Chronology of Archaeological Deposits from the Fort Ross Beach and Native Alaskan Village Sites KENT G. LIGHTFOOT AND STEPHEN W. SILLIMAN HE TEMPORALLY DIAGNOSTIC European/Asian later material culture, the contextual relationship of the 1 artifacts from FRBS and NAVS date primarily to the historic assemblage to remains that predate the Russian Russian occupation of Fort Ross, from A.D. 1812 to 1841. occupation is not so straightforward. Some archaeologi- In his analysis of the ceramic artifacts (chapter 7), cal deposits, such as the clay stratum at FRBS, contain Silliman concludes that most of the diagnostic refined only lithic artifacts that may be prehistoric in age, an earthenwares are handpainted blue, transferprint blue, observation made by Schiff in chapter 9. In other and handpainted polychrome types manufactured deposits characterized by diverse cultural remains, it is primarily in the 1820s and 1830s. The ceramic assem- not clear whether the lithic artifacts found in direct blages of FRBS and NAVS are quite similar, suggesting a association with temporally diagnostic European/Asian very close chronological relationship between the two artifacts were manufactured and used in the early 19th sites. Ross's study of the glass and ceramic beads century or many years before the Russians settled Fort (chapter 8) identifies the majority as inexpensive, drawn Ross. embroidery beads that were manufactured in Europe in Most archaeologists who have worked in or near the the early 19th century. The one notable exception is the Russian Stockade have unearthed lithic artifacts (both Prosser-molded ceramic bead from FRBS that was chipped and ground stone) that appear to be prehistoric in manufactured after 1840 (FRBS 6/15/89-7-be-1). age (see Smith 1974; Treganza 1954:18). The recent Although the sample of beads from FRBS is small excavations of the east wall of the Stockade by Purser et (n=16), most of the FRBS bead types are found in the al. (1990:42-52) unearthed chert and obsidian tools, as larger assemblage at NAVS (n=548), again indicating a well as a rock feature, at depths of 50 to 70 cm or more comparable bead chronology for the two sites. below surface. In using the obsidian hydration chronol- The above findings indicate that European/Asian ogy generated for the southern North Coast Ranges (as materials postdating the Russian Period are uncommon in outlined below), they date the obsidian artifacts from the both FRBS and NAVS archaeological contexts. As Farris Lower Archaic Period (6000 B.C.-3000 B.C.) to the Upper emphasizes in chapter 6, this is an important observation Emergent Period (A.D. 500-A.D. 1500). The hydration because the archaeological deposits unearthed in the band measurements of 13 obsidian artifacts from the Russian Stockade contain historic materials dating to the 1989 surface collection of NAVS also suggest prehistoric later American Period (pOSt A.D. 1846), when the remnant dates ranging from the Upper Archaic Period (1000 B.C.- Stockade buildings were used as a hotel, saloon, dance A.D. 500) to the Upper Emergent Period. Only one hall, and storage facilities. Consequently, the excavations obsidian artifact is probably historic in age (Lightfoot et at FRBS and NAVS are among the very few at Fort Ross al. 1991:109). These findings are consistent with other to have produced a discrete assemblage of historic archaeological sites on raised marine terraces in the Ross materials that date almost exclusively to the Russian Region, where we detected broadly dispersed, low- occupation. density lithic scatters that were dated as early as 8000 to While there is little evidence at either FRBS or 6000 years ago (Lightfoot et al. 1991: 110-12). NAVS for thie mixing of early 19th century artifacts with Smith (1974) and others argue that the majority of 338 The Native Alaskan Neighborhood the chipped stone and ground stone artifacts recovered in Annadel obsidian into prehistoric and historic periods. the archaeological deposits at Fort Ross were produced, The regression equation is based on obsidian readings used, and discarded along the marine terrace before the from radiocarbon dated contexts in six sites in the Russians and their multiethnic work force founded the southern North Coast Ranges. Origer's (1987) study colonial outposL The common occurrence of projectile provides a rough approximation of the hydration band points, flakes, shatter, and grinding and hand stones with measurements of Annadel obsidian in microns for the 19th century artifacts is explained by the mixing of following periods: prehistoric and historic materials in disturbed contexts, such as those created by rodent bioturbation (e.g., Purser Lower Archaic (6000 B.C.-3000 B.C.) 6.6-5.3 microns et al. 1990:42). This observation has serious implications Middle Archaic (3000 B.C.-1000 B.C.) 5.2-4.1 microns for our study of culture change and continuity of the Upper Archaic (1000 B.C.-A.D. 500) 4.0-2.9 microns Native Alaskans and Native Californians at NAVS and Upper Archai (1000 B0C-A.D. 500) 4.0-19 microns FRBS. If the majority of the lithic assemblage described Lower Emergent (A.D. 500-A.D. 1500) 2.-1.7 microns in chapter 9 is not contemporaneous with the Russian UpperEmergent(A.D. 1500-A.D. 1812) 1.61.0microns outpost but dates to prehistoric times, then we may Historic (post A.D. 1812) < 1.0 microns greatly overstate the degree of continuity and/or adoption of traditional Native Californian cultural practices by The Historic Period is further divided into three phases including these artifacts in our analysis of the material that are too fine grained for the current obsidian hydra- culture of the Native Alaskan Neighborhood. tion chronology: This chapter presents the chronology of specific Russian Period (A.D. 1812-l&1) archaeological deposits at FRBS and NAVS and consid- ers the temporal relationship of the lithic artifacts to the Mexican Period (A.D. 1841-1846) early 19th century assemblage of ceramic, glass, and American Perod (post A.D. 1846) metal artifacts. The chronological assessment is based on 172 obsidian hydration measurements, 2 Accelerator The above chronology, we recognize, will continue Mass Spectrometry (AMS) radiocarbon dates, and the to be refined as our understanding of the obsidian contextual association of these dated materials with hydration rates of the five sources, especially under diagnostic European/Asian artifacts. different temperature, moisture, and depositional condi- tions in coastal environments, becomes more sophisti- SOUTHERN NORTH COAST RANGES OBSIDIAN cated (see Lightfoot et al. 1991:67). For the purposes of HYDRATION CHRONOLOGY differentiating historic from prehistoric obsidian artifacts, however, we feel relatively confident that thin hydration The obsidian hydration chronology for the southern band readings of ca. 1.0 microns accurately discriminate North Coast Ranges has been established for four of the young, or possibly historic, artifacts from older and principal obsidian sources in the region: Annadel near clearly prehistoric artifacts with thicker hydration bands Santa Rosa, Borax Lake and ML Konocti near Clear of 2.0 or more microns. Origer's (1990) recent study of Lake, and Napa Valley (see Fredrickson 1987, 1989; debitage from Ishi's obsidian knapping sometime around Jackson 1989; Tremaine and Fredrickson 1988; Tremaine 1915 shows that detectable and measurable hydration 1989; Origer 1987; and Origer and Wickstrom 1982). rims form in less than 100 years. His results suggest that Employing Tremaine's (1989:69-70) "comparison the mean hydration band readings of Napa Valley, ML constants" derived from induced obsidian experiments, Konocti, and Borax Lake obsidians used by Ishi, when hydration band measurements of obsidians from these standardized to the hydration rate of the Annadel flow, different sources can be compared with one another. In vary between .45 to .58 microns. this chapter, comparison constants are calibrated to the The obsidian samples from FRBS and NAVS were hydration rate of the Annadel flow by multiplying Napa analyzed by the Obsidian Hydration Laboratory, Sonoma Valley and Mt. Konocti readings by .77 and Borax Lake State University, under the direction of Thomas Origer. measurements by .62. A fifth obsidian source has Obsidian artifacts were visually identified to one of the recently been defined by Jackson (1986) as Franz Valley above four sources based on macroscopic attributes. located 15 km north of Santa Rosa. A comparison Thin sections were prepared for one or more edges of the constant has not yet been calculated for this source. artifacts, and six measurements of the hydration band Obsidian hydr.ation is generally best used as a were taken at several locations along the edge of each relative dating methiod. However, since our primary thin section. The mean of these six readings was then purpose is to distinguish prehistoric obsidian artifacts calculated; thiis figure was used for chronological from those manufactured and used in the early 19th purposes. century, we employ Origer's (1987:55-59) regression In addition to visual sourcing, a large sample of the equation to assign the hydration band measurements of obsidian specimens was sourced by Silliman using thie Chronology of Deposits 339 energy dispersive x-ray fluorescence (EDXRF) spectrom- rely on macroscopic identifications only for those eter in the Department of Geology and Geophysics at the specimens for which no EDXRF data is available. All University of Califomia, Berkeley. The EDXRF trace hydration readings reported are comparison constants element analysis identified the chemical characterization calibrated to the hydration rate of the Annadel source for each obsidian sample based on trace elements with the exception of Franz Valley specimens. Infonna- [thorium (Th), rubidium (Rb), strontium (Sr), yttrium (Y) tion on the catalog number, unit, level, archaeological and zirconium (Zr) ppm concentrations] that have proven deposit, visual source, chemical source, hydration useful for sourcing northern California obsidians (Jack- reading, and comparison constant for each obsidian son 1986). The geochemical signature for each sample specimen analyzed is presented in appendix 16.1 (FRBS). then was assigned to an obsidian source through compari- The results of the EDXRF trace element analysis are son with source characterization values outlined in presented in appendix 16.2 (FRBS). Jackson (1986). Trace element analysis identified the Table 16.1 presents the summary statistics for the five major sources (Annadel, Franz Valley, Napa Valley, FRBS obsidian artifacts. The majority are sourced as Borax Lake, and Mt. Konocti). Annadel (52%) and Napa Valley (35%), followed by AMS radiocarbon dates were provided by Beta Borax Lake (6%), Mt. Konocti (3%), and Franz Valley Analytic Inc. whose technicians pretreated the charcoal (3%). One specimen's source is unknown. Of the specimens and submitted them for analysis to the sourced obsidian artifacts that yielded interpretable Accelerator Mass Spectrometry Facility at the Lawrence hydration band readings (n=74), designated as "Total Livermore National Laboratory (CAMS) in California. Hydration" in table 16.1, all but one are prehistoric in C13/C12 corrections were applied to the conventional age. The mean hydration readings for the five sources C14 age. The Pretoria Calibration Procedure program range from a low of 1.9 microns (sd=O) for Mt. Konocti was employed by the Beta Analytic staff to convert B.P. to a high of 2.5 microns for Napa Valley (sd=.6) and radiocarbon age determinations into calender years. Franz Valley (sd=O). The chronology of FRBS deposits is outlined below for the five areas of the site (East FORT Ross BEACH SITE Profile, Central Profile, West Profile, Southwest Bench, The sampling design to select obsidian specimens for and East Bench). analysis involved stratifying FRBS into five areas (East Profile, Middle Profile, West Profile, Southwest Bench, EASTPROFILE and East Bench) and then selecting specimens, when Five obsidian artifacts were submitted for hydration available, from the different archaeological deposits band measurements and sourcing from the East Profile (4 represented in each area as described in chapter 2. from P5, 1 from P9). Hydration bands were measured on Samples were usually selected from the stratigraphic four flakes, identified as Annadel (2) and Borax Lake (2) profiles of two to four excavation units in each area that obsidians (table 16.2). Three of the artifacts were contained the largest number of obsidian artifacts. When recovered from the clay stratum and one from the midden multiple specimens were available from the same area deposit The average hydration reading for artifacts from and deposit, we chose a sample that characterized the the clay stratum is 2.1 microns (sd=.4), while the single diversity of lithic categories present. In selecting the hydration band measurement from the midden is 1.6 obsidian samples, we were particularly interested in microns. Given the small sample size, little can be said comparing obsidian readings from the midden and clay about the chronology of the East Profile. The midden deposits. deposit contains diagnostic ceramic and glass artifacts Eighty-four obsidian specimens were submitted to suggesting an early 19th century date, but the one the Obsidian Hydration Laboratory at Sonoma State obsidian reading suggests a prehistoric age. The clay University. Eighty-two were assigned a source through deposit probably predates the establishment of Fort Ross. visual inspection, and of these, 74 exhibited hydration It contains no diagnostic historic remains, and the bands that could be measured with precision. obsidian artifacts present have hydration readings ranging We submitted 64 of the visually sourced specimens from 1.7 to 2.6 microns dating them to the Lower (78%) for EDXRF trace element analysis, the results of Emergent Period (see table 16.2). which indicate that only 6 specimens (7%) had been misidentified. Three specimens macroscopically identi- MIDDLE PROFILE fled as Annadel are from Franz Valley (n=2) and Napa Seventeen obsidian artifacts (16 flakces and 1 biface Valley (n=l1); two believed to be from Napa Valley are fragment) from the Middle Profile (6 from P12, 3 from chemically sourced as Annadel and Franz Valley; while P14, 5 from P15, and 3 from P16) were submitted for one classified as Borax Lake should be Napa Valley. For analysis. Sixteen of the specimens yielded interpretable subsequent analyses of hydration band readings, particu- hydration band measurements (table 16.3). The majority larly the calculation of comparison constants, we use the are sourced as Annadel (12), followed by Napa Valley obsidian sources identified by the EDXRF analysis and (2), Mt. Konocti (1), and EBorax Lake (1). The sample 340 The Naiive Alaskan Neighborhood includes obsidian artifacts from the midden (2), the into this historic age deposit. This stratum was then midden/clay interface (1), the clay deposit (3), and the capped by rock fill produced from the 1920s construction fil (8) and floor (2) of the pit feature. All of the and rerouting of the Fort Ross Cove Road located deposits contain prehistoric lithic material based on directly above the FRBS deposits. hydration readings. The mean hydration readings for The FRBS Pit Feature also contains a mixture of the midden, midden/clay interface, and clay deposits are prehistoric and historic materials. The mean hydration 1.8 (sd=0), 2.3 (sd=0), and 1.9 (sd=.1) microns, readings for the pit fill and floor are 2.5 (sd=1) and 2.1 respectively. Again, the paucity of European/Asian (sd=.3) microns, respectively. The large standard historical materials in the clay deposit and the obsidian deviation in micron readings for the pit fill suggests an hydration readings ranging from 1.7 to 2 microns extensive time span for the obsidian artifacts, dating to as suggest a prehistoric date, probably sometime during the early as the Middle Archaic Period (4.6 microns) and the Upper Emergent Period. The midden stratum probably late Upper Emergent Period (1.3 microns). The recovery was formed in the early 19th century when many types of early 19th century artifacts and faunal remains, of ceramic and glass artifacts dating to the Russian including pinnipeds associated with Native Alaskan occupation were deposited (see chapter 7). The hunting (see chapter 12), suggests that the infilling of the presence of prehistoric lithics with hydration readings pit depression took place during or even after the Russian greater than 1.0 microns in this stratum, however, occupation. A .05 g charcoal sample recovered from the suggests that precontact artifacts were also integrated floor of the pit feature was submitted to Beta Analytical Table 16.1 Fort Ross Beach Site Obsidian Hydration Summary Statistics Source Total Sourced Total Hydration Hydration(inmicrons) # % # % mean sd min max Anradel 43 52 40 53 2.1 .6 1.1 4.6 BoraxLake 5 6 5 7 2.3 .5 1.7 2.9 Franz Valley 2 3 - 2 3 2.5 0 2.5 2.6 Mt. Konocti 2 3 2 3 1.9 0 1.8 1.9 NapaValley 29 35 25 34 2.5 .6 .9 3.7 Uromwn 1 1 Total 82 74 2.3 .6 .9 4.6 Table 16.2 Hydration Readings for the East Profile by Source, Deposit, and Artifact Type COunt % Hydration(inmicrons) mean sd min max Source Annad 2 50 2.1 .5 1.6 2.6 BoraxLake 2 50 1.8 .1 1.7 1.9 Deposit iden 1 25 1.6 0 1.6 1.6 aay 3 75 2.1 .4 1.7 2.6 Type BifaceThinning Flake 2 50 1.6 0 1.6 1.6 InteriorFlake 1 25 1.9 0 1.9 1.9 Secondary Cortical Flake 1 25 2.6 0 2.6 2.6 Chronology of Deposits 341 Inc. for AMS radiocarbon dating. The conventional prehistoric deposits. No European/Asian remains were radiocarbon age is 380 +/- 80 B.P., which, when calibrated recovered here, and, while the sample size is quite small, to a calender date, yields an age of cal A.D. 1410 to 1670 all of the obsidian hydration readings are clearly prehis- (2 sigma, 95% probability). The pit feature was probably toric in age. constructed shortly before or during the Russian occupa- In contrast, the mottled brown clay stratLm was tion and subsequently filled with both prehistoric and probably formed during the Russian Period. It contains a early 19th century artifacts. wide assortment of diagnostic early 19th century ceramic and glass artifacts as well as a diverse range of obsidian WFSitPROFIEa hydration band measurements (.9 to 2.8 microns) Sixteen obsidianpc specin were selected from suggesting that the chronology extends from the Lower across the stratigraphic strata in the West Profile for Emergent to Historic periods. These hydration readings analysis (1 from P20, 10irom P21, 10fromkP26,3 from document that prehistoric obsidian artifacts have been P27, and 1 from P28). Thirteen artifacts (10 fRakes, 1 deposited in the same context as European/Asian shatter, 1 core fragment, and 1 projectile point fragment) materials and at least one historic age obsidian artifact. produced interpretable hydration band measurements The youngest obsidian artifact is a formal tool-a (table 16.4). The majority are identified as Napa Valley notched projectile point fragment (FRBS 6/19/89-17-L- (8), a significant deviation from the predominance of 1)-with a hydration band measurement of .9 microns. Annadel obsidian found in other FRBS locations. Only Notched projectile points are believed to be diagnostic two artifacts are sourced to Annadel, while the remainder markers of the Upper Emergent or Historic periods (see are characterized geochemically as Borax Lake (1), Franz Schiff, chapter 9). Again, as in the Middle Profile, the Valley (1), and ML Konocti (1). Obsidian artifacts were mottled brown clay has been capped with rock fill from selected for analysis from the mottled brown clay (9), the construction and rerouting of the Fort Ross Cove beach gravel (2), and yellow clay (2). Again, all of these Road in the 1920s. deposits contain prehistoric lithics based on hydration readings. The mean hydration reading for the mottled SOuTHWESTBLANCH brown clay is 2.2 (sd=.6) microns, while the mean A large sample of obsidian was selected for analysis hydration measurements for the beach gravel and yellow from across the stratigraphic sequence of the Southwest clay are 2.5 (sd=0) and 2.8 (sd=0), respectively. Bench from two units (14 from 7S, 17W and 24 from 8S, The beach gravel and yellow clay appear to be 19W). Thirty-six obsidian artifacts (31 flakes, 4 pieces of Table 16.3 Hydration Readings for the Middle Profile by Source, Deposit, and Artifact Type GCut % Hydration (in microns) mean sd min max Soure Annad 12 75 2.3 .8 1.6 4.6 BoraxLake 1 6 2 0 2 2 Mt. Konocti 1 6 1.8 0 1.8 1.8 NapaValley 2 13 1.8 .5 1.3 2.3 Dqesit Nden 2 12.5 1.8 0 1.8 1.8 dfdeni/Cay 1 6 2.3 0 2.3 2.3 Cay 3 19 1.9 .1 1.7 2 Pit Feature Fill 8 50 2.5 1 1.3 4.6 Feature Floor Fill 2 12.5 2.1 .3 1.8 2.4 Type Biface 1 6 1.9 0 1.9 1.9 BifaceThinningFlake 6 37.5 2.1 .5 1.7 3.3 EdgeModifiedFlake 1 6 2.4 0 2.4 2.4 InteriorFlake 6 37.5 2.4 1.1 1.3 4.6 SecondaryCorticalFlak