South American Nothochrysinae (Neuroptera, Chrysopidae): I. Description of Nothochrysaehrenbergi sp. nov.

Abstract A new species, Nothochrysaehrenbergisp. nov., is described from Chile; it is the first species of Nothochrysa to be reported from the Southern Hemisphere and only the second from the New World. The genus now contains six extant species as well as two species known from late Oligocene and Miocene fossils. An updated catalog of the valid Nothochrysa species is presented, and three nomina dubia are discussed. The inclusion of the new species in Nothochrysa is well supported by morphological features. However, it and other species currently in the genus also share significant features with Archaeochrysa, an older genus of Nothochrysinae which is known only from the Eocene (Ypresian) to the late Oligocene. It therefore appears that N.ehrenbergi is among the least derived Nothochrysa species, and that the separation of Archaeochrysa from Nothochrysa is open to question and further examination.


Introduction
The family Chrysopidae currently consists of three extant subfamilies. Chrysopinae, with approximately 75% of the known chrysopid genera, is by far the largest (N = ~80 genera). The other two subfamilies combined are much smaller (N =14 genera): Apochrysinae with five genera (Winterton and Brooks 2002) and Nothochrysinae

Systematics of Nothochrysa McLachlan
The genus Nothochrysa has had a tortuous taxonomic history that is well summarized by Tjeder (1966: 264). Briefly, over the years Nothochrysa has included a large number of species that correctly have been moved to other genera, mostly Italochrysa Principi. By the time this study began, the number of species in the genus Nothochrysa had been reduced to only ten -eight extant and two known from fossils (Oswald 2018). However, among the extant species there are three whose validity has been questioned. Thus, with the addition of the new species described here, there are eight confirmed, valid species of Nothochrysa: six extant and two from fossils (Table 1), as well as three nomina dubia (Appendix 1).  Kovanci & Canbulat, 2007 [Eurasia: Turkey] Fossil species praeclara Statz, 1936 [Miocene: Germany] stampieni Nel & Séméria, 1986 [Oligocene: France]

Material and methods
Current usage of terms for veins in neuropteran wings is largely based on the classic studies of tracheal pathways by Tillyard (1916) and Comstock (1918, and his earlier studies with Needham), which were later modified and interpreted by others, e.g., Adams (1967), Kukalová-Peck (1991), Kukalová-Peck and Lawrence (2004), and most recently Breitkreuz et al. (2017). I did not examine tracheal pathways in the current study, and this report uses terminology for veins and cells based on a combination of the above studies. For example, as is customary, the names of the primary veins are abbreviated and capitalized (e.g., C, costa; Sc, subcosta; R, radius; M, media; Cu, cubitus; A1, A2, A3, first, second, and third anal veins; also Psm, pseudomedia and Psc, pseudocubitus). When veins split, I use A and P to indicate the anterior and posterior branches, as proposed by Breitkreuz et al. (2017). In addition, the term "furcation" and its italicized abbreviation "f" are useful in referring to the point on a vein where it forks or splits. Thus, for example, Mf applies to the point on the media where it splits into two branches, the media anterior, MA, and the media posterior, MP.
The names of crossveins are in lowercase, contain a hyphen, and often begin with a number; for example, 1c-sc is the first (basal-most) costal-subcostal crossvein. Cell names are written in lowercase, italicized, and often appended with a number; e.g., csc1 refers to the basal-most cell between C and Sc. For historical and grammatical consistency, I retained the traditional prefix "intra", rather than "inter" (as proposed by Breitkreuz et al. 2017), when referring to cells between two branches of the same major vein. For example, im1 denotes the first "intramedian" cell, and icu3 denotes the third "intracubital" cell. I also reversed the terms "eutriangular" and "pseudotriangular", as used by Breitkreuz et al. (2017) to categorize two types of im1 cells. Their figure 17B, in which the im1 is labeled "pseudotriangular", illustrates a triangular cell with three angles where three entities -two veins (MA, MP) and a crossvein (ma-mp) -intersect. This configuration is a true triangle and should carry the term "eutriangular". Similarly, their figure 17A illustrates another triangular-shaped im1 cell, but this one has two curved sides (MA, MP) and only two angles where the veins intersect. They identified this configuration as "eutriangular", whereas it should be considered "pseudotriangular". The above changes do not affect the authors' interpretation of the venation, nor do they affect figures 17C or 17D. They merely help facilitate grammatical and userfriendly terminology.
The terminal traces of the various major veins were estimated by following the marginal branches basally to their origins on major veins (Fig. 2a, b). In some cases, it is not clear whether a pathway involves actual fusion and/or furcation of longitudinal veins versus the loss and/or insertion of a crossvein. In these cases, marginal veins can be traced to more than one basal origin. Thus, for consistency, the veins within the various areas indicated on Fig. 2a, b are those whose basal-most origin reasonably falls within the indicated field. Given the difficulty in deciphering the fusions and splitting of veins involved in the pseudomedia and pseudocubitus, it is understood that some veins at the margins of each field may stem from more than one basal vein. [Note: For both the forewing and hindwing, I assume that the CuA actually extends distally towards and meets the MP, as opposed to being connected to it via a crossvein. It would be of value to confirm this assumption, via tracheal examination of both wings.] To avoid uncertainty, it is also worthwhile to mention the terms that refer to the orientation of the wing: anterior -toward the elongate margin on the upper (costal) edge of the wing; posterior -toward the elongate margin along the lower edge of the wing; basal or proximal -toward the inner edge of the wing attached to the body; apical or distal -toward the far, outer edge of the wing.
The terminology for other body parts follows common usage. Tauber (Fig. 7f ). This single specimen was found in the CAS collection among the unidentified chrysopids. A subsequent search of the collection did not yield additional examples. Norm Penny's ID label remains on the specimen but was not included in Fig. 7f. It refers to Suarius flavescens, a species that now is placed in Chrysopodes (Neosuarius), and with which the new species shares similar coloration and appearance (see Tauber 2010).

Nothochrysa ehrenbergi
When discovered, the specimen was discolored, and its wings were loosely folded around its body. One pair of wings was removed for study and is now attached with water-soluble hide glue to a card mounted on the pin below the specimen. The other pair fell off and was reattached to the specimen with hide glue. The abdomen was cleared and dissected; it is preserved in glycerin within a genitalia vial attached to the pin.
Diagnosis. Subfamily: This specimen exhibits the following diagnostic features of adult Nothochrysinae (cf. : Tjeder 1966, as Dictyochrysinae;Adams 1967;Brooks and Barnard 1990;Makarkin and Archibald 2013;Breitkreuz 2018): (i) wing-coupling mechanism consisting of a large jugal lobe on the forewing (here, folded ventrally; Fig.  1) and a frenulum on the hindwing (here, broken off); (ii) base of the forewing without tympanal organ (Fig. 1); (iii) forewing (and hindwing) with stem of the media extending basally, adjacent to the radius and not fused with it ( Fig. 1a, b; cf. Breitkreuz et al. 2017: 32); (iv) first intramedian cell triangular, with boundaries formed by the MA, the MP, and the crossvein 1ma-mp ("pseudotriangular", sensu Breitkreuz et al. 2017); (v) pseudomedia ill-defined or appearing to merge with inner (not outer) series of gradates (  Genus placement: The Chilean specimen under study here falls into the genus Nothochrysa on the basis of the following features of its wings (Figs 1, 2): (i) forewing and hindwing having well developed pseudomedia and pseudocubitus; (ii) forewing and hindwing with two regular series of gradate veins (inner and outer); (iii) intramedian cell of forewing triangular, elongate, occupying approximately half the width between the pseudomedia and pseudocubitus; (iv) RP of forewing with 10 or more branches (Adams 1967;Makarkin and Archibald 2013;Archibald and Makarkin 2015;Breitkreuz 2018: 200   than truly quadrangular or triangular (i.e., the MA and MP meet basally at a broadly acute angle); and (iv) the third medial cell (directly below im1, Fig. 2a) is elongate and extends toward the pseudocubitus well beyond the distal edge of first intramedial cell.
Head coloration: Scape cream, with reddish spot on distolateral tip; pedicel, flagellum cream, unmarked; thickset setae in whorls mostly brown, elongate setae pale. Vertex cream, possibly tinged red laterally; dorsal torulus yellow to cream, apparently unmarked. Frons cream, probably with reddish tinge laterally below torulus; torulus cream, unmarked. Clypeus cream, possibly tinged red laterally; basal, distal margins straight. Genal mark dark red/brown throughout, extending to tentorial pit. Labrum probably cream. Palpomeres probably mostly cream, somewhat darkened distally. Thorax (Fig. 4): Cervix not visible. Dorsal thoracic surface with pale longitudinal stripe mesally, probably with broad reddish or brownish stripes or coloration laterally. Prothorax broad, 0.9 mm long, 1.5 mm wide, ratio of length to width = 0.63 : 1; pronotum well sclerotized, with textured surface, transverse fold mesally, few or no setae. Legs elongate, slender, probably cream, unmarked, lacking prominent tibial spurs. Tarsus with basal three tarsomeres appearing coalesced, bearing spurs, setae intermixed along undersurface; middle three tarsomeres with expanded lateral lobes bearing spurs, setae in irregular rows; distal tarsomere narrow basally, enlarged distally, bearing numerous elongate, slender, dark setae laterally, distally, terminus bearing pair of claws laterally, large pad mesally; claw amber, with basal enlargement, acute slender hook terminally.  Wings (Figs 1, 2, 5): Forewing 18.5 mm long, 6.5 mm wide (at widest point); ratio of length to maximum width = 2.9 : 1. Membrane clear, lacking markings; microtrichia present below base of every major vein, pale. Trichosors (sensu Makarkin and Archibald 2013: 140-142) absent. Costal area relatively enlarged; tallest costal cell (7 th from base of wing) 1.8 mm tall, 2.7× width of cell, 0.28× height of wing; costal crossveins simple, six before 1sc-r, twelve after 1sc-r and before stigma, one (very small) after stigma, none within stigma. Sc extending into stigma, fading but not appearing to merge with C or RA; no crossveins in stigma; first sc-r crossvein slightly distal to Rf, slightly basal to Mf; RA with one very short veinlet extending to wing margin after stigma. Radial area between RA and RP with single row of ten closed cells; tallest cell (3 rd from base of wing) 0.6× as tall as wide. Intramedian cell (im1 = mamp1) prominent, elongate, triangular, formed by MA, crossvein 1ma-mp, and two abscissae of MP, occupying approximately half the space between MA and CuA, with Mf broadly acute, long sides (MA, MP) roughly parallel for most of span; crossvein 2m-cu proximal to midpoint of im1. Three medial cells present (mcu, mcua, mpcua), second, third of these elongate, with roughly parallel sides; MP merging into Psc well beyond im1. Two series of gradate veins parallel basally, diverging slightly medially, converging distally. Approximately nine inner gradates in regular, sinuous series, continuing from Psm in zigzag pattern across center of wing; approximately ten outer gradates continuing from Psc in regular, upturned series. RP with nine marginal forks beyond Psc. Cu furcated after m-cu crossvein, with two closed, four open icu cells. CuA with three furcations before meeting MP; CuP furcated below icu2; thus cubital trace having five terminal veinlets (three from CuA, two from CuP). A1, A2, A3 simple, unforked; a1-a2 and a2-a3 crossveins present; distal part of a3 and jugal lobe with dense patch of microtrichia. Jugal lobe large, quadrate, folded beneath third anal cell, without internal vein; margin bearing long, slender setae basally.
Hindwing: 12.4 mm long, 4.2 mm wide. Costal area not enlarged; at least 15 c-sc crossveins before stigma, none within or after stigma. Radial area containing single row of eleven closed cells between RA and RP. Gradate veins in two roughly parallel series, slightly divergent distally; approximately seven inner gradates beyond Psm; approximately 11 outer gradates beyond Psc. Psc with nine marginal forks. MA aligned with RP for approximately one-third length of im1. CuA with two furcations before meeting MP; CuP undivided; thus, wing margin having three cubital veinlets (two from CuA, one from CuP). A1, A2, A3 simple, unforked; a1-a2 and a2-a3 crossveins present. Jugal lobe without internal vein, basal margin bearing long, slender setae.
Coloration of forewing, hindwing (Fig. 5): Membrane clear, somewhat glossy. Stigma slightly opaque, without coloration. Costal, subcostal, radial veins brownish; all other longitudinal veins pale with black marks at intersections and (forewing) at bases of setae. Forewing with posterior veinlets extensively marked black; basal inner gradates pale, others becoming increasingly marked black until entirely black distally; outer gradates mostly black. Hindwing with basal inner gradates pale, marked with black at intersections; outer gradates mostly black.
Gonarcus delicate, slender, broadly arcuate; lateral apodemes slender, quadrate (lateral view), rounded distally, with short, contiguous processes mesally, extending forward. Mediuncus closely attached to dorsal surface of gonarcal arch, flat, recurved into an almost fully circular hood, with two internal sclerotized "rods" extending roughly in parallel from mediuncal base to tip, converging slightly at tip; base of mediuncus quadrate (dorsal view), occupying approximately one-fourth span of gonarcal bridge; terminus of mediuncus with expanded lateral wings, rounded mesal protrusion. Gonosaccus transparent, immediately beneath gonarcal arch and mediuncus, with approximately 32 short setae on distinct setal bases uniformly distributed in two equal patches. Hypandrium internum small, located on delicate membrane extending well below gonosaccus, consisting of paired, curved lateral arms meeting mesally at narrow, rounded apex; comes lightly sclerotized, extending forward beyond apex. Gonapsis, gonocristae absent.
Biology. Nothing is known about the biology or larval morphology of this species. The gut of the N. ehrenbergi specimen did not contain noteworthy contents.
Larval descriptions of several Nothochrysa species are available for comparison if N. ehrenbergi larval specimens were to become available (see Tauber et al. 2014). Nothochrysa larvae generally are considered debris-carriers, but their packets of debris are small, and their morphology is only moderately modified for debris-carrying. In addition, detailed information on aspects of the developmental and reproductive biology of N. californica is available (Toschi 1965).
Known distribution. Currently, this species has only been reported from the type locality, which presumably is the Valle Las Trancas in the region of Ñuble, Chile.
Etymology. This species is named in honor of Ronald G. Ehrenberg, Irving M. Ives Professor of Industrial and Labor Relations and Economics at Cornell University, an esteemed and cherished colleague of the author and her late husband (Maurice J. Tauber).

Characteristics shared with Archaeochrysa species
As shown above, N. ehrenbergi shares many features with other extant Nothochrysa species, and its inclusion in the genus is well supported. However, the species also expresses many features that differ from Nothochrysa and that are shared by at least some of the five species in the fossil genus Archaeochrysa. I discuss four below: First, in the N. ehrenbergi forewing, vein A1 is not forked, whereas it is forked in all other Nothochrysa species (Adams 1967;Aspöck et al. 1980: figs 154, 155;Makarkin and Archibald 2013: 135, 136). The feature is variable in Archaeochrysa specimens where A1 is visible. It is not forked in two species (Adams 1967: 237), forked in two species (Adams 1967: 230, Makarkin andArchibald 2013: 135), and missing from the specimen of the fifth species (Archibald and Makarkin 2015: 363).
Second, in N. ehrenbergi the basal sc-r crossvein arises distal to the furcation of the radius and almost directly above the furcation of the media. Both of these character states are shared with the fossil genus Archaeochrysa (Adams 1967, Makarkin andArchibald 2013), but not with other known Nothochrysa species.
Third, in N. ehrenbergi the distinction between the inner gradate series and the pseudomedia as well as between the outer gradate series and the pseudocubitus is indistinct. Rather, the gradate series and their respective pseudoveins tend to run together more smoothly as a curve, rather than at an angle as in other Nothochrysa species. Again, this feature of N. ehrenbergi is shared most closely with Archaeochrysa species (Adams 1967, Makarkin and Archibald 2013, Archibald and Makarkin 2015. Fourth, currently the primary feature used to distinguish between Nothochrysa and Archaeochrysa is the presence or absence of a crossvein between RP and MA in the basal part of the hindwing. The crossvein is present in all known Archaeochrysa species and is reported to be absent from Nothochrysa (Makarkin and Archibald 2013: 134).
In N. ehrenbergi, MA aligns with RP for about one-third the length of the upper margin of the im1 cell, and no crossvein is present (Figs 1b, 2b). However, even with this character there appears to be a possible exception. Figure 2 accompanying the original description of Nothochrysa turcica Kovanci and Canbulat shows a short crossvein between RP and MA; confirmation of the accuracy of this drawing is necessary.
Given the above, Archibald and Makarkin's (2015) discussion of the phylogeny of Archaeochrysa species is worthy of consideration here. Their paper evaluates how the various Archaeochrysa species express three features; each feature has several conditions ranging from presumably plesiomorphic to more derived. Below, the three features are considered, relative to their expression by Nothochrysa species, especially N. ehrenbergi.
(1) The shape of the im1 cell. Archibald and Makarkin (2015) describe two configurations for this character; N. ehrenbergi expresses the second (more advanced) condition in which the sides of the im1 cell are almost parallel for most of their span and converge basally at a relatively steep angle. The extant species of Nothochrysa, including N. ehrenbergi, share this feature with two species of Archaeochrysa. (2) The position of crossvein 2m-cu. Archibald and Makarkin (2015) list six conditions for this character, each one considered more evolutionarily advanced than the preceding. Nothochrysa ehrenbergi falls into Condition 5, a derived condition in which 2m-cu is located distinctly in the proximal part of im1 (as shown in fig.  2C of Archibald and Makarkin 2015). This character state is typical of at least two Archaeochrysa species, A. creedi (Adams) and A. paranervis (Adams), as well as several other extant genera in Nothochrysinae, including Nothochrysa. (3) The crossveins of Psc. Archibald and Makarkin (2015: 366) describe and illustrate four character states for this feature; interested readers are referred to the original paper. Suffice it to say here, N. ehrenbergi, as well as three Archaeochrysa species but no other Nothochrysa species, fall into the second of the four conditions. This position is considered plesiomorphic among Nothochrysinae, both fossil and extant (Archibald and Makarkin 2015).
On the basis of the above information, it appears that N. ehrenbergi shares a very close phylogenetic relationship with the fossil genus Archaeochrysa. At this point, only one character (the absence of a crossvein between the RP and the MA above the first intramedial cell of the hindwing) supports its exclusion from Archaeochrysa, and this character may have exceptions within Nothochrysa. Indeed, there does not appear to be a synapomorphic character that consistently differentiates Nothochrysa from Archaeochrysa. Thus, given the overall similarity between N. ehrenbergi and the known Archaeochrysa species, I recommend that future studies examine the validity of maintaining the generic separation.