基于个体的空间显性模型和遥感技术模拟入侵植物扩张机制

The dispersal mechanism of invasive plants based on a spatially explicit individual-based model and Remote sensing technology: a case study of Spartina alterniflora

基于个体的空间显性模型和遥感技术,以互花米草为例,模拟了自1997到2010年的种群扩张动态,揭示了土地利用变化与潮间带高程的影响;并通过全局敏感性分析揭示了种子扩散、成体存活率、有性和无性繁殖等种群统计学特征对互花米草种群扩张的相对重要性。研究结果发现:1)有性繁殖与无性繁殖共同决定互花米草种群快速扩张;2)潮间带高程和土地利用变化显著影响模型预测的精度,对互花米草种群扩张有非常重要的影响;3)成体存活率与种子长距离扩散是影响互花米草种群扩张速度最重要的因素;无性繁殖比有性繁殖对种群扩张的影响更大;种子长距离扩散比本地扩散更为重要,同时,小概率的种子长距离扩散事件对种群扩张有非常重要的影响。为了经济有效地控制外来入侵植物的扩张,应该抑制种子的长距离扩散和移除种子长距离扩散形成的位于入侵前沿的小斑块。

Biological invasion is a major cause of biodiversity loss, and poses a growing threat to human health. Therefore, it is necessary to reveal the dispersal mechanisms of invasive plants for biological conservation. Spatially explicit individual- based models account for the life history of a target species, in addition to its dispersal ability; thus, abiotic factors, such as landscape heterogeneity and changes, are widely used for simulating population dynamics. Remote sensing （RS） is an important tool for elucidating changing vegetation patterns associated with broad-scale plant invasions. Therefore, by combining spatially explicit individual-based models with RS and GIS （Geographic Information System） information, it is possible to simulate the population expansion of invasive species accurately in realistic landscapes and to understand their invasive mechanisms. However, few spatially explicit, individual-based models have used RS to investigate spatial and temporal land use changes. In this study, a spatially explicit, individual-based model, combined with remote sensing technology, was built to simulate the range expansion of Smooth cordgrass （Spartina alterniflora） in Yancheng Coastal Wetlands, China, from 1999 to 2010, and to reveal how land use and tideland elevation impact range expansion. Furthermore, a global sensitivity analysis was used to identify the relative importance of adult survival rate, seed dispersal, and sexual and asexual reproduction on range expansion. The results show that removing sexual propagation or asexual reproduction greatly reduces the expansion rate. Thus, rapid expansion requires both sexual and asexual reproduction. Land use changes and tideland elevation reduce simulation accuracy, indicating that they have significant impacts on range expansion. In the global sensitivity analysis, two most important factors influencing S. adult survival rate and long-distance seed dispersal distance represented the alterniflora invasion, while seed production, maximum capacity, and the proportion of long-distance dispersal were the three least important factors. Asexual reproduction contributes much more significantly to quick range expansion than sexual reproduction, except for long-distance dispersal. Seeds dispersed long distances contribute more to quick range expansion than local seed dispersal. Moreover, even a few long-distance dispersal events are enough for quick range expansion. Therefore, invasion control strategies should target a single reproduction mode to decrease expansion rates. Here, realistic strategies for controlling and managing invasion are suggested to remove low density, small patches formed by long-distance seed dispersal at the invasion front before the onset of the growing period, and to limit long-distance seed dispersal during the growing period. The integration of these results to control strategies for S. alterniflora and similar invasive species would considerably reduce management costs, while the likelihood of local or complete eradication would increase. In conclusion, spatially explicit individual-based models combined with RS are verified as being very suitable for simulating the range expansion of invasive species in a realistic landscape, and for revealing dispersal mechanisms.