Limits...
The 'island effect' in terrestrial global change experiments: a problem with no solution?

Leuzinger S, Fatichi S, Cusens J, Körner C, Niklaus PA - AoB Plants (2015)

Bottom Line: Here, we discuss its implications in various global change experiments with plants.We also suggest ways to complement experiments using modelling approaches and observational studies.Ultimately, there is no obvious solution to deal with the island effect in field experiments and only models can provide an estimate of modification of responses by these feedbacks.

View Article: PubMed Central - PubMed

Affiliation: Institute for Applied Ecology New Zealand, School of Applied Sciences, Auckland University of Technology, Private Bag 92006, Auckland 1142, New Zealand sleuzing@aut.ac.nz.

No MeSH data available.


Related in: MedlinePlus

Sensitivity of the island effect in an elevated CO2 experiment. The potential error (Er.) committed in estimating transpiration and NPP with the current RH is shown. RH is reduced by a given factor (fRH), due to hypothetical local or regional feedbacks. The numerical experiment is carried out for two CO2 levels: 400 ppm (ambient, AMB) and 700 ppm (elevated, ELE CO2) and four locations: a tropical forest in Manaus (Brazil), a grassland in Rietholzbach (Switzerland), a deciduous forest near the University of Michigan Biological Station (UMBS) Michigan, USA, and a grassland in California, USA (Vaira ranch). The simulated time series of transpiration and NPP during a characteristic growing season at the UMBS are also shown (subplots E and F); note the dry period between days 2040 and 2050.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4584962&req=5

PLV092F4: Sensitivity of the island effect in an elevated CO2 experiment. The potential error (Er.) committed in estimating transpiration and NPP with the current RH is shown. RH is reduced by a given factor (fRH), due to hypothetical local or regional feedbacks. The numerical experiment is carried out for two CO2 levels: 400 ppm (ambient, AMB) and 700 ppm (elevated, ELE CO2) and four locations: a tropical forest in Manaus (Brazil), a grassland in Rietholzbach (Switzerland), a deciduous forest near the University of Michigan Biological Station (UMBS) Michigan, USA, and a grassland in California, USA (Vaira ranch). The simulated time series of transpiration and NPP during a characteristic growing season at the UMBS are also shown (subplots E and F); note the dry period between days 2040 and 2050.

Mentions: To constrain the potential bias that the island effect introduces in experiments, explicit simulations of changes in VPD using coupled vegetation–atmosphere models could help. As an example, we used a mechanistic ecohydrological model, T&C (Fatichi et al. 2012, 2015; Fatichi and Ivanov 2014), to carry out a sensitivity analysis of the potential errors one can commit assuming that RH does not change in a manipulation experiment, where, instead, a reduction in gs is expected (e.g. elevated CO2). We applied two CO2 levels (400 and 700 ppm), and different reduction factors (fRH) to the observed time series of RH, from fRH = 1.0 (no reduction) to fRH = 0.85 corresponding to a strong (−15 %) reduction of RH. We used four locations characterized by different climates and plant functional types to show the potential variability of the effect (Fig. 4A–D). Simulated errors in the long-term (5–10 years) transpiration may reach −5 to −10 %, but they are generally constrained to less than −2 % for realistic changes in RH (Fig. 4). The negative sign is expected since the island effect leads to an underestimation of transpiration because RH is not reduced. Simulated errors in net primary productivity (NPP) are constrained within ±2 %, except for one case study (UMBS in Fig. 4C), where they can reach up to +10 %. For this location, the changes in transpiration due to a lower RH are sufficient to increase the water stress in the forest stand with comparison to unchanged RH. Changes in soil moisture considerably affect NPP, which would be overestimated in the treated ‘island’. While the long-term expected errors may be small, there are specific seasons or periods where the island effect can be potentially very significant (>100 %), for example the dry period in the middle of the growing season (days 40 to 50, Fig. 4E and F). In this scenario, the difference in transpiration and NPP induced by a 5 % change in RH can affect the system response of a magnitude similar to that of an elevated CO2 treatment. Since many observations in manipulation experiments are typically carried out for limited periods during the growing season, the potential artefacts of the island effect may be significant. The presented results are a sensitivity analysis and likely dependent on model parameters, but they suggest that there may be situations or locations, where a small change in RH can lead to considerably different plant water stress, with potentially large implications in terms of ecosystem stability and composition.Figure 4.


The 'island effect' in terrestrial global change experiments: a problem with no solution?

Leuzinger S, Fatichi S, Cusens J, Körner C, Niklaus PA - AoB Plants (2015)

Sensitivity of the island effect in an elevated CO2 experiment. The potential error (Er.) committed in estimating transpiration and NPP with the current RH is shown. RH is reduced by a given factor (fRH), due to hypothetical local or regional feedbacks. The numerical experiment is carried out for two CO2 levels: 400 ppm (ambient, AMB) and 700 ppm (elevated, ELE CO2) and four locations: a tropical forest in Manaus (Brazil), a grassland in Rietholzbach (Switzerland), a deciduous forest near the University of Michigan Biological Station (UMBS) Michigan, USA, and a grassland in California, USA (Vaira ranch). The simulated time series of transpiration and NPP during a characteristic growing season at the UMBS are also shown (subplots E and F); note the dry period between days 2040 and 2050.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4584962&req=5

PLV092F4: Sensitivity of the island effect in an elevated CO2 experiment. The potential error (Er.) committed in estimating transpiration and NPP with the current RH is shown. RH is reduced by a given factor (fRH), due to hypothetical local or regional feedbacks. The numerical experiment is carried out for two CO2 levels: 400 ppm (ambient, AMB) and 700 ppm (elevated, ELE CO2) and four locations: a tropical forest in Manaus (Brazil), a grassland in Rietholzbach (Switzerland), a deciduous forest near the University of Michigan Biological Station (UMBS) Michigan, USA, and a grassland in California, USA (Vaira ranch). The simulated time series of transpiration and NPP during a characteristic growing season at the UMBS are also shown (subplots E and F); note the dry period between days 2040 and 2050.
Mentions: To constrain the potential bias that the island effect introduces in experiments, explicit simulations of changes in VPD using coupled vegetation–atmosphere models could help. As an example, we used a mechanistic ecohydrological model, T&C (Fatichi et al. 2012, 2015; Fatichi and Ivanov 2014), to carry out a sensitivity analysis of the potential errors one can commit assuming that RH does not change in a manipulation experiment, where, instead, a reduction in gs is expected (e.g. elevated CO2). We applied two CO2 levels (400 and 700 ppm), and different reduction factors (fRH) to the observed time series of RH, from fRH = 1.0 (no reduction) to fRH = 0.85 corresponding to a strong (−15 %) reduction of RH. We used four locations characterized by different climates and plant functional types to show the potential variability of the effect (Fig. 4A–D). Simulated errors in the long-term (5–10 years) transpiration may reach −5 to −10 %, but they are generally constrained to less than −2 % for realistic changes in RH (Fig. 4). The negative sign is expected since the island effect leads to an underestimation of transpiration because RH is not reduced. Simulated errors in net primary productivity (NPP) are constrained within ±2 %, except for one case study (UMBS in Fig. 4C), where they can reach up to +10 %. For this location, the changes in transpiration due to a lower RH are sufficient to increase the water stress in the forest stand with comparison to unchanged RH. Changes in soil moisture considerably affect NPP, which would be overestimated in the treated ‘island’. While the long-term expected errors may be small, there are specific seasons or periods where the island effect can be potentially very significant (>100 %), for example the dry period in the middle of the growing season (days 40 to 50, Fig. 4E and F). In this scenario, the difference in transpiration and NPP induced by a 5 % change in RH can affect the system response of a magnitude similar to that of an elevated CO2 treatment. Since many observations in manipulation experiments are typically carried out for limited periods during the growing season, the potential artefacts of the island effect may be significant. The presented results are a sensitivity analysis and likely dependent on model parameters, but they suggest that there may be situations or locations, where a small change in RH can lead to considerably different plant water stress, with potentially large implications in terms of ecosystem stability and composition.Figure 4.

Bottom Line: Here, we discuss its implications in various global change experiments with plants.We also suggest ways to complement experiments using modelling approaches and observational studies.Ultimately, there is no obvious solution to deal with the island effect in field experiments and only models can provide an estimate of modification of responses by these feedbacks.

View Article: PubMed Central - PubMed

Affiliation: Institute for Applied Ecology New Zealand, School of Applied Sciences, Auckland University of Technology, Private Bag 92006, Auckland 1142, New Zealand sleuzing@aut.ac.nz.

No MeSH data available.


Related in: MedlinePlus