Environment and Subsistence Studies (SGA Research Design)

HOHOKAM ARCHAEOLOGY ALONG THE SALT~GI AQUEDUCT CENTRAL ARIZONA PROJECT Volume I: Research Design Edited by Lynn S. Teague and Patricia L. Crown Contributions by Russell Barber Michael Bartlett Patricia L. Crown Wi IIi am Deaver Suzanne Fish Rona I d Gardner David Gregory Sherry Jernigan Michael Mallouf Charles Miksicek Lynn S. Teague Submitted by Cultural Resource Management Division Arizona State Museum University of Arizona In conjunction with Institute for Conservation Archaeology Peabody Museum Harvard University Prepared for United States Bureau of Reclamation Contract No. 07-0-32-VOlOI 1982 Archaeological Series No. 150 Chapter 8 ENVIRONMENT AND SUBSISTENCE STUDIES Suzanne Fish, Russell Barber, Charles Miksicek Technical analyses related to environment and subsistence will be designed to contribute to an understanding of several major problem orientations for the study area. Four broad headings around which special analyses can be organized are (1) the environmental setting of prehistoric occupation and human modification of the environment, (2) subsistence technology and production, (3) evidence for specialization in subsistence activities, and (4) societal patterns related to subsistence and resources. While the evidence categories pertaining to each of these divisions are not mutually exclusive, the four problem orientations provide foci by which to define relevant data to be gathered and a framework for interpretation. In view of the results from preliminary analyses undertaken in the testing phase of the Salt-Gila Project, information recovery bearing on all four of these problem orientations appears assured. A consideration of requirements for recovery and an identification of problems to be pursued early in project planning have made possible coordinated efforts among the several specialists in environmental studies. Theoretical Orientation Environmental Setting and Human Modifications Modern vegetation in the study area will be described on a zonal basis which has been established through reconnaissance at various times during a year-long interval. In this manner, the presence of annuals as well as perennials may be recorded and data on seasonal availability of resources may be gathered. In spite of the recognition of important recent environmental change, seasonal information may be extrapolated to the prehistoric situation in a more direct manner than the precise location and boundaries of previous plant association. The recognition of important historic environmental change in the study area makes documentation of base line conditions and later trends a crucial aspect of prehistoric interpretation. Major factors 129 130 Fish, Barber, Miksicek differentiating modern vegetation from the historic counterpart include heavy grazing of fragile herbaceous cover, the presence of considerable acreage historically manipulated for agriculture, a drastically lowered water table from ground-water pumping, and the presence of vigorously competitive introduced species. Records from a period prior to that of greatest modification are fortunately available in the form of accounts of early travelers, early photographs, witness trees plotted during surveys as early as 1869, surveyors' notes, and so forth. In addition, physical evidence such as the presence of weathered tree stumps and the age distribution of individual plants in particular communities will be compiled to complement documentary sources. Older residents of the study area will also be queried. A program of modern environmental sampling will be undertaken within each branch of specialized analysis in order to gain a basis for prehistoric interpretation. In the study area, rocky slopes and the upland portions of the 1 ine are more likely to reflect prehistoric conditions. Water-table depths would never have been as important in these settings. Following a reconstruction of historic conditions, modern analog sampling will be extended somewhat beyond the aqueduct right-of-way to include a range of environmental variations suspected in the past. Zones of resource availability for products brought into sites will also be identified. Hohokam modification of the environment in the form of irrigation canals has long been recognized as a hallmark of Hohokam presence and it has been suggested that their hydrological engineering entailed consequences such as salinization of the soil. Recently, Plog (1978) has presented a model of progressive destruction of riverine resources through field clearing in the valley bottomlands, leading to extensions of canal systems. Stein (1979a), on the other hand, has offered an alternative version based on recent Pima practices in which riparian resources are extended out from the drainages along canals and among fields by allowing natural vegetation as well as crops to benefit from irrigation. Birds and small mammals also follow the lines of moister vegetation. The hope of investigating Hohokam effects on the environment rests mainly on an analysis of pollen samples from the period just prior to occupation, throughout the span of occupation, and soon after abandonment. Effects of cultural activity on the landscape away from irrigated land will also be investigated, mainly at the smaller habitation of runoff-agriculture sites. Series of samples may be obtained through the span of occupation or use. Modification in these situations might include removal of trees for fuel and intentional or accidental manipulation of natural plant communities. Environment and Subsistence 131 Subsistence Technology and Production The goal of studies focusing on subsistence technology and production methods is the construction of a systemic view of Hohokam economy during particular periods and of a definition of meaningful changes over time. The environmental biases of the aqueduct right-ofway and the general limitations of dealing with a 1 inear study area make it invalid to assume that we can investigate directly a set of representative subsistence activities related to the support of inhabitants at any one site. Rather, a range of activities can be investigated in some detail to support interpretations of archaeological features recorded in previous work as well as in the current study. Activity at small hunting-and-gathering sites in upland environmental zones is one aspect of subsistence technology only sketchily represented in the SGA sites. Indirect investigation of production will be derived from analyses of samples from features associated with food consumption. House floors, storage features, trash mounds, ground-stone artifacts, and midden levels yield evidence of cultivated plants in use at sites as well as of gathered wild species and fauna. Since the presence of items in these proveniences documents use rather than production directly, an attempt will be made to expand sampling to production loci as well. Hunting and gathering information from small sites represents an aspect of overall subsistence strategy that has only been presumed from remains at large habitation sites. Where site depth and preservation permit, the identification of the plants or animals exploited will be of major concern. Features will be used to identify occupation levels, with samples collected from the immediate vicinity as well. In this way, the evidence for preparation of materials may also be recovered prior to recognition of association with a feature such as a roasting pit. Features and their immediate environs will be treated as discrete units to avoid masking isolated subsistence activities at multiuse sites. Knowledge of resources correlated with specific sites will allow us to draw inferences on seasonality and degree of interaction between sites. Agricultural production using runoff manipulation alone, as opposed to irrigation, is evidenced at various of the smaller sites in the study area as well as in the large contiguous expanses on and adjacent to the right-of-way in the Florence area. Additional features of this nature have been identified during the Arizona State Museum operations, and it is possible that more may be discovered with time. Such sites exemplify an aspect of production likely to differ in crop emphasis and seasonal requirements from canals with sources on permanent watercourses. Habitations, perhaps field houses, associated with such sites should provide an opportunity to compare resource use between dispersed small sites and the larger ones. Samples taken directly from agricultural features are potential evidence of the identity of crops supported by particular kinds of features. A range of topographic situations is represented in known runoff features. 132 Fish, Barber, Miksicek The investigation of irrigation agriculture, other than the configuration of larger systems, involves the location of prehistoric fields. There are a number of ways in which the relevant levels might be located and confirmed, some dependent on specialized analyses. Canal berms, particularly those of the smaller branches, should mark the level of the ground surface corresponding to fields. Occupation surfaces marked by midden or artifact scatters and structural features near canals may also indicate a likely level of fields. Where other devices such as alignments exist, they will present an opportunity to define the sediment associated with previous cultivation. Phosphate analysis, revealing above-normal concentrations of organic material, will be tested for utility in locating fields. The method provides a quick, easily-performed, qualitative indication of relative phosphate content. This substance becomes fixed in soil levels of deposition and does not move upwards or downwards easily. Samples may be taken at regular intervals and immediately subjected to a laboratory procedure requiring only several minutes per sample. Results may then be applied in the field with only minor delay. The suggestion that canals may have had additional roles in subsistence technology beyond the conducting of water is of interest, although virtually undocumented in the Southwest. If the transport of water allowed an extension of riverine plant resources, other riverine resources may also have expanded their ranges. Rea (1979) noted the abundance of fauna attracted to denser, weedy vegetation around fields, and this effect would also extend to the mesic vegetation of canal banks. Such an encouragement of faunal resources may be indicated by faunal remains at sites at a distance from riverine settings, but adjacent to canals. Canals fill with rich alluvial sediments which may be further enriched by a specialized biotic assemblage, including snails, tiny molluscs, and aquatic plants. Canals must also be cleaned fairly regularly to remain efficient. The spoil from cleaning would have a beneficial effect on crop production if used as fertilizer. A test for this practice would involve flotation of samples from suspected field locations to recover any molluscan assemblages present. A successful application of this test has already been carried out in conjunction with Maya agricultural features (Turner 1979: 109-110). Another subsistence resource possibly linked to the presence of canals is fish. Canals in the Southwest may have carried water only seasonally, unlike those in climates with abundant rainfall, such as Southeast Asia, where canals act almost as fish ponds. The potential for obtaining fish might thus be limited to recovery of stranded individuals after periodic high water. Floated samples of canal sediments, already mentioned, would be inspected for fish remains. In flotation of site sediments, an effort would be made to reduce size ranges, since standard harvesting procedures involve the taking of a wide range of fish, including many tiny fish, rather than more selective means of capture. Environment and Subsistence 133 The Hohokam economy has been described as changing over time. The Salt-Gila Aqueduct sites provide an opportunity to detail change in production strategy in the study area. If emphasis on crops or techniques is not constant, the special analyses from Classic sites versus earlier ones in the same vicinity should show direction of change. Similarly, datable small sites can reveal changing patterns of hunting and gathering. Specialization in Subsistence Activities A recent topic of interest in Southwestern archaeology is the possibility of subsistence specialization. This interest has perhaps been stimulated by a concern with the relative complexity of social systems, among them the Hohokam, and with widespread trade networks and cycles. Evidence for specialization is scanty, although systemic benefits for agricultural specialization have been discussed (P. Fish and S. Fish 1977; S. Fish and P. Fish 1978; Lightfoot 1979), and candidates for specialized crops have been proposed (for example, Gasser 1976; Gumerman 1978). Confirmation rests on differential production evidence of gathered resources or crops. Because the study area offers a range of site sizes and environmental settings and because gathering subsistence data of varied sorts is a research priority, the question of specialized kinds of production seems appropriate. Comparison of analysis results using sites as units may also provide data on specialization or at least transport of materials from their probable zone of production or procurement. Some of the approaches described in the previous section apply to interpretations of specialized production. The key to an ability to discuss this topic in other than a general way is the correct recognition of individual production loci. Sites on the aqueduct line provide a variety of contrasting subsistence features and settings which have been demonstrated in the testing phase to contain at least some preserved material with a direct bearing on questions of specialization. Small sites may pertain to hunting and gathering. Runoff features occur on slopes and in more level situations. Canals appear to include both those fed by runoff and those fed by permanent drainages. Reservoirs represent another potential source of water manipulation for agriculture. Any features that might appear to have been hand-watered from canals will be the object of intensive search and sampling. Canal-sediment samples may themselves be relevant. In Belize (Pulestone 1977), cotton cultivation has been documented by relatively large amounts of pollen in canals. Another possible location for specialized crops is in the vicinity of houses in sites where water is immediately available. This would be a convenient place for growing plants that required special or repeated attention. Occupation surfaces outside structures in sites with water sources therefore would also be sampled. 134 Fish, Barber, Miksicek Societal Patterns Related to Subsistence and Resources Differential distribution of resources within a society has implications on various levels. Specialization in production is one aspect of patterned subsistence or resource remains. lntrasite patterns provide referents for village organization. In sites where it may be possible to define more than one house type at a given time, close sampling of each type for a variety of remains should have a bearing on possible differences in access to or utilization of materials. Any structures with suspected communal function would be tested to see whether resource-manipulation or -redistribution might be involved, contrasting with that in individual dwellings. Apparent clustering of structures within sites with associated trash mounds by cluster allows definition of internal patterns and comparison between clusters. Shared storage facilities within a cluster are intriguing possibilities, and any such features will be carefully sampled for the various recovery techniques. Redundancy or lack of it between cluster structures might be important in understanding the nature of relationships between cluster inhabitants. Excavation of the interfeature areas in some clusters should reveal activity areas of a processing nature which, like structure data, can be compared for units within a cluster and between clusters. The distribution of subsistence materials within and between sites which can be identified as gathered versus cultivated, restricted versus redundant, or exotic versus local, should be one of the ultimate products of special analyses. Ethnographic Analogs Analogs with historic Indian subsistence practices should be an indispensable source of insight for the interpretation of prehistoric remains. The emphasis on agricultural evidence, in particular, requires an ethnographic consultant familiar with Pima and Papago cultivation using simpler technologies. In addition, the consultant will be able to suggest appropriate literature. A written contribution by consultant Gary Nabhan is anticipated, dealing with ethnographic parallels and clarification of project results. Another function of the consultant will be to suggest or aid in locating ethnographic situations in which modern analog samples may be obtained and observations made. Such sampling should be most valuable in the case of pollen and flotation interpretation. Structures, processing areas, and fields will be sampled on a limited basis to support interpretation of archaeological data. While these efforts will be modest in terms of time and budget commitment, the added confidence in analysis results should prove most valuable. These efforts are expected to represent one of the more unusual aspects of Salt-Gila special analyses. Environment and Subsistence 135 Research Approaches Pollen Analysis Pollen analysis has been proven feasible with a number of the cultural proveniences and sediment types that were sampled and analyzed during the testing phase. Pollen could not be recovered in only a single instance, a trash mound sample at AZ U:l5:61. Difficulties in analysis were as anticipated: a profusion of tiny charcoal fragments obscuring pollen in one sample from the floor of a burned pit house at AZ U:l4:73 and general difficulty in concentrating recovered pollen. Preservation of grains was good in all instances. Even coarse sediments with many gravel-sized particles, from canals and reservoirs, appear to incorporate preserved pollen. Larger sample sizes will be employed in future phases to concentrate pollen grains further in the extract. A heavy flotation extraction method was employed in this phase, and will be compared with swirl techniques for efficiency and effectiveness. Many of the general approaches outlined for environmental and subsistence studies apply to pollen analysis. Only those which are more strictly 1 imited to palynological study will be elaborated here. A sampling design involving each feature type for both pollen and flotation has also been supplied in table form (Appendix C). Separation of regional aspects of pollen spectra from frequencies influenced by cultural activity is a major problem area in archaeological pollen analysis. There are some possible means or approaching this problem by careful sampling. At small sites, the environment prior to utilization may be represented by surfaces sealed in the construction of features such as rock rings, runoff features, and so forth. These situations have been treated in Appendix C by feature. At larger sites, the problem is greater. Environment in the area is more likely to be heavily modified, so that the base line zonal vegetation exists only at a distance. In these cases, it is as important to perceive hallmarks of the kind of local modification as it is to determine the type of vegetation zone in which the site was situated. Modern analog samples in both more- and less-culturally-modified locales should aid in distinguishing these kinds of information in the pollen record. The two reservoirs may also offer somewhat less biased repositories of regional input, since they are analogous to small ponds. Reservoirs would not reflect vegetation growing directly above the sampling locus or plant handling on the spot. Canals and reservoirs should provide an opportunity for such a chronologically based series of samples. If the berm can be defined, its base would have been resting on the ground surface prior to construction. A period of usage is represented by the sediments in the canal bottom, as well as in spoil thrown on the banks in cleaning. A sequence of postuse samples could be found in the fill accumulating in the canal after abandonment. A similar series of samples might be derived from 136 Fish, Barber, Miksicek sites where constructional features provide a relative time scale. occupation samples should provide an idea of original vegetation. Pre- At small nonagricultural sites, palynology may shed 1 ight on plant materials gathered or processed. Sampling strategies are covered in table form (Appendix C). Evidence would be most recognizable as aberrantly large frequencies of less common pollen types, by presence of aggregates of grains, or by contrasting values for types in identifiable activity loci versus the site as a whole. Sediments from suspected fields will be sampled for pollen content. Spectra will be inspected for cultigen types, documenting particular crops for particular fields. In view of the general rarity of cultigens in the pollen record and the somewhat hit-or-miss nature of field identification, numbers of samples may be scanned for cultigens rather than completely tabulated. It is relevant that pollen of cultivated species tends to be quite localized in distribution as well as rare. Samples tabulated on a more complete basis may be expected to provide insight into possible noncultivated plants allowed or encouraged on a volunteer basis in and near fields as a supplement to planted crops (Rea 1979). Samples taken from modern fields and from the types of Pima fields described by Rea would aid in establishing which species are likely to be abundantly represented as weeds and as additional useful plants. Where fields are not definable, canals themselves provide a sampling milieu more directly related to primary agricultural production than are site features. Fine, damp sediments in canal bottoms would serve as optimal traps and preservation environments for pollen of crops in the immediate vicinity. Soil from agricultural devices other than canals also has a demonstrated potential for revealing pollen of cultivated species. In particular, corn pollen and that of cucurbits has been encountered in a number of studies (for example, Berlin and others 1977; Fish 1978; Fish in press; Martin and Byers 1965). Kitchen gardens near structures will also be investigated by sampling occupation surfaces in the vicinity. It is noteworthy that no pollen of the most common formal cultigen, corn, was encountered in eight preliminary samples from four sites. In the absence of widespread occurrences of the most frequently identified cultigen pollen type, the interpretation of those restricted proveniences where it does occur will be enhanced. The presence of corn pollen in levels sampled as suspected fields, for example, would be more conclusive in view of the pollen's sparse presence in occupation levels within the sites. Another encouraging result from preliminary analysis is the absence to date of huge quantities of Chenopodium-amaranthus (Cheno-am) pollen masking all other types. This type of pollen is often overwhelmingly abundant in Southwestern archaeological sites, but cannot be differentiated between genera that increase as weeds with disturbance and those cultivated or gathered from the same group. The single sample with large amounts of this pollen came from an occupation surface at Environment and Subsistence 137 AZ U:15:19. Cheno-am pollen here appeared economic in nature, with huge aggregates of grains clearly not transported by air. Archaeobotanical Analysis Hohokam sites pose special problems for the recovery and identification of archaeological plant remains. Sites tend to be shallow (and therefore exposed to erosion and weathering), features (and particularly floors of features) are difficult to define, and the use of mesquite and ironwood (which produce very hot, long-lasting fires) as fuel tends to reduce plant materials to a uniform, unidentifiable gray ash. Flotation samples were collected during the testing phase of the Salt-Gila Project to evaluate preservation, determine a standard sample volume for best recovery, provide sampling guidelines for the datarecovery phase of the project, and assess the potential information yield of various archaeological contexts. A standard volume of 4 1 of soil was decided upon which made possible excellent recovery of plant remains (Appendix A). Preservation was good in most contexts and a broad range of materials were identified. Flotation should prove very useful in the following contexts: 1. Providing data about plant species extant at the time of occupation of the sites. The identification of wood charcoal will be an important complement to pollen analysis in this case, especially since most of the conspicuous, woody species of the Sonoran Desert are insect-pollinated and are therefore underrepresented in pollen records from the desert. Mesquite, blue palo verde, little-leaf palo verde, catclaw acacia, whitethorn acacia, creosote, and ironwood have been identified from float samples to date. Evidence of riparian species such as ash, walnut, and cottonwood, which are rare or absent in the survey area today, was also recovered from the flotation samples. The presence of wood charcoal from halophytes such as four-wing saltbush (which is rare in the study area today) makes possible an assessment of salinity problems in Hohokam fields which will supplement the molluscan data. 2. Defining plant-processing activities in and around houses and other features in the sites. Flotation analysis should also be a powerful tool for identifying extramural activity areas. One has only to visit a modern Papago saguaro camp to realize what a small percentage of daily life in the desert actually occurs within the bounds of a house. Most activities occur under ramadas or in archaeologically unrecognizable areas such as under the shade of a mesquite. 3. Identifying construction materials and fuel types. 138 Fish, Barber, Miksicek For example, the ramada from AZ U:15:59 was probably constructed with cottonwood corner posts, ocotillo crosspoles, and grass thatching. The pit house from AZ U:l4:73 had a superstructure composed of creosote, ocotillo, and grass thatching. Changes in the types of wood used for fuel or construction through time should reflect prehistoric human impact on the local environment (field clearance, localized extinction of resources, salinization of fields). 4. Determining a more precise function for more generally defined features such as storage pits or roasting pits. Ethnographically, roasting pits could be used in the preparation of a number of food items including baked squash, green corn, agave hearts, cholla buds, and various types of wild game. Charred plant remains from a feature should indicate its actual last use. For example, the presence of both cholla seeds and numerous cob fragments in the roasting pit from AZ U:l5:59 suggests dual usage. 5. Defining the function of small specialized sites. 6. Describing changes in subsistence strategy through time. ]. Elucidating regional patterns in exchange and interaction. Cotton is an obvious item for trade. Sites that produce cotton will trade with sites that don't grow cotton. The presence or absence of cotton seeds and boll fragments will be important in determining whether cotton is produced locally or imported. By the same token, if only maize kernels and not cob fragments are recovered by flotation from a site, trade in shelled corn could be inferred. Any locally scarce, desirable item could become a trade commodity. For example, saguaro syrup and dried pulp have been historically important exchange items for the Pima and Papago. The detailed analysis of cultigens, especially maize, recovered from flotation should be useful in understanding regional interaction. Maize is extremely variable, occurring in a vast number of distinct and archaeologically recognizable races and cultivars. A high degree of similarity in the maize from a number of contemporaneous and spatially close communities should suggest a high degree of interaction (exchange in seed or food corn) whereas a great deal of variability should suggest little interaction. For this type of analysis large samples of plant remains will have to be recovered. Flotation will probably have minimal direct input into an understanding of agricultural features in the survey area. Even in arid regions such as the Sonoran Desert, plant remains (other than pollen) are seldom preserved in open archaeological sites except by carbonization. Because of the possibility of disturbance of cultural strata by man and burrowing animals and the difficulty in distinguishing old intrusive materials from ancient in situ macrofossils, archaeobotanists are always wary of uncharred remains. Unless fire is used as a tool for field clearing or maintenance (as in swidden agriculture) there should Environment and Subsistence 139 be very little primary burnt material associated with fields or agricultural features. For example, carbonized material could be introduced into canal sediments by any of the following mechanisms: clearing of vegetation growing along canals using fire, slope wash from field clearing (also using fire), or simply erosion from trash dumped along canals. Nevertheless, since we can't dismiss fire as part of the Hohokam agricultural tool kit, agricultural features such as rock piles, checkdams, rock alignments, reservoirs, terraces, and so forth will be sampled in the hopes of getting a few rare instances of good preservation. Canal and reservoir samples analyzed to date have been unproductive. In order to meet as many of the aforementioned research goals as possible, it will be necessary to collect a statistically valid number of samples, from similar contexts, from as many sites as possible, and from all archaeological phases represented in the survey area. Especially useful will be comparing similar samples from the following groups: 1. Large sites adjacent to the floodplain off major drainages such as the Gila River AZ U:15:19, AZ U:15:87, AZ U:l5:85. 2. Large sites adjacent to smaller drainages such as Queen Creek and Weekes and Siphon Washes AZ U:l0:6, AZ U:l4:73, AZ U:15:59, AZ U:l5:61. 3. Small sites on the Pleistocene bench of the Gila River (probably satellite communities for the floodplain sites) AZ U:15:71, AZ U:15:48, AZ U:15:46, AZ U:15:47. 4. Small seasonal gathering or processing camps (often with roast i ng p i t s) AZ U: 10 : 5 , AZ U: 10 : 10 , AZ U: 10 : 12 , AZ U: 10 : 14 , AZ U: 10 : 16 , AZ U: 10: 62, AZ U: 10: 8, AZ U: 10: 11 , AZ U: 10: 13, AZ U: 10: 15, AZ U: 10: 21 , AZ U:l0:64, AZ AA:3:21, AZ AA:3:22, AZ AA:3:20, AZ AA:3:25, AZ AA:3:26. Molluscan Analysis Molluscs, especially land snails, are usually regarded as Illadapted to dry environments such as that of the project area. In reality, however, the adversity of such environments has accelerated evolution, resulting in a large number of species (128 land snail species in the Southwest), most with a narrow range of adaptation and many with a large range of intraspecific variation (Bequaert and Miller 1973: 93-101). The utility of land snail studies to archaeology in the Southwest has already been demonstrated (for example, Antevs 1941). Molluscan analysis can contribute to this project through the detailed reconstruction of microenvironmental zones and the locations and conditions (for example: silting in, salinization) of hydraulic features such as canals and reservoirs. In some cases, it may be possible to assess seasonality by growth-ring analysis of the shells of bivalves collected by the site occupants. Quantitative assessments of 140 Fish, Barber, Miksicek food value from molluscs also. may be possible. Finally, since the harsh conditions of the Southwest usually prevent introduced molluscan species from dispersing beyond the limits of sheltering human occupation (Bequaert and Miller 1973: xi), such adventives may provide hints about trade and other types of interregional flow. The field collection of molluscs falls into two major categories: modern samples and archaeological samples. It is generally acknowledged that modern samples are necessary to "calibrate" archaeological samples if well-founded ecological reconstructions are to be made (Baerreis 1973). As noted above, introduced molluscs fare poorly in the wild in the Southwest, so their impact on modern samples will probably be minor. The collection of modern molluscan specimens must sample the range of environments expected to have existed on the archaeological sites. Accordingly, the variety of ecozones, the range of locations near and distant from modern canals, and the range of conditions within the bottoms of modern canals should all be sampled. At the same time, pertinent environmental characteristics must be noted. The collection of archaeological samples, on the other hand, may be accomplished by excavators trained in proper sampling and collecting procedures. In the testing phase, a very limited amount of molluscan material was recovered. Contexts included a trash mound and a canal. The distribution of molluscan remains in archaeological sites, however, is far from homogeneous. It is quite conceivable that data-recovery operations might retrieve considerable numbers of snails and bivalves from restricted localities. Recovery, indeed, usually tends to be in the form of clusters. Since canals do not appear at this time to be rich in other materials recovered through flotation, larger amounts could be processed with molluscan additions in mind. Areas of eddies or other locations where water is appreciably slowed would also be prime situations for deposition of molluscan remains. [SSt] ..,.. ...., 0 ..,.. "' 0 (1l (1l "0 "0 (1l (1l "'"' "'0"' -, 0 ::J w -, )> "' 00 :e::n :e::n rn n ><- -, ><- -, (1l 3 3 3 (1l ..."' "' rt 0 0 ::J ::J V1 I w "3 3 V1 " Field Number n -, n V1 "'-, (1l 3 ..."' ..."' 0 \..0 0" 0 ::J 0 ::J N ::J Context ..,.. I V1 V1 I V1 V1 I V1 0 "' "' "3 "3 "' "' "3 0 "3 ..,.. w ..,.. V1 w N c: ><- -, (1l V1 ::J V1 I V1 V1 "' ..,.. -....) w I V1 V1 ..,.. \..0 00 0 w N 00 V1 w 0 ..,.. 0. .., <lJ 0 0 (1l ::J rt "' (1l 3 <lJ Sample Weight (gms) ::J "' "'n (1l 0 Maize Cupules < (1l -, (1l w N Maize Kerne 1s 0. Squash Rind -< Cotton Seeds Mesquite Seeds n 0 -, ::J ::J (1l (1l "' "' 3 ~ ..,.. ..,.., <lJ "' ..... -, "'"'"' ~ "' 0 I z ::J 0 3: )> Agave Stem ...::r "'"' Agave Fiber -l rt 0 ::J ,-, 3 (1l Cl> "'rt ::J "' "0 ::r "'"' Prickly Pear Seeds Pigweed Seeds ..,0 (1l rt ::r (1l Tr i anthema Seeds (/} Plantago Seeds rt <lJ I "' "'-, "0 Other P1ant Remains (1l "' " (1l Cho 11 a Seeds 3 ~ n :t: Grass Stems Fish-Hook Cactus Seeds n 0 -, "' )> )> "0 "0 Unid. Legume Pod (see Palo Verde) N "' 3: n ::r -, "'-l Mesquite Pods ..,.. ,"' ...0 0 '-· (1l ...n "' 0 )> n r "'rn 3: )> z (/} (1l 0. X )> [9Sl] )> N 0 N ""N c N--i "'"' "' "' """"' '"'"' 0 0 r 3 "'"' 3<0., "' ., IQJ " ' ::T V"IO I QJ " ' rt :J :;J ., --i rt "' (""') QJ :;J QJ \.() 0 :z 00 ., "'on N N 0 )> ..,. ..,. \.() )> N 00 c c --i N )> N 0 3 "' 00 "' 0 Field Number 0 .,.,"' .,"' "' "' "' "'"' "'"' "0 (""') ... QJ \.() IT 0 ~ 0 :J 0 :;J 0 "' "' 00 I 0 0 0 ·3 I 0 N :;J rt 00 :J 0 0 :;J Context c "' "'., 0. QJ 0. :;J rt 0 3 "'"" 3 QJ 00 N 00 ..,. ""0 ..... N 00 Sample Weight (gms) :J "' "'"' < ..."' 0 0 N N Maize Cupules "" Maize Kernels Squash Rind "' \.() "" "" Cotton Seeds Mesquite Seeds Mesquite Pods Unid. Legume Pod (see Palo Verde) Agave Stem "' "' Agave Fiber Grass Stems Fish-Hook Cactus Seeds Chell a Seeds N Prickly Pear Seeds Pigweed Seeds Trianthema Seeds '"' Plantago Seeds ... "'"' "' "' lO 0. c lO "'"' ,... 0 "' I 0 ., 0. "' "'"'0. "' lO "' "' "'..., < QJ cr QJ "'"' "' 0. 0 Other Plant Remains )> rt "0 "0 rt :;J "' 0 :J ,... 0 3 rt ::T "'--i "'rt"' :J ., ::T "'"' "'0..., rt ::T "' (/) "' rt I "' "'., 0 0 -< "' -< "', ...0 ,...0"' "' "·-· X )> 0 0 :J rt [LS l] ""' ""' ..,.. -<:I: "' 0 ooc:: n"' (1> 00 "' -"' I \.0• 0 ., "' ""' "' ""' ""'"'0 -<::t: \!) -<::t: -<:I: -<:I: -<:I: "' 0 "' 0 "' 0 "' 0 ..,..(1> (1> )> N c Field Number "' 0 ""' " ""' " "'"' "'" "'" "'" "' -<1>"' -<1> "' "'"' "'"' ..,.. "' V1\!) n • 3 ., n 3 0 0 0 0 "' "' 0 o\!) n • 3 0 n \!) 3 • ..,-< "' ..... ., 0 0 "' 0"' 0 ON 3 3 n. rt ., "'..... ., 0 0 .., !!> "' 0 :::0 n • !!> "' !!> cr 0 (1> :T !!> :I: ("") (1> :;,- N (1> Q_ ., Context !!> :::0 !!> rt "' 3 (1> OJ :::0 "' "' \!) 00 00 ..,.. "' "' 00 "' "' 0 "' Sample Weight (gms) "" (1> n 0 < (1) .., Maize Cupules Maize Kernels (1) Q_ V1 Squash Rind "" ., Cotton Seeds rt -< 0 !!> Mesquite Seeds rt Mesquite Pods :::0 Unid. Legume Pod (see Palo Verde) ..,., 0 3 rt :,(1) -< (1) Agave Fiber "' rt Grass Stems "' :::0 <D ., Fish-Hook Cactus Seeds :T !!> "' (1> Chell a Seeds ..., 0 Prickly Pear Seeds rt :;,- (1> Pigweed Seeds Ul !!> Tr ian thema Seeds N rt I G) Plantago Seeds ,., " 0., cr "0 :T !!> n (1> !!> (1> !!> <D 0 cr 3 !!> .~ c: "0 :::0 (1> .,.., c :::0 Q_ (1> (1> "'"3 (1> "' (1> (1> Q_ Other Plant Remains (1> :::0 Q_ 0 Agave Stem "' )> "0 "0 ~. 0 (1> n rt X )> n 0 :::0 rt [SS l] )> N c: "'"' "'"' :I: -o oc (1) (1) .., ., :ZIIl 0 0 \11 (1) < .., I.D \11 I N N 0 n (1) < .., "' (1) N• ..,.. ...., w w w ..,.. ...., 0 n "':::> n "':::> "' "' w"' rno .,-1 ..,-1 \11 .., \11 ...... 00 t:r 0 , \11 "' I 00 0 "' 0 0 \11 3 3 3 (1) ..,"' M ::T .., -i ...... ..,.. \11 n n :I: I N 0 n 3 \11 I.D w "' ..,.. w ..,-1 "'"' ::T .-(1) :::> "' ..,-1 ..,.. )> ..,.. N -I :I: ., 0 c: Field Number n ..,"' \11 -c ""' n"' -Ill .._,N ON n 3 ~ N , 0 :::> (1) 0. M :::> 0 0 .., "' ..,.. t:r 0 ::> c Context "" "'::> (1) 0. M ~ "' (1) \11 3 "':::> 3 3 ...., w 00 n n ...., w N I.D w N ...... w ...... w Sample Weight (gms) "' "' n (1) 0 < (1) ., (1) Maize Cupules 0. Maize Kerne 1s -< t:r Squash Rind , 0 )> M Cotton Seeds "' M "'0 "'0 Mesquite Seeds 0 :::> Mesquite Pods (1) ::> 0. ..,, >< 0 Unid. Legume Pod (see Pa 1o Verde) )> 3 ~ n M ::r 0 :::> (1) Agave Stem rt -1 (1) Agave Fiber Grass Stems Fish-Hook Cactus Seeds Cho 11 a Seeds Prickly Pear Seeds Pigweed Seeds Trianthema Seeds N w "0 "' n :::> "'., "' "' U> c: 3 t:r (1) .., Plantago Seeds :::> (1) "'c3 (1) "'.., "' U> U> M :::> "' ""::r "' U> (1) .., 0 r1- ::T (1) Vl "' M I "' "' ..,"" c: 0. U> 0 Other Plant Remains (1) n M ~ w N )> N 00 c (""") "' V1 ::J "' 00 V1 0 ..,. ..,. ..,. ..,. "" "' "' rn 3 3 3 0 0 (""") ., -o ., ro "'"' -o ., ro "'"' ro "'"' ., ro "'"' -o ::J ::J ::J ::J 00 0 "' w -o ro V1 I N 0 0 0 () n 3 ., 0 ro ro 0 00 V1 V1 0 0 z )> "' "' ., ., 00 3 ro 0 ro n n 3 ro ro rt "' rt "' "'rt 0 0 0 "'0 "' "' "' n 3 ro V1 I w I (""") 0 00 0 V1 3 ., (""") ::J V1 I V1 V1 c (""") 3 ::J ::J "'rt ..,. I V1 0 0 I V1 V1 n n n n :<:: "'" :<:: M· 3 rn M· V1 "" 0 V1 V1 I 3 Field Number ::J V1 V1 I 3 N Context 0 :<:: 0 0 "- 3 (""") :r .,"' n 0 rn 0 "' 0 0 )> )> )> )> w ..,. Relative Amount of Charcoal N 0 00 Mesquite N Blue Palo Verde ..,. "" (/> "' Ironwood n )< )> ~ "'0 n ::J ro rt .., 0 rt "' rt 0 ::J (/> ru 3 -o Walnut "' Unidentified Wood "- 0 Velvet-Leaf Ash Juniper ro ::J ru ., Four-wing Salt Bush )> -o -o I Catclaw Acacia Hackberry x! puadd\f :r ro White-thorn Acacia Cottonwood \f 3 rt "'rt Ocotillo 65! 0 Litt I e-Leaf Palo Verde Creosote N ..,.,"' ro 0 rt "' 0 Appendix A (cont.) -·3: "'-· n Wood Charcoal From the Salt-Gila Project Flotation Samples ..... " '>" -o c ::l '- 0 " E E<ro :z ..... " 'u > -o ..... .0 0 ::l X "c - "u.. ..... " ·-..... .r;;"' ·-,. 0 '-' ::l 0.. 0 ::E -" 6 6 "''-' Q)4- a:: -"' 0 "' <!) ::l co -"' 0 '0 ·-"'u "'u <( ..... 3 0.. c 4- ""' .r;; _J I -.........." -o" '- ·_J <!) > " "' ....."'u ·- "' "' I ..... ·u .r;; u ;3:<( '-' .r;; ..... -"' 0 0 3 c 0 '- -o 01 "0 ..... "'0 - '-'" '- - c 0 3 I '-.J:: ::l 0 (0 "' <( <./) -o 7' "' ::l u..co -..... ·0 u 0 0 0 3 c 0 ..... ..... 0 '-' 7' -o 4- >'""' ''..... ..... " " "> -c -"'u ·-a.c "' :c"' ., >" :3: _J I ::l .0 ::l ·-" 4- ·-..... c " -o -o ·- 0 c 0 ::>;3: AZ U:l5:59 10 Roasting Pit, Tr2a 65-95 em A 12 Trash Lens, Tr2b 25-45 em c Ramada Floor, Tr2a 45-65 em A 1 13 1 7 9 AZ U: 15:61 320 House 20, Roof Fall Tr3b 40 em c 319 House 20, Hearth Tr3b 60 em c 5 316 House 19, Trash Tr5 40 em D 4 315 House 19, Roof Fall Tr5 lOOem A 6 314 House 19, Floor Tr5 125 em B House 21, Floor Tr8c 85-90 em c 1 1 1 3 4 2 4 2 3 )> N c: "' "'..,"' "'< "'"'"' - :r: -o oc < N• ..,"' 0 0 :ZV> CD "'mo """' 0 .., "'"' .., .., \.0 \J1 I I co N N 0 w ..,.. n n :r: ...,.., ..._, ::J ::J [lJ [lJ [lJ [lJ rt "'::r w [lJ ..., .., ..._, [lJ [lJ ..., .., co ::r ..,..., ..._, \J1 0 0 0 \J1 I () [lJ () 3 3 ..,"' 0" 0 () w co ..,.. -h n ..._, w 0 3 N \.n ..,.. 3 .CD ::J "' ...,.., ..,.. [lJ ..,.. w )> ..,.. N c: :r: 0 "' c "' "' "' N N ~ () 0 .., 8.., ..., .., Field Number ::J rt ~ ::J Context c a. "' :<: 0 0 \J1 n () 3 3 a. 0" ..._, n ::r 0 () .., 3 () [lJ m m )> m m m m 0 0 Relative Amount of Charcoal Mesquite Blue Palo Verde Little-Leaf Palo Verde N ..,.., 0 3 ,.., ::r "' ..,-u (ll () Four-wing Salt Bush ,..,0 .., [lJ rt 0 ::J (/1 (l) Hackberry Juniper Unidentified Wood l9l v X!puadd'J X )> n 0 ,.., ::J 0 Creosote Walnut a. "' ,.., Velvet-Leaf Ash CD ::J I Catclaw Acacia Cottonwood "0 "'rt [lJ Ocotillo )> "0 (/1 White-thorn Acacia Ironwood w [lJ 3 "0 (ll "' ~