Long-Term Fertilizer Experiment Network in China: Crop Yields and Soil Nutrient Trends

Results are summarized for the first 15 yr of an eight-site, long-term experimental network in China designed to assess the sustainability of cropping systems in environments representing 70% of Chinese cropland. Systems were wheat-maize double cropping (two crops per year) at four sites, wheat-rice double cropping, rice-based triple cropping, and wheat or maize single cropping. Without fertilizers, wheat yields were mainly <= 1 t ha(-1), and maize yields were >= 2 t ha(-1). With NPK fertilizer (rates averaging 154, 33, and 54 kg ha(-1) per crop of N, P, and K, respectively), wheat yields mainly ranged from 5 to 7 t ha(-1), and maize yields ranged from 6 to 9 t ha(-1). Without P fertilizer, yields declined (up to 4 t ha(-1) less than with NPK), and Olsen-P values in soil declined, although rates differed between sites. Decreasing yields from withholding K usually emerged more slowly. The results emphasize the value of long-term experiments to reveal trends in soil fertility not apparent within a few years and the need for research in these environments to define "critical concentrations" of plant-available P and K in soil for maintaining maximum crop yields. Results with manures show the risk of overfertilization and water pollution with N and P if inorganic fertilizer applications are not decreased to take account of nutrients from manure. At two sites, there was evidence of significant N and P inputs from irrigation water. At one site, the addition of N fertilizer gradually caused soil acidification; this caused inefficient utilization of nutrients and led to crop failure.