| Livestock Research for Rural Development 31 (10) 2019 | LRRD Misssion | Guide for preparation of papers | LRRD Newsletter | Citation of this paper |
Doe reproduction and kid survival performances of Boer and their crosses with Central Highland goats were evaluated in Ethiopia. Survival data on a total of 519 kids born from an on-station Boer x Central Highland goats cross-breeding program recorded within the period 2012 – 2017 were analyzed using Weibull proportional hazard models of the survival kit version 6.0. Reproduction data were analyzed using the General Linear Model and logistic regression procedures of SAS 9.0. The overall percentages of conception, kidding and abortion of does were 48.9%, 37.6% and 10% respectively. Breed group and non-genetic factors influence the studied traits, indicating Boer does had the lowest conception and kidding rate. The overall least-squares means for litter size at birth and at weaning, litter weight (kg) at birth and at weaning were 1.4±0.03, 0.67±0.03, 3.63±0.07 and 10.68±0.28 respectively. The litter size at birth and at weaning were not affected by the doe breed group while the litter weight at birth and at weaning was highest for the Boer does. The overall kid survival up to day 4, 90, 180 and 365 were 73.99%, 53.57%, 47.98% and 40.27% respectively. Kid survival was affected by type of birth, season of birth and year of birth, indicating single born kids had higher survival rates than multiple born kids throughout the studied ages. Kid breed, kid sex and doe parity did not affect survivability at all ages. The attempt which was aimed genetic improvement through crossbreeding with exotic Boer goats in Ethiopia was blocked by their poor reproduction and survival. Therefore, it is advisable to try to bring genetic improvement through within breed selection among the indigenous goat breeds (CHG) in terms of reducing importation cost, conservation and adaptation.
Key words: conception, cross-breeding, risk ratio, weibull
Small ruminants, especially goats, contribute significantly to the economy and food security of developing countries. They are better adapted to harsh environments, thereby help in utilisation of marginal, degraded, sloppy and other lands which are unsuitable for any meaningful agrarian activities (Misra and Singh 2002; Degen 2007). Therefore, for a sustainable utilization, designing a better genetic and phenotypic improvement strategies are the main concern and activities of the farmers and researchers.
Central Highland goats, predominantly found in the central and northern part of Ethiopia, are one of the 12 indigenous goat breeds of Ethiopia which are categorized under the Small East African family (Gizaw 2009). Similar to the other indigenous goat breeds, different genetic improvement programs have been applied to this breed; among them is crossbreeding with Boer goats aiming rapid improvement in growth performances. However, sub-optimal growth performances were achieved from the crossbreds which was below expectation in comparison to the native area of the breed (Mustefa et al 2019). This may be due to different reasons; the exotic breeds need to be well adapted to the climate of the region where they are expected to perform (Mestawet et al 2014). Technically the process of crossbreeding needs to be initiated only after understanding the nicking and complementariness of both the breeds involved in the process of cross breeding to different productive, reproductive and adaptation parameters (Cunningham and Syrstad 1987).
Similarly, introduction of exotic goat breeds (for cross-breeding purposes) have been carried out in the past by both researchers and non-government agencies alike for the purpose of genetic improvement. However, desirable results were not observable and the footfalls of the schemes have since been lost in oblivion, primarily because of lack of follow up and improper records of the achievements (Ayalew et al 2003; Mestawet et al 2014). According to Horst (1983), the concept of ‘productive adaptability’, phenotypic performance is the result of an animal’s true genetic performance and the animals’ ability to cope with environmental stresses, including the burden of common diseases/parasites and heat load. It was also stated that, introduction of the exotic breeds need to be carefully assessed for the adaptive traits prior to initiation of any crossbreeding experiments (Moses 2006).
It has been reported that the Boer goats are less susceptible to heat stress and tolerate heat better than many of the other temperate goat breeds (Lu 1989). They are also moderately tolerant to drought and tannins, and are efficient fiber digesters and they can adapt to various ambient temperature and have lower water turnover rate (Erasmus 2000). Because of their browsing habit, Boer goats are considered less susceptible to infestation by internal parasites (Casey and Van Niekerk 1988). Boer goats are thought to have exceptional ability to withstand and resist diseases such as blue tongue, prussic acid poisoning, and enterotoxaemia (http://studbook.co.za/boergoat/value.html). Similarly, it was reported that the Boer goats can improve reproductive and productive performance of many indigenous breeds where the crossbreds have outperformed the native breeds at many aspects (Casey and Van Niekerk 1988). The traits which showed significant improvement among the crossbreds include prolificacy and fecundity besides carcass quality (Cameron et al 2001). Boer doe also reach puberty at an early age (Greyling 2000). Hence, this study was done to evaluate the kid survival and doe reproduction performances of Boer and their crosses with Central Highland Goats in Ethiopia.
Data were obtained from an on-station Boer x Central Highland goats cross-breeding program carried out at Ataye Research site, Debre Birhan Agricultural Research Center, Ethiopia. The goat flock was a mix of different breed groups including Boer (B), 50% F1 kids (B x CHG), 50% F2 kids (50% F1 x 50% F1) and 75% (B x 50% F1). The site is located 10 o 35’ N latitude, 39o 93’E longitude and 1491 m above sea level altitude. The area is characterized by unimodal rainfall pattern, receives annual rainfall of about 969.3 mm with wet season (March, April, July, August, September and October) and dry season (January, February, May, June, November and December). The average monthly temperature ranges from minimum of 12.6 °C to maximum of 30.1°C. (Girma and Desta, 2007).
A total of 519 kids born from 376 kidding from the cross-breeding program recorded within the period 2012 – 2017 were used for the analysis of survivability from birth to yearling age and reproduction performances. Kid survival from birth to fourth day, to weaning, to six-months and to yearling ages, and reproduction traits including conception, kidding and abortion rates, litter size and weight at birth and at weaning were evaluated under some non-genetic (dam parity, sex, year, season and type of birth) factors in Ethiopia.
Reproduction (time and weight at mating, abortion, stillborn, and kidding) of the does and survival data (birth and mortality date) of the kids were recorded every day. Kids alive up to the recording time were considered as censored, whereas kids dying during the experimental time were considered uncensored. Kids type of birth was categorised into those born as single or multiple. Similarly, doe parity was categorized as parity 1, 2, 3 and 4. Parities 4 and above were merged together. Mating period is the period of time from the date of joining the bucks and does till the end of the mating (removal of the bucks from the does). Mating weight is weight of the does just before joining the bucks.
Based on the data the following reproduction traits were calculated as
lock was managed semi-intensively with grazing and supplement. The supplement includes ad libitum grass hay, chopped pasture (Napier grass, Desmodium spp. and vetch) and concentrate supplement based on their body weight. The health of all the flocks was regularly followed by a veterinarian. Providing strategic deworming, regular case to case follow-up and spraying barns and animals with acaricides to make them free from external parasites were some of the practices exercised. Selected bucks were used for the breeding purpose, and at the end of the mating bucks are removed from the flock of does. The mating was so ensured to minimise inbreeding among the flocks with average buck to doe ratio of 1:20/25.
The data on reproductive performances (litter size and litter weight at birth and at weaning) were analyzed using general linear model (GLM) procedure and means were compared using Adjusted Tukey-Kramer test of SAS 9.0 software. Some of the reproductive traits (conception, kidding and abortion rate) were analyzed using logistic regression of SAS program (2002). Analysis of data on kid survival (survival to fourth day, to weaning, to six months and to yearling) was done using Weibull proportional hazard model of survival kit version 6.0 (Ducrocq et al 2010).
All the fixed effects and covariates were fitted for the analysis of logistic regression, however, the forward selection of steps dropped some of the independent variables which have no effect or association on or with the dependent variables.
The model used for the analysis of conception
Logit (Y) = ln (P/1 – P) = α + β1X1 + β2X2 + β3X3 + β4X 4 + β5X5
Where Y is conception, P is probability of getting conceived, α is the Y intercept, βs are regression coefficients of Xs, X1 is the fixed effect of doe breed group, X2 is the fixed effect of mating year, X3 is the fixed effect of mating season, X 4 is the random effect of mating period and X5 is the random effect of mating weight.
The model used for the analysis of kidding
Logit (Y) = ln (P/1 – P) = α + β1X1 + β 2X2
Where Y is kidding, P is probability of getting kidded, α is the Y intercept, βs are regression coefficients of Xs, X1 is the fixed effect of doe breed group and X2 is the random effect of mating weight.
The model used for the analysis abortion rates was
Logit (Y) = ln (P/1 – P) = α + β1X1 + β 2X2 + β3X3
Where Y is abortion, P is probability of getting aborted, α is the Y intercept, βs are regression coefficients of Xs, X1 is the fixed effect of mating year, X2 is the random effect of mating period and X3 is the random effect of mating weight.
The model used for the analysis of litter size and weight at birth and at weaning was
Yijklm = μ + Ai + Bj + Ck +Dl + Em + e ijklm
Where Y ijklm is an observation, μ is the overall mean, A i is the fixed effect of the ith doe breed group, Bj is the fixed effect of the jth type of birth, Ck is the fixed effect of the kth doe parity, D l is the fixed effect of the lth year of kidding, Em is the fixed effect of the mth season of kidding and eijklm is the random error attributed to the nth kid.
The statistical model for the analysis of kid survival was
λ (t) = λ0 (t) exp (gni + sexj + tb k +prl + yrm + snn + errijklmn)
Where λ (t) is the hazard function (probability of a kid being died) at time t, λ0 (t) is the baseline hazard function which is assumed to follow the Weibull distribution, gni is the fixed effect of kid breed group, sexj is the fixed effect of kid sex, tbk is the fixed effect of type of birth, pr l is the fixed effect of doe parity, yrm is the fixed effect of year of birth, snn is the fixed effect of season of birth and errijklmn is the random error attributed to nth kid.
Due to the non-significant effects of the interactions among the above factors, it was removed from the analysis and results.
Conception, kidding and abortion rates under different categories are presented in Table 1. The overall conception and kidding were 48.85 % and 37.56 % while the abortion was10%.
|
Table 1. Number and percentages of does which conceive, kidded and aborted under the different categories |
||||||||
|
Independent |
Total |
Conception |
Kidding |
Abortion |
||||
|
N |
% |
N |
% |
N |
% |
|||
|
Overall |
1001 |
489 |
48.85 |
376 |
37.56 |
100 |
9.99 |
|
|
Breed |
||||||||
|
Boer |
433 |
163 |
37.64 |
95 |
21.94 |
65 |
15.01 |
|
|
CHG |
480 |
261 |
54.38 |
228 |
47.50 |
27 |
5.62 |
|
|
50%F1 |
88 |
65 |
73.86 |
53 |
60.23 |
8 |
9.09 |
|
|
Mating Year |
||||||||
|
2012 |
323 |
106 |
32.82 |
57 |
17.65 |
47 |
14.55 |
|
|
2013 |
253 |
140 |
55.34 |
101 |
39.92 |
35 |
13.83 |
|
|
2014 |
154 |
85 |
55.19 |
79 |
51.30 |
6 |
3.90 |
|
|
2015 |
162 |
89 |
54.94 |
83 |
51.23 |
3 |
1.85 |
|
|
2016 |
109 |
69 |
63.30 |
56 |
51.38 |
9 |
8.26 |
|
|
Mating Season |
||||||||
|
Wet |
597 |
299 |
50.08 |
242 |
40.54 |
49 |
8.21 |
|
|
Dry |
404 |
190 |
47.03 |
134 |
33.17 |
51 |
12.62 |
|
|
N=number of does |
||||||||
The overall model evaluation and type 3 analysis of effects on the studied traits are presented in Table 2 and 3. The models used were significantly useful in predicting the effects which has strong regression with the traits.
|
Table 2. Overall model evaluation on testing the global null hypothesis BETA=0 |
|||||||||
|
Test |
Conception |
Kidding |
Abortion |
||||||
|
χ2 |
df |
p |
χ2 |
df |
p |
χ2 |
df |
p |
|
|
Likelihood Ratio |
172.64 |
9 |
<.0001 |
138.29 |
5 |
<.0001 |
61.59 |
6 |
<.0001 |
|
Score |
157.98 |
9 |
<.0001 |
132.15 |
5 |
<.0001 |
59.42 |
6 |
<.0001 |
|
Wald |
134.06 |
9 |
<.0001 |
111.14 |
5 |
<.0001 |
49.42 |
6 |
<.0001 |
|
χ2=chi-square, df= degrees of freedom, p= Pr > ChiSq |
|||||||||
Conception rate were affected by all the studied fixed effects and covariates while kidding and abortion rate were affected partially. Doe breed group and mating weight affected kidding rate significantly while abortion rate was affected by mating year, period and weight.
|
Table 3. Logistic regression: Type 3 analysis of effects of on conception, kidding and abortion rates |
|||||||||
|
Effects |
Conception |
Kidding |
Abortion |
||||||
|
df |
Wald χ2 |
p |
df |
Wald χ2 |
p |
df |
Wald χ2 |
p |
|
|
Doe Breed |
2 |
48.13 |
<.0001 |
2 |
130.16 |
<.0001 |
- |
- |
- |
|
Mating year |
4 |
15.57 |
0.0037 |
- |
- |
- |
4 |
22.12 |
0.0002 |
|
Mating season |
1 |
14.59 |
0.0001 |
- |
- |
- |
- |
- |
- |
|
Mating period |
1 |
15.43 |
<.0001 |
- |
- |
- |
1 |
22.46 |
<.0001 |
|
Mating weight |
1 |
64.36 |
<.0001 |
1 |
68.16 |
<.0001 |
1 |
4.95 |
0.0261 |
|
χ2=chi-square, df= degrees of freedom, p= Pr > ChiSq |
|||||||||
The comparisons on conception rate (Table 4) among the breed groups were done with 50% F1 does. The odds ratio estimate for conception show 50% F1 does conception was the highest among the others while Boer does conceive lower than the indigenous CHG breed group does. Does mated in 2015 show higher conception rate which 10.26 times those of 2016. Similarly, does mated at dry season conceive better than those of the wet season. The proportion and odds ratio of does getting conceived increases as the mating period and mating weight advances.
|
Table 4. Analysis of maximum likelihood and odds ratio estimates predicting probabilities of conception |
|||||||||
|
Parameter |
Maximum likelihood estimates |
Odds ratio estimates |
|||||||
|
df |
α, β |
SE |
Wald χ2 |
p |
Parameter |
point |
95% Wald |
||
|
Intercept |
1 |
-3.04 |
0.47 |
41.37 |
<.0001 |
||||
|
Doe Breed |
|||||||||
|
Boer |
1 |
-1.86 |
0.28 |
44.65 |
<.0001 |
vs 50% F1 |
0.01 |
0.003 |
0.040 |
|
CHG |
1 |
-0.80 |
0.32 |
6.18 |
<.0001 |
vs 50% F1 |
0.03 |
0.008 |
0.13 |
|
Mating year |
|||||||||
|
2012 |
1 |
-0.56 |
0.29 |
3.80 |
0.0511 |
vs 2016 |
2.53 |
0.80 |
7.99 |
|
2013 |
1 |
0.41 |
0.18 |
5.30 |
0.0213 |
vs 2016 |
6.71 |
2.05 |
21.98 |
|
2014 |
1 |
0.80 |
0.26 |
9.53 |
0.0020 |
vs 2016 |
9.93 |
2.63 |
37.49 |
|
2015 |
1 |
0.84 |
0.28 |
8.74 |
0.0031 |
vs 2016 |
10.26 |
2.61 |
40.28 |
|
Mating season |
|||||||||
|
Dry |
1 |
-0.41 |
0.11 |
14.59 |
0.0001 |
vs Wet |
0.45 |
0.29 |
0.67 |
|
Mating period |
1 |
0.01 |
0.003 |
15.43 |
<.0001 |
1.01 |
1.01 |
1.02 |
|
|
Mating weight |
1 |
0.12 |
0.015 |
64.36 |
<.0001 |
1.13 |
1.09 |
1.06 |
|
|
χ2=chi-square, df= degrees of freedom, p= Pr > ChiSq, α is the regression coefficients of the intercept, βs are regression coefficients of Xs, SE= standard error of the estimate |
|||||||||
Similar to the conception rate, kidding rate (Table 5) was highest for the 50% F1s when compared to Boer and CHG while Boer goats still kidded at a lower rate than the CHG. The kidding rate also progressed as the mating weight increased.
|
Table 5. Analysis of maximum likelihood and odds ratio estimates predicting probabilities of kidding |
|||||||||
|
Parameter |
Maximum likelihood estimates |
Odds ratio estimates |
|||||||
|
df |
α, β |
SE |
Wald χ2 |
p |
Parameter |
point |
95% Wald |
||
|
Intercept |
1 |
-3.17 |
0.36 |
77.65 |
<.0001 |
||||
|
Doe Breed |
|||||||||
|
Boer |
1 |
-1.76 |
0.16 |
125.35 |
<.0001 |
vs 50% F1 |
0.06 |
0.03 |
0.10 |
|
CHG |
1 |
0.64 |
0.12 |
29.29 |
<.0001 |
vs 50% F1 |
0.62 |
0.39 |
1.00 |
|
Mating weight |
1 |
0.11 |
0.01 |
68.16 |
<.0001 |
1.11 |
1.09 |
1.14 |
|
|
χ2=chi-square, df= degrees of freedom, p= Pr > ChiSq, α is the regression coefficients of the intercept, βs are regression coefficients of Xs, SE= standard error of the estimate |
|||||||||
The kidding rate comparison results (Table 6) show higher abortion on does mated in 2013. On the other hand, as the mating weight increases abortion rate also increases slightly.
|
Table 6. Analysis of maximum likelihood and odds ratio estimates predicting probabilities of abortion |
|||||||||
|
Parameter |
Maximum likelihood estimates |
Odds ratio estimates |
|||||||
|
df |
α, β |
SE |
Wald χ2 |
p |
Parameter |
point |
95% Wald |
||
|
Intercept |
1 |
-4.82 |
0.56 |
74.26 |
<.0001 |
||||
|
Mating year |
|||||||||
|
2012 |
1 |
0.44 |
0.23 |
3.51 |
0.0611 |
vs 2016 |
0.99 |
0.44 |
2.24 |
|
2013 |
1 |
0.99 |
0.23 |
18.46 |
<.0001 |
vs 2016 |
1.72 |
0.78 |
3.81 |
|
2014 |
1 |
-0.82 |
0.38 |
4.78 |
0.0287 |
vs 2016 |
0.28 |
0.10 |
0.84 |
|
2015 |
1 |
-1.04 |
0.49 |
4.62 |
0.0316 |
vs 2016 |
0.23 |
0.06 |
0.86 |
|
Mating period |
1 |
0.02 |
0.004 |
22.45 |
<.0001 |
1.02 |
1.01 |
1.02 |
|
|
Mating weight |
1 |
0.03 |
0.01 |
4.95 |
0.0261 |
1.03 |
1.00 |
1.06 |
|
|
χ2=chi-square, df= degrees of freedom, p= Pr > ChiSq, α is the regression coefficients of the intercept, βs are regression coefficients of Xs, SE= standard error of the estimate |
|||||||||
Litter size and weight at birth and at weaning
The results of litter size and weight by doe breed groups, type of birth, parity, year and season of kidding are presented in Table 7. They show that the litter weight at birth and weaning varied across breed groups. Boer does produce higher litter weight at birth and a weaning than the others. Litter size at weaning and litter weight at birth was higher (p<0.001) among the does which gave multiple births, however no differences were recorded in their litter weight at weaning stage. Year of kidding influenced (p< 0.001) the litter size at weaning with no clear trend. The results also show that litter weight at birth and weaning were higher ( p<0.01 and p<0.05) during the wet season.
|
Table 7. Least square means and pairwise comparison of litter size and weight with standard errors in each category |
||||||||
|
Source of |
N |
Litter size |
Litter weight |
|||||
|
At birth |
At weaning |
At birth |
At weaning |
|||||
|
Overall mean |
376 |
1.40±0.03 |
0.67±0.03 |
3.63±0.07 |
10.68±0.28 |
|||
|
CV (%) |
3.69 |
91.97 |
25.43 |
34.88 |
||||
|
Doe Breed |
p= 0.9352 |
p= 0.4384 |
p <0.0001 |
p= 0.0058 |
||||
|
Boer |
95 |
1.50±0.01 |
0.60±0.07 |
4.28±0.10a |
12.46±0.61a |
|||
|
CHG |
228 |
1.50±0.00 |
0.69±0.05 |
3.62±0.07b |
10.36±0.37b |
|||
|
50% F1 |
53 |
1.50±0.01 |
0.72±0.13 |
3.68±0.19b |
10.06±0.95ab |
|||
|
Kidding Type |
p= 0.0259 |
p <0.0001 |
p= 0.1523 |
|||||
|
Single |
228 |
0.59±0.06 |
3.19±0.09 |
10.56±0.46 |
||||
|
Multiple |
153 |
0.75±0.07 |
4.53±0.11 |
11.36±0.56 |
||||
|
Doe Parity |
p= 0.7362 |
p= 0.6597 |
p= 0.1173 |
p= 0.9605 |
||||
|
1 |
156 |
1.50±0.01 |
0.68±0.06 |
3.67±0.09 |
11.06±0.49 |
|||
|
2 |
102 |
1.50±0.01 |
0.64±0.07 |
3.74±0.11 |
11.06±0.60 |
|||
|
3 |
70 |
1.50±0.01 |
0.61±0.10 |
3.92±0.15 |
11.08±0.76 |
|||
|
4 |
53 |
1.50±0.01 |
0.75±0.11 |
4.12±0.17 |
10.63±0.87 |
|||
|
Kidding year |
p= 0.9250 |
p <0.0001 |
p <0.0001 |
p <0.0001 |
||||
|
2012 |
73 |
1.50±0.01 |
0.61±0.12ab |
3.57±0.18b |
9.01±0.98c |
|||
|
2013 |
24 |
1.50±0.01 |
0.85±0.15a |
3.39±0.22b |
9.14±1.10bc |
|||
|
2014 |
55 |
1.50±0.01 |
0.68±0.10ab |
3.53±0.15b |
7.96±0.81c |
|||
|
2015 |
93 |
1.49±0.01 |
0.92±0.08a |
4.33±0.12a |
12.32±0.66ab |
|||
|
2016 |
72 |
1.50±0.01 |
0.66±0.08a |
3.64±0.12b |
12.65±0.62a |
|||
|
2017 |
64 |
1.50±0.01 |
0.30±0.09b |
4.71±0.13a |
14.68±0.93a |
|||
|
Kidding Season |
p= 0.7351 |
p= 0.0994 |
p= 0.0017 |
p= 0.0114 |
||||
|
Wet |
136 |
1.50±0.01 |
0.60±0.07 |
4.05±0.11a |
11.90±0.63a |
|||
|
Dry |
245 |
1.50±0.01 |
0.74±0.06 |
3.67±0.10b |
10.02±0.50b |
|||
|
Means within a column bearing different superscripts are significantly different. N = number of does |
||||||||
Censored and uncensored records of kids with mean, minimum and maximum number of days are presented in Table 8. Censored kids were those that were alive at the end of observation period. Every failure was associated with death due to different diseases. The proportion of censored kids until weaning and yearling was 53.57% and 40.27% respectively. This study revealed that survival at pre-weaning period was lower than the post-weaning period. This study also shows that 26% of the kids’ loss was occurred before reaching their fourth day after birth.
|
Table 8. Number of censored and uncensored (failure) records of kids from birth to yearling age |
||||
|
Category |
No of kids |
Mean |
Minimum |
Maximum |
|
At 4th day |
||||
|
Censored |
384 (73.99%) |
4 |
4 |
4 |
|
Uncensored |
135 |
1.5 |
1 |
3 |
|
At weaning |
||||
|
Censored |
278 (53.57%) |
90 |
34 |
90 |
|
Uncensored |
241 |
10.7 |
1 |
82 |
|
At six month |
||||
|
Censored |
249 (47.98) |
180 |
34 |
180 |
|
Uncensored |
270 |
23.6 |
1 |
172 |
|
At yearling |
||||
|
Censored |
209 (40.27%) |
365 |
34 |
365 |
|
Uncensored |
310 |
51.5 |
1 |
361 |
|
Censored animals are live animals at the censoring time |
||||
The results of pre-weaning relative risk ratio and survival percentages for effect of kid breed group, kid sex, doe parity, type of birth, year of birth and season of birth are presented in Table 9. Type of birth, year of birth and season of birth had effects (p<0.05) on kid survival at fourth day while survival at weaning was affected by type of birth and year of birth only. On the other hand, pre-weaning survival was not affected by kid breed group, doe parity and kid sex. Multiple born kids had higher risk than the single born kids over the pre-weaning period. Years 2016 and 2017 showed highest risk ratio till the fourth day with 2.47 and 2.53 times to the year 2012 respectively, while kids born in year 2015 were lowly at risk than the kids born in the other years over the pre-weaning period.
Even if there were some numerical differences, the results of the chi-square probability of survivability indicates that there were no differences in pre-weaning survivability among the breed groups studied. Similarly, the results (Table 9) showed that the pre-weaning survivability of the kids (irrespective of sexes) were more or less similar. At the early age stage, kids born during the dry season were at higher risk than kids born at wet season.
|
Table 9. Pre-weaning survival percentages and estimates of risk ratio of kids for genetic and non-genetic categories |
|||||
|
Source of |
N |
At 4th day |
At weaning |
||
|
Survival % age |
RR ± SE |
Survival % age |
RR ± SE |
||
|
Kid Breed group |
p= 0.0661 |
p= 0.7553 |
|||
|
Boer |
129 |
82.17 |
1.00 |
55.81 |
1.00 |
|
50% F1 |
321 |
71.03 |
1.85±0.24 |
53.89 |
1.06±0.16 |
|
50% F2 |
44 |
68.18 |
1.88±0.45 |
47.73 |
1.37±0.33 |
|
75% |
25 |
80.00 |
0.88±0.56 |
48.00 |
1.02±0.36 |
|
Kid Sex |
p= 0.6295 |
p= 0.8776 |
|||
|
Male |
256 |
73.05 |
1.01±0.18 |
53.91 |
0.96±0.13 |
|
Female |
263 |
74.90 |
1.00 |
53.23 |
1.00 |
|
Type of Birth |
p< 0.0001 |
p< 0.0001 |
|||
|
Single |
229 |
82.97 |
1.00 |
66.81 |
1.00 |
|
Multiple |
290 |
66.90 |
2.65±0.21 |
43.10 |
2.50±0.15 |
|
Doe Parity |
p= 0.1315 |
p= 0.7678 |
|||
|
1 |
197 |
78.68 |
1.00 |
54.82 |
1.00 |
|
2 |
141 |
74.47 |
1.14±0.28 |
55.32 |
1.03±0.21 |
|
3 |
98 |
66.33 |
1.18±0.35 |
48.98 |
1.10±0.26 |
|
4 |
83 |
71.08 |
0.66±0.38 |
53.01 |
0.77±0.28 |
|
Year of Birth |
p= 0.0008 |
p= 0.0100 |
|||
|
2012 |
95 |
80.00a |
1.00 |
55.79ab |
1.00 |
|
2013 |
29 |
75.86ab |
1.86±0.52 |
62.07ab |
1.06±0.39 |
|
2014 |
69 |
75.36ab |
2.30±0.40 |
47.83b |
1.64±0.28 |
|
2015 |
133 |
84.21a |
0.50±0.37 |
65.41a |
0.60±0.26 |
|
2016 |
105 |
61.90b |
2.47±0.42 |
45.71b |
1.47±0.31 |
|
2017 |
88 |
64.77b |
2.53±0.42 |
44.32b |
2.04±0.31 |
|
Season of Birth |
p= 0.0229 |
p= 0.6328 |
|||
|
Wet |
180 |
80.00 |
1.00± |
55.00 |
1.00 |
|
Dry |
339 |
70.80 |
2.37±0.25 |
52.80 |
1.32±0.17 |
|
Means within a column bearing different superscripts
are significantly different. N = number of kids born.
|
|||||
The results of post-weaning relative risk ratio and survival percentages for effect of kid breed groups, kid sex, doe parity, type of birth, year of birth and season of birth are presented in Table 10. Kid survival to yearling age was affected (p<0.05) by type of birth and year of birth, whereas kid survival to six-months age was only affected by type of birth. On the other hand, post-weaning survival was not affected by kid breed group, doe parity, kid sex and season of birth. Multiple born kids had higher risk of mortality than the single born kids over the post-weaning period. Kids born in year 2015 showed lowest risk ratio 0.58 (SE = 0.22) till yearling age.
|
Table 10. Post-weaning survival percentages and estimates of risk ratio of kids for genetic and non-genetic categories |
||||||
|
Source of |
N |
At six month |
At yearling |
|||
|
Survival % age |
RR ± SE |
Survival % age |
RR ± SE |
|||
|
Kid Breed group |
p= 0.9008 |
p= 0.1570 |
||||
|
Boer |
129 |
46.51 |
1.00 |
32.56 |
1.00 |
|
|
50% F1 |
321 |
49.22 |
0.94±0.15 |
43.93 |
0.82±0.14 |
|
|
50% F2 |
44 |
45.45 |
1.21±0.31 |
38.64 |
1.28±0.29 |
|
|
75% |
25 |
44.00 |
0.97±0.34 |
36.00 |
1.04±0.31 |
|
|
Kid Sex |
p= 0.9749 |
p= 0.3621 |
||||
|
Male |
256 |
48.05 |
0.98±0.12 |
38.28 |
1.06±0.12 |
|
|
Female |
263 |
47.91 |
1.00 |
42.21 |
1.00 |
|
|
Type of Birth |
p< 0.0001 |
p< 0.0001 |
||||
|
Single |
229 |
60.26 |
1.00 |
51.09 |
1.00 |
|
|
Multiple |
290 |
38.28 |
2.29±0.14 |
31.72 |
2.10±0.13 |
|
|
Doe Parity |
p= 0.9324 |
p= 0.6393 |
||||
|
1 |
197 |
47.21 |
1.00 |
38.58 |
1.00 |
|
|
2 |
141 |
49.65 |
0.98±0.19 |
42.55 |
1.07±0.18 |
|
|
3 |
98 |
45.92 |
1.06±0.24 |
36.73 |
1.21±0.22 |
|
|
4 |
83 |
49.40 |
0.78±0.26 |
44.58 |
0.88±0.25 |
|
|
Year of Birth |
p= 0.2906 |
p= 0.0167 |
||||
|
2012 |
95 |
47.37 |
1.00 |
33.68b |
1.00 |
|
|
2013 |
29 |
55.17 |
1.09±0.36 |
44.83ab |
0.97±0.33 |
|
|
2014 |
69 |
40.58 |
1.60±0.26 |
28.99b |
1.40±0.24 |
|
|
2015 |
133 |
55.64 |
0.69±0.24 |
51.13a |
0.58±0.22 |
|
|
2016 |
105 |
44.76 |
1.32±0.29 |
35.24b |
1.13±0.27 |
|
|
2017 |
88 |
|||||
|
Season of Birth |
p= 0.6258 |
p= 0.6402 |
||||
|
Wet |
180 |
49.44 |
1.00 |
38.89 |
1.00 |
|
|
Dry |
339 |
47.20 |
1.30±0.16 |
41.00 |
1.21±0.15 |
|
|
Means within a column bearing different superscripts are significantly different. N = number of kids born. RR= Risk Ratio (Risk of Mortality), SE=Standard Error |
||||||
The Kaplan – Maier survivor curve (survival function and cumulative hazard) of kids from birth to yearling age is presented in Figure 1. The curve was steeper for the first 25 days, thereafter, the curve showed some stability. Similarly, the cumulative hazard goes the opposite direction and reaches its climax when the kids age increased to yearling. This shows there is high probability of losing the kids depending on the whole data trend.
|
| Figure 1. Kaplan – Maier survival function and cumulative hazard curve of kids from birth to yearling age |
Breed group influenced litter weight, conception and kidding rate, which may be ascribed to the adaptation of the breed groups to a particular agro climate (Mellado et al 2006). Conception and kidding percentages were highest among the crossbreds which can be attributable to the heterosis effect (Gosey 1991). Conception and kidding rate were lower among the Boer does which is in line with the findings of Khanal (2016) who reported that Kiko, Spanish and their cross-bred does have higher kidding rate when compared to the purebreds. Overall higher abortion rate observed in the flock can be ascribed to their adaptation (biotic and abiotic stress related factors). The findings are in close accordance with the reports of Unanian and Silva (1989) who reported that incidences of abortions rates were higher in stressed/ exotic breeds than in native.
The results of adaptation of the does to the agro climate can also be evident from the higher conception and lower abortion rates across the years. The observations are in close accordance with the findings of Mellado et al (2006) who recorded that conception was highest among does which had least stress. Mating season affects conception percentage which is in line with the report of Bushara et al (2016) and Mellado et al (2006). Conception was recorded to be slightly higher for the does mated in dry season, which might be due to the comfortableness of the air condition in dry than in the fall for goats. These results are in line with the findings of Mellado et al (2006) who report that conception was higher in the dry than in the fall. Kidding and abortion rates did not depend on mating season differences as they occur in a different season from the season of mating.
As the mating period extends slight increment in conception rate was observed, however, abortion rate was also growing which might be due to the unnecessary butting by the bucks (Browning 2014). As the doe body weight at mating increases there is also an increment in conception and kidding rate which might be due to body reserves of the heavy does. These reports are in close accordance with those of Mellado et al (2004) who reported that the Does with lower weight cannot provide adequate nutrition to the growing embryo leading to higher incidences of abortion. In contrary to this, Goonewardene et al (1997) indicate that loss of mating weight does not affect estrus response and conception rates in goats. The findings on conception and kidding rates among all the doe breeds from this study were lower than those reported by Đuričić et al (2012) among Boer does.
Litter weight at birth and at weaning differed across the breed groups. The differences can be ascribed to the birth weight of the kids and the effect of their genetic makeup. The findings on litter size and weight at birth among all the doe breeds from this study were lower than those reported by Hamed et al (2009) and Taye et al (2013) among the Zaraibi and CHG does. The results pertaining to the litter weight at birth indicates that it was higher among the kids born as multiple births which is in line with the results of Mellado et al 2011 and Marzouk et al 2000 who reported that litter size and weight are correlated. Increase the former would invariably lead to increase in the later. These observations are in accordance with that of Dadi et al (2008) among the Arsi-Bale goats.
Year and season of kidding significantly affects litter weight at birth and at weaning which is also affected by the growth performance of the kids (which in turn is affected by the kid breed group and non-genetic factors). Year of kidding also significantly affects litter size at weaning which was mainly due to the survivability of the kids. These results are in line with the reports of Hamed et al (2009) among the Zaraibi goats. On the other hand, doe parity did not affect the studied traits which in contrary with the findings of Hamed et al (2009) among the Zaraibi goats who reported that litter weight of does increased with increasing parity, which might cause an increase in the birth weight of their kids.
High kid mortality rate was observed at the pre-weaning period, which is much higher than the findings of Debele et al (2015) and Zeleke et al (2017) among the Boer x Arsi bale, Boer x CHG. This study also shows that relatively high (26%) kids loss was occurred before reaching their fourth day after birth. Therefore, kid management at their earlier age should be focused in order to have better survival. Irrespective of the breed groups the flock had lower survivability which is in line with the findings of Davendra and Burns (1983), who reported that the Boer goats have a lower adaptability under the East African conditions.
The single born kids’ higher survivability (lower risk of mortality) from this study is in close accordance with the findings of Getachew et al (2015) and Devendra and Burns (1983). This may further be ascribed to the low body weight of the kids born as multiple births when compared to those born single Naude and Hofmeyr (1981). It may also be due to poor nursing of the dam and improper management practices to the multiple (Al-Najjar et al 2010). The year of birth too influenced (p<0.05) the survivability of the kids which may be due to the availability of feed and concentrate to the kids and the does over the study years, these results are in line with the findings of Getachew et al. (2015); Perez-Razo et al (1998); Al-Najjar et al (2010) and Vatankhah and Talebi (2009) who reported that, effect of year of birth is significant due to different factors changing over the years. Effect of year of birth is because of variation in incidences of diseases, parasites, climatic conditions, rearing methods and managerial systems (Vatankhah and Talebi 2009; Al-Najjar et al 2010).
The season of birth only influenced the survivability of the kids until 4 th day with the same being higher (p<0.01) among the kids born during the wet season. Effect of season can be attributable to the nutrient availability to the dam thereby influencing the lactal yield (Mukasa-Mugerwa et al 2000; Vatankhah and Talebi 2009). The effect of the season weans off as the animals mature and adapt to the surrounding agro climate (Al-Najjar et al 2010). Getachew et al (2015) indicated that the heritability of survivability is low and survivability in an animal is influenced largely by non-genetic parameters. Therefore, by minimizing the effect of the non-genetic factors (proper planning of mating aiming the kidding to be at feed available season, applying strong selection to increase percentages of single born) probability of kid mortality can be minimized.
Differences were however not recorded across sexes of the kids and the parity of the dam for survivability across all the age categories. The findings are in line with the study of Perez-Razo et al (1998) and partially with Getachew et al (2015) who reported that ewe parity did not affect lamb survival in Menz sheep. while it was contrary to the study by Al-Najjar et al (2010) who reported that parity affects survivability due to an age effect; as dams increased in maturity the survival rate of kids improved, due to the maternal instincts and lactal secretions of the does. The findings of Figures 1 and 2 indicate that the curve for survival function and cumulative hazard are negatively correlated. At day zero the cumulative hazard was zero while their survival function was full however, as the age of the kids proceed as the same time the probability of losing them also increase (fast increment in the cumulative hazard and fast decrement in the survival function curves) due to the above mention environmental factors. These observations were in close accordance with the findings of Getachew et al (2015) among the Menz, Wollo and their crossbreds with Awassi sheep. The results also indicate that, the survival line of the Boer and their F1 crosses cross over somewhere at day 125; showing straight downward of the Boer survival line while some stability was observed for the F1 crosses line. This may be due to the adaptation ability of the F1 crosses gained from the maternal line.
The authors want to thanks the Head and staff of Debre Birhan Agricultural Research Center of Amhara Agricultural Research Institute and School of Animal and Range Sciences, College of Agriculture, Hawassa University.
Alemu Y, Merkel R C 2008 Sheep and goat production handbook for Ethiopia. Addis Ababa, Ethiopia.
AL-Najjar K, Salhab S, AL-Merestani R, Kasem R, AL-Azzawi W, Dawa M, Omed H and Saatci M 2010 Environmental Factors Affecting Kid Mortality in Shami Goats. Kafkas University Veterinary Faculty Derg. 16, 431 – 435.
Ayalew W, Rischkowsky B, King J M and Bruns E 2003 Crossbreds did not generate more net benefits than indigenous goats in Ethiopian smallholdings. Agricultural Systems. 76, 1137–1156.
Browning M L L 2014 Small Ruminant Management: Winter Kidding & Lambing Seasons. Alabama Cooperative Extension System (Alabama A&M and Auburn Universities). Accessed on 12th October 2017 www.aces.edu/urban
Bushara I, El-Gazafey B, Mudalal M O, Mekki D M, Umsalama Ahmed M A and Abu-Nikhaila A M A 2016 Effect of Different Seasons on Taggar Goat Performance. International Journal of Research Studies in Agricultural Sciences 2 (7), 22 – 26.
Cameron M R, Luo J, Sahlu T, Hart S P, Coleman S W and Goetsch A L 2001 Growth and slaughter traits of Boer x Spanish, Boer x Angora, and Spanish goats consuming a concentrate-based diet. Journal of Animal Science 79, 1423-1430.
Casey N H and Van Niekerk WA 1988 The Boer goat I: Origin, adaptability, performance testing, reproduction and milk production. Small Ruminant Research. 1, 291 – 302.
Cunningham E P and Syrstad O 1987 Crossbreeding Bos Indicus and Bos Taurus for milk production in the tropics. FAO Animal Production and Health Paper. No. 68. FAO, Rome.
Dadi H, Duguma G, Shelima B, Fayera T, Tadesse M, Woldu T and Tucho T A 2008 Non-genetic factors influencing post-weaning growth and reproductive performances of Arsi-Bale goats. Livestock Research for Rural Development volume 20, Article no 114. http://www.lrrd.org/lrrd20/7/dadi20114.htm Accessed on 4th October 2017
Debele S, Habta M, Yinessu A and Admassu B 2015 Pre-weaning growth performance of Boer cross with local goat in Gedio zone, Ethiopia. International Journal of Agricultural Sciences and Veterinary Medicine. 3, 14 – 20.
Degen A 2007 Sheep and goat milk in pastoral societies. Small Ruminant Research. 68, 7–19.
Devendra C and Burns M 1983 Goat Production in the Tropics, 2nd ed. Farnham Royal, UK.
Ducrocq V, Sölkner J and Mészáros G 2010 The Survival Kit v6.0 User's Manual 1 – 83.
Đuričić D, Grizelj J, Dobranić T, Harapin I, Vince S, Kočila P, Folnožić I, Lipar M, Gračner G G and Samardžija M 2012 Reproductive performance of Boer goats in a moderate climate zone, Veterinarski Arhiv. 82, 351-358
Erasmus J A 2000 Adaptation to various environments and resistance to disease of the Improved Boer goat. Small Ruminant Research. 36, 179 – 187
Getachew T, Gizaw S, Wurzinger M, Haile A, Rischkowsky B, Okeyo A M, Sölkner J and Mészáros G 2015 Survival analysis of genetic and non-genetic factors influencing ewe longevity and lamb survival of Ethiopian sheep breeds. Livestock Science. 176, 22 – 32.
Girma Y and Desta G 2007 Investigation of salinity problems of Effratana Gidim and Kewot Woredas. Proceeding of the 2nd annual regional conferences on completed Natural Resources Management Research Activities. September 18-19, 2007. (pp.81-90). Bahir Dar, Ethiopia.
Gizaw S 2009 Goat breeds of Ethiopia: A guide for identification and utilization. Technical bulletin no. 27. 1– 9.
Goonewardene L A, Whitmore W, Jaeger S, Borchert T, Okine E, Ashmawy O and Emond S 1997 Effect of pre-breeding maintenance diet on subsequent reproduction by artificial insemination in Alpine and Saanen goats. Theriogenology 48, 151–159.
Gosey J A 1991 Crossbreeding Systems and the Theory behind Composite Breeds. Range Beef Cow Symposium. http://digitalcommons.unl.edu/rangebeefcowsymp/236/ (accessed 10 October 2017).
Greyling J P C 2000 Reproduction traits in the Boer goat doe. Small Ruminant Research 36, 171 – 177
Hamed A, Mabrouk M M, Shaat I and Bata S 2009 Estimation of genetic parameters and some non-genetic factors for litter size at birth and weaning and milk yield traits in zaraibi goats. Egyptian Journal of Sheep & Goat Sciences, 4 (2), 55-64
Horst P 1983 The concept of production adaptability of domestic animals in tropical and subtropical regions. Journal of South African Veterinary Association, 54, 159 – 164.
Khanal Piush 2016 Influence of crossbreeding and non-genetic factors on doe fitness traits of Boer F1 and foundation breeds in South Eastern United States. MSc Thesis, Tennessee State University, United States. 1 – 83.
Lu C D 1989 Effect of heat stress on goat production. Small Ruminant Research. 2, 151 – 162.
Marzouk K M, Feel El, F M R, Hassan H A and Sallam M T 2000 Evaluation of French Alpine goats under Egyptian condition. In proceedings of the 7th International conference on goats, France, 15 – 21 May 2000. France. 236 – 238
Mellado M, Meza-Herrera C A, Arévalo J R, De Santiago-Miramontes M A, Rodríguez A, Luna-Orozco J R and Veliz-Deras F G 2011 Relationship between litter birthweight and litter size in five goat breed groups. Animal Production Science. 51, 144–149
Mellado M, Valdéz R, García J E, López R and Rodríguez A 2006 Factors affecting the reproductive performance of goats under intensive conditions in a hot arid environment. Small Ruminant Research 63, 110-118
Mellado M, Valdez R, Lara L M and Garcia J E 2004 Risk factors involved in conception, abortion, and kidding rates of goats under extensive conditions. Small Ruminant Research 55, 191–198
Mestawet T, Girma A, Ĺdnřy T, Devold T G and Vegarud G E 2014 Effects of crossbreeding and mutation at the αs1-CN gene in the Ethiopian and crossbred goats on casein content, and coagulation properties of their milks; A short review. Small Ruminant Research. 122, 170–175.
Misra A K and Singh K 2002 Effect of water deprivation on dry matter intake, nutrient utilization and metabolic water production in goats under semi-arid zones of India. Small Ruminant Research. 46, 159–165.
Moses J 2006 Goat and Sheep Production and Marketing in the Amhara Region of Ethiopia. AMAREW project. 1–28.
Mustefa A, Gizaw S, Banerjee S, Abebe A, Taye M, Areaya A and Besufekad S 2019 Growth performance of Boer goats and their F1 and F2 crosses and backcrosses with Central Highland goats in Ethiopia. Livestock Research for Rural Development. Volume 31, Article #89. Retrieved August 27, 2019, from http://www.lrrd.org/lrrd31/6/amine31089.html
Naude R T and Hofmeyr H S 1981 Meat Production, in: Gall, C., (Eds.), Goat production. Academic press, London, pp. 285–307.
Perez-Razo M A, Sánchez G F and Meza H C 1998 Factors affecting kid survival in five goat breeds. Canadian Journal of Animal Sciences. 78, 407–411
SAS 2002 Statistical Analysis System, Institute Inc., Cary, NC, USA.
South African Boer Goat Breeders’ Association 2016 http://studbook.co.za/boergoat/stand.html.(accessed 14 October 2016).
Taye M, Deribe B and Hayle-Melekot M 2013 Reproductive performance of central highland goats under traditional management in Sekota district, Ethiopia. Asian Journal of Biological Sciences 6(5): 271 – 276.
Unanian M M and Silva A E D F 1989 Studies association of mal nutrition with abortion in goats in the north eastern region of Brazil. Pesquisa Agrop, Bras. 24, 1221-1228.
Vatankhah M and Talebi M A 2009 Genetic and non-genetic factors affecting mortality in Lori-Bakhtiari lambs. Asian-Australian Journal of Animal Science. 22, 459–464.
Zeleke T, Tilahun M, Deribe B, Lakew M, Belayneh N, Zegeye A and Aychew D 2017 Effects of non-genetic factors on pre-weaning growth, survivability and prolificacy of Central Highland x Boer crossbred goats in North Eastern Ethiopia. Livestock Research for Rural Development. Volume 29, Article #136 Retrieved October 12 2017 from http://www.lrrd.org/lrrd29/7/zele29136.html
Received 29 August 2019; Accepted 18 September 2019; Published 2 October 2019