Devin Kirk’s research while affiliated with Stanford University and other places

Predicting the effects of climate change on plant disease is critical for protecting ecosystems and food production. Here, we show how disease pressure responds to short‐term weather, historical climate and weather anomalies by compiling a global database (4339 plant–disease populations) of disease prevalence in both agricultural and wild plant systems. We hypothesised that weather and climate would play a larger role in disease in wild versus agricultural plant populations, which the results supported. In wild systems, disease prevalence peaked when the temperature was 2.7°C warmer than the historical average for the same time of year. We also found evidence of a negative interactive effect between weather anomalies and climate in wild systems, consistent with the idea that climate maladaptation can be an important driver of disease outbreaks. Temperature and precipitation had relatively little explanatory power in agricultural systems, though we observed a significant positive effect of current temperature. These results indicate that disease pressure in wild plants is sensitive to nonlinear effects of weather, weather anomalies and their interaction with historical climate. In contrast, warmer temperatures drove risks for agricultural plant disease outbreaks within the temperature range examined regardless of historical climate, suggesting vulnerability to ongoing climate change.

The geographical range of schistosomiasis is affected by the ecology of schistosome parasites and their obligate host snails, including their response to temperature. Previous models predicted schistosomiasis’ thermal optimum at 21.7°C, which is not compatible with the temperature in sub-Saharan Africa (SSA) regions where schistosomiasis is hyperendemic. We performed an extensive literature search for empirical data on the effect of temperature on physiological and epidemiological parameters regulating the free-living stages of S. mansoni and S. haematobium and their obligate host snails, i.e., Biomphalaria spp. and Bulinus spp., respectively. We derived nonlinear thermal responses fitted on these data to parameterize a mechanistic, process-based model of schistosomiasis. We then re-cast the basic reproduction number and the prevalence of schistosome infection as functions of temperature. We found that the thermal optima for transmission of S. mansoni and S. haematobium range between 23.1–27.3°C and 23.6–27.9°C (95% CI) respectively. We also found that the thermal optimum shifts toward higher temperatures as the human water contact rate increases with temperature. Our findings align with an extensive dataset of schistosomiasis prevalence in SSA. The refined nonlinear thermal-response model developed here suggests a more suitable current climate and a greater risk of increased transmission with future warming for more than half of the schistosomiasis suitable regions with mean annual temperature below the thermal optimum.

Temperature can influence mosquito-borne diseases like dengue. These effects are expected to vary geographically and over time in both magnitude and direction and may interact with other environmental variables, making it difficult to anticipate changes in response to climate change. Here, we investigate global variation in temperature–dengue relationship by analyzing published correlations between temperature and dengue and matching them with remotely sensed climatic and socioeconomic data. We found that the correlation between temperature and dengue was most positive at intermediate (near 24°C) temperatures, as predicted from an independent mechanistic model. Positive temperature–dengue associations were strongest when temperature variation and population density were high and decreased with infection burden and rainfall mean and variation, suggesting alternative limiting factors on transmission. Our results show that while climate effects on diseases are context-dependent they are also predictable from the thermal biology of transmission and its environmental and social mediators.

The geographical range of schistosomiasis is affected by the ecology of schistosome parasites and their obligate host snails, including their response to temperature. Previous models predicted schistosomiasis’ thermal optimum at 21.7 °C, which is not compatible with the temperature in sub-Saharan Africa (SSA) regions where schistosomiasis is hyperendemic. We performed an extensive literature search for empirical data on the effect of temperature on physiological and epidemiological parameters regulating the free-living stages of S. mansoni and S. haematobium and their obligate host snails, i.e., Biomphalaria spp. and Bulinus spp., respectively. We derived nonlinear thermal responses fitted on these data to parameterize a mechanistic, process-based model of schistosomiasis. We then re-cast the basic reproduction number and the prevalence of schistosome infection as functions of temperature. We found that the thermal optima for transmission of S. mansoni and S. haematobium range between 23.1-27.3 °C and 23.6-27.9 °C (95 % CI) respectively. We also found that the thermal optimum shifts toward higher temperatures as the human water contact rate increases with temperature. Our findings align with an extensive dataset of schistosomiasis prevalence in SSA. The refined nonlinear thermal-response model developed here suggests a more suitable current climate and a greater risk of increased transmission with future warming for more than half of the schistosomiasis suitable regions with mean annual temperature below the thermal optimum. Authors’ summary In this research, we explored the complex interplay between temperature and the transmission risk of schistosomiasis, a parasitic disease currently affecting over two hundred million people, predominantly in SSA. We developed a novel mathematical model accounting for the multiple positive and negative ways temperature affects the free-living stages of the parasite and its obligate, non-human host, i.e., specific species of freshwater snails. Our models show that schistosomiasis transmission risk peaks at temperatures 1-6°C higher than previously estimated. This indicates that the impact of climate change on schistosomiasis transmission might be more extensive than previously thought, affecting a wide geographic range where mean annual temperatures are currently below the optimal temperature. Our model projections are consistent with the observed temperatures in locations of SSA where schistosomiasis is endemic and data on infection prevalence in the human population are available. These findings suggest that the current climate is conducive to schistosomiasis transmission, and future warming could escalate the risk further, emphasizing the need for targeted interventions in these regions.

Climate change has profound effects on infectious disease dynamics, yet the impacts of increased short-term temperature fluctuations on disease spread remain poorly understood. We empirically tested the theoretical prediction that short-term thermal fluctuations suppress endemic infection prevalence at the pathogen’s thermal optimum. This prediction follows from a mechanistic disease transmission model analyzed using stochastic simulations of the model parameterized with thermal performance curves (TPCs) from metabolic scaling theory and using nonlinear averaging, which predicts ecological outcomes consistent with Jensen’s inequality (i.e., reduced performance around concave-down portions of a thermal response curve). Experimental observations of replicated epidemics of the microparasite Ordospora colligata in Daphnia magna populations indicate that temperature variability had the opposite effect of our theoretical predictions and instead increase endemic infection prevalence. This positive effect of temperature variability is qualitatively consistent with a published hypothesis that parasites may acclimate more rapidly to fluctuating temperatures than their hosts; however, incorporating hypothetical effects of delayed host acclimation into the mechanistic transmission model did not fully account for the observed pattern. The experimental data indicate that shifts in the distribution of infection burden underlie the positive effect of temperature fluctuations on endemic prevalence. The increase in disease risk associated with climate fluctuations may therefore result from disease processes interacting across scales, particularly within-host dynamics, that are not captured by combining standard transmission models with metabolic scaling theory.

Predicting effects of climate change on plant disease is critical for protecting ecosystems and food production. Climate change could exacerbate plant disease because parasites may be quicker to acclimate and adapt to novel climatic conditions than their hosts due to their smaller body sizes and faster generation times. Here we show how disease pressure responds to the anomalous weather that will increasingly occur with climate change by compiling a global database (5380 plant populations; 437 unique plant-disease combinations; 2,858,795 individual plant-disease samples) of disease incidence in both agricultural and wild plant systems. Because wild plant populations are assumed to be adapted to local climates, we hypothesized that large deviations from historical conditions would increase disease incidence. By contrast, since agricultural plants have been transported globally, we did not expect the historical climate where they are currently grown to be as predictive of disease incidence. Supporting these hypotheses, we found that disease outbreaks tended to occur during periods of warm temperatures in agricultural and cool-climate wild plant systems, but also occurred in warm-adapted wild (but not agricultural) plant systems experiencing anomalously cool weather. Outbreaks were additionally associated with higher rainfall in wild systems, especially those with historically wet climates. Our results suggest that historical climate affects susceptibility to disease for wild plant-disease systems, while warming drives risks for agricultural plant disease outbreaks regardless of historical climate.

Background West Nile virus (WNV) is the leading cause of mosquito-borne illness in the continental USA. WNV occurrence has high spatiotemporal variation, and current approaches to targeted control of the virus are limited, making forecasting a public health priority. However, little research has been done to compare strengths and weaknesses of WNV disease forecasting approaches on the national scale. We used forecasts submitted to the 2020 WNV Forecasting Challenge, an open challenge organized by the Centers for Disease Control and Prevention, to assess the status of WNV neuroinvasive disease (WNND) prediction and identify avenues for improvement. Methods We performed a multi-model comparative assessment of probabilistic forecasts submitted by 15 teams for annual WNND cases in US counties for 2020 and assessed forecast accuracy, calibration, and discriminatory power. In the evaluation, we included forecasts produced by comparison models of varying complexity as benchmarks of forecast performance. We also used regression analysis to identify modeling approaches and contextual factors that were associated with forecast skill. Results Simple models based on historical WNND cases generally scored better than more complex models and combined higher discriminatory power with better calibration of uncertainty. Forecast skill improved across updated forecast submissions submitted during the 2020 season. Among models using additional data, inclusion of climate or human demographic data was associated with higher skill, while inclusion of mosquito or land use data was associated with lower skill. We also identified population size, extreme minimum winter temperature, and interannual variation in WNND cases as county-level characteristics associated with variation in forecast skill. Conclusions Historical WNND cases were strong predictors of future cases with minimal increase in skill achieved by models that included other factors. Although opportunities might exist to specifically improve predictions for areas with large populations and low or high winter temperatures, areas with high case-count variability are intrinsically more difficult to predict. Also, the prediction of outbreaks, which are outliers relative to typical case numbers, remains difficult. Further improvements to prediction could be obtained with improved calibration of forecast uncertainty and access to real-time data streams (e.g. current weather and preliminary human cases). Graphical Abstract

Background: West Nile virus (WNV) is the leading cause of mosquito-borne illness in the continental United States. WNV occurrence has high spatiotemporal variation and current approaches for targeted control of the virus are limited, making forecasting a public health priority. However, little research has been done to compare strengths and weaknesses of WNV disease forecasting approaches on the national scale. We used forecasts submitted to the 2020 WNV Forecasting Challenge, an open challenge organized by the Centers for Disease Control and Prevention, to assess the status of WNV neuroinvasive disease (WNND) prediction and identify avenues for improvement. Methods: We performed a multi-model comparative assessment of probabilistic forecasts submitted by 15 teams for annual WNND cases in US counties for 2020, and assessed forecast accuracy, calibration, and discriminatory power. In the evaluation, we included forecasts produced by comparison models of varying complexity as benchmarks of forecast performance. We also used regression analysis to identify modeling approaches and contextual factors that were associated with forecast skill. Results: Simple models based on historical WNND cases generally scored better than more complex models and combined higher discriminatory power with better calibration of uncertainty. Forecast skill improved across updated forecast submissions submitted during the 2020 season. Among models using additional data, inclusion of climate or human demographic data was associated with higher skill, while inclusion of mosquito or land use data was associated with lower skill. We also identified population size, extreme minimum winter temperature, and interannual variation in WNND cases as county-level characteristics associated with variation in forecast skill. Conclusions: Historical WNND cases were strong predictors of future cases with minimal increase in skill achieved by models that included other factors. Although opportunities might exist to specifically improve predictions for areas with large populations and low or high winter temperatures, areas with high case-count variability are intrinsically more difficult to predict. Also, the prediction of outbreaks, which are outliers relative to typical case numbers, remains difficult. Further improvements to prediction could be obtained with improved calibration of forecast uncertainty and access to real-time data streams (e.g., current weather and preliminary human cases).

Parasitism is expected to change in a warmer future, but whether warming leads to substantial increases in parasitism remains unclear. Understanding how warming effects on parasitism in individual hosts (e.g. parasite load) translate to effects on population‐level parasitism (e.g. prevalence, R0) remains a major knowledge gap. We conducted a literature review and identified 24 host–parasite systems that had information on the temperature dependence of parasitism at both individual host and host population levels: 13 vector‐borne systems and 11 environmentally transmitted systems. We found a strong positive correlation between the thermal optima of individual‐ and population‐level parasitism, although several of the environmentally transmitted systems exhibited thermal optima >5°C apart between individual and population levels. Parasitism thermal optima were close to vector performance thermal optima in vector‐borne systems but not hosts in environmentally transmitted systems, suggesting these thermal mismatches may be more common in certain types of host–parasite systems. We also adapted and simulated simple models for both types of transmission modes and found the same pattern across the two modes: thermal optima were more strongly correlated across scales when there were more traits linking individual‐ to population‐level processes. Generally, our results suggest that information on the temperature dependence, and specifically the thermal optimum, at either the individual or population level should provide a useful—although not quantitatively exact—baseline for predicting temperature dependence at the other level, especially in vector‐borne parasite systems. Environmentally transmitted parasitism may operate by a different set of rules, in which temperature dependence is decoupled in some systems, requiring the need for trait‐based studies of temperature dependence at individual and population levels.

Temperature can influence mosquito-borne diseases like dengue. These effects are expected to vary geographically and over time in both magnitude and direction and may interact with other environmental variables, making it difficult to anticipate changes in response to climate change. Here, we investigate global variation in temperature–dengue relationship by analyzing published correlations between temperature and dengue and matching them with remotely sensed climatic and socioeconomic data. We found that the correlation between temperature and dengue was most positive at intermediate (near 24°C) temperatures, as predicted from the thermal biology of the mosquito and virus. Positive temperature–dengue associations were strongest when temperature variation and population density were high and decreased with infection burden and rainfall mean and variation, suggesting alternative limiting factors on transmission. Our results show that while climate effects on diseases are context-dependent they are also predictable from the thermal biology of transmission and its environmental and social mediators.