Bee sampling has no effect on bee abundance in montane meadows

Abstract

One of the most talked about potential consequences of climate change is that of phenological mismatches between interacting species, such as flowering plants and their pollinators. While there is ample long term data for plant phenology, there is little data on pollinator phenology. It is becoming more common to monitor pollinator (bee) populations over long periods of time to look for phenological mismatches. The standard method of bee population monitoring uses pan traps, which are small plastic bowls containing water with soap. The bowls are painted bright colors to mimic flowers and attract bees, killing them when they fall in the water. It is unknown if sampling using pan traps affects bee abundance, especially when implemented repeatedly, over long time periods. Thus, I asked two questions: Is there a difference in bee abundance between sites that have been repeatedly sampled using pan traps and sites that have never been sampled? And, how does bee abundance change over two years of repeated sampling? To address the first question I used ten control sites: five dry meadow sites and five wet meadow sites. Control sites were sampled once, and bee abundance was compared between these control sites and six sites (three of each habitat type) that were repeatedly sampled over the 2009 and 2010 flowering season using the same techniques. In total we caught 2060 bees at the repeated sites and 333 bees at the control sites. I found so significant difference in bee abundance at the control versus the repeatedly sampled sites. I also found no significant difference in bee abundance in 2009 versus 2010. I also compared floral abundance at the sites and weather data to ensure that factors other than sampling were not confounding my results. I found no significant difference in flower abundance or temperature between sites. Taken together, these data suggest that sampling using pan traps has no effect on bee abundance.