基于Voronoi图的单粒播玉米产量效果的蒙特卡罗模拟

为了明确在单粒播种条件下田间出苗率对玉米产量的影响,以蒙特卡罗二项分布作缺苗模拟,以Voronoi图分配缺苗土地,通过对产量-密度方程进行数学恒等变换计算产量补偿,模拟了2种单粒播策略的玉米产量随出苗率(75%~95%)的变化。结果表明,缺苗斑数目与出苗率呈单峰曲线关系,缺苗斑大小随出苗率呈负指数下降;单粒播造成0.06%~16.78%的减产,且减产率随出苗率下降而增加;两种单粒播策略的产量大小和产量稳定性均有差异,且差异随出苗率下降而扩大。该研究推荐的玉米单粒播策略是:在千方百计提高出苗率基础上,播种的种子数目应当等于目标种植密度除以出苗率、保证出苗的植株数目达到推荐种植密度。该结论对于在玉米生产上广泛采用单粒精量播种具有重要参考价值。

Monte-Carlo simulation of yield effect under singular seeding strategy in maize based on Voronoi diagrams

Abstract: Maize precision planting with single seeding that saves labor costs for thinning seedlings is becoming one of the most popular technical innovations in China's maize production. Compared with conventional planting, i.e. multiple seeds per hill to tackle the problem of seedling emergence rate less than 100%, which ensures at least one seedling per hill by planting adequate seeds, what are land patches with missing seedlings under planting one seed per hill like? What variation will occur to maize yields? What seeding strategies should be adopted for high maize yields? These are still open questions. The paper put forward a new computer simulation procedure of planting one seed per hill to plant maize. The procedure consisted of 4 key parts. Firstly, Monte-Carlo simulation of binomial distributions was applied to mimic the number of missing seedlings in cornfields. Secondly, a patch of land with missing plants was partitioned among its surrounding seedlings in the way showed as Voronoi diagrams. Thirdly, for the plant occupying a given area of land, its grain yield was computed using a mathematically identical transformation based on the parabola equation, which was extensively used to relate crop yield to plant density for maize. Lastly, the maize crop yield was evaluated as the sum of individual grain yield per hectare. The simulated results, with seedling emergence rates from 75% to 95%, showed that the number of the patches with missing plants in the field varied with seedling emergence rate following an unimodal curve (R2=0.996), with the maximum fleck number of 7.5 per 100 hills at the emergence rate of 82.3%. The patch size decreased with the increasing of seedling emergence rate in a negative exponential function pattern (R2=0.994). Planting one seed per hill reduced maize yield by 0.06%-16.78%, and the reduction rate displayed the steady falling trend with the increasing of emergence rate. For single seeding, Strategy II outweighed Strategy I in the crop yield (P<0.03) and the yield stability (R2=0.850), the seeding rate for Strategy I was equal to recommended planting density and that for Strategy II was equal to recommended planting density divided by emergence rate. Technical approaches to improve seedling emergence rate were discussed. Our recommendation for single seeding strategy is that, after adopting every mean to promote seedling emergence rate, the seed number sowed should be equal to the target planting density divided by the seedling emergence rate in order that emerged seedlings can reach the optimum planting density. The recommendation will play an important role for spreading the practices in planting one seed per hill in maize production.