Sunteți pe pagina 1din 43

Accepted Manuscript

Making the Most of Fragments: A Method for Estimating Shell Length From
Fragmentary Mussels (M. californianus and M. trossulus) on the Pacific Coast of
North America
Gerald G. Singh, Iain McKechnie
PII:

S0305-4403(15)00067-9

DOI:

10.1016/j.jas.2015.02.029

Reference:

YJASC 4359

To appear in:

Journal of Archaeological Science

Received Date: 8 December 2014


Revised Date:

17 February 2015

Accepted Date: 18 February 2015

Please cite this article as: Singh, G.G., McKechnie, I., Making the Most of Fragments: A Method for
Estimating Shell Length From Fragmentary Mussels (M. californianus and M. trossulus) on the Pacific
Coast of North America, Journal of Archaeological Science (2015), doi: 10.1016/j.jas.2015.02.029.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Making the Most of Fragments: A Method for Estimating Shell Length


From Fragmentary Mussels (M. californianus and M. trossulus) on the

M
AN
U

SC

Gerald G. Singh 1 and Iain McKechnie 2,3*

RI
PT

Pacific Coast of North America

TE

Gerald G. Singh
1
Institute for Resources, Environment and Sustainability
University of British Columbia, Vancouver, BC, Canada, V6T 1Z4
Email: geraldsingh@gmail.com

AC
C

EP

Iain McKechnie*
2
Department of Anthropology, 308 Condon Hall, University of Oregon, Eugene, OR,
97403 USA
3
Department of Archaeology, Hakai Institute, Simon Fraser University, Burnaby, BC,
V5A 1S6 Canada
email: iainm@uoregon.edu
* Corresponding author to whom correspondence should be addressed.

ACCEPTED MANUSCRIPT

ABSTRACT

California mussel (Mytilus californianus) are ubiquitous shellfish species in coastal

archaeological sites throughout western North America but are often highly fragmentary when

recovered in small-volume column or bulk samples typically used to quantify shellfish

assemblages. Archaeological research has predominantly focused on evaluating the dietary

contribution of Mytilus but most studies assume an average meat weight or use categorical size

classifications to determine subsistence strategies and harvest profiles. In this paper, we develop

and evaluate a regression-based method for estimating shell length and meat weight for

fragmentary Mytilus remains. Our regressions are based on live-collected M. californianus

SC

RI
PT

specimens from multiple locations in California and British Columbia and provide considerable

11

statistical confidence for predicting length and meat weight. We also apply the same regressions

12

to a collection of M. trossulus and show similar predictive equations, indicating this method can

13

be used in cases where it is not possible to distinguish morphologically between M. californianus

14

and M. trossulus. We demonstrate how these results improve upon previous size-classification

15

methods and discuss the potential for applying these measurements to enhance the relevance of

16

these zooarchaeological data for modern marine conservation and management efforts.

M
AN
U

10

17

18
KEYWORDS

20

Archaeomalacology; Marine Ecology; Dietary Reconstruction; Morphometric Regression;

21

Historical Ecology; Zooarchaeology; Northwest Coast

22

HIGHLIGHTS

23

We develop regressions for estimating length of fragmentary Mytilus californianus.

24

Regressions are based on live-collected shells from California and British Columbia.

25

Regressions improve on existing size classification methods and include M. Trossulus.

26

Strong predictions for length and meat weight from intact umbos can be generated.

27

These improved length estimates extend the interpretive utility of Mytilus assemblages.

28

AC
C

EP

TE

19

ACCEPTED MANUSCRIPT

1. Introduction

30

The California mussel (Mytilus californianus) is a ubiquitous and protein rich shellfish species

31

found in archaeological sites throughout the coast of western North America (e.g., Braje et al.

32

2007; Erlandson 1988; Losey and Power 2005; Porcasi 2011; Sumpter 2005). Mussels are

33

present from the mid-to lower intertidal on wave-exposed rocky coasts from Alaska to Baja

34

California (Schmidt 1999; Suchanek 1979). Judging from the ubiquity and abundance of M.

35

californianus in coastal archaeological assemblages as well as numerous ethnographic accounts

36

of it use, this species was a widely valued marine food regularly harvested and consumed by

37

coastal indigenous people throughout the Holocene (Braje et al. 2012; Ellis and Swan 1981;

38

Erlandson et al. 2008; Jones and Richman 1995; Whitaker 2008). Mussels are considered

39

ecosystem engineers in intertidal ecosystems, providing habitat for many invertebrates and

40

outcompeting intertidal seaweeds (Menge et al. 1994; Paine 1974). M. californianus can be long-

41

lived (>20-50 years) and grow to large sizes (over 20 cm in length) and thus can be

42

overharvested by humans (Seed and Suchanek 1992). Ecologically, M. californianus populations

43

respond to a variety of factors such as climate and oceanography (Menge et al. 2008; Smith et al.

44

2006), temperature (Ford et al. 2010; Harley 2011), and intertidal predation by other marine

45

predators such as sea stars (Pisaster sp.) and sea otters (Enhydra lutris) (Menge et al. 1994;

46

Singh et al. 2013). Given the ecological and economic importance of mussels, further

47

investigation into the characteristics of archaeological assemblages has the potential to enrich

48

understanding of coastal adaptations, human-environment interactions, and paleoecology.

TE

M
AN
U

SC

RI
PT

29

EP

49

Coastal archaeologists have a long history of investigating archaeological shellfish assemblages

51

to detect nutritional, cultural, and ecological patterns and changes over time (Cook 1946; Moss

52

1993; Nelson 1909; Shawcross 1967; Swadling 1976; Waselkov 1987). On the Pacific Coast of

53

North America, the bulk of archaeological research on M. californianus has predominantly

54

occurred in California where analysis has focused on the assessment of dietary importance

55

(Erlandson 1988; Glassow and Wilcoxon 1988; Greengo 1951; Treganza and Cook 1948),

56

harvesting strategies (Jones and Richman 1995; Whitaker 2008), and seasonality (Eerkens et al.

57

2013; Jew et al. 2014; Jones et al. 2008). Archaeologists have additionally utilized

58

archaeological Mytilus sp. for informing environmental and sea level changes (Graham et al.

59

2003) and the impact of long-term sustained human harvests (Botkin 1980; Braje et al. 2012).

AC
C

50

ACCEPTED MANUSCRIPT

60
In contrast, shellfish research on the Northwest Coast has rarely focused on M. californianus

62

despite its ubiquity in archaeological assemblages (Cannon et al. 2008; Clarke and Clarke 1980;

63

Croes 2005; Orchard 2009; Sumpter 2005; Wessen 1988). Rather, the bulk of shellfish research

64

has focused on clams, specifically the seasonality of clam harvesting (e.g., Burchell et al. 2013;

65

Croes 2005:111; Ham and Irvine 1975; Keen 1979; Maxwell 2003; Wessen 1982:144).

RI
PT

61

66

Mussels can exhibit wide size variation that is nutritionally and ecologically meaningful.

68

However, the occurrence of whole shellfish is rare in most depositional settings (Faulkner 2011;

69

Muckle 1985; Wolverton et al. 2010) and this is particularly the case for Mytilus (Ford 1992;

70

Glassow 2000). Mytilus survivorship may be further influenced by excavation and post-

71

excavation sample processing (e.g., wet/dry screening, sorting, and curation). Measurements of

72

complete whole shells is possible given a targeted recovery strategy undertaken during

73

excavation and suitable depositional circumstances such as rockshelter excavations (Croes

74

2005:110). An additional potential taphonomic factor in relying on measurements from whole

75

shells is that larger shells may be more robust and thus preserve more readily. Conversely, it is

76

also possible that smaller shells, with less surface area and or differing shapes, may be less likely

77

to fracture from compaction and trampling (Muckle 1985; Wolverton et al. 2010).

M
AN
U

SC

67

TE

78

While a number of studies have developed regression-based methods for evaluating the size

80

distributions of marine and freshwater mussels (Buchanan 1985; Campbell 2014; Hall 1980;

81

Parmalee and Klippel 1974; Randklev et al. 2009), to our knowledge, the Pacific coast of North

82

America currently lacks a regression-based method for determining M. californianus shell length

83

and no method has estimated meat-weight from fragmentary remains (Bell 2009:25; but see Ford

84

1989:161 for an example specific to M. trossulus). As a result, the majority of research

85

concerning archaeological Mytilus californianus has relied on average meat weight

86

contributions from average sized mussels (Cook 1946; Croes and Hackenberger 1988:36;

87

Kennedy 2004; Mitchell 1988; Moss 1989). In a few cases, length measurements obtained from

88

whole shells have been reported (Braje et al. 2012; Croes 2005:110; Erlandson et al. 2008) but

89

these remain rare relative the more commonly quantified assemblages focused on bulk or column

90

samples (e.g., Glassow 2000; Moss 1993; Sumpter 2005).

AC
C

EP

79

91

ACCEPTED MANUSCRIPT

92

1.1. Categorical Size Classification Methods


The lack of a regression based measurement is significant as the majority of shellfish examined

94

from intensively quantified small-volume bulk samples are fragmentary which constrains the

95

ability to generate robust harvest profiles from these painstakingly calculated proportional data.

96

This circumstance also limits the accuracy of derived secondary information, such as estimates

97

of meat weight calculated from non-repetitive element counts (NRE) (Giovas 2009; Mason et

98

al. 1998) or the affect that shell size may have on the proportional frequency of weighed shell

99

fragments (Glassow 2000).

RI
PT

93

100

Some researchers have provided estimates of Mytilus shell length using size categories to

102

construct cumulative harvest profiles to identify size selectivity (Wessen 1982:151). For

103

instance, research on M. californianus by Whitaker (2008) has built on a size classification

104

method developed by White (1989) and Jones and Rickman (1995) to document the harvest

105

profile for fragmentary mussels from a rockshelter assemblage in Northern California. Whites

106

(1989) method involves comparing archaeological mussels to traced outlines of 2-cm length

107

categories in which researchers place fragmented mussel umbos onto a two-dimensional

108

template with traced outlines of a particular size. Researchers visually match the valve shape and

109

size that best fits the template. The resulting size assessment is used to categorize archaeological

110

valves and then analyzed in a cumulative harvest profile where it can distinguish differences in

111

prey selectivity (e.g., size selective plucking or indiscriminant stripping and variations in

112

between).

EP

113

TE

M
AN
U

SC

101

Bell (2009) subsequently evaluated this classification method and demonstrated considerable

115

statistical uncertainty in this approach, particularly when applied to fragmentary valves, which

116

weakened predictive accuracy. A key recommendation was to develop a more accurate

117

estimation of total shell length based on morphometric relationships from fragmentary remains

118

as has been demonstrated for mussels elsewhere (Campbell 2014; Hall 1980; Randklev et al.

119

2009).

AC
C

114

120
121

In this study, we develop and evaluate a morphometric regression method for deriving total

122

length estimates from incomplete M. californianus shells. Our method offers a variety of

123

measurements that predict the length of fragmentary shells with considerable statistical

ACCEPTED MANUSCRIPT

confidence. Regressions are based on live mussels collected from multiple areas of California

125

and British Columbia and thus relevant for archaeological application across a broad coastal

126

region. Given these regions are subject to varying climatic and oceanographic conditions,

127

incorporating this variability helps the resulting regressions better account for differential

128

growing conditions and climatic changes. We conclude this improves the basis for investigating

129

harvest profiles, measures of Mytilus biomass represented archaeologically, and the historical

130

ecology of this widespread and important shellfish genus.

131

2. Materials and Methods

SC

132

RI
PT

124

2.1. Sample Selection

To derive and evaluate biometric relationships, 132 live M. californianus were collected from

134

numerous intertidal plots within three exposed coastal embayments on western Vancouver

135

Island, British Columbia, Canada as well as ecological collections from Northern and Southern

136

California, USA (Figure 1, supplemental materials). Mussels were obtained from the mid-to-low

137

tide on wave-exposed rocky shores as part of a marine ecology research project seeking to

138

control for growth variation along this latitudinal gradient (Singh 2010; Singh et al. 2013). This

139

broad spatial scale is relevant as it accounts for multiple environmental and growing conditions

140

across space and therefore serves as a proxy for variability in growing conditions across broad

141

temporal scales (e.g., Pickett 1989). Climatic variables, especially temperature and tidal

142

elevation, has a strong effect on mussel growth rate (Harley 2011; Menge et al. 2008; Seed and

143

Suchanek 1992; Suchanek 1981) and different growth conditions can lead to differences in size

144

as well as variation in shell morphology. Thus, our regressions aim to incorporate uncertainty

145

across this large region as such uncertainty is likely to occur over archaeological time scales.

TE

EP

2.2. Measurement Criteria

AC
C

146

M
AN
U

133

147

For each individual mussel used to derive the regression equations, we measured a series of

148

points from the umbo to morphologically distinctive markers on the shell (e.g., the hinge teeth on

149

the inside of the shell) (Figure 2). We measured these attributes because they form the most

150

robust portion of the shell and are more readily recovered in small volume archaeological

151

samples that are typically the focus of detailed quantification efforts.

152

ACCEPTED MANUSCRIPT

The umbo region of mussels is also the oldest site of growth (Ford et al. 2010; Seed 1968). From

154

this point, individuals add layers via secretions from the mantle to the inner side of the shell

155

(from the umbo to the distal end of the shell) (Checa 2000). The robusticity and thickness of the

156

umbo increases as the mussel ages while the ligament (hinge) connecting the two valves of the

157

shell also increases in length as the mussel increases in size. Despite the potential for allometric

158

variation in growth within a particular habitat, these characteristics make the measurements

159

appropriate for morphometric regression with a reasonable degree of accuracy.

RI
PT

153

160

To maximize measurements obtained from a given archaeological assemblage (e.g., fine and

162

coarse mesh fractions), we measured multiple dimensions on individual mussel shells to account

163

for shells in different states of fragmentation (Figure 2). Although this method yields regressions

164

with differing predictive strengths, these uncertainties can be quantified. A particular benefit is

165

that the umbo is used to derive minimum number of individuals (MNI) or non-repetitive

166

elements (NRE), which are especially prevalent in small mesh sizes (Glassow 2000:410). Thus

167

deriving length estimates from these fragmentary specimens can considerably strengthen the

168

analytical potential of these rigorously quantified assemblages and the ecological relevance of

169

data derived from them.

M
AN
U

SC

161

170

We use simple ordinary least squares regression to build the predictive models (Table 1). For

172

fragment samples that are large enough to take multiple measures, we also provide multiple

173

regression model structures for incorporating multiple measures into making an estimate. These

174

multiple regression models explain variation slightly better than the simple models (as indicated

175

by the R2 values in Table 2) and can offer an extra level of confidence in the estimation for

176

particular shells if warranted.

EP

AC
C

177

TE

171

2.3. Meat Weight Estimates

178

Our linear regression models are related to total shell length and the predictive confidence is a

179

corresponding R2 value. Since our regressions are based on live-collected mussels, we

180

additionally calculated dry meat weight for each specimen (distinguishing gonads and non-

181

reproductive tissue). Thus, we also were able to derive a strong shell length-meat weight

182

relationship, which can be applied to a derive minimum weight estimates from individual

183

archaeological M. californianus specimens.

ACCEPTED MANUSCRIPT

184

2.4. Evaluation of Taphonomic Effects


In addition to the live-collected specimens that form the basis of the regression, we conducted

186

these same measurements on 68 M. californianus valves recovered from the upper intertidal

187

wrack zone in two locations in British Columbia (Barkley Sound and Calvert Island, Figure 1,

188

supplemental materials). These beach-collected shells represent natural mortality events induced

189

by wave action during storms, following years (or at least several months) of weathering and

190

erosion as shells are transported to upper intertidal beach deposits. These weathering and wave

191

induced erosional processes can be considered broadly similar to the taphonomic processes

192

affecting archaeological shells over broader time scales (e.g., trampling, erosion of prominent

193

surfaces etc.) (Muckle 1985). While not directly analogous to other post depositional effects such

194

as leaching (Stein 1996), we used the simple regression models to estimate total size of these

195

beach collected shells. This was done in order to assess model performance on weathered shells

196

as well as to evaluate the accuracy of measurements obtained on archaeological shells that may

197

have been subject to weathering and wave erosion following depositional in nearshore contexts

198

(i.e., deposition or re-deposition from a formerly upper intertidal context).

SC

2.5. Mytilus trossulus measurements

M
AN
U

199

RI
PT

185

Despite a considerable maximum size difference, M. californianus is morphologically similar to

201

the Bay mussel (M. trossulus, formerly referred to as M. edulis) and it is often difficult to

202

distinguish the two taxa based on small fragments. Thus, many archaeological analyses

203

cautiously lump all Mytilus specimens into a single category (even umbo fragments) unless local

204

environmental conditions preclude the presence of either. To assess the potential biometric

205

difference and or complementarity between M. californianus and M. trossulus growth, we

206

conducted measurements on 51 M. trossulus shells collected from protected beaches on Quadra

207

Island in the northern Salish Sea (Figure 1, supplemental materials). This inland sea is situated

208

far from the exposed Pacific Coast in habitat unsuitable for M. californianus. Unlike the wave-

209

transported beach collected shells described above, M. trossulus valves were collected at low-

210

low tide from a very protected beach with a lack of wave exposure (Heriot Bay, BC). These

211

shells exhibited a lack of weathering and erosion and appeared to be recently deceased as they

212

were obtained directly below piers or elevated rocky crevices on which live clusters of M.

213

trossulus were growing. The simple regression models developed for M. californianus were

214

applied to these M. trossulus shells to determine how well they perform on these

215

morphologically similar species.

AC
C

EP

TE

200

ACCEPTED MANUSCRIPT

216

2.6. Model Validation


To assess the accuracy and potential inter-observer bias for our measurements and the resulting

218

regressions (e.g., Lyman and VanPool 2009), we conducted double-blind trials that compared

219

differences in the estimates of total mussel length between multiple participants (Table 1). Using

220

a separate population of mussel shells from those used to develop our relationships or our beach

221

collected samples (n=54), we recorded total lengths of individual shells identified with numeric

222

codes and subjected them to destructive fragmentation. Individual umbo fragments obtained

223

from this crushed assemblage were then given to multiple participants to evaluate both our

224

regression method as well as the size classification method developed by White (1989).

225

Accuracy of size predictions was evaluated by comparing predicted lengths with known lengths,

226

with average differences between these measures and average inter-observer differences reported

227

and graphed along a 1:1 line (Table 1, Figure 4).

M
AN
U

SC

RI
PT

217

228

3. Results

230

All regressions based in reference to the umbo showed a strong predictive relationship,

231

indicating a very strong fit of the model to the data and thus a strong relationship between the

232

linear measures and the total shell length (Table 1). The R2 values of the regressions range from

233

0.80 to 0.93, meaning that between 80 and 93% of the variability in mussel lengths is accounted

234

for by the linear regression model, indicating that these models are good predictors. These linear

235

relationships are consistently strong despite regional variability in morphology or growing

236

conditions. We note slight differences in rates of growth, with Southern California being the

237

most distinctly offset (i.e., green dots in Figure 3). However, given that we are interested in

238

incorporating variation in growing conditions as well as paleoclimatic variation, we included

239

data from all regions to generate these predictive models.

TE

EP

AC
C

240

229

241

We also noted that the relationship between total shell length and the dry weights of gonads,

242

non-reproductive body, and shell offer strong relationships, albeit nonlinear (Figure 5). Linear

243

models are more likely to systematically underestimate weight for both small and large shells as

244

well as potentially yield negative weight values for the smallest shells (i.e., there is no negative

245

intercept in our non-linear model). We found that our models explained between 78 and 90% of

246

the variation in body mass, depending on tissue type (reproductive and non-reproductive), while

ACCEPTED MANUSCRIPT

247

relating total length to total dry meat, the predictive model for weight accounts for 90% of the

248

variation (Table 3, Figure 5).

249
250

3.1. Testing Categorical Models


Bell (2009) noted that Whites (1989) method of size categorization has limited predictive

252

power, particularly between estimated length and total shell length. Similarly, our comparison of

253

the difference between estimated shell lengths and actual shell lengths produced by Whites

254

(1989) template indicates an average error of 16 3 mm (Figure 4). Whites template was unable

255

to estimate mussel shells above 100 mm because the template categories did not go above this

256

range. This is indicated in Figure 4a by the wide distributions of actual shell lengths estimated in

257

the 90-100 mm range, though many are larger than 100 mm. Comparatively, the average error

258

between the predicted and actual shell lengths range from 7.3-12.8 mm across shell sizes in the

259

double blind trials using our regression models (Table 1, Figure 4b-e). The model based on the

260

measure from the umbo to the outer hinge teeth was the most precise (Figure 4b), having an

261

average error of 7.3 mm. Our double-blind results demonstrate that there is considerable inter-

262

observer variability between researchers measuring the same specimens using Whites method,

263

which further confounds its predictive power (Figure 4a). Our results indicate the accuracy of

264

Whites method, on average across all participants, was 9.5 1.5 mm between analysts. By

265

contrast, the precision of our estimates, using the umbo thickness model (a modestly performing

266

regression), averaged between all participants, indicates that measurements were off by 5.1 2.3

267

mm between analysts.

SC

M
AN
U

TE

EP

268

RI
PT

251

3.2. Beach Collected Shells

The simple regression models predicted total length of weathered, beach collected shells with

270

similar accuracy compared to the fresh shells, with mean error between 5.4 and 15.4 mm from

271

the actual lengths of shells (Table 4). Models based on the length from the umbo to outer hinge

272

teeth, length of hinge teeth, and umbo thickness were slightly more accurate for beach collected

273

shells, and the model for umbo to inner hinge teeth was slightly more accurate for fresh shells,

274

but in every case the error estimates had 95% confidence intervals that overlapped, indicating no

275

significant difference in performance between fresh shells and beach collected shells.

AC
C

269

10

ACCEPTED MANUSCRIPT

276

3.3. M. californianus and M. trossulus


Evaluating the performance of the four simple regressions developed for M. californianus as

278

applied to M. trossulus indicates that these models predict total length of shells as accurately as,

279

and in some instances, more accurately for M. trossulus (Table 4). The models based on umbo

280

to outer hinge teeth and hinge teeth length performed better for M. trossulus, and the non-

281

overlapping 95% confidence intervals indicate that they perform significantly better than for live

282

collected M. californianus shells, though the 95% confidence intervals overlap with the beach

283

collected shells.

284

4. Discussion

285

Our morphometric regressions provide a more accurate method for confidently estimating the

286

total shell length from mussel umbo fragments relative to the size-class method utilized by

287

Wessen (1982), White (1989), and applied by Whitaker (2008) and others. This is further

288

apparent in a reduction in the variability in length estimates between analysts and an expansion

289

of length estimates of M. californianus on continuous interval scale across their size spectrum

290

(including mussels >200 mm) in contrast to Whites template which appears limited to Mytilus

291

less than 100 mm (White 1989:133). Additionally, incorporating shells from multiple geographic

292

regions subjected to multiple growing conditions is arguably more relevant for application to

293

archeological assemblages across a spectrum of climatic conditions and ecological settings.

M
AN
U

SC

RI
PT

277

TE

294

These models offer more reliable estimates for meat-weight for M. californianus than previous

296

research utilizing average weights and calorie estimates (Croes and Hackenberger 1988:36;

297

Erlandson 1988:104). Due to the high rate of fragmentation for Mytilus (Glassow 2000; Muckle

298

1994), these methods likely also provide a much more representative range of size distributions

299

than measurements of comparatively rare whole Mytilus valves suitable for measurement

300

(Erlandson et al. 2008). Measured shell umbo fragments can provide estimates of the total

301

length, as well as the dry weight (differentiated between tissue types). Combined with

302

archaeological measures such as minimum number of individuals (MNI) or non-repeatable

303

elements (NRE) (Giovas 2009), these models provide the opportunity to refine estimates of

304

meat-weight and Mytilus biomass from archaeological samples. Given the good performance of

305

the length models on M. trossulus as well, it is possible to derive meat weight estimates from

306

fragmentary remains using available shell length to meat weight regressions for M. trossulus

AC
C

EP

295

11

ACCEPTED MANUSCRIPT

307

(e.g., Penney et al. 2008), and thus generate species-specific bounded estimates of meat weight

308

per sample.

309
Surprisingly, our models are similarly accurate in estimating total length for moderately beach-

311

weathered mussel shells, and so still retain strong predictive capacity. This could reflect the

312

robustness of this portion of the shell, and indicate that the morphometric relationships are less

313

affected by taphonomic weathering (i.e. erosion does not strip away shell thickness on the scale

314

of multiple millimeters). These observations provide a measure of confidence that the

315

measurement of archaeological Mytilus assemblages that may have been subject to modest post-

316

depositional wave erosion retain their predictive accuracy. The models also showed strong

317

potential for estimating total length of M. trossulus shells. In some cases the models performed

318

better for smaller mussels, and this could be due to the fact that M. trossulus has a notably

319

narrower growth range and/or their size distribution falls in a zone where more samples fit the

320

model (and therefore the areas of the model that are more certain). This result highlights the fact

321

that both Mytilus species have very similar morphology and morphometric growth relations, and

322

therefore estimates can be combined in cases of taxonomic uncertainty. Overall, the favorable

323

performance of the models across a wide range of specimens (multiple species, various stages of

324

erosion, and locations of collection) highlight the broad applicability of the models and the

325

general approach we advocate.

TE

M
AN
U

SC

RI
PT

310

326

This method is not devoid of uncertainty but represents a significant improvement from size-

328

class data. Based on our trials, our models estimated the maximum error in total lengths as high

329

as 35 mm different from the actual length in individual cases but this is compared to an error of

330

67 mm using Whites (1989) template. Even so, on average, our results indicate that our models

331

produce more certain results based on average error (7.3-12.8 mm with our models versus 15.8

332

mm with the template). More importantly, a regression based approach provides the ability to

333

transform shell length estimates from categorical data (binned size classes) into continuous data

334

therefore increasing the analytical utility through improved quantification of uncertainty and

335

comparison across multiple datasets (e.g., layers within sites, between individual sites, regional

336

trends).

AC
C

EP

327

337

12

ACCEPTED MANUSCRIPT

By sampling over a broad geographic area of the eastern Pacific coast, we have sought to

339

incorporate climatic variation in our models. However, mussels from archeological samples

340

remain likely to have grown in conditions outside what we have sampled for this study (Harley

341

2011). Future research may reduce the uncertainty in our models and/or create regionally specific

342

models should research questions require it. We encourage researchers to contribute to

343

improving this morphometric approach by making their primary measurement data available as

344

weve done here (see supplementary information).

RI
PT

338

345

Which relationship to use will largely depend on how much of the shell remains intact. The

347

model using the measurement from the umbo to the outer hinge teeth had the strongest fit (R2 =

348

0.93) but requires most of the shell to be intact. The model that accommodates the most

349

fragmentary shells (umbo thickness), still yields an estimate of shell length with moderately

350

strong predictive power (R2 = 0.80, Table 1). This is the measure that will likely be most useful

351

for archaeological applications given this is the thickest part of the shell and mostly likely to

352

preserve. While this may only represent a modest improvement in accuracy from existing size

353

classification methods, it is capable of generating length estimates for a larger range of shell

354

sizes and is applicable to a greater range of fragmentary umbos. This method therefore enables a

355

larger number of measurements to be obtained for a given deposit with an greater predictive

356

accuracy. Moreover, if it is possible to obtain more than one measurement on a given shell,

357

multiple regression models can be used to further strengthen length and or meat-weight

358

predictions (Table 2). This latter approach might be of particular interpretive value for an

359

individual shell (e.g., an artifact) or for more confidently establishing an upper size threshold.

M
AN
U

TE

EP

360

SC

346

More important than the statistical power of individual measurements is the interpretive

362

confidence needed for a given research question. In this respect, existing size classification

363

methods may perform adequately for interpretations such as Whitakers (2008) comparison of

364

Mytilus harvest profiles in relation to idealized foraging theory models. However, such

365

measurement data may not be appropriate for statistical analysis that surpasses an ordinal scale

366

(Wolverton et al. 2014). The broader utility of our morphometric approach is an increase in

367

predictive accuracy, the reduction in inter-observer error, and a comparatively more

368

straightforward data output (individual size estimates and an associated uncertainty).

AC
C

361

369

13

ACCEPTED MANUSCRIPT

Collectively, the method can be used to more precisely measure Mytilus size distributions and

371

test for statistically significant differences at different points in time or between sites. Of

372

particular ecological relevance is evaluating the size distribution in relation to the long-term

373

presence of sea otters which have been shown to homogenize mussel sizes through regular

374

harvesting and elimination of larger and older M. californianus (Singh et al. 2013). Sea otters

375

(Enhydra lutris) are voracious consumers of mussels (VanBlaricom 1988), and the methods we

376

provide can potentially be used to evaluate if mussel size distributions were altered when sea

377

otter remains are found in temporally associated midden strata (e.g., Simenstad et al. 1978;

378

Szpak et al. 2012).

RI
PT

370

SC

379

Analysis of shell lengths can also be used to revisit the question of the Mytilus harvesting and

381

cultivation practices of Indigenous peoples as investigated by Whitaker (2008). This method

382

provides a means of refining the statistical characterization of a plucking strategy (the

383

preferential size selection of larger mussels) versus a stripping strategy (indiscriminant removal

384

of a variety of size classes) (Jones and Richman 1995; Whitaker 2008). Alternatively, measuring

385

size distributions may also help characterize periods of change in ocean upwelling as reflected in

386

size distributions before, during, and after known climate changes (e.g., Kennett and Kennett

387

2000), the short term impacts of intensive human harvesting (Croes 1992; Thakar 2011), or the

388

potential restraint exercised in traditional clam management practices (Cannon and Burchell

389

2009; Daniels 2014; Lepofsky et al. in press). While our method provides improved

390

measurement capacity, disentangling such factors will require multiple lines of evidence and a

391

consideration of the causal agents that can produce similar size distributions.

TE

EP

392

M
AN
U

380

While a variety of size-based regressions have been applied to a variety of archaeological

394

shellfish taxa elsewhere (e.g., Jerardino and Navarro 2008; Parmalee and Klippel 1974),

395

additional models for other species are needed on the Pacific Coast in order to refine estimates of

396

dietary content and biomass represented in shellfish assemblages. For instance, clams are another

397

pervasively utilized shellfish that are commonly abundant in shell midden settlements (Cannon et

398

al. 2008) and similar methods have begun to be variously developed for clams at specific sites

399

(Daniels 2014; Ford 1989:161), but this can usefully be expanded to include more species and

400

more geographically representative locations.

AC
C

393

14

ACCEPTED MANUSCRIPT

5. Conclusion

402

This paper has developed and evaluated a method for Mytilus shell length estimation for

403

application to archaeological and ecological research in western North America. This method

404

seeks to improve the precision of zooarchaeological data in order to enhance the interpretive

405

potential and historical ecological significance of archaeological shellfish assemblages (Rick and

406

Lockwood 2013). Future application of this low cost and relatively straightforward method has

407

the potential to increase the number of archaeological assemblages available for metrical

408

comparison and assessment, well beyond the small number of sites where Mytilus length

409

distributions have currently been obtained. Given the importance of shellfish in the

410

archaeological record along the eastern Pacific Coast and the long history of archaeological

411

analysis, improving the capacity to obtain metrical data from existing quantified assemblages

412

will provide for a more thorough integration of ecological and archaeological data. While the

413

method improves coastal archaeologists ability to address existing archaeological questions such

414

as dietary contribution and harvest and management strategies, it also brings zooarchaeological

415

information more directly into conversation with a variety of research in ecology, conservation

416

biology, and resource management (e.g., Erlandson and Rick 2010; Wolverton and Lyman

417

2012). Thus, rather than just another technically challenging and expensive analytical method

418

requiring reallocation of scarce funding resources, this study aims to increase the utility of data

419

derived from conventional shellfish quantification methods and strengthen the analytical effort of

420

thousands of person hours spent counting and/or weighing shell fragments.

TE

M
AN
U

SC

RI
PT

401

AC
C

EP

421

15

ACCEPTED MANUSCRIPT

422
423

Acknowledgements

424

We thank Russ Markel, Chris Harley, Anne Salomon, Dana Lepofsky, RG Matson, Nicole

426

Smith, Margot Hessing-Lewis, and Madonna Moss for their support and encouragement for this

427

project. We thank UC Santa Barbara, Bodega Bay Marine Labs, the Bamfield Marine Sciences

428

Centre, and the Hakai Institute for providing the opportunity to collect and measure Mytilus.

429

Particular thanks to Erin Rechsteiner for collecting and arranging delivery of surf washed shells

430

from Calvert Island. We thank Stefan Dick, Theraesa Coyle, and Gennifer Meldrum for helping

431

with the double blind trials. Thanks to Adrian Whitaker, Dale Croes, and Gary Wessen for

432

providing information during the preparation of this paper and to Steve Wolverton, and the

433

anonymous reviewers for improving this manuscript. Funding for portions of this research has

434

been enabled by the National Engineering and Science council of Canada (GGS), Social

435

Sciences and Humanities Research Council of Canada (IM), and the Hakai Institute for Coastal

436

People, Ecosystems, and Management at Simon Fraser University (IM).

M
AN
U

SC

RI
PT

425

437

AC
C

EP

TE

438

16

ACCEPTED MANUSCRIPT

References Cited

439
440

Bell, A.M., 2009. On The Validity of Archaeological Shellfish Metrics in Coastal California.

442

MA Thesis. Department of Anthropology, California State University, Chico.

443

Botkin, S., 1980. Effects of Human Exploitation on Shellfish Populations at Malibu Creek,

444

California. in: Earle, T.K., Christenson, A.L. (Eds.), Modeling Change in Prehistoric Subsistence

445

Economies, Academic Press, New York, pp. 121139.

446

Braje, T.J., Kennett, D.J., Erlandson, J.M., Culleton, B.J., 2007. Human Impacts on Nearshore

447

Shellfish Taxa: a 7000 Year Record from Santa Rosa Island, California. American Antiquity

448

72(4):735756.

449

Braje, T.J., Rick, T.C., Erlandson, J.M., 2012. A Trans-Holocene Historical Ecological Record

450

of Shellfish Harvesting on Californias Northern Channel Islands. Quaternary International

451

264(20):109120.

452

Buchanan, W.F., 1985. Middens and Mussels: an Archaeological Enquiry. South African Journal

453

of Science 81:1516.

454

Burchell, M., Hallmann, N., Martindale, A., Cannon, A., Schne, B.R., 2013. Seasonality and

455

Intensity of Shellfish Harvesting in the North Coast of British Columbia. Journal of Island and

456

Coastal Archaeology 8(2):152169.

457

Campbell, G.E., 2014. Size prediction in archaeomalacology: the Common Mussel, Mytilus

458

edulis L., as an example. Archaeological and Anthropological Sciences 6(3):305318.

459

Cannon, A., Burchell, M., 2009. Clam Growth-stage Profiles as a Measure of Harvest Intensity

460

and Resource Management on the Central Coast of British Columbia. Journal of Archaeological

461

Science 36(4):10501060.

462

Cannon, A., Burchell, M., Bathurst, R., 2008. Trends and Strategies in Shellfish Gathering on the

463

Pacific Northwest Coast of North America. in: Antczak, A., Cipriani, R. (Eds.), Early Human

464

Impact on Megamolluscs, British Archaeological Reports, Archaeopress, Oxford, pp. 722.

AC
C

EP

TE

M
AN
U

SC

RI
PT

441

17

ACCEPTED MANUSCRIPT

Checa, A., 2000. A new model for perisotracum and shell formation in Unionidae (Bivalvia,

466

Mollusca). Tissue and Cell 32(5):405416.

467

Clarke, L.R., Clarke, A.H., 1980. Zooarchaeological Analysis of Mollusc Remains from Yuquot,

468

British Columbia. in: Folan, W., Dewhirst, J. (Eds.), The Yuquot Project, Volume 2, Parks

469

Canada, Ottawa, pp. 3757.

470

Cook, S.F., 1946. A Reconsideration of Shell Mounds with Respect to Population and Nutrition.

471

American Antiquity 12(1):5153.

472

Croes, D.R., 1992. Exploring Prehistoric Subsistence Change on the Northwest Coast. in: Croes,

473

D., Rebecca Hawkins, and Barry L. Issac (Ed.), Research In Economic Anthropology, JAI Press

474

Inc, Greenwich CT, pp. 337-366.

475

Croes, D.R., 2005. The Hoko River Archaeological Site Complex: The Rockshelter (45CA21),

476

1000--100 B.P., Olympic Peninsula, Washington, Washington State University Press, Pullman,

477

WA.

478

Croes, D.R., Hackenberger, S., 1988. Hoko River Archaeological Complex: Modeling

479

Prehistoric Northwest Coast Economic Evolution. in: Issac, B.L. (Ed.), Prehistoric Economies of

480

the Pacific Northwest Coast, JAI Press, Greenwich, pp. 1985.

481

Daniels, P.S., 2014. Shellfish and Resource Sustainability on the Central Northwest Coast of

482

North America. in: Roksandic, M., de Souza, S.M., Eggers, S., Burchell, M., Klokler, D.M.

483

(Eds.), The Cultural Dynamics of Shell-Matrix Sites, University of New Mexico Press,

484

Albuquerque, pp. 223240.

485

Eerkens, J.W., Byrd, B.F., Spero, H.J., Fritschi, A.M.K., 2013. Stable Isotope Reconstructions of

486

Shellfish Harvesting Seasonality in an Estuarine Environment: Implications for Late Holocene

487

San Francisco Bay Settlement Patterns. Journal of Archaeological Science 40(4):20142024.

488

Ellis, D.W., Swan, L., 1981. Teachings of the Tides: Uses of Marine Invertebrates by the

489

Manhousaht People, Theytus Books, Nanaimo, BC.

490

Erlandson, J.M., 1988. The Role of Shellfish in Prehistoric Economies: A Protein Perspective.

491

American Antiquity 53(1):102109.

AC
C

EP

TE

M
AN
U

SC

RI
PT

465

18

ACCEPTED MANUSCRIPT

Erlandson, J.M., Rick, T.C., 2010. Archaeology Meets Marine Ecology: The Antiquity of

493

Maritime Cultures and Human Impacts on Marine Fisheries and Ecosystems. Annual Review of

494

Marine Science 2:231251.

495

Erlandson, J.M., Rick, T.C., Braje, T.J., Steinberg, A., Vellanoweth, R.L., 2008. Human Impacts

496

on Ancient Shellfish: a 10,000 Year Record from San Miguel Island, California. Journal of

497

Archaeological Science 35(8):21442152.

498

Faulkner, P., 2011. Quantifying shell weight loss in archaeological deposits. Archaeology In

499

Oceania 46:118129.

500

Ford, H.L., Schellenberg, S.A., Becker, B.J., Deutschman, D.L., Dyck, K.A., Koch, P.L., 2010.

501

Evaluating the Skeletal Chemistry of Mytilus californianus as a Temperature Proxy: Effects of

502

Microenvironment and Ontogeny. Paleoceanography 25(PA1203).

503

Ford, P.J., 1989. Molluscan assemblages from archaeological deposits. Geoarchaeology

504

4(2):157173.

505

Ford, P.J., 1992. Interpreting the Grain Size Distributions of Archaeological Shell. in: Stein, J.K.

506

(Ed.), Deciphering a Shell Midden, Academic Press, San Diego, pp. 283325.

507

Giovas, C.M., 2009. The Shell Game: Analytic Problems in Archaeological Mollusc

508

Quantification. Journal of Archaeological Science 36(7):15571564.

509

Glassow, M.A., 2000. Weighing vs. Counting Shellfish Remains: A Comment on Mason,

510

Peterson, and Tiffany. American Antiquity 65(2):407414.

511

Glassow, M.A., Wilcoxon, L.R., 1988. Coastal Adaptations near Point Conception, California,

512

with Particular Regard to Shellfish Exploitation. American Antiquity 53(1):3651.

513

Graham, M.H., Dayton, P.K., Erlandson, J.M., 2003. Ice Ages and Ecological Transitions on

514

Temperate Coasts. Trends in Ecology and Evolution 18(1):3340.

515

Greengo, R.E., 1951. Molluscan Species in California Shell Middens, Department of

516

Anthropology, University of California, Berkeley.

AC
C

EP

TE

M
AN
U

SC

RI
PT

492

19

ACCEPTED MANUSCRIPT

Hall, M., 1980. A method for Obtaining Metrical Data from Fragmentary Molluscan Material

518

Found at Archaeological Sites. South African Journal of Science 76(5):280281.

519

Ham, L.C., Irvine, M., 1975. Techniques for Determining Seasonality of Shell Middens from

520

Marine Mollusc Remains. Syesis 8:363-373.

521

Harley, C.D.G., 2011. Climate Change, Keystone Predation, and Biodiversity Loss. Science

522

334(1124):11241127.

523

Jerardino, A., Navarro, R., 2008. Shell morphometry of seven limpet species from coastal shell

524

middens in southern Africa. Journal of Archaeological Science 35(4):10231029.

525

Jew, N.P., Erlandson, J.M., Rick, T.C., Reeder-Myers, L., 2014. Oxygen isotope analysis of

526

California mussel shells: seasonality and human sedentism at an 8,200-year-old shell midden on

527

Santa Rosa Island, California. Archaeological and Anthropological Sciences 6(3):293303.

528

Jones, T.L., Kennett, D.J., Kennett, J.P., Codding, B.F., 2008. Seasonal stability in Late

529

Holocene shellfish harvesting on the central California coast. Journal of Archaeological Science

530

35(8):22862294.

531

Jones, T.L., Richman, J.R., 1995. On Mussels: Mytilus californianus as a Prehistoric Resource.

532

North American Archaeologist 16(1):3358.

533

Keen, S.D., 1979. The Growth Rings of Clam Shells From Two Pentlatch Middens, Heritage

534

Conservation Branch, Victoria.

535

Kennedy, M.A., 2004. An Investigation of Hunter-gatherer Shellfish Foraging Practices:

536

Archaeological and Geochemical Evidence from Bodega Bay, California. PhD dissertation.

537

Department of Anthropology, University of California, Davis.

538

Kennett, D.J., Kennett, J.P., 2000. Competitive and Cooperative Responses to Climatic

539

Instability in Coastal Southern California. American Antiquity 65(2):379395.

540

Lepofsky, D., Smith, N.F., Cardinal, N., Harper, J., White, E.G., Salomon, A.K., Morris, M.,

541

Puckett, M., Rowell, K., in press. Ancient Shellfish Mariculture on the Northwest Coast of North

542

America. American Antiquity.

AC
C

EP

TE

M
AN
U

SC

RI
PT

517

20

ACCEPTED MANUSCRIPT

Losey, R.J., Power, E.A., 2005. Shellfish Remains from the Par-Tee Site (35-CLT-20), Seaside,

544

Oregon: Making Sense of a Biased Sample. Journal of Northwest Anthropology 39(1):120.

545

Lyman, R.L., VanPool, T.L., 2009. Metric Data in Archaeology: A Study of Intra-Analyst and

546

Inter-Analyst Variation. American Antiquity 74(3):485504.

547

Mason, R.D., Peterson, M.L., Tiffany, J.A., 1998. Weighing vs. Counting: Measurement

548

Reliability and the California School of Midden Analysis. American Antiquity 63(2):303324.

549

Maxwell, D., 2003. Growth Coloration Revisted: Assessing Shellfish Seasonality in Coastal

550

British Columbia. in: Carlson, R.L. (Ed.), Archaeology of Coastal British Columbia: Essays in

551

Honour of Professor Philip M. Hobler, Archaeology Press, Simon Fraser Univeristy, Burnaby,

552

B.C., pp. 175188.

553

Menge, B.A., Berlow, E.L., Blanchette, C.A., Navarrete, S.A., Yamada, S.B., 1994. The

554

keystone species concept: variation in interaction strength in a rocky intertidal habitat.

555

Ecological monographs 64(3):249286.

556

Menge, B.A., Chan, F., Lubchenco, J., 2008. Response of a rocky intertidal ecosystem engineer

557

and community dominant to climate change. Ecology Letters 11(2):151162.

558

Mitchell, D.H., 1988. Changing Patterns of Resource Use in the Prehistory of the Queen

559

Charlotte Strait, British Columbia. in: Issac, B.L. (Ed.), Prehistoric Economies of the Pacific

560

Northwest Coast, JAI Press, Greenwich, pp. 245290.

561

Moss, M.L., 1989. Archaeology and Cultural Ecology of the Prehistoric Angoon Tlingit. PhD

562

Dissertation. Department of Anthropology, University of California, Santa Barbara.

563

Moss, M.L., 1993. Shellfish, Gender and Status on the Northwest Coast: Reconciling

564

Archaeological, Ethnographic, and Ethnohistoric Records of the Tlingit. American

565

Anthropologist 95(3):631652.

566

Muckle, R.J., 1985. Archaeological Considerations of Bivalve Shell Taphonomy. MA Thesis.

567

Department of Archaeology, Simon Fraser University.

AC
C

EP

TE

M
AN
U

SC

RI
PT

543

21

ACCEPTED MANUSCRIPT

Muckle, R.J., 1994. Differential Recovery of Mollusk Shell From Archaeological Sites. Journal

569

of Field Archaeology 21(1):129131.

570

Nelson, N., 1909. Shellmounds of the San Francisco Bay region. University of California

571

Publications in American Archaeology and Ethnology 7(4):310356.

572

Orchard, T.J., 2009. Otters and Urchins: Continuity and Change in Haida Economy during the

573

Late Holocene and Maritime Fur Trade Periods, British Archaeological Reports, Archaeopress,

574

Oxford.

575

Paine, R.T., 1974. Intertidal Community Structure: Experimental Studies on the Relationship

576

between a Dominant Competitor and Its Principal Predator. Oecologia 15(2):93120.

577

Parmalee, P.W., Klippel, W.E., 1974. Freshwater mussels as a prehistoric food resource.

578

American Antiquity 39(3):421434.

579

Penney, R.W., Hart, M.J., Templeman, N.D., 2008. Genotype-dependent Variability in Somatic

580

Tissue and Shell Weights and Its Effect on Meat Yield in Mixed Species [Mytilus edulis L., M.

581

trossulus (Gould), and Their Hybrids] Cultured Mussel Populations. Journal of Shellfish

582

Research 27(4):827834.

583

Pickett, S.T.A., 1989. Space-for-time Substitution as an Alternative to Long-term Studies. in:

584

Likens, G.E. (Ed.), Long-term Studies in Ecology: Approaches and Alternatives, Springer, New

585

York, pp. 110135.

586

Porcasi, J.F., 2011. More on Mollusks: Trans-Holocene Shellfish Exploitation on the California

587

Coast. Journal of Island and Coastal Archaeology 6(3):398420.

588

Randklev, C.R., Wolverton, S.J., Kennedy, J.H., 2009. A biometric technique for assessing

589

prehistoric freshwater mussel population dynamics (family: Unionidae) in north Texas. Journal

590

of Archaeological Science 36(2):205213.

591

Rick, T.C., Lockwood, R., 2013. Integrating Paleobiology, Archeology, and History to Inform

592

Biological Conservation. Conservation Biology 27(1):4554.

AC
C

EP

TE

M
AN
U

SC

RI
PT

568

22

ACCEPTED MANUSCRIPT

Schmidt, D., 1999. A Review of California Mussel Fisheries Biology and Fisheries Programes.

594

Canadian Stock Assesment Secretariat Research Document 99/187, Ottawa.

595

Seed, R., 1968. Factors Influencing Shell Shape in the Mussel Mytilus edulis. Journal of the

596

Marine Biological Association of the United Kingdom 48(3):561584.

597

Seed, R., Suchanek, T.H., 1992. Population and Community Ecology of Mytilus. in: Gosling,

598

E.G. (Ed.), The Mussel Mytilus: Ecology, Physiology, Genetics and Culture, Elsevier,

599

Amsterdam.

600

Shawcross, W., 1967. An Investigation of Prehistoric Diet and Economy on a Coastal Site at

601

Galatea bay, New Zealand. Proceedings of the Prehistoric Society 33(7):107131.

602

Simenstad, C.A., Estes, J.A., Kenyon, K.W., 1978. Aleuts, sea otters, and alternate stable-state

603

communities. Science 200:403-411.

604

Singh, G.G., 2010. Effects of Sea Otters on Nearshore Ecosystem Functions with Implications

605

for Ecosystem Services. MSc Thesis. Resource Management and Environmental Studies,

606

University of British Columbia.

607

Singh, G.G., Markel, R.W., Martone, R.L.G., Salomon, A.K., Harley, C.D.G., Chan, K.M.A.,

608

2013. Sea Otters Homogenize Mussel Beds and Reduce Habitat Provisioning in a Rocky

609

Intertidal Ecosystem. PLoS One 8(5):e65435.

610

Smith, J.R., Fong, P., Ambrose, R.F., 2006. Dramatic Declines in Mussel Bed Community

611

Diversity: Response to Climate Change? Ecology 87(5):11531161.

612

Stein, J.K., 1996. Geoarchaeology and Archaeostratigraphy: view from a Northwest Coast Shell

613

Midden. in: Reitz, E.J., Newsom, L.A., Scudder, S.J. (Eds.), Case Studies in Environmental

614

Archaeology, Plenum Press, New York, pp. 35-54.

615

Suchanek, T.H., 1979. The Mytilus californianus community: Studies on the composition,

616

structure, organization, and dynamics of a mussel bed. PhD Dissertation. Department of

617

Zoology, University of Washington.

AC
C

EP

TE

M
AN
U

SC

RI
PT

593

23

ACCEPTED MANUSCRIPT

Suchanek, T.H., 1981. The Role of Disturbance in the Evolution of Life History Strategies in the

619

Intertidal Mussels Mytilus edulis and Mytilus californianus. Oecologia 50(2):143152.

620

Sumpter, I.D., 2005. An Analysis of Three Shellfish Assemblages from Ts'ishaa, site DfSi-16

621

(204T), Benson Island, Pacific Rim National Park Reserve. in: McMillan, A.D., St. Claire, D.E.

622

(Eds.), Ts'ishaa: Archaeology and Ethnography of a Nuu-chah-nulth Origin Site in Barkley

623

Sound, Archaeology Press, Simon Fraser University, Burnaby, pp. 136172.

624

Swadling, P., 1976. The Implications of Shellfish Exploitation for New Zealand Prehistory.

625

Mankind 11(1):1118.

626

Szpak, P., Orchard, T.J., McKechnie, I., Grcke, D.R., 2012. Historical Ecology of Late

627

Holocene Sea Otters (Enhydra lutris) from Northern British Columbia: Isotopic and

628

Zooarchaeological Perspectives. Journal Of Archaeological Science 39(5):15531571.

629

Thakar, H.B., 2011. Intensification of Shellfish Exploitation: Evidence of Species-specific

630

Deviation from Traditional Expectations. Journal of Archaeological Science 38(10):25962605.

631

Treganza, A.E., Cook, S.F., 1948. The Quantitative Investigation of Aboriginal Sites: Complete

632

Excavation with Physical and Archaeological Analysis of a Single Mound. American Antiquity

633

13(4):287297.

634

VanBlaricom, G.R., 1988. Effects of Foraging by Sea Otters on Mussel-dominated Intertidal

635

Communities. in: VanBlaricom, G.R., Estes, J.A. (Eds.), The Community Ecology of Sea Otters,

636

Springer-Verlag, Berlin, pp. 4891.

637

Waselkov, G.A., 1987. Shellfish gathering and shell midden archaeology. Advances in

638

Archaeological Method and Theory 10:93210.

639

Wessen, G.C., 1982. Shell Middens As Cultural Deposits: A Case Study From Ozette. PhD

640

Dissertation. Department of Anthropology, Washington State University.

641

Wessen, G.C., 1988. The Use of Shellfish Resources on the Northwest Coast: The View from

642

Ozette. in: Issac, B.L. (Ed.), Prehistoric Economies of the Pacific Northwest Coast: Research in

643

Economic Anthropology, Supplement 3., JAI Press, Greenwich, CT, pp. 179-207.

AC
C

EP

TE

M
AN
U

SC

RI
PT

618

24

ACCEPTED MANUSCRIPT

Whitaker, A.R., 2008. Incipient Aquaculture in Prehistoric California?: Long-term Productivity

645

and Sustainability vs. Immediate Returns for the Harvest of Marine Invertebrates. Journal of

646

Archaeological Science 35(4):11141123.

647

White, G.G., 1989. A Report of Archaeological Investigations at Eleven Native American

648

Coastal Sites, MacKerricher State Park, Mendocino County, California. Report on File with

649

California State Parks, Sacramento, CA.

650

Wolverton, S., Lyman, R.L., 2012. Conservation Biology and Applied Zooarchaeology.

651

University of Arizona Press, Tucson.

652

Wolverton, S., Randklev, C.R., Kennedy, J.H., 2010. A Conceptual Model for Freshwater

653

Mussel (family: Unionidae) Remain Preservation in Zooarchaeological Assemblages. Journal of

654

Archaeological Science 37(1):164173.

655

Wolverton, S.J., Dombrosky, J., Lyman, R.L., 2014. Practical Significance: Ordinal Scale Data

656

and Effect Size in Zooarchaeology. International Journal of Osteoarchaeology 10.1002/oa.2416.

657

658

M
AN
U

SC

RI
PT

644

659

TE

660

AC
C

EP

661

25

ACCEPTED MANUSCRIPT

Figure Captions
Figure 1. Map showing locations and sample numbers of M. californianus and M. trossulus
specimens collected for this study. Italicized sample numbers represent beach collected shells
used to evaluate taphonomic effects as discussed in the text.

RI
PT

Figure 2. Measurements made on a shell of M. californianus (main and inset pictures) to create
regressions to predict total shell length. Measures of the lengths of 1-4 were made with the total
length of the shell. The photo shows the inside of the left valve.

SC

Figure 3. Regression model fits to data. Different colored and shaped points correspond to shells
collected in different areas (refer to legend). a) Umbo to far hinge teeth. b) Length of hinge teeth.
c) Umbo width. d) Umbo to near hinge teeth

M
AN
U

Figure 4. Visual comparison of estimated total length to known shell lengths using a) White's
(1989:133) template, b) the regression model based on measuring the length from the umbo to
the outer hinge teeth, c) the regression model based on measuring the length of the hinge teeth,
d) the regression model based on measuring the width of the umbo, and e) the regression model
based on measuring the length from the umbo to the inner hinge teeth. The straight lines are not
regressions in themselves, but instead represent a 1:1 relationship, with the best models falling
along the line. Model predictions that are more accurate fall closer to the straight line, while
those points farther away represent relatively worse predictions.

EP

TE

Figure 5. Relationships between total length and dry weight mass for reproductive, nonreproductive, and total meat weight.

AC
C

662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687

26

ACCEPTED MANUSCRIPT

M
AN
U

SC

RI
PT

Table 1. Regression equations for estimating total length of mussel shells from shell fragments. The accuracy measurements are a
measure of how different, on average across sizes and participants, the estimate is from the actual total shell length, and represents the
mean error as well as the 95% confidence interval.
Measurement Description (x) Regression Equation
R2 Value Average Error ( 95% CI) in double blind trials
umbo to outer hinge teeth
total length = 1.8074 x + 4.947 0.93
7.31.2 mm
length of hinge teeth
total length = 2.3916 x + 4.9349 0.81
11.21.9 mm
umbo thickness
total length = 15.312 x + 7.1701 0.80
10.21.9 mm
umbo to inner hinge teeth
total length = 4.284 x + 28.94
0.80
12.83.0 mm

Table 2. Multiple regression models incorporating multiple measurements. In cases where enough of the shell is intact enough to make
multiple measurements (the two cases are described under the Shell Fragment column), these models can be used to obtain
increased accuracy in length estimates than with simple regressions shown in Table 1.
Shell Fragment
Regression Equation
R2
0.84

total length = -0.9582(umbo) + 2.0493(inner hinge) + 0.8868(hinge length) + 0.6065(far


hinge) + 8.4309

0.94

TE
D

total length = 8.2384(umbo) + 2.2361(inner hinge) + 14.8106

AC
C

EP

Can see hinge teeth closest to


umbo
Can see entire length of hinge
teeth

Table 3. Regression equations to estimate dry body (meat) weight from total shell length for M. californianus.
Estimate
Regression Equation
R2
Dry weight of total body
Dry weight = 0.0006832*total length^1.9678096 0.89
Dry weight of non-reproductive body Dry weight = 0.0006155*total length^1.9249686 0.90
Dry weight of gonad
Dry weight = 0.0001019*total length^2.0887239 0.78

ACCEPTED MANUSCRIPT

AC
C

EP

TE

M
AN
U

SC

RI
PT

Table 4. Mean error estimates of the four different regression models used on beach collected Mytilus californianus shells and shells
of Mytilus trossulus.
Shell Type
Umbo to outer hinge teeth Hinge length Umbo thickness Umbo to inner hinge teeth
Beach collected M. californianus 5.41.4
9.52.3
9.92.3
15.43.4
Mytilus trossulus
4.01.2
5.91.3
9.53.1
15.53.6

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

Quadra Island
n=51 (M. trossulus)

Calvert
Island, n=16

EP

TE

Kyuquot Sound
n=10
Clayoquot Sound
n=20
Barkley Sound
n=38 (52)

AC
C

Pacific
Ocean

Northern California, n=20


Southern California, n=44

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

AC
C

EP

TE

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

AC
C

EP

TE

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

HIGHLIGHTS
We develop regressions for estimating length of fragmentary Mytilus californianus.
Regressions are based on live-collected shells from California and British Columbia.

RI
PT

Regressions improve on existing size classification methods and include M. Trossulus.


Strong predictions for length and meat weight from intact umbos can be generated.

AC
C

EP

TE

M
AN
U

SC

These improved length estimates extend the interpretive utility of Mytilus assemblages.

ACCEPTED MANUSCRIPT

EP

Total
123.37
128.62
118.67
57.03
56.81
121.38
96.04
95.44
50.78
49.38
49.27
61.89
63.93
65.32
128.15
144.82
52.83
126.9
118.21
66.4
55.55
69.59
68.71
117.91
64.74
77.89
88.86
55.44
54.6
52.05
145.64
42.15
38.29
34.45
44.16
37.1
33.27
87.68
69.04
109.29
54.66
124.57
69.52
87.65
67.16
67.62

M
AN
U

SC

RI
PT

Base Hinge Far Hinge Hinge length Umbo


18.26
54.67
38.1 5.35
22.9
59.07
36.87 6.13
27.57
60.81
33.85 9.02
6.37
24.79
18.44 2.75
8.81
21.63
17.94
3.1
36.74
56.53
20.48 8.43
11.23
44.3
35.18 5.91
11.25
45.09
33.69 5.28
8.04
22.84
14.5 3.23
3.56
21.8
18.84 2.21
2.59
23.45
21.84 2.22
9.91
29.24
19.55 2.69
8.36
26.64
19.13 2.19
9.67
27.53
18.17 4.25
15.64
53.61
39.39 5.73
20.48
66.7
47.39 6.44
5.31
22.2
19.11 2.51
23.93
58.22
35.49 9.37
15.88
49.63
34.79 5.57
8.94
26.46
18.76 4.26
5.79
23.28
18.25 2.73
10.54
29.7
21.61 2.64
7.46
27.41
20.25 3.13
17.76
50.55
32.91 5.96
8.37
28.78
21.02 3.85
12.63
35.34
22.95 4.26
10.14
36.03
26.07 4.22
7.51
23.62
18.16 2.67
5.38
23.78
18.53 3.09
4.72
21.68
16.88 2.85
20.65
69.54
51.38
8.3
3.67
20.6
17.04
2.2
3.79
16.98
12.67 1.95
2.7
16.07
13.11 1.82
4.77
19.76
15.29 2.14
2.7
15.84
12.19 1.84
3.04
15.36
12.86 1.58
15.21
38.29
23.23 4.81
7.36
28.89
23.59
2
18.89
46.81
27.78 5.95
5.06
23.78
19.08 2.98
33.29
64.37
33.73 8.59
8.53
29.23
20.95 4.05
14.79
38.63
27.84 4.57
5.95
37.26
30.98
4.3
6.69
38.47
31.17 3.75

TE

Plot
Boathouse 198
Boathouse 202
Boathouse 196
Boathouse 191
Boathouse 189
Boathouse 197
Boathouse 195
Boathouse 195
Boathouse 171
Boathouse 156
Boathouse 159
Boathouse 192
Boathouse 190
Boathouse193
Jalama 49
Jalama 55
Jalama 12
Jalama 54
Jalama 51
Jalama 10
Jalama 8
Jalama 30
Jalama 33
Jalama 51
Jalama 20
Jalama 36
Jalama 43
Jalama 10
Jalama 8
Jalama 7
Jalama J56
Cambria 158
Cambria 181
Cambria 148
Cambria 194
Cambria 191
Cambria 150
Jalama 42
Jalama 32
Jalama 48
Jalama 8
Jalama 53
Jalama 30
Jalama 42
BS 15 2
BS 25 2

AC
C

Region
Site
South California Boathouse
South California Boathouse
South California Boathouse
South California Boathouse
South California Boathouse
South California Boathouse
South California Boathouse
South California Boathouse
South California Boathouse
South California Boathouse
South California Boathouse
South California Boathouse
South California Boathouse
South California Boathouse
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Cambria
South California Cambria
South California Cambria
South California Cambria
South California Cambria
South California Cambria
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
South California Jalama
British Columbia BS
British Columbia BS

side
left
left
right
right
left
right
right
left
left
left
right
left
left
right
right
left
left
right
left
left
left
left
right
right
right
left
left
right
right
right
right
left
right
right
right
right
right
right
right
right
left
right
right
left
left
right

ACCEPTED MANUSCRIPT

EP

19.66 2.39 48.45 left


32.72 3.12 76.63 left
32.39 4.99 74.83 left
25.75 2.41 61.61 right
22.89 3.27 60.51 right
34.89 4.35 77.77 right
29.19 3.32 72.61 right
36.94 5.19 76.81 left
27.89 3.27 55.98 left
18.37 2.66 52.62 right
24.49 3.09 54.08 left
23.42 4.05 56.11 right
22.63 2.77 54.91 right
20.79 4.58 52.54 left
25.2 2.31
55 left
24.68
2.8
54.8 left
40.9 5.04 99.68 left
23.81 3.67 51.96 right
17.57 2.52 47.77 left
30.97 3.08 60.72 right
24.4 3.76 53.87 right
22.14 3.16 54.76 right
25.07 4.22 57.69 right
25.13
3.8 55.29 right
21 3.25 56.95 left
23.24 3.53 55.54 right
26.51
2.8 56.17 left
32.25 5.28 78.68 right
74.77 10.28 191.59 right
56.04 8.46 118.93 left
46.87 6.73 98.28 left
55.73 6.31 126.59 right
39.17 4.25 81.32 left
46.69 5.72 118.93 right
46.95 6.17 95.46 right
33.53
4.6 77.25 left
34.78
4.6
87.9 right
38.06 6.65 102.33 right
24.25 3.66 51.46 left
23.4 3.65 43.65 right
18.32 2.85 39.15 left
20.83 4.35 42.81 right
56.76 6.69 141.83 right
17.34 3.95 40.88 right
18.56 2.35 37.49 left
25.46 3.21
48 left
22.45 3.05 46.49 right

RI
PT

26.87
38.67
41.63
30.76
29.49
42.5
34.95
45.45
32.3
26.02
28.02
29.22
26.46
30.9
28.1
29.03
50.12
29.05
21.58
34.69
30.46
27.2
34.82
30.68
27.45
29.14
31.23
42.91
114.67
70.24
61.39
73.26
45.05
65.44
55.69
46.91
46.14
52.9
30.38
28.58
23.82
23.97
78.72
26.47
21.97
28.88
26.55

SC

7.43
6.82
9.44
4.84
6.31
7.36
8.17
8.61
5.73
9.23
4.21
4.49
4.67
10.4
4.11
5.22
8.43
4.72
4.61
5.46
5.73
6.6
11.02
5.4
6.3
5.49
4.9
10.84
39.95
16.21
15.2
17.58
6.64
21.13
12.38
13.42
11.85
15.36
6.55
4.5
6.13
4.91
22.64
9.98
3.67
4.16
4.53

M
AN
U
D

TE

BS 21 1
BS 16 2
BS 16 1
BS 4 1
BS 21 2
BS 25 1
BS 4 2
BS 15 1
Blun 4 low
Blun 18 low
LAC 0 low
Blun 5 mid
LAC 27 mid
Bart 7 mid
LAC 16 lower
LAC 3 low
Blun 13 low
Bart 5 low
Blun 25
Bart 1 low
LAC 12 mid
Blun 15 mid
Bart 4 low
Bart 11 mid
Blun 4 mid
Blun 11 mid
Bart 22 low
Blun 22 low
Helby Vex 2N 1
Sanf Vex 1N 1
Diana Vex 5N 1
Diana Vex 4N 1
Helby Vex 1N 1
Helby Vex 3N 1
Seppings iB 4 1
KT1 iB 5 1
Helby Vex 4N 1
Seppings iB 6 1
PB Putty 4 2
PB Vex 4N 1
Diana Vexar 3 4
TB Vex 5 3
Helby Vex 5N 1
TB Putty 2 1
TB Vex 2N 3
GDA Vex 1N 1
TB Putty 2 2

AC
C

British Columbia BS
British Columbia BS
British Columbia BS
British Columbia BS
British Columbia BS
British Columbia BS
British Columbia BS
British Columbia BS
British Columbia Blun
British Columbia Blun
British Columbia LAC
British Columbia Blun
British Columbia LAC
British Columbia Bart
British Columbia LAC
British Columbia LAC
British Columbia Blun
British Columbia Bart
British Columbia Blun
British Columbia Bart
British Columbia LAC
British Columbia Blun
British Columbia Bart
British Columbia Bart
British Columbia Blun
British Columbia Blun
British Columbia Bart
British Columbia Blun
British Columbia Helby
British Columbia Sanf
British Columbia Diana
British Columbia Diana
British Columbia Helby
British Columbia Helby
British Columbia Seppings
British Columbia KT1
British Columbia Helby
British Columbia Seppings
British Columbia PB
British Columbia PB
British Columbia Diana
British Columbia TB
British Columbia Helby
British Columbia TB
British Columbia TB
British Columbia GDA
British Columbia TB

ACCEPTED MANUSCRIPT

EP

TE

27.94
41.63
49.9
41.26
50.8
18.4
49.46
10.55
11.88
8
3.82
12.61
7.94
6.36
6.68
6.88
4.36
6.1
3.73
24.31
62.78
47.02
32.45
24.55
26.65
25.55
30.69
26.27
20.65
26.74
23.45
29.35
48.65
27.37
27.4
25.98
23.53
26.52
33.14

4.25
5.29
6.65
5.36
8.25
2.92
5.27
1.59
1.34
1.34
0.98
1.73
0.71
1.22
1.03
1.26
0.74
0.95
0.73
5.3
7.12
5.78
4.71
5.64
3.46
3.93
4.18
4.7
4.75
5.34
2.58
3.31
7.94
3.51
3.52
3.88
3.8
4.25
6.4

49.2
98.3
106.38
102.4
122.47
37.12
110.08
26.99
22.63
18.02
12.64
24.26
15.7
17.12
14.4
18.43
10.27
13.69
9.15
64.55
156.05
103.74
66.42
58.07
63.04
56.14
67.21
57.48
54.21
63.61
51.03
63.77
115.29
56.56
60.11
59.11
61.08
56.78
73.9

RI
PT

32.12
56
60.73
54.84
70.17
22.96
58.46
13.24
13.52
10.46
6.02
14.71
8.74
8.08
7.73
8.83
5.14
7.59
4.56
37.3
84.24
57.68
39.91
33.87
33.76
30.45
36.6
32.8
29.65
35.39
28.12
34.72
62.52
32.64
32.62
33.68
30.13
32.14
44.74

SC

6.03
15.99
13.92
14.9
21.42
4.02
10.08
2.47
1.82
2.39
2.18
1.86
1.05
1.55
1.6
2.11
0.8
1.34
1.22
13.88
23.34
11.89
8.43
11.68
8.53
5.29
7.13
6.76
9.56
8.45
5.76
6.83
15.72
6.21
5.28
7.19
7.11
5.94
12.71

M
AN
U

GDA Vex 3N 1
Bluestone Vex 4N 1 winter
Diana Vex 3N 1 winter
Sanf Vex 5N 2 winter
Sanf Vex 2N 1 winter
KT1 iB1 2 winter
Bluestone Vex 2N 1 winter
Diana s1
Diana s2
Diana s3
Diana s4
Diana s5
Diana s6
Diana s7
Diana s8
Diana s9
Diana s10
Diana s11
Diana s12
Con 5
#1
#11
gray 5
black 8
gray 11
black 7
gray 10
con 6
con 7
black 6
con 8
gray 8
#10
gray 7
black 3
gray 6
con 2
con 11
gray 3

AC
C

British Columbia GDA


British Columbia BS
British Columbia Diana
British Columbia Sanf
British Columbia Sanf
British Columbia KT1
British Columbia BS
British Columbia Diana
British Columbia Diana
British Columbia Diana
British Columbia Diana
British Columbia Diana
British Columbia Diana
British Columbia Diana
British Columbia Diana
British Columbia Diana
British Columbia Diana
British Columbia Diana
British Columbia Diana
North California
North California Mussel pt
North California Mussel pt
North California
North California
North California
North California
North California
North California
North California
North California
North California
North California
North California Mussel pt
North California
North California
North California
North California
North California
North California

left
right
right
left
right
right
right
left
right
left
right
left
left
right
left
left
right
left
right
left
left
right
left
right
right
right
left
right
right
left
right
left
left
left
left
left
right
right
left

ACCEPTED MANUSCRIPT

Whyte est Whyte diff Genn Whyte


Genn diff Theraesa Whyte
Ther diff Stefan Whyte
Stefan diff Stefan umbo
Stefan est
55
-6.79
45
3.21
65
16.79
55
6.79
3.98 69.69981
35
11.52
35
11.52
65
18.48
35
-11.52
3.62 64.04795
55
-3.77
55
-3.77
75
23.77
65
13.77
5.15 88.06834
95
-7.44
75
12.56
95
7.44
95
7.44
3.77 66.40289
95
14.16
95
14.16
95
-14.16
95
-14.16
4.28 74.40969
95
46.86
85
56.86
85
-56.86
75
-66.86
7.31 121.9795
85
-12.42
85
-12.42
85
12.42
85
12.42
4.6 79.43356
85
-12.76
95
-22.76
85
12.76
85
12.76
4.83 83.04447
95
-24.67
45
25.33
65
-5.33
65
-5.33
4.61 79.59056
75
-13.73
55
6.27
75
13.73
75
13.73
3.44 61.22202
95
12.47
85
22.47
95
-12.47
95
-12.47
5.41 92.15024
35
5.87
45
-4.13
65
24.13
55
14.13
3.52 62.47799
95
4.73
95
4.73
95
-4.73
95
-4.73
5.98 101.099
95
31.06
95
31.06
95
-31.06
85
-41.06
6.56 110.2048
75
-22.26
45
7.74
75
22.26
45
-7.74
3.75 66.0889
85
-24.2
25
35.8
35
-25.8
35
-25.8
4.3 74.72368
95
-9.35
55
30.65
65
-20.65
55
-30.65
5.1 87.28336
95
11.33
85
21.33
85
-21.33
95
-11.33
5.47 93.09221
85
4.7
85
4.7
85
-4.7
85
-4.7
5.94 100.471
75
-24.6
65
-14.6
65
14.6
65
14.6
4.26 74.0957
95
-5.49
55
34.51
75
-14.51
85
-4.51
4.41 76.45064
85
-9.57
65
10.43
85
9.57
75
-0.43
4.7 81.00352
85
14.08
85
14.08
95
-4.08
95
-4.08
5.51 93.7202
35
26.09
45
16.09
65
3.91
55
-6.09
3.66 64.67594
35
23.93
45
13.93
45
-13.93
45
-13.93
3.42 60.90803
85
-4.88
75
5.12
85
4.88
85
4.88
4.18 72.83973
65
15.34
55
25.34
75
-5.34
75
-5.34
3.82 67.18787
75
-12.97
75
-12.97
95
32.97
85
22.97
3.67 64.83293
85
-18.54
55
11.46
55
-11.46
55
-11.46
2.45 45.67942
85
-23.63
55
6.37
75
13.63
75
13.63
4.26 74.0957
65
-13.44
45
6.56
75
23.44
55
3.44
3.03 54.78519
95
-19.48
65
10.52
75
-0.52
75
-0.52
3.9 68.44384
55
-1.48
75
-21.48
45
-8.52
55
1.48
2.9 52.74424
85
-10.37
75
-0.37
75
0.37
75
0.37
4.96 85.08542
95
5.26
85
15.26
95
-5.26
95
-5.26
5.7 96.70312
55
8.61
35
28.61
35
-28.61
35
-28.61
4.38 75.97965
75
12.83
55
32.83
75
-12.83
55
-32.83
2.94 53.37222
75
6.17
65
16.17
95
13.83
85
3.83
5.94 100.471
75
17.66
75
17.66
95
2.34
95
2.34
4.1 71.58376
75
-18.9
75
-18.9
75
18.9
55
-1.1
2.84 51.80226
95
-3.69
95
-3.69
95
3.69
95
3.69
6.04 102.041
75
20.37
95
0.37
95
-0.37
95
-0.37
5.31 90.58028
95
60.31
95
60.31
95
-60.31
95
-60.31
6.15 103.7679
95
42.59
85
52.59
95
-42.59
95
-42.59
7.67 127.6313
75
19.35
85
9.35
65
-29.35
95
0.65
6.2 104.5529
55
8.32
25
38.32
45
-18.32
35
-28.32
4.14 72.21174
35
18.4
35
18.4
65
11.6
65
11.6
3.75 66.0889
35
22.87
55
2.87
65
7.13
65
7.13
3.45 61.37902
85
-24.67
35
25.33
65
4.67
35
-25.33
5.19 88.69632
55
-0.89
45
9.11
45
-9.11
45
-9.11
3.72 65.61791
85
-17.15
55
12.85
75
7.15
75
7.15
4.39 76.13664
85
-19.69
75
-9.69
85
19.69
95
29.69
3.57 63.26297
75
-12.58
55
7.42
75
12.58
85
22.58
4.84 83.20146
85
-19.88
45
20.12
45
-20.12
45
-20.12
4.14 72.21174

RI
PT

umbo diff Total


side
-5.81329
48.21 right
-18.9869
46.52 right
-29.4353
51.23 left
20.98128
87.56 left
18.236
109.16 left
12.1979
141.86 right
-4.25743
72.58 left
-9.34399
72.24 left
-2.98578
70.33 left
2.34703
61.27 right
15.01485
107.47 left
-19.2779
40.87 left
2.22205
99.73 right
11.40317
126.06 right
-14.6043
52.74 right
-2.86954
60.8 left
-0.52744
85.65 left
16.32469
106.33 left
-11.4827
89.7 left
-21.9971
50.4 left
16.19422
89.51 left
-1.86677
75.43 left
6.77796
99.08 left
-2.42642
61.09 left
-2.90216
58.93 left
1.445195
80.12 right
7.789795
80.34 right
9.84409
62.03 right
25.91083
66.46 right
-11.6396
61.37 left
-3.38198
51.56 left
8.32882
75.52 right
1.94655
53.52 right
-9.86318
74.63 left
5.967465
100.26 left
-17.2084
63.61 right
-7.84057
87.83 right
-20.0127
81.17 right
17.65996
92.66 right
6.517045
56.1 right
-13.5475
91.31 right
4.59911
95.37 right
11.25509
155.31 right
11.90889
137.59 left
-2.69861
94.35 right
-9.99578
63.32 right
-14.2505
53.4 right
-3.80904
57.87 left
-21.4071
60.33 right
-8.48773
54.11 left
-0.87233
67.85 right
8.83687
65.31 right
-13.4987
62.42 left
-2.2243
65.12 right

SC

umbo est
54.02329
65.50692
80.6653
66.57872
90.92401
129.6621
76.83743
81.58399
73.31578
58.92297
92.45516
60.14789
97.50795
114.6568
67.3443
63.66954
86.17744
90.00532
101.1827
72.39709
73.31578
77.29677
92.30204
63.51642
61.83216
78.67481
72.55021
52.18591
40.54917
73.00955
54.94198
67.19118
51.57345
84.49318
94.29254
80.81842
95.67057
101.1827
75.00005
49.58296
104.8575
90.77089
144.0549
125.6811
97.04861
73.31578
67.65053
61.67904
81.73711
62.59773
68.72233
56.47313
75.91874
67.3443

M
AN
U

bhing diff Far Hinge fhing est fhing diff Hinge length
length est length diff Umbo
-1.63632
28.29 56.07862 -7.86862
22.73 59.30516 -11.0952
3.06
-21.7047
31.47 61.82618 -15.3062
22.45 58.6354 -12.1154
3.81
0.31268
32 62.78411 -11.5541
27.32 70.28444 -19.0544
4.8
28.50308
47.01 89.91333 -2.35333
38.62 97.31404 -9.75404
3.88
35.32328
58.42 110.5359 -1.37588
50.6 125.9702 -16.8102
5.47
29.72432
79.23 148.1481 -6.28808
61.21 151.3493 -9.48932
8
7.0114
40.07 77.36991 -4.78991
32.05 81.5986 -9.0186
4.55
3.71544
41.31 79.6111 -7.3711
32.45 82.5554 -10.3154
4.86
10.15924
43.19 83.00903 -12.679
35.55 89.9706 -19.6406
4.32
6.28288
34.58 67.44723 -6.17723
28.6 73.3462 -12.0762
3.38
20.82412
62.67 118.2174 -10.7474
48.96 122.0473 -14.5773
5.57
-17.4586
26.98 53.71091 -12.8409
19.11 50.64612 -9.77612
3.46
21.69496
56.3 106.7042 -6.97417
45.15 112.9338 -13.2038
5.9
41.51328
72.79 136.5084 -10.4484
61.99 153.2151 -27.1551
7.02
0.06624
34.15 66.67004
-13.93
28.63 73.41796 -20.678
3.93
-3.61192
38.29 74.15272 -13.3527
30.62 78.17804 -17.378
3.69
5.81568
47.87 91.46771 -5.81771
36.5
92.243
-6.593
5.16
25.7674
55.09 104.5172 1.812793
41.4 103.9638
2.3662
5.41
17.9196
52.57 99.96253 -10.2625
43.8 109.7046 -20.0046
6.14
-11.57
33.92 66.25434 -15.8543
26.9 69.2798 -18.8798
4.26
24.06992
43.91 84.31036 5.199637
37.34 94.25228 -4.74228
4.32
-6.4178
35.35 68.83893 6.591067
25.15 65.0938 10.3362
4.58
21.5162
48.83 93.20282 5.87718
37.96 95.73532 3.34468
5.56
7.17388
29.92 59.0247 2.065303
24.03 62.41476 -1.32476
3.68
7.24156
30.2 59.53077 -0.60077
24.48 63.49116 -4.56116
3.57
18.79256
36.44 70.80901 9.31099
29.81 76.24052 3.87948
4.67
-2.87868
35 68.20634 12.13366
23.72 61.67324 18.66676
4.27
-8.76508
26.2 52.30113 9.728868
17.62 47.08204 14.94796
2.94
-2.57864
28.18 55.8798 10.5802
19.44 51.43548 15.02452
2.18
-25.6614
28.16 55.84366 5.526344
15.63 42.32196 19.04804
4.3
0.9854
22.25 45.16186 6.398138
17.84 47.60828 3.95172
3.12
6.05296
35.14 68.45938 7.060623
26.26 67.74892 7.77108
3.92
-1.3386
23.56 47.52957 5.99043
18.19 48.44548 5.07452
2.9
1.65008
35.48 69.0739 5.556103
26.65 68.6818
5.9482
5.05
-11.8328
46.4 88.81081 11.44919
28.71 73.60932 26.65068
5.69
-6.37112
27.17 54.05432 9.55568
18.52 49.23484 14.37516
4.81
8.50976
36.5 70.91746 16.91255
25.49 65.90708 21.92292
5.78
-11.0451
38.7 74.89376 6.276243
25.08 64.92636 16.24364
6.14
13.76816
38.81 75.09257 17.56743
27.63 71.02596 21.63404
4.43
5.13984
25.87 51.70469 4.395313
21.28 55.83676 0.26324
2.77
8.69108
51.39 97.82979 -6.51979
38.51 97.05092 -5.74092
6.38
28.85892
51.95 98.84194 -3.47194
39.88 100.328 -4.95796
5.46
16.35648
84.61 157.872 -2.56195
60.35 149.2922
6.0178
8.94
48.80212
72.44 135.8758 1.71423
58.27 144.3168 -6.72684
7.74
23.89764
53.87 102.3122 -7.96217
45.73 114.3212 -19.9712
5.87
8.8898
36.56 71.0259 -7.7059
30
76.695 -13.375
4.32
-15.8957
28.12 55.77136 -2.37136
18.39 48.92388 4.47612
3.95
6.0102
32.78 64.19389 -6.32389
27.22 70.04524 -12.1752
3.56
-15.1346
34.22 66.79656 -6.46656
24.71 64.04132 -3.71132
4.87
3.40688
29.45 58.17521 -4.06521
25.62 66.21804 -12.108
3.62
7.67924
36.54 70.98975 -3.13975
29.84 76.31228 -8.46228
4.02
2.18328
30.19 59.5127 5.797302
23.01 59.97492 5.33508
3.22
-2.59168
32.85 64.32041 -1.90041
24.78 64.20876 -1.78876
4.49
1.30784
33.13 64.82648 0.293517
25.41 65.71572 -0.59572
3.93

mussel
Base Hingebhing est
vex 2n1
4.88 49.84632
vex 5 4
9.17 68.22468
vex 3 4
5.13 50.91732
vex 4N 1
7.03 59.05692
vex 2n1
10.48 73.83672
vex 5n1
19.42 112.1357
vex 5n1 w
8.55 65.5686
vex 5n3 w
9.24 68.52456
vex 2n1 w
7.29 60.17076
vex 5n1 w
6.08 54.98712
vex 4n 1 w
13.47 86.64588
putty 4 3
6.86 58.32864
vex 3n 1 w
11.46 78.03504
vex 4n 1 s
12.98 84.54672
putty 6 2
5.54 52.67376
vex 2N2 s
8.28 64.41192
vex 1n1 w
11.88 79.83432
vex 5n1 w
12.05 80.5626
vex 5n 4 w
10 71.7804
vex 4n 1 w
7.71 61.97004
P95
8.52 65.44008
P92
12.35 81.8478
P96
11.35 77.5638
P66
5.83 53.91612
P65
5.31 51.68844
P90
7.56 61.32744
V45
12.67 83.21868
V37
9.77 70.79508
V36
9.36 69.03864
V33 s
13.56 87.03144
V12 s
5.05 50.5746
V41
9.46 69.46704
V21
6.05 54.8586
V26
10.28 72.97992
V10
19.41 112.0928
V47
9.58 69.98112
V31
11.76 79.32024
V12 l
14.77 92.21508
V33 l
11.66 78.89184
V24
5.14 50.96016
#13
12.53 82.61892
#12
8.77 66.51108
#2
25.68 138.9535
#8
13.97 88.78788
#14
9.69 70.45236
gray 2
5.95 54.4302
con 9
9.42 69.29568
black 11
5.35 51.8598
con 3
10.86 75.46464
gray 9
5.08 50.70312
black 2
7.29 60.17076
black 9
7.98 63.12672
black 5
8.42 65.01168
black 10
8.14 63.81216

TE

Site
PB
GDA
Beach
Helby
BS
Helby
BS
Helby
Diana
Sanf
Diana
Sanf
Helby
Diana
Beach
PB
Helby
Helby
Sanf
GDA
Piedras
Piedras
Piedras
Piedras
Piedras
Piedras
VAFB
VAFB
VAFB
VAFB
VAFB
VAFB
VAFB
VAFB
VAFB
VAFB
VAFB
VAFB
VAFB
VAFB
Mussel Pt
Mussel Pt
Mussel Pt
Mussel Pt
Mussel Pt

EP

Region
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
S Cali
N Cali
N Cali
N Cali
N Cali
N Cali
N Cali
N Cali
N Cali
N Cali
N Cali
N Cali
N Cali
N Cali
N Cali

AC
C

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54

Stefan diff precision


-21.4898
-0.92
-17.528
0.19
-36.8383
-0.35
21.15711
0.11
34.75031
1.19
19.88052
0.69
-6.85356
-0.05
-10.8045
0.03
-9.26056
-0.29
0.047976
-0.06
15.31976
0.16
-21.608
-0.06
-1.36901
-0.08
15.85522
0.46
-13.3489
0.18
-13.9237
-0.61
-1.63336
0.06
13.23779
-0.06
-10.771
0.2
-23.6957
0
13.05936
-0.09
-5.57352
-0.12
5.359804
0.05
-3.58594
0.02
-1.97803
0.15
7.280272
0.49
13.15213
0.45
-2.80293
-0.73
20.78058
-0.27
-12.7257
0.04
-3.22519
0.09
7.07616
0.02
0.77576
0
-10.4554
0.09
3.55688
-0.01
-12.3696
0.43
34.45778
2.84
-19.301
0.2
21.07624
0.33
4.297736
-0.07
-10.731
0.34
4.789724
0.15
51.54206
2.79
9.958668
0.07
-10.2029
-0.33
-8.89174
0.18
-12.6889
0.2
-3.50902
0.11
-28.3663
-0.32
-11.5079
-0.1
-8.28664
-0.37
2.047028
-0.35
-20.7815
-0.35
-7.09174
-0.21

precision est
-15.6765
1.458963
-7.40304
0.175828
16.51432
7.682624
-2.59614
-1.46048
-6.27478
-2.29905
0.304919
-2.3301
-3.59106
4.452054
1.255395
-11.0541
-1.10592
-3.0869
0.711686
-1.69861
-3.13486
-3.70675
-1.41816
-1.15952
0.924123
5.835077
5.362333
-12.647
-5.13025
-1.08615
0.156792
-1.25266
-1.17079
-0.59224
-2.41059
4.838767
42.29835
0.711686
3.416285
-2.21931
2.816486
0.190614
40.28697
-1.95022
-7.50432
1.104036
1.561625
0.30002
-6.95922
-3.02018
-7.41431
-6.78984
-7.28273
-4.86745

ACCEPTED MANUSCRIPT

EP

TE

M
AN
U

SC

RI
PT

Site
mussel
Base Hingebhing est bhing diff Far Hinge fhing est fhing diff Hinge length
length est length diff Umbo
umbo est umbo diff Total
Dicebox west
1
16.33 98.89772 10.69228
57.29 108.4929 1.097054
43.49 108.9456 0.644416
5.05 84.4957 25.0943
109.59
Dicebox west
2
6.37 56.22908 12.94092
35.1 68.38674 0.78326
29.4 75.24794 6.07794
3.39 59.07778 10.09222
69.17
Dicebox west
3
11.92 80.00528 28.34472
62.9 118.6325 10.28246
50.96 126.8108 18.46084
5.64 93.52978 14.82022
108.35
Dicebox west
4
5.84 53.95856 15.78856
19.1 39.46834 1.29834
13.33 36.81493 1.355072
2.71 48.66562 10.49562
38.17
Dicebox west
5
16.11 97.95524 37.10476
80.88 151.1295 16.06951
65.26 161.0107 25.95072
6.6 108.2293 26.8307
135.06
Dicebox west
6
11.32 77.43488 10.84512
47.58 90.94309 2.663092
37.37 94.30899 6.028992
5.49 91.23298 2.95298
88.28
Dicebox west
7
8.05 63.4262 18.5638
45.45 87.09333 5.10333
37.84 95.43304 13.44304
4.4 74.5429
7.4471
81.99
gilbert
8
31.97 165.8995 1.59948
102.81 190.7658 26.46579
74.02 181.9611 17.66113
11.63 185.2487 20.94866
164.3
gilbert
9
19.43 112.1781 26.72188
76.27 142.7974 3.897398
58.44
144.7 5.800004
10.27 164.4243 25.52434
138.9
gilbert
10
12.12 80.86208 50.04792
68.45 128.6635 2.24647
57.23 141.8062 10.89617
5.94 98.12338 32.78662
130.91
gilbert
11
14.7 91.9148 48.5352
71.06 133.3808 7.069156
57.09 141.4713 1.021344
5.76 95.36722 45.08278
140.45
dicebox east
12
7.78 62.26952 2.75048
30.65 60.34381 4.67619
24.14 62.66812 2.351876
1.91 36.41602 28.60398
65.02
dicebox east
13
12.35 81.8474 23.7974
47.61 90.99731 32.94731
37.52 94.66773 36.61773
5.77 95.52034 37.47034
58.05
dicebox east
14
5.72 53.44448 4.60552
35.14 68.45904 10.40904
30.52 77.92653 19.87653
3.85 66.1213
8.0713
58.05
dicebox east
15
4.69 49.03196 8.07196
22.58 45.75809 4.798092
18.87 50.06439 9.104392
2.84 50.65618 9.69618
40.96
dicebox east
16
4.79 49.46036 12.17964
39.51 76.35737 14.71737
35.23 89.19097 27.55097
3.54 61.37458 0.26542
61.64
dicebox west
17
11.65 78.8486
0.2214
48.85 93.23849 14.16849
38.23 96.36577 17.29577
4.02 68.72434 10.34566
79.07
dicebox west
18
11.36 77.60624 25.11376
57.15 108.2399 5.51991
47.64 118.8707 16.15072
5.72 94.75474 7.96526
102.72
dicebox west
19
11.26 77.17784 15.76216
46.26 88.55732 4.382676
36.2 91.51082 1.42918
4.28 72.70546 20.23454
92.94
dicebox west
20
5.41 52.11644 14.65644
27.81 55.21079 17.75079
23.5 61.1375 23.6775
3.05 53.8717 16.4117
37.46
Diana Islnd Beach 21
13.58 87.11672 8.01672
37.8 73.26672 2.486212
24.18 62.76379 10.8241
4.5 76.0741 21.86672
65.25
Diana Islnd Beach 22
14.18 89.68712 3.703162
38.13 73.86316 5.19594
25.1 64.96406
2.8517
4.3 73.0117 19.52712
70.16
Diana Islnd Beach 23
10.58 74.26472 3.167992
36.08 70.15799 6.392492
28.62 73.38249 5.92166
3.52 61.06834 7.27472
66.99
Diana Islnd Beach 24
9.14 68.09576 0.557558
35.33 68.80244 3.37676
28.35 72.73676 6.14798
3.66 63.21202 1.26424
69.36
Diana Islnd Beach 25
10.75
74.993 3.209944
42.56 81.86994 3.88232
32.45 82.54232 12.84494
3.83 65.81506
3.667
78.66
Diana Islnd Beach 26
16.91 101.3824 3.234404
49.54 94.4856 7.787636
35.54 89.93236 10.62126
5.22 87.09874 3.66244
97.72
Diana Islnd Beach 27
11.97 80.21948 7.556772
39.22 75.83323 10.72499
28.32 72.66501
8.0815
4.45 75.3085 3.17052
83.39
Diana Islnd Beach 28
19.38 111.9639 0.729824
58.76 111.1498 6.879432
41.23 103.5406 5.55934
6.38 104.8607 1.54392
110.42
Diana Islnd Beach 29
12.32 81.71888 6.404628
50.22 95.71463 6.768776
38.11 96.07878 1.13942
5.29 88.17058 7.59112
89.31
Diana Islnd Beach 30
22.05 123.4022 3.910032
60.68
114.62 10.97208
39.64 99.73792 4.31814
6.48 106.3919 12.6922
110.71
Diana Islnd Beach 31
33.52 172.5397 1.334832
62.68 118.2348 38.20816
30.84 78.69184 23.79074
8.72 140.6907 55.63968
116.9
Diana Islnd Beach 32
7.71 61.96964 3.586686
37.61 72.92331
0.7708
30.25 77.2808 3.80454
4.28 72.70546 14.54036
76.51
Diana Islnd Beach 33
23.01 127.5148 5.259118
45.93 87.96088 28.59076
24.96 64.62924
3.5253
5.85 96.7453 34.29484
93.22
Diana Islnd Beach 34
15.44 95.08496 3.786772
50.78 96.72677 1.094356
36.34 91.84564 4.41778
5.89 97.35778 2.14496
92.94
Diana Islnd Beach 35
10.88 75.54992 6.236422
48.03 91.75642 11.56325
38.53 97.08325
2.0381
5.25 87.5581 9.97008
85.52
Diana Islnd Beach 36
8.76 66.46784 2.86659
34.65 67.57341 1.794524
28.14 72.23452 8.75918
3.56 61.68082 3.97216
70.44
Diana Islnd Beach 37
13.58 87.11672 21.86672
37.8 73.26672 8.01672
24.18 62.76379 2.486212
4.5 76.0741 10.8241
65.25
Diana Islnd Beach 38
14.18 89.68712 19.52712
38.13 73.86316 3.703162
25.1 64.96406 5.19594
4.3 73.0117
2.8517
70.16
Diana Islnd Beach 39
10.58 74.26472 7.27472
36.08 70.15799 3.167992
28.62 73.38249 6.392492
3.52 61.06834 5.92166
66.99
Diana Islnd Beach 40
9.14 68.09576 1.26424
35.33 68.80244 0.557558
28.35 72.73676 3.37676
3.66 63.21202 6.14798
69.36
Diana Islnd Beach 41
10.75
74.993
3.667
42.56 81.86994 3.209944
32.45 82.54232 3.88232
3.83 65.81506 12.84494
78.66
Diana Islnd Beach 42
16.91 101.3824 3.66244
49.54 94.4856 3.234404
35.54 89.93236 7.787636
5.22 87.09874 10.62126
97.72

AC
C

Region
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC

side
left
left
left
right
left
left
right
left
right
right
right
left
left
left
left
left
left
right
right
right

ACCEPTED MANUSCRIPT

7.556772
0.729824
6.404628
3.910032
1.334832
3.586686
5.259118
3.786772
6.236422
2.86659
0.960786
7.269912
0.08523
0.600486
0.199476
4.430846
1.534412
3.42098
4.617478
3.534168
17.59757
3.25321
0.139212
2.439912
1.513384
0.48279

28.32
41.23
38.11
39.64
30.84
30.25
24.96
36.34
38.53
28.14
32.55
58.24
42.78
42.21
43.17
37.64
15.94
10.07
11.62
9.69
40.04
61.05
15.34
20.99
9.8
18.6

72.66501
103.5406
96.07878
99.73792
78.69184
77.2808
64.62924
91.84564
97.08325
72.23452
82.78148
144.2217
107.2475
105.8843
108.1803
94.95472
43.057
29.01831
32.72529
28.1095
100.6946
150.9421
41.62204
55.13458
28.37258
49.41866

10.72499
6.879432
6.768776
10.97208
38.20816
0.7708
28.59076
1.094356
11.56325
1.794524
2.57148
7.591684
9.947548
8.864336
13.10973
3.365276
4.617004
4.538312
4.435292
4.429504
16.97456
1.13208
2.882044
4.044584
0.51258
4.85866

RI
PT

75.83323
111.1498
95.71463
114.62
118.2348
72.92331
87.96088
96.72677
91.75642
67.57341
79.24921
143.8999
97.21477
96.41951
121.0905
93.88915
39.97441
27.90098
32.90748
27.21417
101.3176
146.5568
38.60079
53.52991
26.34662
44.07721

SC

39.22
58.76
50.22
60.68
62.68
37.61
45.93
50.78
48.03
34.65
41.11
76.88
51.05
50.61
64.26
49.21
19.38
12.7
15.47
12.32
53.32
78.35
18.62
26.88
11.84
21.65

M
AN
U

3.17052
1.54392
7.59112
12.6922
55.63968
14.54036
34.29484
2.14496
9.97008
3.97216
13.9992
27.62204
32.24588
30.93772
3.35456
20.41388
8.66416
20.3108
14.05892
18.71176
9.45528
44.27208
8.57836
3.59684
15.51708
1.47316

80.21948
111.9639
81.71888
123.4022
172.5397
61.96964
127.5148
95.08496
75.54992
66.46784
66.2108
109.008
65.05412
66.08228
124.6446
77.90612
47.10416
44.7908
42.34892
42.39176
74.26472
105.5379
47.31836
54.68684
43.37708
46.03316

TE

11.97
19.38
12.32
22.05
33.52
7.71
23.01
15.44
10.88
8.76
8.7
18.69
8.43
8.67
22.34
11.43
4.24
3.7
3.13
3.14
10.58
17.88
4.29
6.01
3.37
3.99

EP

Diana Islnd Beach 43


Diana Islnd Beach 44
Diana Islnd Beach 45
Diana Islnd Beach 46
Diana Islnd Beach 47
Diana Islnd Beach 48
Diana Islnd Beach 49
Diana Islnd Beach 50
Diana Islnd Beach 51
Diana Islnd Beach 52
Calvert Islnd 2nd Bch
53
Calvert Islnd 2nd Bch
54
Calvert Islnd 2nd Bch
55
Calvert Islnd 2nd Bch
56
Calvert Islnd 2nd Bch
57
Calvert Islnd 2nd Bch
58
Calvert Islnd 2nd Bch
59
Calvert Islnd 2nd Bch
60
Calvert Islnd 2nd Bch
61
Calvert Islnd 2nd Bch
62
Calvert Islnd 2nd Bch
63
Calvert Islnd 2nd Bch
64
Calvert Islnd 2nd Bch
65
Calvert Islnd 2nd Bch
66
Calvert Islnd 2nd Bch
67
Calvert Islnd 2nd Bch
68

AC
C

BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC
BC

4.45
6.38
5.29
6.48
8.72
4.28
5.85
5.89
5.25
3.56
4.87
9.08
5.31
5.48
7.14
5.09
2.01
1.18
1.51
1.32
4.48
8.48
1.98
3.28
1.2
2.22

75.3085
8.0815
104.8607 5.55934
88.17058 1.13942
106.3919 4.31814
140.6907 23.79074
72.70546 3.80454
96.7453
3.5253
97.35778 4.41778
87.5581
2.0381
61.68082 8.75918
81.73954 1.52954
146.2031 9.57306
88.47682 8.82318
91.07986 5.94014
116.4978 4.79222
85.10818 13.21182
37.94722 0.49278
25.23826 0.75826
30.29122 2.00122
27.38194 3.70194
75.76786 7.95214
137.0159 12.79414
37.48786 1.25214
57.39346 6.30346
25.5445
2.3155
41.16274 3.39726

83.39
110.42
89.31
110.71
116.9
76.51
93.22
92.94
85.52
70.44
80.21
136.63
97.3
97.02
121.29
98.32
38.44
24.48
28.29
23.68
83.72
149.81
38.74
51.09
27.86
44.56

ACCEPTED MANUSCRIPT

EP

TE

M
AN
U

SC

RI
PT

Sample Number
Total Length
1.(mm)
Umbo to2.
outer
length
hinge
of3.hinge
teeth
umbo
teeth
thickness
4. umbo to 5.
inner
greatest
hingeshell
height
teethweight
(cross-section)
outer
(mg) hingelength
prediction
prediction
umbo prediction
inner hingeError
prediction
outerError lengthError umboError inner
HB1
53.25
25.81
24.25
2.66
3.52
10.73
1779 51.59599 62.9312 47.90002 44.01968 1.654006
9.6812 5.34998 9.23032
HB2
54.81
25.36
21.46
3.31
5.89
11.21
2555 50.78266 56.25864 57.85282 54.17276 4.027336 1.448636 3.04282 0.63724
HB3
46.47
20.58
18.59
1.98
2.78
8.53
1320 42.14329 49.39474 37.48786 40.84952 4.326708 2.924744 8.98214 5.62048
HB4
51.74
26.84
24
2.88
3.99
10.2
3523 53.45762 62.3333 51.26866 46.03316 1.717616 10.5933 0.47134 5.70684
HB5
41.15
22.33
19.37
2.27
4.16
9.5
1802 45.30624 51.26019 41.92834 46.76144 4.156242 10.11019 0.77834 5.61144
HB6
32.03
15.84
13.32
1.39
3.2
7.21
670 33.57622 36.79101 28.45378 42.6488 1.546216 4.761012 3.57622 10.6188
HB7
46.05
23.11
13.54
2.45
8.06
9.3
46.71601 37.31716 44.6845 63.46904 0.666014 8.732836
1.3655 17.41904
HB8
44.3
21.27
14.14
2.05
8.28
8.39
43.3904 38.75212 38.5597 64.41152 0.909602 5.547876
5.7403 20.11152
HB9
54.69
18.07
12.1
2.07
10.1
37.60672 33.87326 38.86594
28.94 17.08328 20.81674 15.82406
25.75
HB10
43.52
19.93
18.2
1.5
3.08
7.06
892 40.96848 48.46202 30.1381 42.13472 2.551518 4.94202 13.3819 1.38528
HB11
35.65
19.32
15.15
1.25
3.87
6.39
634 39.86597 41.16764 26.3101 45.51908 4.215968 5.51764
9.3399 9.86908
HB12
39.47
16.83
13.85
1.33
7.39
725 35.36554 38.05856 27.53506
28.94 4.104458 1.41144 11.93494
10.53
HB13
85.8
32.8
28.65
2.62
14.52
6023 64.22972 73.45424 47.28754
28.94 21.57028 12.34576 38.51246
56.86
HB14
38.85
18.61
17.21
2.12
2.71
8
38.58271 46.09434 39.63154 40.54964 0.267286 7.244336 0.78154 1.69964
HB15
37.95
17.66
15.38
2.1
2.84
7.7
36.86568 41.71771 39.3253 41.10656 1.084316 3.767708
1.3753 3.15656
HB16b
34.54
16.05
12.66
1.26
3.58
7.35
829 33.95577 35.21256 26.46322 44.27672 0.58423 0.672556 8.07678 9.73672
HB17b
30.8
15.97
9.1
0.87
8.2
7.13
33.81118 26.69846 20.49154 64.0688 3.011178 4.10154 10.30846 33.2688
HB18a
31.92
13.77
10.38
1.65
4.5
7.08
676 29.8349 29.75971 32.4349
48.218 2.085102 2.160292
0.5149
16.298
HB19b
36.51
16.55
13.29
1.9
4.45
8.08
1134 34.85947 36.71926 36.2629 48.0038 1.65053 0.209264
0.2471 11.4938
HB20b
27.41
13.9
10.33
1.41
2.73
6.3
30.06986 29.64013 28.76002 40.63532 2.65986 2.230128 1.35002 13.22532
HB21
37.81
16.96
11.77
1.91
7.32
935 35.6005 33.08403 36.41602
28.94 2.209496 4.725968 1.39398
8.87
HB22
45.29
23.78
19.29
2.72
2.84
9.44
2188 47.92697 51.06886 48.81874 41.10656 2.636972 5.778864 3.52874 4.18344
HB23
44.74
20.38
15.7
2.68
5.03
9.53
41.78181 42.48302 48.20626 50.48852 2.958188 2.25698 3.46626 5.74852
HB24
51.7
29.4
25.79
2.81
4.15
10.78
58.08456 66.61426 50.19682 46.7186 6.38456 14.91426 1.50318
4.9814
HB25
47.8
23.65
16.94
3.12
6.1
10.07
2804 47.69201 45.4486 54.94354 55.0724 0.10799 2.351396 7.14354
7.2724
HB26
53.39
24.09
21.36
2.44
3.6
9.4
2795 48.48727 56.01948 44.53138 44.3624 4.902734 2.629476 8.85862
9.0276
HB27
48.29
27.15
20.95
2.91
3.14
9.39
2020 54.01791 55.03892 51.72802 42.39176 5.72791 6.74892 3.43802 5.89824
HB28
45.34
23.13
19.1
2.35
3.55
8.46
1764 46.75216 50.61446 43.1533 44.1482 1.412162 5.27446
2.1867
1.1918
HB29
42.23
21.38
18.77
2.13
2.63
7.87
1593 43.58921 49.82523 39.78466 40.20692 1.359212 7.595232 2.44534 2.02308
HB30
51.69
28.76
25.27
2.69
5.81
9.12
56.92782 65.37063 48.35938 53.83004 5.237824 13.68063 3.33062 2.14004
HB31b
70.55
35.89
32.46
3.45
4.38
13.78
69.81459 82.56624 59.9965 47.70392 0.735414 12.01624 10.5535 22.84608
HB32b
92.16
43.91
38.44
4.43
6.04
16.58
84.30993
96.868 75.00226 54.81536 7.850066 4.708004 17.15774 37.34464
HB33b
95.22
43.84
38.54
3.59
4.96
16.84
84.18342 97.10716 62.14018 50.18864 11.03658 1.887164 33.07982 45.03136
HB34
82.02
35.08
27.58
3.73
4.96
14.81
68.35059 70.89523 64.28386 50.18864 13.66941 11.12477 17.73614 31.83136
HB35
72.85
37.9
33.17
3.76
5.07
11.27
73.44746 84.26427 64.74322 50.65988 0.59746 11.41427 8.10678 22.19012
HB36b
82.38
38.32
34.84
3.06
3.64
15.54
7586 74.20657 88.25824 54.02482 44.53376 8.173432 5.878244 28.35518 37.84624
HB37b
47.99
21.45
17.97
2.48
3.81
9.11
2023 43.71573 47.91195 45.14386 45.26204 4.27427 0.078048 2.84614 2.72796
HB38b
61.01
33.56
29.06
2.88
5.43
11.94
4095 65.60334 74.4348 51.26866 52.20212 4.593344 13.4248 9.74134 8.80788
HB39b
18.8
8.26
6.8
1.03
2.02
4.1
166 19.87612 21.19778 22.94146 37.59368 1.076124 2.39778 4.14146 18.79368
HB40
70.43
36.28
27.9
2.6
7.77
14.92
70.51947 71.66054 46.9813 62.22668 0.089472 1.23054 23.4487 8.20332
HB41b
25.62
12.12
9.85
1.49
3.56
5.3
398 26.85269 28.49216 29.98498 44.19104 1.232688 2.87216 4.36498 18.57104
HB42b
23.5
10.36
8.51
1.52
2.89
4.53
283 23.67166 25.28742 30.44434 41.32076 0.171664 1.787416 6.94434 17.82076

AC
C

Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Hyacinthe Bay
Hyacinthe Bay
Hyacinthe Bay
Hyacinthe Bay
Hyacinthe Bay
Hyacinthe Bay
Hyacinthe Bay
Hyacinthe Bay
Hyacinthe Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay
Heriot Bay