The use of Australian native grains as a food: a review of research in a global grains context

Keywords: grassland, purslane, Panicum, Themeda, Acacia, Microlaena.

The use of the grain of grassland species as an energy-rich food source with a long storage life has roots in multiple cultures around the world. Archaeological evidence of domestication of species such as wheat, barley, chickpea and faba bean is present from the Epipalaeolithic period (13 000 years ago) from various sites in the Middle East, and continues into the Neolithic Period in other locations in Asia, Africa and the Americas (Purugganan and Fuller 2009). Today, the grains of wheat, rice, maize and sorghum make up 50.9% of the total weight of human food staples produced globally, with oilseeds, pulses, meat, dairy, fish, sugar and root tubers contributing the rest. Just two species, wheat and rice, make up 41.1% of total staple foods, with ~747 577 Mt of wheat and 511 663 Mt of rice being produced annually (OECD/FAO 2020). The human-mediated process of selection, which has made these annual species so successful, has led to an increase in grain size, loss of seed shattering, ease of post-harvest processing and successful germination in cultivated soil (Purugganan and Fuller 2009).

The grains of native grasses and grassland species in Australia have also been used as a food source for thousands of years, but have been largely overlooked as they were not domesticated according to the definition used by most archaeologists and scientists (see Section Plant breeding management and propagation; Gammage 2012; Pascoe 2014). More than 1100 species of perennial native grass are distributed widely across Australia, with ~42 having records of being used as a food source (Chivers et al. 2015). They are well adapted to Australia’s environment, possessing an ability to establish on poor soils and tolerate stresses, including intermittent droughts and fires (Waters et al. 2001). Aboriginal nations and language groups each had different species preferences and ways of processing the grains according to the climate, soils and culture of their region, although there are many similarities in the tools and methods used (Mildwaters and Clarkson 2020).

Some of the species are native to multiple continents, and have been processed and consumed in ways similar to Aboriginal practices by the local inhabitants. For example, Themeda triandra (kangaroo grass) is also native to tropical and southern Africa where grains are eaten in times of famine (National Research Council 1996). Seed of Portulaca oleracea (purslane) has been used as food and medicine throughout Asia, the Middle East, eastern Europe and more (Amirul Alam et al. 2014). Pods and other parts of Vachellia farnesiana (mimosa bush, formerly Acacia farnesiana) have documented medicinal, food and recreational uses in various parts of Indonesia and Africa (Cribb and Cribb 1981). Species closely related to some of those used on the Australian continent have undergone domestication on other continents, such as Sorghum spp., Oryza spp., Panicum spp., Brachiaria spp. and Dactyloctenium spp. (National Research Council 1996; Klmata et al. 2000; Hunt et al. 2008).

Interest in production of native grains for food in Australia has arisen in the last few years. This has been driven by consumer demand for healthy, natural, ancient and gluten-free grains with purported environmental and/or cultural benefits, as well as Aboriginal communities who desire to bring back productive native grasslands for healing of culture and country, and/or employment opportunities on-Country (Pascoe 2014; Sydney Institute of Agriculture 2020). The market for native grass seed is already growing owing to demand for seed for revegetation, resilient pastures and ecosystem restoration (Reseigh et al. 2008). Conventional crop production in Australia is economically risky; between 1975 and 2011, agriculture was the most volatile sector of the Australian economy (Keogh 2012), and grains and oilseeds were consistently the most volatile subsector. A large contributing factor to this volatility is the effect of drought and climate variability, which reduced average annual cropping farm profits by 35% from 2000 to 2019 (Hughes et al. 2019). In response to the risks associated with a heavy reliance on a narrow range of cereal crops, Chivers (2012) proposed that Australian native perennials could offer a low-risk option for broadening grain production. Diversifying agricultural crop production has been identified as a rational and cost-effective way to build resilience to climate change into agricultural systems by buffering production (Lin 2011). However, it is not known to what extent native grasses might be able to mitigate economic risks through their inherent resilience to adverse climatic conditions as the economic value of native grains in Australian (or global) food markets has not yet been tested.

This synthesis of previous research aims to bring together sources of knowledge of how Australian native grasses have been used as food, and to identify their potential in the context of the current globalised grains market. Much of the previous research has been sourced from ethnobotanical literature, analysis of archaeological artefacts and accounts from early explorers. As a result, current understanding relies heavily on observational, secondary and anecdotal evidence. It is important to note that this information is far from comprehensive. The available evidence offers site-specific descriptions; however, considering the heterogeneity of Aboriginal peoples, it may not be reasonable to extrapolate these findings across the continent.

It is generally accepted that grains are edible seeds from plants. True grains are those from species belonging to the Poaceae family, and include wheat, rice and oats, whereas pseudograins are seeds from different plant families that are nutritionally and functionally similar to true grains. For brevity, this review will mainly focus on the edible seeds of grasses. However, mention of the edible seeds of non-grass species that grow in grasslands, such as Portulaca oleracea (purslane) and small Acacia spp., is unavoidable as the seeds were gathered, processed and consumed in similar ways and for similar purposes. For information on the domestication potential of native herbaceous legumes, see Bell et al. (2011, 2012) and Ryan et al. (2011). Several of the Acacia spp., which are mostly shrubs and trees, have undergone a degree of domestication both in Australia and globally; see Adiamo et al. (2020) for recent review. Australian wild rice species (Oryza spp.), although being members of Poaceae and having similar end-product potential, require flooded growing conditions and, thus, are only briefly mentioned in this review; see Wurm and Bellairs (2018) for a research and commercialisation update.

We stress that the focus of this review is on the role of grains in food, rather than their role in native ecosystems or Aboriginal culture. However, it must be noted that these roles are inseparable, and in practice, the restoration of native grasslands as a source of sustainably produced food must bring together interdisciplinary and Aboriginal cultural knowledge, so that the role of grains in food is intertwined with restoration of native ecosystems and Aboriginal culture in the current economic, environmental and cultural context.

Although seed consumption was not a universal habit of all Aboriginal communities (Chivers et al. 2015), it was widespread, particularly across the arid and semiarid regions. The earliest evidence for seed grinding activity is dated to 65 000 years ago by Clarkson et al. (2017), although this date has been disputed (Bowdler 2017). The discovery of surface starch residue on grindstones by Fullagar and Field (1997) at Cuddie Springs in northern New South Wales offers strong evidence for a seed-grinding economy at that location 30 000 years ago (although see Gillespie and Brook, 2006). The morphology and wear on these stones were also consistent with seed grinding (Gillespie and Brook 2006). Although the site does not have direct evidence of eating breads from these ground seeds, if it were true, it would be earlier than the current earliest evidence of flour baking in the Middle East, dated at 14 400 years ago (Arranz-Otaegui et al. 2018).

Norman Tindale first attempted to map the approximate geographical distribution where Aboriginal seed consumption was considered prevalent, as inferred by the presence of grinding stones, in the mid-20th Century (Tindale 1977). Several more recent studies, including the summary of grindstone studies mentioned in Chivers et al. (2015), plus archaeobotanical studies listed in Table 1 and Fig. 1, widen the geographical area with physical evidence of Aboriginal seed processing beyond Tindale’s Arc, which was originally a wide band across predominately the northern inland of the Australian continent. Protection of sites of Aboriginal and Torres Strait Islander cultural heritage, including grain grinding tools, is now legislated by state and territory governments in Australia. This has led to a rapid expansion of recorded evidence discovered throughout Australia, including locations and basic descriptions of grain-grinding tools. These data are preserved in searchable databases, e.g. the Aboriginal Heritage Information Management System in New South Wales (NSW), but access is restricted to maintain the safety of the sites and artefacts.

Table 1.  Summary of locations and species from key ethnobotanical studies, with source locations displayed in Fig. 1
Only the grassland species mentioned in the original studies are listed here; however, some sources provide more extensive lists of food plants including grasses, herbs, succulents, shrubs, trees, and tuberous plants

Fig. 1.  Location of key ethnobotanical studies, giving detailed evidence of Aboriginal grain-grinding species, mechanisms and/or culture listed in Table 1.

The map and list of nine significant archaeobotanical studies given in Fig. 1 and Table 1 are provided to illustrate the wide variety of native plant species that Aboriginal people used for their seeds and to locate specific studies relative to the maps of Tindale (1977) and Chivers et al. (2015). This list is far from exhaustive. For example, O’Connell et al. (1983) documented 36 plant species of grasses, trees, herbs and succulents from which the Alyawara people of Central Australia collected edible seeds. Latz (1995) stated that ~70 species were utilised across Central Australia for their seeds, with 15 of these being grasses. Of the species mentioned in various reports, two species in particular, namely Acacia aneura, a leguminous shrub native to arid areas, and the grass Panicum decompositum (Native millet), are frequently identified as dietary staples and common sources of seed in arid and semiarid regions.

The species that are common food sources across multiple areas have a combination of wide environmental adaptation, favourable seed size and yield, and seeds that are easy to process (see Section Grain threshing and cleaning). Seed preparation was considered time-consuming, arduous and the most strenuous aspect of traditional subsistence activities. For this reason, and just as on other continents, only a select few grass species have the characteristics required of human food staples (Harris and Hillman 2015), even though many more species are nutritious and edible. The relative importance of seeds in diets also varied considerably among sites (Gammage 2012). This was most heavily influenced by time of year and availability of alternative food sources that were easier to gather or process (Cane 1989), and seeds were probably most important during dry seasons (Edwards and O’Connell 1995). A photograph of a set of typical grain-grinding stones is given in Fig. 2.

Fig. 2.  Examples of grain-grinding stones from north-western NSW. (a) Base plates on display at the National Parks and Wildlife office in Narrabri. (b) Gomeroi woman Kerrie Saunders grinding Microlaena stipoides (weeping grass) on grindstones outside Boggabilla.

The written accounts from early European explorers provide a glimpse into Aboriginal food systems in the period soon after colonisation. However, it is important to consider that these accounts are often limited by a lack of technical understanding, or by bias against Aboriginal people. For example, botanists were not present for many of the earliest expeditions around Australia (Chivers et al. 2015). As a result, explorers’ records often lacked detailed descriptions of food plants, making it difficult to identify the species described in accounts of grain collection and consumption at the time when Aboriginal culture was least altered by European influence. Major Thomas Mitchell (1838), an eminent early explorer, described haystacks ‘extending for miles’ of one species of Panicum grass that was ‘full of seeds’ during an expedition to the Darling River region. Although Mitchell could not explain the purpose of this system, it is now understood that grass seed heads were cut while green, piled up in stacks and left for the grain to ripen, after which the seed that had fallen onto the ground would be collected, cleaned and cooked (Chivers et al. 2015). When Allen (1974) described the seed collection techniques of the Bagundji people of the Darling Basin, Panicum decompositum was mentioned as a staple in this area, and may be the species mentioned in Mitchell’s earlier description.

The value of native grasses was not lost on all colonisers. Fred Turner, a NSW Government botanist in the late 1800s, remarked that

‘It seems a most feasible thing that out of 360 species of grasses found on this continent, some could be cultivated that would yield good grain without its attendant drawbacks, in the way of parasitic fungi, especially in some parts of Australia where wheat and other cultivated cereals are often a precarious crop.’ Turner (1895, p. xvi).

This identification of possible commercial cultivation remained unexplored as conventional agricultural practices became dominant, including breeding wheat cultivars resistant to the fungal diseases mentioned by Turner and expansion of profitable sheep and cattle enterprises across the rangelands of the continent. The result was both replacement of rangelands in areas where soils and climate were suitable for cropping, and the loss of native species suitable for grain production as sheep and cattle grazed out the more palatable plants. Today, several types of native grassland and grassy woodland have been declared endangered ecological communities by the Australian Federal Government or respective state governments, e.g. natural grasslands on basalt and fine-textured alluvial plains of northern New South Wales and southern Queensland (Department of the Environment 2021).

Many of Australia’s native grass species have developed complicated genetic systems in response to surviving in an environment of climatic extremes. This has led to substantial genetic dissimilarity among populations (Whalley et al. 2013), and is of great importance for finding both appropriate species as well as genotypes or populations within those species that have suitable characteristics as a food in both modern markets and agronomic environments. Chivers et al. (2015) identified eight grass species with between three and nine reliable records of historic use by Aboriginal people for food, and a further 42 species with one or two mentions of use. A SWOT analysis was performed on these eight species and recommendations were made for various climatic zones, in addition to a brief review of the species for which some agronomic and/or genetic research has been performed (wild rice, wild sorghum and weeping grass/alpine rice). Of these, the primary species that have had domestication research thus far are Microlaena stipoides and Oryza rufipogon (Davies et al. 2005; Shapter and Chivers 2015; Wurm and Bellairs 2018). This is primarily due to their yield potential per hectare using mechanised (as opposed to human labour and fire) field management tools, and research has focussed on improving their suitability to modern agricultural techniques and demands of markets for large seed size and desirable functional properties in food.

There is very high variability in seed yield and seed size characteristics for wild-collected accessions. Variation in amplified fragment length polymorphism (AFLP) markers ranged from 47% to 65% in a collection of Microlaena stipoides from south-eastern Australia, illustrating a degree of outcrossing (Mitchell et al. 2014). High genetic variability in wild-collected relatives of domesticated food crops such as wheat is also common (Haudry et al. 2007). A 20-fold range in seed yield and five-fold range in seed weight was found among 46 accessions of Microlaena stipoides from Western Australia and NSW (Davies et al. 2005), and even larger ranges of between 8 and 2200 kg ha⁻¹ have been reported for other species (Cole and Johnston 2006). This is significantly less than the yield of wheat, which averages ~10 t ha⁻¹ in the high-rainfall areas most suited to M. stipoides in NSW (Acuña et al. 2011). As for all crops, growing conditions will markedly influence seed yield; however, the optimal agronomic conditions for native grain crops, such as planting rate, row spacing, depth, fertiliser and irrigation vary greatly depending on genotype, species and available equipment (Cole and Johnston 2006). Some of the recommendations for native pasture establishment and management in Australia (e.g. Waters et al. 2001; Local Land Services 2015) and around the world (e.g. Pedrini et al. 2020) will apply to native Australian grain crops. Nevertheless, care should be taken when transferring these findings to growing native grasses for grain production, as the target for these studies was vigorous year-round biomass growth rather than seed production. Furthermore, these practices are far removed from the fire-based methods of grassland management used by Aboriginal people. The impact on seed yield of grass crops under various forms of fire management is a key area for further research.

Global food security in the 21st century is affected by steep growth of the global population, as well as reduced yield of staple crops owing to climate extremes. Although native grasses are likely to be more resilient to climate change and the resultant challenging agronomic conditions for growth of food crops, the yields of introduced crops under heat and drought stress are still likely to exceed those of native crops in the majority of current crop-producing regions (Thistlethwaite et al. 2020). Yet, native grasses may play a role outside current cropping areas, particularly when they can increase local food security for remote communities that struggle to obtain employment or reliable and affordable healthy food items (House of Representatives Standing Committee on Indigenous Affairs 2020). Furthermore, they can provide ecosystem services such as carbon sequestration, water supply and flow regulation, erosion control and pollination support, which support the production of introduced food crops (Bengtsson et al. 2019).

Considering the taxonomic and morphological similarities between native and domesticated millets (Panicum spp.), Allen (1974) questioned why Aboriginal people did not appear to improve food plants through selective propagation. Rainfall is highly unpredictable in arid zones, and the realisation of productivity improvements from breeding depends heavily on rainfall (O’Connell et al. 1983). Aboriginal people moved seasonally between camps, and the advantage of remaining geographically adaptable may have been greater than that gained from domestication of grain species in one location. Importantly, managing perennial grasses pose difficult logistical challenges compared to annual crops, which may have contributed to the lack of evidence for selection of grasses for various productivity traits. For instance, before a new (improved) batch of seeds can be sown, the previous plant and seed source needs to be largely removed, particularly if the traits of interest do not render the improved lines more competitive. Removal of perennial plants and seed banks, even with the use of fire, can take many seasons. In other parts of the world with access to beasts of burden and combined with metal farming technologies such as hoes, ploughs and seeders, plus the prevalence of annual species suitable for food production (Purugganan and Fuller 2009), the removal and resowing of improved lines was considerably more feasible.

There has been some debate regarding the use of the terms ‘farming’ or ‘domestication’ as applied to the management of native grains for food on the Australian continent within its international context. The definitions of these terms relate to the way the species and their growing environments were modified to suit human food needs, particularly when plant selection intentionally or unintentionally increased genetic traits favourable to human food production, such as large seed size and lack of shattering, rendering them uncompetitive in non-cropping environments (Purugganan and Fuller 2009). Some research classifies Aboriginal people as hunters, gatherers and fishers (Keen 2021), whereas others use the terms ‘farming’ and ‘agriculture’ to describe examples of intentional food production by Aboriginal people (Gammage 2012; Pascoe 2014). Despite this debate about the nature of food production practices before the arrival of Europeans in Australia, considerable potential exists for farming and domestication of these grains using modern practices in the future (notwithstanding social perspectives; see Section The importance of Aboriginal perspectives in research and development of native grains for the modern food industry).

In the past 50 years, some plant breeding and selection has occurred for several native grass species with traits of interest to the pasture industry in Australia. Named varieties of Astrebla spp., Themeda triandra and Microlaena stipoides and others, all targeting the pasture seed industry, are available nationally (Cole and Johnston 2006), in addition to other genotypes that are locally available. Priorities for plant breeding and selection include development of synchronous maturity of culms and resistance to shattering, i.e. the dropping of seeds from the head before harvest (Davies et al. 2005), and a better understanding of the natural reproduction methods of each species, including selfing, outcrossing and asexual reproduction (Mitchell et al. 2014). Targeted research to create lines of Microlaena stipoides using mutation which possess traits favourable to domestication has occurred at Southern Cross University in Australia, with multiple lines being identified as candidates for their yield and non-shattering seeds (Shapter et al. 2013). Increasing demand for mine site and other rehabilitation works across Australia have led to research focusing on seed production and establishment (Pedrini et al. 2020). Despite this recent work, the traits of interest to pastoralists and rehabilitators are not identical to those of the food industry, where high yield of large, easy to process seeds are the highest priority.

Modern-day grains have been selectively bred, among many other traits, to be easy to separate from the mother plant and their husk by mechanical methods. This process, termed ‘threshing’, occurs within the header of the mechanical harvester during harvest of most grain crops. Native grains (both historically and in modern times) are difficult to separate from trash of the mother plant as well as their own husk compared to modern grain crops (RIRDC 2014; Mildwaters and Clarkson 2020). It is believed that grassland seed collection and processing were both a time- and labour-intensive operation relative to the calorific value of the resulting food (Section Historic and modern considerations of energy balance in collection of and processing of grains). There is some evidence that particular grass species were targeted as a food source among all the available grasses according to whether the seed shattered or was easy to thresh (Latz 1995).

Although seed preparation practices differed among species and locations, some core processes and implements were shared. After gathering, stacking the plants and ripening of the seeds, it was commonly threshed by trampling in a hole, and some reports specifically mention a square hole (Allen 1974; Latz 1995; Parker 2003). Rocks were added to holes by skilled processors in Central Australia to facilitate husk separation (Latz 1995). To further remove the husks, seeds were either pounded with a log in a round hole or rubbed between the palms of the hands (Cane 1989; Parker 2003). Shaking the grain in long bark dishes sifted the dust and dirt to one end where it could be blown off (Cane 1989). Both winnowing (separation of fragments by density) and yandying (air-assisted removal of fine chaff) operations were essential for producing an edible product, and the skill of the operator was a significant contributor to its efficiency (Latz 1995).

Modern seed processing of edible grasses and pasture grass seeds (i.e. other than the major food grasses including wheat and rice), both in Australia and globally, requires technical knowledge and specialised equipment for success. This is due to the size and appendages on the seed and the production conditions (e.g. impure stands, seeds low to the ground), which make bulk-processing difficult, as has been the case for thousands of years. North America and Europe (particularly the USA) have well developed forage seed industries (Mordor Intelligence 2020) and greater availability to access grass seed processing technology and expertise. In Australia, challenges with processing seeds have been known for many years (Cole and Johnston 2006), and development of processes and technologies to assist seed processing have been identified multiple times as a key research priority in the local pasture seed industry (RIRDC 2014; Oliver et al. 2018), the emerging wild rice industry (Wurm and Bellairs 2018), and by the emerging native grains for food industry (Chris Andrews, pers. comm., Latarnie Mc Donald, pers. comm.). Although the energy ‘cost’ by either humans or machines is rarely considered a major factor in modern food processing, the cost of processing in economic terms certainly is a major factor in the balance between whether a certain species would be viable in modern markets (Sydney Institute of Agriculture 2020).

Analysis of archaeological evidence suggests seed grinding and flour baking has been occurring in Australia for thousands of years, despite the absence of clay or metal implements as was used on other continents. Dry grinding (for softer seeds) and pounding/smashing (for harder seeds) was used with various types of specialist stone tools (Pardoe 2015). The most common and widespread form of seed preparation was grinding seed with water on large flat stones (Fig. 2). The dough was either eaten raw or baked in the ashes of a fire (Allen 1974; O’Connell et al. 1983; Cane 1989).

In contrast, modern commercial grain milling is a highly mechanised multi-stage process, often involving removing most of the bran and germ layers of the grain. Removing these components extends the shelf life of flour and results in a consistent end product, while also altering its nutritional profile and functional properties (Doblado-Maldonado et al. 2012). Dry milling is more commonly used than wet milling because of the preservation of the storage life of the product. Wet milling followed by dehydration is usually used when a highly refined product is required, for example, maize flour or rice flour, and produces a flour of different nutritional and functional properties owing to alterations in protein, lipid, ash and carbohydrate contents (Leewatchararongjaroen and Anuntagool 2016). Differences between traditional Aboriginal grain grinding using stone implements (and often water) and modern industrialised processes will result in differences in the consistency and composition of flours made with native grains. Although modern methods are more efficient, the conversion of traditional to modernised versions of threshing and grinding will incur a loss of cultural practices which may affect Aboriginal communities. For example, women sang songs during traditional grinding and threshing (Curran et al. 2019). Hence, Aboriginal participatory development of processing and grinding technology is vital to successful implementation in communities where such cultural practices are considered important (Dreise and Mazurski 2018).

The association of seed grinding with arid zones is closely linked to relative food source availability. Developing technologies to process plant foods played a role in allowing the settlement in arid environments and were vital in supporting Aboriginal occupation of low rainfall environments (Latz 1995; Fullagar and Field 1997). Compared with tubers and fruits, seeds have a low energy return relative to time spent collecting and processing (O’Connell et al. 1983). There are observational reports in ethnobotanical literature describing the time required at different stages of seed harvesting and processing. O’Connell et al. (1983) described seeing two women pick and remove the husks from ~1 kg of Acacia aneura in less than 2 h, whereas Cane (1989) stated that about 1 kg of seeds could be collected in half an hour. Grinding 200 g of seeds required roughly 1 h, resulting in an energy return of ~350–700 kcal an hour for tree and grass seeds (Cane 1989; Edwards and O’Connell 1995). It would therefore take 7 h a day to process enough food to satisfy half the caloric intake of a family of five, assuming complete reliance on this resource, highlighting the labour-intensive preparation requirements of these seeds (Edwards and O’Connell 1995). In comparison, the mean energy return on edible roots was calculated to be ~4000 kcal an hour and almost 6000 kcal for fruit (O’Connell et al. 1983). This suggests that grass and tree seeds as food sources were selected on the basis of availability, rather than on capturing nutrients most efficiently (Edwards and O’Connell 1995). There is little consideration in current literature on the role that desirable storage capacity, functional properties, taste, aroma and digestibility of grains may have played in the selection of grains as a food source.

There are multiple reports of Aboriginal people using the behaviour of animals to facilitate the collection of seeds and reduce the energy expenditure for seed collection. Ants in Central Australia were known to collect certain grass seeds, and piles of these could be obtained around ants’ nests with ease (Latz 1995). Similarly, certain bird species cannot digest the hard-coated seed of the desert kurrajong (Brachychiton gregorii), and large numbers of seeds can be collected from one area, for example, near a waterhole, after being deposited in bird droppings (Latz 1995).

It is important to note that the low energy return is reflective of the energy expenditure required to process the seeds by hand using stone implements and does not mean that the seeds are nutritionally poor. For example, acacia seeds have been described as of ‘outstanding’ in nutritional content, and are higher in energy, protein and fat content than conventional cereal crops (Brand-Miller and Holt 1998; Adiamo et al. 2020).

Calculations of historic kilocalorie efficiency of seed processing are irrelevant when considering modern grinding on mechanised mills. Unlike threshing and winnowing, for which current technologies are inadequate, existing grinders and mills for globalised crop species can easily handle food-grade native grains, and differences in energy expenditure of grinding domesticated versus native grains are negligible when using electric or otherwise powered mills.

Unsurprisingly, words to describe flour-based bread cooked over hot coals occur in many Aboriginal languages (Table 2). Interestingly, most of these words can also be used to describe a variety of soft foods, with ingredients that can include grains, tubers and other plant-based products. Even though there is no direct evidence, this is suggestive of the generalised role that flour-based foods played in diets before the arrival of Europeans, and perhaps how many variations in ingredients and combinations were used to create foods with a similar dietary purpose, depending on the availability of ingredients by collection batch, location and season. Examples of this variation are described by Edwards and O’Connell (1995) and Gammage (2012). Thus, the English word ‘bread’ is generally a greatly simplified translation of the words for seed-based foods used by Aboriginal people, as the word tends to evoke a specific food product from modern cuisine.

Table 2.  Examples of terms from five languages that refer to the typical form of roasted dough made from ground native grains

Although the main known historical food use of native grains was in flour cakes made with diverse ingredients (e.g. Brand-Miller and Holt 1998), modern food processing equipment and techniques could widen the range of potential uses for these species. Native grains are gluten free, and thus they are unable to form the strong viscoelastic protein network which is required to support leavened bread loaves (Shewry 2019). Nevertheless, they have the potential to be used in blends with wheat flour (see examples in Fig. 3) or combined with ingredients commonly used for gluten-free flours, to make modern loaf breads. Other modern food products that do not require a viscoelastic protein network include muesli, multi-grain bread, bars, pre-cooked rice blends, puffed grain cakes and crackers. These products require processing techniques that are commonly applied to modern grains, such as puffing, rolling, parboiling, kibbling and blending with other grain species, and have the potential to produce high-value health-focussed products from native grains.

Fig. 3.  Food products made with blends of native grain flour and wheat. (a) Basic fire-roasted dough (flour cakes) with ~50% Panicum decompositum flour (left) and ~50% Dactyloctenium radulans flour (right). (b) Sweet sultana biscuits (33% Microlaena stipoides flour). (c) Flat crackers (33% roasted and ground seeds of Atriplex nummularia), presented with Portulaca oleracea tzatziki and Chenopodium nutans greens. (d) Pizza (base of 25% ground Atriplex nummularia seeds, paste a combination of ground Solanum centrale with tomato and garlic, toppings are field-foraged native Australian grassland leaves and fruits with cheese).

Research into the nutritional and functional properties of various non-grass seed sources such as Acacia spp., kurrajong seed, yams and purslane seed has been performed to a greater degree than that of grasses (Thorburn et al. 1987a, 1987b; Brand-Miller and Holt 1998; Liu et al. 2000; Shelat et al. 2019). In general, native grains possess higher proportions of desirable nutrients, such as proteins and minerals, than do domesticated crops, although some potential anti-nutritional factors have also been noted, particularly in Acacia spp. (Adiamo et al. 2020).

Among the native grasses, some functional and nutritional properties have been studied in a small number of species. For example, starch in Australian native rice species is characterised by high amylose content and gelatinisation temperatures, and pasting properties vary among different genotypes (Henry 2019). Native Sorghum spp. and other relatives of domestic grain crops are characterised by a particularly novel morphology in the outer layers of the grain, with a dense protein matrix and sparse starch granules (Shapter et al. 2008, 2009). Protein content, protein digestibility and starch granule size, shape and morphology were all found to vary relative to their domesticated crop relatives. Some of these novel traits may be of interest for introgression into cultivated crop species (Shapter et al. 2008, 2009) or may prove useful when the flour of the original native species is blended with introduced food crop flours in food products.

Post-colonisation, all Aboriginal groups experienced substantial change in subsistence patterns, leading to losses of first-hand knowledge (Clarke 2003). Traditional foods are linked to identity, culture and Country, and remain an integral part of contemporary Aboriginal diets, especially in remote areas (Ferguson et al. 2017). However, the foods that have remained in use have tended to be less labour-intensive than seeds (Clarke 2003), and as a result, seeds are no longer a common food source. Chivers et al. (2015) acknowledged the existence of information that still exists within Aboriginal communities, but is not present in written, verbal or visual records. Although it is unpublished, this Aboriginal knowledge is vital to understanding both the historic and future role of native grains as a food source and to navigating the different ways that food is produced, the land is managed and law around research, plants and food is applied in modern globalised agricultural and food systems.

The increasing interest in native grains in Australia is creating both challenges and opportunities for integration of the knowledge of Aboriginal people into modern research and development. Some of the practices common to commercial cropping that improve food production per hectare have no precedent on this continent until recently. This also means both the advantages and disadvantages of these technologies have been avoided. The case for participation in research by Aboriginal and Torres Strait Islander people in Australia is outlined by Dreise and Mazurski (2018, p. 9). They stated

‘First Nations affairs policy is complex, due to a multitude of factors including history, colonisation, ideology, politics, race relations, geography, and socioeconomic marginalisation. It is therefore reasonable to suggest that Western science and academic research are unlikely, by themselves, to provide a holistic picture or a complete understanding of this inherent complexity or of the pathways necessary to turn Aboriginal marginalisation around.’

It is important that local Aboriginal people are involved early during research design (well before implementation) so that the opportunities and challenges can be navigated in partnership with research institutions. One successful example where Aboriginal people have been involved from the inception of researching a native grain for commercial use is in the development of wild rice (Oryza spp.) in a partnership between Aboriginal enterprises and Charles Darwin University (Wurm and Bellairs 2018). Aboriginal enterprise Black Duck Foods also collaborates widely with multiple Australian research institutions in development of temperate native grains for industry.

Reservations have been raised by some Aboriginal people about how certain crop research and development methods take a plant out of its cultural context (Sydney Institute of Agriculture 2020) and/or are used without the permissions and obligations that come with holding knowledge under traditional law (Whitehead et al. 2006). However, viewpoints on the use of modern technology when applied to native foods vary greatly among individuals, communities and nations. Examples of such technologies that have increased the productivity of native grains but have created discussion about their application in ancient native grain systems include agronomic interventions such as use of high levels of fertilisation and irrigation on monoculture plantations (Cole and Johnston 2006), mutagenesis to create higher-yielding native grain cultivars (Shapter et al. 2013), and making collections of Australian native Sorghum spp. (Shapter et al. 2009) and Oryza spp. (Henry 2019) to cross with commercial crop species to improve their resilience, disease resistance, or other favourable traits for global food security.

In addition, legal issues of Indigenous intellectual property associated with plants globally have been known for a long time, Davis (1997) stated

‘There is a conceptual gap between existing intellectual property systems and the protection and recognition of Indigenous peoples’ rights to their cultural knowledge, products and expressions. Indigenous peoples consider their intellectual property rights are an integral component of a ‘holistic’ cultural heritage, which includes a wider range of subject matter than can be accommodated within existing intellectual property laws.’

Frameworks are being continually proposed and developed to navigate this issue in the Australian context (see e.g. Janke 2018 and Jefferson 2021). The successful development of commercialisation pathways for native grains that ensure protection of Aboriginal interests and access to the benefits of commercialisation will also have implications for the global discussion around successful application of the rights of Indigenous people, including the Nagoya Protocol (United Nations 2007).

For this and many other reasons, there are considerable complexities in development of business models that incorporate Aboriginal cultural knowledge and people in enterprises based on natural resources (Gorman et al. 2020).

With the exception of the acacias, Australian native grains have been studied as food almost exclusively through the lens of historic Aboriginal diets, with modern applications remaining poorly investigated. Native grains have previously been a source of food in areas well outside current agricultural zones, and have considerable potential to increase the area of land used in plant-based food production for a local, national and, potentially, international market. However, a sound understanding of the yield potential, management practices and biochemical properties of key species is required to objectively assess their suitability for commercial applications. Furthermore, urgent improvements in seed processing technologies suitable to Australian native grains (possibly based on technology for forage seeds both nationally and internationally) is required for supply chains to be efficient enough to supply the market. Combined efforts of policy makers, researchers and commercial entities are needed to re-integrate native grains into the Australian landscape, supply chains and human diets in a culturally sensitive way, where economic viability works alongside renewal of people and the environment. Involvement of, and partnership with, Aboriginal people in research and technology development will ensure that their cultural and social views are incorporated into the future native grains industry and will promote the participation of Aboriginal people in native grain enterprises across Australia. This will potentially create flow-on benefits to participation of Indigenous people in agricultural industries across the world.

The authors declare no conflicts of interest.

The writing of this review paper did not receive any specific funding.