Yuninypa tjuta mantjira, warungka pawulku. Palunyanta kapingka puli aata tjunku, yuninypa tjunku. Nyurrilku. Cleanamilalku wirranka, rungkalku, nyuma pawulku, ngalku. Yuwa, mirrka palya ngaatja.

After collecting the seeds, they are passed through flames. The seeds are then softened in water that has been heated with hot stones. After this, the seeds are cleaned in a wooden dish, ground on a millstone, cooked and eaten. Yes, this is really good food.

Yalti Napangarti

In the brevity typical of an Australian Western Desert language, the quote above from Yalti Napangarti articulates one of the most involved processing methods yet reported for an indigenous Australian plant food. The food she describes, an unleavened bread made from the milled seeds of the soft, or gummy, spinifex grass (Triodia pungens R. Br., Pintupi name: tjinalpa), lacks thorough documentation, which is surprising given that it is derived from a grass that thrives across vast regions of Australia. Spinifex (Triodia spp.) grasses are hummock-forming perennial bunchgrasses characterized by long, stiff, needle-like leaves and extensive root systems adapted to withstanding harsh environmental conditions (Jacobs 1984; Lazarides 1997). Triodia pungens is one of the most widespread of these grasses, dominating arid and semi-arid grasslands across inland regions of much of the northern half of Australia. The ecology of the species has been extensively researched (Gamage et al. 2014; Wright 2021), its material properties are well documented (Gamage et al. 2014; Memmott et al. 2009), and, from a pastoral perspective, it is probably the most economically important of the 86 currently described Triodia species (Barrett et al. 2023). Despite all the interest in the species, observations of Aboriginal people using Triodia pungens seed as food are limited, no descriptions exist of the production techniques involved in making the seed cakes, and the nutritional properties of the seeds are unknown.

This article provides a traditional (culturally-informed) description of the processing techniques for Triodia pungens seed cakes, as communicated by Aboriginal ladies from the Kukatja, Pintupi, and Mantjiltjarra language groups at Kiwirrkurra community in the Gibson Desert, Western Australia. Results of laboratory testing on the seeds are also provided, to illustrate the nutritional content of processed Triodia pungens seed flour. In our discussion, we explore the factors contributing to the scant documentation of Triodia pungens seed cakes and delve into the possible health and economic advantages of reintegrating traditional seed foods like tjinalpa into the diets of remote communities.

Methods

Study Area and Study Species

The Gibson Desert is the fifth largest desert in Australia. It covers an area of approximately 500,000 km2 and is characterized by extensive areas of gravel-covered gibber plains, sand plains, and dunefields. The climate of the desert corresponds to the “grassland hot (persistently dry)” climate type of the Köppen classification system. The study species, Triodia pungens, is a long-lived (more than 50 years) hummock-forming grass that grows to approximately 100 cm high. It occurs primarily on oligotrophic (nutrient deficient) substrates in sandplain and dunefield systems, but also it occasionally occurs on river frontages, red earths, rocky hills, and gravelly rises (Jacobs 1984; Wright 2021). Across its distribution, Triodia pungens is a keystone plant species, dominating the ground layer and playing a crucial role in maintaining the structure and function of the grassland and savannah ecosystems in which it occurs.

Seed Processing Workshop

In June 2020, a seed processing workshop was held at Kiwirrkurra community to demonstrate to students from the local Aboriginal primary school the traditional techniques involved in processing seed cakes (Pintupi term: nyuma) from native Australian desert plants. Although various edible native seeds were examined in the workshop, including those of Tecticornia verrucosa Paul G. Wilson (desert samphire, Pintupi name: mungilpa) and Eragrostis eriopoda Benth. (woolly butt grass, Pintupi name: wangunu), the focus of the workshop was on the seeds of the rarely used seed food plant, Triodia pungens. Among the community women that taught the seed processing techniques to students, two were Kukatja speakers (YN, YN) and two were Mantjiltjara speakers (NN, JN). These elders had been born in the bush during the 1950s–1970s and had lived traditional hunting and gathering lifestyles during the early parts of their lives. The dates of the women’s first contact with Europeans ranged from the mid-1960s (NN, JN) to 1984 (YN, YN). In addition to these interviews, a review of anthropological, explorer, and ethnobotanical literature on the use of Triodia seeds by Aboriginal people was conducted.

Laboratory Study

Seed samples of Triodia pungens were sent for analysis to Symbio laboratory, Brisbane, Queensland. At Symbio, standard seed testing protocols were used in accordance with analytical techniques outlined by NATA (National Association of Testing Authorities, (https://nata.com.au/)). Protein levels were analyzed with a LYCO machine and moisture was quantified via direct drying. Furnace-based reduction was used to measure ash, mineral contents were examined using inductively coupled plasma–atomic emission spectrometry, fat profiles were tested using gas chromatography, and high-performance anion-exchange chromatography was used to test carbohydrate levels.

Results and Discussion

Tjinalpa Processing Techniques

During the seed processing workshop, eight distinct stages were identified in the production of tjinalpa seed cakes (ESM 1). The first stage involved the gathering of the seed itself. Methods of spinifex seed collection mentioned by Aboriginal people at Kiwirrkurra included collection from ant nests and direct collection from the inflorescences of the plants. Direct collection involved grasping bunches of inflorescences in the hands and rubbing them vigorously to release seeds into a collecting receptacle. Collection of seed from ants’ nests simply involved robbing the nests of harvester ants and was mentioned as the easiest and quickest way to gather large quantities of Triodia pungens seed. Harvesting of seed from ant nests by humans has been documented for a number of other Australian bush food plant species (e.g., Yakirra australiensis (Domin) Lazarides & R.D.Webster, Poaceae, Fimbristylis oxystachya F.Muell., Cyperaceae, and certain species of Eucalytpus, Myrtaceae) (Latz 2018; Nangala et al. 2019). It has also been documented for several desert staple grasses in Saharan and sub-Saharan Africa (Harlan 1989).

The second stage in the Triodia pungens seed cake production process was passing the seed-filled florets through the flames of a fast-burning, low-intensity fire. This parching process aimed to render the floral bracts (paleas and lemmas) surrounding the grass seeds brittle, facilitating their removal. The fires for this stage were fueled by the semi-woody basal sections of tillers of mature spinifex hummocks. Traditionally, these fires were set on top of hard flat termite mounds (i.e., termitaria, Pintupi term: lintji) or on claypans. Having a hard, clean base beneath the fires ensured that the florets were easy to collect once singed, and that minimal sand or grit entered the seed mix. The spinifex fuel was not allowed to burn down to ash but was removed immediately after the florets passed through the flames. This minimized the need for further cleaning work by reducing the amount of ash that contaminated the seed mix.

Next, the seed-filled florets were placed in a pile and pounded with a stone mortar (Pintupi term: tjungari). This was done to further rupture/damage the floral bracts that enclose the seeds, increasing the potential for seeds to be dislodged from the bracts when rubbed. Following pounding, the florets were rubbed by hand vigorously for several minutes to release the seeds. Two rubbing techniques were employed. The first involved clasping a handful of florets in the palm of one hand and using the thumb of the other hand to roughly abrade the florets. The second method involved taking a handful of florets in one hand and laying the back of the hand down on the ground and using the palm of the other hand to abrade the florets.

Winnowing, the separation of the grain from the chaff by wind, involved taking the seed mix and placing it in a wooden dish. The mix was then tossed into the air, allowing the wind to sweep away the lighter chaff, while the heavier seeds fall back into the dish. Following this step, a dish of water was heated by placing in it small stones, about the size of walnuts, that had been warmed in a fire. The cleaned seeds were then placed in the warmed water and left there for approximately 10 min to soften. As the water cooled, additional hot stones were added to maintain the high (but not quite boiling) temperature of the water. Once the seeds were deemed to have soaked and softened sufficiently, the water was poured out of the dish and the cleaned seeds were sprinkled with water. This moist mix was then ground to a pultaceous mass on a large millstone (Pintupi term—tjiwa) using the same upper mortar stone as had been used in the pounding stage.

Cooking the Triodia pungens seed cakes involved first preparing a bed of hot firm earth and coals for the cakes to be cooked on. The seed slurry was then dripped onto the bed and a flaming fire stick was suspended above the slurry for a couple of minutes so that a crust could develop. Having this crust layer would reduce impurities entering the cake by preventing the adhesion of ash and coals to the cake during cooking. Once a firm crust had developed, hot coals were raked over the top of the slurry using a short stick and the cake was left to bake for approximately 20 min. The bread was removed once it was deemed to be emitting a suitable aroma, reminiscent of roasted crackers, signalling the completion of the cooking process Fig. 1.

Fig. 1
figure 1

Triodia pungens habit and traditional uses. a Characteristic hummock growth form. b YN demonstrating placement of T. pungens seed slurry on hot ashes. c Using burning spinifex leaves to create “skin” on seed cake. d Cooked seed cake placed on tjiwa (grindstone) (photos BRW, Rachel Paltridge)

Full size image

Nutritional Analysis

Laboratory analyses indicated that in comparison with other common bush foods from the Gibson Desert, the seeds of tjinalpa are energy dense (1615 kJ/100 g), low in fat (5.6 g/100 g, of which less than 1 g was saturated, 3 g was monounsaturated, and 1.8 g was polyunsaturated), and extremely high in carbohydrate content (64.7 g/100 g) (Tables 1 and 2). Potassium levels (287 mg/100 g) were also high compared to the other bush foods, though calcium (134 mg/100 g), magnesium (151 mg/100 g), and protein (18 g/100 g) levels were similar compared to most of the other common bush seed foods (Tables 1 and 2). From the laboratory assay, the nutritional content of tjinalpa seeds was similar in terms of energy density and overall macronutrient content compared to commercially harvested grains such as wheat, corn, and rice (Paliwal et al. 2000).

Table 1 Results of nutritional analysis of Triodia pungens seeds. Nutritional information based on serving size of 100 g
Full size table
Table 2 Comparison of nutritional content (g/100 g) of Triodia pungens seeds with seeds of other species of traditional economic importance to Aboriginal people in the Gibson Desert. “Primary sources” column indicates the number of primary source documents that contained reference to the plants being used as a traditional seed food source (see reference list for ESM 2)
Full size table

Traditional Importance of Tjinalpa Seed Cakes as a Food Source

Documentation regarding the utilization of Triodia seed as food is notably scarce compared to many other indigenous Australian seed food species. In a literature search of 48 journal articles and books from primary sources (i.e., Aboriginal informants) that contained inventories of food plants of desert Aboriginal people that occupied Triodia-dominated habitats, only six mentioned the use of Triodia seed as food (ESM 2). Conversely, seeds of species of Acacia were mentioned as being used as food by 36 sources; of the grass genus Eragrostis by 21 sources; of the succulent Portulaca oleracea L. by 34 sources; of the grass genus Panicum by 25 sources; of Eucalyptus spp. by 21 sources; and of the grass species Dactyloctenium radulans (R.Br.) P.Beauv. by 17 sources (ESM 2).

Given that Triodia grasses are edible and dominate grassland and savannah systems across c. 30% of the Australian continent (Gamage et al. 2014) (ESM 3), it is surprising that these grasses have not been more widely documented as a food. A likely explanation for this is that spinifex seed cakes were infrequently consumed because their production is time-consuming and difficult (i.e., there are numerous additional steps required in processing that are not necessary for other seed foods, such as burning the seeds and soaking them in hot water) (Napangarti et al. 2023). However, it is also possible that limited availability of the seeds might have constrained their use as a food source during traditional times. Spinifex grasses exhibit mast seeding behavior, producing minimal or no seeds in most years and only yielding substantial seed crops following infrequent, exceptionally high-rainfall years, occurring perhaps once every 5 to 10 years (Jacobs 1984; Wright 2021; Wright et al. 2022). Since such high rainfall years   also stimulate the growth of other more easily collectible and processable desert fruits and seed staples (e.g., Solanum spp. fruits, Eragrostis eriopoda, Yakirra australiensis (Domin) Lazarides & R.D.Webster, and Fimbristylis oxystachya F.Muell. seeds), the intricate and labor-intensive processing required for spinifex seed cakes may have deterred their frequent production, even during times of abundant rainfall.

Aboriginal Seed Foods in Modern Diets

The collection of wild grass seeds as primary dietary staples was a widespread practice since ancient times in Europe, North America, Africa, and Australia (Łuczaj et al. 2012; Maurizio 1926). Nevertheless, the practise waned considerably across most of these regions during the nineteenth and twentieth centuries due to the advent of modern agricultural techniques, rendering the labor-intensive harvesting of wild grass species obsolete. In Australia, the settlement of tribal Aboriginal peoples from the Western Desert into European urban centers, pastoral stations, and missions during the early-to-late twentieth century led to a rapid replacement of indigenous seeds with easily obtainable refined European flours as the primary sources of dietary carbohydrates (Long 1989; O’dea 1984). This settlement of previously highly mobile hunting and gathering communities coincided with a rapid increase in the prevalence of “lifestyle” diseases, including type 2 diabetes, hypertension, and obesity. This decline in health can be attributed, in part, to reduced levels of physical activity resulting from a more sedentary lifestyle. However, another significant contributing factor was the swift transition from traditional diets, consisting primarily of wild-sourced animal and vegetable foods with high overall nutrient densities and low glycemic indices, to diets predominantly composed of store-purchased European foods with markedly lower overall nutrient densities and significantly higher glycemic indices (O’dea 1984).

The revitalization of the harvest of wild seed foods among desert populations in remote Aboriginal communities has the potential to yield numerous advantageous health and economic outcomes. Seed foods from plants such as Triodia pungens are, comparative to most store-bought foods, absent in trans-fats, rich in healthy omega-3 polyunsaturated fats, high in fiber, and rich in essential micronutrients (vitamins and minerals) (Brand-Miller and Holt 1998). Furthermore, most native seed foods have substantially lower glycemic indices in comparison to store-purchased grain-based foods such as white bread, cakes, and biscuits (O’dea 1984). These nutritional attributes inherent in native seed foods could prove beneficial for individuals with metabolic syndromes, particularly type 2 diabetes, where insulin resistance is a concern. Given the considerable energy expenditure involved in gathering and processing these foods, regular engagement in such activities would naturally integrate high levels of physical activity into daily routines. This, in turn, could have positive health implications for individuals with lifestyle diseases if seed food gathering and consumption were regular practices.

Revitalizing traditional seed-gathering practices has the potential to induce positive psychosocial changes in remote communities, particularly among younger generations. The act of gathering plant foods is inherently social and has the capacity to impart local knowledge and essential bush skills to the youth, potentially fostering a heightened appreciation for the traditions of their elders. The resurgence of seed-gathering cultures in these communities may also yield economic advantages, if the communities were to engage in tourism initiatives centered around teaching seed-gathering and processing skills. Such endeavors are presently under consideration in certain Pintupi-Luritja communities to the east of Kiwirrkurra.

Conclusions

Triodia pungens seed cakes exhibit remarkable nutritional value, characterized by high carbohydrate and micronutrient content alongside low-fat content. Nevertheless, despite their nutritional excellence, the scant literature available on their consumption suggests a likely limited significance in traditional Aboriginal food economies. This limitation is attributable to: (1) the intricate processing techniques of Triodia seed cakes incurring substantial energetic costs on their production; and (2) the irregular mast seed production of Triodia plants, which would have hindered the regular collection of seed quantities necessary for seed cake production. The comprehensive documentation of Triodia pungens seed cake processing techniques presented in this paper contributes essential insights to Australian Indigenous Ecological and Cultural Knowledge. With few remaining Aboriginal individuals who practiced traditional hunter-gatherer lifestyles, it is imperative to gather ethnographic details promptly to preserve knowledge concerning lesser-known seed foods such as tjinalpa.