Phylogenetic path analysis reveals the importance of niche-related biological traits on geographic range size in mammals

摘要: MIGUELA. OLALLA-T ARRAGA, ERIK JOAQU IN TORRES-ROMERO,TALITA FERREIRA AMADO andPABLO A. MARTINEZDepartment of Biology and Geology, Physics Inorganic Chemistry, Rey Juan Carlos University, 28933-Mostoles, Madrid,SpainThe geographic range a species is arguably the basicunit in biogeography macroecology (Brown et al.,1996). In particular, there has been long-standinginterest understanding mechanisms that shapethe immense interspecific variation rangesize, question often framed around Rapoport’s rule(Whitton al., 2012). As an emergent species-level trait,range sizes reflect interplay ecological evolu-tionary processes are utmost importance for pre-dicting speciation-extinction dynamics (Jablonski,2008). Species tend to have higher risk extinction ifthey occupy small (Purvis al.,2000), which also places investigation patternsand species’ asa central applied conservation science.In recent paper, Di Marco & Santini (2015, here-after DM&S) analysed determinants ofrange size extant terrestrial mammals glob-ally. They concluded extrinsic factors (climate andhuman impacts), not intrinsic biological traits, themost influential variables. This study brings tablethe considering anthropogenic effects inmacroecological research. Surprisingly, even thebest-studied taxa such as mammalsand birds, workers traditionally overlooked theinfluence human pressures on observed patterns.We ourselves called attention need toincorporate impact metrics, humanfootprint or accessibility, better understand spa-tial distribution mammal some bio-geographical realms (Torres-Romero Olalla-Tarraga,2015). However, we do agree effectsprevail over traits determining rangesizes mammalian would like call theattention few conceptual methodologicalaspects DM&S’s analyses interpretation arenot at all correct our view.First, DM&S neglect relevance phylogeneticrelatedness species.Intrinsic autoecological features oftenreflect shared ancestry, so closely related specieswill share similar possiblysimilar 1996, Jablonski, 2008).A number comparative methods designedto address potential phylogenetic correlation issues inmodel residuals (as long underlyingrobust hypothesis, case mam-mals). Second, ignore therelationship niche breadth-range biologicalexplanation (Slatyer 2013). The concept ecologi-cal was indeed formalized describe set ofbiotic abiotic conditions where can persistand maintain stable population sizes. Its projection ontogeographic space (i.e. duality niche-biotope) inex-tricably linked spatial speciesand highly relevant analyse biogeographical pat-terns (Colwell Rangel, 2009). only incorporatea simplistic diet category variable classifies mam-mals terms trophic position carnivores, herbi-vores omnivores), but consider breadth.Perhaps more importantly, characterized andincluded realized climatic niches ofeach (see Olalla-Tarraga al.,2011 estimation method), erroneouslyreferred them These measuresof environmental tolerance breadth typically astrong positive relationship with (Slatyeret Third, use random forest regressionmodels allow estimating direct betweeneach predictor response variable, cannotcalculate indirect via other dependent variables.We used recently developed phylogeneticconfirmatory path analysis (von Hardenberg Gonz-alez-Voyer, 2013) included previously untestedvariables provide reassessment themain globally(for methodological details see supplementary infor-mation). We find both nonvo-lant chiropterans strongly depend ontheir thermal hydric niches, biologicalproperty, followed by secondary effect