Spatial patterns of beetle-induced defoliation of invasive tamarisk at the landscape level
The control of exotic tamarisk in riparian areas of the southwest has long been a priority for land managers and ecologists. Diorhabda spp. was introduced as a bio-control agent beginning in 2003 in several states and has since become an inseparable part of tamarisk dominated river systems in the southwest. While anecdotally the beetle seems effective as a means of controlling tamarisk, there is still a limited understanding of the complex effects on tamarisk stands. As of yet, we cannot accurately predict patterns of defoliation in response to beetle release. Also, it is unclear what environmental variables may contribute to the impact of Diorhabda spp. Recognizing the need for greater understanding of the effects of the bio-control beetle, this study explored how spatial modeling can be used to explain variation in defoliation of tamarisk stands exposed to the bio-control beetle. Beetles were released across Grand County, UT between 2004 and 2007 at eleven locations. In 2014 a detailed field survey was conducted at 80 sites across the county to assess tamarisk dieback. We analyzed these data with three spatial models representing either overland or bidirectional watercourse dispersal (Moran’s Eigenvectors Maps) and unidirectional down-river watercourse dispersal (Asymmetric Eigenvector Maps).
Preliminary results showed that the strongest explanatory model was the overland model, explaining 41% of the variation. The bi-directional watercourse model explained 34%. The down-river watercourse model explained 25% of the variation. None of the environmental factors (soil pH, salinity, cattle impact, distance to water source, elevation from water source, beetle presence) tested in the initial study were significant except beetle presence. Beetle presence explained 6% of the variation in live canopy, but merely indicated an abundant food source. Tamarisk canopy is thus spatially structured overland, as well as according to the waterway. Based on this understanding, we must now consider how much of the pattern can be accounted for by beetle movement and how much is the influence of unknown environmental factors. Future work will define the scale and shape of the spatial patterns for each model as well as incorporating potential explanatory environmental variables. These additional variables include type of water source, site slope and river width. This study will help us to understand the ecological impacts of the beetle and ultimately optimize their use as an effective management tool.