For all varieties and both crops (without and with
initial water stress) the highest root yield was
attained at 12 months (Table 2). Nevertheless, a
more dominant influence on root yield was that of
water stress and this was evident at all harvest times
and for all varieties. Root yield of cassava from the
crop with initial water stress was significantly lower
than that of the crop without initial water stress
(PB0.05). Comparisons between the varieties revealed
differences in drought tolerance, CMR 33-
57-81 having the highest root yield at all harvest
times regardless of growth condition. Subjecting
the plants to water stress conditions early in their
development also significantly influenced starch
content (PB0.05, Table 2). Under such conditions
the roots contained a significantly lower starch
content (only 1.2–3.5% at 6 months, Table 2)
compared to those grown without initial water
stress (20.4–25.9% at 6 months, Table 2). This is
in agreement with previous findings showing that
when plants are subjected to water limiting conditions
early in their establishment, root development
is poor. The roots show signs of retarded development
and tend to be immature (Pardales and
Esquibel, 1996). Changing growth conditions, such
that the plants receive sufficient water, will lead to
dramatic gains in root starch content, as the plant
presumably adjusts physiologically to the changing
conditions. The plants form new leaves rapidly
after being released from the initial stress. Depriving
plants of water during the late development
stages does not lead to the severe reduction in root
starch content seen in plants subjected to initial
water stress. Starch content decreases during the
period of rain following the late drought period.
This response typifies both crops. Following removal
of the stress condition plants of the second
crop synthesized increasing amounts of starch, the
content being maximal during the late water stress
at 10 months (except R5) and remaining constant
thereafter. Similarly, for the crop without initial
water stress starch content was maximal at 8
months, coinciding with the late drought period.
For the remainder of the trial starch content
decreased continuously, the reduction being triggered
by the onset of the rainy season.