Spatio-temporal analysis of post-harvest moulds genera distribution on stored durum wheat cultivated in Tunisia

• Advantageous topological and typological statistical analyses for fungal diversity study.
• Organization of fungal diversity into 4 community patterns in stored wheat.
• Significant variations of fungal community patterns in relation to geographical sites.
• Alternated peaks of fungal genera through different periods of year.

Wheat represents a principal ingredient in traditional Tunisian diet including couscous, bread, pasta and biscuits. Northen Tunisia is an important growing area of wheat which after harvest is stored in silos and on farm. The cereal grains can become contaminated by post-harvest moulds during storage in silos under unfavorable conditions leading to a decrease in quality, packing and marketing of wheat. In this study, a mycological survey was undertaken to determine the biodiversity of post-harvest moulds on durum wheat stored in silos localized in five regions of Northern Tunisia and to investigate changes during the storage period. A total of 127 samples were obtained from Oued Mliz, Jendouba, Ksar Mezouar, Mateur and Ghezala silos during 2010–2011 and 2011–2012 wheat seasons. After sampling, seeds were placed on Potato Dextrose Agar medium (PDA) for 7 days of incubation at 28 °C. A total of 6035 strains of filamentous fungi were isolated.The quantitative and qualitative changes on wheat mycoflora during storage were statistically explored by multivariate methods including correspondence and hierarchical cluster analysis. The most predominant post-harvest moulds genera isolated were Alternaria (28%), Fusarium (19%), Penicillium (19%), Aspergillus (14%), Mucor (8%) and Rhizopus (7%). Various genera of fungi imperfecti, including Ulocladium, Geotrichum, Chaetomium, Trichothecium, Paecilomyces, Aureobasidium and Chrysonilia (anamorphic Neurospora), and the Mucorales genera Lichtheiia and Syncephalastrum accounted for the remainder of about 6% of the total. Statistical data analysis revealed six mycological patterns corresponding to six distinct communities as characterized by the prevalence of different moulds. Such patterns clearly showed different spatio-temporal variability indicating that distribution and evolution of moulds during storage was sensitive to geographic location, year of sampling and short or long-term storage.