黄土高原苜蓿-冬小麦轮作系统土壤水分时空动态及产量响应

Spatial and temporal characteristics of soil water dynamics and crop yield response from a 4-year of lucerne and winter wheat rotation system in the Loess Plateau

2 0 0 1～ 2 0 0 3年在甘肃庆阳黄土高原连续 3 a研究了紫花苜蓿 -冬小麦轮作系统中土壤 0～ 3 0 0 cm剖面水分动态特征 ,作物产量及其含 N量。处理包括 4龄苜蓿草地 (L C)、4龄苜蓿草地后茬持续休闲 (L F)、4龄苜蓿草地休闲 4个月后种植冬小麦(L Fl W) ,和 4龄苜蓿草地休闲 1个月后种植冬小麦 (L Fs W)。结果表明 ,种植 4a苜蓿后春季翻挖实施休闲至冬小麦播种(L Fl W)的 4个月期间 ,降雨的入渗深度为 150 cm,而苜蓿秋季翻挖休闲至小麦播种 (L Fs W)的一个月间 ,降雨在土壤内的入渗深度为 90 cm,不同休闲长度对头茬冬小麦土壤 0～ 90 cm水分贮存量无显著影响 ,亦不影响头茬冬小麦的出苗和出苗数。苜蓿后茬完全休闲 (L F)一个生长季后 ,60～ 90 cm土壤水分含量达田间最大重力持水量 (Drainage U pper L imit DU L )的 93 % ,0～ 3 0 0cm剖面土壤贮水量达 670 mm ,是剖面田间最大重力持水量 (DUL )的 78% ;L Fl W和 L Fs W处理下头茬小麦籽粒产量之间差异显著 (P<0 .0 5) ,收获指数和千粒重等指数无显著差异 ;L Fl W和 L Fs W处理中获得的二茬小麦产量无显著差异 ;连续种植苜蓿与种植小麦有接近的生物量 ,但苜蓿地植物总 N的输出量较小麦田高 2～ 3倍。由于黄土高原降雨变率大 ,因此预测土壤含水量动态有?

Legume-cereal rotation is a major component of sustainable dryland farming systems in the Loess Plateau, China. Understanding the water dynamics of rotation after lucerne is therefore critical for crop sequence selection. Field experiments were undertaken from 2001 to 2003 in the Loess Plateau, western Gansu Province, China, to examine the water dynamics in rotation with winter wheat (WW) after a 4 years lucerne ley. The treatments were: lucerne ley (LC), lucerne removed for continuous fallow (LF), and the first winter wheat sowing 4 months after 4 years lucerne removed (LFlW), and the first winter wheat sowing 1 month after 4 years lucerne removed (LFsW). The soil water content from 0～300 cm profile was measured by NMM method in situ. The yields both for winter wheat and lucerne were determined and plant N content was analyzed in lab. In the field plots of LFlW where lucerne was removed, during May and September, the rainfall penetrated to 150 cm in the soil profile, while only reaching 90 cm for the LFsW treatment (ie. only one month fallow before WW sowing). As the fallow duration after lucerne did not affect soil water storage from 0～90 cm profile, winter wheat establishment under 2 fallow lengths after lucerne had a similar performance. The soil water storage for 0～300 cm in the field plots of LF treatment reached 670mm which accounted for 78% of DUL (Drainage Upper Limit) after land fallow for one season. The soil water content at 60～90 cm depth was 93% of the DUL at the same period in the LF plots. A significant difference in winter wheat grain yields between LFlw and LFsw was found, but no significant differences were detected for Harveting Index (HI) and 1000 grain weight in 2002. The second season yields of wheat were not significantly different between the 2 different fallow lengths in 2003. The total biomass for winter wheat and lucerne was 8.7 t/hm^2 and 13.5 t/hm^2 in 2002, 11.4 t/hm^2 and 12.1 t/hm^2 in 2003, respectively. The total nitrogen harvested, however, from lucerne was 2～3 times higher than that from winter wheat. As the Loess Plateau is subject to high rainfall variability, the prediction of soil water change is important to make decisions on crop sequence in rotation systems and to minimise production risk.