I. INTRODUCTION

Artificial insemination (AI) is a proven biotechnique, used globally to improve the genetic makeup and efficiency of reproduction performance in cattle (Noakes, 2009). However, the AI service in Ethiopia was not successful in improving the reproductive performance of the dairy industry (Sinishaw, 2005). Due to poor heat detection, the timing of insemination, embryonic mortalities, inadequate infrastructure and poor management system the efficiency of the AI service in Ethiopia is one of the lowest among the developing countries (Shiferaw et al., 2003). According to Haileyesus (2006), the average number of services per conception (NSPC) among local and crossbreed cows was 2.54 and 2.38, respectively.

One of the ways to overcome the challenges of low conception rates and less efficient AI service is by using hormones for estrous synchronization, thus facilitating the practical use of artificial insemination, and this can positively influence the productive, and reproductive efficiency of dairy cattle (Murugavel et al., 2010). The use of reproductive hormones can help in the planning of AI in a way that coincides with planned parturitions in a specific period. Thus to synchronize the estrus cycle of cattle, PGF2α or its analog is a commonly used hormone, the functionality of which depends on the presence of at least a functional corpus leuteum (CL) in the diestrus stage of the estrous cycle (7 to 17 days of the cycle) (Cordova-Izquierdo et al., 2009).

To overcome such constraints, estrus synchronization combining with artificial insemination (AI) was introduced as a tool to improve the reproductive performance of cattle (Cordova et al., 2009). To this effect, the benefit of using such a technological option is an approach to improve desirable animal genetic materials. Therefore to boost the dairy industries and to alleviate the problems of the aforementioned gaps, evaluation of the effect of Prostaglandin on estrus synchronization of dairy cattle under small holders' conditions is indispensable. Therefore, the objectives of the study were to evaluate the success associated with estrus synchronization and mass insemination at development intervention, re-evaluating the technology through action research, and draw recommendations for the sustainable application of this technology in the study area.

II. MATERIALS AND METHODS

a) Description of the Study Areas

The study was conducted at Arbegona and Bensa districts of Sidama zone, Southern Region. Geographically, Sidama zone is situated between the coordinates of 5045' and 6045' N latitude and 38039' and 38029' E longitude with altitude ranging from 1100 to 3500 m asl. The rainfall pattern of the zone is a bimodal type with a small rainfall amount during February to April followed by the main rainy season from July to September. It has diverse agroecology classified as highlands, midlands, and semi-dry lowlands covering

30%, 60%, and 10%, respectively. The farming system of the zone is characterized as a mixed type of crop and livestock farming and endowed with different livestock resources such as cattle, sheep and goats, equines, and poultry.

b) Evaluation of Synchronization and Mass Insemination in Development Intervention

The performance of the Oestrus synchronization and mass AI program assessed using secondary data from the respective Agricultural Office of Bensa and Arbogona districts. A total of 883 dairy cattle data included in the study. For this evaluation, oestrus response, conception rate, and number of services per conception are taken as dependent variables. Likewise, the independent variables were genotype, parity zero to multiparas, BCS (1 to 9 scale) following Nicholson (1986), age of animals, insemination time, and bull semen. For comparing of action research conducted by Irrigation Value Chains for Ethiopian Smallholders (LIVES) projects of International Livestock Research Institute (ILRI) project and development interventions carried out by each district of BoA compared with results from action research activities conducted in the same area.

c) Evaluation of Estrous Synchronization and Mass AI in Action Research

In the experiment, A total of 126 cows selected from Bensa district based on the presence of receptive corpus luteum and absence of pregnancy upon rectal palpation, parity ( $1^{st}$ to $5^{th}$ level), in good health status, and BSC (1 to 9 scale) following Nicholson (1986), and treated with a single injection of PGF2α (5 ml LutalyseTM) hormone. At the end of the study period, hormone response after injection of PGF2α, conception rate (CR), and the number of services per conception determined (NSPC) as dependent variables. The independent variables/factors were genotype, BCS, age of cows, parity of cows, time of AI, and bull itself.

d) Heat Detection and Insemination Procedure

Before the commencement, farmers were trained about how to detect estrus/heat in their breeding females. The cows were inseminated according to AM/PM method as suggested by Peter (2004). One well-experienced and trained inseminator was assigned for insemination to avoid any ambiguity of differences arising due to the efficiency of the inseminators. The inseminator through rectal palpation assessed the stage of estrus when the animals were brought to the AI center. The cows inseminated after they showed signs of standing estrus (heat) that ranged between 24 to 120 hours post-PGF2α administration with frozen semen (100%) of Holstein Friesian bulls.

e) Data Analysis

The efficiency of estrus synchronization and the association between CR, number of service per conceptions, and determining factors such as location, BCS, parity, breed, age, time of insemination, and bull IDs were analyzed using the chi-square test. The variation between groups was considered significant when the P-value was less than 0.05. CR and NSPC compute according to the method suggested by (Sharifuzzaman et al., 2015).

III. RESULTS AND DISCUSSION

a) Performance of Oestrus Synchronization and Mass Insemination

The rate of oestrous response to hormone treatment, measured as the percentage of females that showed oestrus out of the total treated animals under regular development intervention by the responsive BoA, was comparable to the result obtained in action research activity. The rate of Oestrus response in action research and development intervention presented in Table1. The result of response rate in action research was 90%, somewhat higher than with that of development intervention (87.2%), still, it was lower than those reported by Azage et al. (2012) (97.7%) in Dale district and Tewodros et al.(2015) (98.9%) in Fogera district. The rate of estrus response of action research and development intervention was slightly similar to Adebabay et al. (2013) (89.3%) in Bahir Dar milk shed, however, it was higher than from the reported of Bekana (2005) 82%. The differences in estrus response among the studies attributed to body condition score, age of the animals, breeding season, no-cyclic animals, and disease (Tewodros et al., 2015).

In the current study, interval to oestrus response after PGF2α in action research was shown slightly longer than that of regular development intervention (Table 1). The present finding was higher than those reported by Hamid (2012) 57.9 hrs in Siltie, and Adebebay et al. (2013) 51 and 50 hrs in Bahirdar; however it was lower than those reported by the report of Azage et al. (2012) 13 to 154 hours in Hawassa-Dilla milkshed, and Million et al. (2011) 70.67 hours for Boran cattle. This longer or shorter estrus response interval most likely due to poorly developed ovaries and early stages of CL during the administration of PGF2α.

b) Conception Rate and Number of Service per Conception

As shown in Table 1, there was an enormous variation between conceptions under action research (57.45%) and development intervention (42.2%) of the current study districts. However, the conception rate both in action research and development intervention was better than that reported by Adebabay et al. (2013) 13.3% in Bahir Dar and Desalegn (2008) 27% in Ethiopia. In contrast, the rate of conception was lower with the preliminary results of mass synchronization in SNNPR (62.2%) and Tigray (62%) reported by (Azage, 2012).

The average NSPC in action research was lower than that of development intervention (Table 1). The number of service per conception result under action research was lower than those reported by Hyleyesus (2006) for the local (2.54) and crossbreed (2.38) cattle, but it was higher than what was described by (Menal et al., 2011; Nibret, 2012) 1.25 and 1.3, respectively, but it was in line with the report (1.75) of Sharifuzzaman et al. (2015). Negussie (1992) also indicated that high NSPC is associated with poor semen quality, semen handling, and insemination practices.

Factors affecting the performance of Conception rate

a) Effect of Genotype

In development intervention, the rate of conception (CR) was a significant difference $P\le 0.05$ among the genotypes, but, in action research, there was no significant difference among the local and crossbred cattle. Irrespective of the studies, the conception rate of crossbred is higher than the local cattle. The lower conception rate among the native cattle attributed to the silent heat observed in zebu cattle (Azage et al., 1989). The difference among the reports might be the time of insemination, heat detection efficiency, breeding season, proper semen handling, and intrinsic factors of the cow itself. Similar studies made by Mollal (2011) and Shikder (2011) in his report CR depends mainly on skills of the inseminator, accurate estrus detection, quality and quantity of spermatozoa in semen, proper semen thawing procedure, and placement of semen in the uterus.

b) Effect of Parity and Age

The results of conception in development intervention did not differ (P>0.05) among the parties, but significant (P<0.05) variation observe in action research (Table 2). The performance of conception was higher among the cows aged between 5 to 7 years; this roughly corresponds to the second and third parity of the cows and decreases as the parity number increased. The finding of the results was in line with several other studies that reported that the conception was better among the cows at the second parity (Miah et al., 2004; Khan et al., 2008; Bhattacharyya et al., 2009). The present result showed that as animals' age increases, the CR shows a negative trend, this may be associated with lactation stress, and also that the older cows tend to gain weight, thereby reducing the chances of fertility. Irrespective of all the above conditions about poor CR among the older ages in cows align with the studies by Gebregziabher (2005), it indicated that reduction in the probability of conception with increased age of the cow attributed to the exposure of the cows to different reproductive diseases, older cows and postpartum anoestrus.

c) Effect of Body Condition Scores

Conception rates both in action research and development intervention were significantly $\le P0.01$ affected by the BCS (Table 2). A similar study by Grimard et al. (2003) indicates that the effectiveness of the AI depends on the BCS of the cows. Shamsuddin et al. (2001) and Emebet (2007) also stated that animals with optimum body conditions at insemination had better fertility than those with poor or higher BCS. The variation can be non-genetic factors viz. availability of feed source, season, and year besides assessing the optimal body condition for a particular breed. Studies by Bó et al. (2003) and Hossain (2013) indicated that suboptimal BCS could be a fallout of poor nutritional status that is often observed among the cattle grazing on degraded pastures, especially in the tropics.

d) Effect of Time of Insemination

The results of studies both in action research and mass insemination indicated that conception rate was higher (P<0.05) when the cows were inseminated within 10 to 15 hrs. intervals after the onset of estrus, indicating the validity of AM/PM approach, however, the CR lower when insemination done after 20 hours of estrus (Table 3). Although the appropriate time of insemination in both studies was between 10 to 15 hours of the onset of estrus, the differences as observed could be associated with several non-genetic factors as mentioned ahead. Probably this interval was too short for the time of ovulation. A similar report by Mufti et al. (2010) indicated that the conception was higher when inseminated between 10 to 14 hours after the onset of estrus, but lower CR observed among the cows inseminated earlier or later the above mentioned period. The observations are in close accordance with the findings of Miah et al. (2004) indicate that if insemination is carried out later than 22 hours results in poor conception. The variation in conception rate among different studies could be due to inaccuracy of heat detection, time and season of insemination, skills of the AIT. A Study by Sinishaw (2005) also indicated that animals should inseminate within 24 hours of the onset of heat because late and early insemination may influence the CR of cows.

might be related either due to the level of skill and experience difference of AI technicians on the thawing method, site of semen deposition and heat detection error or quality of the semen, etc. A study by Mufti et al. (2010) indicated that the effect of thawing at different periods of interval influenced the conception rate significantly. Other studies also reported that the quality of frozen semen has a significant result on CR (Shamsuddin et al., 2001; Anzar et al., 2003).

e) Effect of Bull

Conception rate was a significant variation between bulls ID itself inseminated to the responded cows and higher conception rate observed in bull numbers 10-264 and 10-293 (Table 3). The difference

f) Estrus Detection Time

After PGF2α administration most of the cows' estrus sign was observed during the night than days (Figure 1). The observations are in line with the findings of Bekana et al. (2005), who reported that in Fogera cattle, most of the estrus signs observe during the cooler hours of the night. Higher CR also observed when the AI was carried out before noon, indicating that CR was better when AI was conducted at cooler hours of the day (Nordin et al., 2004). Proper detection of estrus is difficult, especially when the cows exhibit during the night hours, thus unless or until the cows monitored regularly to assess the time of standing heat and AI (Aulakh, 2008).

IV. CONCLUSION AND RECOMMENDATIONS

The rate of oestrus response (90% of the case) in action research was slightly higher than the result obtained from development intervention (87.2% of the case). The rate of conception (58.4%) in action research was higher than that of the development intervention (42.2%). Therefore, result from action research and the development intervention AI service with estrus synchronized indicates a possibility to improve the conception rate by 16.2%. The average NSPC under action research was lower (1.7) than that of the development intervention (2.36). The conception rate in action research was higher in the $2^{nd}$ parity, while the same was higher in the $2^{nd}$ parity under the development intervention. Appropriate timing of insemination in action research and development intervention was 10 to 15 hr. after the onset of estrus. However, conception was lower when insemination carried out after 15 hours of AI. In conclusion, the action research result revealed that the efficiency of development intervention carried out by

BoA significantly improved if proper synchronization and AI practices have to be place before the commencement of the program. Animal selection (appropriate age, BCS, and parity), breeding season, heat detection efficiency, and farmers' awareness to detect heat and on-time bringing of cattle for insemination, presence of AI technicians at station should be duly considered for effective synchronization and insemination.

ACKNOWLEDGEMENTS

This research was financially supported by the Livestock and Irrigation Value Chains of Ethiopian Smallholders (LIVES) project, ILRI, and the South Agricultural Research Institute (SARI). A special acknowledgment and credit goes to the principal project investigator, the late Dr. Azage Tegegne, for his restless effort and mentoring of the research work. The authors appreciate the integration and cooperation offered by experts working at Livestock and Fishery Development

Offices of the respective district to assist during the fieldwork. We highly recognize the AI technicians for their unreserved support and full involvement during the mass insemination and data collection. Finally, the authors fully acknowledge and appreciate the research participants, mainly cattle owners of the sample districts.