The adoption of automated milking systems in the Canadian dairy industry remains strong. With that type of milking management, we also have a shift in nutritional management on those farms, with the division of the diet into a partial mixed ration (PMR) and some type of nutritional provision (often in the form of a pellet) at the automated milking system.

Devries trevor
Professor / University of Guelph
Trevor DeVries is a professor and Canada Research Chair in the department of animal biosciences a...

The feed provided at the automated milking system is primarily designed to attract cows to milk; however, there is also potential to vary the amount and composition of that feed to meet the individual nutritional requirements of the cows.

There has been much variation in the dairy industry in how nutritional management occurs on automated milking system farms. In a recent large benchmarking study of automated milking system farms (n=160) from across Canada, we aimed to quantify that variation, describing the diets fed on those farms and also detecting associations of nutritional practices with productive metrics from those farms. 

Our study included farms from Western Canada (n=54), Ontario (n=76), Quebec (n=22), and Atlantic Canada (n=8). Given this distribution and the fact that growing conditions and available feeds vary across the country, we observed high variability in the composition of diets fed on these farms. Across the farms studied, 60% used corn silage as their main source of forage in their PMR. Various types of haylage were the second most common major forage source. In the West, the second most common major forage was barley silage, which is common in the prairie provinces. Corn was the primary energy concentrate ingredient used in the formulated PMR. Fourteen percent of farms nationally had barley grain as their primary concentrate source, but of those farms, 91% were in the West. 

Some commonality was observed in feeding practices at the automated milking systems. The majority (77%) of farms across the country fed a pelleted concentrate within the automated milking system, with the remaining farms feeding some type of texturized feed mixture. This is likely related to the fact that pelleted feed is easier to handle, may have less shrink and is consumed at a faster rate than other forms of concentrate. 

Advertisement

Within the study population, nearly half (45%) of farms used wheat (mostly wheat shorts) as the primary ingredient within their automated milking system concentrate. Corn was the second most common primary automated milking system concentrate ingredient, while barley was the third most common (again nearly exclusively fed in the West).

In terms of primary protein source within the automated milking system concentrate, soybean meal was the most common, while canola meal and dry distillers grains were the next most common. These uses, again, are likely influenced by regional availability, as most of the farms feeding canola meal were from the West. 

Based on the formulated diets, farms in the Atlantic provinces and Ontario fed for the greatest amount of PMR dry matter intake (DMI) (at 22.8 and 21.5 kilograms per day DM, respectively), and the least amount of automated milking system concentrate (at 4.1 and 3.9 kilograms per day DM, respectively). Conversely, farms in the Western provinces and Quebec formulated for lower PMR DMI (at 20.5 and 20.6 kilograms per day, respectively) and higher DMI at the automated milking system (at 4.8 and 4.7 kilograms per day DM, respectively). No regional differences were observed for total formulated DMI (averaging 25.4 kilograms per day DM). Interestingly, across regions, the average predicted automated milking system concentrate DMI, based on the formulated diets, was 4.3 kilograms per day DM. This is slightly lower than some other previous observational studies of automated milking system farms in North America.

The average milking frequency across our Canadian study farms was 2.8 milkings per day, while the average milking visit length was 7.1 minutes. If we do the math, that would mean that the average expected consumption rate of automated milking system concentrate would be around 216 grams per minute DM. That fits well with published values, which suggest that the average dairy cow can consume concentrate at 200 to 300 grams per minute, with higher consumption rates possible (e.g., for higher-producing, higher-intake cows).

In our study, herd-average production was 37 kilograms per day; within that average, the Holstein herds (87.5% of farms) produced on average 38 kilograms per day, while non-Holstein herds (12.5% of farms) produced on average 31 kilograms per day. Interestingly, for Holstein herds, this average production level is higher than the Canadian average for Holstein herds across milking systems, as well as other recent benchmarking studies of North American automated milking system herds.

Given the amount of variation in dietary composition, as well as associated nutrient composition, of the diets fed on all these farms across Canada, it is not surprising that we only detected limited associations of those dietary factors with milk production on these farms. Specifically, we identified that the farms feeding corn silage or barley silage as their primary source of forage had greater milk yield than those farms feeding haylage as their primary forage source. This may be related to difference in forage quality (including digestibility) and nutrient (energy) density. However, these primary forages were also confounded with region (as described above), so we could not conclude if these differences in production were related to the forage types themselves or some other unaccounted-for regional differences.

The only nutrient associated with milk production was the amount of fat in the PMR. There are many research examples of how supplementing fat can have positive effects on milk yield. However, it must be noted that the effect fat supplementation has on milk yield is dependent on source, fatty acid profile and the basal diet being fed. In our study, the majority of farms (61%) fed supplement palm fat, which may be a contributing factor to this difference in production with greater PMR fat content, given palm fat’s known positive impact on production. Unfortunately, a limitation in our study was our lack of ability to determine the concentration of individual fatty acids in the diets across farms, as various products were used with varying, undisclosed fatty acid profiles.

We also investigated how the dietary variation on these farms was associated with the average milking frequency of the herds. Again, few associations were detected, reflective of the high amount of variation in nutritional strategies across the country. One finding was that greater non-fiber carbohydrate (NFC) content of the PMR was associated with fewer milkings per day. It is possible that having more NFC and thus energy in the PMR may, in certain circumstances, reduce the motivation of cows to enter the automated milking system unit to receive the supplemental automated milking system concentrate. This may particularly be the case in our study population since the vast majority (90%) of the farms were set up for free traffic to the automated milking systems, as opposed to only 10% of farms having some form of guided traffic. While we still need more research to understand these effects, having some energy differential between the PMR and automated milking system concentrate does seem to help promote more voluntary milkings, particularly in free-traffic scenarios. 

It is also interesting to note that in our study population, greater forage content of the total diet was associated with fewer average milkings per day. The reasons for this association are not clear. It is possible that with high forage content, total dietary energy may be limiting, which could lead to lesser production, and also be then tied to fewer milkings. That argument would only hold in situations where forage quality is also lacking; unfortunately, we were not able to collect any information on forage quality in this study, so we cannot make any conclusions on that.

What is interesting to note is that some of the other non-dietary factors we recorded on these farms had significant impacts on milking frequency. For example, the free-flow cow traffic farms had, on average, +0.6 milkings more per day than those farms with some form of guided flow traffic. It should be noted that milking permission settings were not able to be recorded and thus considered a limitation of our study. Such settings could allow or prevent cows from accessing the milking robot more frequently and thus would affect the observed milking frequencies regardless of traffic flow.

We also noted that greater feed push-up frequency was associated with greater milking frequency. In previous work, we have demonstrated that not only are voluntary milkings associated with periods of feeding activity in automated milking system cows, but that increasing feed availability, through more frequent pushing up of feed, allows automated milking system cows to optimize their time for feeding rather than searching for feed, and as a result also allows those cows to optimize their resting (lying) time.

Across farms in our study, every 0.1 increase in herd-average successful milkings per day was associated with +0.34 kilogram per day of herd-average milk production. Thus, understanding of those factors associated with voluntary milking, and thus overall milking frequency, are important for promotion of high production in these automated milking system farms.

Overall, the results of our national study on benchmarking the nutritional strategies on Canadian automated milking system farms help to understand how diet variability and management may be associated with milk production and milking behavior. This information can then, in turn, be used to improve how diets are formulated for automated milking system herds as well as how we manage those within-farm.