Peterson A Townsend, Martínez-Campos Carmen, Nakazawa Yoshinori, Martínez-Meyer Enrique
Natural History Museum and Biodiversity Research Center, University of Kansas, Lawrence, KS 66045, USA.
Trans R Soc Trop Med Hyg. 2005 Sep;99(9):647-55. doi: 10.1016/j.trstmh.2005.02.004.
Numerous human diseases-malaria, dengue, yellow fever and leishmaniasis, to name a few-are transmitted by insect vectors with brief life cycles and biting activity that varies in both space and time. Although the general geographic distributions of these epidemiologically important species are known, the spatiotemporal variation in their emergence and activity remains poorly understood. We used ecological niche modeling via a genetic algorithm to produce time-specific predictive models of monthly distributions of Aedes aegypti in Mexico in 1995. Significant predictions of monthly mosquito activity and distributions indicate that predicting spatiotemporal dynamics of disease vector species is feasible; significant coincidence with human cases of dengue indicate that these dynamics probably translate directly into transmission of dengue virus to humans. This approach provides new potential for optimizing use of resources for disease prevention and remediation via automated forecasting of disease transmission risk.
许多人类疾病——仅举几例，如疟疾、登革热、黄热病和利什曼病——是由昆虫媒介传播的，这些昆虫媒介生命周期短暂，叮咬活动在空间和时间上都有所不同。尽管这些在流行病学上具有重要意义的物种的总体地理分布是已知的，但它们出现和活动的时空变化仍知之甚少。我们通过遗传算法使用生态位建模来生成1995年墨西哥埃及伊蚊月度分布的特定时间预测模型。对蚊子月度活动和分布的显著预测表明，预测病媒物种的时空动态是可行的；与人类登革热病例的显著吻合表明，这些动态可能直接转化为登革热病毒向人类的传播。这种方法为通过自动预测疾病传播风险来优化疾病预防和补救资源的利用提供了新的潜力。
