@article{YatesElithLatimeretal.2010,
  author    = {Yates, Colin J. and Elith, Jane and Latimer, Andrew M. and Le Maitre, David and Midgley, Guy F. and Schurr, Frank Martin and West, Adam G.},
  title     = {Projecting climate change impacts on species distributions in megadiverse South African Cape and Southwest Australian Floristic Regions : Opportunities and challenges},
  issn      = {1442-9985},
  doi       = {10.1111/j.1442-9993.2009.02044.x},
  year      = {2010},
  abstract  = {Increasing evidence shows that anthropogenic climate change is affecting biodiversity. Reducing or stabilizing greenhouse gas emissions may slow global warming, but past emissions will continue to contribute to further unavoidable warming for more than a century. With obvious signs of difficulties in achieving effective mitigation worldwide in the short term at least, sound scientific predictions of future impacts on biodiversity will be required to guide conservation planning and adaptation. This is especially true in Mediterranean type ecosystems that are projected to be among the most significantly affected by anthropogenic climate change, and show the highest levels of confidence in rainfall projections. Multiple methods are available for projecting the consequences of climate change on the main unit of interest - the species - with each method having strengths and weaknesses. Species distribution models (SDMs) are increasingly applied for forecasting climate change impacts on species geographic ranges. Aggregation of models for different species allows inferences of impacts on biodiversity, though excluding the effects of species interactions. The modelling approach is based on several further assumptions and projections and should be treated cautiously. In the absence of comparable approaches that address large numbers of species, SDMs remain valuable in estimating the vulnerability of species. In this review we discuss the application of SDMs in predicting the impacts of climate change on biodiversity with special reference to the species-rich South West Australian Floristic Region and South African Cape Floristic Region. We discuss the advantages and challenges in applying SDMs in biodiverse regions with high levels of endemicity, and how a similar biogeographical history in both regions may assist us in understanding their vulnerability to climate change. We suggest how the process of predicting the impacts of climate change on biodiversity with SDMs can be improved and emphasize the role of field monitoring and experiments in validating the predictions of SDMs.},
  language  = {en}
}
@article{DormannElithBacheretal.2013,
  author    = {Dormann, Carsten F. and Elith, Jane and Bacher, Sven and Buchmann, Carsten M. and Carl, Gudrun and Carre, Gabriel and Garcia Marquez, Jaime R. and Gruber, Bernd and Lafourcade, Bruno and Leitao, Pedro J. and M{\"u}nkem{\"u}ller, Tamara and McClean, Colin and Osborne, Patrick E. and Reineking, Bjoern and Schr{\"o}der-Esselbach, Boris and Skidmore, Andrew K. and Zurell, Damaris and Lautenbach, Sven},
  title     = {Collinearity a review of methods to deal with it and a simulation study evaluating their performance},
  series = {Ecography : pattern and diversity in ecology ; research papers forum},
  volume    = {36},
  journal   = {Ecography : pattern and diversity in ecology ; research papers forum},
  number    = {1},
  publisher = {Wiley-Blackwell},
  address   = {Hoboken},
  issn      = {0906-7590},
  doi       = {10.1111/j.1600-0587.2012.07348.x},
  pages     = {27 -- 46},
  year      = {2013},
  abstract  = {Collinearity refers to the non independence of predictor variables, usually in a regression-type analysis. It is a common feature of any descriptive ecological data set and can be a problem for parameter estimation because it inflates the variance of regression parameters and hence potentially leads to the wrong identification of relevant predictors in a statistical model. Collinearity is a severe problem when a model is trained on data from one region or time, and predicted to another with a different or unknown structure of collinearity. To demonstrate the reach of the problem of collinearity in ecology, we show how relationships among predictors differ between biomes, change over spatial scales and through time. Across disciplines, different approaches to addressing collinearity problems have been developed, ranging from clustering of predictors, threshold-based pre-selection, through latent variable methods, to shrinkage and regularisation. Using simulated data with five predictor-response relationships of increasing complexity and eight levels of collinearity we compared ways to address collinearity with standard multiple regression and machine-learning approaches. We assessed the performance of each approach by testing its impact on prediction to new data. In the extreme, we tested whether the methods were able to identify the true underlying relationship in a training dataset with strong collinearity by evaluating its performance on a test dataset without any collinearity. We found that methods specifically designed for collinearity, such as latent variable methods and tree based models, did not outperform the traditional GLM and threshold-based pre-selection. Our results highlight the value of GLM in combination with penalised methods (particularly ridge) and threshold-based pre-selection when omitted variables are considered in the final interpretation. However, all approaches tested yielded degraded predictions under change in collinearity structure and the folk lore'-thresholds of correlation coefficients between predictor variables of |r| >0.7 was an appropriate indicator for when collinearity begins to severely distort model estimation and subsequent prediction. The use of ecological understanding of the system in pre-analysis variable selection and the choice of the least sensitive statistical approaches reduce the problems of collinearity, but cannot ultimately solve them.},
  language  = {en}
}
@misc{ZurellElithSchroederEsselbach2012,
  author    = {Zurell, Damaris and Elith, Jane and Schr{\"o}der-Esselbach, Boris},
  title     = {Predicting to new environments tools for visualizing model behaviour and impacts on mapped distributions},
  series = {Diversity \& distributions : a journal of biological invasions and biodiversity},
  volume    = {18},
  journal   = {Diversity \& distributions : a journal of biological invasions and biodiversity},
  number    = {6},
  publisher = {Wiley-Blackwell},
  address   = {Hoboken},
  issn      = {1366-9516},
  doi       = {10.1111/j.1472-4642.2012.00887.x},
  pages     = {628 -- 634},
  year      = {2012},
  abstract  = {Data limitations can lead to unrealistic fits of predictive species distribution models (SDMs) and spurious extrapolation to novel environments. Here, we want to draw attention to novel combinations of environmental predictors that are within the sampled range of individual predictors but are nevertheless outside the sample space. These tend to be overlooked when visualizing model behaviour. They may be a cause of differing model transferability and environmental change predictions between methods, a problem described in some studies but generally not well understood. We here use a simple simulated data example to illustrate the problem and provide new and complementary visualization techniques to explore model behaviour and predictions to novel environments. We then apply these in a more complex real-world example. Our results underscore the necessity of scrutinizing model fits, ecological theory and environmental novelty.},
  language  = {en}
}