A similar fish subjected to a 12-hour daytime temperature of 30°C and
with a nocturnal non-feeding temperature of 18°C would, in order to grow
by the same 1,000 J/d, have to consume:
( (88.7 x 50 x 0.5) - 10.1 + 1000/0.437 ) + (27.53 x 50 x 0.5) = 8028 J/d
Thus, for individuals from either of the two groups to grow by equal amounts,
those fish maintained at a constant temperature require 1.6 times more food
than the group subjected to a thermal oscillation of 30°C to 18°C. On
testing a variety of combinations in a similar manner, it becomes evident that
as the temperature oscillations become less marked, so the ratio between the
energy of growth and the energy of food consumed (A B/C, Odum's ecol-
ogical growth efficiency) decreases, resulting in a decline in the efficiency of
production.
Estimates of food consumption conducted in (optimal?) T. rendalli
habitats at Lake Kariba during summer indicate that a 50 g fish consumed
an average of 12.6 kJ/d of food, which is equivalent to about 8.2 g/d of fresh
food (maximum measured was 10 g/d). Growth estimates obtained from the
same wild population by tagging indicate an instantaneous growth capability
of 690 mg/d or 3.6 kJ/d. Calculating the ecological growth efficiency (using
energy values) of this population, a value of 0.29 is obtained. If the same size
fish consumed the same amount of food energy but was maintained at 28°C
(mean maximum and minimum of midsummer Kariba temperature oscilla-
tion) then growth would be calculated to be 1.3 kJ/d (derived graphically
from Table 6) with an ecological growth efficiency of 0.10 or approximately
half that calculated for the wild population.
The key factor in this change of growth efficiency is the metabolic energy
demand, since, if food is not limiting, the other parameters of the energy
balance do not vary in sufficient magnitude to seriously affect the balance at
any given temperature. The relationship between food intake and temperature
has, in the laboratory, been shown to be strongly temperature related but
indications are that in the field, a less marked influence is found, although
fish do feed more actively at higher temperatures. This being the case, then,
some influence will be reflected in the growth efficiency, but it still remains
more likely that metabolism is the governing factor. The pattern of change in
metabolic demand with changing temperature may be a complicating factor,
but, on the whole, relationships of the type shown for T. rendalli or S. mos-
sambicus (Figure 3) result in only very subtle variations in the final result.
More important is the effect of fish size on metabolism. The smaller the fish
the greater the benefit of thermal oscillations may be since the metabolic
demand by small fish is greater, per unit mass, than larger fish. Thus, there
may well be a limit whereby larger fish no longer benefit from die1 migrations
to warm shallow water and this, combined with 'food distribution, may be a
reason why juvenile fish are far more abundant in shallow marginal areas of
gradient shorelines.
One is able to draw on limited examples from wild populations to support
the theory that thermal oscillations are beneficial to the growth of juvenile
cichlids. The work of Coe (1966,1967) showed that the resident population
of S. alcalicus (T. grahami) in the thermal springs of Lake Magadi (Kenya)