Using Remote Sensing for Modeling and Monitoring Species Distributions

Abstract

Interpolated climate surfaces have been widely used to predict species distributions and develop environmental niche models. However, the spatial coverage and density of meteorological sites used to develop these surfaces vary among countries and regions, such that the most biodiverse regions often have the most sparsely sampled climatic data. We explore the potential of satellite remote sensing (S-RS) products—which have consistently high spatial and temporal resolution and nearly global coverage—to quantify species-environment relationships that predict species distributions. We propose several new environmental metrics that take advantage of high temporal resolution in S-RS data and compare these approaches to classic climate-only approaches using the live oaks (Quercus section Virentes) as a case study. We show that models perform similarly but for some species, particularly in understudied regions, show less precision in predicting spatial distribution. These results provide evidence supporting efforts to enhance environmental niche models and species distribution models (ENMs/SDMs) with S-RS data and, when combined with other approaches for species detection, will likely enhance our ability to monitor biodiversity globally.

Monitoring biodiversity in the Anthropocene using remote sensing in species distribution models

“…land surface temperature, water availability, topography, land cover, and 3D structure, Section 2). Fewer attempts have been made to take advantage of the existing time series data and the dynamic information contained in remote sensing data products (Fernández et al, 2016;Pinto-Ledezma and Cavender-Bares, 2020), despite the pivotal role that such temporally explicit data play. For instance, long-term time series of remote sensing data are key to test the temporal transferability of SDMs (Yates et al, 2018), a basic requirement to formally guide and inform monitoring strategies in changing environments and make sure that model projections follow the observed trajectories of species.…”

Predicting species distributions and community composition using remote sensing products

“…Biodiversity models, such as the S-SDMs, depend on the reliability of individual species models (SDM) (Guisan and Rahbek, 2011). SDMs represent abstractions of species Hutchinsonian niches, i.e., species distributions are constrained by both, the abiotic environment and biotic interactions with other species (Hutchinson's duality sensu Colwell and Rangel, 2009; see also, Peterson et al, 2011;Pinto-Ledezma and Cavender-Bares, 2020). Our results show that individual oak SDMs constructed using environmental covariates derived from RS-products, have good accuracy (Fig.…”

“…Remote sensing products used as environmental inputs to our models included Leaf Area Index (LAI), obtained from MODIS Terra/Aqua MOD15A2 over a 15-year period (2001 -2015) using the interface EOSDIS Earthdata (Myneni et al, 2002); precipitation from Climate Hazards group Infrared Precipitation with Stations (CHIRPS); and altitude or mean elevation from the Shuttle Radar Topography Mission (SRTM). The 15-year LAI composite provides a representation of the spatial variation of biophysical variables of different components of vegetation and ecosystems over the course of this time frame (Hobi et al, 2017;Pinto-Ledezma and Cavender-Bares, 2020). LAI strongly co-varies with the physical environment; for instance, higher LAI is associated with warmer, wetter and stable environments, whereas lower LAI with cooler, drier and less stable environments (Gower et al, 1999;Myneni et al, 2002;Reich, 2012).…”

Predicting species distributions and community composition using satellite remote sensing predictors

“…SDMs represent abstractions of species Hutchinsonian niches, in which species distributions are constrained by both the abiotic environment and biotic interactions with other species (Hutchinson's duality sensu 40 see also 30,41 ). Our results show that individual oak SDMs constructed using environmental covariates derived from RS-products, have good accuracy (Table S1).…”

“…Remote sensing products (RS-products) have been increasingly used to derive metrics that allow tracking biodiversity from space 18,25,26 , monitoring the state of human impacts 3,27 , as predictors for describing large patterns of species diversity [28][29][30] or to derive Essential Biodiversity Variables, i.e., measures that allow the detection and quanti cation of biodiversity changes 12,31 . Despite their high spatial and temporal resolution, quasi-global coverage and range of data products (e.g., precipitation, plant productivity, biophysical variables, land cover), RS-products have been rarely used as predictors for biodiversity models 17,30,32 . RS-products have been dubbed important "next-generation" environmental predictors in biodiversity models 33 , given that remote sensing continuously captures an increasing range of Earth's biophysical features at global scale 34,35 , avoiding the uncertainty associated with environmental predictors derived from traditional climatic data.…”