Monitoring and Forecasting of Pearl Millet (Pennisetum glaucum L.) Yield in Two Areas of Traditional Rain-fed Agriculture in Western Sudan

Abstract

A study was conducted to test the validity of an FAO method
for monitoring and forecasting pearl millet yield in two traditional rain-
fed areas in western Sudan; namely, Al-Obeid and Al-Fashir. The periods
of crop development were delineated from the graphs of mean normal (30
years) rainfall (Pn) and potential evapotranspiration (PET) versus 10-day
interval (dekad) during the growing season. Annual water satisfaction
indices (SI) were estimated for 10 years from annual water balance for
the period 1987-1996. This index indicates the extent to which the water
requirement of the crop was satisfied in a cumulative way. Highly
significant cubic relationships between Pn or PET and seasonal dekads
were obtained for both areas. For Al-Obeid, Pn for three dekads were
equal to 0.5 PET and for the remaining dekads they were below 0.5 PET,
while for Al-Fashir Pn was below 0.5 PET for all dekads. Thus, according
to the FAO classification system, the two areas have no growing seasons.
Since this conclusion contradicts with acctual practice, a modification of
the system is proposed. According to the modified system, there is a
growing season but it is still very dry in both areas. Trendline equations
for the distribution of Pn and PET were developed for the two areas. An
example for the calculation of SI for millet (dukhn) grown in 1987 in the
two areas was presented and analyzed. The corelation between millet
yield and seasonal rainfall was not significant in Al-Obeid area, but it was
highly significant (r2 =0.947) and cubic in Al-Fashir area. The correlations
between yield and SI were significant and cubic in Al-Obeid (r2 =0.447)