Cow Connectivity - getting closer to your girls

Cow wearables, such as cow collars and ear sensors, are being used more and more on New Zealand dairy farms to help monitor and manage animal performance, health and fertility.

When small herd sizes were the norm, farmers knew each of their girls well. With increases in herd sizes over the last 10 years however, we now need to utilise technology more to help us improve the lives and wellbeing of both farmers and their cows.

Many of us wear Fitbits or smart watches, paired to our phones, thus increasing our connectivity and monitoring of our activity and health. Cow wearables now allow us to become more connected to cows, helping us to understand their behaviour, enabling earlier detection of any health issues and increasing our awareness of herd trends.


Key benefits of cow collars and ear sensors include:

  • Understanding cow’s eating patterns and rumination
  • Knowing cow behaviour, movements and location, helping to detect issues early
  • Heat detection and improved mating results
  • Heat detection materials no longer required
  • Improved pasture management
  • Automated herd movements, saving time
  • Improving cow quality and performance
  • Optimised herd productivity
  • Enabling well informed herd management decisions
  • Reduce on-farm workload, helping improve farmer work-life balance


Here are some real-life case studies from New Zealand dairy farms highlighting how cow wearables have helped.

Case 1: C-Section in a Cow wearing a Cow Collar

At 5am on the 7th of November, Cow 10 flagged up a distress signal, this is when the cow collar noted abnormal behaviour, indicating she was potentially calving. A calf’s head was presented at 10am, and when no progress was made a vet was called and a C-Section was started (the black arrow on the graph below). From this point her rumination and activity dropped sharply till 5am on the 8th of November, at which point it began to track upwards. 12 hours after surgery her rumination and activity had returned to close to normal (blue arrow on the graph below). 72 hours after surgery her eating had returned to normal (orange arrow on the graph below).

The availability of this level of monitoring, combined with the farmer’s observations enabled us to support the recovery with a greater degree of accuracy. Utilising the base line of activity, rumination rate and eating time, any deviation from normal was seen as a ‘red flag’ – pain, infection or other complications. This was a useful in determining if intervention or re-examination was required to make sure the clients received the best outcome for the animal and that the animal had a great recovery.


Case 2: When the vet can’t find anything on a cow wearing a Cow collar

A vet was called out to examine a cow that was “off colour” (black arrow on the dashboard below). She had previously had a ‘distress’ alert 5 days ago and gone “down”. On the vet exam, we couldn’t find anything, but looking at the collar data you could see was following a similar trend in lower total rumination minutes to the last time she went “down” (black oval on the dashboard below) and that her rumination since calving was erratic and she had some distress alerts around calving (orange oval on the dashboard below).

She was given a calcium bolus, a starter drench and placed on once-a-day milking before the blood results were received. The blood sample showed a low blood Magnesium level. From the collar data below, you can see the point of intervention (black arrow on the dashboard below) and the return of her rumination rate (orange oval on the dashboard below) and her cycling.


Case 3: Using Cow collar information for more than non-cyclers, phantom cow scanning and sick cows…

A farm that usually does AI for 10 weeks (6 weeks of Premium pack semen, 4 weeks short gestation Beef Semen), decided to put Premium Pack semen into the top 50% of the herd (Production Worth was the unit of measure) and use short gestation Beef straws on the bottom half of the herd. These cows were put into two groups “Milking cows” and “Milking AI Charolais”.  The top PW cows who were identified as non-cycling premating (utilising the Allflex parameters, >50 days in lactation from calving and no heat event prior to PSM) underwent a non-cycler program at the start of mating. The non-cycler cow criteria was then reduced 4.5 weeks into mating to cows calved >35 days and no heat event shown.

Table 1, below, gives an idea of what the list of non-cycling cows looks like in their Groups (Milking cows and Milking AI Charolaise), Days in Lactation, Lactation Status and Age in Months.

After two rounds of non-cycling cows, phantom cow scanning took place 10 days out from the end of AB using the Allflex pregnancy probability report. At this stage we treated the phantom cows found depending on the value seen by the farmer.

Table 2, below, shows a list of cows that could be pregnant as they have not had a heat for >27 days. This list can be used for phantom cow scans and for pregnancy probability status.

This year the farm's phantom rate was 11%, with a conception rate to phantom cow treatment of 63%. The 1st round of treated non-cyclers produced a conception rate of 60%, while the second round of treated non-cyclers had a conception rate of 50%. The overall results was a 9% MT rate (with 3% showing up as rechecks/suspicious as of March 2022).

We went on to review the MT cows in the system and found that 7 out of the 13 MT cows had transition period issues which was a red flag to us (Transition issues occur -14 days from calving to +30 after calving and show up as distress alerts, health alerts, lumpy or slow to rise rumination rates in the system).

On review of the transition period we found the herds daily rumination rates were low (orange line in the graph below) and there was a decreasing milk protein percent (blue line in the graph below) from the 24th of July through to the 24th of October.

A focus for next year will be the transition period on farm and how we manage individual cows who can’t handle their transitioning but also monitoring the herd and its nutrition over this period, along with metabolic testing.


Case 4: Using wearables data to review reproduction & create a proactive plan

The focus of this case study will be to present aspects of a Reproduction review that was undertaken on a farm using Cow Manager.

Key findings:

1. 59 cows (14% of the herd) flagged as having a heat but did not receive an insemination to that heat during the mating period. This information is found on the ‘Cycling cows, not inseminated report’.

2. 35 cows (8% of the herd) were inseminated without a heat during the AB period. Of these, 26 cows (6% of the herd) were non-cycling cows undergoing hormonal intervention and 9 cows (2% of the herd) did not have a heat but were inseminated. This information will be found in the Fertility Insights section when launched.

3. The 2013 (8 year old), 2014 (7 year old) and 2018 (3 year old) born animals performed poorly compared to the rest of the age groups. They were slower to get in calf (had a wider calving to calving interval) and had a higher MT rate.


Graph 1, below, is the pregnancy rate of the age groups (2013, 2014 & 2018 are highlighted in the legend below the graph).

Table 1, below, shows the cow age, number of animals in each group and their expected calving interval this year. A perfect cow will have a calving-to-calving interval of 365 days (+/- 14 days). Below you can see the 3-year-olds and the 7-year-olds have a higher calving-to-calving interval, and took roughly 60 days extra to get in calf.

*this data can be found by adding the ‘Expected calving interval’ to the ‘All Cows’ tab and exporting it into Excel.

4. The transition period (-25 days to +25 days) was flagged as contributing to the poor in-calf rate of the 2018 born animals. This analysis can be found under the Nutrition module > Behaviour around Event (the event being calving) and Group comparison.

Graph 2, below, looks at the 2018 born animals rumination (blue line) and eating (green line) hours per day around their calving event (Day 0). Pre calving the rumination hours are a lot lower than expected and variable, post calving rumination hours and eating hours are very low. Poor transition period management can lead to increased health issues, cows take longer to start cycling and can produce less milk.

Graph 3, below, shows the 2018 & 2019 born animals rumination hours per day in early lactation (July through to the end of September) compared to the rest of the herd. What you can see is the gradual increase compared to the herd. Feeding in early lactation of the younger cows needs to be monitored to make sure they transition well, and the feed allocated to them is meeting their physiological needs.

Graph 4, below shows the transition period of Cow 444, as an example. The black line is her average rumination + eating hours per day, the red line is the average rumination and eating hours per day for the herd, and the shaded area is the standard deviation of the herd. She was a heifer that calved on the 4th of August and had her first heat 64 days after calving, from which she cycled regularly but was MT. This is an example of an animal that did not transition well (compared to the herd).

Graph 5, below, shows the lactating cows average rumination minutes per day (blue bars) and the bulk milk protein percentage (green line) from July 21 to March 22. There is a significant drop in protein percentage (feed quality or feed quantity) and a matching low rumination rate (>450 minutes is ideal for this farm).

Action point summary for our client:

  • Dry cow monitoring of rumination hours and eating hours will take place 30 days out from calving. Work in with the farm consultant and the vets if < 5 hours per day ruminating and < 6 hours per day eating.
  • Day 0 to Day 4 post calving monitoring of rumination hours. Target a rumination rate of > 6 hours per day per cow. If not reaching this consider the options discussed with your vet on the flow chart created for the farm.
  • Suspicious and Sick cow alerts in the calving period. Utilise the flow chart developed with your vet to triage and provide effective treatment to cows that flag up on the Cow Manager System.
  • Monitoring whole herd rumination hours and eating hours in Early Lactation. This season target keeping rumination hours >7 hours per day. Work in with farm consultant and the vets if this isn’t being met.