Русский

Conference publications

Abstracts

XXVII conference

Quantifying a species ecological niche using SDM (species distribution models)

Orlov M.A., Sheludkov A.V.1

Institute of Cell Biophysics of RAS

1Institute of Geography of RAS

1 pp. (accepted)

Ecological niche represents a basic problem of ecology and biology. It has been addressed actively for a solid century; this yielded multiple dramatically varying definitions. A fruitful approach was proposed by Hutchinson who attributed niche to species rather than geographical space. According to Hutchinson, ecological niche can be drawn in a multidimensional space with ecological factors along the axes. This hypervolume can be transformed into corresponding geographical environment; this two-way transitions represents Hutchinson's duality [1]. The idea behind ecological niche serves as a basis for Species Distribution Modelling (SDM). This type of ecological modelling employs machine learning and uses environment covariates in order to predict species spatial distribution [2].

Here we used SDM as a tool for estimating ecological niche of four most presented (over 50 observations) representatives of Crataegus phylum which inhabit Crimea. The data came from GBIF biodiversity database [3]. These served as a training set. Next, for each of the localities as well as for 100 randomly picked background points environmental variables were taken from Worldclim 2.0 dataset [4]. R free software was used to train models based on Random Forest algorithm (10 for each species); their excellent performance was evidenced by ROC. Finally, we extracted the importance of the environmental variables for each best-performance model. In our suggestion, this impact of covariates reflects real-life species ecological demands and thus may be useful in ecological niche quantitative estimates.

References

1) Colwell, R. K., & Rangel, T. F. (2009). Hutchinson’s duality: The once and future niche. Proceedings of the National Academy of Sciences, 106(Supplement_2), 19651–19658. https://doi.org/10.1073/pnas.0901650106

2) Orlov, M., & Sheludkov, A. (2019). Bioclimatic Data Optimization for Spatial Distribution Models. In Springer Proceedings in Earth and Environmental Sciences (pp. 86–95). https://doi.org/10.1007/978-3-030-11720-7_13

3) https://www.gbif.org/ru/

4) https://www.worldclim.org/



© 2004 Designed by Lyceum of Informational Technologies №1533