Specific trace elements incorporated into the growing surface of the fish o
tolith reflect the physical and chemical characteristics of the ambient wat
er, although not necessarily in a simplistic manner. Since fish which spend
at least part of their lives in different water masses often produce otoli
ths of different elemental composition, the otolith elemental composition (
'elemental fingerprint') can serve as an environmentally induced tag of gro
ups of fish. On the basis of isotope dilution ICPMS (ID-ICPMS) assays of ne
arly 2500 dissolved adult cod (Gadus morhua) otoliths, it has become clear
that cod otolith elemental fingerprints based on the elements Li, Mg, Mn, S
r and Ba are physically stable, reproducible and consistent between left an
d right otoliths. Highly significant differences existed among the fingerpr
ints of all of the spawning aggregations, resulting in a characteristic mar
ker for each aggregation. Long-term stability (4-13 years) of the fingerpri
nts for a given spawning group was not evident, indicating that the fingerp
rint was not a proxy for genetic identity. However, the fingerprint was ver
y stable over the shea-term (up to 1 year), suggesting that it could serve
as a seasonally stable biological tracer, or natural tag, of pre-defined gr
oups of fish, even during situations of extensive stock mixing. As an illus
tration of the tracer approach, a maximum likelihood-based stock mixture an
alysis was applied to feeding (summer) and over-wintering stock distributio
ns, using the fingerprints of the spring spawning aggregations as known-sto
ck reference samples. The results of the summer stock mixture analyses sugg
ested that the mixture analysis was accurate within 1%, while the stock mix
ture analysis of the over-wintering schools produced stock-specific distrib
utions which would have been difficult to obtain using alternative approach
es. While the use of elemental fingerprints as natural tags is not suited t
o all stock mixing situations, suitability can probably be determined befor
ehand on the basis of existing environmental and biological information. (C
) 2000 Elsevier Science B.V. All rights reserved.