angiosperm
; autocorrelation
; harmonic analysis
; invasive species
; kriging
; phenology
; regression analysis
; remote sensing
; spatial distribution
; United States
英文摘要:
Saltcedar (Tamarix spp.) is recognized as one of the most aggressively invasive species throughout the Western United States. Mapping its suitable habitat is of paramount importance to effective management, and thus, becomes a high priority for conservation practitioners. In previous studies, species distribution models (SDMs) have been applied to predicting the suitable habitats of saltcedar at national scale, but at coarser spatial resolution (1 km). Although such studies achieved some success, they are lacking of capability to accommodate fine-scale resolution environmental variables, and therefore, fail to uncover detailed spatial pattern of habitats. The objective of this study was to develop a remote sensing driven SDM so as to characterize suitable habitats of saltcedar at very fine spatial scale (30 m). We exploited several fine-scale environmental predictors through remote sensing images, and utilized the logistic regression model to analyze the species-habitat relationship by identifying influential factors with subset selection criteria. We also incorporated the spatial autocorrelation with regression kriging method. Our results indicated that the model incorporating spatial autocorrelation achieved a higher accuracy than that of regression only model. Among 10 environmental variables, the distance to the river and the phenological attributes summarized by the harmonic analysis were regarded as the most significant in predicting the invasive potential of saltcedar. We conclude that remote sensing driven SDM has potential to identify the suitable habitat of saltcedar at a fine scale and locate appropriate areas at high risk of saltcedar infestation, which could benefit the early control and proactive management strategies to a large extent.
Department of Geography, University at Buffalo, The State University of New York, 105 Wilkeson Quad, Buffalo, NY 14261, United States
Recommended Citation:
Diao C,, Wang L. Development of an invasive species distribution model with fine-resolution remote sensing[J]. International Journal of Applied Earth Observation and Geoinformation,2014-01-01,30(1)