outbreaks of the sagebrush defoliator moth (Aroga websteri
Clarke [Lepidoptera: Gelechiidae]) can cause widespread damage
to rangelands in the western United States. Sagebrush (Artemisia
spp.) is the exclusive larval host of A. websteri and in
high numbers, larvae can kill hostplants and reduce the
production of foliage and flowering by surviving plants for
years. The overall goal of this project was to use habitat data
from sagebrush communities in southeastern Idaho to determine
which variables (e.g., abundance or height of sagebrush;
presence or abundance of other plant species; presence of other
moth and insect species; or land use attributes) most strongly
predict the presence or absence of A. websteri.
Development of a predictive model would be a first step toward
identifying the locations of potential A. websteri
outbreaks. A better understanding of the location, timing, and
pattern of defoliator outbreaks would allow land managers to
better maintain and manage critical sagebrush habitats.
correlation and linear regression were used to test the strength
of relationships between 9 independent variables (IVs) used to
characterize the habitat at each trapping location (table
below). The height of rabbitbrush was eliminated to minimize
effects of colinearity, and logistic regression was used to fit
the presence and absence of A. websteri to the 8
remaining IVs, individually and in combination. Models for
grazed and un-grazed habitats were analyzed separately for 2007
and 2008. All combinations and interactions between IVs were
tested using forward and backward stepwise methods.
Alone or in combination, the
simplistic metrics used to characterize habitat for each
trapping location were not significant in predicting the
presence or absence of A. websteri. Although vegetation
and soil characteristics differed between grazed and un-grazed
locations, the difference in the number of A. websteri
captured in each habitat type (table below) was not significant
and grazing by domestic livestock was also shown to be an
insignificant predictor of the presence or absence of A.
macrolepidopteran specimens representing nearly 30 species
captured in 2007 and 2008 were pinned, labeled, and are being
submitted for identification as qualified taxonomists are
located. Efforts to sort several hundred microlepidopteran
specimens to morphospecies are ongoing.