Genetics and environmental trends in growth performance of Horro (Zebu) and crosses of Holstein Friesian and Jersey cattle breeds

Phenotypic, genetics and environmental trends of birth (BW), weaning (WW) and one year weight (YW), and average daily gain were considered on data collected from Horro cattle and their crosses during the year 1980-2008. Estimated breeding value (EBV) of all animals were generated from a univariate model analysis by the general linear mixed model by residual maximum likelihood (ASREML) while least square means for annual breeding values was calculated by the General Linear Model procedure of the statistical analysis systems (SAS). Direct heritability estimates from bivariate analyses in the present study were 0.61, 0.34 and 0.42 for BW, WW, and YW, respectively. The overall mean predicted breeding value for birth, weaning and one year weight were 0.11±0.06 kg, 0.13± 0.09 kg and1.2 ±1.4 kg, respectively. The results shown that breeding value trends have been improving and there was about 0.016 kg, 0.031 kg, 0.14 kg genetic gain in BW, WW and YW per year. The negative phenotypic and environmental trends were more pronounced for weaning weight and pre-weaning average daily gains in this study. Therefore, selection of in Horro and their crosses at one year weight would be suggested from present result due to high breeding annual value, high genetic correlation and positive environmental regression value.


INTRODUCTION
Genetic and environmental trends are measures of changes that take place in herds (Falconer and Mackay, 1996).They are indicators of genetic and management progresses made and thus are important in evaluation of the efficiency of breeding and determining the success of a selection programme (Musani and Mayer, 1997;Hofgren and Schinckel, 1998;Ebangi et al., 2000).Genetic and environmental trends provide information about herd improvement over time and are a reflection of the herd's progress compared to the breed as a whole (Hofgren and Schinckel, 1998).Positive genetic and environmental trends are indicators of favorable selection methods and good management, respectively.Therefore, to determine the effectiveness of genetic selection, genetic trends in the population can be considered (Van Wyk et al., 1993).Wilson and Willham (1986) reported E-mail: HMaassee@gmail.com.
Author(s) agree that this article remains permanently open access under the terms of the Creative Commons Attribution License 4.0 International License that trend lines are alternative methods of selection and strengthen the selection and management.Genetic and environmental trends are measures of changes that take place in herds (Wasike et al., 2006).Genetics and environment trends were important for selection within the indigenous breeds due to selection depends on correctly identifying animals with the highest true breeding value.Phenotypic, genetic and environmental trend are a quick assessment of a breeder's selection success in previous generations (Wilson and Willham, 1986).
There is scarcity of information on genetic and environmental trends in cattle breed in developing countries, which clearly indicate lack of evaluation of breeding programmes of these herds (Plasse et al., 2002).This study was, therefore, conducted with the objective of segregating the phenotypic value into environmental and genetic values and look into trends during the studies years.Therefore, examination of trends in Horro (Zebu) and crosses of Holstein Friesian and Jersey cattle breeds will provide information for reevaluation of management practices and earlier selection programmes for efficient growth performance of these cattle.

Study site
The data used in this study was generated from Horro cattle and their crosses kept at Bako Agricultural Research Centre during the year 1980-2008.The centre is located at about 250 km West of Addis Ababa at an altitude of 1650 m above sea level.The centre lies at about 09°6'N and 37°09'E.The area has a hot and sub humid climate and receives a mean annual rainfall of about 1220 mm, of which more than 80% falls in the months of May to September.Mean monthly minimum and maximum temperatures are about 14 and 28°C, respectively, with an average monthly temperature of 21°C.The daily mean minimum and maximum temperatures are 9.4 and 31.3°C,respectively.The vegetation cover of the area is woodland and open wood grassland type.The dominant pasture species include Hyperhenia (Hyperhenia anamasa) and Sporobolus (Sporobolus prraminmidalis) grasses and the legume Neonotonia (Ninotonia wighti) (Lemma et al., 1993 as cited by Temesgen, 2010).

Breeding system
Heifers are bred when they are at least two years of age and when they attain a body weight of 200 kg.Heat detection was done visually every day from 06:00 to 08:00 h in the morning and from 17:00-18:00 h in the afternoon by trained inseminator and during the grazing time by the herdsmen.Cows and heifers observed in heat were bred either naturally (local or crossbred bull) or inseminated with frozen semen of Holstein Friesian and Jersey which purchased from the National Artificial Insemination Center, within 24 h after heat.

Animals and managements
Calves were separated from their dams at birth, weighed and fed colostrum from a bucket for the first five days of life.A total of 227 L of milk was fed to each calf and a concentrate mix (49.5% ground maize, 49.5% noug seed cake and 1% salt) were offered until three months of age (weaning).After three months of age, weaned calves were maintained on natural pastures for approximately eight hours a day and supplemented with silage or hay adlibitum during the night and were kept as a group (male and female separately),where concentrate were supplemented to heifer calves only on availability.

Data collection
A total of 2359 calves' records were used in which produced from 184 sires and 710 dams, born from 1980 to 2008 years.Data were recorded from birth weight, weaning weight (age ranges from 61-111 days) and one year weight (age ranges 336-388 day) of Horro and its crossbred animals.

Estimation of genetic parameters
Five variables were analyzed for genetic parameters using a univariate of Animal model of procedure: birth weight (BW), weaning weight (WW), pre-weaning average daily gain (DG), postweaning average daily gain (PDG) and one year weight (YW).Correlations among the different components of the different traits were estimated from bivariate analyses using the direct animal model.The direct animal model was chosen because of limitation of data and inability of the analysis to converge to a global maximum solution when models other than the direct animal were used.
Genetic and environmental parameters were calculated using the (co) variances estimated at convergence.Direct (h 2 a) and the direct additive covariance were calculated as (where σ 2 p is total phenotypic variance) σ 2 a/σ 2 p, and σ 2 am/σ 2 p, respectively.Total heritability were calculated as (σ 2 a +0.5σ 2 m +1.5σam)/ σ 2 p, while direct and maternal additive correlation was expressed as a ratio of the covariance to the square root of the product of direct variance (ram = σam / (σ 2 a).The models are numbered, according to Meyer (1994) as follows: Where: Y is the vector of observations; ß is the vector of fixed effects; X is the incidence matrix that associates ß with Y; a is the vector of breeding values for direct genetic effects; Z1, is the incidence matrices that associate a, with Y; and e is the vector of residual effects.Furthermore, with A, the numerator relationship matrix between animals, In, an identity matrix with order n, the number of dams, and I, an identity matrix with order of the number of records the (co) variance structure of random effects can be described as: V(a) =δ 2 aA, V(e) = δ 2 e I, where δ 2 a is the direct genetic variance; δ 2 e is the residual variance and δam is the genetic covariance between direct and maternal effects.All calculations were done using the options available in ASREML (Gilmour et al., 1999) for parameter and sampling error estimation.

Estimation of breeding value
Estimated breeding value (EBV) of all animals was generated from a univariate model analysis by the ASREML (Gilmour et al., 1999).Animal model where the additive effect of the animal was considered was used.Least square means for annual breeding

RESULTS
The overall mean predicted breeding value for birth, weaning, one year weight; pre-weaning and postweaning average daily gains were 0.11±0.060kg, 0.13± 0.089 kg, 1.2 ±1.4 kg, 0.93±0.57gm and 4.6±1.1gm,respectively.The regression coefficients of breeding value for BW, WW, YW, DG and PDG with year of birth were 0.0157±0.012kg, 0.031±0.02kg, 0.14±0.06kg and 0.16±0.1 g and 0.21±0.11g, respectively.All coefficients were highly significantly different from zero (P<0.001)(Table 1).It was shown that breeding value trends have been improving and there was about 0.0157 kg, 0.031 kg, 0.14 kg and 0.16, and 0.21 g genetic gain in BW, WW, YW, DG and PDG per year.Mean of predicated breeding value for animal born in the different years ranged between, -0.96 to 1.8 kg, -1.4 to 2.5 kg, -3.4 to 11.3 kg, -6.7 to 10.8 g and -6.2 to 22 g, respectively, for BW, WW, YW, DG and PDG.Individual breeding values for the total year ranged from -6.6 to 15 kg, -11 to 14.5 kg, -23.9 to 55 kg, -76 to 87.5 g and -50 to 96 g, respectively, for BW, WW, YW, DG and PDG, respectively.Phenotypic, genetic (breeding values) and environmental trends on year of birth for birth weight (BW), weaning weight (WW), one year weight (YW), pre-weaning average daily gain (DG) and post-weaning average daily gain (PDG) are presented in Figures 1, 2, 3, 4 and 5, respectively (Tables 3 and 4).

DISCUSSIONS
Direct heritability estimates in all models were higher than the usually accepted range of heritability of 0.4 to 0.45 for birth weight in cattle (Woldehawariate et al., 1977).Using animal model and sire model, Mohamed (2004) Meyer (1992).The higher heritability obtained in this study is in agreement with the findings of Schoeman and Jordaan (1999) and Skrypzeck et al. (2000) who found a higher direct heritability estimates of 0.62 and 0.72 for birth weight, respectively.Both authors mentioned that fairly high  heritability, arising from large genetic variances due to the multibreed composition of the herd could have been expected, since the population consists of 15 breeds and this effect was not accounted for by the model.Similarly, in the current study large numbers of genetic groups were categorized into only three groups to have reasonable number of observations in each category.This would create a high level of genetic variability within a group and inflating the estimate of heritability.Direct heritability of weaning weight was lower than birth weight (Table 1).WW, DG, YW and PDG, respectively.These findings were consistent with those of Meyer (1993) and Wasike (2006) on birth weight of Australian Polled Hereford cattle and Boran cattle, respectively.Other studies have reported increase in magnitude of the estimates of genetic variances and heritability from multivariate analyses as compared to parameters in univariate analyses, since univariate analyses used to compare or selected the best model from different model.This was attributed to elimination of selection bias (Mrode, 2000).Wasike (2006) reported higher heritability estimates for weaning weight and one year weight when multivariate models were used.The direct genetic correlations between birth weight and weaning weight, pre-weaning average daily gain, one year weight and post-weaning average daily gain were: 0.75±0.06,0.25±0.11,0.66±0.07and 0.54±0.09,respectively, whereas, genetic correlation between weaning weight and pre-weaning average daily gain, one year weight and post-weaning average daily gain were: 0.84±0.03,0.79±0.07and 0.58±0.11,respectively, in Horro cattle and its crosses.The results are in agreement with the findings of Demeke et al. (2003).There was high genetic correlation between one year weight and post-weaning average daily gain (0.95±0.01).Low and weak genetic correlations between birth weight and other growth performance traits have been reported in other studies (Haile-Mariam and Kassa-Mersha, 1994;Aynalem, 2006).The genetic correlation between birth weight with weaning weight and one year weights were positive and high, which is in agreement with the results obtained by Wasike (2006) but in disagreement with the results obtained by Aynalem (2006).The strong positive correlation may not be desirable since selection, for instance on the basis of weaning weight, could result in calving difficulties due to increased birth weight (Solomon and Gemeda, 2002;Aynalem, 2006).The phenotypic correlation between BW and other growth traits (WW, YW and PDG) were low and range from 0.16±0.03 to 0.37±0.02but negative and low between WW and DG (-0.05±0.02).This is implies managemental problems at weaning and pre-weaning average daily gain.Phenotypic correlation between WW and YW was 0.58±0.02.Phenotypic correlation between YW and PDG were high (0.92±0.01).The high phenotypic correlation values between weaning and one year weight and between one year weight and postweaning average daily gain.With the exception of correlation between WW and DG, genetic correlations between the traits were higher than the corresponding phenotypic correlations.
The annual breeding value trends have been improving and there was about 0.016 kg, 0.031 kg, 0.14 kg and 0.16 g and 0.21 g genetic gain in BW, WW, YW, DG and PDG per year.The possible reason for inconsistent genetic improvement in the present study could be because in the herds from which the current data was used there was no careful selection exercised.However, selection of Horro and crossbreds males calves based on total volume, total count and motility and testicular measurement for natural services at one year weight had been reported (Habtamu et al., 2007).Furthermore, selections of breeding bulls were based on visual observation (body size) and culling of inferior animals was possible reason.Relatively higher genetic gain in one year weight was presented in present results.Similarly, the largest genetic trend which could be because of the heritability of the traits since the genetic trend depends on heritability of a trait (Gray et al., 1999;Shaat et al., 2004;Temesgen, 2010).Thus it was only YW that was subjected to any kind of selection in Horro and crossbred cattle at Bako Agricultural Research Center of livestock farms.
The regression value of environmental effect on year of birth was -0.059±0.03,-0.21±0.06,-0.15±0.22,and -1.6±0.6 and 0.094±0.64for BW, WW, YW, DG and PDG, respectively.These values were very large in magnitude when compared to genetic value and could not be counteracted by the improvement in genetic value.Thus the phenotypic trend in all traits was shown to follow the pattern of the environmental trend.The results are in agreement with findings of Solomon and Gemeda (2002) on Horro sheep at same location.Environmental and phenotypic trends showed a large decline and fluctuating patterns crosses the years (Figures 1-5).
The genetic trends for growth traits indicated a slow but positive rate of progress in growth performance in the Horro and crossbred cattle in the area studied (Figures 1-5).Improved breeding values were observed in this study, particularly between the years 1997 and 2005 for all growth traits due to crossbreeding (AI used).Herds selected on the basis of estimated breeding values have higher genetic trends than those selected based on physical appraisal.This was also the scenario in herds where proven bulls are used against herds where untested bulls are used (Plasse et al., 2002).As shown in Figures 1 and 2, the insignificant change in trends of BW and fluctuations in WW imply, calf BW was independent of environmental influence while WW fluctuated due to the fluctuation in availability feed, health and the availability of supplemented feed.The negative trends in post weaning growth traits up to 1984 were due to a lapse in management on the farms, which improved, in the later years.This shows that management of cattle has a great impact on their growth performance of the calves.The environment provided determines the ultimate growth achieved in Horro and crossbred cattle Bako Agricultural Research Center of livestock farms.
Despite some amount of genetic gain in all the traits, the phenotypic trend has shown significant decline.Wilson and Willham (1986) revealed that there is not necessary to remove environmental trends from phenotypic trends to obtain unbiased estimates of genetic trends.The same authors confirmed that environmental trends could be instructive to a commercial breeder to monitor actual management effects and/or climatic changes.
However, negative phenotypic and environmental trends were more pronounced for preweaning average daily gains in this study.This indicates that attention needs to be given to environmental factors such as nutrition, health and management.Breeding values were almost positive, but the magnitudes varied considerably among the traits.Positive genetic and environmental trends are indicators of favorable selection methods and good management (Plasse et al., 2002).
Similar decline in the phenotypic trend in the presence of genetic gain have been reported for weaning and one year weight in Boran cattle (Haile-Mariam and Philipson, 1996), for milk yield in Sahiwal cattle in Kenya (Rege and Wakhungu, 1992) and for Horro sheep (Solomon and Gemeda, 2002).Solomon and Gemeda (2002) reported that increase in stocking rate, increasing crop land, deterioration of grazing area, fluctuations in climatic conditions, morbidity, the turnover of people who worked in the management of the Horro sheep flock and more use of portion of the flock to stressful experiments might have contributed to the decline and erratic nature of the environmental and phenotypic trend.This explanation is probably true for the present results since the animals reared under the same management and environmental situations.Plasse et al. (2002) observed fluctuating phenotypic trends for 548-day weight, which was attributed to the deterioration in pastures during the study period.From this result it can be suggested that genetic improvement work should be supported by improved environment.

CONCLUSIONS AND RECOMMENDATIONS
This study has also shown higher estimates genetic parameters when animal models used for growth performance of Horro and their crosses.The genetic correlation estimates between various growth traits are strong.The genetic correlation between birth weight with weaning and one year weights were positive and high.This may not be desirable since selection for instance on the basis of weaning weight could result in calving difficulties due to increased birth weight.Similarly, the negative phenotypic and environmental trends were more pronounced for weaning weight and pre-weaning average daily gains in this study.This indicates that attention needs to be given to environmental factors such as nutrition and management.However, breeding values were almost positive, but the magnitudes varied considerably among the traits.The result also suggests that improvement in genetic gain should be accompanied by either improvement of the management.This ensures whatever is gained through genetic means may not be lost due to decline in level of management.Selection of Horro and their crosses at one year weight would be suggested from present result due to high breeding annual value, high genetic correlation and positive environmental regression value.

Figure 1 .
Figure 1.Annual phenotypic (PV), breeding value (BV) and environmental (EV) value trends of birth weight of Horro cattle and their crosses between 1980 and 2008.

Figure 2 .
Figure 2. Annual phenotypic (PV), breeding value (BV) and environmental (EV) value trends of weaning weight of Horro cattle and their crosses between 1980 and 2008.

Figure 3 .
Figure 3. Annual phenotypic (PV), breeding value (BV) and environmental (EV) value trends of one year weight of Horro cattle and their crosses between 1980 and 2008.

Figure 4 .
Figure 4. Annual phenotypic (PV), breeding value (BV) and environmental (EV) value trends of pre-weaning average daily gains of Horro cattle and their crosses.

Figure 5 .
Figure 5. Annual phenotypic (PV), breeding value (BV) and environmental (EV) value trends of post-weaning average daily gains of Horro cattle and their crosses between 1980 and 2008

Table 1 .
(Co)variance components and genetic parameters for growth of Horro and their crosses by using Animal model.

Table 2 .
Estimates of phenotypic (rp12), direct genetic (ra12) and residual (re12) correlations and heritability (h 2 a) birth (BW), weaning weight (WW), pre-weaning average daily gain (DG), one year weight (YW) and post-weaning average daily gain (PDG) BW=Birth weight; WW= weaning weight; DG=pre-weaning average daily gain; YW=one year weight; PDG=post-weaning average daily gain values was calculated by the General Linear Model procedure of the statistical analysis systems(SAS, 2004)and deviations from the mean of the base year (1980) were considered as estimates of annual breeding value.Deviation from least squares means of growth performance from the base year (1980) were assumed to be estimates of annual phenotypic values.Annual environmental values were calculated as the difference between phenotypic and breeding values.Phenotypic, breeding and environmental trends were evaluated by regression of annual values (deviations from the base year) on year.

Table 3 .
Regression coefficients of phenotypic, breeding and environmental trends in birth, weaning and one year weight, pre-weaning and post-weaning average daily gains.

Table 4 .
Range of breeding values of individual Animals born in the different years.