Department of Plant Pathology, University of Georgia, Athens, GA 30602, USA

A casual scan of the Current Contents literature data base using the search string temperature AND (moisture OR humidity OR wetness) reveals more than 70 scientific articles published over the past 6 years that investigated interactions between temperature, moisture and some aspect of plant pathogen or plant disease development. The preoccupation of botanical epidemiologists with temperature and moisture has three likely causes: (i) temperature and moisture strongly affect plant disease development in the real world; (ii) experimental quantification of the effects of these two variables on pathogen or disease development has led to the development of microclimate-based disease warning models for numerous pathogens; and (iii) temperature and moisture can be manipulated elegantly in controlled-environment experiments, leading to reproducible results and the prospect of rapid journal publications.

While most of the aforementioned articles studied interactions between temperature, moisture and pathogen or disease development in growth chambers or incubators with constant temperature and moisture regimes (e.g. fixed durations of leaf wetness or constant soil moisture levels), only few studies examined the effects of variable temperatures, fluctuating soil moisture regimes and/or interrupted leaf wetness periods. However, variable temperature and moisture regimes are more typical of the environment of most plant pathogens and too little is known about differences in pathogen or disease development between constant and variable environments to accept results from constant-environment experiments at face value. For temperature, studies with Bremia lactucae (lettuce downy mildew) and Podosphaera leucotricha (apple powdery mildew) suggested the existence of differences in fungal development times between constant and variable environments with the same mean temperatures. In experiments with B. lactucae, latent periods were shorter at low temperatures and longer at high temperatures under fluctuating temperatures, compared with constant temperatures with the same mean. Possible explanations for this observation, including physiological mechanisms (synergistic action of low-temperature fluctuations) and mathematical properties of nonlinear growth (rate summation effect), will be discussed. Differences in development between constant and variable environments will likely be more important for those components of the disease cycle that require days or weeks for completion (e.g. latent period or infectious period) than for processes with short durations (e.g., spore germination, infection or sporulation).

While point measurements of temperature and moisture can be used readily to predict pathogen dynamics with microclimate-based disease warning models, the scaling-up of these measurements for use as inputs in larger scale models (e.g. for regional disease risk assessment or global change impact assessment) has remained difficult. This is because regional environmental heterogeneity results in a range of conditions that prevents any point measure from having general applicability. Approaches for incorporating environmental variability and uncertainty into large-scale disease models that use regional temperature and soil moisture inputs will be discussed.