The ecology of a Pythium community in relation to the epidemiology of carrot cavity spot

Carrot cavity spot (CCS) is characterised by the appearance of small sunken elliptical lesions on the tap root. It is caused by a complex of Pythium species, but the species diversity and interactions within the complex have never been studied for modelling CCS epidemics. The diversity of a pathogenic Pythium community was assessed during 3 consecutive years in a field experiment after an initial artificial soil infestation with P. violae. 1241 lesions were examined, yielding 728 Pythium isolates. Conventional microbiological methods and restriction polymorphism of the internal transcribed spacer regions of the rDNA of 209 representative Pythium isolates allowed us to identify 655 isolates as belonging to six Pythium species, including P. violae and five indigenous species (P. sulcatum, P. intermedium, P. sylvaticuinlirregulare, P. coloratum, and P. ulthnum). Biological traits, such as pathogenicity, optimum temperature for mycelial growth and saprophytic survival of the inoculum, explained the fluctuations in the composition of the complex over 17 successive samplings during the 3-year period, most notably the prevalence of first P. violae and then P. sulcatum. P. violae and P. sulcatum were occasionally isolated in mixture from single lesions (10.4% and 9.6%, respectively). Other species were more frequently isolated in mixture: 30.8% for P. interniedium, 33.8% for P. sylvaticum/irregulare, 42.9% for P. ultinium, and 66.7% for P. coloratum. A contingency analysis allowed us to define 'major' and 'minor' species on both pathological and ecological criteria (frequency of occurrence in the complex, pathogenicity and ability to induce lesions by themselves), and demonstrated that infection by one 'major' pathogen species (P. violae or P. sulcatum) is not positively correlated with the presence of a second Pythium species. The ratio between 'observed' and 'expected' mixed infection frequency under the assumption of independent infection (mir) was less than l for P. violae, P. sulcatum, P. intermeditan, and P. ylvaticunll irregulare (P < 0.05). For all Pythium species, there was a negative linear relationship between mir and pathogenicity (R-2 = 0.638): the less a Pythium species was pathogenic on carrot, the more often it was isolated from a CCS lesion in mixture with at least one other species. The non- significance of interactions between species during the infection phase suggests that CCS epidemics can be analysed as if they were caused by a single Pythium species. (c) 2006 Elsevier B.V. All rights reserved.