Skip to main content
ARS Home » Pacific West Area » Logan, Utah » Pollinating Insect-Biology, Management, Systematics Research » Docs » Cucurbita » Bee Sampling Protocol

Bee Sampling Protocol
headline bar


The objective here is to identify the more common visitors to flowers of cultivated Cucurbita, and to quantify their relative abundances at Cucurbita flowers without bias. In the Western Hemisphere, we are emphasizing species of specialist squash bees in the genera Peponapis and Xenoglossa, both of which forage early in the day. Data is recorded in Excel spreadsheets as directed, and later compiled by cut-and-paste onto a master spreadsheet that is shared back with participants. For this reason, data must be gathered and reported in a standardized manner.

To begin, each investigator must locate people who are growing Cucurbita locally. These might be full or part-time farmers, market gardeners, or even home gardeners. Before performing a formal bee census, I recommend that one of the more convenient sites be visited for a full morning. This will allow you to collect voucher specimens of the various bee taxa that are observed visiting squash flowers, to learn how to recognize them on the wing, as well as to gain insight into the hours of foraging activity, so that a census can be properly scheduled. Pinned, labelled voucher specimens should be identified to the genus level if possible. You should learn to recognize at least the more common taxa on the wing for later census purposes.

The basic census technique is a walking scan census, very much like the ones used by ornithologists to count birds. The idea is to know, at a given point in time, the densities of the various bee species at a given count of Cucurbita flowers. For the first bee census in your region, it would be useful to census bees at several time periods during the morning so as to fairly represent species that forage at different times of the morning. The census should represent the bee fauna visiting Cucurbita across the entire patch or field. In smaller patches, you might census all plants, but in larger fields, you should be sure to walk to different parts of the field to census (or walk across the field, perhaps along two diagonals). Choose the plants that you will census ahead of you, before you can see if there are bees or not. Count all of the flowers and the bees of each species or genus that you see present in those flowers. Do not wait for bees to arrive or depart; it is meant to be the equivalent of a photographic snapshot of what is present at one moment. Continue walking from plant to plant, keeping a running tally of open flowers present and counts of individuals of different bee species seen at the those flowers. You should continue checking plants and counting bees until you are satisfied that all of the more common bee species are represented in your census, and that their relative abundances are accurately portrayed. As a general rule, we continue until at least 30 bees or 200 plants are included in the census, whichever comes first. There is no problem adding more bees or flowers if you like, as the work is generally quick. Obviously, in small plots with few bees, you will have less, but that is acceptable; do not include flowers twice in the same census.

Because you will also be estimating the average numbers of open flowers per plant, the size of the patch or field, and the number of plants present, it will be possible to calculate the number of bees per plant, bees per unit area, and the total numbers of the different kinds of bees present at squash flowers for the site. These numbers can be compared for sites in your locale that you have surveyed. And, since other collaborators have censused bees at squashes in the very same way as you, you can compare your data with that of other cooperating investigators. How many bees are enough for pollination? Some investigators of us are interested in relating the densities (or visitation frequencies) of the different bee species to satisfaction of the pollination needs of the different species of cultivated squash. Knowing this, it should then be possible to evaluate the degree of pollination service (or shortfall) for every censused site.