Species classified in Penicillium sect. Chrysogena are primary soil-borne and the most well-known members are P. chrysogenum and P. nalgiovense. Penicillium chrysogenum has received much attention because of its role in the production on penicillin and as a contaminant of indoor environments and various food and feedstuffs. Another biotechnologically important species is P. nalgiovense, which is used as a fungal starter culture for the production of fermented meat products. Previous taxonomic studies often had conflicting species circumscriptions. Here, we present a multigene analysis, combined with phenotypic characters and extrolite data, demonstrating that sect. Chrysogena consists of 18 species. Six of these are newly described here (P. allii-sativi, P. desertorum, P. goetzii, P. halotolerans, P. tardochrysogenum, P. vanluykii) and P. lanoscoeruleum was found to be an older name for P. aethiopicum. Each species produces a unique extrolite profile. The species share phenotypic characters, such as good growth on CYA supplemented with 5 % NaCl, ter- or quarterverticillate branched conidiophores and short, ampulliform phialides (< 9 μm). Conidial colours, production of ascomata and ascospores, shape and ornamentation of conidia and growth rates on other agar media are valuable for species identification. Eight species (P. allii-sativi, P. chrysogenum, P. dipodomyis, P. flavigenum, P. nalgiovense, P. rubens, P. tardochrysogenum and P. vanluykii) produce penicillin in culture.

Fusarium basal rot (FBR), which is caused predominantly by Fusarium oxysporum and F. proliferatum, is the main limiting factor of onion crops. Resistant cultivars obtained in other countries do not behave as such in Argentina crop fields. The cultivars Antártica-INTA, Grano de Oro-Seminis, Valcatorce-INTA and TW-2007 (reported as tolerant) were tested with five Fusarium spp. isolates, using four inoculum concentrations. Disease incidence was recorded along 28 days and the area under disease progress curve was calculated. Diverse epidemiological models were fitted to experimental data. There were significant differences in the resistance level among cultivars, with TW-2007 being the most tolerant. Local Fusarium isolates were the most virulent ones. The concentration of 10,000 microconidia/ gram was the most lethal for all isolates. The absence of resistance to Fusarium in the four cultivars tested was confirmed. Inoculum concentration and isolate are critical factors in screening for resistance to FBR. Breeding based on the selection of genotypes against low virulence strains of Fusarium spp. and the presence of more aggressive strains in local fields may be one of the causes why varieties reported as resistant or tolerant behave as susceptible in our environment.

A total of 147 samples of garlic ( Allium sativum ) bulbs affected by blue mould were obtained from a variety of agroclimatic districts between December 1999 and February 2000. Penicillium species were identified using both morphological and chemotaxonomic characteristics. Penicillium allii was the predominant species isolated (81·8%) in this survey and the only species proven to be pathogenic on garlic. Other species were isolated much less frequently: P. chrysogenum (13·7%), P. brevicompactum (2·8%), P. phoeniceum (0·9%), P. aurantiogriseum (0·6%) and P. flavigenum (0·2%). Colonies of P. allii could be classified into four morphotypes and their distribution seemed to be influenced by seed trade and agricultural practices. Penicillium allii isolates were grouped into three aggressiveness phenotypes (low, medium and high) based on their ability to cause disease during field trials on susceptible (Fuego INTA) and less susceptible (Castaño INTA) garlic cultivars. The number of surviving plants at 191 days after planting and postharvest bulb weight contributed the most towards aggressiveness modelling.

Blue mould disease in garlic ( Allium sativum ) is associated worldwide with various Penicillium species, and has been attributed to significant annual crop losses in Argentina; the world's second largest exporter of garlic. To identify the pathogenic Penicillium species responsible, affected plants were sampled in La Consulta station (33 ° 45 ′ S, 69 ° 02 ′ W) and placed in a damp chamber. Characteristic disease symptoms are stunted and chlorotic plants with withered leaves and reduced bulb size. Bulbs are often covered with blue/green conidial masses. Isolations were made from fungal colonies emerging on affected bulbs. Pure cultures (IBT 26466, 26467, 26511 and 26512; CMB collection, BioCentrum-DTU, Denmark) were initially identified by micromorphology as Penicillium allii , and identifications were confirmed by comparing reversed phase-HPLC secondary metabolite profiles with those of P. allii type strains.To confirm pathogenicity, sterilized garlic cloves were injured with a needle and inoculated with 5 µ L P. allii spore suspensions (adjusted to 5 × 10 6 conidia mL − 1 ). Inoculated cloves were planted in a field not previously cultivated with garlic, and the original disease symptoms were produced on inoculated plants while water-inoculated control plants remained healthy. The survival rate of inoculated plants was 68% and P. allii was reisolated from diseased field plants.Penicillium viridicatum was first reported as the causal agent of blue mould of garlic in Argentina (Gatica & Oriolani, 1984) before the characterization of P. allii (Vincent & Pitt, 1989). Penicillium allii is micromorphologically similar to P. viridicatum and both species produce yellow exudates in pure culture. To compare pathogenic ability, standard P. viridicatum strains IBT 16939 and 15053 were inoculated into sterilized garlic cloves and incubated for 12 days. The P. viridicatum strains were not able to sporulate on the garlic cloves.Penicillium hirsutum was recently reported as a pathogen on garlic in Argentina (Cavagnaro et al ., 2005). However, P. allii but not P. hirsutum has been reported as an aggressive pathogen of garlic in comparative pathogenicity trials conducted in damp chambers (Overy et al ., 2005). The results reported here suggest that P. allii , rather than P. hirsutum or P. viridicatum , is the pathogenic species responsible for garlic crop losses due to blue mould rot in Argentina. This is the first report confirming P. allii as a field pathogen of A. sativum.

Fifteen strains representing each Penicillium ser. Corymbifera taxa were compared using phenotypic and chemotaxonomic characters by cluster analysis and discriminant partial least squares regression. Variability in phenotypic expression of species strains resulted in a more fragmented classification compared with secondary metabolite expression. Although the observed phenotypic expression varied for strains cultured upon the same media, it was possible to classify strains into species groupings based only upon a few distinctive phenotypic traits. Data analysis of secondary metabolite profiles generated from HPLC-diode array dectection analysis gave reliable strain classification when more than one media type was employed. Depending on the species, Czapek yeast autolysate agar typically yielded the greatest chemical diversity; however, several metabolites (terrestric acid, corymbiferone, the corymbiferan lactones, and daldinin D) were only produced when strains were grown on either yeast extract sucrose or oatmeal agar. For the classification of strains based on a binary data matrix, application of the Yule coefficient gave the best clustering. Several secondary metabolites, of importance for the classification of ser. Corymbifera strains, were identified by discriminant-partial least squares regression analysis. A diagnostic key based on phenotypic, chemotaxonomic, and pathogenic traits is provided as an aid for species identification.Résumé : Les auteurs ont comparé quinze souches représentant chacun des taxons du Penicillium ser. Corymbifera, sur la base de caractères phénotypiques et chémotaxonomiques, à l'aide de l'analyse par regroupements et la régression des moindres carrés partiellement discriminants. La variabilité de l'expression phénotypique des souches spécifiques aboutit à une classification plus fragmentée, comparativement à l'expression des métabolites secondaires. Bien que l'expression phé-notypique varie chez les cultures cultivées sur les mêmes milieux, il est possible de classifier les souches en regroupements d'espèces, mais seulement sur la base de quelques caractères phénotypiques distinctifs. L'analyse des données des métabo-lites secondaires, générées à partir d'analyses HPLC-DAD, conduit à une classification fiable des souches, lorsqu'on utilise plus d'un type de milieux. Selon l'espèce, la gélose de Czapek avec autolysat de levure donne généralement la plus grande diversité chimique; cependant, plusieurs métabolites (acide terrestrique, corymbiferone, lactones corymbifériennes et daldine D) ne sont produits que lorsque les souches sont cultivées sur gélose à base de saccharose et extrait de levure, ou de farine d'avoine. Pour la classification des souches, basée sur une matrice de données binaires, l'application du coefficient de Yule donne le meilleur regroupement. Plusieurs métabolites secondaires, importants pour la classification des souches de la sér. Corymbifera, ont été identifiées par l'analyse D-PSLR. Les auteurs proposent une clé diagnostique d'identification des espèces, basée...