School of Agriculture and Natural Resourceshttp://ir.mu.ac.ke:8080/jspui/handle/123456789/212024-03-29T12:03:00Z2024-03-29T12:03:00ZImpact of adopting multiple agricultural technologies on nutrition outcome in east africa: a multinomial endogenous switching regression approachk. levy, Kachileihttp://ir.mu.ac.ke:8080/jspui/handle/123456789/71022022-11-21T09:44:46Z2022-01-01T00:00:00ZImpact of adopting multiple agricultural technologies on nutrition outcome in east africa: a multinomial endogenous switching regression approach
k. levy, Kachilei
Majority of the population of East African households are malnourished and much of the
effort to address the problem of stunting, wasting and underweight have focused on
interventions that are designed directly to address its immediate causes. It is expected that the
adoption of multiple agricultural technologies such as improved beans varieties, bio-fortified
maize variety, grafted fruit trees, and garden vegetable techniques can be a means by which
malnourished rural households who may have less access to diverse meals, supplements, and
fortified foods can enhance their balanced diet but malnutrition still remain a salient problem
facing rural households in East Africa. This study determined the factors that affect the
adoption of joint multi-agricultural technologies then analyze the impact of the best four
combinations adopted in East Africa countries that is; improved beans variety, biofortified
maize variety, grafted fruit trees, and use of garden vegetables techniques on the household
nutrition outcome indicators of underweight (WAZ), wasting(WHZ) and stunting(HAZ).
Where TC = base with no technology used, TC1 = Improved beans variety, biofortified maize
variety, and grafted fruit trees, TC2 = Improved beans variety, biofortified maize variety, and
garden vegetable techniques, TC3 = biofortified maize variety, garden vegetable techniques,
and grafted fruit trees, TC4 = Improved beans variety, garden vegetable techniques, and
grafted fruit trees. The study utilized a secondary household panel data of Kenya, Tanzania,
and Uganda that was collected by IFRI for ten waves from 2007 to 2017 and each country
with 500 households. This study utilized multinomial endogenous regression model so as to
casual the impact of technology adoption and to correct for the self-selection bias. It was
conceptualized that the decision to adopt a combination of multiple agricultural technologies
(MATs) is modeled in consumer theory, specifically, a random utility framework. The latent
model (U*jit) which describes the ith farmer’s behavior in adopting MATs j(j=1,...4) at time t
over any alternative MATs combination was utilized in three stages. In the first stage, the
analysis determined the factors for adopting multi-agricultural technologies using a
multinomial endogenous switching regression. In the second stage, the inverse mills ration
generated in stage one is used as linkage between adoption of technologies nutrition outcome,
and on the third stage, the treatment effect was used to establish the relationship between
adopters of the joint multiple agricultural technologies and non-adopters. The results show
that year increase of the education of household head, general participation in community
meetings and barazas increases the adoption of TC1 (45%), TC2 (44%), TC3 (25%), and TC4
(35%) respectively. The 1% percent increase in the adoption of joint technologies, the
prevalence of stunting reduces by 17.4%, wasting 15.4%, and underweight by 16.8%. Results
of the average treatment effects show that the households who adopted joint multiple
agricultural technologies had a positive significant impact (HAZ β= .62, p<0.01), WAZ (β =
.72, p<0.01), and WHZ (β = .74, p<0.01) which improves the nutrition status by HAZ
(103%), WAZ (87%), and WHZ (84%). The best technology combination was TC3 which
impacted all nutrition outcome at the highest percentage HAZ (25.8%), WHZ (24.2%), and
WAZ (25.3%). Kenya(reference) had a higher significant propensity of adoption hence higher
impact on nutrition outcome than Uganda (β = -.128, p<0.01) and Tanzania (β = -.155,
p<0.01). This study concludes that adoption of multiple agricultural technologies improves
household nutrition outcome. The household that adopted the joint multiple agricultural
technologies had systematically higher nutrition outcome than the households who did not
adopt even after controlling for all confounding factors. Among the three countries Kenya has
a higher significant propensity on nutrition outcome. This study offers insight to
policymakers, researchers, and extension workers regarding the advancement of factors
suitable for joint technology combination to be adopted by the East Africa households.
Consequently, this study recommends that household should focus on adopting the multiple
agricultural technologies to improve their nutrition status. And more so focus more on the
combination of TC3 (Biofortified maize variety, garden vegetable techniques, and grafted
fruit trees) since it was the combination with greatest impact on nutrition outcome.
2022-01-01T00:00:00ZBreeding objectives and breeding strategies for small ruminants in the tropicsKosgey, Isaac Sangahttp://ir.mu.ac.ke:8080/jspui/handle/123456789/37082020-12-08T09:37:57Z2004-01-01T00:00:00ZBreeding objectives and breeding strategies for small ruminants in the tropics
Kosgey, Isaac Sanga
Small ruminants (i.e., sheep and goats) are widespread in the tropics and are
important to the subsistence, economic and social livelihoods of a large human
population in these areas. The aim of this thesis was to identify the breeding
objectives for tropical small ruminants, and to develop appropriate breeding
strategies for their improvement. The results indicated that breed substitution and
crossbreeding programmes involving temperate breeds are rarely successful due to
incompatibility of the genotypes with the farmers’ breeding objectives and the
production systems. Within-breed selection programmes utilizing indigenous breeds
are likely to be more sustainable than breed substitution and crossbreeding. In
addition,
they
help
to
maintain
biodiversity.
Indigenous
genotypes
were
predominantly found among pastoral/extensive farmers and mixed crosses among
smallholders. In general farmers perceived crosses less favourably than indigenous
breeds for a range of traits. The effect was studied of including intangible benefits in
the calculation of economic values of breeding goal traits. It resulted in increased
values of traits related to longevity. Litter size and lambing frequency were more
important traits in smallholder and pastoral production. 12-month live weight also
featured prominently in pastoral production. Constraints to small ruminant
productivity included low levels of management, disease and parasite challenge,
inadequate feed and poor marketing. Nucleus breeding schemes are recommended
to optimize the limited available resources. However, ‘interactive cycling screening’
schemes would be more practical under village settings as the farmers are actively
involved in genetic improvement, and minimal recording is required in the commercial
flocks. A single nucleus could serve both the smallholder and pastoral production. In
conclusion, it is prudent to examine the production system holistically, and involve
the producer at every stage in the planning and operation of a breeding programme,
integrating traditional knowledge, practices, behaviour and values.
2004-01-01T00:00:00Za numerical simulation of vascular brain tumor growth using adomian decomposition method.WANJAU PAUL MAINAhttp://ir.mu.ac.ke:8080/jspui/handle/123456789/8562018-03-08T09:53:42Z2017-01-11T00:00:00Za numerical simulation of vascular brain tumor growth using adomian decomposition method.
WANJAU PAUL MAINA
A tumor develops when a single normal cell transforms due to mutations in certain
key genes. To continue growing, it requires new sources of nutrients, hence develops
new blood vessels that continue feeding it from the blood leading to vascularization.
Statistics from World Health Organization (WHO) records shows that incidences of
brain tumors in the year 2014 were already at 1/12,500 persons. The purpose of this
study was to develop a numerical simulation of vascular brain tumor that will help
medical practitioners to predict the size of the tumor for prognosis purposes instead of
exposing patients to radiations through multiple scanning. In this work numerical
simulation was developed from partial differential equations models, whereby cell
nutrients concentration(C) was the dependent variable x, y, z were spatial independent
variables, t was a variable for different time schedules, P was the variable for cells
proliferation, P n was the variable for non- proliferating cells while N was the variable
for necrotic cells. Objectives of study were, to develop a numerical simulation of
vascular brain tumor growth in one, two and three dimensions, to determine the
viable rate of consumption of the nutrients in tumor growth and development, to
present validated results in tabular and graphical form, to determine the period within
which angiogenic inhibitors are viable. In attaining the objectives above results were
generated by Adomian Decomposition Method (ADM) whereby equations are
decomposed into a series of Adomian polynomials. The method generates a solution
in the form of a series whose terms are determined by a recursive relationship.
Results obtained from the simulation of growth and dynamics of malignant brain
tumor (glioma) compares well with those from medical literature. In one dimensional
model, radius of the tumor in different time schedules was obtained, for example
where the rate of diffusion of the nutrients was 11mm/year, in 560 days, simulation
radius was found to be 25.4mm compared to an experimental radius of 25.0 mm. In
two dimensional models, cross section area of the tumor in different time schedules
was obtained, whereby in 560 days, simulation area was found to be 19.02cm 2 ,
whereas analytical area was 19.64cm 2 . In three dimensional models, volume of the
tumor in different time schedules was obtained, whereby in 560 days, simulation
volume was found to be 65.77cm 3 , whereas analytical volume was 65.48cm 3 . Thus
obtained results were found to be consistent with available experimental data, hence
may be used to complement traditional tumor diagnostic. Considering idealized cases
of tumors, ADM gave realistic simulations, which can provide clinical practitioners
with valuable information on the potential effects of therapies in their exact
schedules. However for tumors with multiple distinct clones, current model may not
be reliable thus further studies are needed to address this shortcoming.
2017-01-11T00:00:00Z