Warming, eutrophication, and predator loss amplify subsidies between aquatic and terrestrial ecosystems
The exchange of organisms and energy among ecosystems has major impacts on food web structure and dynamics, yet little is known about how climate warming combines with other pervasive anthropogenic perturbations to affect such exchanges. We used an outdoor freshwater mesocosm experiment to investiga...
Saved in:
| Published in | Global change biology Vol. 18; no. 2; pp. 504 - 514 |
|---|---|
| Main Authors | , , , , , |
| Format | Journal Article |
| Language | English |
| Published |
Oxford
Blackwell Publishing Ltd
01.02.2012
Wiley-Blackwell |
| Subjects | |
| Online Access | Get full text |
Cover
Loading…
| Summary: | The exchange of organisms and energy among ecosystems has major impacts on food web structure and dynamics, yet little is known about how climate warming combines with other pervasive anthropogenic perturbations to affect such exchanges. We used an outdoor freshwater mesocosm experiment to investigate the interactive effects of warming, eutrophication, and changes in top predators on the flux of biomass between aquatic and terrestrial ecosystems. We demonstrated that predatory fish decoupled aquatic and terrestrial ecosystems by reducing the emergence of aquatic organisms and suppressing the decomposition of terrestrial plant detritus. In contrast, warming and nutrients enhanced cross‐ecosystem exchanges by increasing emergence and decomposition, and these effects were strongest in the absence of predators. Furthermore, we found that warming advanced while predators delayed the phenology of insect emergence. Our results demonstrate that anthropogenic perturbations may extend well beyond ecosystem boundaries by influencing cross‐ecosystem subsidies. We find that these changes are sufficient to substantially impact recipient communities and potentially alter the carbon balance between aquatic and terrestrial ecosystems and the atmosphere. |
|---|---|
| Bibliography: | Natural Sciences and Engineering Research Council of Canada ArticleID:GCB2540 ark:/67375/WNG-HWM0TFN7-M Appendix S1. Mean daily water temperature in mesocosms with (warmed) and without (unwarmed) continuous heating with a 300 W submersible aquarium heater, over the course of the 17-month experiment (June 2009-September 2010). Warming increased the daily average water temperature by 2.98 ± 0.07 °C (mean ± SE).Appendix S2. The taxonomic composition of adult insects emerging from the mesocosms over the duration of the experiment.Appendix S3. Mixed effects linear model of the biomass flux of emerging adult aquatic insects (mg × m2 × day) over 7 months. Time refers to each of the fourteen 2-weekly intervals. The model included an AR(1) error structure to account for first-order temporal autocorrelation. Comparisons of models with AIC and plots of temporal autocorrelation indicated the AR(1) error structure significantly improved model fit.Appendix S4. Cumulative proportional aquatic insect emergence for (a) control, (b) nutrient, (c) fish, and (d) nutrients + fish treatments with and without warming. Curves are drawn from mean (± SE) cumulative emergence proportions of five replicate mesocosms at each date. Warming accelerated emergence (steeper curve) in all treatments except nutrients + fish (d), and the effect was strongest in the presence of fish (c). Fish delayed insect emergence [lower slopes in (c) and (d)] especially in the absence of nutrients and warming (c).Appendix S5. Summary of the mixed effects linear model of loge-transformed leaf decomposition rates (k day−1) in spring and summer (Season main effect). Models were run with and without an AR(1) error structure to account for first-order temporal autocorrelation. Plots of temporal autocorrelation indicated weak autocorrelation between seasons (<0.4) which did not change when an AR(1) error structure was included. Consequently, the AR(1) error term was not retained in the final model.Appendix S6. Mixed effects linear model of the mean biomass of primary consumers in leaf packs in spring and summer (Season main effect). Models were run with and without an AR(1) error structure to account for first-order temporal autocorrelation. Plots of temporal autocorrelation indicated weak autocorrelation between seasons (<0.4) which did not change when an AR(1) error structure was included. Consequently, the AR(1) error term was not retained in the final model.Appendix S7. Mean biomass of periphyton (as chlorophyll-a) on benthic tiles. One 25 cm2 terracotta tile was added to each tank on 9 June 2009 and was removed on 4 February 2010. Tiles were brushed and filtered on to a Whatman GF/F filter paper, which was then immersed in 90% acetone to extract pigments. Biomass of chlorophyll-a was estimated with the nonacidification method on a Turner Trilogy Flourometer. Factorial anova revealed strong positive effect of nutrients (P < 0.0001) on periphyton biomass. No other effects were significant. New Zealand Foundation for Research, Science & Technology - No. UBX0901 istex:3F87C46688F78852764A6BA5C52DBDC6A6F42D91 ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
| ISSN: | 1354-1013 1365-2486 |
| DOI: | 10.1111/j.1365-2486.2011.02540.x |