云南热区冬玉米生产潜力预测与种植适宜性评价

Traditional ecological suitability evaluation models for planting maize were poor in rationality and universality, because most of them were based on expert’s experience or statistical analysis. This study would build a mechanism model to predict the potential production and evaluate the ecological suitability for maize. A set of plot experiments and field experiments will be conducted for 3 years in 4 regions of Yunnan located in different climatic zone. The dynamic relationship will be quantified between the environment factors (temperature-T, sun shine-S, and soil water-W) and the increasing process of maize yield. 3 factors are specified by 7 indictors. A suitability evaluation indicator system will be built for maize, in which the indicators are classified clearly and reasonably based on physiology processes. Furthermore, the influential functions of environment factors will be built with factorial method and attenuation coefficient for each major physiology process. These functions include Gen-parameters to show the cultivars difference, it also could be used to explain the interaction effect and the difference between growth phases. Integrating these functions, a new model will be built to predict maize potential production under different condition of T, S and W, and to evaluate the ecological suitability of maize planting. The experiment date will be used to definite model parameters, while the other will be used to test model’s accuracy and sensitivity. Final, using a phenological model of maize built earlier and spatial databases involved climate, farmland using, and soil nature, the model will be used to analyze the spatial distribution pattern of potential production and ecological suitability for maize. This study would find a new method that using growth model to evaluate ecological suitability accurately for crops. It can supply some help for the planting division of maize in Yunnan.

作物生态适宜性评价多基于统计分析或专家经验，机理性和普适性不足。本研究拟构建兼具机理性和实用性的玉米产量形成模型和生态适宜性评价模型。首先基于云南不同气候带4试点、3年分播期的大田试验和小区控制试验，定量分析不同物候期温、光、水3因子40多指标与关键中间变量、以及玉米产量三方面的动态关系，确定主要影响指标及其耦合关系，依据生理过程进行指标分层分类，构建玉米生态适宜性评价指标体系。然后借鉴玉米生长模型算法，引入遗传参数，考虑物候期差异和因子互作，通过析因法和系数化，分别对主要生理过程构建温光水影响玉米产量的衰减效应函数，并综合形成以物候期为步长的玉米产量模拟模型和气候适宜性评价模型，预测玉米光温和光温水生产潜力，应用相互独立的试验数据进行模型参数调试和检验。最后基于已有生育期模型和云南耕地、气象空间数据库，分析玉米光温生产潜力和生态适宜性的空间分布。本研究构建了以物候期为步长的产量形成模型，

目前国内外开展的作物生态适宜性评价研究，多基于统计分析或专家经验，机理性和普适性不足。但近年来随着模型和信息技术发展，在系统的数据和精准模拟基础上进行作物适宜性评价成为可能，因此本研究提出了构建兼具机理性和实用性的玉米产量形成模型和生态适宜性评价模型的构想。经过4年的研究，在云南省热带区域多个试点、3年分播期进行了大田试验和小区控制试验，定量分析不同物候期温、光、水3因子40多指标与关键中间变量、以及玉米产量三方面的动态关系，进而开展了模型研发、生产潜力分析、生态适宜性评价等方面的研究。本研究一方面在田间试验基础上，基于APSIM模型进行了参数调试，确定了云南省主栽的两个玉米品种的遗传参数，结果显示模型对产量的预测精度整体达到预测精度NRMSE为4.61%，分析了温度变化以及不同灌溉和施肥条件下，作物的增产减产情况，例如2009-2018年间每日温度提高1℃情形下,陇川平均增产8.63%、瑞丽平均增产3.68%，芒市平均增产5.95%。有灌溉无施肥的情形下3个站点冬玉米作物平均减产12.1%、17.1%和14.5%，无灌溉无施肥的状态下，陇川平均减产16.2%，瑞丽平均减产21.6%，芒市平均减产18.4%，二是应用APSIM模型，导入云南省热区气象和土壤等环境数据，应用模型进行各地区冬玉米的生产潜力分析，例如2009-2018年间陇川、瑞丽、芒市3个站点9年平均温光生产潜力分别10177kg/hm2、11355kg/hm2、10589kg/hm2，本研究在此基础上应用2021年的气象数据制作热区形成了产量的空间分布专题图；三是应用作物模拟法，比较了分配指数和分配系数方法的预测效果，结果显示分配指数精度更高，进而构建了基于分配指数以物候期为步长的玉米产量形成模拟模型，四是构建了玉米生态适宜性评价指标体系，借鉴模型算法，预测玉米光温和光温水生产潜力，实现了玉米生态适宜性的评价。