Determining spatial units for modeling regional nonnative invasive plant species spread in the southern US forestlands: using the state of Alabama as an example

Abstract

Nonnative invasive plant species (NNIPS) cause significant damage to the native forest ecosystems in the southern United States forestlands, such as habitat degradation, ecological instability, and biodiversity loss. Taking the state of Alabama as an example, we used more than 5,000 permanent United States Department of Agriculture-Forest Service's Forest Inventory and Analysis (FIA) plots measured between 2001 and 2019 over three measurement cycles to test the suitable modeling unit for quantifying invasion patterns and associated factors for regional NNIPS monitoring and management. NNIPS heavily infest Alabama's forestlands, and forestlands plagued with at least one NNIPS have increased over time: 41.1%, 50.8%, and 54.8% during the past three measurements. Lonicera japonica (Thunb.) was the most abundant NNIPS in Alabama, with at least 26% of its forested lands infested. The FIA data were aggregated with multiple spatial units: five levels of hydrological units, three levels of ecological units, and a county level. Invasion indices were calculated for all spatial units based on NNIPS' presence/absence and average cover in each plot. The best modeling unit was identified based on Moran's test, with the county-level modeling unit providing the best Moran's I value over all measurement periods. Influencing factors of invasion were evaluated based on spatial lag models. Our models show that the invasion index decreased with increases in public forest areas in a county. In contrast, the human population density of neighboring counties positively influenced the invasion index.

Keywords

• nonnative invasive plant species (nnips)
• invasion severity
• multiscale modeling framework
• fia
• spatial lag model