Separating Sampling Bias From Abundance Shows That Different Methods Catch Different Wild Bees

Area of Science:

• Ecology
• Entomology
• Biodiversity Research

Background:

• Ecological sampling methods can introduce taxonomic biases, altering the perceived relative abundances of taxa.
• Accurate interpretation of community data necessitates evaluating and understanding these sampling method biases.
• Wild bees are frequently studied using various sampling techniques, but comparisons often confound bias with sampling effort.

Purpose of the Study:

• To compare the biases of pan traps, vane traps, and hand netting for sampling wild bee communities.
• To assess how taxonomic biases differ between sampling methods when controlling for the number of individuals collected.
• To identify specific bee genera that are over- or underrepresented by different sampling techniques.

Main Methods:

• Simultaneous sampling of wild bee communities in the northeastern United States using pan traps, vane traps, and hand netting.
• Comparison of sample richness and composition between paired methods, adjusting for the total number of bees sampled.
• Analysis of relative abundance differences for 21 bee genera across the different sampling methods.

Main Results:

• Hand netting yielded higher species richness per sampled individual compared to pan traps and vane traps.
• Pan traps captured a distinct species pool and showed significant over- or underrepresentation for multiple bee genera compared to netting and vane traps.
• Very large-bodied bee genera were notably underrepresented in pan trap samples relative to other methods.

Conclusions:

• Wild bee community sampling methods exhibit pervasive and significant taxonomic biases.
• Relative abundances of bee genera can be distorted by sampling methods, even when absolute abundance is higher.
• Understanding and quantifying taxon-specific biases is crucial for robust ecological analysis, as bias is an inherent aspect of community sampling.