Iga—the tree that walked

The distribution of the native orange (Capparis mitchellii) in the Northern Flinders Ranges of South Australia is limited generally to a narrow corridor running from Moolawatana Station in the north, to Baratta Springs in the south, on the eastern side of the Ranges. The association of this favoured ‘bush tucker’ species with Aboriginal occupation sites within that area, and an Adnyamathanha narrative tradition which attributes the origin of the tree in this area to direct human intervention, suggests that the species was introduced to this area from north-eastern Australia by Aboriginal visitors in the pre-European period.


Introduction
The native orange, also occasionally called native pomegranate, (Capparis mitchellii), while widely distributed throughout inland and eastern Australia, is generally restricted in the Northern Flinders Ranges of South Australia, to a narrow corridor running north-south from Moolawatana Station in the north to Baratta Springs in the south. Its distribution is confined almost entirely to the eastern side of the Ranges 1 with one isolated specimen reported on John Creek between Parachilna and Copley and another near Iron Knob on Eyre Peninsula, as shown in the online Atlas of Living Australia (2013). In the latter case, the specimen is outside of the traditional territory of the Ranges' Aboriginal (Adnyamathanha) groups. Iron Knob and northern Eyre Peninsula was, however, within the recorded area of travel and cultural association by the Adnyamathanha and their immediate neighbours 2 .
Specimens of the species are relatively common north-east of the Ranges, in association with drainage systems such as Cooper Creek or Strzelecki Creek. A number of trees are recorded in the Atlas of Living Australia within the Innamincka Reserve. The most easterly occurrence of the species I have recorded in the Flinders Ranges area is on Munyalinna Creek on the eastern side of Wooltana Station where two individual specimens are located at about longitude 139° 45'E. According to Adnyamathanha tradition, the species' most westerly appearance in the Ranges is said to be a single specimen in a rocky ridge about two kilometres east of Angepena homestead (at roughly 138° 53'E).
The Adnyamathanha (Adnya-'stone', matha -'mob', plus-nha suffix: 'Hills mob') native title holders and occupants of this region acknowledge the presumed unique location of the specimen near Angepena homestead which they call Iga Vurtu (or

Figure 1. Map of the Flinders Ranges showing named locations (drawn by Jeff Ellis)
Iya Vurtu -'shady Iga') and the ridge in which it is located, Iga Vurtu Marti (see Tunbridge 1988 pp. 47-51).
Within this narrow band of distribution, concentrated on the eastern Ranges foothills and in the gorges associated with the perennial watercourses that flow east into Lake Frome (and less commonly, into Lake Callabonna), the larger stands of the trees are directly associated with historic and pre-historic Aboriginal camping locations. These camp locations are almost exclusively located within a former Yardliyawara matrilineal totemic estate, associated with ngarrandula-mukunha (frog totem) 3 . That estate (yarta) extended south from Wooltana Station to about Martins Well Station and east from the Ranges foothills to the Olary Uplands. The Uplands formed a watershed, dividing the creek flow into Lake Frome on the west and on the east toward the Darling River system. The watershed was apparently also an important cultural boundary, delineating the traditional division of circumcising groups on the west from the non-circumcising groups on the east 4 .
The Yardliyawara, together with other discrete ranges groups, including the Walypi ('Wailpi'), Adnya-Kuyani, Adnya-Barngarla and Pirlatapa later consolidated to become the contemporary Adnyamathanha society. The consolidation followed pastoral occupation of the Ranges, restrictions on their enjoyment of previously available natural resources and a decline in their populations. The amalgamation was deliberate and considered, and was a result of a formal, traditional meeting of Adlya Wirri (Wilyaru initiated men) representing the separate language groups, which was held at Hideaway Well near Blinman in about 1920.
The name Adnyamathanha (or variants of this, such as Mardala which has the same meaning in the Diyari (Dieri) language), had been applied by their neighbours to the several Ranges groups prior to European settlement. That name however, was more particularly applied to the two totemic groups living on Mt Serle and Angepena Stations in the post-European period up to 1931, when they moved to Nepabunna (Nipa-phana Awi-'steeply bedded rocks waterhole'), joining with others who had resided elsewhere in the Ranges before that time. In the period which followed, Nepabunna was operated by the United Aboriginal Mission and functioned as a safe refuge for families and children during the absence of other family members engaged in pastoral work.
Today most of the Adnyamathanha live in major population centres outside of the Ranges, reflecting the declining employment opportunities in the region and the withdrawal of health and government services from the Ranges proper. While Yura Ngawarla (the Adnyamathanha language) is still spoken by older members of the group, the language and many of the traditions are currently under threat, despite efforts by the community to maintain these features of their society.
The historic camping locations on the eastern side of the Ranges, within the former Yardliyawara territory, are normally associated with earlier existing permanent, or semi-permanent springs, soaks or waterholes, many of which have now been replaced by pastoral bores, leading to the failure of those natural water sources.
The camp sites appear to have supported large local populations of up to 40 people for extensive periods of time in the post-European period. Other camping locations on the Frome Plains, which do not host specimens of the tree, were known to have been temporarily occupied by Yardliyawara groups at earlier times. These locations appear to have been occupied in smaller numbers or only opportunistically following rainfall events 5 . In the larger gorges which contain creeks that debouche into Lake Frome, such as Weetootla Gorge (Wirti Urdla Inbiri-'Leafy branch' Gorge) and Italowie Gorge (Ithala Awi Inbiri-'Cave-Water Gorge') the former Aboriginal habitation sites occupy an area of up to half a hectare, with 10-30 mature specimens of the species as markers of the camp location. In these locations the trees tend to stand as an exclusive grouping, separate from other species. They also tend to have little or no undergrowth, perhaps as a result of selective clearing or the previous intensive occupation of the tree stands by Aboriginal residents.
An extensive former occupation site with mature specimens of the species is located just north of the Balcanoona Ranger Headquarters on the Arkaroola Road near Echo Bend (Ngawarla Wami-'Speech Bend') at the entrance to the track to Weetootla Gorge. That gorge itself hosts a number of the trees, estimated to number 36, in six or seven discrete stands, including one stand following a creek at the base of a rocky hill. Italowie Gorge similarly hosts numerous specimens in separate stands. In the narrower gorges which did not host extensive camp locations, such as the western entrance to Mt Chambers Gorge (Marla Wadinha Inbiri-'Boomerang-Broken Gorge') and the eastern end of Waukawoodna Gorge (Wakarla Udnanha Inbiri-'Crow-Excreta Gorge') only single specimens (with limited occupational evidence) are discovered.
The tree and its fruit, which is known as Iga or occasionally Iya 6 in Yura Ngawarla was, and is, a favoured bush tucker source. The trees vary in height from five to eight metres. Their foliage is grey-green to dark green with leaves oval-shaped about 2-6 cm long and 1-3 cm wide. They produce showy, white to creamy-yellow flowers about 5 cm in diameter, usually in late summer in this region, with caper-like, long, projected stamens about 4 cm long. They attract the Caper White Butterfly, the larvae of which may, on occasions, denude the tree of leaves. The fruits that follow flowering are round and green when immature and hang from the branches on long arching peduncles about 6-7 cm long. The round, orange-coloured, mature fruits are about 5 cm in diameter with soft, orange-yellow flesh when ripe. The fruits contain many hard seeds much like the cultivated pomegranate. They are often subject to insect attack leading to subsequent deterioration or consumption of the flesh.
The trees are very slow growing. I have a specimen growing on my property in the Adelaide Hills which is now eight years old. It appears as an untidy, sprawling and spiny bush about 40 cm in height and diameter. They do not abide poor drainageothers I have planted have died as a result of this. Those growing in the Flinders Ranges are single or multi-trunked with dark, deeply cracked bark and occasionally thorny branches, although as they grow older they appear to lose most of their spines. Attempts to propagate the seeds suggest that there is a roughly equal likelihood of the seeds producing multi-trunked forms as well as single trunks (John McGovern 2006, pers. comm., 21 February). Re-growth of the trees in the ranges is not readily apparent with few small trees, perhaps as a result of the impact of goats and stock animals which graze the immature bushes and trees. When young the trees are often found growing in the shelter of Acacia species (John McGovern 2006, pers. comm., 21 February). The association of the species in this area, almost exclusively with prehistoric habitation sites (which were also occupied in historic time), strongly suggests that the distribution of Iga in this area is the result of human intervention, possibly as a consequence of humans consuming the fruit and its many seeds and later defecating in the vicinity of their occupation sites 7 . However, while this may account for the occurrence of the trees in isolated rocky or secluded areas such as that hosting Iga Vurtu, or that at John Creek, or their wider distribution within or between the camping locations, it does not explain the restricted occurrence of the species, almost exclusively located in this part of the Ranges, largely isolated from the other areas where it has been recorded.
The first European discovery and printed description of the species was provided by the English explorer Major Thomas Livingstone Mitchell, after whom the species was named. In his Three expeditions into the interior of eastern Australia: with descriptions of the recently explored region of Australia Felix, and of the present colony of New South Wales (1838) he noted, on page 284: On the bank of the river, at this place, we found beside the native fires, the remains of a fruit different from any I had seen before. It seemed to be of a round shape, with a rind like an orange, and the inside, which appeared to have been eaten, resembled a pomegranate.
Later on pages 310-11 he wrote: Aug. 17…Today I fell in with a tree, of which I saw but a single specimen during my former journey, and had observed only a sickly one during this expedition. It bore a yellow flower, and the fruit resembling a young pomegranate, on a hooked stalk. I had unfortunately omitted to gather specimens of it, when seen by me in flower, in 1831, and now I could not procure any of the seeds, every rind being hollow, and the interior destroyed apparently by insects. I consider this a very remarkable tree, as well as from its rare occurrence, as on account of its fruit, of which the natives appear to make some use. Aug. 18-I was more successful in my search, this morning, for seeds of the fruit above-mentioned, and I was surprised to find many specimens of the tree in the scrub, through which we had previously passed without observing them. On one plant, we found some fruit apparently full grown, but not ripe. And on others perfect specimens of the last year's crop, including, of course, the seeds. The fruit resembles a small lemon but had within, small nuts or stones, enveloped in a soft pulp, and the whole has an agreeable perfume. We also found some specimens of the flower, rather faded*. * My friend Dr Lindley considers this one of the most interesting plants brought home by me… The Adnyamathanha give considerable social importance to economically significant food trees. Their society was organized into two moieties: Ararru and Mathari. All members of the group belong to one or other of the moieties and even today this association is acknowledged. This affiliation they inherit from their mother and they must find marriage partners from among the opposite moiety members if they wish to marry within the group. Each moiety has a totemic affiliation with one of two food producing trees-Urti (Santalum acuminatum) 'quondong' (Mathari) or Minara (Alectryon oleifolius) 'bullock bush' (Ararru). Each plant has special traditions relating to its treatment and natural development. In contrast, Iga has no moiety affiliation, suggesting that while it is an important food species like the others, it stands separately from them, possibly as a consequence of its later introduction into this region.
The Adnyamathanha and their traditions provide an explanation for the tree's limited distribution in their country and perhaps for its lack of social standing. That explanation is contained in a Mura ('traditional narrative' or 'history story') which in its various versions relates a 'Johnny Apple-seed'-type explanation for them. In her publication Flinders Ranges Dreaming (1988 pp. 47-51), Dorothy Tunbridge reproduces three narratives which provide variations of the story of the Iga trees. Essentially the Mura stories describe the trees as Yura ('male humans'-today used exclusively for an Aboriginal man), beginning firstly as a single man, with a woman later joining him. According to that tradition they originated in Queensland and travelled via Nappa Merrie Station to the eastern Flinders Ranges, apparently intent on visiting Italowie Gorge which had special cultural importance to the Adnyamathanha and their neighbours.
According to Dr Tunbridge (and my own record), the trees were originally from Eromanga and travelled via the Wilson River to Cooper Creek between Nappa Merrie and Durham Downs. From here they were 'hunted away' by other human 'tree species', including Coolibah, to Mudluwartana (Moolawatana) from where they travelled south to Wooltana. Wherever they camped, Iga trees grew up, providing shade and fruit. The Iga attempted to travel west from Italowie Gorge 8 but were turned back by 'mulga trees' who were also Yura and who followed them to Mulga View Station, which is believed to be the limit of the range of that Acacia species in this area. They then travelled to Erudina and to Martins Well Stations where they stopped once again leaving Iga each time at those locations. They were again hunted by other trees who stopped them from travelling west. Eventually they travelled to Baratta Springs (probably Vara arta-'forked Xanthorrhoea bush') where trees of a different species again turned them east to Glen Warwick Station. Here the Iga trees settled briefly until they went north, finally returning to their own country.
This narrative and the named locations it records are consistent with the distribution of the species in this area. The temporary occupation of locations east of the Ranges by an Aboriginal group visiting from the north-east is also consistent with ethnographic evidence concerning movements of individuals and local groups into the eastern ranges from New South Wales and Queensland. Such movements occurred both in pre-European time and later. John McEntee (1991) has, for example, recorded visits to the Ranges in historic time by a group identified as 'Vawundi' (who were probably Wilyakali) from Poolamacca Station in New South Wales.
This relatively open and hospitable social environment, within the traditional estate of a Yardliyawara sub-group may be partly explained by the fact that all groups of this region, including those in Western New South Wales and south-western Queensland, shared the same matrilineal, exogamous moieties (although with some variation in names) and a common matrilineal totemic system. While the groups to the east did not, unlike the Yardliyawara, practice circumcision, they shared similar laws and customs which facilitated a degree of accommodation of visiting neighbours. The Yardliyawara in particular had close links with the groups to the east and like all groups of this region they were multi-lingual, speaking many of the languages of their eastern neighbours.
The suggestion that the visiting groups were prevented from residing further west, beyond the eastern side of the ranges and Yardliyawara territory by 'tree species', emblematic of other local Adnyamathanha groups of the central ranges, is also consistent with Adnyamathanha tradition. Their traditions generally claimed their hegemony and resistance to intrusion, and affirmed their exclusive occupation rights in the main Ranges areas. In the area of the eastern Ranges, groups from the east were known to have traditionally visited for the collection of quartzite slabs for use as grindstones and to obtain red ochre, occasionally bringing with them new ritual and ceremonial performances and items of exchange. One notable exchange commodity from the north-east was Pituri (Duboisia hopwoodii) which was highly valued and traded great distances from south-eastern Queensland.
A Barkandji traditional narrative recorded by Dr Jeremy Beckett and Dr Luise Hercus (2009) concerning the travels of two Ngatji ('rainbow serpents') describes their route from the Paroo River region of New South Wales (and that of Aboriginal visitors from this region), to the Flinders Ranges via Lake Callabonna.
Visits to this area by groups from Queensland, travelling down the Cooper Creek, were recorded in the post-European period. These visits were apparently concerned only with cultural revival or innovation and not with relocation of individuals or groups. In about 1898, for example, a travelling ritual performance described by Walter Roth (1897) as appearing in the Diamantina area in 1896 was conducted at Petermorra Creek where the Strzelecki Track today crosses that creek. This ceremony, called by Roth the 'Molonga' and known by the Adnyamathanha as Mudlunga, attracted over two hundred local participants from the ranges groups and their neighbours, some from as far away as Eyre Peninsula. This and an earlier gathering in Diyari (Dieri) country had resulted in expressions of concern from German missionaries who believed that the gatherings were a nativistic uprising planned to cleanse the area of European occupiers of the former traditional lands (Siebert, 1910). In the description provided to me of the Mudlunga (Gordon Coulthard 2009 pers. comm., 22 August) the main participants were a man and woman, who, like the two individuals in the Iga Mura had travelled to the area from Queensland. It seems likely however, that in both cases the visitors were accompanied by others. It seems that prior to the Mudlunga which was performed on Petermorra Creek (Murnpeowie Station) and which was hosted by the Adnyamathanha, a number of Adnyamathanha had attended an earlier performance brought to the area by the Malyangapa (a New South Wales group who traditionally occupied an area around Tibooburra) to Terapinna Waterhole, at the foot of Mt Babbage (Gordon Coulthard 2009 pers. comm., Field Notebook 28 August) and/or at Tjinda Tjindana Bank (Mick Coulthard 2009 pers. comm., Field Notebook, 3 March). In the circumstances it would appear that relations between the Adnyamathanha and the groups east and north-east of this area were cordial, cooperative and probably of long standing. Other groups are also known to have visited the eastern ranges from northern New South Wales. The routes of their travel west to the ranges appear to have been based upon the main creeks (Yandama and Boolkaree) which flow west into Lake Callabonna and Coonee, Billeroo and Eurinilla Creek which generally flow into Lake Frome. I was told by Mark McKenzie (2008 Field Notebook 2 July) that a group known as Vinhatu-mathanha ('Coffin Springs mob'), lived on Martins Well Station, up to the late 1920s. It seems that Coffin Springs and Bamboo Springs, also located on Martins Well station, were two of the locations of the camps of the group. Unlike the Adnyamathanha they were a non-circumcising society, like others from east of the ranges. A map (Figure 4) annotated by Dr Norman Tindale in the period 1920-30 (South Australian Museum, AA338/15/2, section of ) has an arrow drawn to 'Emu Springs' on Martins Well station with a note in Dr Tindale's hand: 'Artesian Well (Natural). Large encampment some 1 mile NE of springs'. That artesian spring, (known as Nilpininha ('Eagle's feather') to the Adnyamathanha), like 'Little Coffin Spring' which is a few miles south-west of Dr Tindale's location, has been filled in and replaced by a pastoral bore and is now known as 'Emu Bore'. I have been unable to discover the eventual fate of the 'Coffin Springs mob'.
The locations occupied by these and other visitors to the area, at least post-European pastoral expansion, are generally on the Frome Plains outside of the Ranges proper. None of these camping locations I have visited have Iga trees in association.
Gordon Coulthard (2009 pers. comm., 14 February) told me that he thought the group on Martins Well Station had moved west to this location from Billeroo Waterhole ('north side'). Mick Coulthard (2009 pers. comm., 3 March) told me that the waterhole is on 'Kunambi Creek', probably Coonee Creek, although modern maps show it on Billeroo Creek on Lakeside Pastoral Lease (Lake Frome Station). In earliest time, Gordon Coulthard told me, Billeroo waterhole was a gathering place for groups travelling through this area to and from New South Wales, including Yardliyawara, Pirlatapa, Walypi and Malyangapa/Wadikali (the latter two now known collectively as Barkandji).
Billeroo waterhole is probably on present-day Frome Downs or Quinyambie Station. I have not had opportunity to visit the location and consequently, do not know if Iga trees are present at this locality, although my inquiries suggest they are not. Gordon Coulthard told me that the Yura Mityi (Adnyamathanha name) of the waterhole is Kwinambi or Kwimbalanha 9 which he thought was the origin of the name Quinyambie Station. The absence of Iga trees at camps associated with visiting groups from northern New South Wales, at least in post-European time, would appear to suggest that the original introduction of the species to this area occurred some time earlier, and to give weight to the observation in the Adnyamathanha Mura that they originated in Queensland and were brought down the Cooper Creek.
It further suggests that their introduction was the result of a deliberate act, or at least a cultural preoccupation, which led to fruit being collected and transported for future use, rather than the result of prior consumption of seeds by visitors to the region whose traditional country may have hosted the species, such as in northern New South Wales. The movement of displaced groups from New South Wales into this area during the post-European period resulted in their camping outside of the main Ranges and on the eastern plains. While they did occupy traditional camp locations in that area, they did not occupy the main Adnyamathanha/Yardliyawara camps on the eastern side of the ranges which are distinguished by the presence of Iga trees.
At about the same time as the Vinathu-Mathanha were recorded on Martins Well Station another group, known to Mark McKenzie (2009 pers. comm., 30 January) as Ngudluwara ('Otherside (as in, 'eastern') mob'-probably Yardliyawara) or 'Wudliwara' (Mountford, 1940), was apparently living at Moorowie Well 10 (Wertaloona Station), although they also had less favoured camps on Bendieuta and Balcoracana Creeks. The main camp at Moorowie Well has several large Iga trees and is today the location of the grave of the former 'headman' of this group (Geraldine Anderson 2009 pers.comm., 12 January). Members of this group participated in the Adnyamathanha Malkarra ('Circumcision Ceremony') and appear to have been considered an Adnyamathanha subgroup. On the death of the 'headman' the group apparently dispersed, possibly joining Yardliyawara at another contemporary camp on Wooltana Station.

Conclusion
It seems likely, on the basis of the evidence I have discussed, that the distribution of the Iga trees in the Northern Flinders Ranges is the result of the introduction of seed by Aboriginal visitors to the area from south-western Queensland. It is also likely that this introduction occurred some time before European arrival since it is discussed in a traditional Adnyamathanha Mura. It would be interesting to map the location of this species elsewhere in Australia in detail in order to determine if the species benefitted from human intervention elsewhere. If it is possible today, it would also be useful to record any cultural association/significance which may be ascribed to the species or individual specimens by Aboriginal groups elsewhere. In the Anangu Pitjantjatjara Yankunytjatjara lands at least one tree in the Northern Territory is attributed by the Anangu as originating from the land of the Adnyamathanha, through an association with a men's-restricted Mura from that southern region. Unfortunately I am unable to discuss that attribution any further.
Did the Iga walk from Queensland to the Flinders Ranges? Obviously, it did not. The species has not literally transferred to, or distributed itself about the Ranges, despite the Iga Mura. As a metaphor for the plant's distribution in this area, facilitated as it was by Aboriginal agents whose own mode of transportation was on foot, it is nonetheless an apt and effective allusion.