The present paper summarizes the results from a long-term experiment setup in 1980 in the Taihu Lake region, China, to address the yield sustainability, the dynamic changes of soil organic carbon (SOC) storage, and soil fertility in the rice–wheat ecosystem. Treatments in three replicates comprising manure-treated and chemical fertilizer-treated groups (two factors), each having seven sub-treatments of different combinations of inorganic nitrogen (N), phosphorus (P), potassium (K), and rice straw, were randomly distributed. Results showed that the treatments of manure (pig manure from 1980 to 1996 and oil rape cake thereafter) + N + P + K (MNPK) and chemical fertilizer + N + P + K (CNPK) produced the highest and the most stable yields for both rice and wheat within the respective fertilizer treatment group. Potassium fertilization was necessary for yield sustainability in the ecosystem. Treatments of straw (as rice straw) + N (CRN) and manure + straw + N (MRN) produced more stable yield of rice but less stable of wheat. It was therefore recommended that straw should be only incorporated during the rice season. SOC contents in all treatments showed increasing trends over the period, even in the control treatment. Predicted SOC in chemical fertilizer-treated plots (mostly yet attainable) ranged from 16 to 18 g C kg−1, indicating the high carbon (C) sequestration potential of the soil as compared to the initial SOC. SOC in manure- or straw-treated plots ranged from 17 to 19 g C kg−1, which had been attained roughly 10 years after the experiment was initiated. Nutrient balance sheet showed that there was P surplus in all P-treated plots and a steady increase in Olsen-P over a 24-year period in 0–15 cm soil, which contributed little to crop yield increases. It was therefore suggested that P fertilization rate should be decreased to 30–40 kg P ha−1 year−1. Comparison of yields among the treatments showed that wheat was more responsive to P fertilizer than rice. Thus P fertilizer should be preferably applied to wheat. Soil pH decrease was significant over the 24-year period and was not correlated with fertilizer treatments. The overall recommendation is to incorporate straw at 4,500 kg ha−1 year−1 during the rice season only, with additional 190 kg N ha−1 year−1, 30–40 kg P ha−1 year−1 mainly during the rice season, and 150–160 kg K ha−1 year−1. Further research on the unusual P supply capacity of the soil is needed.
