The southwest Texas sky is alive with the sounds of bats on the hunt: the rather leisurely echolocation beeps of the search, the more rapid calls of target detection, and the feeding buzzes as bats attack insects. For us, of course, this is a silent drama, the sounds pitched too high for human ears. But to flying insects with hearing tuned to the ultrasonic frequencies of their predators' calls, it is a terrifying symphony - a powerful warning to flee the area or to drop immediately from the sky. Many of them do just that.
But these crop-damaging insects are being deceived. The echolocation calls come not from platoons of feeding bats, but from a plastic box about the size of a coffee mug, a small amplifier, and a bank of ultrasonic speakers mounted on tripods. We call it the Virtual Bat for Environmentally Benign Pest Control, and its first tests suggest enormous promise.
Developed after years of research into bats' vital role in agricultural pest control, much of it funded by Bat Conservation International, the Virtual Bat demonstrated its potential in last year's preliminary study in corn and cotton fields. The simulated bat sounds not only seemed to frighten away many destructive corn earworm moths (also known as cotton bollworm moths), but they also attracted real bats that attacked insects that hadn't flown for cover. We believe this one-two punch may combine for maximum effect.
The idea for the Virtual Bat dates back to the late 1950s and '60s, when researchers explored the impact on insect activity of broadcasting bat calls. Those field and laboratory studies showed promise but were never followed up. The problem with the strategy was that the effects seemed to be short-term: Insects got used to the sound and, over time, simply ignored it.
These early experiments, however, used an unrealistic signal - simple, repetitive pulses that only crudely simulated the search-phase calls of generic bats. For the Virtual Bat to succeed, we needed a complex series of authentic bat calls that would mimic the sounds of bats actually hunting in an area.
The solution came last summer when Paul Schleider, an electrical engineer with the USDA in College Station, Texas, designed and built the Virtual Bat, which allows us to closely mimic the whole array of calls made by Mexican free-tailed bats (Tadarida brasiliensis) as they search for, detect, and feed on insects.
The device is programmed to produce a repeatedly randomized sequence of calls at the frequency, duration, and intensity of the real thing. The result sounds exactly like a large number of bats feeding on swarms of insects.
For our July 2001 research, we went to the Winter Garden area south of Frio Cave, a major roosting site for Mexican free-tailed bats, near Uvalde, Texas. In this bat-rich agricultural area, we set up one-acre plots planted in corn and cotton. Using infrared video cameras, we monitored moth (and bat) activity for seven nights and simultaneously used bat detectors to quantify foraging activity of bats.
At control sites in corn and cotton, we used our monitors but made no attempt to alter moth or bat activity. We placed battery-powered Virtual Bats in the center of experimental plots of each crop type, with the signals fed through an amplifier and broadcast in all directions from a ring of speakers.
We are still processing the mountains of information from the cameras and bat detectors, but we have examined a great deal of the data and the initial results are encouraging to say the least.
Moth activity was decreased substantially - by 50 percent or more in some cases - in the experimental fields as compared with the control plots. At the same time, we saw dramatically increased bat activity, with levels of foraging multiplied several-fold over the experimental plots. The Virtual Bat apparently starts a cascade of impacts as foraging bats hear the simulated activity and fly in to investigate, which increases activity and produces a constant influx of actual bats over these fields.
The corn earworm causes major damage to corn and cotton. In Texas alone between 1995 and 1997, cotton producers treated an annual average 2 million acres with insecticides at a cost of $12.6 million and still lost 182,000 bales of cotton to bollworms (and tobacco budworms). If the Virtual Bat reduces these insects in crops, that may translate not only to increased yields but also to a reduction in the use of pesticides. During this summer's field trials, we plan to collect detailed data on the Virtual Bat's impact on the number of eggs laid, the larvae that infest crops, and overall crop damage.
Additionally, earlier research suggests that the risk of predation, as perceived by insects when they hear bat echolocation calls, can disrupt the mating and reproductive activities of the insects. As each female bollworm moth can lay over 1,000 eggs that grow into larvae and eat crops, inhibiting insect reproduction might provide yet a third hit to the damage caused by these pests.
This summer, we plan to investigate effects on insect reproduction and to continue our work near Uvalde, as well as in agricultural fields near College Station, Texas, where far fewer bats are known to live. Bats are abundant, night-flying predators in all temperate and tropical regions of the world, and we will investigate the Virtual Bat's effectiveness for controlling crop damage in areas that do not have huge cave colonies of bats.
The Virtual Bat, which remains a proprietary invention, is being tested with the echolocation sounds of Mexican free-tailed bats. It could, however, be adapted to mimic virtually any bat species or even a collection of species. We believe this technology could be adapted to combat other insect pests and to protect many other crops, including orchards and row crops such as cabbage or tomatoes.