耦合粘虫胁迫的玉米生长可视化模拟

针对难以定量化模拟虫害影响植物形态结构和生理过程的问题,提出将虫害影响耦合至植物功能-结构模型中的可视化模拟方法。根据粘虫啃食叶片的空间分布特征,改进细胞纹理特征点和基函数,适用于描述粘虫啃食路径,采用单叶被啃食率描述被啃食程度,并以三维可视化形式模拟虫害啃食效果;结合粘虫数量、啃食量以及分布规律,估计各单叶被啃食率,根据单叶虫洞可视化方法,定量化表达粘虫对单株玉米形态结构的影响;根据植物形态结构变化,将粘虫胁迫作用于生物量产量、生物量分配等植物生理过程,确定粘虫对植物形态发育的影响。试验结果表明,单叶虫洞可视化方法能较形象、逼真的仿真不同受灾程度下粘虫对叶片形态的影响,并将虫害影响耦合至功能-结构模型中,实现虫害胁迫下植物生长发育的模拟和仿真,为定量描述和理解灾害程度提供新思路。

Visual simulation of maize growth responding to armyworm (Mythimna separata) attack

Abstract: Insect pest attack has serious consequences for plants growth and development. Biological control of pests is a potential solution, which needs to understand the quantitative interactions among various organisms in the environment for management optimization. However, it is time-consuming and laborious to determine the effect of insect attack on plants through experiments. In this paper, a new method was proposed and developed to simulate the effect of insect pest on leaves and quantify the effect of insect pest on plant growth. An improved cellular texture strategy was used for describing the appearance of leaves eaten by armyworm. The procedures are as follows: (1) to simulate the irregular eating path of pests, we use origination feature point, which is the position where the armyworm begins to eat, and critical feature point to model the pest''s trajectory. The critical feature point is generated randomly within a circle, which is defined by certain distance as radius and the origination feature point as the center, and (2) to simulate armyworm eating habits, we select the closest feature point to a pixel, which is closer to the origination feature point than the distance between the pixel and the origination feature point; (3) a transparent pixel in cellular texture is used to represent the leaf area where pests will not eat. To describe wormhole intuitively and visually, cellular texture values were mapped to color, the pixels in leaf texture will be transparent when the color of corresponding cellular texture pixels are lower than a threshold. To describe the effect of leaf eaten by pest, we used the proportion of being eaten, namely the percentage of eaten leaf area to the whole leaf area. As the proportion of being eaten changes, the number of transparent pixels also changes. Therefore, the appearance of the leaves could represent conditions under various degrees of armyworm attack. Coupling armyworm attack with functional-structural model is able to quantitatively describe the interactions between armyworm and maize and visually simulate the growth of maize. Models of maize architectural development were constructed based on the L-System, which facilitates the simulation of physiological response to damage. The morphological size of each organ was calculated according to their cumulative biomass (fresh weight). For estimating the proportion of being eaten of each leaf, it is necessary to calculate the number of armyworms and the amount of their food intake, and simulate the distribution of armyworms in maize. Based on the literature review, the life cycle of armyworm was divided into 10 different stages, and every stage was further divided into multiple age classes. On the basis of daily effective accumulated temperature at each age, the number and intake of each age class in each stage of the armyworm population was simulated per day, combined with the effect of natural enemies and environmental factors on the survival of armyworms. Ray tracing algorithm was employed to simulate light interception of a canopy, and a photosynthesis model was applied to estimate biomass. To simulate the assimilate partitioning within a maize and quantify the effects of armyworms eating on whole plant structure, Friedlingstein model was used to estimate the partitioning ratio of above-ground and underground assimilate affected by leaf area index, and source-sink model and parameters from GreenLab were used to simulate the distribution of aboveground assimilate. The growth simulation also takes the effects of maize changing, armyworm damage inducing further changes that affect development into consideration. The results showed that simulation could realistic represent the vivid appearance of leaf eaten by armyworms, such as irregularity of wormhole, random selection of eaten areas and armyworms eating habits. The proposed approach can quantify the effects of armyworms attack on maize development and crop yields; it is useful for quantifying and understanding disaster degree in pest management, and it has potential for agricultural technical training and education.