PhD-Thesis
Tree-crop competition is an important concern in agroforestry systems, and has been
the subject of much attention by researchers in recent years. Farmers on the other
hand, often seem to either ignore or accept competition and do little to manage it.
In this thesis, I make the first attempt to reconcile farmers’ practices and beliefs with
experimental data, focusing on the economically important tree: Melia volkensii in
drylands of Kenya. This tree is of particular interest because farmers’ perceptions of
its competivity appear at odds with researchers perceptions: farmers believe it does
not compete with crops while researchers have found it competes.
After conducting a survey on farmers’ fields in which it was confirmed Melia
volkensii does not compete with crops, an experimental programme was conducted to
investigate the causes of this discrepancy in perceptions. The study covered; Kilui,
Embu, Kibwezi and Machakos sites. Melia volkensii provenances from these sites and
Melia plant types used in tree establishment (seedlings and cuttings) were evaluated to
examine whether the reported yield differences were as result of differences in root
architecture between provenances or between plant types used by farmers.
The rainfall confounded the problem of tree-crop interactions; because extreme
rainfall conditions were experienced during the study period. Excessive soil moisture
occurred in the first season while rainfall failed in the second season. Competition
was therefore evaluated when competition was occurring. Results highlights are as
follows; competition index (CI) method was unable to predict tree competitiveness
with crops in conditions where trees are planted closely. The CI values were highly
variable within individuals of a particular Melia provenance or species and therefore
found inconsistent and unreliable to predict tree competition with crops. Surprisingly,
it was able to predict tree-crop competition where trees occurred in isolation. There
were neither differences in root architecture between the Melia volkensii provenances
nor were there difference in soil water or crop yields between provenances’ plots.
However, when root architecture of seedlings and cuttings were compared, cuttings
tended to develop shallow roots than seedling plants and that roots which descended
into the soil at deeper angles developed some secondary root branches which grew to
the surface, thus making them competitive with crops.
Overall, the results support the view that Melia volkensii is competitive with crops in
normal seasons. Isolated Melia trees on farmers’ fields reduced crop yields beyond 15
m radius from trees. However, the value of the timber outweighs the value of the lost
crop and farmers may chose to have Melia trees on the cropland whose value will
compensate the lost yields and still benefit from other tree products e.g. fuelwood and
fodder, or plant trees in separate plots e.g. fallow lands and maximize productivity of
both components. Melia timber fetched >45% more cash than timber products of
other species such as cypress and pines. Results suggest that when promoting the
planting of Melia, seedlings should be used instead of cuttings as they have a less
competitive root architecture, as seed germination is a problem this should be a
priority research area for Kenya seed specialists.