416 
Fishery Bulletin 99(3) 
Table 3 
Summary of results from Mantel’s test and the BIOENV procedure where otolith microchemistry data were compared with the 
microchemistry data obtained from eye lenses, scales, and spines. Comparisons are also made between juvenile and adult fish for 
each structure. The BIOENV procedure (see general text) was done only if significant correlations were found with Mantel’s test. 
Only the combination of elements that contributed to the maximum p overall, as measured by the weighted Spearman correlation, 
are shown for each structure comparison. The significance of the Mantel test statistic is shown by * (P< 0.05) or ** (PcO.Ol). 
Comparison 
Mantel r statistic 
BIOENV-maximum p 
Combination of elements 
contributing to maximum p 
Juveniles 
Otoliths and eye lenses 
0.3402* 
0.127 
Mn, Ba, Hg 
Otoliths and scales 
0.5757** 
0.587 
Sr, Ba 
Otoliths and spines 
0.4097** 
0.403 
Sr, Ba 
Adults 
Otoliths and eye lenses 
-0.0646 
Otoliths and scales 
0.3680** 
0.454 
Mn, Ba 
Otoliths and spines 
0.4440** 
0.512 
Mn, Sr, Ba 
Adults and juveniles 
Otoliths 
0.2000 
Eye lenses 
-0.1357 
Scales 
-0.0136 
Spines 
0.0676 
eralisation (Simkiss, 1974). Despite the possibility of re- 
sorption of some elements, otoliths and scales, and otoliths 
and spines showed a significant correspondence between 
data matrices for both juvenile and adult fish, which may 
indicate that scales and spines provide estimates of stock 
structure that are similar to those obtained from otoliths. 
Differences in elemental fingerprints among structures 
may be due to different l'outes of ion uptake and differen- 
tial abilities of each structure to incorporate elements into 
the organic and inorganic matrix. Calcium and strontium 
in otoliths are primarily taken up by the gills (Simkiss, 
1974), but the route of uptake of trace elements has not 
been identified (Campana and Gagne, 1995). There may 
also be some Sr uptake in otoliths through the diet be- 
cause fish that were fed a Sr-enriched diet showed a de- 
tectable increase in Sr/Ca ratios (Gallahar and Kingsford, 
1996). Other structures, such as scales and bone, may in- 
corporate ions by diffusion across the gills and through 
the skin or by ingestion of food and water (Simkiss, 1974). 
Whether some structures show differential abilities to in- 
corporate ions through different methods of absorption 
and whether ions are resorbed from different structures in 
equal proportions is not known. 
Otoliths and eye lenses showed the greatest differences 
in elemental fingerprints between structures. Otoliths are 
predominantly CaC0 3 (Ca constitutes from 30% to 39% of 
otoliths. Thresher et al., 1994; Dove et al., 1996), although 
small amounts of protein also occur (Degens et ah, 1969) and 
therefore otoliths are likely to incorporate ions that are able 
to substitute for Ca in the CaC0 3 matrix or bind to proteins 
in the organic matrix (Gunn et ah, 1992; Sie and Thresher, 
1992). The spaces between these matrices may also trap 
ions. In contrast, eye lenses are composed largely of water 
and structural protein, the latter of which may be soluble or 
insoluble crystallin (Nicol, 1989). The proteins of eye lenses 
are rich in sulfhydryl groups that may covalently bind with 
metals and the proteins also have specific sites for binding 
with cations (Sharma et ah, 1989). Scales and bone are more 
similar to otoliths in that the dominant ion is Ca (e.g. in 
bone Ca may constitutes from 24% to 37%; Hamada et ah, 
1995), but they vary considerably in that otoliths are pri- 
marily carbonate structures, whereas scales and bone are 
primarily phosphate (or hydroxyapatite) structures. Some 
differences in elemental composition of otoliths and scales or 
spines may therefore be expected. Because otoliths, scales, 
and bone are composed predominantly of a mineral, rather 
than an organic matrix, it is not surprising that these struc- 
tures were similar in their elemental fingerprints and that 
eye lenses presented a different fingerprint. 
Differences in elemental fingerprints between juvenile 
and adult fish were also found for all structures. Eye lenses 
are thought to have no efficient mechanism for removing 
ions, but there may be changes in structural proteins with 
age that possibly alter affinities of proteins for specific ions 
(Dove, 1997). Such ontogenetic effects may result in differ- 
ent elemental fingerprints between juvenile and adult fish 
for eye lenses. Both scales and spines show some evidence 
of resorption or remineralization over time in some species. 
For example, early growth increments in spines of several 
species are known to be destroyed as the core-matrix ex- 
pands (Hill et al., 1989; Gillanders et al., 1997) and there is 
evidence for resorption in scales of fish living in a stressed 
environment (Sauer and Watabe, 1989). 
If an alternative structure to otoliths is required for stock 
identification, as it may be for broodstock, and for rare 
or endangered species, or if removing otoliths decreases 
