Species Distribution Modeling

ATree Species Distribution Modeling Workshop-December 2008, Berkeley

The Workshop, held December 5-6, 2008 on the University of California, Berkeley, campus in the Museum of Vertebrate Zoology and the Geospatial Innovation Facility, was well attended. For the benefit of those who could not attend, below are presentation abstracts, reading lists and the PDF manual for the practical labs. Links to other related resources are also posted below.

In order of presentation:

Kenneth H. Kozak, University of Minnesota
The merging of historical biogeography and spatial ecology
Many biogeographic processes, ranging from speciation to the assembly of regional floras and faunas, are ultimately influenced by environmental variation over space and time. Yet, until recently, many studies in these general areas of evolutionary research ignored or employed only crude proxies for environmental variation (e.g. latitude, distance). With the availability GIS data and new spatial tools, biologists can now readily obtain environmental data for thousands of localities and hundreds of species across the globe, permitting them to more rigorously explore the underlying ecological and evolutionary causes of biogeographic patterns. Using case studies from tropical and temperate plethodontid salamanders, I will explore how these sources of environmental variation and phylogeny-based approaches can be integrated to provide fresh insights on the origin of new species and large-scale gradients in species diversity. Key conceptual and methodological challenges associated with integrating phylogenetic and spatial data over evolutionary timescales will be discussed.

William Collins, Lawrence Berkeley National Laboratory, University of California, Berkeley
Projections of Future Climate Change: What's Past is Prologue

Michelle Koo, MVZ, University of California, Berkeley
Species Distribution Modeling I: Basics, Background and Getting Started
The goal of this workshop is to provide an understanding of the fundamental assumptions and theories of species distribution modeling and to present “best-practice” advice for data and model treatment. An introduction to species distributional modeling will include background on the rise of the natural history collection–based informatics, geospatial data, and the increasing concern for species and habitat predictions. Fundamentals of species modeling will include the theoretical differences between geographic space and environmental space and the varying scenarios that models may present and their respective implications. We will discuss the theoretical factors that affect model performance as well as selection of modeling method. Likewise, the fundamentals of the data necessary to species modeling will be examined closely, specifically understanding where majority of georeferenced specimen data can be found, inherent errors, and data-cleaning, as well as geospatial, climatic variables used, where to acquire them and formatting for modeling. In the lab, we will explore individual datasets and their environmental space using DIVA-GIS, a free GIS program that was designed for species modeling. Preparing datasets for the algorithm and program Maxent will follow.

David Vieites, National Museum of Natural History, Madrid, Spain
Evolution of niches through time: How far can we go back?
Modelling the ecological niches of species has become popular among ecologists and evolutionary biologists. Using current locality and climatic data is possible to develop statistical models to predict the fundamental niche of a species. While the fundamental niche is genetically and physiologically determined, the realized niche includes additional constraints arising from interspecific competition or biogeographic constraints. The combination of genetic data (e.g. phylogenetic or phylogeographic) with bioclimatic or physiological modelling is promising to study the evolution of species and their niches through time. However, there are few studies that investigated the evolutionary dynamics of fundamental niches in a phylogenetic context. Some of these used phylogenies to reconstruct the ancestral states of particular climatic variables. Others correlated environmental and genetic distances to evaluate niche dynamics and niche conservatism in pairs of sister species. These studies do not incorporate past global climatic changes so far, which may have a strong influence in the evolution of ecological niches, and may limit how far back in evolutionary time can we reconstruct them. I will review the state of the art in this field, providing examples about the integration of phylogenetic and bioclimatic data, and discuss the potential uses and limitations of these approaches.

Bill Monahan, Audubon California
Biological considerations when projecting species distribution models
to novel environments
Most studies using ecological niche models to predict changes in species’ distributions over space and time assume that species are in a state of distributional equilibrium with respect to the environment. Under this key assumption, species are expected to track environmentally induced changes in the geographic boundaries of their potential niche. However, an equally parsimonious and non-mutually exclusive hypothesis is that species respond geographically to environmental change by moving into and out of existing areas of potential niche space. These two possibilities raise important questions as to what we are modeling and how we interpret model predictions in both the training and projection environments. In addition to discussing such subjects through case studies, I will present methods for testing the equilibrium assumption and ways for conditioning model results in situations where the assumption simply cannot be tested.

Michelle Koo, MVZ, University of California, Berkeley
Species Distribution Modeling II: Evaluation, Interpretation and Analysis
The ease of generating species distribution models with many of the new algorithms, such as Maxent and DIVA-GIS, often belies the various technical hurdles to visualizing it in GIS for beginners, and the evolving studies on evaluating and interpreting the results. We will step through the process for beginners of GIS using ArcGIS 9 (ESRI) starting with data-acquisition, using time-saving steps for batch processing as well as discussing and performing thresholds for binary models, model performance indices such as Kappa’s and AUC’s, and projections to future climate or other alternative environmental datasets. The focus will be hands-on demonstration, emphasizing the practicalities of modeling with available tools, and whenever possible, participants will use their own dataset.
Some workshop material is adapted from the teaching module for Species Distribution Modeling for Conservation by Richard Pearson of the Network of Conservation Educators and Practitioners (NCEP), Center of Biodiversity and Conservation at the AMNH and from the Geospatial Innovation Facility, UC Berkeley.

Reading & Relevant Reference List
Day 1 Readings, Friday, December 5th:
Ken Kozak:
• Kozak, K.H., C. H. Graham, and J. J. Wiens. 2008. Integrating GIS-based data into evolutionary biology. Trends in Ecology and Evolution 23:141-148.
• Kozak, K. H., and J. J. Wiens. 2006. Does niche conservatism promote speciation? A case study in eastern North American salamanders. Evolution 60:2604-2621.

William Collins:
• Collins, W, R Colman, J Hayward, MR Manning, P Mote. 2007. The physical science behind climate change. Scientific American.

Michelle Koo:
• Elith and Graham et al. 2006 Novel methods improve prediction of species’ distributions from occurrence data. Ecography 29:129-151.
• Graham, C. H., S. Ferrier, F. Huettman, C. Moritz, and A. T. Peterson. 2004. New developments in museum-based informatics and applications in biodiversity analysis. Trends in Ecology and Evolution 19:497-503.
• Guisan, A and W. Thuiller. 2005. Predicting species distribution: offering more than simple habitat models. Ecology Letters 8:993-1009.
• Phillips, S. J., R. P. Anderson, and R. E. Schapire. 2006. Maximum entropy modeling of species geographic distributions. Ecological Modelling 190:231-259.

Day 2 Readings, Saturday, December 6th:
David Vieites:
• Graham, C. H., Ron, S. R., Santos, J. C., Schneider, C. J. and Moritz, C. 2004. Integrating phylogenetics and environmental niche models to explore speciation mechanisms in dendrobatid frogs. Evolution 58:1781-1793.
• Pearman, P. B., Guisan, A., Olivier Broennimann, O. and Randin, C. F. 2007. Niche dynamics in space and time. TREE 23:148-158.
• Yesson, C. and Culham, A. 2006. Phyloclimatic Modeling: Combining Phylogenetics and Bioclimatic Modeling. Systematic Biology 55:785-802.

William Monahan:
• Araújo, M.B. & Pearson, R.G. (2005) Equilibrium of species’ distributions with climate. Ecography, 28, 693–695.
• Hijmans, R.J. & Graham, C.H. (2006) The ability of climate envelope models to predict the effect of climate change on species distributions. Global Change Biology, 12, 2272–2281.
• Kearney, M. 2006. Habitat, environment and niche: what are we modelling? Oikos 115, 86-91.

Additional References of Interest:

Anderson, RP, Lew D, and AT Peterson. 2003 Evaluating predictive modeling of species’ distributions: Criteria for selecting optimal models. Ecological Modelling 162:211-232.
Fielding, A. H., and J. F. Bell. 1997. A review of methods for the assessment of prediction errors in conservation presence/absence models. Environmental Conservation 24:38-49.
Graham, CH, J Elith, RJ Hijmans, A Guisan, AT Perterson, BA Loiselle and the NCEAS Predicting Species Distribution Working Group. 2008 The influence of spatial errors in species occurrence data used in distribution models. Journal of Applied Ecology 45:239-247
Kidd, DM and X Liu 2008 Geophylobuilder 1.0: an ArcGIS extension for creating ‘geophylogenies’. Molecular Ecology Resources 8:88-91.
Peterson, A.T. 2003 Predicting the geographies of species’ invasions via ecological niche modeling. The Quarterly Review of Biology, 78, 419–433.
Phillips, SJ and M Dudik. 2008 Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography, 31:161-175.
Soberon 2007 Grinnellian and Eltonian niches and geographic distributions of species. Ecology Letters 10:1115-1123.
Svenning, J.-C. & Skov, F. 2004 Limited filling of the potential range in European tree species. Ecological Letters, 7, 565–573.

More References and their doi links at:
Species Distribution Modeling References – Center for Biodiversity and Conservation, AMNH