Consumers are increasingly aware of the environmental impacts of their diets and lifestyles, with everything from transport to dinner coming under scrutiny for its footprint. Pets are also coming under the microscope, principally for the role of their diets in the wider food system.
Quantifying the environmental impact of cat and dog food is challenging since the majority of ingredients are by-products of human food production, e.g., bone meal or grain leftovers. In the DELTA Model®, some of these ingredients are classified under “Other uses”, while some fit in the “Inedible portion” class, showing some of the challenges around assessing these commodities. Some studies allocate all impacts of production to the primary product, making the by-product footprint-free, while others allocate impact based on the mass or economic value of the ingredients.
A recent study used the economic approach to calculate that 1-3% of global agricultural emissions are on account of pet food production, with lower percentages for land and water use. Another calculated the impact of the US pet population’s diet as around 25-30% of the human population’s, including land, water, and fossil fuel use.
One estimate stated that around 140 million people could be nourished using the energy currently entering the US pet food system. However, this was purely an energy calculation, and did not include full human nutritional requirements. Moreover, it does not account for the fact that the food sources demanded by people do not match the lower quality ingredients used in pet food. However, there are increasing purchasing trends towards premium products that do include substantial proportions of human edible food.
As the impact of pet food is not negligible, there have been calls to reduce this pawprint. This impact is affected by many of the same issues as the impact of the human diet: food waste, overconsumption (and consequent non-communicable disease), and the differing impacts of different food sources. Thus, similar solutions can be tried, such as minimising waste, correct portion sizing, and inclusion of environmental impact alongside nutrition in ingredient selection.
Pet ownership is on the rise globally. While the benefits of pets are clear to any pet owner, and have measurable benefits for human wellbeing, they cannot be left out of any holistic approach to measuring or reducing environmental impact.
GOAL Sciences have recently launched their new online tool for viewing the movement of food through the current global food system. The visual, interactive nature of the tool is accessible to anyone with an interest in where food comes from, and where it ends up.
The PLANET tool uses data from the Food and Agriculture Organisation on global food production, trade, processing, and end use. The user can examine the flow of food at a global or national level, for either total food mass or total food protein.
For example, a user might be interested in cereal production. They can use the tool to see how much of each cereal crop is produced around the world (nearly 3 billion tonnes), that around half goes to processing into food, a third into animal feed, and the remainder into uses like next year’s seed, biofuel production, or is wasted along the supply chain.
The PLANET tool is complementary to the DELTA Model®: both use the same data as a foundation, so the two can be used in tandem. PLANET allows the user to visually understand how food flows in today’s world, while DELTA shows you the nutritional value of that food to the world, and lets you explore changes to the system in the future.
The impact of agricultural subsidies on climate is challenging to quantify and few such studies exist. In a recent article in Nature Communications, this relationship and its consequences have been examined.
Between 2017 and 2019 global agriculture received over US$553 billion per year in subsidies and support. This equated to approximately 10% of total farm revenue for the six commodity groups examined in this article (rice, other cereals, milk, ruminant meat, pig meat, and poultry meat). The authors coupled economic and production modelling to estimate the difference between current greenhouse gas (GHG) emissions and their state if agricultural support were removed.
The paper demonstrated that, contrary to popular commentary, agricultural subsidies have a very limited impact on GHG emissions. In fact, subsidies had a net negative 1.7% impact on GHG emissions from agriculture in the study period and gave a net 1.1% increase in farm production.
Assuming global emissions from agriculture to be approximately 25% of all emissions, this translates to less than a 0.5% impact on total global emissions. This was the result of incentivising production away from high emission-intensity, low production countries to low emission-intensity, high production countries.
The paper would be strengthened by consideration of the impact of land use change on soil carbon sequestration, which the authors recognise and recommend for future research. It also does not consider the other environmental impacts of agriculture, such as water use, and particularly does not attempt to consider how nutrition and diets would be impacted by changes to subsidies.
The authors recommend that for the purposes of reducing climate impact, use of large sums of agricultural subsidies could be better spent on R&D to improve agricultural production and simultaneously reduce emissions. However, the current purpose of subsidies is not to reduce emissions, but is based on political and economic considerations.
Food security is often accepted in wealthy parts of the world today. But with widespread concerns about trade disruption, what would it actually mean for global nutrition if food trade stopped tomorrow? Could most, some, or any countries adequately feed their populations? What does this mean for how we should think about food production?
The COVID-19 pandemic has resulted in major trade disruption, including of food trade. This and other factors have led to calls for or investigation of greater food self-sufficiency, be it at the household level, regionally, or nationally. The idea of “local food” has long attracted consumer demand for real or perceived benefits to local economies, environmental impact, and food quality, but increasingly has food security connotations. What would the nutritional implications be of restricting food availability to that produced within national borders?
Using data from the DELTA Model’s Nutrient Distribution tool, we investigated which countries and territories produced sufficient quantities of nutrients to meet the requirements of their national population. The Nutrient Distribution tool calculates the quantities of 29 nutrients available from 2015 food production and trade data for 170 countries. As per the DELTA Model approach, these values were adjusted for non-food uses, waste, and bioavailability (in the case of protein and the indispensable amino acids).
There are limitations to this approach: using 2015 data means that these results likely differ from today’s reality. Moreover, the FAO data used covers most, but not all countries, and the quality of this data varies from country to country. It also neglects trade of commodities used exclusively as animal feed, such as soyabean cake. These results should be seen as indicative of today’s dynamics and point to trends, rather than being considered a precise quantification.
National food production cannot meet national nutrient requirements in almost all countries
Only 4 out of 170 countries produced enough of all 29 nutrients to meet the needs of their own populations: Argentina, Kazakhstan, Romania, and Turkey. A further 21 countries could have been self-sufficient for all but 1-3 specific nutrients.
17 did not produce enough of any of the 29 nutrients considered. The majority of these were small, isolated island territories, such as Grenada, or countries with relatively high populations and little agricultural land, such as the United Arab Emirates.
Often, food security studies look at protein and energy as indicative of nutrient sufficiency. Even under this confined approach, only 60 countries could have been self-sufficient for protein and energy, while 71 produced insufficient quantities of both.
The nutrients most commonly falling short of requirement at the country level were the same as those that fall short globally in DELTA analysis: vitamin E and calcium, followed by iron and zinc. Only 22 countries produced sufficient vitamin E for their populations.
At the other end of the scale, those nutrients most commonly produced in excess of national population requirements were the indispensable amino acids, phosphorus and thiamine.
Without changes to food production in almost all countries, nutrient sufficiency in a closed-border world would be possible only for a small minority of countries, without considerable change in domestic production.
These results suggest far more widespread nutrient deficiency than is actually observed around the world. The reason for this is that countries do not rely solely on their own food production to nourish their populations. Next, we take a look at the same data after considering food trade.
Food trade improves nutrient availability
As can be seen from the difference between the shape and colour distribution of the two charts, food trade makes a big difference to nutrient availability. After considering trade, 125 countries were protein and energy sufficient, and no country fell short on availability of every single nutrient.
Nine countries had sufficient availability of all 29 nutrients: Kazakhstan, Romania, and Turkey still made this list, with the addition of Albania, Armenia, Greece, Israel, Italy, and Uzbekistan. Argentina, despite producing sufficient nutrients for its population, no longer featured on this list after consideration of trade due to insufficient calcium, fibre, folate, potassium, and vitamin C.
The DELTA Model shows us that globally, there are only two nutrients that are insufficiently available to meet global requirements: calcium and vitamin E. However, we have seen here how many countries do not produce enough of the right foods to meet their own population requirements for many more nutrients. This highlights the key role of food trade in the delivery of nutrition.
Most countries will overproduce some nutrients compared to their own requirements, while underproducing others. Food trade, and thus nutrient trade, then goes some way to balancing the supply of nutrients between these countries. Clearly it is not perfect, as shown by the second chart. However, the improvement food trade makes to nutrient availability compared to restricting countries to solely their own produce is clear.
How should we think about food production and trade?
The above scenarios teach us about the role of food trade in nutrient supply. The cessation of all international food trade is not realistic and would exacerbate nutrient shortfalls in almost every country in the world. We have simply asked the yes or no question of whether nutrient supply meets requirements in this Thought for Food; what is not shown is how close to sufficiency some countries may be for many nutrients. This more detailed information can be explored in the Nutrient Distribution tool in DELTA.
National food production in most of the world gives very little consideration to the nutrient requirements of the national population, so we should not expect countries to be nutritionally self-sufficient. Nor does nutritional self-sufficiency mean that a country could properly feed its own population on domestically produced food. For example, New Zealand produces far more protein than is needed domestically, but this is dominated by a far less diverse selection of food items than is demanded and required by its population.
Instead, food production globally is largely determined by demand, economics, and local suitability. If the population of Iceland demands coffee, it makes economic and practical sense to source this internationally, rather than relying on domestic production.
A study of infant malnutrition in the West African country of Burkina Faso has established a link between its prevalence and local agricultural conditions.
Climatic conditions play a major role in determining the performance of agricultural land and food production, with local and seasonal variation. The availability of food is an output of this performance, with clear implications for nutrition. In this study, the authors combined household nutrition surveys, malnutrition data and a remotely sensed drought indicator to investigate associations between agricultural conditions and malnutrition.
The Water Requirement Satisfaction Index (WRSI) combines precipitation, temperature, humidity, wind, and solar radiation data to ascertain whether the water needs of an area of cropland are being met. WRSI is a useful risk management tool employed in many parts of the world to study drought and crop yield.
Localised WRSI data was combined with household survey data from Burkina Faso, focusing on 1,721 children aged between 6 months and 2 years old. Around a third of the children were malnourished, as were a quarter of their mothers. Just 15% of the infants had achieved the WHO’s minimum dietary diversity standard over the 24 hours before the survey.
As found by previous studies, undernourished mothers were more likely to have undernourished children in their care. Low dietary diversity, recent diarrhoea, and lower education level of mothers were also associated with malnutrition.
The novel finding of this research was the association between higher WRSI values and reduced chances of malnutrition. These higher values, indicative of regions where the water needs of agricultural production were being met, were also associated with increased infant dietary diversity.
While better agricultural conditions could reasonably be expected to tally with greater food availability and thus reduced rates of malnutrition, studies that consider the interactions between agricultural production due to local climate and nutritional outcomes are rare. These results emphasise the interconnections in our food system – in this case, between rainfall and nutrient adequacy – that have a profound effect on the outcomes for people.
Encouragement from both national and international organisationsof agricultural expansion and intensification as a development strategy is found in many countries. There are economic benefits to increased agricultural production but consequences for environmental impacts. The social outcomes are less often studied. A review in the Journal of Cleaner Production has examined the social impacts of soy production, which has benefited greatly from expansion and intensification in recent decades.
Brazil accounts for a third of global soybean production, with the USA and other South American nations the next largest producers. The review addressed studies of wellbeing related to soy production, finding that the existing literature focussed almost entirely on these countries.
The positive impacts of soy production on wellbeing were mostly related to income, observed both through national statistics and direct interview with individuals in proximity to the soy industry. Increased incomes, reduced poverty, GDP rises, and development opportunities for farmers and rural communities had been found. Knock-on benefits for education and life expectancy in high soy producing areas were also recorded. Not all income impacts were positive however, with some studies finding increased local income inequality as soy production rose.
The negative impacts came in more varied forms: impact on drinking water sources, conflict over land appropriation, and food insecurity due to widespread land conversion to soy for export. Further, the ability of farmers to choose to grow other crops was reduced, as was their sense of security that they would retain their current land holdings in the future. Conversion of land to soy has also displaced families and led to migration away from soy producing regions.
The review next considered the indirect impacts of soy on ecosystem services. The increased use of agricultural chemicals and technology in soy producing areas can increase yields of multiple crops, resulting in greater local food production and availability. However, this result contrasted with evidence of food insecurity found in some of the direct studies. Land conversion had also been found to reduce wild food provision, as well as recreation and tourism opportunities, with corresponding impacts on local economies.
The authors concluded that the positive social impacts of soy production are largely confined to incomes, while the negative impacts are more diverse. Moreover, the recorded impacts of soy production showed variability between countries and even between regions within countries. There remains much to be understood in this area, such as the efficacy of government and trade policy in driving sustainable production and how this translates to social impacts. This review demonstrates the importance of considering the local social outcomes of changes to agricultural production, beyond the value of the food produced.
Dr Bradley Ridoutt, Principal Research Scientist at Australia’s CSIRO and agricultural sustainability researcher, has recently published a commentary on the challenges and opportunities for combining nutritional information with the environmental impacts of food.
Recent years have seen increased efforts to compare the environmental impacts via life cycle analysis (LCA) of food to lead diets and production down more environmentally sustainable paths. Dr Ridoutt highlights that, as yet, no universal definition or best practice for nutritional LCA exist, leading to discussion (including by the FAO) on what the best approach might be.
In a previous article, we discussed the use of LCA in food sustainability research. The main limitations of the nutritional LCA approach were reiterated in the commentary article. A key question remains: what is the function of food? And thus, how should we compare foods?
As stated by Dr Ridoutt, “Foods contain a variety of nutrients, and a healthy diet requires a variety of foods.” Defining an appropriate way to compare the worth of different foods is challenging, which makes incorporating this with environmental impact (fraught with its own challenges of what factors to include), nearly impossible. Dr Ridoutt notes that the further inclusion of the health outcomes of food, often contested, adds to this challenge.
A key quote in the article is “…wrapping environmental LCA results together with nutritional epidemiological findings would appear unlikely to inform wise decision-making and will most likely only benefit individuals and organisations with a social or commercial agenda to promote.” The author concludes that only through separate reporting of the nutritional and environmental impacts of food can trade-offs be identified and assessed. Given that the relative importance of these two impact categories will vary between individual perspectives, this area seems likely to be highly discussed for the foreseeable future.
The future of the food system is a major topic of conversation with various voices advocating the need for a range of changes, with some being quite radical. Historically the focus on the global food system was on providing enough nutrition for those in most need. This has transitioned to many advocating a need for the food system to change to address issues that are a consequence of wider human activities.
When we talk about the food system what are we referring to? What are we talking about changing? And how do we understand the potential impact of changes? This Thought for Food considers three different scales at which we can consider questions around the food system and the work the Sustainable Nutrition Initiative (SNi™) is doing to support this, in line with our recently adopted vision “evidence for sustainable food systems”.
Conversations about the future of food usually fit within one or more of the following areas:
Nutritious, affordable and sustainable diets for individuals
National use of environmental and other resources for food production
Sustainability of global food production and distribution of food to address the nutrient needs of everyone
Each scale has its own unique questions and challenges that need to be understood and addressed.
At the global level…
We will start with the large scale first. How does the world feed the world?
What food production is required to provide adequate nutrition for everyone?
How does this change as global population continues to increase?
What are the resource implications of this at a global level?
What are the impacts of large-scale changes in food production?
How big are the changes needed to be globally relevant?
In one sense the global scale question is easier to frame than the two smaller scales. There is only one Earth – there are no interplanetary trade flows to cover shortfalls or utilize surpluses – everything must balance within the planetary boundary. The total nutrition available is determined from the production of food commodities, and further processing and food formulation is largely a reorganization of the form in which these nutrients are presented to consumers.
The DELTA Model was developed to contribute to this global scale conversation by enabling a user to set the level of production for the major classes of food commodity and see the likely impact of this on nutrient supply for the global population. When we consider making changes to production on the global scale, we must also match this with an appropriate timeline of multiple years. The original version of DELTA provides significant insight on the role of the different primary production systems in providing nutrients to the global population (learn more here). A key conclusion from all our work to date with DELTA is that the global food system is, and must remain, “plant based and animal optimized.”
Further development of DELTA focuses on adding estimates of resource footprints for the selected production systems, starting with cropland use in DELTA 2.0 (learn more here). Consistent with the scenario “what if” concept of DELTA, the model does not enforce a hard constraint on production based on available land, but instead provides feedback to the user on the extent to which their proposed food system is feasible within the land area suitable for crop production.
Beyond 2.0 we will introduce additional resource footprints into DELTA, including greenhouse gases and the use of land for ruminant grazing. At the same time, we are looking at how we make better use of the more detailed food commodity and use data made available by the FAO towards the end of 2020.
An aspect that was part of the original concept of DELTA was to consider economic aspects of the food system. A key challenge that is neglected in the discussion of many high-tech alternatives is the level of capital investment required to establish these technologies at a scale that would make a material impact on global nutrient supply, and the implications of this for the cost of the resulting nutrition (learn more here and here).
At the individual level…
At the other end of the scale is individual dietary choice.
How can individuals within our society access the nutrients they need?
How is this impacted by the price of food items?
How do their choices impact on the environment?
How do choices potentially alter health outcomes?
These are questions that individuals may ask for themselves and are also relevant in the development of public policy. Potential outcomes include changes in dietary guidelines or changes in individual eating habits.
There is considerably more scope for an individual to change their diet than there is for change at the global scale. At the individual level a radical shift in diet may be made quickly and has no discernable impact on the food production system and supply chain. It is only when large collections of individuals make similar changes that this has an impact on the availability and affordability of food items, and the feasibility of these choices (learn more here). Whilst the global nutrition question is constrained by the boundaries of the planet, the individual nutrition question is limited by the range of foods available locally, and the wealth and choices of the individual.
Riddet Institute researchers used linear programming to determine the minimum cost for a nutritionally adequate diet based on a basket of food products in the USA (learn more here). Work in the Netherlands led to the development of Optimeal™, which starts from the Dutch reference diet and explores the impact of restricting the intake of certain foods on the carbon footprint and price through an optimization that seeks to remain close to the reference diet in terms of foods eaten thus addressing some of the cultural aspects of dietary change (learn more here).
Building from these ideas we aim to create an online tool to enable people to explore individual dietary choice, what levels of dietary change may be possible and practical, and the resulting impacts in any country for which the required food product information is available. The aim is to make it easy to customize this for use in any country for which the required food product information is available. Unlike DELTA, where the default time is currently 2018 for the base case and 2030 for the initial projections, the individual model (to be called IOTA) will work in an immediate sense (i.e., If I changed my diet today…).
At the national level…
In between these two extremes sits the national level question:
What is the best use of our natural resources?
How should we best utlise our land?
What should we grow or produce ourselves?
What do we export? What do we import?
How does this influence our environmental impact?
This is more complex than either the global or individual scales due to the movement of food items across borders. Food trade exists at some level for almost every country in the world: there is not the same hard boundary that exists on the global scale (i.e. only one planet) that requires domestic self-sufficiency for any given nation. Nutrient gaps can be addressed through imports and surpluses addressed through exports – although this only partially occurs today. Food trade occurs when one nation has to offer what another wants, so the options are constrained, although wealthier countries are better able to purchase what they want. As the DELTA Model shows at a global level, we already produce enough food energy and protein to meet the needs of the expected 2030 population, however not everyone has equal access.
Our current work at the national level is to understand the present state of food production and trade flows expressed as bioavailable nutrients rather than mass of food items. This provides a human nutrition centric perspective. The following charts from DELTA show, from a food energy perspective, the production, trade flows and consumption for New Zealand, Australasia, and Oceania.
The “Produced” column shows the per capita per day production of energy in the country or region. Oceania is dominated by the food exporting nations of New Zealand and Australia and there is a considerable surplus of food energy at both the regional and sub-regional scales. However, when we just look at New Zealand, we see a significant import of plant-based food energy. Similar charts are available for all 29 of the nutrients considered in DELTA, in all countries. Taking this a stage further, with our collaborators at University of São Paulo we are looking at flows of bioavailable nutrients between exporting and importing countries to show the evolution of the trade network.
Whilst the above helps to understand the current state of nutrient flows, it does not address the questions about most effective use of a country’s resources (environmental, social, economic). To do this, we are developing a proposal for multi-year research to investigate the national and sub-national questions, using New Zealand as a test-case. This will lead to the development of an integrated model framework that connects primary production and food trade decisions with their associated resource and economic implications, through to consumer nutrition and associated health and wellness outcomes. The objective is to develop a model that works for New Zealand and can, through changing the data inputs, be applied to other countries.
As with DELTA, the approach will be scenario-based, with users specifying a potential system and the model estimating the implications of this system in terms of the various capitals. By doing this we hope to contribute additional evidence-based insights to national and local conversations about food systems.
All three scales are important and complementary
The food system must function at all scales from the global, through the national, to the individual level. Each has unique characteristics and challenges that impact on our ability to make changes. Aspects that are easy at one level, create complications at another.
Importantly, no one perspective will deliver a full understanding of the food system and the way forward to greater sustainability.
An individual’s diet that is healthy, affordable and has a low environmental impact will not necessarily be applicable more widely.
National level food preferences and cultures, as well as what can be produced in a country, will mean that variation must exist between healthy and sustainable diets and food production in different parts of the world.
Similarly, it is not sufficient to understand what sustainable dynamics are for a single country, as countries do not operate in isolation: international food trade and global production must also function sustainability.
SNi is developing tools to help people explore food system challenges and provide evidence for sustainable food systems at all three scales.
A report co-produced by the FAO, UNDP and UNEP has analysed current levels of agricultural support (such as subsidies, tariffs, and incentives) in 88 countries. Their finding was that the majority of the total support value could be considered harmful, largely through distorting food prices, indirectly promoting unhealthy diets, or damage to the environment.
Around US$540 billion, or 15% of the total value of agricultural production, is paid out in agricultural support each year. These funds are split almost equally between farmer subsidies (reducing the cost of agricultural inputs or rewarding production of specific commodities) and price incentives (tariffs or subsidies on trade). A small proportion goes towards non-specific, general sector services, such as agricultural training, infrastructure development and product safety.
The authors state that over two thirds of agricultural support has the effect of distorting food prices and increasing the environmental damage of food production. This is often embodied in support for production of foods with low nutritional value (e.g., sugar cane) or high carbon emissions (e.g., ruminant meat).
Most agricultural producer support is currently focused on specific commodities. The foods receiving the greatest support were sugar, animal products and cereals, with cotton a highly supported non-food commodity. Developing countries tend to predominantly subsidise production of staple crops, whereas animal-sourced foods generally receive greater support in developed countries.
The report pushes not for the elimination of agricultural support, but rather its repurposing. This topic was also discussed at the UN Food System Summit in September 2021. Many parts of the food system have evolved some degree of dependency on the current support structures, so any changes must be carefully made to avoid unintended consequences. Consumers need access to affordable healthy diets, but producers also need to be assured of an income. The authors’ general conclusion was that agricultural support should enable environmental, health, and social progress, as well as economic gain.
Earlier this year, the FAO reported the highest global food prices in a decade, driven by a collection of factors. The economics of the food system play an important role in sustainable nutrition, but can be challenging to understand or influence given the complex system that they operate in.
The FAO Food Price Index measures monthly changes in the international prices of a selection of food commodities. The index combines global price data for 23 widely traded food commodities, including cereals, dairy products, meat products, sugar, and a number of vegetable oils, weighted by the average export share of each commodity group. The resulting index score gives an indication of global trends in food prices.
The Food Price Index in October 2021 was up nearly a third on the previous year. This has been largely driven by cereals (22% year-on-year increase) and vegetable oils (reaching an all-time high). Dairy, meat, and sugar were also up 15-40% on 2020 index values.
Cereal use is forecast to slightly exceed production over the coming year, leading to a slight decrease in global cereal stocks and the resulting change in supply/demand ratio that influences price. The forecast also predicts record high levels of international cereal trade, at 478 million tonnes annually, out of a total 2.8 billion tonnes produced.
Changes in international food prices are driven by a variety of inter-related factors. Some recent factors include reduced harvests in major cereal and sugar producing regions, migrant labour shortages in vegetable oil producing countries, and increasing fuel prices.
The complex economics of the food system are challenging to unpick, but play a central role in the delivery of food and nutrition to the global population. It is important to consider the balance of input costs, output value and demand when thinking about the future global food system. Increased trade demand, reduced harvests and external costs all have an impact on retail food prices and availability, with implications for consumers. A sustainable food system must be holistically sustainable, and economic sustainability is a key part of this.