On plotting species abundance distributions.

Area of Science:

• Ecology
• Biodiversity Science
• Statistical Ecology

Background:

• Species abundance distribution (SAD) models are crucial for understanding biodiversity.
• Recent interest in SAD models was spurred by claims of log-normal distribution underestimating rare species.
• The neutral Zero Sum Multinomial (ZSM) distribution was developed to better fit observed species-rich assemblage data.

Purpose of the Study:

• To investigate the impact of different binning methods on the visualization and statistical fitting of SAD models.
• To evaluate the consistency of model fits across various binning strategies using empirical data.

Main Methods:

• Six distinct binning methods were applied to the Barro Colorado Island (BCI) tropical tree data for plotting SADs.
• Visual comparison of SAD curve shapes generated by different binning techniques.
• Assessment of how binning affects the goodness-of-fit for ZSM and log-normal models.

Main Results:

• Different binning methods produced markedly different SAD curve appearances for the same dataset.
• The choice of binning method influenced the statistical fit of SAD models, including ZSM and log-normal distributions.
• No universally agreed-upon binning method for SAD plots currently exists.

Conclusions:

• Binning introduces variability and potential errors in SAD analysis, affecting model evaluation.
• A simple doubling method for binning is suggested for illustrative purposes, or rank-abundance plots should be considered.
• Exact methods for model fitting and testing, which bypass data binning, are recommended to avoid unnecessary errors and improve accuracy.