ENM - CC2

Now what kinds of topics can we address with niche modelling in the context of climate change as we mentioned there are studies in which you look for forcasting ditributions and in order to this you go back in time there is an important difference between these two approaches and it is as Richard mentioned for future we don't have any data to evaluate our resources so what we get is only hypothesis of what can happen in the future on the other hand when we project models back in time we have data to XXXX this information so XXXXX we can get from one study and the other is very different it's much more solid from the XXXX we can learn more from looking at the past that we can gain from the future so let me show you some examples of how these models have been used by different people including us like the XXXXXXXXX once just to understand when you hind test the models XXXXXXXXX the responses of the species to climatic change from the past so in this, what is a part of my PhD dissertattion we modelled the distribution of 23 mammals first we did models in the present and then on past conditions and the reverse to see if the species here were, trcking there niches across space or did they expand their distribution or they XXXXX their distribution the general sort of this is called the XXXXXXX from the 23 mammal species we were able to get data from the last XXXXXXXXXXX and the present one, present in the last climatic change XXXXXX is that for most of them we find a correspondence between model of distribution at present and in the past with XXXXX rate from the present and the XXXXXXX rate from the past this means that those species have been able to make changes in distribution of their niches for XXXXXXXX 6 species results were not solid and for 5 species we find no correspondence between the model and the distributions to the past and the present and these 5 species are large species but in pleistocene those species were not the biggest ones because the megafauna were present, mammals and sabre-toothed tigers so what XXXXXXXXXXX is that these medium sized species in present they are large species found like XXXXX ecological reigns so they can have XXXXXXXXXXXXXXXXXXXXXXXXXX so they can expand across the space well these hand casting has been used also to find pleistocene XXXXXXXXXXXXX they combine different approaches in this case phylogeographical techniques with geological modelling techniques and see how the conclusions match in terms of geographical position of refugia in the pleistocene we have also used phylogeographic information to understand geographic processes and the combination of particular niche models to see looking at the genetic patterns in a species and XXXXXX the genetic structure of populations between species and then we compare that with phylo distributions we can get insight to geographic distributions like bottlenecks XXXXXXXXXXXXX one of the neatest works I have seen using ecological niche modelling for XXXXXX species is this one of XXXXXXXXXXXXXXXXXXXXXXXXX to explore the two hypotheses behind mega fauna in this case the wooly mammoth in one extreme of this XXXX is that mega fauna was driven extinct by human colonization and the other extreme is the people that think the extinction was driven at the end of the pleistocene then XXXXXXXXXXX so what XXXXXXXX his study was to model the distribution of man using fossil beds for different periods of time in the last glacial maximum and in the last inter glcial to see if the distribution tended to shrink in one period and expand in the cold periods in this XXXXX of the pleistocene and what his result gave him is like here presents, the expansion of the first humans XXXXXXXXXXXX so what he observed is that in warm periods the distribution of the mammals was reduced and expanded during the cold period and in the last warming process 10000 years ago humans arrived at the very north of the continent so XXXXXXXXXX that the debate between which of the 2 drivers might have been extinction XXXXXXXX XXXXX combination of XXXXXXXXXXX populations were natural like this warming process and humans came in this process and XXXXX the extinction this is XXXXXXXX that you have in your XXXXXXXXXXXXX then you have this other bulk of studies that project models to the future instead of going to the past and in this case there are XXXXXXXXXXX thousands of papers on modelling one species with differnt aims at the end this is one of the XXXXXX studies XXXXXXXXXXX in which they model XXXXXX in Mexico species of mammals, birds and 4 families of XXXXXX with the idea of taking XXXXX of here extinction turnovers XXXXXXXXX etc then there are also a bunch of studies on XXXX projections to future for conservation purposes like Richard mentioned for his field of conservation planning there have been efforts from different XXXXX on projecting the models to the future to see for example XXXXXXXX protected areas XXXX are useful for future XXXXXXX distribution of XXXXXXXXX and you can find all sorts of studies and projects into the future there is a XXXXXXXXX in this part of the study which is called uncertainty the problem with the models to the future is that we have several sources of uncertainty starting with uncertainties of how the climate is going to be we have the same situation with past that we don't have XXXX climatic models to the past and no fine resolution models to the past and also that the XXXXXXXXXXXXX has a lot of problems in terms of representation in terms of geographic distributions of the species from the from the XXXXX backwards in terms of dating so one main isuue here is the uncertainty of climatic varients it is like a common practise to run models on different climatic scenarios there are hundreds of scenarios out there for the future and much less for past and present so you run models for several scenarios and XXXXXXXXXXXX XXXXXXXX there are techniques for sensing of these models still what you didn't get is an analysis of the uncertainty associated with this type of XXXXX there is also XXXXXXX certainty which Richard mentioned XXXXXXX in which we compared distribution of species of the south XXXXXXXXXXX under the same XXXXXX we project niche model to the same scenarios using 9 different algorithms and these are the results. you can see for some algorithms distribution would expand and for others would collapse but these studies XXXXXX was that the fate of the species depends on the XXXXXXXXXXXXXXXXXX which is XXXXXXXXXXX and XXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXX so differently when we project it to future conditions because that didnt happen for the current conditions all these methods were XXXXXXXXXXXX and modelling the current distributions the problem was in projecting the difference and it has to do with how do each one of these XXXXXXXXXXXX deals with environmental combinations XXXXXX in the calibration scenario and we find them in the projected scenario and you have no environments some of the other readings would extrapolate some others good XXXXXX some others good XXXXXXXXX so it does not take a different decision on how to manage these novel environments so discusses the huge differences among them and this is the problem. How can we know which of these algorithms is doing a good job there is no way to do that for the future. we don't have data but we are doing it in my lab with student of mine and Richard's is that she developed environmental services XXXX of the 20th century so we have XXXXX data for birds, herps and for mammals for the beginning of the 20th century and all along the 20th century so we are using historic data which is XXXX environmental levels to project on to current environmental XXXXXX and through current occurence data which XXXXXXXXXXXXXX and that way we can see that we can harvest like an idea which