PENTOXYLACEAE Sahni, 1948

Pole, Mike, 2019, Middle-Late Jurassic plant assemblages of the Catlins coast, New Zealand, Palaeontologia Electronica (a 51) 23 (3), pp. 1-48 : 21-23

Family PENTOXYLACEAE Sahni, 1948 Genus TAENIOPTERIS Brongniart, 1832

Remarks. Taeniopteris is the second most common taxa of the Catlins coast. The use of Taeniopteris in the present work follows Harris (1932) and Rees and Cleal (2004), who reserve it for material lacking anatomical details, which might otherwise place it in genera such as Macrotaeniopteris, Nilssonia, Nilssoniopteris , or Nipaniophyllum . On this basis, Rees and Cleal (2004) placed their Hope Bay material in Taeniopteris .

Arber (1917) recognised four species of Taeniopteris in the New Zealand Jurassic: T. crassinervis , T. daintreei , T. thomsoniana , and T. vittata . These were based on features of venation — its conspicuousness, spacing, and whether the veins forked near the midrib or within the lamina. Edwards (1934) regarded T. daintreei (McCoy, 1875) as a mis-print for the older (McClelland, 1850) name of T. spatulata . He also suggested that Arber’s (1917) Curio Bay T. vittata were an extreme form of T. spatulata , although made no comment that T. vittata is an even older name. Mildenhall (1970) included both T. daintreei and T. spatulata in his list of valid names, but repeated Edwards’ (1934) point that T. daintreei was probably a misprint for T. spatulata . Since the reviews by Douglas (1969) and then Drinnan and Chamber (1985) of T. daintreei , this name has been widely used for Australasian material, although they did not specifically deal with the issue of T. spatulata .

The most widespread Taeniopteris in the Jurassic of the Catlins Coast (exemplified by material from Curio Bay) has leaf widths ranging from about 5 mm to at least 27 mm, but too few specimens are available from any single locality to recognise a mode. A quantitative study by Blaschke and Grant-Mackie (1976) documented distinctly different Taeniopteris widths in collections from the Jurassic of Port Waikato and Clent Hills. The nar-

POLE: CATLINS COAST JURASSIC rower (mostly 3–13 mm) Port Waikato specimens were included in T. daintreei , and the wider (5–25 mm, mostly 10–22 mm) Clent Hills specimens into T. thomsoniana . Their use of T. daintreei instead of T. spatulata was apparently due to the linking of the later with the Pentoxylales, which they regarded as circumspect for the New Zealand material. However, this ought to be irrelevant for purely impression material. Their use of T. thomsoniana for the Clent Hills material follows Arber (1917). However, there are not clear morphological differences between the species described by Blaschke and Grant-Mackie (1976), and they probably could not be distinguished if they were found together.

As per Arber (1917), the most widespread Catlins Taeniopteris is referable to T. daintreei , whereas in the taxonomic scheme of Blaschke and Grant-Mackie (1976), a simple interpretation could place it into a broadly defined T. thomsoniana .

However, Taeniopteris spatulata has priority and is used in the present work. Taeniopteris vittata (Brongniart, 1828) is an even older name, but is ignored here as a Northern Hemisphere species. Person and Delevoryas (1982) and Rees and Cleal (2004) warned that Taeniopteris can exhibit a very broad range in size and were unimpressed with size as a criterion to distinguish species. The later maintained there was a gradation of Taeniopteris width in their Antarctic material from 9.5 up to 100 mm wide (although they did recognise two forms). This very large range of width encompasses the largest New Zealand Taeniopteris . Some early authors (e.g., Schimper, 1869; Feistmantel, 1877) placed relatively large leaves in Macrotaeniopteris . However, Arber (1905, 1917) maintained that size and shape alone were not sufficient to distinguish this (and other Taeniopteris -like taxa) and kept them all in Taeniopteris . Harris (1932) reserved Macrotaeniopteris for large leaves that had cycadean stomata. Hector (1886) figured a Macrotaeniopteris lata from Mataura Falls, but without a scale. The specimen was seen by Arber (1917) who noted the width as 66 mm, with about 10, mostly simple veins, per 10 mm of length. He placed this specimen, along with other material about 80 mm wide from Mataura, in T. crassinervis . Gupta’s (1986) review of T. crassinervis included Arber’s (1917) material in synonomy. He regarded T. crassinervis as most closely resembling T. lata Oldham (Oldham and Morris, 1863) in terms of size and shape, but gave three characters of the veins to distinguish them. A fragment with a width of about 100 mm from the Drumduan Group of Nelson was illustrated by Johnston et al. (1987) as T. cf. T crassinervis . The largest Taeniopteris specimens found in the current study are two superimposed leaves from Curio Bay that are about 58 mm wide. My personal preference is to continue to draw attention to these larger forms by recording them as T. crassinervis but recognising that they may simply be the extremes of T. daintreei / T. spatulata .

At the opposite end of the width scale, the Waikawa Syncline localities are dominated by a remarkably narrow Taeniopteris . It is often found as overlapping masses on bedding in massive sandstone boulders, making it both difficult to collect and poorly preserved. The widths range from about 2.5–4.0 mm, that is, comparable to only the very narrowest Taeniopteris widths cited by Blaschke and Grant-Mackie (1976), White (1981), and Rees and Cleal (2004) and also to “linear, needle-like” specimens in McLoughlin and Drinnan (1995, fig. 4E). The existence of this very thin morphology in a restricted stratigraphic range has clear stratigraphic and perhaps climatic/edaphic potential. However, here is no clear taxonomic precedent that appears to apply, and for the present, an informal name, T. sp. ‘narrow’ is applied.