RESULTS

Monocot cuticle was found in 35 out of the 121 samples investigated. Seventeen types of fossil cuticle are distinguished in this paper. For two of these, CUT-Mo-EII and CUT-Mo-FEH, monocot affinities are not certain, but are presented here for convenience. For convenient reference the cuticle of the most common, or large New Zealand monocots, is illustrated. An interesting observation is that Cordyline, Phormium, Rhopalostylis, which are all prominent components of the vegetation in various parts of New Zealand today, have not yet been located in the fossil cuticle record. The fossil monocot cuticle includes Astelia, Pandanaceae, Rhipogonum, and Typha (also present as seeds). The identification of Rhipogonum is important because it is one of the few fossils that can be identified as a climber (the only other one in the MG is an unknown genus of the Menispermaceae).

The relatively few specimens reported here certainly represent what would have been a more prominent monocot presence in the vegetation. Most extant monocots have such delicate cuticle that it is difficult to prepare it in the laboratory, and hence would be unlikely to survive both burial and the procedure to obtain dispersed fossil cuticle. The cuticle fragments described here are therefore unusual—the tougher fragments of the tougher species. Typha seeds are abundant in BL-31; unfortunately this locality was a small lens of mudstone which seems to have been destroyed during removal of an adjacent pine tree by the Department of Conservation. This is a notably low-richness locality, which is the type locality for the conifer Retrophyllum vulcanense (Pole 1992). The cuticle fraction is almost entirely Retrophyllum, as well as an extinct, unidentified dicotyledonous leaf, and the Typha seeds. This suggests open, marshy conditions, and is consistent with R. vulcanense being a conifer which favoured standing water or swampy conditions (Hope and Pask's 1998 illustration of extant Retrophyllum growing partially submerged in the Plaine des Lacs in New Caledonia may be a reasonable modern analogy).

Only seven out of 58 samples (12%) in the St Bathans Paleovalley have monocot cuticle, the most in one sample is GL-01 with three taxa. In Southland 22 out of 62 samples (36%) have monocot cuticle, with five taxa occurring in one sample (Sthd-163) and four in another (Sthd-055). In my opinion this is not related to any difference in alteration of the sediments between the two basins which may have preferentially destroyed the more delicate monocot cuticle. In some cases it may be due to fluvial reworking in the St Bathans Paleovalley, which was unlikely to have been a factor in Southland. However, it may be due to a greater prominence of monocots in the vegetation in Southland. In sample Sthd-163, which has the highest monocot richness, there are no other cuticle taxa. This suggests a particular monocot-dominated habitat, probably a particular swamp or wetland habitat. The sample with the second-highest richness of monocots (Sthd-055) had the highest richness of taxa (12) in all the Southland samples. Hence, its high number of monocots may be partly a function of simply being a rich assemblage.

Although identification of these specimens is challenging, they add to the documented biodiversity of the Miocene in New Zealand. Even as unidentified taxa they will help distinguish assemblages stratigraphically and perhaps ecologically.