Early View (EV): 1-EV

Basic ideas for measurement of niche characteristics were fi rst proposed more than four decades ago (Colwell and Futuyma 1971). Initial correlational niche modeling approaches were outlined beginning in the 1970s (Green 1971, Hutchinson 1978), with enormous interest and advancement in the 1990s and 2000s (Peterson et al. 2011). More recently, techniques and toolkits have been developed to make these approaches more tractable (Abrams 1980, R ö dder and Engler 2011); in particular, these approaches have been integrated in tools for model calibration (Phillips and Dud í k, 2008) and for comparative studies of eco-logical niche models in ENMTools (Warren et al. 2008, 2010, Glor and Warren 2011) and NichePy (Bentlage and Shcheglovitova 2012) software packages.

Th e interplay of diff erent factors in geographic ( G ) and environmental ( E ) spaces can be explored via so-called ‘ virtual species ’ . Virtual species are made by software in the form of simulated data that resemble real species, the researcher char-acterizes a priori a limited and known set of causal factors to create a ‘ species ’ of known niche characteristics. Niches and distributions of virtual species can then be explored using modeling techniques; numerous recent studies used virtual species to explore complicated questions in distributional ecology (Hirzel et al. 2001, Elith and Graham 2009, Godsoe 2010, R ö dder and Engler 2011, Barbet-Massin et al. 2012, Saupe et al. 2012, Owens et al. 2013, Meynard et al. 2013, Miller 2014, Moudr ý 2015). Virtual species allow research-ers to avoid problems such as biases related to choice of

Ecography 39: 001–009, 2016 doi: 10.1111/ecog.01961

© 2015 Th e Authors. Ecography © 2015 Nordic Society Oikos Subject Editor: Brody Sandel. Editor-in-Chief: Miguel Ara ú jo. Accepted 7 December 2015

Th e ecological niche has been a central concept in mod-ern ecology (Hutchinson 1957, Leibold 1995, Holt 2009). G. Evelyn Hutchinson presented a formalization of niche concepts more than 50 yr ago (Hutchinson 1978), propos-ing relationships between ecological niches and geographic spaces (which he called ‘ biotopes ’ ; Sober ó n and Nakamura 2009); this linkage is known as the Hutchinsonian dual-ity (Pulliam 2000, Colwell and Rangel 2009). However, tools for visualizing, exploring, and analyzing distributions of species in these linked spaces have remained limited (but see Duan et al. 2015 and Leroy et al. 2015). Recent years have seen massive increases in availability of data on occurrences of species and important environmental dimen-sions, and researchers have developed correlational algo-rithms by which to estimate ecological niches and explore potential distributional areas (Peterson et al. 2011). Th ese correlational models have been referred to as species distri-bution models (Austin 2007, Pearson et al. 2007, Pearson 2010), habitat models (Hirzel and Arlettaz 2003, Guisan and Th uiller 2005), or ecological niche models (Sober ó n and Peterson 2005, Peterson 2006). Sober ó n and Peterson (2005) proposed the biological-abiotic-mobility (BAM) scheme, a heuristic framework linking niche concepts and distributional areas. BAM highlights three important causal factors of species ’ distributions: the geographic distribution of suitable abiotic environmental conditions, the geographic distribution of suitable biotic conditions, and the potential to reach areas by dispersal in relevant time periods.

NicheA: creating virtual species and ecological niches in multivariate environmental scenarios

Huijie Qiao , A. Townsend Peterson , Lindsay P. Campbell , Jorge Sober ó n , Liqiang Ji and Luis E. Escobar

H. Qiao and L. Ji, Key Laboratory of Animal Ecology and Conservation Biology, Inst. of Zoology, Chinese Academy of Sciences, Beijing, 100101 China. – A. T. Peterson (http://orcid.org/0000-0003-0243-2379), L. P. Campbell and J. Sober ó n, Biodiversity Inst., Univ. of Kansas, Lawrence, KS 66045, USA. – L. E. Escobar (lescobar@umn.edu), Minnesota Aquatic Invasive Species Research Center and Dept of Veterinary Population Medicine, Univ. of Minnesota, St Paul, MN 55108, USA.

Robust methods by which to generate virtual species are needed urgently in the emerging fi eld of distributional ecology to evaluate performance of techniques for modeling ecological niches and species distributions and to generate new questions in biogeography. Virtual species provide the opportunity to test hypotheses and methods based on known and unbiased distributions. We present Niche Analyst (NicheA), a toolkit developed to generate virtual species following the Hutchinsonian approach of an n -multidimensional space occupied by the species. Ecological niche models are generated, analyzed, and visualized in an environmental space, and then projected to the geographic space in the form of continuous or binary species distribution models. NicheA is implemented in a stable and user-friendly Java platform. Th e software, online manual, and user support are freely available at < http://nichea.sourceforge.net > .

2-EV

species (Barbet-Massin et al. 2012), biases and error in real-world data sets (Saupe et al. 2012), and unknown true niche dimensions (Broennimann et al. 2012).

Two software packages, SDMvspecies and virtualspecies (Duan et al. 2015 and Leroy et al. 2015), were presented recently to permit researchers to design virtual species by several methods. SDMvspecies software can create virtual species based on ideas from four previous reports: niche syntheses (Hirzel et al. 2001), pick mean method (Jim é nez-Valverde and Lobo 2007), pick median method (Lobo and Tognelli 2011), and artifi cial bell-shaped response method (Varela et al. 2014). In virtualspecies, users can generate virtual species by two approaches – defi ning response func-tions for each environmental variable, and defi ning suitabil-ity from a principal components analysis (PCA; Leroy et al. 2015). Both, virtualspecies and SDMvspecies, were written as R packages (R Development Core Team).

In this paper, we introduced a platform from which to explore Hutchinson ’ s duality in the context of the BAM framework and other niche-related concepts, called Niche Analyst (NicheA). Th is freely available program allows users to create virtual environmental spaces and species, and to analyze ecological niches in both multivariate environmental and geographic spaces, eff ectively linking views of niche and distribution.

Description

NicheA is a cross-platform application released under a GNU Public License (GPL). NicheA is written in Java, combining several other libraries from R (R Development

Core Team), Weka (Mark Hall et al. 2009), JAMA (Hicklin et al. 2012), GDAL (GDAL Development Team 2011), and QuickHull3D (Lloyd 2012; Supplementary material Appendix 1). Th e platform is a window-based, user-friendly application that executes on most common operating sys-tems (Fig. 1), including Microsoft Windows, Mac OS X, and some Linux releases, such as Ubuntu. Th e electronic user manual, which covers all of the functions implemented in NicheA, installation and user support, can be found at < http://nichea.sourceforge.net/ > .

Because biotic interactions introduce enormous complex-ity, and relevant data are generally lacking (Sober ó n 2007, Colwell and Rangel 2009, Holt 2009), NicheA focuses on non-interactive (scenopoetic) variables in Grinnellian niche environments (Sober ó n 2007), thus eff ectively ignoring biotic interactions. Even when ecological niches are related to the environmental space ( E ), the current literature focuses on management, evaluation, and interpretation of ecologi-cal niches in geographic space ( G ). To mitigate biases in G , NicheA allows generating and displaying species ’ niches in E , and visualization of species ’ distributions in linked E and G spaces. Functions and action fl ows in NicheA are illustrated in Fig. 1.

NicheA assumes that species ’ fundamental ecological niches are convex in shape, as available evidence suggests (Birch 1953, Maguire 1967, Hooper et al. 2008, Angilletta 2009, Sober ó n and Nakamura 2009). Niches can thus be operationalized as minimum-volume ellipsoids (Van Aelst and Rousseeuw 2009) or convex polyhedrons (Rissler and Apodaca 2007, Sober ó n and Nakamura, 2009, Lloyd 2012, Monahan and Tingley 2012). NicheA can generate such forms, calculate their volume, density, shape, position, and

Figure 1. Feasible action fl ows in NicheA, showing four function modules. 1) Th e left block is the workfl ow for creating and displaying a background cloud with environmental layers supplied by the user. 2) Th e second block shows generation of occurrence data sets, displaying them in E , and computing their attributes. 3) Th e third block shows a simple function used to design barriers in G . 4) Th e fi nal block contains processes related to interacting with other modeling algorithms. N � virtual niche; BC � background; ENMs � ecological niche models.

3-EV

other attributes, and quantify similarity among multiple niches in terms of overlap in n -dimensional environmental spaces. NicheA generates ecological niche models for virtual species in E in the form of ellipsoids or polyhedrons, and projects results to G . However, in view of the heterogeneous distribution of populations within the niche with high spe-cies abundance in core areas of the niche and low species abundance at the edges (Mart í nez-Meyer et al. 2012, Lira-Noriega and Manthey 2014), NicheA can relax polyhedron assumptions, allowing projection of species distributions based on continuous or logistic functions (Lira-Noriega and Manthey 2014, Leroy et al. 2015).

Niches of virtual species can be generated using three dif-ferent options. 1) Virtual niches can be created based on an ellipsoid drawn manually selecting the shape, position, and size of ellipsoids using the widget in the software interface, which allows ellipsoid manipulation from the computer ’ s mouse or via writing the ellipsoid ’ s dimension for detailed estimations; then, this niche can be exported to the geog-raphy using the ‘ Create a virtual N ’ tool in the ‘ Toolbox ’ of NicheA. 2) Ecological niches can also be generated based on environmental parameters (e.g. temperature range) under diverse suitability distributions established by the user using the ‘ Virtual species – parameter ’ tool. 3) Users interested in generating virtual niches incorporating specifi c geographic areas can create a virtual species based on geographic coordi-nates resembling the site of interest using the ‘ Generate N(s) from occurrences ’ tool.

Furthermore, using NicheA, it is possible to increase complexity of virtual species by, for example, generating dis-persal barriers in G and exploring associated confi gurations in E , to simulate geographic confi gurations of species distri-butions and biogeography. Th us, users are able to generate virtual species of known, user-selected fundamental niche characteristics; NicheA functions are summarized in Table 1. Th ese tools are available in NicheA ver. 3.0 9.14.2015, and steps to develop virtual species are described in Table 2 and Supplementary material Appendix 2.

Environmental variables

Two input data formats are used in NicheA to character-ize environmental landscapes: GeoTIFF and ESRI ASCII raster grid layers. NicheA uses data held in these inputs to draw a background cloud that characterizes the universe of environments available to a virtual species. According to the scientifi c question and user preferences, environ-mental variables can be from original (e.g. temperature), summarized (i.e. principal components; Supplementary material Appendix 3a), or virtual environments. NicheA displays the distribution of the virtual niche in E based on one, two, or three dimensions (Supplementary mate-rial Appendix 3b). One dimension displays are the density and range occupied by the virtual species with respect to each variable, whereas two and three dimensions are visu-alized in the virtual scenario. Users can calibrate models and generate multivariate analyses in NicheA using the n dimensions of E ; however, the cost of overfi tting and redundant data layers in analyses should be considered (Peterson 2007).

Outputs

Diff erent functions in NicheA have diff erent outputs. Th e main output, which is defi ned as the virtual niche, N , in NicheA, is a folder including the following fi les: geographic coordinates of occurrences (ll.txt), plane coordinates (xy.txt), environmental values of the occurrences (values.txt), and the species ’ geographic distribution in both GeoTIFF (present.tiff ) and PNG (present.png) formats. Models can be gener-ated in binary and continuous formats. Continuous models based distribution of suitability using linear and logistic func-tions as in Leroy et al. (2015), but incorporating the ability of applying such functions to all the environmental condi-tions available or restricted to environments falling inside the virtual niche based on a hyperdimensional ellipsoid. Niches from virtual species in the form of ellipsoids can be exported in editable (.elp) and non-editable formats (.mve), that contain the elements of the ellipsoid (the centroid and the covariance matrix) and are readable in NicheA.

Occurrence points drawn randomly from within a vir-tual niche can be exported as a table with geographic coor-dinates mirroring a continuous model with areas of high or low suitability, arranging column labels in formats required by popular ecological niche modeling software packages (e.g. Maxent and OpenModeller; Phillips and Dud í k 2008, Mu ñ oz et al. 2011; Supplementary material Appendix 3c). Models of the virtual species in continuous format from other ecological niche modeling software platforms can be evaluated in NicheA using the Partial-ROC AUC and AIC metrics (Peterson et al. 2008 and Warren and Seifert 2011; Supplementary material Appendix 3d and e). Continuous models in raster format can also be converted to binary based on nine threshold methods (Supplementary material Appendix 3f ). As complementary outputs, NicheA off ers tables and fi gures associated with several analytical tools, including niche overlap measurements in E , and descriptive statistics of environmental variables, occurrence points, and model evaluation (Supplementary material Appendix 3g).

Scientifi c workfl ow management system

Functions in NicheA are connected by the outputs. In other words, an output from one function can be an input of another function. NicheA provides a novel method to man-age connections among functions via a scientifi c workfl ow management system that composes and executes series of computational or data manipulation steps (Fig. 2). Using this tool, users can carry out a complex analysis process, and these workfl ows can be shared, reused, and adapted. Users can download shared workfl ows via the workfl ows reposi-tory (Supplementary material Appendix 3h). Users can also design their own workfl ows, and share them with other scientists to develop replicable ecological niche analyses.

Virtual species example

Th e process to generate virtual species within NicheA is simple, as we will illustrate by means of some worked exam-ple analyses. A basic workfl ow of this example is available as a default in NicheA for new users. First, we created an

4-EV

Table 1. Analytical tools available in NicheA.

Function Tool Description

Background data Principal component analysis Generate a principal component analysis (PCA) from selected raster layers; result is a ‘ background ’ folder to generate a 2D or 3D environmental space. This function also generates fi gures and tables summarizing the analysis and correlation among variables and the R script employed during the analysis.

Draw background – folder Select a background cloud from folder created during the PCA to display a 3D environmental space.

Draw background – fi les Select a background cloud from raster layers selected by the user to display a 2D or 3D environmental space.

Niche simulation Open (N)s Open one or several ecological niche (N) models from NicheA (folders) or other software output (GeoTIFF/ASCII rasters).

Design barriers Open a geographic map that allows the user to create geographic barriers. Areas selected will result in a portion of the virtual niche.

Clear scenario Remove all items from the current environmental space.Create a virtual N Export the virtual niche to the geography based in the current editable

ellipsoid (i.e. white ellipsoid with colorful vertexes).Generate virtual N(s) from

occurrencesDraw a virtual species from a coma delimited fi le (.csv) including the

species ’ name, longitude, and latitude.Generate virtual N(s) from ellipsoid Draw a virtual species from a previous saved minimum-volume ellipsoid

fi le (.mve) not editable.Virtual species – parameter Create a virtual species based on selected physiological parameters.

User can determine the physiological limits of the virtual species and the distribution of the frequency (i.e. uniform, normal, binomial, poisson).

Save current selection Save current ellipsoid selection (.elp) for posterior edition in NicheA.Open a selection Open an editable virtual niche (.elp).Open selections Open more than one virtual niche (.elp).

Niche analysis tools Quantify niche overlap Measure the hypervolume overlap of two or more ellipsoids or convex-polyhedrons.

N attributes Features of the virtual niche.Import/ export Export to ENMs Generate occurrence points from the virtual species to be used in

different ecological niche model software. Occurrences can be drawn from areas with high or low probability based on a logistic function. All occurrences can be exported with different suitability values and then converted from point to raster using the ‘ Convert points to raster ’ tool.

Close N/ export N as continuous raster/ export MVE confi gurations

Close N: remove virtual niches from the background. Export N as continuous raster: generate a species distribution model in

GeoTIFF raster format with continuous values of the distance to the niche centroid.

Export MVE confi gurations: export virtual niche as ellipsoid fi le (.elp) to be open and edit later in an environmental space in NicheA.

Post-ENM analysis Threshold calculator Generate binary maps from continuous models based on 10 different functions. Binary model and occurrences are required. Environmental rasters can also be thresholded to show suitable areas according to each environmental variable.

Occurrence statistics Describe the density distribution in environmental terms of the virtual niche, according to geographic coordinates, a binary niche model, and the environmental values available in the study area.

Change the range of the ENM ’ s result

Normalization technique to convert continuous model from values ranging between 0 and 100 or between 0 and 1.

Calculate AUC value of partial-area ROC approaches

Useful for model comparison. Determines if the continuous models predict independent occurrences better than by chance, considering the area predicted in a binary format. Diverse binary thresholds can be tested. Continuous models and occurrences are required (Peterson 2012).

Calculate AIC/BIC values Useful for model comparison. Determine Akaike information criterion values of Maxent models (Phillips and Dud í k 2008) for model comparisons (Warren and Seifert 2011).

Utility functions Raster conversion Convert raster from GeoTIFF to ASCII fi les and vice versa.Variable statistics Maximum, mean, minimum, and range values of the environmental

variables selected.Read data from multiple raster

layersExtract values from different environmental variables based on one

geographic coordinate.Variable normalization/

standardizationTwo methods available for normalization of environmental variables for

better visualization of the background to create the virtual niche.Set no data value Allows users to change the no data value of raster fi les to fi t with

different software requirements.Convert points to raster Convert occurrence points to raster fi le.

5-EV

Next, we collected 100 occurrences drawn at random from N and generated new virtual niches from this sample. Th e niche estimates that resulted were projected to geog-raphy according to models based on a minimum-volume ellipsoid and a convex polyhedron. A detailed example and explanation are in the online manual (Supplementary mate-rial Appendix 3j). Figure 3 illustrates the completed outputs of the process above.

To display the E and G link, which is a function unique to NicheA, we generated geographic barriers (e.g. ocean, des-ert, river) for the virtual species in the geographic view, and used the ‘ Design barriers ’ tool to divide G into four subsets that could then be displayed in E (Fig. 4, 5). Finally, we mea-sured the overlap between the resulting minimum-volume ellipsoids from the geographic footprint, as divided by the hypothesized geographic barriers. Details of this process can be found in the online manual (Supplementary material Appendix 3k).

Additional analytical tools in multivariate environmen-tal spaces, and the use of NicheA to manage other ENM outputs are described in detail in NicheA ’ s user manual ( < http://nichea.sourceforge.net/ > ). Video tutorials and case studies are available in both English and Spanish ( < http://bit.ly/1O9tmaQ > ).

Discussion

NicheA provides a GUI-based tool by which to generate virtual species based on physiological ranges, geographic coordinates generated by the user, or a manual selection

environmental background using 19 bioclimatic variables from the WorldClim climate data archive (Hijmans et al. 2005) at 10 ’ spatial resolution. In NicheA, we calculated the principal components of the 19 variables, and drew a background cloud using the fi rst 3 principal components in the E viewer. Th en, we created a virtual niche using widgets from the NicheA interface to draw an ellipsoid representing a virtual species ’ niche (Supplementary material Appendix 3i; Fig. 2). To save the virtual species, we used the ‘ Create a virtual N ’ tool. We then visualized the virtual niche, N , in the geographic space based on environmental values inside the minimum-volume ellipsoid we created, and projected the environments corresponding to a convex polyhedron from the same ellipsoid to the geography.

Table 2. Steps to create virtual species using NicheA.

Steps Process Tool

Step 1 Generate background Principal component analysis, draw background – folder

Step 2 Draw virtual species Draw a virtual niche manually using the widget in the left panel of NicheA, generate virtual N(s) from occurrences or virtual species – parameter

Step 3 Import virtual species Save current selection (ellipsoid), export to ENMs (coordinates), export N as continuous raster (map), export MVE confi gurations (ellipsoid)

Step 4 Virtual species analyses Design barriers, quantify niches overlap, N attributes, threshold calculator

Figure 2. NicheA interface. (A) Th e set of tool options by which to generate, analyze, and manage ecological niche models in environmen-tal space. (B) Widget by which to generate virtual species manually using the mouse to move the bars or via entering specifi c dimensions of the niche. (C) Environmental space window where ecological niches can be viewed in a three-dimensional environmental scenario as a background cloud (gray points) with three virtual ecological niches (green, yellow, and red ellipsoids).

6-EV

and geographic spaces. NicheA provides users with tools with which to explore G and E spaces for virtual spe-cies, as opposed to the manipulations in G only typical of

of hypothetical environmental spaces occupied by the vir-tual species. NicheA allows users to explore the Hutchinson duality through visualizations of linked environmental

Figure 3. Process by which to generate a virtual species ’ niche. (A) Virtual species niche based on an ellipsoid generated manually (white ellipsoid) using NicheA ’ s widgets to specify external vertices as environmental semi-axes X, Y, and Z (red points) and the niche centroid (green point). (B) Final virtual ecological niche in the form of a minimum volume ellipsoid (MVE; yellow ellipsoid). (C) Convex polyhe-dron from the virtual ecological niche (CP; green lines). (D) Points falling inside the virtual ecological niche (red points).

Figure 4. Linking environmental and geographic spaces using NicheA. (A) Virtual ecological niche based on a minimum-volume ellipsoid (yellow). (B) Geographic distribution corresponding to the minimum-volume ellipsoid (yellow). (C) Virtual ecological niche based on a convex-polyhedron (green). (D) Geographic distribution corresponding to the convex-polyhedron (green). Notice that niche estimation using minimum-volume ellipsoid generates broader predictions compared to convex-polyhedron.

7-EV

and Leroy et al. 2015), NicheA has a GUI to manage the virtual species ’ niche in a multidimensional environmental space (Table 3), something we found critical for robust study design and accurate interpretation of outputs (Escobar et al. 2014). Beyond creating virtual niches in an environmental space from a variety of possibilities (e.g. manual or via occur-rence data), NicheA allows the user to analyze the features of the virtual niche and to visualize and evaluate continuous model outputs from diverse ecological niche model algo-rithms (e.g. Maxent). A virtual niche from NicheA can also

geographic information systems applications. NicheA gen-erates virtual species based on fi ve approaches: minimum-volume ellipsoids, convex polyhedrons, physiological ranges, and linear and logistic functions (Leroy et al. 2015). Virtual niches are displayed in one, two, or three environmental dimensions, but analyses can be developed in any number of dimensions.

Furthermore, NicheA is a toolkit including several inde-pendent functions with which to analyze the complexity of ecological niche models. Such functions can be arranged, exported, and imported in the form of a workfl ow. By merg-ing diff erent functions via this tool, users can analyze diverse problems creatively, without the limitations of single-func-tion applications. Th us, NicheA is a workbench at which users can address a variety of questions related to species ’ ecological niches and geographic distributions.

Compared to existing tools for creating virtual species including SDMvspecies and virtualspecies (Duan et al. 2015

Table 3. Environmental scenario manipulation in NicheA using the computer ’ s mouse.

Operation Commands

Rotate scenario Left click � mouse movementZoom in/out scenario Alt � left click � mouse movementMove scenario Right click � mouse movement

Figure 5. Linking environmental and geographic space using barriers. (A) Virtual species ’ niche in geographic space showing diff erent por-tions of the original fundamental niche. Th e original virtual species ’ potential distribution split into distinct populations: North America (red), South America (green), South Africa (yellow), and Australia (blue). (B) Each ellipsoid represents a sub-portion of the species ’ funda-mental niche. Notice that although populations of diff erent continents are spatially distant their niches overlap broadly in E .

8-EV

Barbet-Massin, M. et al. 2012. Selecting pseudo-absences for species distribution models: how, where and how many? – Methods Ecol. Evol. 3: 327 – 338.

Bentlage, B. and Shcheglovitova, M. 2012. NichePy: modular tools for estimating the similarity of ecological niche and species distribution models. – Methods Ecol. Evol. 3: 484 – 489.

Birch, L. C. 1953. Experimental background to the study of the distribution and abundance of insects: III. Th e relation between innate capacity for increase and survival of diff erent species of beetles living together on the same food. – Evolution 7: 136 – 144.

Blonder, B. C. et al. 2014. Th e n-dimensional hypervolume. – Global Ecol. Biogeogr. 23: 595 – 609.

Broennimann, O. et al. 2012. Measuring ecological niche overlap from occurrence and spatial environmental data. – Global Ecol. Biogeogr. 21: 481 – 497.

Colwell, R. K. and Futuyma, D. J. 1971. On the measurement of niche breadth and overlap. – Ecology 52: 567 – 576.

Colwell, R. K. and Rangel, T. F. 2009. Hutchinson ’ s duality: the once and future niche. – Proc. Natl Acad. Sci. USA 106: 19651 – 19658.

de Oliveira, S. V. et al. 2013. Potential geographic distribution of hantavirus reservoirs in Brazil. – PLoS One 8: e85137.

Duan, R. Y. et al. 2015. SDMvspecies: a software for creating virtual species for species distribution modelling. – Ecography 38: 108 – 110.

Elith, J. and Graham, C. H. 2009. Do they? How do they? Why do they diff er? On fi nding reasons for diff ering performances of species distribution models. – Ecography 32: 66 – 77.

Escobar L. E. et al. 2014. Potential for spread of white-nose fungus ( Pseudogymnoascus destructans ) in the Americas: using Maxent and NicheA to assure strict model transference. – Geospatial Health 11: 221 – 229.

Escobar L. E. et al. 2015a. First report on bat mortalities on wind farms in Chile. – Gayana 79: 11 – 17.

Escobar, L. E. et al. 2015b. A global map of suitability for coastal Vibrio cholerae under current and future climate conditions. – Acta Trop. 3: 202 – 211.

GDAL Development Team 2011. GDAL – geospatial data abstraction library, version 1.7.0. – < www.gdal.org > .

Glor, R. E. and Warren, D. L. 2011. Testing ecological explana-tions for biogeographic boundaries. – Evolution 65: 673 – 683.

Godsoe, W. 2010. I can ’ t defi ne the niche but I know it when I see it: a formal link between statistical theory and the ecological niche. – Oikos 119: 53 – 60.

Green, R. H. 1971. A multivariate statistical approach to the Hutchinsonian niche: bivalve molluscs of central Canada. – Ecology 52: 544 – 556.

Guisan, A. and Th uiller, W. 2005. Predicting species distribution: off ering more than simple habitat models. – Ecol. Lett. 8: 993 – 1009.

Hicklin, J. et al. 2012. Jama: a Java matrix package. – < http://math.nist.gov/javanumerics/jama/ > .

Hijmans, R. J. et al. 2005. Very high resolution interpolated climate surfaces for global land areas. – Int. J. Climatol. 25: 1965 – 1978.

Hirzel, A. H. and Arlettaz, R. 2003. Modeling habitat suitability for complex species distributions by environmental-distance geometric mean. – Environ. Manage. 32: 614 – 623.

Hirzel, A. H. et al. 2001. Assessing habitat-suitability models with a virtual species. – Ecol. Model. 145: 111 – 121.

Holt, R. D. 2009. Bringing the Hutchinsonian niche into the 21st century: ecological and evolutionary perspectives. – Proc. Natl Acad. Sci. USA 106: 19659 – 19665.

Hooper, H. L. et al. 2008. Th e ecological niche of Daphnia magna characterized using population growth rate. – Ecology 89: 1015 – 1022.

be opened using environmental backgrounds from diff erent geographic areas or past or future climates. Th e ability of NicheA to draw virtual niches from geographic coordinates created by the user, when specifi c areas are selected to be included in the virtual niche, allows importing real coordi-nates from real species. Owing to the versatility of NicheA, the software has already been employed in a diversity of stud-ies using virtual and real species, in fi elds ranging from epi-demiology to evolutionary biology. To date, NicheA has been used to facilitate the study of hantavirus ecology in Brazil (de Oliveira et al. 2013), the potential for the spread of white-nose syndrome in bats (Escobar et al. 2014), species limits in opossums (Tocchio et al. 2014), spatial and environmental distribution of bats (Escobar et al. 2015a), niche divergence in Asian gazelles (Hu et al. 2015), sympatric speciation in butterfl ies (Rosser et al. 2015), eff ects of niche breath on species survival (Saupe et al. 2015), anticipating the global spread of Vibrio cholerae in marine ecosystems (Escobar et al. 2015b), and evaluating performance of ecological niche model algorithms under diverse niche scenarios (Qiao et al. 2015).

NicheA is free, open-source software, available under the GNU Lesser General Public License. All source code can be retrieved via a subversion repository ( < svn://mmweb.animal.net.cn/nichea/trunk > ). Because ecological niche modeling is a rapidly-developing fi eld that is seeing impressive research use in addressing various questions in ecology, NicheA is based in a community platform to meet increasing demand of assistance in the form of forums, videos, an online man-ual, and free remote support. Inclusion of additional model-ling techniques (Blonder et al. 2014) is expected for future NicheA versions. Finally, although NicheA has the capability for developing complex analyses under multivariate environ-mental scenarios; the user-friendly interface makes this soft-ware an ideal resource for teaching in ecology.

To cite NicheA or acknowledge its use, cite this Software note as follows, substituting the version of the application that you used for ‘ version 0 ’ :

Qiao, H., Peterson, A. T., Campbell, L. P., Sober ó n, J., Ji, L. and Escobar, L. E. 2016. NicheA: creating virtual species and ecological niches in multivariate environmental scenarios. – Ecography 39: 000 – 000 (ver. 0).

Acknowledgements – Th e authors thank Sergio Estay for his com-ments on the NicheA installation process. Th is study was supported by the National Natural Science Foundation of China (A New Method to Predict the Species Distributions, 31100390), the Sci-ence and Technology Supporting Project of Ministry of Science, Microsoft Research Grant #47780, and NSF grant DEB-1208472.

References

Abrams, P. 1980. Some comments on measuring niche overlap. – Ecology 61: 44 – 49.

Angilletta, M. J. 2009. Th ermal adaptation: a theoretical and empirical synthesis. – Oxford Univ. Press.

Austin, M. 2007. Species distribution models and ecological the-ory: a critical assessment and some possible new approaches. – Ecol. Model. 200: 1 – 19.

9-EV

Peterson, A. T. 2012. Niche modeling: model evaluation. – Biodivers. Inform. 9: 41.

Peterson, A. T. et al. 2008. Rethinking receiver operating charac-teristic analysis applications in ecological niche modeling. – Ecol. Model. 213: 63 – 72.

Peterson, A. T. et al. 2011. Ecological niches and geographic distributions. – Princeton Univ. Press.

Phillips, S. J. and Dud í k, M. 2008. Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. – Ecography 31: 161 – 175.

Pulliam, H. R. 2000. On the relationship between niche and distribution. – Ecol. Lett. 3: 349 – 361.

Qiao, H. et al. 2015. No silver bullets in correlative ecological niche modeling: insights from testing among many potential algorithms for niche estimation. – Methods Ecol. Evol. 6: 1126 – 1136.

Rissler, L. J. and Apodaca, J. J. 2007. Adding more ecology into species delimitation: ecological niche models and phylogeogra-phy help defi ne cryptic species in the black salamander ( Aneides fl avipunctatus ). – Syst. Biol. 56: 924 – 942.

R ö dder, D. and Engler, J. O. 2011. Quantitative metrics of overlaps in Grinnellian niches: advances and possible draw-backs. – Global Ecol. Biogeogr. 20: 915 – 927.

Rosser, N. et al. 2015. Extensive range overlap between heliconiine sister species: evidence for sympatric speciation in butterfl ies? – BMC Evol. Biol. 15: 125.

Saupe, E. E. et al. 2012. Variation in niche and distribution model performance: the need for a priori assessment of key causal factors. – Ecol. Model. 237: 11 – 22.

Saupe, E. E. et al. 2015. Niche breath and geographic range size as determinates of species survival on geological time scales. – Global Ecol. Biogeogr. 24: 1159 – 1169.

Sober ó n, J. 2007. Grinnellian and Eltonian niches and geographic distributions of species. – Ecol. Lett. 10: 1115 – 1123.

Sober ó n, J. and Peterson, A. T. 2005. Interpretation of models of fundamental ecological niches and species ’ distributional areas. – Biodivers. Inform. 2: 1 – 10.

Sober ó n, J. and Nakamura, M. 2009. Niches and distributional areas: concepts, methods, and assumptions. – Proc. Natl Acad. Sci. USA 106: 19644 – 19650.

Tocchio, L. J. et al. 2014 Niche similarities among white-eared opossums (Mammalia, Didelphidae): is ecological niche mod-elling relevant to setting species limits? – Zool. Scr. 44: 1 – 10.

Van Aelst, S. and Rousseeuw, P. 2009. Minimum volume ellipsoid. – Wiley Interdisc. Rev. Comp. Stat. 1: 71 – 82.

Varela, S. et al. 2014. Environmental fi lters reduce the eff ects of sampling bias and improve predictions of ecological niche models. – Ecography 37: 1084 – 1091.

Warren, D. L. and Seifert, S. N. 2011. Ecological niche modeling in Maxent: the importance of model complexity and the performance of model selection criteria. – Ecol. Appl. 21: 335 – 342

Warren, D. L. et al. 2008. Environmental niche equivalency versus conservatism: quantitative approaches to niche evolution. – Evolution 62: 2868 – 2883.

Warren, D. L. et al. 2010. ENMTools: a toolbox for comparative studies of environmental niche models. – Ecography 33: 607 – 611.

Hu, J. et al. 2015. Niche divergence accelerates the evolution in Asian endemic Procapra gazelles. – Sci. Rep. 5: 10069.

Hutchinson, G. E. 1957. Concluding remarks. – Cold Spring Harb. Symp. 22: 415 – 427.

Hutchinson, G. E. 1978. An introduction to population ecology. – Yale Univ. Press.

Jim é nez-Valverde, A. and Lobo, J. M. 2007. Th reshold criteria for conversion of probability of species presence to either-or presence – absence. – Acta Oecol. 31: 361 – 369.

Leibold, M. A. 1995. Th e niche concept revisited: mechanistic models and community context. – Ecology 76: 1371 – 1382.

Leroy, B. et al. 2015. virtualspecies, an R package to generate virtual species distributions. – Ecography, in press.

Lira-Noriega, A. and Manthey, J. D. 2014. Relationship of genetic diversity and niche centrality: a survey analysis. – Evolution 68: 1082 – 1093.

Lloyd, J. 2012. Quickhull3d: a robust 3d convex hull algorithm in Java. – < www.cs.ubc.ca/ ∼ lloyd/java/quickhull3d.html > .

Lobo, J. M. and Tognelli, M. F. 2011. Exploring the eff ects of quantity and location of pseudo-absences and sampling biases on the performance of distribution models with limited point occurrence data. – J. Nat. Conserv. 19: 1 – 7.

Maguire, B. Jr 1967. A partial analysis of the niche. – Am. Nat. 101: 515 – 526.

Mark Hall, E. F. et al. 2009. Th e weka data mining software: an update. – Sigkdd Explor. 11: 10 – 18.

Mart í nez-Meyer, E. et al. 2012. Ecological niche structure and rangewide abundance patterns of species. – Biol. Lett. 9: 20120637.

Meynard, C. N. et al. 2013. Using virtual species to study species distributions and model performance. – J. Biogeogr. 40: 1 – 8.

Miller, J. A. 2014. Virtual species distribution models: using simulated data to evaluate aspects of model performance. – Prog. Phys. Geogr. 38: 117 – 128.

Monahan, W. B. and Tingley, M. W. 2012. Niche tracking and rapid establishment of distributional equilibrium in the house sparrow show potential responsiveness of species to climate change. – PLoS One 7: e42097.

Moudr ý , V. 2015. Modelling species distributions with simulated virtual species. – J. Biogeogr. 42: 1365 – 1366.

Mu ñ oz, M. E. S. et al. 2011. OpenModeller: a generic approach to species ’ potential distribution modelling. – GeoInformatica 15: 111 – 135.

Owens, H. L. et al. 2013. Constraints on interpretation of ecological niche models by limited environmental ranges on calibration areas. – Ecol. Model. 263: 10 – 18.

Pearson, R. G. 2010. Species ’ distribution modeling for conserva-tion educators and practitioners. – Lessons Conserv. 3: 54 – 89.

Pearson, R. G. et al. 2007. Predicting species distributions from small numbers of occurrence records: a test case using cryptic geckos in Madagascar. – J. Biogeogr. 34: 102 – 117.

Peterson, A. T. 2006. Ecological niche modeling and spatial patterns of disease transmission. – Emerg. Infect. Dis. 12: 1822 – 1826.

Peterson, A. T. 2007. Why not WhyWhere: the need for more complex models of simpler environmental spaces. – Ecol. Model. 203: 527 – 530.

Supplementary material (Appendix ECOG-01961 at < www.ecography.org/appendix/ecog-01961 > ). Appendix 1 – 3.