New Zealand Journal of Geology and Geophysics
ISSN: 0028-8306 (Print) 1175-8791 (Online) Journal homepage: www.tandfonline.com/journals/tnzg20
The zonal arrangement of secondary mineral
assemblages on the tavua goldfield, Fiji
P. Ibbotson
To cite this article: P. Ibbotson (1965) The zonal arrangement of secondary mineral
assemblages on the tavua goldfield, Fiji, New Zealand Journal of Geology and Geophysics, 8:5,
863-868, DOI: 10.1080/00288306.1965.10422127
To link to this article: https://doi.org/10.1080/00288306.1965.10422127
Published online: 17 Jan 2012.
Submit your article to this journal
Article views: 134
View related articles
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=tnzg20
1965] 863
THE ZONAL ARRANGEMENT OF SECONDARY
MINERAL ASSEMBLAGES ON THE TAVUA
GOLDFIELD, FIJI
P. IBBOTSON
Geological Survey of Fiji, Suva, Fiji
(Received for publication 22 June 1965)
(Geological Survey of Fiji: Occasional Paper No.1)
ABSTRACT
Secondary minerals and dike swarms are described from part of an Upper Tertiary
volcano in Fiji. From a study of the lavas and related dike swarms and the
amygdaloidal zeolites in the lavas, it is thought that the species of secondary minerals
deposited in the lavas was controlled by the temperatures developed by the intrusion
of dike swarms.
INTRODUCTION
Several authors including Blatchford (1953), Garretty (undated), Cohen
(1962), and Ibbotson (1963) have described the late Tertiary volcanism of
the Tavua Goldfield of north Viti Levu, Fiji. They show that there a
central volcano of Plio-Pleistocene age, largely made up of olivine basalt
overlain by smaller amounts of trachybasalt, augite trachyandesite, and
biotite trachyandesite, was intruded by monzonite and trachyte; these phases
of volcanicity succeeded one another with only short intervening periods of
quiescence. From the end of the olivine basalt eruption, volcanicity was
accompanied by caldera collapse, and the products of each subsequent
eruptive phase were deposited within the olivine basalt caldera. Subsidence
has produced a basin-like form in most of the trachybasalts and trachy-
andesites though a few trachybasalts retain their horizontal, depositional dip.
Dike swarms related to Tertiary volcanism are found in various parts of
Fiji: in particular an early olivine basalt swarm and a later swarm of
olivine basalt and trachyandesite appear to be related to the central
volcano referred to; these two swarms form the subject of this paper. The
lavas cut by these swarms characteristically contain zeolites and calcite as
amygdale minerals, and the distribution of these minerals is related spatially
and probably also genetically to the dikes.
In addition to the two mentioned, other dike swarms are related to the
Tavua structure, but subsequent to their deposition the secondary minerals
in these areas have been destroyed, or at least influenced, by later mineralising
fluids.
The olivine basalt and trachyandesite dike swarms are shown on the
accompanying map (Fig. 1), which also indicates localities to which
reference is made.
N.Z. J. Geol. Geophys. 8 : 863-68
SECONDARY MINERAL ASSEMBLAGES
..uroK(1 t-~~,:"" - - Dyke _Quartz
R Assemblage boundary
K"O' . . . . 0' . . .
- - LIthologic boundary. c=JCalcite.
-=~~
•u. ----< Creek . C Thomsonlte-mesolite~c8Icite
VITI LEVU F IJ I
~
= Chabazlte-natrollte-ca1cite.
.. ' / Analclte-chabazite-natrolite-
~
L ~ scolecite celcit e. c
C Nat-oht e-vnesohte-calcite.
i"'-----=- J Chabaaite-rneschte with rare
~~. natrolite-calcite.
.:
II ------
-N·
II
FIG. l-Map showing the relation of dike swarms to secondary mineral assemblages in basalt lavas at Vatukoula, Fiji. The dikes are shown somewhat
diagrammatically. The lithological boundary beginning near the south-east corner of the map separates the trachybasalt lavas to the north-east
from the olivine basalts to the south-west; it also divides the Waikumbukumbu dike swarm from the olivine basalt dikes in the Nandele-Nasomo
region.
1965] IBBOTSON - MINERAL ASSEMBLAGES IN FIJI 865
THE LAVAS
The earlier olivine basalt dike swarm cuts through olivine basalt lavas
and pyroclastics of the Mba Series in the Nandele--Nasomo region. Here
lava flows, generally not more than 20 ft thick, are interbedded with pyro-
clastic rocks that make up an estimated 70% of the sequence. Most of the
pyroclastics are volcanic breccias and agglomerates with a matrix of tuff
or crystal debris; tuff or ash bands are less common but seem to be more
persistent than either the lavas or coarse pyroclastics. The sequence, which
also contains a few water-deposited volcanic conglomerates, has a consistent
northerly dip of about 10°.
The olivine basalt and trachyandesite dikes of the later swarm, which is
located in the Waikumbukumbu area, cut through a succession of trachy-
basalt lavas and pyroclastics; there is also, within the trachybasalt, some
trachyandesite (usually containing some fragments of olivine basalt) and a
massive trachybasalt flow having the form of spatter lava erupted from a
local parasitic vent probably active during trachybasalt volcanicity.
THE DIKE SWARMS
The olivine basalt dike swarm is exposed over an area of 2 or 3 square
miles in the numerous creeks flowing northwards between the villages of
Nasomo and Nandele; to the north it is terminated by the partly faulted
olivine basalt-trachybasalt boundary, but to the south the swarm dies out
in the lower slopes of the Koromboya Hills. The members of the swarm
include both olivine basalt and trachyandesite, but while the former pre-
dominates in the central area, the latter is restricted to a few intrusions
on the eastern and western flanks.
The dikes have a general northerly trend, few diverging more than 30°;
from west to east the strike swings a little from north-east to north-west.
IndividuaL dikes are a few inches to 30 ft thick, but average about 4 ft.
Most are vertical and may persist with constant strike for a mile or so.
The swarm is most concentrated in Matekenaka creek where dikes are
estimated to make up more than 3% of the country, although exposures
are not sufficiently good for reliable estimates to be made. It is clear that
the density of the swarm decreases with increasing altitude; few olivine
basalt dikes are seen higher than 700 ft above sea level where most
intrusions are represented by thin stringers of tachylite. Dikes continuing
to a height of more than 1,000 ft are rare.
The upward diminution of the dike swarm is believed to be due to the
presence of a vent at an altitude of about 700 ft; only at times when such a
conduit was blocked would dikes be intruded to a greater height than this.
The intrusive rocks are very similar lithologically to the lavas although
they show a greater variation in texture and grain size. The dominant
type is richly porphyritic basalt bearing phenocrysts of olivine and augite
in varying arnou.nts; usually. augite predominates. Two finer rock types
also occur; one IS a fine-grained black basalt and the other is a vitreous
blue-grey tachylite.
Geology-IO
866 N.Z. JOURNAL OF GEOLOGY AND GEOPHYSICS [Nov.
The second swarm of olivine basalt and trachyandesite dikes crops out
a few miles to the east of the main olivine basalt swarm. The best exposures
are on the south-eastern flank of the volcano from Wainivothe in the north
to Waikumbukumbu and Nandarivatu in the south; the more extensive
outcrops are in the Wainivothe and Waikumbukumbu creek systems. The
augite and biotite trachyandesite intrusions are generally steeply dipping and
sills are rare. Dikes of this group are generally radial to the main volcanic
ring structure, but in the Wainivothe area they are tangential and parallel
to the intrusions at Waikumbukumbu; this consistent strike is believed to
be due to re-intrusion on radial olivine basalt dikes and the root zone of
the trachybasalt volcano.
Many trachyandesite dikes follow pre-existing basalt dikes to give multiple
structures, and some single fissures have been intruded three or four times.
These multiple intrusions are specially common in the Waikumbukwnbu
area, east and south of the road bridge. Many are 30 or 40 ft wide, and
can be followed for several miles, though they vary in form along the
strike and become less complex with altitude.
The trachyandesites are white, yellow, or pale green; many contain
abundant phenocrysts of feldspar and less commonly of augite, biotite, or
augite with biotite. In thin section the groundmass is commonly trachytic,
but they contain much less potash feldspar than the trachytes proper.
Trachyandesite dikes are much less well jointed than basalt dikes but
generally show a better flow lineation.
THE MINERAL ASSEMBLAGES AND THEIR DISTRIBUTION
Zeolites are abundant in the olivine basalt lavas of the volcano except the
collapsed rocks in the vicinity of the Basin Contact. * The amygdales com-
monly measure 1-3 mm, usually partly infilled.
The mineral species identified are shown in Table 1. The minerals were
determined in the field, with optical determinations on representative
samples to give a control on their identity. Five distinct assemblages can
be recognised:
1. thomsonite-rnesolite-calcite
2. chabazite-natrolite-calcite
3. analcite---<:habazite-natrolite---scolecite---<:alcite
4. natrolite-rnesolite-calcite
5. chabazite-mesolite with rare natrolite-calcite
Field work shows that these assemblages occur in well defined areas to
constitute the mappable zones (Fig. 1). Zones 1-3 are superimposed on
the olivine basalt lavas in the form of a wedge having its apex truncated
*The Basin Contact is shown on Fig. I; it is the lithologic boundary separating the
trachybasalts from the later trachyandesites, which display a perfect basin structure.
Near this boundary the lavas contain quartz and other minerals and are commonly
highly altered. The secondary mineralisation of the "basin" is quite unlike the
zeoli tic alteration that is the subject of this paper, and will not be further referred to.
1965] IBBOTSON - MINERAL ASSEMBLAGES IN FIJI 867
TABLE I-Mineral Species
Zeolite Refractive Indices Other Properties
----- -_.--_. - - - ~ - - - - ---------.-----
a f3 y
Scolecite 1'512 1·519 Biaxial negative; moderately low
2V; oblique extinction; length
fast.
?Thomsonite 1'527 1·530 Biaxial positrve; 2V low, estimated
20-30; parallel extinction.
Mesolite 1'501 < 11
a/h
< 1'507 Very low birefringence;
parallel extinction.
almost
Natrolite 11
alh
< 1·483 Biaxial positive; parallel extinction.
Chabazite 1·488 1·491 Very low birefringence, almost iso-
tropic; cleavage not well devel-
oped and fragments rather angu-
lar.
Analcite 11 1·487 Mainly isotropic, though some frag-
ments show very low bire-
fringence.
Stilbite 1·494 1'500 1'503 Biaxial negative; variable extinc-
tion - parallel to oblique.
----- _ .._ - - - - - - -- -- -----
to the north by the area with calcite deposited near the Basin Contact;
the outer margins of this wedge are an extensive development of the
analcite-scolecite--calcite assemblage, which could possibly be split into less
complex zones, though, in the area south-east of Vatukoula where this
assemblage is best developed, the pattern has been complicated by the
mineralisation.
These zeolite zones bear a striking relation to the olivine basalt dike
swarm; where the swarm is most intense the thomsonite-mesolite-calcite
assemblage is developed; where it is less intense chabazite-natrolite--calcite
is found; and beyond the limits of the swarm the analcite---chabazite-
natrolite-scolecite--calcite association is deposited. This last zone may be a
regional zeolite development in the lava pile, or possibly an aureole around
the original olivine basalt vent.
It seems likely that the zeolite zoning may have been produced by the
higher temperatures caused by dike intrusion. Discussing the formation of
zeolites, Walker (1960) states:
"The heat to support zcolitisation is supplied by dykes, by the slow accession of
heat from below and by exothermic reactions in the lavas. The lavas merely supply
the water filled cavities and the required raw materials."
868 NZ. JOURNAL OF GEOLOGY AND GEOPHYSICS [Nov.
There is insufficient evidence to comment fully on the zeolite distribution:
in the trachybasalt lavas. The horizontal lavas and pyroclastics support a
natrolite-mesolite-calcite association, and in the Waikumbukumbu creek
area, where the olivine basalt and trachyandesite dike swarm is quite intense,
the chabazite-mesolite with rare natrolite-calcite association is also exposed;
the two assemblages crop out in zones elongated towards the south-east,
and their formation may be related to temperatures developed on the
south-east-striking Waikumbukumbu dike swarm.
In a recent paper, Crook ( 1963) related secondary mineral assemblages
in Fijian rocks to depth of burial. In early and mid-Tertiary rocks he
described new feldspars, phyllosilicates, and zeolites, but from the Mba
Series (which includes the olivine basalts described in this paper) he
described a zeolite assemblage. The present work suggests that regionally
or locally, and at anyone stratigraphic level, the zeolite assemblage may vary
considerably; such assemblages may possibly be successfully related to
individual volcanic centres and dike swarms.
CONCLUSIONS
The data presented indicate that where the dike swarm is most dense, the
lavas contain a very different zeolite assemblage from where it is less
dense; where the lavas are not cut by dikes, they contain a complex but
persistent assemblage of zeolites with calcite. Since the olivine basalts are
apparently chemically and mineralogically identical, the temperature devel-
oped in the lavas by the intruding dike swarm is the only variable factor
determining the species of zeolite deposited.
ACKNOWLEDGMENTS
I am grateful to Drs W .. Skiba and G. P. 1. Walker of Imperial College, London,
for much discussion and for critically reviewing the manuscript. I am also indebted
to Dr K. A. Phillips and the Geological Survey Department, Fiji, and to Mr E. C.
Cohen and the geological staff of the Emperor Goldmining Co. Ltd., for much
assistance during the field work. This investigation was completed as part of a thesis
project and I wish to record my indebtedness for financial help and the provision of
air fares to the United Kingdom Department of Scientific and Industrial Research.
REFERENCES
BLATCHFORD, A. 1953: Geology of the Tavua Goldfield. Proc, Aust, Inst. Min.
Metall .. No. 168-9.. p. 13.
COHEN, E. C. 1962: Revised Geology of the Tavua Goldfield. Proc, Aust, lnst, Min.
Metall .. No. 204, p. 185.
CROOK, K. A. \1{T. 1963: Burial Metamorphic Rocks from Fiji. N.Z. f. Geol. Geopbys.
6 (5): 681-704.
GARRETTY, M. D. (undated): Geology and Ore Deposits of the Tavua Goldfield.
(Unpublished report in files of Geological Survey of Fiji.)
IBBOTSON, P. 1963: Petrological Aspects of some Volcanic Rocks on Fiji. (Ph.D.
thesis presented to the University of London.)
WALKER, G. P. 1. 1960: Zeolite Zones and Dyke Distribution in Relationship to the
Structures of the Basalts of Eastern Iceland. J. Geol. 68 (5), p. 515.
