Operationalizing expert knowledge in species' range estimates using diverse data types

Frontiers of Biogeography

article

Gonzalo E. Pinilla-Buitrago¹  Jorge Velí1squez-Tibatí1³  Robert P. Anderson¹  Mary E. Blair⁵  Cory Merow⁶  Peter J. Galante⁵  Jamie M. Kass⁷  Matthew E. Aiello-Lammens⁸  Cecina Babich Morrow⁵  Beth E. Gerstner^1,10  Valentina Grisales Betancur^1,11  Alex C. Moore⁵  Elkin A. Noguera-Urbano^1,12 

[1] Department of Biology, City College of New York, City University of New York;Graduate Center, City University of New York;International Alliances Program, National Audubon Society;Division of Vertebrate Zoology;Center for Biodiversity and Conservation;Eversource Energy Center and Department of Ecology and Evolutionary Biology,University of Connecticut;Biodiversity and Biocomplexity Unit, Okinawa Institute of Science and Technology Graduate University;Department of Environmental Studies and Science, Pace University;Spring Health;Department of Fisheries & Wildlife and Ecology, Evolution, and Behavior Program, Michigan State University;Universidad EAFIT;Instituto de Investigación de Recursos Biológicos

关键词: biotic interactions;    dispersal barriers;    ecological niche models;    expert map;    land cover;    land use;    realized distribution;    reproducibility;    species distribution models;   

DOI  :  10.21425/F5FBG53589

学科分类：社会科学、人文和艺术（综合）

来源: International Biogeography Society

【 摘 要 】

Estimates of species’ ranges can inform many aspects of biodiversity research and conservation-management decisions. Many practical applications need high-precision range estimates that are sufficiently reliable to use as input data in downstream applications. One solution has involved expert-generated maps that reflect on-the-ground field information and implicitly capture various processes that may limit a species’ geographic distribution. However, expert maps are often subjective and rarely reproducible. In contrast, species distribution models (SDMs) typically have finer resolution and are reproducible because of explicit links to data. Yet, SDMs can have higher uncertainty when data are sparse, which is an issue for most species. Also, SDMs often capture only a subset of the factors that determine species distributions (e.g., climate) and hence can require significant post-processing to better estimate species’ current realized distributions. Here, we demonstrate how expert knowledge, diverse data types, and SDMs can be used together in a transparent and reproducible modeling workflow. Specifically, we show how expert knowledge regarding species’ habitat use, elevation, biotic interactions, and environmental tolerances can be used to make and refine range estimates using SDMs and various data sources, including high-resolution remotely sensed products. This range-refinement approach is primed to use various data sources, including many with continuously improving spatial or temporal resolution. To facilitate such analyses, we compile a comprehensive suite of tools in a new R package, maskRangeR, and provide worked examples. These tools can facilitate a wide variety of basic and applied research that requires high-resolution maps of species’ current ranges, including quantifications of biodiversity and its change over time.

【 授权许可】

CC BY