ShinyBiomod: A new R application for modelling species distribution

Mr Ian Ondo1, Mr Samuel Pironon1

1Royal Botanic Gardens Kew, London, United Kingdom


Factors shaping species distribution vary in space and time and operate at different geographic scales. Understanding and anticipating changes in species ranges is a major challenge for biogeographers and conservation biologists, especially in the context of global change.

Species Distribution Models (SDMs) are statistical models allowing to predict in space and time potential distribution of species relying on correlation between species occurrences and environmental variables. They have been extensively used last decades across a wide range of disciplines such as biogeography, ecology, evolutionnary biology…etc and for various purposes such as helping to set up prioritization in biodiversity conservation plannings, anticipating the impact of climate change or preventing from biological invasion.

Building SDMs is a multi-step process that requires good knowledge in statistical modelling. Standard practice ignoring potential statistical issues arising along the modelling process can lead to misleading interpretation of the results, and ultimately, reducing our ability to detect species range shifts and their drivers. Hence, SDMs remain often unaccessible.

Here, we propose a new modelling toolbox allowing to perfom Species Distribution Modelling using the commonly used R package biomod2. The tool provides a user-friendly interface more accessible for a large audience not having  sufficient skills in programming and statistics. It incorporates tools to account for sampling bias in observation data; limiting (multi-)collinearity issues among predictors; visualizing and exporting the results. It helps the user to get familiar with good practices in Species Distribution Modelling, which is essential to better understand patterns and processes underlying shifts in species range.


I first obtained a Bachelor degree in « Biology of organisms and ecosystems » at the University of Lorraine, Nancy, France, and then, a Master degree in « Ecology » specialized in « Biostatistics and Modelling » at the University Paul Sabatier, Toulouse, France.

During this Master, I started developping my skills in Biostatistics, Modelling and R programming language and increased my already advanced education in Biology, Ecology and related natural sciences.

From my Master program and my following ~ 3 years of experience as a research assistant in different French research institutions, I have been confronted to a very diverse range of ecological studies related to forest ecology (through the study of tree species mortality and tree species distribution), movement ecology (through the study of bird dispersal and fish species migration), or surface hydrology (through the modelling of soil water availability for plants). I dealt with many types of data (e.g. species occurrence and abundance data, proportions of declining trees, capture-mark-recapture data, time series, etc) collected locally or remotely at coarse to very fine spatial and temporal resolutions.

I collaborated with specialists from a wide range of research fields within different institutions. It has not only strengthened my skills in the processing of spatio-temporal data and the modelling of species distribution, but also developped my ability to work in multi-institutional and -disciplinary environments.

Similar Posts