Dairy is a vital part of the world's food supply. More than people worldwide consume milk and milk products. The dairy industry must continuously improve to ensure it positively impacts people, animals, and the environment. In this post, we will highlight four researchers funded by ̽̽'s Food Systems Research Center (FSRC) that have dedicated their careers to making dairy more sustainable.
Felipe Machado de Sant'Anna: Detecting antimicrobial-resistant genes in the farming process
Mentor: John Barlow
Approximately one in five antibiotic-resistant infections in humans are caused by germs from food and animals. Antimicrobial-resistant (AMR) bacteria are a growing concern worldwide; the dairy industry is no exception.
̽̽ Postdoctoral Research Associate Felipe Machado de Sant'Anna aims to reduce the spread of AMR genes carried by bacteria in the cheese production supply chain. Machado is guided by the concept that we're all interconnected, known as .
"We are seeing many bacteria that used to be related to animals and are now spreading to the human level, and they're carrying a lot of antimicrobial resistance genes. The One Health concept tries to mitigate this not only for human health and the environment but for animals too," Machado de Sant'Anna said.
In a study of six Vermont dairy farms, five cow and one goat, Machado de Sant'Anna and a team of ̽̽ and USDA researchers followed the path of milk production by taking samples from the animal manure, water supply, milk filters, bulk tank, and cheese. They extracted DNA from the samples and used a sequencing platform to make copies and detect AMR genes. The study's results are pending, though a smaller pilot study revealed the milk filter and water supply may have the highest levels of AMR bacteria from pathogens such as Staphylococcus, Listeria, Salmonella, Campylobacter, and E.coli.
"Our theory is that the milk filter accumulates bacteria and other debris. The water the cows and goats are drinking has bacteria from their mouth or nose that is contaminating it," Machado de Sant'Anna said. "If we can assess where the highest amount of AMR genes are within one farm, we can look at ways to mitigate or reduce the spread."
Machado de Sant'Anna says it's important to work with raw milk products because whenever someone eats a raw product, the bacteria within it can potentially spread its genes to other resident bacteria in the human microbiome.
"Humans are a huge amount of cells and bacteria," Machado de Sant'Anna said. "The saying 'you are what you eat' makes sense because what we eat can impact our gut, and we can develop other bacteria more prone to getting these resistant genes."
Though Machado de Sant'Anna, a cheese enthusiast who hails from Minas Gerais, the of Brazil, says people shouldn't worry about eating high-quality raw milk cheese. Minimum ripening times help to ensure food safety, and Sant'Anna's research will look at how to make cheese even safer.
"The ripening process increases the good bacteria in the cheese, but that is just one factor. To get a good raw milk cheese, we have to evaluate how it was made," Machado de Sant'Anna said. "If it was made in an environment that contains an AMR source, how does that affect the cheese ripening? This is one of the questions we expect to answer with our sequencing."
Machado de Sant'Anna hopes his work will open the door to more research.
"Our study might inspire a myriad of studies that could evaluate cheese quality throughout the ripening process, for not only AMR genes but also other aspects of cheese quality," Machado de Sant'Anna said.
Andrew Magnuson: Connecting farm practices with dairy product nutrition
Mentor: Jana Kraft
Approximately one in three Americans have prediabetes and are at risk for developing type 2 diabetes, heart disease, and stroke, according to the . ̽̽ Postdoctoral Scientist Andrew Magnuson and Associate Professor Jana Kraft are working to understand the potential role dairy can play in mitigating this risk.
“We are looking at how agricultural practices influence the nutritional composition of dairy. Namely, how can we enhance the nutritional value and lipid content of dairy for better human nutrition outcomes including dietary prevention of obesity and type 2 diabetes,” Magnuson said.
The study, conducted at the U.S. Dairy Forage Research Center in Madison, Wisconsin, involved collecting milk, ruminal fluid, and blood from a mixture of 60 Jersey and Holstein cows, the two most prolific dairy cow breeds in the U.S. Lipid contents of samples are currently being analyzed to determine how both breed and novel dietary supplements impact dairy nutritional composition.
Magnuson says he hopes that the research will give farmers tools to enhance the nutritional value of dairy, which can not only increase their profitability but also positively benefit society.
“Which will help both dairy farmers and improve the health of dairy consumers. Milk fat has been stigmatized for having a ratio of saturated fatty acids, but recent research had elucidated the health benefits of lipids in dairy products such as prevention of type 2 diabetes,” Magnuson said. “Furthermore, the lipid profile of dairy is influenced by dairy cow breed and agricultural practices which can be controlled by producers. Thus, this work is aimed to further our understanding of how to enhance the nutritional value of dairy to help farmers produce milk with the highest value for human consumers.”
The first results will be presented at the (ADSA) meeting and are expected to be published by the end of 2023.
“Hopefully, this research will further positively change the perception of dairy fat and promote more work to enhance the concentration of health-promoting lipids in dairy foods,” Magnuson said.
Sardorbek Musayev: Technology to predict crop health and productivity
Mentor: Joshua Faulkner
There are a lot of uncertainties in farming, whether it be the weather, the price farmers are paid, or how well crops will grow. ̽̽ Postdoctoral Research Associate Sardorbek "Sardor" Musayev and a team of USDA and ̽̽ researchers are working to make life more certain for farmers. They’re using technology to study farmland to predict crop yields at nearly 50 farms across Vermont, South Dakota, and Pennsylvania.
"We are learning to predict what that crop health and yield will look like in 50, 100 days, and 150 days" Musayev said. "As a crop grows, we study the relationship between remote sensing technologies such as satellites, drones and ground-based data from planters, soil sample analysis, and national ground truth databases."
Musayev is focusing on higher resolution free remote sensing products such as Sentinel-2 (ESA) which provides free images of Earth every five days.
"As we learn the relationship between remote sensing technologies like drones and satellites and ground data through Machine Learning algorithms, we'll look at scenarios with and without drones to see whether we get better results with free satellite images to predict crop health and yield," Musayev said. "If I'm telling a farmer, let's fly the drone, and you buy the drone, the answer is no. But if I say, there is a free satellite product and every five days it takes a picture of the same spot, there is interest. The free product is there for a farmer, and it's underutilized."
Musayev along with the ̽̽ AI team and USDA researchers are developing and testing different Machine Learning algorithms to predict crop yield. Musayev has applied the Random Forest predictor algorithm for crop yield prediction with r-squared 0.77, or 77 percent accuracy. He says he’s learning to improve the performance of the prediction model. The project objectives are to help the farmer reduce farm expenses.
"Nowadays, let's say you have 500 acres of corn. How do you know how the crops are doing? You can drive around...but you cannot drive everywhere. It takes too much time, and you're nervous about how to make decisions," Musayev said.
Another option farmers have is to buy precision agriculture apps from equipment companies, who then sell them equipment to apply nutrients to fields at varying rates using the data, but Musayev says it's problematic.
"Farmers have to pay companies like John Deere, and then the data ownership is not with the farmer," Musayev said. "Farmers, they're not making much money. We should help them to grow."
With the free data, Musayev is optimistic tech developers can create a free app farmers can use to see what areas of their fields need fertilizer, like phosphorus and nitrogen, and then supply the required amounts.
"With precision agriculture, the main idea is to retain the nutrients in the field – it's like a diet, with how much the crop needs, so we make sure it gets enough, and not too much," Musayev said.
As for what's next, Musayev says that their team is also learning to use advanced satellites with hyperspectral sensors that have many bands (up to 400 bands) to detect crop health with more accurate results whereas the current technology he’s using relies on multispectral sensors with only 12 bands.
A data-lover, his enthusiasm brimming, Musayev said, "It's amazing to learn the relationship between crops and remote sensing technologies. It feels like controlling your farm from a high altitude."
Panagiotis Oikonomou: Understanding the actual cost of sustainable practices
Mentor: Asim Zia
One of the significant obstacles to sustainable farming is the cost. Farmers may be reluctant to adopt sustainable practices, as knowing the true cost and return is difficult. ̽̽ Research Assistant Professor, Panagiotis (Takis) Oikonomou, is part of a research team creating a modeling tool to better help dairy farmers understand the cost and benefits of hundreds of practices by combing related operating costs such as energy, water, feed, and generated revenue such as from produced milk or animals sold.
"Let's say a farmer wants to consider implementing a crop or a feed or a manure-related management practice, or even a combination of such practices; sometimes it is not easy to estimate the associated costs and benefits of those practices. The model could provide such information and facilitate towards educated decision-making at the farm level," Oikonomou said.
The tool known as the brings together researchers, among others, from Cornell University, the University of Arkansas, the University of Wisconsin Madison, ̽̽, and the USDA. It is designed to simulate dairy farm production and environmental impact, and it is comprised of different modules (animal, manure, feed storage, soil & crop, energy & emissions, and economics) that simulate key processes taking place in a dairy farm.
"We are close to the release of version 1.0, and it's the first model that would assist with assessing the biophysical along with both ecological and economic considerations together," Oikonomou said.
Oikonomou says version one won't have all the data; labor costs, for example, will come in a future release.
"The model is open-source and written in Python, and there are discussions about making an app and ways it could be more user-friendly. The model will give the flexibility to the user to override default cost values by inserting their own cost of water for example," Oikonomou said.
The ultimate goal is for farmers to be able to look at the benefits and trade-offs for a variety of farming practices.
"It's a tool that can simulate different management practices scenarios, for example, test from different rations composition to different crops to grow, and estimate associated costs and impacts on the revenue side," Oikonomou said. "Of course, no model is perfect. It's an approximation of reality; however, RuFaS will allow us to compare management practices and inform end users in an integrated, holistic way."
Hafedh Ben Zaabza: Studying the impact of inbreeding on milk production
Mentor: Heather Darby
In dairy farming, inbreeding, or the mating of closely related animals, is a significant problem that can cause a decline in herd productivity, increased rates of genetic diseases, and reduced fertility. This phenomenon is called inbreeding depression.
Dairy farmers have struggled for years to minimize inbreeding, but it persists due to the intense selection of bulls in most breeds and the use of a reduced number of sires at the global population level. Calculation of inbreeding relative to a base population that assumes those animals to be unrelated – known as pedigree-based inbreeding, can be done. However, inbreeding can be severely underestimated when pedigree data is incomplete.
Another method, genomic selection, uses DNA information to estimate the inbreeding of genotyped animals. A genomic-based analysis of inbreeding is more accurate, however, there have been few studies that look at the specific genomic region causing common issues among cattle. ̽̽ Postdoctoral Research Associate Hafedh Ben Zaabza is studying another, potentially more accurate, way to detect inbreeding.
"Inbreeding can be calculated using three methods: Pedigree-based inbreeding, genomic-based inbreeding, and then there is homozygosity-based inbreeding," Ben Zaabza said.
Ben Zaabza is focusing on the newer method that characterizes inbreeding based on contiguous regions of the genome known as runs of homozygosity (ROH). Studying unique homozygous genomic regions (ROH) may show negative associations with production, fertility, and disease. Using this method, researchers can pinpoint the locations and the magnitude of different types of genetic effects in very small regions of the genome.
Rather than studying additive effects only, which describe a linear change with the number of copies of a gene effect, Ben Zaabza included the dominance effect into the model. Dominance is believed to play an important role in inbreeding depression. In this situation, two copies of an effect cause twice the difference of an individual having a single copy. There are also “non-additive” effects that include dominance effects. With dominance effects, the difference between the effect of two copies of an effect is not double the effect of a single copy. Statisticians would describe the dominance effect as an “interaction.” The most common type of dominance effect is heterosis or “hybrid vigor.” In heterosis, there is a crossing of different lineages that results in increased levels of different versions of each gene. This is termed an increase in “heterozygosity.” In the opposite situation, individuals that are more similar are mated, and there is an increase in “homozygosity” (and a concomitant reduction in heterozygosity) generating the “inbreeding depression.”
“We hypothesize that the characterization of dominance effects will improve our understanding of the genetic architecture of inbreeding depression and that stretches of homozygosity can provide an effective measure of genomic inbreeding,” Ben Zaabza said. “We are using national data that contains four million genotyped animals and pedigrees representing 90 million animals. Our objective is to characterize the genomic landscape of the regions associated with additive, dominance, and runs of homozygosity effects.”
Ben Zaabza expects to have results soon that will give farmers, researchers, and genetic experts tools to maintain genetic diversity and provide guidance for managing inbreeding.
About Us:
The Food Systems Research Center (FSRC) at ̽̽ (̽̽) uncovers solutions to society’s most pressing issues through the lens of our food system to improve human health, well-being and livelihoods, and environmental sustainability. Visit our website or follow us on .
