PERS_August_2016_Public - page 645

Remote Sensing of Tamarisk Biomass, Insect
Herbivory, and Defoliation: Novel Methods in
the Grand Canyon Region, Arizona
Temuulen Ts. Sankey, Joel B. Sankey, Rene Horne, and Ashton Bedford
Abstract
Tamarisk is an invasive, riparian shrub species in the south-
western USA. The northern tamarisk beetle (Diorhabda carinu-
lata) has been introduced to several states to control tamarisk.
We classified tamarisk distribution in the Glen Canyon National
Recreation Area, Arizona using 0.2 m resolution, airborne mul-
tispectral data and estimated tamarisk beetle effects (overall
accuracy of 86 percent) leading to leaf defoliation in a 49,408
m2 area. We also estimated individual tamarisk tree biomass
and their uncertainties using airborne lidar data (100 points/
m2). On average, total aboveground tamarisk biomass was 8.68
kg/m2 (SD = 17.6). The tamarisk beetle defoliation resulted in
a mean leaf biomass loss of 0.52 kg/m2 and an equivalent of
25,692 kg across the entire study area. Our defoliated tamarisk
map and biomass estimates can help inform restoration treat-
ments to reduce tamarisk. Continued monitoring of tamarisk
and tamarisk beetle effects are recommended to understand
the currently-unknown eventual equilibrium between the two
species and the cascading effects on ecosystem processes.
Introduction
Saltcedar (Tamarix ramosissima), also known as tamarisk, was
originally introduced from Asia to the United States in the
late 1800’s as a decorative tree that provided shade and wind
break, and to prevent soil erosion (Crins, 1989). Tamarisk
eluded the controlled cultivation, and was initially found in
riparian areas close to the cultivation site. By the beginning of
the 20th century, tamarisk was found in almost all of the South-
western riparian areas, where the native woody vegetation
consisted of cottonwood (Populus fremontii), willows ( Salix
spp.), and the western honey mesquite (Prosopis glandulosa)
(Infalt, 2005). It was recently estimated to be spreading along
the arid and semi-arid river systems across the western United
States at a rate of 25 km per year (Nagler et al., 2011).
Tamarisk tree has many biological adaptations that enable
its successful invasion in arid ecosystems: (a) it outcompetes
the native flora for water resources using its large tap roots that
reach deeper sources of water (Hart, 2009), (b) tamarisk has a
high germination rate: its small seeds can germinate within 24
hours of dispersal under wet conditions (DiTomaso, 1998), and
(c) it accumulates salt in its leaves, which the tamarisk drops
annually to maintain salty soil unhospitable for many native
species (Ladenburger et al., 2006; Glenn et al., 2012; Ohrtman
et al., 2012; Merritt and Shafroth, 2012). River management
and water cycle alterations including dam construction, flow
regulation, flow diversion, flood control, and infrastructure for
crop irrigation (DiTomaso, 1998) have further enhanced the
invasive advantages for tamarisk. Many local, state, and fed-
eral agencies have targeted a great deal of management effort
on tamarisk control. Chemical, mechanical, and prescribed
fire control methods have been costly with mixed results
(Jorgensen, 1996; Hultine et al., 2010). A biological control,
known as the northern tamarisk beetle (Diorhabda carinulata
Desbrochers), has been determined to be an effective control
agent (DeLoach et al., 2003; Snyder et al., 2010).
Prior to its introduction, the beetle was carefully consid-
ered for many important parameters (Pattison et al., 2011) to
confirm that it has a narrow host range with impact on the tar-
get species only and is effective in desired climatic conditions
(Dudley et al., 2012). Following many experiments, the beetle
was introduced in 2001 to Colorado, Utah, Wyoming, Neva-
da, California, and Texas. The beetle was then introduced to
Moab, Utah in 2005, and was not expected to travel south past
38 degrees North latitude due to temperature limitations (De-
Loach et al., 2003) and a dormancy cycle, known as diapause,
associated with day lengths (Bean et al., 2007; Dudley, 2005).
In 2009, however, tamarisk beetle was found further south
than anticipated in the Colorado River ecosystem within the
Glen Canyon National Recreation Area and Grand Canyon
National Park. Since then, the tamarisk beetle has spread at a
more rapid rate than expected (Nagler et al., 2014) through-
out many reaches of the Colorado River with visible signs of
herbivory on the tamarisk trees.
The tamarisk beetle preys on tamarisk by defoliating the
leaves during the growing season and multiple times per year
(Paxton et al., 2011; Snyder et al., 2010). Each cycle progres-
sively weakens the plant. After several growing seasons of
the repeated cycle, the tree can eventually die due to carbo-
hydrate reserve depletion. The defoliation can happen over a
large land area as the beetle population grows and disperses
across the tamarisk stands (Paxton et al., 2011; Snyder et al.,
2010). The wide range makes ground monitoring of beetle
impact difficult, and often not very accurate. Remote sens-
ing techniques have, therefore, been used to monitor beetle
impacts (Dennison et al., 2009; Nagler et al., 2012; Nagler et
al., 2014; Meng et al., 2012). As tamarisk leaves are defoliat-
ed, the leaves turn a noticeable brown-orange color, which
is detectable in the visible and the near infrared spectrum in
remote sensing data (Dennison et al., 2009; Fletcher, 2013;
Meng and Dennison, 2015). Furthermore, the leaf defoliation
and the eventual tree mortality result in substantial decrease
in photosynthetic activity and biomass (Bateman et al., 2015), Temuulen Ts. Sankey, Rene Horne, and Ashton Bedford are
with the Informatics and Computing Program, Northern Ar-
izona University, 1298 S. Knoles Drive, Flagstaff, AZ 86011,
(Temuulen.Sankey@nau.edu).
Joel B. Sankey is with the US Geological Survey, Grand Can-
yon Monitoring and Research Center, Southwest Biological
Science Center, 2255 N. Gemini Drive, Flagstaff, AZ 86001.
Photogrammetric Engineering & Remote Sensing
Vol. 82, No. 8, August 2016, pp. 645–652.
0099-1112/16/645–652
© 2016 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.82.8.645
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING August 2016 645