1. Anthropogenic activities have altered historical disturbance regimes, and understanding the mechanisms by which these shifting perturbations interact is essential to predicting where they may erode ecosystem resilience. Emerging infectious plant diseases, caused by human translocation of nonnative pathogens, can generate ecologically-damaging forms of novel biotic disturbance. Further, abiotic disturbances, such as wildfire, may influence the severity and extent of disease-related perturbations via their effects on the occurrence of hosts, pathogens, and microclimates; however, these interactions have rarely been examined.

2. The disease “sudden oak death” (SOD), associated with the introduced pathogen Phytophthora ramorum, causes acute, landscape-scale tree mortality in California’s fire-prone coastal forests. Here, we examined interactions between wildfire and the biotic disturbance impacts of this emerging infectious disease. Leveraging long-term datasets that describe wildfire occurrence and P. ramorum dynamics across the Big Sur region, we modeled the influence of recent and historical fires on epidemiological parameters, including pathogen presence, infestation intensity, reinvasion, and host mortality.

3. Past wildfire altered disease dynamics and reduced SOD-related mortality, indicating a negative interaction between these abiotic and biotic disturbances. Frequently-burned forests were less likely to be invaded by P. ramorum, had lower incidence of host infection, and exhibited decreased disease-related biotic disturbance, which was associated with reduced occurrence and density of epidemiologically-significant hosts. Following a recent wildfire, survival of mature bay laurel, a key sporulating host, was the primary driver of P. ramorum infestation and reinvasion, but younger, rapidly regenerating host vegetation capable of sporulation did not measurably influence disease dynamics. Notably, the effect of P. ramorum infection on host mortality was reduced in recently-burned areas, indicating that the loss of mature host canopies may temporarily dampen pathogen transmission and “release” susceptible species from significant inoculum pressure.

4.Synthesis: Cumulatively, our findings indicate that fire history has contributed to heterogeneous patterns of biotic disturbance and disease-related decline across this landscape, via changes to the both the occurrence of available hosts and the demography of epidemiologically-important host populations. These results highlight that human-altered abiotic disturbances may play a foundational role in structuring infectious disease dynamics, contributing to future outbreaks, and driving biotic disturbance regimes.

This dataset is a subset of a long-term forest monitoring study to examine the impacts of Phytopthora ramorum on the coast redwood and mixed evergreen forests of Big Sur, California. The data was collected over 3 field survey periods, from 2006-7, 2010-11, and 2013-14, and a subset of the included survey plots were impacted by the 2008 Basin Complex and Chalk fires. More detailed information about the variables collected and the field and laboratory methods used to collect these variables can be found in the associated manuscript. The plot-level variables included in this dataset have been aggregated from measurements collected at the stem and tree level from individual tagged trees. The included tree and stem measurements have been subsetted to include only tanoak and oak trees, which are the focus of the associated manuscript.

Metadata, variable descriptions

Relevant abbreviations:
## UMCA = bay laurel, Umbellularia californica
## LIDE = tanoak, Notholithocarpus densiflorus (previous genus = Lithocarpus)
## QUAG = coast live oak, Quercus agrifolia
## 06/2006 = data collected in the 2006-07 surveys
## 2010/10 = data collected in the 2010-11 surveys
## 2013/14 = data collected in the 2013-14 surveys

Variable descriptions:
# Big Sur Plot = plot identification number
# Forest Alliance Type = index for Redwood forest (2) or Mixed Evergreen forest (1)
## tmax = mean 30 year dry-season maximum temperature, derived from PRISM
## totalfire.nb = # of fires experienced since 1950, excluding the Basin complex fire (which occurred before 2006-2007 sampling points)
## burned1950 = whether or not a plot has burned since 1950
## LiveBA0607 = total live woody plant basal area in cm^2 (for stems >1cm dbh) in 2006-2007 samples
## Hosts06 = total host tree basal area in cm^2 (LIDE, Quercus sp, and UMCA stems >1cm dbh) in 2006-2007
## UMCAnonzero = presence/absence of bay laurel at plot level in 2006-2007 samples
## LIDEnonzero = presence/absence of tanoak at plot level in 2006-2007 samples
## UMCA.BA.LIVE.0607 = total bay laurel basal area in cm^2 in 2006-07 samples
## LIDE.BA.LIVE.0607 = total tanoak basal area in cm^2 in 2006-07 samples

Github repository for analysis described in manuscript: 

https://github.com/absimler/postfireepidemiology