P redistribution principles
For P fertiliser use, we developed four redistribution principles: (1) responsibility, (2) capability, (3) equality and (4) food security. Two levels of PB of P were used: a lower limit (6.2 Tg P) and an upper limit (11.2 Tg P) (Steffen et al., 2015) to test the variances of impacts on crop calorie production and inequality in the availability of crop calories to humans under strict or loose PB. P fertiliser application rates or reduction targets are redistributed to each of the 188 studied countries. This is done by considering their shares in the total PB of P or total required reduction targets under each principle. Currently, the global P fertiliser use rate was 19 Tg P per year during 2016 and 2018 (FAO, 2022).
The responsibility principle considered the cumulative soil apparent P surpluses per country from 1961 to 2018, and the P fertiliser reduction target was distributed to each country according to its share in the global cumulative soil apparent P surpluses during the past six decades. The capability principle builds on the ability to invest in advanced technologies to ensure food production at lower levels of P input. The P fertiliser reduction target is proportionally distributed over countries according to their shares in agricultural production but corrected by different discount factors, based on the concept of common but differentiated responsibilities (Pauw et al., 2014; Althor et al., 2016). The equality principle considered equal P fertiliser use per capita for all countries from 1961 to 2050 (i.e., the year in which the PB of P was assumed to be achieved). P fertiliser use per country will be proportionally distributed based on the shares of crop calorie production to global crop calorie production during 2016–2018 in the food security principle. A brief introduction is shown in Table 1, and for details, see the methods. The reference year for all results is 2018 for reasons of data availability.
Description of the synthetic phosphorus fertiliser redistribution principles and categorisation of countries into four different groups.
Redistribution principles or country groups
Descriptions of principles and main rationales
Each country shares the P fertiliser reduction target based on their contributions to cumulative soil P surplus from 1961 to 2018; the premise is that higher P surplus countries have a greater responsibility to reduce P fertiliser use. In addition, higher soil P surplus countries also have greater resilience to no or little P fertiliser input, since crops can utilise P from the soil.
Each country shares the P fertiliser reduction target based on their contributions to total GDP value of crop production during 2016 and 2018, since higher GDP value countries have greater capability to invest in technologies and facilities to recycle P from the food system and to implement new technologies for high yield modern crop production. However, the GDP value of different country groups has been corrected with different accounts, based on their level of economic development.
Each population in each country will share the same amount of P fertiliser per capita from 1961 to 2050, instead of based on current population.
P fertiliser use will redistribute to each country according to their share of total crop calorie production during the period 2016–2018.
(The judgement is based on the average of three years’ data between 2016 and 2018.)
Exporting and large livestock production (ExL) or small livestock (ExS) production countries
The net agriculture-exporting countries, and with the average livestock protein production per capita larger or smaller than the global average.
Importing and large livestock production (ImL) or import and small livestock (ImS) production countries
The net agriculture-importing countries, and with the average livestock protein production per capita larger or smaller than the global average.
Note that not all countries will bear their ideal reduction targets under each principle because the ideal reduction targets, in some cases, exceed the current P fertiliser application rates (since P fertiliser use is unable to be below zero). In such cases, we transferred the reduction targets of these countries to other countries under designated principles. The difference between actual and ideal P application rates has been identified, and the transfer of reduction targets has also been quantified.
To study the impacts of P redistributions on countries with different roles in agricultural product trade and level of livestock production, all the countries were categorised into four groups: i) agricultural P products (this refers to crop and livestock products) net import and small livestock production countries (ImS); ii) agricultural P products net import and large livestock production countries (ImL); iii) agricultural P products net export and small livestock production countries (ExS); and iv) agricultural P products net export and large livestock production countries (ExL) (Table 1). The livestock production level was determined by the average livestock protein production level per capita, larger or smaller than the global average level, based on three years of average data between 2016 and 2018 (Table 1).
Targets of P fertiliser use for different countries
Responsibility principle. P fertiliser reduction targets are disproportionately redistributed among countries because of differences in soil P surplus levels from 1961 to 2018 (Fig. S1). The global current P fertiliser use needs to be reduced by two-thirds under the lower-limit PB, under which the responsibility principle implies targets for ImS, ImL, ExS and ExL countries of 2.7, 0.1, 1.6 and 1.7 Tg P fertiliser, respectively (Fig. 1a). The ImL countries benefit less from the responsibility principle: their P fertiliser target is 87% lower than their current use (Fig. 2). These countries have relatively higher resilience of P fertiliser deficits or lower P input since they have extra availability of P resources embedded in the net imported agricultural products and a large amount of P in livestock manure. ExS countries benefit more from the responsibility principle with a 50% reduction target for P fertiliser use, and this may also help these countries be resilient to future low P input since they were net exports of P and have relatively lower amounts of manure P resources available for recycling (Fig. 2a).
The upper-limit PB of P is 81% higher than the lower limit; however, P fertiliser is not proportionally increased in different countries compared to the lower limit. For ImS, ImL, ExS and ExL countries, the upper-limit PB implies a 76%, 155%, 64% and 98% higher P fertiliser use than the lower-limit PB, respectively (Fig. 1a, b). This is mainly due to the transfer of P reduction targets between different country groups, with more target transfer for the lower-limit PB (Fig. 1c, d). For instance, the ImS and ExS countries together, mainly from Europe, transfer 2.0 Tg P reduction targets to ImL and ExL countries, mainly to Latin America, China and India (Fig. 1c; Fig. S2).
Capability principle. The results for actual P fertiliser use redistribution using the capability principle are similar to those for the responsibility principle (Fig. 1a, b; Fig. S3), probably because the income levels of different country groups are correlated with levels of soil P surplus. The capability principle, however, yields less intensive P fertiliser target transfer between country groups than does the responsibility principle (Fig. 1c and d).
Equality principle. The equality principle results for P fertiliser redistribution differ largely from the other principles. For example, the targets for ExL countries (0.13 Tg P fertiliser) are approximately 1/20 to 1/13 of those under the other principles (Fig. 1; Fig. S3). Small livestock production countries (ImS and ExS) only need to halve their P fertiliser use, while P fertiliser use needs to be reduced by 70% and 98% in ImL and ExL countries under lower-limit PB, respectively (Fig. 2a). This is reasonable since large livestock production is more resilient to low P input due to the large availability of P from livestock manure.
The differences in P fertiliser redistributions between lower- and upper-limit PBs are relatively large under the equality principle. Under the upper-limit PB, the small livestock production (ImS and ExS) countries barely need to reduce their P fertiliser use, with most of the reduction target distributed to countries with large livestock production (Fig. 2b). This is partly due to differences in historic P fertiliser use per capita and partly due to differences in population increases between countries in the future, for example, a more rapidly increasing population in developing countries. There were much larger transfers of P fertiliser reduction targets under the equality principle than under the other principles (Fig. 1c, d). This was due to P fertiliser use per capita being too high for a few countries during 1961–2018, and there is too short a period (from 2019 to 2050) to compensate countries with a long history of low P fertiliser use per capita. Hence, few countries need to reduce their P fertiliser application to zero, for example, most European Union countries (Fig S3c, d). There were still large differences in average P fertiliser use per capita between countries from 1961 to 2050 after massive redistribution under the equality principle, which is against the goal of individual fairness.
Food security principle. This principle largely follows the current geographical distribution patterns of P fertiliser use and yields the smallest differences in the P fertiliser reduction rate across country groups (Fig. 2). Interestingly, most African countries receive similar amounts of P fertiliser distributions compared to the current situation even under the lower-limit PB, in which global P fertiliser use needs to be reduced by two-thirds (Fig. 2). This is probably because these countries currently produce crop calories with low inputs of synthetic P fertiliser or even deplete soil P to support crop production. Targets for countries with lower crop calorie productivity, such as Brazil, China, India and Australia, are redistributed with lower amounts of P fertiliser compared to the current situation (Fig. 2e, j). This is partly due to the unique crop production structure, which is more targeted for protein production than calorie production, such as in Brazil, and in part due to over-fertilisation, such as in China (FAO, 2022). There was no transfer of P fertiliser reduction targets under the food security principle, which is different from other principles (Fig. 1c, d).
Impacts on crop calorie production and availability for people
Impacts on crop calorie production capability. Global crop calorie production could potentially decrease by 0.43–4.0*1015 kcal under the four different redistribution principles at the upper-limit PB, assuming there are no changes in P use efficiencies of crop production in the different countries and no improvement of P management of the entire food production-consumption system. The yield losses may be larger (4.1–6.4*1015 kcal) at the lower-limit PB (Fig. 3a). This means that 3–43% of the current global crop calorie production may be threatened when trying to fairly redistribute P fertiliser under PB guidelines, when there was no comprehensive improvement of P use efficiency and recycling rate. These negative impacts on crop yield could be smaller when assuming a more rational P fertiliser redistribution plan under the upper-limit PB (Fig. 3). The yield loss is smallest for the equality principle at the upper-limit PB (Fig. 3a), as a net effect of increased yields as a result of the additional distribution of P fertilisers to the ImS countries (Fig. 1–2), with much higher P use efficiency for crop production compared to other countries. The largest crop yield losses occurred using the capability principle at the lower-limit PB (Fig. 3a).
Countries react differently to the different redistribution principles and to strict or loose PBs (Fig. 3–4). ExL countries show the largest losses in crop calories for most redistribution principles (Fig. 3). This is mainly because a relatively large part of the P reduction targets are allocated to these countries. Many of the ImS and ExS countries will benefit from the food security principle, with crop calorie production doubling or tripling compared with their current situation (Fig. 4). This is because these countries are currently depleting soil nutrients for food production to some extent. These results are in line with recent studies showing that increasing P inputs to smallholder farmers in Africa could double agricultural productivity and hence reduce the global requirement of land-use change to support an 11 billion population in the future (Mogollón et al., 2021; Langhans et al., 2022).
Impacts on the availability of crop calories for the population. The current production of crop calories is unevenly distributed in the world, especially in countries with serious malnutrition (FAO, 2022). The responsibility and capability principles reduce the inequalities in per capita crop calorie production but at the cost of total crop calorie production (Fig. 3). It seems that there is a strong tradeoff between lowering crop calorie losses and increasing the per capita equality of crop calorie availability. An efficient redistribution of P fertiliser to lower losses in crop calories may increase the inequality of crop calorie production per capita, as shown for the equality and food security principles (Fig. 3).
Options to increase resilience to lower P inputs
The negative effects of P fertiliser redistribution under PB guidelines on the production of crop calories and fair availability to humans need to be alleviated. One option is to increase the international free trade of agricultural products, especially under the equality principle of upper-limit PB, since the GINI coefficient under the equality principle is higher than for the current situation (Fig. 3). Recent studies also show that rational agricultural trade patterns and recycling of P embedded in agricultural trade products in the net importing countries would increase global resilience to the deficit of P (Barbieri et al., 2021).
Our analysis also indicates that fertiliser, pesticides and energy inputs in agriculture have a strong positive correlation with crop calorie productivity in the major countries, which means that maintaining or even sustainably increasing these inputs would help to be more resilient to lower P input (Fig. 5a). Improving the crop production structure may also favour the increase in P use efficiency in a few countries (Fig. 5b). The recycling of P from existing food production systems, such as the recycling of livestock manure, human excreta and food waste, should be increased (Kanter et al., 2020; Schulte-Uebbing et al., 2021). If all countries achieved the livestock manure recycling rates of the top 5% countries in the world and collected and applied human excreta and food loss and waste to the field, an additional 27 Tg P could potentially be recycled (Fig. 5d). This amount is 2.0 or 3.4 times the amount of P fertiliser that needs to be reduced under the lower-limit and upper-limit PB, respectively (Fig. S4, S5). Increased recycling of P would also generate additional benefits for improving P use efficiency for crop production (Fig. 5b).