[Note: Full-size image can be seen here.]

Found at a mercury vapor light on the first night of the summer 2013 BugGuide gathering in AZ. I'd been wanting to see one of these amazing creatures for years!! The last two images of this series show the specimen on a cloth sheet at the lights...the preceding images were taken later on a sheet of paper.

It was fairly straightforward narrowing the ID here down to either Dicromantispa sayi or Leptomantispa pulchella. But how do you distinguish between those two often very similar-looking species...both of which are themselves quite variable, and have distinctively different light and dark forms? Perusing the comments accompanying BugGuide posts for these species will show this question can become a problematic quagmire! I became engrossed in trying to fathom its nuances and understand some of the subtleties involved.

The principal character used to separate the taxa in the Canning's 2006 treatment of Mantisipdae of Canada, and Hoffman's 2002 treatment of Mantispidae of Costa Rica, involves the hairs of the "giraffe-like" necks (or pronota) of these mantispids. In L. pulchella there are short setae ("hairs") distributed uniformly along the length of the pronotum, while in D. sayi the setae are finer still and mostly localized at the anterior and posterior ends of the dorsum of the pronotum (or virtually lacking). A problem with this character is that, even with a good profile angle of view, in most field images the magnification/resolution is insufficient to discern the setae. Luckily, one of my images shows the character fairly clearly. (And I was also able to check the setae under a microscope.) As a secondary character, the Canning brothers also mention that in L. pulchella the (upper surface of the) pronotum is smooth, while in D. sayi it has evident transverse ridges. But it's not clear that how consistent this character is over a large numerical and/or geographical sample.

However, it turns out that in certain cases you can bypass the subtle pronotal setae character. In particular, if your mantispid has tri-colored antennae...with basal 2/3 brown, next 1/6 yellow, and last 1/6 dark brown...then you have Dicromantispa rather than L. pulchella! (See couplet 2 in Hoffman's key). That's the case in my post here. The antennae of L. pulchella, on the other hand, are uniformly brown to black...except for the basal segment which is yellowish.

But it turns out that for many D. sayi specimens, the antennae can be identical to those of L. pulchella,...so the "tri-color antennae" character doesn't always save the day. But then, you might ask, if D. sayi can have uniformly dark antennae, doesn't that contradict Hoffman's main key character for D. sayi ??!!

I believe the answer to that question is no...because Hoffman's key treats Costa Rica, and presumably the form of D. sayi present in Costa Rica has consistently tri-colored antennae. I'm speculating that the tri-colored antennae form ranges from Costa Rica northward into the warmer southern part of the US...which is supported by the (verifiably identified) image data on BugGuide.
[Postscript (8/24/13): I recently obtained Hoffman's excellent dissertation(1), which confirms that "the tricolored antennae occur on specimens from the southern United States southward". The southern range limit given for D. sayi in that work is Panama(1).]

Interestingly, it turns out D. sayi comprises (at least) three different color forms which are largely allopatric (see 2nd paragraph here). Those color forms correspond to three previously named species: Mantispa sayi, M. fuscicornis, and M. uhleri (original descriptions here,  here, and at bottom of pg. 79 here)...all of which were synonymized under D. sayi in Hoffman (1989). The dark forms of D. sayi, from east of the Mississippi River and northward...having much black coloration on the abdomen, pronotum and forelegs (especially in females)...correspond to the former Mantispa uhleri. Lots of interesting life history and other info on M. uhleri (=D. sayi) can be found here. These dark "uhleri" forms of D. sayi can look very much like the the dark forms of L. pulchella...and the light forms of D. sayi (with their dark antennae) can look very much like the light forms of L. pulchella. So what else can be used to tell the two apart ??

Well, in general, L. pulchella is significantly smaller in overall length and size than D. sayi. But precise scale measure is not clear in most images, and besides...adult mantispids within a given species can vary in size by very large amounts, since their final size depends on the size of the spider egg sac the larva parasitizes, and mantispids attack many different spider taxa of many different sizes.

A trick I happened upon (not mentioned explicitly in any references I've found) is to carefully scrutinize the wing venation. The details are a bit cumbersome to describe here, but basically the smaller relative size of L. pulchella translates to fewer cells and veins along the length of the wing than in D. sayi. (In an informal terminology I've adopted, the rubric is: "5 wishbones along the posterior edge of the forewing for pulchella, and 8 wishbones for sayi".) Perhaps I'll write up a forum article with details and diagrams if there's interest. At any rate, I'll just end by encouraging those of you who haven't done so already to read about the fascinating life history, biology, and ecology of these intriguing organisms. Redborg (1998) is an excellent review article. And I'll pose a question: Is this a male or a female? :-)