Ceraeochrysa lineaticornis ( Fitch, 1855 )

Tauber, Catherine A. & Flint, Oliver S., 2010, Resolution of some taxonomic and nomenclatural issues in a recent revision of Ceraeochrysa (Neuroptera: Chrysopidae), Zootaxa 2565, pp. 55-67 : 59-60

Ceraeochrysa lineaticornis ( Fitch, 1855)

Chrysopa lineaticornis Fitch, 1854 [1855]: 795 [MCZ, Holotype].

Ceraeochrysa lineaticornis (Fitch) . First combination in Ceraeochrysa by Adams (1982: 73).

Chrysopa columbiana Banks, 1903: 150 [MCZ, Holotype]. Junior subjective synonym of C. lineaticornis by Bram & Bickley (1963: 16); later, junior subjective synonym of Cer. parvula (Banks) by Freitas et al. (2009: 566). Here: Subjective synonymy with C. lineaticornis Fitch reinstated.

Allochrysa parvula Banks, 1903: 143 [MCZ, Holotype]. Junior subjective synonym of Cer. lineaticornis (Banks) by Adams (1982: 73). Synonymy reversed and Ceraeochrysa (= Allochrysa ) parvula (Banks) recognized as a valid species by Freitas et al. (2009: 566). Here: Subjective synonymy with C. lineaticornis reinstated.

Background: Freitas et al. (2009: 557, 566) recognized Cer. lineaticornis and Cer. parvula (with junior synonym C. columbiana ) as two separate species. They claimed that Banks (1903) considered dark genal markings as an important feature that distinguishes the two species; their description and illustration of Cer. parvula (Fig. 39C in Freitas et al. 2009) include dark genae, whereas pale genae typify Cer. lineaticornis . Furthermore, they mentioned that male genital differences distinguish the two species, but they did not specify what the features are.

Types: Banks (1903: 150) described the genae of all three nominal species as pale and without marks. Moreover, the on-line images of the C. lineaticornis , A. parvula and C. columbiana holotypes [MCZ type database; http://insects.oeb.harvard.edu/MCZ/index.htm] and personal examination of the types in the MCZ (CAT), show that the genae of all three nominal species are pale and without marks.

We (CAT) examined the genitalia of the holotypes (all males) of the three nominal species. Those of C. lineaticornis and C. columbiana are from mature specimens; they are well sclerotized and well preserved. The A. parvula holotype is teneral and the genitalia are less well preserved, but they exhibit some of the relevant characters. Below, we assess the three holotypes relative to possible differences and similarities that we could glean from the descriptions by Freitas et al. [The terminology for the structures is that used by Freitas et al. 2009)

(1) Dorsal apodeme of ectoproct. For Cer. lineaticornis , Freitas et al. (2009) described the dorsal apodeme of the ectoproct as having a “recurved dorsal branch”, whereas that of the Cer. parvula ectoproct as “bifurcate and ventral branch caudally projected”. Freitas et al. ’s drawings of the male terminalia and our examination of the holotypes show that the dorsal apodemes of both nominal species have a dorsally projecting arm that extends behind the callus cerci, and a caudal arm that projects posteriorly. The lengths and degree of sclerotization of the arms of the two species appear to differ on the drawings for the two species (Figs 33D, 39E in Freitas et al. 2009), but the differences shown are well within the continuous range of variation seen amongst Cer. lineaticornis specimens, e.g., in a sample of specimens in the USMN.

(2) Gonarcal arch. Freitas et al. described the “gonarcal arch” of Cer. lineaticornis as “short, thick” and the “gonarcal medial arch” of Cer. parvula as “stout, anterior margin highly curved”; their drawings are consistent with the descriptions. Our examinations of the holotypes and numerous Cer. lineaticornis specimens indicate that there is considerable continuous variation in the curvature and thickness of the gonarcus and that the holotypes of all three nominal species fall well within this range.

(3) Lateral arms of gonarcus. For Cer. lineaticornis , the gonarcal arms are described as “large, hemispherical, ventrally orientated”; the gonarcal arms of Cer. parvula are not mentioned in the description. Our examinations show that they are similar on the holotypes of all three nominal species.

(4) Gonocornua. The Cer. lineaticornis gonocornua are described and figured as “long, arched and acutely pointed apically and with small ventral projection which may be the entoprocesses”. The Cer. parvula gonocornua are described as having an “acute apex and ventral projection”; in Fig. 39G of Freitas et al. (2009), the Cer. parvula gonocornua are shown to be somewhat shorter than those of Cer. lineaticornis (Fig. 33F), and with a slightly larger ventral projection. Here, again, the differences between the types of the nominal species in the lengths of their gonocornua and the associated ventral projections are small and within the range of variation of Cer. lineaticornis specimens.

(5) Arcessus. The arcessus of Cer. lineaticornis is described as having a “long dorsal horn, apical plate trilobed with long acute lobes”; Cer. parvula is described as having a “narrow arcessus, apical plate trilobed, the medial point horn-like, no dorsal processes”. Figs. 33G and 39I in Freitas et al. (2009) are consistent with the descriptions. Our examination of the C. lineaticornis holotype indicates that indeed its arcessus is longer and narrower, the three terminal lobes are more elongate, and the dorsal processes are more prominent than those on the other two holotypes. However, Cer. lineaticornis specimens generally are very variable in these characteristics, and the three holotypes fall well within the range of continuous variation exhibited by Cer. lineaticornis specimens. It is noteworthy that the dorsal processes on the arcessus are often flat and closely aligned to the surface of the arcessus; they can be difficult to see, especially on teneral specimens like the A. parvula holotype.

(6) Gonosaccus. Both species are described and figured as having a gonosaccus with long gonosetae. Indeed, this is the case for the holotypes of all three nominal species.

(7) Gonapsis. The gonapsis of Cer. lineaticornis is not mentioned or figured; that of Cer. parvula is described as “apically forked”. Indeed, the gonapsides of the C. lineaticornis and C. columbiana holotypes are forked at their junction with the tip of S8+9; the free end on both specimens is spatulate. The gonapsis of the teneral A. parvula holotype is difficult to see.

In summary, we were unable to identify any genitalic features that would distinguish the A. parvula or C. columbiana holotypes from Cer. lineaticornis . Thus, on the basis of external (head markings) and internal (male genitalic) characters, we consider that all three belong in the same species.

As an aside, it is appropriate to mention here an independent, biological, character that supports the synonymy of A. parvula with Cer. lineaticornis . The “mass of rubbish” ( Banks 1903: 144) that encloses the cocoon from which the A. parvula type emerged consists largely of spiny trichomes. Cer. lineaticornis larvae are the only Ceraeochrysa larvae known to carry and incorporate spiny trichomes into their cocoons ( Eisner et al. 2002).

Conclusion: It appears that the recognition (by Freitas et al. 2009) of Cer. parvula as a valid species (with Cer. columbiana as a junior synonym) was based on an obvious error and conjecture. We consider that the earlier identifications (by Bram & Bickley 1963 and Adams’ 1982) of A. parvula and C. columbiana as junior subjective synonyms of Cer. (=C.) lineaticornis are correct and should be maintained.

Moreover, the specimen with dark genae illustrated as Cer. parvula in Freitas et al. (2009: 567, Figs. 39C and 39D) is clearly not the holotype of either A. parvula or C. columbiana . Its identity should be re-evaluated. Other than the A. parvula and C. columbiana type specimens, the only specimen of Cer. parvula that Freitas et al. (2009) reported to have examined is a male from Honduras [in the California Academy of Sciences (CAS)]. Thus, we conclude that the illustrations mentioned above were derived from this specimen. Given that it was collected far from the previously reported range of Cer. lineaticornis (and its two synonyms) and given its dark genae, the identification should be re-evaluated.