Dry rainforests: a productive habitat for Australian hunter-gatherers.

Beck, Wendy ; Balme, Jane

Abstract: Most recent work on hunter-gatherer use of rainforests has concluded that, although they are resource-rich, they are difficult environments for people to live in. This is largely because of the inaccessibility of many of the resources. However, most of this research has been on tropical or 'wet' rainforest types. Dry rainforests (seasonal rainforest) have not been so extensively studied, yet they have a wide distribution throughout the world and, in the past, had a much wider distribution. In Australia during the Pleistocene, the distribution was much wider than the present Holocene remnants suggest, especially in the northern and northeastern margin.

A comparison of plant resources from different rainforest types in northeastern New South Wales indicates that dry rainforests, in contrast to wet rainforest types, were potentially productive environments for Indigenous Australians in the past. Many of the species present in Australia's dry rainforests would have been familiar to the first human colonisers. The food resources are easier to access than wet rainforest species and the plant parts available occur in different proportions from those in wet rainforests. Accessible seeds are particularly abundant and so the successful exploitation of the full potential of these forests relies on specialised technologies.

**********

Dry rainforests are a type of rainforest that occurs in areas with a highly seasonal rainfall and, because of this, their vegetation fluctuates seasonally. In this paper we compare the plant resources available to hunter-gatherers in dry rainforests with those in other rainforest types. We have used northeastern New South Wales as our case study and we compare the relative contribution of economic plant species derived from dry rainforests and other rainforest types to characterise the differences between the resource bases of the rainforest types. We then compare strategies for the use of these plants and consider how identified differences between these strategies might be expressed in terms of cultural materials that might survive in the archaeological record. Finally, we consider the role of dry rainforests in Australia's colonisation.

Most recent research on the use of rainforests by hunter-gatherers is about their use of tropical rainforests (e.g. Bailey et al 1989; Bellwood 1985; Cosgrove 1996; Gnecco and Mora 1997; Politis 1996; Roosevelt et al 1996). The main controversy in this work is whether or not hunter-gatherers could live entirely within these forests without agriculture (Bailey et al 1989; Bailey 1990; Bailey and Headland 1991; Brosius 1991; Dwyer and Minnegal 1991; Endicott and Bellwood 1991). In these arguments the protagonists suggest that, although tropical rainforests are resource-rich, most of these resources are difficult to access. Ethnographically recorded groups of hunter-gatherers living in rainforests are said to rely on access to resources outside of the forest either through foraging trips outside or by trade with other groups of people (Bailey and Headland 1991). The main problem, they say, is that wild starchy foods such as yams are scarce and difficult to extract (Headland 1987). There have been many replies to these arguments, for example Townsend (1990), Colinvaux and Bush (1991), Sato (2001) and papers in Mercader (2002).

In a review of the controversy, Headland (2002) says that the question is not as simple as it once seemed, as rainforests are variable in their complexity. Harris (1978:122-5) has shown that Aboriginal population numbers in the rainforests of northeastern Australia varied, with higher densities associated with the more complex forests on the Atherton Plateau than in the neighbouring highlands. The argument about hunter-gatherer occupation of rainforests is also complicated by the fact that most forests today are anthropogenic. In addition, Headland points out that some of those who have attempted to contradict the hypothesis, such as Gardner (1993), have drawn their examples not from evergreen non-seasonal forests but from dry rainforests. The distribution types of resources in dry rainforests are quite different from those in tropical evergreen rainforest.

Rainforests can be described as usually moisture-loving, with 'a more or less closed canopy with one or usually more layers of ... trees and shrubs, generally from a number of different genera and families' (Baur 1957:191). Rainforests are 'distinguished from other closed canopy forests by the prominence of characteristic life forms such as epiphytes, lianas, root and stem structures, and the absence of annual herbs on the forest floor' (Webb 1978:352). However, the form of rainforests varies. In particular, there is a major division between 'wet' rainforests and 'dry' rainforests. Seasonal dry rainforests are usually smaller in stature, with canopy heights of 10 to 40 metres (or 5 to 9 m for vine thickets), compared with 20 to 84 metres for wet rainforests. They are also less complex than wet rainforests, with less diversity of species and fewer canopy layers (usually one or two, compared with three or more for wet rainforests) (Murphy and Lugo 1986:71-4).

Dry rainforests are known by a variety of other names, such as semi-evergreen forests (McDonald 1996; Murphy and Lugo 1986), subtropical or tropical dry forests (Murphy and Lugo 1986; Roth 1999) and deciduous seasonal forests (McDonald 1996). The foliage can be deciduous or, less commonly, evergreen and growth occurs in one or two pulses a year. Beyond this, dry rainforests are a very variable plant community and are difficult to define.

Dry rainforests are important as a resource because of their widespread distribution throughout the world. They occur in Africa, Asia, Australia, Central America and northern South America. Although it is not possible to determine the original extent of dry rainforests, it is likely that they had much wider distributions in the past (Murphy and Lugo 1986:67). In general they occur in tropical and semi-tropical areas where the mean annual temperature is over 17[degrees]C, frost-free and with an annual rainfall of 600 to 1800mm distributed over four to nine months (Murphy and Lugo 1986).

Dry rainforests often occur as small patches within mosaics of other vegetation types. Subtropical rainforests in northeastern New South Wales, for example, contain pockets of dry rainforest. Pockets of dry rainforest also occur in African savannah in vegetation environments argued to be similar to those occupied by early hominids (Sept 1994). Studies of hunter-gatherer exploitation of these areas must understand the contribution that these pockets might offer. This is particularly so because, compared with 'wet' rainforests, dry rainforests are able to quickly recolonise (Murphy and Lugo 1986:78). Thus, in changing climatic periods, they are more resilient and therefore more reliable than other rainforest types. This ability to recolonise rapidly, and their tolerance to thermal extremes and to seasonal lack of water, suggest that dry rainforests may have acted as refuges during the recorded drier periods in the past.

Murphy and Lugo's comparison of tropical and subtropical dry and wet rainforests from a variety of locations around the world identifies a number of differences relating to the lower biomass of dry rainforests compared with that of wet forests. However, as previous anthropological studies have indicated, what matters for successful hunter-gatherer exploitation is the accessibility of the resources, not just their presence. On this it is worth noting that the yams that may be difficult to access in wet rainforests may be more accessible in dry forests. For example, Murphy and Lugo (1986:72, Table 2) record that root biomass as a percentage of the total in dry forests is 8 to 50%, compared with less than 5 to 33% for wet. The figure for dry rainforests may be higher than this because the data are incomplete (Murphy and Lugo 1986:74). These figures indicate that the scarcity of yams identified in the hunter-gatherer rainforest studies is not true for dry forests. This and the other differences between dry and wet rainforests suggest that successful hunter-gatherer exploitation of each of these forests might require quite different foraging strategies.

If differences between different kinds of rainforests demand different strategies for their exploitation, this should be reflected in the associated material culture, especially where the material culture is related to specific plant parts. For example, rainforest seed exploitation is usually indicated by specialised nut-cracking stones and some tuber processing is associated with pestles or graters. Thus, tool assemblages should be a guide to the relative importance of use of rainforest types in the past and should help identify changing distribution of rainforest types.

Significance of dry rainforests for the colonisation of Australia

In Australia, five types of rainforest are recognised by botanists: tropical, warm temperate, cold temperate, subtropical and dry rainforest (McDonald 1996). Of these, dry rainforests have the widest distribution (Figure 1) and, in the Pleistocene, this distribution was considerably more extensive than its present Holocene remnants suggest (Gillison 1987). Although wet rainforests may be penetrated by fire during dry periods, their main susceptibility to fire is along their boundaries with other forests (Hopkins et al 1993). The seasonal nature of dry rainforest makes them more susceptible to firing. Russell-Smith 1985 has suggested that the reduction of dry rainforests may be in part because dry rainforests are sensitive to recent changes in human firing.

[FIGURE 1 OMITTED]

Dry rainforests are important for models of Aboriginal colonisation of Australia because species from the Indo-Malayan floral element are common in them. These genera, including Clerodendrum, Ficus, Erythrina and Terminalia, are also present in subtropical and tropical rainforest and are known to be used by both Indigenous Australians and by the indigenous peoples of Southeast Asia (Golson 1971). Their presence, in addition to the floristic diversity of subtropical and tropical 'wet' rainforest, has been argued by Bowdler (1983) to indicate that these forests were more 'susceptible to successful exploitation' by Aboriginal colonists than cool temperate rainforests of the south that lacked resources. Recent archaeological evidence from northern Queensland (Cosgrove 1996) and Tasmania (Cosgrove and Marshall 1990) lends some support to this interpretation.

If these interpretations are right, the presence of these genera in dry rainforest, with its much wider previous distribution, may indicate that dry rainforest was an important resource for the first Australian human colonisers. However, the lack of research specific to dry rainforests means that we know nothing about the implements associated with them.

Northeastern New South Wales

The area of our case study is shown in Figure 2. The eastern part of this region is a narrow coastal plain that rises steeply to the tablelands at 1000 to 1500 metres elevation. From the tablelands, the land gradually slopes away to the western plains. The four floristic zones defined in Figure 2--the North Coast, Northern Tablelands, Northwest Slopes and Northwest Plains--follow those of the State Herbarium of New South Wales as used by Harden (1990-93), and reflect major differences between vegetation communities in each floristic region.

[FIGURE 2 OMITTED]

The North Coast division is a forest zone consisting of tropical, subtropical, temperate rainforest and sclerophyll eucalypt forest. The Northern Tablelands are predominantly eucalypt forest composed of cold tolerant species. On the western parts of the tablelands, the forests become more stunted, grading into the dry sclerophyll woodlands that continue over the Northwest Slopes. The Northwest Slopes floristic zone corresponds closely to the Brigalow Belt Biogeographic Region which stretches through Queensland and northern New South Wales, and in which subtropical and temperate dry rainforest is relatively well preserved (Thackway and Cresswell 1995). The most western floristic zone in the area, the Northwest Plains, consists of shrubland and open woodland.

Each of these zones is, in fact, a mosaic in which several vegetation types exist. Dry rainforest occurs in each of the four floristic zones as discontinuous outliers and patches (Webb and Tracey 1994:91) which are linked by outliers in gullies, along streams and on hill and mountain tops, where 'fire shadows' occur. The patches of rainforest are often small and often adjacent to other kinds of vegetation such as eucalypt forests. Groves (1994:374) describes the rainforests of northern New South Wales as pockets surrounded by other vegetation types in which visually striking boundaries often occur between rainforest types and more open forests. For example, dry rainforest is particularly obvious in the gorges along the western margin of the North Coast zone. The distribution pattern of dry rainforest make it difficult to assess precisely the relative areas covered by this vegetation type in the four floristic zones; however, wet rainforest is concentrated in the North Coast region. Currently, in northern New South Wales, dry rainforest covers 71 000 hectares and 'wet' rainforests, including subtropical, warm temperate and cool temperate, cover 104 700 hectares (Floyd 1987:97).

Methods

Quantitative analysis of ethnobotanical data has been used previously by Golson (1971), Beck (1985) and others (e.g. Walsh 1987) to define more precisely the productivity of plant resources in an ecological zone and for modelling to test archaeologically. Commonly analysed features (e.g. Chatters 1987; O'Connell and Hawkes 1981) include measures of taxonomic richness (the number of resource species present within the environment), measures of abundance, and measures of energy expended and gained in acquiring resources. This last measure can include such factors as search-and-pursuit time, which in turn are related to the abundance and patchiness of resources as well as the technology used.

In our evaluations we have used only one measure--richness--to assess the productivity of each vegetation zone. The ethnobotanical information for this region derives largely from historical sources that are insufficiently precise to distinguish information about other aspects of productivity. Although our evaluation is not as fine-grained as it might be, it is nevertheless useful for comparing relative productivity between the four vegetation zones in the region.

For our evaluation of the role of different rainforest types in northeastern New South Wales, we asked four questions:

* How many economic plants (food and non-food) are available in each of the four floristic zones?

* Which rainforest types contribute to the food resources in each zone?

* Which rainforest types contribute to the non-food economic resources?

* Which plant parts are used as resources from each of the four floristic zones?

Our database for ethnobotanical resources is that compiled by Gott (nd). Gott's database for New South Wales documents ethnographic records of use of plant species utilised by Aboriginal people in New South Wales. Much of the information is from published sources and early European manuscripts, with particular attention to the earliest references. The database contains 465 forms covering 1457 species, which is about one-third of the plant species of New South Wales, but Gott (nd) considers that they represent only a part of the resource species available.

To answer our first question, we simply counted the economic plants (both food and non-food) for each of the zones. For the second question, we used Appendix 2 of McDonald (1996), which lists the vascular plant species recorded in vine thicket communities (dry rainforest) of the Brigalow Belt Biogeographic Region, to identify those plants which represent dry rainforest. Plants representing other rainforest types were identified from Floyd (1990), and Appendix 1 of Edmonds (1986), which lists subtropical rainforest resources from northern New South Wales from various primary and secondary resources, including Cribb and Cribb (1975), Maiden (1899), Petrie (1983[1904]) and Williams (et al 1984).

For the third question, we identified which of the economic plants in each floristic zone are food resources and then classified these plants into divisions based on the parts of the plant used: tubers, seeds and fruits. This allowed further characterisation of the difference in subsistence organisation associated with each type and gave us some idea of the processing costs associated with exploitation. For example, processing costs are generally highest for seeds and lowest for fruits (Beck 1985).

Finally, we characterised the non-food components of the economic resources by dividing them into medicine, fibre, adhesive, implement and other. To ensure that our comparisons were as precise as possible, genera that were not identified to species were not used in our calculations. This may, however, mean that some species have been omitted from consideration.

Results

The Appendix lists all of the food and non-food plants recorded in each of the four vegetation zones. These data have been summarised for each of the analyses below.

Economic plants available in each zone

Table 1 shows the numbers of food and non-food plants for each of the four vegetation zones, including the percentage that each number represents of the total economic plants. There are a few plants that are used for both food and non-food purposes but so few that they do not affect the percentages shown.

The North Coast zone has the greatest richness of plant resources (297 species) and there are slightly more economic plants in the Northwest Plains zone (214 species) than in the remaining two zones (185 and 208 in the Northern Tablelands and Northwest Slopes, respectively). The proportions of food to nonfood plants are almost identical across all zones.

Forest types contributing to food resources in each zone

Table 2 shows the numbers of dry rainforest and other rainforest food economic plants and the percentages of the total food plants in each floristic zone. It is clear from this table that the proportion of rainforest species' contribution to economic plants in all four zones is similar (ranging between 15% and 20%). Considering the different proportions of rainforest available in each region, this is perhaps surprising. For example, the contribution of resources from 'other rainforest types' is highest in the North Coast zone than in the remaining three zones, but it is not as high as might be expected considering the dominance of 'wet' rainforest types in the area. This is highlighted even further in Table 3, which shows the proportion of total rainforest food plants represented by dry rainforest in each of the four zones. Despite the much greater abundance of 'wet' rainforest types in the North Coast zone, nearly half of the rainforest economic plants derive from dry rainforest. The other figures in Table 3 more closely represent the proportions of different types of rainforest available in each zone.

The highest numbers of dry rainforests occur on the Northwest Slopes, which is also the zone with the highest proportion of dry rainforests. However, dry rainforest occurs on the Northwest Plains only in small remnant patches, yet resources from dry rainforest occur in higher proportions than in the North Coast or Northern Tablelands zones. Thus, the data from Tables 2 and 3 suggest that, in the northern New South Wales region, dry rainforests have a greater richness of food resources than other rainforest types.

Rainforest types contributing to non-food economic resources

Table 4 shows a similar pattern for the uses of nonfood economic plants across all four vegetation zones. Medicine is by far the most common non-food plant use in all zones, and adhesive the least common. The proportions within each zone are roughly similar, although there are comparatively few fibre plants on the Northwest Plains. The paucity of plants used as implements on the Northern Tablelands and North west Slopes is difficult to explain and requires more investigation into the types of implements made with the plant material. The contribution of dry rainforest plants to these non-food resources is surprisingly high, with medicine the most frequently recorded use.

Plant parts used as resources from each zone

The contrast between the two western and two eastern zones is further highlighted in the comparison between the types of food plants shown in Table 5. Both the North Coast and the Northern Tablelands have higher proportions of fruits and root foods and a relatively low number of seeds, while both of the western zones have a higher proportion of seeds. The Northwest Slopes zone is characterised by almost equal numbers of seeds, roots and fruits, whereas the Northwest Plains show a high dependence on seeds, as opposed to roots and fruits. As seeds are associated with greater processing costs (Beck 1985), plant resources in the western two zones would have relatively higher processing costs (with the Northwest Plains being the highest) than those associated with the North Coast and Northern Tablelands.

This difference in food plants is also reflected in the rainforest plants. Table 6 shows plant resources from dry rainforest are most commonly fruit or seed resources rather than roots. Plant resources from other rainforest types are usually fruits. Of these plant foods, seeds are the most costly in terms of processing, followed by roots, which usually require cooking, and then fruit.

Discussion

Dry rainforest economic resources

In our study area, rainforests as a whole provide 16 to 26% of economic plant species used by Indigenous Australians in northern New South Wales (Tables 1 and 4). The contribution of dry rainforests, however, has been particularly neglected in assessments of rainforest subsistence modes of hunter-gatherers. Our comparison of the plant resources available to hunter-gatherers in dry rainforests and other kinds of rainforest in northern New South Wales has shown that dry rainforests, in contrast to other types of rainforest, are potentially productive environments for hunter-gatherers. This is especially important because dry rainforests occur worldwide over a wider geographic range than other rainforest types. In northern New South Wales, dry rainforests make up approximately 40% of the area currently covered by rainforest (Floyd 1987:98), and were probably more extensive in the past. Thus, archaeologists need to understand more clearly the potential of dry rainforest plant resources.

As Tables 2 and 3 show, in comparison to other rainforest types, utilised plant species from dry rainforests contribute disproportionately to the numbers of rainforest economic plants in all the zones (see Appendix for species). Considering the relatively small patches of dry rainforest presently available in each of the floristic zones of northeastern New South Wales, the consistent contribution of dry rainforest economic plant species, both food and non-food uses, across all four floristic zones, is remarkable. In the North Coast zone, the number of food species used from 'dry' and 'other' rainforest types is roughly equivalent (20 species versus 24 species), and as we move further west the relative contribution of dry rainforest species increases while other rainforest species drop out (26 species versus 0 in the Northwest Plains zone).

Table 4 indicates that a relatively high number of species (between 9 and 22 in each zone) are also consistently used as medicines, fibre and implements. Most floristic zones are mosaics of a variety of vegetation types and obviously the relative economic importance of these resources is not just associated with the number of species but also depends on various factors, including the individual species abundance, productivity and accessibility, and the costs and benefits of processing.

Dry rainforest food types

In general, the total picture of carbohydrate-rich food proportions is that, as one moves from the coastal zones and through the tablelands to the inland slopes and plains, the proportion of seeds increases (from 24% to 42%) and roots decrease (from 38% to 27%). The proportion of fruits utilised remains much the same (39% to 31%) (Table 5). This geographic distribution reflects the overall pattern of starch storage in plant parts, as underground storage organs become less frequent in areas with lower, mainly summer rainfall (Pate and Dixon 1982:229), and above-ground starch storage (as seeds) becomes more prevalent.

Although the overall number of species is small, there are also some differences in the proportions of different foods available in the two types of rainforest of northeastern New South Wales (Table 6). There are more roots available in dry rainforest than in other rainforest (six species versus three species), but this difference is not substantial. The greatest difference is the much higher number of seeds in the dry rainforest (15 species) than in other rainforest types (only three species). Fruits are relatively equally contributed by both types of rainforest (18 and 17 species). Again, this may be a reflection of plants storing carbohydrate reserves in seeds or fruits in the more arid areas where dry rainforest can survive.

If we focus here on seeds and fruits as being the most significant dry rainforest food products, then the accessibility of seed and fruit resources should also be taken into account when comparing the forest types. Direct access to the fruits and seeds in dry rainforest should be easier than in other types of forest because of the generally lower canopy height, which averages 4 to 15 metres in this area. Fruits and seeds in subtropical and tropical rainforest, on the other hand, are difficult to access because of the high canopies, which average 20-40 m in this area (Webb and Tracey 1994).

Once plant resources are accessed, the food value of the seeds and fruits in dry rainforest is relatively high, and although costs of processing may also be high, it is likely that dry rainforest species of seeds and fruits are less costly to process than wet rainforest seeds (Cosgrove 1996). Processing is necessary because seeds and roots are relatively indigestible. An earlier study of Indigenous Australian food plants suggests that at least 75% of seeds and roots require processing (Beck 1985:75). Plant organs can be crudely organised on a scale of digestibility, with fruits being the most digestible and seeds and roots the least digestible (Stahl 1984). Roots and seeds often have woody or fibrous components that require processing, such as grinding, roasting or making into cakes or bread, to make them edible. Although the costs of plant processing are high, they are balanced by the reliable food values, seeds and roots having energy values over 400 kilojoules per 100 grams (Brand Miller et al 1993), whereas fruits are often less than that.

Many seeds and fruits in wet rainforests require complex processing to make them palatable or to remove toxins, a technique which raises the processing costs of these foods considerably. For example, seeds from toxic wet rainforest plants such as Castanospermum australe (Black Bean) require cooking for several hours, slicing and leaching for several days (Maiden 1899:281). There are several other examples (Cosgrove 1996). None of the dry rainforest food species in the study area requires complex processing of this kind, but a few (e.g. tree seeds of Pittosporum phillyraeoides, Santalum lanceolatum, and grass seeds, e.g. Sporobolus caroli) are processed by grinding and roasting (Beck 1985: Appendix). Thus, dry rainforest food products are often more accessible and less costly to process than wet rainforest foods, lending support to the proposition that dry rainforests are productive environments.

Implements associated with plant food processing

Artefacts are associated with plant food processing, such as grinding, cooking and pounding. Given the differences in the types of plants obtained from the different rainforest types, with more seed foods being supplied by dry rainforests (but without complex or very specialised processing), would this food processing be identifiable in the associated archaeological artefacts? In our analysis, the two eastern zones are most similar to each other in their relative contribution of dry rainforest (9 to 11%) to other rainforest economic plants and stand in contrast to the two western zones (16 to 17%, Table 2). If the difference in plant use has an archaeological signature, these two zones might be expected to be more similar in the implements relating to plant use than the two eastern zones. The greater proportion of seed foods overall in the western two zones is probably associated with greater processing costs and is expressed materially by the presence of higher frequencies of seed grinders (and other seed processing tools) than in the eastern zones (McBryde 1974:338).

Further ethnobotanical research is needed for each zone to establish the relationship between the particular seed species, especially dry rainforest ones, and their processing technology. In particular, the dry rainforest species are more likely to be tree seeds than grass seeds and therefore may have different kinds of processing tools. In addition, a focus on one tool type such as grindstones may not be enough. Analysis of macro-plant remains, in the manner of McConnell and O'Connor (1997) for Carpenter's Gap, and the identification of starch grains are also needed to fully understand plant use in the past.

Previous archaeological studies for the area have not examined plant processing tools specifically. For instance, the main regional archaeological study for northeastern New South Wales is that of McBryde (1974). In this classic study, McBryde analysed the rainfall, temperature, resource richness and topography of the four zones, together with ethnography and archaeology, and concluded that the Northern Tablelands and Northwest Slopes were less favourable environments for past hunter-gatherers than the North Coast and the Western Plains zones. The less favourable zones were identified as having poorer and more scattered resources, leading to seasonal and mobile occupation that is reflected in the variation between zones of the stone artefact assemblages. The general differences in artefact assemblages identified in northeastern New South Wales by McBryde may reflect differences in the abundance and reliability of resources offered by the different environments. However, as very little is known about the Aboriginal use of forests in each of these zones (Byrne 1987), the contribution of dry rainforest resources to the differences identified for each of the zones is not clear.

Implications for colonisation

As others have suggested (Bowdler 1983; Golson 1971), the familiarity of the Indo-Malayan floral element to the first Australian people may have made rainforests important in their successful adaptation to Australia. However, previous researchers have not differentiated dry rainforest from other kinds of rainforest. The comparatively accessible abundance of resources in dry rainforests might have made them especially important for successful exploitation of the continent. If so, future research into the technologies required to take advantage of the full range of resources is also worth considering.

There is some controversy in Australia about the timing of the introduction of the use of seed-grinders. Smith (1986) argues that they were introduced in the mid-Holocene while, at the other extreme, Fullagar and Furby (1997) and Gorecki et al (1997) argue for a Pleistocene introduction. If the latter is true, it increases the potential importance of dry rainforests in the colonisation of Australia. However, if the introduction of grinders was more recent, on our figures, dry rainforest might have less resource importance for hunter-gatherers until appropriate grinding technology was available for dry rainforest seeds. These forests need to be more specifically examined in future research. Further evidence on the role of rainforests in colonisation could be acquired by re-examining older Australian sites for additional direct and indirect evidence of dry rainforest utilisation. For example, many of the sites in the Central Queensland region fall within the vine-thicket vegetation distribution zone of dry rainforest.

Conclusion

Dry rainforests are a neglected part of the study of Australian hunter-gatherer exploitation of rainforests. Previous studies have focused only on the wet rainforests of temperate and subtropical and tropical areas. Dry rainforests are an important historical resource zone because they cover significant areas of the rainforest regions of Australia and elsewhere. Although their current distribution is patchy, dry rainforests were probably much more widely distributed in the past. In the study area in northern New South Wales, dry rainforests make up 40% of the rainforest area, and they are productive zones for hunter-gatherers.

Dry rainforests contribute consistent numbers of economic plants, both food species and those used for medicine, implements and fibres. The food plants, especially the seeds, are higher in energy, more accessible and less costly to process than carbohydrate resources in other rainforest zones in the area, as well as occurring in areas, such as the Northwest Slopes, which are not conventionally considered rainforest zones. Such rainforest outliers may be important sources of food and medicinal plants.

Some parts of this paper have been speculative and to explore more fully the role of dry rainforests in the colonisation of Australia we need to collect more data on the relative abundance and nutritional values of the various types of rainforests species and on the archaeological signatures of dry rainforest exploitation.

APPENDIX

Fruits NC NT NWS NWP

Acacia farnesiana * * * *
Acacia verniciflua *
Acmena smithii * * *
Acronychia oblongifolia *
Acrotriche spp. * * *
Alangium villosum *
Alpinia caerulea *
Amyema & related spp. * * * *
Astroloma humifusum *
Austromyrtus dulcis *
Billardiera spp. * *
Brachyloma spp. * * * *
Canthium latifolium * *
Capparis spp. * * * *
Carissa ovata * * *
Carpobrotus glaucescens *
Cassytha spp. * * * *
Cissus antarctica * *
Cissus hypoglauca * *
Citriobatus pauciflorus * *
Commelina ensifolia * *
Coprosma hirtella *
Coprosma quadrifida * *
Coprosma sp. * *
Cucumis melo ssp. * *
 agrestis
Cymbidium canaliculatum * * * *
Cyttaria sp. * *
Davidsonia pruriens *
Dianella spp. * * * *
Diospyros australis *
Diploglottis australis *
Dodonaea viscosa ssp. * * *
 angustissima
Einadia spp. * * * *
Elaeocarpus spp. * *
Enchylaena tomentosa * * *
Eremocitrus glauca *
Eremophila debilis * * * *
Eremophila desertii * * *
Eremophila longifolia * *
Eupomatia laurina * *
Eustrephus latifolius * * *
Exocarpos aphyllus * *
Exocarpos cupressiformis * * * *
Exocarpos latifolius *
Exocarpos nanus *
Exocarpos strictus * *
Ficus coronata * * *
Ficus fraseri *
Ficus macrophylla ssp. *
 macrophylla
Ficus rubiginosa * * * *
Gaultheria appressa *
Grewia latifolia *
Hicksbeachia pinnatifolia *
Hovea longipes * *
Leucopogon spp. * * * *
Lissanthe spp. * * * *
Lysiana spp. * * *
Malaisia scandens *
Marsdenia australis *
Maytenus cunninghamii * *
Melastoma affine *
Melichrus spp. * * * *
Melodinus australis *
Microcitrus australasica *
Monotoca spp. * *
Muehlenbeckia spp. * * * *
Myoporum spp. * * * *
Nitraria billardieri *
Owenia acidula * *
Passiflora species * * *
Persoonia spp. * * * *
Physalis minima * *
Pimelea microcephala * *
Piper novae-hollandiae *
Planchonella australis *
Podocarpus elatus *
Polyalthia nitidissima *
Polyscias sambucifolia * *
Psychotria simmondsiana *
Randia sp. *
Rauwenhoffia leichhardtii *
Rhagodia spp. * * * *
Rhyncharrhena linearis *
Rubus, native spp. * * *
Sambucus spp. * * * *
Santalum acuminatum * *
Santalum lanceolatum * *
Santalum obtusifolium * *
Sarcozona praecox *
Scaevola spinescens *
Schizomeria ovata * *
Sida corrugata * * * *
Solanum americanum * *
Solanum aviculare * * * *
Solanum cleistogamum *
Solanum coactiliferum *
Solanum ellipticum *
Solanum esuriale * *
Solanum laciniatum * *
Solanum, native spp. * * * *
Solanum opacum * * * *
Solanum stelligerum * *
Solanum vescum * * * *
Styphelia spp. * * * *
Syzygium spp. *
Trochocarpa spp. * *
Zygophyllum * *
Totals 85 54 53 50

Roots NC NT NWS NWP

Acianthus spp. * * *
Actites megalocarpa *
Adenochilus nortonii *
Alisma plantago-aquatica * *
Alocasia brisbanensis *
Alpinia caerulea *
Arthrochilus spp. * *
Arthropodium milleflorum * * * *
Arthropodium minus * * *
Blechnum spp. *
Boerhavia sp. * * * *
Bolboschoenus spp. * *
Brachychiton populneus * * * *
Bulbine bulbosa * * * *
Burchardia umbellata *
Caesia spp. * * * *
Caladenia spp. * * * *
Calandrinia spp. * * * *
Caleana spp. * * *
Calochilus spp. * * * *
Calostemma purpureum *
Casuarina pauper * * *
Cheirostylis ovata *
Chiloglottis spp. * * *
Cissus opaca * * * *
Clematis microphylla * * *
Clerodendrum floribundum *
Codonocarpus cotinifolius *
Convolvulus erubescens * * * *
Corybas spp. * *
Crinum spp. * * *
Cryptostylis spp. *
Cucumis melo ssp. * *
 agrestis
Curculigo ensifolia *
Cyperus bulbosa *
Cyperus victoriensis * *
Cyrtostylis reniformis * *
Dendrobium spp. * * *
Dichopogon spp. * * *
Dicranopteris linearis *
Dioscorea transversa * *
Dipodium spp. * * * *
Diuris spp. * * *
Doryanthes excelsa *
Eleocharis dulcis *
Eleocharis sphacelata * * *
Epipogium roseum *
Eriochilus spp. * * *
Erodium crinitum * * *
Erythrina vespertilio *
Eucalyptus coolabah *
Eucalyptus dumosa *
Eustrephus latifolius * * *
Gastrodia spp. * * *
Genoplesium spp. * * *
Geodorum densiflorum *
Geranium spp. * * * *
Glossodia spp. * * *
Hakea leucoptera *
Hibiscus heterophyllus * *
Hibiscus tiliaceus *
Hypoxis spp. * * *
Ipomoea pes-caprae ssp. *
 brasiliensis
Lepironia articulata *
Lyperanthus spp. *
Marsdenia australis * *
Marsdenia spp. * *
Microseris lanceolata * * * *
Microtis spp. * * * *
Murdannia graminea * * * *
Neopaxia australasica *
Nymphaea gigantea *
Nymphoides crenata * *
Oreomyrrhis spp. * *
Orthoceras strictum * * *
Oxalis perennans * * *
Pelargonium australe * *
Phragmites australis * * * *
Portulaca oleracea * * *
Prasophyllum spp. * * * *
Pteridium esculentum * *
Pterostylis spp. * * * *
Rhizanthella slateri *
Rumex crystallinus *
Schoenoplectus litoralis *
Schoenoplectus validus * * * *
Sonchus oleraceus * * * *
Spiranthes sinensis * *
Sterculia quadrifida *
Thelymitra spp. * * *
Thysanotus spp. * * * *
Trichosanthes subvelutina *
Triglochin procerum * * *
Typha spp. * * * *
Typhonium brownii * *
Vigna sp. * * *
Wurmbea spp. * * * *
Xanthorrhoea spp. * * * *
Zeuxine oblonga *
Totals 83 57 52 44

Seeds NC NT NWS NWP

Abutilon otocarpum *
Acacia aneura *
Acacia binervata * *
Acacia brachystachya *
Acacia farnesiana * * * *
Acacia implexa * * * *
Acacia ligulata *
Acacia longifolia * *
Acacia melanoxylon * * *
Acacia murrayana * *
Acacia oswaldii * *
Acacia pycantha *
Acacia salicina * * * *
Acacia sophorae *
Acacia stenophylla * *
Acacia tetragonophylla *
Acacia verniciflua *
Alectryon oleifolius * *
Ammania multiflora *
Astrebla spp. * *
Atalaya hemiglauca * *
Atriplex spp. * * * *
Avicennia marina *
Brachychiton populneus * * * *
Bromus arenarius * * *
Calandrinia spp. * * * *
Calostemma purpureum *
Canavalia rosea *
Castanospermum australe *
Chenopodium spp. * * * *
Cleome viscosa * * *
Cucumis melo * *
Dactyloctenium radulans * * * *
Danthonia sp. *
Digitaria coenicola *
Diploglottis australis *
Dysphania rhadinostachya *
Enchylaena tomentosa * * *
Enneapogon avenaceus *
Eragrostis dielsii *
Eragrostis eriopoda *
Eragrostis parviflora * * * *
Eriochloa pseudoacrotricha * * * *
Erodium crinitum * * *
Eucalyptus camaldulensis * * *
Eucalyptus coolabah *
Eucalyptus largiflorens *
Flindersia maculosa *
Floydia prealta *
Gahnia spp. * * * *
Grevillea spp. * * * *
Grevillea striata * *
Heliotropium asperrimum *
Hicksbeachia pinnatifolia *
Hovea longpipes * *
Juncus spp. * * * *
Linum marginale * * * *
Lysiana spp. * * *
Lysiphyllum carronii * *
Macadamia tetraphylla *
Macrozamia sp. * * *
Malaisia scandens *
Marsilea spp. * * * *
Maytenus cunninghamii * *
Monachather paradoxa *
Mucuna gigantea *
Muehlenbeckia spp. * * * *
Nymphaea gigantea *
Pandanus tectorius *
Panicum native spp. * * * *
Paspalidium jubiflorum * *
Pittosporum phylliraeoides * *
Planchonella australia *
Plantago cunninghamii * * * *
Podocarpus elatus *
Polygonum plebeium * * *
Portulaca oleracea * * *
Rumex crystallinus *
Salsola kali * * *
Santalum acuminatum * *
Santalum lanceolatum * *
Schoenoplectus *
dissachanthus
Schoenoplectus validus * * * *
Senna artemisioides * * *
Sesbania cannabina * * *
Sida corrugata * * * *
Sporobolus actinocladus *
Sporobolus caroli * *
Sterculia quadrifida *
Themeda avenacea * *
Tragus australianus * * *
Triodia spp. * * * *
Urochloa sp. * * *
Xylomelum pyriforme *
Totals 52 29 55 70

Non-food species NC NT NWS NWP

Acacia cambagei *
Acacia doratoxylon * *
Acacia falcata * *
Acacia pendula * * *
Acacia penninervis * * *
Acacia rigens *
Adriana glabrata var. * * * *
 subglabra
Aegiceras corniculatum *
Ajuga australis * * * *
Alloteropsis semialata * * *
Alphitonia excelsa * * *
Alstonia constricta * * *
Araucaria cunninghamii *
Atriplex nummularia *
Baumea rubiginosa * *
Bruguiera gymnorhiza *
Bursaria spinosa * * * *
Calamus muelleri *
Callitris spp. * * * *
Canarium australasicum *
Carex tereticaulis *
Centaurium spicatum * *
 (?introduced)
Centipeda cunninghamii * * *
Centipeda minima * * * *
Centipeda thespidioides * *
Chamaesyce drummondii * *
Chenopodium cristatum * *
Chrysopogon fallax * * *
Cleome viscosa * * *
Commersonia bartramia *
Corchorus cunninghamii *
Cymbopogon obtectus * * * *
Cyperus gymnocaulos/ * * * *
 vaginatus
Dendrocnide excelsa * * *
Dendrocnide photinophylla *
Digitaria brownii * * * *
Dodonaea truncatiales *
Duboisia hopwoodii *
Duboisia myoporoides * *
Enteropogon species * * *
Eragrostis australasica * *
Eremophila gilesii *
Eremophila mitchellii * *
Eucalyptus agglomerata *
Eucalyptus crebra * * *
Eucalyptus obliqua *
Eucalyptus sideroxylon * * *
Evolvulus alsinoides * * *
 ?var. villosicalyx
Excoecaria agallocha *
Flagellaria indica *
Geijera parviflora * *
Goodenia spp. * * * *
Gratiola pedunculata * * *
Haemodorum planifolium * * *
Hedycarya angustifolia * *
Heteropogon contortus * * * *
Imperata cylindrica * * *
Indigofera australis * * * *
Jagera pseudorhus *
 var. pseudorhus
Kunzea ericoides *
Lechenaultia divaricata *
Lepidosperma elatius * *
Lepidosperma laterale * * *
Lophostemon suaveolens *
Macaranga tanarius *
Melanthera biflora *
Melicope octandra *
Nicotiana sp. * * *
Pandorea pandorana * * * *
Petalostigma pubescens * * *
Phebalium squameum *
Philydrum lanuginosum * *
Phyllanthus lacunarius *
Petalostigma pubescens *
Pluchea dentex/tetranthera *
Poa spp. * * * *
 (large tussock spp.)
Polyscias murrayi *
Pomaderris aspera * *
Prostanthera lasianthos * * *
Prostanthera striatiflora *
Pseudognaphalium * * * *
 luteo-album
Pseudoweinmannia *
 lachnocarpa
Psoralea patens *
Pterocaulon sphacelatum *
Ripogonum album *
Rulingia dasyphylla * * *
Sarcostemma australe *
Schoenus melanostachys * *
Spartothamnella juncea * * *
Spinifex sericeus *
Stephania japonica * * *
 var. discolor
Stylidium spp. * * * *
Syncarpia glomulifera *
Tabernaemontana *
 pandacaqui
Tephrosia brachyodon * * *
Themeda triandra * * * *
Tinospora smilacina *
Tricoryne anceps *
subsp. pterocaulon
Verbena officinalis * * * *
 (introduced)
Wikstroemia indica *
Zieria arborescens * *
Zieria smithii * *
Totals 77 45 47 50

Table 1 Number of economic plants in each zone and the percentage of
each as the total economic plants for each zone

 NC NT NWS NWP

Food species 220 (74%) 140 (76%) 160 (77%) 164 (77%)
Non-food species 77 (26%) 45 (24%) 48 (23%) 50 (23%)
Totals 297 185 208 214

Note: NC=North Coast, NT=Northern Tablelands, NWS=Northwest Slopes,
NWP=Northwest Plains.

Table 2 Number of rainforest food plants and the percentage that
number represents of the total food plants identified for each zone

 NC NT

Dry rainforest food species 20 (9%) 15 (11%)
Other rainforest food species 24 (11%) 12 (9%)
Non rainforest food species 190 (80%) 127 (81%)

 NWS NWP

Dry rainforest food species 28 (17%) 26 (16%)
Other rainforest food species 6 (4%) 0 (0%)
Non rainforest food species 133 (79%) 148 (84%)

Table 3 Number and percentage of total rainforest food plants in
each zone represented by dry rainforest plants

 NC NT

Total rainforest food species 44 27

Dry rainforest food species 20 (45%) 15 (56%)

 NWS NWP

Total rainforest food species 34 26

Dry rainforest food species 28 (82%) 26 (100%)

Table 4 Number of non-food economic plants and percentage
contribution of dry rainforest plants in each zone

 NC NT

Medicine 35 23
Implement 22 10
Adhesive 6 1
Fibre 22 16
Other 7 5
Totals 92 55
No. and % of total of dry 22 (24%) 9 (16%)
rainforest non-food plants

 NWS NWP

Medicine 28 30
Implement 10 11
Adhesive 2 2
Fibre 12 5
Other 5 7
Totals 57 55
No. and % of total of dry 15 (26%) 14 (25%)
rainforest non-food plants

Table 5 Number of root, fruit and seed food plant species used in
each zone

 NC NT NWS NWP Total

Roots 83 (38%) 57 (41%) 52 (33%) 44 (27%) 99
Fruit 85 (39%) 54 (39%) 53 (33%) 50 (31%) 107
Seeds 52 (24%) 29 (21%) 55 (34%) 70 (42%) 94
Total 220 140 160 164

Note: The total in the right-hand column is less than the total
of all columns because some species occur in more than one zone.

Table 6 Number of dry rainforest and other rainforest plants in each
food plant category for all zones

 Dry rainforest Other rainforest Total food species
 food species food species in all zones

Roots 6 (6%) 3 (3%) 99
Fruit 17 (16%) 18 (17%) 107
Seeds 15 (16%) 3 (3%) 94

ACKNOWLEDGMENTS

We would like to thank Beth Gott for allowing us to use her New South Wales plant database for this work. Heather Burke helped to compile the plant tables and prepared the illustrations. Mike Roach drew the final illustrations. Three anonymous reviewers provided useful comments on an earlier draft of this paper.

REFERENCES

Bailey, R.C. 1990 Exciting opportunities in tropical rainforests: a reply to Townsend, American Anthropologist 92:747-8.

Bailey, R.C., G. Head, M. Jenike, M. Owen, R. Rechtman and E. Zechenter 1989 Hunting and gathering in tropical rainforest: is it possible? American Anthropologist 91:58-82.

Bailey, R.C. and T. Headland 1991 The tropical rainforest: is it a productive environment for human foragers? Journal of Human Ecology 19:261-85.

Baur, G.N. 1957 Nature and distribution of rainforests in New South Wales, Australian Journal of Botany 5:190-233.

Beck, J. 1985 Technology, toxicity and subsistence: Australian Aboriginal plant food processing, doctoral thesis, La Trobe University, Melbourne.

Bellwood, P. 1985 Prehistory of the Indo-Malaysian Archipelago, Academic Press, Sydney.

Bowdler, S. 1983 Rainforest: colonised or coloniser? Australian Archaeology 17:59-66.

Brand Miller, J., K.W. James and P.M.A. Maggiore 1993 Tables of Composition of Australian Aboriginal Foods, Aboriginal Studies Press, Canberra.

Brosius, J.P. 1991 Foraging in tropical rainforest: the case of Penan of Sarawak, East Malaysia, Human Ecology 19:123-50.

Byrne, D. 1987. The Aboriginal and archaeological significance of the New South Wales rainforest, a report to Forestry Commission of New South Wales and Australian Heritage Commission, Forestry Commission of New South Wales.

Chatters, J.C. 1987 Hunter-gatherer adaptations and assemblage structure, Journal of Anthropological Archaeology 6:336-75.

Colinvaux, P.A. and M.B. Bush 1991 Rain-forest ecosystem as a resource for hunting and gathering, American Anthropologist 93:153-60.

Cosgrove, R. 1996 Origin and development of Australian Aboriginal tropical rainforest culture: a reconsideration, Antiquity 70:900-12.

Cosgrove, R.J.A. and B. Marshall 1990 Paleo-ecology and Pleistocene human occupation in south central Tasmania, Antiquity 64:59-78.

Cribb, A.B. and J.W. Cribb 1975 Wild Foods in Australia, Collins, Sydney.

Dwyer, P. and M. Minnegal 1991 Hunting in lowland, tropical rainforest: towards a model of non-agricultural subsistence, Human Ecology 19:187-212.

Edmonds, V. 1986 Subtropical rainforest: Archaeological evidence of Aboriginal use in the border ranges, S.E. Queensland, Master of Letters dissertation, University of New England, Armidale.

Endicott, K. and P. Bellwood 1991 The possibility of independent foraging in the rain forest of peninsular Malaysia, Human Ecology 19:151-85.

Floyd, A.G. 1987 Status of rainforest in northern New South Wales. In G. Werren and P. Kershaw (eds), The Rainforest Legacy: Australian national rainforests study, Australian Heritage Commission Special Heritage Publication Series 7(1), AGPS, Canberra, 95-103.

--1990 Australian Rainforests in New South Wales, vol. II, Surrey Beatty and Sons, Sydney.

Fullagar, R.L.K. and J. Furby 1997 Pleistocene seed-grinding implements from the arid zone, Antiquity 71:300-7.

Gardner, P.M. 1993 Dimensions of subsistence foraging in South India, Ethnology 32:109-44.

Gillison, A.N. 1987 The 'dry' rainforests of Terra Australis. In G. Werren and P. Kershaw (eds), The Rainforest Legacy: Australian national rainforests study, vol. 1, Australian Heritage Commission Special Heritage Publication Series 7(1), AGPS, Canberra, 305-44.

Gnecco, C. and S. Mora 1997 Late Pleistocene/Early Holocene tropical forest occupations at San Isidra and Pena Roja, Colombia, Antiquity 71:683-90.

Golson, J. 1971 Australian Aboriginal food plants: some ecological and cultural-historical implications. In J. Golson and D.J. Mulvaney (eds), Aboriginal Man and Environment in Australia, Australian National University Press, Canberra, 196-238.

Gorecki, P., M. Grant, S. O'Connor and P. Veth 1997 The morphology, function and antiquity of Australian grinding implements, Archaeology in Oceania 32:141-50.

Gott, B. nd NSWUSE Database, unpublished report (copy held by AIATSIS). (last updated 1999)

Groves, R.H. 1994 Present vegetation types. In R.H. Groves (ed.), Australian Vegetation, 2nd edn, vol. 1, Cambridge University Press, 369-401.

Harden, G.J. (ed.) 1990-93 Flora of New South Wales, vols 1-4, University of NSW Press, Sydney.

Harris, D.R. 1978 Adaptation to a tropical rainforest environment: Aboriginal subsistence in northeastern Queensland. In N.B. Jones and V. Reynolds (eds), Human Behaviour and Adaptation, Taylor and Francis, London, 113-34.

Headland, Thomas N. 1987 The wild yam question: how well could independent hunter-gatherers live in a tropical rain forest ecosystem? Human Ecology 15:463-91.

--2002 Could 'pure' hunter-gatherers live in a rainforest? A1999 review of the current status of the wild yam question (written March 1999, updated May 2002), <www.sil.org/~headlandt/ wildyam.htm> [accessed 2 September 2003].

Hopkins, M.S., J. Ash, A.W. Graham, J. Head and R.K. Hewett 1993 Charcoal evidence of the spatial extent of eucalyptus woodland expansions and rainforest contractions in north Queensland during the Late Pleistocene, Journal of Biogeography 20:357-72.

Maiden, J.H. 1899 Native food plants, Agricultural Gazette of New South Wales 10(2):117-30, 279-90, 618-39, 730-40.

McBryde, I. 1974 Aboriginal Prehistory in New England: An archaeological survey of northeastern New South Wales, Sydney University Press.

McConnell, K. and S. O'Connor 1997 40,000 year record of food plants in the southern Kimberley ranges, Western Australia, Australian Archaeology 45:20-31.

McDonald, W.J.F. 1996 Spatial and temporal patterns in the dry seasonal rainforests of southern Queensland, doctoral thesis, University of New England, Armidale.

Mercader, J. (ed.) 2002 Under the Canopy: The archaeology of tropical rainforests, Rutgers University Press, New Jersey.

Murphy, P.G. and A.E. Lugo 1986 Ecology of a tropical dry forest, Annual Review of Ecology and Systematics 17:67-88.

O'Connell, J. and K. Hawkes 1981 Alyawara plant use and optimal foraging theory. In B. Winterhalder and E. Alden Smith (eds), Hunter-Gatherer Foraging Strategies: Ethnographic and archaeological analysis, University of Chicago Press, 99-125.

Pate, J. and K. Dixon 1982 Tuberous, Cormous and Bulbous Plants: Biology of an adaptive strategy in Western Australia, University of Western Australia Press, Perth.

Petrie, C.C. 1983 (1904) Tom Petrie's Reminiscences of Early Queensland, Angus & Robertson, Sydney.

Politis, G.G. 1996 Moving to produce: Nukak mobility and settlement patterns in Amazonia, World Archaeology 27:492-511.

Roth, Linda C 1999 Tropical dry rainforest change, Journal of Biogeography 26:739-59.

Roosevelt, A.C., M. Lima Da Costa, C. Lopes Machado, M. Michab, N. Mercier, H. Valladas, J. Feathers, W. Barnett, M. Imazo Da Silverira, A. Henderson, J. Silva, B. Chernoff, D.S. Reese, J.A. Holman, N. Toth and K. Schick 1996 Paleoindian cave dwellers in the Amazon: the peopling of the Americas, Science 271:373-84.

Russell-Smith, J. 1985 Studies in the jungle: people, fire and monsoon forest. In R. Jones (ed.), Archaeological Research in Kakadu National Park, Special Publication no. 13, Australian National Parks and Wildlife Service, Canberra, 241-67.

Sato, H. 2001 The potential of edible wild yams and yam-like plants as a staple food resource in the African tropical rain forest, African Study Monographs, Supplementary issue 26:123-34.

Sept, J.M. 1994 Beyond bones: archaeological sites, early hominid subsistence, and the costs and benefits of exploiting wild plant foods in east African riverine landscapes, Journal of Human Evolution 27:295-320.

Smith, M.A. 1986 The antiquity of seed grinding in arid Australia, Archaeology in Oceania 21:29-3.

Stahl, A.B. 1984 Hominid dietary selection before fire, Current Anthropology 25(2):151-68.

Thackway, R. and I. Cresswell (eds) 1995 An Interim Biogeographical Regionalisation for Australia (version 4.0), Australian Nature Conservation Agency, Canberra.

Walsh, F. 1987 The influence of the spatial and temporal distribution of plant food resources on the traditional Martujarra subsistence strategies, Australian Archaeology 25:88-101.

Webb, L.J. 1978 A general classification of Australian rainforests, Australian Plants 9:349-63.

Webb, L.J. and J.G. Tracey 1994 The rainforests of northern Australia. In R.H. Groves (ed.), Australian Vegetation, Cambridge University Press, 87-130.

Williams, J.B., G.J. Harden and W.G.F. McDonald 1984 Trees and Shrubs in Rainforests of New South Wales and Southern Queensland, University of New England, Armidale.

Wendy Beck is an academic at the University of New England, and an archaeologist who has a special interest in how people use spaces, both in the past and the present.

Archaeology, School of Human and Environmental Studies, University of New England, Armidale NSW 2351 <wbeck@metz.une.edu.au>

Jane Balme is an academic at the University of Western Australia who has been teaching and researching the archaeology of Indigenous Australia for many years. Her interests are broad but most of her research has been about the relationships between people and their changing geographical and social environments.

Centre for Archaeology, University of Western Australia, Crawley WA 6009 <jbalme@cyllene.uwa.edu.au>