Résumé
Crosses of Asian rice,
Oryza sativa, with African rice,
O. glaberrima, aim at reducing the tradeoffs between yield potential, weed competitiveness and drought resistance. Physiological studies were initiated to develop a crop model that simulates the performance of plant-types combining the traits of the two species. The present study characterizes the control of stomatal conductance and leaf rolling under drought in one
O. glaberrima cultivar (CG14), one
O. sativa tropical-japonica cultivar (WAB56-104), one inter-specific descendant of the two cultivars, and the
O. sativa indica semi-dwarf-type Bouaké 189. An upland field experiment (Exp. I) with two sowing dates was conducted during the 1997 wet season under naturally occurring drought at Mbé, Côte d'Ivoire. Patterns of growth and yield were observed, and five cycles of diurnal (0800–1600
h) measurements of leaf water potential (LWP), stomatal conductance (
g
s) and leaf rolling score (LRS) were conducted in drought situations. In a second experiment at the same site during the 1997 dry season, irrigation was suspended for 35 days, beginning 15 days after sowing. During and after the drought period, LWP,
g
s, LRS and soil water tension of the topsoil were observed. Soil-matric potential (SMP) was determined from soil water content using a pF curve. Values of
g
s had no stable relationship with LWP across different times of day but were roughly constant during the day and depended on the LWP observed at 0800
h. LWP measured in the morning was similar to SMP as determined from pF curves, but soil tensiometers gave lower absolute values. Leaf rolling was closely related to LWP regardless of the time of day. The cultivars did not differ in the
g
s versus soil water tension or
g
s versus LWP
Morning relationship, but differed significantly in the LRS versus LWP relationship. It is concluded that
g
s was controlled by a soil-moisture-dependent root signal, whereas LRS was controlled directly by leaf water status. Simple models to simulate these relationships in drought-sensitive crop models are discussed.