Key to the Bee Families of the World
Laurence Packer and Claudia Ratti
Packer Collection at York University
4700 Keele St., Toronto, ON., M3J 1P3 CANADA
Welcome. This website is intended to allow the user to identify a bee from anywhere in the world to family level. It is comprised of three independent keys: 1) a key to determining the sex of a bee 2) a key to family level for both male and female bees 3) a key to family level for female bees only.
Use of the Key for Non-English Speakers
(If characters do not display properly, change character encoding setting to "Unicode (UTF-8)")
It should be possible to obtain a correct identification of a bee simply by looking at the images.
At the top of each page, there is a "couplet" - two different statements. Every specimen will be described by one or the other of these statements. For any one couplet, a specimen that matches the images on the left leads you to the next couplet or the family noted at the end of the first half of the couplet; those on the right lead to the next couplet or family noted at the end of the second half of the couplet (Figure 1 ).
Is it a Bee or Not?
For a beginner, this is surprisingly difficult to decide.
Bees can be differentiated from almost all other Hymenoptera on the basis of their possessing branched hairs somewhere on their bodies (Figure 2 ). These are most often seen around the pronotal lobe (Figure 3 ) or on the propodeum (Figure 4 ). Sometimes the hairs are merely woolly (Figure 5 ) rather than being truly branched and many bees have unbranched hairs somewhere on their bodies (Figure 6 ). Some bees have mostly unbranched hairs, or they have hairs with such tiny branches that they appear merely crinkled (Figure 7 ) although the branches are often clear at high magnification (Figure 8 ). This means that, when in doubt, it is important to look particularly carefully to be certain whether an insect is a bee or not.
The only other Hymenoptera with branched hairs are some groups of velvet ants - family Mutillidae (Figure 9 ). Bee females are easily distinguished from those of Mutiliidae because the velvet ant female is wingless (Figure 10 ). Males (Figure 11 ) can be differentiated by numerous characters, perhaps the easiest of which to be certain of is the shape of the pronotum. In bees (as well as the apoid wasps - families Ampulicidae, Sphecidae and Crabronidae), this narrows ventrally and forms a complete ring around the prothoracic foramen (Figure 12 , Figure 13 ). In Mutillidae (as well as other non-apoid wasps) the pronotum ends as a blunt angle on the side of the thorax and does not extend towards the midline ventrally (Figure 14 , Figure 15 , Figure 16 , Figure 17 ).
Sexing a Bee
Once you have determined that your specimen is indeed a bee you next need to determine the sex of your bee.
For most bees, the presence of pollen collecting structures (scopae on the hind leg or on the ventral surface of the metasoma) is sufficient to identify an individual as a female, although sometimes males have brushes of long hairs on legs or metasoma which likely serve some sexual function.
With a few exceptions, male bees have seven metasomal terga and 13 antennal segments whereas females have 6 metasomal terga and 12 antennal segments. The number of metasomal terga is sometimes difficult to establish if the more apical segments are telescoped within the more basal ones. Consequently, the Key to Determining the Sex of a Bee serves to identify the sex of a specimen whether its metasoma is telescoped or not. Almost all of the really problematic specimens will be females of bees that do not possess a scopa.
At end of the sexing key, when you arrive at the sex of your bee, select the sex and you will be directed back to the key selection page. There you can choose which key you would like to use to identify your bee (Key to Males and Females or Key to Females Only).
Bee Family Identifications
Bees are notoriously difficult to identify to the family level for beginners. There are three main reasons for this.
First, family level identification of bees is most easily done using characteristics of the mouthparts which are usually hidden in specimens.
Second, there are numerous exceptions even to the mouthpart characteristics.
Third, male bees, cleptoparasitic bees (cuckoo bees and social parasites) and some bees that carry pollen internally do not have or have lost those structures associated with pollen collection that make keys based on non-tongue characters easier to use.
Once one becomes accustomed to bee diversity, identification of the genera found in a particular region can become quite manageable. But for beginners, a key to families is a necessary evil that serves as an entry point to keys at lower taxonomic levels.
Our keys are modifications of those provided by Michener (2007). The modifications are either aimed at making it as easy as possible for a beginner to identify a bee specimen to the correct family or permit the identification of the, usually rare, exceptions to the main characteristics of bee families.
Notes on Some of the Tricky Characteristics
Because mouthpart characteristics are required for identification in many cases, it is recommended that the beginner use ethyl acetate to kill specimens in the field. This is because this killing agent causes the bees to extrude the tongue, making it possible to see the important characteristics without relaxation and manipulation of the specimen. Irrespective of the killing agent used, we recommend that the beginner extrude the tongue in freshly killed specimens before drying (see paragraph below on extruding the tongue).
If you have to use the mouthpart components of the keys to identify a specimen for which the mouthparts are hidden, then it will be necessary to relax the insect. This is best done at room temperature in an almost airtight container with a small amount of water and with absorbent tissue placed between the lid and the rest of the container so that condensation does not build up and drop onto the bee inside (Figure 20 , Figure 21 ). 24 hours of relaxation should be enough for a small bee or a large one that has been collected in the past few years years (assuming that it was not killed in absolute ethanol or other dehydrating agents). Large bees, old bees and bees that have been kept in ethanol will require perhaps 48 hours to become sufficiently relaxed for the tongue to be extruded. If you leave the specimen in a damp environment for much longer than 48 hours (or if the room in which you are working is of tropical heat temperature) the specimen will likely go mouldy. This can be avoided by placing a few crystals of chloro-cresol in the relaxing jar.
Once a specimen has been relaxed, proceed as follows. It is best to hold the head of the insect between finger and thumb of one hand and manipulate the mouthparts using fine forceps held in the other. It is usually necessary to open the mandibles first. This can be done by inserting a pin between the crossed mandibles such that they become somewhat separated. Then the fine forceps can be used to open the mandibles more fully. Next, it is useful to rotate the labrum forwards and upwards. This can be done using the end of a robust pin (size #3 perhaps). [This is also useful for differentiating Megachilidae from Apidae in the keys as this sometimes relies on the relative length to breadth of the labrum and/or whether its base articulates with the clypeus for its entire basal width or for a smaller proportion.] The remaining mouthparts should then be visible and their extrusion made possible through careful use of forceps and pin (Figure 22 , Figure 23 , Figure 24 , Figure 25 ).
Making the labial palps, stipes, lorum and proboscidial tube visible should be relatively straightforward if the aforementioned procedure is followed. Although often more difficult to find, the lacinia should also be made visible using the above method. The relative size and orientation of the labial palpomeres and/or the presence of the stipital comb and cavity are relatively easy to see, often even in unmanipulated specimens.
Sometimes it is necessary to assess the shape of the glossa. This will be straightforward if the mouthparts are extruded (at death or after relaxation) but may not be easy if the glossa itself is folded or shriveled. Careful massaging of the apex of the tongue, after the mouthparts have been pulled forward, will often enable the shape of the apex of the glossa to be ascertained.
Note that when facial foveae are referred to in the keys, they have to be distinctly depressed with a clear "step" between the fovea and surrounding area at least on one side. Many bees that should key out as not having a facial fovea have an area of different colour (Figure 26 ), or different surface sculpture where the fovea would be, but unless it is clearly depressed, or with a "step" it is considered absent.
The episternal groove is usually easy to see, although in densely hairy bees the hairs may have to be removed (with the end of a pin) or pushed aside to see it. The episternal groove may be difficult to see in species that have very strongly sculptured surfaces. For example, the episternal groove may appear merely as a row of unusually large or unusually deep pits in a surface that is covered in slightly smaller pits (Figure 27 ).
Subantennal sutures may also require the removal of the bee's facial hairs. Rotating the bee under the light of the microscope will aid in detection of these sometimes fine sutures which might be visible at certain angles but not others.
A corbicula is a bare, or almost bare, space surrounded by scopal hairs (Figure 28 , Figure 29 , Figure 30 ) forming what can be referred to as a pollen basket. Either tibial or femoral corbiculae are referred to in the key. The hairs surrounding the bare space are sometimes surprisingly sparse on the tibial corbicula.
The location of the apex of the stigma in relation to the position of other veins is important at one point in the key. The apex of the stigma is at the point where the vein on the anterior margin of the wing in the marginal cell starts to thicken. The stigma perpendicular is at right angles to the anterior margin of the forewing at the apex of the stigma.
Additional Notes on the Use of the Keys
In the lateral views of specimens, the bees are all oriented with their heads to the left.
Note that couplet 4 in the Key to Males and Females is reached from both couplets one and two. This is intentional.
Eventually, identifications will be linked to images of bees in that family (below family level in instances where subgroups within a family are keyed out). For some groups that are exceptions to normal structures for a family, the individual you have should look very similar to the whole animal images provided when you get to the identification. But bees within a family are extremely diverse (this is one of the reasons why family level identifications are so difficult). Nonetheless, we are gradually building up an image bank of almost all bee genera on the linked website, so you should be able to find something that looks similar to your specimen.
These keys were made possible primarily by a collaborative agreement between the Food and Agriculture Organization of the United Nations and the Packer Lab at York University.
Laurence Packer's research and collections are primarily supported by the Natural Sciences and Engineering Research Council of Canada.
The images were taken using a Visionary Digital BK Plus imaging system using a Canon EOS 40D digital SLR camera and processed with Adobe Photoshop. This system was purchased with funds from the Canadian Fund for Innovation and the Ontario Research Fund through Canadensys. Thanks to Roy Larimer for assistance with imaging.
We thank Dave Cheung (University of Guelph, Guelph) for invaluable technical assistance.
Some of the specimens imaged in this key were obtained as loans. We thank Terry Griswold (USDA bee lab., Logan UT), Jerry Rozen and John Ascher (AMNH, New York, NY) and Jason Gibbs, Lincoln Best and Cory Sheffield (PCYU, Toronto). We are grateful to Denis Michez (Mons, Belgium) for allowing us to use images of Eremephanta convolvuli.
The keys are based upon Michener (2007) The Bees of the World, John Hopkins University Press. We are grateful to Dr. Michener for permission to adapt his work.