**Background:** The COVID-19 pandemic, with a death toll exceeding 6.5 million people by December 2022, has underscored the global necessity of mitigating the underlying drivers of zoonotic spillover events. Zoonotic diseases, caused by pathogens of wildlife or livestock origins that spill over into human populations, account for more than half of human-susceptible pathogens. The incidence of zoonotic outbreaks and the diversity of their source species have increased in recent decades. Despite expanded pandemic preparedness policies, national and international efforts do not adequately address reducing emergent infectious disease risks, and one specific omission is the global food system. This omission is concerning because of agriculture's outsized spatial dominance over tropical and temperate ecosystems and its consequent potential roles in zoonotic outbreaks. The authors argue that food systems vary widely in production methods, product varieties, governing policies, and stakeholders, and developing a general understanding of the zoonotic disease risks generated from various food systems is imperative.
**Methods:** The authors used a scenario framework based on an expert-driven approach to draw a qualitative blueprint of the variable impacts different agricultural production methods, land allocation patterns, and diets can have on zoonotic spillovers. They categorised ten direct food-system-related drivers of zoonoses: (1) biodiversity loss; (2) land fragmentation; (3) pesticide use; (4) water use; (5) fertiliser application; (6) antibiotics use; (7) wildlife hunting; (8) aquaculture; (9) livestock densities; and (10) farmworker densities. Using land use extent (continuous) and agricultural practices (categorical) as the two axes spanning the most variance in these drivers, they constructed four archetypal food system scenarios: industrial animal-based, agroecological animal-based, industrial plant-based, and agroecological plant-based. Based on expert opinion, they ranked the potential risk for zoonotic spillover for each of the ten individual food-system drivers on a scale of 1 (very low) to 5 (very high) in each of the four scenarios. The overall risk of a scenario was calculated as the sum of all its individual risks, with smaller sums corresponding to low-risk scenarios and larger sums corresponding to high-risk scenarios.
**Key Results:** The industrial animal-based scenario (exemplified by the USA or Australia) carries the largest zoonotic spillover risk, driven by high livestock production volume, monocultures, and intense use of water, antibiotics, pesticides, and fertilisers over large land areas. The agroecological animal-based scenario (e.g., mixed cropping systems in southeast Asia, silvopastoral systems in South and Central America) carries medium risk potential, with lower risk than the industrial animal-based scenario due to lower livestock numbers and densities, but with elevated risk from habitat fragmentation and farmworker density. The industrial plant-based scenario (e.g., large palm oil plantations in Malaysia and Indonesia, large-scale wheat production in the USA) carries medium overall risk, with intensive agricultural practices but lower fragmentation due to larger contiguous farms and reduced on-farm human activity from automation. The agroecological plant-based scenario (e.g., contemporary agroecological farms in Latin America and Africa) has the lowest risk for zoonotic spillover events, minimising land use through plant-based diets and employing agroecological practices that optimise agrobiodiversity. The authors note that because animal-based foods are the largest determinant of land occupation, reducing animal-based products and promoting plant-based diets will reduce global land use markedly without necessarily compromising nutritional outcomes.
**Clinical Implications:** The authors argue that prophylactic measures to restrict the emergence of zoonotic diseases are closely linked to diets and food policy, and are location-specific and context-specific due to differing ecological, climatic, cultural, and socioeconomic realities. Promotion of plant-based diets will probably be essential for zoonotic spillover risk reduction, as they can potentially free large areas of currently human-appropriated lands for rewilding, reducing livestock densities, and reducing agricultural lands. In low-income and middle-income countries where a planetary health diet is unaffordable and people's nutrition depends on animal-based foods and wildlife hunting, promoting agroecological practices and increasing region-specific agrobiodiversity alongside natural habitats is beneficial. Policy directions include subsidising plant-based food consumption, taxing red meat, promoting agronatural landscapes and ecosystem services, improving hygiene measurements, and optimising the location of wet markets. The authors emphasise that many knowledge gaps connecting specific agro-practices and agrobiodiversity measures directly to zoonotic risks remain, and future research should identify region-specific low-zoonosis-risk agricultural practices that yield economically viable, nutritious, and culturally acceptable diets while simultaneously reducing zoonotic risks and explicitly promoting agrobiodiversity. The key message is that zoonotic disease risks are inextricably linked to dietary choices and the structure and functioning of food systems, and proposed transformation towards sustainability should carefully, sensitively, and explicitly consider these risks in their suitable societal and cultural context.