Fig 1.
Location of the studied ash stands (black dot) relative to the year of first ash dieback report.
Data from the Département de la Santé des Forêts. Am, Fréchencourt, Be, Sarrance, Ch1-6, Champenoux (6 plots), Co, Colombier, Fe, Ance-Féas, Gr, Gremeçey, La, Landivisiau Lu, Lupé, Ol1-2, Oloron-Sainte-Marie (2 plots), Po, Pont-de-Buis, Ro, Roche-sur-Grane, Sa, Salignac-sur-Charente Sc, Scaër, Se, Seichamps). The region border shapefile can be uploaded at https://geoservices.ign.fr/telechargement.
Fig 2.
Impact and ability of H. fraxineus to complete its cycle on ash saplings with either no or severe dieback.
(A) % leaves with necrosis in autumn 2014 (likelihood Chi-square of 0.7084, p = 0.40). (B) % of rachis length with presence of H. fraxineus apothecia in April 2015 (likelihood Chi-square = 1.19, p = 0.28). (C) % of rated shoot with bark infection in summer 2015 (likelihood Chi-square = 6.99, p = 0.01).
Fig 3.
Modelled and observed shoot mortality for ash trees of different level of dieback.
(A) Site with different dieback status (with dieback, score 0–1 or without, score of 2–3). (B) Sites with different tree cover. Points represent the observed data, pooled by increasing 10% leaf necrosis frequency. For tree cover, blue dots represent sites with 9–25 tree cover while red dots represent sites with 75–100% tree cover. Points representing less than 10 shoots are not represented.
Table 1.
Relationship between leaf necrosis and shoot mortality at the shoot level.
Parameters of the hierarchical Bayesian model.
Table 2.
Relationship between leaf necrosis or shoot mortality and climate parameter (plot level).
Parameters of the hierarchical Bayesian model.
Fig 4.
Modelled and observed leaf necrosis and shoot mortality depending on climatic parameters and sites conditions.
The curves represent the predicted response for a given parameter, all other parameters being fixed at the average value, except for the number of year of ash dieback presence that is set to 1. The orange area (grey in printed version) represent the 95% confidence interval. Dots represent oberved values.
Table 3.
Colonisation of rachises at leaf fall in November.
Fig 5.
Ability of H. fraxineus to produce apothecia on the ash rachises overwintered in common garden (fructification assay).
(A) Relation with leaf necrosis observed in the same plot two months before (pvalue = 0.304). (B) Relation with climate (pvalue = 0.006, Table 4).
Table 4.
Influence of climatic parameters on the ability of H. fraxineus to produce apothecia in the fructification assay.
Fig 6.
Map of climate suitability for H. fraxineus ability to produce apothecia on the ash rachises or symptoms on leaf and shoot of ash trees.
Safran meteorological data, 2010–20. (A). Proportion of ash rachises collected in fall producing apothecia in laboratory conditions after overwintering in a common garden (fructification assay). (B) Leaf necrosis probability. (C) Shoot mortality probability in open canopy (20% tree cover). (D) Shoot mortality probability in forest conditions (100% tree cover).
Fig 7.
Hypothesised impact of studied environmental parameters on H. fraxineus life cycle.
Higher summer rainfall and higher spring temperatures are associated with higher rates of leaf necrosis leading to higher shot necrosis especially in high tree cover. Higher summer temperatures are associated with low leaf necrosis frequency at the end of the summer, but not to low colonisation of rachises at leaf-fall during autumn. Low autumn temperatures are associated with high shoot mortality (better transfer from leaves to shoot or lesion development in shoots?). The part of the cycle needed for inoculum production is indicated in blue while red arrows and text refers to symptoms development that does not lead to inoculum production. Cycle adapted from [41].