The Biology and Culture of Tilapias

nor suggest a possible explanation for the overwhelming incidence of mater-
nal bearer species beyond the following comments, as this issue is not central
to our discussion.
All female-uniparental Sarotherodon forms are lek species in courtship and
breeding (Fryer and Iles 1972; Lowe-McConnell 1975; Loiselle and Barlow
1978). Field observations on reproductive behavior of male-uniparental
Sarotherodon forms are insufficient to draw any firm conclusions from
(Lowe-McConnell, Trewavas, pers. comm.). This by itself is not a remarkable
correlation, since a number of other teleosts also have lek breeding systems,
but with uniparental male custodial care (e.g. Centrarchidae, Belontiidae,
Gasterosteidae; Loiselle and Barlow 1978; Keenleyside 1979). Such com-
munal mating displays (leks) have been noted (Southwood 1976) as typical
of "K-strategists" but irrespective of whether this is a valid generalization,
we shall see that it does not necessarily restrict these Sarotherodon forms to
a strictly K-selected strategy.
Nor is it mouthbrooding (bearing) which is responsible for the parental
roles in Sarotherodon. The general pattern of parental roles in mouthbrood-
ing (bearing) teleosts does not appear significantly different from that
in parental teleosts in general (nonbearers) (Oppenheimer 1970), although
the data are very sketchy (Breder and Rosen 1966). There must be some
aspect of the ecology and/or life history of the Sarotherodon forms which
predisposes them to, or necessitates, the predominant role for the female in
parental care (see also Maynard Smith 1977). Neither their position as
cichlids, nor as lek breeding species nor as bearers can account for the female
parental role. The proposal by Barlow (1974) of an evolutionary progression
from biparental care, to polygyny and uniparental female care certainly
agrees with the major trend in tilapias, assuming selection favors parental
care.
If we assume that these species were likely adapted for riverine existence
in their ancestral form(s) (Fryer and Iles 1972; Liem 1973; Greenwood
1974), they would be better adapted (over an evolutionary time scale) as
altricial rather than precocial forms. The evidence of such fluvial species is
that they tend to be adapted to breed seasonally, often in flooded or inun-
dated habitats with strong seasonal and/or yearly fluctuations (Lowe-McCon-
nell1975; Welcomme 1979a). But if conditions did not permit guarding (e.g.,
unsuitable substrate for a nest, with danger of exposure, etc.) and hence
bearing were favored, maternal bearing (mouthbrooding) would most likely
be the end result of an extended evolutionary progression (Barlow 1974;
Loiselle and Barlow 1978; Balon 1981b; but see Timms and Keenleyside
1975). Perrone and Zaret (1979) hypothesized that lakedwelling cichlids,
because of the rather constant environmental conditions, with extended or
continuous breeding seasons, should be maternal mouthbrooders (bearers).
They suggested that riverine cichlids should have a greater incidence of
male parental care, including biparental guarding, because of stronger seasonal
changes and more restricted breeding seasons. The correlation may be a
general one, but may also be difficult to apply rigorously to all tilapias(e.g.,
Lowe-McConnell, this volume).
Bearing and its attendant adaptations are clearly an advanced, precocial
life style, seemingly ill-suited to the kind of habitat we have described for a