4 Site Formation Processes at the Native Alaskan Neighborhood HEATHER A. PRICE IN CONJUNCTION WITH the archaeological investiga- tions at the Native Alaskan Village Site (NAVS) and the Fort Ross Beach Site (FRBS) introduced in chapters 2 and 3, a multi-scalar analysis is undertaken to elucidate the history and processes of site formation on and along the base of this California coastal terrace. The analysis includes a field-based geomorphological study and a series of sedimentological studies carried out in the laboratory facilities of the Archaeological Research Facility at U. C. Berkeley. Qualitative observations taken in the field are designed to address three related issues. These are: the origin of the site sedimentary matrix (location and lithology of the parent rock); the way in which sediment travels from sources to deposits (including weathering and tansport processes); and the degree of and kinds of alterations undergone by the sediments once deposited. This information enables the archaeologist to determine both how the cultural materials were deposited at the sites and the relative integrity of the archaeological deposits; that is, whether the culturl materials were relatively undisturbed once deposited, or whether and to what extent they have been disturbed post-depositionally. Qualitative and quantitative analyses carried out in the lab complement the geomorphological study by providing further information regarding post-depositional processes and the history of soil formation at both sites. Lab analyses include determination of color, pH, organic carbon content, calcium carbonate content, and structure. These analyses help us understand the nature and degree of human impact on these sites from the time of deposi- tion through the present. Before proceeding with the geomorphological report, I will briefly summarize observations of the stratigraphic successions exposed at both sites during the field seasons of 1988-89 and 1991-92. The stratigraphic successions are treated in greater detail in chapters 2 and 3 and in the sedimentary section of the current chapter. FORT Ross BEACH SITE The Fort Ross Beach Site extends for roughly 30 m across the colluvial toe at the southeastern base of the coastal terrace. The profile is divided into three seg- ments-West, Middle, and East based on topographical and archaeological characteristics. The West Profile is nearly 2 m deep and has 4 strata: a fill consisting of yellow compacted clay and large angular siltstone fragments, compacted brown mottled clay, a localized yellow clay lens, and a beach gravel base. The Middle Profile centers around a feature composed of a red clay- lined basin roughly 2 m across that was partially filled with large rocks. This profile consists of topsoil, fill of yellow compacted clay and large angular siltstone fragments, a midden (garbage area), compacted brown mottled clay, and an underlying clay stratum. The midden is sometimes divided into midden 1 (upper) exhibiting numerous bone and shell fragments, and midden 2 (lower) exhibiting the same dark-brown-to- black loosely packed sediments with a significantly lower concentration of bone and shell. The East Profile is less than 1 m deep and consists of the topsoil, a midden, and a clay stratum. The beach gravel base and the lowest clay strata of FRBS yield charcoal and obsidian tools and debitage from lithic production. Historic Period artifacts such as glass, glass beads, and ceramics are limited to the upper portion of the profile, including the compacted brown clay stratum, the midden (where they are quite common), Site Formation Processes 97 and the topsoil. Bone and shell fragments of all sizes have been recovered from the top meter or so of the profile. In general, the West Profile has only a sparse concentration of cultural materials while the Middle and the East profiles yield more artifacts. These are primarily associated with the midden and the pit feature. The West Profile of the site is steeply sloping. The Middle Profile is moderately sloping, and the East Profile forms almost a bench, although it too has a gentle slope. NATIVE ALASKAN VILLAGE SITE The Native Alaskan Village Site extends across the top of the terrace between the Fort Ross Stockade and the ocean. The excavation trenches and test units placed across the site expose a basic stratigraphic succession including the topsoil, a midden (identified as dark sandy loam), and a clay substratum. In some places there are additional strata that appear to be related to cultural features. For example, in unit lIOS, 11W there is a layer of rock rubble between the dark sandy loam and the clay substratum. In other units, such as 125S, 22W, a lens primarily composed of bones, fire-cracked rock, and artifacts (the bone bed) is seated within the dark sandy loam. Beneath this stratum is a concentration of rock rubble, then a dark pit fill associated with a pit feature, and beneath the pit feature a discontinuous layer of mottled dark sandy loam. This stratum is succeeded by the final clay substratum that rests on the marine sedi- mentary bedrock of the terrace. This terrace top deposit is less than one meter deep. GEOMORPHOLOGICAL METHODS Geomorphological methods include reference to geological maps in association with field inspection of bedrock exposures, stream and road cuts, erosional features such as gullies, and evidence of modem day rodent activity. Geological literature describing local processes (such as wave erosion) is consulted. Historical documents, including maps and photographs housed in the Fort Ross Interpretive Association Library at the Fort Ross State Historic Park, help to locate old roadbeds, determine the date of road building and other ground disturbing activities, trace the rate and path of migration of the Fort Ross Creek, and locate Historic Period structures on or near the sites. The excavation profiles at FRBS and the excavation units and trenches of NAVS present vertical and horizontal windows into the succes- sion of sediment deposition at the two sites. The source of sediments is determined by comparing characteristics of the deposit to local bedrock. The lithology of the sedimentary parficles indicates the possible source materials. Relative particle angularity and size indicate the relative distance the particles have been transported, as well as the most likely agent of transport. For example, the more angular the sediments, the less distance traveled from their parent rock. Smoothed particles are more likely to have been water transported. Rock composition is taken into account, as it affects the amount of weathering and alteration any given particle may be susceptible to (Selby 1982). Once the source is determined, the exposed bedrock is examined for evidence of predominant weathering processes. Local weather patterns and the rock structure are considered in creating a scenario for the decomposi- tion of the bedrock into sediment. The transport pro- cesses are ascertained by looking for a checklist of possible agents such as earthworms and rodents, land- slides, rock falls, sheetwash, and so on. Rodent activity is revealed by heaps of churned up sediment on the ground surface, and by the presence of abandoned and fllled in rodent burrows in the profile. Landslides are identified by scarp scars, and rock falls are identified by the concentration of large particle sizes in the colluvial profile. Post-depositional alteration includes any physical or chemical processes that changed the sediments once they were deposited. Evidence for such alteration includes vertical horizonation and soil development, animal burrowing, leaching, color change, or erosion. The profile is examined for any evidence of structure within the colluvial deposit, such as size graded lenses, particle orientation, and for any signs of disturbance such as mixing. GEOMORPHOLOGICAL RESULTS BACKGROUND The Native Alaskan Village Site is situated on the relatively flat terrace top that gently slopes upward away from the ocean. The Fort Ross Beach Site is situated along the beach front of this southeast facing coastal terrace. The terrace beach front ranges from a nearly vertical escarpment at the ocean to a moderate hillslope inland. At the base of the hillslope is the Fort Ross Creek. The creek bed has migrated across the cove through time, and its current position up against the north edge of the cove (where FRBS is located) was reached after the mid-1800s. Both the terrace top and hillslope are crisscrossed by the remains of a road built by the Russians as well as by the currently used dirt road that was constructed in the early 1920s. The hillslope bears these scars as well as small scale, localized erosional features such as gullies and slumps. Large sandstone boulders (riprap) litter the base of the terrace and the beach. These originated from the terrace above. The parent material can be seen protruding in isolated clumps towards the edge of the terrace closest to the ocean. The Soil Survey of Sonoma County, California (USDA 1972) assigns the local soils to the Rohnerville series. These soils are moderately well-drained loams of varying slopes. The subsoil consists mainly of sandy 98 The Native Alaskan Neighborhood clay material weathered from the soft sandstones on the marine and beach terraces. The Mean Annual Precipita- tion is from 30 to 45 inches, while the Mean Annual Temperature ranges from 52 to 540 F. The vegetation consists primarily of annual and perennial grasses and legumes. ORIGINS OF THE SEDIMENT The Fort Ross coastal terrace is composed of Monterey Formation marine sandstones, siltstones, shale, and thin-bedded chert laid down during the Tertiary (Wagner and Bortugro 1982). These are most likely turbidites formed by turbidity or density currents in the ocean, in this case, on the continental shelf. Variation in the currents and therefore in the suspended sediment load resulted in distinct beds of different grain size and thickness (Walker 1979). These beds are parallel sided and laterally extensive. Exposed bedrock near the top of the terrace shows parallel beds of siltstone of varying thickness. The color varies from light brown to light grey. This may be an effect of weathering since only exposed beds were observed, and marine sedimentary beds are known to change color with weathering (Bedrossian 1974). These beds have been tilted to an almost vertical attitude (dipping west), probably during subduction activity on the ocean floor (cf. ibid.). The thicker, more consolidated beds range from 4 to 10 cm wide. Laminae of roughly 1 cm are clustered together in bedsets up to 15 cm thick. Both types of beds have sharp contacts and uniform composition (cf. Collinson and Thompson 1989). The siltstones in these almost vertical beds exhibit joints or cracks horizontal to the present surface. These joints serve as foci for weathering processes, both chemical and physical. Water enters the rock through these joints, and solution and cracking due to wetting and drying lead to loss of material and rock fall (Selby 1982; Small and Clark 1982). The thinner laminae appear to break up more frequently and in smaller fragments. At the exposed surface, the thinner bedsets are more recessed, suggesting that they erode at a faster rate than do the thicker, more consolidated beds. The terrace top sediments and the colluvial deposits at the base of the terrace exhibit many features of this bedrock. The color is light brown to light yellow, larger particles are angular, predominantly sandstone and siltstone, and the matrix is composed primarily of clay and silt. Occasional inclusions of sand, pebbles, and large rounded cobbles match those of the beach below. TRANSPORT OF SEDIMENTS The sediments forming the terrace top and the colluvial terrace base originate primarily from the local mudstone, siltstone, and sandstone, yet the two settings have a significant difference in slope. Although both loci are currently vegetated in coastal grasses, the hillside appears less stable. The southeastem side of FRBS is a colluvium-covered hillside cut at the base by the Fort Ross Creek. Modem roads and footpaths crisscross the hill's slope. Numerous small slip and slump scars are evident. A recently formed gully extends down to the beach from the dirt road. Just above the FRBS archaeo- logical excavation, roughly 2 m above the beach base, a recent scar extends across the colluvium. It is unclear whether this scar is due to slumping or storm wave erosion. Sediment transport both on top of the terrace and downhill is often initiated and speeded up by rodent burrowing. The rodents break up subsurface sediments and bring them to the surface where they are susceptible to downhill movement by gravity as well as by rainfall impact. There is ample evidence in the excavation profiles of both FRBS and NAVS of old, filled-in rodent burrows. Many piles of churned-up sediment deposited by an active rodent community of the present day cover the land surface. In fact, rodents are a major agent in the slow process of sedimentation and downslope soil erosion along the entire northern California coastline (Black and Montgomery 1990; Bocek 1992; Erlandson 1984; Thom 1978). Even without help from rodents, tfhe fragments produced by differential weathering of the marine sedimentary rocks are gradually transported downhill by rainfall impact and gravity, and more quickly during events such as earthquakes or major storms (Griggs and Johnson 1979). Slumping and gullying also cause downhill move- ment of sediments. At least one slump episode is evident in the westernmost section of the FRBS profile. Gullies of different widths and depths are obvious on the present- day slope surface of FRBS. The effects of slumps and gullies on NAVS have been negligible. Finally, humans have caused (and continue to cause) the transport of sediment during road building, road use, and foot traffic on both sites. The uppermost stratum of some of the excavation units of FRBS exhibits a very large grained, angular, and loosely packed structure. This sediment dates to a major rerouting of the road in the 1920s. A new cut was made into the siltstone bedrock above the site, and the material was pushed over the edge of the terrace. EROSiON OF SEDIMENTS ONCE DEPOSITED Predominant local erosive agents include wind and episodic water action by the creek and ocean at the base of the colluvial deposit and wind. Our attention was initially drawn to the site because cultural materials were eroding out of the base of the terrace after major stonns. The terraces along the California coast erode at differing rates due to four major factors. The four are: the available wave energy (affected by the absence or Site Formation Processes 99 presence of a protective beach); the lithology of the sea cliffs (their resistance to erosion); the geologic structure including joints and folding; and the height of the sea cliff or terrace edge (Griggs and Johnson 1979). The southeastern portion of the Fort Ross terrace is relatively protected from major storm wave impact by a wide beach. Nonetheless, the presence of the large sandstone boulders along the south base of the terrace suggests that the area is a high-energy littoral environment These boulders were transported by wave action after having eroded out of the cliffside above. The erosive action of the creek is limited to episodic and probably only severe flood events. It is difficult to deternine the extent to which the terrace colluvium and archaeological deposits have been disturbed through wave and creek impact, and how much material has been lost through erosion. The FRBS excavations were conducted at the foot of the existing colluvium. The natural contour of the more gradual, eastern slope suggests that less than a meter of the colluvial toe has been eroded away. By contrast the western slope is cliff-like. Nonetheless, it would appear to have been more impacted by erosive agents since it seems to have lost at least 2 m of the colluvial toe. As mentioned above, terraces and hillsides along the northern California coastline are unstable and frequently exhibit small gully and slump scars. The Fort Ross Beach slope is no exception. At least two gully scars are evident on the present-day surface, and the FRBS excavation profiles show a distinct slump (filled with culturally sterile yellow clay with angular structure) on the western end, prior to Historic Period occupation. This feature occurs roughly a meter and a half below the present surface. This natural process is most likely responsible for transport of sediments from above to FRBS below. The terrace top shows no evidence of this form of erosional disturbance. In sum, wave and creek action have removed cultural deposits from across the length of FRBS. In addition, rodents and localized slumping and gullying have removed and/or redeposited small portions of deposits within this site. On the terrace top, NAVS is continu- ously subject to human traffic on foot and by car, strong coastal winds, and to a lesser extent, the gopher and rainfall activity observed on the hillside. WEATHERING OF THE SEDIMENTS IN SITU The archaeological profile across the length of FRBS offers additional information. The sediments in the upper strata are primarily siltstones and sandstones, clays and silts. These materials are particularly vulnerable to chemical and physical break down. The lowest stratum shows some additions of the beach gravels upon which the archaeological deposits rest. With the exception of these beach gravels, the majority of the particles through- out the profile are angular, showing little evidence of chemical weathering or long distance transport. These observations support the conclusion that sediments that form the colluvial deposit have not traveled very far from their origin, and in fact, derive from the terrace above. They have been deposited gradually, with the exceptions of one localized event of rapid deposition (road building) and another localized slump. Excluding the road building event in the 1920s and localized slumping and gullying, both NAVS and FRBS appear to have been formed by gradual accumulation of local sediments from the marine sedimentary bedrock and the beach gravels. PEDOGENIC ANALYSES GoALs The geomorphological study provides an understand- ing of the processes that have acted on the beach terrace and how the sedimentary matrix has accumulated. The pedogenic study is designed to explore questions of differential site use and development through time. I have chosen a series of qualitative and quantitative characteristics of the sediments that together form a comparative basis for identifying the relative affects of anthropogenic and non-anthropogenic pedogenic pro- cesses acting on the sites. No single measurement leads to clear-cut interpretive results. Rather, it is the combina- tion of measures taken together with the geomorphologi- cal and archaeological observations that add to an overall understanding of the site development. The chosen measures include color, alkalinity (pH), percent of organic carbon, percent of calcium carbonate, and structure. BACKGROUND AND EXPECTATIONS The following expectations derive from discussions in Birkeland (1984) and Courty et al. (1989). The Fort Ross soil is a relatively young, cumulative soil in a coastal setting. The soil of FRBS is somewhat thicker than that on the terrace top due to its position at the base of a steep slope and the consequently more rapid buildup of colluvium from above. In both site locations, color should be darkest in the top stratum as this is where organic matter is the most concentrated. Typically organic matter is limited to the upper stratum and causes dark-brown-to-black colors. However, middens or the organically rich deposits created by humans are also dark-brown-to-black and may occur in strata located beneath the top stratum. The presence of a substratum as dark or darker than the topsoil strongly suggests the presence of a midden. Relative soil acidity or alkalinity may not vary much between strata. This characteristic tends to develop over a longer period of time than we know the Fort Ross soil to have existed. However the pH is crucial for determin- ing the existence of preservation bias for faunal remains. 100 The Native Alaskan Neighborhood A neutral sedimentary matrix (pH in a range of 6.6 - 7.3) offers the best conditions for the preservation of organic matter such as bones and shell. Because we have already identified the presence of both middens and distinct bone beds, it is important to assess the pH across the site and at all depths to determine the extent to which these features reflect a past depositional event, or are simply a result of differential preservation. As with color, the organic carbon content should be highest in the top stratum. This is where micro-organ- isms live and process plant and animal remains as they are added to the soil. Relatively high organic carbon content elsewhere in the strata would suggest past human activity. Of course there are other potential causes, such as rodents burrowing down through lower strata and allowing organic carbon rich topsoil to be transported to a lower level. Thus organic carbon values must be interpreted in the context of profile drawings to help distinguish between filled-in rodent burrows and true middens. No significant leaching and reprecipitation of carbonates is expected in this area of low annual rainfall. Consequently in such a young soil, the calcium carbonate concentration should show little variation throughout the profile, though the fragmented shell present in the midden zones possibly causes calcium carbonate values to be enriched locally compared to those around the rest of the site. Pedogenic structure should grade from non-struc- tured or granular near the top to progressively more structured towards the base of the profile. This means that the top stratum should be composed of loose sediments with very few or no aggregates of sediments. Lower strata should exhibit aggregates of sediments. Anthropogenic strata, such as middens, generally show less structure than strata that have not been affected by human actions. Because FRBS is located at the base of a hillside and is part of a dynamic geologic and pedogenic system, new sediments are constantly being added from upslope. Compared to NAVS soil of similar age located on top of the terrace, the entire profile should be thicker (including the topsoil), percent of clays will be greater (as clays are not only forming from pedogenesis but are added by erosion from the sediments above), and the distribution of organic matter should be more constant throughout the profile. Of course, erosion and removal of some of the top stratum is also evident, and this should be kept in mind when interpreting the results of analysis. PEDOGENIC SAMPLING In order to get a representative picture of variation across both sites, samples are analyzed from several locations. Within each sample location, discontinuous systematic samples are taken. That is, at each location selected, a representative sample is taken from each identifiable stratum in the profile. Eight locations were sampled from FRBS. These include, from west to east, profile units P29, P23, P16, P15, P14, P6, and P2. The eighth unit (7S, 19W) comes from a 3 by 2 m2 block located in the Southwest Bench upslope from the Middle Profile. Seven units are sampled from NAVS. These include the South Central Test Unit (llOS, 11W); two units from the West Central Trench (75S, 16W and 75S, 20W); two units from the East Central Trench (75S, OE and 75S, 1E); and two units (125S, 21W and 125S, 24W) from the South Trench. These samples are chosen to represent a variety of depositional contexts across the terrace top site. The units from the West Central Trench represent the simplest depositional succession (topsoil, dark sandy loam, clay). The South Central Test Unit (lIOS, 11W) exhibits a relatively simple stratigraphic succession, including a layer of rock rubble between the dark sandy loam and the basal clay stratum. The East Central Trench and South Trench units represent perhaps the most complex deposits, including pit features, at least two midden strata, and lenses of concentrated faunal debris and artifacts (the bone beds). A Jones sample splitter is used in the lab to obtain smaller, representative samples of those collected in the field. Roughly a kilogram of sediment was collected for each sample in the field. All stages of the lab analyses together used a maximum of 10 grams from each sample. The remaining sediment is stored at the Archaeological Research Facility, U.C. Berkeley for future analyses. PEDOGENIC METHODS COLOR Sediment color, when compared throughout a profile, indicates the degree and type of pedogenic as well as anthropogenic processes that have acted on the deposits. The longer a soil has developed, the more distinctive the differences between horizon color. In a relatively undeveloped soil, color mainly serves to identify the parent material and relative amounts of organic matter. Given the marine sedimentary parent material and the lack of strong leaching, the expected range of colors is in the yellow-to-brown shades, with darker browns near the surface. In an anthropogenic soil, darker browns and blacks indicate the locations of middens, or concentrations of organic refuse that was deposited as either a direct or indirect effect of human behavior. The FRBS archaeo- logical profile shows an extensive midden that corre- sponds to the Russian occupation. The same is true for NAVS on the terrace top. Comparing color across the site should help to delineate the extent of this midden as well as to identify areas of greater or lesser concentration. All colors are determined in the lab with moistened Site Formation Processes 101 samples using the standard Munsell color chart. SoIL ACIDITY AND ALKALINITY The degree of acidity or alkalinity of a sediment (the pH) provides another basis for comparison. There is no one value that can be linked to a single cause or condi- tion, yet the relative acidity or alkalinity can serve as a guideline for judging preservation bias of bones or other organic materials. That is, when delimiting the extent of a midden, the possibility of differential preservation can be assessed by comparing the pH of the sediments. Sediment pH is measured in the lab using an automatic Beckman 32 pH meter. Samples are brought to stability in a 1:1 ratio with distilled water (20 g sample in a 20 ml distilled water solution). The pH value is determined when three successive readings are within .1 unit of each other. ORGANIC CARBON CONTENT A higher concentration of organic carbon is expected in the top stratum, as it is continuously added through roots, organisms living in the soil, tree leaf litter, etc. As organic carbon travels downwards through the profile it is broken down into a nonorganic form of carbon. The combination of constant addition at the surface and constant migration and breakdown beneath the surface maintains a gradient in a nonanthropogenic soil. Anthro- pogenic soils and especially middens tend to be enriched in organic carbon. The concentration of organic carbon within the same horizon across the site should help to delimit the extent of midden deposits. Organic carbon content is measured by the loss-on- ignition method as described in Dean (1974). The weight of a sample is recorded before and after exposure to 5500 C for one hour. This temperature is high enough to burn off organic carbon without burning carbon attached to inorganic compounds (calcium carbonate, for example). The difference between the before and after weights is expressed as a percentage of the total weight of the sample to give the percent of organic carbon present in the sediment. CALCIUM CARBONATE Concentration of calcium carbonate can be an indicator of the relative intensity of pedogenic processes. For example, if a soil has experienced intensive leaching, there should be a concentrated stratum of calcium carbonate just below the greatest depth to which water regularly saturates the profile. Because the Fort Ross soil is both relatively young and has experienced relatively little leaching, pedogenic differences in calcium carbon- ate are not expected. Variance in the calcium carbonate percentage across the site is more likely to reflect varying concentrations of shell associated with debris generated by human inhabitants of the sites. Therefore, this value is expected to be most useful in delimiting areas of concen- trated refuse disposal or some form of shell processing, especially when the shell has not been preserved as concentrations of large particles obvious to, and noted by, the observer in the field. Calcium carbonate content is measured by the loss- on-ignition method (Dean 1974). The same principle applies as for organic carbon, but the temperature necessary to burn off the calcium carbonate is 800? C. The samples are exposed to a temperature of 1000' C for one hour to ensure that the calcium carbonate is com- pletely burned off. The difference between the post-550? C burn weight and the post-G000' C burn weight is expressed as a percentage of the original total weight of the sample. This value is then divided by 0.44 to obtain the percentage of calcium carbonate present in the sample. PEDOGENIC STRUCTURE AND PARTICLE ANGULARITY Structure is a qualitative variable. Structure is described in the field and lab as the percentage of sediments that are aggregated, and the relative sizes and compaction of those aggregates. Particle angularity is described in the lab with the aid of microscopic inspec- tion. Structure is one way of assessing the degree to which pedogenesis has taken place. As a general rule, the uppermost stratum of a developed soil shows little structure and the sediment is rather loose and granular. Lower strata show aggregates of sediment. Anthropo- genic soils, even when they are in lower strata, tend to develop fewer aggregates. Particle angularity is a qualitative indication of the degree to which individual particles have been exposed to weathering processes such as chemical solution or physical abrasion. Both the colluvial deposit and the terrace top have two major sources of sediments, the siltstone bedrock and the beach gravels. In this study a qualitative assessment of the angularity of particles serves to corroborate conclusions made as to the source of sediments, and also as a baseline for comparison of post-depositional alteration across the site. PEDOGENIC RESULTS The results from each analysis are presented in table format for comparison between units and strata. The sample units are grouped on the basis of similar strati- graphic successions, but these successions are not always identical. The basic stratigraphic succession in both FRBS and NAVS includes topsoil, midden, clay, and bedrock. Topsoil refers to the uppermost stratum of sediment including the root zone. Midden is defined as "a deposit of artifactual debris and garbage in a primary archaeological context and/or sediment that is predomi- nantly made up of anthrogenic materials in a secondary 102 The Native Alaskan Neighborhood archaeological context" (Kolb et al. 1990:216). Clay has been used to describe the stratum between midden and bedrock composed primarily of clay particles and exhibiting little or no cultural and organic inclusions. The individual strata are described more thoroughly in the preceding chapters. For comparison between sample units the following explanation of discrepancies is provided. The 7S, 19W mottled brown clay was divided into midden 1 and midden 2 in the field on the basis of cultural inclusions. Midden 1 exhibited a much higher concentration of visible shell, bone, charcoal, and artifactual fragments. The highly mottled clay stratum of unit 7S, 19W was subdivided in analysis to look for any differences that might have developed vertically in this relatively thick stratum. The first value in the clay segment corresponds to 120 cm below surface and the second corresponds to 160 cm below surface, just above the bedrock. Unit P15 was not excavated below the midden. Units P6 and P2 did not exhibit a fill stratum. Due to the erosional action of storm waves, the westem units of FRBS had no topsoil remaining. The mottled clay stratum began just beneath the topsoil, and there was no evidence of a midden. The yellow clay of unit P23 corresponds to a lens of yellow clay with large angular fragments that appears to corre- spond to a localized slump event. No topsoil samples were taken for the NAVS units 75S, 16W and 75S, 20W. Only two units, 75S, OE and 75S, 1E, had bone bed strata between the midden and the clay. Hence, examination of the bone bed deposit characteristics occurs only in these two units of the East Central Trench. COLOR The hues are uniform across both sites (IOYR) due to the similarity of the parent material (table 4.1). The darker colors of NAVS and to the east at FRBS suggest a more intensive human use of those portions of the sites. They correspond to the midden zones that were identified in the field on the basis of darker color, more loosely packed sediments, as well as obvious bone, shell, and charcoal inclusions. SEDIMENTACIDITY AND ALKALINITY: PH In FRBS samples there appears to be a trend from west to east of slightly acidic sediments to neutral or slightly alkaline sediments (table 4.2). Bone is best preserved in aLkaline conditions, yet the difference in pH value from west to east is not great enough to suggest differential preservation of bone across the site. Variance in bone and shell concentrations observed in the field can then be interpreted as representing spatially distinct differences in the use or deposition of material at the site. In the NAVS samples there is a distinct correlation between concentrations of faunal remains and soil pH. The zones that exhibited a high concentration of faunal remains, such as the bone bed, show a neutral pH. The 75S, 16W and 75S, 20W units exhibit somewhat surpris- ingly acidic pH values, even those in association with the midden. In terms of differential use of the site, this suggests that in the West Central Trench area of NAVS a lesser concentration of organic debris was deposited. ORGANIC CARBON CONTENT Each unit sampled shows similar patterns of decreas- ing organic carbon content from the top of the profile to the bottom (table 4.3). This is to be expected in pe- dogenic development. Perhaps the more informative pattern is the variance in values from one unit to another Table 4.1 Color Values of Neighborhood Sediment Samples FRBS Units P29 P23 P16 P15 P14 P6 P2 7S, 19W Topsoil * * IOYR3/2 10YR3/2 IOYR3/2 1OYR2/2 10YR2/2 1OYR3/3 Fill * * 1OYR3/2 10YR3/3 IOYR3/3 * * 1OYR3/3 Midden * * IOYR2/2 1OYR2/2 IOYR2/2 1OYR2/1 1OYR2/1 IOYR2/1 Clay * * IOYR2/2 * IOYR3/2 10YR2/2 lOYR2/1 lOYR3/2 Mottled Brown Clay II lOYR3/2 10YR3/2 * * * * * * Yellow Clay IOYR3/2 1OYR4/3 * * * * * * Beach Gravel IOYR2/2 1OYR22 * * * * * * NAVS Units 75S, OE 75S, 1E l1OS, 11W 125S, 21W 125S, 24W 75S, 16W 75S, 20W Topsoil 1OYR2/2 1OYR3/1 IOYR2/1 IOYR2/1 10YR2/1 * * Midden 1OYR2/1 1OYR2/1 IOYR2/2 1OYR2/1 10YR2/1 10YR2/2 1OYR2/1 Bone Bed 1OYR2/1 1OYR3/1 * * * * * Clay 1OYR3/3 1OYR2/2 1OYR3/3 1OYR2/2 IOYR3/3 1OYR2/2 1OYR2/2 * This strata either did not exist or was not sampled. Munsell color key is as follows: 10YR4/3 = brown; 10YR3/1 = very dark grey; 1OYR3/3 = dark brown; 10YR2/2 = very dark brown; lOYR3/2 = veTy dark grayish brown; 10YR2/1 = black. Site Formation Processes 103 Table 4.2 pH Valuesfor Neighborhood Sediment Samples FRBS Units P29 P23 P16 P15 P14 P6 P2 7S, 19W Topsoil * * 5.9 5.5 5.8 7.0 6.2 5.6 Fill * * 6.6 5.9 6.6 * * 5.9 Midden * * 7.0 7.5 7.7 7.7 7.1 7.1 Clay * * 7.3 * 7.4 7.8 7.1 7.2 Mottled Brown Clay II 6.5 6.8 * * * * * * YellowClay 6.9 6.5 * * * * * * Beach Gravel 6.8 6.5 * * * * * * NAVS Units 75S, OE 75S, 1E 11OS, 11W 125S,21W 125S,24W 75S, 16W 75S, 20W Topsoil 5.9 6.1 6.4 6.6 6.8 * * Midden 6.0 6.5 6.8 7.3 6.7 4.9 4.9 Bone Bed 7.0 7.0 * * * * * Clay 6.1 6.4 6.8 7.1 7.0 5.3 5.4 * This strata either did not exist or was not sampled. The pH values are as follows: 4.5 - 5.0 = very strongly acid; 6.1 - 6.5 = slightly acid; 5.1 - 5.5 = strongly acid; 6.6 - 7.3 = neutral; 5.6 - 6.0 = medium acid; 7.4 - 7.8 = mildly alkaline. across the two sites. For example, P29, the westernmost unit in FRBS, shows relatively high organic carbon content for the clay strata. These values are not only high compared to other clay strata across the site, they are comparable to the organic carbon content of some midden samples. This suggests that although bone and shell and other human-introduced organic constituents are not immediately visible in the field, this steep section of FRBS is relatively rich in organics. This may be due to the hypothesized steady disposal of debris over the edge of the terrace from the inhabitants of NAVS above. CALCIUM CARBONATE CONTENT Calcium carbonate content varies little across FRBS (table 4.4). Nearly all of the samples range between 2.2 and 3.5%. Two samples, midden 1 of unit 7S, 19W and the midden of unit P14, yielded values nearly twice as high as the others, at 5.2% and 5.0% respectively. The higher concentration in the midden strata of these units is most likely attributable to the visibly higher concentra- tion of shell fragments. NAVS shows a base level of percent calcium carbonate at from 2.1 to 3.0 %. The elevated values associated with the midden and bone bed zones in the neighboring 75S, OE and 75S, lE units are attributable to noticeably higher concentrations of marine shell debris. PEDOGENIC STRUCTURE Initial inspection of the samples showed that they contained various proportions of three major categories of particles. These include rounded grains of sizes matching those found in the beach sands and gravels, Table 4.3 Organic Carbon Content of Neighborhood Sediment Samples FRBS Units P29 P23 P16 P15 P14 P6 P2 7S, 19W Topsoil * * 11.2 13.0 12.4 6.7 6.2 9.4 Fill * * 7.1 6.9 7.0 * * 6.8 Midden * * 6.4 6.5 6.4 4.8 6.5 6.8 Clay * * 4.8 * 4.2 5.0 5.8 5.9 Mottled Brown Clay II 7.6 6.6 * * * * * * Yellow Clay 7.3 5.5 * * * * * * Beach Gravel 6.9 4.8 * * * * * * NAVS Units 75S, OE 75S, 1E 11OS, 11W 125S, 21W 125S, 24W 75S, 16W 75S, 20W Topsoil 9.7 10.4 9.5 12.7 11.8 * * Midden 9.3 8.7 7.8 10.3 11.8 8.3 9.0 Bone Bed 7.8 8.1 * * * * * Clay 5.5 6.2 5.9 8.8 5.4 5.8 5.2 * These strata either did not exist or were not sampled. Values are percentages of total sample. 104 The Native Alaskan Neighborhood predominantly composed of quartz. The remaining particles are composed of siltstone and sandstone particles of varying sizes. Some show evidence of rounding through weathering. All topsoil samples show a few small aggregates but are primarily composed of loose sediments. The FRBS clays, from the mottled clay of the west to the below midden clay of the east, show predominance of aggre- gates that are composed primarily of angular colluvial fragments and are quite compact or hardened. The beach gravels show few aggregates and are composed predomi- nantly of rounded particles. The fill of the FRBS profile to the East Profile shows few aggregates and a predomi- nance of large angular siltstone and shale fragments, reflecting the rapid deposition of debris during the road building event in the 1920s. The middens of the East Profile show some aggregates, inclusions of bone and shell, and a mixture of angular and rounded particles. This suggests human-related transport of gravels and sands from the beach. Unit P2 diverges from the general trends described above. Large aggregates and particle rounding are found in the top stratum. This could be due to erosion and removal of the uppermost layer. The midden has yielded no visible bone or shell. The aggregates in the clay stratum are the most hardened of all clays across the site. The position of this unit, at the most level and most accessible portion of the site, probably has allowed heavier use and even vehicular traffic by humans during and since the time of the Russian occupation. With the exception of the topsoil, the strata of NAVS are also quite compacted. This is most likely due to the heavy and continuous traffic across this relatively level terrace top, both before and during the Russian occupa- tion and up to the present day. The only exception are the loosely packed and granular zones of rodent activity. SUMMARY AND DISCUSSION FRBS is located within a colluvial deposit that has gradually accumulated at the foot of the primary beach terrace immediately south of the Stockade. The parent materials are of predominantly local origin, including siltstone, sandstone, and beach sands and gravels. Other materials include bone, shell, and artifacts of stone, ceramic, glass, and metal from prehistoric through present day occupations of the terrace above and FRBS below. The predominant erosional forces acting on FRBS include gopher activity, winds, storm-driven ocean waves, and flood-level creek waters. The winds and water have also contributed to weathering of the sedi- ments. Both the terrace top and the colluvial soils are relatively young and have not undergone significant pedogenesis. Sedimentary analysis supports the observa- tion and delineation of a midden stratum running across FRBS and a midden in some portions of NAVS. This midden varies in color, pH, organic carbon content, and calcium carbonate content across the sites. In chapter 2, Lightfoot and Schiff suggest that the western, near-ocean segment of FRBS (represented here by P29 and P23) represents the gradual disposal of culturally generated debris over the edge of the cliff-like terrace. The artifactual and faunal materials are sparse and randomly dispersed throughout the steeply sloped colluvium. The results of the present analyses suggest that the pit feature (in units P16, P15, and P14) is in primary context, while the remainder of the eastern portion of FRBS (units P6 and P2) consists of a concen- trated and thick midden deposit that could have been formed by more intensive dumping of debris from the NAVS occupation above. This midden also may have been formed through direct occupation of this more Table 4.4 Calcium Carbonate Content of Neighborhood Sediment Samples FRBS Units P29 P23 P16 P15 P14 P6 P2 7S, 19W Topsoil * * 2.6 2.3 2.4 3.2 2.2 2.7 Fill * * 2.8 2.8 2.7 * * 3.0 Midden * * 2.8 2.8 5.0 3.0 3.6 5.2 Clay * * 2.6 * 2.4 2.9 2.6 2.7 Mottled Brown Clay II 2.8 2.8 * * * * * * Yellow Clay 3.0 3.6 * * * * * * Beach Gravel 2.8 2.7 * * * * * * NAVS Units 75S, OE 75S, 1E 110S, 11W 125S, 21W 125S, 24W 75S, 16W 75S, 20W Topsoil 2.3 2.3 3.0 2.9 2.8 * * Midden 3.4 5.7 2.6 4.6 3.8 2.1 2.1 Bone Bed 4.4 3.2 * * * * * Clay 2.1 2.2 2.7 3.3 2.4 2.1 2.0 * These strata either did not exist or were not sampled. Values are percentages. Site Formation Processes 105 gently sloping portion of the terrace base colluvium. The geomorphological and sedimentological analyses support these interpretations of FRBS. The extent of the midden was better defmed through color and organic carbon content analyses. The relatively uniform pH values establish that bone and shell preserva- tion was more or less unifonn throughout the site, and that the differential concentrations can be interpreted as anthropogenic in origin. Both the fill and midden of the eastern portion of FRBS show relatively little structure and the sediments include a higher percentage of artifac- tual materials, shell, and bone. They are also less compacted than the mottled clays of the westem end of the site. The eastern clay, or lower stratum, is relatively enriched in organic materials, most likely transferred from the midden above by natural pedogenic processes through time, such as leaching. The differences across the site are not just due to differences in natural pe- dogenic development caused by the topographic varia- tion. Rather these differences are due to differential human use of the FRBS, and in the differential distribu- tion of materials from NAVS above. The NAVS values support the archaeological interpretation of intra-site differences. The horizontal concentration of fire-cracked rock, bone, and shell debris in the bone bed stratum (20 cm to 30 cm below the surface) of 75S, OE and 75S, 1E is correlated with darker color, neutral pH, and relatively enriched levels of calcium carbonate. Although evidence of gopher activity and even partial gopher skeletons exist, the bone bed feature does show integrity. That is, this deposit shows little pedogenic evidence of significant post-depositional disturbance. For similar reasons, the same can be said for the pit feature at FRBS observed in units P16, P15, and P14. CONCLUSION Since the Fort Ross Archaeological Project is ongoing, it has been the purpose of this study to establish a baseline description and understanding of local pro- cesses as well as to reconstruct the depositional history of FRBS and NAVS. A second goal has been to establish baseline measurements of depositional and pedogenic variables across both sites so that tfie degree and type of deposition could be better understood. Most importantly, the study has determined that the sediments and archaeo- logical materials have not experienced differential preservation due to differential soil acidity. The diverse yet complementary analytic approaches employed in the present study answer specific questions about site formation, especially with regard to the Russian occupation. The relative contributions of natural processes-pedogenic, geomorphological, and biologi- cal-to the formation of the site were not immediately obvious. During the Russian occupation, the entire FRBS profile exposed in the archaeological excavations shows evidence of past human activities. Although the western portion of FRBS shows little more than intermit- tent dumping from the terrace above, the pit feature and the eastem portion were more intensively and more directly used. The bone beds of NAVS attest to an equally or perhaps more intensively used area of the Native Alaskan Neighborhood. Although there is ample evidence of rodent disturbance, the concentrations of bone and other debris can be attributed more securely to past human actions rather than to differential preserva- tion. Future archaeological investigation in the immediate area of the Stockade will benefit from this initial study. For example, the results of the present study suggest that bone preservation in the immediate area is quite good, that the soils are relatively undeveloped and have experienced very little alteration due to pedogenic processes, but that rodent activity along with gullying and slumping have been and will continue to be disrup- tive to the archaeological deposits. ACKNOWLEDGEMENTS I would like to acknowledge the technical guidance of Dr. Kent Lightfoot, Dr. William Dietrich, and Dr. Ronald Amundson. Materials studied were collected during the course of the 1988, 1989, 1991, and 1992 U. C. Berkeley Field School seasons. Denise Boyce, Elizabeth Carson, Darren Moore, and Stacy Richardson carried out some of the sediment analyses as students in Lightfoot's Archaeological Laboratory Analysis course (Anthropology 134). Sheelagh Frame assisted in the geomorphological field analysis and served as field photographer. Aron Crowell gave invaluable advice in the collection of samples. Ann Schiff provided helpful editorial suggestions. REFERENCES Bedrossian, T. L. 1974 Geology of the Marin Headlands. California Geology April 1974:75-88. Birkeland, P. W. 1984 Soils and Geomorphology. Oxford University Press, Oxford. Black, T. A., and D. R. Montgomery 1990 ms. Sediment Transport by Pocket Gophers, Marin County, California. Bocek, B. 1992 The Jasper Ridge Reexcavation Experiment: Rates of Artifact Mixing by Rodents. American Antiquity 57(2):261-70. 106 The Native Alaskan Neighborhood Collinson, J. D., and D. B. Thompson 1989 Sedimentary Structures, 2nd Ed. Uniwin and Hyman, London. Courty, M. A., P. Goldberg, and R. MacPhail 1989 Soils and Micromorphology in Archaeology. Cambridge University Press, Cambridge. Dean, W. E. Jr. 1974 Detemination of Carbonate and Organic Matter in Calcareous Sediments and Sedimentary Rocks by Loss on Ignition: Comparison with Other Methods. Journal of Sedimentary Petrology 44(1):242-48. Erlandson, J. M. 1984 A Case Study in Faunalturbation: Delineating the Effects of the Burrowing Pocket Gopher. American Antiquity 49:785-90. Griggs, G. B., and R. E. Johnson 1979 Coastline Erosion: Santa Cruz County. California Geology April 1979:67-75. Kolb, M. F., N. P. Lasca, and L. G. Goldstein 1990 A Soil-Geomorphic Analysis of the Midden Deposit at the Aztalan Site, Wisconsin. In Lasca, N. P., and J. Donahue, editors, Archaeological Geology of North America. Special edition of the Geological Society of America, vol. 4, pp. 199-230. Selby, M. J. 1982 Hillslope Materials and Processes. Oxford University Press, Oxford. Small, R. J., and M. J. Clark 1982 Slopes and Weathering. Cambridge University Press, Cambridge. Thorn, C. E. 1978 A Preliminary Assessment of the Geomorphic Role of Pocket Gophers in the Alpine Zone of the Colorado Front Range. Geografiska Annaler 60 A (3-4):181-87. United States Department of Agriculture 1972 Soil Survey of Sonoma County, California. Sheet #60. Wagner, D. L., and E. J. Bortugro 1982 Geologic Map of the Santa Rosa Quadrangle, California, 1:250,000. Williams and Heinz Map Corpora- tion, Maryland. Walker, R. G., ed. 1979 Facies Models. Ainsworth Press Limited, Ontario.