Specialized equipment used for harvesting distant food resources was left afield where needed . Equipment depots of this kind are discussed below. Of food caches, ceramic ollas hidden away in rock crevices and small shelters would be the most readily recoverable archaeologically and the most readily recognizable as actual caches. We would expect them to be uncommon occurrences at field camps or other locations far distant from residential sites. The woven platform granaries at residential sites would leave little recognizable evidence in the archaeological record. The term “forager” is applied to hunter-gatherers who moved about the landscape identifying exploitable food resources, mapping onto those resources, and seasonally moving the base camp to the place where the resources occurred. Residence lasted as long as the resources held out or untD other resources became available elsewhere. Foraging peoples probably were more keyed into the location and timing of food resources than any other hunter gatherers in prehistory. They knew where, when, and how to get food; had they not, such groups would have gone extinct, or at least gone hungry, with their first major error in subsistence planning. Ethnoarchaeological studies suggest that caching of foodstuffs by foragers was discouraged because of the constant need to move the base camp; it merely exacerbated the already uneven distribution of resources in time and space. Caching in such contexts ensured that the food would be where the people were not,vertical growing towers and someone would have to go back and get it later.

Caching of food would have been favored, however, if the foraging group intended to loop back and reoccupy the base camp for another purpose later in the season . In such cases, stored foods could augment those obtained on a subsequent visit. Concealment, such as in a rock-lined cache pit in a rock shelter, may have been important if the site was expected to be used by other groups during the interim. Such joint use of sites by two or more groups may have occurred in areas where water sources were few. Nomadic hunter-gatherers following a forager strategy occupied most of the California desert region in ethnographic times. These groups included the Panamint Shoshoni , Kawaiisu , and Southern Paiute-Chemehuevi , and they typified a lifeway common to almost the entire desert area for thousands of years. Little specific information is available on the caching behavior of California desert foragers. What information is available suggests that they adopted a collector strategy when resources were abundant and then regularly cached food. They cached autumn resources such as pinenuts collected to sustain the group over winter. At such times, the forager strategy definitely was abandoned. Steward reported that a single family might harvest at least 450 kg. of pine nuts in a good year. Storage on this scale meant that people often were compelled to spend the winter at or near the place of storage. If a decision was made to remain in the mountains where the nuts were gathered, the field camp became the winter village. Often, however, the winter village was located some distance away, preferably at a lower elevation that offered a more favorable climate and other amenities. In such cases, some of the nuts were transported to the village and the remainder were cached where collected. Descriptions of pine nut caches vary according to region and author, and are sketchy at best. Some accounts say the cones or nuts were buried in the ground and covered with grass and rocks; others suggest they were buried but not in a pit; still others suggest they were cached and covered with brush and earth within a circle of rocks .

AH agree the caching occurred outdoors in the pinyon groves, rather than in rock shelters. A thorough and systematic study of such caches has not been accomplished, but they should leave a conspicuous and decipherable record on the landscape. Field gear, such as pinyon hooks and poles, seemingly would have required little maintenance and special care, and would have been of little value for anything but harvesting pine nuts. Pinyon harvesting gear would have been left, usually in trees, where last used. Pointed or chiselended hardwood poles for extracting agave hearts would have been useful for little else, and likely would have been left near where agave was obtained. We have observed large block milling stones left in plain view as site furniture on the surface in patches of wild bunch grass at various locations in southern and western Nevada and eastern California. These milling stones mark the resource patch and also the field camp or work area from which the resource was harvested. Elsewhere in the same region we have seen similar milling stones around ephemeral ponds or lakes, perhaps left there as site furniture to be used when aquatic resources appeared at irregular intervals. Equipment of this nature likely would have been available for use by anyone in the group. The place it was left thus would not have been secret, and the equipment would have been used on subsequent visits to harvest the same resources. Such equipment would have been too heavy to carry away readily and almost indestructible anyway if someone else chanced upon it and used it. Because site furniture often tended to consist of large objects or facilities that remained at the site, anyone who used the place was likely to bring them into service. The situation can be described no more clearly than thus: “Upon arrival at a known site, one generally searches for the ‘furniture’ and pulls it ‘up’ out of its matrix for reuse. This means that large items of site furniture get continuously translated ‘upward’ if a deposit is forming” . A distinction must be made, however, between visible site furniture that is publicly accessible, and concealed site furnishings not intended for use by others.

Highly specialized and maintained personal equipment such as snare or trap bundles of the kind found at Ord Shelter, San Bernardino County, California , many similar examples reported by Janetski from across the American West, and a cache of dead fall triggers from Fortymile Canyon, Nye County, Nevada , are examples of equipment caches. Hidden equipment of this nature would have been retrieved from the cache and put to use only by the person that owned and cached it. The place of concealment would not have been generally known, and such gear would enter the archaeological record upon the death of its owner. A cache of basketry, ceramic, and metal containers was reported by King from Joshua Tree National Monument. This equipment cache was secreted in a small rock shelter. Whether it represents the activities of foragers, who left it for use on a subsequent trip to the same area, or whether it represents a field camp of collectors, cannot be determined on the basis of available information. Whatever the case,container vertical farming the equipment appears to be of too specialized and personal a nature to be considered site furniture. In few or no cases involving equipment of these kinds do we believe specialized caching or storage pits would have been used. Highly cared-for personal equipment would have been hidden away where it was safe; site furniture would have been given only the necessary care to ensure its presence the next time it was needed. Put another way, personal equipment was cached ; site furniture simply was abandoned . The models of hunter-gatherer organizational strategy discussed with reference to aboriginal life in the California deserts were derived largely from observations on remnant hunter gatherer populations in various parts of the world. In no case was the group being studied as residentially stable as those we have characterized as collectors in southeastern California, nor were they part-time horticulturalists. The models are idealized extremes that are of limited value to characterize the actual range of variation in aboriginal settlement and subsistence behavior in the California deserts. They tend to characterize extremes beyond the actual range of variation displayed by any group. Many ethnographic groups that were primarily foragers employed a combination of the two strategies, or switched from one strategy to the other and back again as circumstances changed, even in the course of a single year. Their organizational strategy therefore changed, and in some cases they may have occupied the same site more than once in a given year, each time with different agendas . The models are heuristic devices that we employ to force us to consider the nature of hunter gatherer organization and how it might help us characterize site function. WTien used in that way, they are useful; they simply cannot be taken too seriously or they lose their intended utility. Construction of better theoretical models is necessary for understanding the role and significance of food caches and the full range of food caching behavior in prehistory, and using such insights for anticipating other aspects of the archaeological record. Improved models will come from reanalysis of ethnographic accounts and from additional ethnoarchaeological observations, but the real contribution must come from carefully collected data from prehistoric contexts. AH hunter-gatherers probably manipulated their environments to one degree or another. Horticulture is merely one form of environmental manipulation, and it is an integral part of prehistoric cultural adaptations on parts of the Colorado Plateau, where during early Anasazi times extensive use was made of rock-lined cache pits that apparently were concealed in rock shelters.

Models that attempt to explain caching behavior should therefore not be limited strictly to non-horticultural hunter-gatherers, or they become even less useful for broad application to understanding prehistoric adaptations in the California deserts, the Southwest, and elsewhere. Indian Hill Rock shelter is a large overhang in Anza-Borrego Desert State Park of southeastern California . This region Ues m ethnographic Kamia territory on the eastern slope of the Peninsular Ranges and west of the Imperial Valley just north of the international border. Elevations in the desert range from below sea level to over 1,000 m. To the west, the Peninsular Range rises to nearly 2,000 m. Extensive excavations disclosed that aboriginal use of Indian Hill Rock shelter probably began as early as 5,000 years ago. Fairly continuous but non-intensive use of the site occurred into the historic period. Although ethnographic records for the region document irrigation of crops by the Kamia at nearby Jacumba, higher in the mountains to the southwest , no evidence of horticulture was found in the excavations. AH available evidence recovered from the shelter suggests a non-horticultural, hunter-gatherer adaptation, although it is recognized that the same group that used the shelter may have been involved in horticultural pursuits elsewhere in their seasonal round, at least in very recent times. The archaeological deposit is nearly two meters deep. The artifact assemblage includes abundant manos and metates, projectile points, hammer stones, and debitage. The assemblage suggests use of the shelter by groups engaged in hunting and plant gathering, but industries such as basketry and skin working are poorly represented. In short, the assemblage suggests the activities of task groups, not the residue of daily living. Excavations were conducted both within the shelter and in the exposed area to the front. Rock-lined cache pits or cists were encountered throughout the lower levels of the deposit, but only in the sheltered area behind the dripline. The upper levels ^ contain little evidence of cache pits, but instead contain ceramic sherds. Although radiocarbon analyses are still in progress, we anticipate that the levels containing abundant ceramics span no more than the last 1,000 years. Eleven rock-lined cache pits were exposed in the sandy deposits at Indian Hill Rock shelter. None of these was covered ; all had been opened and emptied of their contents in antiquity . Thus, what remained of many of the cache pits were the rock-lined or paved floors and some portions of the rock-lined walls. For this reason, no complete measurements for diameter, depth, or volume are meaningful. Several other features also may have been cache pits, but were lacking enough structural detail to be certain; these were recorded as “rock clusters.” No evidence of burning was found, so the features were not hearths. Three general methods of construction were noted during excavation: overlapping slabs; slabs placed in a mosaic fashion; and large, irregular rocks and/or milling stone fragments, sometimes chinked with smaller rocks.