presence appears to be a factor limiting cichlid populations. Underwater
observations have confirmed that cichlid distribution in the openwaters of
the main rivers is limited to those tilapia which have managed to reach a size
large enough (c.19 cm TL) to withstand predation by Hydrocynus. In the
Kafue Hepsetus odoe is the common openwater predator, but in the Upper
Zambezi Hepsetus is apparently restricted by competition to habitats not
frequented by Hydrocynus, such as lagoons, backwaters and smaller tribu-
taries. Bell-Cross (1974) commented that in years of poor rainfall, when the
main river is confined between sandy banks for most of the year and the
isolated floodplain pools, which serve as reservoirs for tilapia stocks, dry up,
tke predation on juvenile and small adult cichlids must be extremely severe.
Preimpoundment studies on. +Ae Kafue floodplain suggested that predation is
particularly high at two periods a year: for tilapia in the rivers in the dry
season, and for juvenile fishes as the floodplains dry up, a time when pisci-
vorous fishes (Hepsetus and Clarias) and the numerous birds and other
animals can catch them very easily (Williams 1971).
The preimpoundment survey for Lake Kariba in the Middle Zambezi
showed S. mortimeri (then called S. mossambicus) to be the commonest
tilapia, but not nearly as abundant as Labeo, Distichodus, Hydrocynus and
Alestes. The tilapia hugged the shore or lived in vegetation rather than in
openwater, the juveniles in very shallow water. Some T. rendalli (then called
T. melanopleura) lived in weedy shallows. In an attempt to provide a more
openwater-living species the new Kariba lake was stocked with S. macrochir,
known in Lake Mweru to live in openwater, but in the event stock was taken
from the nearby Kafue, so was what was later known to be the 'volcano-nest'
subspecies, not the 'star-nest' one used to openwater conditions. In the four
months after dam closure in 1959 over 11,000 fingerlings of S. macrochir and
T. rendalli were stocked. For many years these S. macrochir were not seen;
many of the stocked fish were apparently consumed by Hydrocynus in the
new lake, a piscivore not previously encountered by the Kafue fishes (Van der
Lingen 1973). It was the indigenous S. mortimeri which thrived and underwent
a population explosion in the new lake, becoming very abundant within three
years. Recently S. macrochir have reappeared and are slowly increasing in
numbers (Bowmaker et al. 1978). In Lake Kariba, after the flooded terrestrial
vegetation rotted (by 1963) there was a hiatus before the establishment of
rooted macrophytes, and tilapia numbers then fell until these were established
(Bowmaker et al. 1978). Perhaps these recently found S. macrochir came
from upstream, for S. andersonii also appeared for the first time (in 1971)
when conditions had been altered by the new lake and there was cover for
the fish.
In Lake Kariha, Donnelly (1969) found that juveniles less than 10 cm TL
of both T. rendalli and S. mortimeri lived in water less than 30 cm deep,
the 'primary nurseries' on gently sloping shorelines. Fish of 11 to 19 cm
TL (age one year) tend to leave these for adult habitats, though they are
not sexually mature till at least a year later. S. mortimeri over 20 cm TL
live in a variety of habitats and in water up to 15 m deep. T. rendalli is
confined to macrophyte beds, though adults return to shallows to breed, or
during a rise in lake level to feed on decomposing grasses. These Kariba
tilapias breed throughout the year, but with spawning peaks from October to
sean pound
(Sean Pound)
#1