A simulation study comparing common methods for analyzing species–habitat associations of plants

Hesselbarth, M. & Wiegand, K. (2024). A simulation study comparing common methods for analyzing species–habitat associations of plants. Journal of Vegetation Science 35 (2) e13243. 10.1111/jvs.13243.

Preview

Text J Vegetation Science - 2024 - Hesselbarth - A simulation study comparing common methods for analyzing species habitat.pdf - Published Version Available under License Creative Commons Attribution Non-commercial. Download (2MB) | Preview

Abstract

Question
Species-specific habitat associations are one of several processes that lead to a clustered spatial pattern of plant populations. This pattern occurs in tropical and temperate forests. To analyze species–habitat associations, four methods are commonly used when determining species–habitat associations from spatial point pattern and environmental raster data. Two of the methods randomize the spatial point pattern of plants, and two randomize the raster data of habitat patches. However, the strengths and weaknesses of the four methods have never been analyzed in detail.

Methods
We conducted a simulation study to analyze the strengths and weaknesses of the four most used methods. The methods are the gamma test, pattern reconstruction, the torus-translation test and the randomized-habitats procedure. We simulated neutral landscapes representing habitat patches and point patterns representing fine-scale plant distributions. We built into our simulations known positive and negative species–habitat associations.

Results
All four methods were equally good at detecting species–habitat associations. Detected positive associations better than negative ones. Furthermore, correct detections were mostly influenced by the initial spatial distribution of the point patterns, landscape fragmentation and the number of simulated null model randomizations.

Conclusions
The four methods have advantages and disadvantages, and which is the most suitable method largely depends on the characteristics of the available data. However, our simulation study shows that the results are consistent between methods.

Actions (login required)

View Item