Check out these crocata found on the Texas/Oklahoma border. The red mark almost but doesn't quite reach the spinnerets.

Examining Reiskind 1969 for additional characters, I see it going into detail for descripta but having little to say about coloring one way or another for crocoata. That probably reflects the author's short supply of specimens and may not reflect the species.

I'd be curious to know where you got your research, to help me with my IDs.

Joe, those images are indeed crocata. I am working on the genus in Kansas and have discovered that although descripta has a variable-sized red mark, it always is wider or parallel toward the rear, never tapering as in crocata. The red spot in descripta has a faint darker stripe in the anterior end. It looks like a smudge, while crocata never has such a darker mark inside the red area.

Great to hear from you Hank! I have also been working on Castianeira and have some really interesting results still to share. So according to your findings, the following is C. descripta because the red mark does not taper towards the spinnerets:



That's a great field mark. I'll try to confirm that down here. I believe I currently have a C. crocata/descripta in a jar I'm raising. And it started out looking more like a C. longipalpa!

a while back. I think we used Reiskind 1969 heavily. He has a lengthy description on page 200, and we relied heavily (which is the part we are unsure of) on the drawings provided. C. crocoata drawing found on page 293. This is where we got the idea of separating them by the posterior ending point of the red dorsal mark. It's certainly not 100% and we sorely need some voucher specimens for the pages.

Okay, I see. I think using these revisions is like using an illustrated field guide. I came to spiders by way of butterflies, so I know how things go with butterflies. Usually newbies will page through a field guide looking for the image that appears to best match the critter. When they've found it, they think they've ID'd the critter. This is kind of the way BugGuide works and is probably my biggest issue with BugGuide.

However, these field guides have descriptions of the critter too. Particularly the butterfly field guides will have both descriptions and arrows on the images pointing to key characters. Looking like an image is only a clue to a critter's identity. The person who wrote the guide has seen far more specimens than can be shown in the guide and is the best person to actually say what he or she found to be common to all of them. So the guide will explicitly say what is clearly common. It seems that only the veteran butterfliers pay much attention to these diagnostic characters.

I think something similar may be going on here with how you were using Reiskind's paper. He can only draw one or a few specimens, and he has to choose which specimens to draw. He makes up for that by describing in the text what he found to be the full range of variability. Many of these papers, particularly Reiskind's, will identify the same figure for several different species, saying in the text that the species has a form similar to that of the figure. The only characters that matter, though, are those that the text explicitly identifies as being diagnostic. The images are there just to help and are only sometimes diagnostic -- only when the text says so.

There's another point too, which Kevin Pfeiffer has helped me learn over the past two years. Even most of the text has to be largely ignored! All these specimens have descriptions that say the dimensions and colors of things and such. The text is attempting to document the range of characters that the author noticed for the species. The problem is that the author has lots of specimens for some species and few for others, and also, regardless of how many specimens they have, they effort they put into the description varies from species to species. When an author thinks they have found what characters distinguish the species, that's what they focus on, and they may not put the energy into fully documenting the species.

So it comes down to just those characters that the author specifically indicates are diagnostic for the species, according to the full palate of specimens they examined. For most spiders, these characters are limited to the palps and epigyna.

There just isn't any way for us to make inferences about species from the few representative samples that an author presents for illustrative purposes. We are dependent on what the author explicitly declares to be common among the species.

(And unfortunately, BugGuide doesn't put diagnostic characters front and center, so it further perpetuates the idea that critters can be ID'd just by comparing images.)

with all your points above. What John B. and I try to do when we focus on a genus of spiders is to read all the available material and do our best to use that to place images posted to the guide. In many cases we find that we are unable to separate species because the ranges and descriptions are too similar (see Tetragnatha, the whole family of Lycosidae, etc.). We’ve tackled and given up on many genera as we’ve gone along. However, when we feel that the species can be separated by field markings and range information we’ve moved forward.

On the info page for Castianeira we've written "... C. occidens can be separated from the other two by the white stripe on the carapace. At this point we aren't sure we can separate C. crocata & C. descripta. It's possible that the red mark goes all the way to the rear in C. descripta and stops short in C. crocata.” I look at it as taking a "you can't prove me wrong" approach. If the description fits, the range fits and we have some field marking to separate it from the other species then we move forward. I don't feel that we're misrepresenting the species here... when using BugGuide as a field guide.

Certainly we have room to improve and it would be great if we had experts helping from each area that could send us verified images, but until then I think it’s best to have a few images as a place-holder to let people know this species is a possibility for this range.

I agree we need to do more to make the diagnostic characters front and center. I've been watching Dick Walton do some great work with the jumping spiders here. I really like the way he has removable red arrows & labels pointing right at the field markings. I wish we could do something like that here.

Dick Walton's site is very nice! I have dreams of creating a wiki-like site for creating a key to the spiders based on field/photo-observable characters.

It does help to have a representative of crocata present to make sure people know that there are alternatives to descripta. We just can't use BG to infer field characters if we label things despite being unsure. I think the problem is more with BG than with how were using it, though. If I could just finish some of my other projects, I might get around to creating that wiki key.

It paused just this once for a picture and then went quickly on its way.