Individualistic species vs. ecosystem responses in under changing climate and CO2 conditions

Background/Question/Methods

Different modelling approaches to assessing the impacts of climate and atmospheric change on ecosystems and their component species yield projections that may be difficult to compare and to reconcile. Two such contrasting approaches are dynamic global vegetation modelling (DGVM), which seeks to simulate vegetation structure and function through projections of dominant plant functional types and their carbon, water and nitrogen relations, and bioclimatic niche-based modelling (NBM), a species-focused approach that derives correlative relationships between mainly climatic variables and the likelihood of individual species occurrence. I review applications of a number of variants of these two approaches in a range of ecological contexts and ask how progress might be made through comparing insights from both approaches, and what opportunities exist to combine them.

Results/Conclusions

Species-focused NBM approaches produce spatially detailed and complex projections of diversity change based on climatic (and sometimes substrate) preferences. But apart from inherent assumptions in the correlative methodologies used, they lack a population dynamics aspect, they tend to ignore habitat and disturbance relationships, and they are blind to important ecological functional aspects. Nonetheless, there is some evidence that the approach can produce credible outputs in terms of species occurrence and projected range changes. DGVM approaches provide useful insights into the time-dynamics of ecosystem structural shifts and changes in disturbance regime, but are mute on diversity changes.  DGVMs allow important questions to be asked about the potential structure and function of ecosystems, and the role of disturbance and atmospheric CO₂ in driving structural and functional changes through demographic processes.  DGVMs project significant ecosystem structural changes under future climate and especially CO₂ change that could lead to widespread habitat changes in many regions. But these projections have seldom been used to infer diversity changes, and explicitly compared with NBM-based projections. I conclude that there is potential to meld the outputs from DGVM approaches with NBM efforts, especially in disturbance-driven ecosystems.