Artificial Insemination of Dairy Cows - Do Automated Systems Improve Reproductive Performance?

Introduction

In North America, the use of timed artificial insemination has become prevalent, and approximately 75% of US herds and 21% of Canadian herds utilize an estrus or ovulation synchronization program to inseminate their cows. The use of synchronization protocols, a method for timing and synchronizing ovulation through a series of hormonal treatments, has allowed farmers to control ovulation - the phase when an egg is released from the ovary and fertilization can occur. The protocols used for ovulation synchronization allow all eligible cows in the herd to be inseminated at the same time, while simultaneously decreasing labour resources and potentially economic costs.

While this practice is widespread, concern around the use of hormones and the subsequent effects on animal welfare and human health are rising. A potential alternative for at least some of these hormone interventions is the use of activity monitors, such as accelerometers and pedometers. These monitors can detect the expression of estrus - the phase of the reproductive cycle when cows are in ‘heat’ and ready to accept insemination. Using these monitors allow farmers to determine when to inseminate their cows, instead of relying entirely on timed artificial insemination. However, it is not yet clear what factors affect the efficiency of these monitors, and subsequently their effectiveness compared to timed artificial insemination protocols.

The aim of this study was to compare two reproductive programs - timed artificial insemination only and the use of activity monitors paired with timed artificial insemination - on pregnancy success. In addition, researchers were also interested in how animal health factors, such as body condition and leg health affect the detection of the expression of estrus by activity monitors. They also examined how these same factors affect the success of each reproductive program.

Research Process

The study took place between September 2012 and July 2014 at the University of British Columbia’s Dairy Education and Research Centre (farm A) as well as a local commercial dairy farm (farm B), both located in British Columbia. In total, the study included 918 Holstein dairy cows: 466 at farm A and 452 at farm B. Cows at both farms were producing similar levels of milk and had similar housing set ups. Cows at farm A were milked twice daily, whereas cows at farm B were milked three times daily, both in conventional milking parlours. During the study, all cows were examined for body condition and leg health twice, two weeks apart, and ovaries were examined for cyclicity on the same schedule (at the time of pre-synchronization injections). Pregnancy diagnosis was determined by ultrasound 36 (+/- 7) days after artificial insemination.

The figure below highlights the two different treatment protocols that cows were randomly assigned to. In the timed artificial insemination treatment (TAI), all cows underwent a full Presynch-Ovsynch protocol and were bred by timed artificial insemination. In the activity monitor treatment (ACT), all cows underwent the first half of the protocol (Presynch), and then were monitored for 12 days for the expression of estrus. If estrus was detected, cows would be bred by regular artificial insemination, but if after 12 days no estrus was detected, cows then underwent the second half of the protocol (Ovsynch) and were bred by timed artificial insemination. Activity monitors on both farms were checked twice daily for cows in estrus. 

Physical activity was continuously monitored on both farms and the threshold of activity to be considered an estrus event was determined by the maximum change in activity and duration of the estrus event. Activity data from farm A was used to check for the effects of body condition, leg health, parity and milk production on the expression of estrus. 

Results

The study found that pregnancy per artificial insemination did not differ between using activity monitors within a timed artificial insemination program compared with using only timed artificial insemination. However, while overall treatment did not affect pregnancy success, cows bred at timed AI were 1.48 times more likely to get pregnant than those bred at estrus. Interestingly, the days from parturition to conception were equal between treatments, suggesting that in the long term there was no advantage to using only the timed artificial insemination protocol.

There was an interaction between treatment and parity, where primiparous (first lactation) cows which were on the activity treatment had 2.36 greater odds of pregnancy than multiparous cows. Conversely, there was no difference in the odds of pregnancy between primiparous and multiparous cows in the timed artificial insemination treatment.

There were distinct differences in factors that affected pregnancy success between farms, suggesting that the use of activity monitors is greatly affected by individual farm variation. Factors affecting the success of pregnancy on farm A included parity, cyclicity, and body condition. Factors affecting pregnancy success on farm B were body condition, milk production, and leg health. Estrus detection also differed between farms, with 65.1% of cows detected in estrus at farm A and 34.4% at farm B. Farm also had an effect on overall pregnancy rates, along with parity and body condition.

This study also demonstrated that lameness (poor leg health) increases the days to conception and decreases the odds of pregnancy. However, both leg health and poor body condition had no effect on the maximum change in activity at estrus or the duration of estrus, as measured by the activity monitors. Primiparous cows had more intense estrus episodes. Cyclicity also affected estrous expression, with cows who began ovulating earlier after parturition having a greater chance of being detected in estrus.

Implications

Overall, the study showed that while pregnancy success was similar between both treatments, the success of activity monitors is prone to individual farm variations. Factors that negatively affect pregnancy success and days to conception, such as parity, cyclicity, and leg health, should be monitored and screened for, especially when considering the economic implications between the two systems. This study further illustrates that if farms can achieve good conception rates when inseminating cows at estrus, it is more profitable than completing a full timed artificial insemination protocol. The results presented here illustrate that in order to maximize both profit and reproductive performance on the farm, a combination of breeding at estrus through the use of activity monitors and timed artificial insemination is most effective.