J.M. FRANZJ.M. FRANZ is a member of the Institut für Biologische Schädlingsbekämpfung, Biologische Bundesanstalt für Land und Forstwirtschaft, Darmstadt, Germany.
INTERNATIONAL TRAFFIC and exchange of goods have increased the hazard of new introductions of pest organisms into forests in various parts of the world. In addition, many newly and hastily created uniform forests bear the risk of mass outbreaks of native or introduced pest organisms. The situation in forest protection is, therefore, not at all satisfactory, in spite of preventive (quarantine) measures and of the considerable array of modern silvicultural or chemical control techniques. In order to counteract effectively the increasing menace to our forests, it will be necessary to review the possibilities for forest pest control offered by biological measures.
In general terms we may say that biological control is the utilization by man of natural enemies for the (regulative) reduction of pest populations. Biological control, like silvicultural control, is applied ecology. A forest, even an artificial plantation, represents a relatively undisturbed and lasting environment in which a complex network of regulating biotic factors can develop. If these are naturally insufficient, alterations such as the colonization of beneficial species will have time to express their full impact. Foresters also favour biological control because of the low costs involved as compared with repeated chemical treatments. The low monetary value of the annual increment of trees necessitates economical control methods. Fortunately, the level of pest damage that can be tolerated by healthy forests is usually rather high. The fact that relatively great numbers of pest organisms can be tolerated makes possible the survival of sufficient natural enemies that depend on these organisms for food supply. Foresters are therefore more interested in economical, possibly self-perpetuating control rather than in reduction to extremely low pest densities or in eradication. Population dynamics of forest pests have been studied intensively because forest ecosystems are long-lasting and offer ideal conditions for such investigations. Results obtained have a direct bearing on applied problems (Waters, 1969).
From the ecological point of view, chemical control of forest pests is frequently regarded as an emergency method. Despite recent advances in the techniques of aerial spraying, drastic disturbances of the natural composition of the fauna are seldom avoidable. This consequence of large-scale application of nonspecific pesticides becomes more important where a great deal of the value of forests shifts from timber production to more general and social benefits. Conservation of natural areas, water reserves, and soil cover and the increasing need for recreational areas become more dominant features of forest policy with the growth of human population. The quality of the environment can be endangered by the application of chemical biocides; therefore, the avoidance of contamination and the reservation of undisturbed areas taken on a value beyond that which can be measured in monetary scales.
In addition to such problems of general welfare, the possibility of the appearance of strains of pest organisms which are resistant to pesticides is a definite threat to any repeated chemical treatment. The recent discovery that the North American spruce budworm (Choristoneura fumiferana) is showing an increase in DDT tolerance (Randall, 1963) clearly demonstrates the dangerous consequences of repeated chemical applications in the same area. Foresters must remember that resistance to insecticides is not limited to agricultural pests. It would be wise not to abuse one of the strongest weapons available against forest pests and to consider other control methods early enough to avoid drastic emergency measures.
The following brief review of techniques useful in biological and integrated control of forests also includes some new possibilities for which operational tests of efficiency are still lacking. For several reasons, the major subject of discussion is the biological control of pest insects. For a more exhaustive treatment of the subject, articles and books are recommended such as: Balch, 1960; Beirne, 1962; DeBach, 1964; FAO, 1966; Franz, 1961a and 1961b, 1964; McGugan & Coppel, 1962; Sweetman, 1958; Turnbull & Chant, 1961; Voûte, 1964.
IMPORTATION OF ENTOMOPHAGOUS ARTHROPODS AND VERTEBRATES
The classic technique of importing from the country of origin natural enemies of previously introduced pests is well known. It is based on the assumption that natural enemies play a vital role in the natural regulation of pests and that introduced insects often become pests because these essential limiting factors are not present in the new environment (Belch, 1960). Although it is not practical to review here the basic theories of the principles of population regulation by natural enemies, it can be stated that some of these biotic agents show a remarkable tendency to act in a density-dependent way, others do not, or not sufficiently. Correspondingly, some of the numerous efforts to establish parasites or predators of pest insects in foreign areas were completely successful and led to economic control; other attempts yielded only partial control, and still others produced no practical results whatsoever.
Unfortunately, there is a great difference between standardized testing of uniform insecticides and introducing (i.e., testing) of entomophagous insects. Almost everything is variable in the second case; the qualities of the insect species involved, such as density relationships, searching capacity, fecundity, and ecological requirements as well as the environmental conditions in the test area are all important. Identical repetitions are never possible.
After reviewing the approaches to introducing entomophagous insects, various aspects of planning pest control will be discussed. Here it will suffice to emphasize the basic and insurmountable difference between the " LD50 approach " to pest control and efforts to manipulate populations. In the latter, only probabilities of success can be accumulated and variables selected. Techniques might be refined and new tricks played, but all are limited by the ever-changing nature of living organisms. As far as the introduction of entomophagous insects into new areas is concerned, several refinements can be added to the basic technique which was demonstrated for the first time as economically feasible by Koebele in California about 80 years ago. This technique is to ship enemies of a pest insect from its native faunal region to the problem area and release them there.
Collections of parasites or predators have traditionally first been made in the country of origin of the pest and have consisted mainly of the more conspicuous species. Recently, this technique has been refined, and the area of search widened so that collections frequently now also include:
(a) less conspicuous and less abundant species, which are difficult to find;
(b) natural enemies from all parts of the distribution area of the hosts;
(c) enemies of related host species;
(d) enemies which have accepted or adapted themselves to a newly imported host pest and are later colonized in the original faunal region of the pest (" adaptation importation ");
(e) intra-areal transfers.
Comments and examples:
(a) The choice of beneficial species -for introduction is an intricate problem and cannot be discussed fully. The usual method has been to include as many parasitic or predacious species as possible as long as they do not have harmful side effects and prefer the target host. By this method, for instance, the parasites of the European spruce sawfly (Diprion hercyniae) were brought from Europe to Canada. Some became numerous and were effective at high densities, before the virus epizootic influenced the gradation, and others appeared later and acted better at low host densities (e.g., Exenterus vellicatus and Palexorista bohemica). The final outcome of this experiment, the complete and permanent suppression of the sawfly by combined action of parasites and the virus disease (Bird and Elgee, 1957) could not have been predicted, as both groups of effective parasites are rare in Europe.
The abundance of entomophagous insects in their natural situation is in itself a poor measurement of their practical efficiency. Low abundance may be caused by competition (Pschorn-Walcher and Zwölfer, 1968). Some species may appear to be rare because they have a short active period, have their centre of distribution elsewhere, or because they live only in well-hidden habitats (e.g., the tips of trees). Night activity has often led to claims of a species being rare.
For instance, the derodontid beetle, Laricobius erichsonii, was formerly considered to be so rare that almost each collected specimen was published in central Europe. Now, after discovery of its short activity period and ecological affiliation with the balsam woolly aphid Adelges piceae (Franz, 1958), it can be collected abundantly and has been shipped to North America, where it has become established. In general, the selection of natural enemies for biological control importations will lose some of its empirical elements when the behaviour and ecological requirements of beneficial agents are better understood. The recent colonization in Canada of two effective parasites of the winter moth Operophtera brumata (Embree, 1965), both introduced from Europe as was their host, simultaneously with the development of population models for the pest, may be a starting point for the full utilization of simulation studies. These may contribute in future to the judicious selection of promising biotic control agents.
(b) The inherent differences among populations of one species, that is, the existence of biologically differing strains of natural enemies, have often been used to select forms best suited to the new ecological niche (Simmonds, 1963). The best approach to a search for beneficial biotic agents is to cover their total distribution area. By shipping samples of parasites or predators to the receiving laboratory in the problem area and by subsequent field tests, the most effective strain may be found. The speed and range of modern transportation, detrimental at is may be for the distribution of pests, facilitates the exchange of beneficial organisms all over the world. New techniques have been found for successfully shipping tiny parasitic wasps by air (Bartlett, 1962). Nowadays the technical problems of transportation are no longer limiting factors for biological control; the difficulties lie elsewhere, in the shortage of qualified, ecologically trained entomologists and in the lack of funds to keep them on a project for a sufficiently long time.
Many pest insects occur naturally over huge areas of the world. The larch sawfly (Pristiphora erichsoni), the gypsy moth (Lymantria dispar) and the European pine sawfly (Neodiprion sertifer) are to be found not only in Europe, but also in the far eastern part of the Palaearctic region. Nowadays, the search continues for natural enemies of the gypsy moth in northern India and of the two sawflies in Japan, where a rich and predominantly new fauna of natural enemies is available and offers more possibilities for additions to the parasite complex already available in the United States and Canada (Commonwealth Institute of Biological Control, Annual Reports).
Another example of increasing the range of the search for natural enemies is the Sirex project in Australia and New Zealand. Although useful work had been done previously and parasites from England established in New Zealand (Hanson, 1939), the exploratory work has been extended considerably since Sirex noctilio has become established in Australia. Investigations of parasites of siricids in conifers are being undertaken in different parts of the northern hemisphere, including Europe, Pakistan, Japan, and North America. An Australian unit has been set up in England (F. Wilson), cooperating for the other areas with the Commonwealth Institute of Biological Control. The work aims at discovering species or races of parasites able to regulate Sirex numbers effectively in the southern hemisphere (Taylor, 1967).
(c) The utilization of parasites and predators from related host species and genera has been practiced for many years. The review by Pimentel (1963) points out the possibility that this procedure has a certain superiority over the traditional approach because no ecological homeostasis has yet evolved and the association between parasite or predator and host is new. At any rate, this additional source of beneficial insects has not yet been fully exploited. Recent projects against the European balsam woolly aphid, now a pest in several parts of North America, extend to the Himalaya mountains and to Japan and Argentina, where only other related adelgids occur (CIBC, 1963). The idea is to find predators with a potential host range that may also include Adelges piceae. So far, it has been possible to establish in North America only predators from the native area of the pest. Their limited success depends probably on the greater susceptibility of host trees (Abies) occurring in North America as compared to the European fir, Abies pectinata (Balch, 1960) and also perhaps on the imperfect density-dependent reaction of the predator complex to host outbreaks (Eichhorn, 1969). On the other hand, some classic cases of biological control using this method have demonstrated that the search for new enemies from allied species is not a matter of blind chance, but can be successful if the necessary ecological data are known. An example of this type of deliberate inclusion of a parasite from an allied host is found in the famous book by Taylor (1937) on the control of a palm leaf mining beetle (Promecotheca reichei) by transfer of parasites from a related beetle of the same genus. Pimentel (1963) lists 27 other examples in the review.
(d) " Adaptation importation " is a new term which designates a special case of the utilization of entomophagous insects with a wide host range. If a native parasite or predator successfully attacks a newly imported pest, chances are that it will do the same after importation into the original faunal region of the pest. Of course, the stage of the host attacked should not already be " occupied " by an effective native enemy. To my knowledge, no example is known of successful application of adaptation importation against forest insects. A project is under way, for instance, on the study of the European pine shoot moth Rhyacionia buoliana (Franz, 1968a), a Palaearctic pest which was unintentionally introduced into other continents. Several native parasites accept the species readily in North America. Three laboratories in Germany and one in Poland are now engaged in an effort to establish permanent rearings and, later on, to colonize two promising hymenopterous parasites (Itoplectis conquisitor and Elachertus thymus) in central Europe, the original faunal region of the pest. Other possibilities and different reactions by parasites are discussed by Zwölfer and Pschorn-Walcher, 1968.
(e) Intra-areal transfer of indigenous natural enemies within the same continental area may help to speed up their natural spread. Although dispersal might occur unaided, this often takes too much time or is hampered by geographical barriers. Some projects of this type are under study in the Mediterranean area. One, the investigation of natural enemies of processionary caterpillars of the genus Thaumatopoea within this region, has as its final aim the general colonization of such agents that are lacking in some areas. More advanced is the transfer of parasites of the larch casebearer (Coleophora laricella) from the eastern to the western United States. Agathis pumila, a braconid imported originally from Europe, as was the pest itself, has been established in Idaho. As it is known that the parasite is most successful when coexisting with the eulophid Chrysocharis laricinellae, the outcome of this transfer experiment will be of great interest.
Another project is the transportation and colonization of predatory forest ants of the Formica rufa group, particularly in Germany, Italy, and Spain. These ants have proven to be aggressive predators - at least at times - of several forest caterpillars and sawflies in Europe. New colonies have been established in forests very susceptible to repeated outbreaks of these pests. The transfer of ant colonies (Formica lugubris) from the Alps to the Apennine mountains in central or southern Italy exemplifies the technical possibilities of such intra-areal shipments (Pavan, 1961). Time will show how effective such newly founded colonies will be, first in establishing aphid colonies as sources for staple food, and secondly in suppressing new outbreaks of pests. Care has to be taken that no harmful aphids will be protected by them. One interesting side benefit of such transfers is already certain: Formica ants stimulate honey-dew production of relatively harmless aphids (Lachnidae) on coniferous trees, thereby enabling beekeepers to harvest, on an average, at least twice as much of this valuable forest-type honey than if ants were absent (Wellenstein, 1960). This is one of the few cases of direct economic side effects of biological control measures.
Earlier experiences show that intra-areal transfers of indigenous parasites or predators of pest insects can be remarkably successful, for instance, in Canada (McLeod, 1962), in the United States Clausen, 1956), and in the U.S.S.R. (Rudnew and Telenga, 1958). Obviously, a large continental area offers such possibilities. These can best be explored and utilized if close cooperation exists between all experts on the same continent. The most recent carefully investigated example of an intra-areal transfer of an insectivorous vertebrate is the successful introduction of the masked shrew (Sorex cinereus cinereus) into Newfoundland (Canada) since 1958 to augment the natural enemies of the larch sawfly (Pristiphora erichsoni) there. Although predation on the sawfly cocoons was less than anticipated, the net gain in natural limitation of the pest is considerable (Buckner, 1966).
Summarizing this discussion of new trends in the importation of entomophagous arthropods and vertebrates, it can be said that this field has remarkably expanded since Koebele, in 1888, shipped the first vedalia beetles.(Rodolia cardinalis) from Australia to California. In spite of some arguments on the possibility of excluding the empirical element in biological control and the danger of multiple importations, all experts agree on the necessity to investigate as many details as possible of the qualities of beneficial organisms which are candidates for importation (Zwölfer, 1967). There is no other way, finally, than trying; and the sequence of introductions of predators or parasites does not seem to have the great influence sometimes anticipated (Huffaker and Kennett, 1969).
MICROBIAL CONTROL
The utilization of disease-causing microorganisms has become an important field of biological control. Although efforts in this direction are about 80 years old, development has depended largely on a better knowledge of such pathogens as viruses, bacteria, protozoa, and fungi. Economically important results have been obtained mainly with viruses and bacteria during the last two decades. Larvae of sawflies and Lepidoptera were the objectives of such operations, but this is not the place to list them. As in the foregoing section on importation, I propose to discuss only some new points of view in microbial control, which will refer partly to new techniques and partly to new concepts. The similarity of application between pathogens and insecticides prompted the name " microbial insecticide." This produced the impression that pathogens are essentially more selective insecticides. However, some of the fundamental differences between these two tools of forest protection have been obscured by this nomenclature. Thus, in addition to techniques, the basic differences between chemicals and biotic agents will be dealt with under the following headings: spread and persistence; vectors; selectivity and safety; host resistance; evaluation of mortality figures.
Insect pathology and microbial control are rapidly developing fields of science; there are new and comprehensive reviews or books available which should be consulted (Burges and Hussey, 1970; Cameron, 1963; Franz, 1961a; Krieg, 1961; van der Laan, 1967; Maramorosch, 1968; Müller-Kögler, 1965; Steinhaus, 1963; Tanada, 1967).
Spread and persistence. The artificial spread of pathogens by dusting and spraying is very similar to the application of insecticides. Ground and aircraft equipment producing sprays, fogs, and dusts have been found useful depending on the special situation. Thorough studies have revealed the possibility of using oil instead of water as a carrier for bacteria and viruses (Smirnoff and Juneau, 1963) or of mixing certain insecticides with pathogen preparations without reducing the effect of either component (McEwen et al., 1960; Jaques, 1963). Care must be taken, however, to select compatible agents. Recently, also the ultra-low-volume spray equipment has been used successfully to apply, for example, virus polyhedra of the Douglas fir tussock moth (Hemerocampa pseudotsugata) in Oregon (Anonymous, 1969).
The artificial spread of pathogens is recommended where natural spread will be insufficient, either because of shortage of time or because the density of the host population is too low to allow satisfactory natural spread. The availability of insect pathogens that lend themselves to mass production for use on a commercial scale has shown remarkable progress. B. popilliae and B. lentimorbus, as well as the fungus Beauveria bassiana, are examples of the interest shown by industry in the preparation of materials of benefit to the forester. Polyhedral and granular viruses, well protected in inclusion bodies, have been used for 30 years in the forest. Production is relatively easy and can be recommended for developing countries. Rearing and infecting of larvae for virus production does not require expensive equipment. Nonpurified suspensions showed particularly high effectiveness in the field against the gypsy moth Lymantria dispar (Magnoler, 1968). If a stock suspension is available, substitute hosts can sometimes be used for pathogen production. Examples for this are: Cytoplasmal polyhedral viruses of Thaumatopoea pityocampa produced in Arctia caja (Sidor, 1965); the microsporidium Telohania pristiphorae of Pristiphora erichsonii, produced in larvae of tent caterpillars Malacosoma (Smirnoff, 1968). The possibility of treating one hectare of forest with the polyhedra from one or two dozen larvae makes this method very economical. Larger operations, however, would need commercially available quantities of virus preparations. Industry is preparing them also for several forest pests such as Neodiprion sertifer, L. dispar and Hemerocampa pseudotsugata in the United States. Specifications have been developed for safety evaluation as well as for the data required on virulence against the target insect, specificity, processing and purification procedures (W.E. Waters, pers. comm.). Production techniques have improved so much that costs are competitive with chemical insecticides (Ignoffo, 1968).
In addition to the traditional mechanical methods of application of pathogens, new methods using biological processes have been successfully tested. Females of a butterfly (Colias eurytheme) were released in a cage after contamination of their genital armature with a paste containing nuclear virus polyhedra. Each time the female deposited an egg, it contaminated the egg shell or leaf surface with the pathogen. Over 60 percent of the larvae became infected (Martignoni and Milstead, 1962). Results of similar experiments on natural virus dissemination by Trichoplusia ni in field cages were less encouraging (Elmore and Howland, 1964). An example of the same type of natural dissemination was a test on the pine sawfly, Neodiprion swainei (Smirnoff, 1962). Cocoons containing larvae given a sublethal infection during an earlier instar were distributed under pine trees. Emerging females laid infected eggs and the subsequent epizootic spread to the natural healthy population. Obviously, this method has advantages because only small amounts of the pathogen and no costly equipment for field application are needed. Also, the dispersal of pathogens is achieved by the insect itself and soon spreads throughout the host population. Particularly where forested areas are widely distributed, the spread of pathogens through the release of contaminated adults as healthy carriers, or the distribution of latently infected individuals might be a very useful method (Knipling, 1960). These possibilities demonstrate the great differences between chemical and microbial control.
Repeated artificial spread is needed where pathogens do not persist naturally. Host and pathogen, both influenced by their environment, determine the effect and persistence of a disease in insect populations. Several instances are known where diseases persist and are self-perpetuating, either unaided by man or after applied biological control. Two possibilities enter into such long-lasting effects: " environmental persistence " and " biological persistence. "
Environmental persistence. The pathogen itself is highly resistant as, for instance, viruses in inclusion bodies or bacteria and microsporidia in spores. This type of persistence against the influences of the environment might be compared to the long-lasting effect of such insecticides as chlorinated hydrocarbons, which remain active in the soil for years. Systematic experiments by Jaques (1967) have shown that nuclear polyhedral viruses (NPV) of the cabbage looper (Trichoplusia ni) remain viable in the soil for several years. The same can be assumed for NPV of forest insects Industry has greatly improved several virus and B. thuringiensis preparations by specific formulations which utilize buffers or coating substances to increase their environmental persistence. It may depend on the specific condition and the stage of infection whether or not a pathogen is able to survive for some time. B. thuringiensis is, for example, decidedly short-lived in agriculture; but it can persist in the carcasses of the Siberian pine moth (Dendrolimus sibiricus) attached to the needles of coniferous trees and thus cause infection of larvae of the subsequent generation (Talalaev, 1958).
Biological persistence. The other type of persistence results from vector activity or from transmission of the pathogen from one generation to the next (vertical transmission). Such transmission occurs frequently with viruses and with microsporidia, more rarely with bacteria. The viruses of some sawflies and the microsporidiosis of the spruce budworm (C. fumiferana) are well-known examples. Biological persistence is possible when the pathogen is able to pass through some stages of the host in an active condition. Such latently infected individuals are still able to mate and to lay eggs. Females (and rarely males) transmit pathogens by such means and cause infections in the offspring or after several generations. This type of transmission, where it occurs, again demonstrates the great differences between living pathogens and chemical insecticides.
Owing to the differences in persistence of pathogens in the environment and in the host population, the duration of microbial control effects varies greatly. The extremes are frequently called short-term and long-term control, but transitions occur. B. thuringiensis usually obtains short-term control. For economic reasons, its application in forestry can on]y be recommended where high-valued forest crops are endangered, if the outbreak does not last too long, and when special circumstances affect the quality of the environment and public health. An example was the control of the green tortrix (Tortrix viridana) in Germany. B. thuringiensis was preferred near the city of Hanau because of its harmlessness to man and animals; it obtained sufficient foliage protection (~80%), avoided the growth of epicormic branches, and ensured good oak-mast (Franz et al., 1967).
Other examples of short-term microbial control in Europe and in the Soviet Union using B. thuringiensis preparations have been recently reviewed (Franz and Krieg, 1967). Satisfactory results were achieved in the field against: Dendrolimus sibiricus, Tortrix viridana Selenophora lunigera, Thaumatopoea pityocampa and T. wilkinsoni, Zeiraphera griseana and Lymantria dispar (partially, only medium efficiency). Lymantria monacha, the nun moth, is not susceptible. The fir budworm, Choristoneura murinana, could not be controlled satisfactorily, which corresponds to the experiences of North American entomologists so far published concerning the spruce budworm Choristoneura fumiferana). Trials with forest insects in North America were also partially positive against the jack-pine budworm Choristoneura pinus (Cameron, pers. comm.) and against the black-headed budworm Acleris variana (Morris, 1969). Other examples are reviewed by Falcon in the chapter " Use of bacteria in microbial control " in Burges and Hussey, 1970.
The examples of medium and long-term microbial control in the forest referring to viruses are too numerous to be listed here. The pest insects, in addition to those mentioned in this review, include the sawflies (Diprionidae) and the Lepidoptera (genera Dendrolimus, Malacosoma, Hyphantria, Lymantria, Thaumatopoea and Kotochalia). Pilot experiments using viruses against the spruce budworm (C. fumiferana) and its European parallel species (C. murinana) were recently successful for the first time in Canada (Cameron, pers. comm.) and in Germany (Schönherr, 1969). Further examples are listed in recent reviews and books as mentioned above.
In the case of alien pests, an effort should be made to find out whether any diseases play a natural limiting role in the country of origin. The persistent effect of the imported virus diseases of the European spruce sawfly (D. hercyniae) and, to a lesser degree, also of the milky disease of the Japanese beetle (Popillia japonica), are encouraging results of partly unintentional, partly planned introductions of pathogens to new areas. A recent example of the rare case that a new pathogen can be imported to a faunal area previously invaded by a pest insect is the colonization of a virus (Rhabdionvirus oryctes) and its subsequent dramatic suppression of the Indian rhinoceros beetle (Oryctes rhinoceros) on palms in Samoa and surrounding islands (Huger, 1966, and pers. comm.). Even if diseases are not effective enough to suppress populations permanently, they may be valuable additions to the complex of mortality factors, as in the case of the gypsy moth (L. dispar) in North America, where regular outbreaks of the " wilt " disease (nuclear polyhedrosis) prevent long-lasting mass outbreaks of the pest.
As with entomophagous insects, diseases are not always restricted to one host. Experience with the American fall web-worm (Hyphantria cunea) in Europe has shown that, after ten years in the new environment, the web-worm harboured. three microsporidian and three virus diseases which are not known in North America and probably do not occur there. It is suspected that these diseases have been acquired from other hosts, probably from lepidopterous larvae living in the same habitat in Europe (Weiser, 1956).
The general conclusion seems to be warranted that any search for new natural enemies in other areas should automatically include a search for pathogens from the target host and related species. This is particularly true if entomophagous insects alone are not able to keep the pest at a sufficiently low level.
Vectors. Consideration of factors aiding the persistence of biological controls is as important for introduced as for indigenous pests. In the transmission of diseases, predators or parasites may play a role as vectors which is easily overlooked. As a matter of course, we judge the value of biotic agents by the direct mortality they produce. Vectors which naturally spread a disease are not easily assessed as to their efficiency. We know that the passage of some pathogens even through the intestinal tracts of such predators as voles, birds, and insects does not inhibit their infectivity. Parasitic Hymenoptera can transmit microorganisms into hosts, either when feeding or during egg deposition. The spread of pathogens from dead carcasses, from excrement, or in several other ways shows how an epizootic may be furthered by vector activity. Although there is little exact information on the total effect of vectors, except for the excellent study on the European spruce sawfly in Canada (Bird and Burk, 1961), this facet of microbial control should not be neglected in the future. We will discuss below the ways in which populations of parasites and predators can be enriched.
A special case of vector activity is that of the nematode Neoaplectana carpocapsae, commonly designated as DD-136 (Dusky, 1959; Poinar, 1967). These worms carry insect-pathogenic bacteria. Apparently, the relationship is symbiotic. A wide range of insects is attacked by the nematode and subsequently by the septicemic infection. Several qualities of this nematode make it ideal for colonization in new areas - its broad temperature tolerance, the easy way in which it can be propagated and stored, its resistance to many insecticides, and the simple means of application using conventional spray equipment. Several forest insects are susceptible, including mining species (Schmiege, 1963; Nash and Fox, 1969). The nematodes enter feeding tunnels and are ingested by the host. On the other hand, their moisture requirements are high and survival depends on the availability of humid places. Some evaporation retarding additives to the watery spray emulsion may prolong the humidity required (Webster and Bronskill, 1968). However, recent appraisals of practical experiments do not show much promise (Niklas, 1967, 1969). As frequent application of nematode suspensions would probably be too expensive in forestry except in special situations, again the need for providing a variety in fauna and flora habitat is emphasized. The possibility of creating by cultural measures centres for nematode colonization in plantations and highly managed forests deserves future study.
Selectivity and safety. Some of the insect pathogens are highly specific and limited to one host species, as are several viruses of sawflies of coniferous trees. Others, like Bacillus popilliae, causing the milky disease of Japanese beetle grubs, or B. thuringiensis, infect a group of related species. B. thuringiensis primarily attacks open-feeding lepidopterous larvae (Angus, 1968; Heimpel, 1967; Heimpel and Angus, 1960; Krieg, 1967). This broader specificity is advantageous when it allows us to combat several pests with one pathogen. At the same time, beneficial insects are not affected and the natural balance is not disturbed as with long-lasting broad-spectrum insecticides. It will be necessary to watch closely the effect of preparations containing Beauveria fungi, which are becoming available in America (Dunn and Mechalas, 1963) and in Europe (Martouret, 1969; Samsinakova, 1964).
The protection of a regulative biocoenosis after the immediate reduction of the pest population remains of primary importance in forestry - much more so than in agriculture, in which the economic threshold is lower and often only the immediate consequences of treatments are considered. Furthermore, all microbial preparations so far used against insects are safe to men and domestic animals. This removes the problem of toxic residues and of unintentional poisoning of vertebrates (including man) by careless handling of such preparations. This again contrasts with most insecticides.
Host resistance. The appearance of pests that are more or less tolerant to insecticides is now understood as the selection of genetic strains better able to survive the action of insecticides. The question arises whether such changes could also result from microbial control (Franz, 1966b). During the interplay of a chemical compound and a pest insect, the insect alone is capable of genetic change; however, in the struggle between pathogen and host, both have the possibility of changing - toward greater pathogenicity or greater tolerance, respectively. In addition, selection by pathogens is probably never as extreme as that by insecticides, because other natural enemies continue to exert pressure. Thus, theoretically, increased tolerance against pathogens should occur more rarely than against insecticides, or not at all. Very few insect pathogens have been studied long enough to permit any general statement. Actually, observations have shown that tolerance against some microorganisms can be increased by continuous laboratory selection. An example is the large white butterfly Pieris brassicae, which lost much of its original susceptibility to a granular virus after continuous laboratory rearing and selection (David and Gardiner, 1960).
In nature, however, where the selection is neither so one-sided nor so enduring, reduced susceptibility of insect populations was never observed to be the permanent effect of high mortality by a pathogen, although short cyclic changes in host tolerance are probable in some cases (Martignoni and Schmid, 1961). None of the repeated field applications of insect pathogens seems to have caused decreased efficiency of the biotic agent, either introduced or native. The most thoroughly studied example is the persistent high effect of the virus of the European spruce sawfly mentioned above. Twenty years after the introduction of the virus to Canada, no change in the mortality rate has been detected and qualities of host as well as of pathogen seem to be unaltered (Bird and Burk, 1961).
Although there is reason for an optimistic outlook, care has to be taken when pathogens like B. thuringiensis are cultured in an artificial medium for an immense number of generations. The counterbalance between pathogen and host is disrupted, and mutants of different pathogenicity may appear. Permanent supervision of such processes is required, to assess both their effect on insects and their harmlessness to warm-blooded animals.
Evaluation of mortality figures. The differences between biological and chemical control methods are abundantly clear following application, when the effect of the treatment is measured. The interval between spraying and mortality is usually longer with pathogens. An extreme case was the artificial spread of a nuclear virosis of the Great Basin tent caterpillar (Malacosoma fragile) in aspen forests of New Mexico (Thompson, 1959). The population of this pest suffered 100% mortality as late as two years after infection. In forests able to withstand some years of damage, the early initiation of even slow-acting epizootics may be recommended if their effect proves to be more persistent than insecticide application. Immediate mortality cannot be used in this case to judge the value of the method.
Not only the speed, but also the degree of mortality is sometimes lower after applications of microbial pathogens than after insecticides. As we have already seen, latently infected individuals are frequently needed to carry over the epizootic from one generation to the next. Therefore, after application of some pathogens (e.g., sawfly viruses) total mortality is not desirable. The survival of a portion of the population ensures transmission of the disease to the subsequent generation. The level of the pest insect that can be tolerated depends, of course, on the economic threshold. It is determined by several biological and economic factors. Here again, the principles of application and evaluation of mortality differ greatly from the conventional approach in insecticidal control. Therefore, any comparison between insecticides and microbial control agents may not express the full truth if it is only concerned with such characteristics as the slope of the probit line in a dosage-mortality curve or immediate mortality. The applicability of the usual correcting formulae after Abbott or (identical) Schwerdtfeger is limited in microbial control. With slow-acting pathogens the assessment of total leaf losses in treated and untreated plots may yield results of greatest significance (Franz, 1968).
More insight into mechanism of regulation of natural populations helps to achieve the same economic result with less effort and expense. Other examples could be quoted to prove that the mere numerical comparison of mortality figures does not disclose all the consequences of either chemical or microbial control.
An unexpected possible consequence of microbial control concerns the sex ratio. The relative abundance of sexes is basic to a population's reproductive potential. For example, female pine sawflies (Neodiprion sertifer and N. pratti) have one more larval instar than males. Hence, they are exposed to infection longer than males when viruses are applied artificially and males reach the relatively unsusceptible prepupal stage earlier. Therefore, the virus application causes not only heavy mortality of the sawflies, but also a greater loss of female than male larvae (Bird, 1961; McIntyre and Dutky, 1961). The selective elimination of females resulting in a shift of the sex ratio is again a special characteristic of pathogens.
In conclusion, we may say that microbial pathogens offer several new possibilities for forest-insect control. Most needed are more research, more chances to test new preparations without the obligation to succeed in the first trial, and a more ecological approach to pest control problems.
GENETIC MANIPULATION
Natural populations exist in a state of continuous change. Not only numbers, sex ratios, age classes, and other phenotypic changes, but also the genetic composition is modified by each individual that drops out or is born into the group. Artificial manipulation of populations has been tried in two directions: (a) to improve them; or (b) to make them less fit for survival.
Selection of better strains. The breeding of varieties better suited to a certain purpose has long been exercised with domestic animals, including silkworms and honeybees. The problem of improving the efficiency of predators or parasites by selection has been discussed in the literature (DeBach, 1958; Franz, 1961a; Simmonds, 1963). Four steps have been envisaged for this procedure:
1. determination of characters that need improvement;
2. provision of adequate genetic variability;
3. satisfactory selection procedures;
4. maintenance of the integrity of the new strain in the field.
As mentioned above, the accumulation of as much genetic variability as possible is the modern basis for the importation of beneficial insects. From this pool, natural selection will preserve the strains that are superior under the prevailing environmental conditions. A good start, however, is necessary and the numbers released should be able to compete with the existing population (Wilson, 1965).
The selection in the laboratory of certain improved qualities for field populations has not yet been successful, probably because any selection in confinement is necessarily one-sided. This bias may be acceptable in breeding improved domestic animals or for beneficial insects mass-reared for one single use only in the " inundation method." It is often overlooked that continued mass production in the laboratory leads to selection of strains well adapted to this procedure. The necessity of mass rearing under more natural conditions has therefore been emphasized where the stock is intended for field release (Stein and Franz, 1960).
The case is well known of the imported parasite (Mesoleius tenthredinis) of the larch sawfly (Pristiphora erichsonii) that lost its former efficiency in central North America because most of the host larvae developed the ability to encapsulate the parasite embryo (Muldrew, 1953). Since all imported sawfly parasites were formerly collected in England, where larch forests are not indigenous, the reason for this local failure may be the low variability of the original parasite stock. Introductions of other biotypes from other parts of the distribution area of the parasite (Bavaria; Muldrew, 1964) have led to establishment, and encapsulation begins to decrease. In addition, the ichneumonid Olesicampe benefactor has been introduced from Europe to Manitoba; it reached over 90% parasitization on the release spot and has been successfully transferred to other (Canadian provinces (Simmonds, 1969). The rarity of such cases of defense mechanisms developing in the host against the parasite demonstrates that usually host and effective parasite exist in a sort of balanced relationship, probably advantageous for the continuing existence of both (Doutt, 1960). If this balance has been disrupted, the biological control method is able, by genetic manipulation and/or introduction of new parasites (as the example shows), to reestablish the efficiency of the parasite complex.
In future, the selection of strains more tolerant to pesticides might be profitable for release in greenhouses, in isolated areas, and for beneficial species like predatory mites which have a low power of dispersion. In many situations the maintenance of the integrity of new stocks will be endangered and the desired qualities lost unless selection pressure for this new quality is usually low and other, nonselected individuals can invade the release area.
Beneficial biotic agents having, like pathogens, low powers of dispersion are particularly suitable for selective breeding. Promising successes have been reported of nuclear polyhedrosis virus strains selected as being more virulent against the Swaine jack-pine sawfly (Smirnoff, 1961). Similar results were obtained in the field against the gypsy moth (Orlovskaja, 1962) and in the laboratory with the wax moth (Galleria mellonella) using viruses (Veber, 1964). Selection was accomplished by experimentally transmitting viruses from those larvae that died first and showed the clearest symptoms. This same heterogeneity of virus material allowed the adaptation of a virus of the poplar sawfly (Trichiocampus viminalis) to a closely related willow sawfly (T. irregularis) in three passages (Smirnoff, 1963). The observation that virus stock from larvae of the wattle bagworm (Kotochalia junodi) in Africa which had been collected in a distant outbreak zone acted more strongly upon a test population (Ossowski, 1960) proves again that enough variability of viruses (and probably also of other pathogens) occurs to warrant a more systematic study of the problem. The hope for more instances of increased virulence and host range seems well founded; this may enlarge the future field for microbial control.
Autocidal control. The control of an insect through its own actions requires production of genetically inferior strains. It is, therefore, the antithesis of selective breeding for superior strains as mentioned above. The method has gained much publicity after the successful eradication of the screw-worms (Cochliomyia hominivorax) from Curaçao and the southeastern United States.
Following are several ways of utilizing the autocidal principle of insect control which have been proposed for exploration and further research (Knipling, 1960):
1. Release of males that have been made sexually sterile by gamma radiation or other physical means.
This is the only method already tested successfully in a large-scale experiment. Probably the first example using a forest pest is the eradication of the cockchafer (Melolontha vulgaris) from a limited area in Switzerland (Horber, 1963). Many male beetles were collected from the field, and after sterilization by X-rays they were released in an isolated valley. The local population of cockchafers disappeared after the second flight and release period.
Horber (1969) shows conclusively that, under special conditions, use of the sterile-male technique could be the ideal complement to conventional control methods where long-term regulation is desired.
The gypsy moth (L. dispar) is the second forest pest which was used in the field to test autocidal control in the United States. As mass production remained difficult owing to the interference of diseases, at present only limited numbers of adults for release are envisaged to test the applicability of the method to prevent spread of a marginal population (Knipling, 1969). For further exploratory studies on forest insects see Lynn, 1967.
2. Release of males sterilized by chemical means or the induction of sterility in populations by chemicals.
Recent intensive research, particularly in the United States, has shown that chemosterilants can be developed to sterilize males or both sexes without reducing their general vigour or sexual competitiveness (Smith et al., 1964). The great advantage of this method lies in the possibility of sterilizing populations without rearing and mass release. Some of these compounds produce sterility simply by contact; others must be ingested. Chemosterilants that are too toxic for general field use may find practical applications when used in combination with attractants.
3. Development or detection and release of strains of insects that possess genetically inferior or lethal factors or that are incompatible with the local strain.
For the gypsy moth, Downes (1959) suggested the importation to the United States of males from the Asiatic strain which, when crossed with the European strain now present in the United States, produces sterile female intersexes. In addition to genetic sterilization, selection and genetic manipulation could be used to produce inferior strains of destructive insects. Such individuals should be released during favourable periods in order to dominate the natural population. Inferior qualities transferred into a population in sufficient frequency would lead to its destruction because the population would become more susceptible to environmental factors.
The following types of genetic deficiencies have been discussed: inability to diapause or to fly; absence of glands that produce glue for the fixation of eggs to the food plant; deformed mouth parts of larvae; and several others. One dominant, or at least three independent recessive, lethal characteristics might be as effective as induced sterility with monogamous insects and probably more effective with polygamous species. The boll weevil (Anthonomus grandis) has been used as a basis to calculate the feasibility of insect control through release of specimens with heritable lethal factors (LaChance and Knipling, 1962; Klassen et al., 1970).
All these methods of autocidal control require individuals that retain their sexual vigour and remain competitive with normal individuals. This is difficult to obtain by radiation or chemicals. Recent research at the United States Department of Agriculture laboratories has shown that substerilizing doses can also lead to a level of sterility in surviving F1 progeny that is higher than the sterility level in the parent generation. This is apparently caused by the induction of large numbers of reciprocal chromosome translocations. The unique chromosomal structural feature of Lepidoptera - the occurrence of diffuse centromers - makes this group of insects particularly susceptible to this very promising new technique for control (Knipling, 1969; North and Holt, 1968,).
All investigators emphasize that some of these considerations are still theoretical. The necessary data on population dynamics, dispersal, and particularly on the rate of increase are frequently lacking. However, a series of pilot experiments mainly with Lepidoptera and Diptera have shown the general principle to be sound and have stimulated research in fields hitherto neglected.
Such experiments and theoretical considerations have demonstrated one great advantage of autocidal methods as compared with conventional control methods. When insects are appropriately genetically altered or sterilized, by whatever means, population control is obtained in two ways: the insect that is sterilized cannot reproduce; thus we obtain the same effect as by killing. In addition, these sterile individuals remain in the population and compete with the normal ones during mating. This produces an additional effect comparable with that achieved by releasing sterile individuals. This "bonus effect" represents the chief advantage of the autocidal method over other control methods. It produces a quick effect that leads to at least local eradication much faster than other methods. Because sterile or otherwise deficient insects help to reduce their own population level, less effort and expense is required to achieve the desired low level of the pest.
In forests, two types of pest infestation lend themselves most clearly to autocidal control methods: the appearance of a new introduced pest; and the control of pests in isolated areas such as islands, valleys, or in new plantations in nonforested areas. Obviously, isolated populations are advantageous, although not an absolute prerequisite, for autocidal control methods. Calculations have demonstrated that in some cases the repeated release of sterile males or the repeated induction of deficiencies, including sterility, may be more economical than the repeated application of insecticides (Knipling, 1960, 1969). A period of low population level is necessary for beginning the autocidal treatments; this may occur naturally or be obtained by other control techniques.
In conclusion, we may consider the autocidal control method as one of the great new achievements in biological control. Its applicability is being considered in all fields of pest control. Autocidal control is part of biological control because living organisms are utilized. The self-destruction of a population may end with suppression or with eradication, whereas all other biological control methods at best create a new and low level of the pest that must continue if the regulation is to persist.
CONSERVATION
There are three aspects of conservation (Beirne, 1963a): (a) protection of the natural enemies from silvicultural practices that destroy them; (b) encouragement of the natural enemies by not removing requisites needed for survival and increase; (c) augmentation of the natural enemies by deliberately providing these essential requisites.
Examples of these phases of conservation can best be understood by defining the aims of modern silviculture. The ideal is no longer simply the maximum yield of forest products in the minimum time; the concept of the sustained yield has gained more and more acceptance in forest management. If we are to maintain over the years adequate sources of forest products, some of our current procedures must be modified. This is particularly true for newly planted forests of introduced tree species, as in South America and New Zealand (Rühm, 1964).
The pest-control aspect of this approach is to realize that, for instance, uniform and even-aged forests frequently are more susceptible to pest outbreaks than others having a more diversified flora and a range of age classes. The thorough study of Morris and co-workers (1963) in Canada demonstrated this, and also why monocultures of balsam fir provide optimum conditions for spruce budworm outbreaks. Promoting the mixed forest type is part of silvicultural control and, simultaneously, of biological control, as can be seen in the following examples. Several important pest insects have parasites which need alternate hosts. These frequently live on other plants. Radical weed control automatically reduces the numbers of beneficial species. Adults of many beneficial insects require food from wild flowers or other plant sources. Without hollow trees, very few titmice (Paridae), important insectivorous birds, are able to nest. Many predacious insects must have alternative prey on which they may survive temporary scarcities of the pest. More vectors of disease-causing microorganisms will be available in a diversified fauna. To protect such requisites, to avoid unnecessary destruction of these essential facets of the habitat is the conservation phase of biological control.
Another facet is the deliberate augmentation of such requisites. Examples are: growing of plants required by adult parasites as food sources necessary for reproduction (Beirne, 1963b; Scepetilnikova, 1963); providing the hiding places (shrubs) or nesting boxes required by insectivorous birds (Bruns, 1960); planting tree or shrub species that produce additional winter food for temporarily insectivorous birds such as titmice to survive during a critical period (Gibb, 1960); protecting predatory ants of the Formica rufa group against commercial overexploitation by man (Gösswald, 1958).
Some birds, mammals and social insects have shown particularly good qualities for the suppression of local outbreaks because of their strong behavioural response to increasing host populations (Buckner, 1966; Hassell, 1966).
Obviously, such "planned augmentation" of essential requisites of beneficial organisms needs a thorough understanding of the key factors regulating the abundance of the pest and its enemies in an area. Because such detailed knowledge is frequently lacking, the following general principle - one that is well substantiated by observation and experiment (Pimentel, 1961) - may guide us. In pest control, we are dealing with an interacting complex, the ecosystem; in this complex, one part cannot be changed without altering other parts; species diversity and complexity of the biocoenosis play an important role in preventing pest outbreaks; therefore, wherever possible in the cultivation of forests, faunal and floral variety should be encouraged and protected.
The second part of this review, including the bibliography, will appear in Unasylva Volume 25 (1) Number 100.
