Appendix B. Stable isotope analysis of lizard tails, leaves, and seaweed, and estimation of percentage of marine-based prey consumed by lizards using mixing models.
Sample preparation and analysis
Lizard tails, terrestrial plant leaves, and seaweed were dried at 55 °C for 48 h. Plant leaves and seaweed were ground to fine powder. Lipids were extracted from tail tips using methanol-chloroform (2:1 by volume). Isotope and leaf nitrogen content analysis was performed with a continuous-flow isotope ratio mass spectrometry using an ANCA-GSL elemental analyzer interfaced to a GEO 20-20 IRMS (Sercon Ltd., Cheshire, UK) at Faculty of Engineering at University of Yamanashi, Japan. The standard deviation of replicates of organic standard was 0.14‰ for δ13C. All stable-isotope values were reported in the d notation: δ13C = ([13C/12Csample / 13C/12Cstandard] - 1) × 1000. No significant difference in δ13C was found between the tail and whole body of the Anolis lizards (Takimoto et al. 2008).
Baseline values and mixing models
To obtain baseline values for terrestrial and marine resources, we conducted carbon stable isotope analysis on samples of leaves and seaweed collected in each experimental plot. Baseline δ13C values did not differ significantly between treatments (Fig. B1; overall treatment effect in repeated measures ANOVA on terrestrial plants, F1,5 = 0.25, P < 0.64; on seaweed, F1,5 = 0.89, P < 0.39). These values were used in a simple mixing model (Post 2002) which estimates the percentage of marine-based prey consumed by lizards in each plot as follows: % Marine diet = 100((δ13CL – ΔC) – δ13CT) / (δ13CS– δ13CT), where δ13CL, δ13CT and δ13CS are the mean carbon isotope ratios for lizards, terrestrial plants and seaweed, respectively, and DC is the trophic fractionation of carbon. There is some debate over carbon fractionation. Post (2002) found that it was minimal (~0) but more recent studies found that it can be substantial for herbivorous insects (McCutchan et al. 2003, Langellotto et al. 2005, Mooney and Tillberg 2005, Spence and Rosenheim 2005). We conducted stable isotope analysis on Conocarpus erectus leaves and herbivorous beetles collected on C. erectus leaves from 3 different islands. Mean δ13C values were 3.0 to 4.6 units higher for beetles than for leaves (Fig. B2). We used the average difference (3.8 units) as ΔC in the mixing model which assumes that fraction of all terrestrial lizard prey is this value and no fractionation in marine prey or in lizards. We also computed estimates using the same mixing model but with no fractionation (ΔC = 0). The results show that the difference between treatments is very similar with and without fractionation, except the estimates of percent marine-based diets are higher for the model without fractionation (Fig. B1); this pattern is very similar to that of the raw δ13C values for lizards presented in the manuscript (Fig. 3C). Results of the model with fractionation are obviously better because some estimates in the model without fractionation were greater than 100%. Estimates with fractionation are reasonable but we caution that they may be inaccurate due to a lack of information on all of the potential sources of prey in the lizard diets. The main purpose of this presentation is to verify that there was indeed a shift in the lizard diets in the subsidized plots even when accounting for variation among plots in the baseline isotopic values.
Results of stable isotope analysis of all plots
One of the three experimental islands was omitted from the analysis given in the manuscript because it did not have at least 3 lizards in each plot during the first half of the experiment; we opted to present the results for only the other two islands for which we had a complete time series of data, as we did for all of the other response variables. Lizards became more abundant on all islands during the second half of the experiment. Here we present the data for all plots during this period. In May and September 2007, lizard δ13C values were significantly higher in subsidized plots than in removals (Fig. B3).
FIG. B1. Carbon stable isotope analysis of terrestrial plants and seaweed and estimates of percentage of marine prey in lizard diets in experimental plots with seaweed added and removed (mean ±1 SE). * P < 0.05, significant within-date ANOVA treatment effect (stepdown Bonferroni adjusted).
FIG. B2. Carbon stable isotope analysis of C. erectus leaves and herbivorous beetles collected on three different islands. Symbols and bars are the means ±1 SE.
FIG. B3. Carbon stable isotope analysis of lizard diets in experimental plots with seaweed added and removed (mean ±1 SE). Significance level of within-date ANOVA treatment effect (stepdown Bonferroni adjusted): * P < 0.05, ** P < 0.01.
Langellotto, G. A., J. A. Rosenheim, and M. R. Williams. 2005. Enhanced carbon enrichment in parasitoids (Hymenoptera): a stable isotope study. Annals of the Entomological Society of America 96:205–213.
McCutchan, J. H. Jr., W. M. Jr. Lewis, C. Kendall, and C. C. McGrath. 2003. Variation in trophic shift for stable isotope ratios of carbon, nitrogen, and sulfur. Oikos 102:378–390.
Mooney, K. A., and C. V. Tillberg. 2005. Temporal and spatial variation to ant omnivory in pine forests. Ecology 86:1225–1235.
Post, D. M. 2002. Using stable isotopes to estimate trophic position: models, methods, and assumptions. Ecology 83:703–718.
Spence, K. O., and J. A. Rosenheim. 2005. Isotopic enrichment in herbivorous insects: a comparative field-based study of variation. Oecologia 146:89–97.
Takimoto, G., D. A. Spiller, and D. M. Post. 2008. Ecosystem size, but not disturbance, determines food-chain length on islands of the Bahamas. Ecology 89:3001–3007.