Fungicides continue to be essential for the effective control of plant diseases. New classes of fungicides with novel modes of action are being developed in the 1990s. These include the strobilurins, phenylpyrroles, anilinopyrimidines, phenoxyquinolines, and compounds that trigger defense mechanisms in the plant. For the foreseeable future, new toxophores will be identified through a process of random screening, with natural products representing a rich source of fungicide leads. Progress is being made in the development of high-throughput screens comprised of target enzyme sites or cell-based assays; these techniques will improve the probability of discovery. Following the identification of suitable leads, biorational design is used to optimize specific properties. In vivo glasshouse screens and field trials are expected to remain the dominant methods for characterizing new compounds. Low toxicity to humans and wildlife, low environmental impact, low residues in food, and compatibility with integrated pest management (IPM) programs are increasingly important considerations in the selection of fungicides for development.

No abstract

: Insecticides, fungicides and herbicides are critical to successful crop production, but the development of pesticide resistance is a continual threat, especially to many of todayÏs selective toxophores with speciÐc binding sites. In order to manage resistance e †ectively, an assessment of genetic, ecological and operational risk factors is required, which must then be translated into meaningful local strategies that can be implemented through appropriate labelling of products and education of end users. Assessing resistance risk is a fundamental part of the development process for new molecules and is increasingly becoming a requirement of registration alongside toxicological and environmental risk data. Laboratory studies, including elucidation of target sites and metabolic degradation pathways, mutagenesis, computer models and cross-resistance tests, and Ðeld studies, including establishment of baseline sensitivities and evaluation of anti-resistance strategies, all play a part in such assessment. The challenge is then to devise management strategies which are relevant to local practice and actually reduce selection pressure to a point where product life is preserved. A preventative strategy should be in place at time of launch and for most pesticides, regional co-operation between all interested parties, of the kind advocated by the Resistance Action Committees of GCPF (Global Crop Protection Federation), increases the chance of success. Implementation of strategies via a universal product labelling system, already practised in some herbicide markets, is seen as a key way to improve product usage patterns. Monitoring resistance levels in Ðeld populations after product launch enables any Ðne tuning of tactics over time, for example in response to new technologies such as transgenic varieties being introduced. The limited successes in resistance management in Australia, Zimbabwe, Europe and USA are not so easily achieved in small-holder farming in developing countries, as exempliÐed by continuing problems in parts of India and China. Emphasis must be given to the education of growers and dealers in IRM and IPM, and improved extension services, in order to bring about a more sustainable approach to crop protection.1998 Society of Chemical Industry ( Pestic. Sci., 54, 435È446 (1998)

From 1997 onward, the strobilurin fungicide azoxystrobin was widely used in the main banana-production zone in Costa Rica against Mycosphaerella fijiensis var. difformis causing black Sigatoka of banana. By 2000, isolates of M. fijiensis with resistance to the quinolene oxidase inhibitor fungicides were common on some farms in the area. The cause was a single point mutation from glycine to alanine in the fungal target protein, cytochrome b gene. An amplification refractory mutation system Scorpion quantitative polymerase chain reaction assay was developed and used to determine the frequency of G143A allele in samples of M. fijiensis. Two hierarchical surveys of spatial variability, in 2001 and 2002, found no significant variation in frequency on spatial scales <10 m. This allowed the frequency of G143A alleles on a farm to be estimated efficiently by averaging single samples taken at two fixed locations. The frequency of G143A allele in bulk samples from 11 farms throughout Costa Rica was determined at 2-month intervals. There was no direct relationship between the number of spray applications and the frequency of G143A on individual farms. Instead, the frequency converged toward regional averages, presumably due to the large-scale mixing of ascospores dispersed by wind. Using trap plants in an area remote from the main producing area, immigration of resistant ascospores was detected as far as 6 km away both with and against the prevailing wind.