Vegetation and climate of the New Zealand Jurassic

Abstract

New Zealand's Jurassic plant fossils are known from non-marine settings in two terranes: the Murihiku forearc basin and from a small area on the Rakaia Terrane, an accretionary prism. These terranes lay in relatively high latitudes along the Gondwana margin but their precise location, latitude and position with respect to each other is unclear. The flora was dominated by conifers, ferns, bennettitaleans, pentoxylaleans and locally equisetaleans, but it is relatively depauperate, perhaps reflecting a high latitude position. Most climate indicators suggest a warm temperate climate with rainfall that was not especially high and that might have been seasonal at times. Revision of the New Zealand Jurassic flora is overdue, and while new genera may be discovered, species numbers may well decrease.

Keywords:

• New Zealand
• Jurassic
• Murihiku
• Rakaia
• conifers
• biodiversity

Introduction

New Zealand may be the key source of information for the vegetation of high southern latitudes during the Jurassic, despite some uncertainty about its palaeogeographic position. Research began with the description of fossil wood by Unger (1864), then came the now-classic work by Kidston & Gwynne-Vaughan (1907-1910) on the structure of tree fern stems, and a taxonomic apogee was reached with the work of Arber (1917) and Edwards (1934). More than a century from the earliest study, Mildenhall (1970) evaluated the taxonomic legitimacy of all New Zealand plant macrofossils listing many Jurassic species as invalid. Recent work on New Zealand's Jurassic floras has included an unpublished MSc thesis (Broekhuizen 1984), a field-trip guide (Grant-Mackie 1995), a review by Grant-Mackie et al. (2000), the work of Pole (1999, 2001) and Thorn (2001, 2005) on fossil forests, including palynology and sedimentology, and a concerted effort to develop a palynological zonation (de Jersey & Raine 1990, 2002). Raine et al. (2008) have also provided an atlas of palynomorphs for the New Zealand region encompassing Jurassic forms.

The purpose of this paper is to review the taxa currently known from New Zealand and to summarise climatic indications and recent advances in understanding palaeogeography.

Palaeogeography

Any discussion on the Jurassic vegetation of New Zealand must address the issue of where New Zealand was located. Or more specifically, where were the components of New Zealand? New Zealand is known to be a collage of tectonic terranes of diverse origins (Bishop et al. 1985; Mortimer & Campbell 1996), although Kamp (2001) cautioned that the time has come to consider their potential inter-relationships more than their differences. By the mid-Cretaceous, the components had essentially amalgamated and New Zealand lay in high, polar latitudes, but in the Jurassic at least three units have been identified and their relative positions at that time are still not clear.

The Murihiku Terrane (MT) was a forearc basin that was active from the Triassic into the Cretaceous (Ballance & Campbell 1993; Landis et al. 1999 for a more recently recognised occurrence of Jurassic MT), and is the source of most of New Zealand's Jurassic plant fossils (Fig. 1). The youngest Jurassic strata in New Zealand constitute the Huriwai Formation of Port Waikato, which is a part of the MT. It was initially thought to be Neocomian (Early Cretaceous) by Arber (1917), but is now regarded as Tithonian (Edwards 1934; McQueen 1955; Norris 1968; Norris & Waterhouse 1970; Stevens & Speden 1978). In the broadest sense, the sediments of the MT accumulated somewhere along or off the margin of south-eastern Pangaea/Gondwana. The exact position has potential ecological implications, for instance, in determining whether the terrane and associated flora lay in a polar position or not. Pole (1999) regarded the Middle Jurassic (Bathonian–Callovian, late Temaikan local stage; based on palynology, J.L. McKellar, J.I. Raine, and N. de Jersey, pers. comm. 1997; see also Speden (1958), for marine faunal control of dating similar-aged leaf beds at Owaka Creek) Curio Bay fossil forests of the MT, southernmost New Zealand, as growing near to the southern polar circle, but had insufficient data to decide on which side. Thorn (2001) claimed the Kawhia Harbour forests, of similar age to those at Curio Bay, grew at approximately 75–78°S, “well within the polar circle” and cited Stevens (1980, 1985) in support of a warmer climate at the time, a poleward extension of the Tethyan faunal realm, and an absence of a distinct polar realm. However, the point that Stevens tried to make was that if New Zealand lay at high latitudes at the time, it would most likely have had a distinct cool-temperate “Palaeoaustral” faunal realm – to mirror the Boreal fauna that occurred in high northern latitudes. Stevens (1980) concluded that New Zealand lay in mid-latitudes during the Middle Jurassic.

Fig. 1 Locality map of Jurassic sites in New Zealand, showing terranes (modified from Bishop et al. 1985) and localities mentioned in the text.

Adams et al. (2007) examined the ages of detrital zircons in New Zealand's terranes; most MT zircons were contemporary and there was little in the way of an older or “continental” source. For this reason, Adams et al. (2007) argued that the MT was an “isolated volcanic environment” and suggested that it existed in its present New Zealand position; they did acknowledge a source of older zircons in the Middle Jurassic, which might indicate a continental source. Their reconstruction of the Gondwana coastline shows the MT at around 70°S in the Late Triassic and lying at around 85°S by the Late Jurassic.

Outboard of the MT was a series of accretionary prisms including the Rakaia Terrane. This complex includes abundant Jurassic rocks, but most have either undergone significant metamorphism to schist (e.g. Mortimer 1993), or are marine. However, a small fluvial, plant fossil-containing, Late Jurassic (Tithonian) deposit in the Clent Hills (Canterbury region) accumulated on this terrane as it was being imbricated against the New Zealand margin and uplifted (Graham & Mortimer 1992; Kamp 2001).

Attempts to show the palaeolatitudinal positions of all New Zealand terranes in the Jurassic have been made by Grant-Mackie et al. (2000) and Adams et al. (2007). However, from a palaeoecological point of view (for instance, their position relative to the palaeopolar circle), it is important to realise that these are estimates.

The macrofossil taxa

New Zealand's Jurassic macrofossil taxa are summarised below. For convenience, they are grouped into likely or generally accepted higher taxa, but in the absence of cuticular and fertile material for most New Zealand specimens, these are provisional.

Clubmosses

Lycopodiaceae

Lycopodites arberi, based on an impression of a small, herbaceous shoot, is the only lycopod described from the New Zealand Jurassic (Edwards 1934).

Ferns

Osmundaceae

Arber (1917, 31) regarded Cladophlebis australis as the “most abundant of all species, without exception” in the New Zealand Jurassic. Since then, Frenguelli (1947) placed the New Zealand Jurassic forms into Cladophlebis patagonica and Retallack (1983) argued that C. australis should be restricted to Triassic forms that have doubly forking veins. At least four other species of Cladophlebis occur in the New Zealand Jurassic (Arber 1917; Edwards 1934): Cladophlebis sp. cf. C. albertsi, C. antarctica, C. denticulate and C. reversa. Cladophlebis is generally accepted as the foliage associated with the fossil axes that were first described by Kidston & Gwynne-Vaughan (1907-10) as Osmundites (see also Marshall 1926). The New Zealand species have since been placed into Osmundacaulis (Miller 1967), and Millerocaulis and Ashicaulis (Tidwell 1994). The size of these axes indicates that the plant was a small tree fern.

Dicksoniaceae

The foliage Coniopteris hymenophylloides recorded by Arber (1917) was probably a ground fern. Edwards (1934) also recorded C. lobata.

Dipteridaceae

Arber (1917) listed two species of Dictyophyllum from New Zealand (D. acutilobum and D. obtusilobum), although Edwards (1934) regarded them as specifically indeterminate and possibly the same species. Oliver (1950) recorded a third species, D. rugosum. Following Herbst's (1975) work on the family, Rees (1993) reviewed the dipterid ferns of Antarctica and New Zealand and described a new species from the Clent Hills as Goeppertella cf. woodii, the first record of this genus from New Zealand. He pointed out that it would be difficult to differentiate Goeppertella from Dictyophyllum based on fragmentary material (i.e. if it was not clearly bipinnate). This suggests that Arber's records may need re-evaluation.

Schizaeaceae

Arber (1917) recognised six species of Sphenopteris, including the cosmopolitan S. (Ruffordia) goepperti, S. currani previously known from Australia (Tenison-Woods 1883), three endemic and one undescribed species.

Equisetaceae

Two endemic species of Equisetum are validly described (as Equisetites) from the MT: E. hollowayei (Edwards 1934) and E. nicoli (Arber 1917). They are represented by small axes and leaf whorls and probably represent herbaceous plants.

Gymnosperms

Pteridosperms

Microphyllopteris was instituted by Arber (1917) on New Zealand material for ferns that resemble Gleichenia and belong to the Gleicheniaceae, but Nagalingum & Cantrill (2006) have proposed that it should be restricted to ferns that “cannot be ascribed conclusively to Gleicheniaceae”. Arber (1917) described one species M. pectinata, and figured another. McLoughlin et al. (2002) transferred M. pectinata to a new genus, Rintoulia, which they suggested was a seed fern based on its similarity to leaves with robust cuticle from the Victorian Cretaceous.

At least four species of Thinnfeldia are present according to Arber (1917), all of them being widespread species. Barbacka (1994) transferred T. indica (Feistmantel) (from India) and T. nordenskioeldii Nathorst (from Europe and possibly China) to a new genus, Komlopteris, and it may be that the New Zealand Thinnfeldia also belong to this new genus.

Bennettitales

The Bennettitales are probably represented by three species in three form genera. These include the widespread (though uncommon within assemblages) Ptilophyllum acutifolium and the endemic Pterophyllum matauriensis (Arber 1917), and Edwards (1934) added Otozamites.

Pentoxylales

The most common Taeniopteris species in New Zealand and Australia, T. daintreei, is generally accepted as the foliage of the pentoxylaleans (Drinnan & Chambers 1985), and this is supported by the presence in New Zealand of pentoxylalean cones, Carnoconites (Harris 1962) and permineralised stems known as Pentoxylon. Pole (1999) illustrated in situ tree stumps in the fossil forest at Curio Bay, which show the distinctive polyxylic architecture of Pentoxylales. The group (or something closely related) is also clearly present in a private collection; R. Jackson, Christchurch, collected from the Slope Point area, south of Curio Bay; M. Pole, pers. obs. 2000. The widespread T. arctica, T. crassinervis, T. vittata and an endemic species T. thomsoniana are also present in New Zealand (Arber 1917), but it is not clear whether these were also pentoxylaleans since similar leaves were also produced by some cycads, bennettites and ferns. Taeniopteris is a very simple leaf morphogenus, leaving little to distinguish the species other than a range of dimensions, the approach used by Blaschke & Grant-Mackie (1976) for delimiting Taeniopteris species at Port Waikato and the Clent Hills. A study of large collections would give a clearer idea of the species limits and could well reduce the number of taxa recognised from New Zealand. Edwards (1934) adopted a broad species view and combined T. arctica, T. daintreei and T. vittata with T. spatulata (he regarded T. daintreei as a synonymy of T. spatulata).

Cycadales

The cycadales are possibly represented by the widespread Nilssonia compta and the endemic N. elegans as identified by Arber (1917). However, in light of recent disagreement among authors on what constitutes Nilssonia or Taeniopteris (e.g. Gee 1989 versus Rees & Cleal 2004), these species may yet be placed in Taeniopteris. Arber (1917) also recorded a Cycadites sp. from Curio Bay.

Conifers

The most common type of conifer foliage by far is a widespread form that has flattened and spreading single-veined leaves, and is commonly referred to Elatocladus conferta (Arber 1917). This happens to be the only Jurassic plant in New Zealand for which the cuticular detail is known from specimens in the Clent Hills locality (Townrow 1967); this information, together with associated reproductive organ data, allowed him to describe a new genus and species, Mataia podocarpoides (Townrow 1967). However, extreme caution should be exercised before automatically placing specimens described as E. conferta into M. podocarpoides because the former is a rather generalised morphospecies that may cover several species. An Elatocladus with longer leaves, E. plana, also occurs in the MT, in some cases in association with E. conferta (Bartrum 1921; pers. obs, Fig. 2.).

Fig. 2 Bedding surface of siltstone (False Island Formation, probably Bathonian–Callovian), with Elatocladus sp. and thin-leaved Taeniopteris from the coast north of Curio Bay, New Zealand. Rock hammer for scale.

Other conifers include forms with imbricate scale leaves identified as Brachyphyllum, Cryptomerites and Pagiophyllum (Arber 1917). The multiveined, endemic Podozamites gracilis is probably referable or related to the Araucariaceae, as are the similar Nageiopsis longifolia (Arber 1917) and Desmiophyllum (Edwards 1934). These latter taxa might be the foliage of the plants that produced Araucarites cutchensis and Araucarites grandis ovulate scales (Arber 1917).

Two form genera of conifer wood are known, Araucarioxylon or Dadoxylon and Mesembrioxylon (Crié 1889; Arber 1917; Edwards 1934), but the affiliation of the New Zealand woods needs to be updated given recent nomenclatural revisions (e.g. Bamford & Philippe 2001), and there is no doubt that more species are awaiting discovery and description. Certainly, palynological work indicates that the actual diversity of conifers in the region was much higher than that indicated by macrofossils (de Jersey & Raine 1990; Thorn 2001).

Palissyaceae

The distinctive but enigmatic cone Palissya bartrumi (Edwards 1934) is common in some localities in the MT. It was earlier assumed that these structures were produced by conifers, but a recent review that also documents Australian occurrences has suggested otherwise (Parris et al. 1995). On the basis of this work, Miller (1999) ruled out a coniferous origin.

Palynology

Early palynological investigation of the Jurassic was carried out by Couper (1953, 1960), Norris (1968) and Norris & Waterhouse (1970). Detailed work by de Jersey & Raine (1990) documented a close relationship between the Early Jurassic palynoflora of New Zealand and that of Queensland, Australia. De Jersey & Grant-Mackie (1989) and Grant-Mackie et al. (2000) have also noted that the broader palynological sequence of the New Zealand Jurassic parallels that of Australia, New Caledonia, New Guinea, and perhaps more broadly, with Antarctica and Argentina. They noted that the New Zealand palynological record follows a similar sequence of a Cheirolepidiacean phase, followed by an Araucariacean phase, and finally by a Podocarpacean phase. Thorn (2001) documented palynological assemblages from Kawhia Peninsula and Curio Bay.

The Forests

The Temaikan fossil forest at Curio Bay (actually one of several at that location and part of a broad sequence of plant fossil-rich sediments; Pole 2004) has long been known (Hector 1886; Park 1887; Arber 1917; Guy-Ohlson 1974; Raine & Pole 1988). The structure of the most completely exposed forest was documented by Pole (1999) who regarded it as lying close to the palaeo-South Polar circle, but without sufficient data to determine on which side. The forest was probably of first generation, with a low canopy of about 10 m high, where most trees were about 70 years old, plus a few large emergents around 200 years old rising to about 30 m. There was a well-developed undergrowth of small gymnosperms (young conifers and Pentoxylon and osmundaceous “tree ferns”). There was nothing unique about the structure or density that might suggest that the forest grew in unusual polar conditions. The forest grew on a floodplain adjacent to a volcanic upland where intermittent, flashy and poorly confined floods overwhelmed at least 10 forests in about 40 m of section (Pole 2001).

In the North Island, Thorn (2001) documented the Temaikan (Middle Jurassic MT) vegetation of Kawhia Harbour. She identified several plant communities that had grown in a braided river environment. Elatocladus was the only type of conifer foliage present. This taxon is generally assumed to belong to the Podocarpaceae, although interestingly, Thorn (2001) did note that the local palynoflora was dominated by Araucariacites. Two types of coniferous wood were present: the Dadoxylon–Araucarioxylon type and Pentoxylon. Palynology suggested that several species of Araucariaceae or relatives were present, together with Podocarpaceae and possibly Taxodiaceae. Beyond this, only the most common Jurassic macrofossil genera in New Zealand were present: Taeniopteris and at least two species of Cladophlebis, although there were possible bryophytes, and some seeds and cones.

The Curio Bay and Kawhia forests both grew in relatively unstable environments and Ballance (1988) envisaged a similar landscape for the Late Jurassic Huriwai braidplain. Lower energy, meandering fluvial systems did exist, for instance, along the Otara-Fortrose coastline in the southernmost MT (Pole 2004) but these await more complete documentation.

Climate

The Jurassic is recognised as a relatively warm period globally, with a low equatorial–polar temperature gradient. Although the evidence for any “notable” global climate change over its time span is inconsistent, variation certainly occurred (Parrish 1991; Hallam et al. 1993; Grant-Mackie et al. 2000).

Major advances have been made in recent years in techniques for quantitative estimates of the palaeoclimate of angiosperm-dominated macrofossil assemblages (e.g. Wolfe 1995). These techniques rely on the shapes and sizes of angiosperm leaves, which are very responsive to climate, or on the climatic requirement of nearest living relatives. Unfortunately, these techniques are not so applicable to the pre-angiosperm floras of the Jurassic. Non-angiosperm vegetation has a more conservative morphology with respect to climate, and at least until now, no one has devised a similar scheme for the Mesozoic or older. For this reason, most estimates of Jurassic climate have been qualitative. In broad terms, New Zealand lay in relatively high latitudes where growth rings in wood were prominent (Creber & Chaloner 1985).

Rees et al. (2000) used multivariate analysis to illustrate a global floral gradient of genera (or form genera) for the Early Jurassic. They were able to show that microphyllous cycadophytes (in which they included bennettites) and conifers plotted together at one end of an ordination axis, while macrophyllous conifers and ginkgophytes plotted together at the other. They interpreted these two distinct floral associations in terms of hot and dry, and seasonally cool and/or dark (i.e. deciduous), respectively, and they went further to develop a “gradient score” for each Jurassic genus. This allowed them to “assign a value to any Jurassic plant locality, by averaging the scores of its constituent leaf genera” (Rees et al. 2000, 303), and thus place a locality on the ordination axis. For New Zealand, the relevant genera and their scores are: Equisetites, 76; Podozamites, 72; Coniopteris, 72; Cladophlebis, 70; Nilssonia, 55; Elatocladus, 52; Sphenopteris, 51; Taeniopteris, 50; Pterophyllum, 45; Pagiophyllum, 28; Ptilophyllum, 13; Brachyphyllum, 9 and Otozamites, 9. This gives an average of 46 and fits broadly with mid-latitudes (approximately 40–70°). The addition of the unpublished Hausmannia would raise the score slightly to 49 (meaning a move away from the pole).

But do microphyllous conifer leaves indicate a hot and dry climate? The conifer leaves of the New Zealand Jurassic are small and are of similar size to those living now in cool, ever-wet and evergreen New Zealand forests, particularly those living in the south, and in similar regions in Tasmania and Patagonia. Even the largest New Zealand Jurassic conifer leaves, those of Podozamites, are smaller than those of Agathis australis, which lives today in northernmost New Zealand. This suggests broadly similar temperatures to present-day mid- and southern New Zealand, and cooler than the far north. These figures generally agree with those determined for sea water from belemnite guards (Clayton & Stevens 1967; Stevens 1971; Stevens & Clayton 1971). With respect to T. daintreei, Drinnan & Chambers (1985, 92) wrote “It is tempting to suggest that the abundance of leaf remains in some deposits may represent autumnal banks of leaves and that the species was deciduous.” Whereas this might suggest seasonally very cold conditions, a deciduous habit might also reflect a long polar night in high-latitude conditions.

For the Curio Bay forest environment, Pole (2001) suggested that the climate was less humid, although not arid, with seasonal rainfall. Coal is virtually absent from the system. By contrast, Thorn (2001) noted fossil forest horizons at Kawhia Harbour that were rooted in coal, and she interpreted conditions as comparable to the modern mesothermal zone – with humid, warm conditions, high average rainfall and with large storms.

Biodiversity

Based on foliage macrofossils, New Zealand as a whole has a relatively low diversity of Jurassic plants (with a total of about 43 species); also the per-site diversity is strikingly low. Many assemblages contain little more than Cladophlebis and Taeniopteris, with the Elatocladus group of conifers the next most abundant (Fig. 2). Most other taxa are much less common. For instance, in more than 25 years working in the MT Jurassic, the author has collected only one Ptilophyllum. The New Zealand flora shared species with surrounding Gondwanan areas, such as Australia, Antarctica and Patagonia (e.g. Grant-Mackie et al. 2000), but there is no evidence yet of stronger ties with any particular one of those areas. The New Zealand Jurassic seems to have been essentially a depauperate version of the Australian flora. If the identification of Hausmannia is correct, then an assemblage from the Otara coast correlates with that described from Queensland by McLoughlin & Drinnan (1995).

New Zealand's low-diversity Jurassic flora might be partly a function of a very unstable landscape, but is more likely a reflection of a suboptimal climate at high latitudes. Some low-diversity communities are likely to reflect specific habitats rather than climate per se, and such associations would match similar low-diversity communities elsewhere on the globe. For instance, in the southernmost MT the author has found exclusively Equisetum-dominated zones (M. Pole, pers. obs.), which might be similar to the extensive Equisetum stands envisaged by Batten (1974) for the English Wealden in areas of shallow standing water, such as delta tops.

As noted above, Rees et al. (2000) assigned Jurassic leaf genera globally to 10 morphological categories and found that maximum morphological diversity was at mid-latitudes in both hemispheres, peaking at 9–10 “morphocategories”. The New Zealand Jurassic had at least four morphocategories (sphenophytes, ferns, microphyllous cycadophytes, and unassigned and microphyllous conifers) and lacked ginkgophytes and macrophyllous conifers. This was, therefore, somewhat below the maximum achievable morphodiversity.

The total macrofossil diversity of New Zealand is comparable with that found in just the single Mid-Jurassic (cf. Hunter et al. 2005 for discussion of the age) locality of Hope Bay, on the Antarctic Peninsula, where 43 species were accepted by Gee (1989) and 37 by Rees & Cleal (2004). Continuing research may add a few genera to New Zealand's Jurassic flora, but it may also reduce the species list. In any case, the diversity of most localities is likely to remain low.

Conclusions

There is an abundant record of New Zealand's Jurassic vegetation. The combined flora was relatively low in diversity, dominated by conifers, ferns, bennettitaleans, pentoxylaleans and locally, equisetaleans. Other groups, such as ginkgophytes, which are important elsewhere, were absent. Some of the species documented from New Zealand, such as the cosmopolitan Coniopteris hymenophylloides, were already known from elsewhere and some described from New Zealand have since been discovered elsewhere, as was P. bartrumi in Australia, but many of the species from New Zealand remain endemic taxa. However, since these species were described, the Jurassic floras in surrounding lands, such as Antarctica, have been revised and in some cases re-revised. Revision of New Zealand's Jurassic taxa is long overdue and it might be found that there is less endemicity and stronger links with other regions. However, New Zealand's palaeolocation remains unclear. All of its component pieces lay in relatively high southern latitudes, but the question of whether they were supra- or sub-polar might benefit from further geophysical evidence. With revision of the flora and a clearer picture of where it grew, New Zealand provides a key piece of the global Jurassic picture.

Acknowledgements

The author gratefully acknowledges the helpful comments of S. McLoughlin, S. Turner and V. Vajda. This is a contribution to IGCP 506: Marine/Non-marine Jurassic correlation and was performed within the framework of the Lund University Centre for Studies of Carbon Cycle and Climate Interactions (LUCCI).