Table 2. Summary of minimum adequate linear mixed effects (lme) model results for upper lethal temperature (ULT) as a function of log10(experimental rate of temperature change) and habitat (subtidal, low intertidal and upper intertidal).
Fixed-effects | df | Contrast coefficient | Standard error | t-value | P-value | Random-effects | % variance |
reference | 49 | 40.41 | 0.54 | 75.13 | <0.0001 | S (77) | 62.40 |
rate | 49 | −2.66 | 0.27 | −9.94 | <0.0001 | Residual | 37.60 |
habitat(LIT) | 31 | 2.45 | 0.78 | 3.15 | 0.0036 | ||
habitat(UIT) | 31 | 6.54 | 0.78 | 8.40 | <0.0001 | ||
rate: habitat(LIT) | 49 | −0.13 | 0.39 | −0.33 | 0.75 | ||
rate: habitat(UIT) | 49 | −1.57 | 0.41 | −3.85 | 0.0003 |
Treatment contrasts indicate the effect of each parameter level on the reference level (subtidal). Species (S) was retained as a random effect on the intercept. Effect types are intercept (unshaded) and slope (shaded). The model-averaged coefficient estimates and 95% confidence intervals for all parameters included in the full model are in Table S3.
lme(UTL∼rate*habitat,random = S)
AICc = 335.14
LIT = lower intertidal, UIT = upper intertidal, AICc = Akaike's information criterion corrected for finite sample sizes, df = degrees of freedom.