Abstract:
Nationwide survey was carried out with the objectives to characterize and describe native
Ethiopian equines populations (horse and donkey) at both phenotypic and molecular levels.
Standard breed identification procedures were employed to identify distinct populations in
both species. Accordingly, we identified six distinct domestic donkey populations namely
Abyssinian, Afar, Hararghe, Ogaden, Omo and Sinnar but rejected a previously misidentified
Jimma donkey as it failed to meet standard breed identification criteria. Twelve
morphometeric variables were measured on a total o f569 adult donkeys (289 jacks and 280
jennets) that belonged to the six identified populations. The result showed that Sinnar donkeys
were significantly taller (P<0.05) (109.78 cm) for height at wither (HW) while Abyssinian
donkeys were shorter (93.87 cm) than all the remaining donkey populations. Principal
component analysis (PCA) showed that four of the morphometeric variables including HW,
height at back (HB), height at rump (HR) and body length (BOL) accounted for about 87.5%
of the total variations among the donkey populations. Percent assignment of discriminant
function analysis showed that, on avera g e 64.83% of individuals w ere reclassified back to
their respective populations. Based on squared Mahalanobis distances, the six indigenous
Ethiopian donkey populations were clustered into four major morphological clusters viz.
l)Abyssinian; 2) Sinnar; 3) Omo and 4) Afar, Hararghe and Ogaden. Using similar
procedure, seven distinct domestic and one feral horse populations were identified in Ethiopia. These include Abyssinian, Bale, Bor an, Horro, Ogaden, Kafa, Selale and Kundido
feral horse. Seventeen morphometeric variables were measured on a total of 503 native
Ethiopian horses that belonged to five of the eight identified horse populations. Comparison of
Least squares means showed that Selale horses were taller (P<0.05) (131.2 cm) and longer
bodied (133.6 cm) than the rest o f native Ethiopian horse populations. PCA showed that HW,
HB, HR, BOL, back length (BAL) and barrel length (BEL) jointly account for about 80% per
cent of the total variations o f the 17 measured morphometeric variables. Percent discriminant
function analysis showed that 77.05% of individuals were correctly classified into their
respective populations. Cluster analysis based on squared Mahalanobis distances clustered
the five Ethiopian horse populations into at most three and at least two major morphological
groups viz. 1) Selale and Kafa; 2) Bale; and 3) Abyssinian and Horro.
Molecular characterization on the basis o f maternally inherited mitochondrial DNA (mtDNA)
D-loop sequence was carried out in 39 randomly selected donkeys sequenced for 378 base
pairs (bp). The 39 sequences were characterized by 29 polymorphic sites that further
collapsed into 19 different haplotypes. Average haplotype diversity was high (0.903±0.032)
ranging from 0.400 in Ogaden to 0.933 in Omo donkeys. Average nucleotide diversity was
also high (0.020±0.003) ranging from 0.002 in Ogaden to 0.029 in Sinnar donkeys showing
the least and highest heterozygous population, respectively. Reynolds ’pair-wise genetic (Dp)
distances between Ethiopian donkey populations showed that Sinnar donkeys were
consistently distant from the rest of Ethiopian domestic donkey populations suggesting it might
be descended from a different ancestral trunk than the progenitor/s that gave rise to the rest of
Ethiopian donkey populations. Hence, based on mtDNA data, Ethiopian domestic donkeys are
clustered into two major genetic groups: 1) Sinnar and 2) Abyssinian, Afar, Hararghe,
Ogaden and Omo. On the other hand, genetic relationship studies based on microsatellite loci
variation clusters Ethiopian donkeys into highland and lowland population, which is
concordant with the traditional donkey classification systems in Ethiopia. Moreover,
Reynolds' genetic distance estimate from mtDNA sequence polymorphism and population
differentiation estimates based on Fst estimates obtained from microsatellite marker loci both
support that donkeys descended from two ancestral lineages co-exist in Ethiopia. Phylogenetic
analysis using the newly sequenced sequences and previously published worldwide sequences
retrieved from the GenBank showed that Ethiopian donkey haplotypes formed the center of the network suggesting that Ethiopian haplotypes are the most likely founding lineage for most of
the worldwide domestic donkeys. We also investigated the genetic diversity and relationship
among eight native Ethiopian horse populations using 46 randomly selected mtDNA D-loop
sequences sequenced for 454 bp. The 46 mtDNA D-loop sequences were characterized by only
15 variable sites that further shrunk into five distinct haplotypes. The average haplotype
diversity was 0.706 ranging from 0.333 in Boran to 0.800 in Abyssinian, Ogaden and Selale
horses. The average nucleotide diversity was 0.014 ranging from 0.0026 in Kundido feral
horses to 0.0167 in Abyssinian horses. The average number of nucleotide differences was
6.320 ranged from 1.200 in Kundido feral to 7.600 in Abyssinian domestic horses. Only Boran
domestic horses and Kundido feral horses were genetically distant from the rest o f native
Ethiopian horse populations. This indicates that native Ethiopian horses have a narrow
genetic background. Based on mtDNA diversity, native Ethiopian horses are clustered into
three major genetic groups 1) Kundido feral horses; 2) Boran and Bale 3) Abyssinian, Horro,
Kafa, Ogaden and Selale horse populations. Phylogenetic analysis carried out using network
on the combined Eihio-Eurasian mtDNA D-loop sequences showed that haplotypes derived
from native Ethiopian horses formed separate clusters indicating continental sequence
grouping in domestic horses. This suggests that native Ethiopian horses may trace their
matrilineal genetic origin back to other native horses that might had been independently
domesticated in Africa or emerged from the interbreeding between native African and
introduced Eurasian type horses. Thus, the genetic origin of native Ethiopian horses remained
enigmatic. In any case however, African domestication of domestic horses cannot completely ruled out. This warrants further multidisciplinary studies.