Estimating thermal performance curves from repeated field observations.

Area of Science:

• Ecology
• Physiology
• Climate Change Biology

Background:

• Estimating thermal performance is vital for predicting species' responses to climate change.
• Laboratory studies often fail to capture real-world environmental and individual variations affecting thermal performance.
• Existing models have limited predictive ability for field conditions.

Purpose of the Study:

• To develop and validate a novel model for estimating thermal performance curves using field observations.
• To incorporate environmental and individual variation into thermal performance models.
• To compare the accuracy of field-derived versus laboratory-derived thermal performance models.

Main Methods:

• Developed a Bayesian framework model for estimating thermal performance curves from repeated field observations.
• Utilized Markov Chain Monte Carlo (MCMC) sampling for parameter estimation and uncertainty propagation.
• Validated the model using simulated data and applied it to individual growth data from wild trout.

Main Results:

• The field-based model demonstrated accurate, precise, and unbiased parameter estimates.
• Field-derived models showed higher out-of-sample predictive ability compared to laboratory-derived models.
• Laboratory models produced biased predictions for field performance and potentially underestimated climate change impacts.

Conclusions:

• Field-based thermal performance models provide more realistic estimates of organismal performance under environmental conditions.
• Laboratory-derived models may overestimate thermal tolerance and underestimate the negative effects of warming temperatures.
• The developed model advances thermal ecology by providing a more accurate approach to predicting climate change effects on wild populations.