Daily fluctuations in leaf temperature modulate the development of a foliar pathogen

Highlights

• The daily amplitude of air and leaf temperatures differed in a wheat field.

• Pathogen fitness is impacted by the short-term fluctuations of leaf temperature.

• The higher daily leaf thermal amplitude resulted in a longer latent period.

• Implications could be high for epidemiology modeling, in the climate change context.

Abstract

Thermal ecology studies on the ecophysiological responses of organisms to temperature involve two paradigms: physiological rates are driven by body temperature and not directly by the environmental temperature, and they are largely influenced not only by its mean but also its variance. These paradigms together have been largely applied to macro invertebrates and vertebrates but rarely to microorganisms. According to these paradigms, foliar fungal pathogens are expected to respond directly to the fluctuations in leaf temperature, rather than in air temperature. We determined experimentally the impact of two patterns of leaf temperature variation of equal mean temperature, but differing in their daily amplitude, on the development of Zymoseptoria tritici, a fungus infecting wheat leaves. The highest daily thermal amplitude resulted in two detrimental effects for the pathogen fitness: an increase in the length of the latent period, i.e. the ‘generation time’ of the fungus when infecting its host plant, and a decrease in the density of fruiting bodies on the leaves. We discussed these empirical results, mainly the impact of both the daily thermal amplitude and the fluctuation frequency on the pathogen development in planta, in the light of the mathematical effect of the integration of non-linear functions. We concluded that it is necessary to take into account daily leaf temperature amplitudes to improve our understanding and prediction of the development of foliar fungal pathogens and other micro-organisms living in the phyllosphere in the climate change context.

Introduction

Most organisms grow and evolve in fluctuating environments (Du and Ji, 2006). This is particularly true regarding temperature, a factor that affects many, if not all, physiological processes involved in growth and development (Angilletta, 2009). Temperature fluctuates at various time scales, from minute to years, with periodicity over daily and yearly periods, and each time scale can matter for different physiological processes from thermal tolerance (small time scales) to diapause/quiescence strategies (long time scales) (Dillon et al., 2016). The exchange of energy (radiation, convection, conduction, latent heat) between an organism and its environment generates temperature deviations between the body of organisms and their surrounding air. In ectotherms, the body temperature, subjected to fluctuating solar radiation and wind, and to night sky radiation cooling, is expected to fluctuate more intensively and rapidly than air temperature (Gates, 1980). This paradigm also applies to plant surfaces, which generate a particular thermal environment for the large variety of organisms living on them, from phytophagous arthropods to bacteria and fungal pathogens.

Leaf dwelling organisms experience variations in the temperature of the leaf surface rather than in ambient air (Pincebourde and Woods, 2012; Pincebourde et al., 2021; Scherm and van Bruggen, 1993). Temperature heterogeneity in space and time over leaf surface can depart largely from ambient air, with deviation of up to 20 °C between a specific leaf area and ambient air (Saudreau et al.. 2017). The potential effects of large thermal amplitude on biological processes within the leaf envelope have been considered in studies on the effect of climate variations on arthropod development (Bradshaw et al., 2000; Kingsolver, 1979; Pincebourde and Casas, 2006; Potter et al., 2009; Pincebourde and Suppo, 2016). The influence of the ‘phylloclimate’ (the microclimatic conditions occurring in the phyllosphere; Chelle, 2005) is presumed to be high on the whole leaf microbiota (Pincebourde and Woods, 2012; Vacher et al., 2016). Many studies focused on the impact of (constant) temperature on plant disease cycles (incubation, latent period, senescence, etc.), both by experimental and modeling approaches (de Wolf and Isard, 2007), but the impact of the fluctuation of leaf temperature on the development of leaf fungal pathogens has never been studied, except partly by Bonde et al. (2012). Some studies focused on entire living plants or detached leaves (e.g. Scherm and van Bruggen, 1994a; Xu, 1996; Shakya et al., 2015) but were based on air temperature fluctuations, while some others were carried out on artificial media instead of leaf (e.g. Zhan and McDonald, 2011; Boixel et al., 2019). This last experimental approach has two main drawbacks: (i) a Petri dish or wells of a microplate does not have the same energy budget as a living leaf; (ii) only the direct effect of temperature on the fungus can be observed, while the complex interrelationship between environmental temperature, leaf temperature, and the foliar and fungal responses are ignored. Therefore, the use of artificial media limits our ability to apply growth predictions to more natural situations.

Mean temperature can be an uninformative, even misleading, descriptor of a fluctuating thermal environment (Kingsolver et al., 2004). Thermal fluctuations likely matter for the growth of fungal pathogens. Scherm and van Bruggen (1994b) were the first to demonstrate theoretically that the difference between growth of plant fungal pathogens at constant versus fluctuating temperatures is maximized when the mean temperature is close to one of the three cardinal temperatures (minimal, optimal and maximal temperatures) and/or when the temperature range over which the growth response is approximately linear is narrow. As an example, the use of daily mean temperature to predict the incubation period (for plant pathologists, the time needed for the first symptoms to appear) and the latent period (the ‘generation time’, i.e. the duration between inoculation and the appearance of fruiting bodies releasing contaminating spores) of a fungal pathogen under field conditions may result in errors when the underlying rate function is non-linear (Xu, 1996). A higher resolution in temperatures is therefore required, and several studies showed that the hourly time step increased the accuracy of predictions (Narouei-Khandan et al., 2020; Salotti and Rossi, 2021). The model developed by Narouei-Khandan et al. (2020) to simulate effects of daily amplitudes on the development of late blight highlighted a significant interaction between average air temperature and amplitude in their effects on the area under the disease progress curve (AUDPC) as predicted from growth chamber data on a single infection cycle. Greater effects of amplitudes were observed at the extreme temperatures (including the optimal temperature), and no amplitude effect at the inflection point of the optimal temperature curve. The importance of daily temperature fluctuations was also demonstrated by Salotti and Rossi (2021) for the development of Ascochyta blight on chickpeas. Environmental sampling rate, such as the frequency of temperature recording (duration of time step), relative to the frequency of leaf temperature changes, is therefore crucial when predicting organism fitness, yet few studies have quantified its importance.

The objective of this study was to assess the relevance of using leaf temperature rather than air temperature as climatic driver by (i) comparing daily amplitudes of leaf and air temperatures in field conditions, and (ii) comparing the in planta development of a foliar fungal pathogen under two leaf temperature regimes of equal mean but differing in their daily leaf thermal amplitude (DLTA). As a case study, we used the fungus Zymoseptoria tritici (formerly Mycosphaerella graminicola), the causal agent of Septoria tritici blotch disease on wheat. Present wherever wheat is grown and developing throughout the wheat growing season (Suffert and Sache, 2011), the pathogen is exposed to a wide range of mean and amplitude temperatures across its geographical distribution (Suffert et al., 2015). Finally, we used a simple mathematical model based on a non-linear relationship of fungal performance to temperature (see Supplementary Materials) to provide additional support for discussing on the importance of daily temperature amplitude relative to both the shape of the nonlinear growth curve and the frequency of temperature recordings.