In the Dairy Cattle Biology and Management Lab at Cornell University, rumination and activity monitors were implemented by Matias Stangaferro, DVM, M.S. and Ph.D. student, and Julio Giordano, DVM, M.S., Ph.D., to study and identify health disorders in dairy cows.
“Early postpartum disorders in dairy cattle have been associated with significant economical losses for the dairy industry,” said Stangaferro during a recent webinar presented about the research. “For that reason, many dairy farms have developed and implemented intensive health monitoring programs, which are usually costly, time-consuming and require qualified labor.”
However, there is substantial variation across farms in the U.S. in how this is accomplished. According to Stangaferro, new technologies such as rumination and activity monitors could be used to reduce or eliminate the burden associated with intensive health monitoring programs and help to identify cows with health disorders earlier than traditional methods.
Monitoring systems available in the U.S. include the SCR Heatime HR System neck collar, the CowManager eartag and the DVM TempTrack rumen bolus. Normally these products are used for reproductive management but can also be used for health monitoring because they record data on individual cow rumination, activity, body temperature and feeding time.
Previous studies have shown how disease onset and the onset of calving correlate with changes tracked by these monitoring systems, explained Stangaferro. However, there is little information about the ability of a particular system to identify cows with health disorders and the rumination and activity patterns of those cows suffering from health disorders.
For their research, Stangaferro and Giordano enrolled 1,121 commercial dairy cows from November 2013 to October 2014 and used the SCR Heatime HR system to track every cow. Data was collected from each cow every two hours from four weeks before calving until 80 days in milk. The SCR system records rumination and activity data, and uses internal algorithms to calculate a health index (HI) score ranging from 0 to 100. If a cow’s HI drops below 86, the cow is flagged by the system.
During the research trial, Stangaferro received a report every day by email that showed which cows had a HI less than 86. This information was not shared with the farm personnel who made clinical diagnoses for the study.
Stangaferro found that the HR system had a 98 percent rate of detecting cows with a displaced abomasum and that those cows were flagged by the system three days earlier than clinical diagnosis by farm personnel. The HR system also had a 91 percent rate of detecting ketotic cows 1.5 days earlier than clinical diagnosis and was 90 percent accurate at detecting indigestion half a day earlier than farm personnel.
Stangaferro next evaluated why some cows were found by farm personnel but not flagged by the system. His data showed that the milk production, rumination and activity scores of cows not flagged by the system were similar to healthy cows. However, cows flagged by the system decreased in milk production and had lower rumination, activity and HI scores compared with healthy cows.
The HR system had a moderate ability to detect metritis and mastitis at a rate of 55 percent, flagging cows approximately one day before the clinical diagnosis by farm personnel. To figure out why almost half the cows diagnosed with metritis and mastitis were not flagged by the system, Stangaferro assessed the severity of each case based on the treatment protocol used by farm personnel. This information showed that the most severe cases were the ones flagged by the HR system.
“In conclusion, we can say that the HR system from SCR was most effective for identifying cows suffering from metabolic and digestive disorders, with a sensitivity of more than 90 percent,” Stangaferro said. “A relatively lower sensitivity to identifying cows with mastitis can be explained by less severe illnesses and the type of mastitis-causing pathogen.”
According to Stangaferro, this study generates new opportunities for earlier detection and treatment, which can improve the animal’s response and prevent associated health disorders. However, Stangaferro said using the system is a challenge because treatment decisions must be made based on the absence of clinical signs because low HI scores do not tell which specific disorder farm personnel should treat for.
Stangaferro said implementing an automated monitoring system could reduce time, labor and antibiotic treatment costs because only cows with an HI score below 86 would be checked. And farms with little to no ability to check cows could use this system to help identify more cows with health disorders and improve treatment, thus improving animal welfare.
The goal of Stangaferro’s next phase of research is to use more data such as milk components, somatic cell count, cow age and stage of lactation to create algorithms to identify and treat diseases more effectively. PD
The webinar from which this information was originally presented is available at the Cornell University website.
Audrey Schmitz is a student at Kansas State University, studying agricultural communications and journalism.