Abstract:
Agricultural production in the Sahelian region of Africa is
among the lowest in the world. Cyclical infestations of locusts and
grasshoppers have contributed to this low productivity. The infestations
appear to be triggered by heavy rainfall at the beginning of a crop season
after extended drought. Because locusts and grasshoppers are
migratory, their control on a local and farm by farm basis is difficult.
To combat the problem of periodic outbreaks USAID has
provided direct assistance through aerial spraying of grasshoppers in a
number of African countries.' This includes a control program in the
north African country of Chad in 1987. From the beginning concern
arose as to when it is economically desirable to control grasshoppers.
The International Plant Protection Center (IPPC) at Oregon State
University was awarded a contract for economic evaluation of spraying in
Chad which included development of a grasshopper crop loss simulation
model (GHLSIM) driven by climatic factors for grasshopper and crop
growth. GHLSIM's posterior analysis of the 1987 anti-grasshopper
campaign produced results with a wide range of B/C ratios across the 13
sites which were evaluated. The B/C ratios ranged from 0 to 6.9.
To test GHLSIM's use beyond weather conditions of a single
year (1987), a Monte Carlo simulation technique was used to generate
long term (100 year) rainfall patterns at selected sites and a series of
expected benefits generated. The effect of rainfall variability upon
economic benefits was assessed. So also were changes in grasshopper
population densities and output prices. Results are presented
graphically under a number of such conditions to understand the biophysical
relationships among grasshoppers, crop, natural vegetation and
rainfall as modeled by GHLSIM.
The model is sensitive to within season and to some extent
between season weather variation. Rainfall appears to be the dominant
variable controlling crop production, regardless of grasshopper densities.
When weather is the overriding factor, changes in grasshopper density
levels do not play a dominant role in affecting B/C ratios. When weather
is not a constraint for crop production, grasshopper density changes can
affect benefits significantly.
In its present form the GHLSIM model cannot be used as an
early warning indicator. While the model appears to predict crop
production quite well, model refinement which will change the qualitative
prediction of grasshopper populations to a quantitative one throughout a
crop season is still needed. When this is completed and both the
GHLCROPS and GHLOSE submodels performance are validated against
field observations, the GHLSIM model can be used as an early warning
predictor. With these accomplishments GHLSIM can then be considered
for use in the public policy arena for assessing what areas to spray,
when and what distributional effects may be involved.
