6. Significance in Terms of E. ferox Exploitation at GBY

The Indian record of E. ferox seed consumption indicates that there are several ways to harvest and process E. ferox nuts. Clearly, the major difficulty to be overcome is the presence of prickles/spines that make processing the fruit difficult. Further difficulties involve collection of seeds after the fruit bursts. The Bihar method described here overcomes both of these problems through adoption of underwater collection procedures. Key points emerging from investigation of traditional methods from Bihar in the context of the E. ferox remains from GBY are summarised as follows: a) gathering of E. ferox seeds takes place after they ripen and sink to the bed of the water body; b) gathering by diving is a necessity, as the plants grow in still waters and seeds are not washed to the edges of the water body; c) the work necessitates observation of the lifecycle of the plant and of the prime time for gathering seeds; d) drying and popping seeds was done at a distance from the water body, where fire and dry land facilitated later stages of processing; e) roasting and popping are both procedures requiring the technology of fire and that of anvils and hammers; and f) a well-established division of labour was associated with each stage of gathering and processing.

Different methods of E. ferox consumption are cited here, showing that the seeds are a highly valuable component in the diet of wetland communities. It is clear from literature and our own data that common processing methods of E. ferox seeds in India are those that include the use of fire. Clearly, the prickly nature of the plant parts is an important trait that local communities have to consider and overcome. We believe that the traditional ethnobotanical procedures involved in the gathering and multistage processing of E. ferox in Bihar are instructive parallels for the interpretation of the GBY archaeological data. Thus, the archaeological record of GBY, which includes the use of fire and the presence of pitted stones, anvils and hammerstones in association with E. ferox seeds, strongly supports the use of analogies with traditional modes of gathering and processing, such as that practiced by communities in Bihar. Studies of the GBY archaeological record provides information on the co-occurrence of a range of finds that may be compared with the ethnographic data. 1) In each of the archaeologically rich horizons there were spatial concentrations of burned flint microartefacts. Analysis of these concentrations suggests the presence of phantom hearths, the earliest evidence for the control and continual use of fire in western Eurasia (Alperson-Afil and Goren-Inbar 2010). High-resolution data from excavations enables estimation of the size of these hearths, which were around 0.49m long and 0.35m wide (Alperson-Afil and Goren-Inbar 2010, 74, table 4.1). 2) Pitted stones and hammerstones, as well as the newly identified thin basalt anvils, were also found in each of these horizons (Figure 10). 3) In all archaeological horizons, remains of T. natans and E. ferox were discovered as well. The pristine taphonomic context of the archaeological horizons at GBY, along with the significant patterns of association noted between various find categories discussed above, provide a background for our discussions of the spatial patterning of past activities. Spatial analysis of these associations and analyses of Layer II-6 Levels 2 and 6 provide further insight into the proximity of hearths and pitted stones (on both blanks and blocks) (Table 1; Alperson-Afil and Goren-Inbar 2010, 91, figs 4.8, 4.9; Alperson-Afil et al. 2009, 1678, fig. 2). This correlation of nuts, phantom hearths and pitted stones at GBY leads us to suggest that some key aspects of the methods of collecting and processing noted in Bihar, which includes roasting and subsequent popping of the seeds, may be of greater relevance for the GBY data than those described from elsewhere in India. Greater precision in spatial associations between the nuts and other features in the vicinity of the paleo-lake is not possible owing to the light weight of the seeds. However, common aquatic taxa in the Upper Jordan Valley (Lake Hula), the Acheulian site of GBY and Bihar (India) (Table 2) reflect the extent of ecological similarity, despite their great biogeographical distance. The habitat and surrounding environment of paleo-Lake Hula was a rich and diverse Mediterranean one, as evidenced by the identification of an array of 60 edible taxa recorded at GBY (Melamed 2003, table 3), as well as a wealth of fish, crustaceans, birds and mammals.

E. ferox grows in water bodies with water depths ranging from around 0.3m to 1.5m, up to a maximum of around 2.5m (Jha et al. 1991; Mishra et al. 2003; Mandal et al. 2010) or up to around 3.5m as noted in the study region. Although Acheulian hominins may have consumed seeds raw, this would have entailed considerable effort in harsh conditions owing to the prickly nature of the plants. With the technology enabling them to process nuts using fire, anvils and percussive tools, hominins could avoid the difficulties posed by exploitation of raw seeds.

The fluctuating water level of paleo-Lake Hula would have been an obstacle to adopting simpler methods for gathering nuts, in view of the plant's life cycle and the water depths and geochemistry required for its growth and survival. A desiccation scenario of fluctuating lake levels would have resulted in the death of plants, unable to regenerate as germination occurs under water. Exposure to atmospheric conditions would have resulted in the complete decomposition of the macrobotanical remains found at the site. The entire issue of organic preservation is based on anaerobic conditions (and hence inappropriate conditions for bacteria responsible for the decomposition of organic material). Irrespective of the depth of the water, hominins would have had to collect nuts from beneath the lake surface, entailing some amount of time spent under water.

It is important to note that we do not suggest that Acheulian hominins followed modes of collecting or processing that were identical to those practiced today, particularly in the case of elements dictated by modern economic conditions (use of bamboo poles to demarcate underwater areas for collecting nuts, sieves for sorting nuts for sale or even gender-based division of labour).

Such cognitive procedural abilities of planning and performance in aquatic habitats, particularly when combined with exploitation of fish (Zohar and Biton 2011) have not previously been reported for Acheulian hominins. Ethnographic analogies demonstrate that exploitation of E. ferox nuts is performed by communities of fishermen in water bodies that are also used for fishing (Jha et al. 1991). The most abundant fish species currently exploited in habitats associated with E. ferox include three families of air-breathing fish: Cyprinidae (carps), Clariidae (catfish) and Bagridae (catfish) (Table 2). At GBY, remains of Cyprinidae and Clariidae were recovered, predominantly Cyprinidae (mainly the large Barbus sp. and Barbus longiceps). Interestingly, the cyprinids remains were recovered in association with living floors excavated in Area B (Table 2) (Alperson-Afil et al. 2009; Zohar and Biton 2011). The archaeological association between E. ferox nuts, large quantities of cyprinid remains and other cultural activities documented at GBY presents novel evidence for intensive exploitation of the aquatic fauna and flora of paleo-Lake Hula.