An Analysis of Capture Rates of Native Bees by Bowl and By Net
A Summary of Relative Capture Rates of Bees in Bowls and Those Caught While Netting Using Selected Data from the USGS BIML Laboratory
Internet Age Executive Summary: Capture rates of bees can vary quite strongly by genus and by technique. The results are dispersed and complicated enough that you need to scan the tables below to get a feel for the details. An Excel file of the raw data and complete summary tables is available from Sam Droege (sdroege@...).
Written by Sam Droege 7/21/11
An Introduction to the Problem at Hand
Bees have be captured many ways. Two common ones are the use of pan/bowl traps which attract bees to a soapy container colored in some shade of blue, white, or yellow, and the traditional hunting down and capturing of bees using insect nets. The USGS BIML lab has used both techniques extensively. As part of the creation of a genus level workbook on bees we decided to explore these data to look for patterns in the captures of bees by genus to see if it would illuminate anything of interest.
Characteristics of the Dataset
These data span from 2000 to the fall of 2010 and only captures from bowl traps or netting were included.
In any particular collecting event, be it by bowl or net with few exceptions all specimens captured were included in the database.
The figure below shows the extent of collecting localities of the BIML lab, though it includes some additional 2011 collection sites.
These data were collected by a variety of people and researchers with major contributions coming from individuals Virginia/West Virginia/DC/Delaware/Illinois /Massachusetts (Netted Bombus only).
Much of the netting data, however, is from Sam Droege’s efforts. His pattern of collecting bears some scrutiny. In general, he would collect any bee he saw, but would generally only take one each of Xylocopa, Apis, and each Bombus species at any individual site.
All the seasons are covered and reasonably represented with the data heavily weighted toward the Mid-Atlantic area, though anything East of the Mississippi was included with perhaps a few stray odds and ends from the West, but only an insignificant amount.
A summary of all the follow data are available in an excel spreadsheet that you can obtain from Sam Droege (sdroege@...).
A total of 87,482 bees were included in the analysis… of those 14,557 were captured using nets and 72,925 were captured using bowls.
Table 1. Top 10 Genera Captured While Netting
Figure 2. Top 10 Genera Captured by Bowls
Greatest Shifts in Rank between Netting and Bowls
One way to look at how capture technique impacts the numbers of captures of bees is to calculate the difference between each genus’ rank order of capture between the two groups. If there is very little difference in the rank order then both bowls and nets capture similar proportions of bees of that genus in comparison to other genera. This is not the case in this dataset and some genera make huge changes in rank order between capture techniques.
For example the following 10 genera made the largest changes in rank order in favor of being caught in nets:
The Lithurguare a bit of an artifact from a single collecting trip made to specifically net collect Lithurgus chrysurus where it occurs in its currently limited (introduced) range in New Jersey. However, note that five of the genera with the largest shifts in favor of netting (and avoiding bowls) are from the Family Megachilidae. The Bombus numbers are likely affected by a large study in Massachusetts that only net captured Bombus.
While the next group of 10 genera showed the greatest shift towards bowl captures:
Note that six of these genera are in the Family Apidae.
Adjusting Figures so that Both Netting and Bowl Captures Show the Same Total Number of Bees Captured
Because about 5 times as many bees were captured in bowls as compared to nets one can multiple the captures in nets by 5 and then compare netting captures to bowl captures using standardized values. Below are tables similar to those that show change in rank but in this case these tables now show the percentage of standardized captures.
Table 3. Top 10 genera with highest capture rates using netting (percentages are percent of bees captured using nets).
Table 4. Top 10 Genera with highest capture rates using bowls (percentages are percent of bees captured using nets).
Some thoughts and observations
1. All 38 genera that had more than 25 total captures had individuals that were captured both in bowls and in nets.
2. The values presented here could all be radically altered by changing how and when netting takes place and in changing the number of bowls set out and where they are placed and, as such, this analysis represents a landscape approach to capturing bees and should not be seen as unalterable Gospel .
3. Most of the bees netted in this database were netted by Sam Droege who fancies himself to be pretty good at wrangling bees with the net. It is his perhaps slightly egotistical observation that most people (particularly beginners) are quite poor at catching and handling bees in nets and thus a similar test of net versus bowl captures using the unskilled masses would skew the results strongly towards the use of bowls.
4. It would be useful to also standardize captures by time rather than simply by the numbers caught. But this presents problems in that bowls can be deployed anytime while netting must be done during the active foraging hours of bees. So, in most circumstances, it makes sense to deploy bowls during time periods that bees cannot be captured with going netting during the sunny mid-day hours… if maximizing captures or a complete inventory Is one of your objectives.
5. Circumstances of a study will also obviously dictate which technique is deployed and in many cases the limitations of circumstances forces you to choose one and only one approach.
6. Malaise, glycol, vacuums, etc. are all completely legitimate and useful ways of catching bees but were not included in this analysis simply because the BIML group doesn’t employ them that often, though that will soon change with our emphasis now on long-term glycol traps.
Sam Droege sdroege@...
w 301-497-5840 h 301-390-7759 fax 301-497-5624
USGS Patuxent Wildlife Research Center
BARC-EAST, BLDG 308, RM 124 10300 Balt. Ave., Beltsville, MD 20705
2 is not equal to 3 -- not even for large values of 2