Advertisement

  • Loading metrics

Open Access

Peer-reviewed

Short Reports

Short Reports offer exciting novel biological findings that may trigger new research directions.

See Journal Information »

Excess mortality of infected ectotherms induced by warming depends on pathogen kingdom and evolutionary history

  • Jingdi Li ,

    Roles Conceptualization, Data curation, Formal analysis, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing

    jingdi.li@biology.ox.ac.uk

    Affiliations Department of Biology, University of Oxford, Oxford, United Kingdom, Department of Zoology, University of British Columbia, Vancouver, Canada

    ⨯
  • Nele Guttmann,

    Roles Data curation, Investigation, Writing – review & editing

    Affiliations Department of Evolutionary and Integrative Ecology, Leibniz Institute of Freshwater Ecology and Inland Fisheries (IGB), Berlin, Germany, Department of Biology, Chemistry, Pharmacy, Institute of Biology, Freie Universität Berlin (FU), Berlin, Germany

    ⨯
  • Georgia C. Drew,

    Roles Conceptualization, Data curation, Validation, Writing – review & editing

    Affiliations Department of Biology, University of Oxford, Oxford, United Kingdom, Collegium Helveticum, The joint Institute for Advanced Studies (IAS) of the ETH Zurich, The University of Zurich, &The Zurich University of the Arts, Zurich, Switzerland

    ⨯
  • Tobias E. Hector,

    Roles Validation, Writing – review & editing

    Affiliation Department of Biology, University of Oxford, Oxford, United Kingdom

    ⨯
  • Justyna Wolinska,

    Roles Conceptualization, Funding acquisition, Supervision, Validation, Writing – review & editing

    Affiliations Department of Evolutionary and Integrative Ecology, Leibniz Institute of Freshwater Ecology and Inland Fisheries (IGB), Berlin, Germany, Department of Biology, Chemistry, Pharmacy, Institute of Biology, Freie Universität Berlin (FU), Berlin, Germany

    ⨯
  • Kayla C. King

    Roles Conceptualization, Funding acquisition, Supervision, Validation, Writing – original draft, Writing – review & editing

    Affiliations Department of Biology, University of Oxford, Oxford, United Kingdom, Department of Zoology, University of British Columbia, Vancouver, Canada, Department of Microbiology & Immunology, University of British Columbia, Vancouver, Canada

    ⨯

Abstract

Climate change is causing extreme heating events and can lead to more infectious disease outbreaks, putting species persistence at risk. The extent to which warming temperatures and infection may together impair host health is unclear. Using a meta-analysis of >190 effect sizes representing 101 ectothermic animal host–pathogen systems, we demonstrate that warming significantly increased the mortality of hosts infected by bacterial pathogens. Pathogens that have been evolutionarily established within the host species showed higher virulence under warmer temperatures. Conversely, the effect of warming on novel infections—from pathogens without a shared evolutionary history with the host species—were more pronounced with larger differences between compared temperatures. We found that compared to established infections, novel infections were more deadly at lower/baseline temperatures. Moreover, we revealed that the virulence of fungal pathogens increased only when temperatures were shifted upwards towards the pathogen thermal optimum. The magnitude of all these significant effects was not impacted by host life-stage, immune complexity, pathogen inoculation methods, or exposure time. Overall, our findings reveal distinct patterns in changes of pathogen virulence during warming. We highlight the importance of pathogen taxa, thermal optima, and evolutionary history in determining the impact of global change on infection outcomes.

Citation: Li J, Guttmann N, Drew GC, Hector TE, Wolinska J, King KC (2024) Excess mortality of infected ectotherms induced by warming depends on pathogen kingdom and evolutionary history. PLoS Biol 22(11): e3002900. https://doi.org/10.1371/journal.pbio.3002900

Academic Editor: Lauren Buckley, University of Washington Seattle Campus: University of Washington, UNITED STATES OF AMERICA

Received: April 29, 2024; Accepted: October 14, 2024; Published: November 18, 2024

Copyright: © 2024 Li et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The data that support the findings of this study are openly available in Figshare at https://doi.org/10.6084/m9.figshare.22060646.v7.

Funding: JW received funding from Beethoven Life-1 (grant no. WO 1587/9-1) grant, funded by the German Science Foundation (DFG, https://www.dfg.de/en). KCK received funding from the Natural Environment Research Council (NE/X000540/1, https://www.ukri.org/councils/nerc/), a Philip Leverhulme Prize, and an NSERC Canada Excellence Research Chair. JL received funding from Pembroke College Oxford Graduate Scholarship. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Abbreviations: AICc, Akaike’s information criterion corrected for small sample size; ES, effect size; GLM, generalized linear model; RR, risk ratio; VIF, variance inflation factor

Introduction

Climate change is resulting in more extreme heating events [1]. Temperature affects all aspects of biology, directly or indirectly impacting the physiology and life-history of all organisms [2]. Typically for ectotherms, environmental warming changes their body temperatures, influencing their physiological performance and fitness [3]. Shifting global temperature patterns are also changing the geographic distribution of infectious diseases, generating novel transmission opportunities (i.e., host jumping) [4,5].

Positive relationships between warmer temperatures and disease severity (or pathogen virulence) have been observed in a diversity of systems [6–11]. Rising temperature can compromise animal immune responses, leading to increased susceptibility to infections [12–15]. Moreover, by causing stress and disrupting homeostasis, pathogen infection can reduce the upper temperatures tolerated by hosts, contributing to worsened health outcomes under warming [16]. Temperature can also directly mediate pathogen virulence [17–21]. Higher temperatures can up-regulate the expression of virulence factors in the coral pathogen Vibrio coralliilyticus [20]. For the human pathogen Shigella sonnei, higher temperatures enhance its immune evasion abilities by increasing virulence protein synthesis [21]. Warming may also enhance pathogen metabolism, resulting in faster within-host growth and greater host harm. Such virulence-enhancing effects might be especially true for many bacterial pathogens, as their population growth rates can increase exponentially with temperature, up to their thermal limits [22]. Whether warmer temperatures result in sicker hosts across host–pathogen systems remains unclear [23,24]. In some host–pathogen systems, virulence is unaffected or even reduced by warming [23–28], emphasizing the complex interactions between temperature and disease outcomes.

Both host and pathogen responses to temperature change depend on their thermal tolerance. This trait is characterized by the critical thermal maximum, minimum, and optimum—the upper, lower, and optimal temperatures at which the organism can function [3,29,30]. A mismatch in host and pathogen thermal optima can shape infection outcomes. Cold-adapted pathogens might be less harmful to warm-adapted hosts at higher temperatures [31,32]. Also, eukaryotes generally have a lower thermal tolerance compared to prokaryotes [33], a difference which might make eukaryotic pathogens (e.g., fungi, nematodes) more sensitive to warmer temperatures.

Understanding the broad ecological consequences of elevated temperature on pathogen virulence (or infection-induced host death herein) is crucial for projections of future wildlife disease dynamics and species persistence. If higher temperatures generally exacerbate the severity of infection, climate change could worsen disease outcomes for wild animal populations. Forecasts of species persistence in a changing world might need to account for rates of infection and disease severity [8,34]. However, if virulence is broadly unaffected or context-dependent [23], the health risks posed by warming and infectious disease might need to be evaluated on a case-by-case basis. Any general relationship between warming and virulence would have profound consequences for animal conservation [34,35] and potentially human health [4,36].

In this study, we conducted a meta-analysis to investigate the generality of the relationship between warming and pathogen virulence. We searched the published literature for experimental studies and included a breadth of ectothermic animal host–pathogen systems. We used the most widely reported quantitative form of virulence: pathogen-induced host death [37]. We focused on host death under infection and warming as this measure is pertinent to projections of species extinction and biodiversity loss [38]. We calculated effect sizes via relative risk ratios of death at experimentally high and low temperatures. We separately collected thermal optima data for all included pathogens. We tested whether biological traits (e.g., host type, pathogen type, host–pathogen evolutionary history, pathogen thermal optimum), the scale of temperature increase, and experimental variables (pathogen inoculation method, pathogen dosage, infection exposure/measurement time) could explain variation in the temperature-virulence relationship.

Materials and methods

Literature search and data collection

A first-round literature search was performed in Leibniz Institute of Freshwater Ecology and Inland Fisheries (IGB, Germany) in March 2021 on Web of Science Core Collection database. We used search query [’parasit*’ OR ’pathogen*’ AND ’temperature’ AND ’virulen*’] within “Topic” field (encompassing title, abstract, keyword plus, and author keywords), to identify studies reporting the effect of temperature on pathogen virulence. Then a second-round search using the same keywords was conducted in University of Oxford in April 2022 to update the data set. We used host mortality to quantify pathogen virulence. We excluded studies on parasitoids as host mortality under infection would always be 100%. Our main inclusion criteria were: (i) the pathogen was tested in living animal hosts rather than on cells or host tissues; (ii) more than 1 temperature was used in host infection experiment and temperature was the only changing variable in the tested comparisons; (iii) pathogen virulence was measured by assessing host mortality rate under different temperatures; (iv) host mortality data (mean mortality rate and sample size, where sample size is defined as the number of individual host used in each infection treatment) were available publicly or shared by the authors by May 2022.

Following our inclusion criteria, we retrieved data from a total of 60 research articles (see PRISMA flow-chart in S1 Fig and PRISMA-EcoEvo Checklist in S1 Text), published between 1996 and 2021 (summary of studies and moderator variables are provided in S1 Table). In individual articles, data on host mortality were extracted manually from main text, tables, supplementary information, from figures by using WebPlotDigitizer [39], or raw data made available from the authors by request. To maximize the ecological relevance, we imposed a series of restrictions to exclude data from extreme high or low temperatures and extreme high pathogen dosages that are unlikely to happen in nature. Firstly, if more than 5 temperatures were tested, data were extracted from the second lowest and second highest temperatures (when between 2 to 4 temperatures were tested, data were extracted from the highest and lowest temperatures). Secondly, if multiple pathogen dosages were used, data of the middle value (or middle minus one value, in case of even number) was extracted. Thirdly, host mortality data was extracted from the end time point of the assay. However, if mortality began to appear in controls, data were extracted from the time point with control mortality close to or at zero. We excluded certain experimental conditions to capture the most substantial changes in host fitness by selecting conditions with the largest temperature changes, while avoiding the confounding effects of extreme temperatures and dosages. In studies that included multiple temperature levels, the extreme lowest and highest temperatures often represent conditions that are rarely encountered by hosts and pathogens in natural settings, typically resulting in host death or loss of pathogen infectivity confounding the infection effects. We ensured that the exclusion of all these extreme treatments were based on the following criteria: (i) host mortality was primarily caused by temperature rather than infection; (ii) no pathogen growth or infectivity at these extreme temperatures; and (iii) the original study explicitly showed that the extreme temperatures were outside the realistic range for the host–pathogen system (details in S2 Table). As a result, we focused on infection-induced host mortality, rather than death due to temperature alone. Hosts not exposed to pathogens had no mortality (e.g., Jiang and colleagues [7]), very low mortality (<10%, e.g., Brand and colleagues [10]), or host mortality in the infection treatment was corrected for control mortality following Abbott’s (1952) formula (e.g., Bugeme and colleagues [11]). Fifteen effect sizes from 6 studies were discarded due to large host mortality (>10%) in control groups.

When host mortality was expressed as a percentage, we estimated the number of alive and dead individuals in both temperature groups by multiplying percentage of host mortality by the number of host individuals tested. We otherwise obtained the number of alive and dead hosts from the raw data sent by authors. These data were used to calculate the risk ratio (RR) effect size, using escalc function in “metafor” R package [40]. RRs represent the relative risk of warming on infected hosts’ survival. An overall total of 192 effect sizes from 60 studies were used across 56 pathogens and 50 host species (101 host–pathogen combinations, S2 Fig).

Moderator variables

To evaluate the impact of various factors on pathogen virulence in response to rising temperatures, we included several moderators in our analysis, including host, pathogen, and experimental variables. To explore whether pathogen type could predict effect size estimates, we recorded whether the pathogen used in the study was bacteria, fungi, nematode, virus, or protist. As the number of effect sizes in protist was low (1 ES for protist), our analysis focused on bacteria, fungi, nematode, and virus pathogens.

We collected data on the optimal growth temperature (Topt) for well-studied pathogens as virulence can correlate with pathogen growth rates [41]. These data were collected for 39 pathogen species included in the meta-analysis from published literature (S3 Table, Topt for in vitro growth/fitness was measured for nonviral pathogens). We ordered the experimental temperatures from low to high. Using known thermal optima for these pathogens, we defined whether experimental warming was towards or away from their Topt. “Towards Topt” was used when the temperature range in the study was below Topt. “Away from Topt” was used when the study temperature range was above Topt. When the low and high temperatures in the study straddled Topt, we used “Range includes Topt”. Specific details on pathogen acclimation to their thermal optima prior to the experiments were generally not provided within individual articles, although pathogens were shown to be cultured within their known thermal ranges.

We tested whether infection outcomes at increased temperature depended on host–pathogen evolutionary history. We defined a moderator based on whether the interaction was “established” (system originally collected from the wild and with shared evolutionary history, see Bally and Garrabou [42] as an example), “semi-established” (host infected by the pathogen in the wild, but combination in the study were not co-isolated, e.g., Wekesa and colleagues [43]), and “novel” (no known record of an infection between the pathogen and the host in nature, e.g., Ekesi and colleagues [44]). The “novel” associations might represent opportunistic infections or a host shift for the pathogen.

We also tested whether host traits—including host life-stage (“adult,” “larva,” “pupae,” and “juvenile”) and host type with different immune complexity (“vertebrate” or “invertebrate”)—were relevant moderator variables. The availability of host thermal optima data was limited, as well as information on whether hosts were acclimated to their thermal optima or other environments prior to the infection experiment. In most instances, hosts were either recently collected from the field or reared in the laboratory for multiple generations. To account for the potential acclimation effects, we included “host source” as a moderator, distinguishing between wild-collected and lab-reared hosts. Lastly, we collected and evaluated experimental variables, including temperature span (difference between high temperature and low temperature used in the study), infection exposure duration (measured in days), pathogen dosage (within each pathogen type), and pathogen inoculation method (“injection” or “not injection”). These experimental protocols were evaluated given their potential for a bias in extrapolations to nature [45].

Statistical analysis

To assess the influence of sample size and correction methods on effect sizes, we fitted sample size or correction method as a moderator, and checked that neither sample size nor the methods used to correct for control mortality influenced the effect sizes (sample size p = 0.9630, correction method p > 0.52 for all levels). All categorical moderators were assessed for multicollinearity by calculating the variance inflation factor (VIF, VIF function in “regclass” R package) when fitted together in a regression model and no multicollinearity was observed (S4 Table, GVIF^(1/(2*Df)) < = 2.5 for all moderators included). Correlation of continuous variables were assessed for Pearson correlation (cor.test function in R) and the possibility of multicollinearity was excluded (S4 Table, p > 0.05 for all pairwise correlations).

We coded the data accordingly with RR>1 (logRR>0), indicating greater host mortality at higher temperatures (higher risk associated with higher temperatures), or RR<1 (logRR<0) with lower host mortality at higher temperatures. No relationship between temperature and virulence was indicated by RR = 1 (logRR = 0). Summary effect sizes were calculated by establishing a multilevel random effects meta-regression model using rma.mv function in “metafor” R package. To account for taxonomic non-independence, we included species-level random effects for both hosts and pathogens. These random effects were correlated according to a taxonomic correlation matrix derived from a taxonomic tree obtained from the NCBI Taxonomy common tree [46]. When species-level taxonomy was unavailable, genus-level information was used instead. A taxonomic tree was used instead of a phylogenetic tree because, in many cases, species-level phylogenies could not be reliably resolved from the published literature, and substituting species with their relatives would likely introduce bias. Additionally, we included non-taxonomic species-level random effect to capture heterogeneity in effect sizes arising from differences between species that are not associated with evolutionary relatedness. This approach was initially proposed by Schmidt and colleagues [47] and has shown to be a best practice by Cinar and colleagues [48]. To address statistical non-independence due to repeated measures within studies, we included study ID as a random effect in the model.

To determine the importance of moderators, we first fitted a full single joint random effects model without interaction terms, with effect sizes as the response variable. Model selection was performed using glmulti function in glmulti R package, with the fitfunction customized to implement the multilevel meta-analytic random-effects model. We selected the best-fitting models using Akaike’s information criterion corrected (AICc) for small sample size. Models with a ΔAICc < 2 were considered to have substantial support and were included in the final model set. We performed model averaging across this set of top models to account for model uncertainty by calculating the relative importance of each moderator. Additionally, model-averaged estimates and their 95% confidence intervals were obtained for each moderator to assess their significance. Moderators with low importance values (model-averaged importance <0.25 and 95% CI included zero) were considered to have little support for inclusion in the model.

The remaining moderators were fitted in a joint model and interaction terms were included. Similarly, model selection was performed, and the top model set (ΔAICc < 2) was obtained. Since model averaging is generally not recommended with interactions in the model [49], the best model with the simplest structure and the lowest AICc was reported in the main text. The alternative top models can be found in S4 Table. The significance of a moderator was determined based on its consistent significance across all top models. To maximize the number of retained effect sizes, moderators including pathogen thermal optima, infection dosage, and exposure time, which had missing data in some studies, were not included in the single joint model. Instead, multiple regression models were employed to explore the heterogeneity they explained. For multiple regression models, both taxonomic and non-taxonomic structures for host and pathogen species were included as random effects, as well as Study ID. Interactions between these moderators were tested based on a priori hypotheses (all hypotheses and modeling results in S4 Table). Lastly, host mortality data was fitted in generalized linear model (GLM) with binomial distribution to compare host baseline mortality among host–pathogen systems with different evolutionary histories.

Results and discussion

Increased host mortality due to bacterial infections at elevated temperatures

We firstly tested whether higher temperatures generally increased host mortality during infection (i.e., the proportion of dead individuals in the infected host population). A total of 1,649 studies were identified and screened through Web of Science. After title and abstract screening, 1,363 studies were excluded. We then assessed 286 articles in full text, of which 219 were excluded based on our criteria. Ultimately, 67 studies met all the inclusion criteria, having conducted infection experiments at different temperatures. However, 7 of these studies were excluded due to data unavailability despite attempts to contact the authors. As a result, data from 60 studies were included in the final analysis (S1 Fig). These studies included 50 ectothermic animal host species and 56 pathogen species, resulting in 101 host–pathogen combinations (see S2 Fig), and provided 192 effect sizes (ESs). Our data set contained a variety of pathogen types including bacteria, fungi, nematodes, and viruses from diverse geographic regions (S3 Fig). We included diverse host–pathogen systems spanning terrestrial (42 ESs) and aquatic habitats (150 ESs). Terrestrial hosts were predominantly Insecta (125 ESs), and aquatic hosts were predominantly fish (17 ESs) and mollusca (14 ESs) (S2 Fig). Notably, bacterial pathogens were predominantly studied in aquatic animals, while fungal and nematode pathogens were mainly tested in insect hosts (S2 Fig).

Under warming, uninfected host mortality across studies remained zero or low (<10%). Infected hosts did not suffer significantly greater death rates with increase in temperature (Fig 1A, summary RR = 1.64, 95% CI [0.91, 2.95], p = 0.1004). Given the substantial heterogeneity among the effect sizes (I2 = 98.74, p < 0.0001), we investigated whether biological and abiotic factors contributed to the observed variation in effect sizes. We specifically tested whether the relationship between temperature and virulence varied by (i) pathogen types (bacteria, fungi, nematoda, viruses); (ii) host–pathogen evolutionary history (established versus novel); (iii) host types and immune complexity (vertebrate versus invertebrate); (iv) host life-stage (e.g., adult versus larva); (v) wild-collected versus lab-reared hosts; (vi) pathogen inoculation method (injection versus not injection); and (vii) the span of temperature changes.

thumbnail
Download:
Fig 1. Results of the meta-analysis on the impact of warming on pathogen-induced host death.

(A) Summary ES (RR) overall is not significantly >1 (logRR>0, indicating positive relationship between temperature and virulence). Bold black diamond with 95% confidence interval represents summary ES, and individual effect size are displayed as jittered points. Point size indicates sample size (number of host individuals in the infection treatment) used for effect size calculation. Effect sizes differed by (B) pathogen type and (C) host–pathogen evolutionary history. Star shown on the bottom right corner indicate significance of moderators. In (B) and (C), jittered points are colored by pathogen type. The data and code needed to generate this figure can be found in https://doi.org/10.6084/m9.figshare.22060646.v7. ES, effect size; RR, risk ratio.

https://doi.org/10.1371/journal.pbio.3002900.g001

The best model incorporated fixed effects of pathogen type, host–pathogen evolutionary history, host life stage and temperature changes, along with interactions between host–pathogen evolutionary history and pathogen type, as well as between host–pathogen evolutionary history and temperature changes. We found that the impact of higher temperatures on host mortality varied depending upon the type of pathogen infecting. Effect sizes calculated from bacterial pathogens were larger than from viruses (virus es = −3.0, 95% CI [−4.58, −1.41], p = 0.0002). Subgroup analyses showed that among all types of pathogens, only bacteria exhibited a significant positive summary effect size (Fig 1B, bacteria RR = 2.25, p = 0.0002; fungi RR = 0.96, p = 0.8153, nematoda RR = 1.61, p = 0.0716, virus RR = 1.28, p = 0.6868). As most of the effect sizes for host–bacterial systems were calculated from aquatic hosts, our results suggest that bacterial pathogens might pose significantly greater risk during warming, particularly to aquatic animal species. Increased disease outbreaks for aquatic animals, especially fishes, has been an overwhelming issue for the aquaculture industry, causing considerable economic damage and threats to human health [50]. Bacterial diseases are common in aquaculture. We suggest that future disease management should also consider the aggravating effects of warming on the severity of these epidemics.

Our results showed that warming increased host mortality to a larger extent in established host-pathogen systems, compared with semi-established or novel systems (Fig 1C, semi-established es = −4.31, 95% CI [−6.84, −1.77], p = 0.0009; novel es = −4.03, 95% CI [−6.49, −1.58], p = 0.0013). Interestingly, in novel systems, the average host mortality was already higher even at low/baseline temperatures (Fig 2A, novel versus established es = 1.56, p = 0.0378), but not at high temperatures (Fig 2B). This high level of virulence might be due to pathogen maladaptation during host shifts [51], suggesting that emerging infections could be highly harmful at present temperatures, and less likely to be worsened during global climate change. On the contrary, the increased severity of endemic infections should be considered when predicting wildlife health under warming. We further showed that the lack of overall virulence-enhancing effect by warming for fungal pathogens was not because they caused higher host mortality at baseline temperatures, though fungal pathogens were mostly from novel systems (S4 Table, p = 0.1308). We also observed a significant interaction between pathogen type and evolutionary history (novel:virus es = 5.08, 95% CI [2.38, 7,78], p = 0.0002). Given that novel viral infections were represented by only a few effect sizes (N = 2) in our data set, generalizing this interaction effect requires caution.

thumbnail
Download:
Fig 2.

Host killing at (A) low/baseline temperatures and (B) high temperatures, for host–pathogen systems with different evolutionary histories. Under low/baseline temperatures (A), significantly higher host killing is observed in “novel” compared to “established” systems. The data and code needed to generate this figure can be found in https://doi.org/10.6084/m9.figshare.22060646.v7.

https://doi.org/10.1371/journal.pbio.3002900.g002

We found that the effect of warming on host mortality was mediated by the extent of temperature change in novel host–pathogen systems. The larger the temperature change, the higher the difference in host mortality between experimental temperatures. A significant interaction was observed between host–pathogen evolutionary history and the span of temperature change (S5A Fig, novel:temperature_span es = 0.30, 95% CI [0.07, 0.52], p = 0.009; semi-established:temperature_span es = 0.36, 95% CI [0.11, 0.60], p = 0.004). This finding was further supported by a re-analysis of a single pathogen species—Beauveria bassiana, comprising 25 effect sizes (all from novel/semi-established systems), which demonstrated that higher level of host mortality during infection by this fungal pathogen was associated with greater temperature changes (S5B Fig, p = 0.0011).

Experimental temperature increases included in our analysis ranged from 3°C to 25°C. This range is relevant to the upper boundary of the predicted temperature rise on Earth: 4°C increase in annual mean surface temperature by 2100 and a 14.1°C increase by 2300 [52]. If no actions are taken to alleviate climate change, our findings suggest that the outcomes of emerging or novel infectious disease could worsen to a greater extent in regions of intense average warming and more frequent extreme thermal events [53]. Future research should explore how host responses, such as evolutionary adaptation in shorter-lived species, acclimatization in longer-lived species, and migration strategies under warming, could mitigate these adverse effects [54].

We found little evidence that the magnitude of the temperature-host death association was influenced by host life-stage or pathogen inoculation method (S6 Fig and S4 Table, p > 0.05 for all individual and interactive effects). We did not find a consistent effect of exposure time on warming-induced changes in host mortality, across host–pathogen systems (S4 Table, hypothesis 2). Other moderators, such as host immune complexity and host source, were excluded from the model selection process due to their low importance and wide confidence intervals that spanned zero (S4 Fig and S4 Table). However, a significant interaction was observed between pathogen dosage and host–pathogen evolutionary history (S7 Fig, dosage:semi-established es > 0, p = 0.007), suggesting that higher pathogen dosage was associated with increased host mortality in semi-established systems under warming, compared to established systems.

Direction of temperature change is vital for fungal infection outcomes

We hypothesized that warming-induced increase in host mortality observed in certain sub-groups, such as bacterial infections, may be driven by faster pathogen growth or higher within-host loads. To test this hypothesis, we estimated pathogen thermal optima (Topt) for 39 well-studied species, based on their in vitro growth rate data (S3 Table). However, our analysis showed no significant effect of temperature change directions on effect sizes for bacterial infections (S4 Table, p = 0.4626). Expanding the analysis to other pathogen types, we found that the direction of the temperature shift relative to Topt was vital for outcomes of fungal infections (Fig 3, Towards Topt es = 0.97, p = 0.037). Unlike bacterial pathogens, fungi generally did not cause higher host mortality under warmer temperatures (Fig 1B). This outcome supported the observation that fungi are more heat-sensitive than bacteria [55]. However, host death caused by fungi increased when the temperature was shifted towards fungal Topt (Fig 3). Notably, as 92% of fungal effect sizes were calculated from novel host–fungi systems, our results might help predict patterns of emerging fungal diseases. Fungal species often have limited growth at elevated temperatures which constrains their ability to infect mammals [56]. However, our finding suggests that if temperatures shift towards pathogen Topt, fungi could become more virulent. These results are consistent with observations that Batrachochytrium dendrobatidis (Bd)—a deadly fungal pathogen driving the extinction of amphibian species—is more prevalent at highland localities where temperatures approach the pathogen’s Topt [57]. At higher latitudes, persistently warmer temperatures caused by climate change are contributing to the ongoing expansion of fungal pathogens such as Coccidioides, Blastomyces, Histoplasma, and Sporothrix [58,59]. With continued warming in these regions, emerging fungal infections could spread faster and be more severe, posing a greater threat to endangered ectotherms including amphibians and reptiles [60].

Host thermal tolerance might also play an important role in shaping infection outcomes under warming conditions [31]. Unfortunately, we were unable to obtain data on host thermal optima primarily due to lack of experimental evidence. Host immune response is another critical factor. For example, host immunopathology induced by infection can be heightened under elevated temperatures, causing host cell and tissue damage and worsening infection outcomes [61,62]. It can be difficult to distinguish the adverse effects of host immunopathology from direct effects of pathogens under warming [63]. Given these complexities, further experiments on the impact of temperature on immune activation and immunopathology are needed for disentangling the host and pathogen-imposed adverse effects. Tackling the impact of global climate change on wildlife health requires an understanding of the mechanisms that might increase the likelihood of host mortality when periods of extreme warming occur.

thumbnail
Download:
Fig 3. RR varied with direction of temperature change.

(A) Schematic illustration of the classification for directions of temperature shift related to pathogen thermal optima (Topt). The curve represents a generic thermal performance curve with Topt as the optimal growth temperature. The direction of a temperature shift is represented by arrows. “Towards Topt”: the temperature range in the study was below Topt. “Away from Topt”: the temperature range was above Topt. “Range includes Topt”: the low and high temperatures in the study straddled Topt. (B) Comparisons between effect sizes across temperature shift directions, within each pathogen type. Individual effect sizes are displayed as jittered points. Effect sizes for fungal pathogens have significant larger increase from “away from Topt” to “towards Topt” (p = 0.0372). The data and code needed to generate this figure can be found in https://doi.org/10.6084/m9.figshare.22060646.v7.

https://doi.org/10.1371/journal.pbio.3002900.g003

Conclusions and future perspective

Our study underscores the need for a broader range of host–pathogen systems to be included in future experimental tests of the temperature–virulence relationship. The current research landscape is skewed, while invertebrate hosts included in our analysis ranged from terrestrial (e.g., insects, mite, etc.; 78.1% of the effect sizes) to aquatic (e.g., corals, oysters, etc.; 12% of the effect sizes), there were fewer experimental studies involving aquatic vertebrate hosts (fishes, amphibians, 9.9% of the effect sizes) and none with terrestrial vertebrates. This bias is likely caused by a high prevalence of studies related to agriculture (e.g., used Arthropoda hosts) or aquaculture (e.g., used fish hosts). Our study focused on ectotherms in which both the host and pathogens are directly exposed to warming. Whether endothermic hosts—with more complex thermal regulation mechanisms—have similar responses requires further study. Indeed, most emerging human pathogens originate in other vertebrates, especially mammals [64–66]. Further research on the impact of higher temperatures in vertebrate–pathogen systems is essential for robust predictions on the outcomes of infections with zoonotic potential under climate change.

Both ecological and evolutionary perspectives are essential for understanding the impacts of rising temperatures on ecosystems. Our study revealed the effect of temperature on infection outcomes within a host’s lifetime. Over longer evolutionary timescales, the difference between host and pathogen ability to adapt with thermal changes may be increasingly important to infection outcomes. Shorter generation times might help pathogens adapt faster to temperature shifts than their hosts [67], leading to increased pathogen growth within hosts. To test whether the pattern observed on ecological timescales will remain the same across generations, longitudinal studies (e.g., experimental evolution) could be used to track the potential for host and pathogen adaptation under temperatures relevant to climate change projections. Slow temperature increases may accommodate host acclimation (in ecological time) or adaptation (in evolutionary time), potentially mitigating the negative effect of warming on infection outcomes [68].

Parasites and pathogens are ubiquitous. It is thus critical to understand how warming might impact virulence to better understand the consequences for animal species during global climate change [69–73]. Our meta-analysis provides evidence that warming, and the degree and direction of temperature increase, can impact mortality in infected animal populations. We took a conservative approach by only focusing on host mortality. However, if morbidity were also considered, there might be more widespread deleterious effects of warming on infected animals. Our findings highlight distinct consequences of warming on bacterial and fungal infections, as well as for hosts in different habitats. If concomitant with bacterial pathogen epidemics, warming could exacerbate aquatic animal population declines. While if temperatures create more optimal conditions for pathogen growth, emerging fungal diseases in terrestrial animals are more of a concern. There are important consequences in this temperature–virulence relationship for species at risk and ecosystems sensitive to biodiversity loss. Human activities are warming the Earth at an unprecedented rate [74] and leading to an increased frequency and intensity of temperature extremes [52]. The current global effort to restrict warming to 1.5°C above pre-industrial levels by 2050 [52] may also have the added benefit of limiting the lethal impacts of infectious diseases.

Supporting information

S1 Text. (Word document).

List of studies included PRISMA-EcoEvo Checklist.

https://doi.org/10.1371/journal.pbio.3002900.s001

(DOCX)

S1 Fig. PRISMA-style flow diagram of screening and inclusion of journal articles.

The 1,649 records from a Web of Science search were screened based on our criteria, resulting in the 60 total articles included in the meta-analysis; n represents the number of articles.

https://doi.org/10.1371/journal.pbio.3002900.s002

(PDF)

S2 Fig. The evolutionary history of the 101 host–pathogen combinations included in the meta-analysis.

Each is colored by the evolutionary history of the host–pathogen system (established, semi-established, and novel as defined in the main text). The data and code needed to generate this figure can be found in https://doi.org/10.6084/m9.figshare.22060646.v7.

https://doi.org/10.1371/journal.pbio.3002900.s003

(PDF)

S3 Fig. Geographical locations of pathogens with known collection sites.

Some pathogens were collected years ago and may have adapted to lab environment. Different colors represent pathogen taxa (shown on genus-level when available). Collection sites data were variable in their geographic specificity, ranging from precise coordinates to broader regions such as cities and countries, depending on details reported in the original literature. The proportions in each pie chart represents proportion of pathogen taxa from each collection site. Map lines delineate study areas and do not necessarily depict accepted national boundaries. The data and code needed to generate this figure can be found in https://doi.org/10.6084/m9.figshare.22060646.v7. Base layer of the map is from the Natural Earth 1:50 m map (https://www.naturalearthdata.com/), under a non-exclusive license.

https://doi.org/10.1371/journal.pbio.3002900.s004

(PDF)

S4 Fig. Model-averaged importance of moderators.

Moderators including “host immune complexity” and “host collected field or lab” with low importance values (model-averaged importance <0.25 and 95% CI included zero) were considered to have little support for inclusion in the model. The data and code needed to generate this figure can be found in https://doi.org/10.6084/m9.figshare.22060646.v7.

https://doi.org/10.1371/journal.pbio.3002900.s005

(PDF)

S5 Fig. Association between the span of temperature change and effect sizes.

(A) Temperature change is interacting with host–pathogen evolutionary history to influence effect sizes. (B) Increases in effect sizes are associated with larger temperature changes across studies of the pathogen Beauveria bassiana (k: 25 effect sizes across multiple hosts). The data and code needed to generate this figure can be found in https://doi.org/10.6084/m9.figshare.22060646.v7.

https://doi.org/10.1371/journal.pbio.3002900.s006

(PDF)

S6 Fig.

Effect sizes were not impacted by (A) host life-stage, or (B) pathogen was inoculated by injection or other means. Individual effect sizes are displayed as jittered points. The data and code needed to generate this figure can be found in https://doi.org/10.6084/m9.figshare.22060646.v7.

https://doi.org/10.1371/journal.pbio.3002900.s007

(PDF)

S7 Fig. The impacts of pathogen dosage on effect sizes.

Positive interaction effect between pathogen dosage and semi-established system is observed. The data and code needed to generate this figure can be found in https://doi.org/10.6084/m9.figshare.22060646.v7.

https://doi.org/10.1371/journal.pbio.3002900.s008

(PDF)

S1 Table. Summary of included studies and collected moderator variables.

https://doi.org/10.1371/journal.pbio.3002900.s009

(XLSX)

S2 Table. Effect sizes calculated for meta-analysis.

https://doi.org/10.1371/journal.pbio.3002900.s010

(XLSX)

S3 Table. Pathogen in vitro Topt data collected from published literatures.

https://doi.org/10.1371/journal.pbio.3002900.s011

(XLSX)

S4 Table. Results from modeling and statistical analyses.

https://doi.org/10.1371/journal.pbio.3002900.s012

(XLSX)

References

  1. 1. Dosio A, Mentaschi L, Fischer E, Wyser K. Extreme heat waves under 1.5°C and 2°C global warming, ENVIRONMENTAL RESEARCH LETTERS, ISSN 1748-9326, 2018, 13 (5), p. 054006, JRC108180.
  2. 2. Lagerspetz KY. Temperature effects on different organization levels in animals. Symp Soc Exp Biol. 1987;41:429–449. pmid:3332496
  3. 3. Huey RB, Kearney MR, Krockenberger A, Holtum JAM, Jess M. Williams SE. Predicting organismal vulnerability to climate warming: roles of behaviour, physiology and adaptation. Philos Trans R Soc Lond B Biol Sci. 2012;367:1665–1679.
  4. 4. Williams ES, Yuill T, Artois M, Fischer J, Haigh SA. Emerging infectious diseases in wildlife. Revue scientifique et technique (International Office of Epizootics). 2002;21(1):139–157. pmid:11974625
  5. 5. Roberts KE, Hadfield JD, Sharma MD, Longdon B. Changes in temperature alter the potential outcomes of virus host shifts. PLoS Pathog. 2018;14(10):e1007185. pmid:30339695
  6. 6. Vezzulli L, Previati M, Pruzzo C, Marchese A, Bourne DG, Cerrano C. VibrioSea Consortium Vibrio infections triggering mass mortality events in a warming Mediterranean Sea. Environ Microbiol. 2010;12(7):2007–2019. pmid:20370818
  7. 7. Jiang Q, Shi L, Ke C, You W, Zhao J. Identification and characterization of Vibrio harveyi associated with diseased abalone Haliotis diversicolor. Dis Aquat Organ. 2013;103(2):133–139. pmid:23548363
  8. 8. Paull SH, Johnson PT. Experimental warming drives a seasonal shift in the timing of host-parasite dynamics with consequences for disease risk. Ecol Lett. 2014;17:445–453. pmid:24401007
  9. 9. Mayers TJ, Bramucci AR, Yakimovich KM. Case RJ A Bacterial Pathogen Displaying Temperature-Enhanced Virulence of the Microalga Emiliania huxleyi. Front Microbiol. 2016;7:892. pmid:27379036
  10. 10. Brand MD, Hill RD, Brenes R, Chaney JC, Wilkes RP, Grayfer L, et al. Water Temperature Affects Susceptibility to Ranavirus. Ecohealth. 2016;13(2):350–359. pmid:27283058
  11. 11. Bugeme DM, Knapp M, Boga HI, Wanjoya AK, Maniania NK. Influence of temperature on virulence of fungal isolates of Metarhizium anisopliae and Beauveria bassiana to the two-spotted spider mite Tetranychus urticae. Mycopathologia. 2009;167(4):221–227. pmid:18987988
  12. 12. Moriyama M, Ichinohe T. High ambient temperature dampens adaptive immune responses in influenza A virus infection. Proc Natl Acad Sci U S A. 2019;116:31183125.
  13. 13. Regnier JA, Kelley KW. Heat- and cold-stress suppresses in vivo and in vitro cellular immune responses of chickens. Am J Vet Res. 1981;42:294–299. pmid:7258780
  14. 14. Lecchi C, Rota N, Vitali A, Ceciliani F, Lacetera N. In vitro assessment of the effects of temperature on phagocytosis, reactive oxygen species production and apoptosis in bovine polymorphonuclear cells. Vet Immunol Immunopathol. 2016;182:89–94. pmid:27863557
  15. 15. Lacetera N, Bernabucci U, Scalia D, Ronchi B, Kuzminsky G, Nardone A. Lymphocyte functions in dairy cows in hot environment. Int J Biometeorol. 2005;50:105–110. pmid:15991017
  16. 16. Hector TE, Hoang KL, Li J, King KC. Symbiosis and host responses to heating. Trends Ecol Evol. 2022;37:611–624. pmid:35491290
  17. 17. Laine AL. Pathogen fitness components and genotypes differ in their sensitivity to nutrient and temperature variation in a wild plant–pathogen association. J Evol Biol. 2007;20:2371–2378. pmid:17956398
  18. 18. Vale PF, Little TJ. Measuring parasite fitness under genetic and thermal variation. Heredity. 2009;103:102–109. pmid:19455181
  19. 19. Allen DE, Little TJ. Dissecting the effect of a heterogeneous environment on the interaction between host and parasite fitness traits. Evol Ecol. 2011;25:499–508.
  20. 20. Kimes NE, Grim CJ, Johnson WR, Hasan NA, Tall BD, Kothary MH, et al. Temperature regulation of virulence factors in the pathogen Vibrio coralliilyticus. ISME J. 2012;6(4):835–846. pmid:22158392
  21. 21. Lam O, Wheeler J, Tang CM. Thermal control of virulence factors in bacteria: a hot topic. Virulence. 2014;5(8):852–862. pmid:25494856
  22. 22. Qiu Y, Zhou Y, Chang Y, Liang X, Zhang H, Lin X, et al. The Effects of Ventilation, Humidity, and Temperature on Bacterial Growth and Bacterial Genera Distribution. Int J Environ Res Public Health. 2022;19(22):15345. pmid:36430064
  23. 23. Lafferty KD, Mordecai EA. The rise and fall of infectious disease in a warmer world. F1000 Res. 2016;5:2040.
  24. 24. Karvonen A, Rintamaki P, Jokela J, Valtonen ET. Increasing water temperature and disease risks in aquatic systems: Climate change increases the risk of some, but not all, diseases. Int J Parasitol. 2010;40(13):1483–1488. pmid:20580904
  25. 25. Delisle L, Petton B, Burguin JF, Morga B, Corporeau C, Pernet F. Temperature modulate disease susceptibility of the Pacific oyster Crassostrea gigas and virulence of the Ostreid herpesvirus type 1. Fish Shellfish Immunol. 2018;80:71–79. pmid:29859311
  26. 26. De Silva PM, Chong P, Fernando DM, Westmacott G, Kumar A. Effect of Incubation Temperature on Antibiotic Resistance and Virulence Factors of Acinetobacter baumannii ATCC 17978. Antimicrob Agents Chemother. 2017;62(1):e01514–17. pmid:29061747
  27. 27. Kerchev IA, Kryukov VY, Yaroslavtseva ON, Polovinko GP, Tokarev YS, Glupov VV. The first data on fungal pathogens (ascomycota, hypocreales) in the invasive populations of four-eyed fir bark beetle Polygraphus proximus Blandf. Russ J Biol Invasions. 2017;8:34–40.
  28. 28. López JR, Lorenzo L, Marcelino-Pozuelo C, Marin-Arjona MC, Navas JI. Pseudomonas baetica: pathogenicity for marine fish and development of protocols for rapid diagnosis. FEMS Microbiol Lett. 2017;364(3):10.1093/femsle/fnw286. pmid:28011698
  29. 29. Chown SL, Hoffmann AA, Kristensen TN, Angilletta MJ Jr, Chr. Stenseth N, Pertoldi C. Adapting to climate change: a perspective from evolutionary physiology. Climate Res. 2010;43:3–15.
  30. 30. Hoffmann AA, Sørensen JG, Loeschcke V. Adaptation of Drosophila to temperature extremes: bringing together quantitative and molecular approaches. J Therm Biol. 2003;28:175–216.
  31. 31. Cohen JM, Venesky MD, Sauer EL, Civitello DJ, McMahon TA, Roznik EA, et al. The thermal mismatch hypothesis explains host susceptibility to an emerging infectious disease. Ecol Lett. 2017;20(2):184–193. pmid:28111904
  32. 32. Cohen JM, Sauer EL, Santiago O, Spencer S, Rohr JR. Divergent impacts of warming weather on wildlife disease risk across climates. Science. 2020;370:abb1702. pmid:33214248
  33. 33. Hickey DA, Singer GA. Genomic and proteomic adaptations to growth at high temperature. Genome Biol. 2004;5(10):117. pmid:15461805
  34. 34. Pounds JA, Bustamante MR, Coloma LA, Consuegra JA, Fogden MP, Foster PN, et al. Widespread amphibian extinctions from epidemic disease driven by global warming. Nature. 2006;439(7073):161–167. pmid:16407945
  35. 35. Boyd IL, Freer-Smith PH, Gilligan CA, Godfray HCJ. The consequence of tree pests and diseases for ecosystem services. Science. 2013;342:1235773. pmid:24233727
  36. 36. Patz JA, Campbell-Lendrum D, Holloway T, Foley JA. Impact of regional climate change on human health. Nature. 2005;438:310–317. pmid:16292302
  37. 37. Cressler C, Mcleod D, Rozins C, Van Den Hoogen J, Day T. The adaptive evolution of virulence: A review of theoretical predictions and empirical tests. Parasitology. 2016;143(7):915–930. pmid:26302775
  38. 38. Fisher MC, Henk DA, Briggs CJ, Brownstein JS, Madoff LC, McCraw SL, et al. Emerging fungal threats to animal, plant and ecosystem health. Nature. 2012;484:186–194. pmid:22498624
  39. 39. Rohatgi A. WebPlotDigitizer (Version 3.9) [Computer software]. 2015. Available from: http://arohatgi.info/WebPlotDigitizer.
  40. 40. Viechtbauer W. Conducting Meta-Analyses in R with the metafor Package. J Stat Softw. 2010;36(3):1–48.
  41. 41. Pandey A, Dawson DE. The shapes of virulence to come. Evolution, medicine, and public health, 2019(1), 3. pmid:30740224
  42. 42. Bally M, Garrabou J, Thermodependent bacterial pathogens and mass mortalities in temperate benthic communities: a new case of emerging disease linked to climate change. Glob Chang Biol. 2007;13:2078–2088.
  43. 43. Wekesa VW, Moraes GJ, Ortega EM, Delalibera I Jr. Effect of temperature on sporulation of Neozygites floridana isolates from different climates and their virulence against the tomato red spider mite, Tetranychus evansi. J Invertebr Pathol. 2010;103(1):36–42. pmid:19836399
  44. 44. Ekesi S, Maniania NK, Ampong-Nyarko K. Effect of Temperature on Germination, Radial Growth and Virulence of Metarhizium anisopliae and Beauveria bassiana on Megalurothrips sjostedti. Biocontrol Sci Techn. 1999;9:177–185.
  45. 45. Hairston NG. Ecological experiments: purpose, design and execution. Cambridge University Press, Cambridge, UK. 1989.
    • 46. Schoch CL, Ciufo S, Domrachev M, Hotton CL, Kannan S, Khovanskaya R, et al. NCBI Taxonomy: a comprehensive update on curation, resources and tools. Database: the journal of biological databases and curation, 2020, baaa062. 2020. pmid:32761142
    • 47. Schmidt JP, Davies TJ, Farrell MJ. Opposing macroevolutionary and trait-mediated patterns of threat and naturalisation in flowering plants. Ecol Lett. 2021;24(6):1237–1250. pmid:33786974
    • 48. Cinar O, Nakagawa S, Viechtbauer W. Phylogenetic multilevel meta-analysis: A simulation study on the importance of modelling the phylogeny. Methods Ecol Evol. 2022;13(2):383–395.
    • 49. Banner KM, Higgs MD. Considerations for assessing model averaging of regression coefficients. Ecol Appl. 2017;27(1):78–93. pmid:27874997
    • 50. Dayana Senthamarai M, Rajan MR, Bharathi PV. Current risks of microbial infections in fish and their prevention methods: A review. Microb Pathog. 2023;185:106400. pmid:37863271
    • 51. Longdon B, Hadfield JD, Day JP, Smith SCL, McGonigle JE, Rodrigo Cogni R, et al. The causes and consequences of changes in virulence following pathogen host shifts. PLoS Pathog. 2015;11:e1004728. pmid:25774803
    • 52. Arias PA, Bellouin N, Coppola E, Jones RG, Krinner G, Marotzke J, et al. Technical Summary. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Masson-Delmotte V, Zhai P, Pirani A, Connors SL, Péan C, Berger S, et al., editors. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. 2021:33−144. https://doi.org/10.1017/9781009157896.002
      • 53. Rangwala I, Sinsky E, Miller JR. Amplified warming projections for high altitude regions of the northern hemisphere mid-latitudes from CMIP5 models. Environ Res Lett. 2013;8(2):024040.
      • 54. Bennett JM, Sunday J, Calosi P. et al. The evolution of critical thermal limits of life on Earth. Nat Commun. 2021;12:1198. pmid:33608528
      • 55. Alster CJ, Weller ZD, von Fischer JC. A meta-analysis of temperature sensitivity as a microbial trait. Glob Chang Biol. 2018;24:4211–4224. pmid:29888841
      • 56. Garcia-Solache MA, Casadevall A. Global warming will bring new fungal diseases for mammals. MBio. 2010;1(1):e00061–e00010. pmid:20689745
      • 57. Bosch J, Carrascal LM, Durán L, Walker S, Fisher MC. Climate change and outbreaks of amphibian chytridiomycosis in a montane area of Central Spain; is there a link?. Proc Biol Sci. 2007;274(1607):253–260. pmid:17148254
      • 58. Friedman DZP, Schwartz IS. Emerging fungal infections–New patients, new patterns, and new pathogens. J Fungi. 2019;5(3):67. pmid:31330862
      • 59. Gadre A, Enbiale W, Andersen LK, Coates SJ. The effects of climate change on fungal diseases with cutaneous manifestations: A report from the International Society of Dermatology Climate Change Committee. J Clim Chang Health. 2022;6:100156.
      • 60. Cohen JM, Civitello DJ, Venesky MD, McMahon TA, Rohr JR. An interaction between climate change and infectious disease drove widespread amphibian declines. Glob Chang Biol. 2019;25:927–937. pmid:30484936
      • 61. Evans SS, Repasky EA, Fisher DT. Fever and the thermal regulation of immunity: the immune system feels the heat. Nat Rev Immunol. 2015;15(6):335–349. pmid:25976513
      • 62. Dittmar J, Janssen H, Kuske A, Kurtz J, Scharsack JP. Heat and immunity: an experimental heat wave alters immune functions in three-spined sticklebacks (Gasterosteus aculeatus). J Anim Ecol. 2014;83:744–757. pmid:24188456
      • 63. Graham AL, Allen JE, Read AF. Evolutionary Causes and Consequences of Immunopathology. 2005;36:373–397.
      • 64. Cleaveland S. Diseases of humans and their domestic mammals: pathogen characteristics, host range and the risk of emergence. Philos Trans R Soc Lond B Biol Sci. 2001;356:991–999. pmid:11516377
      • 65. Woolhouse MEJ, Gowtage-Sequeria S. Host range and emerging and reemerging pathogens. Emerg Infect Dis. 2005;11:1842–1847. pmid:16485468
      • 66. Taylor LH, Latham SM, Woolhouse MEJ. Risk factors for human disease emergence. Philos Trans R Soc Lond B Biol Sci. 2001;356:983–989. pmid:11516376
      • 67. Gillooly JF, Brown JH, West GB, Savage VM, Charnov EL. Effects of size and temperature on metabolic rate. Science. 2001;293:2248–2251. pmid:11567137
      • 68. Schaum C-E, Buckling A, Smirnoff N, Yvon-Durocher G. Evolution of thermal tolerance and phenotypic plasticity under rapid and slow temperature fluctuations. Proc R Soc B. 2022:. pmid:35919998
      • 69. Garamszegi LZ. Climate change increases the risk of malaria in birds. Glob Chang Biol. 2011;17:1751–1759.
      • 70. Altizer S, Ostfeld RS, Johnson PT, Kutz S, Harvell CD. Climate change and infectious diseases: from evidence to a predictive framework. Science (New York, N.Y.). 2013;341(6145):514–519. pmid:23908230
      • 71. Smith KF, Acevedo-Whitehouse K, Pedersen AB. The role of infectious diseases in biological conservation. Anim Conserv. 2009;12:1–12.
      • 72. Rohr JR, Cohen JM. Understanding how temperature shifts could impact infectious disease. PLoS Biol. 2020;18(11):e3000938. pmid:33232316
      • 73. Harvell CD, Mitchell CE, Ward JR, Altizer S, Dobson AP, Ostfeld RS, et al. Climate warming and disease risks for terrestrial and marine biota. Science (New York, N.Y.). 2002;296(5576):2158–2162. pmid:12077394
      • 74. IPCC, 2021: Summary for Policymakers. In: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Masson-Delmotte V, Zhai P, Pirani A, Connors SL, Péan C, Berger S, et al., editors. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. p. 3−32. https://doi.org/10.1017/9781009157896.001