Geographic Information Systems as a Tool for Analyzing Intrasite Spatial Variability on San Nicolas...

97

6

Geographic Information Systems as a Tool for Analyzing Intrasite Spatial Variability on San Nicolas Island, California

Richard B. Guttenberg, René L. Vellanoweth, William E. Kendig,

Rebekka G. Knierim, and Steven J. Schwartz

!e use of geographic information systems (GIS) in archaeological research has greatly increased in recent years (Conolly and Lake "##$; Kvamme %&&&; Lock and Stancic %&&'; Wheatley and Gil-lings "##"). It is now widely used and has be-come an integral part of the archaeological tool kit (Katsianis et(al. "##); Knabb "##); Lawson "##*; Lie+ "##,; McCoy and Ladefoged "##&; Sharon et(al. "##-); however, much of the archae-ological use of GIS focuses on regional analysis of landscapes and settlement patterns (A.. "###; Kennett "##'; Kvamme %&&&; %&&'; McCoy and Ladefoged "##&; Pletka "##'; Teeter et(al., Chap-ter &). Less common is the application of GIS on an intrasite scale. In this chapter, we examine use of GIS as an e+ective tool to conduct an intrasite spatial analysis on San Nicolas Island.

In the following discussion we present prelim-inary results of a long-term research project that focuses on the spatial modeling of human activ-ities that took place at the late Holocene village site of CA-SNI-"' on San Nicolas Island. We used GIS and cluster analysis to further examine the associations of artifacts, ecofacts, and features at the site. !e statistical clustering of sandstone saws and circular shell .shhooks, for instance, led us to con.rm the associations of these arti-facts demonstrated by recent experimental and analytical methods (Kendig et(al. "#%#). We also examined the pattern of distribution of red ochre within one locus of the site that shows evidence of ceremonial activity. Since ochre is o/en asso-

ciated with ritualized behavior, we focused on its distribution around ceremonial features at CA-SNI-"'. !is approach has allowed us to shed light on the possible linkages of artifacts, as well as patterns of spatial and temporal variability in technology, subsistence, and behavior at the site.

Background

San Nicolas is the most remote of California’s eight Channel Islands (Figure $.%). It lies approx-imately %"# km (*' mi) southwest of Los Angeles and is approximately &) km ($# mi) from the nearest point on the mainland (Martz "##)). Al-though it is one of the four southern Channel Is-lands, it is centrally located between the northern and southern island groups. San Nicolas is rel-atively small, measuring only %-.' km long and '( km wide, and has a maximum elevation of "**(m. !e island is characterized by .ve topo-graphic zones: the large, relatively 0at central pla-teau; the northern, relatively steep southern cli+s; the northern and southern coastal terraces; and the dune .elds of the west end (Martz "##").

Geologically, the island is composed primar-ily of alternating beds of Eocene sandstones, silt stones, and interbedded marine sediments consisting of a conglomerate of metavolcanic and meta sedi mentary cobbles within a matrix of sand stone and mudstone (Vedder and Norris %&$,: %,). Periodic exposures of these conglomer-ate beds and sandstone shingle beaches serve as primary sources of tool stone.

Jazwa_text_1.indd 97 4/25/13 10:36 AM

98 GUTTENBERG ET AL.

San Nicolas Island sandstone is o/en highly indurated, dense, and extremely hard (!omas- Barnett "##-), making it useful as a raw ma-terial for both ground stone and expedient 0ake stone tools (Kendig et(al. "#%#). Constant cortical weathering of sandstone shingles within the is-land’s surf zone produces an abundance of nat-urally polished sandstone cobbles thought to be used in the production of the sandstone saws dis-cussed below (Kendig et(al. "#%#; Rogers %&,#).

!e island’s size, isolation, lack of diverse hab-itats, and relatively arid landscape support a lim-ited variety of 0ora, primarily coastal sage scrub, chaparral, and grasses, with few edible endemics (Junak "##)). Terrestrial native fauna are likewise limited to the island fox, deer mouse, island night lizard, land snails, and a variety of insects. None are particularly useful as sources of raw material or sustenance.

In contrast to its sparse terrestrial resources, San Nicolas is surrounded by highly productive and diverse marine habitats. !e island’s sur-roundings are home to the most extensive kelp forests per capita among the California Channel Islands and support a wide variety of shell.sh, .sh, marine mammals, and seabirds (Mariani "##%; Pondella et(al. "##'; Schoenherr et(al. %&&&: ,-'–,-$).

Culture History

Although the Channel Islands’ oldest archaeo-logical sites are believed to have been inundated by rising sea levels a/er the last glacial maximum (Inman %&),; Nardin et(al. %&)%; Curray %&$'), ''% sites have been identi.ed throughout the island (Martz "##", "##'). A radiocarbon date ()-## cal 12) from a single Mytilus sp. shell from CA-SNI-,,& represents the oldest known human hab-itation on San Nicolas Island. Other radiocarbon dates from sites representing villages, camps, seafood processing locations, and 0ake stone re-duction areas indicate an increase in human hab-itation from the early to late Holocene (Martz "##':$'). At the time of contact, the Nicoleño spoke a language from the Takic family within the Uto-Aztecan linguistic group, associating them with Gabrielino (Tongva) groups on the southern islands (Munro %&&&).

Tule Creek Village (CA-SNI-25)

!e Tule Creek site (CA-SNI-"') is a large late Holocene village situated on the north end of the central plateau (Figure $.%). !e site over-looks Corral Harbor, a calm inlet that would have served as a favorable canoe launch (Bowers %)&#: '*). !e site is adjacent to Tule Creek and ap-proximately " km southeast of !ousand Springs,

FIGURE 6.1. San Nicolas Island, showing topographic zones and CA-SNI-25.


GIS as a Tool for Analyzing Intrasite Spatial Variability on San Nicolas Island, CA 99

both of which are excellent sources of fresh water. Malcolm Rogers (%&,#) conducted the earliest formal investigations of CA-SNI-"', characteriz-ing it as a large village site with house pits, a com-munal structure, at least one cemetery, 0aked stone tools, and a dense midden with numerous features that are well preserved (Rogers %&,#). Be-cause of these features, the Tule Creek site serves as an excellent example for exploring intrasite or-ganization and use of space in both daily and cer-emonial contexts.

More recent .eld investigations at CA-SNI-"' began in "##% and were completed in "##&. Our .eldwork was focused primarily on two loci: East Locus and Mound B. We identi.ed and tested sev-eral other loci and have incorporated these data into our analysis as well; however, we chose East Locus and Mound B because they were sampled using contiguous block excavations, which are conducive to identifying behavioral patterns (fea-tures) in the archaeological record.

We found numerous features at East Locus that are likely associated with ritualized behavior and ceremonial events. !e space appeared to be organized and included several dog burials,

multiple hearths, discrete pits, balancing stone caches, ochre, and other features and artifacts (Cannon "##$) (Figure $."). At Mound B, how-ever, we encountered more domestic activities that were deposited during at least three major periods of habitation. We used more than *# radio carbon dates on well-preserved single pieces of shell, bone, and charcoal to establish a chronology for the site, suggesting it was inhab-ited between about '### cal 12 and '## cal 12. Although the site was occupied intermittently for about ',### years, the major village occupa-tion occurred between about )## and '## cal 12 (Kendig et. al "#%#) (Table $.%).

Spatial Studies and GIS in Archaeology

Over the last '# years, archaeologists have in-corporated di+erent methods of spatial analysis into archaeological research designed to under-stand the distribution of sites, features, and ar-tifacts (Hodder and Orton %&*$; Dacey %&*,; Bartlett %&*-; Whallon %&*,, %&*-). Spatial statis-tics (e.g., nearest neighbor analyses) developed by researchers in other .elds have been adopted and applied to a variety of archaeological settings

FIGURE 6.2. Open excavation of CA-SNI-25, showing East Locus hearths and pits.


TAB

LE 6

.1. R

adio

carb

on d

ates

from

CA-

SNI-2

5

Sam

ple

#U

nit

and

Con

text

Stra

tum

/Le

vel

Dep

th

(cm

)M

ater

ial

Unc

orre

cted

14

C A

ge (

BP)

Cal

ibra

ted

Age

Ra

nge

(cal

BP)

, 1

Sigm

aa

Cal

ibra

ted

Age

Ra

nge

(BP)

, 2

Sigm

aa

East

Loc

us b

OS-

5441

1U

nit 8

, Fea

ture

6 (

hear

th)

IV/2

70C

harc

oal

175

± 30

220–

170

225–

73

OS-

5456

2U

nit 7

, Fea

ture

9 (

hear

th)

V/1

60–7

0C

harc

oal

395

± 70

510–

430

530–

306

OS-

5435

5U

nit 7

H, fi

shin

g ta

ckle

kit

feat

ure

II/1

18H

. cra

cher

odii

1090

± 3

552

0–45

059

6–39

7

OS-

5439

7U

nit 7

L, F

eatu

re A

(pi

t)IIB

/163

M. c

alifo

rnia

nus

5700

± 3

559

10–4

780

5966

–572

1

OS-

5533

6U

nit 8

E2, l

eft-

hand

ed fi

shho

okII/

4—

Nor

risia

nor

risi

1180

± 3

551

2–59

863

7–48

6

OS-

5546

5U

nit 8

A, s

addl

e be

adI/

1—

Oliv

ella

bip

licat

a11

80 ±

35

512–

598

637–

486

OS-

6678

9Pi

t 7Q

II/

812

5H

. cra

cher

odii

900

± 30

274–

379

439–

245

OS-

6691

0D

og b

uria

l—

—Ca

nis

fam

iliar

is68

0 ±

2557

0–67

067

7–56

3

Mou

nd B

b

OS-

5435

4U

nit 1

1, to

p of

mou

ndII/

346

H. c

rach

erod

ii88

0 ±

3036

0–26

042

8–12

8

OS-

5441

3U

nit 5

2, s

hell

and

lithi

c fe

atur

eI/

363

H. c

rach

erod

ii22

5 ±

3552

0–46

056

4–40

0

OS-

5436

0U

nit 5

8, p

it fe

atur

eIIB

/193

H. c

rach

erod

ii47

50 ±

35

4810

–467

048

35–4

569

OS-

5435

8U

nit 5

7, h

eart

h fe

atur

eI/

394

H. c

rach

erod

ii48

00 ±

30

4840

–477

049

00–4

623

OS-

5435

7U

nit 5

5, p

it fe

atur

eII/

3–IIB

/185

H. r

ufes

cens

4890

± 3

549

40–4

830

5030

–479

6

Mou

nd A

c

Beta

-116

352

Inde

x un

it-

90–1

00C

harc

oal

130

± 40

147–

6215

2–54

Beta

-116

920

Inde

x un

it-

20–3

0C

harc

oal

550

± 50

631–

522

651–

509

Beta

-116

351

Inde

x un

it-

50–6

0C

harc

oal

650

± 90

672–

553

733–

513

a All

date

s w

ere

calib

rate

d us

ing

Cal

ib 6

.02

(Rei

mer

et a

l. 20

09)

and

appl

ying

a !

R of

225

± 3

5 fo

r all

shel

l sam

ples

(se

e Ke

nnet

t et a

l. 19

97).

b Dat

es p

rovi

ded

by D

r. Re

né V

ella

now

eth,

Dep

artm

ent o

f Ant

hrop

olog

y, C

alifo

rnia

Sta

te U

nive

rsity

, Los

Ang

eles

.c D

ates

pro

vide

d by

Dr.

Patr

icia

Mar

tz, D

epar

tmen

t of A

nthr

opol

ogy,

Cal

iforn

ia S

tate

Uni

vers

ity, L

os A

ngel

es.



(Clark and Evans %&'-; Cox %&)%; Getis %&$-; Getis and Franklin %&)*; Getis and Ord %&&"; Hodder and Orton %&*$; Ord %&*'; Pinder et( al. %&*&). Researchers began using these methods to bet-ter interpret spatial patterning in the distribu-tion of sites, features, and artifacts (Bartlett %&*-; Dacey %&$,, %&*,; Hodder and Orton %&*$; Whal-lon %&*,, %&*-).

It is important to note that we do not imply that the spatial statistics performed by GIS are a new development. In fact, the operations o+ered by GIS are well established and are not intended as a replacement for nonspatial statistics per-formed by other so/ware programs (Sharon et(al. "##-). Rather, we demonstrate that GIS can be used in tandem with nonspatial statistics and other forms of analysis as a method for testing and con.rming hypotheses about relational and spatial associations. !e intent of our research is to utilize GIS to enhance nonspatial statistical methods on data that are the subject of ongoing research at CA-SNI-"'. We further contend that these methods may be e+ectively used in archae-ological research throughout the Channel Islands and in other contexts where large, open-area ex-cavations have yielded an abundance of data.

GIS in Archaeology on San Nicolas Island

Other studies using GIS and spatial analysis have been conducted on San Nicolas Island (A.. "###; Casaus %&&); Martz "##"; Merrill "##-). !ese studies focused on island-wide landscape analysis, comparing site and resource distribu-tions. In the %&&#s, Martz ("##") conducted the initial GIS work, which involved pedestrian sur-veys of the island and site mapping using a global positioning system (GPS). !is work led to the creation of GIS data for the island’s archaeologi-cal sites and features. Data from that initial sur-vey was incorporated into our current research. Additional work used the compiled GIS data to describe settlement patterns on the island (A.. "###). More recent spatial studies have been con-ducted at CA-SNI-"' using mathematical models as well as statistical analysis (Merrill "##-; Can-non "##$).

Methods

For our work at CA-SNI-"', we constructed a map series of the site showing distributions and statistical clustering of excavated data using Arc-

GIS spatial autocorrelation and hot spot analysis (Getis and Ord %&&"; Hodder and Orton %&*$; Mitchell "##&). We .rst composed a base map using analog survey data collected in "##- com-bined with GPS coordinates of datum points and unit locations recorded in the .eld in "#%#. Loca-tional data were collected using a Trimble GeoX with a Zephyr antenna, providing an accuracy of ±%% cm a/er di+erential correction. We tied in datum points from the "##- survey (Merrill "##-) with additional coordinates of unit corners and feature locations to increase the accuracy of our .eld data.

We transposed the "##- survey data by calcu-lating the o+set coordinates and combining them with recent georeferenced data, which converted our original analog data into a digital format. !is process provided a base map showing the datum points and the outline of the unit grids in each locus (Figure $.,). We then digitized unit grids using the Delta XY editing function and created individual polygons of exact size. !e result was a direct representation of locus grids with accu-rate unit dimensions. !e data were summarized and generalized to the unit grids of each respec-tive locus. Data summarized into units (mostly %(3 % m) are treated in the same manner as popu-lation data in census tracts. Distance is measured from the unit centroid and must be analyzed in the context of scale, which in this case is on the locus level (Mitchell "##&).

With the locational data reconciled and fully digitized, we performed a data visualization analysis and conducted statistical operations us-ing the ArcGIS &., spatial statistics toolbox. Data visualization consisted of displaying the summa-rized data in distribution maps of features across each locus. Data were symbolized and displayed stratigraphically. Statistical operations included Moran’s I spatial autocorrelation and Getis-Ord Gi* hot spot analysis (Getis and Ord %&&"; Hod-der and Orton %&*$; Mitchell "##&).

Spatial autocorrelation is an operation that shows statistical clustering at de.ned distance intervals. Hodder and Orton (%&*$) discussed early applications of spatial autocorrelation and described the method in terms of scale and mea-sure of distance between features (see also Whal-lon %&*,, %&*-). In ArcGIS, the Moran’s I spatial autocorrelation tool is designed to run at di+er-ent parameters to determine the distance interval



(Mitchell "##&). Our replicative studies infor-mally show debris .elds from various tool manu-facturing activities at approximately a " m radius (Kendig et(al. "#%#). !us, we expected our dis-tance interval to be around " m. !e results from the spatial autocorrelation met our expectations, and our distance intervals were set between ".' and , m.

Hot spot analysis is based on the Getis- Ord- Gi* statistic (Getis and Ord %&&"; Mitchell "##&). !is method shows the patterns and locations of high and low values of statistical clustering based on the distance interval de.ned by the spatial au-tocorrelation tool. !e values of clustering are ac-tually a z-score that is produced as a result of the operation (Mitchell "##&). Statistically positive z-scores show visual clustering of features on the map, and small or negative z-scores show areas with little or no clustering (Figure($.-).

Both statistical operations are based on Eu-clidean distance and nearest neighbor statistics that imply spatial context between features (Clark and Evans %&'-; Dacey %&$,, %&*,; Getis %&$-; Getis and Franklin %&)*; Getis and Ord %&&"; Mitchell "##&). Other authors have described the problems and limitations of nearest neighbor sta-tistics in archaeological research (Merrill "##-; Merrill and Read "#%#). We maintain, however, that these techniques are useful in archaeology,

especially when combined with other methods such as replicative experiments and nonspa-tial statistics that suggest relationships between variables.

Results

Our sample from CA-SNI-"' is ideal for intrasite GIS analysis. Large, open-area excavations allow for spatial dissection of the site across a temporal plane. Good preservation, excellent stratigraphic integrity, and numerous radiocarbon dates pro-vide a solid foundation for spatial studies. By ex-amining the distribution of artifacts as well as pigments, we highlight the utility of intrasite GIS as a tool to understand a variety of past cultural phenomena.

For instance, we examined the distribution of sandstone saws relative to shell .shhooks and hook making debris to con.rm spatially what we derived experimentally about the use of sand-stone saws as abraders for manufacturing shell .shhooks (Table $.") (Kendig et(al. "#%#). Doc-umenting the spatial distribution of saws and hooks, and then analyzing their clustering across the site, allowed us to make inferences about the organization of shell .shhook production, one of the most important technological innovations of the late Holocene. To examine the role of pig-ments within a ceremonial context, we followed

FIGURE 6.3. CA-SNI-25 site map.



the distribution of red ochre (hematite) relative to ceremonial features such as pits, hearths, dog burials, and balancing stones. Understanding the distribution of ochre at East Locus has allowed us to address questions about how pigments were used in the past and how they were incorporated into ritualized activities, topics of global signi.-cance considering the great antiquity of pigment use by humans. In the discussion section we ex-plore the application of these methods to data sets collected years ago and housed in museums, uni-versities, and government agencies worldwide, but .rst we present the results of our study at CA-SNI-"'.

Saws and Hooks

Once all spatial data were digitized and standard-ized, we performed a spatial autocorrelation and hot spot rendering of the distributions of sand-stone saws and .shhooks across CA-SNI-"'. In the stratigraphic visualization analysis of our data, we observed the greatest distribution of saws and hooks in Stratum II, which is consistent with the density of materials found throughout the site, particularly at East Locus (Table $.") (Can-non "##$). Saws were made of extremely hard sandstone cortical 0akes with proximal edges thinned and straightened through the removal of striking platforms. Many of the saws were

FIGURE 6.4. Example of output from spatial autocorrelation tool in ArcGIS showing areas of statistical significance.

TABLE 6.2. Stratigraphic distribution of fishhooks, ochre, and saws at East Locus

StratumFishhooks

(n)

Ochre Sandstone Saws (n)Count Weight (g)

Stratum I 190 129 169.43 25

Stratum II 56 814 510.47 47

Stratum III 0 0 0

Undifferentiated/Features 26 89 226.49 9

Totals 272 1,032 906.39 83



utilized on all margins and both faces, suggest-ing their use as cutting implements and abrad-ers (Figure $.'a). Red abalone (Haliotis rufescens), once common in the deeper waters around San Nicolas, were the predominant material used for .shhook manufacture. Hooks were also cra/ed from black abalone (Haliotis cracherodii), mus-sel (Mytilus californianus), and Norris’ top-shell (Norrisia norrissi) (Strudwick %&)$) (Figure $.'b).

!e autocorrelation function indicated posi-tive statistical clustering of both saws and hooks. !e .shhooks showed clustering at a distance interval of ".' m, and the saws were clustered at ,.#(m. !e distance intervals provided by the spa-tial autocorrelation were the parameters used to conduct the hot spot analysis. !e hot spot map of saws shows clustering in roughly the same area of East Locus as the clustering of .shhook materials. !e areas around the hearth and pit features show hot spots with positive z-scores greater than ".#. Although both saws and hooks occur at Mound B and other loci, the pattern of distribution of these two artifact types is stronger at East Locus (Figure $.$).

Ochre

Red ochre is found in the island’s sedimentary deposits and occurs throughout the site, but for this study we focused on the distribution of ochre around areas at East Locus with evidence of ceremonial activity (Cannon "##$). !e visual analysis of red ochre revealed that it occurs in

Stratum I at East Locus and Locus D, but the greatest distribution of ochre is clearly seen in Stratum II of East Locus (Figure $.*). Here we see ochre distributed around the pit features at the north end of East Locus, as well as along the east side of the largest hearth.

We analyzed the occurrence of ochre at East Locus based on both the spatial distribution of the weight and number of individual pieces. !e spatial autocorrelation operations performed for ochre suggested statistical clustering by count at a distance interval of ".' m, but its distribution by weight was randomly dispersed. !e di+er-ence between clustering patterns by weight and total count are likely indicative of a problematic characteristic of the ochre sample. !e numerous ochre fragments contribute to a greater distribu-tion than the comparative measure of ochre by weight (Hodder and Orton %&*$:-%). !e sample may either demonstrate a positively skewed pat-tern of clustering or be indicative of the wide-spread use of ochre across East Locus. !e hot spots of red ochre at East Locus show clustering in areas around features believed to be associated with ceremonialism, such as the hearths, pits, and dog burials (Figure $.*) (Bartelle et(al. "#%#; Can-non "##$; Vellanoweth et(al. "##)).

Discussion

Our approach to using intrasite GIS provides a visual and interactive platform conducive to spa-tial studies in archaeology. By digitizing spatial

FIGURE 6.5. Sandstone saw morphology (left) and fishhook production sequence.


FIGURE 6.6. Hot spots of hooks and saws at East Locus in Stratum II (top). Hooks are clustered with a distance interval of 2.5 m, saws with a distance interval of 3 m; (bottom) hot spots of total distribution of hooks and saws at East Locus, all strata. Both hooks and saws are clustered at a distance interval of 2.5 m.


FIGURE 6.7. Distribution of ochre by weight (g) and total count at East Locus in Stratum II (top); hot spots of ochre by count in Stratum II at a distance interval of 2.5 m (bottom).



data from .eld notes and sketches, transit surveys and maps, and other analog forms, combining these with new georeferenced points, reconciling scales, and analyzing the distribution of archaeo-logical .nds, we created a powerful tool for exam-ining the archaeological record. !e correlations between saws and hooks, as well as ochre rela-tive to ceremonial features, were made clear by GIS analysis. Although connecting spatially as-sociated artifacts (saws/hooks) to functionality requires a variety of descriptive and analytical ap-proaches, the results of our intrasite GIS analysis suggest that it is e+ective at .nding and display-ing in-ground relationships. Documenting and analyzing in situ spatial associations of objects and features form the basis for making behav-ioral inferences about how people organized and divided up their space to perform individual, so-cial, and communal activities.

Our spatial analysis of CA-SNI-"' con.rms previous studies on the association between saws and hooks, as well as assumptions about the role of ochre among Native Americans. Kendig et(al. ("#%#), for example, suspected an association be-tween saws and hooks based on .eld observa-tions and literature searches, and then tested it with a series of replicative experiments, use-wear analyses, nonspatial statistics, and other analyti-cal studies. !e intrasite exploratory analysis de-scribed above strengthens the hypothesis that these two artifact types share not only a spatial association, but a functional one as well. Further-more, the examination of red ochre across East Locus showed some interesting characteristics and raised questions about di+erences in the dis-tribution of small fragments compared to weight. Our results suggest that although the use of ochre at East Locus was widespread, there was signif-icant clustering of pigment in association with hearths and pit features.

Our focus on the potential use of sandstone saws in circular and J-shaped shell .shhook man-ufacture has additional implications for Channel Islands research. !e introduction of the circular shell .shhook on San Nicolas and other Channel Islands was a signi.cant factor in the rise of social and economic complexity observed in archaeo-logical and ethnographic records (Arnold "##%; Kennett "##'; Strudwick %&)$). !e innovation of the shell .shhook expanded the productivity of

o+shore .sheries used by Native Americans oc-cupying the Channel Islands as well as the coastal mainland. !e development of this technology was related to reliance on .shing as an e+ective subsistence strategy and contributed to other cul-tural developments observed in the archaeologi-cal record on San Nicolas and the other islands.

Evidence of pelagic .sh remains and cere-monial feasting at CA-SNI-"' suggests an in-creased production from .shing harvests and a potential for food surplus (Bartelle et(al. "#%#). Martz ("##') argues that this development par-allels a rise in population levels and densities on San Nicolas Island. Furthermore, this technology is believed to have set the stage for the develop-ment of cra/ specialization, the accumulation of personal wealth, and elite-based social strati.ca-tion. !ese developments go hand in hand with the emergence of the plank canoe (tomol) as a ve-hicle of food procurement, trade, and travel be-tween the islands and the mainland, perhaps the most signi.cant factors in0uencing the rise of so-cial complexity documented throughout coastal and insular Southern California.

Inter-Island Implications

!ere are many implications for further research using GIS elsewhere on the Channel Islands. !e application of GIS exploratory analysis allows for the visual display of spatial patterns that may not be immediately apparent in large data sets. Furthermore, intrasite GIS provides a de.nitive direction for additional research on previously analyzed data. It may also prove useful to com-bine this type of intrasite approach with regional landscape analyses to examine source locations of materials and networks of trade and exchange. For instance, the spatial patterning of exotic trade materials (e.g., obsidian) on an intrasite scale may shed light on the nature and 0ow of trade goods between the islands and the mainland (Arnold "##%; Cannon "##$; Rick et(al. "##%; Vellanoweth "##%).

Channel Islands sites excavated prior to the development of GIS could bene.t from the ap-plication of the methods used in our study. If sites excavated many years ago have not been analyzed, or only partially analyzed, then intra-site GIS can be used to predict and guide fruit-ful avenues of research. Sites such as Lemon Tank



(CA-SCLI-%'"-) and Eel Point (CA-SCLI--,) on San Clemente Island and the El Montón site (CA-SCRI-,,,) on Santa Cruz Island are good candi-dates for intrasite spatial analysis using GIS and updated georeferenced data.

!e Lemon Tank site, for example, has excel-lent potential for exploratory analysis in compar-ison to the Tule Creek site. Both sites have similar chronologies and have produced an abundance of data indicative of ceremonial activity. Hale (%&&') has documented numerous features at the Lemon Tank site found within a ceremonial complex similar to those at East Locus, including dog burials, caches of ochre and seeds, artifacts made of exotic stone such as chert and steatite, hearths, and .shing kits (see Perry, Chapter )). Many of these features exist within a ceremonial enclosure, which Hale has posited as being sim-ilar to those described in the toloache ceremony, or boy’s puberty rite. Such ceremonial com-plexes are characteristic of Chingichngish, a re-ligion practiced by Southern California peoples (see Boscana %&,,; DuBois %&#); McCawley %&&$; Raab et(al. "##&). A similar sacred enclosure, or yovaar, is described at the Isthmus Cove site (CA-SCAI- ,&) on Santa Catalina Island in the ethno-historic account of Sebastian Vizcaino’s %$#" expedition (McCawley %&&$). We have consid-ered these possibilities as well at East Locus. !e abundance of comparable features and artifacts at Lemon Tank and East Locus, coupled with a spatial analysis of both sites, would undoubtedly shed new light on possible religious ties between the island peoples, as well as connections with their mainland neighbors.

Perry ("##*) described archaeological evi-dence of ritual activity on the northern Channel Islands from a landscape perspective. Sacred lo-cales and shrines were viewed as part of a geo-graphic and cultural landscape, ideal for using GIS at a regional scale. Perry (Chapter )) also de-scribed intrasite ritual correlates on San Clemente Island similar to what we observed at CA-SNI-"'. Our intrasite spatial analysis at CA-SNI-"' pro-vided insight into archaeological signatures of rit-ual activity that can be applied elsewhere on both the northern and southern islands.

Teeter, Martinez, and Richardson ( Chapter &) have also applied GIS to indigenous trail sys-tems across Santa Catalina Island. Using ethno-

graphic data and a digital elevation model to produce a least-cost path, the authors suggested that trails functioned as more than just conduits of goods and information; trails also maintained social bonds, alliances, and geographic connec-tions to sacred locales. !is research on Catalina Island and Perry’s work on Santa Cruz Island are both excellent examples of regional landscape approaches that can be used in tandem with our intrasite GIS work on San Nicolas Island. Under-standing patterns of personal and social space on the Channel Islands requires both intrasite and landscape approaches, which are not mutually exclusive but complementary.

Conclusions

Perhaps one of the least understood aspects of the archaeology of the Channel Islands is the or-ganization and use of space by native peoples in both domestic and ritualized contexts. !e types of data necessary for such research demand large contiguous samples, excellent stratigraphic and chronological control, and the application of spa-tial analysis. It is also necessary to develop frames of reference (Binford "##%) for interpreting the spatial association of objects and features, and linking them to the dynamic behavioral and cul-tural processes that formed them (Binford %&)#). Data sets derived from experimental, replicative/use-wear analyses, and ethnohistoric approaches, as well as computer modeling, provide excellent frames of reference and will prove useful in fu-ture studies on the Channel Islands. Despite the abundance of well-preserved sites on the Channel Islands, few have been excavated and analyzed using these approaches.

By combining the types of spatial analysis pre-sented in this chapter with existing site-speci.c and regional data (see Perry "##* and Chapter )), it will be possible to model the use of space by native peoples through time on the Channel Is-lands. Perry’s ("##* and Chapter )) survey and synthesis of ceremonial sites on Santa Cruz Island provides an excellent backdrop against which site-speci.c spatial data can be compared. Mod-eling site-speci.c, island-wide, and regional data regarding the use of space for ceremonial and every day purposes has the potential to illuminate di+erential land use patterns between the north-ern and southern Channel Islands.



We are hopeful that the new directions of in-trasite spatial analysis will help to dispel some of the previous critiques of the use of GIS in archae-ology. O/en described as overgeneralizing and deterministic, GIS analysis has fallen under at-tack by critics implying that digital methods of analysis dehumanize human behavior (Conolly and Lake "##$; Kvamme %&&&; Lock and Stancic %&&'; McCoy and Ladefoged "##&; Wheatley "##-). Although these authors make some valid points, we argue that new technology is seeking to revive and enhance the spatial archaeology of the %&$#s and %&*#s (Holdaway and Wandsnider "##$; Jones and Munson "##'; McCoy and Lade-foged "##&; Wheatley and Gillings "##":",$). Our approach of integrating intrasite GIS with multiple methods of analysis retains the hu-manistic component of archaeological research necessary to properly interpret past human be-havior (Challis and Howard "##$; Kvamme "##$; Kvamme et(al. "##$; Rigaud and Simek %&&%).

If shown to be successful, predictive models may be integrated into policies for managing cul-tural resources on the Channel Islands. !is is particularly relevant to other sites on the islands where large, open-area excavations have yielded an abundance of data. At some point, we expect that georeferenced data will be generated for all Channel Island sites investigated in detail, allow-ing for intra-island and inter-island comparisons. Salient topics of interest — such as the develop-ment of innovative technologies, trade networks, cra/ specialization, the rise of social complexity, population demographics, and a host of other questions relevant to both the northern and southern islands, as well as to hunter-gatherer so-cieties elsewhere in the world — can be integrated. By incorporating northern and southern island data sets, we can better understand the complex-ity and dynamics of the prehistoric interactions that took place on the California Channel Islands.

AcknowledgmentsWe thank the U.S. Navy for its continued support for and access to the archaeological resources on San Nicolas Island. Excavation and lab work was funded by NAVAIR Weapons Division, Range Sustainability O4ce, Point Mugu, California, as well as Humboldt State University (HSU) and California State University, Los Angeles (CSULA). We thank Lisa !omas-Barnett (NAVAIR) and Catherine Girod at the Naval Base of Ventura County for their assistance and logistical sup-port. Ray Corbett (Santa Barbara Museum of Natural History) and Jennie Allen (CSULA) assisted with ed-iting early versions of this chapter, and Kevin Smith (CSULA) o+ered insight and recommendations for our analysis. We also thank Mark Raab and an anonymous reviewer for their comments. Finally, thanks to Chris Jazwa and Jennifer Perry for all of their hard work in assembling this volume.

References CitedA.., A."### Prehistoric Settlement Pattern on San Nicolas

Island, California: A GIS Application. MA thesis. Department of Anthropology, Califor-nia State University, Los Angeles.

Arnold, J. E. (editor)"##% !e Origins of a Paci"c Coast Chiefdom: !e

Chumash of the Channel Islands. University of Utah Press, Salt Lake City.

Bartelle, B. G., R. L. Vellanoweth, E. S. Netherton, N. W. Poister, W. E. Kendig, A. F. Ainis, R. J. Glenn, J. V. Marty, L. !omas-Barnett, and S. J. Schwartz"#%# Trauma and Pathology of an Articulated

Dog(from San Nicolas Island, California, U. S. A. Journal of Archaeological Science ,*(%"):"*"%–"*,-.

Bartlett, M. S.%&*- !e Statistical Analysis of Spatial Pattern.

Advances in Applied Probability $("):,,$– ,').

Binford, L. R.%&)# Willow Smoke and Dogs’ Tails: Hunter-

Gatherer Settlement Systems and Archaeo-logical Site Formation. American Antiquity -'(%):-–"#.

"##% Constructing Frames of Reference: An Analyti-cal Method for Archaeological !eory Building Using Hunter-Gatherer Data Sets. University of California Press, Berkeley.

Boscana, G.%&,, Chinigchinich: A Revised and Annotated Ver-

sion of Alfred Robinson’s Translations of Fa-ther Gerónimo Boscana’s Historical Account of the Belief of Usages, Customs and Extrav-agancies of this Mission of San Juan Capist-rano Called the Acagchemem Tribe. Fine Arts Press, Santa Ana, CA.



Bowers, S.%)&# San Nicolas Island. In California State Mining

Bureau Ninth Annual Report of the State Min-eralogist for the Year Ending December #, #$$%. Sacramento, CA.

Cannon, A. C."##$ Giving Voice to Juana María’s People: !e

Organization of Shell and Exotic Stone Arti-fact Production and Trade at a Late Holocene Village Site on San Nicolas Island, California. MA thesis. Department of Anthropology, Humboldt State University.

Casaus, K. R.%&&) Use of GIS for Natural and Cultural Resource

Management: A Computerized Rule-Based Activity Planning System on San Nicolas Island, Point Mugu Naval Air Weapons Sta-tion. MA thesis. School of Renewable Natural Resources, University of Arizona, Tucson.

Challis, K., and A. J. Howard"##$ A Review of Trends Within Archaeological

Remote Sensing in Alluvial Environments. Archaeological Prospection %,:",%–"-#.

Clark, P. J., and F. C. Evans%&'- Distance to Nearest Neighbor as a Measure of

Spatial Relationships in Populations. Ecology ,'(-):--'–-',.

Conolly, J., and M. Lake"##$ Geographical Information Systems in Ar-

chaeology. Cambridge University Press, Cambridge.

Cox, T. F.%&)% Re0exive Nearest Neighbors. Biometrics

,*("): ,$*–,$&.Curray, J. R.%&$' Late Quaternary History: Continental

Shelves of the United States. In !e Qua-ternary of the United States, edited by H. E. Wright and D. G. Frey, pp. *",–*,'. Princeton University Press, Princeton, NJ.

Dacey, M. F.%&$, Order Neighbor Statistics for a Class of Ran-

dom Patterns in Multidimensional Space. Annals of the Association of American Geog-raphers ',(-):'#'–'%'.

%&*, Statistical Tests of Spatial Association in the Locations of Tool Types. American Antiquity ,)(,):,"#–,").

DuBois, C. G.%&#) !e Religion of the Luiseño Indians of South-

ern California. University of California Publi-cations in American Archaeology and Ethnol-ogy )(,):$&–%)$.

Getis, A.%&$- Temporal Land-Use Pattern Analysis with

the Use of Nearest Neighbor and Quadrat Methods. Annals of the Association of Ameri-can Geographers '-(,):,&%–,&&.

Getis, A., and J. Franklin%&)* Second-Order Neighborhood Analysis of

Mapped Point Patterns. Ecology $)(,):-*,–-**.Getis, A., and J. K. Ord%&&" !e Analysis of Spatial Association by Use

of Distance Statistics. Geographical Analysis "-(,):%)&–"#$.

Hale, A.%&&' !e World in a Basket: Late Period Gabrieliño

Ceremonial Features from the Lemon Tank Site, San Clemente Island, CA. MA thesis. California State University, Northridge.

Hodder, I., and C. Orton%&*$ Spatial Analysis in Archaeology. Cambridge

University Press, Cambridge.Holdaway, S. J., and L. Wandsnider"##$ Temporal Scales and Archaeological Land-

scapes from the Eastern Desert of Australia and Intermontane North Africa. In Confront-ing Scale in Archaeology: Issues of !eory and Practice, edited by G. Lock and B. Moly-neaux, pp. %),–"#". Springer, New York.

Inman, D. L.%&), Application of Coastal Dynamics to the

Reconstruction of Paleocoastlines in the Vi-cinity of La Jolla, California. In Quaternary Coastlines and Marine Archaeology: Toward the Prehistory of Land Bridges and Conti-nental Shelves, edited by P. M. Masters and N. C. Flemming, pp. %–-&. Academic Press, London.

Jones, G., and G. Munson"##' Geophysical Survey as an Approach to the

Ephemeral Campsite Problem: Case Studies from the Northern Plains. Plains Anthropolo-gist '#(%&,):,%–-,.

Junak, S."##) A Flora of San Nicolas Island, California.

Santa Barbara Botanic Garden, Santa Bar-bara, CA.

Katsianis, M., S. Tsipidis, K. Kotsakis, and A. Kousou-lakou"##) A ,D Digital Work0ow for Using Intra-Site

Research in GIS. Journal of Archaeological Science ,':$''–$$*.

Kendig, W. E., K. N. Smith, R. L. Vellanoweth, J. A.( Allen, C. M. Smith, and A. M. Points"#%# !e Use of Replicative Studies to Understand

the Function of Expedient Tools: !e Sand-stone Saws of San Nicolas Island, California. Journal of California and Great Basin Archae-ology ,#("):%&,–"%#.



Kennett, D. J."##' !e Island Chumash: Behavioral Ecology of

a Maritime Society. University of California Press, Berkeley.

Knabb, K. A."##) Understanding the Role of Production

and Cra/ Specialization in Ancient Socio- Economic Systems: Toward the Integration of Spatial Analysis, ,D Modeling and Virtual Reality in Archaeology. MA thesis. Depart-ment of Anthropology, University of Califor-nia, San Diego.

Kvamme, K. L.%&&' A View From Across the Water: !e North

American Experience in Archaeological GIS.(In Archaeology and Geographical In-formation Systems, edited by Gary Lock and Zoran Stancic, pp. %–%-. Taylor and Francis, London.

%&&& Recent Directions and Developments in Geographic Information Systems. Journal of Archaeological Research *("):%',–"#%.

"##$ !ere and Back Again: Revisiting Archaeo-logical Site Locational Modeling. In GIS and Archaeological Site Location Modeling, edited by Mark w. Meher and Konnie L. Wescott, pp. ,–,). Taylor and Francis, Boca Raton, FL.

Kvamme, K. L., J. K. Johnson, and B. S. Haley"##$ Multiple Methods Survey: Case Studies. In

Remote Sensing in Archaeology: An Explicitly North American Perspective, edited by J. K. Johnson, pp. "'%–"$*. University of Alabama Press, Tuscaloosa.

Lawson, K. S."##* De.ning Activity Areas in the Early Neolithic

Site at Foeni-Salas(Southwest Romania): A Spatial Analytic Approach with Geographi-cal Information Systems in Archaeology. MA thesis. Department of Anthropology, Univer-sity of Manitoba, Winnipeg, Canada.

Lie+, S."##, Applications of Geographic Information Sci-

ence in the Archaeological Research of the Fincastle Kill Site (DIOx ') Alberta, Canada, and Tel Beth-Shemesh, Israel. MA thesis. De-partment of Geography, University of Leth-bridge, Alberta, Canada.

Lock, G., and Z. Stancic%&&' Archaeology and Geographical Information

Systems. Taylor and Francis, London.Mariani, R. R."##% Middle and Late Holocene Fishing Strategies

on San Nicolas Island, California. M.A thesis. Department of Anthropology, California State University, Los Angeles.

Martz, P. C."##" San Nicolas Island Prehistoric Archaeological

Sites Mapping and Recordation Project. Pre-pared for Naval Air Weapons Station, China Lake, CA. On .le, South Central Coastal Ar-chaeological Information Center, California State University, Fullerton.

"##' Prehistoric Context and Subsistence on San Nicolas Island. Proceedings of the Sixth Cal-ifornia Islands Symposium, edited by D. K. Garcelon and C. A. Schwemm, pp. $'–)". Institute for Wildlife Studies, Arcata, CA.

"##) &''' Years on Ghalas-At, Part One of the San Nicolas Island Index Unit Program. Prepared for Naval Air Weapons Station, China Lake, CA. !rough a Cooperative Agreement with California State University, Los Angeles, N$)*%%-##-LT-##-%.

McCawley, W.%&&$ !e First Angelinos: !e Gabrielino Indians

of(Los Angeles. Malki Museum Press, Ban-ning, CA.

McCoy, M. D., and T. Ladefoged"##& New Developments in the Use of Spatial

Technology in Archaeology. Journal of Ar-chaeological Research %*:"$,–"&'.

Merrill, M. L."##- Intrasite Spatial Analysis and Interpretation

of Mapped Surface Artifacts at CA-SNI-"', San Nicolas Island, California. MA thesis. Department of Anthropology, California State University Los Angeles.

Merrill, M. L., and D. Read"#%# A New Method Using Graph and Lattice

!eory to Discover Spatially Cohesive Sets of Artifacts and Areas of Organized Activity in Archaeological Sites. American Antiquity *'(,):-%&–-'%.

Mitchell, A."##& !e ESRI Guide to GIS Analysis, Vol. ": Spa-

tial Measurements and Statistics. ESRI Press, Redlands, CA.

Munro, P.%&&& Takic Foundations of Nicoleño Vocabulary.

In Proceedings of the Fi)h California Islands Symposium. U.S. Minerals Management Ser-vice, Camarillo, CA. Published version of %&&- manuscript.

Nardin, T. R., R. H. Osborne, D. J. Bottjer, and R. C. Scheidermann%&)% Holocene Sea-Level Curves for Santa Mon-

ica Shelf, California Continental Borderland. Science "%,:,,%–,,,.

Ord, K.%&*' Estimation Methods for Models of Spatial



Interaction. Journal of the American Statisti-cal Association *#(,-&):%"#–%"$.

Perry, J. E."##* Chumash Ritual and Sacred Geography on

Santa Cruz Island, California. Journal of Cal-ifornia and Great Basin Anthropology "*("): %#,–%"-.

Pinder, D., I. Shimada, and D. Gregory%&*& !e Nearest-Neighbor Statistic: Archaeo-

logical Application and New Developments. American Antiquity --(,)-,#–--'.

Pletka, N."##' Late Holocene Gabrielino Settlement Pat-

terns in the Newport Coast Area of Cali-fornia: A Geographic Information Systems Based Approach. MA thesis. Department of Anthropology, California State University, Long Beach.

Pondella, D. J., B. E. Ginter, J. R. Cobb, and L. G. Allen"##' Biogeography of the Nearshore Rocky-Reef

Fishes at the Southern and Baja California Islands. Journal of Biogeography ,":%)*–"#%.

Raab, L. M., J. Cassidy, A. Yatsko, and W. J. Howard"##& California Maritime Archaeology: A San

Clemente Island Perspective. Altamira, Lan-ham, MD.

Rick, T. C., C. E. Skinner, J. M. Erlandson, and R. L. Vellanoweth"##% Obsidian Source Characterization and

Human Exchange Systems on California’s Channel Islands. Paci"c Coast Archaeological Society Quarterly ,*(,):"*–--.

Rigaud, J., and J. F. Simek%&&% Interpreting Spatial Patterns at the Grotte

XV: A Multiple Method Approach. In !e Interpretation of Archaeological Spatial Pat-terning, edited by E. M. Kroll and T. D. Price, pp.(%&&–""#. Plenum, New York.

Rogers, M. J.%&,# Field Notes: %&,# Expedition to San Nicolas

Island. Transcribed and edited by S. J. Schwartz, %&&-. On .le, Naval Air Weapons Station, Point Mugu, CA.

Schoenherr, A. A., C. R. Feldmeth, and M. J. Emerson%&&& Natural History of the Islands of California.

University of California Press, Los Angeles.

Sharon, I., Y. Dagan, and G. Tzionit"##- !e [Awful?] Truth about GIS and Archae-

ology. In British School at Athens Studies, Vol.(%%: Archaeological Field Survey in Cyprus: Past History, Future Potentials, pp. %'%–%$". British School at Athens, Greece.

Strudwick, I. H.%&)$ Temporal and Areal Considerations Re-

garding the Prehistoric Circular Fishhook of Coastal California. MA thesis. Department of Anthropology, California State University, Long Beach.

!omas-Barnett, L."##- A Sandstone Bowl Quarry Site on San Nico-

las Island, California. MA thesis. Department of Anthropology, California State University, Los Angeles.

Vedder, J. G., and R. M. Norris%&$, Geology of San Nicolas Island, California.

Professional Paper ,$&. U.S. Geological Sur-vey, Menlo Park, CA.

Vellanoweth, R. L."##% AMS Radiocarbon Dating and Shell Bead

Chronologies: Middle Holocene Trade and Interaction in Western North America. Jour-nal of Archaeological Science "):&-%–&'#.

Vellanoweth, R. L., B. G. Bartelle, A. F. Ainis, A. C. Cannon, S. J. Schwartz"##) A Double Dog Burial from San Nicolas Is-

land, California, USA: Osteology, Context and Signi.cance. Journal of Archaeological Science ,'(%"):,%%%–,%",.

Whallon, R., Jr.%&*, Spatial Analysis of Occupation Floors, I: Ap-

plication of Dimensional Analysis of Vari-ance. American Antiquity ,)(,):"$$–"*).

%&*- Spatial Analysis of Occupation Floors,(II: !e(Application of Nearest Neighbor Analysis. American Antiquity ,&(%):%$–,-.

Wheatley, D."##- Making Space for an Archaeology of Place.

Internet Archaeology %'. http://intarch. ac. uk / journal/issue%'/wheatley_index. html.

Wheatley, D., and M. Gillings"##" Spatial Technology and Archaeology: !e Ar-

chaeological Applications of GIS. Taylor and Francis, London.


Geographic Information Systems as a Tool for Analyzing Intrasite Spatial Variability on San Nicolas...

Documents

Transcript of Geographic Information Systems as a Tool for Analyzing Intrasite Spatial Variability on San Nicolas...