Certain species of fish have recently become important model systems i
n comparative genomics and in developmental biology, in certain instan
ces because of their small genome sizes (e.g., in the pufferfish) and,
in other cases, because of the opportunity they provide to combine an
easily accessible and experimentally manipulable embryology with the
power of genetic approaches (e.g., in the zebrafish). The resulting ac
cumulation of genomic information indicates that, surprisingly, many g
ene families of fish consist of more members than in mammals. Most mod
ern fish, including the zebrafish and medakka, are diploid organ isms;
however, the greater number of genes in fish was possibly caused by a
dditional ancient genome duplications which happened in the lineage le
ading to modern ray-finned fishes but not along the lineage leading to
tetrapods. Since these two lineages shared their last common ancestor
(in the Devonian about 360 million years ago) individual duplicated m
embers of gene families were later lost in fish. Interestingly, compar
ative data indicate that, in some cases, genes in mammals even serve s
omewhat different functions than their homologues in fish, highlightin
g that the degree of evolutionary relatedness of genes is not always a
reliable predictor of their evolutionary conservation and their simil
arity of function. Since fish are phenotypically probably not more com
plex than mammals, it is possible that evolution took alternative path
s to the ''economics of genomics'' through alternative solutions to ge
ne regulation. It is suggested that the more complex genomic architect
ure of fish permitted them to adapt and speciate quickly in response t
o changing selective regimes. (C) 1998 John Wiley & Sons, Inc.