I modelled plant community changes for eastern South Africa using environmental domains as an alternative to species distribution modelling. You may get the paper from me directly if you are interested or look it up - Jewitt et al 2015 Climate-induced change of environmentally defined floristic domains: a conservation based vulnerability framework. Applied Geography 63: 33-42.
Predicting changes in biodiversity in different climatic scenarios is not straight forward especially since you need to know the biodiversity factor of the area/region you are investigating. If the different climatic variables are known you can use a number of different mathematical models and/or formulae to actually get the biodiversity/biological factor as a % of the total regional biodiversity. This would require identification of the all the number of species within all kingdoms. Once you have all the variables accounted for, a number of different scientific articles released mathematical or computational models to calculate or represent what you are looking for.
One way is through species distribution modelling which has already been mentioned. Another way is through risk analysis. I will attach a paper that might be useful to you that identified the risk of global biodiversity due to future land-use change and climate change scenarios. You can also use the same approach on a smaller local scale.
Dear colleague. This is depend about what type of data you have available and on what period. You have to chose a specific set of indicator species and habitats ecologicaly assess them and monitor them. The results should be modeled in an integrated manner with climatic significant elements (distribution modeling, population structure trends, metapopulations dynamic, risk analysis, etc). No general formula have absolute wining potential, a specific site/case formula should be created and implemented. It is a medium-long term approach. You can contact me and my team at [email protected] SUCCEES:)
Climate modelling is complex enough, I'd keep it simple and limited to as few variables as possible. But that's probably because I'm not a great modeller! Modelling future change is obviously a best-guess using your variables and climate scenarios, and undeniably limited. That does not mean it does not have value and is not of interest. The above suggestions are good. How will climate change impact biodiversity - i.e., who wins, who loses? e.g., Altered rainfall patterns and CO2 fertilisation may benefit some species. Anyway, the attached paper (on Researchgate, you may be familiar with it) might be of help, concerning climate change impacts on mostly endemic bird species in the region in which you're studying. It has been quite widely cited too. Good luck.
Try also C. McAlpine's work (no relation, via Researchgate, google scholar etc) on climate modelling in Australia, another region that is predicted to experience some of the most severe effects of future climate change.
Article Modelling relationships between species spatial abundance pa...
The paleoecological record provides information on species distribution under different climatic conditions (e.g. the ecology of the last interglacial which was warmer than the current interglacial), particularly for taxa that are temperature sensitive. I have attached a publication on this subject.
I maintain an annotated llnks webpage of papers that are important for using the conservation tool of "assisted migration" of species in anticipation of climate change. You will find key papers on paleobiological and modelling methods on that page:
The only readily-available program for doing that is Jabowa IV, by Dan Botkin. But, that program grows only trees, not shrubs or animals. Here is the URL to his webpage: http://www.danielbbotkin.com/jabowa/, Dan invented this approach in 1970, I did research with it 1987 -1994. The latest version for Windows 7 in 32-bit or 64-bit format is only $150. I used Jabowa II to project forest growth changes by species under simulated global warming conditions for 90 years into the future, driven by the GISS climate model data.