Thousands of nursing Holstein cows in multiple states are supplying continuous data that is being evaluated to enhance breeding for improved feed efficiency in the US and worldwide as you read this.
Higher feed efficiency, which implies that less feed is required to produce the same quantity of milk, is clearly a win for everyone, not just because it reduces environmental effect, but also because it helps farmers cut feed costs.
Unrivalled quantity of cow performance data gathered
Breeding for improved feed efficiency (achieving valid ‘genomic breeding values for this characteristic) has been limited up to this point by a lack of sufficient, trustworthy data. That is why, around 5 years ago, a big research centred at multiple US colleges was launched. Not only is the quantity of cows unusual, but this research is also notable for collecting vast volumes of cow performance data of various sorts using automated sensors.
The project is being funded by the Foundation for Food and Agriculture Research (FFAR), a government-affiliated organisation that creates public-private partnerships to fund research projects that help solve critical food and agriculture challenges, complementing the USDA’s research agenda. In this example, FFAR provided a $1 million grant to Michigan State University, which was matched by a $1 million grant from the Council on Dairy Cattle Breeding.
Dr. Michael VandeHaar of Michigan State University is conducting the project, which includes a core group of six geneticists and four nutritionists from the University of Wisconsin, Iowa State University, the University of Florida, and the USDA Animal Genomics Improvement Laboratory.
The initiative will result in more accurate genomic feed efficiency forecasts, as well as the ultimate creation of a feed intake index that employs sensors to estimate feed intake on individual cows. The researchers will also look at whether genetic predictions of feed efficiency may help reduce methane emissions.
Sensors on the ears – a feed intake indicator that predicts feed intake on individual cows using sensors. Dr. James Koltes, Iowa State University.
Sensors on the ears – a feed intake indicator that predicts feed intake on individual cows using sensors. Dr. James Koltes, Iowa State University.
Data gathering
Because the emphasis is on feed efficiency, the data collected includes not only individual cow feed intake and milk output, but also body weight – and from sensors, body temperature, eating behaviour, and movement data. Milk from the research cows will also be tested.
“We currently have data on feed intake, body weight, and production from 8200 cows, milk spectra from 3250 cows, body temperature from 1200 cows, and activity (locomotion and feeding behaviour) from 1800 cows,” VandeHaar explains.
The team also fulfilled its objective of genotyping all new cows last year. They also measured methane emissions from 147 cows, which was three times the intended amount.
The diets of the research cows are normal commercial diets that differ amongst participating farms. All of the cows are at least two years old.
Although the research is proceeding well, there are some difficulties. The availability of cows to use in the study, for example, has been a challenge. Furthermore, VandeHaar claims that gathering quality data on feed consumption of individual nursing cows and integrating it to the database is costly and time-consuming.
Supplementing current dairy cow data
This initiative expands on a database built more than a decade ago by Michigan State scientists, who discovered in a USDA-National Institute of Food and Agriculture study that breeding for more feed-efficient cows may save the US dairy industry $540 million USD per year with no decrease in milk supply.
“We developed a database of 3950 US cows in that project,” VandeHaar continues. “Those cows had their feed intake, body weight, and production data recorded for at least 28 days and, in most cases, 42 days.” They were all genotyped as well.”
Individual cow feed intake and milk production are being collected, as well as body weight – and from sensors (here around the neck), body temperature, eating behaviour, and movement data. Dr. James Koltes, Iowa State University.
Individual cow feed intake and milk production are being collected, as well as body weight – and from sensors (here around the neck), body temperature, eating behaviour, and movement data. Dr. James Koltes, Iowa State University.
Sensor use in cow feed intake
VandeHaar and his colleagues are now assessing the efficacy of sensors for forecasting feed intake in their present study.
“Our goal was to have sensor data from a subset of the 3600 cows, mostly from Michigan State, Iowa State, and the University of Florida,” he says, “and we are currently at 1800.” To present, sensors and spectral data seem promise for forecasting certain changes in Residual Feed Intake (RFI, a measure of feed efficiency) among cows within a farm, but establishing equations that operate across farms will be tough.”
The team has previously released assessments of sensor data from individual stations and has begun studying data from several sensor stations. They have already begun examining sensor phenotype combinations for forecasting feed intake, although much more work need to be done.
Indeed, it is too early to tell what can be achieved on this front at this time.
“We can already use a cow’s body weight, body condition score, parity, and milk production to predict if she eats more or less than the other cows in her feeding group,” VandeHaar says. “Then, if we know the group’s average intake, we can predict her intake.” However, we know that certain cows consume more or less than anticipated, as predicted by RFI. Our objective is that data from sensors (particularly activity, feeding time, and rumination time) and milk spectra data may provide further information that can assist us in predicting RFI. But we don’t know how well that will work at the moment.”
VandeHaar observes that if it proves to be effective, organisations that sell sensors would most likely utilise this method of forecasting RFI using their own sensors.
Developing mathematical equations for farm management software
“Perhaps we’ll be able to develop equations for farm management software that combine activity and rumination time with milk spectra data, genomics data, automated body weight and body condition data, milk production, and group intakes from TMR feeding systems to predict intake of individual cows,” he adds. These might be utilised to estimate feed efficiency and aid in breeding and culling choices.”
Looking forward, he and his colleagues intend to publish equations to forecast feed intake of individual cows in group-fed circumstances in 2024 or 2025, which would be utilised in culling and breeding choices.
Feed Efficiency – Feed Savings
Because of the project’s first-year findings, the trait known as ‘Feed Saved’ (a feed efficiency trait based on RFI and body size) was added in the Net Merit Index in 2021, one of 39 qualities that have appeared in the Index since that time for Holsteins.
(The Net Merit Index is a 1994 US breeding index that forecasts net profit throughout the lifespan of the typical dairy cow or bull’s daughter. VandeHaar and his colleagues say that it’s frequently used to select sires for use on commercial farms across the US, and that these rankings will have a long-term impact on Holstein genetics throughout the globe.)
Increasing the number of novel RFI phenotypes
Last year, the researchers measured RFI phenotypes to match with genotypes of unique cows for 1170 cows, bringing the project’s total of novel RFI phenotypes to 4200.
However, the team also observed last year that the new Feed Saved trait’s dependability is lower than expected. “As a result,” they write, “we intend to continue phenotyping as many cows as possible in year 5 (2023) to maximise the effectiveness of the Feed Saved trait.”
