Bee abundance estimates vary by collection method and flowering richness

Monitoring bee populations is becoming increasingly important and commonplace, but do current methods produce reliable estimates of bee communities? Authors Marirose Kuhlman and Philip Hahn explore this question in their latest research.

Wild bees are the main pollinators in nearly all terrestrial ecosystems and are essential to the reproductive cycles of many native plants, agricultural crops, and to the success of habitat restoration projects. Because they are essential for maintaining ecosystem functions, conservation practitioners are often interested in quantifying how bee communities respond to landscape degradation, urban or agricultural intensification, or if bee communities are positively influenced by restoration or conservation efforts.

In 2014, as part of our conservation goals at MPG Ranch, a private conservation property in western Montana, USA, we deployed a large bee sampling initiative following the Bee Inventory Plot – a sampling design that could help us achieve our goals and was also appropriate for conservation land managers.

Bee bowl at Pt 180
Bee bowls, also known as pan traps or bowl traps, are a common sampling method

The Bee Inventory Plot provides a basic protocol for combining bowl trapping (passive sampling) and netting (active sampling) that has become widely adapted for quantifying bee community composition around the world. Bowl traps, also known as pan traps or bee bowls, are appealing as a sampling method because they don’t require a skilled collector (although accurate bee identification is conducted by a taxonomist), deployment is simple and quick, and traps can be left unattended for many hours. Like all sampling methods, bee bowls are known to have some biases, though many of these are not well understood or well documented.

A benefit of working in restoration and conservation is that we get to spend a lot of time in the field to observe natural history and phenology.

Early in the spring before many flowers were blooming, we would find that the bowl traps would be filled with bees . However, with few flowers blooming, we would net few or no bees during these periods. From late spring into early summer, the grasslands of western Montana bloom with a diversity of wildflowers and bees were found actively foraging during these periods. We often spent the entire net sampling period at each plot catching bee after bee. Surprisingly, the bowl traps did not seem to have high capture rates during this peak bloom.

Field crew member Mackenzie Sebastian hunts for bees at one of the plots in a late July survey

Based on these natural history observations, we hypothesized that bee capture rates using bee bowls were inversely related to flowering richness. If true, this would suggest that bee bowls would not accurately reflect true bee abundance or community composition, casting doubt on the effectiveness of this widely used method for evaluating conservation and restoration outcomes.

In our latest research, we used our spatially and temporally robust bee monitoring data, combined with flowering phenology data, to test this hypothesis.

Bee taxonomist and co-author Skyler Burrows identifies MPG bees at the USDA-ARS Pollinating Insects Research Unit in Logan, Utah, USA © Skyler Burrows

With bee taxonomist and co-author Skyler Burrows, we identified over 230 bee species in 38 genera, including 34 new occurrence records for Montana. We found that, as hypothesized, bee abundance in bowl traps negatively correlated with flowering richness, whereas bee abundance in net samples was positively correlated with floral richness. Additionally, when examining community composition based on bee abundance, bowl and net surveys had very different communities and that bowl-trapped bee communities were highly sensitive to changes in floral richness.

This is concerning in that it suggests that bowl traps do not accurately represent bee community composition when using abundance-based metrics. However, we also examined community composition using presence-absence metrics and found that composition was much more similar between bowl-trapped and netted bee communities, and both changed similarly with floral richness.

Western Montana grasslands bloom with a colourful diversity of flowers from mid-May through June

The lesson for researchers and land managers is clear: when quantifying bee community composition from bowl traps, using presence-absence based metrics should be preferred to abundance-based metrics.

These natural history observations were key to making the discovery that when many plant species are in bloom the bee bowls are less attractive to bees. Our research highlights the importance of how careful observations and a solid understanding of natural history can translate into actionable practices used by conservation and restoration practitioners.

Read the full research: “Relative bee abundance varies by collection method and flowering richness: Implications for understanding patterns in bee community data” in Issue 2:2 of Ecological Solutions and Evidence.

One thought on “Bee abundance estimates vary by collection method and flowering richness

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s