forum debris-carrying in larval chrysopidae: unraveling ... › neuropterida ›...

FORUM

Debris-Carrying in Larval Chrysopidae: Unraveling ItsEvolutionary History

CATHERINE A. TAUBER,1,2 MAURICE J. TAUBER,1 AND GILBERTO S. ALBUQUERQUE3

Ann. Entomol. Soc. Am. 107(2): 295Ð314 (2014); DOI: http://dx.doi.org/10.1603/AN13163

ABSTRACT Larval debris-carrying, which occurs in many insect taxa, is a remarkable behavioraltrait with substantial life history signiÞcance. For the Chrysopidae, information on the topic isscattered, and the habitÕs diversity and evolutionary history are unassessed. Here, we compile acomprehensive, annotated catalog on chrysopid debris-carrying and its associated larval morphology,and we identify emerging systematic patterns of variation, from larval nakedness to the constructionof elaborate packets. Then, we examine these patterns in the context of available phylogenies withtwo objectives: 1) to begin unraveling the evolutionary history of chrysopid debris-carrying and 2)to evaluate the current and potential role of larval morphology (including debris-carrying modiÞ-cations) in classiÞcation and phylogeny of this family. Debris-carrying: the literature revealed thatdebris-carrying occurs in the chrysopoid fossil record and in all three extant chrysopid subfamilies,including those proposed as basal (Nothochrysinae, Apochrysinae). Nevertheless, the familyÕs an-cestral state remains unresolved. The habit may have evolved at least once in Nothochrysinae or beenlost several times.Larvae fromonlyonegenusofApochrysinaeareknown, and theyaredebris-carriers.Each of the four tribes of the third subfamily, Chrysopinae, has distinctive debris-carrying charac-teristics. In ankylopterygine larvae, debris-carrying modiÞcations seem relatively conserved. Amongthe ant-associated belonopterygine genera, debris-carrying is either highly evolved or, in one case,possibly absent. Within the large chrysopine tribe, nakedness and debris-carrying appear to haveevolved independently numerous times; also, some reversals may have occurred. With one possibleexception, leucochrysine genera have debris-carrying larvae. Larval morphology: scrutiny of theliterature showed that all chrysopid genera whose larvae are known exhibit characteristic suites ofanatomical structures related to carrying debris. Moreover, larval morphology provides strong (sy-napomorphic) evidence for the monophyly of four of the seven suprageneric chrysopid taxa: thesubfamily Nothochrysinae and three of the four tribes of Chrysopinae (Ankylopterygini, Belonop-terygini, and Leucochrysini). Larval morphological and debris-carrying characteristics appear tosupport some, butnot all, previouslyproposedgeneric relationshipswithin the tribeChrysopini.Giventhe demonstrated potential advantages for including larval characters in chrysopid phylogeneticstudies, it is essential to enhance the pool of available larval data. Therefore, we propose thatcitizen-scientists be involved in gathering veriÞable data and that systematists develop comprehensivedata matrices for comparative larval studies.

KEY WORDS camoußage, phylogeny, larval morphology, trash-carrying, Neuroptera

Among insects, the larval habit of camoußaging orshielding the body with exogenous organic or inor-ganic material (� debris-carrying) is an elaboratebehavioral attribute with far-reaching life history andevolutionary consequences (e.g., see Principi 1943,1946; Eisner et al. 1978; Milbrath et al. 1993, 1994;Tauber et al. 1995b; Brandt and Mahsberg 2002; Bu-cheli et al. 2002). Examples exist in a wide range of

insect orders,4 including Hemiptera, Trichoptera,Lepidoptera, Coleoptera, and Neuroptera. Within theChrysopidae, this behavior occurs in the majority(�65%) of genera for which larvae are known (Ap-pendix 1), and it appears closely associated with the

1 Department of Entomology, Comstock Hall, Cornell University,Ithaca, NY 14853-2601 and Department of Entomology, University ofCalifornia, Davis, CA 95616.

2 Corresponding author, e-mail: [email protected] Laboratório deEntomologia eFitopatologia,CCTA,Universidade

Estadual do Norte Fluminense, Campos dos Goytacazes, RJ, Brazil28013-602; e-mail: [email protected].

4 Hemiptera: Reduviidae (Butler 1923, Brandt andMahsberg 2002);Trichoptera: most families (Wiggins 1977); Lepidoptera: Geometri-dae (Ferguson 1985), Coleophoridae (Bucheli et al. 2002, Falkovitsh2003), Psychidae (Rhainds et al. 2009); Coleoptera: Derodontidae(Brown and Clark 1962), Curculionidae (Gressitt et al. 1968),Chrysomelidae (Chaboo et al. 2008, Chaboo 2011), Endomychidae(Leschen and Carlton 1993), Sphindidae (McHugh and Kiselyova2003); Neuroptera: Ascalaphidae (Henry 1977); Myrmeleontidae(Robert Miller and Lionel Stange, personal communication), Sisyri-dae (Killington 1936, 1937), Chrysopidae (e.g., Smith 1926; Killington1936, 1937; Catalog, Supp Table 1 [online only]).

0013-8746/14/0295Ð0314$04.00/0 � 2014 Entomological Society of America

http://dx.doi.org/10.1603/AN13163ST1

evolutionary diversiÞcation of the family (Tauber andTauber 1989, Tauber et al. 1993).

Much information on debris-carrying within theChrysopidae has accumulated over the past �90 yr.During this period, systematic and natural historystudies documented a wide range of variation inchrysopid debris-carrying and concomitant modiÞca-tions in larval anatomy. However, the information hasbeenneither compiled nor examined in detail, and theevolutionary history of debris-carrying has not been ex-plored phylogenetically. To assess current knowledgeand provide a foundation for future work, we: 1) assem-ble, in a systematically based catalog, the dispersed lit-erature on the larval morphology and debris-carryinghabitsofchrysopidsandtheirchrysopoidancestors(Cat-alog, Supp Table 1 [online only]); 2) search for evolu-tionary patterns contained within the current literature;3) highlight signiÞcant Þndings for speciÞc taxa; and 4)identify taxa whose poorly known larvae might yieldespecially important comparative information.

Finally, despite the proven value of immatures inphylogenetic analysis in other insect groups (e.g., seeOberprieler et al. 2000, Meier and Lim 2009), wedocument their almost complete absence fromchrysopid phylogenies. In doing so, we also provide arationale for using larval characters, including thoserelated to debris-carrying, in phylogenetic analyses ofthe Chrysopidae, and we offer recommendationswhereby specialists and citizen-scientists can cooper-ate to accelerate the process.

Background

Debris-Carrying. Both the behavioral and morpho-logical components of larval debris-carrying inchrysopids exhibit valuable taxon-speciÞc variation.Although the behavior involved in constructing a dor-sal cloak or packet (often referred to as “loading be-havior”; see Fig. 2F) is described for only a very smallnumber of species, many reports have focused on thepacket as an indicator of the underlying behavior. Forexample, in nature, larvae ofmost species appear to beselective in the material they use for their coverings,and as a consequence, the composition of the packetcan be taxonomically signiÞcant. Depending on the spe-cies, thismaterialmayincludewaxyßocculencefromthesternorrhynchan prey; arthropod exoskeletons, exuviae,snail shells, or fragments thereof; small pieces of driedleavesorwood, trichomes,or lichens; silkenthreadsfromspiders or mites; and sand or soil (e.g., see Smith 1922,Eisner et al. 2002, Anderson et al. 2003).

Other aspects of the larval packet also vary amongspecies, for example, the amount (weight) of the ma-terial carried, the size of the packet relative to thebody, the mode of attachment, and the degree of thepacketÕs cohesiveness. Any of these features could betaxonomically informative; regrettably, their value isunknownbecause reports in the literature are few andinterspeciÞc comparisons are meager. Typical cover-ings include a small number of scattered pieces ofmaterial loosely held by hooked setae on the dorsum,light layers of material intertwined with long setae,

loosely to tightly constructed cloaks that partially orcompletely cover the larval body, and very denseshields that are tightly attached and intertwined withsilken strands and that extend well beyond the outlineof the larval body.

Furthermore, larvae show interspeciÞc variation inthe frequency with which they carry debris: somespecies carry debris occasionally or during a speciÞcinstar, which is usually the Þrst instar (“occasionaldebris-carriers”). Generally, the occasional debris-carriers carry scattered pieces of material. Ideally, ifsufÞcient data were available, we could categorizeeach species (e.g., “debris-carrier” if all instars con-sistently have a distinct covering of material; “lightdebris-carrier” if all instars usually carry a few, scat-tered fragments of material; “occasional debris-car-rier” if only some instars or some individuals carrymaterial; or “naked” if all instars are usually withoutdebris). In reality, for most species, data are onlysufÞcient to differentiate “naked” (without debris)from “debris-carrying” in the broad sense of the term(see Figs. 1Ð3 for examples).

In contrast to the behavioral variation above, thenumerous features of larval morphology implicated indebris-carrying are relatively well-documented, and abroader rangeof taxa is included.Table 1describes thefeatures that are typically associatedwith eithernakedlarvae or with debris-carriers. What is especially strik-ing is that larvae express great variation in these traitsand yet no speciÞc combination of features deÞneseither of the two lifestyles.

Literature. Larval debris-carrying in Chrysopidaehas been documented for �275 yr (since RéaumurÕs[1737] report). Many subsequent papers referred tothe behavior, but often anecdotally and with varyingamounts of detail. For some taxa, the informationreßects numerous observations accumulated overmany years; in others, only one observation may beavailable. Our catalog here (Catalog, Supp Table 1[online only]) includes all the references to larvaldebris-carrying and larval morphology that we foundfor each chrysopid species.

Larval morphology has long been a key element inneuropteran systematics and phylogeny at the ordinallevel (initiatedbyWithycombe[1922, 1924]andTilly-ard [1926], advanced by Ellis MacLeod [1960, 1964],and continuing to the present by Aspöck and Aspöck[2007] and Monserrat [1996, 2008]). Moreover, at thefamily level, comparative larval morphology has con-stituted a substantial element in the classiÞcation ofthe two largest neuropteran families: Myrmeleontidae(e.g., Stange and Miller 1990, Stange 2004, Miller andStange 2012) and Chrysopidae (see below).

For the family Chrysopidae, comparative descrip-tive studies that deal with larvae began mainly duringthe 1920s and 1930s, with studies by Smith, Lacroix,Withycombe, and Killington (Appendix 1; Catalog,Supp Table 1 [online only]). Later researchers, in the1940s through the 1970s (Principi, Tjeder, Tauber,Monserrat, Tsukaguchi, and others), identiÞed thevalue of larval morphology in deciphering the identityand relationships of chrysopid taxa especially at the

296 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 107, no. 2

http://dx.doi.org/10.1603/AN13163ST1

generic level; during the 1940s and 1950s, the work ofProfessor Maria Matilde Principi set the “gold stan-dard” in morphological detail, clarity, and esthetics.

The last three decades saw a considerable increasein information on immature stages from various re-gions (Appendix 1; Catalog, Supp Table 1 [onlineonly]): Africa (Hölzel and colleagues), Asia (Tsuka-guchi), Australia and Micronesia (Adams; New;Boros), Hawaii (Tauber and Tauber), Europe (Geppand colleagues; Monserrat and Dṍaz-Aranda), and theNew World (Tauber, Tauber, Albuquerque, and col-leagues; Freitas). Publications by these and other au-thors provide the basis for this review. The recentcompilation of research on the Iberian fauna (Mon-serrat and Dṍaz-Aranda 2012) takes an in-depth viewof the fauna, from subfamilies to species, and it em-phasizes the great advantage of using immature stages(especially Þrst instars) in chrysopid systematics at allhierarchical levels. It constituted a valuable resourcefor this work.

Our bibliographic record (Catalog, Supp Table 1[online only]) is presentedwithin a taxonomic frame-work. Each listing identiÞes the type of informationthat is contained in the reference cited: whether lar-vae are described and, if so, which instar(s); whetherdrawings or photos are provided and, if so, whichinstar(s); and,whether additional information is avail-able (e.g., the type of material carried, the instar[s]carrying debris). We restricted our citations on thetype of material used to those packets that larvaeconstructed in their natural habitat because in thelaboratory, larvae often carry material not used innature. Despite our efforts, we may have missed somepublications.

Although the chrysopid larval stage includes threeinstars (L1, L2, and L3), most current informationrefers only to the Þrst or third. The Þrst instar (Sema-phorontA)differsmarkedly in structure, setation, andoften coloration from the other two instars (Sema-phoront B), which differ from each other only inminor ways that are largely related to size and setalnumbers. Thus, for systematic purposes, Þrst and thirdinstars constitute themost useful andphylogeneticallyinformative larval stages for evaluation; they are theones that we use here (see Wheeler 1990 for a dis-cussion of phylogenetically relevant comparisonsamong instars).

Phylogenetic Framework. As a basis for our analy-sis, we used the three phylogenetic studies that areavailable for the Chrysopidae; all three offer provi-sional and restricted, but useful, results. One is basedon morphological characters; it includes all genera inthe family, but the phylogenetic comparisons arelargely informative only at the subfamilial and triballevels (Brooks and Barnard 1990, expanded by Brooks1997). It is noteworthy that all but one of the mor-phological characters in this study stem from the adultstage. The other two phylogenetic studies (Wintertonand Freitas 2006, Haruyama et al. 2008) are based onmolecular data; their analyses extend to the genericlevel, but each included a very different range of taxaandeachhas signiÞcant limitations in thediversity andnumber of species it sampled.

Given the disparities among each of the above stud-ies, it is not surprising that they differ in their supportfor the monophyly and relationships among the threechrysopid subfamilies (Apochrysinae, Chrysopinae,and Nothochrysinae). In the morphologically based

Fig. 1. Examples of naked chrysopid larvae: all Chrysopinae, Chrysopini. (A) Chrysoperla rufilabris (Burmeister), (B)Meleoma signoretii Fitch, (C) Chrysopa oculata Say, and (D) Anomalochrysa hepatica McLachlan. Photos by: Lyle Buss (A),Stephen A. Marshall (B and C), and M.J.T. (D). (Online Þgure in color.)

March 2014 TAUBER ET AL.: CHRYSOPID DEBRIS-CARRYING 297

phylogeny (Brooks and Barnard 1990, Brooks 1997),all three subfamilies are considered monophyletic,andNothochrysinae is sister andbasal to theother twosubfamilies.

In comparison, theÞrstmolecular study(Wintertonand Freitas 2006) recovers only two of the subfamilies(Nothochrysinae and Chrysopinae) as monophyleticclades, and a weakly supported Apochrysinae (exclu-

sive of one exceptional genus) is found to be sister toNothochrysinae � Chrysopinae. This study involved18 of the �80 known chrysopid genera; all suprage-neric taxa included at least one representative.

The secondmolecular study (Haruyama et al. 2008)included more genera (n � 24) than the Þrst. Mostwere in Chrysopinae (tribe Chrysopini); tribe Leu-cochrysini was excluded. In this study, Chrysopinae is

Fig. 2. Examplesofdebris-carryingchrysopid larvae. (A)Apochrysinae:Apochrysamatsumurae(Okamoto)(identiÞcationnotconÞrmed), carrying ßocculence and material from male “cocoons” of the coccoidean Drosicha corpulenta (Kuwana); (B)Chrysopinae,Ankylopterygini:Semachrysamatsumurae(Okamoto)(identiÞcationnotconÞrmed),carryingexuviaeandarthropodparts; (C) Chrysopinae, Belonopterygini: Italochrysa italica (Rossi) (identiÞcation by R. A. Pantaleoni), carrying pieces of woodymaterial; (D) Chrysopinae, Chrysopini: Cunctochrysa albolineata (Killington) (identiÞcation by R. A. Pantaleoni), carryingßocculence, miscellaneous animal material; (E) Chrysopinae, Chrysopini: Pseudomallada sp. (identiÞcation not conÞrmed); (F)Chrysopinae, Chrysopini: Chrysopa slossonae Banks, Þrst instar engaging in “loading behavior,” with ßocculence from wooly alderaphids (Prociphilus tesselatus [Fitch]). Photos by Yusei Hara (A and B), Roberto A. Pantaleoni and Carlo F. Cesaroni (C), BrianP. Valentine (D), Gilles San Martin (E), and Thomas and Maria Eisner (F). (Online Þgure in color.)


the only subfamily with support for monophyly; it wasrecovered as sister to an unresolved Apochrysinae �Nothochrysinae. Thus, the early evolutionary historyof the Chrysopidae remains largely enigmatic.

Despite the above limitations and differences, eachof the three phylogenies presents some credible phy-logenetic hypotheses for evaluating currently knownpatterns of chrysopid debris-carrying.

Evolutionary Patterns of Larval Debris-Carrying

Family Chrysopidae

At least some larvae are known from each of thechrysopid suprageneric taxa (Appendix 1; Catalog,Supp Table 1 [online only]). As a result, severalgeneral observations regarding debris-carrying inthe Chrysopidae are apparent: 1) debris-carrying

occurs within all the major lineages of Chrysopidae,including the three subfamilies and all four tribes ofChrysopinae. 2) The larvae of each chrysopid genusstudied to date express a unique suite of anatomicalfeatures, most of which may be associated with thedebris-carrying habit. 3) Because the available lar-val data and the phylogenetic trees lack sufÞcientsampling of basal taxa, it is not yet known whetherdebris-carrying or nakedness is the ancestral statefor the family. 4) Larval synapomorphies supportthe monophyly of the subfamily Nothochrysinaeand three of the four tribes in Chrysopinae (Anky-lopterygini, Belonopterygini, and Leucochrysini).However, they have not yet been identiÞed forsubfamilies Apochrysinae and Chrysopinae or thetribe Chrysopini. 5) The larvae of each chrysopidgenus express a unique suite of characteristics; in

Fig. 3. Examples of debris-carrying chrysopid larvae. (A)Chrysopinae, Chrysopini:Ceraeochrysa smithi (Navás) carryingsmall fragments of dried plant material, (B) Chrysopinae, Chrysopini: Ceraeochrysa lineaticornis (Fitch) carrying trichomesfrom sycamore leaves, (C) Chrysopinae, Leucochrysini: Leucochrysa (Leucochrysa) insularis (Walker) carrying snail shellsand living snails, and (D)Chrysopinae, Leucochrysini:Leucochrysa (Nodita) pavida (Hagen) carrying lichens (ventral view).Photos by Thomas and Maria Eisner (A, B, and D) and Daniel C. Dourson (C). (Online Þgure in color.)


some cases, individual character-states appear au-tapomorphic for the genus.

Our speciÞc, taxon-based analysis begins below,with the two small, presumably basal, chrysopidsubfamilies: Nothochrysinae and Apochrysinae. Thelarger subfamily Chrysopinae and its four tribes fol-low, and the analysis concludes with the chrysopoidfossil record and a brief summary of the resultingtaxon-based patterns.

Subfamily Nothochrysinae

Despite its small size, the subfamilyNothochrysinaeexhibits a broad range of variation in its debris-carry-ing habits. Larvae are known from four of its ninegenera (Catalog, Supp Table 1 [online only]), ofthese, three are naked (Hypochrysa, Dictyochrysa, andKimochrysa), and one is a debris-carrier (Notho-chrysa).

None of the naked larvae in the subfamily are re-ported to have any of the usual morphological mod-iÞcations for debris-carrying (New 1981; MonserratandDṍaz-Aranda 2012;C.A.T., unpublisheddata). Theknown larvae (L2 and L3) of Hypochrysa and Kimo-chrysaaregreenandverycryptic against abackgroundofgreen foliage(Duelli et al. 2010), and theyhaveveryunusual “bacilliform” (� short, rodlike, blunt toslightly clavate) setae (Monserrat and Dṍaz-Aranda2012, C.A.T., unpublished data). Those of Dictyo-chrysa (L1) arebuff-coloredwithbrownpatches; theytoo have short, blunt dorsal setae (New 1981).

All of the known larvae of the fourth genus, Notho-chrysa (n � 3 of 7 species), are debris-carriers (Cat-alog, Supp Table 1 [online only]). Typically, theirpackets are substantial to moderate in size, and theycontain few, but relatively large, fragments of woodyor leafy(dried)plantmaterial, lichens, and frass.Theirmorphological modiÞcations for debris-carrying aremodest; for example, thebodies are only slightly thick-

ened, the lateral tubercles of the thorax and abdomenare small and carry few setae, and laterodorsal tuber-cles are absent from the abdomen (Toschi 1965, Mon-serrat andDṍaz-Aranda 2012).Nevertheless, all instarshave numerous hooked and Þliform setae on the dor-sumof the abdomen.The threedescribedNothochrysaspecies exhibit some variation in the morphologicalstructures associated with carrying debris (e.g., in thedegree of development of lateral tubercles and in thenumber of large setae associated with thoracic andabdominal tubercles).

Nothochrysinae is generally considered a relictgroup based on the retention of plesiomorphic con-ditions; as yet, adultmorphologyhas providednoclearsynapomorphies to unite the group (e.g., Adams 1967,Adams and Penny 1992, Brooks and Barnard 1990,Brooks 1997).5 However, several larval synapomor-phies (based on specimens from four genera) are nowproposed (Appendix 1; also Monserrat and Dṍaz-Aranda 2012, C.A.T., unpublished data). All of themare on cranial appendages and are not associated in anobvious way with debris-carrying.

A cladogram for thenine genera ofNothochrysinae,based on adult (wing and genital) characters (Fig. 1 inBrooks 1997), depicts one well-supported clade con-

5 As an aside, we propose that an overlooked possible source ofinformative characters (adult stage) is the unique form of reproduc-tive behavior currently documented for two genera in Nothochrysi-nae(Principi 1956,Toschi 1965).Aftermating,Hypochrysa andNotho-chrysa females retain the spermatophore externally, at the tip of theabdomen, for �1 hr; during this time the abdomen pulsates and thecontents of the spermatophore appear to be drawn into the femaleabdomen; subsequently the female eats the shell of the spermato-phore. In addition, during mating, a mass of frothy, granular material(distinct from the spermatophore) is transferred from the male to thefemale: onto the lateral surface of the female abdomen (Nothochrysa)or associated with the spermatophore (Hypochrysa). It would be ofgreat interest to learn whether any of these reproductive features areshared by the other seven genera in the subfamily, and if they haveassociated anatomical modiÞcations that would be pertinent to sys-tematic study.

Table 1. Major morphological features that are typically associated with debris-carrying and naked chrysopid larvae

Structure Debris-carrier Naked larva

BodyShape Globose, humped Fusiform, ßatSetae Numerous, mid-length to long, hooked Few to numerous, short, acute

HeadSetae Textured Smooth

ThoraxLateral tubercles (LTs) Cylindrical, digitiform Absent, hemisphericalSetae on LTs Elongate, textured, hooked, multipronged Mid-length, smooth, acuteSetae on posterior T3 Numerous, robust, textured, hooked Few, slender, smooth, acute (rarely modiÞed)Laterodorsal tubercles (LDTs) Present AbsentDorsal setae Sometimes numerous, often modiÞed Usually sparse, unmodiÞed (acute, blunt)

AbdomenA1: LDTs Present, sometimes large Absent or very smallA1: setae on LDTs Numerous, long, textured, hooked Few, small, smooth, acuteA2ÐA5: LTs Papilliform, digitiform Absent, hemispherical, sphericalA2-A5: setae on LTs Elongate, textured, hooked Mid-length, smooth, acuteA2ÐA5: LDTs Prominent Absent or very smallA2ÐA5: setae on LDTs Numerous, long, textured, hooked Few, small, acuteSubmedian setae (SMS) Midlength to long (stout), often textured,

hooked, or modiÞedShort to midlength (slender), smooth, acute or

blunt, rarely modiÞed

Larvae usually do not express all of the individual features associated with either of the life styles; in reality, there is a full range of variationbetween debris-carrying and nakedness.


taining Þve of the genera; relationships among theremaining four genera (including Kimochrysa) areunsupported. Our recent studies of Kimochrysa larvaeindicate that this genus is very closely related to Hy-pochrysa, which falls within the supported clade(C.A.T., unpublished data). Given that the ancestrallarval condition of Nothochrysinae is unknown, thiscladogram is consistent with two alternative hypoth-eses regarding the evolution of debris-carrying in thesubfamily. The most parsimonious assumes nakednessas the basal state, in which case debris-carrying wouldhave evolved at least once (Nothochrysa). However, ifdebris-carrying were ancestral, then nakedness wouldhave evolved independently two times (Dictyochrysaand Hypochrysa � Kimochrysa).

Subfamily Apochrysinae

Although there are six valid genera in the Apochry-sinae (Winterton and Brooks 2002), larvae are knownfrom only one, Apochrysa (Appendix 1; Catalog, SuppTable 1 [online only]). The third instar of Apochrysamatsumurae (Okamoto) was described and reportedto be a debris-carrier (Tsukaguchi 1995). Photos (oneby P. Duelli in Aspöck and Aspöck [2007], and an-other, Fig. 2A here) show Apochrysa third instars car-rying ßocculence from sternorrhynchan prey.

The anatomy of the described Apochrysa larva istypical of debris-carriers or occasional debris-carriers.For example, the abdomen is moderately thickened,and it bearswell-developed lateral tubercles and smalllaterodorsal tubercles (Tsukaguchi 1995). The setaearenot especiallymodiÞed for debris-carrying, but arenumerous, with some weakly hooked and many rela-tively long.

The monophyly of Apochrysinae seems well-sup-ported by the adult morphology (Brooks and Barnard1990). However, the few known larvae of the grouphave not been shown to express features that distin-guish them from debris-carrying larvae in the subfam-ily Chrysopinae. It is possible that larval support forApochrysinae is absent because very few larval char-acters andvery few taxa fromApochrysinaehavebeenexamined. Thus, additional studies are needed.

In this regard, we noted in TsukaguchiÕs illustrationof the A. matsumurae third instar a character that isworthy of exploration, that is, the mesothoracic andmetathoracic lateral tubercles appear longer thanthose of the prothorax. This condition has not beenreported for other chrysopid taxa.

Subfamily Chrysopinae

Currently, the large and diverse subfamily Chrysopi-nae is distinguished by two relatively strong apomor-phies from adult morphology (antennal setae in fourrings, alar tympanal organ enlarged; Brooks and Bar-nard 1990, Brooks 1997). Also, a unique suite of larvalfeatures distinguishes the subfamily; however, the in-dividual features that comprise the suite are sharedwith either Apochrysinae or Nothochrysinae, and the

taxa comprising Chrysopinae, as yet, have no con-Þrmed larval synapomorphies (Appendix 1).

The three phylogenies differ with regard to themonophyly and the relationships of the tribes in-cluded within Chrysopinae. First, with regard tomonophyly, Brooks (1997) provides some morpho-logical evidence for the monophyly of Belonoptery-gini, Leucochrysini, and Ankylopterygini, but thesingle apomorphy proposed for Chrysopini is incon-sistent among the included taxa. The results of Win-terton and Freitas (2006) provide evidence for themonophyly of each of the four tribes; those of Har-uyama et al. (2008) support monophyly for Ankylop-terygini and for Belonopterygini, but not for Chryso-pini (Leucochrysini was not tested).

Second, with regard to the relationships among thefour tribes, the morphological study places a monophy-letic grouping of Belonopterygini � Leucochrysini in asister relationship with a monophyletic grouping of An-kylopterygini�Chrysopini; however, the featurespro-posed as apomorphies in support of these groupingsare highly variable. One of the molecular studies(Winterton and Freitas 2006) places a monophyleticChrysopini as sister to the monophyletic grouping of(Belonopterygini � [Leucochrysini � Ankyloptery-gini]). In the secondmolecular study(Haruyamaet al.2008), the Chrysopini is paraphyletic, and its relation-ship with the other tribes is unclear. The chrysopinegenera that were examined fell into seven clades, allwith various degrees of support.

Larval morphology provides support (synapomor-phic character states) for the monophyly of two of thefour tribes (Belonopterygini andLeucochrysini). Lar-vae in the other two tribes (Ankylopterygini andChrysopini) are in need of further comparative study.

Tribe Ankylopterygini. Larvae from three of theÞve known ankylopterygine genera (Ankylopteryx,Parankylopteryx, Semachrysa) are classiÞed as debris-carriers (Appendix 1; Catalog, Supp Table 1 [onlineonly]). Detailed larval descriptions are available fortwoof the threegenera (Hölzel et al. 1990,Tsukaguchi1995). These larvae share a suite of anatomical fea-tures related to debris-carrying: their bodies are mod-erately humped; they have long, cylindrical lateraltubercles on the thorax and well-developed, hemi-spherical lateral tubercles on the abdomen; and thedorsal setaeonabdominal segmentsA2 throughA6arehooked. In contrast, laterodorsal tubercles appear tobe absent from abdominal segments A1 through A5,and setaeon the lateral tubercles and thoracicnota arerelatively sparse and unhooked. The two genera differin the number of hooked setae, as well as their patternof distribution.

Based on the small number of described larvaewithin the two genera, ankylopterygine larval debris-carrying modiÞcations do not appear to be highlymodiÞed. At this time, one larval feature has beenproposed tentatively as unique for Ankylopterygini(Tsukaguchi 1995; also see Appendix 1). More studiesof the larvae are needed.

Tribe Belonopterygini. The ant-associated life his-tory of one the belonopterygine species, Italochrysa


italica (Rossi), is well documented (Principi 1943,1946). The few additional observations and unpub-lished notes that are available for larvae of thischrysopid tribe also mention an association with antcolonies (e.g., Weber 1942; E. G. MacLeod, personalcommunication; C.A.T. and S. L. Winterton, unpub-lished data).

The morphology and behavior of I. italica larvaeindicate a high degree of modiÞcation for debris-car-rying (Principi 1943, 1946). For example, the body ofthe third instar I. italica is extremely humped andcompact, the thoracic and abdominal tubercles arestout and highly setose, and the dorsal surface of thelarva, including the head, is covered with rough-sur-faced setae and numerous small secondary setae.Moreover, the legs are short and the empodia andclaws are relatively large; such features may enablelarvae to retain purchase on the substrate and avoidattack directed at the ventral surface. Larvae alsoappear highly adept at maneuvering their bodies totake advantage of their dense protective covering.Also, the mode of moving debris along the surface ofthe body (through peristaltic contractions of the ab-dominal segments [as described by Principi], as op-posed to lifting the head and abdomen and using themandibles [as occurs in larvae of otherChrysopinae])reduces exposure of the vulnerable ventral surface.

The Þrst instars of two belonopterygine genera(Italochrysa and Calochrysa) are unique amongchrysopids in that they express some traits that typifylater-stage larvae, that is, second or third instars. Forexample, unlike the 2Ð3Ð3 pattern of setae (LS) on thepro-, meso-, and metathoracic lateral tubercles that istypical of Þrst instars in all other chrysopid taxa, theÞrst instars of these two genera have from six to nineLS on each of the thoracic tubercles (Fig. 4A; New1983, 1986; Dṍaz-Aranda and Monserrat 1995; Mon-serrat and Dṍaz-Aranda 2012). The Þrst instars of an-otherbelonopteryginegenus (Vieira) retain the2Ð3Ð3LS pattern on the thoracic lateral tubercles, but the

prothorax also has an anterior row of seven pairs ofrobust, hooked setae, and the mesonotum and met-anotum each have a pair of laterodorsal tuberclesbearing approximately nine robust, hooked setae (Fig.4B; Tauber et al. 2006). This pattern of thoracic seta-tion and notal tubercles is unique among chrysopidtaxa, and its presence may indicate that the extraor-dinarily large number of setae on the thoracic lateraltubercles of Italochrysa and Calochrysa Þrst instars arederived from secondary setae that became enlargedand were shifted from the notum to the tips of thelateral tubercles.

Italochrysa and Calochrysa Þrst instars also have anexceptionally large number of dorsal abdominal setae,relative to Þrst instars in other chrysopid taxa (New1983, 1986; Dṍaz-Aranda and Monserrat 1995; Mon-serrat and Dṍaz-Aranda 2012). Again, the characterstate resembles that of second and third instars inother taxa. It appears that, inÞrst instar Italochrysa andCalochrysa, natural selection for defensive modiÞca-tions against antsmay have resulted in the acceleratedexpression of features that typically occur in laterinstars. This pattern of accelerated development is inkeeping with the relatively large eggs reported forItalochrysa and Calochrysa (Principi 1946; New 1983,1986; Monserrat and Dṍaz-Aranda 2012; their eggshave volumes between 3 and 10 times greater thanthose of Chrysopa or Chrysoperla [see Tauber et al.1991]).

The above shows that belonopterygine Þrst instarsexpress an extraordinary range of intergeneric varia-tion in larval structures related to debris-carrying.WeberÕs (1942) intriguing note on the New Worldbelonopterygine genus Nacarina indicates that thevariation may be even broader. He reported Þnding agroup of chrysopid larvae that resemble ant brood, inassociation with Camponotus ants. He reared one ofthe larvae to the adult stage and identiÞed it as Nadiva(� Nacarina). Based on his Þndings, we suggest thatnaked Nacarina larvae, like the debris-carrying larvae

Fig. 4. Morphological variation in thoracic structure of belonopterygine Þrst instars. (A) Italochrysa stigmatica (Rambur)(modiÞed from Fig. 24 in Monserrat and Dṍaz-Aranda 2012); (B) Vieira elegans (Guérin-Méneville): upper, prothorax; lower,mesothorax (modiÞed fromFigs. 6 and 7 inTauber et al. 2006). (Abbreviations: LDT, laterodorsal tubercle; LS, setae on lateraltubercles; LT, lateral tubercle; row, row of secondary prothoracic setae).


of other belonopterygines, may use ant brood as theirprey. Perhaps they have a mechanism other than cov-ering themselves with debris, e.g., defensive chemi-cals, forprotection frombrood-tendingants (seeDett-ner and Liepert 1994). ConÞrmation of WeberÕs 1942report would be of great value to understanding theevolution of debris-carrying in Belonopterygini.

At this time, the monophyly of tribe Belonoptery-gini is well-supported by synapomorphies (adult:Brooks and Barnard 1990; larval: Appendix 1), but thetaxonomic diversity of larvae that have been studied isrelatively small.

Tribe Chrysopini. Not surprisingly, this tribe, thelargest and most diverse within Chrysopinae, has abroad range of variation in larval debris-carrying hab-its and in associated morphological modiÞcations. Ofthe 27 genera of Chrysopini whose debris-carryingstatus are known, 17 are classiÞed as debris-carriers,whereas naked larvae or occasional debris-carriers arereported for 10 genera (Appendix 1; Catalog, SuppTable 1 [online only]); larvae are unknown from theother nine chrysopine genera.

Chrysopini With Naked Larvae. Larvae in the gen-era Brinckochrysa, Chrysoperla, Peyerimhoffina, Atlan-tochrysa, andNineta are reported consistently as beingnaked or mostly naked; the rarely seen larvae of Nip-ponochrysa, have also been reported not to carry anydebris (Catalog, Supp Table 1 [online only]).Chrysoperla species sometimes inhabit the ßowers ofcomposites (Asteraceae), where they become “liber-ally sprinkled with pollen” (Killington 1928). Speciesof Chrysopa, Plesiochrysa, Anomalochrysa, and Me-leoma also are considered to have naked larvae, butwithin these genera the degree of nakedness variesamong species and instars.

Naked larvae usually have bodies that are narrowand relatively ßat, occasionally moderately thickenedlateral tubercles are absent (Brinckochrysa), small(Anomalochrysa, Chrysoperla, Peyerimhoffina, and At-lantochrysa), hemispherical (Chrysopa, Plesiochrysa,and Meleoma), or well-developed and almost cylin-drical (Nipponochrysa); see Figs. 1 and 5A. Abdominallaterodorsal tubercles may be present (Chrysopa,Plesiochrysa, Meleoma, and Peyerimhoffina), reduced(Chrysoperla and Nineta), or absent (Brinckochrysaand Nipponochrysa). Dorsal abdominal setae (subme-dian setae, SMS, of Tsukaguchi 1995) are smooth andusually relatively short, straight, and acutely tipped(sparse: Chrysoperla, Nineta; dense: Nipponochrysa).However, there are notable exceptions to the above:the dorsal abdominal setae on Peyerimhoffina are clav-ate; those on Brinckochrysa have star-shaped tips; inChrysopa they are long and sometimes broadly curvedand hooked; while those of Meleoma are straight andhooked.

Surprisingly, the third instar of one very unusualspecies of Chrysoperla has hooked dorsal SMS on ab-dominal segments A1 to A5 (Tsukaguchi 1995:Chrysoperla furcifera [Okamoto]). It is noteworthythat this speciesÕ unusual larval features are consistentwith its exclusion from the Chrysoperla clade in one ofthe molecular phylogenies (Haruyama et al. 2008).

Also, unusual larval features may support that studyÕsexclusion of Chrysoperla suzukii (Okamoto) fromChrysoperla. The larva in TsukaguchiÕs (1995) draw-ings (see his Fig. 93i and j) appears more robust andsetose than those of other Chrysoperla species.

Most species in the large genus Chrysopa are re-ported to have naked larvae or larvae (usually Þrstinstars) that occasionally carry scattered pieces ofdebris (e.g., Smith 1922, 1926; Killington 1936, 1937;Tsukaguchi 1978, 1995; Gepp 1983, 1988); however,there is considerable variation and one notable ex-ception. The larvae of Chrysopa slossonae Banks arehighly prey-speciÞc; they feed on a single species ofaphids and they have a conspicuous propensity tocover themselves with the wax that their prey secrete.In nature, this covering serves as a defense againstseveral species of ants that tend the aphids (Eisner etal. 1978, Tauber and Tauber 1987).

In the laboratory, the debris-carrying C. slossonaewas shown tobe capable of interbreedingwith its nearrelative, Chrysopa quadripunctata Burmeister, a spe-cies that has a wider prey range and larvae that areonly occasional debris-carriers (Tauber and Tauber1987). Comparative studies of the two species, theirhybrids, as well as geographic populations of C. quad-ripunctata, revealed genetically based morphologicalandbehavioralmodiÞcations for larval debris-carrying(Tauberet al. 1993, 1995a,b;Milbrathet al. 1994).Theyalso showed that environmental factors can stronglyinßuence the degree of debris-carrying behavior andthat under some conditions larvae of the occasionaldebris-carrier express a strong tendency to apply de-bris on their dorsa.

Wehave collected larvae of fourMeleoma species inthe Þeld: Meleoma emuncta (Fitch), Meleoma dolich-arthra (Navás),Meleoma signoretiiFitch, andMeleomakennethi Tauber, and all were either naked or carriedonly a few pieces of debris (Toschi 1965, M.J.T. andC.A.T., unpublished data). However, as in Chrysopa,Meleoma larvae have well-developed lateral tubercleson the thorax. Moreover, their dorsal abdominal setaeare numerous, with small hooks or knobs at the tips,and apparently are held somewhat ßat on the abdom-inal surface (Tauber 1969, C.A.T. and T. de León,unpublished data). We suggest that some Meleomalarvae may be debris-carriers or that the dorsal ab-dominal setae that usually subserve debris-carrying inother chrysopid larvae may have other, as yet unrec-ognized, functions in Meleoma.

Chrysopini With Debris-Carrying Larvae. Amongthe debris-carrying genera of Chrysopini, all larvaehave thickenedor globose bodies andwell-developed,but not especially elongate, thoracic and abdominallateral tubercles (Figs. 2D, 2E, 3A, 3B, 5B, and 5C).Typically, the metathorax has a row of four or moreprominent,modiÞed setae, and eachof the anterior sixto seven segments of the abdomen (A1 through A7)has two to three rows of dorsal setae that are hookedor otherwise modiÞed for holding material.

Despite the above similarities among the debris-carrying genera of Chrysopini, striking differencesexist. For example, the shape of the larval body (spe-


ciÞcally the degree of thickening and the height of themesothorax and metathorax, as well as the size ofthe Þrst abdominal segment) varies notably amongdebris-carrying genera; Chrysopodes, Yumachrysa,and Ceraeochrysa illustrate the variation well. For ex-ample, the posterior section of the Chrysopodes met-anotum rises steeply above the anterior section, andthe row of robust metanotal setae extend from largechalazae along the top of the raised posterior section(Tauber 2003, Silva et al. 2013). Each of these largechalazae has an oval marking that extends onto theintegument in front of the seta. In Ceraeochrysa, themetathorax rises gradually above the mesothorax, butthe Þrst abdominal segment is constricted and the

second segment iswide, a situation that gives the bodysomewhat of an hourglass appearance (Tauber et al.2000, Tauber and de León 2001). In Yumachrysa, theÞrst abdominal segment is enlarged and bears sub-stantial setose laterodorsal tubercles (Tauber 1975).

The broad range of variation in larval chaetotaxyamong the debris-carrying genera of Chrysopini isexempliÞed by the following: 1) Setae on the lateral orlaterodorsal tubercles often are smooth or slightlygranular (Cunctochrysa, Pseudomallada, Rexa, and Ti-tanochrysa), but they may be thorny in Chrysopodes,Ceraeochrysa, and Yumachrysa. 2) The posterior re-gion of the metanotum usually has a transverse rowof long setae, but the mesonotum varies in having

Fig. 5. Morphological variation among Þrst instar Chrysopidae. (A) Typical naked larva: Chrysoperla externa (Hagen)(from Tauber 1974), (B) Typical debris-carrier: Yumachrysa apache (Banks) (from Tauber 1975), (C) Typical chrysopineprothorax: Ceraeochrysa (debris-carrier) (modiÞed from Fig. 10 in Tauber et al. 2000), and (D) Typical leucochrysineprothorax:Leucochrysa (Nodita) (debris-carrier) (modiÞed fromFig. 22 inMantoanelli et al. 2011). (Abbreviation: LT, lateraltubercle).


such a row (present: Cunctochrysa, Chrysopidia, andCeraeochrysa; absent: Chrysopodes, Mallada, Aperto-chrysa, Pseudomallada,Rexa, Yumachrysa, and Suarius;Fig. 5B and C). 3) Dorsal setae on the abdomen maybe long (Yumachrysa and Suarius), intermediatelength (Ceraeochrysa, Chrysopodes, Pseudomallada,and Cunctochrysa), or short (Titanochrysa). Also, theymay range from smooth, hooked, and without aßattened tip (Ceraeochrysa and Chrysopidia), tosmooth,with a long, laterally ßattened terminal hook(Chrysopodes), or broadly hooked with a spatulate orslightly spatulate tip (Cunctochrysa and Pseudomal-lada). In Titanochrysa, the posterior setae on the dor-sum of each segment, i.e., A2 through A4, are short,stout, and hooked. 4) Finally, the presence or absenceof small dorsal tubercles on various abdominal seg-ments and the number and characteristics of the setaethey bear varies considerably among genera. The lar-vae of most debris-carrying genera have the tuberclesonly on segments A6 and A7. Indeed, among the de-bris-carrying genera, Pseudomallada and Yumachrysaare exceptional in having the tubercles on each ofabdominal segments A5 through A7; Yumachrysa alsohas an exceptional, large dorsal tubercle on the Þrstabdominal segment.

Phylogeny of Debris-Carrying in the Chrysopini. Be-low, we discuss the diversiÞcation of larval featuresamong the chrysopine genera in relation to the twomolecular-based chrysopid phylogenies.

The Þrst study (Winterton and Freitas 2006) in-cluded only 9 of the 36 genera in Chrysopini; all 9,except 1 (Mallada), were found to be monophyletic.Moreover, relationships among some of the generaseemed reasonably well-resolved. Chrysopodes (a de-bris-carrier) was recovered as the sister genus to theother eight genera, of which Ceraeochrysa (a debris-carrier) is sister to Chrysoperla (naked) and the Ple-siochrysa–Chrysopa clade (naked) is sister to theApertochrysa–Mallada–Glenochrysa–Pseudomalladaclade (all debris-carriers: Catalog, Supp Table 1 [on-line only]).

When larval morphology is viewed within the con-text of the above phylogeny, we Þnd agreement withsome of the proposed relationships. First, the sisterrelationship between Chrysopodes and the eight othergenera is in concordance with larval morphology; in-deed, Chrysopodes larvae express several larval fea-tures that are not reported for any other chrysopidgenus whose larvae are known. These distinctive fea-tures include the structure of the metathorax and therobust metathoracic setae arising from modiÞed cha-lazae (Tauber 2003, Silva et al. 2013, Catalog, SuppTable 1 [online only]). It is noteworthy that the phy-logeny based on adult morphology (Brooks 1997)identiÞed Chrysopodes as having “unknown afÞnities,”a status that also may attest to a distant relationshipwith other chrysopine genera.

Second, larval characters strongly support the dis-tinction and the close relationship between Chrysopaand Plesiochrysa (shown by the two molecular studiesand the adult morphological data). Larvae of the twogenera share many morphological features (see Tsu-

kaguchi 1978, 1995; Tauber et al. 2001). (Note: Therelationship between Chrysopa and Plesiochrysa hasbeen fraught with misinterpretation. 1) Wintertonand Freitas [2006] apparently misread the Þndings ofTauber et al. [2001], instead of differences existingbetween the molecular and larval-based studies, bothsets of data concur that the twogenera aredistinct, butclosely related. 2) Also, Monserrat and Dṍaz-Aranda(2012) erroneously dismissed the relationship be-tween Chrysopa and Plesiochrysa. Their conclusionwas based on a larva that Adams [1959] described andwhose identity as Plesiochrysa had been shown to bein error [Tauber et al. 2001]).

Third, the Apertochrysa–Mallada–Glenochrysa–Pseudomallada clade in the phylogeny of Wintertonand Freitas (2006) is not contradicted and is tenta-tively supported by shared larval features, many ofthem associated with debris-carrying. The larvae of allfour genera are debris-carriers, and the described lar-vae of three of these genera (Tsukaguchi 1995, Mon-serrat and Dṍaz 2012, Catalog, Supp Table 1 [onlineonly]) express a relatively broad range of overlappingfeatures (Apertochrysa, Mallada, and Pseudomallada,especially the latter two). Larvae of Glenochrysa areundescribed.

Fourth, the proposed sister relationship betweenChrysoperla andCeraeochrysaÞnds little or no supportin current larval data. Similarly, larval support forBrooksÕ grouping of Chrysopa � Plesiochrysa withCeraeochrysa is not forthcoming. These hypothesizedrelationships warrant further examination, includingdetailed comparisons of the larval anatomy.

To summarize our thoughts on the evolution ofdebris-carrying in Chrysopini in relation to the Win-terton and Freitas (2006) phylogeny: 1) The mostparsimonious interpretation is that debris-carryingrepresents the basal state for the tribe. 2) Nakednessapparently evolved independently at least twiceamong the taxa tested: once in Chrysoperla and oncein theChrysopa–Plesiochrysaclade. Inevaluating theseconclusions, it should be remembered that only one-fourth of the chrysopine genera are included in thephylogeny.

In comparison, the second molecular study (Har-uyama et al. 2008) examined a slightly broader rangeof taxa in the tribe Chrysopini (approximately one-third of the genera), and it recovered seven reason-ably well-supported clades. Relationships among theclades are not well-resolved, but some hypothesizedgeneric groupings are noteworthy. First, the closerelationshipbetweenChrysopa andPlesiochrysa is cor-roborated, and as shown above, this relationship isstrongly supported by morphological (larval andadult) data. Second, the study provides additional ev-idence for a close association between the debris-carrying genera Pseudomallada and Apertochrysa, asdo both the earlier molecular study and larval mor-phological data (see above). Third, a relationship be-tween Chrysoperla and Peyerimhoffina was recovered;it, too, is supported with larval data (Monserrat andDṍaz-Aranda 2012) and by adult morphology (Brooks1997). Finally, although adult morphological data sup-


port a close relationship between the above two gen-era (Chrysoperla and Peyerimhoffina) and Mallada(Brooks 1997), supporting larval data are not obviousat this time; this proposed relationship deserves fur-ther study.

To summarize the evolution of debris-carrying inChrysopini in relation to the phylogeny of Haruyamaet al. (2008), two alternate patterns are apparent: 1) inthe most parsimonious interpretation (i.e., when de-bris-carrying is the presumed basal state of the tribe)nakedness evolved independently within the testedchrysopine genera approximately six times, with onereversal (Clades 1, 3, 4 [and one reversal], 5 [twoevents], and 6). 2) In contrast, if nakedness were thebasal state, then debris-carrying would have evolvedindependently approximately eight times (Clades 2, 3,4 [two events], 5 [two events], 7, and Borniochrysa).

TribeLeucochrysini.Thedebris-carryinghabits (innature) of only a few species in one leucochrysinegenus have been reported, but those that are knownexhibit a diverse and thought-provoking range of de-bris selectivity (Fig. 3C and D). One species, Leuco-chrsya (Nodita) pavida (Hagen), largely carries li-chens (Skorepa and Sharp 1971), and careful analysisof the dorsal packets shows that the larvae are highlydiscriminating in the lichen species and the speciÞcparts of the lichens they use for covering themselves(Wilson and Methven 1997). The larvae may serve asimportant dispersal agents for the lichens, and as aresult the disparate organisms may have a mutualisticrelationship (Slokum and Lawrey 1976).

Another species, Leucochrysa (Leucochrysa) insu-laris(Walker), carries very small terrestrial snail shells(sometimes living snails!) in its packet (Jones 1929,1941). Here again, the larvae appear to be narrowlyselective as to the types and sizes of snails they choose.Other Leucochrysa species are reported to carry ster-norrhynchan ßocculence, but these reports are un-conÞrmed.

Larvae have been described from most leuco-chrysine genera (Leucochrysa, Gonzaga, Santocellus,and Berchmansus; Appendix 1; Catalog, Supp Table 1[online only]), and they all express strikingly elon-gate, digitiform lateral tubercles on the thoracic seg-ments (Figs. 3D and 5D; Appendix 1). Because ofthese elongate tubercles and other features (e.g.,humped body shape, papilliform abdominal lateraltubercles, and hooked abdominal SMS) leuco-chrysines generally are assumed to be debris-carriers(Mantoanelli et al. 2011).

However, recently Berchmansus larvae were shownto have a fusiform and somewhat ßat body, elongatelateral tubercles on the abdominal and thoracic seg-ments, and spatulate (not hooked) thoracic and ab-dominal setae (Tauber andTauber 2013).Given theseattributes, it is now apparent that leucochrysinelarval anatomy has a much larger range of interge-neric variation than previously assumed. Moreover,the Berchmansus larval features lead us to questionwhether they carry debris. It is likely that in naturethe somewhat ßat but setose larvae of Berchmansus

either are naked or carry only a sparse packet ofmaterial.

Regrettably, the phylogenetic relationships amongleucochrysine genera have not been examined witheither morphological or molecular approaches. Nev-ertheless, based on larval characters included in cur-rent descriptions, it appears that Santocellus andBerchmansus are characterized by strong larval apo-morphies (Santocellus: relatively short mandibles anduniquely shaped labial palpomeres; Berchmansus:elongate abdominal lateral tubercles and knobbeddorsal abdominal setae; Tauber et al. 2008b, TauberandTauber 2013). In addition,Leucochrysa (both sub-genera) and Gonzaga appear to be more closely re-lated to one another than to either of the two abovegenera (Tauber et al. 2008a,b); indeed, consistent fea-tures differentiating the two genera have not beenidentiÞed.

Ancient Chrysopid Relatives (Superfamily:Chrysopoidea)

Recently, a very large, debris-carrying chrysopoidlarva, Hallucinochrysa, was discovered in early Creta-ceous amber from Spain; the presence of fossilized“debris” (fern trichomes) entwined in its elongatelateral setaeoffersevidence that the larvawasadebris-carrier (Pérez-de la Fuente et al. 2012). This fossildemonstrates that debris-carrying behavior, with itsassociated morphological modiÞcations, has deeproots in the chrysopoid lineage. Furthermore, thelarva has a remarkable set of structures that are unlikeany known among the extant chrysopids. These in-clude extremely long, slender lateral and sublateraltubercles on each thoracic segment, equally long andslender lateral tubercles on the abdomen, anduniquely shaped setae (LS) extending from the lat-eral tubercles. The unusual anatomy of the Hallu-cinochrysa larva underscores that the ancient trait ofchrysopid debris-carrying has evolved through mul-tiple pathways.

Amoremodern fossilizedchrysopid larva (Mioceneamber, second or third instar, Dominican Republic)also presents features that the discoverers attributedto debris-carrying (Engel and Grimaldi 2007). Theseinclude well-developed lateral tubercles with elon-gate setae and a relatively ßat body with a laterallyupturned dorsal surface. Together, these structureswere described as comprising a “corbicula,” or basket,for carrying debris. However, we, and others (seePérez-de la Fuente et al. 2012), propose that thisspecimen could have been either naked or a debris-carrier. First, unlike the Hallucinochrysa fossil above(which was embedded in amber with debris entwinedin its setae), this larva was not reported to be associ-ated with debris. Second, the anatomical characteris-tics of the larva also occur in naked larvae, especiallythoseofChrysopa; they share similarlywell-developedlateral tubercles, relatively long lateral setae, and ßat-tened bodies. Also, it is noteworthy that the Miocenelarva has well-developed lateral tubercles on abdom-inal segmentA1; among extant species, only the larvae


of a few Chrysopa species are known to have lateraltubercles on this segment (Principi 1940, Tsukaguchi1978).

Apart from the question of the Miocene larvaÕs de-bris-carrying status, we noted that the description andillustration of the larva (Engel and Grimaldi 2007)include two unusual traits: a pair of small setose tu-bercles on the sides of the prothorax behind or belowthe large lateral tubercles and the absence of dorsalsetae on the prothorax, mesothorax, and abdominalsegments posterior toA4. These features attracted ourattention because they are not known to occur in anycurrently described extant chrysopids.

General Taxon-Based Patterns in ChrysopidDebris-Carrying

Because of sparse information on the larval be-havior and morphology of many chrysopid taxa, weemphasize that most statements (including our ownbelow) regarding the origin and evolutionary pat-terns in chrysopid debris-carrying are, at best, sub-ject to conÞrmation. Given that caveat, we summa-rize our thoughts on the topic below with the goalof stimulating discussion and additional research.

Based on existing data fromextant and fossil chryso-poids, it is now known that debris-carrying and itsassociated larval morphology express a relatively highdegree of evolutionary plasticity. The data also indi-cate that this plasticity may be subject to signiÞcantconstraints. SpeciÞcally, we conclude that:

1. Debris-carrying is an ancient characteristic that hasarisen or been lost repeatedly during the evolutionof the Chrysopoidea.

2. The morphological modiÞcations that underlie thediverse forms of chrysopoid (and chrysopid) de-bris-carrying are numerous and diverse.

3. Among the extant chrysopids, the morphologicaladaptations that support debris-carrying involve arelatively large, well-deÞned group of anatomicalstructures, each of which apparently can be en-hanced, reduced, or modiÞed independently overevolutionary time (Table 1). Prominent amongthese structures are the shape of the body; thethoracic and abdominal lateral tubercles and lat-erodorsal tubercles, as well as the setae they bear;the dorsal setae of the thorax; and the transverserows of submedian abdominal setae. The legs, ce-phalic appendages, and cephalic setae may also beinvolved. Debris-carrying larvae often show exten-sive modiÞcation of some but not all of these struc-tures; moreover, the alteration of one structuredoes not always appear to be dependent on thealteration of others.

4. Given the evolutionary ßexibility implied in thestatements above, it also appears that the morpho-logical modiÞcations for chrysopid debris-carryingevolve within a deÞned and constrained frame-work. Evidence for this restriction stems from theconsistent pattern in which basic larval structuresare retained in all genera but are modiÞed to sub-

serve debris-carrying. For example, chrysopid lar-vae have one pair of relatively large lateral tuber-cles on each thoracic segment and on eachabdominal segmentA2 throughA7. The size, shape,orientation, and setation of these tubercles varyamong genera. Indeed, in a few cases the basicstructures (e.g., the lateral tubercles themselves)appear to have been lost, but in no case have ad-ditional ones been added. Similar statements couldbe made for many other larval features involved indebris-carrying (e.g., thoracic and abdominal lat-erodorsal tubercles, thoracic and abdominal setae,and thoracic folding). Thus, it seems that each ge-nus evolved a unique set of modiÞcations to a fun-damental, shared, ground plan of anatomical struc-tures, and in each genus these modiÞcationssupport a speciÞc habit of debris-carrying, occa-sional debris-carrying, or nakedness.

We have found only one exception to the fourthpoint above, and that is the case of two belonoptery-gine genera Italochrysa and Calochrysa in which thetypical chrysopid pattern of Þrst-instar setation is al-tered (discussed above). However, this deviationcould be explained by the accelerated developmentand displacement of secondary thoracic setae, and notthe de novo evolution of LS (Tauber et al. 2006).

Integration of Debris-Carrying, Larval Morphology,and Chrysopid Systematics

Several authors have expressed some misgivingsabout using larval characters related to debris-carry-ing in phylogenetic studies of Chrysopidae, becausethey suspected that the repeated evolution of larvaldebris-carrying may result in excessive homoplasy(Tauber 1975, Monserrat and Dṍaz-Aranda 2012,Pérez-de la Fuente et al. 2012). However, in view ofthe comparative morphological studies of chrysopidlarvae during the last two decades, we believe thatthese reservations now need reevaluation.

The following recentquote fromPérez-de laFuenteet al. (2012): pg. 2, Supporting Information) providesa useful focus for our discussion of the issue:

“It is interesting to note that the current high-levelclassiÞcation within Chrysopidae is almost solelybased on adult characters (12) [Brooks and Barnard1990], because parallelisms and convergence in larvaltraits are rampant across the family (28) [Tauber1975].However, larval anatomy andmorphology haveproven tobeofgreat importance for recovering lower-level relationships within the family (29, 30) [Tauberet al. 2008a,b].” (References in square brackets arethose that the authors cited by number).

Below, we review whether this type of well-inten-tioned cautionary statement could be misinterpretedand lead to the unfortunate marginalization of larvalmorphology as a source of characters for phylogeneticstudies. We evaluate three assertions in the quotation:

1. Indeed, the current higher-level classiÞcation ofChrysopidae is based almost solely on adult char-


acters (Brooks and Barnard 1990, Brooks 1997).However, it is now known that larval characters alsoare systematically very informative at suprage-neric levels (see Appendix 1). Nothochrysinae,Belonopterygini, Leucochrysini, and Ankylop-terygini provide signiÞcant examples (Dṍaz-Aranda and Monserrat 1995, Tsukaguchi 1995,Monserrat and Dṍaz-Aranda 2012, Tauber andTauber 2013). Thus, the informative value of lar-val characters should also be emphasized in state-ments that evaluate the sources of characters forphylogenetic study.

2. As with any other characters, larval traits can besubject to parallelism and convergence, especiallyif they are not chosen and deÞned carefully. Forexample, if larvae were scored on the basis of “de-bris-carrying present or absent,” “humped abdo-men present or absent,” “well-developed thoracictubercles present or absent,” “dense, long, setaepresent or absent,” the characters undoubtedlywould introduce homoplasy into the data set(Tauber 1975). However, speciÞc larval morpholog-ical features that support debris-carrying (if preciselydeÞned and presented with alternative characterstates) would be no more subject to parallelism orconvergence than any other type of well-delineatedfeature. Examples include “posterior section ofmetathorax raised above anterior section abruptly orgradually,” “laterodorsal tubercles on abdominal seg-ment A1 (or A5) present or absent,” “clavate SMSpresent or absent.” Thus, the careful choice and def-initionof characters is crucial, but this noteof cautionapplies equally to both adult and larval characters(e.g., see Miller and Wenzel 1995).

3. Larval morphology has proven to be highly signif-icant for recovering lower-level (i.e., generic) re-lationships within the family; in fact, all currentlyrecognized chrysopid genera (whose larvae areknown) are distinguished by a unique set of larvalfeatures. We suggest that the acquisition, modiÞ-cation, and loss of debris-carrying may have leftvisible and useful markers along a retrievablecourse of evolutionary change. Because of theirdiversity andgeneric-level consistency, these larvalmorphological features may play a major role inunraveling the tangled branches that currentlycomprise the chrysopid phylogenetic tree.

Recommendations

While preparing this article, we noted opportunitieswhere modest but focused efforts could accelerate in-vestigations of chrysopid debris-carrying. One involvesincreasing the larval database. The other focuses on de-veloping comprehensive larval data matrices.

Gathering Larval Data: Citizen-scientists andProfessionals

The lack of specimens with useful data is one of themost signiÞcant reasons why immature insects aregenerally not included in phylogenetic studies (Meier

and Lim 2009). This problem reßects the situationwith chrysopid larvae. However, given popular Websites, including social media, it is now possible foramateur scientists, photographers, and hobbyists toshare images and information with specialists in aquick and efÞcient manner (see Figs. 1Ð3). Such im-ages, if associatedwith thepreserved specimens, couldprovide valuable information for systematic and otherstudies. Thus, we recommend that professionals en-courage and work with citizen-scientists to retainvoucher specimens for their photographs.6 For exam-ple, such interactions, initiated through social mediaand followed upwith exchange of voucher specimens,led to the description of a new chrysopid species fromThailand (Winterton et al. 2012), and the discovery ofa unique pattern of oviposition in a North Americanchrysopid species (Tauber et al. 2013).

Developing Larval Data Matrices: ChrysopidSystematists

Chrysopid systematists have made considerableprogress in promoting uniformity and clarity in theterminology for larval characters. A similar but morefocused communal effort would help develop an ex-tensive analysis of chrysopid larvae and their inclusionwithin a comprehensive phylogenetic study of thefamily. Using immature stages in phylogenetic studiesrequires that comparisonsbemadeacross homologouslarval characters on equivalent semaphoronts (equiv-alent developmental stages) of all included taxa (seeWheeler 1990, Meier and Lim 2009). Thus, we suggesttwo areas in need of discussion and cooperationamong chrysopid systematists: 1) identiÞcation ofmorphological and behavioral homologies among lar-vae across the full range of chrysopid taxa and 2)development of an extensive taxonÐcharacter datamatrix (with precise, well-described, and well-illus-trated character states) for systematists worldwide touse in evaluating chrysopid larvae in any taxon. Wepropose that the recentcompendiumonthechrysopidlarvae of the Iberian Peninsula (Monserrat and Dṍaz-Aranda 2012) offers a valuable prototype upon whichto build an extensive taxonÐcharacter data matrix.

Acknowledgments

The following were generous in allowing us to use theirÞne photographs: T. Eisner (deceased) and M. L. Eisner(both Cornell University, Ithaca, NY), L. Buss (University ofFlorida, Gainesville, FL), D. C. Dourson (Lincoln MemorialUniversity, Harrogate, TN), G. San Martin y Gomez (Wal-loon Agricultural Research Center, Gembloux, Belgium),S. A. Marshall (University of Guelph, Guelph, ON), R. A.Pantaleoni and C. F. Cesaroni (Istituto per lo Studio degliEcosistemi, Sardinia, Italy), B. P. Valentine (Sussex, Eng-

6 For voucher specimens of chrysopid larvae, we recommend pres-ervation in 95% ethyl alcohol (larvae and packet), unless the packetincludes waxy material. In that case, the packet should be dry-mounted on a point or stored in a gelatin capsule. All images andspecimens should carry speciÞc locality and collection data. Preser-vationof larval specimens in alcohol andpromptmailing to a specialistwill facilitate identiÞcation via DNA procedures.


land), and Y. Hara (Hyogo, Japan). We acknowledge infor-mation on speciÞc aspects of the literature from H. and U.Aspöck (Medizinische Universität Wien, NaturhistorischesMuseum Wien, Austria; early historical records) and C. S.Henry (University of Connecticut; Chrysoperla). We thankQ. D. Wheeler (Syracuse University) for commenting on anearlier draft of the manuscript and A. J. Tauber, the anony-mous reviewers, and the editor for their constructive sug-gestions. We acknowledge J. D. Oswald (Texas A&M Uni-versity), for developing and maintaining the onlineNeuroptera Bibliography (Oswald 2013); it was very helpfulduring this study. Finally, we appreciate the continuing helpof the librarians at the Peter J. Shields Memorial Library(University of California, Davis) and Albert R. Mann Library(Cornell University). Our research received support fromthe National Science Foundation (Grants INT-9817231,DEB-0542373), the National Geographic Society, the U.S.Department of AgricultureÐNational Research Initiative(Competitive Grant 9802447), Regional Project W-3185,Conselho Nacional de Desenvolvimento CientṍÞco e Tec-nológico (CNPq), Cornell University, and Universidade Es-tadual do Norte Fluminense.

References Cited

Adams,P.A. 1959. Neuroptera:MyrmeleontidaeandChrysopi-dae. Insects of Micronesia 8: 13Ð33.

Adams, P. A. 1967. A review of the Mesochrysinae andNothochrysinae (Neuroptera: Chrysopidae). Bull. Mus.Comp. Zool. 135: 215Ð238.

Adams, P. A., and N. D. Penny. 1992. Review of the SouthAmerican genera of Nothochrysinae (Insecta: Neurop-tera: Chrysopidae), pp. 35Ð41. In M. Canard, H. Aspöck,and M. W. Mansell (eds.), Current Research in Neurop-terology. Proceedings of the Fourth International Sym-posium on Neuropterology (24Ð27 June 1991, Bagnères-de-Luchon, Haute-Garonne, France). Privately printed,Toulouse, France.

Anderson, K. L., J. E. Seymour, and R. Rowe. 2003. Inßu-ence of a dorsal trash-package on interactions betweenlarvae of Mallada signata (Schneider) (Neuroptera:Chrysopidae). Aust. J. Entomol. 42: 363Ð366.

Aspöck, U., and H. Aspöck. 2007. Verbliebene veilfalt ver-gangener blüte. Zur evolution, phylogenie und biodiver-sität der Neuropterida (Insecta: Endopterygota). Denisia20: 451Ð516.

Brandt, M., and D. Mahsberg. 2002. Bugs with a backpack:the function of nymphal camoußage in the West Africanassassin bugs Paredocla and Acanthaspis spp. Anim. Be-hav. 63: 277Ð284.

Brooks, S. J. 1997. An overview of the current status ofChrysopidae (Neuroptera) systematics. Deutsche Ento-mologische Zeitschrift, Berlin (N. F.) 44: 267Ð275.

Brooks, S. J., and P. C. Barnard. 1990. The green lacewingsof the world: a generic review (Neuroptera: Chrysopi-dae). Bull. Br. Mus. Nat. Hist. (Entomol.) 59: 117Ð286.

Brown, N. R., and R. C. Clark. 1962. Studies of predators ofthe balsam woolly aphid, Adelges piceae (Ratz.) (Ho-moptera: Adelgidae) X. Field identiÞcation of Laricobiuserichsonii Rosen. (Coleoptera: Derodontidae). Can. En-tomol. 94: 191Ð193.

Bucheli, S., J.-F. Landry, and J. Wenzel. 2002. Larval casearchitecture and implications of host-plant associationsfor North American Coleophora (Lepidoptera; Coleo-phoridae). Cladistics 18: 71Ð93.

Butler, E. 1923. A biology of the British Hemiptera-Het-eroptera. H. F. and G. Witherby, London, United King-dom.

Chaboo, C. S. 2011. Defensive behaviors in leaf beetles:from the unusual to theweird, pp. 59Ð69. In J.M.Vivancoand T. Weir (eds.), Chemical Biology of the Tropics.Signaling and Communication in Plants 8, Springer, Ber-lin, Germany.

Chaboo, C. S., C. G. Brown, and D. J. Funk. 2008. Faecalcase architecture in the gibbosus species group ofNeochlamisus Karren, 1972 (Coleoptera: Chrysomelidae:Cryptocephalinae: Chlamisini). Zool. J. Linn. Soc. 152:315Ð351.

Dı́az-Aranda, L. M., and V. J. Monserrat. 1995. Aphidopha-gous predator diagnosis: key to genera of Europeanchrysopid larvae (Neur.:Chrysopidae).Entomophaga 40:169Ð181.

Dettner, K., and C. Liepert. 1994. Chemical mimicry andcamoußage. Annu. Rev. Entomol. 39: 129Ð154.

Duelli, P., H. Hölzel, and M. W. Mansell. 2010. Habitat andlarvae of the enigmatic genus Kimochrysa Tjeder (Neu-roptera: Chrysopidae) in South Africa, pp. 153Ð158. In D.Devetak, S. Lipovsek, and A. E. Arnett (eds.), Proceed-ings of the Tenth International Symposium on Neurop-terology (22Ð25 June 2008, Piran, Slovenia). University ofMaribor, Maribor, Slovenia.

Eisner, T., K. Hicks, M. Eisner, and D. S. Robson. 1978.“Wolf-in-sheepÕs-clothing” strategy of a predaceous in-sect larva. Science 199: 790Ð794.

Eisner, T., J. E. Carrel, E. van Tassell, E. R. Hoebeke, and M.Eisner. 2002. Construction of a defensive trash packetfrom sycamore leaf trichomes by a chrysopid larva (Neu-roptera: Chrysopidae). Proc. Entomol. Soc. Wash. 104:437Ð446.

Engel, M. S., and D. A. Grimaldi. 2007. The neuropteridfauna of Dominican and Mexican amber (Neuropterida:Megaloptera, Neuroptera). Am. Mus. Novit. 3587: 1Ð58.

Falkovitsh, M. I. 2003. On the classiÞcation of the casebear-ers (Lepidoptera, Coleophoridae), with descriptions ofthe new taxa. Entomol. Rev. 83: 827Ð847.

Ferguson, D. 1985. Geometroidea: Geometridae (in part),pp. 1Ð131. In R. B. Dominick, T. Dominick, D.R. Davis,D.C. Ferguson, J.G. Franclemont, R.W.Hodges, E.G.Mu-nroe, and J.A. Powell (eds.), TheMoths ofAmericaNorthof Mexico, Fasc. 18.1. The Wedge Entomological Re-search Foundation, Washington, DC.

Gepp, J. 1983. Schlüssel zur Freilanddiagnose mitteleu-ropäischer Chrysopidenlarven (Neuroptera, Chrysopi-dae). Mitteilungen des Naturwissenschaftlichen Vereinsfür Steiermark. 113: 101Ð132.

Gepp, J. 1988. Zur Ökologischen differenzierung derpräemaginalen stadien baumbewohnender chrysopidenim alpenraum (Planipennia: Chrysopidae). Sitzungs-berichte der Österreicheschen Akademie der Wissen-schaften, Mathematische-Naturwissenschaftliche Klasse197: 1Ð73.

Gressitt, J. L., G. A. Samuelson, and D. H. Vitt. 1968. Mossgrowing on living Papuan moss-forest weevils. Nature217: 765Ð767.

Haruyama, N., A. Mochizuki, P. Duelli, H. Naka, and M.Nomura. 2008. Green lacewing phylogeny, based onthree nuclear genes (Chrysopidae, Neuroptera). Syst.Entomol. 33: 275Ð288.

Henry, C. S. 1977. The behavior and life histories of twoNorth American ascalaphids. Ann. Entomol. Soc. Am. 70:179Ð195.

Hölzel,H.,M. Stelzl, andP.Ohm. 1990. Chrysopidae (Neu-roptera) aus Senegal und Gambia. I: Ankylopterygini.Neuroptera International 6: 63Ð73.


Jones, D. T. 1929. A snail-collecting aphid-lion larva. A pre-liminary study of external features. Marietta College Res.Publ. (Zool.) 1: 1Ð9.

Jones, D. T. 1941. Further notes on the snail-collectingaphid-lion larva (Neuroptera: Chrysopidae). Entomol.News 52: 39Ð44.

Killington, F. J. 1928. Larvae of Chrysopa vulgaris, Schn.(Order Neuroptera) on ßower-heads. Entomologist 61:234.

Killington,F. J. 1936, 1937. Amonographof theBritishNeu-roptera, vols. I, II. Ray Society, London, United Kingdom.

Leschen, R.A.B., and C. E. Carlton. 1993. Debris cloaking inendomychidae: a new species from Peru (Coleoptera).Zool. J. Linn. Soc. 109: 35Ð51.

MacLeod, E. G. 1960. The immature stages of Boriomyiafidelis (Banks) with taxonomic notes on the afÞnities ofthe genus Boriomyia (Neuroptera: Hemerobiidae).Psyche 67: 26Ð40.

MacLeod,E.G. 1964. Acomparativemorphological studyofthe head capsule and cervix of larval Neuroptera (In-secta). Ph.D. dissertation, Harvard University, Cam-bridge, MA.

Mantoanelli, E., C. A. Tauber, G. S. Albuquerque, and M. J.Tauber. 2011. Larvae of four Leucochrysa (Nodita) spe-cies (Neuroptera: Chrysopidae: Leucochrysini) fromBrazilÕs Atlantic coast. Ann. Entomol. Soc. Am. 104: 1233Ð1259.

McHugh, J. V., and T. G. Kiselyova. 2003. First descriptionsfor larval stages of Eurysphindus (Coleoptera: Cucu-joidea: Sphindidae). The Coleopterists Bull. 57: 17Ð25.

Meier, R., and Lim, G. S. 2009. Conßict, convergent evolu-tion, and the relative importance of immature and adultcharacters in endopterygote phylogenetics. Annu. Rev.Entomol. 54: 85Ð104.

Milbrath, L. R., M. J. Tauber, and C. A. Tauber. 1993. PreyspeciÞcity in Chrysopa: an interspeciÞc comparison oflarval feeding and defensive behavior. Ecology 74: 1384Ð1393.

Milbrath, L. R.,M. J. Tauber, andC.A. Tauber. 1994. Larvalbehavior of predacious sister-species: orientation, molt-ing site, and survival in Chrysopa. Behav. Ecol. Sociobiol.35: 85Ð90.

Miller, J. S., and J. W. Wenzel. 1995. Ecological charactersand phylogeny. Annu. Rev. Entomol. 40: 389Ð415.

Miller, R. B., and L. S. Stange. 2012. The cave mouth ant-lions ofAustralia (Neuroptera:Myrmeleontidae). InsectaMundi 250: 1Ð65.

Monserrat, V. J. 1996. Larval stages of European Nemop-terinae, with systematic considerations on the familyNemopteridae (Insecta, Neuroptera). Deutsche Ento-mologische Zeitschrift 43: 99Ð121.

Monserrat, V. J. 2008. Nuevos datos sobre algunas especiesdeNemopteridaeyCrocidae (Insecta:Neuroptera).Het-eropterus: Revista de Entomologia 8: 1Ð33.

Monserrat, V. J., and L. M. Dı́az-Aranda. 2012. Los estadioslarvarios de los crisópidos ibéricos (Insecta, Neuroptera,Chrysopidae), nuevos elementos sobre la morfologṍa lar-varia aplicables a la sistemática de la familia. Graellsia 68:31Ð158.

New, T. R. 1981. Some early stages of Dictyochrysa Esben-Petersen (Neuroptera, Chrysopidae). Neuroptera Int. 1:136Ð140.

New, T. R. 1983. The egg and Þrst instar of Italochrysa in-signis (Neuroptera, Chrysopidae). Aust. Entomol. Mag.10: 29Ð32.

New, T. R. 1986. Some early stages of Calochrysa Banks(Neuroptera, Chrysopidae). Aust. Entomol. Mag. 13: 11Ð14.

Oberprieler, R., A. Calder, and C. Meyers. 2000. Inverte-brate taxonomy, advances in beetle systematics and bi-ology. Invert. Taxonom. 14(6): 719Ð958.

Oswald, J. D. 2013. Bibliography of the Neuropterida: aworking bibliographyof the literature on extant and fossilNeuroptera, Megaloptera and Raphidioptera (Insecta:Neuropterida) of the World. Version 10. (http://insects.tamu.edu/research/neuropterida/neur_bibliography/bibhome.html).

Pérez-de la Fuente, R., X. Delclòs, E. Peñalver, M. Speranza,J. Wierzchos, C. Ascaso, and M. S. Engel. 2012. Earlyevolution and ecology of camoußage in insects. Proc.Natl. Acad. Sci 109: 21414Ð21419.

Principi, M. M. 1940. Contributi allo studio dei NeurotteriItaliani. I. Chrysopa septempunctata Wesm. e Chrysopaflavifrons Brauer. Bollettino dellÕIstituto di Entomologiadella R. Università degli Studi di Bologna 12: 63Ð144.

Principi, M. M. 1943. La Nothochrysa italica Rossi ed i suoisingolari costumi (Neuroptera - Chrysopidae). Bollettinodella Società Entomologica Italiana 75: 117Ð118.

Principi, M. M. 1946. Contributi allo studio dei Neurot-teri Italiani. IV. Nothochrysa italica Rossi. BollettinodellÕIstituto di Entomologia della Università degli Studi diBologna 15: 85Ð102.

Principi, M. M. 1956. Campagna di ricerche dellÕIstituto diEntomologia dellÕUniversità di Bologna nella “ForestaUmbra” (Gargano). IV. Reperti etologici su di un raroneurottero crysopide, lÕHypochrysa nobilis Schneider.Atti dellÕAccademiadelle ScienzedellÕIstitutodiBologna,Rendiconti, Classe di Scienze Fisiche (11)3: 152Ð154.

Réaumur, R.A.F. 1737. Mémoires pour servir a lÕhistoire desinsectes, p. 532, vol. 3. Paris, France.

Rhainds, M., D. R. Davis, and P. W. Price. 2009. Bionomicsof bagworms (Lepidoptera: Psychidae). Annu. Rev. En-tomol. 54: 209Ð226.

Silva, P. S., C. A. Tauber, G. S. Albuquerque, and M. J.Tauber. 2013. Larvae of Þve horticulturally importantspecies of Chrysopodes (Neuroptera, Chrysopidae):shared generic features, descriptions and keys. ZooKeys262: 39Ð92. (http://www.pensoft.net/J_FILES/1/articles/4119/4119-G-3-layout.pdf).

Skorepa, A. C., and A. J. Sharp. 1971. Lichens in “packets” oflacewing larva (Chrysopidae). Bryologist 74: 363Ð364.

Slocum, R. D., and J. D. Lawrey. 1976. Viability of the epiz-oic lichen ßora carried and dispersed by green lacewing(Nodita pavida) larvae. Can. J. Bot. 54: 1827Ð1831.

Smith. R. C. 1922. The biology of the Chrysopidae. CornellUniv. Agric. Exp. Stn. Mem. 58: 1287Ð1372.

Smith, R. C. 1926. The trash-carrying habit of certain lacewing larvae. Sci. Mon 23: 265Ð267.

Stange, L. A. 2004. A systematic catalog, bibliography andclassiÞcation of the world antlions (Insecta: Neuroptera:Myrmeleontidae). Mem. Am. Entomol. Inst. 74: 1Ð565.

Stange, L. A., and R. B. Miller. 1990. ClassiÞcation of theMyrmeleontidae based on larvae (Insecta: Neuroptera),pp. 151Ð169. In M. W. Mansell and H. Aspöck (eds.),Advances in Neuropterology. Proceedings of the ThirdInternational Symposium on Neuropterology (3Ð4 Feb-ruary 1988, Berg en Dal, Kruger National Park, SouthAfrica). South African Department of Agricultural De-velopment, Pretoria.

Tauber, C. A. 1969. Taxonomy and biology of the lacewinggenus Meleoma (Neuroptera: Chrysopidae). Univ. Calif.Publ. Entomol. 58: 1Ð94.

Tauber, C. A. 1974. Systematics of North Americanchrysopid larvae: Chrysopa carnea group (Neuroptera).Can. Entomol. 106: 1133Ð1153.


Tauber, C. A. 1975. Larval characteristics and taxonomicposition of the lacewing genus Suarius. Ann. Entomol.Soc. Am. 68: 695Ð700.

Tauber, C. A. 2003. Generic characteristics of Chrysopodes(Neuroptera: Chrysopidae), with new larval descriptionsand a review of species from the United States and Can-ada. Ann. Entomol. Soc. Am. 96: 472Ð490.

Tauber, C. A., and T. de León. 2001. Systematics of greenlacewings (Neuroptera: Chrysopidae): larvae of Ceraeo-chrysa from Mexico. Ann. Entomol. Soc. Am. 94: 197Ð209.

Tauber, C. A., and M. J. Tauber. 1987. Food speciÞcity inpredacious insects: a comparative ecophysiological andgenetic study. Evol. Ecol. 1: 175Ð186.

Tauber, C. A., and M. J. Tauber. 1989. Sympatric speciationin insects: perception and perspective, pp. 307Ð344. In D.Otte and J. A. Endler (eds.), Speciation and its Conse-quences. Sinauer, Inc. Sunderland, MA.

Tauber,C.A., andM. J. Tauber. 2013. Anunusual chrysopidlarva: identiÞcation, description, and taxonomic implica-tions. Ann. Entomol. Soc. Am. 106: 729Ð740.

Tauber, C. A.,M. J. Tauber, andM. J. Tauber. 1991. Egg sizeand taxon: their inßuenceon survival anddevelopment ofchrysopid hatchlings after food and water deprivation.Can. J. Zool. 69: 2644Ð2650.

Tauber, M. J., C. A. Tauber, J. R. Ruberson, L. R. Milbrath,and G. S. Albuquerque. 1993. Evolution of prey speci-Þcity via three steps. Experientia 49: 1113Ð1117.

Tauber, C. A., J. R. Ruberson, and M. J. Tauber. 1995a. Sizeand morphological differences among the larvae of twopredacious species and their hybrids. Ann. Entomol. Soc.Am. 88: 502Ð511.

Tauber, C. A., M. J. Tauber, and L. R. Milbrath. 1995b.Individual repeatability and geographical variation in thelarval behaviour of the generalist predator, Chrysopaquadripunctata. Anim. Behav. 50: 1391Ð1403.

Tauber, C. A., T. de León, N. D. Penny, and M. J. Tauber.2000. The genus Ceraeochrysa (Neuroptera: Chrysopi-dae) of America north of Mexico: larvae, adults, andcomparative biology. Ann. Entomol. Soc. Am. 93: 1195Ð1221.

Tauber, C.A., M. J. Tauber, and G. S. Albuquerque. 2001.Plesiochrysa brasiliensis (Schneider) (Neuroptera:Chrysopidae): larval stages, biology, and taxonomic re-lationships. Ann. Entomol. Soc. Am. 94: 858Ð865.

Tauber, C. A., M. J. Tauber, and G. S. Albuquerque. 2006.Berchmansus elegans (Neuroptera: Chrysopidae): larvaland adult characteristics and new tribal afÞliation. Eur. J.Entomol. 103: 221Ð231.

Tauber, C.A., G. S. Albuquerque, and M. J. Tauber. 2008a.Gonzaga nigriceps (McLachlan) (Neuroptera: Chrysopi-dae): descriptions of larvae and adults, biological notes,

and generic afÞliation. Proc. Entomol. Soc. Wash. 110:417Ð438.

Tauber, C. A., M. J. Tauber, and G. S. Albuquerque. 2008b.A new genus and species of green lacewings from Brazil(Neuroptera: Chrysopidae: Leucochrysini). Ann. Ento-mol. Soc. Am. 101: 314Ð326.

Tauber, C. A., F. Sosa, G. S. Albuquerque, and M. J. Tauber.2013. Adults and larvae of two Leucochrysa (Leuco-chrysa) species (Neuroptera:Chrysopidae);descriptions,biologicalnotes, and relationships.Zootaxa3750: 101Ð129.

Tillyard, R. J. 1926. The insects of Australia and New Zea-land, 1st ed, p. xi 560. Angus and Robertson, Sydney.

Toschi, C. A. 1965. The taxonomy, life histories, and matingbehavior of the green lacewings of Strawberry Canyon(Neuroptera, Chrysopidae). Hilgardia 36: 391Ð433.

Tsukaguchi, S. 1978. Descriptions of the larvae of ChrysopaLeach (Neuroptera, Chrysopidae) of Japan. Kontyû 46:99Ð122.

Tsukaguchi, S. 1995. Chrysopidae of Japan (Insecta, Neu-roptera). S. Tsukaguchi, Aioi-cho 6Ð14-102.Nishinomiya-shi, Hyogo, 662 Japan.

Weber, N. A. 1942. A neuropterous myrmecophile, Nadivavalida Erichs. Psyche 49: 1Ð3.

Wheeler, Q. D. 1990. Morphology and ontogeny of postem-bryonic larval Agathidium and Anisotoma (Coleoptera:Leiodidae) Am. Mus. Novit. 2986: 1Ð46.

Wiggins, G. B. 1977. Larvae of the North American cadd-isßy genera (Trichoptera). University of Toronto Press,Toronto, Canada.

Wilson, P. J., and A. S. Methven. 1997. Lichen use by larvalLeucochrysa pavida (Neuroptera: Chrysopidae). Bryolo-gist 100: 448Ð453.

Winterton, S. L., and S. J. Brooks. 2002. Phylogeny of theApochrysine green lacewings (Neuroptera:Chrysopidae:Apochrysinae). Ann. Entomol. Soc. Am. 95: 16Ð28.

Winterton, S., and S. de Freitas. 2006. Molecular phylogenyof the green lacewings (Neuroptera: Chrysopidae). Aust.J. Entomol. 45: 235Ð243.

Winterton, S. L., H. P. Guek, and S. J. Brooks. 2012. Acharismatic new species of green lacewing discovered inMalaysia (Neuroptera, Chrysopidae): the conßuence ofcitizen scientist, online image database and cybertax-onomy. ZooKeys 214: 1Ð11.

Withycombe, C. L. 1922. Notes on the biology of someBritish Neuroptera (Planipennia). Trans. R. Entomol.Soc. Lond. 70: 501Ð594.

Withycombe, C. L. 1924. Some aspects of the biology andmorphology of the Neuroptera. With special reference tothe immature stages and their possible phylogenetic sig-niÞcance. Trans. R. Entomol. Soc. Lond. 72: 303Ð411.

Received 15 October 2013; accepted 23 January 2014.


Appendix 1. Chrysopid larval habits and larval descriptions: summary of current information

I. Subfamily Nothochrysinae (2% of chrysopid species)Nine genera, larvae from four known

Genera with detailed larval descriptions (no. � 3)Dictyochrysa (1 species, L1) nakedHypochrysa (1 species, L1, L3) nakedNothochrysa (3 species, L1, L3) debris-carrier, occasional debris-carrier

Larval synapomorphiesSeveral shared, unique larval features (enumerated by Dṍaz-Aranda and Monserrat 1995, Monserrat and Dṍaz-Aranda 2012)Antenna: terminal segment short (approximately 10 times shorter than remainder of antenna); terminus with group of small apical

setae, not an elongate seta (L3)Labial palpus: terminal segment with more than three lateral sensillaa

Larval features with intergeneric variationHabit: naked to debris-carrierBody: fusiform, elongate, ßattened dorsoventrally to slightly thickenedHead: setae (L1) clavate or Þliform, (L3) blunt or ÞliformThorax: lateral tubercles (LTs) absent to very small with a few (2Ð3) setae (LS), or fairly well developed with numerous straight or

hooked LS, and if present, not more developed than abdominal LTsAbdomen: LTs absent to small; laterodorsal tubercles (LDTs) absent or very small, with only one seta (LDS)Types of setae: thoracic notal setaeÐshort, bacilliform or Þliform; abdominal submedian setae (SMS) (A1ÐA6)Ðshort, bacilliform or

elongate, hooked or elongate, blunt without hook

II. Subfamily Apochrysinae (1% of chrysopid species)Six genera, larvae known from one

Genera with detailed larval descriptions (no. � 1)Apochrysa (1 species, L3 only) debris-carrier or occasional debris-carrier

Larval synapomorphiesNone currently identiÞed (Tsukaguchi 1995; possible apomorphy: mesothoracic and metathoracic LTs longer than those on prothorax)Larvae of subfamily currently distinguished by set of features shared with Chrysopinae or Nothochrysinae

Distinguishing larval characteristics: based solely on ApochrysaBody: elongate, weakly humpedHead: jaws longer than head capsule, antenna with terminal segment long (only approximately three times shorter than remainder of

antenna), antennal terminus with elongate apical setaThorax: LTs larger than those on abdomen; LTs considerably longer than broad (especially on T2 and T3); dorsum without tuberclesAbdomen: dorsum of A1ÐA7 with small LDTs, each with more than one setaTypes of setae: thoracic notal setae Þliform, abdominal SMS of A1ÐA6 moderately hooked

III. Subfamily Chrysopinae (97% of chrysopid species)Larvae known from all four tribes

L

forum debris-carrying in larval chrysopidae: unraveling ... › neuropterida ›...

Documents