thi s fracture system is the locus of a serpen en echelon to the run of the island. There is tine belt (there are other serpentine bodies on some cross-folding, usually at about 40 to 60 Guadalcanal which are related to major frac degrees. The folds show cascading and other tures, especially the NE-trending set). The evidence of deformation before consolidation so serpentinous ultrabasics on Choiseul are flat-ly that, in general, the folding is considered to be ing, in places almost horizontal. This relatively of Bmchfalten type, brought about by movement unusual attitu de is explained as the result of of blocks in the faulted basement, during and the foundering of a thrust sheet. The erosion following sedimentation. The faults trend mainly pro ducts of the serpentin ites first appear in WNW (ro ughly 300 degrees) and NE, para llel Lower Miocene sediments, so that an initial to the cross-folding. The general structure and emplacement in the Oligocene is probable. On history is portrayed, as it were in model form, Santa Isabel they were remobilized in the late by the magnificent series of marine seismic pro Tertiary and Quaternary and were easily weath files obtained east of Malaita by Woollard and ered to yield serpen tinous sediment which flows his colleagues (1967). and blankets large areas. Deformation on Malaita has probab ly been The Pacific Province contrasts markedly with intermittently continuous from early Tertiary the rest of the Solomon Block. It shows best on time to the present (topographic and structural Malaita. H ere there are highly basic basal lavas highs coincide), but there were at least two ma (postulated as original seafloor) which are over jor phases, one in the Eocene, another in the lain concordantly, but irregula rly, by deep-water Upper Miocene. foraminiferal oozes of Upper Cretaceous (Se The marg inal areas of Santa Isabel, and, pos nonian) age (Coleman, 1966b). These, in turn, sibly, northeastern San Cristobal and the north are overlain by about 1,300 meters of foraminif western tip of Choiseul, have thinner sequences eral organogenic calcilutites of Cainozoic age, which are generally more contaminated by ter without apparent break. Some of the Eocene rigenous erosional detr itus. sediments were intruded by contemporaneous The present-day proximity of Malaita to submarine lavas. They include highly mafic Guadalcanal, with their severely contrasti ng types akin to alnoite and ankaratrite; similar stratigraphies and structures, poses a problem. exotic lavas have been reported from the Lihir In particular, it is surprising that effluent from group at the western end of the group (Allen the great sedimen t-filled trough in eastern Gua and Deans, 1965) . dalcanal (t he Aola Trough: Coleman and Day, The older sediments consist of calcareous 1965; and Fig. 6) has not registered more clay (up to 80 percent), acid-insoluble clay (up strongly in the Malaitan succession. In genera l, to 20 percent) , and planktonic foraminifera the contrast between the Pacific and Central (up to 20 percent) with occasional arenaceous provinces at least suggests that the two may benthonic foramin ifera, radiolaria, and shards once have occupied different relative positions. of volcanic glass. In some layers, usually asso- The postulate is that Malaita, at least, began a 300 PACI FIC SCIEN CE, Vol. 24, July 1970 westward movement relative to the Central W ithin the Solomons the highest gravity Province, probably in the Upper Miocene. values are found along the southern edges of Taking the Block as a whole, a few other individual islands (M alaita is an exception : points are worth emphasis. The Pacific Province here, the gravity highs tend to coincide with has the oldest rocks, the Volcanic Province the lava cores of the larger anticlines). youngest. There is a grada tion in age of lavas The distribution of the more obvious ultra from the Pacific flank to the Austral ian one, basics is clearly indicated by high anomalies, as each Province having distinctive lava suites. The also of the associated major fractures- for exam oldest (early Miocene) terrigenous sediments ple, the Korigole Thrust on Santa Isabel and are found widespread in the Central Province. the great fracture zone which defines south The problem of source areas for these great vol eastern Choiseul and continues east to Santa Isa umes of sediment, filling deep troughs in the bel (defining, incidentally, the northwest edge Guadalcanal area, is a difficult one. Th e present of the N ew Georgia Sound rhombochasm) . mountain axis to the south of the island is too In a general way, the oldest rocks on the vari young to have served as source. San Cristobal, ous islands are marked by high gravity values essentially a planed-off lava block, has been sug (exceeding 200 mgals on Guad alcanal) , despite gested as providing the great volume of vol the usual high topographic position of these caniclastics required (Coleman, in Rose et al., rocks. Thi s may result in extremely high gra 1968) . The re is no real evidence for this nor dients. For example, in eastern Guadalcanal the for the earlier suggestion that the source area is difference in levels between the base of the sedi now foundered in deep water off southern mentary section in the Aola T rough (with per Guadalcanal (Coleman, 1965). Despite its con haps 6,000 meters of sediment ) and the top of notation of "yo-yo" tectonics, the earlier idea is the adjacent ridges of basement rocks (onto again favored: there is increasing evidence that much of which the sediments lap) is more than the Solomons-New Hebrides region is one in 8,000 meters. Thi s difference is expressed over which there have been extremely rapid vertical a horizontal distance of about 25 km. The movements, exceeding 3,000 meters, of large gravity gradient here is one of the highest ever crustal blocks. Th is idea, incidentally, highl ights recorded. the importance of futur e detailed bathymetric The composite-block character of the islands studies around and about the islands, for this is well reflected, especially by the gravimetric movement is probably still going on. pattern on Choiseul. The collapsed volcano com The correspondence between the geology and plex of which the Russell Islands are now the the pattern of gravity values is generally close. remaining northern part is also indicated clearly. On a regional scale, the Solomon Block is a Other correlations between geology and gravity gravity high with superimposed positive anom are well illustrated in Grover ( 1968) and in alies of up to 250 milligals on land. These sug Laudon (1968). gest that the Block is in a state of isostatic imbalance with great vertical displacement of GEO LOGY OF THE NEW HEBRIDES crustal and subcrustal blocks. New Georgia Sound is marked by a comparative low, suggest The regional geology of the N ew Hebrides ing, as might be expected, a considerable thick Archipelago is described in a paper by Mitchell ness of sediment ( 2 km : Kroenke, in Rose et and Warden (1 970) , to which the reader is al., 1968) . The largest land values are found referred for particular detail. along the trench or the Australian margins of The N ew Hebrides Block, as with the Solo the southern islands. These values increase still mons, is an autochthonous geological entity further to the south, and their distribution is although there is some slight evidence suggest reflected by the bathymetry and its suggestion ing a closer association with the Fiji group, in of a SW-trending connection, between the mid the early Tertiary, than would appear from their dle part of the Solomon Block and the Louisiade present geograph ic separation. It also is intensely Pocklington-Rennell Ridge, by a set of sub fractured, the islands having been shaped by marine horsts and graben (Figs. 1 and 4 ). faults. Folding is minor and incidental. The Geology of Solomons and N ew Hebrides- CoLEMAN 301
elongated shapes of the islands reflect the dom chain, W arden ( 1967) stressed this variability inance of a fault set with strong meridional in composition both in space and time. In gen trends, acting along with a W-E set. The dis eral, he concluded that the basal par ts of the position of some of the minor volcanic cones volcanic sequences are essentially basaltic, pass and the trends of magnetic anomalies suggest ing up into mainly andesitic pyroclastics and that the latter is a tensional set and of special flows. Recent eruptions at several vents included interest because it so clearly cuts across the dacitic pumice and clasts of rhyolitic composi "grain" of the Block. This, in turn, may reflect tion. Although not all volcanic piles show it, a subcrustal arching of the Block. there is thus a tendency for the volcanics to Alth ough lacking the clear-cut character of become less basic with time. the Solomons subdivision, the New Hebrides There are at least eight active centers: Tina Block is made up of three areas of distinctive kula (Santa Cruz group) in the north; Vanua geological style : the Western Belt, the central Lava and Gaua (Banks group); Ambrym (a volcanic chain, and the Eastern Ridge. splendid caldera) and Lopevi; two submari ne The Western Belt and Eastern Ridge are the volcanoes-one off eastern Epi, the other off oldest. Their history began probably in the early northeastern Erromango ; and Tanna in the Tertiary but most of the rock successions are south. T inakula, Vanua Lava, Gaua, and Am younger than Oligocene and, in the Wes tern brym have passed through a caldera-forming Belt, have close similarities with those of the stage; Ambrym is a fine example (Stephenson Cent ral Province in the Solomons. The central et al., 1968) . Yasour, on Tanna, is a very volcanic chain is the equivalent of the Solomon young cone, but it represents the recrudescence Volcanic Province and also had its beginning in of an older extremely large volcano which ex the Pliocene. There is no equivalent to the ploded paroxysmally. Lopevi and the submarine Pacific Province. vents are young and at the cone-building stage. The central volcanic chain is made up of the Some of the associated inactive volcanoes of the Santa Cruz (i n par t only) and Banks groups in centra l por tion of the chain have dissected cones the north ; Aoba, Ambrym, and the Shepherd but most are well preserved : Aoba is an essay gro up in the center; and Efate, Erromango, in vulcano logical features (Warden, 1968), and Ta nna, and Aneityum in the southern "tail." has a marked E-W elongation as well (Fig. 8). These are volcanic islands and island clusters The central chain is intensely fractured. The disposed along a line parallel to, but a littl e east major fractures are aligned roughly NNW-SSE of, the axis of the New Hebrides Block. and E-W, the volcanic vents being disposed As in the Solomon Volcanic Province, the along one or the other of these, with a con larger islands are made up of coalesced accumu centration of activity where they intersect. Mala lations of volcanic effusives and detritus and hoff's aeromagnetic data- show this relationship reefal sedimen ts emergent from sea depths as clearly, the influence of the E-W trend ing frac great as 2,000 meters. The older of these have tures standing out par ticularly well. The princi been block-faulted and partly upl ifted , resulting pal volcanoes in the central chain are associated in horst and grabe n structures. The smaller with bipole magnetic anomalies of up to 1,000 islands are usually single volcanoes and consist gammas and suggest the presence of magma entirely of subaerial extrusives. chambers relatively near the surface and having In the nort hern part of the chain the volcanics a tendency to E-W elonga tion. are dominantly basaltic, those of Utupua and Unlike the Solomons, the volcanic central Vanikoro (southern Santa Cruz group) includ chain is not marginal but is placed between two ing ankaramites and olivine-labradorite-basalts. areas of differing and older geology. It can, In the central portion and south they are more nevertheless, be considered a version of the variable and include, as the main representa Wes tern Belt as the latter was in the early tives, feldsparphyric olivine basalts and basaltic M iocene: just as the Solomon Volcanic Province and hypersthene andesites. Some eruptions have is a youthful extension of its Centra l Province. included both basaltic and andesit ic rocks at the same time. In describing the central part of the 4 Ma lahoff, in prep aration. 111 7".0' 117 ".' 1.7"'0' E• •' r:Jf G' ""wICh Ilr,,' 168 0 00 ' \,JJ o ,.. ~ N GEOLOGICAL SKETCH MAP C ~ OF o I AOBA '~ c 1 1 C I I f ,.. I ~~ r L.(S ~: I
V r tJ s 0
LEGEND ~'~ I r D R.cent ••h d . po li h . I "• M A I N C".T IIlA L VOLCANO "0 ~P o .t ~ c . ld . , . p)" lcl••t •• , >n -"!'l n 0..... 11 • - lItock _.. nC/) t'!'l ItOi OlfCAl (, .." 'ou) - ~ ------
Santa Cruz Island, the Torres group, Espiritu The basal volcanics are overlain disconform Santo, and Malekula are the islands of the ably and unconformably by a sedimentary pile, Western Belt. Santo is the largest island in the perhaps over 3,000 meters thick and varying in Archipelago and may serve as a model for the age from early Miocene through Qu aternary (as geology of the Belt (Robinson, 1968; 1969). It with the Central Province of the Solomons). is divided neatly into two tectonic elements, the There are no quartzose sediments. The older western mobile element and the eastern shelf sediments are calcareous, dominantly algal/ (F ig. 9). foraminiferal reefal limestones (n ot coralline) , In western Santo, the oldest rocks are pre which are overlain by, and in the younger parts Miocene andesitic and basaltic lavas (o ften pass laterally into, wacke-type sediments and autobrecciated), flow breccias, pill ow lavas, and volcanics. The wackes, in volume and area, are occasional pyroclastics. An older limit to these the dominant sediments on Santo and in places volcanics has not been set, nor whether they are occupy most of the column. They span the truly basal. If they are, the older age limit is Miocene and were clearly derived from rapid probably Eocene. Associated with the volcanics erosion of a mountainous source area similar in are stock-like intrusions of dioritic and gabbroic composition to the basal volcanics. rocks which may be marginally younger. The Miocene sedimentary sequence is charac-
LEGEND
~Uplift ed r eef Qua ternary ~ platforms
Pliocene I t~~:1 Calcilutite
L.M iocQnQt o rrmm InterroQf L.PliocQnQ W!lW tro uqh sees.
' o c ~ n ~ ~ A l t Q m a t i ng beos M . M I ~ ~ ~ l st n Q. vol c ·s . q'wcckes,
L :-M.Miocene l:·.~~ .:.; 1 Greywocke
Ear ly to M.M iocene B Reef complexes
Olig. rv:-::lVolcanics s,: MMiocenQ ~ int r usives
-- Faults
5 15 5 o 10! 2,0 I ~ I score. miles
SKE TCH MAP
GEOLOGY O F
E SPIRITU SANTO.
NEW HEBRIDE S.
FIG. 9. G eol ogical sketch map of Espir i tu Sant o, W estern Belt, New H ebrides ( after Robinson, 1968) . 304 PACIFIC SCIENCE, Vol. 24, July 1970 terized by complex facies changes. Particularly They may represent a distinct set of intrusio ns in the central part of the Cumberland Peninsula, . ( one such body is reported to be serpentinized) . the dominant wackes repeatedly pass laterally The Western Belt is intensely fractured by into andesit ic volcanics including interbedded systems of faults . Folding is incidental and flows. Elsewhere, especially on both sides of the minor. The faults are in three main sets, trend Peninsula and in southwest Santo, the wackes ing roughly N, NW, and W, respectively. become more calcareous and organogenic, with Those with Nand W trends seem to have been occasional irregular reef bodies (probably mar the dominant shaping sets. ginal to the old volcanic highs) . Between these Interaction between the faults resulted In marginal reefs, the sequence is one in which differenti ally uplifted blocks, tending to be beds of calcarenite alternate with calcilutite beds elongated meridionally, and the formation of rich in planktonic foraminifera, indicating free horst and graben. At least until the Middle access to open ocean. These off-shore, inter -reef Miocene some of the highs were centers of beds range through the Pliocene but then give vulcanism. Movement of the blocks was con way, probably unconformably, to a Plio-Quater tinuous throughout the Tertiary, so that the sites nary coralgal reef limestone which covers part accumulating sediment were not constant but of western and all of eastern Santo. shifting. In a general way, the Western Belt was The volcanics mentioned, associated both with built up from west to east. the oldest, early Miocene reef limestones, and This picture of sedimentation from local the Miocene greywackes, represent vulcanism sources into transient fault troughs explains the and intrusion intermittent through the Miocene. complicated facies changes, lateral and vertical, They probably represent the sporadic continua and rapid thickening and thinning of individual tion, in time and space, of the pre-Mi ocene form ations, so characteristic of the Santo column vulcanism. (see also the Malekula section, Fig. 10) . The dominant lava types are feldsp arphyric The eastern shelf of Santo is covered by basalts and andesites, the basalts tending to bulk Pliocene ( ? Upper) and Pleistocene reefal, flat larger in the pre-Miocene and early Miocene lying sediments which have lapped onto mar parts of the sequence. Massive flows and intru ginal areas of the western mobile element. W hat sions are not volumetrically domin ant. The most underlies these sediments is not known, but common are autobrecciated lavas, which often Robinson (1968) suggests a basement of vol pass laterally into volcanic rudites, then into canics of late Oligocene or early Miocene age, peperites, and so into volcanic lithic calcarenites. that is, an extension of the basal rocks of the Fringi ng-reef bodies may also be a part of this western element. Most of the upli ft proceeded gradation from volcanic to shallow marine shelf during the Pleistocene and continues, as evi environments. The coral content of the reef denced by the high incidence of shallow-focus masses does not become prominent unt il well earthquakes today. The uplift has been in stages, into the Pliocene. the periods between being represented now by As Robinson emphasizes, there is great varia exceptiona lly well developed reef terraces. tion in the chemical and mineralogical content The fracture pattern over the eastern shelf is of the lavas and associated volcanics, so that essentially the .same as in the west, but Nand only in a general way can they be described as NW trends are dominant. calc-alkaline. The hornblende diorites of south The geology of Malekula, the second largest western Santo are more definitely calc-alkaline. island in the Archipelago, is essentially similar There are several large, stock-like bodies of to that of western Santo (Mitchell, 1966; these rocks. They range in probable age from 1968) . In the basal (early Miocene) part of the late Oligocene throug h Miocene. Mitchell and sequence, flows and pillow lavas are rarer than W arden (1970) relate them genetically to the they are in Santo, but thick, turbiditic, coarse andesitic volcanics. volcanic arenites and rudites are more common, The gabbros occur in small, isolated, usually as also are intrusive dikes and sills. Massive fault-bounded bodies which have not been suc reef bodies are not as well developed (Figs. 10 cessfully related to the associated volcanics. and 11) . Geology of Solomons and New Heb rides-CoLEMAN 305
The turbiditic greywacke sediments are es stones, and minor limestones deposited in a pecially prom inent in the Middle Miocene, but fault-trough and inter-reef environment. Al they are not as regularly interspersed with plank though heavily eroded, a flat-lying Quaternary tonic foraminiferal silts. Mitchell (1966) places plateau limestone, similar to the Santo occur a major unconformity, not apparent in Santo, rence and with equally fine terraces (Mitchell, between Middle Miocene and Pliocene sedi 1969), completes the Malekula succession. ments. The latter consist of conglomerates, sand- There is no equivalent to the eastern Santo shelf.
DIAGR AMMATIC CRO SS-SECTION THROUGH SOUT H M AL E KU LA ILLUSTRATING PROBABLE GEOLOGICALHIS T OR Y
(V t RTle"'lElAGGERATIO"lII3)
10,000:
L AT E L O WER MIOCE NE
( 0) EA RLY LOW ER MIOC EN E
LEGEND
W," ' UA ~Reefllme SI O"e FORMATION ~ ~ Volc.anidutic sed"" e nh PLIO AND <) PLEISTOCEN E PLATEAU "'AUNUI Ca lcAreo ..., sa nd'lone and muds tone } Pyroelu tit ? de pOSit, . 1::;::,:;::1 LIMESTONES i Wi GROUP ) Reef and de t rital liMltslone. ~=~ San d , tone end muddone ...... in1y ep,el!lltie,} ,, "'1B MIDDLE - - POR T ; ~ MIOCENE ~ S" NDWICH 9 ~ Hor nblende di oert e s and al'loe,,'. s. ..:~ Reef And det rital I,me !olonl' FORMATIO N
-- Fault
FIG. 10. A scheme of development for South Maleku la, Western Belt, New Hebrides (from Mitchell, 1967). 306 PACIFIC SCIENCE, Vol. 24, July 1970
Mitchell and Warden (1970) quote thick similar values result if the maximum thicknesses nesses of 6,000 meters for the Lower Miocene, of all units are added, but after allowing for 2,300 meters for the Middle Miocene, and over likely repetition by faul ting and rapid lateral 500 meters for the Plio-Quaternary sequences. th ickening and thinning of each formation, a The first two values are remarkable. In Santo, total of 3,000 meters seems more likely.
SKETC H MAP GEOLOGY OF SOUTH MALEKULA
LEGEND
RECENT[ill] Reef and alluvium.
6J' Uripiv Isla n d . P LI O-PLEISTOCENE PLATEAU LIMESTO NES &wlNTUAFORMA TION. M IDDLE MI OCENE PORT SANDWrCN FORMATiON. +5, 000'
LOWER M IOC ENE MU ANUI GROUP ;2 0.000'
~ Zone of hornblende andesit e end ~ d iori t e int rus ions.
- - -- Geolog ical boundary.
-- Fault. ___<.:: River.
be a rin g and e sit ic dy ke s . I S' Dip. : ~. S $
Dix on Reef.
S CALE
OR....w t.! 8 '1' It NAIlAY" '"
FIG . 11. Geological sketch map of South Malekula, Western Belt, N ew Hebrides (from Mitchell, 1967) . Geology of Solomons and New Hebrides-Co LEMAN 307
The Torres group, to the north of Santo, has steep-sided slabs shaped by faulting, and have not yet been surveyed. Preliminary work sug much of their central and northern areas ob gests that the islands consist of andesitic lavas scured by a probable Quaternary, reefal lime (one of which gave a radiometric age of 39 ± 5 stone cover, upl ifted over 800 meters in places. m.y.) and pyroclastics and some marine volcanic Det ailed work is still going on (Liggett, 1967; arenites, possibly of Lower Miocene age. Mallick, 1969) (Fi gs. 12 and 13). Santa Cruz Island deserves special mention Maewo has the oldest sediments so far dated for , until very recently, it, along with the other in the eastern belt. These are early Miocene islands of the Santa Cruz group, was considered turbiditic wackes and conglomerates, possibly to be a Qu aternary volcanic center. It has now overlying basalts, with clasts of varying litholo been surveyed by the Solomon Islands Geologi gies, including volcanics and Upper Eocene lime cal Surve y, and a large area and volume of stones (remarkably similar in fauna and lith marine volcanic and calcareous sediments have ology to those found in the W ainimala Group been discovered. Samples of these sediments of Viti Levu, Fiji) , neither of which has been have been examined by the writer and found to found ill sit» anywhere in the New Hebrides. range back in age to the early Miocene. With These old sediments crop out along the eastern the Director's permission, P. M. Craig, Solo coast. In the center and north of Maewo, the mons Survey geologist, has generously provided basal sediments, perhaps 800 meters thick, give the following preliminary summary of the way to finer grained wackes and calcilutites of geology: about the same total thickness and of Miocene age. The exposed volcanic basement of Santa Cru z Is The voIeanic content increases, in a general land consists of two adjacen t piles. In both, re worked pyroclastics which conta in an early Mio cene way, to the south, for over the southern half of microfaun a are predominant wit h basaltic lava the island there are large areas of basaltic lavas, flows becoming less common upwards. The flows of breccias and associated pyroclastics, and inter the eastern pile tend to be more numerous, thinner , beds of calcilutites. Both massive flows and pil and mor e porp hyritic ,Overlying both piles are cal low lavas occur; dikes and other intru sions are careous sediments. The western pile passes rapidly up into a uniform succession of limestones and fairly rare. These volcanics are at best several mudstones. The eastern pile, however, passes grad thousand feet thick and, while the bulk of them ually upwards into calcareous grits which, in tu rn, is probably Miocene, the oldest may pre-date pass upwards into finer grained, more calcareous the Miocene and thus form the basement rocks horizons. An unconformable mantle of reef lime stone obscures much of the eastern pile, the coasta l of the island. margins of the weste rn pile, and forms a large up The Pliocene (about 300 meters) is repre lifted plateau still further west. sented by calcilutites rich in planktonic forami Vert ical faulting appears to be the main struc tura l nifera, and calcarenites which appear to be con control, there being little evidence for lateral fault formable on the Miocene. Intrusions of basalt moveme nts or of folding. D irections of major faults tend to parall el the long and short axes of the have been recorded . To the north, these pass island. into banded beds with an increased terrigenous content. The early Miocene (Tertiary upper 'e") fauna The Quaternary limestone cover (up to 330 referred to is the " Spiro clvpens-Enlepidine" or meters thick) has an irregular base, probably "San Jorge" fauna (Coleman, 1963), wide of varying age, but over large areas it takes on ranging from northern coastal N ew Guinea the form of a single sheet. It is coralgal, but the through the Solomons, New Hebrides to Fiji. coral is scattered and there do not seem to be The eastern pile on Santa Cruz Island appears any reef frameworks structured in coral. Apart to have a more complete section than the west from its obscuring physical presence, it is re ern, includ ing younger fauna. The dominance sponsible for much mantling of the older rocks of faults trending Nand W fits the general with travertine. It is obviously but a part of a New Hebrides pattern . much larger original area of reefal carbonate Th e Eastern Ridge consists of Maewo and sedimentation. Pentecost islands. Both are narrow, elongate, The sedimentary succession has a region al 308 PACIFIC SCIENCE, Vol. 24, July 1970 westerly dip of 20 to 30 degrees, remarkably degrees approximately, the other westerly, 265 consistent in view of the fracturing and uplift to 290 degrees. Great scarps with these orienta which have taken place. The fracture pattern is tions are spectacular features of the present made up of two sets of near-vertical normal topography. faults, one with a northerly trend, 350 to 020 The geology of Maewo suggests that it was
168'"E
GEOLOGICAL SKETCH MAP OF MAEWO
SC".LE MIL.ES .2.' -,==",",---====2, 4' MILES
LEGEND
RECENT 1\~jW~ri@ ~1 Holocene deposits.
_ Nasawa beds. ? PLEISTOCENE { g Limestone...... -.. UNCONFORMIT Y. ? MIOCENE ~ Betarara beds.
~':=-=~j Lolaruk beds.
I .; f..; ;.~~q... Ker embei beds.
MIOCENE
-- Geological boundary. ___ Major scarp.
-""C:::: Ri ver or c r e ek.
... Prominent peak.
R .N. GOPA l .
FIG. 12. Geological sketch map of Maewo, Eastern Ridge, New Hebrides (from Liggett, 1%7) . Geology of Solomons and New Hebrides-CoLEMAN 309
GEOLOGICAL SKETCH""'MAP OF SOUTHERN PENTECOST
Bay
Ranbout:or
Banmatmat
seA LE o I 2 : : ,
LEGEND RECENT k;i./J Holocene depo si ts. -- Geological boundary. ?PLIO - p:;:r:;:q . -- Fault. PLEISTOCENE b::C:d Lime s-rone, Contours at 200m, ?PLIOCENE Volcan:-c1astic s ediments . 1·.·.···1• • • . and pillow lavas. -- int e rv als . ~ Peridotite and ~ Fr i;'ging reef. ~ serpentinite. 168·1~'E .. ViliaSe.
FIG. 13. Preliminary geological sketch map of southern Pentecost, Eastern Ridge, N ew Hebrides (from Mallick , 1969) . once a part of a larger entity. The eastern part (1969) . Southern Pentecost is especially note has foundered ; the western edge has been up worthy because it alone of the New Hebrides lifted , not as a single slab or sliver but as a Islands has large areas of ultrabasics. These set of blocks at different rates, as reflected in the rocks are largely enstatiteolivinites and are ex irregul ar base of the Quaternary limestone and tensively serpentinized (there is a record of a the larger exposures of lavas in the south. The small serpentinite body in central west Santo) . uplift probably began in the Pliocene but most They are similar to the Solomon Islands ser of it took place in the late Quaternary and may pentinites, especially those of Santa Isabel and still be proceeding. Choiseul, and like them, they weather rapidly Pentecost (or Raga, the indigenous name) is to a clayey serpentinous sediment which drifts something of an anomaly. The oldest rocks are readily and obscures the relationships of the exposed in southern Pentecost. A provisional ultrabasics to the associated rocks. The latter statement on the geology is given by Mallick consist of submarine volcanics (apparently in- 310 PACIFIC SCIENCE, Vol. 24, July 1970
truded in part by the ultrabasics) which include is shown separate from the eastern shelf. Robin submarine basaltic lavas, breccias and lith ic son's postulate of a basement volcanic complex volcanic rudites, and arenites. The finer grained beneath the shelf is supp orted . The increasing of these sediments may range into calcarenites thickness of sediment to the east of western and, especially, calcilutites. There are inter Santo is well shown . spersed beds of calcilutites which are extremely Th e intensity of the block faulting is not re rich in planktonic foraminifera (amounting to flected, due prob ably to the wide spacing of the pelagic oozes) but have little volcanic material. gravity measurements, but the fracture trends There are lavas as well, particularly pill ow lavas, are suggested, especially the north and north which occur at various levels within the sedi westerly ones. Some of the volcanic centers are ments. The oldest sediments so far dated are clearly delineated by local anomalies, for ex Upper Miocene in age. Some of them include ample, Lopevi and Tongoa. The graben structure serpentinite detritus which puts the age of em of Efate is indicated, as is also the remnant char placement of the serpentinites before this time. acter of both Erromango and Tan na. No Lower Miocene terrigenous sediments of the On the Eastern Ridge, the highest values are Maewo or Santo kind have been found up to the found along the eastern margins of Maewo and time of writing, although the field work (by Pentecost. If the Solomons are a guide, still Mallick, on a 1: 50,000 scale) is now nearly higher anomalies will be found further out to complete.s sea. The ultr abasics on Pentecost are only vague N orthern Pentecost is only now being sys ly indicated, but areas of maximum Quaternary tematically surveyed. It includes a similar basal uplift on both Maewo and Pentecost relate well mixture of basaltic volcanics and calcilutites rich to the gravity pattern. in planktonic foraminifera, indicating an Up per In the Solomons, the gravity values rise steep Miocene or younger age. ly as the southwestern trench areas are ap The fracture pattern on Pentecost is similar to proached. The N ew Hebrides cover is sparse, that of Maewo, and is shown best over the but shows a parallel case in the Torres Trench southern half where large near-vertical fractures, area to the north. Perhaps significantly, the trending 0 to 20 degrees, are especially con western central portion of the New Hebrides, spicuous. occupied by Santo and Malekula, is a compara It would appear that Pentecost has experi tive gravity low: the offshore area here lacks a enced little subaerial erosion during its develop trench, separating as it does the Torres and New ment and that it has had only one major episode Hebrides deeps. of uplift, probably during the Quaternary. For the New Hebrides, the gravity picture is much less detailed; shipboard studies have been OUTSTA N DIN G FEATURES, COMPARISON S, AND confined to the northern end, and only over the DEVELOPMENT central portion of the group may inter-island As composite fractured island chains, the extrapolations be considered. Nevertheless, there Solomons and the New Hebrides are highly are some interesting results (Malahoff and distinctive entities and as such are difficult to Woollard, 1969). compare meaningfully with other island arcs, The Block is also a positive high with land outside the Melanesian Re-entr ant. The Carib values up to 200 mgals, the largest tending to bean complex may be an exception: Donnell y's be on the eastern, Pacific edge, the lowest, on description of the regional geology of the Virgin the western, continental side, in direct contrast Islands hints at a parallel (Donnelly, 1964 ) . to the Solomons. In the W estern Belt, Santo Discussion of this and much closer parallels and Malekula show a N-S polarization of high within the Melanesian Re-entrant is outside the positive anomalies, with a pole-in-common be purpose of this paper and is left to another tween the two islands. On Santo, the western time. part, in particular the Cumberland Peninsula, The Solomons and N ew Hebrides have the following noteworthy characteristics in common : 5 April 1970 : These sediments have now been found. 1. Both are crustal blocks, 20 to 30 km in Geology of Solomons and New Hebrides-CoLEMAN 311 thickness, and isolated from neighb oring blocks. have strongest affi nities with the so-called cir Both are autochthonous geological systems which cumocean suite. began their histories in the late Mesozoic or 7. Serpentinites, after eustatite olivinites and earliest Tertiary. Their generalized stratigraphic hartzburgites, occur in both blocks, the larger columns are remarkably similar and complete, and better defined area, amounting to a serpen without major break. These do not account for, tine belt, occurring in the Solomons. nor suggest the nature of, the remaining major 8. Granitoid rocks of diorite type, ranging in part of the estimated crustal thickness of each age from Oligocene to Plio-Pleistocene, are com block. The transition from oceanic crust to that paratively rare. There are no granites nor quartz composing each block appears to be more ose sediments . abrupt on the trench side than it is on the north 9. Terrigenous sedimentation, widespread in east or Pacific side. both areas, began in the early Miocene and was 2. "Reverse" island arc features are shared in fairly continuous thereafter. Although algalf common : deep linear trenches on the continental foraminiferal reefal sediments are a common side (not continuous but with a break in the element in their sedimentary columns, ranging middle); general absence of arcuate features; back to the early Miocene, the coral content is abnormally placed Holocene volcanic belts; sub surprisingly sparse; coral-structured reef masses marine platforms on the Pacific side (Ontong do not become at all prominent until the Qua Java Platform and Fiji Plateau); "planes of ternary. seismicity" which are either vertical or dip steeply away from the continental side. These similarities are important; but there are 3. They are areas of high gravity values, with differences: large superimposed positive anomalies, on the 1. The Solomons have a folded terrain-the flanks of an even greater regional high which is Pacific Province-with a highl y individual strati covered by the Coral and Solomon seas. They graphic succession, including at least 1,000 are highly seismic, the New Hebrides Block meters of late Cretaceous to Lower Miocene, especially so. The distribution of hypocenters is deep-water organogenic oozes. There is no direct irregular, more so in the Solomons, with a sug equivalent in the N ew Hebrides ; Pentecost, with gestion of a gap between about 200 and 400 km. its comparative lack of terrigenous sediment and 4. Deep fractur ing is the dominant structural its essentially submarine history, comes nearest. theme, and taphrogenesis, large and small scale, 2. The distribution of hypocenters in the N ew is its expression. Differential uplifts of up to Hebrides is more regular and, in a sense, more 6,000 meters over short lateral distances are "normal," than it is in the Solomons. recorded . These are associated with extremely 3. The Solomons show a marked gradational high gravity gradients. In general, surface ten development from the Pacific to the Australian sional stress seems to have prevailed during the margins . The N ew Hebrides lack this. Tertiary and Quaternary. The geologically re 4. Strike-slip movement along faults is more cent expression of this stress is a set of tensional evident in the Solomons. But the vertical differ faults which trends normal to the axis of each ential movement between blocks, amounting to block. several thousand meters, within a time span of a 5. Both blocks-at least over their central few million years, is shown rather better by positions- are linked with the Louisiade-Pock New Hebrides examples. lington-New Caledonia "high" by a complex of 5. The Solomons experienced at least a mild submarine features suggestive of horsts and form of regional, shearing, metamorphi sm in graben. the early Tertiary. Regional metamorphics have 6. The volcanic rocks include a great variety not been found in the N ew Hebrides. of basic and intermediate lavas, including highly 6. The New Hebrides stratigraphic column mafic types such as picrite-basalt, ankaratite, does not include pre-Tertiary sediments. The alnoite, and ankaramite. Basalts and andesites early Miocene sediments, however, include de are in about equal bulk, the oldest being the rived Upper Eocene shallow-water calcarenite more basic. Petrogenet ically, the lavas appear to clasts which could not have been transported 312 PACI FIC SCIENCE, Vol. 24, July 1970 far. There is no shallow-water Upper Eocene ocean. Great uplift of sediment-filled troughs sediment in the Solomons. and foundering of eroded blocks, togeth er with 7. The highest positive gravity anomalies in strike-slip movement (dextral) along NW-SE the Solomon land areas are found on the trench trending faults (most marked in the Solomons) or continental flank. In the New Hebrides they has taken place in the Uppe r Pliocene/Quater occur on the Pacific or Fiji Plateau side (further nary. This was also the time of marginal accre detailed gravity work in the New H ebrides may tion around most islands with upl ift of reef remove this contrast). slabs; in the Solomons there was a shift of The similarities quite outweigh the differ intensified vulcanism to the southwest, and the ences, so that the same generalized scheme of intrusion of dioritic stocks; in the New Hebrides development will serve for both areas (the New there was intensified vulcanism along the axis of Hebrides lagging behind the Solomons in the the Block. beginnings of the earlier major events) . The Pacific Province has its own history. It They began their history with the submarine began with crustal arching , block faulting, and extrusion or injection of basaltic lava piles be intrusion/extrusion of basaltic lavas beneath neath a cuirass of deep-water organogenic oozes deep-water, Upp er Cretaceous (Senonian) pe during late Cretaceous (early Tertiary in the lagic oozes. Th is was followed by accumulation New Hebrides) . The relationship of this em of the same sediments with faulting of the lava placement to any underlying (? ocean floor) pile and initiation of folds of the Brucbjnlten rocks is unknown. Build-up by way of deep type. There was then minor intrusion of highly tensional fractures continued, producing rhorn basic lavas and carbonate sedimentat ion (Eocene boidallaval welts elongated more or less parallel to Upper Miocene; up to 1,000 meters) most of to the axis of each block: in terms of their it deep-water, but with at least one shallow orientation today, this is NW-SE in the Solo water intercalation in the Eocene. The Eocene mons, N-S in the New Hebrides. By the Lower was also a time of emphatic fault-block move Eocene, some of the Solomon blocks were above ment and widespread production of Bm cbjalten. sea level and receiving pyroclastics. A large part Th e terrigenous (volcanic) content of the sedi of the Solomons was then regionally meta ments increased in late Miocene and Pliocene, morphosed (albite-epidote-amphibolite facies and there was yet another major episode of up but no evidence of this in the New Hebrides). lift and folding in the late Tertiary. Mobility Further vulcanism, upward movement of gabbro continued to the present, with some cross-fold and peridotite, differential uplift of blocks and ing and faulting, so that topographic and struc intrusion by stocklike masses of acid segrega tural highs are coincident. The shift of this tions from andesitic magma, all took place during Province, probably from the east relative to the approximately Oligocene time (peridotite em other parts of the Solomon Block, is thought to placement in the New Hebrides was probably have begun in the late Miocene. later). This was followed by a quiet stage with The possible shift of both the Solomon and extensive build-up of reef and reefal sediments New Hebrides blocks, from positions in an (not especially coralline) in the Lower Miocene outer arc bordering continental Australia in the (accompanied by vulcanism in the New He early Terti ary, is briefly suggested in an earlier brides) . Serpentinites were first exposed to paper (Coleman, 1967) . It was prop osed that erosion in the Solomons at this time. Th ere was this arc is now represented by the rim of the then renewed vulcanism and erosion of elevated Melanesian Re-entrant. blocks, with major uplift probably in the Mid dle/Upper Miocene, to produce great thick LITERATURE CITED nesses of wacke-type sediments with occasional ALLEN, J. B., and T. DEANS. 1965. Ultrabasic still-stages permitting the growth and accumula erupt ives with alnoitic-kimberlitic affinities tion of reefal sediments. The latter have in from Malaita, Solomon Islands. Mineralogical creased coral content and a considerable admix Magazine (London), vol. 34, pp. 16-34. ture of warm-water planktonic foraminifera, BLAK E, D. H. , and Y. MIEZITIS. 1967. Geology suggesting free access of island waters to nearby of Bougainville and Buka islands, New Geology of Solomons and New H ebride s-CoLEMAN 313
Gu inea. Australia Bureau of Mineral Re southwest Pacific. Publications du Bureau sources, Bulletin 93 (PNG 1), 56 pp . central sismologique interna tional, 1960A, no. CAREY, S. W . 1958. The tectonic approach to 22, pp. 137- 149. continental dr ift. In: Continental drift-A FURUMOTO, A. S., D. M. H USSONG, J. F. symposium, pp . 177- 355. Geology Depart CAMPBELL, G. H. SUTTON, A. MALAHOFF, ment , University of Tasmania, Hobart. J. C. ROSE, and G. P. W OOLLARD. 1970. - -- 1963. The asymmetry of the earth . Crustal and upper mantle structure of the Australian Journ al of Science, vol. 25, pp. Solomon Islands as revealed by seismic refrac 369-383, 479-488. tion survey of N ovember-December 1966. COLE MA N, P. J. 1962. An outline of the geology Pacific Science, vol. 24, pp . 315-332. ( In this of Choiseul, Solomon Islands. Journal of the issue.) Geological Society of Australia, vol. 8, pp. GROVER, J. C. 1968. The Brit ish Solomon Is 135-158. lands: Some geolog ical impli cations of the --- 1963. Tertiary larger foraminifera of gravity data, 1966. In: The crust and upper the British Solomon Islands , southwest Pacific. mantle of the Pacific area, pp. 296-306. Geo Micropaleontology, vol. 9, pp. 1-38. physical Monograph 12, American Geophysi --- 1965. Stratigraphical and structural cal Union, W ashington, D.C. notes on the British Solomon Islands with - -- 1969 . Record of gravity, magnetic, reference to the First Geological Map, 1962. bathymetric, aerial and crustal surveys in the British Solomon Islands Geological Record, British Solomon Islands, 1963-1966. British vol. 2 (1959-1962) , pp . 17-31. Geological Solomon Island s Geological Record , vol. 3 Survey, H oniara. (1963-1967) . In press. - -- 1966a. The Solomon Islands as an is HACKMAN, B. D . 1969. The geology of east land arc. N ature, vol. 211, no. 5055, pp . and central Gu ada1canal: A preliminary state 1249-1251. ment. British Solomon Islands Geological --- 1966b. Upper Cretaceous (Senonian ) Record, vol. 3 (1963-1967). In press . bathyal pelagic sediments with Globotruncana KRAUSE , D . C. 1966. Bathymetry and submarine from the Solomon Islands . Journal of the geology of the northern Tasman Sea-Coral Geological Society of Australia, vol. 13, pp . Sea-Southern Solomon Sea region of the 439-447. southwestern Pacific Ocean. New Zealand - -- 1967. A possible resoluti on of the Mela Oceanographic Institute, Memoir 41. nesian Re-entrant. Upper Mantle Project, LAUDON, T . S. 1968 . Land gravity survey of Second Australian Progress Report, 1965 the Solomon and Bismarck islands. In: The 1967, pp. 192-194. Australian Academy of crust and upper mantle of the Pacific area, Science, Canberra. pp . 279- 295. Geophysical Mon ograph 12, COLEMAN, P. J., and A. A. DAY. 1965. Petro American Geophysical Union, Washington. leum possibilities and marked gravity anom alies in north-central Guada1canal. British LIGGETT, K. L. 1967. Preliminary report on Solomon Islands Geological Record, vol. 2 Maewo. New Hebrides Condominium Geo (1959-1962), pp . 112-119. logical Survey, Annual Report (1965) , pp . 8-12. COL EMAN, P. J., J. C. GROVER, R. L. STANTON, and R. B. THOMPSON. 1965. A first geologi MALAHOFF, A. In preparation. Gravity and cal map of the British Solomon Islands. magnetic studies of the New Hebrides island Briti sh Solomon Islands Geological Record, arc. N ew H ebrid es Condominium Geological vol. 2 (1959-1962), pp . 16-17. Survey Report. DONNELLY, T. W. 1964. Evolution of eastern MALAHOFF, A., and G. P. WOOLLARD. 1969. Greater Antillean Island Arc. Bulletin of the The New Hebrides Islands gravity netw ork. American Association of Petr oleum Geolo Hawaii Institute of Geophysics Report 69-14, gists, vol. 48, pp . 680-696. 26 pp. FAIRBRIDGE , R. W . 1961. The Melanesian Bor MALLICK , D. I. J. 1969. Prel iminary report on der Plateau, a zone of crustal shearing in the southern Pentecost. N ew Hebrides Condo- 314 PACIFIC SCIENCE, Vol. 24, July 1970
minrum Geological Survey, Annual Report ROSE, J. c., G. P. WOOLLARD, and ALEXANDER (1967), pp. 20-25. MALAHOFF. 1968. Marine gravity and mag MAWSON, D. 1905. The geology of the New netic studies of the Solomon Islands. In : Hebrides. Proceedings of the Linnean Society The crust and upper mantle of the Pacific of New South Wales, vol. 30, pp. 400-484. area, pp . 379-410. Geophysical Monograph MENARD, H . W . 1964. Marine geology of the 12, American Geophysical Un ion, Washing Pacific. New York, McGraw-Hill. 271 pp. ton, D.C. MITCHELL, A. H. G. 1966. Geology of South RUE, A. de lao 1937. Le Volcanisme des Nou Malekula . New Hebrides Condominium Geo velles-Hebrides (Melanesie) . Bulletin vol logical Survey, Report 3, 40 pp. canologique, ser. 2, vol. 2, pp. 79-142. --- 1967. Geological history of South SAGATZKY, J. 1959. Contribution a l'etude Malekula . New Hebrides Condominium Geo geologique de I'Ile Espiritu Santo (Nouvelle logical Survey, Annual Report (1965), pp. Hebrides) 1. Roches sedirnentaires. Bulletin 34-39. de la Societe geologique de France, ser. 6, --- 1968. No rth Malekula. New Hebrides vol. 8, pp . 501-509. Condominium Geological Survey, Annu al Re --- 1960. Contribution aI'etude geologique port (1966), pp. 28-34. de I'Ile Espiritu Santo (Nouvelle Hebrides) --- 1969. Raised reef-capped terraces and 2. Roches volcaniques. Bulletin de la Societe Plio-Pleistocene sea-level changes, North geologique de France, ser. 7, vol. 1, pp. 588 Malekula, New Hebrides. Journal of Geol 593. ogy, vol. 77, pp.56-67. STANTON, R. 1. 1961. Explanatory notes to ac MITCHELL, A. H . G., and A. J. WARDEN. 1970. company a first geological map of Santa Geology of the New Hebrides island arc. Isabel, British Solomon Islands Protectorate. In press. Overseas Geology and Mineral Resources, vol. MIYAMURA, S. 1968. Seismicity of island arcs 8, pp. 127-149. and other arc tectonic regions of the circum STANTON, R. 1., and J. D. BELL. 1969. Vol Pacific zone. In : The crust and upper mantle canic and associated rocks of the New of the Pacific area, pp . 60-69. Geophysical Georgia Group , British Solomon Islands . Monograph 12, American Geophysical Union, British Solomon Islands Geological Record, Washington, D.C. vol. 3 (1963-1967) . In press. OBELLIANE, J. M. 1958. Contribution ala con STEPHENSON, P. J., G. J. H. MCCALL, R. W . naissance geologique de I'Archipel des Nou LE MAITRE, and G. P. ROBINSON. 1968. The velles Hebrides (Iles Vate, Pentecote, Ma Ambrym Island Research Project. New He ewo, Santo) . Science de la Terre, vol. 6, pp . brides Condominium Geological Survey, An 139-368. nual Report (1966), pp . 9-15. OFFICER, C. B. 1955. Southwest Pacific crustal WARDEN, A. J. 1967. The geology of the structure. Transactions of the American Geo Central Islands. New Hebrides Condominium physical Union, vol. 36, pp. 449-459. Geological Survey, Report 5, 106 pp. RICHARDS, J. R., J. A. COO PER, A. W. WEBB, - -- 1968. Geology of Aoba volcano. New and P. J. COLEMAN. 1966. Potassium-argon Heb rides Condominium Geological Survey, measurements of the age of basal schists in Annual Report (1966), pp. 15-27. the British Solomon Islands. Nature, vol. 211, WOOLLARD, G. P., A. S. FURUMOTO, J. C. no. 5055, 1251-1252. ROSE, G. H . SUTTON, A. MALAHOFF, and ROBINSON, G. P. 1968. Biostratigraphy and 1. KROENKE. 1967. Cruise report on 1966 structure of Espiritu Santo, New Hebrides seismic refraction expedition to the Solomon Archipelago , southwest Pacific. Unpublished Sea. Hawaii Institute of Geophysics Report doctoral dissertation, University of Western 67-3, 31 pp. Australia. WRIGHT, P. C. 1969. The pyroxene diorite in - -- 1969. Geology of north Santo. New trusion of Mount Kanekolo , central-east New Hebrides Condominium Geological Survey, Georgia. British Solomon Islands Geological Regional Report no. 1, 77 pp. Record, vol. 3 (1963-1967) . In press.