The climatology of droughts over parts of Ethiopia and their impacts on crop production with special reference to the impact of drought on the production of Barley and Maize
Abstract
A predominantly subsistence economy like that of Ethiopia is very sensitive
to drought hazards. Whenever drought occurs, the damage it inflicts on the economy
is often disastrous particularly to the subsistence agricultural sector of the economy.
A wide spectrum of adverse impacts (economic, social and political) are experienced,
especially, when the droughts in question are severe, continuous, frequent and
widespread. Even during the rainy season, crops may encounter drought (moisture
stress) if the soil moisture supply falls below the crop moisture requirement. The
subtle impacts of such imperceptible droughts are detected when crop yields are suboptimal.
The thesis contains an analysis of the spatial and temporal behaviour of
droughts over parts of Ethiopia and their impacts on the production of annual crops,
particularly the ISO-day variety of barley, the l80-day, and 125-day varieties of
maize.
The method of Principal Component Analysis (PCA), was used to divide the
study area into nine homogenous rainfall regions and this formed the basis of the
selection of 30 representative rainfall stations for detailed analysis of drought
characteristics.
In order to investigate the temporal characteristics of the seasonal rainfall, the
rainfall data (at the 30 stations) for the period 1957-87 was subjected to harmonic
analysis. The phase angles used to detect possible year to year shifts in the duration,
onset and withdrawal dates of the rainy seasons did not indicate any significant
changes in the seasonal rainfall characteristics over the study area. Instead, a year to year fluctuation in the amplitude (magnitude) of rainfall of the major
harmonics, was observed at many locations. The high fluctuation in the amplitudes
of rainfall, from year to year, indicated a high degree of below normal and above
normal seasonal rainfall occurrences.
A randomization test applied to the standardized rainfall anomalies indicated
that the positive and negative rainfall anomalies were random (chance) occurrences
except at Gore, Nekempte, and Jijiga where the test indicated non-randomness of the
rainfall departure from the mean. An examination of the frequency of positive and
negative anomalies and their runs indicated that dry areas « 600 nun per annum)
experience more frequent and successive negative departures from the mean, than wet
areas (> 600 nun per annum). The difference between the wet areas and the dry
areas, in the frequency of negative departures was, however, found to be very small.
The dependable amount of rainfall (expected three out of four years), for
Spring (March to May), Summer (June to August), Autumn ( September to
November) and for the year, was obtained using the ranking order method. The
amount of dependable rainfall generally decreased towards the low lying areas of the
south, southeast, east and northeast. The magnitude of the dependable rainfall was
- higher and covered larger areas in Summer (Kiremt) than in Spring and Autumn. The
magnitude and the spatial spread of dependable rainfall decreased from Summer to
Spring and to Autumn in that order.
The Rainfall Anomaly Index, used to describe severity levels of meteorological
drought indicated that most of the drought occurrences were mostly mild and
moderate. Severe and very severe droughts were also experienced at isolated
localities scattered over the study area. The probability occurrence of severe and
very severe droughts ranged between 3% and 6% except at Koka where the
probability occurrence was 10%. Results from the frequency analysis of seasonal
droughts indicated that seasonal droughts were more frequent and severe over the
drier areas than over wetter areas. The southeastern and eastern parts of the study
area which normally get rains in Spring and Autumn encountered lower drought
frequencies in those seasons than in Summer. Hardly any drought occurred, except
mild ones, in Summer over the western and highland parts of the study area.
The impacts of drought on the cultivation of annual crops were examined (with
special reference to barley and maize which are the two most important crops in the
study area) by using different methods. Hargreave's Moisture Availability Index
(MAl), which is the ratio of dependable rainfall to the potential evapotranspiration
(PET), indicated that those areas characterized by more frequent meteorological
droughts had very deficient and moderately deficient moisture conditions. The
observed MAl values for these locations were below 0.67 during the growing periods
for annual crops. On the other hand, those areas of relatively lower frequencies of
meteorological droughts were characterized by adequate or excessive moisture during
the growing seasons with MAl values above 1.0. A comparison of dependable
rainfall with the maximum crop evapotranspiration (ETm) of barley and maize, on
a monthly basis, showed that dependable rainfall equalled or exceeded the maximum
evapotranspiration of the two crops in most of the months during the growing season
over many parts of the study area.
The estimated relative evapotranspiration loss for the two crops (barley and
maize) was found to be above 57% for all the stations for the growing seasons. The
growing seasons at all of the stations were, therefore, classified as suitable or very
suitable in as far as the moisture supply during the growing season was concerned.
Some months within the growing season were, however, found marginally suitable
at a few locations only. In general, moisture stress (drought) during the growing
season was not significant enough to limit the growth of barley and maize in the study
area.
The length of the growing seasons and the mean temperatures during the
growing period were, however, found to be the main limiting factors for precluding
the growth of the two crops at many locations. The cold highlands in Arsi, Bale and
northern Shewa preclude the growth of maize due to low temperatures during the
growing period. In most of the lowlands (in the west, southwest, south, southeast
and northeast) the growing periods were too short for maize of the I80-day variety
to grow. Unlike the I80-day variety of maize, the medium (12S-day) variety of
maize can grow in southern and western lowlands of adequate length of the growing
period. The growth of barley of 150-day variety was limited to the interior highlands
due to high growing season temperatures and short growing periods in the
surrounding areas. In short, the area where the late maize variety could be grown
was limited to the interior and relatively wetter areas at intermediate altitudes. Thegrowing
periods in the western and southwestern lowlands were long enough to allow
the growth of medium (125-day) maize variety. Most of the lowlands were found to
be suitable for early maize varieties.
A comparison of the relative evapotranspiration loss with the relative yield
ratio suggested that in some regions up to 79% of the yield of barley was affected by
non-climatic factors. As much as 66% of the yield of maize was also found to have
been affected by non-climatic factors in some regions.
From the various analyses, rainfall and soil moisture regimes were delineated.
The results indicated that there was no moisture problem for the cultivation of barley
and maize in areas of 1500 m and above. In mountainous areas where the growing
season temperatures are below the operative range of maize, unfavourable
temperatures preclude the cultivation of maize; whereas the lowland areas where the
high growing season temperatures added to the short growing season preclude the
cultivation of barley.
It may be concluded that the study region is, by and large, characterized by
favourable moisture and thermal conditions during the growing period, and provides
a favourable ground for a successful agricultural activity so long as the appropriate
crops are grown. However, as indicated earlier, non-climatic factors seem to affect
crop yield, substantially, in the study area.
It was, finally, recommended that attention be focused on the line of
improving the efficiency of the agricultural systems by adopting better land
management practices such as the provision of subsidised commercial fertilizers,
improved farm implements, seed selection, cross-breeding, careful use of weed killers
and insecticides and selection of appropriate sowing time. The creation of favourable
environment which allows -the peasants to concentrate on the improvement of the
agricultural production even with the means already available to them is believed, by
the author, very crucial for the Ethiopian peasants.
Citation
Ph.D Thesis 1994Sponsorhip
University of NairobiPublisher
Depatment of Geography, University of Nairobi
Description
Ph.D Thesis