A century ago it was common knowledge among historians and anthropologists with research interests in northwestern Mexico that Indians had inhabited four large islands in the Gulf of California. Three of these islands—Cerralvo, Espíritu Santo, and San José—are in the far southern Gulf off the coast of Baja California. The fourth island, Isla Tiburón, is located on the mainland side more than 400 km to the northwest, in the constricted region of the Gulf known as the Midriff. At the turn of the twentieth century, Isla Tiburón was well known as the homeland of the Seri Indians and was the only Gulf island still inhabited by native people. Since 1900, ethnohistorical and archaeological research has greatly expanded our knowledge of Indians on both sides of the Gulf. Much of that information, however, pertains to the people living on the peninsula and mainland, and touches on the islands only incidentally. Consequently, few historians or anthropologists are aware that Indians made use of many islands in addition to Cerralvo, Espíritu Santo, San José, and Tiburón. Scholars in other fi elds may not even know that there were Indians on the Gulf islands other than Tiburón. This is particularly unfortunate for ecologists because native people have been in the region for thousands of years and may have played a signifi cant role in shaping the island ecosystems we see today. Although indigenous humans were by far the largest native terrestrial mammal on all but two islands, and the most voracious omnivore on all of them, Homo sapiens does not appear on island species lists, and the potential effects of native people on insular ecosystems have seldom been considered. Discoveries made just in the past decade show that researchers in all fi elds have seriously underestimated the extent to which native people made use of the islands. Reports of early Spanish navigators have established the presence of Indians on many islands in addition to the four listed above. Recent archaeological research on several islands, along with fortuitous observations on others, have revealed evidence of native people on islands with no known documentary record of Indians. Chronological data from the southern Gulf establishes a time depth for indigenous people spanning at least ten millennia. New information from Seri oral history alludes to Seri voyages far beyond Isla Tiburón and greatly expands the picture of indigenous people in the Midriff region. Collectively, these results show that the traditional assumption, that most islands were beyond the range of native people, is dead wrong. It is now clear that Indians knew and exploited nearly every signifi cant island in the Gulf. This study reviews the evidence of native people on each of 32 major Gulf islands. The list includes all 22 islands larger than 2 km2 and 10 smaller islands for which some data exist (summarized in Table 3.2). The data are drawn from historical documents, oral history, and the archaeological record. To the extent possible these data are given as quotations from the original sources. Collectively, the evidence suggests that native people were familiar with at least 29 of the 32 islands. For 19 of the islands the evidence can be considered unequivocal, consisting of unambiguous historical documentation, credible oral history, and/or a clear archaeological record. For ten islands there is some evidence of human use, but it is limited, weak, or equivocal, and therefore, in need of corroboration. There are no data of any kind for two islands, and one small island has produced no evidence of native people despite a comprehensive archaeological survey (these data are summarized in Table 4.1). Of course, Indians made greater use of some islands than others. In general, large islands, with a greater diversity of resources (including fresh water), were exploited more than small islands, and several supported permanent or seasonal communities of people. However, native visitors may have been drawn seasonally or intermittently to even very small islands with special resources, such as nesting seabirds, sea lion colonies, concentrations of cactus fruit, and abundant seed crops that appear after a rain. Historic and ethnographic sources show that islands did not need permanent water to sustain native visitors, who were quite capable of bringing water with them, subsisting on temporary water in tinajas, or utilizing water substitutes. Distance was apparently no barrier to native use of islands. The cane balsa was the universal watercraft, and historic sources suggest that balsa traffi c was extensive throughout the Gulf. In the hands of the Indians, the balsa was a swift and seaworthy craft, and navigation was no doubt facilitated by expert knowledge of winds and currents. All but two islands were within a day’s paddle, and in most cases the overwater distance to the more remote islands could have been reduced by island hopping. Although we now know that native people exploited nearly every signifi cant island, we need much more information about the time span over which those visits took place. Indians were certainly making use of many islands during the seventeenth century when European navigators began keeping careful records. Seri oral history fi rmly places Seri ancestors on several Midriff Islands during the nineteenth century and conceivably earlier. Archaeology is potentially capable of extending island chronologies into the prehistoric past, but there has been only limited progress on this front. Fewer than half of the 32 islands considered in this study have been systematically surveyed, and controlled excavations have been conducted on only three islands. Unfortunately, the archaeological record for many islands consists only of surface remains, and includes few or no time-sensitive artifacts or structures, and little or no organic material suitable for radiocarbon analysis. Clovis projectile points from both the Sonora mainland and the Baja California peninsula indicate that humans were present in the Gulf region by 13,000 years ago, which means that native people have been potential island visitors since that time. So far, radiocarbon dates have been secured for only five islands, and for four of these islands the few dates that are currently available all post-date AD 700. However, a series of 179 radiocarbon dates from 40 sites on Isla Espíritu Santo have clearly established this island’s long occupational sequence, which extends from about 9000 BC to the fi fteenth century AD. Although one site on this island produced spectacular radiocarbon ages on shells from approximately 36,550 to greater than 47,500 BP, these shells were probably already ancient when people collected them. During the late Pleistocene and early Holocene it is likely that not all of today’s islands existed, or existed as islands. Some small volcanic islands may not yet have emerged from the sea, while some of today’s islands were connected to the shore by land bridges because sea level was much lower than it is today. Any island without a landbridge connection at that time would have been accessible only by watercraft. These non-landbridge islands should be considered prime places to search for early evidence of navigation. This in turn raises the question of whether initial human entry into the Americas took place by boat along a Pacifi c coastal route, and whether subsequent dispersal involved the Gulf. In most coastal entry and dispersal scenarios, it is assumed that coastally-adapted boat people arriving at the southern tip of Baja California crossed the mouth of the Gulf to the Mexican mainland, making landfall in Sinaloa or even farther south. However, as R. James Hills has pointed out, this is highly unlikely. People arriving at the southern tip of the peninsula would have seen seemingly endless ocean in all directions except along the peninsular coast leading into the Gulf. There would have been no reason for people adapted to coastal resources to set out into an apparently empty sea. Instead, they would have followed the coastline into the Gulf, presumably exploring the islands along the way. The fi rst reasonable place to cross the Gulf would have been the Midriff, where they could have island-hopped to the mainland with no overwater distance exceeding 17 km. In Hills’ scenario, the Gulf would occupy a pivotal position in human dispersal in the Americas, and it is possible that evidence of this process has been preserved on some of the Gulf islands.

This seventh volume in the nine-volume series showing results of archaeological studies along the Salt-Gila Aqueduct, focuses upon studies of environmental conditions and subsistence practices at the 45 Hohokam sites investigated by the project. These represent an important element of project research and a level of attention to these studies unprecedented in Hohokam archaeology. It is in this volume that the final results of botanical, faunal, and palynological work are reported. In addition, there are summary statements on SGA Project work related to agricultural technology, broader agricultural strategies, and strategies for the exploitation of natural resources of the Sonoran Desert. This work, taken as a whole, reflects the extraordinary diversity and flexibility of Hohokam subsistence strategies. While drought and floods are the inevitable enemies of agricultural populations, the Hohokam appear to have developed the means of coping with their environment early in their history.

For ten weeks during the late spring and summer of 1976, the Arizona State Museum conducted data recovery operations at seven archaeological sites as part of the Cyprus-Bagdad Project. These sites were located within the right-of-way of a pipeline to be built by the Cyprus-Bagdad Copper Company, and were investigated in an effort to mitigate adverse impacts from pipeline construction. Research conducted within the project was directed primarily toward problems involving prehistoric adaptation to the local environ- ment. The analyzed data were applied to the testing of hypotheses regarding the relationship of site locations to local availability of water and to the locations of economically significant resources. Also tested were hypotheses intended to assess the importance of agriculture as a mode of subsistence for the area's prehistoric inhabitants. The data were also used to determine the functions of individ- ual project sites. Analysis indicates that these sites represent differing functions ranging from specialized activities such as 1ithic raw material procurement and wild plant food procurement and proces- sing to long-term habitation. At least one site possessed material remains that indicated its inhabitants practiced agriculture. Evi- dence from the project sites also suggests that the availability of water was perhaps the primary consideration of the area's prehistoric inhabitants in determining site location. The history of previous anthropological research and the cul- ture history of the project area are briefly discussed. All seven project sites are described in terms of their condition when discov- ered; morphology; environment; architectural and agricultural features; and chipped stone, ground stone, and ceramic assemblages. The appendices to this report discuss the cr i teria used in evaluating the various artifact assemblages and the location and composition of local source areas of lithic raw material.

This volume is the last in a series of nine reporting the work of the Salt-Gila Aqueduct, Central Arizona Project Archaeological Data Collection Studies and Supplemental Class 3 Survey Project (SGA). This study was funded by the U.S. Bureau of Reclamation (Contract No. 0-07- 32-V0101) to mitigate potential adverse impacts of Central Arizona Project construction on cultural resources in the aqueduct right-of-way. Data recovery was conducted at 45 Hohokam sites along a 93 km (58 mile) transect extending from Apache Junction to a point southwest of Coolidge and near Picacho, Arizona (Fig. Intro. 1). This is the largest of the Central Arizona Project archaeological studies to date, although it may fall short of being the largest that will be conducted under the program. (excerpt from Introduction)

This study focuses on one of the few hallmark artifacts from the Hohokam Pre-Classic period: palettes. Previous research on palettes is synthesized and the nature of the prehistoric production, distribution, and use-lives of these artifacts is explored. The term "palette" is found to inadequately and inaccurately describe the many ways in which these artifacts were used within Hohokam society. In its place, the author suggests that the term "tablet" be used for the artifact class. Important byproducts of this study are life-size and highly accurate line drawings of over one thousand palettes, as well as a fully searchable relational database.

The mobile hunters and gatherers of the Archaic Southwest were members of at least three different kinds of social groups: bands, endogamous marriage groups and a risk-sharing economic network. By comparing the geographic distributions of iconological and technological style in cordage, sandals and projectile points, it is possible to distinguish marriage groups from the larger economic networks. Using artifacts from Bat Cave, Tularosa Cave and Cordova Cave in the New Mexico Mogollon and from Presnal Shelter in the Tularosa Basin, this research was able to demonstrate that technological style in fiber artifacts is more geographically constrained than iconological style in textiles or projectile points indicating that although groups using these rock shelters came from different bands, they belonged to the same marriage group, yet participated in different risk-sharing economic networks.

This report describes the nature, extent, and results of archaeological excavations conducted at Mound 8 of the Las Colinas Ruins Group, Phoenix, Arizona. The excavations, undertaken in 1968, and subsequent analyses carried out during the following decade, were supervised by Laurens C. Hammack, then of the Arizona State Museum, University of Arizona. The Arizona State Museum excavations were authorized by the Arizona Department of Transportation as part of an annual Statewide Archaeological Contract between the two agencies. Las Colinas is a Classic period (A.D. 1100-1450) Hohokam site located between Interstate 17 and 27th Avenue, south of McDowell Road, in Phoenix. The site consists of a specially constructed platform mound, habitation structures, and assorted features. These prehistoric remains were damaged by the construction of a historic adobe house in the 1880s and by vandalism through the following years. Mound 8, and the remains in the immediate vicinity, are threatened by the proposed construction of a segment of Interstate 10 known as the Papago Freeway. Excavations focused on Mound 8, although the flat area east of the mound was tested; a single cremation area was discovered there. Habitation structures on and around the mound were excavated also. It was determined that the mound was composed of post-reinforced, adobe-walled cells and encircling walls that formed the core of the structure. Various additions were made to this core, the most notable of which was a massive, solid-adobe wall. The top of the mound was capped with a layer of adobe on four different occasions. At least 22 habitation structures, both pit houses and houses with solid-adobe walls, were discovered during the excavations. The range of structure morphology for any given period of occupation is much greater than that previously reported for other Classic period Hohokam sites. These and numerous other architectural features are discussed in this report. Substantial collections of ceramic, chipped stone, and ground stone artifacts were recovered. These assemblages were thoroughly analyzed. The methods and results of these analyses are reported in this volume. In addition, specialized analyses were performed on a wide range of materials recovered from the excavations. These analyses, described in appendices, pertain to human osteological remains, disposal of the dead, mammalian remains, bone artifacts, avian remains, shell artifacts, pollen identification, charcoal identification, and historic artifacts. A final appendix lists Arizona State Museum catalogue numbers for many of the artifacts and illustrations found in the volume.

Los Morteros (AZ AA:12:57 [ASM]) is a large Hohokam village site located at the northern end of the Tucson Mountains along the Santa Cruz River. Named for the bedrock mortars located near the center of the site, Los Morteros has a rich and varied history. Several phases of prehistoric occupation are suggested by ceramics representing the Colonial, Sedentary, and Classic periods (A.D. 500 to 1450). The range of features present is considerable, including cremation pits, pit houses, roasting pits, mounds, canals, petroglyphs, hill-side terraces (trincheras), a ballcourt, and the bedrock mortars. Spatially, th~ site covers a large area, but most features are clustered in a limited core area (Fig. 1.1). The area around Los Morteros has also been used during historical times (see Stein 1982). In this report reference to the historical period is limited to those events that bear directly on the site of Los Morteros, in particular the history of archaeological interest in the site. Our major concern is with the prehistoric components at Los Morteros. (excerpt from Introduction)

In April of 1980, the Cultural Resource Management Section (CRMS) of the Arizona State Museum (ASM) contracted with Arizona State Parks to provide an archaeological assessment of a portion of the proposed Catalina State Park lands. Approximately 9.7 square km (3.75 square miles) of the proposed park, which is located roughly 22 km (14 miles) north of Tucson, Arizona, were surveyed, and a records check and literature search were undertaken for the area. Two previous CRMS projects, the Rancho Romero reconnaissance (Roubicek, Cummings, and Hartmann 1973) and the Canada del Oro assessment (Brew 1975), had studied portions of the present project area, but to acquire the precise archaeological information required by Arizona State Park planners for the first phase of park development for public use, additional study of the area, especially in the form of survey work, was deemed mandatory. The archaeological survey that was conducted in conjunction with preparing the archaeological assessment resulted in the location of four previously unrecorded prehistoric sites (two camp sites, one resource- processing locality, and one agriculture-related water-control system); the relocation of known sites, with the correction of some erroneous site loca- tions; the definition of zones of moderate artifact density that will require additional investigation; and photographic documentation of all recent his- toric structures encountered in the project area. The literature search revealed seven previously recorded sites in the project area, some additional historical information on ranching activities in the vicinity, and sources that should produce more in-depth data for future studies. This report provides both a summary of the available information on the prehistoric and historic resources contained in the study area and recommendations regarding their management, and discusses the eligibility of various sites for inclusion on the National Register of Historic Places and the Arizona State Register of Historic Places.