The fermented oil fruit of southwestern Australia

Aim: The aim of our reconstructive anthropological experiments was to gain an understanding of why Noongar people, unlike other Aboriginal groups in Australia, processed and consumed only the sarcotesta (outer fleshy layer) of Macrozamia discarding the carbohydrate-rich seed (endosperm).  It is our assumption that the sarcotesta was processed for a number of reasons, most importantly to improve its palatability, digestibility, increase its nutritional value and to facilitate its easy removal from the seed. We would also suggest that the anaerobic pit process further provided a form of storage to delay spoilage and to prolong the consumption of this valuable oil-rich seasonal food resource. 

In this paper we have tried to reconstruct the traditional methods used to process the outer fruity rind (sarcotesta) of Macrozamia seed. To achieve this we have relied on ethnohistorical accounts, contemporary Noongar views and some anthropological imagination. Never before, to our knowledge, have such experiments been conducted to assess the food value of processed Macrozamia sarcotesta.  As research anthropologists it has been a mystery to us as to why the nutritional value and composition of this unique ancient food has never been evaluated by food scientists, anthropologists and archaeologists alike. 

It is well established that the kernel (or endosperm) of Macrozamia species found in southwestern Australia is highly toxic and contains harmful glycosides including macrozamin that was first isolated from the kernel of Macrozamia riedlei by Lythgoe and Riggs in 1949 but Noongar people traditionally consumed only the processed fleshy outer layer or sarcotesta. And yet there has been no scientific evidence that conclusively establishes whether this thin fruity fleshy layer is toxic or not.  Why such definitive research has not been carried out in the name of public safety by our local universities and government authorities is unfathomable in this day and age, given the presence of these plants in our remnant urban bushland areas and many council landscaped gardens. When we asked veterinary specialists about instances of Macrozamia poisoning in dogs, they said it was a problem but they could not be certain which part of the ingested seed or seed coat was responsible for what is known as cycad toxicosis.

Until it has been scientifically verified that the seed coat is non-toxic, we would caution against anyone consuming Macrozamia sarcotesta – ripe or unripe, processed or unprocessed – for safety reasons. 

Methods and Material

Seeds from the large cones of Macrozamia fraseri were used for our experiments. This cycad species, according to Barrett and Tay (2005:28) ‘occurs from Jurien Bay to Mandurah, and is common in the bushlands of Kings Park and Bold Park.’¹  Florabase WA describes Macrozamia fraseri as

‘tree or (cycad), trunk variable; dull strongly keeled leaves with narrow to medium leaflets; large, broad cones.’ 

According to the Cycad pages M. fraseri is distinguished by its ‘robust and often arborescent habit, keeled leaves, densely woolly crowns and large cones.’ Macro derives from macros, Greek, meaning large or long (see Sharr 1996: 159). The seed cones can be large and long, sometimes containing up to 160 seeds, each seed measuring about 3-5 cm long and 2-3 cm wide, often ovoid or egg-shaped. Of the three species of Macrozamia found in southwestern Australia, M. fraseri was the first to be named.  It was named in 1842 by the Dutch botanist Miquel to honour Charles Fraser (1788 -1831) the colonial botanist of New South Wales who accompanied Captain James Stirling in his expedition up the Swan River in 1827 to observe the soil, botany and geology for an assessment of the region’s suitability for a colonial settlement. ²

Since colonial times the Macrozamia plant of southwestern Australia has commonly been known as the “zamia palm” –  although it is not a true palm but a cycad. Cycads are seed plants (cycad meaning ‘palm like’).³  They develop seed cones (or strobili) not flowers. It is interesting to note that Macrozamia was the first toxic plant to be recorded for Western Australia.

In replicating indigenous Macrozamia sarcotesta processing methods, traditional implements such as the wanna (digging stick) and dtappa (stone knife) were not used. A spade was employed for digging the pits, a pocket knife for separating the firmly adhering sarcotesta from the seed and plastic buckets were used for collecting and soaking the seeds.  A sandy patch in our back garden in Perth was the location for our reconstructive processing experiments which were carried out over three successive seasons – March 2008, March 2009 and March 2010.  For the “soaking” experiments saline water was sourced from the Avon River in 2008 and from the Swan River in 2009 and 2010 in an attempt to replicate the salty or brackish water traditionally used. At the time of traditional Macrozamia seed processing (late summer/ autumn), fresh water was a scarce resource in southwestern Australia. Most dried-out river pools were brackish or saline, often containing high levels of microorganisms.

In 2008 and 2010 we collected ripe seeds from a private property in Toodyay and in 2009 the seeds for our experiments were sourced from Bold Park bushland by staff from the Botanical Garden Parks Authority (BGPA) in our presence to ensure that only “ripe” fruit with their sarcotesta intact were collected. The results of our 2009 processing experiments were later analysed by the Chem Centre WA.

In all our experiments we collected seeds of Macrozamia fraseri when they were ripe and possessed a strong characteristic odour. The seed had a bright red or orange-red thin fleshy outer covering which enclosed a stony shell (sclerotesta) enveloping the kernel.  Throughout this paper we cannot emphasise enough the importance of ripeness in the collection and processing of Macrozamia sarcotesta. 

The accuracy of our anthropologically reconstructed experiments relied heavily on 19th century documented accounts. Often these were vague, some possibly deriving from second hand sources (colonial hearsay or newspaper references), although we acknowledge that without these descriptions our investigative research would not have been possible. 

Indigenous food processing has a long tradition of trial-and-error and has undergone thousands of years of testing and fine-tuning, adjusting to localised climatic and environmental conditions in southwestern Australia.  Noongar women would have been skilled technologists in the processing and preparation of Macrozamia sarcotesta.  

Preliminary experimentation carried out in March 2008

A pilot study was carried out in March 2008 during the Macrozamia fruiting season to test the viability of the three historically described methods of processing sarcotesta by the Noongar people of southwestern Australia.  These techniques were (i) water-soaking, (ii) burying (iii) soaking and then burying the seeds. Moore (1835 in Schoobert 2005: 424), who provides one of the earliest descriptions of baio processing, describes it as having been ‘steeped so long in water, or buried in the earth, as to arrive at a state approaching decay.’ He describes the flavour of the sarcotesta as ‘something like that of medlar, or the taste of old cheese.’ In a later publication (1842) he describes soaking the nuts for only ‘a few days’ and then burying. This is consistent with Grey’s (1841) description of soaking for “several days” then burying for ‘about a fortnight.’  Drummond’s early work vaguely refers to the “boyas” or Zamia fruit being steeped either in water or buried in the earth ‘for some weeks.’

The fruit of the “Zamia” is variously recorded in the ethnohistorical literature as biyoo (Lyon 1833), baio (Moore 1835), bayoo (Bunbury 1836), by-yu (Grey 1840:22), by-yu (Symmons 1841), boyas (Drummond 1842), boyah (Edwards 1894), bay-i-o (Ward and Fountain 1907: 211) and boyoo (Hammond 1933). In an earlier publication Grey (1840:22) records “by-yu” as ‘the nut of the Zamia tree, when enveloped with pulp’ and in his Exploration Journal he writes:

‘this name is applied to the pulp of the nut of a species of palm.’ (Grey 1841:295)

Moore (1842) in his Descriptive Vocabulary changes his original spelling of baio to conform to Grey’s by-yu.  

In the southern region it was commonly known as quenine (also spelt quinine, kwinin, kween-een, gwineen and quinning).  Grey (1840: 17,72, 114) records gwineen at King George Sound as ‘the common stock of food.’ He is probably referring to its abundance when in season.  Writing much earlier at King George Sound, Captain Barker in 1830 records the name as quinine or quenine and describes it as being buried

‘in the earth for about a moon when it becomes fine eating’ (Barker 1830 in Mulvaney and Green 1992: 302). 

Barker derived this information from his indigenous informant Maragnan. Drummond (1862) also refers to burial for a month at King George Sound. Bunbury (1836) referring to the Vasse and Perth areas refers to bayoo (as it was known there) being buried for ‘a month.’ 

Selection of preparation method and timing would have varied depending on weather factors, availability of water, individual and group food and taste preferences, seasonal crop yield, the availability of other foods, variations in regional practice and other cultural factors not known to modern day researchers.

All three methods that we trialled proved viable and successful in rendering the sarcotesta (outer covering) soft, pliable, oily and easy to remove from the seed. Our preliminary study indicated that anaerobic pit processing may have involved dry fermentation and leaching, and also a means short-term fruit preservation and storage.  

Replicative experimentation in March 2009 to assess food value 

Macrozamia fraseri seeds collected from Bold Park Bushland

In early March 2009 we conducted further processing experiments, on the basis of the success of the above trials, using 40 Macrozamia fraseri seeds sourced from Bold Park bushland. These were collected by staff from the Botanic Gardens and Parks Authority (BGPA) under our advice and observation to ensure that only ‘physiologically ripe’ seeds were collected. Several days before harvest we had identified a couple of disintegrating female cones to ensure that only “ripe” fruit were targeted for collection. We observed some “fresh” seeds but as these did not appear ripe, that is, they lacked the distinctive all-pervading ripeness odour, we did not collect them.  Seeds that were collected displayed a bright red seed coat, possessed an odour and showed no obvious signs of animal predation or decay, although in some cases insect activity, native and non-native, was present (see Plates 7-9).

Variability in Macrozamia Seed Production

Seed and foliage samples were collected by BGPA personnel for two herbarium vouchers to be lodged in the BGPA and the State Herbarium, respectively. These were verified as Macrozamia fraseri (see Appendix). The seeds were collected in Bold Park (Reserve A45409) under conditions relating to our Bold Park Research Permit BPRP 09-002. At the time of collection we noticed (and so did the BGPA seed collectors) that there were many fewer Macrozamia strobili fruiting in the bushland than in the previous season. Whether this was the result of seasonal fluctuations in weather patterns, fire history or the opportunistic and sporadic seasonal variability of Macrozamia seed production is beyond the scope of our expertise to comment.

When we consulted a group of Noongar Elders from the Swan Valley Community they explained that

‘there are good years and bad years… when it’s a good year, the plant uses up all its energy so there are fewer the next.  It happens with other bush tucker too.’

The Elders stated that ‘the old people’ knew that burning the country regenerated the plants.⁴

‘The smoke and ash was like food for the country.’

Indigenous semi-cultivation

‘Firing country’ as a form of land and indigenous resource management is well-recognised in the archaeological literature (e.g. Beaton 1982; Hallam 1975; Smith 1982:117 and Gott 1982).  This practice was confirmed by Noongar spokespersons who stressed that anthropogenic firing regimes were once instrumental in the sustained production of Macrozamia fruit as well as other bush tuckers. It is probable that by this means Noongar people were able to regularise or synchronise the cycle of bayoo production, increasing its annual fruitage. We would suggest that this was a form of indigenous semi-cultivation.

Macrozamia fruit production is variable. Only a small subset of female plants usually produce seeds within a localised population during any one season. Ripening phases also vary from plant to plant and within the same plant the cones may exhibit differential maturation and dehiscence rates. Seed production is usually staggered, except in the event of fire-induced masting occurrences.  We are unclear about the frequency of traditional anthropogenic firing regimes on plant colonies but the literature suggests that a patchwork or mosaic pattern of ‘fire-stick farming’ was practised (e.g. Rhys-Jones 1978, Hallam 1975 and Gammage 2012) and this would have facilitated the indigenous management of Macrozamia fruit production, ensuring a predictable supply of this crop, subject to rainfall and other seasonal factors being favourable.

The Noongar were pyro-engineers, well acquainted with the effects of smoke, ash and fire on indigenous vegetation and its natural productions.

Seasonal fruit such as Macrozamia sarcotesta deteriorates soon after ripening from oxidative spoilage and predation from birds and animals. For this reason there must have been a sense of urgency to collect this valuable fat-rich food resource before it rotted away or was stripped from the seed by rival nonhuman sarcotesta -raiders.  We are convinced that the soaking of seeds in brackish water and pit burying were a practical and energy conserving means of stockpiling ripe fruit and a conscious strategy for extending its seasonal consumption and enhancing its nutritional value.

“Control” sample: ripe “unprocessed” sarcotesta delivered to the Chem Centre WA for analysis 

40 seed specimens were transported to our backyard in Perth where processing pits were being prepared. The sarcotesta of 9 ripe seeds was removed from the sclerotesta using a sharp pocket knife as the thin seed cover was firmly attached and extremely difficult to remove from the seed by hand. The pared strips of sarcotesta were placed into a plastic container, chilled overnight and then couriered to the Chem Centre WA the following day for analysis. This untreated sarcotesta was to be used as our “control” sample.

Reconstructive Processing Method 1 –  Soaking and burying 

Bunbury (1836), Grey (1841), Moore (1842), Drummond (1862) and others refer to the method of soaking and burying the Macrozamia fruit. As noted by Grey (1841, Vol 2: 296):

The native women collect the nuts from the palms [Macrozamia] in the month of March, and having placed them in some shallow pool of water, they leave them to soak for several days.  When they have ascertained that the by-yu has been immersed in water for a sufficient time, they dig, in a dry sandy place, holes which they call mor-dak; these holes are about the depth that a person’s arms can reach, and one foot in diameter; they line them with rushes, and fill them up with the nuts, over which they sprinkle a little sand, and then cover the holes nicely over with the tops of the grass-tree…’

Stage 1: Soaking

12 ripe Macrozamia seeds were placed in a plastic bucket and covered by 3 litres of brackish water sourced from the Swan River.  The idea was to replicate the brackish water that we assume was used at this time of the year (end of summer/ early autumn) to process the Macrozamia sarcotesta.  We left the brackish water to settle for a day, then added the seeds to the water and soaked them for six days. They were observed and monitored regularly to check daily water temperature variations (17-24 degrees C.) and any discernible changes to the sarcotesta.  The average daily shade temperature varied from 30-35 degrees C. After three days the water started to become cloudy and the sarcotesta covering the seeds showed signs of swelling.

After six days the water was opaque with an oily slick on the surface and visible evidence of an unidentified mould or fungal growth (see Plate 20). When the seeds were extracted from the water the sarcotesta showed no visible signs of any fungal growth. The outer covering was firm, moist and had a glossy orange-red colour.

Stage 2: Burying – reconstructing Macrozamia sarcotesta earth-processing pits

The 12 seeds were removed from the water medium after 6 days and then buried in a dry sandy anaerobic processing pit approximately 30 cm radius x 50 cm deep for two weeks. This pit was lined with Xanthorrhoea fronds somewhat resembling a bird’s nest (see Plate 26). These fronds would have helped to waterproof the food resource, protect it from sand contamination and create an anaerobic compartment for fermentation.  This together with the water repellent sandy soils of the Swan Coastal Plains would have protected the fruit from external moisture and unseasonal rains. We surmise that depth of burial chamber (approximately an arm’s length, according to Grey 1840) was important for a number of reasons, including to create a relatively stable anaerobic environment and to protect against fruit-predation and spoilage by insects and burrowing animals.

After two weeks burial in a Xanthorrhoea-lined pit the seeds were extracted and cleaned by hand to remove any sand or debris (see Plates 29 & 30). The sarcotestas had an earthy smell that soon dissipated after cleaning. Grey (1840: 20) records the Noongar term for ‘rubbing’ or ‘cleaning by rubbing’ as bur-ang-yu-yur-ung-ween which he notes ‘is usually applied to rubbing By-yu nuts.’ After soaking, burying and cleaning, the fermented fruit were orange-reddish and shiny and the sarcotestas were swollen and partially exposed at the hilum (see below). The flesh was soft, spongy, rather oily and very easy to remove by hand from the sclerotesta using a thumb nail. The oil was a bright orange-red colour and had a mild, not unpleasant, odour. The processed fleshy sarcotesta was removed and frozen for later analysis by the Chem Centre WA (see Table 1).

Grey (1841: 296) states with respect to soaking “for several days” and burying for two weeks that:

‘in about a fortnight the pulp which encases the nut becomes quite dry, and it is then fit to eat, but if eaten before that it produces the effects already described.’

Grey’s (1841:296) view on the processed sarcotesta pulp having to be ‘quite dry’ in order to be ready for consumption totally contradicts the views of Bunbury and Drummond, and also Moore who writes that the Macrozamia fruit must be

‘steeped so long in water, or buried in the earth, as to arrive at a state approaching decay’ (Moore 14th April 1835 in Schoobert 2005: 424).

We start to wonder what Grey was eating because all of our experiments showed that processed sarcotesta after soaking and/or burying was soft, pliable, pulpy and oily. Even Drummond describes the flavour of processed boya as tasting like ‘train oil.’  Grey’s (1841) theory leaves us with the question of how do you dry an oil seed in an anaerobic pit?

Reconstructive Processing Method 2 – Soaking using saline river or seawater (no burying)

An early reference to the soaking of Macrozamia or “bayoo nuts” is provided by Bunbury (1836) who states:

‘The quickest method of ripening the Bayoo nuts is to bury them in a hole of water at the edge of a swamp or river when they become fit to eat in a few days but in this way they acquire a very strong bad smell & unpleasant taste, so I recommend all who are willing to wait a month for such delicacies, to bury them in the dry ground rather than in water.’ (Bunbury 1836 in Cameron and Barnes 2014:136)

We are unsure whether Bunbury is referring here to human constructed soaking pools on the edge of rivers and swamps which at this time of year were mostly brackish or dried up. The few freshwater sources that remained in permanent pools would have been protected as a valuable resource for human usage rather than risking contamination through soaking toxic Macrozamia nuts. Both Edwards (1894: 233) and Hammond (1933: 28) describe salt water as being used in the preparation of boyah or boyoo.  Edwards describes how he observed boyah being soaked “in shallow brooks” and also sometimes in bags ‘suspended by a string attached to a stake on the sea beach.’

Hammond (1933: 28) also notes that

‘When the fruit was ripe it was picked and then submerged in salt water for a couple of weeks, after which it became soft and mealy. Only the fleshy part, which resembled a tomato in colour and taste, was eaten. The nut was not eaten.’

Using these scant descriptions provided by Bunbury (1836), Edwards (1894), Hammond (1933) and information gained from contemporary Noongar consultants, we attempted to replicate the method of Macrozamia sarcotesta processing using saline water.  Twelve Macrozamia fraseri seeds were placed in a plastic bucket to which 3 litres of saline water sourced from the Swan River was added.  The bucket was placed in a shady position for 14 days to replicate the ‘couple of weeks’ period suggested by Hammond (1933:28).  The seeds were observed and monitored regularly to check daily water temperature variations and any discernible changes to the sarcotesta.  The average daily shade temperature varied from 30-35 degrees C.

After seven days of soaking the seeds were still a red-orange colour, having lost only a minimal amount of their original colour intensity. The sarcotesta was still intact on all the seeds and showed no visible signs of decay. The salt water solution was cloudy with an oily slick on the surface and emitted a rank smell, possibly the result of micro-organism activity in the water.

After a further seven days of soaking the seeds were again examined. The sarcotestas had swollen while shrinking at the hilum end, exposing the area where the seeds were formerly joined to the cone. We assumed that this shrinking possibly indicated that the flesh was ready for eating. It had an over-powering stench that may have resembled what Bunbury (1836) described as a ‘very strong bad smell.’ This fetid smell was likely a by-product of the stagnant saline water and possibly a ferment of microbial activity.  Olfactory as well as gustatory perceptions are always subjective and culturally relative and what one person or group considers malodorous or bad tasting may be a mouth-watering signal to someone from another culture.

Was this ancient technique of soaking Macrozamia seeds in saline water analogous to such fermentation practices as soaking olives in brine?

Reconstructive Processing Method 3 – Burying for a “moon” (no soaking)

Barker (1830) provides the earliest account of Macrozamia sarcotesta processing. Based on information from his indigenous informant Maragnan, Barker describes quinine (or quenine) – as it was called in the southern region – as follows:

‘This is the great country for ‘Quinine’, the fruit of the low fan leaved palm which after gathering they bury in the earth for about a moon when it becomes fine eating.’  (Barker 1830 in Green and Mulvaney 1992:302). 

Bunbury (1836 in Cameron and Barnes 2013: 136), referring to the Swan Coastal Plain area, also states that ‘they must be left a month, i.e. from moon to moon.’  Drummond (1842, 1862) too points out that Macrozamia nuts were buried for about ‘a moon’ (approximately 28 days, a lunar cycle). Following these descriptions we buried 12 Macrozamia seeds in a processing pit and left them undisturbed for 28 days. We dug a hole in the same manner as already described for Experiment 002 (Burying before soaking).  When the seeds were unearthed after a month, they were bright red in colour and the outer skin which showed signs of shrinkage at the hilum had increased in thickness and become more fleshy or voluminous.  The sarcotesta was not dry but soft and oily and on the verge of decomposition.  It had a mild Macrozamia odour but not overpowering.  The processed fleshy covering was easy to remove by hand without the need for a sharp instrument. It was removed from the seed, frozen and at a later date couriered to the Chem Centre for analysis. Like all other samples, it contained a characteristic bright orange paste, suggestive of a high beta-carotene content

in March 2010 we were curious to test whether extra time within a controlled dry anaerobic environment would affect the condition of the sarcotesta and its storage life. Ripe seeds of Macrozamia fraseri were sourced from a private property in Toodyay and they were buried for a six week period without soaking. This was two weeks longer than the ethnohistorically prescribed “one moon” or four week period by Barker (1830), Bunbury (1836)  and Drummond (1862).  After six weeks the sarcotesta showed signs of shrivelling and decay (see Plate 34) and at this time we assumed that the consumption of the decaying and somewhat rancid looking sarcotesta may have been deleterious to health.  However, a number of recorders including James Backhouse have noted that the Noongar

‘buried or macerated the nut until the rind became half-decomposed in which state they ate the rind rejecting the kernel.’ (Backhouse 1843: 542) 

Drummond also emphasises that only the fermented red rind or covering was consumed by the Noongar:

“The nuts, when deprived of their red covering, are not used by the natives as food.” (Drummond May 1842). 

Fermented sarcotesta was a seasonally valued fat source possibly consumed at different stages of processing/ fermentation, depending on individual or group taste preference and circumstance. We wonder, could the sarcotesta in this semi-decayed condition have once been a Noongar gourmet delicacy equivalent to our highly esteemed Stilton or even a ripe camembert?

Variability in burying regimes

There is no fixed or agreed upon time for the processing of Macrozamia sarcotesta by means of soaking and/or burying in early ethnohistorical accounts. Grey (1840, 1841) refers to burying the nuts for ‘about  a fortnight’ (after soaking) to render the pulp ‘quite dry’ whereas others, such as Barker (1830), Bunbury (1836) and Drummond (1862), refer to burying (without soaking) the nuts for a month or a “moon.”  Grey’s two week period possibly takes into consideration that the sarcotesta was already soft and partially water-fermented and had it been left buried in the earth for longer than two weeks, it would have decayed and led to fruit spoilage.  If soaking was absent, then a four week period of pit processing may have been optimal, although the precise timing regimes would have varied depending on region, climate, soil type and cultural considerations.

We would suggest that a part of the surplus resource would have been buried successively to accommodate the variable fruit ripening phases from plant to plant and from grove to grove, enabling some processing pits to be ready before others. We do not understand the specifics of these ancient burial regimes. However, we do believe that the variability in production and  staggered ripening times may have been advantageous as it would have extended the season of consumption. This may well explain Moore’s (1842: 24) comment that the by-yu ‘is in full season in the month of May.’ He is referring to the indigenous consumption of this highly coveted fat-rich food.

Grey (1840:16) refers to April and May as ‘the season for eating by-yu’ or “By-yu ngannoween.” Moore records the name of the season corresponding to April and May as geran (nowadays called “jeran”). This translates as “fat” (jerang, jerrung, cherung) and corresponds to the time of the year around autumn when it was mandatory for Noongar people to build up their reserves of sub-cutaneous body fat to ensure their survival through the long cold dark wet lean season of makuru (known as mokkar or makur in the Albany region).  A wide range of fat-rich foods were consumed at this time including bardi, kuya (frog), yakkan (turtle), kalda (mullet) and salmon (http://anthropologyfromtheshed.com/project/the-ancient-practice-of-macrozamia-pit-processing-in-southwestern-australia/)

Our replicative experimentation of Macrozamia pit processing and storage would suggest that the duration of the burial method was limited by the onset of heavy autumn rains, although we wonder whether the use of Xanthorrhoea fronds to line and cover the processing chambers may have served as a form of waterproofing thatch to prevent fruit spoilage.  Salvado (1851 in Stormon 1977: 161), unlike the other recorders, refers to burying the whole Macrozamia seed cone or what he calls the “flower” in the ground:

‘The shell is red, and of fine texture, with no pulp. In order to make them fit for eating, the natives bury the flower [cone] together with the nuts for a certain time a couple of feet deep in the ground.  The heat of the earth makes them swell as if to germinate a new plant, and they are then cooked on hot coals to form a substantial food with a pleasant taste.’

Salvado (1851 in Storman 1977) is the only recorder to describe this method of interring the whole cone in the processing pit.  The female cone may contain up to 150 or more large seeds. As we did not try this experiment, we cannot assess the method but we believe it is highly feasible that such a method of burying the whole seed cone may have been used by Noongar people living in the kwongan or sand plain country to the north of Perth, the district to which Salvado was referring. The large pineapple-like cones of the female plant may weigh up to 10 kilos or more. According to the early botanist Von Huegel, who observed “Zamia” fruit at the Swan River area in the early 1830’s, each fruit “spike” is described as weighing up to 18-22 kilograms:

‘I was astounded by the huge fruit of the Zamia.  They grow fairly close together here and several had more than one spike of fruit each weighing between 40 and 50 pounds’ (Von Huegel 1833 in Clark 1994: 28)

Some early colonial accounts refer to the dense stands of Macrozamia vegetation that once grew in the Swan River colony, often in clusters or colonies, especially after fire. It is highly probable that traditional indigenous firing regimes contributed to the observed large size and weight of these culturally valued fruit.

Chem Centre sarcotesta analysis results – 1938 & 1939

As early as 1938 when the Chemical Branch of the WA Mines Department tested the seeds and sarcotesta of Macrozamia fraseri to determine their toxicity, they found that only the seeds were toxic.

‘Parts of the fruit of the zamia plant (Macrozamia Fraseri, Miq.), common locally, are poisonous to cattle, producing gastro-intestinal irritation. The fruit consists of a hard seed or nut containing a starchy endosperm surrounded, when ripe and fresh, by a thin fleshy layer of mesocarp with an orange-red epicarp. Only the seeds were found to be poisonous – a toxic principle not precipitated by lead acetate, and not extracted by immiscible solvents, being present. The fruit pulp surrounding the seeds contained 14 per cent of a bright orange-coloured oil, which appears to contain a considerable amount of carotene and closely resembles palm oil in its physical and chemical constants. Further work is in progress.’ (Annual Report of the Chemical Branch, Mines Department, for 1938). 

The above noted similarity between the carotene-rich oil of Macrozamia and that of palm oil is interesting. According to Clegg (1973):

“Palm oil is a fruit‐coat fat that is low in sterols and rich in vitamins A and E. Up to 50% of its fatty acids are unsaturated, and linoleic acid constitutes up to 11% of the total acids. Its composition makes it an edible oil of nutritional importance…’

Tests on the nutritional composition of the seed coat of Macrozamia fraseri conducted in 1939 by the Chemical Branch of the Mines Department revealed:

‘The fleshy layer (a) contained: water, 6.5; oil, 28.2; protein, 8.0; fibre, 8.9; ash, 2.6; nitrogen-free extract (by diff.), 45.8 per cent. The orange-red oil extracted with petroleum spirit from the fleshy layer had the following constants: Saponification value, 213.0; acid value, 7-2; iodine value (Wijs), 61.1; n/d/25 1.4630; n/d/40 1.4573; n/d/60 1.4500; it contained a considerable amount of carotene. The 45.8 per cent of nitrogen-free extract consisted of: Starch, 5.08 ; sugars, 3.00; acid (as malic acid), 1.34 ; mucilage, pentosans, hemicellulose, etc. 36.33 per cent.’  (Annual Report of the Chemical Branch, Mines Department, for 1939).

These nutritional analyses reveal high levels of oil or fat (14% and 28.2% respectively) in untreated Macrozamia sarcotesta.

Chem Centre sarcotesta analysis results – (Macintyre & Dobson experiments 2009)

With the exception of our own work in 2009, all chemical analyses of the nutrient content of Macrozamia sarcotesta (e.g. WA Chemical Branch 1938-1939, Thieret 1958 and Ladd et al 1993) have used “untreated” sarcotesta.  It was for this reason to gain an understanding of the nutritional composition of “treated” Macrozamia seed coat that we trialled different historically-recorded indigenous processing methods. After each experiment the processed sarcotesta was removed from the hard sclerotesta, placed in a plastic container, frozen and at a later date transported to the ChemCentre WA where a basic analysis was carried out to determine levels of sarcotesta fat (oil), ash, moisture, NDF (non-digestible carbohydrate, fibre), NFE (digestible carbohydrate, free sugars), protein and kilojoule energy content, pre- and post- processing (See Table 1). The untreated sample served as the “control.”

Table 1: Treated and untreated sarcotesta analysis on a dry basis (%db)

Fat

All three processing techniques increased the oil-rich content of the Macrozamia sarcotesta rendering it equal to an oil seed.⁵

Lipid levels’ significantly increased by 12.5% on a dry basis (Chem Centre WA 2009).   

A bright orange-red oil was extracted from our samples. This was not tested although the deep orange-red colour of the lipid extract suggested a high level of carotene. Similarly, in 1939 the Chemical Branch of the Mines Department had identified the presence of ‘a considerable amount of carotene’ in the outer fleshy layer.  Could these lipid-soluble carotenoids have traditionally provided Noongar people with an essential of source of Vitamin A and possibly other vitamins with strong anti-oxidant properties?  It is well known that carotene is the precursor of Vitamin A which is required for healthy skin, vision, eye health and immune function. 

Could these semi-decayed fruits have once been a Noongar gourmet delicacy equivalent to an expensive Stilton or ripe camembert?

NFE (digestible carbohydrates)

There was a noticeable decrease (%dry basis) in free sugars (NFE) after processing, possibly as a result of fermentation. When we applied a crude iodine test to treated and untreated sarcotesta samples, these both showed the presence of starch (see Plates 37 & 38).

NDF (non-digestible carbohydrates)

There was a slight increase in NDF (nitrogen free extract) as a result of processing, most noticeably in the ‘buried only’ specimens with a slight loss in the ‘soaked only’ specimens.

Moisture

Ken Dods (food scientist, Chem Centre WA, 2010) who facilitated all the chemical analyses of our Macrozamia sarcotesta samples stated that:

‘Moisture levels indicate this is not a drying process. However, increases in ash levels initially after soaking and before burying indicate an uptake of soluble salts which may provide beneficial preservation through anti-microbial activity and lowering of AW – free moisture levels, particularly during the first phases of burial duration. (Chem Centre WA 2010)

There appears to be a leaching process enacted by the burying possibly an interaction between the soil environment and the food. Both available sugars and soluble salts have reduced during the burying process. It is difficult to determine if the loss of available sugars (NFE) is through leaching or a fermentation process.’ (Chem Centre WA 2010)

Protein

There was a slight increase in protein levels as a result of processing, especially after ‘soaking and burying.’

Energy

‘There was not much change in the total energy availability, a slight increase 5% in metabolisable energy for the soaked/ buried sample’ (Chem Centre WA). 

The conscious processing of Macrozamia sarcotesta was an important evolutionary development in the nutritional diet of Noongar people in that it significantly increased the calorific value of the food.  In hunter-gatherer cultures energy rich foods, especially fats and sugars, were highly valued and much sought after.  The cost-benefit of Macrozamia sarcotesta processing involved a low to moderate labour input with a value-added high energy output food product.

Discussion

Our experiments demonstrated that changes to nutrient levels in Macrozamia sarcotesta as a result of processing involved elements of fermentation and leaching, either alone or in combination.  Drummond (1839) describes the processing of “boyas” as ‘a sort of fermentation.’

‘Fermentation is one of the oldest forms of food preservation technologies in the world…..The first fermented foods consumed probably were fermented fruits.’  (Battcock and Azam-Ali 1998, FAO).

Fermentation is the “slow decomposition process of organic substances induced by micro-organisms, or by complex nitrogenous substances (enzymes) of plant or animal origin.” (Walker, 1988).  It can be described as a biochemical change, which is brought about by the anaerobic or partially anaerobic oxidation of carbohydrates by either micro-organisms or enzymes. (Battcock and Azam-Ali 1998, FAO).  

The soaking of seeds in saline water may have had a beneficial effect on preserving the sarcotesta by slowing down microbial activity, especially during the first stage of anaerobic burial.  The lowering of free sugars (NFE) as a result of processing may suggest fermentation and the burying in dry sandy soil may have provided a leaching medium causing the NFE and ash (soluble salts) to reduce. It is likely that elements of fermentation and leaching were occurring at different stages of sarcotesta processing.

It was not a drying process

Our results do not support Grey’s (1840, 1841) view that the Macrozamia sarcotesta is ready for eating when it is ‘quite dry.’ Even we were convinced by his assertion when carrying out our first processing experiments and we were surprised to find that the processed sarcotesta was not “dry” but rather soft, pulpy and oily similar to the state of near-decomposition described by Moore (1835), Bunbury (1836) and Drummond (1839, 1862).  We then realised how easy it was to be taken in by an explorer’s account without questioning it, and we were left wondering how many other early colonial ethnographic descriptions and assumptions involving Noongar culture, such as food processing and consumption, have never been questioned, tested or debated,  With the exception of Ethel Hassell, most 19th century recorders were male and they would not have been privy to Aboriginal women’s domestic economic activities, including the often lengthy processing and preparation of foods such as by-yu (Macrozamia), yanjet (bulrush), twotta (root bark) and warrain (yam).  Also problems in cross-cultural communication and the lack of eyewitness accounts often resulted in misunderstandings of the local language and culture.  Many of the early descriptions and interpretations of indigenous culture derived from secondary sources, including early explorer’s journals and newspaper accounts, sometimes from the other colonies, intermixed with colonial hearsay.

Preserving and storing sarcotesta fruit

A further experiment was conducted to test the shelf life and storability of processed sarcotesta.  After the fruit had been soaked and buried (in the soaking and burying experiment 0002), the sarcotesta was removed from the seed by hand and threaded ‘kebab style’ onto a strand of prepared native flax (Dianella revoluta) in a similar fashion to that described by Ethel Hassell (1975: 24).  The specimen was stored in a sealed dry container for 12 months after which time it was delivered to the Chem Centre WA for analysis. It showed no signs of decay or drying.

The chemical analysis of this specimen showed a decrease in free sugars (NFE 14.8 -8.1% ar), possibly suggesting continued fermentation after processing. This may explain the significant increase in protein from 2.5 to 6.5 (% ar).

‘Increase in protein level is significant 15% on a dry basis.’ (Chem Centre WA 2010)

This experiment convinced us that it would have been possible and highly probable that sarcotesta was once stored. For how long, it is impossible to say. The stringing and dry storage of the processed and de-seeded Macrozamia sarcotesta would have been advantageous in transportation and trading of this item. According to the Chem Centre WA analysis, storage for 12 months in such dry conditions did not significantly affect ash, fat or moisture content but significantly increased the protein value of the food.

The question of toxicity of Macrozamia sarcotesta 

Throughout our practical experiments we were continually mindful of a feeling of uncertainty surrounding the question of Macrozamia sarcotesta toxicity. We, like everyone else, were taken in by the colonial assumptions of Grey and others that the red fruit coat – the only part eaten – was toxic and that the indigenous rationale behind processing was principally for detoxification. This was further underscored by the results of a one-off study by Ladd et al (1993) which reported, as we have already discussed in a previous work, that the sarcotesta of M. riedlei seed contained ten times more toxin (macrozamin) than the kernel. See http://anthropologyfromtheshed.com/project/the-ancient-practice-of-macrozamia-pit-processing-in-southwestern-australia/ Ladd et al’s (1003) study contradicts the 1938 findings of the Government Chemical laboratory which stated that only the seeds were poisonous:

‘The fruit consists of a hard seed or nut containing a starchy endosperm surrounded, when ripe and fresh, by a thin fleshy layer of mesocarp with an orange-red epicarp. Only the seeds were found to be poisonous…’ (Annual Report of the Chemical Branch of the Mines Department for 1938). 

Unsure what to believe, but concerned about Ladd’s (1993) toxic findings, we took extra precautions at all times when handling Macrozamia sarcotesta and also when sampling the final processed products. Although careful, we continually asked ourselves why would the Noongar people consume such a highly toxic substance and then discard the less toxic nutritious starch-rich kernel? Also, why would a plant poison its animal and bird vectors with a toxic sarcotesta at the critical time of  ripening and seed dispersal?  It didn’t make sense to us and it still doesn’t.

Burbidge and Whelan (1982: 66) from animal studies and personal communication with J.R. Cannon subscribe to the idea that the sarcotesta of Macrozamia riedlei is non-toxic. They note that:

‘In Macrozamia seeds, the poison is confined to inside the stony layer (J. R. Cannon, pers. comm.) and would not affect animals which swallow seeds whole, and fail to break them up in the gut.’ 

‘The edible outer integument of seeds, combined with the toxic inner integument, attracts the attention only of animals able to strip off the outer layer or ingest the seeds whole.’   

A compelling study by Hall and Walter (2014) found that the sarcotesta of Macrozamia miquelii from Eastern Australia was non-toxic.⁶

“There is Toxic Cycasin in the Seeds of Cycads, but Not Their Sarcotesta “Fruit” ‘(Hall and Walter 2014: 862)

Hall and Walter (2014: 860) propose that Macrozamia sarcotesta was probably an ancient adaptation that served as ‘a reward for cycad seed dispersal fauna.’

Asmussen (2008: 94) cites a number of researchers who refer to Macrozamia sarcotesta as providing a “nutritionally valuable starchy food reward” for seed dispersers. She states:

Raw Sarcotesta of Macrozamia fraseri fed to rats showed no adverse effects

From observations in bushland, as well as in our back yard at the time of our experiments, we have seen evidence of non-native animals (rats or mice) having eaten quantities of ripe Macrozamia sarcotesta seemingly without harmful effects. This prompted us to conduct our own animal experiments in 2011 using two white male rats.  Initially we tried feeding them a pure mashed Macrozamia sarcotesta diet (100%). However, they were reluctant to consume this food, so we blended the raw sarcotesta (60%) with mashed banana (40%) and fed them this mixture (Plate 40) twice a day over a 10 day period.  They readily accepted this food. We cannot be sure whether their initial reluctance was due to the distinctive smell or taste of the raw sarcotesta or because it was outside the range of their normal diet of fruit, vegetables and rat pellets.

The rats’ health was monitored every day and they showed no obvious adverse effects in faecal matter, fluid consumption or behaviour as a result of ingesting the untreated Macrozamia sarcotesta over a ten day period.  If anything, they seemed to thrive on it.  However, the results of this simple rat experiment cannot be extrapolated to human ingestion of Macrozamia sarcotesta. Until a definitive scientific analysis is carried out to determine whether fully ripened Macrozamia sarcotesta is toxic or not, we would strongly recommend against human ingestion, especially in its raw state.

Our research suggests that bayoo (by-yu, baio, boyoo, bi-yoo) was an essential pre-winter food that would have added an important high energy source of lipid-rich nutrient to the traditional Noongar diet. In our study we have explored the historically-recorded methods of indigenous processing of Macrozamia sarcotesta which included soaking, burying or soaking and burying. All of these methods appear to have produced a palatable and culturally valued end product which was considered a delicacy, either cooked or raw (after processing).

Sarcotesta processing increases the fat content of bayoo

Much of the current knowledge relating to the processing of this traditional food derives from the observations and assumptions of early colonial recorders, especially the influential accounts of Grey (1840) and Moore (1842) who both posited that Macrozamia sarcotesta processing was carried out for the purpose of detoxification. However, it is our view that the primary function of processing the “fruit” or sarcotesta was to enhance its nutritional content, improve its taste, digestibility and to facilitate its easy removal from the seed by hand. Pit processing also enabled the short term storage of this highly perishable food and prolonged its seasonal consumption. 

Until further scientific research is conducted, it is difficult to ascertain the exact chemical processes that were taking place. The soaking of Macrozamia sarcotesta in saline water may have involved elements of leaching, salting and possibly fermentation. Additional earth pit processing of the fruit in a dry anaerobic environment may have involved further elements of fermentation and leaching. Our findings did not support Grey’s (1840, 1841) description of pit processing as a drying treatment. In all our experiments the processed sarcotesta was soft, moist, pliable, oily and almost approaching a state of decay. These results were consistent with Moore’s earlier descriptions and also Bunbury (1836), Backhouse (1843) and Drummond (1862) who describe the end product as being consumed when approaching a state of decomposition. Individual or group taste preference may have determined at which stage of fermentation the food was consumed. 

The results of our reconstructive experiments demonstrated that Macrozamia sarcotesta processing significantly increased the lipid (fat) content of this highly valued seasonal food. To our knowledge the Noongar are the only documented group in the world known to have processed cycad sarcotesta using anaerobic earth pits. We have no doubt that their processing of Macrozamia sarcotesta was a conscious strategy to preserve and maximise the fat quality of this traditional delicacy.

We caution readers not to experiment with consuming any part of the Macrozamia seed or sarcotesta for safety purposes.  The seed is well known to contain toxins while the toxicity of the sarcotesta when fully ripe is yet to be definitively assessed.