中文摘要：

红松属小兴安岭地区地带性植被优势种，该地区也是其分布的北缘。在景观尺度上开展红松的分布格局研究有利于进一步了解红松分布机理、未来迁移过程等问题，对其经营和保护有重要意义。将景观指数法与点格局分析法结合，设定8个空间尺度，利用红松存在／不存在数据，通过计算各空间尺度上红松聚集程度和景观指数，分析小兴安岭地区红松种群在多尺度上的分布格局。研究结果表明，小尺度上红松聚集分布明显，随机分布区多处于其聚集分布区的边缘，均匀分布区则散布在其聚集分布区内。景观指数研究表明，通过景观指数可判断红松聚集分布格局趋势，而不能判断均匀分布、随机分布格局趋势，因为它们在多尺度下景观指数波动大，不能用景观指数来描述分布格局。研究得出如下结论：1）红松主要分布在其分布区的核心区域内，在分布区边缘和过渡带上呈随机分布，2）存在／不存在数据能够用来分析种群的多尺度空间分布格局。3）空间尺度的变化会引起树种分布格局的变化，随机分布随尺度增加，边缘化程度加强．4）单一尺度上。景观格局指数不能完全描述种群分布格局；而在多尺度上，变化趋势稳定的景观指数表明聚集分布存在，而波动剧烈的景观指数常与随机分布和均匀分布联系在一起，5）地形因子中，红松对坡度和海拔两个因子变化敏感。

英文摘要：

Spatial patterns reflect the distribution characteristics of populations and the adaptions of populations to environmental resources and ecological processes. Dispersal limitation, inter- and intraspecific competition, and habitat heterogeneity all influence species distribution. In addition, species are affected by local temperatnre, precipitation, and terrain. Therefore, the relationship among species distribution patterns, landscape indices, and environmental factors are important in population studies affected by changing climates. To reveal these relationships, Korean pine, the dominant species in the Small Khingan Mountains, at the northern boundary of its distribution range, was selected to analyze its distribution patterns at multi-spatial scales. The point pattern analysis and landscape pattern metrics were calculated at eight spatial scales （90 m x 90 m resolution） using R package （Spatstat 1.42-1） and FRAGSTATS （Version 4.2） , based on presence/absence data derived from forest inventory maps and attributed data. The results from the point pattern analysis showed that Korean pine was aggregated in small scales, random distribution areas surround an aggregated pattern, and uniform distribution areas were often mosaicked in aggregated pattern areas. However, the distribution pattern of Korean pine was more aggregated with an increasing spatial scale because the aggregated distribution area increased, especially in the core area of this region. Meanwhile, random distribution areas occurred at the edge of aggregated distribution areas, and uniform distribution areas disappeared from the core area. The results from the landscape pattern analysis indicated that landscape pattern indices could be used to describe an aggregated distribution at multi-spatial scales since the landscape metrics were most stable, or changes were logarithmic; whereas landscape pattern indices dramatically fluctuated with changing spatial scales for the others. The analyses also revealed that Korean pine was sensitive to