Randy Hamilton, Rick Foster, Tim
Gibb, and Larry Biehl
Introduction
White
grubs are the most destructive and economically injurious pests of turfgrass in
the Midwest and eastern United States. In Indiana, the larvae of Japanese and
masked chafer beetles cause the most severe damage in turfgrass. Adult beetles
lay their eggs in the spring of the year, preferring well-maintained turfgrass
for their egg-laying sites. After hatching, the larvae feed on grass roots just
below the soil surface. By late summer, if populations are high, the grubs can
cause extensive damage to a turfgrass stand, which may require costly
renovation. Unfortunately, because the grubs live underground, they often go
unnoticed until the effects of their feeding (thinning, wilting, chlorosis, and
eventually death of the turfgrass plants) occur. At this stage, turfgrass
damage is irreversible and reseeding or resodding is inevitable. To avoid this,
turfgrass managers generally apply a preventive pesticide to their turfgrass
stands. Such an approach disregards the fact that grubs are often spotty in
distribution and sporadic in occurrence. Recent studies have confirmed that
more than 70% of pesticide applications for grubs are unnecessary because they
are applied when few if any grubs are present. Furthermore, repeated preventive
pesticide applications can lead to environmental pollution, detrimental
non-target effects, and reduction in effectiveness through pesticide
resistance. A method to detect localized areas of grub feeding in advance of
irreversible damage to the turfgrass is needed.
In forests and agricultural crops, remote sensing
(measuring visible and infrared light reflected or emitted from the plant
vegetation) has proven to be an effective way to detect physiological stresses.
We suggest that this technology might be used to monitor and detect white grub
infestations in turfgrass before damage becomes otherwise visible to the naked
eye. If grub infestations can be detected before extensive damage occurs,
turfgrass managers could potentially adopt a site-specific pesticide
application program to target only those areas where pesticide application is
truly needed. Adopting such a management program could greatly reduce
unnecessary pesticide applications, reduce potential environmental
contamination and human exposure, and maintain pesticide effectiveness.
During the past two years several studies were
conducted to investigate the possibility of using remote sensing for early
detection of grub infestations in turfgrass.
Simulated
Grub Injury
In one study, we simulated grub
feeding by cutting strips of turfgrass with a sodcutter (Fig. 1). Throughout
the day, at approximately 40-minute intervals, we collected reflectance
measurements over the strips of grass using a field spectrometer (Geophysical
& Environmental Research Corporation [GER] 1500) and an infrared
temperature gun (Fig. 2). These measurements were compared with visual ratings
of the differences between cut and uncut strips. The spectrometer data yielded
significant differences between cut and uncut grass approximately 40 minutes
before the differences became visually apparent. Surface temperatures allowed
us to detect differences between treatments approximately 1.5 hours before the
differences became visually apparent.
Although
the root pruning caused by the sodcutter is more abrupt than that caused by
grubs, the physiological response of the grass is similar in both situations.
Because of this similarity, the results of this study suggest that grub damage
may be discernible using remote sensing before it becomes visually apparent.
Natural Grub Injury
To further verify this assertion,
we conducted a study to test the feasibility of using imagery and field
spectrometer data to detect early damage resulting from natural grub
infestations. Large turfgrass plots (40x40 ft) were established having low,
intermediate, and high grub densities. From late August through early October
in 2001 and 2002, multi-spectral aerial imagery (Agri-Vision, Columbus, IN) and
field spectrometer data (GER 2600) were collected every 7-10 days (Figs. 3, 4).
Using field spectrometer data, we identified significant differences between
treatments by September 5th in both 2001 and 2002. Using
imagery, significant differences between treatments were identified by August
21st in 2002. Although these differences are difficult, if
not impossible, to visually detect (see Fig. 4), statistical differences in the
red (and often near infrared) band were present. In 2001, we were not
successful in detecting differences using imagery until late in the season.
Visual damage was difficult to discern until mid September.
Conclusions
It was found, in both the sodcutter
grub simulation and the natural grub experiment, that differences in canopy
reflectance could be detected before visual differences became obvious. Because of
this, we feel that site-specific grub management programs, using remote sensing
techniques, may be possible in the near future.