Crockford: Archeological evidence of Thunnus thynnus off British Columbia and northern Washington 
15 
Table 2 
Calculated fork lengths (cm) and estimated weights (kg) of comparative specimens — USNM catalog numbers 269001, 269004, 
268964, 269002 (Nankai collection numbers 1, 2, 3, 6) National Museum of Natural History (NMNH), Smithsonian Institution. 
Nankai 1 
269001 
Nankai 2 
269004 
Nankai 3 
268964 
Nankai 6 
269002 
Skull length (cm) 
26.2 
25.3 
23.5 
28.5 
Total of vertebral lengths (1-39 cm) 7 
148.1 
124.1 
121.2 
146.9 
Total skeletal length (SL) (cm) 
174.3 
149.4 
144.7 
175.4 
Estimate of intervertebral cartilage — 40 spaces 
20.0 
20.0 
20.0 
20.0 
Estimate of snout and tail flesh (cm) 
10.0 
10.0 
10.0 
10.0 
Estimated fork length (cm) 
204.3 
179.4 
174.7 
205.4 
Estimated weight (kg) 2 
184 
130 
121 
187 
1 All measurements available from the author or NMNH, Smithsonian Institution. 
2 Foreman and Ishizuka, 1990; 184. 
Recent excavations at four locations along Toquart 
Bay in Barkley Sound on the west coast of Vancouver 
Island have recovered relatively large numbers of 
both vertebral and nonvertebral bluefin tuna skel- 
etal remains. Full analysis of this material is still in 
progress: only a few of the nonvertebral remains have 
been examined thus far. All vertebrae, however, are 
included in this study. 
Modern skeletal samples 
In order to estimate the size of fish represented by 
isolated vertebrae from archeological samples, it was 
necessary to determine the size relationship between 
individual vertebrae and the corresponding fork length 
in modem samples of the fish. Measurements taken 
from the vertebrae of modem skeletal specimens of 
known-size fish of comparable size were used for this 
purpose (Casteel, 1976; Wheeler and Jones, 1989). 
Recent skeletal specimens of large ( 160 cm TL and 
over) Pacific bluefin tuna were found to be extremely 
rare, and the only known specimens had, unfortu- 
nately, no corresponding size data (length or weight); 
therefore fork lengths (snout to fork of the tail) had 
to be estimated for these specimens as well. Fortu- 
nately, these four recent specimens of bluefin tuna 
(loaned by B. Collette, Museum of Natural History, 
Smithsonian Institution, Washington, D.C.) have 
skulls that are still articulated, and it was possible 
to determine a “skeletal length” for these specimens 
(Table 2). The skeletal length is defined as the basal 
length of the skull plus the combined lengths of all 
39 vertebrae. The vertebral column of the compara- 
tive specimens had been sawed into sections during 
skeletal preparation, sometimes by cutting through 
a centrum. Vertebra no. 30, either by itself or with 
portions of no. 29 and no. 31 attached, was appar- 
ently removed from the specimens at some point and 
not returned. Estimates of the length measurements 
of all three of these vertebrae were used in the re- 
gression equations. These four fish appear to be the 
only disarticulated skeletal specimens of large Pa- 
cific bluefin tuna available for analysis (however, 
several museums have reconstructed skeletal speci- 
mens of large individuals on display). All raw data 
for these specimens are available on request from 
the author and are also on file at the National Mu- 
seum of Natural History, Smithsonian Institution. 
In order to estimate a fork length from the skel- 
etal length for these comparative specimens, I as- 
signed a value of 0.5 cm to the intervertebral carti- 
lage (40 spaces, 20 cm total) and an additional 5 cm 
each for flesh on the snout and the tail. These values 
consistently added 30 cm to the measured skeletal 
length and yielded an estimated fork length. This 
method was chosen so that if a more accurate deter- 
mination of the “soft tissue” component of the fork 
length of bluefin tuna is subsequently developed, the 
estimates given in this report can be easily adjusted. 
The vertebral centrum length and breadth mea- 
surements from the four comparative specimens 
(Fig. 2) were used in single (least-squares) regres- 
sion equations for each of the 39 vertebrae in the 
spinal column by using logarithmic transformations 
of vertebral and skeletal length measurements to de- 
termine their linear relationship. Because the size 
and shape of vertebrae change (sometimes quite dra- 
matically) over the length of the fish, it was neces- 
sary to calculate a separate algorithm for each ver- 
tebra in the spinal column. 
