Bradley Gotto, Undergraduate Student and Dr. Kirby Krogstad, Assistant Professor, Department of Animal Sciences, The Ohio State University
Introduction
Rumination is an important process in cattle and is associated with a high functioning rumen. As a cow ingests and chews its feed, the feed particles enter the rumen where microbial degradation begins. The feed will then be regurgitated for the cow to further chew and swallow again. The chewing of “the cud”, or the re-chewing of regurgitated feed, is what we call rumination. Rumination is impacted by numerous factors on the farm, including temperature-humidity index (THI), feed particle size, feed nutrient composition, and cow comfort.
Monitoring Rumination
Monitoring rumination is helpful for dairy farmers to detect illness, stress, or nutritional challenges in their herd. There are many methods used to monitor rumination. An “old school” approach to rumination monitoring is to conduct “cud counts”, where the number of cows ruminating in a pen are counted, with the goal being 60% of lying cows ruminating at any one time. With technological improvements in the dairy industry, activity collars, ear tags, and rumen boluses have become more common for monitoring rumination. These technologies track different data like head and jaw movements (collar or ear tags) to track time spent ruminating. Rumen boluses are administered orally and track data from inside the cow's reticulorumen.
An ongoing challenge for these technologies is to utilize the information generated by rumination monitoring technologies for economically impactful decisions. Our goal was to improve how to use and apply the data gathered from these rumination tracking technologies. We also wanted to understand whether rumination time and rumination variance were affected by different farm factors.
Our Study Objectives and Hypothesis
The objective was to determine associations of rumination time and rumination variance with environmental and dietary factors that were measured on the farm (Krauss Dairy Farm, Wooster, OH). Our initial hypothesis was that increasing THI would have a negative association with daily rumination time and would have a positive association with rumination variance. We expected that when performing the Penn State Particle Separator (PSPS) tests on each pen’s ration, an increased amount of the ration in the 19 mm sieve would be negatively associated with daily rumination time, and an increased amount of the ration in the 8 mm sieve would have a positive association on the daily rumination time. We also expected that an increased variance of particle size across the bunk would be positively associated with rumination variance measured for each pen.
Methodology
From 5/12/2025 to 7/31/2025, we gathered data from a collar-based rumination monitoring system (SCR, Allflex US, Madison, WI) at Krauss Dairy Farm. The animals were housed in a 4-row freestall barn with a 1:1 bed to cow stocking density, providing a non-competitive environment for the animals to access feed and lay down. The farm milks approximately 190 cows twice daily and feeds once daily. The Holstein and Jersey cows were housed in different barns.
Daily pen averages for rumination time, rumination variance, days in milk, milk yield, and activity data were collected from the herd management system. Daily pen dry matter intake (DMI), feed delivery time, and feed refusal rate were collected from the farm’s feeding program. Every Monday, footage from time lapse cameras in each barn were collected to analyze out of feed time and feed pushup occurrence. The THI data were collected from sensors placed in the center of each freestall barn. Twice each week, the PSPS test was performed on 3 locations across the feed bunk for each pen. The PSPS was also used to assess the particle size of each forage fed in the milk cow ration (corn silage, haylage, and baleage). Temperatures were taken from 5 locations across the feed bunk for each pen using an infrared temperature sensor. Temperatures were also taken from 5 locations on each silage face (corn silage and haylage).
We then calculated Pearson’s correlation coefficients for rumination time and rumination variance with each farm factor to identify potential factors that are related to rumination time and rumination variance.
Correlations
Correlations are a simple statistical method that we used to assess the strength of the relationships between different variables. The greater the correlation is between two factors, the stronger the relationship is. A positive correlation signifies that as one factor increases, the other will also increase. A negative correlation signifies that as one factor increases, the other will decrease. As the value approaches 1.0, the strength of the correlation is increasing. Table 1 lists the cutoffs we used for describing correlations.
Table 1. Strengths of correlation and adjectives used to describe the correlation.
| Correlation | Description |
| 1.00 | Perfect |
| 0.70-0.99 | Very High |
| 0.50-0.69 | Substantial |
| 0.30-0.49 | Moderate |
| 0.10-0.29 | Low |
| 0.01-0.09 | Negligible |
Results
Rumination is Higher in Holstein Cows
Through the summer, we observed that rumination time was greater in Holstein than Jersey cows. Pen 1, the early lactation Holstein cows, averaged 535 minutes/day of rumination. Pen 4, the late lactation Holstein cows, averaged 559 minutes/day of rumination. Pen 5, the peak lactation Holstein cows, averaged 546 minutes/day of rumination. Pen 8, the first lactation Holstein cows, averaged 526 minutes/day of rumination. Pen 21, the Jersey cows, averaged 418 minutes/day of rumination. Figure 1 shows the daily average rumination time for each group throughout the data collection period.
Figure 1. Average daily rumination time by pen at Krauss Dairy Farm.
Rumination Variance Differs by Management Group
Rumination variance was greatest in pen 1, averaging 98 min/day. Pen 5 was the next highest, averaging 76 min/day, and pen 8 averaged 68 min/day. Pen 4 averaged only 52 min/day. Pen 21, the Jersey cows, averaged 52 min/day. Figure 2A and Figure 2B show the daily rumination variance on a min/day and as a percentage of total rumination for each pen.
Figure 2. Average daily rumination variance for each pen. Panel A is in variance on a min/day, while panel B displays variance as a proportion of the total rumination time.
Rumination Time is Associated with Greater Milk Production
When analyzing milk yield versus rumination time, we observed that a greater rumination time is associated with a greater milk yield. Our data suggest that for every 33 minutes/day of additional rumination time, we would expect a 1.0 lb/day increase in milk yield (Figure 3). The relationship between milk yield and rumination time was strongest in the early and peak lactation groups. Alternatively, when analyzing milk yield versus rumination variance, we observed that a greater rumination variance was associated with reduced milk yield. Our data suggest that for every 3 minute/day increase in rumination variance, we expect a 1.0 lb/day decrease in milk production (Figure 4).
Figure 3. Increasing rumination time is associated with increased milk yields.
Figure 4. Increasing rumination variance is associated with reduced milk yields.
Correlations of Rumination with Management Factors
When analyzing the correlations, a substantial positive correlation exists between rumination time and milk yield (r = 0.64, P = <0.001; Table 2). We also observed a very high positive correlation between dry matter intake (DMI) and rumination time (r = 0.78, P = <0.001). A very high negative correlation exists between activity and rumination time (r = -0.85, P = <0.001). Conversely, there was a moderate negative correlation between rumination variance and days in milk (r = -0.47, P = <0.001; Table 3).
Table 2. Correlations of daily pen-level production variables with rumination time (CI = confidence interval).
| Variable | Correlation | 95% CI | P-value |
| Days in milk | 0.15 | 0.05-0.24 | 0.004 |
| Activity | -0.85 | -0.87- -0.82 | <0.001 |
| DMI | 0.78 | 0.75-0.82 | <0.001 |
| Milk | 0.64 | 0.58-0.69 | <0.001 |
| FE | 0.26 | 0.17-0.35 | <0.001 |
| Adj. FE | 0.40 | 0.32-0.48 | <0.001 |
| Refusal rate | 0.12 | 0.03-0.22 | 0.01 |
| Out of feed time | -0.13 | -0.23- -0.01 | 0.02 |
Table 3. Correlations of daily pen-level production variables with rumination variance (CI = confidence interval).
| Variable | Correlation | 95% CI | P-value |
| Days in milk | -0.47 | -0.55 - -0.39 | <0.001 |
| Activity | -0.09 | -0.19-0.00 | 0.06 |
| DMI | -0.12 | -0.22- -0.02 | 0.01 |
| Milk | 0.18 | 0.09-0.28 | <0.001 |
| FE | 0.33 | 0.24-0.42 | <0.001 |
| Adj. FE | 0.14 | 0.04-0.23 | 0.006 |
| Refusal rate | 0.0 | -0.10-0.10 | 0.93 |
| Out of feed time | -0.21 | -0.30- -0.10 | <0.001 |
Correlations of Rumination with Dietary and Environmental Factors
Upon evaluating correlations between rumination time and dietary characteristics, a moderate negative correlation was found between the feed retained on the 8 mm (%), and rumination time (r = -0.43, P = <0.001; Table 4). We also observed moderate positive correlation between 8 mm (%), and rumination variance (r = 0.37, P = <0.001; Table 5) and a moderate negative correlation between 19 mm (%), and rumination variance (r = -0.41, P = <0.001)
Table 4. Correlations of daily pen-level dietary variables with rumination time.
| Variable | Correlation | 95% CI | P-value |
| 19 mm, % | 0.14 | -0.9-0.35 | 0.25 |
| 8 mm, % | -0.43 | -0.60- -0.22 | <0.001 |
| 4 mm, % | -0.28 | -0.48- -0.06 | 0.01 |
| Pan, % | 0.40 | 0.19-0.58 | <0.001 |
| 19 mm CV | 0.07 | -0.16-0.29 | 0.60 |
| 8 mm CV | 0.07 | 0.00-0.43 | 0.05 |
| 4 mm CV | 0.11 | -0.12-0.33 | 0.33 |
| Pan CV | 0.13 | -0.10-0.35 | 0.27 |
| Feed bunk temperature | 0.03 | -0.20-0.26 | 0.79 |
Table 5. Correlations of daily pen-level dietary variables with rumination variance.
| Variable | Correlation | 95% CI | P-value |
| 19 mm, % | -0.41 | -0.59- -0.19 | <0.001 |
| 8 mm, % | 0.37 | 0.14-0.55 | <0.001 |
| 4 mm, % | 0.04 | -0.19-0.27 | 0.76 |
| Pan, % | 0.23 | 0.00-0.44 | 0.06 |
| 19 mm CV | 0.06 | -0.18-0.28 | 0.65 |
| 8 mm CV | -0.08 | -0.31-0.16 | 0.54 |
| 4 mm CV | 0.10 | -0.14-0.32 | 0.44 |
| Pan CV | 0.12 | -0.11-0.35 | 0.31 |
| Feed bunk temperature | 0.30 | 0.06-0.51 | 0.01 |
We did not observe significant correlations between environmental variables with rumination time (Table 6). There was a low-to-moderate positive correlation between rumination variance and THI. As THI increased, rumination variance also increased (Table 7).
Table 6. Correlations of daily pen-level environmental variables with rumination time.
| Variable | Correlation | 95% CI | P-value |
| Average temperature | 0.07 | -0.02-0.17 | 0.17 |
| Minimum temperature | 0.06 | -0.03-0.15 | 0.22 |
| Maximum temperature | 0.09 | -0.01-0.19 | 0.07 |
| THI | 0.06 | -0.03-0.16 | 0.21 |
| Minimum THI | 0.04 | -0.06-0.14 | 0.44 |
| Maximum THI | 0.08 | -0.01-0.17 | 0.10 |
Table 7. Correlations of daily pen-level environmental variables with rumination variance.
| Variable | Correlation | 95% CI | P-value |
| Average temperature | 0.29 | 0.20-0.38 | <0.001 |
| Minimum temperature | 0.27 | 0.18-0.36 | <0.001 |
| Maximum temperature | 0.28 | 0.19-0.37 | <0.001 |
| THI | 0.31 | 0.22-0.40 | <0.001 |
| Minimum THI | 0.30 | 0.20-0.38 | <0.001 |
| Maximum THI | 0.31 | 0.22-0.40 | <0.001 |
Conclusions
We observed that diet particle size may be associated with rumination time, but it was not as we expected. The 8 mm screen of the PSPS was negatively associated with rumination time (r = -0.43, P = <0.001) and positively associated with rumination variance (r = 0.37, P = <0.001). We suspect that this may be related to how our cows were grouped and the diets they were fed. Pen 4 was fed a high amount of baleage in their ration, which reduced the proportion of feed retained on the 8 mm sieve and increased the proportion on the 19 mm sieve, while also being the pen with the highest rumination time and rumination variance.
Perhaps our most intriguing finding was that the environmental factors had negligible correlations with rumination time. However, we observed that as THI and temperature increased, rumination variance also increased. Perhaps, rumination variance may be used as an indicator of heat stress on dairy farms.
Rumination time and variance are associated with milk production, and this can be helpful to look at while managing your herd. Increased rumination variance may suggest that something is wrong with the cows and that you may be leaving milk on the table. While rumination time and variance are useful data points, they should not be used or viewed in isolation. They should be used within the constellation of other data points at your disposal to determine what may be going on with a group of cows.
Works Cited
Antanaitis, R., et al. The impacts of heat stress on rumination, drinking, and locomotory behavior, as registered by innovative technologies, and acid-base balance in fresh multiparous dairy cows. Animals: An Open Access Journal from MDPI, U.S. National Library of Medicine, 13 Apr. 2024, pmc.ncbi.nlm.nih.gov/articles/PMC11047379/.
Grant, R.J. Chewing behavior of dairy cows: Practical perspectives on ..., www.appliedanimalscience.org/article/S2590-2865(23)00027-7/pdf. Accessed 11 Aug. 2025.
MacPherson, L. The benefits of rumination monitoring: Helping farmers in Scotland. FAS, 19 Dec. 2023, www.fas.scot/article/the-benefits-of-rumination-monitoring/.