Except for control of bird depredations, practical uses of recorded communication signals are potentialities, not actualities. We can, therefore, only suggest what possible practical uses may come in the future as studies on animal communication continue. The area of greatest future promise, besides the already developing bioacoustic bird control, would seem to be in control of insect pests. Concern over possible hazards from insecticide residues has led to an intensified interest in insect control other than by insecticides. Since communication chemicals, for instance, are generally of low toxicity and can be used in very low concentrations, their use for insect control would eliminate the residue hazard. Sounds, of course, if they could be used, leave no residues at all. Certainly the great potential benefit to be derived from developments of new and improved insect control should lead in the future to extended studies of all facets of communication by these animals.
Johnson (1948) suggested that sounds produced by marine animals could be used to locate schools of fish, or certain types of sea floor, for many marine animals are characteristically found over certain substrates and not others. Since then there has been some interest in the possibility that recorded communication signals of marine and freshwater fish might be used to attract these to fishermen, thus increasing the catch. The first possibility, that is, using sounds produced by animals to detect areas of concentration, is now in use. The second, that is, attracting fish by sounds, has been tried, but so far results have not been as satisfactory as might be desired. Recent attempts to attract fish by sounds are reviewed by Moulton (1964).
An area that would seem to have considerable promise, but has been almost totally unexploited, is the possible use of communication signals of domesticated animals to influence their behavior. Recent studies of the behavior of domesticated animals are reviewed in a book edited by Hafez (1962). As is pointed out there by Hale (pp. 21-53), domesticated animals usually have been selected by man because they have favorable communication signals. Thus, they are generally not specifically colored, for genetic manipulation in domestication is apt to result in a variety of color patterns. Instead they use acoustic, chemical, or postural communication signals in reproduction, and these tend to be relatively simple and unchanging. So it should be possible, by recording or collecting chemicals for controlled release, to affect the behavior of these animals.
So far, except for isolated anecdotal reports, there has been little reported along this line. Signoret et al. (1960) have reported, for swine, increased mating readiness when recordings of sounds made by the animals during mating are played to them. It is well known that many domestic animals are attracted by sounds and odors of the opposite sex during the mating season. All these suggest possible future practical uses for communication signals with domestic animals.
Students of human communication hope that studies on animals may give them information about basic communication processes. So far, few of the studies on animal communication have practical implications in human communication. In general, when comparisons between human and animal communication have been essayed, they have involved the question of whether animals have language or not. The answer to this, of course, depends upon one’s definition of language. In other words, the direction of comparison has been from man to animal, not generally from animal to man. Certainly studies of such elaborate communication systems as those of bees in tactile channels, and of other insects in chemical channels, might lead to a better understanding of human communication in nonvocal channels, but so far there is little to show in the way of practical applications.
Some years ago, von Frisch (1946, 1953, 1965) foresaw the possibility that one might apply his discoveries about the communication system of honey bees in bee management. Some of his discoveries on sensory physiology of bees had almost immediate use, e.g., painting hives with different colors, thus making it easier for a bee to find her own hive in a group. Von Frisch has suggested that, since honey bees communicate in part by odors, either their own or of flowers in which they have found nectar, the application of these odors to bees in the hive might direct others to food sources.
The signaling system of scout bees involves waggling movements on the comb, and it is not easy to see offhand how these might be used to inform bees of a food source before scout bees have found it. It is by no means impossible, however. The recent discoveries of the place of sounds in honey bee communication (reviewed in this volume) could mean that one might play back recordings of these sounds to honey bees in the hive and thus communicate the distance from the hive to the food source.
When honey bees are ready to swarm, they produce certain characteristic sounds, as beekeepers have known from antiquity. An electronic device (Woods, 1959) utilizes this fact to enable beekeepers to prevent swarming or to have available an alternate hive for the bees. It certainly is possible to place a microphone in each hive, record or sample the sounds coming from it, and thus determine conditions in the hive. So far, these procedures have not proved reliable and cheap enough that they have become standard.
Many insects produce acoustic communication signals (Haskell, 1961), and these might be used for control. In many Orthoptera, mature males produce characteristic sounds that call in the females (reviewed elsewhere in this volume). This suggests that recordings of the sounds might be used to attract the females to places where they can be captured or destroyed. So far, tests have been small-scale, and they generally have indicated that reproduction of the sound must be of high fidelity and therefore rather expensive.
Some moths respond to ultrasonic sounds similar to the pulsed ultrasounds that are used by bats to find and capture the moths in flight. A general reaction is avoidance by dropping from the air. Belton and Kempster (1962), using bat sounds, have tried to set up an acoustic fence around an orchard to cause moths to drop out of the air. They reported an increased productivity of these orchards as a result of decreased infestation rate. So far, however, these tests have been only what might be called pilot-scale tests. Other practical possibilities with insects are reviewed by Haskell (1961) and Frings and Frings (1962, 1965). Certainly the wide variety of insect sound producers and the wide distribution of sound receptive mechanisms in insects suggest that there may be practical possibilities here, but some sort of conceptual breakthrough may be necessary if practical uses are to come.
CHEMICAL COMMUNICATION SIGNALS OF INSECTS
For well over a hundred years it has been known that the sexes of many insects are attracted to each other by special odors. These chemical attractants, or communication signals, may be active over long distances under appropriate conditions. About seventy years ago the suggestion was made that these materials, collected from the animals and purified, might be used to attract pest species to their destruction. Collins and Potts (1932) prepared a relatively pure attractant from the female gypsy moth and used it for censusing by trapping the attracted males. This method for censusing by attractant odors is now widely employed.
Actually, sex attractants are only one class of chemical communication signals that are widespread in insects, integrating various aspects of social life. For example, in the honey bee colony, queen substance and other materials are all involved in the organization of the economy of the hive (Butler et al., 1961; Butler and Fairey, 1963, 1964; Callow et al., 1964). Among ants there are chemical materials that are involved in the communication of alarm through the colony (Wilson, 1963b; Maschwitz, 1964). In both ants and bees, odors are used in the designation of location of food sources. Barnhart et al., (1963) have shown that, in flies, addition of odorous substances from the bodies to materials, such as sugar, which have no odor in themselves attracts other flies to the food.
The wide distribution of these chemical signals and the variety of effects that they have upon insects led to the coining of a special term to designate them—pheromones (Karlson and Lüscher, 1959; Karlson and Butenandt, 1959; and Karlson, 1960). The name purposely resembles the word hormone, because it was thought that these substances were hormone-1ike in action. Pheromones, like hormones, can have various actions—exciting to activity, as with the sex attractants, or inducing chemical or metabolic changes, as with queen substance of bees, which affects the development of reproductive organs. The coining of this term has given considerable impetus to research in the field (Wilson, 1963a; Wilson and Bossert, 1963) and as such should be considered valuable. It is also unfortunate, however, for it artificially sunders these communication signals from other types—acoustic, visual, etc.—as if they were separate, whereas all are similar in effects. The statement that pheromones affect metabolic or developmental processes does not differentiate them, for acoustic signals, at least, may also have effects upon growth or maturation. It would seem preferable, if a single term is useful, to designate communication signals in general (e.g., as semants) and to have specific classes designated as part of this generality (e.g., phonosemant, chemosemant, etc.). However, the term, pheromone, has met with wide acceptance, even among workers who ordinarily are not interested in animal communication, and it is probably here to stay.
Wright (1964b) has also objected to the term because, unlike hormones, the effective chemical communication signals are usually complex, not single materials. Furthermore, they usually induce behavioral changes rather than call forth metabolic actions in the receiver. Wright (1964c, 1965) has suggested that chemical communication signals, at least, might have value in pest control, and he has coined the word metarchon to designate this class of pest control agents. Metarchons are defined as external stimuli artificially introduced into the environment of the organism for the purpose of modifying its behavior by eliciting inappropriate responses or inhibiting appropriate ones. Wright (1964a, b, 1965) has suggested that it might be possible to flood a whole area with the sexual attractants of females of a pest species and thus completely confuse the males, for they would have no directional information by which to find the females. Such overloading of a communication system—essentially an increase in the noise level to the point that the signal-to-noise ratio becomes insignificantly small—might be used also with sounds.
A number of recent reviews of insect sex attractants have appeared (Beroza, 1960; Jacobson and Beroza, 1963; Jacobson, 1963, 1965). We can only summarize the studies very briefly here. In general, sex attractants are produced more by females than vice versa. Jacobson (1965) lists 159 species in which females lure males, but only 53 species in which males lure females, 40 of them Lepidoptera. A number of the sex attractants of insects have been extracted in pure form and identified; some have been synthesized. In some cases, the synthetic chemicals have proved to be just as attractive, or even more attractive, than the natural materials. These are not all from a single organic group, but represent different classes of compounds. Chemical sex attractants are usually quite specific in action. In fact, in at least one species, the antennal receptors of the male seem to respond only to the chemical communication signal of the female, whereas the antennal receptors of the female do not respond to this at all.
The use of insect sex attractants in control has not been as successful as early workers thought it might be. We have already mentioned that the most successful use has been for censusing insects, particularly the gypsy moth. This gives a check on population changes, and allows one to apply controls when the insects are present. Seemingly the more obvious use of insect sex attractants would be to attract pests to their death in some form of trap. Where this has been tried, however, it has not been effective in reducing populations, for generally not all individuals are caught, and those that are left are capable of producing enough fertilized eggs, with the high reproductive potential of insects, to maintain the population. It has been suggested that a better method might be to attract males to an area where they can be brought into contact with chemosterilants, sterilizing but not killing them. These sterile males would then mate with normal females, producing no offspring. In some species, a female mates either with only one male or with relatively few males, so that, if the majority of males in an environment were sterilized, the majority of matings would be sterile. As in the case of the now famous control of screwworm fly, this might result in effective reduction of the population.
Saturation techniques, such as has been suggested by Wright, have not so far been tried on a large scale, and where they have been tried at all, they have not given as satisfactory results as might be hoped. Certainly, as Wright (1965) points out, the amount of chemical sex attractant needed to saturate an area may be very small. With Gyplure, the sex attractant of the gypsy moth, he calculates that only 3.5 pounds per acre would be needed to protect an area for 100 days. This is based, of course, on a number of assumptions that he makes as to the effectiveness of the material, needs for retreatments, wind movement, etc. Jacobson (1965) points out, however, that the material may have to be present all the time, thus not allowing economies that are postulated, and that resistant strains may arise.
In general, therefore, for insect sex attractants—metarchons, or pheromones, as one likes—there is a large body of fundamental knowledge about the production, synthesis, and effects of these materials on insects. It would certainly seem that practical uses, besides censusing, should be possible, but so far large-scale tests have not given as satisfactory results as might be desired.
BIOACOUSTIC CONTROL OF PEST BIRDS
The use of acoustic communication signals for controlling the movements of pest birds dates from approximately 1953; this is therefore a relatively recent development. Previous to that time, sounds had been used for bird control, but they were generally mere noises, and the birds rather readily adapted and failed to respond. Furthermore, the noises were often produced by fireworks, with some fire hazard. The use of recorded communication signals to influence bird movements has certain advantages over the use of noises in practical pest control: (1) The signals are effective at intensities as low as 3 dB above ambient noise levels; (2) habituation or adaptation is slow, for communication signals are part of the social structure of bird populations, and (3) the signals may be specific, allowing one to control a particular species, or nonspecific, allowing broad-spectrum control. An extended review of this field is given by Frings and Frings (1967), with earlier reviews, in the form of symposium reports, edited by Busnel and Giban (1960, 1965), and Giban (1962).
Types of Calls Used for Bird Control
Among a variety of call notes of birds (Thorpe, 1961; Brémond, 1963) two have been used almost exclusively for pest control—distress calls and alarm calls. Distress calls, by all odds, have been the most widely studied. The term, distress call, unfortunately, has been used for more than one type of call. As used here, it refers to the call given by a captive bird when held in the hand or maltreated. Alarm call refers to a call given by a bird that is free itself when it sights a captive member of its group or potential danger, such as a predator or human being.
The distress call of the European starling was the first used for practical control of pest birds (Frings and Jumber, 1954). It is obtained by holding a starling upright by the legs without injury. Some workers claim that holding the bird upside down increases the effectiveness of the call, but variations from bird to bird are so great that this is not certain. Interestingly enough, distress calls given by caged starlings are essentially ineffective when played back to wild starlings in the field. The distress call of most birds is usually fairly raucous, as one might imagine. Reactions to distress calls vary with the species. Generally starlings are repelled by the distress call. On the other hand, gulls and crows show a complex series of reactions, which have been studied in great detail by Busnel, Giban, Gramet, and Brémond (cf. reviews and colloquium proceedings). Usually, in the case of gulls and crows, the birds are first attracted and later dispersed. It is thus obvious that the distress call may not in all circumstances be an appropriate call for practical use if one wishes to remove birds from a particular area.
Alarm calls are usually different from distress calls, and birds react to them differently. In general, birds leave an area in which an alarm call is broadcast, but this is not always so. Herring gulls, for instance, are first attracted and then dispersed. There have been far fewer practical studies on reactions to alarm calls than to distress calls. The evidence that is available so far, however, suggests that alarm calls are more effective than distress calls.
The first practical tests, using the distress call of the European starling (Sturmis vulgaris) to drive these birds from tree roosts, were made in the summer of 1953 by Frings and Jumber (1954). These tests, using rather primitive equipment, proved so promising that other trials were made in the summer of 1954, again with starlings in tree roosts, and reported by Frings et al. (1955).
In the meantime, in France, Busnel and Giban had begun a series of studies on bioacoustic bird control that were to become the most extensive in the world. They were chiefly interested in corvids and, later, gulls. Their work has been reported in many individual publications and summarized in the proceedings of the three symposia referred to above. In the summer of 1954, the method was extended in the United States to a study on the herring gull (Larus argentatus), showing the effectiveness of the alarm call, and in 1955 to the eastern crow (Corvus brachyrhynchos). These studies led to a series of cooperative studies between the French and United States workers which demonstrated the existence of dialects in these calls (Busnel et al., 1957; Frings et al., 1955a, b, 1958; Frings and Frings, 1959). This means that specific calls from specific areas are needed to control pests in those areas. These discoveries further explained the failure of reactions when recordings of calls were shipped from one country to another.
In Germany the first tests were made in 1955 and were strikingly unsuccessful. For a few years, the German workers could seemingly obtain no reactions with birds, using either distress or alarm calls. In 1959, however, they discovered that their earlier failures resulted from poor fidelity in recording and broadcasting equipment (Bruns, 1959, 1960; Schmitt, 1959). Since 1959, there have been rapid developments in Germany in the use of recorded acoustic signals to control starlings. In Switzerland, Müller and Gerig (1957) were successful in moving starlings and blackbirds from areas where they were not desired, and they later applied the technique in olive orchards in Tunisia (Buchman and Müller, 1957). In Holland, after some disappointment, Hardenberg (in Colloquia) reported control of gulls that were infesting airports. Attempts in the United States and England to control gulls at airports by using recorded communication signals have so far proved unsatisfactory; the reasons for this are obscure.
Most work so far has been on the European starling in the United States and Germany and on corvids and larids in France. Certainly these birds become important pests, but there are many other important pest birds that have not been studied at all: quelea, bobolink, American robin, many species of blackbirds, ricebirds, and even wild ducks and geese. Since all of these have communication systems and their call notes can now be easily recorded and broadcast, studies on bioacoustic control seem warranted.
Equipment and Informational Needs
There is now available a wide variety of recording and broadcasting equipment suitable for use with birds. It is rather interesting that in some cases birds seem to tolerate almost any sort of distortion and still respond to the recorded communication signals, whereas in others they require rather high fidelity. The most striking case of this is found in the European starling. In the United States, the original recordings were of low fidelity and the projecting equipment was of poor fidelity; yet the birds responded well. In Germany, on the other hand, starlings did not respond to recordings until they were of sufficient fidelity. The fidelity needed varies from species to species, and even within the same species. Generally, equipment that allows one to record and play back the repertoire of calls of a bird is easily available, but in some cases it may need to be checked for requisite fidelity.
But merely finding a call to which birds react is not control. If this seems to be trite, one should remember that there have been many persons—and not all nonscientists—who were confused about this. In the early days of testing, 1954-1956, we received disappointed reports from people who had tried the starling distress call to chase the birds. It was obvious that to these people the call was all that was necessary, and they expected that by merely playing the call they would achieve control. But it should be obvious that one does not expect even the most potent insecticide to offer a solution to all problems. In fact, the synthesis of an insecticide and proof of its activity against insects in the laboratory is merely the beginning. In one of the earliest papers on this subject (Frings et al., 1955, p. 21) the following statement was made:
It is hoped that this information on our tests will put to rest belief that this method is simple and fool-proof. Certainly the use of broadcast recorded calls is safe and relatively cheap. But the clearance of an objectionable roost requires attention to equipment, knowledge of the habits of starlings, coordination during the treatments, and persistence in the work. No one expects to cure a disease with one easy treatment, nor to eliminate pest insects or rodents without continued effort. No one should expect to clear starlings from objectionable roosts without intelligent and persistent work.
It is absolutely necessary, if one wishes to have effective control, that he consider all facets of behavior and ecology of the bird. We may categorize these under six headings: (1) identification of the species, (2) nature of the problem, (3) population size and structure, (4) environmental conditions, (5) behavior patterns of the animals involved, and (6) possible factors interfering with control operations. We shall discuss these only very briefly.
The identification of the species may seem to be so elementary as to be not worth mentioning, but this is by no means true. Obviously, if one is to use communication signals in everyday pest control, the methods must work in the hands of relatively untrained individuals. And here the problem of identification is encountered; for many persons are not familiar with birds. They often lump together all black birds as blackbirds or starlings. Since communication signals may be variably specific, it is absolutely essential to know what species is involved in a particular problem.
There are four general types of problems created by pest birds: (1) their mere presence may at certain places be objectionable to some people; (2) they may roost in large number in trees or on buildings; (3) they may rest or loaf in areas of critical traffic, such as at airports; and (4) they may devour food that man wants. Each of these requires special handling which we can not go into here.
In general, the most difficult of bird pest problems are those that involve feeding. Birds are not usually going to be scared into starving to death, although it has been reported in some cases that birds kept away from food by the distress call have actually starved. Roosts may be difficult to clear, particularly winter roosts, when alternatives are not possible; they may be rather easy to clear if alternative roosts are present. Birds that are merely resting are usually easiest to move if other areas are available. In general, it is easier to move birds at the beginning of an activity cycle—e.g., as they form a roost or start to feed—than it is to move them when they are fully established.
The size of the population is obviously important, determining how large an area may need to be treated. If one is using sound trucks or many loudspeakers, this will determine the number needed. Also, the behavior of the birds varies with the numbers of individuals in the population. Generally, it is easier to move a large group of birds than a small one. The flight of a few individuals in large groups may create a sort of mass hysteria in the whole flock. Where the birds are more scattered, the flight of one may have little influence on the others. The composition of the population is, of course, important. For instance, young individuals may respond less or more readily than do adults.
Environmental conditions, such as topography and weather, need to be considered in planning control programs. The topography of the area determines the distribution of sound-projecting equipment or sound trucks to be used. Thus, if one projects sounds beneath trees, which are quite sound-absorbent, more power may be needed than if he is broadcasting against buildings. Furthermore, buildings may reflect sounds or create sound shadows, necessitating movement of speakers or sound trucks. Weather may be important not only in causing cancellation of outdoor tests but also in determining distances to which sounds travel adequately.
Behavior patterns of animals are obviously important. First of all, one has to find a reaction to a communication signal if he is to use it as the basic unit for a control program. Migrations and movements of birds over a period of time are also important. For instance, European starlings in the eastern United States breed in March to May and are then scattered and not pests. In early summer, they accumulate in small tree roosts; by July or August a city or town may have a number of relatively small roosts (10,000 or fewer). While these are small in relation to the over-all population, they may be very disturbing to people nearby. These roosts by late August or September usually fuse into large autumn tree roosts which remain until leaf fall. The birds now redistribute themselves for the winter, some inside shelters, some to the south, and some taking up their abode on buildings in cities, again becoming pests. It is much easier to chase these birds from areas where they are not desired if one starts when they are just organizing the roosts. Thus, to clear a building the best time is as the roost begins to form in the fall.
Generally birds have regular diurnal movements from place to place. Thus, starlings in tree roosts enter and leave the roosts at definite light intensities. They do not come directly from feeding into the roosts, but instead stage in through a series of preroosts. The latter are usually numerous, and apparently the birds are willing to substitute one preroost for another rather readily. So, if one broadcasts a call near a preroost, it may seem to chase the birds successfully, whereas actually there has been no effect on the true roost itself.
Interfering factors may be various in practical bird control. An obvious one is the presence of background noise. When one broadcasts calls in a city or alongside a busy highway, there is a high level of background noise, and the calls must be loud enough to reach the birds. If one broadcasts calls at high intensity inside a hollow building, such as a hangar, reverberation may result in distortion. So, practical uses of recorded communication signals for bird control involve both biological and acoustic knowledge. It may be necessary to place speakers differentially or to plan movements of sound trucks so that the proper intensity and fidelity reach the bird.
Probably the most disturbing factor, to a scientist, in testing acoustic control of birds is the reactions of human beings to the tests or to the calls. The calls must usually be broadcast at a high enough level to reach birds in the roosts. This may mean that near the source of the sound the intensity is fairly high, and this can produce problems in public relations.
Planning and Executing Bird Control Operations
It should be obvious that the method of application of communication signals is the most important factor determining success or failure. It is not enough merely to use a signal, no matter how effective, without knowledge of the behavior and ecology of the birds.
These tools must be used in a coordinated program to achieve the objectives. As with drugs, when used clinically, the patient is not usually satisfied with failure just to discover some new information. An agriculturist generally wants his crops protected, not a demonstration of how not to do the job. A town council wants the birds cleared from the trees or buildings, not data on sound intensities too low to do this. Unfortunately, at least in the U.S., there are no experimental bird infestations maintained just for testing purposes. Thus, while scientifically one should keep all factors constant, practically he seldom can, and a constant feature of tests of acoustical bird control is changing the conditions as data accumulate. Consequently, reports of tests tend to be long and drearily detailed, hard to summarize briefly, often raising more questions than they answer (Frings and Frings, 1967, p. 426).
The first practical point in applying communication signals for pest control is the selection and deployment of equipment. We have already noted that it is necessary to have equipment of the proper fidelity and power, and we need not belabor this point. Deployment of sound sources is an important consideration. Actually, this is an acoustic problem, not a biological problem, but biology is involved in that one must know the intensity and fidelity needed. These determine how many speakers or how many sound trucks need be used. This may involve both acoustic engineers and biologists for planning. In some experiments in the past, problems arose, for sound engineers did not appreciate the biological aspects, and biologists did not appreciate the acoustic aspects.
Wherever possible, it is simpler to use fixed installations, but this may be expensive if the infested area is extensive. For a single building or a single group of trees, one may use a few speakers so placed as to get correct coverage. For a tree-roosting area, which may cover several city blocks, placement of many speakers may be prohibitively expensive. In this case, it is much cheaper to use a sound truck. In all, the ultimate objective with equipment is to have it as reasonable in cost as possible, consistent with required fidelity, mobility, and power.
The second practical point is the selection of signals to be used. Selection may be made, of course, before one even considers practical applications, but one may need to change the signal during the operation. So far, most studies have been made with distress calls of birds. They seem to be quite effective, but the species that have been studied are all social species—starlings, corvids, gulls. For nonsocial species, distress calls may be relatively ineffective.
Alarm calls generally seem to be more effective than distress calls, and in Germany alarm calls of starlings are used rather than distress calls. In some cases, e.g., attempts to chase gulls from airports in Britain and the United States, it is surprising that alarm calls, which have been reported to be effective, have not been used instead of distress calls, which are generally not nearly so effective.
So far there have been no practical studies on the use of attractive calls to draw birds away from areas where they are not desired. There are calls, such as the assembly call of the eastern crow and food-finding call of the herring gull, which might be used for this purpose (Frings et al., 1955; Frings and Frings, 1957). Actually, as long as distress or alarm calls are effective, there is probably little reason to shift to these calls, but studies should be made. Other sounds, such as calls of predators, or calls given when birds sight predators, remain to be tested.
Application methods are primarily determined by the type of problem that one has. As noted earlier, there are three basic types of problems: roosting, resting, and feeding. Each presents some peculiar aspects that must be taken into account when sounds are applied.
Usually roosts are at one spot, in trees or on buildings, and are occupied only at night and ordinarily only at certain times of the year. This means that times of application can be short, as the birds come into roosts, and only during certain seasons. If the roosts are small, e.g., a single tree or one building, one may use a single loudspeaker or a small group of speakers as fixed installations. If the roosts are extensive, e.g., tree roosts in town, it may be necessary to use one or more sound trucks.
For roosting starlings, the sound is generally broadcast when the birds come into roosts, as darkness falls. Ordinarily, the sound is used as little as possible: only for twenty to forty seconds at a time and only while the birds are in the roost. So the total application time may be only ten to twenty minutes, spread over a period of about one to one and one-half hours. Usually three to five consecutive evenings of sound treatment are necessary to achieve a clearance. On the first evening, the sound is generally applied at the definitive roost, and it seems best to wait until most of the birds have entered the roost in the dusk. As soon as the birds have flown away, the sound is turned off, and it is not turned on again until they try to return. As darkness falls, the birds become more persistent, and finally they drive into the roost regardless of the sound. At this time, the sound is discontinued. Superficially, it looks as if nothing has happened, for there seem to be about as many birds at the end as at the beginning. It can be noted, however, if one has studied the roost carefully ahead of time, that the birds have been disturbed. They do not occupy their usual positions but instead have to take any position they can find as darkness comes on and they can no longer see.
On the second evening, the same procedure is used, except that usually the sound emissions are started earlier and the birds are driven back toward the preroosts and held there as long as possible. Ordinarily, the second evening shows a slight reduction in the number of birds. On the third evening, the sound is moved out to the preroosting areas. If one has mobile units, the birds are irradiated with sound as soon as they try to come into the preroosts, and they are driven centrifugally from the definitive roost. If this is done correctly, one disperses fifty to seventy-five per cent of the birds, and they do not come back.
Further treatments are like those on the third evening and may extend to five or six nightly periods. With large and persistent roosts, there may be a residue of five to ten per cent of the birds originally in the roost still returning after these nights. It is best to leave these birds, for they generally join the major flock later.
This is obviously a method for dispersing birds; it does not destroy them. Therefore, it is often necessary to direct the birds to alternative roosting areas where they may be acceptable. For building roosts, this may mean driving them completely out of the city into surrounding areas, where they may find protection. For small towns, it is often possible to drive the birds into woodlots outside of town. Busnel and Giban (1960) have shown, with French corvids, that one can reduce the population by treating communal nesting sites. In this case, the birds gather in trees to nest and raise their young. The French workers treated these roosts with broadcasts of recorded distress calls in the middle of the night. This chased the birds, and, in the dark, they could not find their way back to the eggs. The chilled eggs did not hatch, and a whole generation from that nesting area was destroyed.
The most publicized problems with resting birds are those along or on runways of airports. Considerable research has gone into efforts to try to keep these birds from becoming hazards to aircraft (Busnel and Giban, 1965). For bioacoustic control, sound trucks or fixed installations of speakers along the runways are used. Just prior to takeoffs and landings of aircraft, the recorded signals are broadcast, driving the birds away. It is obvious that, if airplane movements are numerous, the sound may be almost continous, and this could create a problem of habituation. This can be obviated by changing the calls, by moving the sound source, or by adding other frightening situations, e.g., firecrackers, moving persons, etc.
Feeding problems are tremendously varied and little can be said in general about them. Usually only short times of control are necessary, for birds only eat fruit or vegetables at certain stages of maturity, so habituation to the signals is not likely to be a problem. The most practical and reasonable type of installation is a series of fixed speakers so placed that complete coverage of a field is assured. Automatic timers may be used to broadcast the sound at definite time intervals. So far there have been few studies on time intervals needed for effective control of pest birds; little can be said in general about this. It is usually necessary to have someone present, at first, to find the necessary time intervals. If the fields are extensive, fixed speakers may be expensive. In this case, a mobile unit may be used; this requires a driver and thus adds to the cost. Comparative costs, therefore, determine the type of equipment to be used.
Correct timing of broadcasts of recordings is still uncertain for most species. Indeed, different workers have reported equal success with different time intervals; some using the sound as little as possible, others having the sound on almost all the time. Offhand, one would think that the sound should be broadcast as little as possible, for there is evidence that birds may come to recognize the recordings. But it must be said that, where long-continued applications have been tried, e.g., in Germany and in the United States, habituation did not occur. In France, Busnel, Giban, and co-workers have developed equipment that operates automatically on average time schedules, with random intervals between broadcasts. But, they point out, it is sometimes necessary to change the schedules depending upon the “pressure,” that is persistence, of the birds.
So far, where control has been achieved, the applications have been made by trained biologists, who are sensitive to behavior patterns of birds and adjust the applications to suit particular conditions. There may be a question, therefore, whether the method, unless automated, will be suitable for public use. Up to now, however, most of the tests have been in the nature of experiments. Where commercial pest control operators have been instructed carefully, they too have been successful. Wrights’ (1964b) suggestion of saturating the environment with communication signals to jam the animals’ systems has not been tried with birds. Conceivably, one might flood an area with calls or territorial songs of birds and thus shatter the social structure. It is obvious, however, that covering any large area with sound is an expensive operation.
Problems in Uses of Communication Signals for Pest Control
We have mentioned, off and on, the problem of habituation or adaptation. In the past, a major difficulty with the use of sounds, usually loud noises, for bird control has been that the birds soon stopped responding. Birds roost or nest near noisy human operations and quickly learn to ignore these artificial noises. With recorded distress or alarm calls, in spite of some experimental evidence of habituation if the distress call were played at too low an intensity (Frings, et al., 1955) habituation has not occurred in practical operations. As a matter of fact, there is evidence now, in agricultural applications, that the birds may learn to avoid an area in which the signals are broadcast, and the sound may be unnecessary after a while (Bruns, 1960; Boudreau, 1962, 1964).
Birds that are chased by recorded communication signals are not killed, so this is no means of population control. Buchman and Müller (1957), however, reported that chasing starlings from olive orchards resulted in deaths of some birds which apparently lacked sufficient food to live overnight; so some population control might be achieved indirectly, but this would be relatively small. We have already noted that Busnel and Giban (1960), by scaring nesting birds, destroyed unhatched babies, but this too involved special circumstances. Generally, the birds are driven from one place to another. If they go from a place where they are pests to another where they are equally pests, one has accomplished little. However, it is often possible to disperse large roosting or feeding flocks and bring them to manageable size. Furthermore, one can often select a place where the birds can stay and drive the birds there. If this is a wild area, there may then be increased predator pressure on them, thus reducing the population. They may also adopt habits that are not objectionable, such as feeding on wild fruits or vegetables. It does not follow, therefore, that birds driven from one problem area always go somewhere else where they will become a problem once again.
From a practical standpoint, the expense involved in applying communication signals for pest control must be comparable with that of other methods. Sound equipment is by no means cheap, but it is generally fairly permanent. Also, amplifiers, etc. can often be used otherwise in a factory or town, so that the total cost need not be applied against pest control. Generally, a farmer knows fairly well his losses and can make an accurate judgment on comparative expense. In urban situations, where the birds are merely a nuisance, however, cost is not clear-cut, and control usually becomes a political matter, often far removed from scientific considerations.
The discovery that birds have dialects, and conversely that some bird calls produce reactions in more than one species, suggests that problems might arise here. It may not be possible, for instance, to use recordings made in one laboratory or one country in another. It is usually necessary for distress calls to be recorded where they are to be used. This may be advantageous, for it allows precision in control. Interspecificity, conversely, reduces precision but increases efficiency, if one wishes to get rid of all birds in an area. Thus, whether specificity and interspecificity are problems or advantages depends upon the population structure and particular problem.
When broadcasts of recorded distress calls or alarm calls are used in cities or towns, some people are disturbed by the sounds. Luckily, most birds are active in daylight, but even then loud sounds can be irritating to sensitive people. It would certainly be advantageous if appropriate calls were ultrasonic, but unfortunately there is no evidence that birds hear ultrasounds or that ultrasounds form an important part of their communication signals (Schwartzkopff, 1955; Frings and Slocum, 1958; Frings and Cook, 1964). Generally the broadcasts do not disturb domesticated animals and are specific enough that, when played near chickens or ducks, they do not disturb those birds.
These problems, however, must be balanced against some practical advantages of communication signals for pest control. There is no fire or soiling hazard, as with firecrackers or insecticides. Sound generally can be kept within bounds and does not drift out of a treated area as do insecticides. And, possibly most significantly, there are no undesired residues.
Much further work is needed on many more species of birds before recorded communication signals will take their place in the armament of practical pest control. Considerable practical progress has been made in this field, enough to indicate the possible future value. Communication signals, whether synthesized or recorded, however, are only tools. If they are to be used successfully for pest control, one must study all aspects of the behavior and ecology of the animals involved. There are no panaceas among any control techniques.
Barnhart, C. S., and L. E. Chadwick, 1953. A “fly factor” in attractant studies. Science, 117:104-105.
Belton, P., and R. H. Kempster, 1962. A field test on the use of sound to repel the European corn borer. Ent. Exp. and App., 5:281-288.
Beroza, M., 1960. Insect attractants are taking hold. Agr. Chem., July 1960, pp. 37-40.
Boudreau, G. W., 1962. Summary report of Western House Finch control experiments, Delano, California. Mimeo., Phoenix, Arizona.
Boudreau, G. W., 1964. Bird pests in Arizona spring lettuce. Mimeo., Phoenix, Arizona.
Brémond, J. C., 1963. Acoustic behaviour of birds. In: R.-G. Busnel (Ed.). Acoustic behaviour of animals. Elsevier Publ. Co., Amsterdam, pp. 709-750.
Bruns, H., 1959. Erfolgreiche Starenabwehr mittels Leuchtkugeln, Raketen und Tonband. Pflanzenschutz, Heft 5, pp. 79-81.
Bruns, H., 1960. Starenabwehr mittels pyro- und phonoakusticher Methoden. Tagesberichte No. 30 (Probleme der Angewandte Ornithologie) der Deutschen Akademie der Landwirtschaftswissenschaft zu Berlin, 30:105-110.
Buchmann, E., und O. Müller, 1957. Das Tonbandgerät im Dienste des Olivenbaues in Nordafrika. Schweiz. Zeit. Obst- und Weinbau, 66:575-580.
Busnel, R.-G., J. Giban, P. Grämet, H. Frings, M. Frings, and J. Jumber, 1957. Interspécificité de signaux acoustiques ayant une valeur sémantique pour des Corvidés européens et nord-américains. C. R. Acad. Sci., 245:105-108.
Busnel, R.-G., et J. Giban, 1960. Colloque sur la protection acoustique des cultures et autres moyens d’effarouchement des oiseaux. Jouy-en-Josas les 26 et 27 novembre 1958. 246 pp.
Busnel, R.-G., et J. Giban, 1965. Colloque le problème des oiseaux sur les aérodromes. Nice les 25, 26 et 27 novembre 1963. Inst. Nat., Rech. Agron., Paris. 326 + xx pp.
Butler, C. G., R. K. Callow, and N. C. Johnston, 1961. The isolation and synthesis of queen substance, 9-oxodec-trans-2-enoic acid, a honeybee pheromone. Proc. Roy. Soc. B, 155:417-432.
Butler, C. G., and E. M. Fairey, 1963. The role of the queen in preventing oogenesis in worker honeybees. Jour. Apicultural Res., 2:14-18.
Butler, C. G., and E. M. Fairey, 1964. Pheromones of the honeybee: Biological studies of the mandibular gland secretion of the queen. Jour. Apicultural Res., 3:65-76.
Callow, R. K., J. R. Chapman, and P. N. Paton, 1964. Pheromones of the honeybee: Chemical studies of the mandibular gland secretion of the queen. Jour. Apicultural Res., 3:77-89.
Collins, C. W., and S. F. Potts, 1932. Attractants for the flying gipsy moths as an aid in locating new infestations. U.S. Dept. Agric. Tech. Bull. No. 336, 43 pp.
Frings, H., and Betty Cook, 1964. The upper frequency limits of hearing in the European Starling. The Condor, 66:56-60.
Frings, H., and M. Frings, 1957. Recorded calls of the Eastern Crow as attractants and repellents. Jour. Wildlife Management, 21:91.
Frings, H., and M. Frings, 1959. The language of crows. Sci. American, 201(5): 119-120, 122-124, 126, 128, 130-131.
Frings, H., and M. Frings, 1962. Pest control with sound. Part I. Possibilities with Invertebrates. Sound, 1:13-20.
Frings, H., and M. Frings, 1965. Sound against insects. New Scientist, 26:634-637.
Frings, H., and M. Frings, 1967. Behavioral manipulation (visual, mechanical and acoustical). Ch. 11, In: Pest control. Biological, physical, and selected chemical methods. W. Kilgore and R. Doutt (eds.). Academic Press, N.Y., pp. 387-454.
Frings, H., and J. Jumber, 1954. Preliminary studies on the use of a specific sound to repel starlings (Sturnus vulgaris) from objectionable roosts. Science, 119:318-319.
Frings, H., J. Jumber, and M. Frings, 1955. Studies on the repellent properties of the distress call of the European Starling (Sturnus vulgaris). Occasional papers from The Department of Zoology and Entomology, Pennsylvania State University, No. 55, 21 pp.
Frings, H., M. Frings, B. Cox, and L. Peissner, 1955a. Recorded calls of Herring Gulls (Larus argentatus) as repellents and attractants. Science, 121:340-341.
Frings, H., M. Frings, B. Cox, and L. Peissner, 1955b. Auditory and visual mechanisms in food-finding behavior of the Herring Gull. Wilson Bull., 67:155-170.
Frings, H., M. Frings, J. Jumber, R. G. Busnel, J. Giban, and P. Gramet, 1958. Reactions of American and French species of Corvus and Larus to recorded communication signals tested reciprocally. Ecology, 39:126-131.
Frings, H., and B. Slocum, 1958. Hearing ranges for several species of birds. Auk, 75:99-100.
Frisch, K. v., 1946. Die “Sprache” der Bienen und ihre Nutzanwendung in der Landwirtschaft. Experientia, 2:397-404.
Frisch, K. v., 1953. Aus dem Leben der Bienen. Springer-Verlag, Berlin, vii + 182 pp.
Frisch, K. v., 1965. Tanzsprache und Orientierung der Bienen. Springer-Verlag, Berlin, vii + 578 pp.
Giban, J., 1962. Colloque sur les moyens de protection contre les espèces d’oiseaux commettant des dégats en agriculture. Ann. Epiphyties, 13(No. Hors Série): 1-258.
Hafez, E. S. E., 1962. The behaviour of domestic animals, xiv + 619 pp. Baillière, Tindall and Cox, London.
Haskell, P. T., 1961. Insect sounds, viii + 189 pp. Quadrangle books, Chicago.
Jacobson, M., 1963. Recent progress in the chemistry of insect sex attractants. In: Hall, S. A. (Ed.). New approaches to pest control and eradication. Advances in Chem. Ser. No. 41, pp. 1-10.
Jacobson, M., 1965. Insect sex attractants. xi + 154 pp. Interscience Publ. New York.
Jacobson, M., and M. Beroza, 1963. Chemical insect attractants. Science, 140:1367-1373.
Johnson, M. W., 1948. Sound as a tool in marine ecology, from data on biological noises and the deep scattering layer. Jour. Marine Res., 7:443-458.
Karlson, P., 1960. Pheromones. Ergebn. Biol., 22:212-225.
Karlson, P., and A. Butenandt, 1959. Pheromones (Ectohormones) in insects. Ann. Rev. Ent., 4:39-58.
Karlson, P., and M. Lüscher, 1959. “Pheromones”: a new term for a class of biologically active substances. Nature, 183:55-56.
Maschwitz, U. W., 1964. Alarm substances and alarm behaviour in social Hymenoptera. Nature, 204:324-327.
Moulton, J. M., 1964. Underwater sound: biological aspects. Oceanogr. Marine Biol. Ann. Rev., 2:425-454.
Müller, O., and L. Gerig, 1957. Das Tonbandgerät im Dienste des Obst- und Weinbaues. Schweiz. Zeit. Obstu. Weinbau, 66:34, 51-53.
Schmitt, N., 1959. Das Tonband im Dienste der Schadvogel-Bekämpfung. Gesunde Pflanzen, 11(2):32-38.
Schwartzkopff, J., 1955. Schallsinnesorgane, ihre Funktion und biologische Bedeutung bei Vögeln. Acta XI Congr. Int. Orn. 1954, pp. 189-208.
Signoret, J. P., F. du Mesnil du Buisson, et R. G. Busnel, 1960. Rôle dun signal acoutique de verrat dans le comportement réactionnel de la truie en oestrus. C. R. Acad. Sci., 250:1355-1357.
Thorpe, W. H., 1961. Bird-song. xxii + 143 pp. Cambridge University Press, Cambridge.
Wilson, E. O., 1963a. Pheromones. Sci. American, 208(5):100-106; 108; 110; 112; 114.
Wilson, E. O., 1963b. The social biology of ants. Ann. Rev. Ent., 8:345-368.
Wilson, E. O., and W. H. Bossert, 1963. Chemical communication among animals. In: Recent progress, 19:673-716. Academic Press, N.Y.
Woods, E. F., 1959. Electronic prediction in bees. Nature, 184:842-844.
Wright, R. H., 1964a. Insect control by nontoxic means. Science, 144:487.
Wright, R. H., 1964b. After pesticides—what? Nature, 204:121-125.
Wright, R. H., 1964c. “Metarchon”: a new term for a class of non-toxic pest control agents. Nature, 204:603-604.
Wright, R. H., 1965. Metarchons: insect control through recognition signals. Bull. Atomic Sci., 21(3):28-30.