Nitrogen (N) and phosphorus (P) loss through contemporary dairy production systems contribute to water quality deterioration, algal bloom and direct toxicity. Annual losses are reported to vary from 5 to 200 kg N ha–1 and from 0.5 to 10 kg P ha–1, depending on climate, soil, landscapes and farm management practices.

Observed N and P losses from 156 sites across 17 countries, although a preponderance of data exists from New Zealand, the United States and Australia, utilizing 3 dairy production systems — grazing (forage production on-farm and grazed livestock outdoors), hybrid (combination of housing, a few months of outdoor grazing and forage preservation) and confined (minimal outdoor grazing, housed animals with most feed harvested and brought indoors) — suggest no clear advantage of one system over another for reducing nutrient losses to water. Modelling the three systems in New Zealand, northeast United States and the Netherlands shows N losses were lower from hybrid and confined systems when expressed in terms of productivity, but no differences across dairying systems were found when N losses were expressed on an area basis. P losses were lower from grazed systems compared with hybrid and confined systems on both productivity and area basis.

See McDowell et al.

Poor air quality is considered one of the greatest environmental mortality risk factors. While progress has been made on air pollution from the industry, transportation, solid waste management and household sectors, air quality has been largely absent from the discussion of food systems, and of human and planetary health. The development of tools for estimating air pollution from food systems have not kept pace with other sectors.

The European Commission’s Emissions Database for Global Atmospheric Research (EDGAR) is now used to develop an inventory of NH3, NOx, N2O, SO2, CO, non-methane volatile organic compounds, and particulate matter emitted in the production (including land-use change), distribution, consumption and waste of food. Air pollution emissions from food systems have risen over the past 50 years, corresponding to more than half of total nitrogen emissions and more than a third of particulate matter globally. More than a fifth of deaths due to poor air quality are related to pollution from food systems.

See Crippa et al. and Balasubramanian and Babbar

There are many viewpoints on how cellular agriculture technologies can benefit or hinder sustainable food system transformations.

This focus issue takes stock of the field from an interdisciplinary perspective. Our contributors comment on sustainability, food justice, corporate power and potential for greenwashing, virtue ethics, scaling for impact and antimicrobial resistance, and examine tensions and opportunities for moving forward.

See Editorial, Comments by Tuomisto, Ellis et al., Broad and Chiles, Howard, Bomkamp, Holmes et al. and Alvaro, World View by Friedrich, and Feature by Gruber

An exercise of accounting the flow of nutrients from farm to fork has been undertaken to assess the adequacy of the supply of nutrients to the world population. Thirty-six bioactive components in all farm commodities recorded by the Food and Agriculture Organization of the United Nations were traced from production through trade, processing, cooking, loss and waste — and accounting for bioavailability and age–sex distributions of requirements — to supply the nutrient needs of populations for 173 countries over almost 60 years.

Nutrient Balance Sheets add to our understanding of food security. Mechanisms to close the nutrient gaps between supply and requirements within populations are given greater granularity by the Nutrient Balance Sheets, identifying where supply can be enhanced by food production, trade, fortification, supplementation or other policy approaches.

See Lividini and Masters and Joy and Kumssa

Salmon aquaculture is projected to grow by 2–3% per year to meet the increasing demand for aquatic foods. Wild-caught, marine-derived resources in the form of fish meal and fish oil are key sources of protein and lipids in salmon aquafeed formulations, but with wild fish stocks stagnating, this resource presents a limiting factor for future sector growth. Global salmon production potential was modelled, incorporating a 1–3% growth rate and exploring a variety of fish oil and fish meal utilization scenarios. Incorporating 3% fish oil and 3% fish meal in aquafeed could permit 2% per year production growth until 2100 — independent of novel aquaculture feeds that are currently being utilized.

Aquatic foods, such as salmon, form an important dietary source of long-chain omega-3 fatty acids. Based on these findings, 300 grams of salmon per week would provide almost all the recommended weekly intake of eicosapentaenoic acid and docosahexaenoic acid — demonstrating that finite marine resources, when used judiciously, can contribute to salmon aquaculture sector growth and healthy diets.

See Rocker et al.

The yield gap is the difference between the potential yield of a crop under management that minimizes yield losses from biotic and abiotic stresses and the actual farm yields under dominant management practices and soil conditions. A wheat simulation model was used at 53 study sites across the world under optimum local wheat cultivar management practices to estimate potential yield (globally, 6 dry matter tonne per hectare (DM t ha–1)), and the difference between the current mean global wheat yield (3 DM t ha–1) indicates a global yield gap of 50% due to sub-optimal crop and soil conditions.

Genetic yield potential represents the yield that could be achieved in target environments through genetic improvements compared with yields from current local cultivars. Eight wheat traits relating to canopy structure, phenology, root water uptake and drought tolerance that are considered important for yield improvement, large available genetic variation, high heritability and breeder friendliness were optimized in designed wheat ideotypes based on current cultivars. The global genetic yield gap is estimated to be 51%, ranging from 30% (lowest, New Zealand) to 70% (highest, Australia and Kazakhstan). The genetic yield gap quantifies opportunity for increasing productivity through genetic improvement.

See Senapati et al. and Hasegawa and Wilson

‘Food miles’ indicate the carbon footprint of food transport from points of production to consumption. Measured in tonne-kilometres (tkm), estimates can vary widely by employing different calculation methods and system boundaries. A global multi-region accounting framework estimates food miles and associated greenhouse gas emissions over the entire supply-chain network to account for 27% of total freight-mile emissions. Food miles were found to correspond to approximately 20% of food systems emissions when transport, production and land-use change are considered, and 30% when land-use change is excluded from the calculation. Domestic and international food-miles emissions are shown per country and region, economic sector and transportation mode; about 30 million direct trade connections are included in the analysis.

Agricultural and food trade has more than doubled since 1995, and has a key role in food security. However, large food-miles-related emissions are embedded in supply chains within and between affluent countries, and although high-income countries represent about 12.5% of the global population, their food trade represents 46% of global food-miles emissions.

See Li et al. and Pradhan

Nutrient production and accumulation in food result from the biochemical reaction networks that characterize living organisms. Nutrient distributions may, therefore, be predicted from biochemical first principles and are now shown to display consistent statistical behaviour across foods. Our understanding so far of the composition of food has not taken this universality into account.

Utilizing the concept of universality, and the computational approaches that support it, to elaborate our knowledge of food composition has advantages. As nutrients are shown to follow common patterns across foods, missing quantities in food composition tables may be estimated, the concentration of unquantified chemicals of food may be predicted, and resources associated with analytical laboratory approaches to determining chemical concentrations in food may be reduced.

See Menichetti and Barabási

Indigenous microbial consortia harnessed from grape, vine and fermentation vessel attribute favourable characteristics to wine. Yet, diversity of microbial communities contributes to vintage variation from one year to the next and, with adverse future scenarios that may impact the industry, the prediction of fermentation outcomes is challenged.

Synthetic biology enables the engineering of microbial communities for new purposes. Applied to wine-making, yeast consortia may be engineered for model systems that facilitate ‘plug-and-play’ fermentation, allowing the winemaker to generate — and maintain — desirable characteristics at will, with greater predictability.

See Walker and Pretorius

Antimicrobial use and overuse in livestock production is a key driver of global antimicrobial resistance (AMR) development. Although different antimicrobials are typically used in humans and animals, resistance genes can spread horizontally in the environment — even to bacterial populations with no previous exposure to antimicrobials.

Now, a genomic analysis of Escherichia coli samples taken during the past 50 years from livestock production sites in China reveals patterns of AMR spread and identifies plasmids that conferred increasing resistance for critically important veterinary and human antimicrobials. Policy actions informed by the biology underlying AMR are urgently needed to curtail the use of antimicrobials in livestock production and limit the global spread of AMR.

See Yang et al.

Adoption of the EAT–Lancet reference diet across the world will be challenged by heterogeneity in food systems, dietary patterns, socio-economic development and environmental boundaries. Current domestic food supplies will need to transition for populations to achieve healthy, sustainable diets — and that endeavour varies in complexity across food groups and countries. The food supply of countries highest in the socio-economic development index, at present, is characterized by animal-based foods, fats and sugars in excess of the reference diet. Countries of lower socio-economic development have excess domestic supply of cereal and starchy root foods, and all countries have inadequate supply of legumes, nuts and fruits to achieve the reference diet.

It has previously been reported that adoption of the EAT–Lancet diet will decrease agricultural greenhouse gas emissions globally but increase them from some, primarily low- and middle-income, countries. Now, transition of food supply towards the EAT–Lancet reference diet is shown to reduce the global water footprint by 12%, but increase that of 54 low- and middle-income countries, representing 40% of the world’s population.

See Tuninetti et al.

Crop pest and disease (CPD) dynamics under climate change have been challenging to ascertain. The scale and extent of CPD dispersal cannot be well represented by field and laboratory experiments alone; models facilitate understanding of large-scale impacts but have simplified representations of CPD occurrence, highlighting the need to study CPD occurrence with long-term observational data.

The fall armyworm (Spodoptera frugiperda) spread across 26 provinces in China in 2019 infesting more than 112 million hectares of cropland, and such pests pose a threat to China’s crop production and food security in the future. The analysis of historical data on CPD occurrence from 1970 to 2016 indicates that climate change has already been responsible for one-fifth of the increase in CPD occurrence observed, most prominently driven by warmer night-time temperatures.

This trajectory is predicted to develop. Scenario analysis to the end of this century indicates that China will experience increasing CPD occurrence and cropland damage. Appropriate management and application of technology may minimize impacts, should a policy environment support these advances.

See Wang et al.