A taxonomic monograph of the assassin bug genus Zelus Fabricius (Hemiptera: Reduviidae): 71 species based on 10,000 specimens

Guanyang Zhang; Elwood R Hart; Christiane Weirauch

doi:10.3897/BDJ.4.e8150

. 2016 Jul 8;(4):e8150. doi: 10.3897/BDJ.4.e8150

A taxonomic monograph of the assassin bug genus Zelus Fabricius (Hemiptera: Reduviidae): 71 species based on 10,000 specimens

Guanyang Zhang ^‡,^§,^✉, Elwood R Hart ^|, Christiane Weirauch ^¶

PMCID: PMC5019016 PMID: 27651730

Abstract Abstract

The New World assassin bug genus Zelus Fabricius, 1803 (Insecta: Hemiptera: Heteroptera: Reduviidae: Harpactorinae: Harpactorini) is revised based on more than 10,000 specimens. Seventy-one species are recognized and twenty-four described as new: Zelus aithaleos sp. n., Zelus amblycephalus sp. n., Zelus antiguensis sp. n., Zelus auralanus sp. n., Zelus bahiaensis sp. n., Zelus banksi sp. n., Zelus casii sp. n., Zelus championi sp. n., Zelus cordazulus sp. n., Zelus fuliginatus sp. n., Zelus gilboventris sp. n., Zelus gracilipes sp. n., Zelus grandoculus sp. n., Zelus kartaboides sp. n., Zelus lewisi sp. n., Zelus panamensis sp. n., Zelus paracephalus sp. n., Zelus rosulentus sp. n., Zelus russulumus sp. n., Zelus spatulosus sp. n., Zelus truxali sp. n., Zelus umbraculoides sp. n., Zelus umbraculus sp. n., and Zelus xouthos sp. n. Five species, Zelus araneiformis Haviland, 1931, Zelus gradarius Bergroth, 1905, Zelus modestus (Stål, 1862), Zelus subfasciatus Stål, 1860 and Zelus vittaticeps Stål, 1866, are removed from Zelus and placed incertae sedis within Harpactorini. Nine new synonyms are recognized (senior synonym in parentheses): Zelus atripes Champion, 1898 syn. nov. (=Zelus conjungens [Stål, 1860]), Zelus dispar Fabricius, 1803 syn. nov. (=Zelus pedestris Fabricius, 1803), Zelus formosus Haviland, 1931 syn. nov. (=Zelus laticornis Herrich-Schaeffer, 1853), Zelus obscuridorsis (Stål, 1860) syn. nov. (=Zelus pedestris), Zelus pallidinervus Haviland, 1931 syn. nov. (=Zelus kartabensis Haviland, 1931), Zelus personatus Berg, 1879 syn. nov. (=Zelus versicolor Herrich-Schaeffer, 1848), Zelus trimaculatus Champion, 1898 syn. nov. (=Zelus means Fabricius, 1803), Zelus trimaculicollis (Stål, 1855) syn. nov. (=Zelus means), and Zelus tristis Haviland, 1931 syn. nov. (=Zelus laticornis). Zelus conjungens (Stål, 1860) stat. rev. Is resurrected from junior synonymy with zealous armillatus (Lepeletier & Seville, 1825). Zelus ambulans Stål, 1862 stat. rev. and Zelus cognatus (Costa, 1862) stat. rev. are resurrected from synonymy with Zelus exsanguis Stål, 1862. Iquitozelus Bérenger syn. nov. is synonymized with Zelus and its only species transferred to Zelus, hence resulting in a new combination, Zelus couturieri (Bérenger, 2003) comb. nov. Lectotypes, paralectotypes or neotypes are designated for a number of species. Habitus images, illustrations of male genitalia, distribution maps and measurements are provided for nearly all species. The three previously recognized subgenera of Zelus are found to be based upon superficial characters and these divisions do not reflect natural groupings. Using sets of characters, especially those of the male genitalia, eleven species groups are proposed. It is also hypothesized that Zelus is closely related to three other New World genera: Atopozelus Elkins, Ischnoclopius Stål and an undescribed genus "Hartzelus" [manuscript name]. Zelus is endemic to the New World, occurring naturally in the Caribbean and all but one of the continental countries, with introductions to Pacific islands, Europe and Chile.

Keywords: Harpactorinae , Heteroptera , natural enemy, Nearctic, Neotropical, new species, Reduviidae , synonymy, systematics, taxonomic revision, Zelus

Introduction

Zelus Fabricius, 1803 is one of the largest reduviid genera (Maldonado 1990) and the largest New World genus in the tribe Harpactorini (Reduviidae: Harpactorinae). Zelus is endemic to and widely distributed throughout the New World, ranging from southern Canada through central Argentina. One species, Zelus renardii Kolenati, 1856, has been introduced to Hawaii (Kirkaldy 1902, Zimmerman 1948), and was recently found in Chile (Curkovik et al. 2004, Elgueta and Carpintero 2004), Greece (Davranoglou 2011, van der Heyden 2015 and Spain (Vivas 2012) (reviewed in Weirauch et al. 2012). Species of Zelus, among several other genera (e.g., Arilus Hahn, Sinea Amyot & Serville, and Montina Amyot & Serville), have been explored and studied as natural enemies in the Americas (Cohen and Tang 1997, Cogni et al. 2002). Species of Zelus prey on a wide range of insects in cotton, corn, soybean, alfalfa crops and fruit trees in California and elsewhere (Ali and Watson 1978, McPherson et al. 1982, Cisneros and Rosenheim 1998, Virla et al. 2015), may reach population densities of up to 50,000 to 75,000/ha, and prevent outbreaks of lepidopteran larvae (Ables 1978). Hart (1972) conducted a taxonomic revision of Zelus with descriptions of twenty-five new species and twenty-six new synonyms, most of which remained unpublished (see Hart 1986, Hart 1987 for treatments of twenty Canadian, US, northern Mexican and Caribbean species). The current state of taxonomy of Zelus remains unsatisfactory and impedes further research on the evolution and ecology of this group. Species identification is difficult in many instances, and misidentifications may arise. For example, Z. renardii was misidentified as Zelus cervicalis Stål, 1872 when it was reported as having been introduced to Chile (Curkovik et al. 2004). This project was thus undertaken to provide a taxonomic monograph of the genus Zelus.

In the current study seventy-one species are treated and twenty-four described as new. Five species are removed from Zelus and placed incertae sedis within Harpactorini. Nine new synonyms are recognized. Three species are resurrected. Iquitozelus Bérenger is synonymized with Zelus. Habitus images, illustrations of male genitalia, distribution maps, and identification keys are provided. This work evaluates and maintains most of the manuscript new species names proposed in Hart (1972)'s systematic revision of Zelus. Four additional new species are discovered and described herein, some based on specimens more recently collected, but some arising from different views of species boundaries. The vast areas of the Amazon and many mountainous regions of Central America and South America remain poorly sampled and should represent an immediate frontier for new species discoveries in this genus.

Review of taxonomic history

The taxonomic history of Zelus is complex and the generic limit of Zelus has undergone constant fluctuations. The first species of Zelus, Z. longipes (Linnaeus), was described by Linnaeus in the 12th edition of Systema Naturae (Linnaeus 1767) under Cimex, a genus in which he also included various other Heteroptera that are now classified within a number of families. Fabricius (1775) transferred Z. longipes from Cimex to Reduvius, a genus that was established to accommodate most of the Reduviidae known at the time. It was again Fabricius who later in the first comprehensive treatment of Hemiptera (Fabricius 1803) erected the genus Zelus. In this work for each genus Fabricius selected one species for which he repeated the short generic description, expanded the species description and italicized the terms referring to the morphological structures described. Zelus longipes was treated by Fabricius this way for the genus Zelus. Therefore, many workers assumed that Fabricius had indicated Z. longipes as the type species of the genus (Kirkaldy 1900a).

Lepeletier and Serville (1825) expanded the limit of Reduvius to include nearly all then described Reduviidae and described many new species; several of them would now be considered members of Zelus. An erroneous designation of Zelus festinans as the type species of Zelus was made by Laporte (1832). He erected new genera and removed some species from Zelus, somewhat changing the generic limit of this genus. Perty (1834) treated Reduvius similarly to Lepeletier and Serville and described several new species of Zelus, placing them in Reduvius.

Burmeister (1835) modified the classification of Reduviidae and divided members of Zelus into two genera: Euagoras Burmeister, 1835 and Arilus Hahn, 1831. The limits of Zelus were expanded by Brullé (1836) to include part of Reduvius as defined by Fabricius, Harpactor and Prionotus as defined by Laporte and Myocoris, Evagoras (for Euagoras), Notocyrtus and Arilus as used by Burmeister. Blanchard (1840) again changed the limit of Zelus. It was to include some species of Cimex of Linnaeus and Stoll (1788), Reduvius of Fabricius and Wolff and Prionotus of Laporte. Amyot and Serville (1843) erected a new genus Diplodus to accommodate both described and new species of some Zelus. Herrich-Schaeffer (1848), Herrich-Schaeffer (1853) described several new species of what would be considered as members of Zelus by subsequent workers. Signoret (1862) described a new species of Diplodus which would eventually be transferred to Zelus.

A series of works by Stål greatly changed the generic limits of Zelus (Stål 1855, Stål 1859, Stål 1860, Stål 1861, Stål 1862, Stål 1866). Among those, Stål (1862) redefined Zelus as containing three subgenera: (1) Zelus of Fabricius, which also contained Euagoras Burmeister, characterized by the posterior pronotal lobe unarmed and the humeral angles rounded. One previously described species was listed, four species being therein synonymized as part of a total of nine varieties listed for that species. (2) Diplodus of Amyot and Serville, characterized by the disc of posterior pronotal lobe unarmed and the humeral angles armed with a tooth or spine. Ten new species were described therein, one of these with three varieties, one with two varieties and another with four varieties. (3) Pindus Stål, recognized by the posterior pronotal lobe armed with two spines on the disc and a spine on each humeral angle. One new species was described.

However, the subgeneric groups were raised to the generic rank by Stål (1866) in a key to the genera of the New World reduviids. Later Stål (1872) again recognized Zelus, Diplodus and Pindus as subgenera of the genus Zelus, with fifteen, thirty and two species respectively. A Fabrician species was not assigned to a subgenus. He also provided a key to the genera, subgenera and species and a list of synonymy. The geographical distribution of Zelus, according to this study, is restricted to the New World.

Uhler listed a species of Diplodus from one of the U. S. Geological Survey expeditions (Uhler 1872a), described a new species of Pindus (Uhler 1872b) and compiled a checklist of North American Hemiptera (Uhler 1886). This checklist recognized six species of Zelus, sixteen species of Diplodus and one species, which in reality should have been assigned to Zelus, had been incorrectly described as a Darbanus by Provancher (1872). Based on the framework established by Stål (1872), Berg (1879) described two new species of Zelus in his faunal study of the Hemiptera of Argentina.

Lethierry and Severin (1896), in their catalogue of the Hemiptera-Heteroptera of the world, adopted Stål (1872)'s definition of Zelus. Fifty-two species, nine synonymized names and eleven varieties were recognized. Champion (1898) recorded eighteen species within Mexico and Central America, eight of which were described as new species. A short discussion of the genus, a key to the eighteen species, drawings, descriptions, redescriptions and species discussions were included.

Kirkaldy (1900a), Kirkaldy (1900b), Kirkaldy (1901), Kirkaldy (1903) reviewed the nomenclatural validity and the history of the types, genera and subgenera of the new species, discussed the possible synonymy of some others and synonymized several species. As Diplodus Amyot and Serville was proven to be preoccupied, Kirkaldy (1900b) proposed Diplacodus as a new name for this subgenus. However, this name was also preoccupied and most following taxonomists used the name Diplocodus.

The generic and subgeneric definitions of Stål were also used by Fracker (1913). In North America, including Mexico, twenty species were included in a key. In a list of the geographic distribution, three additional species for which no specimens were available were discussed. One of these is noted as being a probable synonymy and another as probably belonging to Castolus Stål.

Van Duzee (1916)'s checklist of Hemiptera of the United States included Zelus as a genus of the tribe Zelini, subfamily Harpactorinae. Stål’s three subgenera were recognized. Kirkaldy's Diplacodus was accepted for Diplodus, but was here and after spelled as Diplocodus, no explanation being given for the change. There were nine species recognized in the genus and five names were listed in synonymy. The known references for the genus, subgenera and these nine species were compiled in Van Duzee (1917)'s catalogue of North American Heteroptera.

In his study of the Heteroptera of eastern North America, Blatchley (1926) used the subfamily name Zelinae to include Zelus and related genera. Zelus was redescribed and a short discussion of the distribution and habits of assassin bugs in the genus was included. Each of the six species of this area was keyed, redescribed, and short notes on its biology and distribution made. Readio (1927) made studies of the biology of the ten species which he recognized in North America north of Mexico. These species were keyed and the descriptions and distributions of each were included. A major addition to the number of species of Zelus was made in by Haviland (1931). A total of nine species, including six new species, were reported from British Guiana. Wygodzinsky (1949) published a checklist of the Reduviidae of the Americas. He included sixty-six species, most of which are Neotropical in distribution, as valid members of the genus. Zayas (1960) described a new species from Cuba. Elkins (1954) removed two species.

In his Ph.D. dissertation Hart (1972) provided a systematic revision of Zelus, in which he described new species, proposed new synonyms, created a species group classification, illustrated male genitalia of nearly all species, and provided an identification key. Most of this work remains unpublished, except for two regional treatments of species occurring in Canada, US, northern Mexico and the Caribbean (Hart 1986, Hart 1987), in which he dealt with twenty species, described four new species, proposed fifteen new synonyms and provided identification keys. Maldonado (1990) in the "Systematic Catalogue of the Reduviidae of the World" listed sixty valid species names and did not use subgeneric divisions within the genus Zelus.

Recent taxonomic activities on Zelus spp. are scant. Jadin et al. (2002) described a new species, Zelus josephpaulusi Jadin et al., which was later found to be a synonym of Zelus araneiformis Haviland, 1931 by Baena (2010), a species we do not consider as belonging to Zelus. Bérenger (2003) described a new genus, Iquitozelus, based on a single species, Iquitozelus couturieri Bérenger, 2003. This genus is synonymized with Zelus in the current study. Gil-Santana (2008) synonymized Zelus nigrispinus (Herrich-Schaeffer, 1848) with Zelus versicolor (Herrich-Schaeffer, 1848). Gil-Santana and Forero (2015) placed Zelus iopterus (Perty, 1832) in synonymy with Aristathlus imperatorius Bergroth 1913 and thus created a new combination Aristathlus iopterus (Perty, 1832), removing that species from Zelus.

Materials and methods

Specimens, databasing and georeferencing

During the course of this study, 10,626 specimens were examined and databased. Among those, 4,833 are males, 5,626 are females and the remainders are immatures or with sex undetermined (usually because of missing abdomen). Specimen loans were kindly provided by museums or collections listed in Table 1. Each specimen was affixed with a Unique Specimen Identifier (USI) label (see Schuh 2012). This label has a catalogue number that has a prefix indicating the institution that performed the databasing and almost all specimens examined in this study have the prefix "UCR_ENT" on their USI labels, which means that they were databased at the University of California, Riverside (UCR) or by people associated with that institution. The prefix "UCR_ENT" is followed by an eight digit number. The acronym of the owner or specimen-depositing museum or collection is also indicated on the USI label, e.g., USNM. Label information, including locality and collecting event, was entered into the PBI (Plant Bug Planetary Biodiversity Inventory) instance of the Arthropod Easy Capture database (Schuh 2012) (https://research.amnh.org/pbi/databases/locality_database.html) hosted at the American Museum of Natural History. Specimen records are publicly accessible at the 'Heteroptera Species Page'. When coordinates were not provided on specimen labels, which is the situation for the great majority of the specimens, the locality was geo-referenced using gazetteers. These included the USGS Geographic Names Information System (mainly for US localities), Google Earth, Global Gazetteer and Fuzzy Gazetteer. For localities outside the USA, higher level administration divisions included only country and state (province or department), and the county-level administrative division was not used. Localities consisting of only a US county, but not more detailed information, were given the coordinates of the county seat. Specimen information was then exported from the PBI database as Excel files. Data description terms used in PBI were manually changed to equivalent Darwin Core Terms as specified in the "Darwin Core Terms: A quick reference guide" maintained by the Biodiversity Information Standards - TDWG (2013). Specimen records representing holotype, allotype, lectotype, neotype, paratype and paralectotype of both valid species and synonyms are provided in the main text, and all specimen records including both type and non-type specimens in Suppl. material 1.

Table 1.

List of museums/collections

Acronym	Museum/Collection	Manager/Curator
AMNH	American Museum of Natural History, New York, USA	Ruth Salas/Randall T. Schuh
BMNH	Natural History Museum, London, UK	Mick Webb
BPBM	Bernice Pauahi Bishop Museum, Honolulu, USA
CAS	California Academy of Sciences, San Francisco, USA	Norman D. Penny
CUIC	Cornell University Insect Collection, Ithaca, USA	Rick Hoebeke/James Liebherr
FMNH	Field Museum of Natural History, Chicago, USA	James Boone
FSCA	Florida State Collection of Arthropods, Florida Department of Agriculture, Gainesville, USA	Susan Halbert
ICN	Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Bogotá, Colombia	Carlos Sarmiento
IEXA	Instituto de Ecologia, Xalapa, México	Luis Cervantes
INBIO	Instituto Nacional de Biodiversidad, San José, Costa Rica	James Lewis
KU	Snow Entomological Museum, University of Kansas, Lawrence, USA	Zachary H. Falin
LACM	Natural History Museum of Los Angeles County, California, USA	Weiping Xie/Brian Brown
MEFLG	Museo Entomológico Francisco Luis Gallego, Medellín, Colombia	John Albeiro Quiroz
NHMW	Natural History Museum of Vienna, Vienna, Austria	Herbert Zettel
NHRS	Sweden Museum of Natural History, Stockholm, Sweden	Gunvi Lindberg
RMNH	Nationaal Natuurhistorisch Museum (formerly Rijksmuseum van Natuurlijke Historie), Leiden, Netherlands	Yvonne van Nierop
TAMU	Texas A&M University Insect Collection, College Station, USA	Edward G. Riley
UCB	Essig Museum of Entomology, University of California, Berkeley, USA	Cheryl Barr
UCD	Bohart Museum of Entomology, University of California, Davis, USA	Steve Heydon
UCR	Entomology Research Museum, University of California, Riverside, USA	Douglas Yanega
UMSP	University of Minnesota Insect Collection, St. Paul, USA	Philip J. Clausen
UNAB	Museo Entomológico, Facultad de Agronomía, Universidad Nacional de Colombia, Bogotá, Colombia	Francisco Serna
UNAM	Universidad Autonoma de México, Instituto de Biología, México	Harry Brailovsky
USNM	United States National Museum of Natural History, Washington DC, USA	Michele Touchet/Thomas Henry
ZMAN	Zoological Museum Amsterdam, Amsterdam, Netherlands	Willem Hogenes
ZMUC	Copenhagen University Zoological Museum, Copenhagen, Denmark	Henrick Enghoff

Open in a new tab

Type specimens

Information of type specimens of described, valid species, when available, were reported in the 'Materials' section of each species, as holotype, lectotype or neotype. Hart (1972) examined nearly all primary type specimens of valid names and synonyms. Some of these were also examined in the present work and associated with USI labels, but some were not examined nor attached with USI labels. Holotypes of newly described species were designated and reported. Holotypes and lectotypes of junior synonyms, either previously established or newly designated in the current study, were also reported, but included in the 'Other materials' part of the 'Materials' section under the Darwin Core term 'occurrenceRemarks'. This is because no Darwin Core term specifically denotes the type status of a junior synonym. The primary type specimen stays as a type of a species name even if the name becomes a junior synonym and should be indicated as such. However, it would be confusing and incorrect to include the type specimen of a junior synonym as part of the primary type material of the valid name. It was thus decided to report the type specimens of junior synonyms with the aforementioned method. Dikow and Leon (2014) has used a similar method. Lectotypes, paralectotypes and sometimes allolectotypes of some described species (valid or synonym) were newly designated in this study and indicated with the term 'New Designation' in the materials section for the respective species.

Hart (1972) designated numerous type specimens including holotypes, allotypes and paratypes of his manuscript new species names and lectotypes, allolectotypes, and paralectotypes of some previous names. Most of these type designations were adopted in the present study, but the status "allotype" was not used, and the specimens were instead designated as paratypes. These pertained exclusively to new species. Hart's type designation labels remain attached to specimens, and new designation labels are also affixed to indicate the type status of the specimen designated in the current study.

Distribution and mapping

Distributions were based on specimen records captured in the current study. We have gathered the largest samples ever known of all species, which represent the best available knowledge of the distributions of the species of Zelus. Maps were created using the Simple Mapper tool through the PBI website (http://research.amnh.org/pbi/maps) based on the geo-referenced locality data. Because accuracy and error of geo-referencing are highly variable, distribution records shown on the maps are at best indicative. Besides, ambiguous localities and other localities provided only at the country or state level were not geo-referenced and thus not reflected on the distribution maps. It is, therefore, advisable to look up actual specimen data for locality information (Suppl. material 1). Interactive specimen mapping is available using the Global Mapper module of "DiscoverLife" (http://www.discoverlife.org/mp/20m?act=make_map).

Morphological methods – dissection, observation, imaging and measurement

Dissection. Male genitalia, including the eighth abdominal segment and pygophore with phallus enclosed, were removed, cleared in heated 10% potassium hydroxide (KOH) solution for 5–10 minutes, washed in distilled water, and stored in glycerol. To remove the genitalia, a specimen was softened by soaking the abdomen in water. This was achieved by stationing a pinned specimen on play clay with its abdomen pointing down and immersed in water, while the rest of the specimen was not submerged. This method avoided soaking the whole specimen or removing the entire abdomen. A 'genitalia hook' was made by melting the tip of a glass Pasteur pipette with a minuten insect pin inserted and fixed into it. The pygophore was carefully removed by holding the softened abdomen with forceps on one hand, and inserting the genitalia hook into the membranous connection between the seventh and eighth segments, breaking that membrane and pulling off the pygophore, a series of actions performed by the other hand.

Observation. Observations were made using a Nikon stereo dissecting microscope SMZ1500, illuminated by a Nikon NI-150 High Intensity Illuminator. Initial observations of morphological characters were made based on typically a small number of specimens (one to five) and intraspecific variations were subsequently examined based on a larger selection of geographical representation. Genitalia were observed in glycerol. Structures in this medium may look different from their dry state, especially for soft cuticles. For example, the apices of the parameres of Z. cervicalis, Zelus luridus Stål, 1862 and many other species usually appear shriveled in dry specimens, but are fully expanded in glycerol or 70% ethanol. The orientation of the dissected structures shown in illustrations does not necessarily reflect their natural condition.

Imaging and illustration. Dorsal and lateral habitus images of specimens were taken with a Microptics-USA system (now Visionary Digital, http://www.duninc.com/index.html) with a K2 lens and CF-2 or CF-4 objectives connected to a Canon EOS 1D digital SRL. Images were edited in Photoshop (CS3/CS4) to adjust levels and sharpness. Image background was removed and replaced using CorelDRAW or Photoshop. For most species illustrations of male genitalia were adapted from Hart (1972) except for several new species not included in that work.

Measurement. Measurements were made on a dissecting scope equipped with a two-axes movable stage (Mitutoyo Corp.), with the aid of two digital micrometers (Boeckeler®), which were connected to a Microcode II RS-232 digital readout (Boeckeler®). Most measurements were done in dorsal view, but various orientations were necessary for measuring appendages. Typically, five to ten specimens were measured for each species, but the number may be fewer for species without enough properly preserved specimens. All measurement values reported here are in mm, unless otherwise stated. In Suppl. material 2 measurement values are divided into length values and width values. Suppl. material 2 reports not only average values for each species but also values of all individual specimens. A total of twenty-nine measurements were captured as listed in the following.

Length measurements

Total length: length of body from clypeus to apex of hemelytron
Clyp-Abd: Clypeus-abdomen (length from clypeus to apex of abdomen)
Head (length of head from clypeus to collar of anterior pronotal lobe)
AntOc: Anteocular (length of anteocular region of head, from clypeus to anterior margins of eyes)
PostOc: Postocular (length of postocular region of head, from posterior margins of eyes to collar of anterior pronotal lobe)
AntPron: Anterior pronotal lobe (length from collar to transverse sulcus of pronotum)
PostPron: Posterior pronotal lobe (length from transverse sulcus of pronotum to posterior margin of posterior pronotal lobe)
Scut: Scutellum (only exposed part measured, from posterior margin of pronotum to apex of scutellum)
Scap: Scape
Ped: Pedicel
Antn3: Antennal segment 3/Basiflagellomere (the basiflagellomere tends to be curled and in that case two or several consecutive measurements were taken and their sum was used)
Antn4: Antennal segment 4/Distiflagellomere
Profem: Profemur
Protib: Protibia
Mesofem: Mesofemur
Mesotib: Mesotibia
Metafem: Metafemur
Metatib: Metatibia
Lb1: 1st visible labial segment (this is actually homologous to the second labial segment in other heteropteran insects, and Lb2 and Lb3 are homologous to the third and fourth segments. See Weirauch 2008b)
Lb2: 2nd visible labial segment
Lb3: 3rd visible labial segment

Width measurements

Head (width from outer margin of one eye to that of the other)
InterOcDi: Interocular distance (width from inner margin of one eye to that of the other)
AntPron: Anterior pronotal lobe (width across the widest part)
PosPron: Posterior pronotal lobe (width between humeral angles, not including processes)
Abd: Abdomen (measured at the widest part of the abdomen)
Profem: Profemur (measured at median point)
Mesofem: Mesofemur (measured at median point)
Metafem: Metafemur (measured at median point)

Descriptive taxonomy

Description. Observations were recorded with the software DEscriptive Language for TAxonomy (DELTA) (Dallwitz 1980, Dallwitz et al. 1999). Natural descriptive language was exported and edited. Observations and descriptions were done primarily based on male specimens, descriptions of females were restricted to non-genitalic characters different from males. Four major character systems were described: coloration, vestiture, structure (non-genitalic), and male genitalia. Description of abdominal vestiture was restricted to the ventral surface. Description of a body part usually started with the overall appearance of that body part such as length, width, and general shape. Structural components of that body part were then described in the order from anterior to posterior, medial to lateral, dorsal to ventral, and proximal to distal. Ratios were determined by comparing mean values across specimens measured. Ratio between the segments of labium or antenna was in reference to the respective first segment. Unless otherwise stated, all measurement values reported in the text are mean values. For closely related or very similar species, a full description was provided for one representative and only variable characters were described for other species. The singular form was usually used for paired structures except when referring to spatial relationships between these structures, e.g., "struts (of phallus) separate, sub-parallel", or when referring to the different pairs of legs at the same time, e.g., "femora and tibiae with alternating yellow and dark brown bands". Descriptions of coloration were mainly based on observations of dried specimens, but notes were made if images or observations of live specimens were available. We note that many dried specimens show browning or fading of colors. This was particularly evident in lightly or brightly colored species. We observed in a number of species that the ventral outline of abdomen is curved, a condition apparently resulting from the folding of segments three to six (sometimes two as well). It was not clear if this was a preservation artifact or a natural condition, but it has been consistently seen in several species. Thus, this character was described as observed.

Description of intraspecific variations. Intraspecific variations were described and indicated by terms or phrases as the following: sometimes, occasionally and in some specimens. When variations in coloration can be roughly delimited to several patterns, they were described and the frequency of the patterns sometimes mentioned as well. Intraspecific variations in male genitalic structures were usually not described or documented unless they are important for species diagnosis and identification, which is usually the case only for closely related species.

Association of males and females. For the majority of species, males and females show limited sexual dimorphism in size and coloration, and could be readily associated based on external morphology, corroborated by collecting data. However, sexual dimorphism is pronounced in a number of species. Males and females differ drastically in size, body configuration and coloration. Association of sexes for these species was based mainly on locality data and series of specimens of both sexes. Observations of mating reported in the literature were also consulted and used as corroborative evidence whenever available.

Terminology and abbreviation. External morphology and genitalic terminology followed Forero and Weirauch (2012), Davis (1966), Weirauch (2008a), Weirauch (2008b). The term ‘posterior margin of foramen’ was used to refer to the posterior margin of the foramen of the dorsal phallothecal sclerite. The foramen is located at the anterior part of the dorsal phallothecal sclerite and is an area that lacks sclerotization and surrounds the struts. The following abbreviations are used: BL = body length; C.A. = Central America; cat. = catalogue; comb. nov. = combinatio nova; Cu = Cubitus; descr. = description; desig. = designation; f = female; fig. = figure; L = length; m = male; M = median (vein), orig. = original; Pcu = Postcubitus; preocc. = preoccupied; R = radius, S.A. = South America; SD = standard deviation; sp. n. = species nova; syn. = synonym; syn. nov. = synonymum novum; stat. rev. = status revisus; W = width. Terms used to describe male genitalia are illustrated in Fig. 1.

Figure 1. — Male genitalic structure terms (*Zelus* *errans* Fabricius, 1803 is shown in the illustrations)

Nomenclature (Annotated synonymic list). The synonymic list comprises abbreviated synonymies and included those names that previously appeared in the taxonomic literature or have affected the taxonomy of the species. Citation to ecological, agricultural or other non-taxonomic literature was presented when appropriate, but not meant to be exhaustive. Historical taxonomic publications were briefly annotated to indicate kind of taxonomic information or nomenclatural acts such as .orig descr., checklist, cat., note, fig. and key. When a species name is followed by the original author and year, there is no colon (:) separating the name and the author. A species name followed by a colon indicates that the author of the work is not the author of the name.

Electronic publication

This publication is registered in ZooBank. In accordance with the 2012 Amendment to the International Code of Zoological Nomenclature regarding electronically published works (Krell and Pape 2015), all new species names have been registered in ZooBank and their lsid values provided.

Taxon treatments

Zelus

Fabricius, 1803

Zelus Fabricius, 1803, p. 281, orig. descr.; Latreille, 1804, p. 260, list; Latreille, 1807, p. 129, list; Latreille, 1810, p. 433, type desig.; Lepeletier and Serville, 1825, p.815, list and descr.; Laporte, 1832, p. 9, type desig.; Burmeister, 1835, p. 225, descr.; Brulle, 1836, p. 316-317, descr.; Blanchard, 1840, p. 100, descr. and note; Blanchard, 1845, p. 433, 438, list and note; Herrich-Schaeffer, 1848, p. 88, descr. and note; Kolenati, 1857, p. 458-459, descr.; Stål, 1861, p. 148, descr.; Stål, 1862, p. 449-454, key and subgeneric descr. (with subgenus Zelus); Carpenter and Westwood, 1863, p. 565, note; Mayr, 1866, p. 138, list; Stål, 1866, p. 296, list; Stål, 1868, p. 107, restriction of definition; Stål, 1872, p. 69, 88, key and cat. (with subgenus Zelus); Walker, 1873, VII., p. 49, key, VIII., p. 131-136, cat.; Berg, 1879, p. 150, list (with subgenus Zelus); Uhler, 1886, p. 24, checklist; Provancher,1887, p. 179, note; Lethierry and Severin, 1896, p. 151, cat.; Champion, 1898, p. 251, cat. and note; Kirkaldy, 1900a, p. 263, type verification; Kirkaldy, 1900b, p. 242, syn.; Kirkaldy, 1902, p. 149, note; Fracker, 1913, p. 223, 238-240, key and note (with subgenus Zelus); Van Duzee, 1916, p. 30, checklist (with subgenus Zelus); Van Duzee, 1917, p. 258-259, cat. (with subgenus Zelus); Blatchley, 1926, p. 567-568, key, descr. and note (with subgenus Zelus); Readio, 1927, p. 167, 168-169, key, descr. and note; Zimmerman, 1948, p. 137, note; Wygodzinsky, 1949a, p. 48, checklist; Fracker and Usinger, 1949, p. 277, key to nymphs; Alayo, 1967, p. 5, 35, list, key and note; Hart, 1986, key to North American species; Hart, 1987, key to Caribbean species; Maldonado, 1990, p. 325-332, cat.
Reduvius Fabricius, 1775 (type by subsequent desig., Cimex personatus Linnaeus, 1758), Lepeletier and Serville, 1825 (in part), p. 272, descr.; Perty, 1834 (in part), p. 173, list of species.
Arilus Hahn, 1831 (type by subsequent designation, Cimex carinatus Forster, 1771); Burmeister, 1835 (in part), p. 227-228, descr.; Herrich-Schaeffer, 1848 (in part), p. 33-35, descr.
Euagoras Burmeister, 1835 (type by subsequent designation, E. stollii Burmeister, 1835) (in part), p. 226, descr.; Amyot and Serville, 1843 (in part), p. 368, descr. (as Evagoras); Herrich-Schaeffer, 1848 (in part), p. 43-44, descr.; Stål, 1855 (in part), p. 189, list (as Eccagoras); Stål, 1861, p. 148, (in part) junior syn. of Zelus Fabr.; Mayr, 1866, p. 139, list; Walker, 1873, p. 49, 117, key and cat.; Provancher, 1887, p. 182, descr. (as Evagoras); Kirkaldy, 1900b, p. 242, junior syn. of Zelus Fabr.; Kirkaldy, 1903, p. 215-216, note.
Diplodus Amyot and Serville, 1843, p. 370, descr.; Burmeister, 1853, p. 91, list (included in Euagoras Burm.); Stål, 1860, p. 74, list; Stål, 1862, p. 450, descr. (as subgenus of Zelus); Stål, 1866, p. 296, key; Stål, 1872, p. 90, list (as subgenus of Zelus); Walker, 1873, VII., p. 49, VIII., p. 123, key and cat. (as Diploda); Berg, 1879, p. 151, list (as subgenus of Zelus); Uhler, 1886, p. 24, checklist; Provancher, 1887, p. 179, key and descr.; Kirkaldy, 1903, p. 232, note; Fracker, 1913, p. 239, 240, key and list (as subgenus of Zelus).
Darbanus Amyot and Serville, 1843 (type by monotypy, D. nigrolineatus); Provancher, 1872, p. 106, species descr.; Uhler, 1886, p. 24, checklist; Provancher, 1887, p. 179, 181, key and note; Van Duzee, 1912, p. 324; Fracker, 1913, p. 241, note.
Pindus Stål, 1862, p. 454, orig. descr. (as subgenus of Zelus); Stål, 1866, p. 296, key (as genus); Stål, 1872, p. 92, list and cat., as subgenus of Zelus); Walker, 1873, VII., p. 66, list and cat. (as genus); Berg, 1879, p. 150, list (as subgenus of Zelus) ; Thierry and Severin, 1896, p. 151, cat.; Fracker, 1913, p. 223, 240, key and list; Van Duzee, 1916, p. 30, checklist; Van Duzee, 1917, p. 261, cat.; Blatchley, 1926, p. 569, key.
Diplacodus Kirkaldy, 1900b, p. 242, new name for Diplodus A. and S. (preocc.).
Diplocodus Van Duzee, 1916, p. 30, checklist (new name for Diplacodus Kirkaldy, preocc.); Van Duzee, 1917, p. 260, cat.; Blatchley, 1926, p. 569, key.
Iquitozelus Bérenger, 2003, p. 23, orig. descr., syn. nov. (current study).
Zelus Cimex longipes Linnaeus, 1767

Description

Male: Small to large, total length 8-25 mm (Suppl. material 2), with most of moderate sizes (11-18 mm); usually slender (length/width = 4.0-5.0), some species relatively robust (<3.5) to rather slender (>6.0). COLORATION: Colors and patterns of preserved specimens variably yellowish-brown, reddish-brown, orange-brown, and brownish-black to black, with most species uniformly colored. VESTITURE: Most species with moderately dense or dense, fine, short, recumbent and short, long, erect setae; some species with short, spine-like setae on head and pronotum; few species nearly glabrous. Setation on legs sparse in most species; profemur and tibia with dense sundew setae in some species. STRUCTURE: Head: Length much greater than width across eye. Postocular lobe usually longer than anteocular, tube-like posteriorly in most species. Ocellus raised, directed somewhat laterally. Eye variably sized, not protruding above or below dorsal or ventral surfaces of head, with one exception (Zelus grandoculus sp. n.). Antenna: Scape and basiflagellomere long and subequal in length, usually longer than head and pronotum combined; pedicel and distiflagellomere short and about 1/3 length of scape. Scape thickest; basiflagellomere usually thicker than pedicel, subequal in some species. Labium: Segment II longest, 1.3-2.2x length of segment I; segment III shortest, usually 0.5x length of segment I; variably curved between segments I and II. Thorax: Anterior pronotal lobe about 1/2 to 3/4 length of posterior lobe; anterolateral angles of pronotal collar rounded, with or without tuberculate protrusion; medial dorsal longitudinal sulcus usually shallow at collar, deepening through posterior 1/2; sometimes with subtuberculate elevation near posterior margin laterad to medial sulcus. Posterior pronotal lobe rugulose (not conspicuous in species with dense setation); slightly or greatly wider than anterior lobe; disc of most species elevated above humeral angle and posterior margin of lobe; humeral angle with tuberculate to long spinous lateral process, rounded and unarmed in small number of species. Scutellum in most species with angulate apex, slightly produced and projected upward in some species. Legs: long, slender in most species; femoral diameters generally subequal; pro- and metafemoral lengths subequal, greater than mesofemoral length. Hemelytron: Attaining or surpassing apex of abdomen, by large proportion in some species. Quadrate cell small to large; median vein conspicuous in some species and not visible in many. Cu and M of cubital cell subparallel in most species, converging in some. Abdomen: Lateral margins subparallel; ventral outline usually straight, in some species somewhat concave and abdomen appearing arched (see "Material and methods" for discussion of this character). Genitalia: Segment eight usually short, less than 1/2 length of pygophore; posterior margin generally slightly concave, straight in some species, never convex. Pygophore: ovoid to elongated; slightly to greatly expanded laterally close to base of paramere; dorsal bridge short to long. Medial process single, not bifurcating, of variable length and shape; triangular or cylindrical as most common configuration; apex blunt or with hooklike process. Paramere generally cylindrical, often swollen and bending apically, length variable. Phallus: Dorsal phallothecal sclerite generally semi-cylindrical, broad and shield-like in many species, elongated in some; dorsal surface lacking armature in most species, with projection, process or elevation in some species; lateral margins usually straight or convex, constricted or recurved in some species; apical part keeled in middle and/or curved dorsad; apex usually rounded or truncate, with or without medial emargination. Struts attached to dorsal phallothecal sclerite in majority of species; apical part recurved dorsad and often semi-circular; bridge connecting two sides in many species. Basal plate arm slender to heavy, separate or fused; basal plate bridge present, variable in width and degree of sclerotization; basal plate extension short, often extended onto basal plate arm.

Female: Larger than male. Coloration usually similar to that of male and more variable in some species, but may differ between sexes dramatically in certain species. Eye and ocellus smaller than in male in some species. Basiflagellomere not swollen and about equal diameter as or smaller than pedicel. Lateral process on humeral angle, if present, usually more produced and longer than in male. Mesofemur slightly swollen in many species. Lateral margins of abdomen expanded in some species.

Diagnosis

This genus is distinguished from other genera of the New World Harpactorini by the cylindrical head, the length of the head being at least 1.9X its width; the unarmed antenniferous tubercles; the second labial segment being at least 1.3x the length of the first segment; the long scape and basiflagellomere that are subequal in length and the short pedicel and distiflagellomere; the generally unarmed (i.e. no tubercles or spines) disc of the posterior pronotal lobe (except in Zelus tetracanthus Stål, 1862, Zelus lewisi sp. n. and Zelus minutus Hart, 1987); the humeral angle with or without process, and if present, usually not prominently projected; the legs with sundew setae and sticky glands (Zhang and Weirauch 2013); the profemur subequal in length and diameter to the metafemur; and the medial process of pygophore single, not bifurcating. Zelus is apparently closely related to three other genera, Atopozelus Elkins, Ischnoclopius Stål and "Hartzelus" [manuscript name], that share many of the aforementioned characters. It is separated from Atopozelus by the presence of paramere (lacking in Atopozelus). Ischnoclopius is distinguished from Zelus by its rather slender body form (length:width ratio greater than seven), the very long profemur, at least 0.6x of body length, and the very short paramere. An undescribed genus, "Hartzelus" (Gil-Santana and Berenger, pers. comm.), which would be in part based on species removed by us from Zelus, differs from Zelus in having a bifurcating medial process of the pygophore (single in Zelus) and generally more slender legs. No Old World species of Harpactorini are similar or appear to be closely related to Zelus. Confusion may potentially arise with members of genera that show a similar slender body form and slender legs (e.g., Euagoras and Vestula Stål), but these are distinguished from Zelus based on the characters listed above.

Distribution

Native to (except for Chile) and throughout the New World, including the Caribbean, with highest diversity in the Neotropics. One species (Z. renardii) has been introduced to Hawaii, the Polynesian islands, Jamaica, Philippines, Spain, Greece and Chile.

Biology

We provide a non-exhaustive account of the biology of various species of this genus. As with other harpactorines, species of Zelus generally do not show associations with or preferences of host plants, probably due to their generalist habits. However, two recent studies have found two species of Zelus that have both nymphs and adults occurring in the same plant species in relatively large quantities. In Gil-Santana and Alves (2011), based on a multi-year study the authors observed forty-seven females, twenty-seven males and fifteen nymphs of Z. versicolor from Bidens rubifolia Kunth (Asterales, Asteraceae) in a single site in the city of Nova Friburgo, Brazil. Interestingly, they did not see individuals of the same species in other plants in the same site, which can be seen as evidence for host plant preference. In French Guiana Revel et al. (2010) counted as many as 405 individuals of Zelus annulosus (Stål, 1866) and its egg masses from several pubescent plant species, including (but not limited to) Hirtella physophora Mart. & Zucc. (Chrysobalanaceae), Cordia nodosa Lam. (Boraginaceae) and Tococa guianensis Aubl. (Melastomataceae); all three are myrmecophytes. They hypothesized an intriguing tri-party mutualistic relationship between the assassin bug, an ant (Allomerus decemarticulatus Mayr) and the plants.

Several species of Zelus are possibly mimics of various other insects. Zelus errans Fabricius, 1803, Zelus vespiformis Hart, 1987 and to some extent Zelus vagans Fabricius, 1803 and Zelus gracilipes sp. n. have wing and body color patterns similar to many braconid wasps, an intriguing form of mimicry seen also in a number of other Neotropical harpactorine genera. Zelus vagans shows areas of black and orange colors, however, the posterior pronotal lobe is medially dark and laterally orange. Zelus gracilipes also shows a uniformly orange posterior pronotal lobe, but the hemelytron is uniformly dark and lacks the banding pattern typical to a wasp mimic. Zelus nigromaculatus Champion, 1899 has an appearance similar to that of a typical vespid, the only species in this genus with that kind of color pattern. Zelus laticornis (Herrich-Schaeffer, 1853), Zelus grassans Stål, 1862 and Zelus ruficeps Stål, 1862 have red and dark markings on abdomens and orange or reddish dorsal surfaces, a pattern found in many species of pyrrhocorids (e.g., Dysdercus spp.) and coreids (e.g. Hypselonotus spp.). Interestingly, in Z. laticornis, it is only the females showing this coloration. Certain color forms of Z. longipes are possibly mimics of the milkweed bug, Oncopeltus fasciatus (Dallas).

Weirauch et al. (2012) studied predatory and mating behaviors of Z. renardii and Z. tetracanthus and discussed a possible link between biological attributes and invasion potential. Law and Sediqi (2010) experimentally demonstrated that sticky substance derived from egg mass coating improves predation success and substrate adhesion ability of Z. renardii first instar.

Taxon discussion

The generic limit of Zelus is now relatively well defined and the genus can be separated from all other but one genera of New World Harpactorini based on characters discussed in the diagnosis. Based on a molecular phylogeny, Zhang and Weirauch (2014) recovered the monophyly of Zelus, Atopozelus and "Hartzelus" (which includes Z. araneiformis, a species we remove from Zelus). In that analysis, Ischnoclopius was represented by a single species and placed as sister to Atopozelus. The genera Atopozelus, "Hartzelus", Ischnoclopius and Zelus together constitute a monophyletic group in the same study, and we here refer to this group the "Zelus clade". Without a cladistic analysis, questions remain if the characters used to diagnose Zelus are synapomorphies of that genus. It is almost certain that the unbifurcating medial process represents a symplesiomorphic state as that character can be seen in Atopozelus, Ischnoclopius and many other Neotropical harpactorines. The unarmed antenniferous tubercles are also plesiomorphic to Zelus, since all other genera of the Zelus clade exhibit that condition, but may be synapomorphic to the Zelus clade. We agree with Forero (2012) that Zelus is defined mainly by the absence of apomorphies seen in other genera. Future research should illuminate this issue by studying the distribution and the polarity of characters with a formal cladistic framework.

The genus that we are uncertain about its relationship with Zelus is Pronozelus Forero, erected by Forero (2012) to accommodate a new species, Pronozelus schuhi Forero, 2012. This species appears to possess all the characters diagnostic of Zelus, but also shows some peculiar characters. The principal characters separating Pronozelus from Zelus include the laterally expanded posterior pronotal lobe, the prominent, greatly expanded posterolateral rim of pygophore lateral to paramere socket, and the posterior pronotal lobe greater than 2.2x length of the anterior lobe. The conspicuous lateral expansion of the posterior pronotal lobe is not observed in any species of Zelus, but this character appears to be autapomorphic in the Zelus clade and does not support P. schuhi being phylogenetically separated from Zelus. We have not done an extensive survey of the condition of the posterolateral rim of the pygophore and cannot determine the distribution or polarity of the lateral prominence as exhibited in P. schuhi. In Zelus rosulentus sp. n., the posterolateral part of the pygophore also appears to be expanded, although not as prominent as that seen in P. schuhi. Finally, according to measurements done in this study, in Zelus spp. the posterior pronotal lobe frequently exceeds 2.2x length of the anterior lobe, thereby negating the use of that character as a basis for placing P. schuhi outside Zelus. Despite the foregoing discussion, we have opted to not transfer P. schuhi to Zelus or synonymize Pronozelus with Zelus. The polarity of the characters diagnostic to either genus has not been clearly defined. There remains a possibility, although we think a small one, that P. schuhi represents a lineage sister to Zelus.

Bérenger (2003)'s new species, Iquitozelus couturieri, exhibits all the characters diagnostic of Zelus, except for those of the male genitalia as the known specimens are all females. The main character that Bérenger used as the basis for erecting a new genus, i.e., the "foliaceous expansion of the VI connexivum segment", appears to be autapomorphic within the Zelus clade. Synonymy of Iquitozelus with Zelus is warranted and established here. We further postulate that I. couturieri is most closely related to Zelus amblycephalus sp. n., Zelus umbraculus sp. n. or Zelus umbraculoides sp. n. Further discussions regarding the status of Iquitozelus and the specific membership of Zelus couturieri syn. nov. (Bérenger, 2003) are presented in the treatment of that species.

Maldonado (1990) considered two unpublished, manuscript names invalid, and they are "Diplodus armiger" and "Diplodus melanophthalmus". They appeared in Dohrn (1860). We follow this treatment.

Except for several pairs or complexes of closely related species, identification of males can be almost always unambiguously performed based on exposed genitalic structures such as paramere and medial process, further corroborated with phallic structures, external morphology and coloration. Identification of females of many species, where females appear to be as distinct as males, is straightforward based on coloration and external morphology. However, identification can be difficult for closely related species, where females are indistinguishable based on external morphology. In these cases, association of males and females and identification of females were primarily based on collecting event information. Sexual dimorphism presents another special challenge. While most species show limited sexual dimorphism that does not go beyond minor size and coloration differences, some species exhibit pronounced differences between the sexes (see Material and Methods for discussion of association of male and female specimens). Based on the observation that species in closely related genera do not exhibit strong sexual dimorphism, we here hypothesize that pronounced sexual dimorphism is a derived condition within Zelus.

Species groups

We find here that previous subgeneric groups are based on superficial resemblance and these are not adopted. Instead, we recognize eleven species groups in the current study, based primarily on characters of the male genitalia, but also on non-genitalic external morphology if those characters can be applied to both sexes. Several species for which only females are known are therefore not assigned to a species group. Although the groupings proposed here are not based on a cladistic analysis, they show a degree of congruence with the relationships recovered in the phylogenetic analysis based on molecular data in Zhang and Weirauch (2014) and many of the characters are putative synapomorphies of the groups. A brief discussion of the species groups is presented below.

1. Zelus tetracanthus species group.

Zelus minutus Hart, 1987, Zelus prolixus Stål, 1860, Zelus rosulentus sp. n. and Zelus tetracanthus Stål, 1862.

Members of this group have a rather broad, indistinct medial process, the base of which is nearly continuous with or inseparable from the ventral rim of the pygophore. We speculate that this character represents a plesiomorphic condition as it is seen in several other genera of the New World Harpactorini and thus the condition of the medial process does not necessarily support the monophyly of this group. Zelus tetracanthus and Z. minutus also both have tubercles on the disc of the posterior pronotal lobe, which are more pronounced in the former. Comparative views of male genitalia are shown in Fig. 2.

Figure 2. — *Zelus* *tetracanthus* species group, male gentitalic structures

2. Zelus luridus species group.

Zelus ambulans Stål, 1862, Zelus antiguensis sp. n., Zelus exsanguis Stål, 1862, Zelus grandoculus sp. n., Zelus luridus Stål, 1862 and Zelus spatulosus sp. n.

This is a group of species with primarily a North American distribution, with some species extending to northern Central America. The males show an apically expanded paramere and a triangular medial process that has a protrusion at the base but lacks any apical modifications. Notably, Z. spatulosus has a slender medial process, deviating greatly from the remainders of the group. It is placed in this group mainly because of the apically expanded paramere and the uniform coloration. Zelus ambulans and Z. exsanguis have the humeral angle elevated to about same level of and nearly continuous with the disc of the posterior pronotal lobe, a condition rarely seen in the genus. The coloration is quite homogenous among members of this genus, most of which have a uniform greenish (in live specimens) or dull brownish (in preserved specimens) habitus, with only Z. ambulans showing variable patterns or banding on the pronotum or legs. Comparative views of male genitalia are shown in Fig. 3.

Figure 3. — *Zelus* *luridus* species group, male genitalic characters

3. Zelus mimus species group

Zelus inconstans Champion, 1898 and Zelus mimus Stål 1862.

Members of this group, consisting of only two species, exhibit a highly unique paramere and a medial process of the pygophore. The paramere is slender and apically curved dorsad at an angle of nearly ninety degrees. The medial process, as is especially evident in Z. inconstans, possesses a simple posterior liplike fold at the apex; its lateral margins are subparallel and not broadened significantly at the base. Both have a semi-cylindrical dorsal phallothecal sclerite which is modified by a fold running obliquely toward the base from the middle of the lateral margins. Both species, being quite small, exhibit the usual reduction of setal tracts common to nearly all small species. Comparative views of male genitalia are shown in Fig. 4.

Figure 4. — *Zelus* *mimus* species group, male genitalic structures

4. Zelus nugax species group.

Zelus grassans Stål, 1862, Zelus illotus Berg, 1879, Zelus impar Kuhlgatz, 1902, Zelus nugax Stål, 1862 and Zelus pedestris Fabricius, 1803.

This is a group of smallish species with quite variable distributional ranges. The defining characters include a slender, laterally compressed medial process that is curved or recurved, and an acute apex of the dorsal phallothecal sclerite (except in Z. grassans). Zelus nugax has one of the widest distribution ranges in this genus, ranging from much of Mexico to northern South America. Zelus grassans is found primarily in Central America and the remaining two species mainly in northern South America. Comparative views of male genitalia are shown in Fig. 5.

Figure 5. — *Zelus* *nugax* species group, male genitalic structures

5. Zelus puertoricensis species group.

Zelus bruneri De Zayas, 1960, Zelus puertoricensis Hart, 1987, Zelus subimpressus Stål, 1872 and Zelus zayasi Bruner and Barber, 1937.

Members of this group are restricted to the Caribbean. They can be easily recognized by the rather slender body form. The posteriorly directed, robust medial process with a somewhat blunt apical protrusion is also distinctive of this group. The basal plate arms are widely separate and diverging and these features are rare in other species in the genus. They show resemblance to species of the Zelus renardii species group, especially to Z. cervicalis. Zelus bruneri was not physically examined, but the rather slender body form as seen in the original illustration places it within this group. Comparative views of male genitalia are shown in Fig. 6.

Figure 6. — *Zelus* *puertoricensis* species group, male genitalic structures

6. Zelus renardii species group.

Zelus cervicalis Stål, 1872 and Zelus renardii Kolenati, 1856.

The two members of this group are very likely sister species since they share a number of unique characters: the apex of the medial process is greatly bent ventrad and hooklike, the lateral margin of the dorsal phallothecal sclerite is recurved dorsad and the basal part of the strut is absent. Both species are mainly distributed in North and Central America, but Z. cervicalis extends to northern South America. Comparative views of male genitalia are shown in Fig. 7 .

Figure 7. — *Zelus* *renardii* species group, male genitalic structures

7. Zelus armillatus species group.

Zelus amblycephalus sp. n., Zelus annulosus (Stål, 1866), Zelus armillatus (Lepeletier & Serville, 1825), Zelus conjungens (Stål, 1860), Zelus janus Stål, 1862, Zelus leucogrammus (Perty, 1833), Zelus lewisi sp. n., Zelus litigiosus Stål, 1862, Zelus ruficeps Stål, 1862, Zelus sulcicollis Champion, 1899, Zelus umbraculoides sp. n. and Zelus umbraculus sp. n.

This is one of the two largest groups in the genus (the other being the Zelus panamensis species group). Species in this group are generally robust and large-sized (15-25 mm), and some are among the largest in the genus. The most distinctive character is that of the medial process, which has the apex slightly projected into two minute small lateral prongs or processes. This condition is different from that in several species groups listed below, where the apex of the medial process is hook-like and more strongly projected. The lateral spine of the humeral angle tends to be pronounced and somewhat broadened into a dentate effect. The pygophore is large, rounded, and somewhat shortened relative to the total length of the individual. The dorsal phallothecal sclerite having dorsolateral expansions or projections close to the basal arm is also unique to some species of this group. This condition, however, is not seen in Z. amblycephalus, Z. umbraculus, or Z. umbraculoides, which appear to be divergent from the remainders of the group, but the features of the medial process unambiguously place them in this species group. Comparative views of male genitalia are shown in Figs 8, 9.

Figure 8. — *Zelus* *armillatus* species group, male genitalic structures - pygophore

Figure 9. — *Zelus* *armillatus* species group, male genitalic structures - phallus

8. Zelus longipes species group.

Zelus bahiaensis sp. n., Zelus errans Fabricius, 1803, Zelus longipes (Linnaeus, 1767), 1803 and Zelus vespiformis Hart, 1987.

This and the next species group (Zelus vagans species group) possess dense, spine-like setae on the head and pronotum, and a rounded, unarmed humeral angle, both characters rather unique in Zelus and probably synapomorphies uniting the two groups. The former character is possibly homoplastic as it is also seen in two species in the Zelus armillatus species group. The medial process is slender and cylindrical and this condition is among the most extreme in the genus. It is semi-erect and posteriorly directed. The paramere exceeds the apex of the medial process. The dorsal phallothecal sclerite has subparallel margins and lacks obvious modifications or ornamentations (except for small lateral folds in Z. longipes). Some individuals of Z. errans and Z. vespiformis appear to be wasp mimics. Comparative views of male genitalia are shown in Fig. 10.

Figure 10. — *Zelus* *longipes* species group, male genitalic structures

9. Zelus vagans species group.

Zelus aithaleos, Zelus championi sp. n., Zelus fuliginatus sp. n., Zelus gracilipes sp. n. and Zelus vagans Fabricius, 1803.

Species of the Zelus vagans group share two characters also present in the preceding group (Zelus longipes species group): spinelike setae and rounded humeral angle. However, they differ in the structure of the male genitalia in significant ways. The medial process shapes like a somewhat laterally flattened cone. It is relatively broad at base, narrowing toward the apex, and is laterally compressed. The medial process is posteriorly directly, nearly horizontal. The paramere is removed from or barely reaching apex of the medial process. Furthermore, the phallus is elongated and slightly constricted toward the apex (not conspicuous in Z. gracilipes). Zelus vagans and Z. gracilipes also resemble wasps to some extent, but both not as perfectly as seen in Z. errans and Z. vespiformis. Comparative views of male genitalia are shown in Fig. 11.

Figure 11. — *Zelus* *vagans* species group, male genitalic structures

10. Zelus panamensis species group.

Zelus banksi sp. n., Zelus cordazulus sp. n., Zelus filicauda Bergroth, 1893, Zelus gilboventris sp. n., Zelus korystos Hart, 1986, Zelus nigromaculatus Champion, 1899, Zelus panamensis sp. n., Zelus truxali sp. n., Zelus varius (Herrich-Schaeffer, 1853) and Zelus xouthos sp. n.

This is another large group with ten species. Interestingly, most (seven) are new species. It is characterized by having an acute apical modification usually in the shape of a hook on the medial process and the conspicuous medial carination of the apical part of the dorsal phallothecal sclerite. The condition of the apical modification of the medial process differs from that in the Zelus armillatus species group in that it is much more prominent, usually acute and sometimes extending further ventrally. Rugulosity of the posterior pronotal lobe is highly pronounced relative to the other groups. Sexual dimorphism is pronounced in some species in this group (e.g., Z. gilboventris and Z. truxali). Most species in this group are concentrated in southern Central America and northern South America. Comparative views of male genitalia and habitus images are in Figs 12, 13.

Figure 12. — *Zelus* *panamensis* species group, male genitalic structures - pygophore

Figure 13. — *Zelus* *panamensis* species group, male genitalic structures - phallus

11. Zelus erythrocephalus species group.

Zelus auralanus sp. n., Zelus casii sp. n., Zelus chamaeleon Stål, 1872, Zelus erythrocephalus Fabricius, 1803, Zelus kartabenoides sp. n., Zelus kartabensis Haviland, 1931, Zelus laticornis (Herrich-Schaeffer, 1853), Zelus mattogrossensis Wygodzinsky, 1947, Zelus paracephalus sp. n., Zelus russulumus sp. n. and Zelus versicolor (Herrich-Schaeffer, 1848).

Two diagnostic characters identify members of this group. The medial process possesses a broad ridge-like projection or carina that initiates from the apex and extends ventrally or is removed from apex. The second feature is the apically oriented lateral sharp processes or projections on the dorsal phallothecal sclerite. These are not to be confused with the lateral expansion seen in the Zelus armillatus species group, where the direction of the expansion is laterad. In Z. auralanus and Z. versicolor, this process is short and somewhat dorsally directed, rather than apically directed. Three species, Z. kartabenoides, Z. kartabensis and Z. chamaeleon lack this structure. Their placement in this group is primarily based on the configuration of the medial process and the absence of characters of other groups. Also, the longitudinal ridge-like elevation or hook on the medial process is similar to the condition in another species, Z. laticornis, although the latter has a short modification. In this species group the parameres are usually somewhat bulbous and curved medially with moderate to long erect setae on the apical 1/2. The medial process is broadened at base, and usually anteroposteriorly compressed. Furthermore, the basal plate of the phallus is strongly curved in some members of this species group. Pronounced sexual dimorphism is seen in some species of this group. Notably, three species, Z. erythrocephalus, Z. paracephalus and Z. russulumus have purple, blue or greenish iridescence on the membrane of the hemelytron. Species of this group show a predominant southern South American distribution, with a few found only from the Amazons. Comparative views of male genitalia and habitus images are in Figs 14, 15.

Figure 14. — *Zelus* *erythrocephalus* species group, male genitalic structures - pygophore

Figure 15. — *Zelus* *erythrocephalus* species group, male genitalic structures - phallus

Because of the heavy emphasis on male genitalic characters for grouping species, four species described only from females are not placed in any of the species groups defined in the above. These are: Zelus fasciatus Champion, 1899, Zelus plagiatus (Signoret, 1852), Zelus sphegeus Fabricius, 1803 and Zelus means Fabricius, 1803. Zelus fasciatus is similar to the females of some of the species in the Zelus panamensis species group and also occurs in an overlapping geographical region (southern Central America). Zelus plagiatus and Z. sphegeus show resemblance to the females of Z. versicolor, which is in the Zelus erythrocephalus species group. Zelus means, by possessing a rounded humeral angle and spinelike setae, aligns most closely with the Zelus vagans species group and the Zelus longipes species group. A future cladistic analysis, including morphological and molecular data, is needed to test the monophyly of these species groups and may also have the potential to place these female-based species.

Species removed from Zelus

Five species are removed from Zelus: Z. araneiformis, Zelus gradarius Bergroth, 1905, Z. modestus (Stål, 1862), Zelus subfasciatus Stål, 1860 and Zelus vittaticeps Stål, 1866. These species represent an undescribed genus "Hartzelus" and will be treated in a separate study. They will be listed as Harpactorini incertae sedis until their generic placement is formally clarified.