A new University of Michigan study concludes that nearly half of the food waste, about 620 million metric tons, could be eliminated by fully refrigerated food supply chains worldwide.

At the same time, fully refrigerated supply chains, or “cold chains,” could cut food waste-related emissions of climate-warming greenhouse gases by 41% globally, according to the study published online May 28 in the peer-reviewed journal Environmental Research Letters.

Sub-Saharan Africa and South and Southeast Asia have the greatest potential for reductions in both food losses and related emissions through increased cold-chain implementation, according to the study.

South and Southeast Asia could see a 45% reduction in food losses and a 54% decrease in associated emissions under an optimized refrigeration scenario. Sub-Saharan Africa has tremendous opportunities for both food loss (47%) and emissions (66%) reductions under optimized refrigeration conditions, the study shows.

And in many situations, developing more localized, less industrialized “farm-to-table” food supply chains may yield food savings comparable to optimized cold chains, according to the study.

“I was surprised to find the scale of our opportunity for reducing food loss and waste globally,” said study lead author Aaron Friedman-Heiman, a master’s student at U-M’s School for Environment and Sustainability and Ross School of Business. “Approximately half of the roughly 1.3 billion tons of food that goes to waste annually can be solved through food supply-chain optimization.”

The other author is Shelie Miller, a professor at U-M’s School for Environment and Sustainability and at the College of Engineering.

Food losses produce an estimated 8% of human-caused greenhouse gas emissions. The new U-M study focuses on food losses in the post-harvest to retail stages of the food supply chain and does not address on-farm or at-home losses.

The study accounts for the greenhouse gases emitted during food production. It does not include emissions tied to refrigeration or other supply-chain operations and does not include emissions from food waste in landfills.

The study, funded in part by Carrier Global Corp., found that:

• The greatest opportunity to improve food losses in less industrialized economies is the supply chain between the farm and the consumer. But in North America, Europe and other more industrialized regions, most food loss happens at the household level, so cold chain improvements would not have a major impact on total food losses.
• Reinforcing previous research, the U-M study highlights the importance of meat-related food losses. While the amount of fruit and vegetable losses is much higher, by weight, throughout the world, the climate-related emissions associated with meat losses are consistently greater than those associated with any other food type — due mainly to the high greenhouse gas intensity of meat production.
• Unlike previous studies of this topic, the U-M researchers compared the benefits of globalized, technologically advanced food-supply chains with those of localized “farm-to-table” food systems. “Hyper-localized food systems resulted in lower food losses than optimized global, refrigerated supply chains,” Friedman-Heiman said. “The results help quantify the value of maintaining and supporting local food chains.”

For the study, the researchers built a food-loss estimation tool to assess how improved access to the cold chain could impact food loss and its associated greenhouse gas emissions for seven food types in seven regions. They used data from the U.N. Food and Agriculture Organization and other sources.

By modeling food losses at each stage of the supply chain, the study highlights where the cold chain can be optimized to reduce food losses and emissions. The researchers analyzed the effects of moving from the current state of inconsistent and variable-quality cold chains throughout the world to an optimized system, defined as one with high-quality refrigeration across all stages.

The study estimates that poor cold-chain infrastructure could be responsible for up to 620 million metric tons of global food loss annually, resulting in emissions of 1.8 billion tons of carbon dioxide equivalents, the equivalent of 28% of U.S. annual greenhouse gas emissions.

The researchers say their adaptable, easy-to-use tool will be of use to anyone involved in the food supply chain, including farmers, grocery retailers, government officials and nongovernmental organizations.

“Although cold chain infrastructure is rapidly increasing worldwide, an optimized cold chain will likely develop at different rates and in different ways across the globe,” Miller said. “This analysis demonstrates that while increased refrigeration should lead to improvements in both food loss and greenhouse gas emissions associated with food loss, there are important tradeoffs associated with cold chain improvements by food type and region.”

She said Investment decisions will need to be prioritized to maximize the desired outcomes and impacts. For example, if an NGO’s top priority is ending hunger, then cold-chain upgrades that provide the greatest overall food-loss reductions may best meet that objective.

But organizations that prioritize climate action may choose to focus on reducing meat losses specifically, rather than total food losses.

The study found that meat accounts for more than 50% of food loss-related greenhouse gas emissions, despite accounting for less than 10% of global food losses by weight. Optimized refrigeration of meat could result in the elimination of more than 43% of emissions associated with meat loss, according to the study.

The researchers emphasize that the actual amount of greenhouse gas emissions savings will depend on the efficiency of cold-chain technologies and the carbon intensity of local electrical grids, since climate emissions associated with refrigeration can be significant.

The U-M study was supported by the U.S. National Science Foundation and by Carrier Global Corp., a global leader in intelligent climate and energy solutions.

“The study highlights that we are all mosaics,” said Korbel, who is Senior Scientist in the Genome Biology Unit and Head of Data Science at EMBL Heidelberg. “Even so-called normal cells carry all sorts of genetic mutations. Ultimately, this means that there are more genetic differences between individual cells in our bodies than between different human beings.”

Both Korbel and Sanders, Group Leader at the Max Delbrück Center study how genetic structural variation — deletions, duplications, inversions, and translocations of large sections of the human genome — contributes to the development of disease. In the cancer field, it is well known that genetic mutations can cause cells to grow out of control and lead to the formation of a tumour, explained Sanders. “We are applying similar concepts to understand how non-cancerous diseases develop,” she added.

The discovery was enabled by a single-cell sequencing technology called Strand-seq, a unique DNA sequencing technique that can reveal subtle details of genomes in single cells that are too difficult to detect with other methods. Sanders is a pioneer in the development of this technology. As part of her doctoral research, she helped develop the Strand-seq protocol, which she later honed with colleagues while working as postdoctoral fellow in Korbel’s lab.

Strand-seq enables researchers to detect structural variants in individual cells with better precision and resolution than any other sequencing technology allows, Sanders said. The technology has ushered in an entirely new understanding of genetic mutations and is now being widely used to characterise genomes and to help translate findings into clinical research.

“We are just recognising that contrary to what we learned in textbooks, every cell in our body doesn’t have the exact same DNA,” she said.

Genetic mosaicism is common

The study represents the first time anyone has used Strand-seq technology to study mutations in the DNA of healthy people. The researchers included biological samples from a range of age groups — from newborn to 92-years-old — and found mutations in blood stem cells, which are located in the bone marrow, in 84% of the study participants, indicating that large genetic mutations are very common.

“It’s just amazing how much heterogeneity there is in our genomes that has gone undetected so far,” said Sanders. “What this means in terms of how we define normal human ageing and how this can impact the types of diseases we get is really an important question for the field.”

The study also found that in people over the age of 60, bone marrow cells carrying genetic alterations tended to be more abundant, with populations of specific genetic variants, or sub-clones, more common than others. The frequent presence of these sub-clones suggests a possible connection to ageing.

But whether the mechanisms that keep sub-clones from proliferating in check break down as we age, or whether the expansion of sub-clones itself contributes to diseases of ageing is not known, said Korbel. “In the future, our single cell studies should give us clearer insights into how these mutations that previously went unnoticed affect our health and potentially contribute to how we age.”

“The whole program is like day and night — there is life in the normal state and life in the diapause state, and the way this happened was by reshuffling or re-wiring the regulatory region of a whole set of genes,” says senior author and molecular biologist Anne Brunet of Stanford University.

African turquoise killifish mature faster than any other vertebrate species, and adults live for only around 6 months, even in captivity. The fish reproduce rapidly before their watery homes disappear, but their embryos remain behind in the dry mud, ready to hatch when the next year’s rains come. Embryonic diapause also occurs in other vertebrate species, including fish, reptiles, and some mammals, but killifish diapause is remarkably extreme because it lasts for such an extended period (8 months on average and up to 2 years in the lab) and because killifish embryos enter suspended animation much later in development than other animals.

“It’s roughly in the middle of development, and many organs are already formed by that stage — they have a developing brain and a heart which stops beating in diapause and then starts again,” says first author Param Priya Singh of the University of California, San Francisco. “Killifish are the only vertebrate species that we know of that can undergo diapause so late in development.”

To understand diapause evolution, the team first characterized the gene expression of the African turquoise killifish (Nothobranchius furzeri) during different developmental stages. They focused on duplicated copies of genes called “paralogs,” because gene duplication is one of the primary mechanisms by which new genes originate and specialize. Overall, the researchers identified 6,247 paralog pairs that exhibited specialized gene expression patterns during diapause. Surprisingly, they estimated that most of the diapause-specialized genes were “very ancient” paralogs, having originated more than 473 million years ago.

“Even though diapause evolved relatively recently, the genes that are specialized in diapause are really ancient,” said Brunet. “We found that most of the genes that specialize for diapause in killifish are very ancient paralogs, which means that they were duplicated in the common ancestor of all vertebrates.”

Since diapause also occurs in some other species of killifish, the researchers compared gene expression between embryos of the African turquoise killifish, the South American killifish (Austrofundulus limnaeus), which also undergoes diapause, and two killifish species that do not undergo diapause, the red-striped killifish (Aphyosemion striatum)and lyretail killifish (Aphyosemion austral).

They found significant overlap in gene expression patterns between the African turquoise and South American killifish, which evolved diapause independently of each other, but not in the two non-diapausing species. Likewise, the researchers found significant correlation in the gene expression patterns of house mouse (Mus musculus) embryos during diapause and showed that diapause-specialized genes in mice also have very ancient origins.

“This suggests that the same mechanisms that enable diapause have been repeatedly co-opted for the evolution of diapause across distantly related species,” says Singh.

Next, the researchers explored how these diapause-specialized genes are regulated in the killifish. They identified several key transcription factors that control the altered gene expression patterns seen during diapause, including REST and FOXO3, which are known to be expressed during hibernation (a different form of suspended animation) in mammals. Notably, several of these regulatory genes are involved in lipid metabolism, which has a distinctive profile during diapause.

“One of the key elements of diapause is this special lipid metabolism,” said Brunet. “During diapause, they seem to have much higher levels of triglycerides and very long chain fatty acids, which are forms of storage and also perhaps aid with long-term protection of the organism’s membranes.”

The researchers plan to continue investigating how different species regulate diapause and to dig deeper into the role of lipid metabolism during diapause and other types of suspended animation.

“It’s such a complex state that I think we are just scratching the surface,” said Singh. “We want to go deeper into specific aspects of how lipid metabolism is regulated during diapause, and we are also interested in examining the role of specific cell types during diapause.”