Proteaceae leaf fossils: phylogeny, diversity, ecology and austral distributions.
"I ... confess to a desire for more work on informative macrofossils." (Johnson, 1998; p. 253)
Proteaceae (~1700 spp.) are central to interpreting the biogeography of the Southern Hemisphere. Each of the vegetated continents and major island groups hosts representatives of the family, and fossils of pollen and sometimes flowers, fruits, wood and leaves are known from across the region, including Antarctica.
Major advances in understanding the relationships of Proteaceae both to other taxa and within the family have been made over the last 15 years or so, mostly through the acquisition of DNA sequence data and their analyses using powerful algorithmic methods. Thus it is now clear that Proteaceae are sister to Platanaceae (APGIII, 2009), and Weston and Barker (2006) recently proposed a new classification of Proteaceae with resolved clades of genera at the ranks of subtribe, tribe and subfamily. This scheme was based on the comprehensive treatment of the family by Johnson and Briggs (1963, 1975, 1981). Improved understanding of phylogenetic relationships at the genus and higher levels continues. For instance, Dryandra is now included in Banksia (Mast & Thiele, 2007), and a new phylogeny for Macadamieae, the most widespread tribe of Proteaceae has been inferred (Mast et al., 2008).
Importantly, the new molecular-based phylogenies can also provide a framework for interpreting the evolution of morphological traits. With respect to leaves, Drinnan et al. (1994) noted the interesting similarity of the pinnate/pinnatifid architecture of many extant Proteaceae to that of certain Cretaceous platanoids, and Carpenter et al. (2005) proposed that distinctive donut-like trichome base scars that often encompass more than one epidermal cell represented a synapomorphy linking Platanaceae and Proteaceae. The common ancestor of Proteaceae was reconstructed by Johnson and Briggs (1975) to have had paracytic stomata, and Carpenter et al. (2005) proposed that (brachy)paracytic stomata were synapomorphic for the family.
The overall potential of leaves as a valuable source of taxonomically and phylogenetically significant traits has long been recognised, but often under-exploited. Stace (1965) stressed the importance of epidermal (cuticular) features, and Hickey and Wolfe (1975) presented a comprehensive synthesis of leaf architectural evolutionary trends across angiosperms within the Cronquist (1968) and Takhtajan (1969) taxonomic systems that prevailed at the time. Recently, Doyle (2007) updated Hickey and Wolfe's (1975) work in the context of molecular phylogenies, and found that many of their views were supported, and in fact that they had anticipated some of the molecular data. With respect to Proteaceae leaves, there is increasing evidence that there are cuticular features that have taxonomic and perhaps phylogenetic significance. For instance, Carpenter (1994) described the leaf cuticular features of two putatively new species of Orites from North Queensland rainforests as respectively "quite distinct from that of the other Orites species" (p. 259), and suggestive of "no particularly close relationship to any existing Oriteae, Knightieae or Embothrieae" (p. 260). The first of these taxa [now Nothorites megacarpus (A.S.George & B.Hyland) P.H.Weston & A.R.Mast] has subsequently been recognised using DNA evidence as in fact not even a close relative of Orites (i.e. Macadamieae: Mast et al., 2008), and the second (now Megahertzia amplexicaulis A.S.George & B.Hyland) appears to be sister to Hollandaea and Helicia (Sauquet et al., 2009a) in subtribe Heliciinae of tribe Roupaleae. Also, Carpenter (1994) concluded that a third taxon of unknown affinity at the time was "most likely to be allied to Helicieae" (p. 291). This taxon was subsequently described as a new species of Hollandaea, H. riparia B.Hyland (Hyland, 1995).
The potential importance of leaf and cuticular traits for informing the past history of Proteaceae is magnified because recent parsimony analyses of morphological characters of pollen (Sauquet et al., 2009a) provide support for only a few of the proposed close relationships between fossil pollen species and extant taxa (e.g. Martin, 1995; Dettmann & Jarzen, 1998). Fossil leaves with preserved cuticles usually possess many more and diverse assessable characters than pollen grains. These characters range from leaf architectural features such as overall size and shape and the nature of venation and tooth types to cuticular features that can be analysed using light and scanning electron microscopy, such as the nature of stomatal form and distribution, trichomes, glands and inner and outer surface ornamentation.
An increasingly important role for well-dated fossils is their use as calibration points in molecular phylogenetic work that seeks to reconstruct the timing of lineage diversifications and their biogeographical consequences. However, one of the potentially severe weaknesses of this approach is that fossils may be included without proper assessment of the validity of their identifications (see Gandolfo et al., 2008). Of particular interest is whether or not such fossils belong to the crown group of a clade, or to extinct stem taxa. In reality, it may rarely be possible to provide conclusive, unequivocal evidence for the position of a fossil. No fossil possesses the full suite of molecular and morphological evidence that is available from an extant plant: fossils are usually only represented by detached organs (i.e. unconnected pollen, leaves, seeds etc. that may or may not co-occur in the same assemblage) and a leaf taxon may only consist of one or a few specimens. However, it is reasonable to propose placement of a fossil in a stem group of a clade based on putative synapomorphies of that clade, and in the crown group of that clade if the fossil possesses putative synapomorphies of a subclade of the crown group (see Magallon, 2004).
Leaf fossils may be significant not only as evidence of past distributions of taxa, bur also as environmental indicators. This significance revolves around the fact that leaves representa large interface with their surroundings, and are sensitive to factors including light, carbon dioxide concentration and humidity. In addition, the anatomy and morphology of leaves are in part determined by edaphic qualities such as nutrient availability. Notably, for instance, it is well known that the remarkably high extent of sclerophylly in Australian Proteaceae is associated with the ancient weathered nature of Australian soils.
In this report I aim to present the current state of knowledge of the leal fossil record of Proteaceae. The report follows the brief review of Proteaceae macrofossils by Weston (2006) and previous assessments of the Australian, New Zealand and Argentinean records by Hill et al. (1995), Pole (1998) and Gonzalez et al. (2007) respectively. I focus on the potential for leaf fossils of Proteaceae to calibrate molecular dates and on what Proteaceae leaf fossils can tell us about vegetation and biogeography in the Southern Hemisphere through rime.
Materials and Methods
No comprehensive family-wide cladistic assessment of leaf and cuticular characters (or of other morphological traits) has been undertaken for Proteaceae. However, leaf specimens and cuticular preparations of numerous species and from all 81 genera [80 genera recognised by Weston and Barker (2006) in their new classification system of the family, minus Dryandra (Mast & Thiele, 2007), and plus two new genera described by Mast et al. (2008)] were available for comparison with fossils. These are housed in the paleobotanical collection at the University of Adelaide. Proteaceae leaf fossils that have been formally or otherwise described in the literature from the Upper Cretaceous, Paleogene (Paleocene, Eocene and Oligocene) and Neogene (Miocene, Pliocene and briefly, the Pleistocene, where the record is mostly of extant species only), as well as some new fossils from Australia and New Zealand are considered here. Fossils with cuticular preservation are emphasised, because most of the many late 19th and early 20th century records are too poorly preserved to yield adequate details for taxonomic determinations (see Hill, 1988 for discussion and Australian references).
A brief review of the chronological record of fossils is presented. Then, the fossils are listed in the context of a scheme of extant subfamilies, tribes and genera and their generic and species diversities (Weston & Barker, 2006; Mast et al., 2008; Sauquet et al., 2009a) and evaluated with respect to characters that may be phylogenetically informative. Finally, I present a discussion of how Proteaceae leaf fossils are relevant to interpreting past community assembly, diversity and ecology.
This review focuses on foliar (leaf and cuticular) fossils, in part because most macrofossils are leaves, but readers should be aware that a range of other Proteaceae organs appear in the fossil record. Thus, apart from the extensive literature on the important Proteaceae pollen record (e.g. Dettmann & Jarzen, 1998; Askin & Baldoni, 1998), there are also fossils of Musgraveinae inflorescences from Australia (Christophel, 1984; Christophel & Greenwood, 1987), Proteaceae wood from Argentina (Pujana, 2007) and several reports of fruits from Australia and elsewhere. These fruits include described Orites follicles from Patagonia (Gonzalez et al., 2007) and diverse taxa from Australia, including Banksia infructescences (Cookson & Duigan, 1950; McNamara & Scott, 1983), Grevillea-like follicles (Dettmann & Clifford, 2005) and fruits closely comparable with those of Eidothea (Rozefelds et al., 2005) and Macadamieae (Rozefelds, 1990, 1992; Dettmann & Clifford, 2010). Many of these fossils were described in the late 19th and early 20th centuries but not recognised as Proteaceae (see the above references for original sources).
Results and Discussion
Chronology of Leaf Fossil Records
So far, no leaf fossils are known from the Mesozoic that are convincingly proteaceous. This is unfortunate because fossils of the sister group of Proteaceae, Platanaceae, are widespread in the Northern Hemisphere and date from the Lower Cretaceous (Crane et al., 1993). The sister relationship implies that the Proteaceae lineage is also this old, and any fossils would be of great interest with respect to evolution in Proteales.
Lower-mid Cretaceous Australian material (from the Eromanga Basin) that is known to contain angiosperms shows no evidence of leaves that are assignable to Proteaceae (McLoughlin et al., 1995), nor dispersed cuticles (Pole, 2000). This is consistent with the palynological evidence which suggests an Upper Cretaceous arrival of ancestors in Australia (Dettmann & Jarzen, 1998). This arrival was followed by a massive diversification and expansion of Proteaceae in the Australian-Antarctic region during the Upper Cretaceous (Santonian to Maastrichtian) from about 85 million years ago (Dettmann & Jarzen, 1998). However, the potential for the presence of Upper Cretaceous leaf fossils in Australia is low because of a scarcity of appropriately aged exposures. Upper Cretaceous leaf fossils would be of interest with respect to the apparent diversification in the family and the degree to which Proteaceae spread across the Southern Hemisphere at a time when Australia, New Zealand and South America were directly connected via Antarctica, and when there may have been at least a stepping stone route for biota between Australia and Africa via the Kerguelen Plateau (see Ali & Aitchison, 2009).
Interestingly, leaf fossils in an Upper Cretaceous deposit near Dunedin in New Zealand show no evidence of Proteaceae cuticles despite the close proximity of the region to southeastern Australia at that time, and the presence of at least 15 other angiosperm cuticle taxa (Pole, 1998). This lack of Proteaceae may reflect the fact that fossil pollen from New Zealand Cretaceous sediments do not show anywhere near the same extent of proteaceous diversification as in southeastern Australia (Pole, 1998). However, latest Cretaceous--Paleocene sediments in New Zealand potentially contain the oldest known Proteaceae leaves (Pole, 1998; Pole & Vajda, 2009) which are Banksia-like in form and described as Dryandra comptonieaefolia Ett. Useful cuticle has not been recovered from these leaves, and although dispersed cuticles with Proteaceae-type trichome bases also occur in the sediments, the cuticles that have been recovered so far lack stomata that might strengthen an argument for the family (Pole & Vajda, 2009). Proteaceae pollen is also known from the Upper Cretaceous of South America (Antarctic Peninsula and Patagonia), but as in New Zealand, is not highly diverse compared with coeval floras in Australia (Askin & Baldoni, 1998), and the sites do not appear to host convincing leaf fossils (Gonzalez et al., 2007).
The oldest definite Proteaceae leaves currently known come from late Paleocene sediments in southeastern Australia. There, the Lake Bungarby site in New South Wales contains Banksieaephyllum taylorii R.J.Carp., G.J.Jord. & R.S.Hill leaves as well as at least four other taxa of Proteaceae (Carpenter et al., 1994). Palynologically coeval (Taylor et al., 1990) fossils from a nearby site at Cambalong Creek have also been referred to several taxa of Proteaceae (Vadala & Greenwood, 2001), including Banksieaephyllum (Vadala & Drinnan, 1998). There is also abundant and widespread evidence of diverse Proteaceae in the form of dispersed cuticular remains in central Australian Eyre Formation (Alley, 1998) sediments of late Paleocene to early Eocene age (R.J. Carpenter, unpubl.). Overall, these fossil leaf data support the palynological evidence that the family was well differentiated in Australia prior to the Eocene, and were probably important components of the vegetation (Dettmann & Jarzen, 1998). Proteaceae also occur in the Paleocene of New Zealand, and include an incomplete but cuticle-bearing pinnatifid leaf described by Pole (1997) as Lomatia novaezelandiae.
Leaves of Proteaceae and cuticles with features of the family are typically abundant and diverse in all Australian Eocene and Oligocene sediments that have been studied, with the apparent exceptions of the middle Eocene Nerriga site in New South Wales (Hill, 1982; Christophel, 1994), and the early Oligocene Little Rapid River site in Tasmania (Jordan et al., 1998). Some of the fossil material shows pronounced similarities to extant species, whereas the overwhelming majority of taxa are clearly extinct. Parafatsia subpeltata Blackbum from the middle Eocene Maslin Bay assemblage of South Australia (Blackburn, 1981) is a notable example of an extinct leaf species (Carpenter et al., 2006), because it is palmately lobed, a condition that does not occur in extant Proteaceae leaves, but that does occur in species of Platanus. Other cuticle types from central Australian Eocene sediments can clearly be regarded as of extinct species of Proteaceae because these cuticles show sometimes bizarre forms of surface ornamentation that do not occur in any extant species. Interestingly, it has previously been emphasised that the Australian Eocene hosted a remarkable abundance and diversity of Proteaceae pollen types (summarised by Dettmann & Jarzen, 1998), including large, ornate pollen types that are extinct. These pollen types and the unusual leaves and cuticles may well have belonged to stem lineages of Proteaceae.
Proteaceae leaves have also been recognised in New Zealand and South American Eocene and Oligocene assemblages, but so far very little cuticular material has been reported. Thus, lobed leaves alike some extant Proteaceae are known from the Eocene of New Zealand (Pole, 1998), and three fossil leaf species of Lomatia, one of Roupala and one indeterminate Proteaceae taxon are recognised in Argentinean assemblages, based on well-preserved leaf architectural features (Gonzalez et al., 2007). Gonzalez et al. (2007) considered that the numerous Paleogene records of Proteaceae leaves from the Antarctic Peninsula and elsewhere in this region must be regarded as doubtful. I am not aware of any convincingly proteaceous leaves from the Paleogene anywhere else in the Antarctic area, or from Africa, where Proteaceae pollen has been reported from the latest Cretaceous to Paleocene Arnot Pipe of South Africa (Scholtz, 1985).
Leaf fossils of Proteaceae are also common in the Miocene of Australia and New Zealand. Several dispersed leaf cuticle types from New Zealand appear remarkably similar to Australian rainforest taxa (Carpenter, 1994; Pole, 1998), and there are recent records of Banksia (Carpenter et al., 2010a) and Persoonieae (Carpenter et al., 2010b) from an Oligo-Miocene site in southern New Zealand. Banksia and Persoonieae are now overwhelmingly Australian in distribution.
Pole (2007a) also noted the presence of Proteaceae leaf material in the Pliocene of New Zealand, including at least one taxon that subsequently became extinct there. Numerous early-middle Pleistocene records from western Tasmania suggest the presence of several extant Tasmanian species by that time, as well as extinct species of extant genera and perhaps also an extinct genus (Jordan, 1995).
Leaf Fossil Records of Extant Clades
There are no fossil records of this subfamily, now represented by only Bellendena montana R.Br., a species endemic to Tasmanian mountains. Carpenter et al. (2005) showed that many stomata in Bellendena are not brachyparacytic, but have a more or less laterocytic (or anomocytic) arrangement and are thus similar to Platanus stomata. Considering the evidence at the time that Bellendena was sister to the rest of Proteaceae (Hoot & Douglas, 1998), Carpenter et al. (2005) raised the possibility that the stomatal types of Bellendena and Platanus may represent a shared ancestral state. However, the most recently published molecular phylogenies of Proteaceae, including the supertree synthesis of Weston and Barker (2006), resolve Bellendena as sister to Persoonioideae, so that the stomata in Bellendena could be independently derived with respect to other Proteaceae.
Apart from their unusual stomata, the leaves of B. montana are distinctive in being amphistomatic and in having variable concentrations of often conical papillae on their surfaces. However, the leaf laminae are entirely glabrous (Weston, 1995), meaning that B. montana cuticles also lack trichome bases, and any fossils would be difficult to recognise as having affinity to Proteaceae.
There are no convincing records of Persoonioideae (five genera; 104 spp.) from Australia, despite Persoonia now being one of the largest Australian genera of the family with 100 species. However, Carpenter et al. (2010b) described two species in the new genus Persoonieaephyllum from the Oligo-Miocene of New Zealand. An apparent synapomorphy for subfamily Persoonioideae is the presence of very large stomata, sometimes with guard cell lengths as much as 100 [micro]m (typical lengths elsewhere in the family and in angiosperms generally are less than half this). The New Zealand fossils also showed stomata aligned more or less parallel with the long axis of the leaves and undulate cell wall outlines, other features typical of Persoonioideae. Carpenter et al. (2010b) argued that the leaves belonged to tribe Persoonieae because they conformed to those of extant taxa and were distinctly different from leaves of Placospermum coriaceum C.T.White & W.D.Francis, the only other representative of the subfamily. In particular, where preserved, the fossil leaf specimens were quite linear and exhibited more or less parallel venation, apparently derived states with respect to the generally large leaves with brochidodromous venation of Placospermum. Carpenter et al. (2010b) noted that the fossils could belong in the crown group of Persoonieae, but assignment at that level was not possible on the available evidence.
Other than leaf fragments of Agastachys odorata R.Br. from Pleistocene sediments of western Tasmania (where this species is now endemic) (Jordan et al., 1991; Jordan, 1995), there are no fossil records of Symphionematoideae (two genera; three spp.). The leaves (and cuticles) of Agastachys odorata R.Br. are not particularly distinctive, but the two species of Symphionema typically have pinnately to tripinnately dissected leaves (Telford, 1995) that would attract attention if found as fossils.
There is only a scant leaf fossil record of Proteoideae, despite this being a large subfamily (25 genera; >600 spp.). An important feature of Proteoideae is its great extant diversity in the Mediterranean-climate heathlands of South Africa and southwestern Australia, and the fact that there are several examples of proteoid taxa that have disjunct sister groups in these regions. For example, the African genus Aulax is strongly supported as sister to the Australian genus Petrophile (Hoot & Douglas, 1998; Barker et al., 2002, 2007; Weston & Barker, 2006; Sauquet et al., 2009a,b).
Apparently all species of Proteoideae show an absence of both cuticle surface striations and inner surface granulations, and all with the exception of Faurea and Dilobeia have trichome bases associated with only one (very rarely two) basal epidermal cell(s) (Carpenter & Jordan, 1997). Two major clades of Proteoideae were recognised in the most recent molecular phylogenetic analysis of the subfamily (Sauquet et al., 2009b). All species of Proteoideae II (that comprises the South African Cape Proteaceae plus the Australian genera Franklandia, Petrophile, Adenanthos and Isopogon), with the exception of Faurea also show more or less parallel-aligned (with the long axis of the leaf) stomata and amphistomaty. The genera of Proteoideae I (plus its possible sister Eidothea) with the exception of Stirlingia and Conospermum differ from those of Proteoideae II in that their stomata are generally not aligned more or less parallel.
Several Proteoideae have pollen types that have probable phylogenetic significance within Proteaceae (Sauquet & Cantrill, 2007). Beauprea pollen is of special interest because it has distinctly colpoid apertures, a state seen nowhere else in extant members of the family, but recognised in fossil pollen species from across the Southern Hemisphere (Pocknall & Crosbie, 1988; Milne, 1998). It is reasonable to assume from this fossil evidence and cladistic assessment of pollen characters that the Beauprea lineage, now represented by only Beauprea in New Caledonia, dates to the Upper Cretaceous (Sauquet et al., 2009b), and that therefore at least some recognisable foliar material could also have become fossilised in some region of the Southern Hemisphere from that time onwards. However, it seems unlikely that the leaves of Beauprea can be defined by any synapomorphies, and they lack distinctive features. In fact, trichomes or trichome bases are effectively absent from extant Beauprea leaves, meaning that any fossils linked to the genus can even be questioned as belonging to Proteaceae. Despite its apparently long history, leaf fossils that could belong to Beauprea on the basis of likeness to extant species are scant, and are only found in much younger sediments than the oldest pollen records. These possible Beauprea fossils are a single, poorly preserved leaf from the Oligocene of Tasmania (Carpenter & Jordan, 1997), cuticular material from the Pliocene of New Zealand assigned to the genus by Pole (2007a), and some new specimens from the OligoMiocene Gore Lignite Measures at Newvale Mine, southern New Zealand (R.J. Carpenter, J.M. Bannister & D.E. Lee, unpubl.). These Newvale specimens are closely comparable with extant Beauprea in available leaf form and cuticular details (compare Figs. 1 and 2) and co-occur with Beauprea-type pollen, which is found in relatively high frequencies in the Gore Lignite Measures (Pocknall & Crosbie, 1988).
There are no convincing or adequately published records of other Proteoideae fossils from the pre-Quaternary, despite the crown group of the subfamily dating to the Upper Cretaceous (Sauquet et al., 2009b). Leaf fossils could be inferred as belonging to Proteoideae through the combined presence of amphistomaty, parallel-aligned stomata, trichome bases associated with a single basal epidermal cell, and an absence of both outer cuticle surface striations and inner surface granulations. Lange (1978) proposed that some Western Australian Eocene cuticular fossils belonged to Synaphea, but Carpenter and Pole (1995) strongly doubted this. The taxon "aft. Conospermum" was listed as a rare macrofossil in the Yallourn coals of Victoria, Australia by Blackburn and Sluiter (1994), but not illustrated. Pole (1992) assigned a toothed leaf fragment with well-preserved cuticle from a mid--late Eocene site in Tasmania to Cenarrhenes nitida Labill. However, the affinities of this fossil could just as well lie elsewhere, including with Beauprea (Carpenter & Jordan, 1997). Cenarrhenes nitida leaves do occur in Pleistocene sediments from western Tasmania, where the species is extant (Jordan, 1995).
Grevilleoideae is the largest and most diverse (48 genera; > 1100 spp.) extant subfamily of Proteaceae, and has by far the best leaf fossil record, especially when compared with the other large subfamily, Proteoideae. The Grevilleoideae fossils include records of taxa from each of the four tribes; i.e. Roupaleae, Banksieae, Embothrieae and Macadamieae. Numerous other fossils have been suggested as belonging to Grevilleoideae, and many of these have been assigned to Euproteaciphyllum, a form genus which was erected to encompass Proteaceae leaves of unknown affinity with preserved cuticular details (Carpenter & Jordan, 1997; Jordan et al., 1998). Inner cuticle surface granulations (often limited to the subsidiary cells) and outer surface striations are characteristic features of Grevilleoideae (Carpenter & Jordan, 1997).
a. Tribe Roupaleae (13 genera; ~170 spp.). Paleogene leaf fossils of Orites (subtribe Roupalinae) have been described from Tasmania (Carpenter & Jordan, 1997; Jordan et al., 1998) and can be considered as belonging to crown group Orites on the basis that each has synapomorphies for a subclade of the genus. Among the extant clades recognised using molecular evidence are sister group relationships between O. excelsus R.Br. (including O. fragrans F.M.Bail.) and O. diversifolia R.Br., and the Tasmanian species O. acicularis R.Br. and O. milliganii Meisn. (A. R. Mast & P.H. Weston, unpubl.). Orites excelsoides was described from the early Oligocene Leven River site by Carpenter and Jordan (1997) and can be regarded as belonging to the lineage now represented by only O. excelsus. The leaf fossils are difficult to distinguish from the extant species, and in particular share a uniquely derived feature in Orites of having wax on the abaxial cuticle surface that completely obscures the positions of the stomata (Carpenter, 1994; Carpenter & Jordan, 1997). Two other species from the early Oligocene, Orites milliganoides G.J.Jord., R.J.Carp. & R.S.Hill and O. scleromorpha G.J.Jord., R.J.Carp. & R.S.Hill belong to the clade comprising O. acicularis and O. milliganii (Jordan et al., 1998). This is because these species are the only Orites species to share a distinctive type of sclerified hypodermis, and this type is not found elsewhere in Proteaceae (Jordan et al., 1998, 2005). The Orites fossils described above have so far not been used in dated phylogenies. Jordan (1995) also reported leaves of three of the four extant Tasmanian Orites species in Pleistocene sediments from western Tasmania, and described another distinctive leaf type as O. truncata G.J.Jord.
Christophel et al. (1987) reported lobed leaves of Proteaceae from the late middle Eocene Anglesea site of Victoria, Australia that they compared closely with the then undescribed extant taxon now known as Megahertzia amplexicaulis. They drew particular attention to what they described as an auriculate leaf base on the fossils. As discussed previously, Megahertzia appears to be sister to Helicia and Hollandaea in Roupaleae (Sauquet et al., 2009a). Sessile and auriculate (or amplexicaul) leaf bases are a highly distinctive feature of M. amplexicaulis (George & Hyland, 1995) and are probably autapomorphic within at least extant Roupaleae. Rowett and Christophel (1990) showed one illustration of the cuticle of the Anglesea fossil leaves and described it as having prominent cuticular striations and large multicellular hair bases. The cuticle of Megahertzia was described in detail by Carpenter (1994) (as Orites code #752). There is no doubt that fossil and extant cuticles are closely comparable. In particular, both taxa share prominent fine striations on both outer surfaces, relatively rare trichome bases that are associated with numerous basal epidermal cells, sinuous to buttressed anticlinal cell wall cuticle and granular inner cuticle surfaces. Although more work is required, the Anglesea taxon is presently regarded as belonging to Megahertzia (R.J. Carpenter and D.C. Christophel, unpubl.). Interestingly, the nearest extant affinity of the Eocene fossil pollen species Proteacidites latrobensis Harris is considered to be Megahertzia (Macphail, 1999). Pollen grains of both fossil (Harris, 1966; Stover & Partridge, 1973) and living (Sauquet et al., 2009a) taxa have distinctive small spines (or "scrobiculi": Harris, 1966) scattered on the sexine. Supratectal spines were scored as a derived character state in the pollen morphological matrix of Sauquet et al. (2009a), and were otherwise only recorded by the authors in tribe Embothrieae, subtribe Embothriinae and in the more distantly related Agastachys (subfamily Symphionematoideae) and Aulax (subfamily Proteoideae).
A number of other fossils have been ascribed to taxa within tribe Roupaleae, but none have been adequately justified. For instance, Carpenter and Pole (1995) proposed that cuticles from the late middle Eocene of Western Australia were closely comparable with Darlingia ferruginea J.F.Bail., and Pole (1998, 2007a) assigned cuticles from the Miocene and Pliocene of New Zealand to Helicia and Knightia respectively. However, in the absence of demonstrable apomorphies and without whole leaf specimens, no confirmatory evidence for these proposals can be provided. The early Eocene Argentine fossil Roupala patagonica Durango de Cabrera & E.J.Romero is represented by imparipinnately compound foliage with toothed leaflets (Gonzalez et al., 2007). However, as acknowledged by Gonzalez et al. (2007), the Australian genus Neorites can have similar leaves to Roupala, and moreover, the fossils were described without the support of cuticular details.
b. Tribe Banksieae (3 genera; ~170 spp.). Banksia and relatives are widespread in the Australian fossil record (e.g. Cookson & Duigan, 1950; Hill & Christophel, 1988; Jordan & Hill, 1991; Carpenter et al., 1994; Jordan, 1995; McLoughlin & Hill, 1996; Carpenter et al., 2011), and can be the most readily recognised of all Proteaceae leaf fossils. In the first study of Proteaceae fossil leaves that incorporated cuticular evidence, Cookson and Duigan (1950) erected the new genus Banksieaephyllum to house fossils with affinity to Banksia and Dryandra, which at the time were the only representatives of tribe Banksieae: hence the genus name. Later, subtribe Musgraveinae (comprising only two small genera confined to rainforests of northern Queensland) was added to Banksieae by Johnson and Briggs (1975), and many other leaf fossils have also been referred to the tribe.
On current understanding, the pre-Quaternary Banksia-like fossils comprise a very interesting group of species that includes at least stem taxa of both Banksia and Musgraveinae and also a group of species that is only doubtfully proteaceous. Recently, Carpenter et al. (2010a) established the presence of Banksia leaves in the Oligo-Miocene of New Zealand, and summarised evidence for five foliar synapomorphies for the tribe Banksieae. These authors also proposed that within Banksieae, Banksia could be distinguished from other taxa by the presence of cuticular papillae. Many Banksia leaves also differ from other Banksieae by being quite small and having lobes cut to the primary vein. The most important example of such leaves is Banksieaephyllum taylorii, from the late Paleocene of eastern Australia (Carpenter et al., 1994). These leaf fossils were used in the dating study of Crisp et al. (2004) to constrain the base of the stem age of Banksia. Other leaf fossils that are architecturally similar to some deeply-lobed Banksia species of Western Australia (Hill & Christophel, 1988) probably belong to stem group Musgraveinae (Carpenter et al., 1994; 2010a), and perhaps the most notable example is Banksieaephyllum cuneatum R.S.Hill & Christophel. This leaf species has been recovered from two separate middle Eocene rainforest assemblages that also have fossil inflorescences of Musgraveinae (Christophel, 1984; Christophel & Greenwood, 1987). It also shows features that do not occur in extant Banksia, including elongate leaf bases and cuticular surface striations (Hill & Christophel, 1988).
Another late Paleocene species, Banksieaephyllum praefastigatum Vadala & Drinnan, was used together with the earliest diporate Banksieae pollen grains by Barker et al. (2007) to constrain the age of the stem lineage of Banksieae, and the leaf was justified on the basis of architectural features. However, these leaves may not belong to Proteaceae since the stomata and trichome bases illustrated by Vadala and Drinnan (1998) are very unconvincing for the family, let alone tribe Banksieae. For the same reasons, B. obovatum Cookson & Duigan, B. pinnatum Cookson & Duigan, B. fastigatum H.Deane (Cookson & Duigan), B. attenuatum R.S.Hill & Christophel and B. regularis R.S.Hill & Christophel must also be considered only doubtfully proteaceous (see Carpenter & Jordan, 1997).
c. Tribe Embothrieae (12 genera; >560 spp.). Paleogene leaf fossils attributed to Embothrieae occur in Australia, New Zealand and southern South America. The most convincing of these leaves have distinctive architectural and cuticular features that are shared uniquely within the range observed in an extant species. Indeed, at present, this type of evidence is the most compelling for identifying fossils to Embothrieae, since unique leaf and cuticular synapomorphies for extant genera appear to be lacking. One possible synapomorphy for the tribe is an absence of obvious leaf surface striations, since across this tribe striations have only been observed in some species of Stenocarpus (Carpenter, 1994).
Carpenter and Hill (1988) described Lomatia xeromorpha from Cethana in Tasmania, a site of probable early Oligocene age (Macphail et al., 1994). This species has extremely reduced, bipinnatisect foliage that conforms only to that of the extant Tasmanian species Lomatia tinctoria R.Br. Carpenter and Jordan (1997) also placed fossils, including incomplete but pinnatisect leaves with well-preserved cuticle, from the same site in the extant species L. fraxinifolia F.Muell. ex Benth. Apparently identical cuticular remains have been recovered from other sites, including late middle Eocene core material from Western Australia (Carpenter & Pole, 1995). Overall, these fossils provide evidence of the evolution of crown group taxa by the early Oligocene, but strongly inferred synapomorphies are lacking. Weston and Crisp (1994) excluded leaf shape (i.e. simple vs dissected) as a character from their cladistic analysis of Lomatia because of doubt concerning homologies. However, all the extant species that show variously dissected to compound leaves are confined to their subclade of Lomatia that comprises all Australian species plus L. ferruginea (Cav.) R.Br. from southern South America. Many other Australian leaf fossils are consistent with toothed, dissected-leaved Lomatia, including several species of the fossil genus Euproteaciphyllum from early Eocene and Oligocene sites in Tasmania that Carpenter and Jordan (1997) and Jordan et al. (1998) reasoned could not be placed in the extant genus. This was because there were apparently no apomorphic leaf features uniting extant Lomatia, and similar toothed, dissected leaf forms were noted in Beauprea and Neorites. Similarly, the toothed, apparently pinnatisect leaf fossil from the Paleocene of New Zealand that Pole (1997) described as Lomatia novae-zelandiae has architectural and cuticular features that are consistent with Lomatia, but without demonstrated apomorphies. Also, none of the three species of Lomatia recognised from early Eocene, middle Eocene and late Oligocene-early Miocene sediments in Argentina (reviewed by Gonzalez et al., 2007) were described with the support of cuticular evidence, and leaves with similar bipinnatisect/imparipinnate architecture and venation occur elsewhere in Proteaceae. Lomatia fossils attributed to the extant Tasmanian species L. tasmanica W.M.Curtis occur in Upper Pleistocene sediments from near the limited current range of the species in western Tasmania (Jordan et al., 1991).
Subtribe Embothriinae comprising Embothrium, Telopea, Alloxylon and Oreocallis is recognised with some confidence in the fossil record because its pollen has the synapomorphy of spinulate sculpturing (Feuer, 1990; Weston & Crisp, 1994; Barker et al., 2007). Further, biporate pollen with such sculpturing is only found in the southern South American species Embothrium coccineum Forst. (Johnson & Briggs, 1975; Feuer, 1990) and the fossil pollen species Granodiporites nebulosus A.D.Partr. (Macphail et al., 1994), which is first recorded in Australian sediments of latest Eocene age (Stover & Partridge, 1973).
Leaf fossils of Telopea were recognised from the early Oligocene of Tasmania by Carpenter and Jordan (1997), and assigned to extant T. truncata (Labill.) R.Br. on the basis of leaf architectural and venation features, as well as cuticular details. This is another case where fossils were assigned to an extant taxon because the available morphological details were well preserved, relatively extensive, and considered to be uniquely within the range of that extant taxon. Leaf evolution and characters are currently difficult to interpret in Telopea, in part because molecular studies have so far not resolved phylogenetic relationships in the genus (P.H. Weston, pers. comm.). However, morphological evidence places T. truncata as sister to the two closely species T. oreades F.Muell. + T. mongaensis Cheel (Weston & Crisp, 1994). In terms of leaf characters, these three species differ from the other two in having entire leaf margins, not toothed.
Jordan (1995) and Carpenter and Jordan (1997) also found that T. truncata differs from the other species of Telopea in that its stomata are situated on a plane well below the level of prominent, compound cuticular papillae. Embothrium coccineum has similar structures that were reasoned by Jordan et al. (2008) to have most parsimoniously evolved separately from those in T. truncata. An apparent difference between the leaves of these two species is that the ledges associated with the stomata in E. coccineum are not well developed (and do not stain darkly with Safranin O) compared with those in all Telopea species and in the Cethana fossils (Carpenter & Jordan, 1997). Telopea truncata leaves are also lanceolate to obovate, whereas Embothrium leaves tend to be more broadly ovate (Jordan, 1995; Gonzalez et al., 2007). Overall, the Tasmanian fossils are currently considered as evidence for the evolution of crown group Telopea belonging to the T. truncata lineage by the early Oligocene. Telopea leaf fossils including T. truncata also occur in Tasmanian Pleistocene sediments (Jordan, 1995). Dispersed cuticles from the early Eocene of Tasmania (Pole, 2007b) and the late middle Eocene of Western Australia (Carpenter & Pole, 1995) also exhibit likenesses to Telopea, although these cuticles are not as conclusive as the Oligocene and Pleistocene leaf specimens. Gonzalez et al. (2007) regarded the records of Embothrium leaves from the middle Eocene of Argentina to be doubtful. A number of other leaf fossils from Australia and New Zealand may well belong to tribe Embothrieae, and are especially reminiscent of Alloxylon and Stenocarpus (Blackburn, 1985; Hill & Merrifield, 1993; Pole, 1998; R. J. Carpenter, unpubl.).
The large, lobed leaf described as Maslinia grevilleoides Blackburn from the middle Eocene of South Australia was regarded as a member of the clade comprising Grevillea, Hakea, and Finschia by Blackburn (1981). However, his illustrations and descriptions of the fossil do not establish any close relationship with Grevillea. Indeed, apart from leaf fragments of Hakea from Pleistocene sediments in Tasmania (Jordan et al., 1991; Jordan, 1995), there are so far no known fossil leaves that could be assigned to this extremely speciose clade (>500 spp.) of subtribe Hakeinae with any confidence, although there are some leaves (lacking cuticle) that appear similar to some grevilleas in leaf form and venation (McLoughlin & Hill, 1996; Pole & Bowman, 1996; Carpenter & Jordan, 1997; Carpenter et al., 2011). Foliar synapomorphies for this group could include trichomes with two-branched terminal cells (Johnson & Briggs, 1975) which are associated with only a single basal epidermal cell (Carpenter, 1994), and rather broad, oval-shaped complexes of guard and subsidiary cells (Carpenter, 1994). All species of Hakea have terete or similifacial leaves (Barker et al., 1999) and all appear to be amphistomatic. Hakea species also appear to be united by the presence of stomata in individual crypts (Jordan et al., 2008).
d. Tribe Macadamieae (18 genera; ~90 spp.). As currently understood, the best leaf fossil evidence for this tribe comprises several leaf and dispersed cuticular specimens that have been proposed as members of subtribe Gevuininae. This subtribe comprises a trichotomy of clades with an overall distribution in Australia-New Guinea, New Caledonia, Fiji and South America, plus the sister of this group, the Australian genus Cardwellia (Mast et al., 2008). Carpenter (1994) noted that a distinctive type of trichome base occurs in representatives of the trichotomy, including in Hicksbeachia, which at the time was included in a different subtribe by Johnson and Briggs (1975). This type of trichome base may be of quite large diameter, and typically appears as a deeply staining (Safranin O), donut-like ring associated with a thickened base that has a serrated or 'crimped' margin. Inner cuticle surfaces are also obviously granular. Leaf material with cuticle of this type has been recorded from the late middle Eocene of Western Australia (Carpenter & Pole, 1995), the early Oligocene of Tasmania (Carpenter & Jordan, 1997), and the Miocene of New Zealand (Carpenter, 1994; Pole, 1998, 2008). Mast et al. (2008) considered that the trichome base type was a derived feature of the taxa in the trichotomy, it not being reported in Cardwellia (Carpenter, 1994), and used the Western Australian fossils as age constraints with respect to subtribe Gevuininae in their dated phylogeny of tribe Macadamieae.
Carpenter (1994) also reported cuticular fossils from the Miocene of New Zealand that he compared closely with Macadamia, and Pole (1998) assigned these to the genus on the basis of similarity only. It seems probable that these fossils do belong to tribe Macadamieae, but as for other taxa, much more work is required to establish leaf and cuticular characters and their states in the tribe. Ideally, apart from light microscopy, this requires scanning electron microscopy of the inner and outer surfaces of cuticle from both sides of the leaves of all extant species, and cladistic analysis. Distinctive features of the fossils that are shared in extant Macadamia are that the guard cell pairs have an overall round shape, and the anticlinal cell walls are tightly sinuous (to buttressed) (Carpenter, 1994). These features are potential synapomorphies for the clade comprising subtribes Macadamiinae, Malagasiinae and Virotiinae, since, as discovered by Carpenter (1994), they also occur widely among the species of Virotia, Catalepidia and Heliciopsis, as well as in Nothorites megacarpus and the five species of Macadamia that were transferred to Lasjia by Mast et al. (2008). Thus the New Zealand fossils are presently best viewed as representing this clade only, not any one genus within it.
Proteaceae Leaf Fossils and Diversity
Current centers of Proteaceae diversity are the well known 'Mediterranean hotspot' heathland regions in southwestern Australia and the South African Cape, as well as the small rainforested region of the Wet Tropics of northeastern Australia. Indeed, this latter region has a much higher generic diversity of Proteaceae (26 genera) than the 15 and 14 genera in the two Mediterranean climate regions (tallies based on Weston & Barker, 2006; Mast et al., 2008 and Sauquet et al., 2009a).
What does the foliar (leaf and cuticle) fossil record say about the past diversity of Proteaceae? An important point here is that leaf fossils are almost always likely to only provide direct evidence of minimum local species diversity, i.e. only that of the plants growing near a potential fossil site, and then usually only those that are trees with sufficiently robust leaves. With this in mind, virtually all of the Cenozoic fossil sites in Australia and New Zealand that have been studied to any extent contain foliar evidence of Proteaceae. So far, the most diverse site is the early Oligocene Cethana site in Tasmania. Here, there are 15 described species of Proteaceae with wellpreserved cuticle (Carpenter & Jordan, 1997; Jordan et al., 1998), as well as several more taxa without (Hill & Christophel, 1988; Carpenter, 1991; Carpenter & Jordan, 1997). High diversity is also evident in the middle Eocene Maslin Bay assemblage (Hill et al., 1995) and sediments of the Eyre Formation (latest Paleocene-middle Eocene: Alley, 1998) of central Australia, where core samples of only a few [cm.sup.3] may each contain discrete assemblages of ~10 taxa of Proteaceae cuticles (R.J. Carpenter, unpubl.). Significantly, Jordan et al. (1998) estimated that the Cethana fossils signalled a regional (northwestern Tasmanian) proteaceous diversity of at least 200 species, in part because nearby palynologically coeval sites do not share any of the same taxa as at Cethana. It has also been demonstrated that at least Banksia was more diverse than at present in Tasmania as recently as the Upper Pleistocene (Jordan & Hill, 1991). Extinctions were probably related to the extreme effects of glaciation events in the region.
Only two species of Proteaceae in two genera are native to New Zealand, but if New Caledonia is considered as part of the same region of continental crust (see Ladiges & Cantrill, 2007), then this rises to a total of 11 genera (or 12 with the addition of Turrillia, the ancestor of which arose in New Caledonia: Mast et al., 2008). There is so far no evidence that early Paleogene leaf assemblages of New Zealand were anywhere near as diverse as in Australia, but at least six species occur later in the Oligo-Miocene Newvale lignites (Carpenter et al., 2010a,b; R. J. Carpenter, unpubl.), and a total of eight or nine species (as dispersed cuticles) in sediments of the early Miocene Manuherikia Group (Pole, 1998, 2008). Overall, the history of Proteaceae in New Zealand reflects that of numerous other biotic elements in showing a marked decline in diversity during the Neogene (e.g. Lee et al., 2001).
In South America there are only eight extant genera in total, and only Euplassa, Panopsis and Roupala are relatively species-rich (Sleumer, 1954; Prance & Plana, 1998). Fossil evidence of Proteaceae leaves in South America is so far limited to specimens without cuticle. However, this evidence is consistent with the pollen record (Askin & Baldoni, 1998) which suggests much lower Cenozoic proteaceous diversity than in New Zealand and particularly Australia. This lower diversity is exemplified by the early Eocene Laguna del Hunco macrofossil assemblage of Argentina, where only four species of Proteaceae, including two of Lomatia, are recognised among >3,500 vouchered specimens and 130 angiosperm taxa (Wilf et al., 2005). Moreover, only the Lomatia species can be considered as relatively common at Laguna del Hunco (Wilf et al., 2005), the other taxa being represented by only a few specimens each (Gonzalez et al., 2007). Proteaceae pollen is also found in southern South American assemblages that are more recent than Eocene (e.g. Miocene: Barreda & Palazzesi, 2007; Hebel & Torres, 2009), but there is no fossil evidence that suggests that the family ever approached Australasian levels of diversity. Overall, this relatively low diversity is not inconsistent with the observation (Johnson & Briggs, 1975; Prance & Plana, 1998) that Euplassa, Panopsis and Roupala have all radiated only recently. Also, molecular phylogenetic evidence suggests that the ancestors of at least Euplassa and Panopsis arrived relatively recently (during the Neogene) in South America via long-distance dispersal (Mast et al., 2008).
Proteaceae Leaf Fossils and Ecology
It is well known that Proteaceae are very strongly associated with soils deficient in nutrients, especially phosphorous, and many taxa, including those of rainforests exhibit sclerophylly and xeromorphy (Beadle, 1966). These soils include those formed on siliceous sandstones, quartzites, "acid" granitic or even volcanic rocks, ultrabasic igneous rocks, fossil laterites and aeolian sands, and peats in oligotrophic waterlogged situations (Johnson & Briggs, 1975). Often, such as in southwestern Australia, the landforms themselves are ancient and highly weathered, and not surprisingly, many authors have hypothesised that the association of Proteaceae with such environments is also ancient. Indeed, it is clear that several taxa of Paleogene Proteaceae fossils do exhibit scleromorphy, as has previously been discussed with respect to Banksia, highly reduced leaf types in Lomatia and Lomatia-like species, and Orites species with sclerified hypodermes.
The most prominently sclerophyllous extant taxa are those that occupy open, often heathland habitats with seasonal water deficits. Johnson and Briggs (1975) considered that the leaves of these typically shrubby taxa show a range of derived, xeromorphic features including revolute margins, tereteness, reduced "quasi-parallel" venation, thick cuticles, sunken stomata and the presence of sclerotic hypodermes. Jordan et al. (2005) used phylogenetic evidence to argue that thick cuticles and hypodermal structures were in fact more likely to have evolved primarily for protection from light damage in open habitats. Oligo-Miocene species of Banksieaephyllum from Victoria show development of stomata sunken in areolar depressions (Cookson & Duigan, 1950), but this does not approach the state seen in many extant Banksia species of subgenus Banksia (Mast & Thiele, 2007), where the stomata are deeply encrypted in balloon-like structures. This type of deep encryption, in addition to stomatal protection afforded by grooves, tightly revolute margins or strongly overarching cuticle has repeatedly evolved in Proteaceae in response to drying climates (Jordan et al., 2008). So far, no pre-Quaternary Proteaceae fossils show such levels of protection, although several, notably including some of the Banksieaephyllum species, can be interpreted as showing xeromorphic trends in that direction (Hill, 1998). Overall, this is broadly consistent with the development of drying climates in southern Australia associated with the final separation of Australia from Antarctica in the mid-Cenozoic.
Some further comments can be added to these discussions, in part based on new fossil material from New Zealand. Firstly, Carpenter et al. (2010a,b) documented the presence of Persoonieae and Banksia as part of an Oligo-Miocene oligotrophic mire community at Newvale (see also Ferguson et al., 2010). The Persoonieae leaves show more or less parallel venation, and leaves of two further Proteaceae species are amphistomatic (Fig. 3). Johnson and Briggs (1975) included amphistomaty as a derived state frequently found in sclerophyllous Proteaceae, and more generally, amphistomaty is regarded as a convergent feature of plants growing in full-sun (open) environments without water stress (Mott et al., 1982). Moreover, proteoid roots (Purnell, 1960), which are specialised structures that markedly increase the surface area of the root system and serve to maximize P acquisition from soils of low fertility (Lambers et al., 2006), are abundant in the lignites (Fig. 4). This evidence, in combination with the presence of taxa such as very small-leaved epacrid (Ericaceae) species (Jordan et al., 2010), strongly implies the existence of a richly diverse, relatively open, sclerophyllous community during the Oligo-Miocene of New Zealand, and provides evidence for such a community in a region that now does not have any Proteaceae-rich habitats. Swampy habitats with proteoid roots and many species of Proteaceae including Banksia were also present in southeastern Australia at the same time (Blackburn, 1985; Blackburn & Sluiter, 1994). A much older occurrence of amphistomaty, and thus inference of open habitats, is possible given the Upper Cretaceous age of crown Proteoideae (Sauquet et al., 2009a,b) and widespread amphistomaty among its extant clades.
Secondly, it is emphasised that a number of Australian Paleogene taxa exhibited thick cuticles with slightly sunken stomata and pronounced, elaborate surface features such as striations, rugulations and numerous trichome bases (Figs. 5, 6) (Carpenter et al., 2004, 2006). This combination of features is not or rarely encountered among extant species, where such features are actually statistically weakly associated with wet climates, not dry (Jordan et al., 2008). The extreme forms of cuticular ornamentation may have resulted from plant investment in sclefification and protection from, for instance, epiphytic fungi in wet environments (e.g. Hill, 1998), and been enabled under conditions of extremely high C[O.sub.2] concentrations. Jordan et al. (1998) and Carpenter et al. (2004) have also observed that many fossil Proteaceae had much smaller stomata than those across the extant members of the family, a factor that may also be related to greater C[O.sub.2] availability in the past. Notably, this is apparent in taxa that can be placed in extant genera, especially Banksia (Jordan et al., 1998). Also, most stomata of the Cethana Telopea truncata-type fossils are smaller than in extant T. truncata (R.J. Carpenter, unpubl.).
Jordan et al. (1998) referred to the possibility that edaphic differences between sites could in part explain differences in the abundance and diversity of Proteaceae fossils in Tasmania. Pole (1996) found no Proteaceae leaves in an otherwise well-preserved and diverse assemblage of leaves from the early Miocene Foulden Hills volcanic maar site in New Zealand, and implied that this could be attributed to the family being generally less competitive on the rich basalt soils that were likely to have been present locally. In a wider Southern Hemisphere context it is probable that the preponderance of sclerophyllous and to some degree xeromorphic Proteaceae leaf fossils in only Australia attests to the extraordinary age of landscapes in that continent, and general lack of edaphic rejuvenation.
In Australia, with the exception of Banksia and perhaps Lomatia, lineages that now occur in Mediterranean-climate heathlands, and that are sometimes extremely diverse and abundant there, have essentially non-existent fossil records. This is despite the fact that the overwhelming majority of these extant taxa are from subfamily Proteoideae, Persoonioideae tribe Persoonieae and Grevilleoideae tribe Embothrieae subtribe Hakeinae, and molecular evidence suggests that the Australian genera of these taxa belong to lineages that diverged from their sister groups in the Paleogene or earlier (Sauquet et al., 2009a). However, this evidence also suggests that at least Persoonia and Grevillea+ Hakea have undergone rapid, recent diversification in southwestern Australia, consistent with the relatively recent development of drier climates in the Neogene. Thus, although taphonomic reasons could largely explain absences of Mediterraneanclimate lineages from the fossil record (Carpenter et al., 1994; Jordan et al., 1998), it is plausible that these absences might reflect that such lineages were not abundant (and/or perhaps not diverse) in the ancient past. In the Cape Floristic Region of Africa, the only Proteaceae represented belong to subfamily Proteoideae. However, although there is molecular evidence that some diversifications in Proteoideae preceded the late Neogene development of the Mediterranean climate-type in this region, the known fossil record of Proteaceae is not helpful in interpreting this history (Sauquet et al., 2009b).
Proteaceae Leaf Fossils and Austral Distributions
A clear signal from the fossil record is that taxa of Proteaceae were formerly present in regions now very remote from their current distributions. This is particularly well established with respect to the marked contraction of Australia-wide wet forest vegetation types to disjunct sites along the eastern coast of Australia (e.g. Carpenter, 1994; Greenwood & Christophel, 2005). The fossil record (of macrofossils and pollen) confirms that this contraction saw the extinction of numerous taxa of Proteaceae, and the high number of present-day rainforest genera with one or very few species is probably also testament to this (Johnson & Briggs, 1975).
New Zealand holds a special place in interpreting Southern Hemisphere biogeography, including the relative roles of vicariance and dispersal in the development of biotas (Pole, 2001). The presence of Banksia in the Oligo-Miocene of New Zealand demonstrates that this famous lineage of Australian plants once had a distribution in a region of the Southern Hemisphere that has been isolated from Australia for probably at least 55 million years (Carpenter et al., 2010a). Vicariance could explain this distribution given that the Paleocene age of Australian Banksia fossils (Carpenter et al., 1994) overlaps with the existence of land connections between Australia and New Zealand via the Lord Howe Rise (see Ladiges & Cantrill, 2007). However, especially if the New Zealand region was completely inundated by seawater in the latest Oligocene to earliest Miocene (Waters & Craw, 2006; Landis et al., 2008) and the New Zealand Banksia fossils are younger than this event, interesting questions arise about the dispersal ability of a group not normally considered to disperse over long distances, and certainly not over seawater gaps.
Recent studies using dated molecular phylogenies have challenged the long-held hypothesis that the distribution of Proteaceae is fundamentally the result of vicariance and the breakup of the southern continents (Barker et al., 2007; Mast et al., 2008; Sauquet et al., 2009a, b). Interestingly, there are no well-justified leaf fossil records that strongly contradict any of the proposed times of lineage divergences in these studies, and several that are consistent with them. Thus, Orites has extant species in Australia and southern South America, proposed crown group Tasmanian leaf fossils of early Oligocene age, and an estimated stem group age based on a UCLN (BEAST) analysis of molecular data of 34.9 million years [lower (Lb) and upper (Ub) bounds of 95 % confidence interval, 19.3 and 52 million years respectively; Sauquet et al., 2009a]. There are also described Orites follicles from the early Eocene of Patagonia (Gonzalez et al., 2007). These lines of evidence are all consistent with the possibility of a trans-Antarctic spread of Orites, perhaps largely across ice-free terrestrial land during the Paleogene. The first major glaciations of Antarctica occurred around the Eocene-Oligocene boundary (maximum at 33.6 Ma; Coxall et al., 2005), and although there is evidence for the presence of seawater gaps between the South Tasman Rise (south of Tasmania) and Antarctica and South America and Antarctica prior to this (Lawver & Gahagan, 2003), recent evidence indicates that the Antarctic Circumpolar Current did not develop until the late Oligocene (Lyle et al., 2007). Lomatia is a parallel case to that of Orites, but with an even older estimated stem age [49.7 million years (Lb=35.9, Ub=63.2); Sauquet et al., 2009a], and also the evidence of Lomatia or Lomatia-like leaves from the Paleocene of New Zealand and the early Eocene of both Tasmania and Argentina. Embothriinae were estimated by Sauquet et al. (2009a) to have mean stem and crown ages of 66.6 (Lb= 56.5, Ub= 75.6) and 41.1 (Lb=35.4, Ub=48.1) million years respectively, and so it is not unreasonable that crown group Telopea closely related to T. truncata should have existed in the early Oligocene of Tasmania.
A further point is that because the likelihood of fossilisation of any plant organs is extremely low, the relatively high abundance and/or wide spatial and/or temporal distribution of at least Banksia, Orites and Lomatia fossils itself probably reflects that these plants were important components of the vegetation, and had traits that enabled their lineages to become widely established.
Finally, there is increasing fossil evidence to support the notion of a high latitude Gondwanic element that was widespread across Antarctica between Australia and South America in the Paleogene, and perhaps able to disperse as a more or less discrete community across land. The component plants of this group included the fossil examples of Araucariaceae, Podocarpaceae, Cupressaceae, Akania (Akaniaceae), Casuarinaceae, Nothofagus (Nothofagaceae) and Eucalyptus (Myrtaceae) discussed by Hill and Carpenter (1991). To this list of fossils can be added the Proteaceae discussed here and Lygodium (Schizaeaceae), fossils of which are found in Chile (Halle, 1940), Australia and New Zealand (Rozefelds et al., 1992). Also, wellpreserved fossils of Papuacedrus (Cupressaceae) (Wilf et al., 2009), Gymnostoma (Casuarinaceae) (Zamaloa et al., 2006) and Eucalyptus (Gandolfo et al., 2011) have recently been described from the early Eocene Laguna del Hunco site in Argentina. These genera are now endemic to Australasia. Conversely, fossils of the Patagonian endemic Fitzroya (Cupressaceae) have been reported from Tasmania (Hill & Whang, 1996). This Paleogene austral vegetation apparently provided habitat for a range of animals, including Chulpasia, a genus of marsupial that appears to have been moreor-less contemporaneously distributed in Australia, Antarctica and South America in the late Paleocene to earliest Eocene (Sige et al., 2009).
This review should serve to highlight the enormous potential of the fossil record of Proteaceae for understanding the phylogenetic, ecological and biogeographical history of the family. On the basis of inferred foliar synapomorphies there is very good evidence for at least the presence of stem taxa of Persoonieae in the Oligo-Miocene, many stem taxa of Banksieae including Banksia dating from the late Paleocene, and for crown group Orites species by the early Oligocene. Probable crown group Lomatia and Telopea species were also present in the Oligocene, but this is only based on compelling foliar features in common with extant species, and it may not be possible to support this with the evidence of convincing non-homoplasious synapomorphies. There is also foliar evidence for Megahertzia and subtribe Gevuininae and other Maeadamieae, but all these fossils require more comparative work with extant taxa. Several Proteaceae leaf fossils are suitable for inclusion in molecular dating studies, but it is recommended that these studies should acknowledge the inherent limitations and difficulties of paleobotanical research (Gandolfo et al., 2008). Only very rarely have detailed leaf (and cuticular) anatomical and morphological studies been undertaken on extant groups that can enable adequate cladistic assessment of trait evolution and fossil placement. Usually such work has only been undertaken by individual paleobotanists themselves. The ontogenetic sequences of Proteaceae leaves among taxa, and their various forms of venation, tooth types and lobing were addressed by Johnson and Briggs (1975). In addition to cuticular traits, assessment of these gross leaf features with respect to modern phylogenies might prove rewarding, as might an understanding of trichome development and shedding. Further studies related to other key Southern Hemisphere angiosperm families that often occur with Proteaceae and that also appear to be relatively well represented in the fossil record would also be of interest. These families include Sapindaceae, Lauraceae, Cunoniaceae, Elaeocarpaceae and Myrtaceae.
Acknowledgments I thank M. C. Zamaloa, M. A. Gandolfo and N. R. Cumeo for the invitation to speak at the VII Southern Connections Congress, Bariloche, Argentina, and to submit this paper. I am very grateful to R. S. Hill, G. J. Jordan and D. E. Lee for encouraging discussions and funding assistance over many years, in part through the Australian Research Council, an Otago Research Grant from the University of Otago, and the Marsden Fund (New Zealand). Funds from Australian Research Council Discovery Project 110104926 assisted in completing this paper.
Published online: 27 July 2012
Ali, J. R. & J. C. Aitchison. 2009. Kerguelen Plateau and the Late Cretaceous southern-continent bioconnection hypothesis: tales from a topographical ocean. Journal of Biogeography 36:1778 1784.
Alley, N. F. 1998. Cainozoic stratigraphy, palaeoenvironment and geological evolution of the Lake Eyre Basin. Palaeogeography, Palaeoclimatology, Palaeoecology 144: 239-263.
APG, III (The Angiosperm Phylogeny Group). 2009. An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG I11. Botanical Journal of the Linnean Society 161: 105-121.
Ask, in, R. A. & A. M. Baldoni. 1998. The Santonian through Paleogene record of Proteaceae in the southern South America Antarctic Peninsula region. Australian Systematic Botany 11 : 373-390.
Barker, N. P., P. H. Weston, J. P. Rourke & G. Reeves. 2002. The relationships of the southern African Proteaceae as elucidated by internal transcribed spacer (ITS) DNA sequence data. Kew Bulletin 57: 867-883.
--, F. Rutsehmann & H. Sauquet. 2007. Molecular dating of the 'Gondwanan' plant family
Proteaceae is only partially congruent with the timing of the break-up of Gondwana. Journal of Biogeography 34: 2012-027.
Barker, R. M., L. Haegi & W. R. Barker. 1999. Hakea. Pp 31-170. in: A. J. G. Wilson (ed). Flora of Australia, Vol. 17B. CSIRO-ABRS, Canberra.
Barreda, V. & L. Palazzesi. 2007. Patagonian vegetation turnovers during the Paleogene-early Neogene: origin of arid-adapted floras. Botanical Review 73:31-50.
Beadle, N. C. W. 1966. Soil phosphate and its role in molding segments of the Australian flora and vegetation, with special reference to xeromorphy and sclerophylly. Ecology 47: 992-1007.
Blackburn, D. T. 1981. Tertiary megafossil flora of Maslin Bay, South Australia: numerical taxonomic study of selected leaves. Alcheringa 5: 9-8.
-- 1985. Palaeobotany of the Yallourn and Morwell coal seams. Palaeobotanical Project Report 3. State Electricity Commission of Victoria, Melbourne.
-- & I. R. K. Sluiter. 1994. The Olig-Miocene coal floras of southeastern Australia. Pp 328-367.
In. R. S. Hill (ed). Australian vegetation history: Cretaceous to Recent. Cambridge University Press, Cambridge.
Carpenter, R. J. 1991. Palaeovegetation and environment at Cethana, Tasmania. PhD thesis, University of Tasmania, Hobart.
--1994. Cuticular morphology and aspects of the ecology and fossil history of North Queensland rainforest Proteaceae. Botanical Journal of the Linnean Society 116: 249-303.
--& R. S. Hill. 1988. Early Tertiary Lomatia (Proteaceae) macrofossils from Tasmania, Australia. Review of Palaeobotany and Palynology 56: 141-150.
--& M. Pole. 1995. Eocene plant fossils from the Lefroy and Cowan Paleodrainages, Western Australia. Australian Systematic Botany 8:11021154.
--& G. J. Jordan. 1997. Early Tertiary macrofossils of Proteaceae from Tasmania. Australian Systematic Botany 10: 533-563.
--, -- & R. S. Hill. 1994. Banksieaephyllum taylorii (Proteaceae) from the Late Paleocene of New South Wales and its relevance to the origin of Australia's scleromorphic flora. Australian Systematic Botany 7:385-392.
--, R. S. Hill & G. J. Jordan. 2005. Leaf cuticular morphology links Platanaceae and Proteaceae.
International Journal of Plant Sciences 166: 843-855.
--, -- & L. J. Scriven. 2006. Palmately lobed Proteaceae leaf fossils from the Middle Eocene of South Australia. International Journal of Plant Sciences 167: 1049-1060.
--, --, D. R. Greenwood, A. D. Partridge & M. A. Banks. 2004. No snow in the mountains: Early Eocene plant fossils from Hotham Heights, Victoria, Australia. Australian Journal of Botany 52: 685 718.
--, J. M. Bannister, G. J. Jordan & D. E. Lee. 2010a. Leaf fossils of Proteaceae tribe Persoonieae from the Late Oligocene-Early Miocene of New Zealand. Australian Systematic Botany 23: 1-15.
--, M. P. Goodwin, R. S. Hill & K. Kanold. 2011. Silcrete plant fossils from Lightning Ridge, New South Wales: new evidence for climate change and monsoon elements in the Australian Cenozoic. Australian Journal of Botany 59:399-425
--, G. J. Jordan, D. E. Lee & R. S. Hill. 2010b. Leaf fossils of Banksia (Proteaceae) from New Zealand: an Australian abroad. American Journal of Botany 97: 288-597.
Christophel, D. C. 1984. Early tertiary Proteaceae: the first floral evidence for the Musgraveinae. Australian Journal of Botany 32:177-186.
-- 1994. The early Tertiary macrofloras of continental Australia. Pp 262-275. In: R. S. Hill (ed). Australian vegetation history: Cretaceous to Recent. Cambridge University Press, Cambridge.
-- & D. R. Greenwood. 1987. A megafossil flora from the Eocene of Golden Grove, South Australia. Transactions of the Royal Society of South Australia 111 : 155 162.
--, W. K. Harris & A. K. Syber. 1987. The Eocene flora of the Anglesea locality, Victoria. Alcheringa 11:303-324.
Cookson, L C. & S. L. Duigan. 1950. Fossil Banksieae from Yallourn, Victoria, with notes on the morphology and anatomy of living species. Australian Journal of Scientific Research, Series B, Biological Sciences 3: 133-165.
Coxall, H. K., P. A. Wilson, H. Pailike, C. Lear & J. Baekman. 2005. Rapid stepwise onset of Antarctic glaciation and deeper calcite compensation in the Pacific Ocean Nature 433:53 57.
Crane, P. R., K. R. Pedersen, E. M. Friis & A. N. Drinnan. 1993. Early Cretaceous (early to middle Albian) platanoid inflorescences associated with Sapindopsis leaves from the Potomac group of eastern North America. Systematic Botany 18: 328-344.
Crisp, M., L. Cook & D. Steane. 2004. Radiation of the Australian flora: what can comparisons of molecular phylogenies across multiple taxa tell us about the evolution of diversity in present-day communities? Philosophical Transactions of the Royal Society of London, Series B 359:1551-1571.
Cronquist, A. 1968. The evolution and classification of flowering plants. Houghton Mifflin, Boston. Dettmann, M. E. & D. M. Jarzen. 1998. The early history of the Proteaceae in Australia: the pollen record. Australian Systematic Botany 11: 401-438.
-- & H. T. Clifford. 2005. Fossil fruit of the Grevilleeae (Proteaceae) in the Tertiary of eastern Australia. Memoirs of the Queensland Museum 51: 359 374. --&--.2010. Fossil fruit of the Macadamieae (Proteaceae) in the Tertiary of eastern Australia: Eureka gen. nov. Memoirs of the Queensland Museum 55: 147-166.
Doyle, J. A. 2007. Systematic value and evolution of leaf architecture across the angiosperms in light of molecular phylogenetic analyses. Courier Forschungsinstitut Senckenberg 258:21-37.
Drinnan, A. N., P. R. Crane & S. B. Hoot. 1994. Patterns of floral evolution in the early diversification of non-magnoliid dicotyledons (eudicots). Plant Systematics and Evolution Suppl. 8: 93-122.
Ferguson, D. K., D. E. Lee, J. M. Bannister, R. Zetter, G. J. Jordan, N. Vavra & D. C. Mildenhall. 2010. The taphonomy of a remarkable leaf bed assemblage from the Late Oligocene-early Miocene Gore Lignite Measures, southern New Zealand. International Journal of Coal Geology 83:123-181.
Feuer, S. 1990. Pollen morphology of the Embothrieae (Proteaceae) II. Embothriinae (Embothrium, Oreocallis, Telopea). Grana 29: 19-36.
Gandolfo, M., E. J. Hermsen, M. C. Zamaloa, K. C. Nixon, C. C. Gonzalez, P. Wilf, N. R. Cuneo & K. R. Johnson. 2011. Oldest known Eucalyptus macrofossils are from South America. PLoS ONE 6: e21084.
Gandolfu, M. A., K. C. Nixon & W. L. Crepet. 2008. Selection of fossils for calibration of molecular dating models. Annals of the Missouri Botanical Garden 95: 34-42.
George, A. S. & B. P. M. Hyland. 1995. Megahertzia. Pp 354-355. In: P. McCarthy (ed). Flora of Australia, Vol. 16. CSIRO-ABRS, Melbourne.
Gonzalez, C. C., M. A. Gandolfo, N. R. Cuneo, P. Will & K. R. Johnson. 2007. Revision of the Proteaceae macrofossil record from Patagonia, Argentina. Botanical Review 73: 235-66.
Greenwood, D. R. & D. C. Christophel. 2005. The origins and Tertiary history of Australian 'Tertiary' rainforests. Pp 336-373. In: E. Bermingham, C. Dick, & C. Moritz (eds). Tropical rainforests: past, present and future. Chicago University Press, Chicago.
Halle, T. G. 1940. A fossil fertile Lygodium from the Tertiary of South Chile. Svensk Botanisk Tidskrifl 34: 257 264.
Harris, W. K. 1966. Proteacidites latrobensis. Taxon 15:332-333.
Hebel, I. & T. Torres. 2009. Fossil pollen on King George Island (Antarctica) and continental Chile with affinity to current Chilean Proteaceae. Anales Instituto Patagonia (Chile) 37: 7-22.
Hickey, L. J. & J. A. Wolfe. 1975. The bases of angiosperm phylogeny: vegetative morphology. Annals of the Missouri Botanical Garden 62:538-589.
Hill, R. S. 1982. The Eocene megafossil flora of Nerriga, New South Wales, Australia. Palaeontograpbica AbtB 181: 14-77.
-- 1988. Australian Tertiary angiosperm and gymnosperm leaf remains - an updated catalogue. Alcheringa 12: 207-219.
-- 1998. Fossil evidence for the onset of xeromorphy and scleromorphy in Australian Proteaceae. Australian Systematic Botany 11:391-400.
-- & R. J. Carpenter. 1991. Extensive past distributions for major Gondwanic floral elements: macrofossil evidence. Pp 239-247. In: M. R. Banks, S. J. Smith, A. E. Orchard, & G. Kantvilas (eds). Aspects of Tasmanian botany. Royal Society of Tasmania, Hobart.
-- & D. C. Christophel. 1988. Tertiary leaves of the tribe Banksieae (Proteaceae) from south-eastem Australia. Botanical Journal of the Linnean Society 97: 205-227.
-- & H. E. Merrifield. 1993. An early Tertiary macroflora from West Dale, southwestern Australia. Alcheringa 17:285-326.
--& S. S. Whang. 1996. A new species of Fitzroya (Cupressaceae) from Oligocene sediments in north-western Tasmania. Australian Systematic Botany 9: 867-875.
--, L. J. Scriven & G. J. Jordan. 1995. The fossil record of Australian Proteaceae. Pp 621-630. ln: P. McCarthy (ed). Flora of Australia, Vol. 16. CSIRO-ABRS, Melbourne.
Hoot, S. B. & A. R. Douglas. 1998. Phylogeny of the Proteaceae based on atpB and atpB-rbcL intergenic spacer region sequences. Australian Systematic Botany 11 : 301 320.
Hyland, B. P. M. 1995. Hollandaea. Pp 391 393. In: P. McCarthy (ed). Flora of Australia, Vol. 16. CSIRO-ABRS, Melbourne.
Johnson, L. A. S. 1998. Proteaceae where are we? Australian Systematic Botany 11:251-255.
--& B. G. Brlggs. 1963. Evolution in the Proteaceae. Australian Journal of Botany 11: 21-51. --&--. 1975. On the Proteaceae--The evolution and classification of a southern family. Botanical Journal of the Linnean Society 70:83-182.
--&--.1981. Three old southern families Myrtaceae, Proteaceae and Restionaceae. Pp 429469. In: A. Keast (ed). Ecological biogeography of Australia. Junk, The Hague.
Jordan, G. J. 1995. Early-middle Pleistocene leaves of extinct and extant Proteaceae from western Tasmania, Australia. Botanical Journal of the Lirmean Society 118: 19-35.
--& R. S. Hill. 1991. Two new Banksia species from Pleistocene sediments in western Tasmania. Australian Systematic Botany 4:499-511.
--, R. J. Carpenter & R. S. Hill. 1991. Late Pleistocene vegetation and climate near Melaleuca Inlet, south-western Tasmania. Australian Journal of Botany 39:315-333.
--&---. 1998. The macrofossil record of Proteaceae: a review with new species. Australian Systematic Botany 11:465-501.
--, R. A. Dillon & P. H. Weston. 2005. Solar radiation as a factor in the evolution of seleromorphic leaf anatomy in Proteaceae. American Journal of Botany 92: 789-796.
--, P. H. Weston, R. J. Carpenter, R. A. Dillon & T. J. Brodribb. 2008. The evolutionary relations of sunken, covered, and encrypted stomata to dry habitats in Proteaceae. American Journal of Botany 95:521 530.
--, J. M. Bannister, D. C. Mildenhall, R. Zetter & D. E. Lee. 2010. Fossil Ericaceae from New Zealand: deconstructing the use of fossil evidence in historical biogeography. American Journal of Botany 97:59-70.
Ladiges, P. Y. & D. Cantrill. 2007. New Caledonia Australian connections: biogeographic patterns and geology. Australian Systematic Botany 20: 383-389.
Lambers, H., M. W. Shane, M. D. Cramer, S. J. Pearse & E. J. Veneldaas. 2006. Root structure and functioning for efficient acquisition of phosphorus: matching morphological and physiological traits. Annals of Botany 98: 693-713.
Landis, C. A., H. J. Campbell, J. G. Begg, D. C. Mildenhall, A. M. Paterson & S. A. Trewick. 2008. The Waipounamu Erosion Surface: questioning the antiquity of the New Zealand land surface and terrestrial fauna and flora. Geological Magazine 145: 173-197.
Lange, R. T. 1978. Some Eocene leaf fragments comparable to Proteaceae. Journal of the Royal Society of Western Australia 60:107-114.
Lawyer, L. A. & L. M. Gahagan. 2003. Evolution of Cenozoic seaways in the circum-Antarctic region. Palaeogeography, Palaeoclimatology, Palaeoecology 198:11-37.
Lee, D. E., W. G. Lee & N. Mortimer. 2001. Where and why have all the flowers gone? Depletion and turnover in the New Zealand Cenozoic angiosperm flora in relation to palaeogeography and climate. Australian Journal of Botany 49: 341-356.
Lyle, M., S. Gibbs, T. C. Moore & D. K. Rea. 2007. Late Oligocene initiation of the Antarctic Circumpolar Current: evidence from the South Pacific Geology 35: 691-594.
Macphail, M. K. 1999. Palynostratigraphy of the Murray Basin, inland southeastern Australia. Palynology 23: 199-242.
--, N. F. Alley, E. M. Truswell & I. R. K. Sluiter. 1994. Early Tertiary vegetation: evidence from spores and pollen. Pp 189-261. In: R. S. Hill (ed). Australian vegetation history: Cretaceous to Recent. Cambridge University Press, Cambridge.
Magallon, S. A. 2004. Dating lineages: molecular and paleontological approaches to the temporal framework ofclades. International Journal of Plant Sciences 165(4 Suppl.): $7-$21.
Martin, A. R. H. 1995. Palaeogene proteaceous pollen and phylogeny. Alcheringa 19: 27-40.
Mast, A. R. & K. Thiele. 2007. The transfer of Dryandra R.Br. to Banksia L.f. (Proteaceae). Australian Systematic Botany 20: 63-71.
--, C. L. Willis, E. H. Jones, K. M. Downs & P. H. Weston. 2008. A smaller Macadamia from a more vagile tribe: inference of phylogenetic relationships, divergence times, and diaspore evolution in Macadarnia and relatives (tribe Macadamieae; Proteaceae). American Journal of Botany 95: 843-870.
McLoughlin, S. & R. S. Hill. 1996. The succession of Western Australian Phanerozoic terrestrial floras. Pp 6180. In: S. D. Hopper, J. A. Chappill, M. S. Harvey, & A. S. George (eds). Gondwanan heritage: past, present and future of the Western Australian biota. Surrey Beatty & Sons, Chipping Norton, Australia.
--, A. N. Drinnan & A. C. Rozefelds. 1995. A Cenomanian flora from the Winton Formation, Eromanga Basin, Queensland, Australia. Memoirs of the Queensland Museum 38: 273-13.
McNamara, K. J. & J. K. Scott. 1983. A new species of Banksia (Proteaceae) from the Eocene Merlinleigh Sandstone of the Kennedy Range, Western Australia. Alcheringa 7: 185-193.
Milne, L. A. 1998. Beaupreaidites and new Conospermeae (Proteoideae) affiliates. Australian Systematic Botany 11 : 553-603.
Mott, K. A., A. C. Gibson & J. W. O'Leary. 1982. The adaptive significance of amphistomatic leaves. Plant, Cell and Environment 5: 455-460.
Pocknall, D. T. & Y. M. Crosbie. 1988. Pollen morphology ofBeauprea (Proteaceae): modern and fossil. Review of Palaeobotany and Palynology 53: 305-327.
Pole, M. S. 2001. Can long-distance dispersal be inferred from the New Zealand plant fossil record? Australian Journal of Botany 49: 357-66.
Pole, M. 1992. Eocene vegetation from Hasties, north-eastern Tasmania. Australian Systematic Botany 5: 431-475.
-- 1996. Plant macrofossils from the Foulden Hills Diatomite (Miocene), Central Otago, New Zealand. Journal of the Royal Society of New Zealand 26:1-39.
-- 1997. Paleoeene plant macrofossils from Kakahu, south Canterbury, New Zealand. Joumal of the Royal Society of New Zealand 27: 371-400.
-- 1998. The Proteaceae record in New Zealand. Australian Systematic Botany 11:343-372.
-- 2000. Dicotyledonous leaf macrofossils from the latest Albian-earliest Cenomanian of the Eromalaga Basin, Queensland, Australia. Paleontological Research 4: 39-52.
-- 2007a. Plant-macrofossil assemblages during Pliocene uplift, South Island, New Zealand. Australian Journal of Botany 55:118-142.
-- 2007b. Early Eocene dispersed cuticles and mangrove to rainforest vegetation at Strahan-Regatta Point, Tasmania. Palaeontologia Electronica 10.3.15A.
-- 2008. Dispersed leaf cuticle from the Early Miocene of southern New Zealand. Palaeontologia Electronica 11.3.15A.
-- & D. M. J. S. Bowman. 1996. Tertiary plant fossils from Australia's 'Top End'. Australian Systematic Botany 9:113-126.
-- & V. Vajda. 2009. A new terrestrial Cretaceous-Paleogene site in New Zealand - turnover in macroflora confirmed by palynology. Cretaceous Research 30:917-938.
Prance, G. T. & V. Plana. 1998. The American Proteaceae. Australian Systematic Botany 11:287-299.
Pujana, R. R. 2007. New fossil woods of Proteaceae from the Oligocene of southem Patagonia. Australian Systematic Botany 20:119-125.
Purnell, H. M. 1960. Studies of the family Proteaceae. I. Anatomy and morphology of the roots of some Victorian species. Australian Journal of Botany 8: 38-50.
Rowett, A. L & D. C. Christophel. 1990. The dispersed cuticle profile of the Eocene Anglesea clay lenses. Pp 115-121. In: J. G. Douglas & D. C. Christophel (eds). Proceedings, 3rd IOP conference, 1988. A-Z Press, Melbourne.
Rozefelds, A. C. 1990. A mid-Tertiary rainforest flora from Capella, Central Queensland. Pp 123-136. In: J. G. Douglas & D. C. Christophel (eds). Proceedings, 3rd IOP conference, 1988. A-Z Press, Melbourne.
--1992. The subtribe Hicksbeachiinae (Proteaceae) in the Australian Tertiary. Memoirs of the Queensland Museum 32:195-202.
--, D. C. Christophel & N. E Alley. 1992. Tertiary occurrence of the fern Lygodium (Schizaeaceae) in Australia and New Zealand. Memoirs of the Queensland Museum 32: 203-22.
--, M. E. Dettmann & H. T. Clifford. 2005. Xylocaryon lockii F.Muell. (Proteaceae) fruits from the Cenozoic of south eastern Australia. Kanunnah 1: 91-102.
Sauquet, H. E & D. J. Cantrill. 2007. Pollen diversity and evolution in Proteoideae (Proteales: Proteaceae). Systematic Botany 32:271 316.
Sauquet, H., E H. Weston, C. L. Anderson, N. E Barker, D. J. Cantrill, A. R. Mast & V. Savolainen. 2009a. Contrasted patterns of hyperdiversification in Mediterranean hotspots. Proceedings of the National Academy of Sciences, USA 106:221-225.
--,--, N. P. Barker, C. L. Anderson, D. J. Cantrill & V. Savolainen. 2009b. Using fossils and molecular data to reveal the origins of the Cape proteas (subfamily Proteoideae). Molecular Phylogenetics and Evolution 51: 31-43.
Scholtz, A. 1985. The palynology of the upper lacustrine sediments of the Amot Pipe, Banke, Namaqualand. Annals of the South African Museum 95: 1-109.
Sige, B., M. Archer, J.-Y. Crochet, H. Godthelp, S. Hand & R. Beck. 2009. Chulpasia and Thylacotinga, late Paleocene-earliest Eocene trans-Antarctic Gondwanan bunodont marsupials: new data from Australia. Geobios 42: 813-823.
Sleumer, H. 1954. Proteaceae americanae. Botanische Jahrbiicher 76:139-211. Stace, C. A. 1965. Cuticular studies as an aid to plant taxonomy. Bulletin of the British Museum of Natural History, Series E 4:1 78.
Stover, L. E. & A. D. Partridge. 1973. Tertiary and Late Cretaceous spores and pollen from the Gippsland Basin, southeastern Australia. Proceedings of the Royal Society of Victoria 85: 237286.
Takhtajan, A. L. 1969. Flowering plants: origin and dispersal. Smithsonian Institution, Washington, D. C.
Taylor, G., E. M. Truswell, K. G. MeQueen & M. C. Brown. 1990. The Early Tertiary palaeogeography, landform evolution and palaeoclimates of the Southern Monaro, NSW, Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 78:109-134.
Telford, I. R. H. 1995. Symphionema. Pp 133-135. In: P. McCarthy (ed). Flora of Australia, Vol. 16. CSIRO-ABRS, Melbourne.
Vadala, A. J. & A. N. Drinnan. 1998. Elaborating the fossil history of the Banksiinae: a new species of Banksieaephyllum (Proteaceae) from the late Paleocene of New South Wales. Australian Systematic Botany 11 : 439-463.
-- & D. R. Greenwood, 2001. Australian Paleogene vegetation and environments: evidence for palaeo-Gondwanic elements in the fossil records of Lauraceae and Proteaceae. Pp 196-221. In: I. Metcalfe, J. M. B. Smith, & I. Davidson (eds). Faunal and floral migrations and evolution in SE AsiaAustralasia. Swets & Zeitlinger Publishers, Lisse.
Waters, J. M. & D. Craw. 2006. Goodbye Gondwana? New Zealand biogeography, geology, and the problem of circularity. Systematic Biology 55:351-356.
Weston, P. H. 1995. Bellendena. Pp 125-127. In: P. McCarthy (ed). Flora of Australia, Vol. 16. CSIROABRS, Melbourne.
-- 2006. Proteaceae. Pp 36-404. In: K. Kubitzki (ed). The families and genera of vascular plants, Vol. 9. Springer, Berlin.
-- & M. D. Crisp. 1994. Cladistic biogeography of waratahs (Proteaceae: Embothrieae) and their allies across the Pacific. Australian Systematic Botany 7: 225-249.
-- & N. P. Barker. 2006. A new suprageneric classification of the Proteaceae, with an annotated checklist of genera. Telopea 11 : 314-344.
Wilf, P., K. R. Johnson, N. R. Cuneo, M. E. Smith, B. S. Singer & M. A. Gandolfo. 2005. Eocene plant diversity at Laguna del Hunco and Rio Pichileuffi, Patagonia, Argentina. American Naturalist 165: 634-650.
--, S. A. Little, A. Iglesias, M. C. Zamaloa, M. A. Gandolfo, N. R. Cuneo & K. R. Johnson. 2009. Papuacedrus (Cupressaceae) in Eocene Patagonia, a new fossil link to Australasian rainforests. American Journal of Botany 96:2031-2047.
Zamaloa, M. C., M. A. Gandolfo, C. C. Gonzalez, E. J. Romero, N. R. Cuneo & P. Wilf. 2006. Casuarinaceae from the Eocene of Patagonia, Argentina. International Journal of Plant Sciences 167: 1279-1289.
Raymond J. Carpenter (1,2)
(1) School of Earth & Environmental Sciences, University of Adelaide, Adelaide, South Australia 5005, Australia
(2) Author for Correspondence; e-mail: firstname.lastname@example.org
|Printer friendly Cite/link Email Feedback|
|Author:||Carpenter, Raymond J.|
|Publication:||The Botanical Review|
|Date:||Sep 1, 2012|
|Previous Article:||New Zealand Eocene, Oligocene and Miocene macrofossil and pollen records and modern plant distributions in the Southern Hemisphere.|
|Next Article:||Allotetraploids in patagonia with affinities to western North American diploids: did dispersal or genome doubling occur first?|