The 1981 Autumn Symposium (held at the Zoological Society, London, on 31 October) was a new departure for the British Society for Parasitology in that Cambridge University Press have taken over publication of the proceedings. The style and layout of the contributions have been changed in order to conform with the format currently used in the journal Parasitology. The symposia proceedings will now be published as separate issues of Parasitology but with their own distinctive colour and cover design. The proceedings will be available by subscription to the journal or separately. Furthermore, the meeting organizers will in future be responsible for editing the symposia: assistance will be by Drs Taylor and Muller during a transition period.

The population dynamics of insect–pathogen interactions are examined with the aid of simple mathematical models. Three concepts of central importance to the interpretation of population behaviour are discussed, namely the ability of the pathogen to persist within its host population, the ability to regulate and depress host population abundance, and the ability to induce non-seasonal cyclic changes in host density. The selection of pathogen species or strains to depress pest population growth is discussed and the optimal characteristics are shown to be intermediate pathogencity combined with an ability to reduce infected host reproduction, high transmission efficiency, including elements of vertical as well as horizontal transmission stages. When the pathogen plays a significant role in the regulation of host population growth, it is argued that many insect–pathogen interactions will exhibit non-seasonal oscilations in host and pathogen abundance. Mathematical models are used to explore the patterns of population behaviour that result from the continual introduction of a pathogen into a target pest population. It is shown that there exists a critical introdution rate, above which the eradication of the pest is theoretically possible. Significant reductions in pest population abundance will not occur until the introduction rate approaches this critical value, whereupon the oscillatory behaviour of the interaction between host and pathogen population will be suppressed.

A general dicussion is given of the problems arising from the combined use of chemical agents and pathogens for the control of pest species, and the evolutionary pressures acting on host and pathogen populations.

Virus diseases have been reported from more than 800 species of insects and mites. Isolates of the baculovirus and cytoplasmic polyhedrosis virus groups have biological properties which should lead to their successful use as microbial control agents in integrated pest management programmes. These viruses infect the larval stages of many lepidopterous and hymenopterous pests, producing a chronic or lethal infection and the release of large quantities of relatively stable infective inclusion bodies (IBs). The IBs serve as the means by which the viruses are transmitted and persist outside the host. Younger larvae are more susceptible to infection than older stages, and this difference influences the timing of application and doses of virus needed for practical pest control. The high degree of host specificity of many isolates reduces their potential ecological hazard but also limits their use, particularly on crops where a complex of pests is established. Environmental persistence is also a limiting factor as virus is rapidly inactivated by ultra-violet light even when contained within IBs. The viruses persist for longer periods when transmitted within the host population, a feature of virus infections restricted to the insect gut.

The practical use of insect viruses in horticulture and agriculture does not utilize their full epizootic potential, but takes advantage of their high pathogenicity and specificity. The baculoviruses of codling moth, and Heliothis spp. provide satisfactory pest control, but for their most cost-effective use it is important to determine the minimum dosage rates of virus required. It is encouraging that studies of the virus control of Pieris spp. have suggested that control achieved by the insecticidal use of a virus can be closely predicted from information on dosage-mortality responses, larval feeding rates and virus persistence. The stability of forest and grassland, and their high economic thresholds makes them ideal candidates for longer-term control. Viruses of the coconut rhinoceros beetle and european spruce sawfly provide examples of classical biological control where the viruses persist for long periods, are efficiently transmitted and act as natural regulators of their hosts. Virus control of pasture, and some forest, pests may be possible by manipulating enzootic viruses without the need for direct control measures. More frequently insecticidal applications are needed, providing control of limited duration which requires periodic ‘topping-up’.

Few viruses are commercially-available; their selectivity and often small potential market, may limit industrial interest. However, improvements in virus production, formulation and a better understanding of virus epizootiology should lead to an increasing role for this group of insect pathogens in biological control.

All bacteria in microbial insecticides are species of Bacillus and form spores since they have to survive in the environment and on the shelf. They can be formulated as wettable powders, suspensions and dusts for application with conventional pest control machinery. All are safe to man and virtually all non-target organisms. Development costs are relatively low, but host specificity greatly restricts markets, the largest being ca. 2000 tons per annum in the West for B. thuringiensis. All act only after ingestion, a disadvantage because there is no contact action and usually only larvae are attacked. Three main groups have special features that determine their commercial success.

The B. popilliae group is produced only in vivo which limits production by three small firms. The Japanese beetle has been controlled in grassland in the warm parts of the USA by single applications of spores in heaps, spaced 2 m each way. The bacterium spreads slowly to untreated areas, is very persistent and kills only by infection.

The B. thuringiensis group kills larvae of Lepidoptera, mosquitoes and blackflies, mainly by gut poisoning with a protein crystal toxin. It rapidly paralyses mouthparts and gut, stopping crop damage. It is readily produced by deep liquid fermentation, but does not persist and needs repeated application during the pest season. Products containing no beta exotoxin can be applied at unlimited dosage to food crops up to harvest. Only one application is needed for stored grain. After 20 years' use of strains against Lepidoptera, a different strain is now used commercially against mosquitoes and blackflies (only 5 years after its discovery), although improvements in formulation for aquatic application are needed. A recent new product based on the beta exotoxin is used in Finland only against flies in pig houses because it has some vertebrate toxicity.

The B. sphaericus group is similar to B. thuringiensis, except that its proteinaceous toxin is different, is situated in the spore wall in strain 1593, and attacks only mosquitoes. Now at the pilot production stage, its commercial future depends on whether it is more potent than B. thuringiensis against certain species and whether it can recycle to give effective extended mosquito control in some environments.

Intensive selection from natural isolates has improved potency 100 to 600 fold. This selective effort must be maintained and improved by genetic manipulation, which can be used to develop greater potential, particularly since DNA coding for the crystal toxin is carried on plasmids. This also gives speculative hope that the toxin may be incorporated into natural aquatic bacteria for mosquito control and into plants for protection against lepidopterous larvae. A great advantage is that these bacteria do not harm beneficial fauna to cause pest resurgence. At present, the main use lies in integrated pest control systems, although bacteria are not likely to supplant chemical insecticides on a large scale in the near future.

Of the unicellular eukaryotes, formerly Protozoa, now considered to belong to five separate phyla, only the neogregarines and microsporidia are serious contenders for a role in biological control of invertebrate pests. Ciliates of the genus Lambornella, which penetrate their hosts via cuticular cysts, have potential in mosquito control but have not been investigated in depth.

‘Protozoa’ generally kill their hosts by overwhelming numbers, destroying the normal function of organs or depleting the host of essential reserves. Because they are slow-acting pathogens they cannot be used on their own when pests have already reached a high level of abundance nor can they be relied upon when the damage threshold of pests is low. Their principal use will be as the slow-acting component of a 2-pathogen or pathogen-plus-chemical formulation, which can be used when a degree of damage is tolerable.

A comparison is made between two microsporidia in lepidopteran hosts, Vairimorpha necatrix and Nosema pyrausta. The former causes high mortality in a wide range of hosts, when bacterial septicaemia ensues after disruption of the gut wall by the microsporidian invasion process. Some larvae may survive this period and live to damage crops, but none survives to adulthood. There is no transovarial transmission and the parasite is rarely found in natural populations. V. necatrix could be used as a microbial pesticide for short-term control. N. pyrausta is restricted to a single host, the European corn borer. It has low pathogenicity, causing some larval mortality especially under conditions of environmental stress. Most hosts survive to adults but show reduced longevity and fecundity. The parasite is transmitted transovarially and is highly prevalent in the field. It is not considered pathogenic enough to be used as a microbial pesticide but is an important factor in regulating natural populations. These examples illustrate the inverse relationship between pathogenicity and prevalence and show how cycles of host population abundance may be driven by pathogens of moderate to low pathogenicity.

Two kinds of transovarial transmission mechanisms are discussed. With the microsporidia of winter moth, vegetative stages and spores, even when abundant in egg yolk, do not gain access to larval tissues but are confined to the meconium in larvae just before eclosion. Larvae are not infected when they hatch but the spores are carried over in the eggs to the next period of larval feeding activity. In contrast, some genera of microsporidia in haematophagous diptera, e.g. Amblysopora in mosquitoes, actually infect the cells of developing larvae, which are already infected when they hatch.

The prospects for biological control with ‘Protozoa’ are reviewed for vectors of medical importance and for pests of pasture, field crops, forests and stored products. Particular attention is given to the use of microsporidia in combination with low concentrations of compatible chemical insecticides and with other pathogens (e.g. viruses).

Spores for field application can be produced in natural or experimental hosts by feeding or intrahaemocoelic inoculation. Yields vary according to the species of parasite and host. Examples are Nosema locustae in Melanoplus bivittatus yielding 3·9 × 109 spores/grasshopper enough to treat more than 1 hectare of rangeland, and Vairimorpha necatrix in Heliothis zea, yielding 1·67 × 109 spores per larva, with 2·5 × 1012 spores/hectare required for field application. In vitro culture is at present a laboratory tool only, with yields too low for economic returns.

Spores can be stored, according to species, dry or in distilled water with antibiotics at 4 °C. This gives good survival for months or years. In field applications feeding-bait formulations are more efficacious than sprays because they concentrate the spores for uptake by the target species and give the spores some protection from harmful ultraviolet radiation.

Pheromone lures have been used for the introduction of spores by males into pest infestations in stored grain. Males are lured to sites dusted with spores and return to the grain after removal of the lure, to contaminate females and larvae. The use of these lures, first as traps to monitor pest population increases, then to effect a controlled pest growth curve by introduction of pathogens, is an attractive innovation. Protozoa are considered safe for field application on the limited evidence available.

There are two ways in which trematodes can be used in the control of other trematodes which transmit diseases of medical or veterinary importance. Either they can interfere with the reproductive capacity of the host species of snails, or they can exert an antagonistic effect against the larval stages of the target trematode species inside the snail. Often the two effects act together.

There are six essential criteria for the selection of a suitable species of snail for control: (1) complete and permanent sterilizing effect; (2) clear dominance over target trematodes; (3) strong infectivity to molluscs of all ages; (4) high egg productivity in the definitive host; (5) lack of pathogenicity towards man and domestic animals; (6) ease and low cost of maintenance of the life-cycle.

Various field trials are discussed and it is considered that trematode species producing rediae are usually dominant over those producing only sporocysts (such as the schistosomes). Most species used in control trials to date have belonged to the family Echinostomatidae as they have rediae, and eggs that can be produced in large numbers in a laboratory host such as the rat.

The mathematical basis and feasibility of control schemes are discussed and it is concluded that much more information is necessary before their potential can be evaluated.

This review highlights published research on the principal entomogenous nematodes that have potential as biological control agents of insects. The life-cycles and status of promising members of the families Allantonematidae, Diplogasteridae, Heterorhabditidae, Mermithidae, Neotylenchidae, Rhabditidae, Sphaerulariidae, Steinernematidae and Tetradonematidae are discussed. Emphasis is placed on attempts to control insect populations with these nematodes. Mass propagation of the Steinernematidae and Mermithidae are also discussed, including current in vivo and in vitro rearing systems. A number of these nematodes show promise as biological control agents but commercial development has been slow because of environmental limitations, host specificity and inherent problems associated with manipulation of living organisms. Although some of these nematodes have been considered for commercial preparation, it appears that they will remain ‘on the shelf’ until the need for such biological control provides small businesses with the incentive to make them available for general use.

There are many species of fungi attacking terrestrial and aquatic insects of agricultural and medical importance. Of these, few species have received much attention with a view to their use as biological control agents. The mechanisms of infection have been comparatively well studied, although many questions remain to be answered. The knowledge so far obtained has shed some light on the reasons for host specificity. Temperature, above all relative humidity, and their interactions are the most important physical factors influencing infection of terrestrial insects by entomopathogenic fungi and these are discussed in relation to epizootiological events. In aquatic environments, temperature, salinity and organic pollution are the important factors. In addition to these physical factors, numerous biotic factors at the level of the insect and the fungal pathogen influence both infection and spread of disease in insect populations. The complexity of the interactions of the biotic and abiotic factors makes it extremely difficult to study the influence of any one of these. Virulence of fungal pathogens and its measurement are discussed, together with the shortcomings of present bioassay systems; virtually no laboratory bioassay system exists which has been designed to yield data meaningful in the field. Mass-production techniques are described as are their inherent problems and those of formulation and storage. Finally, the achievements of fungal control of insects in the field are reviewed. Thus far, several species are mass-produced and are in widespread use, two of which, Verticillium lecanii and Hirsutella thompsonii, have been commercialized. More studies, ecological, fundamental and developmental are required in this field to realize fully the potential of other candidate fungi.

This review begins with a description of the parasitoid life-style and the ecological and evolutionary factors which generate the remarkable diversity of insect parasitoids. We then describe the various ways that parasitoids have been used in the biological control of insect pests, and survey their success to date. The use of parasitoids remains largely an art, aided by past experience of success and failure. A more fundamental approach, involving basic research and theory, has not as yet contributed significantly to practical biological control. We explore the potential for such a science of parasitoid use and review basic research on parasitoid ecology and evolution which is of particular relevance to biological control. Mathematical models are used to identify and examine those parasitoid and host attributes which lead to successful biological control. Factors such as parasitoid foraging behaviour, fecundity, larval survival and sex ratio are shown to be important in influencing the depression of host populations and/or the stability of host–parasitoid interactions after depression. Multiple release is discussed and a model for inundative release of parasitoids is explored.

A comprehensive account is given of the evolution of registration guidelines and safety testing procedures for microbial insecticidal agents. Particular emphasis is given to the use of viruses for pest control and the various guidelines developed to assess their possible hazards. The likely gains and risks associated with using viruses are discussed. Several meetings have been held in the last 10 years to assess the hazards of virus insecticides. Some of these meetings have produced recommendations, some have developed guidelines for safety testing. These meetings are reviewed. The various guidelines developed for safety testing are critically evaluated and the UK Registration Criteria for Biological Agents used as Pesticides are reproduced in full. Examples of viruses that have been safety tested and registered for use are given and the criteria used are described. Conclusions are drawn on the merits of the guidelines presently available and the likely future development of safety testing schemes is considered.