Phytopathologia Mediterranea (Jan 2016)

Potential distribution of <em>Xylella fastidiosa</em> in Italy: a maximum entropy model

  • Luciano BOSSO,
  • Danilo RUSSO,
  • Mirko DI FEBBRARO,
  • Gennaro CRISTINZIO,
  • Astolfo ZOINA

DOI
https://doi.org/10.14601/Phytopathol_Mediterr-16429
Journal volume & issue
Vol. 55, no. 1

Abstract

Read online

Species distribution models may provide realistic scenarios to explain the influence of bioclimatic variables in the context of emerging plant pathogens. Xylella fastidiosa is a xylem-limited Gram-negative bacterium causing severe diseases in many plant species. We developed a maximum entropy model for X. fastidiosa in Italy. Our objectives were to carry out a preliminary analysis of the species’ potential geographical distribution and determine which eco-geographical variables may favour its presence in other Italian regions besides Apulia. The analysis of single variable contribution showed that precipitation of the driest (40.3%) and wettest (30.4%) months were the main factors influencing model performance. Altitude, precipitation of warmest quarter, mean temperature of coldest quarter, and land cover provided a total contribution of 19.5%. Based on the model predictions, X. fastidiosa has a high probability (> 0.8) of colonizing areas characterized by: i) low altitude (0–150 m a.s.l.); ii) precipitations in the driest month < 10 mm, in the wettest month ranging between 80–110 mm and during the warmest quarter < 60 mm; iii) mean temperature of coldest quarter ≥ 8°C; iv) agricultural areas comprising intensive agriculture, complex cultivation patterns, olive groves, annual crops associated with permanent crops, orchards and vineyards; forest (essentially oak woodland); and Mediterranean shrubland. Species distribution models showed a high probability of X. fastidiosa occurrence in the regions of Apulia, Calabria, Basilicata, Sicily, Sardinia and coastal areas of Campania, Lazio and south Tuscany. Maxent models achieved excellent levels of predictive performance according to area under curve (AUC), true skill statistic (TSS) and minimum difference between training and testing AUC data (AUCdiff). Our study indicated that X. fastidiosa has the potential to overcome the current boundaries of distribution and affect areas of Italy outside Apulia.

Keywords