Elevated pan traps to monitor bees in flowering crop canopies

Bee behavior is strongly directed by visual cues in thevisible and ultraviolet wavelength ranges, with response tocolor and pattern controlling location of owers (Chittka& Menzel, 1992; Kevan et al., 1996; Gumbert, 2000). Theirattraction to color can be exploited for monitoring beepopulations through the use of passive colored traps lledwith soap and water, which provide an inexpensive andeasily replicated method to capture bees (Aizen F Leong & Thorp, 1999; LeBuhn et al.,2002; Toler et al., 2005; Baum et al., 2006). The use of pantraps to monitor bee communities is a relatively recentdevelopment, but it is increasingly used as a sampling toolby researchers in urban, pastoral, and natural landscapes(Leong & Thorp, 1999; McIntyre & Hostetler, 2001;Williams et al., 2001; Russell et al., 2005; Toler et al., 2005;Brosi et al., 2007).Pan trapping has important advantages compared tomore traditional bee collection methods. It can eliminatecollector bias, which is particularly important whencomparing data across different studies or when usingmultiple collectors in the same study, and it can be easilyreplicated for consistent sampling intensity by collectorswith minimal training at multiple sites. Specimens collectedusing this method can be accurately identied to speciesunder microscope by trained personnel. As with anysampling method, there are some drawbacks. Trappedinsects are removed from the population. Pan traps may bebiased toward attracting some bee taxa over others (Roul-ston et al., 2007) and may underestimate the abundance oflarge bees. Unless pollen can be obtained from specimensin the trap, one cannot determine which owers werevisited. However, in comb ination with other methods pantrapping can be very effective for estimating bee abundance(Williams et al., 2001).The majority of native-bee monitoring studies that usepan traps have been conducted in habitats with littleoverhead plant canopy, where traps were placed directly onthe ground among open, or low-growing vegetation. Thisdeployment strategy is appropriate for many naturalhabitats where bees are likely to forage and so is generallyrecommended (e.g., LeBuhn et al., 2002). However, if thistechnique is to be applied successfully in agriculturalhabitats where there is increasing interest in bee conservation,it is important to consider that many bee-pollinated cropshave dense, vertical vegetation and oral displays. The cropstudied here, highbush blueberry [