Node-based measures of connectivity in genetic networks.

Area of Science:

• Ecology
• Genetics
• Conservation Biology

Background:

• At-site environmental conditions significantly influence genetic connectivity, particularly during dispersal's immigration and settlement phases.
• However, landscape genetic analyses rarely incorporate these crucial at-site processes.
• Network approaches offer a framework to model site-level effects on connectivity.

Purpose of the Study:

• To compare the performance of seven node-based genetic connectivity metrics using simulated genetic networks.
• To identify metrics suitable for assessing the impact of at-site habitat conditions on genetic connectivity.
• To apply a selected metric to understand American marten genetic connectivity in Ontario, Canada.

Main Methods:

• Simulated genetic networks with varying node connectivity by adjusting migration rates and recipient nodes.
• Evaluated seven node-based genetic connectivity metrics for their responsiveness to simulated connectivity changes.
• Applied the average inverse edge weight metric to empirical genetic data from American martens in Ontario.

Main Results:

• Average edge weight and average inverse edge weight metrics showed a linear relationship with simulated genetic connectivity.
• Node degree was found to be an unreliable measure of connectivity.
• Highly connected nodes for American martens corresponded to high-quality habitat (deep snow, coniferous/mature forests) and greater distance from the range edge.

Conclusions:

• Node-based metrics, specifically average edge weight and average inverse edge weight, are recommended for modeling at-site influences on genetic connectivity.
• These metrics can effectively capture the effects of local habitat conditions on the immigration and settlement stages of dispersal.
• The study highlights the importance of habitat quality and range position in shaping genetic connectivity for American martens.