Figure 3a).
In a second plant assay, the five bacterial cultures (i.e., the five strains) of each species were assembled to constitute two experimental populations and each was inoculated on 15 plants. A mix of the two species (10 strains) was also inoculated on 15 plants. The assay was duplicated (2 x 15 plants per treatment). Kruskal-Wallis test revealed differences between the three treatments, i.e., D. solani strain mixture, D. dianthicola strain mixture and species mixture (k=6.1; DF=2; p=0.04). Pairwise comparison of the symptom class counts (Post-hoc Tukey test) showed that the aggressiveness of D. solaniwas higher than that of D. dianthicola (F=-2.5; p= 0.03) and that of the species mixture (F=1.9; p= 0.10). However, symptom incidence caused by the species mixture did not differ from those of the D. dianthicola mixture (F=-0.5; p= 0.86). Post-hoc Tukey p-values and mean value (± SE) of the percentage of symptomatic plants inoculated by theD. dianthicola (53% ± 7), D. solani (23% ± 3) and mixed (47% ± 0) populations are presented in the figure 3b .
In a subsequent experiment, qPCR allowed to quantify pathogen loads. When the two species mixtures were inoculated separately, D. dianthicola and D. solani reached the mean values ± SE of 2 x 1011 ± 1 x 104 cells and 3 x 109 ± 1 x 102 cells per gram of symptomatic stem tissues, respectively. In eight co-infected symptomatic plants, quantifications of the pathogen abundance revealed an excess ofD. dianthicola with a CI median value of 10-5(Figure 3c ). The eight CI values differed from one (Kruskal-Wallis test, k=9.5; DF=1; p= 2 x 10-3), meaning that D. dianthicola outcompeted D. solani in blackleg tissues. Altogether, these data revealed a fitness advantage ofD. dianthicola in terms of multiplication within lesions in the aerial parts of the host.
Pathogen mixtures were also inoculated in pots containing plants with wounded roots. Symptomatic and asymptomatic plants were counted among 15 plants per treatment. The assay was duplicated (2 x 15 plants per treatment). Kruskal-Wallis test revealed differences between the three treatments, i.e., D. solani strain mixture, D. dianthicolastrain mixture and species mixture (k=29.2; DF=2; p=4.6x10-7). Pairwise comparisons (post-hoc Tukey tests) showed that the number of plants with symptoms differed following inoculation by D. solani from inoculation by D. dianthicola (F=-4.5; p=1.9x10-5) and by a mixture ofD. dianthicola and D. solani (F=6.1; p=2.7x10-9). Mean ± SE of the percentage of symptomatic plants reached 33% ± 0 for D. solani , 83% ± 10 forD. dianthicola and 93% ± 7 for a mix of D. dianthicolaand D. solani , indicating a stronger aggressiveness of D. dianthicola as compared to D. solani . In 27 co-infected plant tissues, qPCR quantification of the pathogens revealed an excess ofD. dianthicola with a CI median value of 10-5. These CI values were different from one (Kruskal-Wallis test: k=19.5; DF=1; p=10-5; Figure S3) .
Altogether these experimental data revealed fitness differences betweenD. solani and D. dianthicola for infecting and exploiting potato plants, the resident being fitter than the invader in all conditions and measured traits.