Identifying Ground Crab Spiders (genus Xysticus)

I’ve slowly been learning to identify spiders native to my part of the continent (Great Lakes region of N. America). Recently I decided to tackle the Ground Crab Spiders, genus Xysticus, since they have a relatively high ratio of “Needs ID” to “Research Grade” observations (about 20:1 in N. America). I found some great resources, like “The Crab Spiders of Canada and Alaska” and BugGuide, and set about learning to ID some of the 20+ species that occur near me. (There are 67 species in the US and Canada, according to BugGuide.)

So I started identifying a few really obvious ones (to me, at least) on iNaturalist. I’ve gotten a lot of immediate blind agreements (ugh), but also a really interesting conversation with someone who doesn’t believe anyone can identify any Xysticus to species without dissection, and has therefore knocked the ID back to genus. https://www.inaturalist.org/observations/747886

So I’m wondering if anyone reading this has or would be willing to track down some photos of live Xysticus spiders already paired with photos of their palps (male) or epigyne (female), in other words photos that have been identified to species with as much certainty as possible, so I can attempt to ID the full-body photos to species without seeing the palp or epigyne photos. Collection locations would have to be in N. America, preferably in the north-east, but I’m willing to attempt southern and western ones as well.

I’m reasonably confident I can accurately ID X. punctatus, and I’d like to measure my success rates for other species as well. If I’m overconfident, I want to destroy my overconfidence. And if I’m not overconfident, then I’d like to create some independent evidence of that. Either way, I’ll post the results of the test here.

5 Likes

Great discussion in that thread. A good point you made in there:

Genetalia are usually used for ID because preservation in alcohol washes out color from specimens: structural features are the only thing available to taxonomists working with museum specimens. Good color photographs of live specimens are a relatively recent thing, and it’s pretty easy to break new ground in spider identification by looking at photos of live specimens paired with photos of their genetalia

and cheins makes a good counterpoint

Until more work like this is done, we’re stuck using the old taxonomic treatments which focus on genitalia (sometimes ignoring other observable characteristics)

Now I have no experience with spiders, but from my work on other taxa I do agree that in this era of high quality digital photography there is definitely room to break new ground on potentially diagnostic color pattern characters which are often ignored in the literature. But my approach is this: even if a specimen looks exactly like how species X should look, do we know for certain that there are not other species which also happen to look superficially exactly the same, especially if the patterns of those other species are not adequately described? The best way to know one species, really, is to know them all.

3 Likes

Personally I find Xysticus all to look rather similar in coloration and many species tend to have variable patterning so I tend to stick to genus level ID myself

Someone, in an observation on Salticidae, I forgot which, made the suggestion to put the spider in a glass container or a petri dish and press it flat with a soft material or sponge. This will not be good photos because of possible light reflection. But it will allow keeping the spider alive to be released. I am not sure if these photos will be sufficient in many families, but it is worth trying.

I think it’s worth pasting Joe Lapp’s comment on the original iNat observation here as part of this discussion, since Joe is one of the co-authors of the crab spider chapter in the Spiders of North America Identification Manual and a national expert on this family - he comments:

" I am also tempted to call a spider to species when I see a perfect photo match with a specimen identified by genitalia, but there is a case where this doesn’t work: when other species have identical or nearly identical color patterns. This happens a lot with Mecaphesa, which is the group I primarily study. I no longer believe it’s even possible to distinguish some Mecaphesa by color pattern.

In any case, before we can identify any Xysticus by color pattern, we need to know what all species of Xysticus in the region look like. Suppose we were to have 20 species of Xysticus in the region and knew what 19 species looked like. We might find a specimen that exactly matches one of the 19 species by color pattern, but we can’t know that the specimen doesn’t also match the 20th species.

It’s also extremely difficult to learn what the color patterns of any one species are. One species of Mecaphesa can have many different color pattens, and sometimes those color patterns are indistinguishable from those sometimes available to another species.

To know what the color patterns of a species are, we have to have a huge sample of the species. That sample includes photographs of the specimen live before dissection to definitively ID. Only then can we have a clue which color patterns are available to which species. I haven’t studied Xysticus enough to know if this happens for them, but it’s possible that some of the color patterns are shared across species. There aren’t many of us doing this sort of work, but @kschnei, @cheins1, and I are a few of those people. It’s going to be a while before we can do this confidently. We largely limited to calling species by photo in areas where only few are found".

3 Likes

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.