When key biological statistics (growth parameters, mortality rates, etc.) o
f a large number of fish species are plotted against other variables (notab
ly maximum size and mean habitat temperature), clear trends or patterns ten
d to emerge. These patterns are usually not perceived as requiring explanat
ion, as reflected in the rarity of pertinent hypotheses in the literature,
relative to the ubiquity of the patterns. A theory is presented which has t
he potential to explain the propensities behind many of these observed patt
erns. This theory, based on the author's earlier work, is built around the
notion that gill area severely constrains the respiration and hence the gro
wth of fishes (and other water-breathing animals). The patterns generated b
y growth-related processes, such as mortality, reproduction or food consump
tion can then also be explained. In this theory, tropical fishes are not di
fferent from other fishes, once account is taken of the fact that tropical
fish are, not metaphorically but literally, in hot water. (C) 1998 The Fish
eries Society of the British Isles.