(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



Excess mortality of infected ectotherms induced by warming depends on pathogen kingdom and evolutionary history [1]

['Jingdi Li', 'Department Of Biology', 'University Of Oxford', 'Oxford', 'United Kingdom', 'Department Of Zoology', 'University Of British Columbia', 'Vancouver', 'Nele Guttmann', 'Department Of Evolutionary']

Date: 2024-11 

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 mostly consisted of Insecta (125 ESs), and aquatic hosts mostly consisted of 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.

PPT PowerPoint slide

PNG larger image

TIFF original image 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.

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.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3002900

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/