MEASURES OF SIZE AND SHAPE USING PRINCIPLE COMPONENT
ANALYSIS IN WEST AFRICAN DWARF AND RED SOKOTO GOATS
REARED IN SELECTED LOCAL GOVERNMENT AREAS OF ANAMBRA
STATE NIGERIA.
BY
OJUKWU CHUKWUEBUBE MERCY
2018814024
DEPARTMENT OF ANIMAL SCIENCE AND TECHNOLOGY
FACULTY OF AGRICULTURE
NNAMDI AZIKIWE UNIVERSITY
AWKA.
FEBRUARY 2024.
ABSTRACT
This study was conducted in Ayamelum, Awka North and Orumba North Local Government
areas as sample areas in Anambra state. The objectives of this study were to estimate size and
shape of West African Dwarf (WAD) and Red Sokoto (RS) goats using Principal Component
Analysis. The populations of goats studied were two hundred and thirty comprising of one
hundred and forty nine West African Dwarf and eighty one Sokoto Red goats. The parameters
measured were body length (BL), hearth girth (HG), wither height (WH), hind leg length (HLL),
fore leg length (FLL), hip height (HH), paunch girth (PG), canon bone circumference (CBC) and
body weight (BW), ear length (EL). Statistical analysis used were stepwise multiple regression,
correlation coefficient, coefficient of determination (R), adjusted coefficient of determination
(adj) and standard error. In the male WAD, the average mean parameters for all the traits were
larger than that of the females with the exception of the CBC (6.73±0.92). In the male RS, all the
mean average parameters were larger than that of the females with the exceptions of BL
(57.58±2.78), HH (62.90 ±2.22), RH (16.35±1.01), PG (73.57±1.86) and BW (30.15±1.99).
Phenotypic correlation for the male WAD was highest between BW and HG (r=0.97) and lowest
between BW and HLL (r=0.54), female WAD was highest between BW and HG (r=0.95) and
lowest between BW and CBC (r=0.52). For the male RS, the phenotypic correlation was highest
between BW and BL (r=0.96) and lowest between BW and HH (r=0.53). The female RS
however, was highest between BW and BL (r=0.98) and lowest between BW and HH (r=0.68).
The Kaiser Meyers Olkin measures of sampling adequacy was recorded as 0.73, 0.80, 0.86 and
0.58 for the male WAD, female WAD, male RS and female RS respectively. Percentage variance
for male WAD for PC1 and PC2 accounted for 80.67% of body weight and the rest of the
percentage variance accounted for PC1 alone whereas the female accounted for 63.11%. Male
RS was at 74.48% and female RS at 86.04% for percentage variance. Principal component score
coefficient for BW estimation was highly negative for PC1 and PC2 in male WAD while PC1 for
female WAD, male RS and female RS were all positive. In male RS, stepwise multiple
regression of BW accounted for 98.5% when BL, HG, HLL and PG were included in the model.
For female RS, variation in body weight was highest when BL, HG, RH, WH and HLL was
included in the model (r=100%). In male WAD variation for BW was highest when HG, BL and
WH were included in the model (r=99.6%). Female WAD had the highest BW variation when
HG, BL, HLL, RH and FLL were included in the model(r=99.1%). The observed and predicted
weight of the animals didn’t vary much. The interaction of BW and morphological body traits
may be useful in selection criteria for genetic influences. The use of multivariate analysis using
PC is useful in the prediction of BW using orthogonal body traits. PC1 obtained from PCA of ten
LBPs contained only positive coefficients. Results from the present investigation will assist
farmers in determining the BW of their animals and will also assist farmers in improving specific
traits in goats.
