A taxonomic-based joint species distribution model for presence-only data

Area of Science:

• Ecology
• Biodiversity Science
• Computational Biology

Background:

• Species distribution models (SDMs) are crucial for mapping taxa and addressing biodiversity loss.
• Traditional SDMs often require presence-absence data, which is scarce globally; most available data are presence-only.
• Existing presence-only SDMs make assumptions about absences and typically focus on single species, limiting their utility with large biodiversity databases.

Purpose of the Study:

• To develop a novel Bayesian-based SDM that directly utilizes presence-only data for multiple species.
• To overcome the limitations of traditional SDMs by inferring absences and sampling effort from presence-only occurrences.
• To enhance the utility of global biodiversity databases like GBIF.

Main Methods:

• Developed a Bayesian SDM framework that models the joint distribution of ecological processes.
• The model directly processes presence-only data, inferring both ecological niches and sampling effort per taxon.
• Applied the model to two case studies: diverse taxa in central Mexico and the Cactaceae family in continental Mexico.

Main Results:

• The Bayesian SDM successfully identified ecological and sampling effort processes for each taxon using only presence data.
• The model demonstrated effectiveness in handling presence-only data for multiple species simultaneously.
• Case studies confirmed the model's ability to infer critical distribution and sampling information.

Conclusions:

• The new Bayesian SDM offers a powerful alternative for analyzing presence-only data in biodiversity research.
• This method enhances the use of large-scale biological occurrence datasets, improving spatial mapping of taxa.
• The approach provides a robust framework for understanding species distributions and sampling biases.