Mecaphesa asperata (Northern Crab Spider) – quick notes on geographic range and identification
For reasons unknown to me, the crab spider species Mecaphesa asperata (formerly known as Misumenops asperatus, common name Northern Crab Spider) has become the AI/automated suggestion of choice for crab spider images in the western US, the area where I normally review spider observations. Although this species is known to be widespread in North America, to my knowledge it has not been accurately documented from California and it's unclear to me if it is known from Arizona, Oregon or Washington state. In addition, with only one exception I am aware of, crab spiders in the genus Mecaphesa are difficult or impossible to ID to species without examining the genitalia of an adult spider, usually a collected specimen.
This is all to say that I encourage everyone NOT to accept the automated suggestion of M. asperata for crab spiders photographed in California and neighboring states and, if folks feel comfortable, to disagree with IDs of this spider in that area. I've included more information below.
From Dondale and Redner 1978 - Crab Spiders of Canada and Alaska:
Range: “New Mexico to Florida, northward to British Columbia, Quebec, and Massachusetts".
ID: Discuss genitalia, no comments on coloration being diagnostic. Collected from foliage and blossoms.
From Gertsch 1939:
Range: “United States and Canada. The species become increasingly rarer toward the south and- at the present time there seem to be no authentic records from Mexico or the West Indies”. Gertsch 1939 does indicate a single record of a male collected from Claremont, California – BUT see below.
From Schick 1965:
“Asperatus group: At least three species of this group occur in the United States, M. asperatus (Hentz), M. verityi, new species, and M. devius Gertsch, the last two from California. Asperatus is cited from California from a single record in the 1939 revision of Gertsch (Claremont, Los Angeles County), but this is a doubtful record". (note: Schick produced a detailed monograph on the crab spiders of California in 1965 and probably understood the ID and ranges of these spiders in CA better than anyone before or since that time).
M. asperata/asperatus not listed in the Johnson/Lew checklist found here:
https://www.ocf.berkeley.edu/~stevelew/soc.html
M. asperata/asperatus is not listed in Rod Crawford’s 1988 Annotated Checklist of the Spiders of Washington
תגובות
@arachnologus : Dondale and Redner say M. asperata occurs in British Columbia. Have you found this species in WA state since the publication of your 1988 checklist?
I have one single specimen of asperata from Washington, this one being from the extreme northern Cascades a few miles south of the BC border. It would have to be considered extremely rare in the state, and definitely should not be the AI's default species in this group. The common Mecaphesa here are sierrensis, lepidus, celer. I have about 3 records of importunus. iNaturalist photos of misumenoid spiders are pretty easy to ID to genus if the resolution is good enough to show macrosetae, since the body is pretty well covered with these in local Mecaphesa whereas they are lacking on the abdomen and the rear 2/3 of the carapace in Misumena. The only thomisid genera I know of in WA are Misumena, Mecaphesa, Xysticus, Coriarachne, Bassaniana, Ozyptila, Tmarus. There are old records of Misumenoides but I have never found one.
Thanks Rod! Much appreciate that input!
The "Computer Vision" is problematic when it comes to small arthropods, to say the least. Mecaphesa asperata is also the default suggestion for several common Xysticus species in Texas as well as any Mecaphesa and several other Thomisids. Misumena vatia is also a common suggestion even though it is rare in Texas. I think there are a few widespread species like Mecaphesa asperata and Misumessus oblongus which are not easily confused with other species in certain parts of their range, making them a safe ID in some areas. So there is an abundance of "Research Grade" photos of these common species, which leads to excessive weighting by the Computer Vision.
To make matters worse, it seems the various apps (iNaturalist Android/iOS apps and Seek), which are very popular with observers, will often ignore geographical distribution entirely. I now regularly see Texas spider observations identified as species endemic to New Zealand/Australia, South Africa, Western Europe, etc. and they are invariably from one of the apps. As the apps gain popularity, this seems to be happening more.
This is not just a problem with spiders, FWIW - many other identifiers of small cryptic arthropods have the same complaints. It has been requested several times that iNat include some way to confer that certain taxa are unlikely to be identified to species from photos, and to stop the CV from suggesting species. Something like a taxon-level curator toggle to tell the CV to leave it at genus/family, or maybe at least present a friendly warning to the user. But iNat staff don't seem interested because the CV admittedly does work very well for plants, mammals, birds, butterflies, etc. which is the majority of iNat's activity.
I don't think there are any real solutions to any of this except to try and recruit more identifiers, which will spread the frustration out across more people and make identifying more tolerable :)
הוספת תגובה