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

Bellotti W.D
DOI: https://doi.org/10.3321/j.issn:1000-0933.2004.03.035
2004-01-01
Acta Ecologica Sinica
Abstract: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.
What problem does this paper attempt to address?