INTRODUCTION

The Lauraceae has been regarded as a component of New Zealand plant fossil assemblages at least since Holden's (1982) work on the Early Miocene of Murchison. Pole described Lauraceae leaves from the Manuherikia Group (Pole 1993a) and from the Foulden Hills Diatomite (Pole 1996), also of Early Miocene age. In the case of Holden, and for some of Pole's work, these were impression fossils, and the criteria that authors have used to identify leaf impressions as Lauraceae are generally poorly defined (although in Holden's case, likely correct). For instance, entire-margined leaves with an acrodromous development of venation (sensu Pole 1991), particularly if the higher order venation is strong and percurrent, have often been placed in the family. While this kind of leaf architecture can be found in other families, the cuticle has a distinctive structure (e.g., Bandulska 1925; Pal 1978; Avita and Inamdar 1981; Hill 1986; Bakker et al. 1992), and if it is preserved, not only can it help confirm the family identification, but it opens up the possibility of identification to generic level. Small fragments of cuticle mean that Lauraceae may be identified even without knowledge of the leaf architecture.

Cuticular preservation has shown that Lauraceae are an almost ubiquitous component of dispersed cuticle assemblages in New Zealand extending back to at least the Late Cretaceous (Pole 1993a and unpublished data). As now, Lauraceae seem to have been a common component of most rainforests. They also have relatively robust cuticle—if cuticle is preserved at all in a fossil assemblage, it is likely to include Lauraceae. Lauraceae are an excellent example of the macrofossil record complimenting the palynological record, as the pollen has thin exine, or has extremely limited production, so are essentially invisible palynologically (Erdtman 1952; Macphail 1980). Generic names applied to impressions of Lauraceae-like leaves must be suspect, as identification of extant material from leaf morphology alone is difficult, if not impossible (e.g., Hyland 1989). With the addition of cuticular information, identification of fossils to the Lauraceae has become routine. Hill (1986) noted that extant Australasian Lauraceae cuticle has a distinctive structure whereby the guard cells are typically overarched by a pair of subsidiary cells (a paracytic structure), and a cuticularised flange develops between the two cell types. This structure uniquely distinguishes Lauraceae from all other genera in the Laurales. However, as pointed out by Upchurch and Dilcher (1990), one other family includes species with this structure – the Myristicaceae. They also stated that Myristicaceae possess trichome bases that are distinct from those in Laurales. Based on my own cuticle reference collection covering about 4000 species in more than 1500 genera and 280 families, this can be confirmed. Myristicaceae have multi-celled trichome bases, a morphology that shares some similarity to Proteaceae and Platanaceae (e.g., Carpenter et al. 2005). My reference collection and all published illustrations (e.g., Hill 1986; Christophel and Rowett 1996; Christophel et al. 1996; Vadala and Greenwood 2001) agree that extant Lauraceae have simple, poral trichome bases.

While identification of well-preserved fossils to family level is routine, identification to generic level is challenging. Workers who have identified fossils as Lauraceae with the aid of cuticle have generally described them using broad organ-genera, such as Laurophyllum Goepp. (e.g., Weyland and Kilpper 1963; Kvaček and Bek 1966; Hill 1986; Kvaček 1988; Conran and Christophel 1998). In a few European cases they have been placed into extant genera (for example; Ferguson 1974; Bek et al. 1996; Uzunova and Stojanova 1999). However, Christophel and Rowett (1996) and Christophel et al. (1996) have demonstrated that Australian Lauraceae genera can be distinguished based on cuticular morphology. This information has since been applied to Australian Lauraceae fossils (Vadala and Greenwood 2001). The purpose of this paper is to document the range of Early Miocene Lauraceae, and Lauraceae-like, cuticle morphology from two regions in southern New Zealand, the Manuherikia Group, and the Gore Lignite Measures.