Implications of a freshwater radiocarbon reservoir correction for
the timing of late Holocene settlement of the Elk Hills,
Kern County, California
Brendan J. Culleton a,b,*
a Department of Anthropology, University of Oregon, Eugene, OR 97403, USA
b Pacific Legacy, Incorporated, 1525 Seabright Avenue, Santa Cruz, CA 95062, USA
Received 1 October 2005; received in revised form 21 January 2006; accepted 24 January 2006
Abstract
Uncertainties regarding the magnitude of freshwater radiocarbon reservoir effects can introduce random errors into dates on archaeological
freshwater carbonates. As a result, many archaeologists avoid dating freshwater shells unless no other datable materials are available. The
chronology of prehistoric occupation of the former Naval Petroleum Reserve No. 1 (NPR-1) at Elk Hills, Kern County, has been established
with 50 radiocarbon dates on freshwater mussels (Gonidea and Anodonta sp.). Characterization of any freshwater radiocarbon reservoir effect
is crucial for the accurate interpretation of inferred settlement and subsistence changes on the Elk Hills. Paired charcoal and freshwater mussels
sampled from closely associated contexts were dated to identify a freshwater reservoir effect. Paired Anodonta and Gonidea sp. shells were dated
to investigate interspecific differences in fractionation. Results indicate that a 340 C620 14C yr correction should be applied to conventional 14C
dates on freshwater carbonates in the Buena Vista Basin before calendar calibration. Evidence of interspecific differences is inconclusive. Dates
recalibrated with the reservoir correction indicate that widespread occupation of the Elk Hills is correlated with increasing precipitation towards
the end of the Medieval Climatic Anomaly and during the Little Ice Age, suggesting that slough resource exploitation may have been driven by
regional population pressure rather than drought-related declines in aquatic productivity.
C211 2006 Elsevier Ltd. All rights reserved.
Keywords: Freshwater Radiocarbon Reservoir; Anodonta sp.; Gonidea angulata; Medieval Climatic Anomaly; Elk Hills; Southern San Joaquin Valley
1. Introduction
Radiocarbon dating provides archaeologists with absolute
temporal controls for developing regional cultural and
environmental chronologies. Archaeologists are also well-
positioned to address the potential interpretive problems asso-
ciated with dating various materials from different radiocarbon
reservoirs [59]. Terrestrial plant carbon is usually the most
reliable material, though the ??old wood effect?? can introduce
errors [40,46], and differences in isotopic fractionation
between C3 and C4 plants must be considered [13,53]. Marine
radiocarbon reservoirs vary geographically and temporally,
potentially introducing random errors that can only be
accounted for by direct 14C-dating of pre-bomb shells or
comparison of archaeological marine shell and terrestrial
charcoal [5,22,31,32,39,60]. Characterization of freshwater
reservoir effects has received less attention from archaeolo-
gists, and it has been suggested that shells of freshwater organ-
isms should only be dated if no other materials are available,
unless a reservoir study has been conducted [59:53]. Such has
been the case with archaeological investigations conducted on
the former Naval Petroleum Reserve No.1, Elk Hills, in the
southern San Joaquin Valley of California, where the shells
of two freshwater mussel taxa Gonidea angulata and
Anodonta sp. are the most visible archaeological remains at
* Current address: Department of Anthropology, University of Oregon, Eu-
gene, OR 97403, USA.
E-mail address: bculleto@uoregon.edu
0305-4403/$ - see front matter C211 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jas.2006.01.013
Journal of Archaeological Science 33 (2006) 1331e1339
http://www.elsevier.com/locate/jas
many sites and datable charcoal samples are typically scarce
[18,19,33].
Differences in radiocarbon content between terrestrial and
freshwater systems was first theorized by Godwin [24], who
suggested that bicarbonate dissolved from ancient limestone
could introduce random errors in carbonates precipitated in
freshwater environments. Spurred by observed age discrep-
ancies between marl and terrestrial plants in New Zealand
[8], Deevey et al. [21] demonstrated the effect of 14C-depleted
bicarbonate in modern aquatic organisms in soft- and hard-
water lakes in Connecticut and New York state. Broecker and
Walton [12] presented a formal geochemical model describing
the radiocarbon budget of freshwater systems, noting that the fac-
tors governing 14C concentrations in lakes are the initial ratio of
dissolved carbonate to silicate, and the rate of CO2 exchange
with the atmosphere. Subsequently, Keith and coworkers
[35e37] argued for decomposition of old humus in rivers and
lakes as a main source of 14C-depleted carbon in freshwater
systems, a view rejected by Broecker [9]. Several methodolog-
ical studies during the 1950s and 1960s dated modern freshwa-
ter mussel shells (Family Unionidae) to estimate the magnitude
of freshwater reservoirs in the Great Lakes and Great Basin re-
gions, in some cases specifically to address archaeological
chronologies [10,11,14e17,28,45]. More recently, Berger and
Meek [4] proposed a genus-specific radiocarbon correction
(i.e., one due to vital effects) of 450 14C yr for Anodonta in
the Mojave River Basin determined from pre-bomb museum
specimens. In the region of the present study, Sutton and Orfila
[57] have calculated a 300 14C yr correction for Anodonta using
co-associated turtle and artiodactyl bone from an archaeologi-
cal site north of Buena Vista Lake, CA-KER-4220.
The potential for a freshwater reservoir effect in Buena Vista
Slough was noted by Jackson et al. [33] during analyses of ma-
rine shell beads and freshwater mussels at site CA-KER-5404
on the Elk Hills. Calibrated dates on eight beads manufactured
from the marine gastropod genus Olivella were consistently
w300 years later than samples of freshwater mussel (G. angu-
lata) from a shell dump feature at the site [33:131,134e136].
Jackson et al. [33] assumed that the deposition of the beads
and the mussel shells was contemporaneous, and suggested
that either old carbon was present in the freshwater mussels,
or that an inappropriate local marine reservoir correction
(DR) had been used to calibrate the Olivella beads. Noting
that DR in the Santa Barbara Channel (the presumed source
of the Olivella shells) had been shown to fluctuate during the
Holocene [39], the authors concluded that the local correction
applied (i.e., DR ?225 C635 14C yr; [5]) was incorrect. The is-
sue was raised again by Culleton and Jackson [18] in the inter-
pretation of dates on freshwater mussel shell and Olivella beads
from two other Elk Hills sites, CA-KER-5376 and CA-KER-
5955, where the same pattern was noted.
Cultural responses to environmental change in California
and the Great Basin are increasingly a focus of archaeological
research (e.g., [1,34,38,41,44,47]), but inferring those re-
sponses depends on accurate archaeological and climatic chro-
nologies. Accounting for a freshwater reservoir effect allows
for clearer comparison of the Elk Hills archaeology with
California?s late Holocene environmental records (e.g.,
[20,25,29,30,42,49,50]). Initial mussel shell dates indicated
that the vast majority of sites on the Elk Hills were occupied
during the Medieval Climatic Anomaly (MCA; [50]) from
1000 to 600 cal BP; only 3 of 23 dates from 16 sites dated
before ca. 1450 cal BP [33:150]. Most of these sites represent
short-term camps for processing and consuming foods gath-
ered and hunted from the former Buena Vista Slough [33].
Aside from the mussel shell, bones of fish, turtles, waterfowl,
and rabbits e but rarely larger mammal e were recovered.
The widespread and sudden appearance of people on the north
flank of the Elk Hills at this period was interpreted by Jackson
et al. [33:150] as a response to the desiccation of Buena Vista
Lake during the prolonged MCA droughts. They hypothesized
that the well-established lakeside village sites (cf. [61]) were
abandoned as year-round occupation sites as people were
forced to move greater distances in search of resources. De-
clining productivity and availability of terrestrial and aquatic
habitats led to the intensive exploitation of lower-ranked
slough resources, including freshwater mussels. Characteriz-
ing a freshwater reservoir effect in the Buena Vista Basin
could alter interpretations of the role of climate change in
the timing of settlement shifts in the late Holocene Buena
Vista Basin. This paper presents radiocarbon data on paired
Anodonta and Gonidea shells to identify genus-specific reser-
voir corrections, and paired charcoal and shells of these two
genera to characterize the magnitude of the effect in the late
Holocene Buena Vista Slough.
2. Methods
Freshwater mussel shell (Anodonta sp. and G. angulata)is
found scattered on the surface or in concentrations represent-
ing dumping events, on sites along the north flank of the Elk
Hills adjacent to the now-reclaimed Buena Vista Slough
(Fig. 1). Samples were selected from two excavated sites,
CA-KER-3079/H and CA-KER-5404, to characterize the
freshwater reservoir effect (Rf) and to investigate interspecific
differences between Anodonta and Gonidea. To test whether
the two genera of freshwater mussels date differently, a pair
of individual Anodonta sp. and G. angulata valves were se-
lected from each of two small, discrete shell lenses at CA-
KER-5404 (units SS-14 and SS-18). KER-5404 is located
roughly 1.5 km from the former Buena Vista Slough on a small
knoll near the transition to the incised interior of the Elk Hills.
Each lens measured approximately 30 cm in diameter and con-
tained the valves of 50e60 individual mussels [19]. The small
amount of meat provided by each bivalve and the discrete
concentration of the shells suggest that these lenses represent
single episodes of mussel collection and consumption by indi-
viduals, rather than diachronic accumulations of dietary debris
by larger groups [19].
To estimate the magnitude of the freshwater reservoir effect
in the former Buena Vista Slough, six freshwater mussel/char-
coal pairs were recovered from small hearth features and mid-
den excavated at CA-KER-3079/H Locus A. The locus is on
a small knoll about 500 m from the old Buena Vista Slough,
1332 B.J. Culleton / Journal of Archaeological Science 33 (2006) 1331e1339
and characterized by dense concentrations of freshwater mus-
sel shell, midden and Late Period Olivella beads (e.g., Benny-
hoff and Hughes? [3] Classes G, J, H, and K). Individual pieces
of freshwater mussel shell were removed from hearth fill for
AMS dating. The shell was too fragmented to allow genus-
level distinction between Gonidea and Anodonta. Charcoal
samples were chosen from these same contexts. Generally,
smaller charred twigs were chosen to avoid the ??old wood
effect?? [40,46], though most of the local plants are relatively
short-lived scrub. More pertinent is the ubiquity of Atriplex
(saltbush) on the Elk Hills. This C4 plant efficiently incorpo-
rates heavier carbon isotopes (i.e., 13C and 14C) during photo-
synthesis, resulting in high d13C values that can significantly
skew radiocarbon calibrations [13]. The problem is obviated
by directly measuring rather than estimating d13C during
AMS dating. The measured ratios reveal that five of the six
charcoal samples are indeed Atriplex, judging by d13C values
ranging from C25510.8 to C25512& typical of C4 plants [13,53].
Samples were submitted to Beta Analytic, Inc. for AMS
dating, who conducted pretreatment of the specimens (acid
etch for carbonates, and acid/alkali/acid baths for charcoal).
Measurements of d13C were made for all samples, and used
to correct measured radiocarbon ages [53]. Radiocarbon ages
were calibrated with CALIB 4.3 using the INTCAL98 decadal
atmospheric curve for all samples [54e56].
3. Results
3.1. Magnitude of the freshwater reservoir
The radiocarbon results for the six shell/charcoal pairs from
the hearths at CA-KER-3079/H Locus A are presented in
Table 1. The freshwater reservoir correction (Rf) is calculated
as the difference between the conventional age of the mussel
shell sample and that of its paired charcoal sample.
The conventional ages of the charcoal samples are remark-
ably consistent with each other, suggesting a relatively short oc-
cupation in the late Holocene. The freshwater mussel shells
show more variability in their conventional ages, perhaps due
to habitat or species effects between Anodonta and Gonidea at
CA-KER-5404. Despite these variations, Rf for each pair is of
a similar magnitude and is consistently in the same direction,
i.e., all of the shells have greater apparent ages than their paired
charcoal sample. These samples indicate that Rf averaged
340 C620 14C yr from 720 to 680 C640 14C yr BP (calculations
follow Stuiver et al. [52:983]). This Rf corresponds to a 300-
year disparity between calibrated freshwater carbonate and ter-
restrial charcoal dates during this period. A roughly 50-year
disparity remains between these corrected dates and those on
marine shell beads at CA-KER-5404, perhaps due to differences
between the atmospheric and marine calibration curves, or to
late Holocene DR fluctuations in the Santa Barbara Channel.
Below, fractionation effects between shellfish genera are evalu-
ated with Anodonta/Gonidea pairs from site CA-KER-5404.
3.2. Interspecific effects
Radiocarbon data for two Anodonta/Gonidea pairs from
CA-KER-5404 are presented in Table 2. Comparison of the
d13C-corrected conventional ages indicates an offset between
the two genera, with Gonidea dating 70 14C yr and 100 14C
yr older than Anodonta in SS-14 and SS-18, respectively.
The 2s calibrated ages of each pair overlap, but less in the
pair from SS-14. The age discrepancies between the taxa are
unlikely to be the result of misinterpreted archaeological con-
text, but may be due to variability in atmospheric 14C content
Fig. 1. Location of Elk Hills study sites in relation to major hydrographic features of the Buena Vista Basin, Kern County, California. Basemap Source: USGS
Seamless NED Database.
1333
from ca. 1000 to 550 cal BP, species-specific carbonate metab-
olism, habitat differences between Anodonta and Gonidea,or
some combination of these factors. As to habitat, Anodonta
prefers slow-moving waters with muddy substrates such as
lakes, whereas Gonidea is more commonly found in faster wa-
ters with sandier substrates such as streams and rivers [58].
The observation that the fluvial Gonidea is more depleted in
14C than is the lacustrine Anodonta is consistent with the find-
ings of Keith and coworkers [35e37], who identified a similar
trend between lake and river pelecypods in their data and that
of Broecker and Olson [10,11].
The variation in the apparent ages of Anodonta and Goni-
dea may partly result from variations in the atmospheric cali-
bration curve during the last ca. 1000 14C yr BP. Applying the
340 C620 14C yr correction to the four dates from CA-KER-
5404 renders the 2s calibrated ranges for each pair statistically
indistinguishable, despite the differences in conventional 14C
ages (Table 3). In sum, the question of genus-specific 14C cor-
rections for these two species (i.e., differences arising from
metabolic fractionation rather than habitat differences or vari-
able atmospheric 14C content) remains open. Anodonta/Goni-
dea pairs from periods of more stable 14C production are
required to definitively resolve the issue.
4. Discussion
The estimated value Rf ?340 C620 14C yr between ca. 720
and 680 14C yr BP is comparable to offsets reported for
Anodonta by Sutton and Orfila [57] of 300 14C yr north of
Buena Vista Lake, and Joan Scheinder (written communica-
tion 1989, cited in Berger and Meek [4]) of 350 14C yr in the
Lake Manix Basin. Berger and Meek [4:581] report a Rf of
450 14C yr derived from pre-bomb Anodonta specimens
collected from Los Angeles and Yermo, California, with
measured 14C age of 480 C660 14C yr, and a known age of
ca. 30 14C yr (i.e., collection in AD 1920). This value did
not account for the proportions of the three specimens that
the sample comprises, or the apparent age of the atmosphere
in AD 1920. Using the reported sample weights and assum-
ing complete homogenization of all the shells, the average
age is closer to 20 14C yr, or collection in AD 1930, making
the offset 460 C660 14C yr. The apparent age of the atmo-
sphere in AD 1930 was approximately 150 14Cyr[56].
Deducting this value from the measured age gives
Rf ?310 C660 14C yr, which is similar to the range of values
cited above. The general consistency of these offsets may in-
dicate a genus- or family-level (Unionidae) 14C fractionation
effect, as proposed by Berger and Meek [4], though the pres-
ence of a geologic source of 14C cannot be ruled out in any
of these settings.
4.1. The source of depleted radiocarbon
in the Buena Vista Basin
The classical explanations for the presence of 14C-
depleted carbon in freshwater mussels include dissolution
Table 1
14C Data for paired freshwater mussel and charcoal samples from CA-KER-3079/H Locus A
Provenience Sample number
(Beta-#)
Material Measured 14C
age BP
d13C
(& PDB)
Conventional 14C
age BP
Rf
S3.2/W1.3 Feature 1 180840 Charcoal 630 C640 C25520.4 710 C640 430 C650
180839 Freshwater Mussel 830 C650 C2556.4 1140 C650
S3.195/W1.305 Feature 1 180842 Charcoal 480 C640 C25510.8 710 C640 340 C660
180841 Freshwater Mussel 720 C660 C2554.9 1050 C660
S2.78/W1.65 Feature 1 180843 Charcoal 480 C640 C25512.0 690 C640 340 C640
180844 Freshwater mussel 760 C640 C2558.5 1030 C640
S2.8/W1.65 Feature C 180846 Charcoal 490 C640 C25511.0 720 C640 280 C640
180845 Freshwater mussel 680 C640 C2555.5 1000 C640
S2.41/W1.53 Point
Provenience 4
180847 Charcoal 450 C640 C25510.8 680 C640 350 C640
180848 Freshwater mussel 740 C640 C2557.3 1030 C640
S2.41/W1.53 Point
Provenience 2
180849 Charcoal 470 C640 C25511.4 690 C640 320 C640
180850 Freshwater mussel 710 C640 C2556.9 1010 C640
Weighted mean Rf 340 C620
Conventional radiocarbon ages are d13C-corrected with measured values [53]. Rf is calculated as a weighted mean with unequal uncertainties following Stuiver
et al. [52:983]. Uncertainty in Rf is taken as the maximum of the ??scatter?? sigma in unweighted mean (i.e., s/On ?20.1 14C yr) and the weighted mean sigma (i.e.,
1=
C16 ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiP
1=32p
C17
? 17:7 14C yr, where e is the measurement error) after Stuiver et al. [52:983].
Table 2
14C Data for Anodonta/Gonidea shell pairs, CA-KER-5404
Sample number
(Beta-#)
Provenience Material Conventional 14C
age BP
d13C(& PDB) 2s Calibrated age
range (cal BP)
180855 SS-14 Shell lens/20e30 cm Single Anodonta valve 680 C640 C2556.1 680e560
180856 SS-14 Shell lens/20e30 cm Single Gonidea valve 750 C640 C2554.0 740e650
180853 SS-18 Shell lens/25e35 cm Single Anodonta valve 890 C640 C2555.8 920e700
180857 SS-18 Shell lens/25e35 cm Single Gonidea valve 990 C640 C2556.0 970e790
Conventional radiocarbon ages are d13C-corrected with measured values [53]. Dates were calibrated with Calib 4.3 using the Intcal98 curve [54e56].
1334 B.J. Culleton / Journal of Archaeological Science 33 (2006) 1331e1339
of geologic carbonates [8,9,12,21,23,24,45]; inputs of old soil
humus [35e37]; residence time [12,23]; and vital effects
[4,37]. Some or all of the old carbon in the Buena Vista
Basin is likely derived from geologic sources in the Kern River
watershed, and is then stored and concentrated in the ground-
water aquifer. Stretches of the Kern River flow over pre-
Cretaceous limestones in the Sierra Nevada, and through
Miocene marine sediments in the Sierra foothills before set-
tling into the Buena Vista Lake and Slough system [43,48].
In general, the present groundwater chemistry of dissolved
solids in California?s Central Valley grades from bicarbonate
anions on the east side to chloride and sulfate anions on the
west side, reflecting mineral inputs from the Sierra Nevada
and the Coast ranges, respectively [6]. In the Buena Vista
Basin itself, surface waters of Buena Vista and Kern lakes,
and Buena Vista Slough, were likely to have been dominated
by NaHCO3, as well as lesser amounts of CaCO3 and
Mg(HCO3)2, as the groundwater underlying those now-relict
lake beds show [6:A38]. Water samples from the Kern River
and Kern Lake analyzed by the pioneering agronomist E. W.
Hilgard in 1880 demonstrated that compared to rivers to the
north, such as the Kings and San Joaquin, the Kern River
contained dangerously high concentrations of NaHCO3 that
threatened to destroy the agricultural productivity of the
southern San Joaquin Valley [27]. The extreme concentra-
tions of NaHCO3 in Kern Lake in 1880, about 18 months af-
ter it had nearly dried out and caused a mass die-off of fish
and turtles, suggest that bicarbonate was probably also peri-
odically concentrated in the lakes by evaporation, perhaps
contributing to a type of carbon reservoir described by
Geyh et al. [23] in Laguna Lej?a (Chile?s Atacama Desert).
This also suggests that the concentration of 14C-free carbon
in the lakes of the Buena Vista Basin probably varied
through time in response to fluctuating surface area to vol-
ume ratios, which reflect time-dependent hydrologic, climatic
and geologic factors [12,21,23].
4.2. Archaeological implications for the
late Holocene Buena Vista Basin
Recalibrated freshwater mussel shell dates from Elk Hills
sites are presented in Table 4 with Rf ?340 C620 14C yr sub-
tracted from the conventional 14C age. Procurement of fresh-
water mussels appears to be most intense after ca. 700 cal
BP, and the few dates earlier than 1200 cal BP (i.e., only 5
of 50 dates) may indicate that slough resources were not a sub-
sistence focus on Elk Hills during earlier periods. Fig. 2
depicts the Rf-corrected 2s calibrated ranges of the 45 post-
1500 cal BP dates on freshwater mussel from 17 Elk Hills
sites, plotted against inferred climate from the drowned
stumps in the Mono Basin [50], and tree-rings from the south-
ern Sierra Nevada [25] and the White Mountains [29,30,42].
Comparison with the regional climate records indicates a gen-
eral correlation of mussel procurement on the Elk Hills with
cooler, wetter conditions towards the end of the Medieval
Climatic Anomaly and during the Little Ice Age. The parallel
between the termination of drought in the Mono Basin and
cluster of freshwater mussel 2s-ages ca. 700e550 cal BP ap-
pears to be the strongest relationship, though the dates also
show good agreement with LaMarche?s [42] data, and the
broader trend of increasing precipitation after ca. 700 cal BP
in all the records. It has been argued that the differences in
timing and severity of droughts revealed in these proxies e
specifically, apparent contradictions between the tree-ring
records and the drowned stumps in the Mono Basin e under-
mine hypotheses of widespread social reorganization in
response to climate change in the late Holocene of western
North America (e.g., comments on Jones et al. [34]:
[2,7,26]). Bearing those criticisms in mind, appearance of
abundant freshwater mussel shell across the Elk Hills at the
termination of MCA droughts in the Mono Basin [50] is com-
pelling because both indicate the state of the hydrologic sys-
tem directly. Unlike dendroclimatic records, the Mono Basin
stumps show the end result of extended temperature and pre-
cipitation anomalies on the hydrologic landscape from which
prehistoric peoples made their living [51:627].
Intensive procurement of freshwater mussels (and other
slough resources) appears to have begun at the termination
of the MCA and during the Little Ice Age (post-ca. 700 cal
BP) as aquatic habitats were expanding and renewing in the
Buena Vista Basin, rather than shrinking and declining in
productivity during the MCA, as proposed by Jackson et al.
[33]. In light of this new chronology, the forces that drove
the intensive gathering of mussels and other slough foods
may need to be reconsidered. For example, greater
productivity of post-MCA aquatic habitats may have allowed
regional population to increase, necessitating increased diet
breadth or a larger foraging range that included the Elk Hills
to support the permanent village sites at Buena Vista Lake
[61].
Table 3
Rf-corrected 14C data for Anodonta/Gonidea shell pairs, CA-KER-5404
Sample number
(Beta-#)
Provenience Material Rf-corrected
conventional 14C
age BP
d13C(& PDB) Rf-Corrected 2s
Calibrated age
range (cal BP)
180855 SS-14 Shell lens/20e30 cm Single Anodonta valve 340 C640 C2556.1 490e310
180856 SS-14 Shell lens/20e30 cm Single Gonidea valve 410 C640 C2554.0 520e320
180853 SS-18 Shell lens/25e35 cm Single Anodonta Valve 550 C640 C2555.8 650e510
180857 SS-18 Shell lens/25e35 cm Single Gonidea valve 650 C640 C2556.0 670e550
Conventional radiocarbon ages are d13C-corrected with measured values, and reservoir-corrected by subtracting 340 14C yr from the conventional age [53]. Dates
were calibrated with Calib 4.3 using the Intcal98 curve [54e56].
1335
Table 4
Rf-corrected freshwater mussel 14C dates from Elk Hills Sites
Sample
(Beta-#)
Site number Provenience Material Conventional 14C
age BPa
d13Cb 2s calibrated
age range (BP)
Reference
111362 CA-KER-5397 SCA-1/surface Bulk mussel 50 C680 C2556.9 280e0 Jackson et al. [33]
112706 CA-KER-5367 SP-3/0e50 cm bs Bulk mussel 140 C660 C2557.8 280e0 Jackson et al. [33]
108269 CA-KER-5401 STP-26/20e40 cm bs Bulk mussel 190 C660 nr 310e0 Jackson et al. [33]
190870 CA-KER-3080 Locus A Unit SW1/0e20 cm bs Single Gonidea 290 C640 C2556.1 460e290 Culleton et al. [19]
180851 CA-KER-3082 Unit 277920E/3911460N
Feature 1/19e30 cm bs
Single mussel 350 C640 C2554.5 490e310 Culleton et al. [19]
180855 CA-KER-5404 SS-14/20e30 cm bs Single Anodonta 340 C640 C2556.1 490e310 Culleton et al. [19]
190868 CA-KER-3080 Locus A Unit 104.03/0e20 cm bs Single Gonidea 340 C640 C2553.9 490e310 Culleton et al. [19]
111364 CA-KER-5398/H SP-24/0-50 cm bs Bulk mussel 400 C660 C2555.3 520e310 Jackson et al. [33]
180856 CA-KER-5404 SS-14/20e30 cm bs Single Gonidea 410 C640 C2554.0 520e320 Culleton et al. [19]
112711 CA-KER-3082 RRU-21/0-40 cm bs Bulk mussel 400 C670 C2555.6 530e310 Jackson et al. [33]
190869 CA-KER-3080 Locus A Unit 105.03/0e20 cm bs Single Gonidea 430 C640 C2553.9 540e330 Culleton et al. [19]
116688 CA-KER-3077 Locus A EU-3/0-20 cm bs Bulk mussel 440 C660 C2554.6 550e320 Jackson et al. [33]
116689 CA-KER-3077 Locus A EU-6 Feature 1 Bulk Anodonta 460 C670 C2555.1 630e310 Jackson et al. [33]
111363 CA-KER-5396 SP-4/0e50 cm bs Bulk mussel 470 C660 C2555.5 630e320 Jackson et al. [33]
168084 CA-KER-3397 CU-2/0e10 cm bs Single mussel 520 C640 C2555.0 640e500 Culleton and Jackson [18]
108268 CA-KER-5401 STP-3/20e40 cm bs Bulk mussel 510 C670 nr 650e340 Jackson et al. [33]
180853 CA-KER-5404 SS-18 Shell
lens/25e35 cm bs
Single Anodonta 550 C640 C2555.8 650e510 Culleton et al. [19]
180854 CA-KER-5404 SS-18 Feature 18eC/26 cm bs Single Anodonta 550 C640 C2554.4 650e510 Culleton et al. [19]
168090 CA-KER-5955 STU-1/0e20 cm bs Single mussel 560 C640 C2553.3 650e520 Culleton and Jackson [18]
168087 CA-KER-5955 CU-1/0e30 cm bs Single mussel 600 C640 C2553.1 650e540 Culleton and Jackson [18]
106600 CA-KER-5404 TU-3/20e30 cm bs Bulk Gonidea 600 C660 C2556.3 660e530 Jackson et al. [33]
112707 CA-KER-3079 Locus C TEU-1 Feature 1 Bulk Anodonta 620 C670 C2558.0 670e540 Jackson et al. [33]
112708 CA-KER-3080 Locus C EU-7/0e20 cm bs Single Gonidea 640 C650 C2556.6 670e550 Jackson et al. [33]
180836 CA-KER-5373/H Locus C Unit 30 Feature 1/5e25 cm bs Single mussel 660 C640 C2556.0 670e550 Culleton et al. [19]
180838 CA-KER-5373/H Locus C Unit 103 Feature 2/13e28 cm bs Single mussel 650 C640 C2556.8 670e550 Culleton et al. [19]
180845 CA-KER-3079 Locus A Unit S 2.8/W 1.65 Feature
C/12e22 cm bs
Single mussel 660 C640 C2555.5 670e550 Culleton et al. [19]
180850 CA-KER-3079 Locus A Unit S 2.41/W 1.53 Point
Provenience 2; 10e20 cm bs
Single mussel 670 C640 C2556.9 670e550 Culleton et al. [19]
180857 CA-KER-5404 SS-18 Shell
lens/25e35 cm bs
Single Gonidea 650 C640 C2556.0 670e550 Culleton et al. [19]
108266 CA-KER-3167 TU-1/0e10 cm bs Bulk mussel 630 C670 nr 680e520 Jackson et al. [33]
180844 CA-KER-3079 Locus A Unit S 2.78/W1.65/0e10 cm bs Single mussel 690 C640 C2558.5 690e560 Culleton et al. [19]
180848 CA-KER-3079 Locus A Unit S 2.41/W 1.53 Point
Provenience 4; 10e20 cm bs
Single mussel 690 C640 C2557.3 690e560 Culleton et al. [19]
112709 CA-KER-3080 Locus C EU-7/0e20 cm bs Bulk mussel 680 C660 C2554.5 710e550 Jackson et al. [33]
116691 CA-KER-5373/
H Locus C
CU-1/30e40 cm bs Bulk mussel 690 C670 C2555.0 730e540 Jackson et al. [33]
116690 CA-KER-5404 RRU-25/0e40 cm bs Bulk mussel 700 C660 C2556.0 730e550 Jackson et al. [33]
188992 CA-KER-3085 Unit 45.03/15e20 cm bs Single Anodonta 720 C650 C2556.6 730e560 Culleton et al. [19]
180841 CA-KER-3079 Locus A Unit S 3.195/W 1.305 Feature
1/10e20 cm bs
Single mussel 710 C660 C2554.9 740e550 Culleton et al. [19]
168088 CA-KER-5955 RRU-2 0e40 cm bs Single mussel 780 C640 C2556.5 760e660 Culleton and Jackson [18]
107055 CA-KER-3168 TU-2/0e20 cm bs Bulk mussel 760 C660 C2557.1 790e560 Jackson et al. [33]
106601 CA-KER-5404 TU-3/20e30 cm bs Single Gonidea 760 C650 C2556.6 790e570 Jackson et al. [33]
180839 CA-KER-3079 Locus A Unit S 3.2/W 1.3 Feature
1/17 cm bs
Single mussel 800 C650 C2556.4 880e660 Culleton et al. [19]
112710 CA-KER-3080 Locus C EU-6/20e40 cm bs Bulk mussel 770 C660 C2556.3 890e560 Jackson et al. [33]
107056 CA-KER-3168 TU-2/40e60 cm bs Bulk mussel 790 C670 C2556.4 910e570 Jackson et al. [33]
111365 CA-KER-5392 Locus A SP-37/0e50 cm bs Bulk mussel 810 C660 C25510.7 910e660 Jackson et al. [33]
108270 CA-KER-3080 Locus D STP-2/0e20 cm bs Bulk mussel 1120 C660 nr 1170e930 Jackson et al. [33]
168085 CA-KER-3397 CU-2/10e20 cm bs Single mussel 1160 C640 C2556.2 1170e970 Culleton and Jackson [18]
168086 CA-KER-3397 RRU-2 Single mussel 1920 C640 C2559.2 1950e1740 Culleton and Jackson [18]
188993 CA-KER-3085 Unit 83.03/120e125 cm bs Single Anodonta 2050 C640 C2556.7 2120e1900 Culleton et al. [19]
116692 CA-KER-5392 Locus B CU-1/30e40 cm bs Bulk mussel 2740 C650 C2556.0 2950e2760 Jackson et al. [33]
116693 CA-KER-3166/H CU-1/50e60 cm bs Bulk mussel 4500 C6100 C2555.8 5450e4860 Jackson et al. [33]
108267 CA-KER-3166/H TU-2/40e50 cm bs Bulk mussel 4400 C660 nr 5280e4845 Jackson et al. [33]
a Conventional (13C-corrected) 14C ages reported with 340 C620 14C yr reservoir correction subtracted, and calibrated with Calib 4.3 using the Intcal98 atmo-
spheric curve [53e56].
b nr ?not reported by Jackson et al. [33].
1336 B.J. Culleton / Journal of Archaeological Science 33 (2006) 1331e1339
Annual Precipitation (cm)
22
21
20
19
18
2000-yr average
ca. 19.8 cm
3
2
1
-1
-2
0
Ring width departure
upper treeline
(temperature)
lower forest
border
(moisture)
Cool-dry Cool-dry
Warm-moistWarm-moist
Warm-dry Cool-moistInferred ClimaticAnomaly
Winter precipitationanomalies (cm)
+40
-40
0
Extended drought
periods (shaded)
168085
168088
108270
108266
108268
106600
168087
168084
168090
112710
112709
112708
111363
111362
116689
116688
112711
112706
111364
190869
108269
190870
190868
112707
180839
180841
180848
180857
180850
180845
180838
180836
180854
180853
180856
180855
180851
180844
188992
106601
107056
107055
111365
116690
116691
cal BP
0
cal BP
0
Beta-# Site CA-#
KER-3397
KER-5955
KER-3080 D
KER-3167
KER-5401
KER-5404
KER-5955
KER-3397
KER-5955
KER-3080 C
KER-3080 C
KER-3079/H A
KER-5396
KER-5367
KER-3077 A
KER-3077 A
KER-3082
KER-5397
KER-5398/H
KER-3080 A
KER-5401
KER-3080 A
KER-5404
KER-3079/H C
KER-3079/H A
KER-3079/H A
KER-3079/H A
KER-5404
KER-5373/H C
KER-5373/H C
KER-3080 C
KER-3079/H A
KER-5404
KER-5404
KER-5404
KER-3082
KER-3080 A
KER-3079/H A
KER-3085/H C
KER-5404
KER-3168
KER-3168
KER-5392 A
KER-5404
KER-5373/H C
Drought DroughtWetInterval
Medieval Climatic Anomaly Little Ice Age
during MCA
Mono Lake conditions
[50]
A
B
C
D
Southern
Sierra Nevada
Precipitation
[25:254; Fig.4]
White Mtns.
Bristlecone Pine
Intersite Comparison
[42:1047; Fig.6]
White Mtns.
Bristlecone Pine
Precipitation
[29,30]
Rf-Corrected
Mussel Dates
from Late Holocene
Elk Hills Sites
Fig. 2. Reservoir-corrected 1s (closed) and 2s (open) calibrated dates on freshwater mussel shell from Elk Hills archaeological sites post-1500 cal BP compared
with: A. Mono Basin drought [50]; B. Southern Sierra Nevada precipitation anomalies (redrawn from Graumlich [25: Fig. 4]); C. White Mountains intersite bris-
tlecone pine comparison (redrawn from LaMarche [42: Fig. 6]; D. White Mountains bristlecone pine precipitation record (data from Hughes and Graumlich
[29,30]. The intensive exploitation of freshwater mussels coincides with generally cooler, wetter conditions at the end of the Medieval Climatic Anomaly [50]
and during the Little Ice Age.
1337
5. Conclusions
Paired radiocarbon dates on charcoal and freshwater mussel
shell (Anodonta and Gonidea sp.) samples from archaeological
sites on the Elk Hills in the Buena Vista Basin of the southern
San Joaquin Valley indicate a freshwater reservoir effect (Rf)
of 340 C620 14C yr between 720 and 680 14CyrBP.ARf of
this magnitude is consistent with values reported for Anodonta
by other researchers. Paired dates on Anodonta and Gonidea
sp. shells from discrete lenses may indicate genus-specific off-
sets from metabolic or habitat difference, but further work is
required to resolve the issue. The source of 14C-depleted car-
bon Buena Vista Basin may be bicarbonate dissolved from
pre-Cretaceous limestone in the Kern River watershed, which
is then concentrated and stored in groundwater aquifers. Reca-
libration of Rf-corrected mussel shell dates from the Elk Hills
shows a correlation of intensive use of the Buena Vista Slough
resources with periods of increased precipitation toward the
end of the Medieval Climatic Anomaly and during the Little
Ice Age. Slough resource exploitation on the Elk Hills may
have resulted from growing regional population pressure as
aquatic habitats became more expansive and productive, rather
than from declining resource availability as previously sup-
posed. Characterizing the freshwater radiocarbon reservoir in
the Buena Vista Basin allows for better articulation of the re-
gional cultural chronology with California?s climate record.
Acknowledgements
The field work and analyses reported here were conducted
by Pacific Legacy, Incorporated, under contract to the US De-
partment of Energy, Naval Petroleum Reserves of California
(Contract No. DE-AC01-01FE66862). Thanks go to Thomas
L. Jackson, Principal Investigator, and the members of Pacific
Legacy?s 2003 field crew. This manuscript was improved by
extensive comments from Doug Kennett and two anonymous
reviewers. Paula Reimer of Queen?s University Belfast and
John Southon of the UC Irvine Keck Carbon Cycle AMS Fa-
cility generously provided guidance on reservoir calculation
issues.
References
[1] J.E. Arnold, Complex hunteregathererefishers of prehistoric California:
chiefs, specialists, and maritime adaptations of the Channel Islands,
American Antiquity 57 (1992) 60e84.
[2] M.E. Basgall, Comment on: environmental imperatives reconsidered, by
T.L. Jones et al. (1999), Current Anthropology 40 (2) (1999) 157e158.
[3] J. Bennyhoff, R. Hughes, Shell bead and ornament exchange networks
between California and the Great Basin, Anthropological Papers of the
American Museum of Natural History 64 (2) (1987).
[4] R. Berger, N. Meek, Radiocarbon dating of Anodonta in the Mojave
River Basin, Radiocarbon 34 (3) (1992) 578e584.
[5] R. Berger, R.E. Taylor, W.F. Libby, Radiocarbon content of marine shells
from the Californian and Mexican West Coast, Science 153 (1966) 864e
866.
[6] G.L. Bertoldi, R.H. Johnston, K.D. Evenson, Groundwater in the Central
Valley, California e A Summary Report, USGS Professional Paper 1401-
A, 1991.
[7] R.L. Bettinger, Comment on: environmental imperatives reconsidered,
by T.L. Jones et al. (1999), Current Anthropology 40 (2) (1999) 158e
159.
[8] M. Blau, E.S. Deevey Jr., M.S. Gross, Yale natural radiocarbon measure-
ments, I. Pyramid Valley, New Zealand and its problems, Science 118
(1953) 1e6.
[9] W.S. Broecker, Radiocarbon dating: a case against the proposed link be-
tween River Mollusks and soil humus, Science 143 (1964) 596e597.
[10] W.S. Broecker, E.A. Olson, Lamont radiocarbon measurements VI,
American Journal of Science: Radiocarbon Supplement 1 (1959) 111e
132.
[11] W.S. Broecker, E.A. Olson, Lamont radiocarbon measurements VIII,
Radiocarbon 3 (1961) 176e204.
[12] W.S. Broecker, A. Walton, The Geochemistry of C14 in freshwater sys-
tems, Geochimica et Cosmochimica Acta 16 (1959) 15e38.
[13] D.L. Browman, Isotopic discrimination and correction factors in radio-
carbon dating, Advances in Archaeological Method and Theory 4
(1981) 241e295.
[14] H.R. Crane, University of Michigan radiocarbon dates I, Science 124
(1956) 664e672.
[15] H.R. Crane, J.B. Griffin, University of Michigan radiocarbon dates IV,
American Journal of Science: Radiocarbon Supplement 1 (1959) 173e
198.
[16] H.R. Crane, J.B. Griffin, University of Michigan radiocarbon dates VII,
Radiocarbon 4 (1962) 183e203.
[17] H.R. Crane, J.B. Griffin, University of Michigan radiocarbon dates VIII,
Radiocarbon 5 (1963) 228e253.
[18] B.J. Culleton, T.L. Jackson, Draft Final Cultural Resources Report for the
Elk Hills Power Project (99-AFC-1), Pacific Legacy, Incorporated, Santa
Cruz, CA, 2003.
[19] B.J. Culleton, T.L. Jackson, S. Brewer, Cultural Response to Environ-
mental Change in the Buena Vista Basin: Archaeological Data Recovery
at Eight Sites on the Former Naval Petroleum Reserve No. 1 (Elk Hills),
Kern County, California, Pacific Legacy, Incorporated, 2005.
[20] O.K. Davis, Rapid climate change in coastal Southern California inferred
from pollen analysis of San Joaquin Marsh, Quaternary Research 37
(1992) 89e100.
[21] E.S. Deevey Jr., M.S. Gross, G.E. Hutchinson, H.L. Kraybill, The natural
C14 contents of materials from hard-water lakes, Proceedings of the
National Academy of Sciences, USA 40 (1954) 285e288.
[22] M. Fontugne, M. Carre, I. Bentaleb, M. Julien, D. Lavalle, Radiocarbon
reservoir age variations in the South Peruvian upwelling during the
Holocene, Radiocarbon 46 (2) (2004) 531e537.
[23] M.A. Geyh, U. Schotterer, M. Grosjean, Temporal changes of the 14C
reservoir in lakes, Radiocarbon 40 (2) (1998) 921e933.
[24] H. Godwin, Comments on radiocarbon dating for samples from the Brit-
ish Isles, American Journal of Science 249 (1951) 301e307.
[25] L.J. Graumlich, A 1000-year record of temperature and precipitation in
the Sierra Nevada, Quaternary Research 39 (1993) 249e255.
[26] J. Haas, W. Creamer, Comment on: environmental imperatives reconsid-
ered, by T.L. Jones et al. (1999), Current Anthropology 40 (2) (1999)
160.
[27] E.W. Hilgard, Appendix no. 1: alkali soils and irrigation waters of the
San Joaquin Valley, in: E.W. Hilgard (Ed.), Report of the Professor in
Charge to the Board of Regents, Being a Part of the Report of the Board
of Regents [Annual 1880], University of California, Berkeley, 1881,
pp. 12e36.
[28] C.L. Hubbs, G.S. Bien, H.E. Suess, La Jolla natural radiocarbon mea-
surements IV, Radiocarbon 7 (1965) 66e117.
[29] M.K. Hughes, L.J. Graumlich, Climatic variations and forcing mecha-
nisms of the last 2000 years, in: Multi-Millenial Dendroclimatic Studies
from the Western United States. NATO ASI Series, vol. 141 (1996) pp.
109e124.
[30] M.K. Hughes, L.J. Graumlich, Multi-millennial Nevada Precipitation
Reconstruction. International Tree-Ring Data Bank. IGBP PAGES/World
Data Center e A for Paleoclimatology Data Contribution Series
#2000-049, NOAA/NGDC Paleoclimatology Program, Boulder CO,
USA, 2000.
1338 B.J. Culleton / Journal of Archaeological Science 33 (2006) 1331e1339
[31] B.L. Ingram, Differences in radiocarbon age between shell and charcoal
from a Holocene shellmound in Northern California, Quaternary
Research 49 (1998) 102e110.
[32] B.L. Ingram, J.R. Southon, Reservoir ages in Eastern Pacific coastal and
estuarine waters, Radiocarbon 38 (3) (1996) 573e582.
[33] T.L. Jackson, L.A. Shapiro, J.H. King, Prehistoric Archaeological Re-
sources Inventory and Evaluation at Naval Petroleum Reserve No. 1
(Elk Hills), Kern County, California, Pacific Legacy, Incorporated, Santa
Cruz, California, 1999.
[34] T.L. Jones, G.M. Brown, L.M. Raab, J.L. McVickar, W.G. Spaulding,
D.J. Kennett, A. York, P.L. Walker, Environmental imperatives reconsid-
ered: demographic crises in Western North America during the medieval
climatic anomaly, Current Anthropology 40 (2) (1999) 137e170.
[35] M.L. Keith, G.M. Anderson, Radiocarbon dating: fictitious results with
mollusk shells, Science 141 (1963) 634e637.
[36] M.L. Keith, G.M. Anderson, Radiocarbon dating of mollusk shells:
a reply, Science 144 (1964) 890.
[37] M.L. Keith, G.M. Anderson, R. Eichler, Carbon and oxygen isotopic
composition of mollusk shells from marine and freshwater environments,
Geochimica et Cosmochimica Acta 28 (1963) 1757e1786.
[38] D.J. Kennett, The Island Chumash: Behavioral Ecology of a Maritime
Society, University of California Press, Berkeley, 2005.
[39] D.J. Kennett, B.L. Ingram, J.M. Erlandson, P. Walker, Evidence for tem-
poral fluctuations in marine radiocarbon reservoir ages in the Santa Bar-
bara Channel, Southern California, Journal of Archaeological Science 24
(1997) 1051e1059.
[40] D.J. Kennett, B.L. Ingram, J.R. Southon, K. Wise, Differences in 14C age
between stratigraphically associated charcoal and marine shell from the
Archaic period site of kilometer 4, Southern Peru: old wood or old water?
Radiocarbon 44 (1) (2002) 53e58.
[41] D.J. Kennett, J.P. Kennett, Competitive and cooperative responses to cli-
matic instability in coastal Southern California, American Antiquity 62
(2) (2002) 379e395.
[42] V.C. LaMarche Jr., Paleoclimatic inferences from long tree ring records,
Science 183 (4129) (1974) 1043e1048.
[43] R.A. Matthews, J.L. Burnett, Geologic Map of California: Fresno Sheet.
State of California, Division of Mines and Geology, San Francisco, 1966.
[44] L.M. Raab, D.O. Larson, Medieval climatic anomaly and punctuated cul-
tural evolution in Coastal Southern California, American Antiquity 62 (2)
(1997) 319e336.
[45] M. Rubin, D.W. Taylor, Radiocarbon activity of shells from living clams
and snails, Science 141 (1963) 637.
[46] M.B. Schiffer, Radiocarbon dating and the ??Old Wood?? problem: the
case of the Hohokam chronology, Journal of Archaeological Science
13 (1986) 13e30.
[47] A. Schimmelmann, C.B. Lange, B.J. Meggers, Palaeoclimatic and ar-
chaeological evidence for a w200-yr recurrence of floods and droughts
linking California, Mesoamerica and South America over the past
2000 years, The Holocene 13 (5) (2003) 763e778.
[48] A.R. Smith, Geologic Map of California: Bakersfield Sheet, State of
California, Division of Mines and Geology, San Francisco, 1965.
[49] S. Stine, Late Holocene fluctuations of Mono Lake, Eastern California,
Paleogeography, Paleoclimatology, Paleoecology 78 (1990) 333e381.
[50] S. Stine, Extreme and persistent drought in California and Patagonia
during medieval times, Nature 369 (1994) 546e549.
[51] S. Stine, On the medieval climatic anomaly, Current Anthropology 41 (4)
(2000) 627e628.
[52] M. Stuiver, G.W. Pearson, T. Braziunas, Radiocarbon age calibration of
marine samples back to 9000 Cal Yr BP, Radiocarbon 28 (2B) (1986)
980e1021.
[53] M. Stuiver, H.A. Polach, Discussion: reporting of 14C data, Radiocarbon
19 (1977) 355e363.
[54] M. Stuiver, P.J. Reimer, Extended 14C database and revised CALIB 3.0
14C age calibration program, Radiocarbon 35 (1) (1993) 215e230.
[55] M. Stuiver, P.J. Reimer, CALIB 4.3 Radiocarbon Calibration Program,
Quaternary Isotope Lab, University of Washington, 2000.
[56] M. Stuiver, P.J. Reimer, E. Bard, J.W. Beck, G.S. Burr, K.A. Hughen,
B. Kromer, F.G. McCormac, J. van der Plicht, M. Spurk, INTCAL98 ra-
diocarbon age calibration 24,000e0 cal BP, Radiocarbon 40 (3) (1998)
1041e1083.
[57] M.Q. Sutton, R.S. Orfila, A radiocarbon correction factor for freshwater
shell for the Lower Kern River/Northern Buena Vista Lake Area, South-
ern San Joaquin Valley, California, Society for California Archaeology
Newsletter 37 (2) (2003) 23e24.
[58] D.W. Taylor, Freshwater mollusks of California: a distributional check-
list, California Fish and Game 67 (3) (1981) 140e163.
[59] R.E. Taylor, Radiocarbon Dating: An Archaeological Perspective,
Academic Press, Orlando, Florida, 1987.
[60] R.E. Taylor, R. Berger, Radiocarbon content of marine shell from the
Pacific Coast of Central and South America, Science 158 (1967)
1180e1182.
[61] W.R. Wedel, Archaeological investigations at Buena Vista Lake, Kern
County, California, Bureau of American Ethnology, Bulletin 130
(1941) (Washington, DC).
1339