Andrena (females) subgenera of the Pacific Northwest

 

There are a couple of routes to finding an ID for an Andrena collected in Oregon or Washington. You could begin with the excellent DiscoverLife guide, an interactive key that encompasses the 150 (or so) species in the region. In my experience, this will generally yield a short list that may include the actual answer (though not always!). But the final step – nailing the species—is usually just out of reach.

 

There is a very helpful shortcut in DiscoverLife: If you can name the Andrena subgenus (for experts only!), you will usually winnow your hits to a manageable list. If you know you have Diandrena, there will be only 5 species to consider.

 

But knowing subgenus has another, critical benefit: it tells you where to go in the primary literature for a species-level key. This is a necessary step for many (especially western) taxa that are incompletely scored in DiscoverLife. One additional, less obvious advantage: if you are stymied by a difficult diagnosis, you can at least assign morphospecies within the subgenus. In the collections I work with, we have ~40 Andrena species. To name them Andrena morphopsecies X is far less useful than Andrena (Micrandrena) cf. chlorogaster.  

How to use the key

The menu should be mostly self-explanatory. Rather than create a detailed description of the features, I encourage you to just play around with the options. A few general suggestions:

 

·      Educate yourself about the characters. If you are not intimately familiar with Andrena, you might start by reviewing the Guide to the DiscoverLife Andrena key. Look at a set of specimens and decide what traits you think you can reliably identify. Super easy: integument, dark or metallic; submarginal cells 2 or 3; hair bands complete, or not. Many other characters are daunting at first, but fairly straightforward if you have the photo references. Then there are a few characters that you would consider last. E.g., to say whether a pronotal ridge (often hidden by the head) is weak or absent is valuable, but often undetectable feature.

·      Educate yourself about the key. Even in advance of evaluating a specimen, try selecting choices for a number of traits. You will get a feel for which traits might be the best discriminators. E.g., “T2 hair bands edge-edge” will match half of the options. Add “posterior margin of pronotum angled” and you cut the choices by another half.
You will also find some characters that select for just one or two matches, E.g., “short e vein,” “bidentate labral process,” “completely scultured propodeal triangle.” Each of these is easily evaluated, so might be among those things you look for first.
Recognize that the key generates probabilities – in the single-column key, the cumulative score for each subgenus is on the left. Consider high scores as evidence that you have a match. You will also want to look at the scores for individual characters, which may argue against a match. These will be cases where you’ve interpreted the character wrong, you have an atypical specimen, or where I referenced a source that was wrong. All of these will occur!

 

o   After using the key to generate a short list of likely matches, select the side-by-side comparison for these. This provides every point of similarity or difference for the selected subgenera.

o   Select full taxon details to see a more comprehensive description, and images. Use the cursor on images to zoom in—you can often see the plumosity of hairs, or the lengths of antennal flagella.he key is the default view on the www page. At the bottom of the text description, you will see the list of species recorded for Oregon or Washington. These names are links – they will take you to the corresponding DiscoverLife page.

o    

·      Test the results in this key against DiscoverLife, or other source. Examine the specimen images in the key, or look for others. I have used only a subset for those from the Laurence Packer Lab, and from  the USGS Bee Inventory and Monitoring Lab

 

Method

I started with a character matrix in Identikit software, scoring each subgenus with any of the traits for the PNW species in that subgenus. I then adjusted these based on the subgenus treatments in primary literature.

 

Some further considerations:

 

-       I am starting with females, simply because I have better resources for describing them. Of course males should be added—eventually.

-       I included most of the subgenera likely to occur in Oregon or Washington, according to DiscoverLife. My key may of course fail for a specimen collected elsewhere.

-       I assigned weights to each character. E.g., I weighted “length of r-vein” highly, because it is easy to evaluate, and it is a good diagnostic for the few taxa with short veins. I gave a low weight to “clypeus shape” because I consider this difficult to evaluate. I expect to revisit the weighting scheme – tell me where I should tweak this.

-       I described characters and states mostly using language and images borrowed from the DiscoverLife key.

-       The PNW key is the first iteration. There is no reason why the data architecture can’t be extended to other regional versions. I will share this with anyone.

 

Acknowledgements:

-       The Institute for Applied Ecology (Corvallis OR) gave me the entrée into the field of pollination biology. Since 2019, we have been characterizing pollinator communities in the Willamette Valley.

-       Quamash EcoResearch (PI Susan Waters) supplied many specimens, and her in-house Andrena enthusiasts Marisa Fisher and Cody Blackketter made the case for the creation of this key.

-       Sam Droege and Clare Maffei, USGS, my connections to the DiscoverLife resources, have been generous with their time, and specimens and photos.

-       The Oregon State Arthropod Collection (Chris Marshall) allowed me to photograph specimens in the collection. Many of these have determinations by LaBerge, obviously the gold standard for Andrena.

-       I’ve included links to many images created by the USGS Bee Inventory and Monitoring Lab, and  the Packer Collection at York University. These are critical to evaluating a diagnosis, and beautiful besides.

 

--David Cappaert <> cappaert@comcast.net