Discovery of Permian Fossils in the Eglinton Volcanics, Southland

Transactions of the Royal Society of New Zealand : Geology, Volume 4, Issue 6, 21 July 1966, Page 139

Discovery of Permian Fossils in the Eglinton Volcanics, Southland

C. A. Landis

J. B. Waterhouse

and

[Received by the Editor , 17 August 1965.]

Abstract

The Permian brachiopod Attenuatella is reported in member EE 2 , near the base of the Eglinton Yolcanics, hitherto reported unfossiliferous. The Eglinton specimens are close to the Kazanian species A. incurvata Waterhouse and probably identical with Aktastinian (lower Artinskian) Attenuatella in the Takitimu Group. Stratigraphic and faunal considerations suggest that the lower Eglinton Yolcanics are probably Aktastinian, and therefore somewhat older than considered by Wood (1962). Recent views on the age of the Eglinton Yolcanics are discussed and the lithology and comparisons of the fossiliferous band are described. Formation names are proposed for the units recognised in the Eglinton Yolcanics by Grindley (1958), and the Eglinton Yolcanics plus the underlying Skippers Formation of Wood (1962) are referred to a new Alabaster Group.

Earlier Views on the Age of the Eglinton Yolcanics

The name Eglinton Yolcanics was applied by Grindley (1958) to a strip of chiefly volcanic rock exposed along the Eglinton Valley, Southland (Table 1). In the absence of fossils, Grindley (1958, p. 27) assigned a Carboniferous age, remarking that the volcanics appeared to lie conformably below Maitai ( = Bryneira) strata at Lake Alabaster 1 and that they were probably correlative with

the Takitimu Group of Southland and the Brook Street Volcanics of Nelson. The correlations were based on generalised lithology, petrography and structural considerations, and the Carboniferous age was assigned when the Maitai was thought to be Artinskian and the Takitimu Group probably Carboniferous (e.g., Wellman, 1956; Fleming, 1957; Mutch, 1957). The same year as Grindley’s bulletin was published Waterhouse (1958) showed that faunas from the Takitimu Group were Lower Permian (Sakmarian and Artinskian), not Carboniferous. In 1959 Grindley (in Grindley, Harrington, and Wood, 1959: 43) therefore referred the Eglinton Volcanics to the Lower Permian, but it was too late to change the Carboniferous age preferred in the 1958 Geological Map of New Zealand. In a number of places Grindley, Harrington, and Wood (1959: 42, 70, 86) still referred to the Eglinton Volcanics as Carboniferous, possibly due to the difficulties of correcting a bulletin in press.

Further indirect support for a Lower Permian age for the Eglinton Volcanics came with the discovery by Bruce (1962) of fossils in the presumed correlative Brook Street Volcanics. These fossils were described by Waterhouse (in Bruce, 1962) as the bivalve Atomodesma, limited to the Permian, and the gastropod Peruvispira, chiefly Permian but reported from Carboniferous beds in Australia. Later Waterhouse (1963 a: 590) pointed out that the Carboniferous records of

Peruvispira were of doubtful value and that the genus was probably limited to the Permian. Both forms indicated an Upper Sakmarian or more probably lower Artinskian age (Waterhouse, 1963 a: 595, fig. 1).

In discussing the New Zealand Permian stratigraphy and faunas Waterhouse (1963 b; 1964 a) has referred the Eglinton Volcanics tentatively to the Lower Permian, with a Sakmarian to Kungurian age. This requires modification, for Wood (1962) has erected the new Skippers Formation (Table I) for strata previously included by Grindley (1958) and Grindley, Harrington and Wood (1959) in the Eglinton Volcanics. The formation is described in more detail by Mutch (1965). It is lithologically similar to the Eglinton Volcanics, and we here propose to place both units in a new group, the Alabaster Group 2 . The name is based on Lake Alabaster, in NW Otago; both units are exposed in the Skippers Range immediately west of the lake. We offer this name with some diffidence, as a device to retain a usage consistent with the definition (but not the map) of the Eglinton Volcanics by Grindley (1958), and completely consistent with the maps and definitions by Wood (1962) and Mutch (1965). Alternatively the Eglinton Volcanics could be expanded to take in the Skippers Formation. In any case we urge that the Skippers Formation and Eglinton Volcanics be treated as a single major unit. Regional mapping in the Eglinton area (Grindley, 1958) has shown that seven (numbered) units recognised in the Eglinton Volcanics are areally extensive and persistent enough to be treated as formations; we therefore propose to relegate the Eglinton Volcanics to subgroup status and name the constituent formations as set out in Table 11.

With one exception these names are based on geographic features in the area of the type section (the Earl Mountains and Eglinton Valley). Owing to an inadequate number of named topographic features in the Earl Mountains, the fossil-bearing siltstone unit (EE 2 ) is designated the Wesney Siltstone after probably correlative outcrops in the valley of Wesney Stream, which flows west from the Tapara Range into the Eglinton River.

The Skippers Formation underlies the Eglinton Volcanics and is likely on Waterhouse’s correlations to occupy at least the Sakmarian part of the column, so that the Eglinton Volcanics are likely to be Artinskian-Kungurian. A much shorter time-span was preferred by Wood (1962), amounting to no more than

the Kungurian Stage (Table III). Wood correlated the entire Eglinton Volcanics with the Livingstone Volcanics, on the opposite limb of the Key Summit Syncline, and correlated the Skippers Formation bands with the Red Mountain ultramafites, which were probably emplaced no earlier than Upper Artinskian or lower Kungurian (Waterhouse, 1963 b; 1964 a). Wood’s correlation was not accepted by Waterhouse (1964 c; and in press), for it seemed more likely that the Skippers serpentinite was equivalent to serpentinites of possible Sakmarian age on the east limb, well below the Red Mountain Ultramafites, and represented by the Groisilles Volcanics in Nelson. The Artinskian-Kungurian correlation, indicated by correlation with the Brook Street Volcanics of Nelson, was preferred, as in Table 111.

The Discovery of Fossils in the Eglinton Volcanics

The first direct evidence on the age of the Eglinton Volcanics is provided by the discovery of well-preserved external moulds of a brachiopod in the gorge of the East Branch of the Eglinton River. The fossils have been found in an extensive siltstone horizon, mapped by Grindley (1958) as EE 2 , near the base of the volcanics. The exposure is approximately one mile east of the Te AnauMilford road bridge, on the north side of the East Eglinton Gorge, at a low cliff separated from the river by a flood plain about 100 ft wide, covered by beech and fuchsia (grid reference 009655, S 131, Map 1 of Grindley, 1958).

Lithology

The cliff exposes dusky red mudstones and siltstones, massive, as defined by McKee and Weir (1953: 383), and too intensely sheared and shattered to yield data on attitude or direction of younging. Nearby volcanic layers are subvertical. The sheared surfaces are often veneered with chlorite and the rocks are veined by stilbite and, less commonly, prehnite.

The fossil-bearing rock is a poorly sorted, tuffaceous, silty mudstone. Petrographically, it is composed of silt grains embedded in a red and green matrix. Grain to grain contacts are occasionally continuous, but more often floating. The silt fraction is composed of angular to subangular grains of plagioclase and quartz, with less abundant positive chlorite and detrital clinopyroxene, and minor sericite, spilitic rock fragments, sphene, amphibole, and hematite. A thick jacket of hematite covers many silt grains. Many grains show cuspate shapes and unrounded acute intersection of edges, suggesting derivation from volcanic ash. Grain size rarely exceeds 0.05 mm and is commonly much finer.

The silt grains are set in a pale green matrix which is spottily overprinted with a conspicuous rusty red. The matrix is composed predominantly of chlorite (14A) and finely comminuted quartz and albite; hematite and additional 10A clay mica are also common.

Numerous thin sheets of augite basalt and ankaramite cut the exposure, at various steep angles. These rather distinctive rocks contain abundant brightgreen chrome diopside phenocrysts set in a groundmass of augite and plagioclase (Anss.ys). The rocks have been slightly altered and contain laumontite and chlorite. Some altered mafic minerals may have been originally olivine.

Elsewhere in the Tapara Range Grindley (1958: 26) has described EE 2 as “ dominantly dove-grey to pale-green, extremely hard, massive mudstone without trace of bedding or cleavage ”; he also notes minor beds and lenses of palegreen breccias.

No exactly similar lithology is known in Lower Permian fossiliferous rocks of the South Island, although fossils come from a diversity of rock types; massive tuffaceous siltstones and mudstones, volcanic conglomerate and breccia, at least four types of volcanic greywacke (medium- and coarse-grained, conglomeratic, tuffaceous, and calcareous), and limestone, such as the feldspathic biosparrudite (Folk, 1959) at G 56323, Productus Creek Group. The fossil-bearing Lower Permian sediments have a volcano-clastic nature and this volcanic detritus in the few slides examined shows mild erosive rounding, evidence of redeposition, incipient sorting, and only minor amounts of shard relicts and completely unrounded crystals. These data suggest that the fossiliferous sediments were deposited in shallow water close to the source of the sediment, and that the seafloor was not overwhelmed by completely fresh broken-up volcanic material such as is seen in the Kaka volcanic breccias of Nelson.

The Atomodesma-hearing rocks from the Grampian Formation recorded as a boulder not in situ by Bruce (1962), and found by one of us (J.8.W.) extensively in situ at Flaxmore Hill, Nelson, are coarse volcanic greywackes. The Peruvispira is found in a dense tuffaceous sandstone (Bruce, 1962). Associated non-fossiliferous rocks typical of the Grampian Formation are petrographically and to some extent lithologically comparable to the EE2 Wesney Siltstone. These are dense, fine-grained, grey and black, calcareous tuffaceous beds, particularly well exposed at Greville Harbour, D’Urville Island. The black beds are composed of subangular grains of quartz, feldspar, and iron ore in a slightly calcareous and tuffaceous matrix (Bruce, 1962: 164), and somewhat resemble EE2. Interbedded grey layers are similar, but more calcareous (Bruce, 1962). Thus, although petrographically similar, the Grampian Formation appears to differ from the fossiliferous EE2 band in colour, grain-size, and calcareous matrix. Neither intrusions of ankaramite nor chrome-diopside-bearing volcanics have been described from the Grampian beds, Bruce (1962) reporting only rare keratophyre and spilite.

Identity of the Fossils

The fossils from the EE 2 beds belong to a single species, preserved chiefly as natural external moulds with the ornament well preserved and with few details of the interior. One is a specimen with valves conjoined; two others! are dorsal valves and the remainder ventral valves. All are kept at the Geology Department, University of Otago, where they are registered as 0U2440-48, 2467-2468. They belong to the distinctive spiriferoid brachiopod Attenuatella Stehli (1954), a genus restricted to the Permian and reported from Russia, United States, Mexico, and New Zealand. Specimens have also been recently found in New Caledonia and Australia (Waterhouse, in press). The Eglinton specimens can be readily distinguished from all overseas species apart from the New Caledonian ones, which, not yet examined in detail, could be conspecific with A. incurvata Waterhouse. They are more inflated and with a more incurved ventral umbo than the type species A. texana Stehli (1954), from the Leonard of Texas, and have a narrower ventral umbo, narrower delthyrium, and more concave dorsal valve than in A. attenuata (Cloud, 1944) from the Waagenoceras Zone of Southern Mexico. The two Russian species, A. stringocephaloides (Chernychev and Liharev, in Liharev and Einor, 1938) and A. taimyrica Chernjak (1963), are more transverse with a better defined ventral sulcus and more massive ventral umbo. The Australian species found by Professor A. H. Voisey in beds above the Monilopora horizon in the Drake Group is close in

its elongated outline and extended ventral umbo, but has spines nearly twice as fine as those of the Eglinton form. Undoubtedly the closest species so far described is A. incurvata Waterhouse (1964 b) from the Kazanian faunas of New Zealand. A. incurvata is of similar size, shape and inflation, has an identical narrow ventral sulcus and spines of almost identical density, 10 to 13 occurring per mm usually, although spines are occasionally fewer on the Eglinton specimens and more numerous on the Kazanian ones. Details of the hinge, interarea, delthyrium, and the external posterior face of the cardinal process also appear to be identical as far as they can be compared, but a limit is imposed by distortion and by paucity of material. There is usually a low dorsal fold, either extending for the length of the valve or restricted to the anterior margin, as well as low dorsal costae, in the Eglinton specimens, whereas these features are rarely developed in A. incurvata. No consistent difference can be found, however, between the specimens from the two horizons.

As noted by Waterhouse (1964 a; in press), Attenuatella is also found widely in the middle Takitimu Group. The best preserved specimens come from high in the group and are indistinguishable in most respects from A. incurvata apart from internal differences in the dorsal valve. As in the Eglinton specimens, a fold and costae lie on the dorsal valve, features less commonly found in Kazanian A. incurvata. Other Attenuatella from lower horizons of the Takitimu Group are internal moulds of ventral valves, and although similar in shape and size are too incomplete to be identified with certainty.

Age

The similarity of the Eglinton Attenuatella to the Takitimu specimens suggests that an Aktastinian (Lower Artinskian) age is most likely for the EE2 Wesney horizon of the volcanics. A Kazanian age is possible on objective consideration of the fossils alone, but appears to be ruled out by the position of the volcanics below the Kazanian Annear Sandstone and Howden Limestone. An Upper Artinskian-Kungurian age is unlikely, because, although descendants from the Takitimu Attenuatella undoubtedly persisted, not one specimen is known from the prolific upper Artinskian-Kungurian faunas in New Zealand. Faunas of this age are cold-water ones of east Australian affinities (Waterhouse, in press), and so it is probable that Attenuatella moved temporarily to warmer waters farther north.

Acknowledgments

We thank Mr S. N. Beatus, New Zealand Geological Survey, for his photographs of Attenuatella, and Professor D. S. Coombs and Mr J. D. Campbell, University of Otago, for reading and discussing the manuscript.

References

American Commission on Stratigraphic Nomenclature, 1961, Code of Stratigraphic Nomenclature. Bull. amer. Assoc. Petrol. Geol., 45(5): 645-665.

Bruce, J. G., 1962. The Geology of the Nelson City Area. Trans, roy. Soc. N.Z. Geol. 1(11): 157-81.

Cloud, P. E., 1944. In King, R. E., Dunbar, G. 0., Cloud, P. E., Miller, A. K., Geology and Paleontology of the Permian Area NW of Las Delicias, SW Goahuila, Mexico. Geol. Soc. America spec. Pap. 52: 49—67.

Fleming, G. A., 1957. Trans-Tasman Relationships in Natural History. In Science in New Zealand, Wellington: A. J. Reed, pp. 228-46.

Folk, R. L., 1959. Practical Petrographic Classification of Limestones. Bull. amer. Assoc. Petrol. Geol. 43: 1-38.

Grindley, G. W., 1958. The Geology of the Eglinton Valley, Southland. N.Z. geol. Surv. Bull. n.s. 58: 1-68.

land, 1:2,000,000. N.Z. geol. Surv. Bull. n.s. 66: 1-111.

McKee, E. D., Weir, G. W., 1953. Terminology for Stratification and Gross-stratification in Sedimentary Rocks. Geol. Soc. America Bull. 64: 381-89.

Mutch, A. R., 1957. Facies and Thickness of the Upper Paleozoic and Triassic Sediments of Southland. Trans, roy. Soc. N.Z. 84: 499-511.

1965. Martin’s Bay, Sheet 105, 1: 63,360. N.Z. geol. Surv.

Stehli, F. G., 1954. Lower Leonardian Brachiopoda of the Sierra Diablo. Bull. amer. Mus. nat. Hist. 105 (3): 261-358.

Chernyak, G. E., 1963. In Ustritskia, V. 1., Chernyak, G. E., Biostratigraphy and Brachiopods of the Upper Paleozoic of Taimyr. Trud. Arkt. Inst. Geol. 134: 1-139. 47 pis. (In Russian.)

Waterhouse, J. 8., 1958. The Age of the Takitimu Group of Western Southland. N.Z. ]. Geol. Geophys. 1: 604-10.

cluded). N.Z. J. Geol Geophys. 6(4): 587-622.

10(1): 165-76.

Bull. n.s. 72: 1-101.

35: 1-287.

Abstracts: 142.

N.Z. J. Geol. Geophys. 8: 1239-40.

(in press, a). The Permian of New Zealand. 22nd internat. geol. Congr. 1964 India, Section 9, Gondwanas.

(in press, b). A new Species of Attenuatella (Brachiopoda) from Permian Beds near Drake, New South Wales. Rec. aust. Mus.

Wellman, H. W., 1956. Structural Outline of New Zealand. N.Z. Dep. sci. indust. Res. Bull. 121: 1-36.

Wood, B. L., 1962. Geological Map of New Zealand, Sheet 22, Wakatipu, 1:250,000, N.Z. Geol, Surv.

New Zealand Geological Survey, 1958. Geological Map of New Zealand. Wellington: D.S.I.R.

G. A. Landis, Geology Department, University of Otago, Dunedin.

Dr J. B. Waterhouse, N.Z. Geological Survey, P.O. Box 368, Lower Hutt.

2 In order to insure consistency and avoid possible misunderstandings, stratigraphic terminology used in this paper is based on the Code of Stratigraphic Nomenclature set forth by the American Commission on Stratigraphic Nomenclature (1961).

l lt appears dubious that this contact is conformable; as far as we know no actual contact can be observed; it is only inferred. The Maitai (Bryneira) strata appear to belong to the upper Annear Sandstone. The basal Howden Limestone seems to be missing —probably due to faulting rather than facies change, because limestone on the opposite limb of the Key Summit Syncline is over 2,000 ft thick in this neighbourhood (Waterhouse, 1964 a, p. 56), requiring a very drastic and unlikely facies change for the limestone to pass westwards into upper Annear Sandstone. Moreover, in the Tapara Range, a few miles to the south, limestone has been found by one of us (C.A.L.) on both limbs of the syncline, and it is actually thicker on the western limb. In further support of a faulted contact, it may be noted that the Pyke Fault has been already established along part of the east shore of Lake Alabaster by Grindley (1958), between infaulted Tertiary (Arnold) coal-measures and the Bryneira Group. A thin sliver of coal-measures has been found by one of us (C.A.L.) between the volcanics and Bryneira Group south of the supposed conformable contact, and an infaulted Jurassic outlier lies along the strike to the north (Mutch, 1965). At the neighbourhood of the supposed conformable contact Grindley proposes that the Pyke Fault suddenly swings westwards into the volcanics. The fault may branch in this neighbourhood—it is concealed by alluvium—but we suggest that one branch with a fairly straight course remains between the volcanics and Bryneira Group.

Wood (1962) has omitted the volcanic outcrops from the eastern shore of Lake Alabaster from his map, but they are present approximately as shown by Grindley (1958).

World Standard Stages New Zealand Stages (Waterhouse, 1965, in press, a) Formations and Groups and Groups Tatarian Makarewan a* P 0 Pi Gountess formation O Winton Formation Waiitian BRYNEIRA GROUP Countess Formation Waiitian Winton Formation Pi i—i Tapara Formation Tanara Formation w £ Kazanian Puruhauan oi Annear Sandstone pq Annear Sandstone Howden Limestone Kungurian Braxtonian •—) PU P o (West) Nurse Formation Oh (East) Livingstone Volcanics eg ■n j -%/r • 0 Pi O pq Birley Andesite Earl Formation Largs Porphyrite Wesney Silts tone ALABASTER GROUP Knobs Porphyrite EGLINTON SUB-GROUP (West) Nurse Formation Birley Andesite Earl Formation Largs Porphyrite Wesney Silts tone Knobs Porphyrite Eglinton Bluffs 0 Pi i\cu jlvxuuiiuiiu Ultramafites Artinskian Mangapirian o Pi C3 CO § o h Grindley, 1958; Wood, 1962— faulted out Waterhouse, 1964a Ultramafics of M H H £ HH Q >A Routeburn Valley Keratophyre HUMBOLDT GROUP (East) Livingstone Volcanics Red Mountain Ultramafites Artinskian Grindley, 1958; Wood, 1962— faulted out Waterhouse, 1964a Ultramafics of Routeburn Valley Mangapirian Telfordian CO C pq O w Keratophyre 0 PQ VH Sakmarian <3 Skippers Formation Skippers Formation h) <M

Table I.—Permian Formations in the Eglinton Valley, Southland.

Map Symbol Formation Geographic Feature EE 6 Nurse Formation Nurse Creek EE 5 Birley Andesite Birley Pass EE 4 Earl Formation Earl Mountains EE 3 Largs Porphyrite Largs Peak EE 2 Wesney Siltstone Wesney Stream EE, Knobs Porphyrite Knobs Flat EE 0 Eglinton Bluffs Keratophyre Eglinton Bluffs

Table II

Period Stage Substage Eglinton Volcanics, plus Skippers Volcanics Volcanics Eglinton Kungurian Baigendzinian Grindley, Waterhouse, 1963c; Wood, 1962 Waterhouse, 1964c; and This in paper ,rsv < ** Grindley, Waterhouse, 1963c; Wood, 1962 Waterhouse, 1964c; and This in paper Permian Artinskian Aktastinian Grindley in press et al.. 1964a Sakmarian 1959 Carboniferous Grindley, 1958 Wood, in Grindley et

Table 111. —Correlations of The Eglinton Volcanics,