Predicting growth responses to climate change in the barnacle Balanus glandula: a test of temperature sensitivity using a Dynamic Energy Budget model

Background/Question/Methods

 Accurate predictions of responses to temperature change require mechanistically modeling the physiological effects of temperature on an organism.   Dynamic Energy Budget (DEB) models have been used to successfully predict the growth of many species, including several marine invertebrates. DEB models calculate energy balance as a function of body size, temperature, and food supply, and predict growth by tracking somatic and reproductive energy investments.  However for these models to accurately predict responses to climate change, they must correctly capture the thermal dependence of physiological processes within the organism.  Current DEB models assume a single temperature dependence applies to all physiological processes, including feeding and respiration, but this may not be realistic.  We developed a DEB model for the barnacle Balanus glandula.   We tested the assumption of a single temperature dependence factor by fitting the model to growth data from a 37-day field experiment conducted at two different water temperatures (11 and 14 ^oC ). The majority of model parameters were calculated from empirical laboratory observations, but feeding and respiration parameters were fit iteratively and allowed to vary between the two temperatures.   Based on laboratory measurements, we hypothesized that respiration would be more sensitive to warming temperatures than feeding.

Results/Conclusions

Model fits were within 1 standard error of empirical length and mass measurements at both temperatures.  Comparison of the best fit feeding and respiration parameters between the two temperature treatments indicated that both physiological processes were thermally sensitive, and that the degree of sensitivity varied by process.  The respiration parameter for the best model fit at 14 ^oC was roughly 2.5 times greater than the respiration rate in the model fit to the 11 ^oC  treatment, corresponding to a Q-10 of 10.  However, the relative increase in the feeding parameter was 3.3, indicating a Q-10 of 13.2.  Our results suggest that the assumption of a unified temperature dependence for all processes within the organism may be an oversimplification.  Future studies should explore this question in a wider range of organisms and over a broader range of temperatures.