Attaining Reproductive Solutions through Activity and Health Monitoring

For a variety of reasons estrus detection rates in North American dairy herds based solely on visual observations are usually near the 50 percent range. Most large dairy operations use tail chalk/paint and once daily observation which is usually superior to visual observations but still allows room for improvement with service rates typically in the mid-60 percent range. Low detection rates led to the successful and profitable adoption of systematic timed A.I. programs (see dcrcouncil.org for industry recommended protocols). Despite the effectiveness of timed A.I. programs, many producers desire to breed cows based on estrus behavior and have mentioned their dislike of frequent injections that are required with timed A.I. protocols or concern about public perception with the use of reproductive hormones.

Increased physical activity is recognized as being associated with estrus, and various automated systems have been developed to detect standing to be mounted or increased activity either as steps or neck movements (Nebel et al., 2000; Firk et al., 2002). Numerous research reports reveal ample evidence that activity systems (pedometers or accelerometers) are able to accurately identify the majority of cow/heifers that are in estrus (van Erdenburg, 2008; Hockey et al., 2010; Kamphuis et al., 2012; Lovendahl and Changunda, 2010; Neves et al., 2012). As a result of technical progress in monitoring cows with the use of computers, automated detection systems have become a practical reality. Results of efficiency and accuracy of detection varied depending on the threshold value to determine when to declare high activity, the number of cows studied, housing and cow comfort, the method of time series analysis, and the number of days and time comparison for establishing a baseline of activity to determine when to declare high activity that should be associated with estrus. Across numerous herds using activity systems that we have evaluated the detection rates obtained range from the mid-70 to 80 percent range.

Affordable advanced computer technology that includes mini printed circuit boards that contain microelectronic circuits that function as an on-board micro-computer that tracks and transmits data, either using radio frequency or infrared technology, to either an on-farm computer based or Web based software package has driven an explosion of available activity systems. Activity monitoring systems allow for individual cow management with unique data collection and interpretation practically in real time. Initially, systems were developed for the detection of estrus but today systems are available that monitor rumination, resting time, temperature, and many other events associated with animal well-being. Activity monitoring has many different approaches, from pedometers that measure walking activity, to accelerometers that measure head movements, and ear tags that monitor activity associated with estrus and rumination and inner ear temperature. Proprietary complex algorithms allow comparison of both individual baseline and for a few systems a group baseline to identify individuals that deviate from normal or expected levels of activity to determine which animals are outside the desired population confidence interval and require management attention and or action. Each system (sensor) records different types of activity. This can greatly affect accuracy. One of the key differentiators between systems is that their level of accuracy and false positives of detection algorithms.

All systems include three or more basic components, the sensor on each cow, the hardware receiver to collect the data from the sensors and computer software. Presently sensors are in the form of either an ankle mounted pedometer, collar mounted monitor, an ear tag, or a rump mounted transmitter. All sensors transfer data either wirelessly using radio frequency or infrared technology to some configuration of a reader that transmits the data, usually in binary code, to a coordinator that translates and decodes the signal. The software is either located on an on-farm computer or a server that receives the information via Web based technology where the proprietary algorithms sort the information and determine which individuals need attention. Web apps, email alerts, and smart phone program downloads are available with many of the current systems. Table 1 summarizes the systems available in North America.

Farris (1954) was the first to report using pedometers to measure activity associated with estrus in dairy cows. The data on six cows that had AM and PM pedometer data showed an average increase in activity during estrus of 218 percent. More than twenty years elapsed before research revived interest in pedometry as a practical tool for detection of estrus of dairy cattle (Kiddy, 1977). It was noted that the daily activity for each cow must be monitored and activity associated with estrus compared to that obtained during the other stages of the estrous cycle for pedometry to be most effective in identifying estrus. A second significant finding was that individual cows differed significantly in the amount of activity expressed under the same conditions. The average increase in activity at the time of estrus was 393 percent. For 93 percent of the estrus periods the activity was three standard deviations above the mean activity during diestrus. These two studies were the basis for the development of activity systems. Over the next thirty-five years, numerous scientific studies reported on various properties of activity systems from environmental factors that affect accuracy rates to the ideal timing of insemination. Activity systems transfer data either during milking, using a walk-through portal or with a reader at each stall, or periodically during the day using radio frequency technology. The latest systems are using Internet Cloud service to alleviate the necessity for on-farm software and allow for remote access from almost anywhere.

In general, activity data, however it is measured, is examined mathematically to identify cows that deviate from a pre-determined baseline. High or low activity is a reflection of the baseline activity for each cow and some systems account for routine herd movement. Results are available for individual cows and lead to three possible conclusions: cows in estrus, cows that need examination for other signs of estrus, or no action needed. The major limiting factors with activity systems are the type and level of activity used in the mathematical equations that are employed that determine detection and accuracy rates. It is well known that individual cows differ significantly in the amount of activity expressed during estrus even under the same farm conditions. All systems determine an individual baseline of activity that is used to compare current activity to decide if the real-time information is above a preset threshold. This is where the mathematical equations become important to declare the cow in high activity. An adjustment in the threshold level requires a balance between high detection rates and low error rates. All systems have a default threshold level based on the research for that system. After a period of time threshold levels should be adjusted according to specific farm conditions such as size of the pens or corrals, number of milkings per day, walking distance from the barn to the parlor and other activities. The recording of cows not detected by the system and cows detected but not believed to actually be in estrus, is important for establishing an accurate threshold that should optimize service rates and minimize error rates.

In many studies different traits have been analyzed for inclusion in activity systems for the detection of estrus. Results of estrus detection varied depending on the used threshold value, the number of cows, housing and treatment of cows and the methods of time series analysis. The detection rate of most investigations is sufficiently high at 80 to 90 percent. Error rates between 17 and 55 percent and specificities between 96 and 98 percent indicate a large number of false positive high activity alerts. A primary goal of most activity systems was to reduce the false positive alerts. In recent years several systems have combined different traits with the objective of improving detection rates. De Mol and Woldt (2001) stated that a detection rate of 100 percent in combination with no false estrus alerts on the basis of computer aided management devices will not be possible.

The basic training of new users and continuous support of farm personnel are important to maximize the benefits of an activity system. Experience and training to differentiate a “true” peak of activity caused by a cow in estrus from a “false” peak in activity due to other causes for an increase in activity is closely related to improve reproductive performance, and there is a learning curve with all systems. Getting cows pregnant is complex and requires a comprehensive approach. Reproductive success is measured by confirmed pregnancies. One of the first responses that are noticed by producers is increased palpation pregnancy rates or more pregnant cow at herd checks. With the growing public interest in the living conditions of food producing animals, and the growing objection to any unnecessary use of pharmaceuticals, activity systems have the potential to be useful tools in the reproductive management of dairy herds. There are a growing number of producers who have purchased and use an activity system as a vital component of their reproductive management program. The real proof on how a system improves the reproductive performance is visiting dairy operations or questioning producers who use a system to learn from their experiences.

The costs of activity monitoring systems vary considerably depending on the system, operation needs, barn size, need for a computer, and number of monitors needed. For instance if a producer is just looking to use activity monitoring for estrus detection, monitors can be switched out and only used in 40 to 50 percent of the herd. Producers should compare the cost of activity monitoring to the results of their current reproductive system to help understand how it could impact profitability. The biggest consideration is that these systems are a tool to assist in decision making and should minimize labor associated with identifying cows in heat. There will continue to be enhancements to motion-sensing and health monitoring systems, so it is important the select the system that best fits your objectives. There are some options that are outdated and will not provide enough accuracy to rely on as a sole means of estrus detection. Conversely, there are some options that have not been thoroughly tested in the wide variety of North American dairy operations and may not be a good fit for your operation.

Deciding whether to implement an activity/health monitoring system into the health and reproductive management tool-box, as well as which type of system, is an important decision with several factors to consider. The investment of an activity/health monitoring system will have a different level of return on that investment for each operation. If you are going to invest in an activity system, it is critical that you work with someone who has the expertise to help you manage the system properly including how to best utilize estrous synchronization protocols in combination with the activity system technology to maximize your results. There will be a percentage of cows that will need some hormone intervention to assist them in overcoming anestrous status and a high-performing activity system that provides accurate estrus detection will help you identify the cows that are cycling, so you can focus individual intervention on the cows that need it. That is why it is critical to purchase an activity system from a reproductive specialist that can maximize the investment in the same way that a milking equipment specialist can help maximize milk harvest with the investment in new milking equipment, so work with specialists within their field of expertise. 

De Mol, R.M., Woldt, W.E., 2001. Application of fuzzy logic in automated cow status monitoring. J. Dairy Sci. 84, 400–410.

Farris, E.J. 1954. Activity of dairy cows during estrus. J. American. Vet. Med. Association. 125:117-120.

Firk, R., E. Stamer, W. Junge, and J. Krieter. 2002. Automation of oestrus detection in dairy cows: A review. Livest. Prod. Sci. 75:219-232.

Hockey, C., J. Morton, S. Norman, and M. McGowan. 2010. Evaluation of a neck mounted 2- hourly activity meter system for detecting cows about to ovulate in two paddock-based Australian dairy herds. Reproduction in Domestic Animals. 45:e107-e117.

Kamphuis, C., B. DelaRue, C.R. Burke, J. Jago. 2012. Field evaluation of 2 collar-mounted activity meters for detecting cows in estrus on a large pasture-grazed dairy farm. J. Dairy Sci. 95:3045-3056.

Kiddy, C. A.1977. Variation in physical activity as an indication of estrus in dairy cows. J. Dairy Sci. 60:235-243.

Lovendahl, P. and M.G.G. Chagunda. 2010. On the use of physical activity monitoring for estrus detection in dairy cows. J. Dairy Sci. 93:249-259.

Nebel, R.L., M.G.Dransfield, S.M. Jobst, and J.H. Bane. 2000. Automated electronic systems for the detection of oestrus and timing of AI in cattle. Animal Reprod. Sci. 60-61:713-723.

Neves, R. C., K.E. Lesile, J.S. Walton, and S.J. LeBlanc. 2012. Reproductive performance with an automated activity monitoring system versus a synchronization breeding program. J. Dairy Sci. 95:5683-5264.

Van Eerdenburg, F.J.C.M. 2008. The pedometer, an automated aid in the detection of estrus. Vet. Quarterly. 30 (Suppl 1):49-57.