Ecophysiological behavior of major Fusarium species in response to combinations of temperature and water activity constraints

Abstract

Fusarium head blight (FHB) is a devastating fungal disease affecting cereals, caused by Fusarium species that can produce harmful mycotoxins. Fusarium species coexist within the same ecological niche during infection, with their population dynamics and associated mycotoxin patterns strongly influenced by the environment. This study provides a comprehensive investigation of the ecophysiological responses of the major Fusarium species causing FHB under varying abiotic factors. We assessed growth and mycotoxin production of different isolates of Fusarium avenaceum, Fusarium graminearum, Fusarium langsethiae, Fusarium poae, and Fusarium tricinctum under 24 combinations of temperature (θ = 15, 20, 25, 30°C) and water activity levels (a_w = 0.99, 0.98, 0.97, 0.96, 0.95, 0.94). Our findings indicated that θ, a_w, and their interaction have a main significant impact on species behavior. Thanks to innovative statistical approaches using fungal growth data from optical density measurements and mycotoxin quantification, we demonstrated significant inter- and intra-specific differences in environmental responses. Growth and mycotoxin production of F. graminearum and F. avenaceum appeared favored under high temperature (≥25°C) and high water activity (≥0.97), whereas lower a_w levels (≥0.95) were also conducive for F. poae and F. tricinctum. A specific and unique behavior of F. langsethiae to lowest temperatures (≤20°C) was highlighted. Understanding the ecophysiological requirements of Fusarium species is crucial in the context of climate change, which is expected to worsen disease outbreaks. This study provides valuable knowledge for improving the reliability and robustness of FHB prediction models and anticipating the associated mycotoxin risk.IMPORTANCEFusarium species pose a significant threat to major cereal crops, particularly wheat, by reducing yields and producing mycotoxins that are harmful to animals and humans. The prevalence of each Fusarium species is strongly influenced by environmental conditions, and climate changes have already been reported as responsible for shifts in pathogen populations, leading to changes in mycotoxin patterns. This study revealed distinct ecophysiological behaviors, including growth and mycotoxin production, of the five major Fusarium species infecting small grain cereals when exposed to varying temperature and water activity conditions. Our findings provide a valuable foundation for a deeper understanding of mycotoxin risk and for developing more effective mitigation strategies in the near future.

Keywords: adaptive response; climate change; combined effects; ecological niche; fungal pathogens; inter/intraspecific diversity.

Fig 1

Growth frequency (f_g) and growth parameters (𝜏, K, r, and Vmax) for the five Fusarium species, according to environmental variation. f_g (expressed without unit) and growth parameters (𝜏 expressed in days, K without unit, r, and Vmax in OD_unit.days⁻¹) in lines were estimated on five different isolates per species (F. avenaceum, F. graminearum, F. langsethiae, F. poae, and F. tricinctum, on the x-axis). Four different θ (15, 20, 25, and 30°C, shown in columns) and six levels of a_w (0.94, 0.95, 0.96, 0.97, 0.98, and 0.99) were tested. Each dot represents the mean frequency of the six replicates for each isolate, and the boxplots show the intra-specific diversity within the five species.

Fig 2

Modeling growth probability with logistic regression for the five Fusarium species, according to environmental variation. Growth probability (p_g on the y-axis) was estimated by the optimal logistic regression model (logit(p_g) = β₀ + (β₁ + β_1i) · θ_i + (β₂ + β_2i) · a_wi + β_3i · fs_i) against four θ (15, 20, 25, and 30°C, on each box) and at six levels of a_w (0.94, 0.95, 0.96, 0.97, 0.98, and 0.99; on the x-axis). Each dot represents the p_g of the five Fusarium isolates for each species. Confidence intervals (95%) are represented by shading around the curves.

Fig 3

Response surface plots of growth parameters values for the five Fusarium species, according to combined abiotic conditions (a_w and ϴ). Contour plots of growth parameters (𝜏, K, r and Vmax in line) were estimated on grown replicates using quadratic models for each Fusarium species (F. graminearum, F. avenaceum, F. langsethiae, F. poae and F. tricinctum in column), as a function of a_w (y-axis) and ϴ (x-axis). Each response surface has its own scale, this is indicated by numbers on the lines of the contour plot. The contour plot of K for F. tricinctum is shaded because the model did not fit correctly the response curves for this species.

Fig 4

Variance partitioning of growth parameters (𝜏, K, r and Vmax) according to species, environmental factors (a_w and ϴ) and their interactions. The percentage of variance explained (y-axis) by different factors (in colors) for each growth parameter (x-axis) was estimated on grown replicates only. Environmental factors (a_w and ϴ), their interaction (a_w × ϴ), species factor and its interactions with environmental factors (Species × a_w and Species × ϴ) are represented in colors. The isolate factor was nested within the species factor.

Fig 5

Variance partitioning of growth parameters (𝜏, K, r and Vmax) according to intraspecific and interspecific diversity. Four different ϴ (15, 20, 25 and 30°C, x-axis) and six levels of a_w (0.94, 0.95, 0.96, 0.97, 0.98 and 0.99, in column) were tested to assess the effect of interspecific and intraspecific diversity on the total variance (y-axis).

Fig 6

Variance partitioning of growth parameters (𝜏, K, r and Vmax) for the five Fusarium species, according to isolate, environmental factors and their interactions. The percentage of variance (y-axis) explained by different factors (in colors) for each Fusarium species studied: (a) F. graminearum, (b) F. langsethiae, (c) F. poae, (d) F. avenaceum and (e) F. tricinctum. Environmental factors (a_w and ϴ), their interaction (a_w × ϴ), isolate factor and its interactions with environmental factors (Isolate × a_w and Isolate × ϴ) are represented in colors.

Fig 7

Mycotoxin production by the five Fusarium species under environmental variation. TCTB for F. graminearum are represented in box 1, TCTA for F. langsethiae and F. poae in boxes 2 and 3, respectively, and ENNs for F. avenaceum and F. tricinctum in boxes 4 and 5, respectively. Mycotoxins produced (log-transformed, y-axis) were analysed at four ϴ (15, 20, 25 and 30°C, x-axis) and three a_w levels (0.95, 0.97 and 0.99, in colors). Some replicates did not produce mycotoxins at a_w = 0.95. Each dot represents one replicate and the boxplots show the intraspecific diversity in terms of mycotoxin production within each Fusarium species. 0 to DL corresponds to the number of grown replicates in which the mycotoxins were not detected (concentration below the limit of detection (DL)). DL to QL corresponds to the number of grown replicates in which the mycotoxins were detected but not quantified because they were below the lower limit of quantification (QL). Quantified corresponds to the number of grown replicates where the mycotoxins were quantified.

Fig 8

Probability of mycotoxin production for the five Fusarium species, according to environmental variation. The probability of mycotoxin production (y-axis) was estimated by the logistic regression model (logit(p_p) = β₀ + (β₁ + β_1i)·ϴ_i + β_2i·a_wi + β_12i·a_wi·ϴ_i + β_3j·fs_j + β_ij·a_wi·fs_j) against four ϴ (15, 20, 25 and 30°C, on the x-axis) and at two levels of a_w (0.97 and 0.99, in column). Each dot represents the probability of mycotoxin production of the five Fusarium isolates for each species. Confidence intervals (95%) are represented by shading around the curves. Data for a_w = 0.95 are not shown because very few grown replicates were able to produce mycotoxins above the quantification threshold.

Fig 9

Principal component analysis of mycotoxin and growth parameters (𝜏, K, r and Vmax) for the five Fusarium species. The ranks of growth parameter and mycotoxins for F. avenaceum, F. graminearum, F. langsethiae, F. poae and F. tricinctum are presented at four ϴ (15, 20, 25 and 30°C) and three a_w levels (0.95, 0.97 and 0.99). Ranking was performed separately for each group of mycotoxins associated with a species : TCTB (DON + 15-ADON) for F. graminearum, TCTA (DAS + T-2 + HT-2) for F. poae and F. langsethiae, and ENNs (ENNA + ENNA1 + ENNB + ENNB1) for F. avenaceum and F. tricinctum. For each species, first biplot is for (Dim1, Dim2) and second biplot is for (Dim1, Dim3). Each dot corresponds to an isolate growing at specific a_w and ϴ.