Graduation date: 2007
Although the Pacific Northwest has the least proportion of non-native plant species in relation to other regions of North America, exotic species continue to spread into mountainous areas, including the Cascade Range. In a forested landscape, road networks can act as corridors for exotic plant dispersal and establishment, helping species to overcome environmental barriers to expansion. Observation of the expanding, contracting, and possibly static spatial patterns of exotic species distributions over an extended time period can provide insight into biogeographical factors influencing invasion.
The objective of this study was to compare the results of a road survey conducted in 2005 to a previous road survey conducted in 1994, in order to describe the current distribution and eleven-year change in the spatial pattern of eight exotic plant species in and around H.J. Andrews Experimental Forest in Blue River, Oregon, and to highlight the role of road networks and timber harvest activity in facilitating exotic plant spread over time. Specific objectives were to (1) describe the distribution in 2005 of eleven exotic species on the road network, (2) quantify the change in distribution from 1994 to 2005 of eight exotic species on the road network, (3) determine the spatial pattern of distributions in 2005 as well as the change in spatial pattern from 1994 to 2005, (4) analyze species distributions and change over time in relation to four environmental and four human-influenced landscape factors, and (5) assess the degree to which landscape factors are correlated on the road network. Presence/absence and relative abundance on the road network was mapped using GIS. Spatial pattern was characterized using standardized semivariograms. CART (classification analysis and regression tree) modeling was used to evaluate landscape factors and species distributions.
Target species covered as little as 3.7% (C. maculosa - spotted knapweed) to as much as 90.2% (H. perforatum – St. John’s wort) of the entire network sampled in 2005. Species increased on much as 14.7% of the road network (C. leucanthemum – ox eye daisy) and decreased on much as 20% of the road network (C. vulgare – bull thistle) from 1994 to 2005 on the roads at the H.J. Andrews Forest. Spatial analysis of standardized semivariograms for exotic species distributions in 2005 did not reveal any preferred scale of patch size for the species evaluated, and species distributions were heterogeneously distributed at the landscape scale. In CART analysis, elevation was the landscape factor most associated with exotic species presence/absence in 2005, as well as increase in exotic species from 1994 to 2005. Age of adjacent clearcut patch was associated with decrease in exotic species form 1994 to 2005. However, the high degree of correlation between landscape factors on the road network makes direct causal relationships between exotic species distributions and single landscape attributes difficult to isolate.
The results of this study suggest that, for some species, exotic species distributions are closely related to areas of the network that had been clearcut in the 1950s and 1970s, and that change in exotic species distributions over time is a function of dispersal processes along road corridors subjected to high vehicle traffic, as well as changes in light availability due to closing canopy cover in aging clearcut patches. Hotspots of exotic plant diversity and abundance also occurred at sites that receive traffic from researchers, public visitors, and in some cases in conjunction with recent logging treatments. This paper outlines a conceptual model of exotic plant invasion over time as a function of explanatory landscape factors, both environmental and anthropomorphic in nature, which are correlated in space and heterogeneously distributed across the landscape.