STUDIES ON GROWTH CURVE PARAMETERS OF SIROHI GOAT UNDER FIELD CONDITION
Loading...
Date
2021
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
RAJASTHAN UNIVERSITY OF VETERINARY AND ANIMAL SCIENCES, BIKANER (Rajasthan)
Abstract
The present study was assumed to estimate growth curve
parameters. The detailed information of all the animals regarding
growth trait of Sirohi goat was composed over the period from year
2009 to 2017 of the database, maintained at All India Co-ordinated
Research Project (AICRP) on goat improvement, Livestock Research
Station, Vallabhnagar, Udaipur district of Rajasthan.
The average value of body weight for both sexes at birth to 360
Days as 2.54±0.01, 5.60±0.03, 8.52±0.06, 11.81±0.09, 13.13±0.10,
14.44±0.10, 15.80±0.12, 17.50±0.15, 18.00±0.16, 18.50±0.16,
20.50±0.20, 21.30±0.20 and 22.10±0.20 Kg, respectively. There were
32.33 per cent male and 67.77 per cent female used for present
investigation of data from at birth to 12th months of age.
The coefficient of variation (CV) for both sexes was observed for
body weights at birth to 360 Days of age were ranged from 19.99 to
32.50 per cent. The range of CV was observed for body weight for high
range (19.72-36.59 per cent) for male kids and (18.11-29.76 per cent)
for female kids, this may be due to different environmental factor
causes variation in body weight of individual. Descriptive statistics of
live body weights of Sirohi goat ranged from minimum 1.10 Kg to
maximum 60.00 Kg for male kids.
170
The four growth curve models (Brody, Gompertz, Logistic and
Bertalanffy) were fitted to the body weight data of 1015 kids of Sirohi
goat form birth to 360 days of age. The best non-linear regression
growth curve model was determined by considering the following
goodness of fit statistics viz., R2, Radj.
2 , MAE, MAPE, MSE, RMSE, AICC
and chi-square (X2).
The Brody non-linear growth curve model gave values for
parameter A (26.66±1.15, 24.81±1.16 and 25.38±1.14Kg), B
(0.88±0.01, 0.89±0.01 and 0.89±0.01) and K (0.14±0.01, 0.14±0.01
and 0.14±0.01) for male, female and both sexes, respectively. The
goodness of fit statistics was revealed as R2 (99.46, 99.35 and 99.35
per cent), ����.
2 (99.35, 99.22 and 99.26 per cent), MAE (0.449, 0.506
and 0.501), MAPE (0.040, 0.042 and 0.040), MSE (0.26, 0.27 and
0.27), RMSE (0.147, 0.168 and 0.168), AICC (-7.89, -7.13 and -7.41)
and Chi-square (0.300, 0.376 and 0.360) for male, female and both
sexes, respectively.
The Gompertz non-linear growth curve model provided values
for parameter A (23.48±0.91, 21.90±0.91 and 22.39±0.91Kg), B
(1.79±0.12, 1.80±0.12 and 1.80±0.12) and K (0.28±0.03, 0.28±0.03
and 0.29±0.03) for male, female and both sexes, respectively. The
goodness of fit statistics was revealed as R2 (98.62, 98.39 and 98.47
per cent), ����.
2 (98.34, 98.06 and 98.16 per cent), MAE (0.614, 0.622
and 0.621), MAPE (0.067, 0.075 and 0.071), MSE (0.66, 0.68 and
0.68), RMSE (0.205, 0.208 and 0.200), AICC (4.35, 4.60 and 4.60) and
Chi-square (0.765, 0.852 and 0.740) for male, female and both sexes,
respectively.
The Logistic non-linear growth curve model gave values for
parameter A (22.49±0.88, 20.99±0.88 and 21.46±0.87Kg), B
(3.84±0.58, 3.87±0.63 and 3.89±0.62) and K (0.41±0.05, 0.41±0.05
and 0.42±0.05) for male, female and both sexes, respectively. The
goodness of fit statistics was revealed as R2 (97.58, 97.28 and 97.37
171
per cent), ����.
2 (97.09, 96.73 and 96.84 per cent), MAE (0.762, 0.716
and 0.799), MAPE (0.097, 0.103 and 0.103), MSE (1.17, 1.15 and
1.16), RMSE (0.265, 0.264 and 0.261), AICC (11.63, 11.47 and 11.62)
and Chi-square (1.389, 1.420 and 1.387) for male, female and both
sexes, respectively.
The Bertalanffy non-linear growth curve model provided values
for parameter A (24.41±0.93, 22.47±0.94 and 22.97±0.93Kg), B
(0.46±0.02, 0.47±0.02 and 0.47±0.02) and K (0.24±0.02, 0.24±0.02
and 0.24±0.02) for male, female and both sexes, respectively. The
goodness of fit statistics was revealed as R2 (98.95, 98.76 and 98.82
per cent), ����.
2 (98.74, 98.51 and 98.58 per cent), MAE (0.544, 0.540
and 0.536), MAPE (0.059, 0.062 and 0.058), MSE (0.50, 0.52 and
0.52), RMSE (0.175, 0.178 and 0.177), AICC (0.80, 1.25 and 1.17) and
Chi-square (0.548, 0.585 and 0.553) for male, female and both sexes,
respectively.
Four non-linear growth curve models viz., Brody, Gompertz,
Logistic and Bertalanffy fitted to body weights data of male, female and
both sexes of Sirohi goat. All the four models were used for estimate
growth curve parameters (A, B and K). The growth curve models were
compared by using goodness of fit statistics viz., R2, Radj.
2 , MAE, MAPE,
MSE, RMSE, AICC and X2 (Chi-Square) values to identify the best
growth curve model in explaining the body weights of male, female and
both sexes of Sirohi goat.
The best growth curve model was defined which had properties
viz. highest values for growth curve parameters (A, B, and K), highest
R2 value, highest ����.
2 value , lowest MAE value, lowest MAPE value,
lowest MSE value, lowest RMSE value and lowest AICC value. Thus,
Brody growth curve model was found to be best model for body weight
of male, female and both sexes of Sirohi goat due to the highest R2
value, highest ����.
2 value, lowest MAPE value, lowest MSE value,
lowest RMSE value and lowest AICC value.
172
The least-squares analysis of variance observed that the
random effect of sire was highly significant (P≤0.01) effect on
parameters (A) and (B) however, significant (P≤0.05) was revealed on
parameter (K).
The least-squares analysis of variance observed that effect of
year of birth had highly significant (P≤0.01) effect on A, B and K
parameters.
The least-squares analysis of variance revealed that effect of
season of birth was non-significant effect on parameters (A) and (B) of
growth curve however, significant (P≤0.05) effect was on parameter
(K).
The least-squares analysis of variance revealed that effect of
type of birth was significant (P≤0.05) effect on parameter (A) and highly
significant (P≤0.01) effect was on parameter (B). In present study nonsignificant
effect of type of birth was revealed on parameter (K).
The least-squares analysis of variance observed that effect of
parity was non-significant effect on parameters (A), (B) and (K) of
growth curve model.
The least-squares analysis of variance revealed that effect of
sex of kid was highly significant (P≤0.01) on parameter (A) while the
effect of sex of kid was revealed non-significant on parameter (B) and
(K) of growth curve.
The least-squares analysis of variance observed that highly
significant (P≤0.01) effect of cluster was on parameter (A) and (B) of
growth curve whereas effect of cluster was revealed non-significant
effect on parameter (K).
The least-squares analysis of variance observed that effect
regression of dam’s weight at kidding was non-significant for
parameters (A), (B) and (K) of growth curve model.
173
The heritability estimates of parameters (A), (B) and (C) in the
present study was estimated as 0.40±0.05, 0.13±0.03 and 0.02±0.01,
respectively.
Estimates of genetic correlations between parameters of growth
curve of Sirohi goat in present investigation ranged from -0.67±0.33
between parameter (B) and (K) to 1.00±0.04 between parameter (A)
and (B). The genetic correlations between parameter (A) and
parameter (B) were estimated high value as 1.00±0.04. The genetic
correlations between parameters (A) and parameter (K) were valued
highly negative value as -0.96±0.31. The genetic correlation between
parameters (B) and parameter (K) were estimated moderately negative
value as -0.67±0.33. The negative correlation between parameter (B,
related to initial weight) and parameter (K) indicate that selection of
higher maturity rate could result lower birth weight.
Estimates of phenotypic correlations between parameters of
growth curve of Sirohi goat in present investigation ranged from -
0.14±0.00 between parameter (B) and parameter (K) to 0.45±0.00
between parameter (A) and parameter (B).
The phenotypic correlations between parameter (A) and
parameter (B) were estimated moderately positive value as 0.45±0.00.
The phenotypic correlations between parameters (A) and parameter
(K) were estimated moderately negative value as -0.43 ±0.00. The
negative phenotypic correlation between parameters (A) and (K)
indicates that selection was used for increase asymptotic body weight
could leads to decrease maturity and growth rate. The phenotypic
correlation between parameters (B) and parameter (K) were estimated
negative value as -0.14±0.00.
Description
STUDIES ON GROWTH CURVE PARAMETERS OF
SIROHI GOAT UNDER FIELD CONDITION