PhD. Thesis
The Small East African (SEA) goat breed in Tanzania is comprised of a number of strains
which vary considerably. The strains are identified and named mainly based on a few
morphological characteristics, ethnic communities keeping them or geographic location
where they are kept. The SEA goats are generally characterized by low production and
reproduction capacity, consequently several crossbreeding programmes have been
implemented in order to improve their performance. Many exotic goat breeds, especially
dairy goats have been introduced in the country since early 1960s and used to upgrade the
SEA goats for the purpose of increasing milk production. There is fear that unplanned
crossbreeding can lead to loss of adaptive characteristics and genetic purity of the strains
of SEA goats. Unfortunately, the SEA strains have not been adequately characterized,
hence, specific genetic information for each of the strains is lacking.
The overall goal of this study was to determine phenotypic and genetic distinctiveness of
four strains of the SEA goats in Tanzania. To achieve the overall goal, four specific
studies were conducted. Study I focused on phenotypic characterization of four strains of
SEA goats and involved measurements of quantitative and description of qualitative
characteristics. Study II and III involved analyses of DNA samples from the four strains to
assess genetic structure, diversity and maternal origin using eight microsatellite markers
and sequences of the hyper-variable region I (HV1) of the mitochondrial DNA (mtDNA)
Displacement loop (D-loop), respectively. Study IV involved sequencing of Myostatin
(MSTN) gene to determine variations in the gene responsible for growth among the four
strains. The goats used in this study belonged to Sukuma, Pare, Gogo and Sonjo strains of
the SEA. The animals were sampled from different regions in Tanzania, including
Mwanza region, northern Tanzania (Sukuma strain), Kilimanjaro region, north-eastern
Tanzania (Pare strain), Dodoma region, central Tanzania (Gogo strain), and Arusha region
(Sonjo strain). Boer goats were included in studies II, III and IV as a reference breed and
were sampled from Ngerengere Livestock Multiplication Unit (LMU) located in
Morogoro region, eastern Tanzania. In study IV, DNA samples and growth data were
obtained from Pare, Sonjo and Blended goats which are kept at West Kilimanjaro
Livestock Research Centre, and from Boer goats kept at Ngerengere LMU.
Results on body measurements in study I indicated that Pare goats were the heaviest with
mean body weight of 29.8 ± 0.5 kg, followed by Gogo goats (25.1± 0.6 kg), Sonjo goats
(22.8 ± 0.48 kg) and Sukuma goats which were the lightest with the mean weight of 22.3 ±
0.5 kg. The results on qualitative traits showed that Pare (39.6%) and Gogo (40.2%) goats
had predominantly white coat colour while Sonjo were red in colour (85.4%) and the
Sukuma goats had mixed black and white colour (66.7%). Almost all goats from the four
strains of SEA were horned and did not have wattles and the majority of them had
medium sized and horizontally oriented ears. The squared mahalanobis distance based on
morphological traits was largest between Pare and Sukuma goats (5.45) and least between
Pare and Gogo goats (0.94). Cluster analysis based on morphological characteristics
revealed two separate groups; the Pare goats in one cluster and another cluster comprised
of the Gogo, Sonjo and Sukuma goats. Based on morphological characteristics, 75.3% of
Sonjo, 70.5% of Sukuma, and 67.9% of Pare goats were correctly assigned to their source
populations while 51.1% of the Gogo goats were mis-assigned to other populations.
In study II within population genetic diversity was determined using observed
heterozygosity, expected heterozygosity and mean number of alleles per locus while
genetic diversity between populations was assessed using DA genetic distance, genetic
differentiation coefficient, Neighbour-Joining phylogenetic tree and principal componentiv
analyses. The observed heterozygosity (H O ) ranged from 0.583 ± 0.04 (lowest) in Sukuma
to 0.659 ± 0.030 (highest) in Gogo while expected heterozygosity (H E ) ranged from 0.632
± 0.16 in Sukuma goats to 0.716 ± 0.16 in Boer goats. Five microsatellite markers
deviated from Hardy–Weinberg equilibrium (HWE) across the populations. The mean
number of alleles per locus ranged from 4.75 ± 1.58 in Pare strain to 6.88 ± 3.00 in
Sukuma goats. Among the four SEA goat strains, differentiation coefficient (F ST ) was
highest (0.078) between Sukuma and Gogo goats and lowest (0.008) between Gogo and
Sonjo goats. The largest DA genetic distance (0.451) was found between Sukuma and
Gogo goats while the smallest (0.031) was between Gogo and Sonjo strains. The
Neighbour-Joining phylogenetic tree of breed relationships revealed three clusters, the
first cluster comprised of Pare, Gogo and Sonjo while the second and third clusters were
made of Sukuma and Boer goats, respectively. However, in the principal component
analysis only Pare and Sonjo grouped together, the rest of the breeds (Gogo, Sukuma and
Boer goats) formed their own independent clusters.
In study III, the maternal origin, demographic history and genetic variability of the SEA
goats were studied using mitochondrial DNA (mtDNA) analyses.
A total of 154
polymorphic sites and 102 haplotypes were revealed in the HV1 region of the mtDNA D -
loop and among these, 94 were unique and eight were shared among different goat
populations. The number of haplotypes ranged from 12 to 27 in Boer and Sonjo goats,
respectively. Haplotype diversity ranged from 0.873 ± 0.039 in Boer to 0.994 ± 0.013 in
Gogo goats. The Goats were divided into A, B1 and G maternal haplogroups; haplogroup
A being the most predominant. The largest number of nucleotide diversity was found in
Pare population (0.030 ± 0.007) while the lowest was observed in Gogo and Sonjo (0.023
± 0.003). The mtDNA sequences of Tanzanian goats in this study were compared with
those of the goats from other African countries including Kenya, South Africa, Namibia, Zimbabwe, Mozambique, Algeria, Egypt and Nigeria. The median joining (MJ) network showed that the Tanzanian and Kenyan SEA goats shared haplotypes with those of the Southern African countries. No sharing of haplotypes was detected between SEA goats and goats from West or North Africa. Demographic expansion was detected in all SEA goat strains whereas background selection occurred only in Pare goats.
In study IV the polymorphisms of intron 2 and exon 3 of MSTN gene and its association with growth performance in two SEA strains and Blended goats was assessed. No variation was detected in the sequenced part of the MSTN gene in all strains. A singleton polymorphic site was detected in Boer goats which resulted into two genotypes TT and TC with frequencies of 0.97 and 0.03, respectively. Concerning growth performance, it was revealed that Blended goats were heavier at all stages of growth compared to Pare and Sonjo goats. However, due to lack of polymorphism of the MSTN gene in the three goat populations, association between alleles of the MSTN gene and growth performance was not studied.
Based on the results of the four studies it is concluded that;
i. There is high phenotypic variation among the SEA goats. Pare goats have the
highest values for body dimension whereas Sonjo and Sukuma goats have the
lowest.
ii. Genetically, the SEA goats are highly diverse but admixed and, therefore, none of the strains could distinctively be regarded as a separate breed.
iii. The SEA goats have multiple maternal origins.
iv. There is variation in growth performance among the Blended, Pare and Sonjo
goats but the variation is not associated with the intron 2 and exon 3 of of the
MSTN gene.