TY - JOUR T1 - Species Invasion History Influences Community Evolution in a Tri-Trophic Food Web Model JF - PLoS ONE Y1 - 2009 DO - 10.1371/journal.pone.0006731 A1 - Mougi, Akihiko A1 - Nishimura, Kinya SP - e6731+ KW - aliens KW - ecology KW - invasion KW - models AB - Recent experimental studies have demonstrated the importance of invasion history for evolutionary formation of community. However, only few theoretical studies on community evolution have focused on such views. PB - Public Library of Science VL - 4 UR - http://dx.doi.org/10.1371/journal.pone.0006731 ER - TY - BOOK T1 - Ecological Models and Data in R Y1 - 2008 A1 - Bolker, Benjamin M. KW - ecology KW - models AB - \_Ecological Models and Data in R\_ is the first truly practical introduction to modern statistical methods for ecology. In step-by-step detail, the book teaches ecology graduate students and researchers everything they need to know in order to use maximum likelihood, information-theoretic, and Bayesian techniques to analyze their own data using the programming language R. Drawing on extensive experience teaching these techniques to graduate students in ecology, Benjamin Bolker shows how to choose among and construct statistical models for data, estimate their parameters and confidence limits, and interpret the results. The book also covers statistical frameworks, the philosophy of statistical modeling, and critical mathematical functions and probability distributions. It requires no programming background–only basic calculus and statistics. * Practical, beginner-friendly introduction to modern statistical techniques for ecology using the programming language R * Step-by-step instructions for fitting models to messy, real-world data * Balanced view of different statistical approaches * Wide coverage of techniques–from simple (distribution fitting) to complex (state-space modeling) * Techniques for data manipulation and graphical display * Companion Web site with data and R code for all examples PB - Princeton University Press SN - 0691125228 UR - http://www.worldcat.org/isbn/0691125228 ER - TY - JOUR T1 - Niches, models, and climate change: Assessing the assumptions and uncertainties JF - Proceedings of the National Academy of Sciences Y1 - 2009 DO - 10.1073/pnas.0901639106 A1 - Wiens, John A. A1 - Stralberg, Diana A1 - Jongsomjit, Dennis A1 - Howell, Christine A. A1 - Snyder, Mark A. SP - 19729–19736 KW - biodiversity KW - climate KW - climate-change KW - conservation KW - models KW - sdms AB - 10.1073/pnas.0901639106 As the rate and magnitude of climate change accelerate, understanding the consequences becomes increasingly important. Species distribution models (SDMs) based on current ecological niche constraints are used to project future species distributions. These models contain assumptions that add to the uncertainty in model projections stemming from the structure of the models, the algorithms used to translate niche associations into distributional probabilities, the quality and quantity of data, and mismatches between the scales of modeling and data. We illustrate the application of SDMs using two climate models and two distributional algorithms, together with information on distributional shifts in vegetation types, to project fine-scale future distributions of 60 California landbird species. Most species are projected to decrease in distribution by 2070. Changes in total species richness vary over the state, with large losses of species in some ” hotspots” of vulnerability. Differences in distributional shifts among species will change species co-occurrences, creating spatial variation in similarities between current and future assemblages. We use these analyses to consider how assumptions can be addressed and uncertainties reduced. SDMs can provide a useful way to incorporate future conditions into conservation and management practices and decisions, but the uncertainties of model projections must be balanced with the risks of taking the wrong actions or the costs of inaction. Doing this will require that the sources and magnitudes of uncertainty are documented, and that conservationists and resource managers be willing to act despite the uncertainties. The alternative, of ignoring the future, is not an option. VL - 106 UR - http://dx.doi.org/10.1073/pnas.0901639106 ER - TY - JOUR T1 - Differences in spatial predictions among species distribution modeling methods vary with species traits and environmental predictors JF - Ecography Y1 - 2009 DO - 10.1111/j.1600-0587.2009.05883.x A1 - Syphard, Alexandra D. A1 - Franklin, Janet SP - 907–918 KW - gis KW - mapping KW - maps KW - models KW - sdms AB - Prediction maps produced by species distribution models (SDMs) influence decision-making in resource management or designation of land in conservation planning. Many studies have compared the prediction accuracy of different SDM modeling methods, but few have quantified the similarity among prediction maps. There has also been little systematic exploration of how the relative importance of different predictor variables varies among model types and affects map similarity. Our objective was to expand the evaluation of SDM performance for 45 plant species in southern California to better understand how map predictions vary among model types, and to explain what factors may affect spatial correspondence, including the selection and relative importance of different environmental variables. Four types of models were tested. Correlation among maps was highest between generalized linear models (GLMs) and generalized additive models (GAMs) and lowest between classification trees and GAMs or GLMs. Correlation between Random Forests (RFs) and GAMs was the same as between RFs and classification trees. Spatial correspondence among maps was influenced the most by model prediction accuracy (AUC) and species prevalence; map correspondence was highest when accuracy was high and prevalence was intermediate (average prevalence for all species was 0.124). Species functional type and the selection of climate variables also influenced map correspondence. For most (but not all) species, climate variables were more important than terrain or soil in predicting their distributions. Environmental variable selection varied according to modeling method, but the largest differences were between RFs and GLMs or GAMs. Although prediction accuracy was equal for GLMs, GAMs, and RFs, the differences in spatial predictions suggest that it may be important to evaluate the results of more than one model to estimate the range of spatial uncertainty before making planning decisions based on map outputs. This may be particularly important if models have low accuracy or if species prevalence is not intermediate. VL - 32 UR - http://dx.doi.org/10.1111/j.1600-0587.2009.05883.x ER - TY - JOUR T1 - Bayesian Regularization for Normal Mixture Estimation and Model-Based Clustering JF - Journal of Classification Y1 - 2007 DO - 10.1007/s00357-007-0004-5 A1 - Fraley, Chris A1 - Raftery, Adrian E. SP - 155–181 KW - classification KW - cluster KW - models KW - taxonomy AB - Abstract  Normal mixture models are widely used for statistical modeling of data, including cluster analysis. However maximum likelihood estimation (MLE) for normal mixtures using the EM algorithm may fail as the result of singularities or degeneracies. To avoid this, we propose replacing the MLE by a maximum a posteriori (MAP) estimator, also found by the EM algorithm. For choosing the number of components and the model parameterization, we propose a modified version of BIC, where the likelihood is evaluated at the MAP instead of the MLE. We use a highly dispersed proper conjugate prior, containing a small fraction of one observation's worth of information. The resulting method avoids degeneracies and singularities, but when these are not present it gives similar results to the standard method using MLE, EM and BIC. VL - 24 UR - http://dx.doi.org/10.1007/s00357-007-0004-5 ER -