We predict that a significant part of the answer to the effectiveness of UW85 and other biological control agents will be the influence the agent has on the composition and dynamics of microbial communities. We are also expecting that the identification of functions provided by the plant that support effective biological control will point to some further ecological principles on which a better understanding of systems can be built. Thus, we are expecting that it will be the integrative effects on communities of micro- and macroorganisms that will truly explain biological control.
We envision the kind of work described here as being the infancy of new approaches in which biological control will be based on mimicking the workings of nature and be more responsive to management of the soil, perhaps not even requiring amendment or supplementation with nonindigenous organisms. In limited ways such approaches, termed organic or biodynamic, have worked empirically. Can we understand these approaches well enough at the ecological level that the salient features of their successes could be implemented in the mainstream? We need to learn about the microbial world around us and gain a better understanding of its dynamics and regulation. We must find new tools with which to "see" the microbial world and study its intricacies and learn how to employ emerging tools in agricultural research. We must be able to address fundamental questions about microbes in the agroecosystem. Which organisms are the important ones in supporting healthy plant growth? What in their repertoire of biological activities should be encouraged and what discouraged in a productive agricultural system? How do plants and rhizosphere microflora communicate? An understanding of microbial populations and communities offers tremendous potential for developing knowledge from which we will be able to develop new approaches to disease control for the future.
We have identified recently antibiotic B as the aminoglycoside, kanosamine ( Milner et al. 1996a). A method for detecting zwittermicin A-producing isolates, based on PCR-amplification of a zwittermicin A resistance gene ( Milner et al. 1996b) has successfully identified zwittermicin A-producing isolates from diverse soils ( Raffel et al. 1996). Studies with tomato have demonstrated cultivar differences in responsiveness to biological control in the presence of Pythium ( Smith et al. 1996). Research by other workers published after April 1995 has not been incorporated into the text.
We are indebted to Greg Gilbert for conceptualizing the camouflage hypothesis and enriching our thinking about microbial community ecology, to Eric Stabb for shaping our thinking about genetic diversity of microorganisms, and to Craig Grau for contributing to our thinking about the role of the host genotype in biological con-