Pesticides Research, 91 – Biological Control of Weevils (Strophosoma Melangrammum and S. capitatum) in Greenery Plantations in Denmark

Biological Control of Weevils (Strophosoma melanogrammum and S. capitatum) in Greenery Plantations in Denmark

7 Discussion and Conclusions

7.1 Population studies
7.2 Molecular characterization
7.3 Bio-assays and strain selection
7.4 Field releases and biocontrol experiments
7.5 Non-target effects
7.6 Conclusions

In the preceding chapters a discussion on the specific obtained results were included and will not be repeated here. Below is a brief discussion on the more general, conceptual elements of the methods in the studies performed.

7.1 Population studies

Population ecology is a core element in the discipline ecology, and insect population ecology has over the past been subjected to many studies, and results have been compiled in classical books like Price (1975). The classical methods for population studies include method for absolute and relative population estimates (Southwood, 1980). In this book, which has been published in many editions, standard devices for such studies are described. Our emergence traps and funnel traps are a well-know device for absolute population estimates and our design for the traps were based on the recommendations in Southwood (1980). Methods for absolute population estimates are to be preferred over relative estimates, whenever possible, since the results from the former can be calculated to an estimate on number of individuals per surface square unit. The present project managed to obtain such data, which are very valuable to understand the phenology and development of Strophosoma species.

The knowledge and understanding of pathogen-insect interaction is generally not as well elucidated as the interactions between insect populations and predators and parasitoids. Examples of the latter interactions have to a much greater extend been adopted in general textbooks on ecology (for example Begon et al. (1995). The methods employed for the assessment of fungus prevalence in our studied insect populations were based on the conceptual framework defined by Fuxa and Tanada (1987). In their book, the interaction between insect pathogens and insect populations are seen in an epidemiological context, defining the terms `prevalence' and `incidence' for insect pathology. The present project thus gave data on the prevalence on fungal diseases in the insect populations in a correct context.

7.2 Molecular characterization

The use of DNA-based methods for characterization and identification of a wide range of organisms is a rapidly developing area, and every year, new methods or modification of existing methods become available. The importance of genetic studies in biological control is outlined in Ehler et al. (2004), although none of the cases in their book are from insect-pathogen interaction studies. We used molecular methods in two ways: 1) in order to validate the phenological data on the two included Strophosoma species, 2) in order to evaluate the effects and the persistence of a released fungus.

With respect to 1), this project managed to validate the phenological data, thus allowing for the first time a more full understanding of the life cycle of S. melanogrammum and S. capitatum. It can be expected in the future that DNA-based methods will be used to a much greater extent in studies of species complexes in insects, and this study is a valuable contribution to the studies in the family Curculionidae.

With respect to 2), the fungus M. anisopliae was not present at detectable level in the A. procera stands used in this study before the application. To some extent, the site could therefore be regarded as an `agent free' environment. This gave the advantage that our characterization did not aim to discriminate between a range of natural genotypes and the applied genotypes, but solely aimed to verify that the applied genotype was the causal agent. We managed to do this, and especially as concerns non-target effects, it is desirable that such verification is possible.

7.3 Bio-assays and strain selection

Our approach to perform bio-assays with insect pathogenic fungi was based on the classical concept (Lacey, 1997a): first, the pathogenicity of the fungus isolates was proven. Then, quantitative bio-assays were performed in order to discriminate between the levels of virulence. Virulence is `the disease producing power of a microorganism' (Lacey, 1997a), and is in most cases calculated from data on either dose-response relationships or experiments providing data on the lethal time or survival time. We used both approaches and obtained a set of data sufficient to select the best isolate on these premises. It should, however, be stressed that the most virulent isolate in laboratory studies is not necessarily the isolate with the best ecological fitness.

The next step is then either to go straight to the field or to add semi-field experiments to obtain information about ecological fitness. In the present project it was decided to go directly from laboratory experiments to field experiments, although additional laboratory experiments were carried out after the initiation of the field experiments. In practice, it is almost never possible to `wait' for all needed information before the field experiments to test the ecological fitness. In this case, a perennial cropping system and pest insects with an expected life cycle of more than one year made it indispensable to postpone field experiments. The selected and later applied strain of M. anisopliae, BIPESCO 5, was known for its high ecological fitness in other cropping systems, so it was expected that this strain would be ecologically fit also in our system. The outcome of the field experiments confirmed this hypothesis.

7.4 Field releases and biocontrol experiments

The studies on efficacy after field releases were based on the expectation of an inundative effect. Inundation biological control is performed when large amounts of an agent are released with the expectation that control will be achieved with the released organisms themselves (Eilenberg et al., 2001). Inundation is thus complementary to inoculation biological control. In the case of inoculation, low amounts of an agent are released with the expectation that the released organisms will multiply. In many cases, however, the final outcome of an inundation biocontrol seems to be an additive result of the `real' inundation and some inoculation effects so these two strategies are often together listed as augmentation (Hajek, 2004). In the field experiments performed under the present project were done by inundation.

Further studies may show if there in addition to the inundation effects are also inoculative effects, for example by production of conidia from infected hosts, which result in infection of uninfected hosts.

Hand-operated sprayer for application of insect pathogenic fungi has been used in many earlier described studies aiming to test field efficacy of insect pathogens (Lacey and Kaya, 2000), in relatively small areas of application. The application of an aqueous solution of conidia by one person by a hand-operated sprayer (Bateman et al., 2000) allows a relatively precise dosage. The environment at the field sites used in the resent studies was an uneven ground surface, with stems and decaying twigs in clusters (major clusters of twigs were, however, removed before application). Much care was therefore needed to ensure that spraying was exact in dosage and distribution. The data on density of the M. anisopliae from sprayed plots indicates that we managed to do this. Obviously, large-scale experiments and routine usage require a further development of the application technique.

The assessment of efficacy made benefit of the data from the population studies. As a basic rule, efficacy by an applied insect pathogen should be based on both 1) prevalence assessments, 2) population estimates of targets, 3) crop yield and/or quality. In the present project, data on 1 and 2 were obtained, while the project did not aim to assess 3). Additional studies are needed to obtain such data.

7.5 Non-target effects

Nowadays, studies on non-target effect are regarded as an integral element in biocontrol studies, both due to the demands from authorities before a registration in EU is possible (Anonym, 2003a) and due to the intention of developing biological control as environmentally sound plant protection.

It is essential to discriminate between the ecological and physiological host range (see chapter 6 on non-target effects). The present study aimed to pay full attention to the ecological host range. The most used method for ecological non-target studies is based on the collection of selected non-target species followed by incubation in the laboratory and diagnosis of possible pathogens in case of death of the incubated non-target (Hokkanen and Hajek, 2003). Results obtained in this project documented that the applied fungus was able to cause infections in non-target arthropods. The ecological host range thus gave valuable data on the expected events in a field situation. However, even though infection was seen in the sampled non-target arthropods the effect of the population development of these non-targets is not known, since assessment of population effects would require more laborious studies than the resources available in this study. A general drawback of field studies is, however, that the sampling of non-targets depends on the practical circumstances: only if a sufficient number of a specific non-target species is available, such data are achievable. If non-target studies are to be performed at a number of localities it cannot be guaranteed that the same non-target species can be collected on each site. The same is to be said about studies on the same locality over several years.

In recent years, the attention to the so-called indirect effects of non-targets has increased. The basic assumption is that the direct effect on non-target species is not the only expected population effect. If a non-target species decreases in numbers, natural enemies of that species will also decrease etc. (Wajnberg et al., 2001). We did not address such questions in our studies; even though they can potentially be present.

7.6 Conclusions

The aim of the projects was to clarify the potential of insect pathogenic fungi for microbial control of Strophosoma spp. The conclusions in the chapters can be summarized as follows:

Isolates of the fungi Metarhizium anisopliae, Beauveria bassiana, Paecilomyces farinosus and Verticillium lecanii were all pathogenic to Strophosoma melanogrammum and S. capitatum.
Based on virulence studies in the laboratory, one isolate of M. anisopliae was selected for field studies.
Molecular methods were successfully implemented to assist in the identification of Strophosoma spp. larvae as a prerequisite in the elucidation of the life cycles of Strophosoma melanogrammum and S. capitatum.
Adults of both S. melanogrammum and S. capitatum feed on the greenery in spring and autumn.
Eggs of S. melanogrammum and S. capitatum are deposited in the canopy and the small larvae drop to the ground to complete their development to adults. This behaviour makes it possible to specifically target a microbial control towards the small larvae.
Both Strophosoma species have a development time of 15-17 months,
although S. capitatum may have a greater flexibility in its development time from egg to adult.
Field application of M. anisopliae in summer against larvae of Strophosoma spp. resulted in population reductions the following year.
It was also possible to obtain a reduction of population levels by treating the adult Strophosoma spp. during spring prior to reproduction.
The applied fungus, M. anisopliae, was able to infect non-target coleopterans, hemipterans and ticks in the field plots.
M. anisopliae BIPESCO 5 persisted in the soil for more than one year after application.
In overall, the potential of microbial control of Strophosoma spp. in Danish A. procera stands using insect pathogenic fungi is regarded as high.
We suggest, based on the present knowledge, the future strategy of application is a prophylactic treatment of the greenery stands prior to harvest of the decoration green.