中文摘要：

为探究落叶松毛虫种群数量与气象因子的关系,利用大兴安岭地区落叶松毛虫发生年度监测数据与气象数据构建线性回归模型,以赤池信息量准则为评价依据,通过逐步回归的方式从拟合优度和复杂度两方面权衡选取最优模型.采用相对权重法确定影响落叶松毛虫种群数量的关键气象因子,并量化了关键气象因子对松毛虫种群数量的影响.结果表明：低龄幼虫期和繁殖期是影响落叶松毛虫种群数量的关键时期,而中龄幼虫期和高龄幼虫期的影响较弱;低龄幼虫期日相对湿度均值、低龄幼虫越冬期日均温小于-22℃积温和繁殖期降水总量是影响落叶松毛虫种群数量的关键气象因子;低龄幼虫期日相对湿度均值和繁殖期降水总量的标准差每增加1倍,落叶松毛虫种群数量的标准差分别减少62%和35%;低龄幼虫越冬期日均温小于-22℃积温的标准差每增加1倍,落叶松毛虫种群数量的标准差增加40%.表明全球气候变暖趋势下落叶松毛虫的爆发特点等可能呈现出新的形式,建立长期的种群数量监测体系尤为重要.

英文摘要：

To explore the relationship between larch caterpillar population and meteorological fac- tors, a suite of linear regression models were developed. We used a stepwise regression approach to obtain the best model based on the Akaike information criterion （AIC）. We also identified the key meteorological factors based on relative weight, and analyzed their marginal influences on larch ca- terpillar population. Our modeling results showed that meteorological conditions during the young larva stage and breeding stage played a key role in impacting larch caterpillar population. In con- trast, meteorological conditions during the middle larva stage and old larva stage had a weaker effect. The mean daily relative humidity during young larva stage, the accumulated daily tempera- ture less than -22℃ during young larva＇ s overwintering stage, and the total rainfall in breeding stage were the key meteorological factors affecting the population of larch caterpillar. With the in- crease of one standard deviation from the mean daily relative humidity during young larva stage and the total rainfall in breeding stage, the larch caterpillar population would be reduced by 62% and 35% of standard deviation, respectively. In contrast, one standard deviation increase of the accumu- lated daily temperature less than -22℃ during young larva＇ s overwintering stage would increase larch caterpillar population by 40% of standard deviation. Our study suggests that the larch caterpil- lar population in the future may explode in response to global wanning, and its infestation could ex- hibit a new pattern. It is therefore very important to establish a long-term population monitoring sys- tem.