基于不同水肥组合的春玉米相对根长密度分布模型

Normalized root length density distribution model for spring maize under different water and fertilizer combination

根系分布是模拟作物生长与土壤水分和养分运移,以及制定合理的灌溉制度和作物管理不可缺少的参数之一。现有的根系分布模型大多是于单一灌水和施肥条件下建立的,因此研究不同水肥组合下滴灌玉米的根系分布模型更具有实际意义。利用2a的田间小区试验,以春玉米"强盛51号"为试验材料,设置4个灌水水平,2015年和2016年分别为I_(60)(60%ET_c)、I_(75)(75%ET_c)、I_(90)(90%ET_c)、I_(105)(105%ET_c)和I_(60)(60%ET_c)、I_(80)(80%ET_c)、I_(100)(100%ET_c)、I_(120)(120%ET_c),ET_c为玉米需水量;4个N-P_2O_5-K_2O kg/hm^2施肥水平:F60(60-30-30)、F120(120-60-60)、F180(180-90-90)和F240(240-120-120),共16个处理。在春玉米灌浆期对根长密度(root length density,RLD)进行了测定,建立了不同水肥供应条件下相对根长密度(NRLD,normalized root length density)分布模型。结果表明:春玉米NRLD与土壤剖面相对深度呈现显著的三阶多项式函数关系,且三次项参数(R0)与灌水量和施肥量呈现二元二次多项式函数关系,决定系数(R^2)为0.84;验证结果显示,I_(60)、I_(75)、I_(90)和I_(105)灌水水平下NRLD模拟值与实测值均方根误差(RMSE)分别为0.20、0.16、0.16和0.17,标准化均方根误差(n-RMSE)分别为32%、27%、14%和17%,且R^2达0.95以上;基于NRLD分布模型估算各相对深度RLD分布比例,估算结果表明地表至相对根系扎深1/3处根长占总根长比例平均达73.6%,地表至相对根系扎深1/2处根长占总根长比例平均达82.8%。该模型对于指导大田春玉米灌溉施肥管理具有重要的理论意义。

Spring Spring maize is one of the main grain crops in Hexi region of China.Water and fertilizer are generally two of the most important factors in determining the spring maize productivity and yield. Roots are important vegetative organ of crop, and root distribution is one of the indispensable parameters for simulating crop growth, soil moisture and nutrient transport, as well as for developing rational irrigation systems and crop management. Root length density（RLD） is a highly wanted parameter for use in crop growth modeling but difficult to measure under field conditions. The existing root distribution models are mostly established under the single irrigation and fertilization condition. Different RLD distributions were created under varying row and drip line spacing, irrigation volume and fertilizer application rate. The model of root distribution for maize under different water and fertilizer supply was studied, which provided the practical significance for maize irrigation and fertilization management. The test site was in Shiyanghe Experimental Station of China Agricultural University, which is located in Wuwei City, Gansu Province, Northwest China. Using spring maize Qiangsheng 51 as the test cultivar, the field experiment was performed with 4 water supply levels, 60% ETc（I60）, 75% ETc（I75）, 90% ETc（I90） and 105% ETc（I105） in 2015, and 60% ETc（I60）, 80% ETc（I80）, 100% ETc（I100） and 120% ETc（I120） in 2016（ETc is the average annual crop evapotranspiration）, in interaction with 4 fertilization levels based on different ratios of N-P2O5-K2O, i.e. 60-30-30, 120-60-60, 180-90-90 and 240-120-120 kg/hm^2. There were 16 treatments in total. The RLD was measured at the grain filling stage, and the normalized root length density（NRLD） distribution model was established. The results showed that under the same fertilization amount, the distribution of RLD in soil profile increased first and then decreased with the increase of irrigation amount. Under the same irrigation condition, the distribution of RLD in soil profile increased first and then decreased with the increase of fertilizer amount. There was a cubic polynomial relationship between the NRLD of spring maize and relative soil depth, and the parameters of the cubic term showed a quadratic polynomial function with the irrigation amount and the fertilization amount, with a coefficient of determination of 0.84. Evaluation results showed that under the condition of I60, I75, I90 and I105, the root mean square error（RMSE） between simulated and observed NRLD values was 0.20, 0.16, 0.16, 0.17 and standardized root mean square error（n-RMSE） was 32%, 27%, 14%, 17%, respectively. The determination coefficient of linear relationship between simulated and observed RLD was 0.95. The distribution of RLD of spring maize and its proportion in different soil depths could be estimated under different water and fertilizer conditions. The vertical distribution of RLD estimated by the model was in good agreement with the observed pattern in the field, with the coefficient of determination of over 0.87. When the irrigation levels were the same, with the increase of fertilization amount, the ratio of root length over the upper 1/2 soil depth to that of the whole root system first decreased and then increased, with the average values of 87.6%, 79.8%, 79.6% and 84.3%, respectively. At the same fertilization level, with the increase of irrigation amount, the ratio of root length over the upper 1/2 soil depth to that of the whole root system first decreased and then increased, with the average values of 87.3%, 81.3%, 79.6% and 83.2%, respectively. On the whole, the root length over the upper 1/3 soil depth accounted for 73.6% of that of the whole root system, while the corresponding value was 82.8% for the root length over the upper half soil depth. This study is of great theoretical significance to guide the management of irrigation and fertilization in the field.