Sunday, August 8, 2021

Nitrogen fertilizer is not a climate solution

It is often claimed from proponents of conventional farming that the use of chemical fertilizers, in particular nitrogen, is a prerequisite for binding more carbon in soils (see for instance the critique of regenerative agriculture by the World Resource Institute).

This is normally founded on two different arguments: 1) Through the use of nitrogen fertilizers total photosynthesis is increasing, thus more carbon is bound by plants. Some of this carbon is taken away from the land in the form of higher yields, but there will also be more straw, roots and residues left in the field; 2) There is a rather narrow relationship (stoichiometry) between carbon and nutrients like nitrogen and phosphorus in soil organic matter and, therefore, one have to supply concomitant quantities of those nutrients to increase carbon content.

Below, I discuss both of the arguments.

It stands beyond doubt that the supply if nitrogen fertilizers, in reasonable quantities, increases yields. It is often assumed that this also corresponds to more straw, roots and other organic matter being left in soils. This seems to be an oversimplification though. To begin with, in parallel to the increased use of fertilizers, the harvest index (the share of the above ground biomass that is the desired products, e.g. wheat kernels) of major agriculture crops has also changed a lot so that a bigger share of the carbon is allocated to the harvested crop and less to other parts of the plant. In addition, through the use of herbicides the biomass in weeds has also been substantially reduced, biomass that otherwise would contribute to soil organic matter.  

The harvest index is also influenced by management and experiments show that
increased availability of nitrogen as well as irrigation will increase the harvest index even more. Research in both Switzerland and Denmark demonstrate that in organic farming (by definition without the use of synthetic nitrogen fertilizers) a substantially higher share of the biomass is allocated to roots than in conventional farming.

This is even more important as recent research (e.g. Kätterer et al 2011 and Villarino et al 2021) show that roots and root exudates (carbon rich substances released from the roots to the soil, sometimes referred to as the liquid carbon pathway, or the microbial carbon pump) are much more important for building soil organic matter than plant litter and straw. 

There are many review articles and meta-analyses published on the topic and they come to varying conclusions. Poeplau concludes that there is a general positive relationship between the use of nitrogen fertilizers and soil carbon in grassland, but there are also research showing the opposite, e.g Sochorova 2016. A synthesis of Bolinder et al (2020) claims that 1 kg of N can increase soil C with 1 kg compared to plots which are totally unfertilized. They conclude, however, that this is not a linear relationship, i.e. there is higher effect with low nitrogen supply than with a high supply. They also say that to compare with plots that are not fertilized at all has limited practical value as farmers, in the absence of chemical fertilizers, will try to supply nitrogen in other ways, such as with biological nitrogen fixation or use of organic amendments. They point out that many other methods than increased use of fertilizer to increase soil carbon is much more important. Another synthesis from 2021 concurs.  Therein, Alexandra Tiefenbacher and colleagues show that adding N can work both ways depending on conditions. It can stimulate the decomposition of organic matter and thus reduce the carbon stock, but it can also increase primary production and thereby the supply of carbon.

Based on the research above it seems quite clear that use of N fertilizer is no shortcut to soil carbon sequestration and that its effects are uncertain, and in any case very small. If we lift our perspective this should be apparent already from the fact that most soils have been losing carbon all throughout the era of increased use of nitrogen fertilizer. On a systems level, use of nitrogen fertilizers is clearly not a pathway to increased soil organic matter.


Let’s now look at the other argument. Do you have to supply a certain amount of nitrogen to increase the carbon stocks in soils? This is based on the observation that a C:N ratio of around 10 is a benchmark for organic matter in soils, i.e. for each kg of carbon there is 100 gram of nitrogen. In a very simplistic way this is interpreted into that in order to increase soil carbon stocks with, say, 1 ton we need to supply 100 kg of nitrogen.  

But while these relations indeed have some value one can’t draw the arguments thus far. The C:N ratio differs a lot between different types of organic matter and even within the organic matter itself. In plants it vary the most while the range is narrower for microorganisms. Recent research has shown that dead microorganisms make up a substantial part of the soil organic matter. But the micro life will also adapt itself to soil conditions such as the availability of nitrogen. Fungi has typically more than two times higher C:N ratio than bacteria for instance. Also within a certain group of organisms species thrive under different conditions. A high C:N ratio will stimulate bacteria which preserve nitrogen (and reduces emissions of nitrous oxide) while N- fertilization will favor the opposite.

Again, if we look through a wider lens we can see that organic soils have almost 5 times higher C:N ration than mineral soils and that there is also a marked difference between arable soils and grasslands. In addition the C:N ratio changes with soil depth. Summing up, the stoichiometric argument for nitrogen fertilizers is weak.

Both arguments are based on a static view of agriculture systems. If you start with the prevailing agriculture system, which is adapted to the regular supply of nitrogen, and just cut out nitrogen fertilizer, you will get all kinds of problems, of which reduced yields is the dominating one. But farmers will respond to changed conditions by changing management, crop and variety selection and crop-livestock integration. It is therefore not really possible, or at least not meaningful to discuss one component of an agriculture system within the framework of ceteris paribus, all else being equal.

In addition, already today, nitrogen use efficiency in the agriculture system is low and more than half of the nitrogen supply is lost in the fields. There is no research showing that increased use of nitrogen fertilizers would be a good pathway for increasing soil carbon stocks. Even with the assumed carbon storage efficiency of nitrogen fertilizer of one to one (or even slightly better) it would still be a “climate loss” to increase soil carbon through the use of nitrogen fertilizer because the whole life cycle emissions (in the range of 10 kg CO2e) from 1 kg of N surpasses the climate value of 1 kg of C (which is 3.66 kg CO2). Few, if any, of the proponents of nitrogen fertilizers as an important factor for carbon sequestration, are actually suggesting that nitrogen should be supplied in higher rates to cater for that. The effect on yields will still determine nitrogen supply.


The nitrogen argument for carbon sequestration could, therefore, be seen as a distraction or possibly a method of diverting focus away from the substantial greenhouse gas emissions associated with their production, transportation and use.

Nitrogen fertilizers are not just a technicality, it is a major building block in the industrial, global and capitalist agriculture system. As such they both drive and enable the increasing metabolic rift between human society and the ecosystem that sustains it. It is hard, but definitively possible to feed the global population without them, but it will need changes in the food system and in society at large.  

Tuesday, July 13, 2021

The nitrogen challenge for organic agriculture

Three new research articles wrestle with the question of nitrogen (N) availability in organic farming. All three are based on the indisputable fact that nitrogen availability is the factor that makes the biggest difference between organic and conventional farming. It is often claimed that half of the global food supply is made possible by the use of chemical fertilizers, but is it true that we will starve if all farms in the world were organic?

First, let us be clear that in the current situation starvation is widespread despite tremendous increase in crop production (see graph). In essence, who gets to eat and who doesn’t has very little to do with brut production and everything to do with power, equity, justice and empowerment. Of course, there are limits for how many people the planet can feed and that we shouldn’t appropriate a too big share of the primary productivity of the ecosystems. No agriculture system can feed a population that keeps on growing infinitely.

Still, it is prudent for those who promote a wide scale conversion to organic to make plausible that it is indeed possible for organic farming systems to have global adoption and feed the (growing) global population.

Anyway, back to the new research.

Pietro Barbieri and colleagues concludes, in an article in Nature food, that in a fully organic world food production would go down with 36 % of dietary energy needs with nitrogen shortage the key determining factor. Protein would not be a problem though.

Turning to the paper Agroecological measures and circular economy strategies to ensure sufficient nitrogen forsustainable farming by T.G. Morais and colleagues in Global Environment Change, the picture changes. They conclude that with sufficient mitigating measures 100 % organic is still a feasible option on a global scale.

A third paper on the possibility to reshaping the European agro-food system published in One Earth by a team headed by Gilles Billen claims that Europe can be self sufficient in a fully organic diet, with a huge reduction in the consumption of animal products.

All three papers are, quite naturally, based on modelling and the results are basically determined by the input data as well as the assumptions made and constraints imposed. Below I compare some of the main features of the papers using the surname of the main author as id. In order to help the reader, I occasionally put in brackets (Global organic impossible) for Barbieri, (Global organic possible) for Morais and (Europe organic possible) for Billen. Obviously I have to make a lot of simplifications, especially as Barbieri and Morais have multiple scenarios in their models. For instance, Barbieri has scenarios for 20, 40,60 and 80 percent organic as well as waste reduction and diet change, in total 216 combinations and Morais explores a set of improved practices as well as eight different diets in various combinations. 

Population and diet. Barbieri has the current population of ~7,3 billion getting sufficient energy and protein, Morais the global population 2050 with 8 different diets. Billen explores a European (Ex Ukraine, Belarus and Russia) population of 601 million eating a “healthy diet” limited by protein intake of 5 kg N per person per year (corresponding to approximately 85 gram per person per day) and a cap of 25% of protein from animals, compared with current 55%. 

Land constraints. Billen: No changes in land cover for food agriculture. Morais: 5 different cropland scenarios from current acreage to 70% expansion, no grassland expansion. Barbieri: cropland and grasslands as in year 2000.

Localization of production. Morais (Global organic possible) define production regions and allows trade between then. Barbieri (Global organic impossible) studies the production in grid cells which are independent of each other so that all production as well as the inputs come from the same area, food can be traded however. Billen (Europe organic possible) works with regionally adapted production systems where livestock is fed locally. Trade within Europe can take place when local demand is satisfied. Extra European trade is very limited and Europe is a net exporter compared to being a net importer today.   

Production scenarios and yields. Barbieri assumes that yields in both organic and conventional is a direct function of N supply and that organic and conventional yields differ mainly in areas with high yields while the yields are similar in permanent grasslands and crop lands with low level of N inputs. Morais assumes that organic yields are 80% of conventional and that 25% of the area is used for biological nitrogen fixation. Billen uses a generalization of four prevailing crop rotations for organic and uses the relationship between yield and nitrogen supply to determine organic yields.

Recycling of human excreta. Billen assumes 75 % recycling. Barbieri has no recycling in the main 100% organic scenario, but also a scenario with 10 % recycling. Morais assumes considerable recycling of human excreta and waste from the food system.

Other relevant constraints, assumption. Billen sets a maximum of 35 kg of N surplus per hectare and limits to the use of human edible feed to livestock as well as no feed imports.  Barbieri assumes more than a doubling of grazing from permanent grasslands, not by expansion of grasslands but by increasing grazing intensity.

Knowing these assumptions and constraints it becomes considerably easier to understand the difference in results and conclusion. The main difference between Barbieri (Global organic impossible) on the one hand and Morais (Global organic possible) and Billen on the other is that Barbieri assumes fewer and less effective mitigating measures. Billen (Europe organic possible) assumes substantial recycling of waste from food system and a tripling of biological nitrogen fixation. Morais count on a rather radical improvement in nitrogen use efficiency (NUE) covering around half of the anticipated nitrogen shortfall and improved/increased biological nitrogen fixation for most of the rest. Without these improvements organic can only feed the world through a considerable expansion of cropped area, up to 70 % depending on the assumed diet.

Even if Barbieri concludes that organic can’t feed the world they conclude that in some food insecure developing countries a full scale conversion to organic would improve food supply compared to the current situation.

All papers discuss the change of diet as a result of the limited N-supply. In Billen this is already set as a constraint. Barbieri concludes that an organic diet would mean a doubled quantity of dairy and four times more pulses and big reduction in the consumption of meat and vegetables. Nutritional composition of diets are projected to improve. 

Morais’ paper shed quite some light on the trade-offs between diets and production methods when it comes to environmental impact.  Fully optimized organic production, i.e. after implementation of all improvement strategies, is associated with less GHG emissions and a lower N surplus than conventional production, regardless of dietary choices. They conclude that “Overall, there is no combination of diet variant and production variant that is significantly better than all the others in all indicators (land used, GHG  emissions and N surplus).“ The paper also demonstrates that strategies to spare land through intensification of (conventional) product will increase green house gas emissions as well as nitrogen pollution.

Both Billen and Morais point to that livestock, particularly ruminants play an important role in organic production systems. This is not because they supply manure per se (manure is after all just recycled crops), but because they can use co-products and by-products, they can utilize permanent grassland which means that N is imported to the food system and they offer better crop rotations which can improve also crop production.

The focus on all three studies is the provision of food to the human population, but they also draw some conclusions regarding health and environment, in particular regarding nitrogen. Reactive nitrogen is not only an essential nutrient for plants and the building stone of proteins, it is also an environmental threat of global proportions almost on par with greenhouse gas emissions (to which it also has strong links). Barbieri project a 77% reduction of N losses to the ecosystems while the main focus of Billen is to reduce N losses to the biosphere and their results show a reduction to the half of current emissions. According to Morais, N losses can be reduced with 70% and greenhouse gas emissions can be halved in a fully organic scenario. 

Even though it is not the main focus of the articles, they also address trade, directly or indirectly. For instance Billen (Europe organic possible) writes: “While not excluding food trade when required, the scenario privileges local food supply. This implies giving up land specialization in favor of a multifunctional conception of land planning.”

In my view, all three papers have some weaknesses (it would be strange if they hadn’t). Barbieri (Global organic impossible) applies a far too static perspective on the response of farmers and society. This is unrealistic and undermines their main conclusion – that organic can’t feed the world – not credible. It is quite obvious that large-scale recycling of waste from the food system, including human excrements would be a necessity in a 100 % organic scenario. Also, their projected increase in biological nitrogen fixation is just 4%, why so little is not explained. Their assumption of a huge increase of grassland contribution to the food system should have been better explained to be credible.

Morais’ (Global organic possible) scenario for vegetarian food assumes that the animals are not eaten. While I respect that some people simply don’t want to eat meat, I believe that a vegetarian scenario where cows and calves are just left to die instead of eaten is largely unrealistic as well as undesirable. It is not even working in India. Interestingly, they discuss this in the final part of the paper where they also conclude that a “ovo-lacto vegetarian” diet where people also eat the used hens, cows and their calves would perform equal to the vegan diet. Nitrogen use efficiency tend to decrease with increasing supply, so in that sense it is quite obvious that in an organic world nitrogen use efficiency will be higher. Nevertheless, the expectations on the possible improvements in nitrogen use efficiency in Morais paper is not explained well, instead they just make reference to another article supposedly demonstrating that. Taking into account that the improvement in nitrogen use efficiency is the cornerstone in their paper, that is disquieting.

In the paper of Billen I am a surprised by the very high share of fruit and vegetables (assumingly including potatoes) included in the diet, 15 % of the nitrogen is supposed to come from them. I don’t think it is likely at all that consumption of fruit and vegetables can increase substantially in an all-organic scenario (the conclusion of Barbieri is that it will shrink with 65%). They are very susceptible to pests and they are spoiled plants requiring a very ample supply of nutrients. As far as I can see their substantial nitrogen demand – and losses - are not considered in the research. By and large it would have been interesting if their analysis had not been limited to a prescribed diet. Another question mark is that it doesn’t include the opportunity to expand grassland areas or intensify grassland use. In many parts of Europe huge tracts of grasslands have been abandoned the last century and in many parts grazing is underutilized because it is cheaper to buy (feed) feed than to use grazing and to breed monogastric animals instead of ruminants. Unfertilized permanent grasslands are hot spots for biodiversity and abandonment of them pose major bio-diversity threats. They assume instead a 46% decrease in the production from grasslands.  That may be a realistic assumption for grasslands which today are fertilized with synthetic N with no other changes in management. But in many European countries just a small fraction of the permanent grasslands receives any fertilizers.    


What if? is a human way of thinking. Earlier this thinking was in the realm of culture and imagination, but with increasing computational power we can make more and more sophisticated models. My reading of the three articles demonstrate quite well that models of this kind shouldn’t be referred to as evidence of that one or the other opinion is correct. Food and agriculture systems are far too complex to allow for such conclusion. One change, such as the elimination of synthetic nitrogen triggers a cascade of changes, which in turn cause new changes.


Having farmed organically for more than forty years, I can agree with the basic tenet of all three articles; that N-supply is a major challenge for organic. Of course, if you grow vegetables, fruits and berries, pests can be a bigger challenge some years for some crops. Under some circumstances phosphorus or potassium supply can also be a problem.

By and large, all three papers demonstrate that a 100 % organic scenario requires substantial changes in the food system from production to consumption. This should be no surprise. The availability of a rather unlimited supply of nitrogen has clearly steered the food system in a certain direction that goes all the way from how we (don’t) handle human waste to enormous increase of the consumption of chicken (per capita consumption of chicken has increased more than five times since 1961, while beef production has remained the same). It has also been a prerequisite for the globalization of food as it made possible the constant export of nutrients from one region to another. Clearly it is totally impossible to take this system and just exclude nitrogen fertilizers. But farmers and societies will adapt and adjust to new conditions. The changes needed include but are not limited to:

  • Recycling of organic waste from all parts of the food system, including human excrements.
  • Reducing food waste.
  • Integration of livestock and crop production.
  • Considerable use of permanent, non-fertilized grasslands. This doesn’t necessarily mean expansion of grasslands, but rather better use.
  • Expanded use of biological nitrogen fixation through, among others, the cultivation of leguminous plants. This is not limited to peas and beans for direct human consumption but also clover and alfaalfa for forage, the growing of leguminous plants as cover crops or living mulches and leguminous trees in forest gardens, permaculture, silvopastoral or agroforesty systems.
  • Adaptation of diets to what works well in an organic production system. Notably an appropriate organic diet is not the same in all parts of the world. The notion that there is one good global diet with largely the same composition is socially, culturally and ecologically inappropriate. On the contrary, they diet should be adapted to what can readily be produced locally.

Instead of describing it as a set of technologies or practices it can better be seen as a change in the whole food system based on the integration of food production and consumption into the local ecological context. As I explained earlier these changes will trigger new changes. In the particular case of synthetic nitrogen, we do have history to learn from. Many of the measures needed have been practices for centuries. A circular food economy is nothing new but rather standard practice for centuries, now discarded by the capitalist market-economy. Global flows of nutrients are largely incompatible with the closing of the nutrient cycles. That doesn’t only apply to trade in feed for animals but equally for food to people. This points to a relocalization of the food system, and ultimately, at least a partial reruralization of societies.