Figure 2. Historical and hypothesized P budget and PUE trends. (a) PUE versus P yield from 1961 to 2014. PUE larger than 2 (200%) is not shown. The greyscale indicates P surplus (kg P ha-1). (b) PUE versus accumulated P residual from 1961 to 2014. PUE larger than 2 (200%) is not shown. High- and middle-income economies’ (The World Bank 2017) data are in blue and green, respectively. Malawi’s data are in red. The pink area indicates soil mining (negative P surplus). Data have been smoothed by ten years to limit the data variation influenced by weather conditions. Seven data points represent the data in 1961, 1970, 1980, 1990, 2000, 2010, and 2014. (c) The hypothesized relationship between P surplus and GDP per capita. (d) The hypothesized relationship between PUE and GDP per capita.
Different from the global trends, the historical patterns of regional P budget and PUE vary by country, and their historical trajectories suggest a typical pattern of PUE as countries increase their P yield, accumulated P residual, and income levels. Malawi, as a low-income country, has had minimal success in increasing its yield in the past 50 years, and it relied mainly on mining native soil P (Fig. 2a). Middle-income economies like Brazil, China, and India, have had increased crop yield at the expense of nutrient use efficiency and managing P surplus (Fig. 2a). However, their accumulated P residual was still lower than that of France since 1961 (Fig. 2b). The relatively high P residual accumulated in France, lower accumulation in the U.S., and the increasing accumulation in China and India due to large P surplus in recent years were also found in a previous study estimating P residual since 1900 (Zhang et al. 2017). Our estimation differs from theirs on the starting point of estimation and assumptions for initial values. As middle-income economies continuously improve their soil fertility and yield, they may still have low PUE and high P surplus in the next few years or even decades. High-income economies such as the U.S. and France, have experienced yield enhancement in the 1960s, relying on increased fertilizer input. They transited to the second phase of yield enhancement and surplus reduction in the 1970s and 1980s, which benefited from improved PUE (Fig. 2a).
Based on those observations (Fig. 2a and 2b), we hypothesized that the relationship between GDP per capita and PUE follows the Environmental Kuznets Curve (EKC) (Dinda 2004, Zhang et al. 2015a). EKC suggests that P pollution first increases with GDP per capita at the early development stage. It starts to level off and decreases at a later stage due to technological development and increased environmental awareness. Accordingly, as the economy develops, the P surplus first increases and PUE first decreases, leading to more P accumulated in the soil or lost to the environment (Fig. 2c and 2d). P surplus then decreases, and PUE improves, as more P inputs are converted to crop products (Fig. 2c and 2d). We consider that the PUE of a county (e.g., the U.S. and France) may exceed 100% during the sustainable intensification, as the crop production may mine some accumulated residual P before reaching a relatively steady state (Fig. 2d). For each country, the whole EKC process may take a longer time than current databases have recorded, and many other factors (such as social and climate change, and global collaboration) can affect its development and change rate. Thus, the suggested U-shape curve is clearly seen on a global scale but is not obvious in individual countries (Fig. 2a and 2b).
The historical data of P budget and PUE also indicate low-efficiency countries and crops policymakers need to prioritize when improving P management. Comparing the PUE of 12 regions in 2010, China and India had relatively low PUE (<40%) and high P inputs, contributing over half of the P surplus to the global environment (Table 3). On the global scale, rice, fruit, and vegetable had relatively low PUE (<40%) and high P inputs, leading to a global P surplus at 9 Tg P yr-1 in 2010 (Table 3).
Table 3. P budget and PUE in crop production by region and crop group in 2010 (average of 2005-2014) and in 2050 in the three scenarios (BAU, MPA, and HPA). The unit of each budget term except for PUE is Tg P yr-1. The unit of PUE is %. The definitions of the 12 regions are in Table S17. The definitions of the 11 crop groups are in Table S12. Value 0.0 means it is smaller than 0.05.