Invasive plant species and climate change are considered to be chief interacting threats to biological diversity in the national parks of the United States. Because invasive species show a logistic growth response over time, with a colonization period preceding a period of rapidly irruptive establishment and spread, early detection is essential. Targeted searches are the preferred methods used by managers for early detection, yet emphases on non-probabilistic sampling and mapping of known infestations without collection of absence data has hampered understanding of invasion risk in park landscapes.
We developed a monitoring protocol that focuses on early detection of infestations of prioritized species in a probabilistic sample of road and trail segments in six National Parks in northern California and southern Oregon. In summer 2009, we tested the protocol collecting data on invasive species abundance at 170 randomly selected road and trail segments (395 km total). Our objective is to report on the implications of the protocol to managers of protected areas trying to minimize impacts of invasive plants.
Results/Conclusions
We detected and mapped 225 separate infestations of 26 invasive species to be controlled by managers, and also collected site-specific environmental data at 152 of these infestations and in 788 random plots that were not infested. Invasive species infestations tended to occur at low elevations, within 6 m of a road or trail, in areas with low overstory cover, and significantly more often along roads than trails. Our findings underscore the importance of dispersal and disturbance to invasive colonization in protected settings. The finding of no invasive species in the random plots, suggests a wide-ranging, rapid assessment technique centered on roads and trails may be the most efficient use of limited resources to manage invasives in protected areas. Yet a hybrid approach such as this, which supplements data from infestations with data from randomly located plots, yields information appropriate to support parkwide inferences and invasive species modeling.
