2.1 Site description and fertilization strategies
Soil samples were collected in April 2015 (before early rice transplanting) from long-term fertilization experiment stations at Ningxiang in the subtropical region of the Hunan Province, China (111°54’−112°18’E, 28°07’−28°37’N). The mean annual temperature and rainfall of the site are 17°Cand1681 mm, respectively. The soil was derived from river alluvium, Quaternary red clay, and plate shale, and the cropping system used was milk vetch–rice–rice. The main physical and chemical properties of the soil at the start of the long-term field fertilization trials were as follows: pH 5.8, SOC17.1 g kg−1, total N 1.76 g kg−1, available N 144mg kg−1, and Olsen-P 12.8 mg kg−1 (Yuan et al., 2012).
The experiment started in 1986 with the following four treatments: no fertilizer (CK), mineral fertilizer (NPK) (urea, superphosphate, and potassium chloride), rice straw combined with mineral fertilizer (NPK + ST), and 70% NPK +30% chicken manure (NPK +OM). The rice straw produced from the straw plots was returned to the corresponding plots in July and November after the early and late rice harvests at rates of 2775 kg ha−1 a−1 and 3600 kg ha−1 a−1, respectively. The experiment design ensured that all fertilizer treatments received the same amount of N, phosphorus pentoxide (P2O5), and potassium oxide (K2O). The total amount of N, P2O5, and K2O was the sum of the chemical fertilizer plus that from the rice straw residue or manure, corresponding to143 kg N ha-1, 23.2 kg Pha-1, and 52.3 kg K ha-1; and 158.0 kg N ha-1, 18.6 kg P ha-1, and 67.2 kg K ha-1per fertilized treatment for the early and late rice seasons every year, respectively. The total amounts of fertilizer applied in each fertilized treatment during the milk vetch season were identical to those for the late rice season.