Quantifying climatic control of fungal spore phenology with Bayesian statistical modeling

Background/Question/Methods

: Airborne spores from fungi are a major source of allergy, causing respiratory illness in humans. They are also biological aerosol particles that can influence climate such as through cloud formation. In order to inform public health and atmospheric modeling, it is important to understand spore phenology, or the timing of spore release, and how it is controlled by environmental conditions. Previous local studies suggested spore abundance is affected by meteorology, increasing during periods with greater air temperature, relative humidity, or precipitation. Nevertheless, it is unclear how climatic factors drive fungal spore phenology on a large scale.We address this gap using a continental-scale long-term fungal spore count dataset from the US National Allergy Bureau. We built and fitted a Bayesian model based on a double logistic curve to characterize the change of fungal spore abundance over a year. We employed a hierarchical model structure to allow interannual variations while keeping generally consistent phenology in each site. Parameters from the model, such as maximum and minimum spore abundance, the start and end of spore season, and the rates of spore increase and decrease, were explained with climatic variables including mean annual temperature and total annual precipitation, by fitting generalized linear models.

Results/Conclusions

: The novel hierarchical Bayesian model characterized fungal spore phenology with high accuracy (Pearson correlation > 0.9). There are large variations in fungal spore phenology over space and time. These differences can partly be explained by differences in temperature and precipitation. Taking the dominant family Cladosporiaceae as an example, a higher temperature is correlated with lower maximum spore abundance, earlier start and later end of spore season, and slower decline at the end of season, while a higher precipitation is correlated with earlier end of season and faster increase at the start of season. The influence of climate seems to explain the geographic pattern of fungal spore phenology across sites better than explaining interannual variations within a site.Fungal spore phenology is an interesting yet under-studied aspect of phenology. With large observational data across the continental US and the creative use of a phenology model, we statistically described the temporal variations of fungal spore abundance. Moreover, we revealed how climatic factors drive spatial variations in fungal spore phenology. On one hand, such climate-phenology coupling allows extrapolation to areas without extensive spore monitoring. On the other hand, ecological understandings of fungal spore phenology help us anticipate climate change impacts on public health.