The major Objective Of this study is to present an agroecologically ba
sed system of forecasting yield potentials and cropping risks for defi
nded crops in the semiarid drylands of SE-Kenya by means of atmosphere
-ocean-teleconnection (El Nino Southern Oscillation Phenomenon = ENSO)
. Three distinct rainfall conditions during the short rains, i.e. ENSO
, Anti-ENSO and Normal were identified from calculations with a rainfa
ll simulation model. Under ENSO rainfall conditions, rainfall was foun
d to be above normal while the reverse was found for the Anti-ENSO con
ditions. These results from the rainfall simulation model were then us
ed to estimate yield potentials under the three rainfall conditions by
means of an agroecological simulation model (MARCROP) that was calibr
ated on the results of laboratory (University of Trier, Germany) and f
ield (off- as well as on-farm) experiments at and near KARI/NRRC Kibok
o, Kenya. From this analysis, it was found that farmers should already
be informed as early as in August (before the onset of the short rain
s) about the expected rainfall conditions so that they are able to dec
ide on the selection of crops and crop varieties for cultivation early
enough. Drought resistant. short-cycle Tepary beans (Phaseolus acutif
olius) turned out to be the best suited for Anti-ENSO and Normal condi
tions when the cropping risk for high yielding but less adapted bean a
nd maize varieties is too high (e.g.: 1 out of 2 seasons is failing fo
r Mwezi moja beans - Phaseolus vulgaris, GLP 1004). However. in ENSO y
ears soil moisture conditions are sufficient to bring up Mwezi moja be
ans and maize (Katumani Composite B).