Relation between body condition score and conception rate of Japanese Black cows

Objective This study analyzes interactions of body condition score (BCS) with other factors and the effect of BCS on estimates of genetic paremeters of conception rate (CR) in Japanese Black cows. Methods Factors affecting CR were analyzed through the linear mixed model, and genetic parameters of CR were estimated through the threshold animal model. Results The interactions between BCS and each season and the number of artificial inseminations (AI) was significant (p<0.05), but that between BCS and parity showed no significance for CR. High CR was observed with BCS 3 in autumn (0.56±0.01) and BCS 4 in summer (0.56±0.02). The highest CR with BCS 3 (0.56±0.02) and BCS 4 (0.55 ±0.01) was observed at first AI. With BCS 5, however, the highest CR (0.55±0.08) was observed at second AI. Conclusion The model with BCS was notably conducive to the estimation of genetic parameters because of a low deviance information criterion of heritability that, nevertheless, was slightly lower than the model without BCS.


INTRODUCTION
In the recent decade, the price of Japanese Black calves has doubled because of the high cost of calf production in the Japanese cowcalf operating system. High production cost is primarily due to a decline in reproductive efficiency. The low efficiency in calfproducing operations is mostly influenced by the rate of artificial insemination (AI) success [1], which is controlled by numerous factors, including sperm quality [2], season [3], AI technician [4], parity [5], health, nutrition and heatdetection methods for the herd [6]. Conception rate (CR) is one of the indicators commonly used for evaluating the success of AI in cattle.
Body condition score (BCS) is a method for the management and monitoring of the amount and accumulated relative fitness of cattle [7]. Body condition of cows after calving is closely related to AI efficiency, i.e., success of first insemination, CR, and interval from first insemination to gestation [8,9]. Assuming that the success of AI is related to BCS, the magnitude and change of BCS after calving needs to be investigated to determine wheth er nutritional goals are achieved and to identify potential problems associated with failure of AI [10]. Cows with poor BCS may have longer time to express estrous after calving. On the other hand, cows with high BCS are usually phenotyped as obese or overweight animals requiring plural servicings for conception and as exhibiting inefficiency of the normal re productive cycle [11,12]. The interaction between BCS, seasonal changes and feeding regimes direcly affects the reproductive performance of cows [3]. The aim of this study was to analyze the interactions between BCS with other fac tors and the effect of BCS on estimates of genetic paremeters of CR in Japanese Black cows.

Data collection
Insemination and pedigree records of Japanese Black cows were obtained from Artificial Insemination Center of Northern Okinawa. A data set comprised 6,034 records of AI carried out on 2,114 cows by eight AI technicians. Collected be tween January 2015 and December 2017 from 114 farms, the records comprised 5,003 cows with and 1,031 without BCS data. The total number of animals in the pedigree was 16,334. BCS was judged at the time of insemination and scaled using 5point system from one (1) for attenuated animals to five (5) for obese animals [13,14]. CR was coded 1 if the cow was inseminated and subsequently concieved successfully, otherwise 0. Cow was declared as conceived after two evalu ations: nonreturn at 21 days after AI, and palpation per rectum at 35 to 40 days after AI. Approval from the committee on the care and use of animals was not sought because data used in this study were collected from field records; no field experiments were conducted.
GLIMMIX procedure with contrasts of statistical analysis system 9.3 (SAS) [15] was used to analyze the effect of the interaction between BCS and parity, BCS and season, and BCS and AI number. Farm was treated as a random effect. The linear model was as follows: where y ijklmn is the observation of CR, B i the ith fixed effect of BCS, P j the jth fixed effect of parity, S k the kth fixed effect of season, A l the lth fixed effect of AI number, (BP) ij , (BS) ik , and (BA) il are effects of interactions between ith of BCS with jth effect of parity, kth effect of season and lth effect of AI num ber, recpectively, c m is the random effect of cow/animal, f n is the random effect of farm and e ijklmn the random residual of y ijklmn .

Estimate of genetic parameters
Genetic parameters of CR were estimated through single animal model, in matrix notation the mixed linear model was: where, y is the a vector of CR, b is a vector of fixed effect, X is an incidence matrix for the fixed effect, a is the vector of random genetic additive effect, pe is the vector of permanent environmental effect, Z 1 and Z 2 are incidence matrix, and e is the residuals vector.
This analysis was repeated twice; the first with BCS as fixed effect and the second without BCS to know the impact of BCS on genetic parameters estimated.
The Bayesian procedure [16] assuming that the normal distribution with density as: where, y is the a vector of CR, b is a vector of fixed effect, X is an incidence matrix for the 104 is the vector of random genetic additive effect, pe is the vector of permanent environm 105 and are incidence matrix, and e is the residuals vector.

106
This analysis was repeated twice; the first with BCS as fixed effect and the second wit 107 know the impact of BCS on genetic parameters estimated.

108
The Bayesian procedure [16] assuming that the normal distribution with density as:

108
The Bayesian procedure [16] assuming that the normal distribution with density as:

108
The Bayesian procedure [16] assuming that the normal distribution with density as: the additive genetic variance, permanent environmental variance and residual variance, respectively. The variance covariance components were esti mated through Bayesian procedures THRGIBBS1F90 [17], that yielded a period of data collection of 1,000,000 itera tions after a burnin period of 100,000 iterations. Values of standard error were estimated through POSTGIBBS1F90 of the Geweke diagnostic test [18]. A posteriori distribution was obtained with 4,000 samples which were taken for every 250 cycles. This estimation was repeated twice; the first with BCS and the second without BCS analysis. Geweke criteria and error of Monte Carlo chain (MCE) were obtained by calculating the variance of samples for each component di vided by the number of samples to monitor the convergence.

RESULTS
The reproductive performance of cows from 114 farms is summarized in Table 1. BCS, and parity were statistically significant (p<0.05) for CR, whereas season and AI num ber were not. The interaction between BCS and between season and BCS and AI number was significant, however, that between BCS and parity was not significant (Table 2). Accordingly, multiple comparisons between BCS, and season; BCS, and AI number were tested afterwards.
In terms of the interaction between BCS and the season for CR, the BCS showed no significant difference among the seasons. Nevertheless, significant seasonal differences were observed in BCS 3 and 4. A high CR was seen with BCS 3 in autumn and with BCS 4 in summer ( Table 3). Table 4. The effect of BCS, AI number and their interactions were significant differences. Significant differences of AI num ber were observed in BCS 3, BCS 4, and BCS 5. The highest CR with BCS 3 was observed in cows at first AI, and CR de creased as the AI number increased. With BCS 5, the highest CR was seen in cows at second AI. Also, significant differ ences among the BCS were observed only at the second and third AI.

The interaction between BCS and AI number is shown in
In comparing estimated heritability between animals with and without BCS, the latter was slightly higher than the for mer (Table 5). Heritability with BCS showed a low posterior standard deviation. Values for Geweke (pvalue), MCE, and deviance information criterion (DIC) for estimated herita bility with BCS were lower than that without BCS.

DISCUSSION
The mean CR of the Japanese Black cow in this study (0.52±     BCS, body condition score; AI, artificial insemination; LSM, lease squares means; SE, standard error. a,b Different superscripts within a column shows significant differences (p < 0.05).
x,y Different superscripts within a row shows significant differences (p < 0.05). 0.01) was slightly higher than previously reported (0.48±0.02) [19]. Compared with other breeds, CR in Japanese Black was slightly lower than the estimates of beef cows in New Zealand (0.55) [20] and of local zebu cows (0.57) [21]. Although fac tors for the lower CR of Japanese Black cows are still unclear, some studies have indicated that a decrease in reproductive performance of this breed might be due to genetic improve ment programs that have heavily focused on meat quantity and meat quality, especially on beef marbling [122]. In this study, BCS and parity have a significant effect (p< 0.05) on CR. Significance of parity in Japanese Black cows has also been described [23], but without showing any sig nificant effect of BCS on CR. Furthermore, in the current study, the interaction between BCS and each of season and AI number was significant (p<0.05); however, no significant effect of the interaction between BCS and parity has been shown in the aforementioned study. In contrast, a statitically significant interaction between BCS and parity for CR has been described in Florida beef cattle [24].
Cows with BCS 5 showed the lowest CR in spring and summer. On the other hand, cows with BCS 4 showed the highest CR in summer. Low CR of cows serviced in spring is consistent with the previous studies in terms of seasonal effect on reproductive performace of Japanese Black cows, whereas low CR has been described in cows serviced in spring and winter [19]. Japanese Black cows need more servicings for conception in spring [25].
In terms of AI numbers, CR showed a significant differ ence between the second and third servicing. As the AI number increased, the CR of cows with BCS 3 and 4 de creased. The lowest CR of cows with BCS 5 was observed in the third AI. This result also agrees with a study on the num ber of servicings for CR of Japanese Black cattle [19], where a low CR is shown at a second or later insemination. Various factors have been reported for low CR on repeated AI: ge netic [26], nutritional [27] and hormonal causality [28,29]. Cows with low BCS that fail to conceive at the first AI need nutritional treatment to improve CR in subsequent AI.
In the present study, search of the literature produced no estimates of heritability of CR in Japanese Black cows. The estimated heritability of CR (0.012 and 0.021) was lower than that reported for Holstein cows (0.027 to 0.049) in Japan [30]. Low DIC of the model with BCS indicated that the model was better than the model without BCS [31]. The statistical significance of BCS for CR comfirms that BCS should be in cluded in the genetic analysis of CR records.

CONCLUSION
The present study showed that Japanese Black cows with a moderate BCS had a good conception rate. The model with BCS fitted well with the estimation of genetic parameters because of low DIC of heritability that, albeit, was slightly low.

CONFLICT OF INTEREST
We certify that there is no conflict of interest with any financial organization regarding the material discussed in the manu script.

FUNDING
The authors received no financial support for this article.