Daniel S. Meatte
(taken from the Prehistory of the Western Snake River Basin (1990) pp.63-70)
Julian Steward's ethnographic model of Great Basin Shoshone (1938) is frequently used in conjunction with Jesse Jenning's (1957) archaeological model of a Desert Culture concept to frame the archaeological record in southern Idaho (Butler 1968; J. Green 1972; Plew 1976a; Webster 1978). The borrowing of these concepts is a natural assumption based on the known presence of Numic populations in southern Idaho at contact times (Kroeber 1939; Steward 1938; Stewart 1970), and the recognition that material culture from excavated sites in southern Idaho exhibited many similarities to archaeological assemblages in the Great Basin (Butler 1968).
This has resulted in a deeply imbedded perception by archaeologists that anadromous fisheries occupied a minor role in the prehistoric diet (see Pavesic 1978b for discussion of this problem). Only recently has attention been focused on the identification and examination of evidence for fishing in the study area (Meatte 1982, 1983, 1986a, 1987; Pavesic 1978b, 1986a; Pavesic et al. 1987; Plew 1980f, 1981c, 1983c, 1987a). Summaries of recovered fish remains and fishing equipment from southern Idaho are documented by Plew (1983c) and Pavesic (1986a). Currently, anadromous fish remains identified at excavated sites include the Clover Creek Site, 10-EL-22, (Butler 1983a:4); Three Island Crossing 10-EL-294, (Meyer and Gould 1987:4); Givens Hot Springs, 10-OE-1689, (Thomas Green, personal communication 1988); Nahas Cave, 10-OE-1674, (Plew 1980f, 1986b:97, 1987a), the Bliss site, 10-GG-1 and 10-TF-352 (Huelsbeck 1981:270-284); Cave #1, 10-OE-240 (Schelibach 1967; Pavesic et al. 1987); Deer Creek Cave, 26-EL-25, (Follett 1963:31-32); and Pence-Duerig Cave, 10-JE-4, (Randall Schalk, personal communication 1988). A single vertebrae recovered from Dry Creek Rockshelter, 1 0-AA-67, (Webster 1978:18), may also be salmonid, but identification is indeterminate.
Non-anadromous fish remains identified from excavated sites include: Lydle Gulch site, 10-AA-72, (Brown 1981:238); Kanaka Rapids locality, 10-GG-273, 278, (Butler and Murphey 1983:42); Dirty Shame Rockshelter, 35-ML-65, (Hall 1977:7); Bliss site, 10-GG-1 and 10-TF-352 (Huelsbeck 1981:270-275); and the Crutchfield site (Murphey and Crutchfield 1985:149).
Additional fish remains have been recovered from several other sites, but these remains have not been identified. They include Swan Falls Dam site 10-AA-17 (Ames 1983:27); 10-TF-129 (Bucy 1971a:8); the Braden site, 10-WN-1 17 (Harten 1975:48); and Hagerman National Fish Hatchery (Pavesic and Meatte 1980:4 1).
What is known about fishing in the study area? The ethnographic record for the region clearly documents the use of anadromous and non-anadromous fishes by most groups occupying the region (Liljeblad 1957, 1959; Murphy and Murphy 1960, 1986; Steward 1938, 1941, 1943; Stewart 1941). This documentation also includes several ethnographic groups from outside the region (Northern Paiute, Ft. Hall Shoshone) that traveled to the area with the specific intent of harvesting fish resources (Hultkrantz 1956:20-21; Liljeblad 1957:63, 84-85). The inventory of fishing equipment included a vast array of gear such as fish hooks, bipointed and barbed spears, harpoons with detachable composite points, dip nets, lifting nets, seine nets equipped with floats, weirs, basket traps, and fixed platforms (Hewes 1947; Liljeblad 1957; Murphy and Murphy 1960; Rostlund 1952; Steward 1938; Stewart 1941). Equipment associated with the processing of fish included drying racks, split roasting sticks, fish skin bags, deep bowl mortars, cache pits and framed storage sheds attached to house structures (Liljeblad 1957; Murphy and Murphy 1960; Steward 1938; Stewart 1941). This equipage together with first-hand observations made by early travelers in the region, evidence an efficient, well-defined adaptive pattern of fish exploration (Pavesic 1986a).
Verifying the time depth of this pattern (or any other) in the archaeological record has proven to be a difficult task (Pavesic 1986a). Despite the perishable nature of fish remains and the material culture associated with it, the absence of fish remains and tackle in the archaeological record is understandably deemed negative evidence of fishing activities. A more heightened awareness of fishing activities on the part of archaeologists working in the region (Pavesic 1978b; Plew 19801) has led many to employ more refined collection strategies at excavations (Ames 1983; T. Green 1982a; Plew 1981c) as well as more innovative ways of examining the archaeological record (Meatte 1982; Pavesic 1986a).
Faunal information recovered from dated contexts indicate the use of chinook salmon (Oncorhynchus tshaHytscha) as a food resource by 4,200 years B.P. at the Givens Hot Springs locality on the main stem of the Snake River (F. Green 1982a; Tom Green, personal communication 1987). This date can probably be extended to ca. 7,000 years B.P. based on the recovery of chinook salmon at Bernard Creek Rockshelter in Hells Canyon along the Idaho-Oregon border (Casteel 1977; Randolph and Dahlstrom 1977). A large sample of chinook (Oncorhynchus tshawytscha) remains were recovered there with two radiocarbon dates of 7,250 ± 80 years B.P. and 7,190 ± 135 years B.P. (Sheppard 1977:96). Though Bernard Creek Rockshelter is downstream from the study area, it would seem safe to speculate chinook salmon were able to ascend the remaining portion of the Snake River up to the natural barrier at Shoshone Falls at those times (Pavesic 1986a).
Steelhead trout (Salmo gairdnerii) were available in the area by 2,920 ± 70 years B.P. as evidenced by their recovery from excavated contexts at Nahas Cave (Plew 1987a:19). Plew's assertion that the steelhead trout remains date between 2,920 and 4,990 years B.P. is, however, misleading (Plew 1987a:19). Plew indicates that Steelbead trout (Salmo gairdnerii) remains were recovered from arbitrary level 50-60 cm (1987a:19). A radiocarbon date from a hearth in this arbitrary level is dated at 2,920 ± 70 (TX 3637) years B.P. (Plew 1986b:29, 1987a:19). No fish remains of Steethead trout (Salmo gairdnerii) are reported by Plew for the remaining nine intervening 10 cm arbitrary levels (Plew 1986b, 1987a). In arbitrary level 150-160 cm, Plew reports a radiocarbon date of 5,990 ± 170 years B.P., but no Steelhead trout (Salmo gairdnerii) are documented for this level (Plew 1986b, 1987a:19). Clearly, there is no documented evidence at Nahas Cave to suggest the age of steelhead trout (Salmo gairdnerii) extends or ranges to 5,990 ± 170 years B.P.
Understanding the overall relationship between the local anadromous fishery and the indigenous subsistence system(s) means we must draw from a larger faunal record that is well secured in time with radiocarbon dates. We should remember the Western Snake River Basin represents the upstream limits of the Columbia River anadromous fishery in the southern Plateau. Efforts to understand the importance of this anadromous fishery to prehistoric lifeways are hindered by the borrowing of conceptual frameworks from Great Basin archaeology (Jennings 1957; Steward 1938). To understand the local prehistoric subsistence system(s) linked to this anadromous fishery, we must examine other prehistoric subsistence systems linked to this same anadromous fishery. This means the archaeological record found along the upstream edge of the Columbia River watershed will be of considerable importance to interpreting the archaeological record in the Western Snake River Basin (Swanson 1965). At present, the prehistoric record in the Western Snake River Basin is too steeped in definitions and conceptual frameworks borrowed from the Great Basin. Until this record is understood as a regional system, linked to the Columbia Plateau by way of subsistence, it will not provide us with an accurate measure of the archaeological record (Pavesic 1978b).