Corrigendum

In the 2008 paper, “Predicting natural mortality rates of plants and animals” in Ecology Letters, 11, 710–716, data for a small number of fish and invertebrate species were inadvertently misreported. This mistake does not change the overall patterns observed, or the conclusions of the paper. It does, however, have minor effects on the parameter estimates. Here we present the corrected figures (Figs 1, 2a,b, and 3) and statistical results. We also have corrected the online supporting information table—Appendix S1.

Figure 1.  Temperature dependence of body mass-corrected natural mortality rates, ln(Zm^1/4), for invertebrates and fishes. The solid line indicates the estimated slope and intercept, and the dashed line indicates the predicted slope of −0.65 eV for animals based on the average activation energy of respiration.).

Figure 2.  Body mass dependence for temperature-corrected natural mortality rates, ln(Z.e ^{[E / k [1 / T−1 / T20oC]}), for: a. invertebrates, b. fish. Solid lines indicate the estimated slope and intercept, and the dashed lines indicate the predicted slope of −1 / 4 based on the mass dependence of individual metabolic rate.

Figure 3.  Body mass dependence for temperature-corrected natural mortality rates, ln(Z.e ^{[E / k [1 / T−1 / T20oC]}), for plants and animals. Solid line indicates the estimated slope and intercept and the dashed lines indicates the predicted slope of −0.25 based on the mass dependence of individual metabolic rate.

Prediction 1: For both fish and invertebrates, the natural logarithm of mass-corrected natural mortality rate is significantly negatively correlated with inverse absolute temperature. The common slope estimated for the two groups gives activation energy of -0.69 eV (95% CI = −0.78 to −0.60 eV), which is close to the predicted value of −0.65 eV (Table 1; Fig. 1). Fitting the data with the predicted slope of −0.65 explains 35% of the variation in body mass-corrected natural mortality for the pooled data and 45% and 17% of the variation in for fish and invertebrates respectively.

Table 1.   Parameter estimates and confidence intervals estimated via analysis of covariance for the data shown in Figs 1–3

Independent variable	Groups	Predicted Slope	Fixed Slope Intercept	Fitted Slope	Fitted Intercept
Numbers in parentheses represent lower and upper bounds of the 95% confidence intervals.
Temperature, (1/k [1/T-1/T20oC])	Fish	−0.65	0.58(0.45 to 0.72)	−0.83(−0.95 to −0.72)	0.74(0.56 to 0.91)
	Invertebrates	−0.65	0.23(0.07 to 0.39)	−0.47(−0.76 to −0.18)	0.02(−0.42 to 0.47)
Body mass ln(m)	Fish	−0.25	0.58(0.45 to 0.72)	−0.27 (−0.35 to −0.20)	0.72(0.34 to 1.09)
	Invertebrates	−0.25	0.23(0.07 to 0.39)	−0.18 (−0.24 to −0.12)	0.45(0.16 to 0.76)

Prediction 2: The natural logarithm of the temperature-corrected mortality rate for all plants and animals (pooled) is negatively correlated with the natural logarithm of body mass (Regression: F = 3761, P < 0.0001), with a fitted slope (−0.22, 95% CI = −0.23 to –0.21) that is very close to the predicted value of –0.25 (Fig. 2a,b and Fig. 3). Moreover, the model, assuming a slope of −1/4 explains 63% of the variation in natural mortality rates for the pooled data. Within taxonomic groups, the slopes for fish and invertebrates are changed from those reported in McCoy and Gillooly (2008) but fall within the range of −0.18 to −0.27 (Table 1). The model, assuming a slope of −1/4 explains 36% and 46% of variation in natural mortality rates for fish and invertebrates respectively.