7.8 Climate change and agriculture

Climate-change impacts on agriculture

Overall, climate change has reduced growth in agricultural productivity world-wide,^[1] although cropland area has increased (Fig 1). As one might expect, reductions in productivity growth and increases in cropland area occur together, particularly in South America and Africa.^[2] In the absence of climate change, one study suggests cropland area might have decreased over 1992-2020.^[3] Instead, climate change is a significant factor in reduced productivity growth, which is reducing progress towards food security goals, globally.^[4]

Impacts on productivity – changes in average weather variables

Average changes in temperature, water availability, and carbon dioxide have major impacts on plant productivity generally.^[5] Plants have preferred temperatures, like other organisms. Lacking strong metabolic processes, they are dependent on atmospheric temperatures to function. In addition, when water is frozen, plants cannot function, so that sets a firm bottom temperature for production, although many plants are not adapted even to survive at temperatures near freezing. Warming temperatures create a longer growing season in temperate and polar regions, but the tropics do not get this advantage, as that region already has growing temperatures year-round. Warmer temperatures will be advantageous for some species, but not for others; many areas that grow corn are already near the upper limit of temperature for that crop, whereas areas that grow wheat still have some room for warming. Agricultural suppliers are breeding and engineering crop varieties adapted to warmer temperatures to allow crops to continue to be grown in areas where they have historically produced well.

Shifts in average precipitation are better predicted than shifts in average temperatures. In temperate areas, rainfall is increasing overall, but is shifting into the non-growing season, leaving a higher risk of drought during the growing season.

In mountainous areas, more precipitation is falling as rain, rather than as snow. Snowpack is a form of water storage that releases water downhill and downstream long after the snow falls, potentially delivering water throughout the summer season. Rainfall has no such delaying mechanisms and quickly moves downstream or evaporates, reducing water supplies during the growing season. Warmer air and warmer soils lead to faster evaporation, so increases in rainfall during the growing season, when warmer temperatures are most in evidence, are often offset by increased evaporation.

The increase in CO₂ that is causing warming has the potential to provide a “fertilizing effect” in agriculture. For plants, CO₂ is a raw material used, along with water and sunlight, to create sugars that are the basis for the world’s food webs. More CO₂ could translate to more food. However, the negative effects of warming, drying, and changes in nutrient cycling generally cancel the fertilization effects except in wet areas, leaving overall negative impacts. The more severe climate-change scenarios result in greater losses of production.

Indirect effects of climate change on productivity can arise through changes in invasive species, pest and disease loads. Outside of the tropics, warming typically increases numbers and densities of invasive species, crop pests, and disease organisms. These reduce agricultural productivity. Breeding and engineering of new crop varieties and consideration of new crop species are constantly working to counter these new problems. As we have seen, adoption of new practices is slow in agriculture; new crop species will take longer to catch on than new varieties of existing crops. Other indirect effects are occurring through increased energy needs for fertilizer and pesticide production, irrigation, and other aspects of agricultural practice needed to cope with changes in weather and other crop threats.

Modeling climate-change impacts on agricultural productivity – limitations

Note the number of ways that climate change can affect agriculture. The models used to predict future climate and future agricultural productivity cannot reflect all of the latest science. Some findings are still insufficiently precise to permit mathematics to be developed to incorporate them into models. Other findings are too recent to permit time to incorporate them into models. As a result, crop-production predictions are fuzzier (lower precision of included variables) and less accurate (missing variables) than they may become in the future. Decisions often cannot wait on hoped-for future predictions, but must proceed with an understanding of the potential weaknesses of models and the areas in which research is still needed.

Impacts on productivity – changes in weather variability and extremes

Temperature and precipitation averages are changing, but variability is also increasing, which decreases predictability for producers and increases stress on plants. Crops tend to do best with intermediate temperatures and moisture. Increased variability means more time away from intermediate values and also increasing extremes, particularly hotter heatwaves, more intense rainfall and hail, larger floods, and longer droughts. Weather variability both reduces agricultural productivity and increases variability in yields.^[6] Unpredictable food supplies in turn reduce food security.

Extreme weather events do not merely stress plants but may destroy crops locally or over large areas. One late freeze that kills flower buds is enough to wipe out a fruit crop for a year. A hailstorm, driving rain event, or flood just before harvest can similarly eliminate production for a season.

Mongolia has a weather pattern called a “white dzud” in which a summer drought is followed by a severe winter with heavy snows, resulting in starvation of livestock. In 2024, an especially severe dzud resulted in the deaths of more than 7 million animals, more than 10% of the country’s livestock.

Changes in the severity of extreme weather events are harder to predict than changes in averages simply because of sample size. All the weather in a year contributes to our understanding of the average temperature or precipitation. But only the highest temperature or the total rainfall or the maximum rainfall per hour provides our understanding for the maxima for that year. Because extreme events occur rarely, we have fewer data from which to predict their behavior. Yet extremes have a greater chance of causing great harm – of killing 10% of livestock or destroying an entire harvest. So, as climate change increases weather variability, it also decreases our ability to accurately predict the occurrence and outcome of extreme events.

Increasingly, severe weather events are entirely outside of recorded weather – they represent new highs or lows, or combinations of events. Science cannot prove that climate change caused a particular severe-weather event. However, it can document trends in weather severity since climate change began, and, where sufficient data are available, science can determine how much climate change raised the odds of a severe event occurring. For example, researchers have calculated that extreme weather conducive to forest fires has become 88% to 152% more likely across the world, in forested areas, compared to the 1851-1900 climate.^[7]

Finally, extreme weather events can co-occur, bringing more than one severe stressor into play, simultaneously. Co-occurrence of heat and extremes in precipitation (both drought and flood) can cause more harm than either stressor alone, but in some cases, co-occurrence may balance one stress against the other. More frequent occurrence of these compound events challenges farm managers to understand crop responses – responses that work for a single stressor may be counterproductive for compound events.^[8]

Adaptation

Adaptation to climate change is already going on in agriculture, with mixed effects. For example, farmers in northern China have switched from soybeans to more valuable maize, which can now grow there.^[9] However, China imports more than 90% of its soybeans, presently, so the move does not contribute to food security.

A study of 6 major staple crops suggests that adaptation can help to mitigate about 23% of worldwide production losses by 2050, under the most pessimistic (but perhaps most realistic) climate-change scenario. But climate change impacts are projected to be large in current producing regions for staple crops, and adaptation has been slow to occur in part because these areas are experiencing relatively moderate climate-change impacts, to date. Maize is projected to decline by 40% in temperate zones, but only by 15% closer to the tropics, where precipitation will be greater and more reliable. Soybeans decline by 50% in the US but gain 20% in the wet tropics of Brazil.^[10]

A number of methods are available for adapting agricultural techniques to climate change, and these are typically grouped under the heading of climate-smart agriculture. As a group, these approaches seek to increase crop and soil resilience, maintain or increase productivity under climate change, and reduce emissions from agriculture to diminish harm from climate change.

Diversification can help to reduce losses from pests and extreme events. In a good year, farmers could lose money by not having all their fields in the most valuable crop. But in a bad year, diversification could provide yields that would not otherwise be available. Crop insurance, widely available throughout the agricultural world, even in developing countries, can work against diversification if insurers won’t write policies for diverse crops or new crops. Policies that support diversity and innovation are becoming more common, and could help promote these adaptive practices.^[11]

Precision agriculture can help to optimize resource use – water, fertilizer, pesticides, etc. The combination of drone imagery and GPS locations can pinpoint use of these resources. Drip irrigation can deliver water in smaller quantities, more precisely. These are high-tech solutions, but approximations might use lower technology. Training could help farmers understand the significance of variation in crop height, leaf color, and other variables, and a drone shared within a community could provide the subtler information available with advanced sensors. In smaller fields, precision use of fertilizer and pesticides might mean walking to the right location.

Regenerative agriculture provides a suite of practices designed to improve farm resilience by increasing carbon uptake, improving soil conditions, benefiting natural pest predators and pollinators, balancing soil drainage with soil water-holding capacity to best fit local water conditions, etc. Although not all of these are aimed specifically at climate change, some are, and plants that have the strong, general support systems are more likely to withstand all stressors, including climate change.

Getting the greatest food yield out of existing land will reduce carbon costs and biodiversity loss of clearing yet more land. Food-bearing cover crops, crop rotations, and intercropping can be used where possible. High-yield varieties are useful if they are designed for local conditions. These are forms of diversification – another common recommendation (see above).

Breeding/engineering resilient crops suited to specific conditions will improve producers’ ability to maximize yields. We have already seen that organic farmers lack crops bred to thrive under organic practices. Similar problems exist for agroforestry as well as for producers with soils or climates that vary from the most common conditions. Agricultural companies are breeding and engineering crops to withstand climate change and increased climate variability, but the wider range of soil and climate conditions and agricultural approaches is not yet being addressed well.

One drawback to private development of crop varieties is that they are legally protected in the way any privately developed product might be. Farmers that buy hybrid and engineered seeds are legally prevented from keeping seed of the resulting crops, so that they are locked in to buying seed every year. In the developed world, this is standard practice, but in the developing world, farmers often hold back some seed from a harvested crop to sow for the next year. With privately-developed crop varieties, not only are they prevented from sowing seeds from their crops, but offspring of hybrid crops do not retain the characteristics of the parent plants (although offspring of GMO crops will retain GMO characteristics), so that the benefits of breeding are lost.

Resilience can be extended to infrastructure, supply chains, and design of farm equipment and structures. All the parts of agrifood systems can be improved to survive increases in variability and extreme conditions. Simple improvements like windbreaks or sturdier design of irrigation systems can help, but more and better shelters for farm equipment, food storage areas, transportation systems for moving food towards processing and market are also appropriate.

Education, training, knowledge sharing and networking of farmers, support services, and policy-makers can increase adoption of successful practices, increase food yields, and encourage development of useful agricultural policies such as the crop-insurance policies discussed above. Farmers are typically described as conservative in their approaches to crop choices and practices because they face many risks and often have most of their money sunk into equipment they cannot sell if the agricultural fashions change. Both the farmers and the policy-makers can benefit from frequently updated information about new equipment, new varieties, better practices, and demonstrated successes in their regions in order to adapt to the rapidly changing conditions they face.

The food-water-energy nexus 

Agriculture, water, and energy come together in several ways. Obviously, undertaking mechanized agriculture or using synthetic fertilizers and pesticides requires all three, as we have already seen. In addition, climate-change impacts on agriculture are increased by the current use of food crops for first-generation biofuel (Chapter 6), which subtracts agricultural land from food production in favor of using it for energy production. Water use for biofuel also subtracts resources from food production, where water is limiting. Water to cool energy facilities also subtracts water from food production, although some cooling water is returned to its source in relatively high proportion. Renewable energy reduces water use and the need for biofuels, but will not be a majority share of energy production for many years.

Agricultural impacts on climate change

Agriculture presently accounts for 10-15% of GHG emissions, second only to energy production. Including deforestation, transportation, and other aspects of food systems brings the proportion up to about 30% – almost a third of world GHG emissions (Fig 2).^[12] Regionally, emissions totals are highest in Asia, followed by the Americas (Fig 3), but Indonesia and Brazil lead the world in emissions per dollar of product. Their lead is clear in farm-to-gate emissions. But in emissions due to land-use change – largely from deforestation, agriculture in these two nations accounts for the vast majority of such emissions, globally (Fig 4). Note that Figures 2 and 3 include all aspects of agrifood systems, including pre-and-post production, which is outside the scope of agriculture for the discussion here.

Palm-oil agriculture for food oils and for biofuels is responsible for the conversion of large swaths of tropical forest, particularly peatland forests. Conversion, alone, of these forests is responsible for approximately 25% of Indonesia’s total GHG emissions, and agriculture on the remaining soils also contributes significantly. Tropical peatlands are  immensely carbon rich and their conversion to oil-palm plantations, primarily in Southeast Asia, causes annual GHG emissions similar in magnitude to the C that may be lost from boreal forests in the future – 70-117 tons of CO2-eq per hectare per year.^[13] Loss of these forests is also linked to extensive biodiversity loss and economic impacts to local farmers.

The magnitude of harm caused by converting peatland forests to palm-oil plantations created considerable controversy around palm oil beginning in the early 2000s, particularly as some of it is used (especially in Europe) as a biofuel, which is supposed to reduce GHG emissions. A certification program – the Roundtable for Sustainable Palm Oil – was created to provide sustainably sourced palm oil, but after 20 years, it accounts for only 20% of available palm oil.^[14] Generating demand for sustainable palm oil has been more difficult than creating sustainable palm oil.

Agriculture promotes climate change by producing all three of the major greenhouse gases (Fig 5).^[15] Land conversion contributes most to carbon dioxide, as do some soil-amendment practices. The fact that some carbon is sequestered in agricultural land offsets a portion of these CO₂ emissions. The majority of methane from agriculture is from livestock, with manure, the moist or inundated soil of rice paddies, and burning of crop residues contributing the rest. Nitrous oxide comes primarily from use of nitrogen fertilizers, with a small proportion from use of manure and burning of crop residues. Note that nitrous oxide, which has the greatest global warming potential and the longest lifespan in the atmosphere is produced in the largest quantity of the three GHG.

As Figure 5 makes clear, the greatest gains in GHG reductions from agriculture come from reducing nitrous oxide emissions from agricultural soils and methane burps from livestock. Not only are these two gases produced in larger amounts than CO₂, but they also have higher warming potentials.

The most common recommendation for reducing nitrous oxide emissions from agricultural soils is to reduce the use of nitrogen fertilizers and to reduce the conditions that allow fertilizers to become N₂O. Precision agriculture helps farmers to put the right amount of fertilizer in the right place at the right time, which can reduce use. Placing the fertilizer at the correct soil depth for plant roots can also be helpful. In the midwestern US, reducing fertilizer application in fall would allow more of it to be taken up by plants and less to be left where it can become N₂O.

Soil compaction can contribute to N₂O emissions by holding water on the field and creating the oxygen-poor conditions that support the bacteria that convert nitrogen fertilizer to N₂O. Methods to reduce soil compaction vary by soil type, but can include no-till and use of cover crops with stout roots (e.g., radishes) that can break up compacted layers.

The most obvious way to reduce methane emissions from livestock is to reduce the number of livestock. Feed suppliers are working on ways to reduce burping by modifying animal feeds and introducing feed additives. Researchers are working to produce livestock breeds with digestive systems that produce less methane. But reducing livestock numbers would be most effective.

Manure management also contributes to methane production. In larger operations such as CAFOs, methane can be captured from manure containment areas and used for fuel. Keeping manure dry and aerated also reduces methane production, but is difficult to do if manure is piled. Livestock that are out on pasture at relatively low density deposit manure in the field where it can dry and decompose naturally, much of the time, but as we have seen, only a small proportion of livestock, globally, are on pasture.

Agriculture’s contributions to climate-change mitigation

Climate-change mitigation efforts seek (1) to reduce GHG emissions and to also (2) to remove existing GHG from the atmosphere and sequester it away from the atmosphere. Both pathways work to slow climate change. A suite of nature-based solutions to climate change has been proposed to aid in this effort, including reforestation and afforestation.

Among agricultural activities, reduced land clearing (CO₂), protection of and increase in organic material in topsoil (CO₂), reduced farming in wet conditions (N₂O and CH₄), reduced and better targeted use of N fertilizers (N₂O) and other approaches discussed above are recognized ways of reducing GHG emissions from agriculture. However, efforts to actively sequester C in agricultural soils are less well understood.

Many agricultural soils exhibit a downward trend in C levels owing to the history of agricultural practices, so even if organic material is increased, the trend may only slow, rather than reversing – soils may continue to lose C but at slower rates. .^[16] This would contribute to reducing GHG emissions but would not reduce GHG in the atmosphere – no net C sequestration would occur. Even where C increases occur, it may take many years to restore the C that was once in the soil; still, any C sequestration will reduce GHG emissions. Recent modeling suggests, however, that attempts to use agricultural soils for climate-change mitigation – as a means of removing C from the atmosphere – rarely also improve crop production.^[17]

It’s important to understand C changes throughout the soil column to assess impacts of soil practices. Many soil studies sample only the top 10-30 cm (4-12 inches) of the soil column. It’s now clear that climate change is causing loss of C in deep soils, so that experiments that study changes in organic material in upper levels of soil are not able to report the overall picture of soil C levels, and cannot accurately determine whether overall gains in sequestering C are occurring. Our understanding of deep soil C is still in its infancy and recent reviews of the dynamics of soil C under climate change refrain from making definitive statements and call for additional research.^[18] While that aspect of agricultural impacts on C sequestration is being better studied, efforts to reduce GHG emissions are certainly still appropriate!

Climate change and the agricultural land base 

Under climate change, climate in the boreal zones of Canada, Alaska, Russia, and northern China is creating weather conditions suitable for agriculture. Hundreds of millions of hectares of land currently suited to forestry is projected to become suitable for agriculture by 2100.^[19] In addition to being forested, much of this land also has peat soils, which are the largest natural carbon sink on Earth. The loss of the carbon in the tree biomass and the peat soils would significantly increase GHG levels at the global levels. As of 2018, only about 2% of the boreal peatlands – wet, boggy forests – had been drained for farming, leaving most of the region’s carbon stocks intact, but ongoing warming will make these areas increasingly attractive for food production. A 2020 study estimated that converting northern areas to farming could convert 177 Gt on C in peatland soils into CO₂ – an amount equal to more than 100 years of present US GHG emissions (Fig 6).^[20]

Peat soils are highly acidic in nature, as well as containing large amounts of organic material.^[21] Preparing these lands for agriculture requires extensive road building, deforestation, and draining. The acidic soils that result are low in nutrients and retain added nutrients poorly. To be productive, high fertilizer use will be needed, which is likely to result in nutrient pollution of surrounding waters. Weather extremes, particularly unseasonal freezing and flooding, are likely to decrease productivity.  Labor and supplies are generally less available. Lastly, many of these northern areas belong to, or are used by, indigenous people who rely on the existing forests for food, building materials, and cultural services.

At the same time that boreal lands are becoming more climatically suitable for farming, some tropical areas are becoming less suitable for both timber production and agriculture, as heatwaves, drought, and wildfires increase.^[22][23] Current predictions suggest that increases in agricultural yields in the northern land base will not replace what is lost from the tropics.

Knowledge Check

Take a moment to complete the short quiz below to assess your understanding of this section. Read each question carefully and refer to the section content as needed. This quiz is not graded – it’s simply an opportunity for you to reflect on what you’ve learned and reinforce key concepts.