Abstract: Based on geographic information system and geostatistics (GIS), the spatial structure of weed population at different periods was analyzed in Lanxi district of Heilongjiang province. The spatial distribution of weed population was also simulated by ordinary Kriging (OK) interpolation method. Results showed that the weed population of different periods was correlated spatially in the study area. The semivariograms of weeds could be described by exponential model indicating an aggregated spatial arrangement. The degree of aggregation showed the tendency of equably-assembly-equably along with sowing date. The dynamic change of weed population at different periods could be analyzed on the simulated maps of the spatial distribution. The sowing date of bothⅠand Ⅵ treatment was suitable for weed control.

Key words: Alfalfa, Geostatistics, Geographic information system, Spatial structure, Simulation

摘要： 运用地理信息系统(GIS)和地统计学方法对不同播期苜蓿(Medicago sativa L.)人工草地入侵杂草种群相对多度进行分析,并采用普通克里格(OK)插值法模拟入侵杂草种群空间分布,为有效防止苜蓿草地杂草入侵发生提供依据。结果表明:不同播期处理间,苜蓿苗期入侵杂草种群存在空间相关性,其半变异函数曲线均为指数型,主要呈聚集格局,随着6个设置播期的推迟,聚集程度呈现均匀-聚集-均匀的变化趋势。通过比较不同播期处理杂草种群相对多度的模拟分布,播期Ⅰ和Ⅵ处理有利于控制苗期苜蓿杂草的发生,是兰西地区当年建植苜蓿人工草地的适宜播期。

关键词: 苜蓿, 地统计学, 地理信息系统, 空间结构, 模拟