penvs

In wallace: A Modular Platform for Reproducible Modeling of Species Niches and Distributions

Component: Process Environmental Data

ORIENTATION

Environmental data may be processed in many ways for niche/distributional modeling: reclassification (re-categorizing the original cell values), resampling (changing the cell size), masking (cropping the grid based on a specified shape), etc. At present, Component: Process Environmental Data addresses one critical step: selecting a study region for analysis by masking the predictor variable grids. This determines the spatial extent for model building, including the environmental characteristics of the data used for comparison with records of the species' presence (e.g., the 'background' data; Peterson et al. 2011 chap. 7). By doing so, it also sets the range (and combinations) of environmental conditions associated with model building, allowing later flagging of 'non-analogous' conditions (e.g., when applying a model to another region or time period; Williams and Jackson 2007).

Selection of a study region is critical for niche/distribution modeling approaches that compare the environments associated with species presence data with those of some comparison dataset (including the Maxent approach; see Component: Build and Evaluate Niche Model). This decision defines the extent of grid cells for the comparison dataset (absences provided by the user, or pseudoabsence or background points sampled by the algorithm; Anderson 2013). Various, sometimes conflicting, principles for selecting a study region have been set forth in the literature (Peterson et al. 2011 chap. 7). Many researchers emphasize the geographic or environmental domain over which to build the model, whereas others suggest identification of areas where the species is likely to be at environmental equilibrium with the predictor variables (Vanderwal et al. 2009; Anderson and Raza 2010; Barve et al. 2011; Saupe et al. 2012; Franklin 2010 chap. 4; Anderson 2013; Merow et al. 2013). For example, one critical guiding principle (when estimates of suitability are desired, especially for use in other places and times) is that the study region should not include geographic areas that the species does not inhabit due to dispersal barriers. Inclusion of such areas will send a false negative signal, biasing the model's estimated response to the environment (Anderson 2015). Despite these theoretical suggestions, operational selection of the study region usually still depends on expert opinion and available natural history information (Acevedo et al. 2012; Gerstner et al. 2018).

Wallace provides several alternative ways for delimiting a study region. The first three options are based on shapes around the occurrence data, found under Module: Select Study Region: 1) a rectangular “bounding box” around occurrence localities, 2) a convex shape drawn around occurrence localities with minimized area (minimum convex polygon), or 3) buffers around occurrence localities. Users also have the option of drawing a study region by using the polygon drawing tool (Module: Draw Study Region. Alternatively, users can upload a polygon (Module: User-specified Study Region).

Any of these options can then be buffered by a user-defined distance in degrees.

After choosing a way to delimit the study region (Step 1), Wallace samples background points (= pixels; Step 2) from the environmental data according the number provided by the user. If that number is smaller than the total in the environmental data (within the chosen background extent), some environmental conditions may be missed in the sample. Depending upon the variables used in the final model, such a situation may lead to the need for environmental extrapolation in order to make a prediction to the full background extent (see Components: Build and Evaluate Niche Model, Visualize Model Results, and Model Transfer; Guevara et al. 2018). Conversely, if the user-specified number is higher than the total in the environmental data, a warning message advising the use of a smaller number will appear, including the maximum number of points allowed (the number of background points cannot be higher than total available in the chosen background extent).

REFERENCES

Acevedo, P., Jiménez‐Valverde, A., Lobo, J.M., & Real, R. (2012). Delimiting the geographical background in species distribution modelling. Journal of Biogeography, 39(8), 1383-1390. DOI: 10.1111/j.1365-2699.2012.02713.x

Anderson, R.P. (2015). El modelado de nichos y distribuciones: no es simplemente "clic, clic, clic." [With English and French translations: Modeling niches and distributions: it's not just "click, click, click" and La modélisation de niche et de distributions: ce n'est pas juste "clic, clic, clic"]. Biogeografía, 8, 4-27. pdf

Anderson, R.P. (2013). A framework for using niche models to estimate impacts of climate change on species distributions. Annals of the New York Academy of Sciences, 1297(1), 8-28. DOI: 10.1111/nyas.12264

Anderson, R.P., & Raza A. (2010). The effect of the extent of the study region on GIS models of species geographic distributions and estimates of niche evolution: preliminary tests with montane rodents (genus Nephelomys) in Venezuela. Journal of Biogeography, 37(7), 1378-1393. DOI: 10.1111/j.1365-2699.2010.02290.x

Barve, N., Barve, V., Jiménez-Valverde, A., Lira-Noriega, A., Maher, S.P., Peterson A.T., Soberón J., & Villalobos F. (2011). The crucial role of the accessible area in ecological niche modeling and species distribution modeling. Ecological Modelling, 222(11), 1810-1819. DOI: 10.1016/j.ecolmodel.2011.02.011

Franklin, J. (2010). Mapping Species Distributions: Spatial Inference and Prediction. Data for species distribution models: the biological data. In: Mapping species distributions: spatial inference and prediction. Cambridge: Cambridge University Press. DOI: 10.1017/CBO9780511810602

Gerstner, B.E., Kass, J.M., Kays, R., Helgen, K.M., & Anderson, R.P. (2018). Revised distributional estimates for the recently discovered olinguito (Bassaricyon neblina), with comments on natural and taxonomic history. Journal of Mammalogy, 99(2), 321-332. DOI: 10.1093/jmammal/gyy012

Guevara, L., Gerstner, B.E., Kass, J.M., & Anderson, R.P. (2018). Toward ecologically realistic predictions of species distributions: A cross-time example from tropical montane cloud forests. Global Change Biology, 24(4), 1511-1522. DOI: 10.1111/gcb.13992

Merow, C., Smith, M.J., & Silander, J.A. (2013). A practical guide to MaxEnt for modeling species' distributions: what it does, and why inputs and settings matter. Ecography, 36(10), 1058-1069. DOI: 10.1111/j.1600-0587.2013.07872.x

Peterson, A.T., Soberón, J., Pearson, R.G., Anderson, R.P., Martinez-Meyer, E., Nakamura M., & Araújo M.B. (2011). Modeling Ecological Niches. In: Ecological Niches and Geographic Distributions. Princeton, New Jersey: Monographs in Population Biology, 49. Princeton University Press. DOI: 10.23943/princeton/9780691136868.003.0005

Saupe, E.E., Barve, V., Myers, C.E., Soberón, J., Barve, N., Hensz, C.M., Peterson, A.T., Owens, H.L., & Lira-Noriega, A. (2012). Variation in niche and distribution model performance: the need for a priori assessment of key causal factors. Ecological Modelling, 237-238, 11-22. DOI: 10.1016/j.ecolmodel.2012.04.001

Williams, J.W., & Jackson, S.T. (2007). Novel climates, no-analog communities, and ecological surprises. Frontiers in Ecology and the Environment, 5(9), 475-482. DOI: 10.1890/070037

VanDerWal, J., Shoo, L.P., Graham, C., & Williams, S.E. (2009). Selecting pseudo-absence data for presence-only distribution modeling: How far should you stray from what you know?. Ecological Modelling, 220(4), 589-594. DOI: 10.1016/j.ecolmodel.2008.11.010

wallace documentation built on Sept. 11, 2024, 9:16 p.m.