INTRODUCTION

New Zealand's Miocene sediments have provided a small number of monocot macrofossils. One of the earliest to be described was a palm frond, Seaforthia zeelandica (von Ettingshausen 1887, 1891). This probably came from the same locality (near Cromwell) from which Pole (1993a) described palm fronds, fruits, and flowers and redescribed S. zealandica as Phoenicites zeelandica. Phoenicites is a morphogenus, the usage of which was clarified by Read and Hickey (1972) to apply to one of the limited range of palm frond shapes. As such, it is not a genus in any comparable way to extant genera, and I would no longer use the concept. Phoenicites zeelandica has small differences from New Zealand's single extant species of palm, Rhopalostylis sapida, which is the most southerly palm in the world (Wardle 1991). Impressions of the distinctive net-veined monocot, Rhipogonum, were also described from the same region (Pole 1993b), and Rhipogonum leaves with cuticle were later documented from the Miocene of the Foulden Hills Diatomite near Middlemarch (Pole 1996). Holden's (1982) Cinnamomum miocenicum from Murchison, and Oliver's (1936) Coriaria latepetiolata from Dunedin are likely to also be Rhipogonum. Fossil "coconuts" from the far north of the North Island were identified as Cocos by Berry (1926), but have since been suggested to have a closer relationship with Parajubaea (Endt and Hayward 1997).

These fossils are 'macrofossils'-in the classic sense that they are apparent to the naked eye on the bedding surfaces of splitting sediment, or in the case of the 'coconuts', can be picked from the sediment or even, in the case of the fossil "coconuts" can be picked up from strand lines on the present beach. But another collecting rationale, in the appropriate sediments, involves searching for leaf cuticle, the resistant material which covers the leaf epidermis and which preserves an impression of the epidermal cell morphology. Although fragments of cuticle can be large (as large as the original leaf), the essential features can only be observed with a microscope.

Stebbins and Khush (1961) were probably the first to present a broad review of extant monocot cuticle but by far the greatest contribution has been that of Tomlinson (1960, 1961, 1965, 1969, 1971, 1974, 1982), who included descriptions of epidermal features as part of his broader treatment of the monocots. Dahlgren and Clifford (1982) included a useful comparative treatment of monocot stomatal construction. Typical monocot epidermis can be recognised primarily by having rows of longitudinally oriented stomatal complexes (although Butterfass 1987, included some monocots in his list of plants with transversely oriented stomata) and epidermal cells, a result of the typically parallel-veined leaves. The stomatal complex morphology, in the sense that they typically have a pair of distinct polar and lateral subsidiary cells, overlaps that of many conifers, although the topography and other details are quite different. For instance, in conifers the guard cells are typically embedded below overarching subsidiary cells, and the ends of the guard cells are often partially surrounded by distinct 'polar extensions'. Like other angiosperms, monocots have outer stomatal ledges, projecting from the guard cells, which are absent in conifers. Conifer cuticle is mostly much more robust, and tends to be preserved as recognizable leaves (often small and single-veined). Dunn et al. (1965) also pointed out that monocot stomatal complexes within a leaf are of equal size, or at least do not fall into distinct size classes, which they often do in non-monocot angiosperms. The epidermis of the reticulate-veined monocots is distinct from typical monocots, because the stomata and epidermal cells are not arranged in rows. Identification of taxa in this group relies on direct comparison of the fossils with an extant species.

Over some 15 years of research on Miocene cuticle in southern New Zealand, some small and typically rare fragments of monocot cuticle have been found, as well as some seeds. The purpose of this paper is to describe the cuticle and seed types as a basis for further studies of past biodiversity, ecology, and biogeography.