Interactions between genetics and the environment determine tolerance to ash dieback disease in an ecosystem model

Pathogens, insects, and parasites (PIPs) are a common disturbance in many forest ecosystems that lead to reduced tree growth and increased mortality over time. While native PIPs can be key to maintaining biodiversity, PIP outbreaks are becoming more common and devastating to forests in a changing climate. Outbreaks of invasive PIPs pose a particular risk as host trees may have no evolved defences. Process based models are ideal for studying and predicting forest responses to disturbance as they can be used to test hypotheses about processes, test conditions that are challenging to experimentally create, and can inform further experimentation. PIPs have been incorporated in some forest ecosystem models to determine the effects of a PIP on tree growth and mortality. However, the host response to PIPs, including defences that lead to resistance or tolerance of a particular pest, have not been represented in process-based models, despite their demonstrated role in determining the resulting severity of PIP impacts on tree growth and mortality. Modelling both sides of the host-PIP interaction will provide more accurate forecasts of tree mortality and growth in the face of disturbance and allow us to test hypotheses about host defence processes and tolerance to disease. We develop a process-based model to quantify the impacts of pathogen infection on tree growth and function, while incorporating host defence and tolerance mechanisms, to simulate the effects of the widespread invasive pathogen, Hymenoscyphus fraxineus, on ash (Fraxineus excelsior) across Europe. H. fraxineus, is the causal agent of ash dieback disease that has led to the steep decline of native ash trees in Ireland, UK and Europe, killing up to 85% of trees in some areas. A small percentage of trees are genetically tolerant to the disease, but tolerance levels are variable and environmental conditions, tree age, and pathogen load may all further influence the level of susceptibility. Combining the model with ash tree trial data, we show that disease tolerance has a genetic component, but even among genetically tolerant trees, high disease pressure in wet environments may outweigh genetic tolerance. Further the effects of the environment and site characteristics on disease severity are mediated primarily through effects on pathogen abundance rather than tree growth. In addition to providing insights into drivers of ash dieback tolerance, our study showcases the power of process-based models combined with field trial and genetic data to reveal aspects of plant function that cannot be inferred from data alone.