Input variable selection with a simple genetic algorithm for conceptual species distribution models: A case study of river pollution in Ecuador

Species distribution models (SDMs) have received increasing attention in freshwater management to support decision making. Existing SDMs are mainly data-driven and often developed with statistical and machine learning methods but with little consideration of hypothetic ecological knowledge. Conceptual SDMs exist, but lack in performance, making them less interesting for decision management. Therefore, there is a need for model identification tools that search for alternative model formulations. This paper presents a methodology, illustrated with the example of river pollution in Ecuador, using a simple genetic algorithm (SGA) to identify well performing SDMs by means of an input variable selection (IVS). An analysis for 14 macroinvertebrate taxa shows that the SGA is able to identify well performing SDMs. It is observed that uncertainty on the model structure is relatively large. The developed tool can aid model developers and decision makers to obtain insights in driving factors shaping the species assemblage. An approach to improve conceptual species distribution models (SDMs) is presented.Alternative SDM formulations are found with input variables selection (IVS).The use of genetic algorithms as a tool for IVS improves SDM performance.Uncertainties in the structure of the identified models are present.The proposed approach increases insight in freshwater ecosystem functioning.

[1]  Trevor H. Booth,et al.  bioclim: the first species distribution modelling package, its early applications and relevance to most current MaxEnt studies , 2014 .

[2]  Aaron M. Ellison,et al.  Bayesian inference in ecology , 2004 .

[3]  E. Domínguez,et al.  Macroinvertebrados bentônicos sudamericanos: sistemática y biología , 2009 .

[4]  E. Martínez‐Meyer,et al.  Using Ecological‐Niche Modeling as a Conservation Tool for Freshwater Species: Live‐Bearing Fishes in Central Mexico , 2006, Conservation biology : the journal of the Society for Conservation Biology.

[5]  Arthur E. Mynett,et al.  A decision framework for wetland management in a river basin context: The “Abras de Mantequilla” case study in the Guayas River Basin, Ecuador , 2013 .

[6]  Patrick Van Damme,et al.  Application of genetic algorithm and greedy stepwise to select input variables in classification tree models for the prediction of habitat requirements of Azolla filiculoides (Lam.) in Anzali wetland, Iran , 2013 .

[7]  C. E. SHANNON,et al.  A mathematical theory of communication , 1948, MOCO.

[8]  Paolo Vezza,et al.  Random forests to evaluate biotic interactions in fish distribution models , 2015, Environ. Model. Softw..

[9]  D. Hering,et al.  www.freshwaterecology.info – An online tool that unifies, standardises and codifies more than 20,000 European freshwater organisms and their ecological preferences ☆ , 2015 .

[10]  R. Gregory,et al.  Improving species distribution models: the value of data on abundance , 2014 .

[11]  Paolo Vezza,et al.  Comparing four methods for decision-tree induction: A case study on the invasive Iberian gudgeon (Gobio lozanoi; Doadrio and Madeira, 2004) , 2016, Ecol. Informatics.

[12]  Peter Goethals,et al.  Fuzzy knowledge-based models for prediction of Asellus and Gammarus in watercourses in Flanders (Belgium) , 2006 .

[13]  Peter L. M. Goethals,et al.  Data-driven habitat analysis of the Ponto-Caspian amphipod Dikerogammarus villosus in two invaded regions in Europe , 2013, Ecol. Informatics.

[14]  M. Araújo,et al.  Modelling distribution in European stream macroinvertebrates under future climates , 2013, Global change biology.

[15]  Ans Mouton,et al.  Ecological relevance of' performance criteria for species distribution models , 2010 .

[16]  Steven Hamblin,et al.  On the practical usage of genetic algorithms in ecology and evolution , 2013 .

[17]  Tom Fawcett,et al.  An introduction to ROC analysis , 2006, Pattern Recognit. Lett..

[18]  Holger R. Maier,et al.  Data-driven modelling approaches for socio-hydrology: opportunities and challenges within the Panta Rhei Science Plan , 2016 .

[19]  G. Carpenter,et al.  DOMAIN: a flexible modelling procedure for mapping potential distributions of plants and animals , 1993, Biodiversity & Conservation.

[20]  P. Goethals,et al.  Modelling occurrence of roach “Rutilus rutilus” in streams , 2014, Aquatic Ecology.

[21]  Peter Goethals,et al.  Genetic algorithms for optimisation of predictive ecosystems models based on decision trees and neural networks , 2006 .

[22]  D. Hering,et al.  A comparison of modelled and actual distributions of eleven benthic macroinvertebrate species in a Central European mountain catchment , 2015, Hydrobiologia.

[23]  Jorge Soberón,et al.  Niches and distributional areas: Concepts, methods, and assumptions , 2009, Proceedings of the National Academy of Sciences.

[24]  Kalyanmoy Deb,et al.  A Comparative Analysis of Selection Schemes Used in Genetic Algorithms , 1990, FOGA.

[25]  John Bell,et al.  A review of methods for the assessment of prediction errors in conservation presence/absence models , 1997, Environmental Conservation.

[26]  Avi Ostfeld,et al.  Evolutionary algorithms and other metaheuristics in water resources: Current status, research challenges and future directions , 2014, Environ. Model. Softw..

[27]  Holger R. Maier,et al.  Using characteristics of the optimisation problem to determine the Genetic Algorithm population size when the number of evaluations is limited , 2015, Environ. Model. Softw..

[28]  R. Raleigh,et al.  Habitat Suitability Index models and Instream Flow Suitability curves: brown trout , 1984 .

[29]  Bernard De Baets,et al.  Knowledge-based versus data-driven fuzzy habitat suitability models for river management , 2009, Environ. Model. Softw..

[30]  A. Guisan,et al.  An improved approach for predicting the distribution of rare and endangered species from occurrence and pseudo-absence data , 2004 .

[31]  A. Townsend Peterson,et al.  Novel methods improve prediction of species' distributions from occurrence data , 2006 .

[32]  Carsten F. Dormann,et al.  Towards novel approaches to modelling biotic interactions in multispecies assemblages at large spatial extents , 2012 .

[33]  Minar Naomi Damanik-Ambarita,et al.  Ecological water quality analysis of the Guayas river basin (Ecuador) based on macroinvertebrates indices , 2016 .

[34]  Peter Reichert,et al.  The method matters: a guide for indicator aggregation in ecological assessments , 2014 .

[35]  J. Heikkinen,et al.  Testing hypotheses on shape and distribution of ecological response curves , 2010 .

[36]  R. Zarkami,et al.  Modeling habitat preferences of Caspian kutum, Rutilus frisii kutum(Kamensky, 1901) (Actinopterygii, Cypriniformes) in the Caspian Sea , 2015, Hydrobiologia.

[37]  David R. B. Stockwell,et al.  Induction of sets of rules from animal distribution data: a robust and informative method of data analysis , 1992 .

[38]  Damaris Zurell,et al.  Collinearity: a review of methods to deal with it and a simulation study evaluating their performance , 2013 .

[39]  A. E. Eiben,et al.  From evolutionary computation to the evolution of things , 2015, Nature.

[40]  Peter L. M. Goethals,et al.  Development and assessment of ecological models in the context of the European Water Framework Directive: Key issues for trainers in data-driven modeling approaches , 2013, Ecol. Informatics.

[41]  N. LeRoy Poff,et al.  Landscape Filters and Species Traits: Towards Mechanistic Understanding and Prediction in Stream Ecology , 1997, Journal of the North American Benthological Society.

[42]  Bernard De Baets,et al.  Data-driven fuzzy habitat suitability models for brown trout in Spanish Mediterranean rivers , 2011, Environ. Model. Softw..

[43]  Holger R. Maier,et al.  Review of Input Variable Selection Methods for Artificial Neural Networks , 2011 .

[44]  Sovan Lek,et al.  Analysis of macrobenthic communities in Flanders, Belgium, using a stepwise input variable selection procedure with artificial neural networks , 2007, Aquatic Ecology.

[45]  W. Jetz,et al.  Effects of species’ ecology on the accuracy of distribution models , 2007 .

[46]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[47]  Steven Broekx,et al.  Evaluation and comparison of data-driven and knowledge-supported Bayesian Belief Networks to assess the habitat suitability for alien macroinvertebrates , 2015, Environ. Model. Softw..

[48]  M. Termansen,et al.  The use of genetic algorithms and Bayesian classification to model species distributions , 2006 .

[49]  Peter Goethals,et al.  Species distribution models grounded in ecological theory for decision support in river management , 2016 .

[50]  Andrew P. Bradley,et al.  The use of the area under the ROC curve in the evaluation of machine learning algorithms , 1997, Pattern Recognit..

[51]  David R. Anderson,et al.  Model selection and multimodel inference : a practical information-theoretic approach , 2003 .

[52]  M. Austin.  Species distribution models and ecological theory: A critical assessment and some possible new approaches , 2007 .

[53]  Omri Allouche,et al.  Assessing the accuracy of species distribution models: prevalence, kappa and the true skill statistic (TSS) , 2006 .

[54]  Andrea Castelletti,et al.  An evaluation framework for input variable selection algorithms for environmental data-driven models , 2014, Environ. Model. Softw..

[55]  B. Baets,et al.  Fuzzy rule-based macroinvertebrate habitat suitability models for running waters , 2006 .

[56]  S. Manel,et al.  Evaluating presence-absence models in ecology: the need to account for prevalence , 2001 .

[57]  Peter Goethals,et al.  Impact of abundance data errors on the uncertainty of an ecological water quality assessment index , 2016 .

[58]  A. Hirzel,et al.  Habitat suitability modelling and niche theory , 2008 .

[59]  W. D. Kissling,et al.  The role of biotic interactions in shaping distributions and realised assemblages of species: implications for species distribution modelling , 2012, Biological reviews of the Cambridge Philosophical Society.

[60]  Alain F. Zuur,et al.  A protocol for data exploration to avoid common statistical problems , 2010 .

[61]  Peace Liz Sasha Musonge,et al.  Bayesian belief network models to analyse and predict ecological water quality in rivers , 2015 .

[62]  Antoine Guisan,et al.  SESAM – a new framework integrating macroecological and species distribution models for predicting spatio‐temporal patterns of species assemblages , 2011 .

[63]  Niels De Pauw,et al.  Method for biological quality assessment of watercourses in Belgium , 2004, Hydrobiologia.

[64]  J. Lennon,et al.  Incorporating uncertainty in predictive species distribution modelling , 2012, Philosophical Transactions of the Royal Society B: Biological Sciences.

[65]  Joseph H. A. Guillaume,et al.  Assessing certainty and uncertainty in riparian habitat suitability models by identifying parameters with extreme outputs , 2014, Environ. Model. Softw..

[66]  B. De Baets,et al.  Abundance versus presence/absence data for modelling fish habitat preference with a genetic Takagi–Sugeno fuzzy system , 2012, Environmental Monitoring and Assessment.

[67]  Peter Goethals,et al.  Multimetric Macroinvertebrate Index Flanders (MMIF) for biological assessment of rivers and lakes in Flanders (Belgium) , 2010 .

[68]  Wilfried Thuiller,et al.  Accounting for dispersal and biotic interactions to disentangle the drivers of species distributions and their abundances. , 2012, Ecology letters.

[69]  Peter L. M. Goethals,et al.  Comparison of modelling techniques to predict macroinvertebrate community composition in rivers of Ethiopia , 2010, Ecol. Informatics.

[70]  David E. Goldberg,et al.  Designing a competent simple genetic algorithm for search and optimization , 2000 .

[71]  J. Grinnell.  The Niche-Relationships of the California Thrasher , 1917 .

[72]  Antoine Guisan,et al.  Predictive habitat distribution models in ecology , 2000 .

[73]  P. Goethals,et al.  Use of genetic algorithms to select input variables in decision tree models for the prediction of benthic macroinvertebrates , 2003 .

[74]  W. Thuiller,et al.  Predicting species distribution: offering more than simple habitat models. , 2005, Ecology letters.

[75]  P. Hernandez,et al.  The effect of sample size and species characteristics on performance of different species distribution modeling methods , 2006 .

[76]  T. Hastie,et al.  Bias correction in species distribution models: pooling survey and collection data for multiple species , 2014, Methods in ecology and evolution.

[77]  P. Goethals,et al.  Effect of water quality on blackflies (Diptera: Simuliidae) in Flanders (Belgium) , 2014 .

[78]  Holger R. Maier,et al.  A genetic algorithm calibration method based on convergence due to genetic drift , 2008, Inf. Sci..

[79]  Bernard De Baets,et al.  Habitat prediction and knowledge extraction for spawning European grayling (Thymallus thymallus L.) using a broad range of species distribution models , 2013, Environ. Model. Softw..

[80]  M. Austin.  Spatial prediction of species distribution: an interface between ecological theory and statistical modelling , 2002 .