February 2003

Site-Specific Management of White Grubs Using Remote Sensing

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.

Purdue Research

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 5^th in both 2001 and 2002. Using imagery, significant differences between treatments were identified by August 21^st 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.