Sequences of selected marker loci have been widely used for the identification of specific pathogens and the development of sequence-based diagnostic methods. Although such approaches offer several advantages over traditional culture-based methods for pathogen diagnosis and identification, they have their own pitfalls. These include erroneous and incomplete data in reference databases, poor or oversimplified interpretation of search results, and problems associated with defining species boundaries. In this letter, we outline the potential benefits and drawbacks of using sequence data for identification and taxonomic deduction of plant-pathogenic fungi and oomycetes, using phytophthora as a primary example. We also discuss potential remedies for these pitfalls and address why coordinated community efforts are essential to make such remedies more efficient and robust.

The mycobiota of fresh and ensiled maize was studied with culturing techniques and a DNA sequence-based approach. Freshly chopped and ensiled maize were collected for 2 y from 12 farms in Pennsylvania. Samples were plated on selective media and isolates identified by morphology and sequences of the internal transcribed spacer regions of rDNA, 800-900 bp of the 5' end of the translation elongation factor 1-alpha gene and a portion of the rodA gene (Aspergillus fumigatus only). ITS regions were amplified from total silage DNA, cloned, sequenced and compared to fungal ITS sequences in GenBank with the BLAST-N algorithm. For samples analyzed by both methods, the molecular technique detected a greater number of species than selective plating. Plating recovered several Penicillium and Fusarium species and Aspergillus fumigatus, while molecular analysis detected Alternaria, Penicillium and Fusarium species. Data from both methods found that Fusarium and Penicillium were the dominant mycotoxigenic fungi in silage, while yeast made up the majority all fungi recovered or detected. Known mycotoxigenic species often accounted for 50% or more of the total number of species isolated or detected at each site. Viable Fusaria were not isolated from or detected in ensiled maize, suggesting that Fusarium species do not survive the ensiling process. Results from this study suggest that given the numerous species of fungi present in silage with mycotoxin producing ability, there is a strong possibility that silage may be contaminated with multiple toxins simultaneously.

Toxins produced by Penicillium species are reported in maize silage and have been associated with health problems in cattle. Our objectives were to evaluate the prevalence and dynamics of patulin (PAT), mycophenolic acid (MPA), cyclopiazonic acid (CPA), and roquefortine C (ROC) in fresh and ensiled maize. To achieve these objectives we developed a high-performance liquid chromatography method coupled with mass spectrometry to detect all four toxins simultaneously in silage. In addition we collected weather data, information on agronomic practices, and silage fermentation characteristics for each study site. Silage was collected at harvest and after ensiling in 2001 and 2002 from 30 Pennsylvania dairies. The average concentration of toxins (range in parentheses) was: PAT 0.08 microg/g (0.01 to 1.21), MPA 0.16 microg/g (0.02 to 1.30), CPA 0.12 microg/g (0.02 to 1.43), and ROC 0.38 microg/g (0.01 to 5.71). ROC was the most frequently detected toxin (60%), followed by MPA (42%), CPA (37%), and PAT (23%). Of 120 samples tested, 15% contained no detectible levels of toxin, 25% were contaminated with one toxin, 32% with two, 18% with three, and 10% with all four toxins. All four mycotoxins were found in freshly harvested material, contradicting the belief that Penicillium toxin formation occurs exclusively during storage. We observed that weather conditions during specific growth stages of the crop affected the final concentration of toxins in freshly harvested maize. In ensiled material, PAT levels were affected by concentrations of propionic and isobutyric acids. Based on our data, Penicillium mycotoxins can form while the crop is in the field and after ensiling, suggesting that preventative measures should begin prior to ensiling.

Phytophthora spp. represent a serious threat to agricultural and ecological systems. Many novel Phytophthora spp. have been reported in recent years, which is indicative of our limited understanding of the ecology and diversity of Phytophthora spp. in nature. Systematic cataloging of genotypic and phenotypic information on isolates of previously described species serves as a baseline for identification, classification, and risk assessment of new Phytophthora isolates. The Phytophthora Database (PD) was established to catalog such data in a web-accessible and searchable format. To support the identification of new Phytophthora isolates via comparison of their sequences at one or more loci with the corresponding sequences derived from the isolates archived in the PD, we generated and deposited sequence data from more than 1,500 isolates representing the known diversity in the genus. Data search and analysis tools in the PD include BLAST, Phyloviewer (a program for building phylogenetic trees using sequences of selected isolates), and Virtual Gel (a program for generating expected restriction patterns for given sequences). The PD also provides a customized means of storing and sharing data via the web. The PD serves as a model that easily can be adopted to develop databases for other important pathogen groups.

This study was initiated to investigate the possible role of Phytophthora species in white oak decline (Quercus alba) in southern Ohio at Scioto Trail State Forest. Surveys demonstrated the presence of four species of Phytophthora including one novel species. By far, the most common species was P. cinnamomi; P. citricola and P. cambivora were isolated infrequently. In few instances, P. cinnamomi was isolated from fine roots and necroses on larger roots. No special pattern of incidence was found, but P. cinnamomi was more commonly isolated from greater Integrated Moisture Index values suggesting moist lower bottomlands favour this Phytophthora species. When tree crown condition was examined relative to the presence of Phytophthora, no significant association was found. However, roots of declining P. cinnamomi-infested trees had 2.5 times less fine roots than non-infested and healthy trees, which was significantly different. The population densities of P. cinnamomi from declining trees were significantly greater than from healthy trees, suggesting increased pathogen activity that has the potential to cause dieback and decline and possibly the cause of a reduced fine root amount found on declining trees.