Xia Bing, Ren Xiaoming, Du Mingwei, Xu Dongyong, Yin Xiaofang, Tian Xiaoli, Li Zhaohu
High plant populations and multiple applications of the plant growth retardant mepiquat chloride (MC) are required in the production of mechanically harvested cotton (monocropping) in the Yellow River Valley. Cotton-stalk recycling to the field is one of the main approaches to avoid the further increase in fertilizers during cotton production. However, the optimum nitrogen (N) rate and whether N use efficiency would increase under these conditions remains unclear. We conducted experiments involving cotton-stalk recycling to the field at Hejian in Hebei Province (38°41′N, 116°09′E, and 11 m elevation) during the 2013 and 2014 growing seasons. In 2013, there was a considerable amount of precipitation (279.1 mm) in July (flowering period); however, in contrast, there was only 42.4 mm precipitation in the same month of 2014. We investigated the effects of plant density, MC application and N rate on cotton yield and its components, accumulation and distribution of biomass and N, as well as N use efficiency. In comparison with the low plant density treatment (6.75 × 104 plant·hm-2, LPD), the seed cotton yield and Npartial factor productivity (PFPN) under the high plant density treatment (11.25 × 104 plant·hm-2, HPD) significantly increased by 8.1% and 7.4% in the drier year (2014). In addition, the N apparent recovery efficiency (REN) of HPD also significantly in- creased to 41.6% in 2014. The MC application significantly increased the seed cotton yield (39.2%), PFPN (43.3%), and agronomic efficiency (212.8%) relative to the water control (CK) in the wetter year (2013). The REN of MC was also greater than that of CK in 2013 but this difference was not significant. The HPD showed increased dry matter, but a decreased harvest index (HI) in 2013. Although less dry matter was produced, the MC increased HI in both the wetter and drier years. The N rate ranged from 0 to 315 kg·hm-2 did not significantly affect the seed cotton yield, but there was only a tendency that the low N rate (105 kg·hm-2, LN) had a higher yield compared with the medium (210 kg·hm-2, MN) and high N rates (315 kg·hm-2, HN). Therefore, the LN treatment had the highest PFPN, which was 24.5 and 54.4 kg·kg-1 in 2013 and 2014, respectively. The REN of LN was 45.2% and 41.0% in 2013 and 2014, also significantly greater than those of MN and HN. In summary, the combination of HPD and MC application was in favor of the biomass accumulation and HI increase in addition to the yield stability of mechanically harvested cotton in the Yellow River Valley, and the optimum N rate could be reduced to 105 kg·hm-2 when stalks are returned to the field.