Review and phylogeny of Lutzsimulium ( Diptera : Simuliidae )

Lutzsimulium d'Andretta Jr & Vulcano, 1947 is an enigmatic South American genus with four species: L. flavopubescens Lutz, 1910, L. hirticosta Lutz, 1909, L. pernigrum Lutz, 1910 and L. simplicicolor Lutz, 1910. It can be diagnosed by median arms of furcasternum with projections; subbasal tooth of the claw reduced; wing basal cell absent; spermatheca with net-like structure; apex of trichomes coiled (pupa); gill with two main trunks (pupa); antennomere 3 equal to or longer than 1+2 (larva); hypostomal teeth reduced (larva); postgenal cleft deep (larva). A morphological cladistic analysis under equal weights, with the four Lutzsimulium species and six outgroups, resulted in two most parsimonious trees, with 81 steps, CI = 0.61 and RI = 0.68. The monophyly of the genus is corroborated, supported by 15 synapomorphies, therefore it is proposed that Kempfsimulium Py-Daniel & Nunes de Mello, 1982 is synonymous of Lutzsimulium. Also the status of Araucnephia Wygodzinsky & Coscaron, 1973 and Araucnephioides Wygodzinsky & Coscaron, 1973 are revalidated, because they do not form a monophyletic group with Lutzsimulium. All the species of Lutzsimulium are revised, with redescriptions, illustrations and identification keys for adults, pupa and larva. The male and larva of L. flavopubescens are described for the first time.

ordered (FITCH 1971).When two or more states of a character were found in one terminal taxon, the character was treated as polymorphic and both states were coded in the matrix.For the maximum parsimony analyses under equal weights the program NONA (GOLOBOFF 1993) was used.The searches were carried out through exact algorithms and the sequence of commands was: "hold 100000 search mswap+".
The parsimony analyses including the representatives of the ingroup and outgroup resulted in unrooted networks.Character optimization and mapping, and comparisons among the trees obtained were carried out with the WinClada program (NIXON 2002).Branch support was estimated by bootstrap (FELSENSTEIN 1985), via WinClada (NIXON 2002), with 3,000 replications.The decay index (BREMER, 1994) of each branch was calculated by Autodecay (ERIKSSON 2001) using PAUP* (SWOFFORD 1998).
The matrix proposed by PY-DANIEL (1982: 297) was adapted and a hypothetical ancestor was included to polarize the matrix.The data set was analyzed under the same parameters of my main analysis.

CLADISTIC ANALYSIS
The data matrix was composed of 11 terminal taxa and 53 characters, obtained from adult (24), pupa ( 16) and larva of last instar (13) (Appendix 2).The characters used in the present analysis are listed below.The discussion of characters and their consistecy indexes (ci) are given for one of the two shortest trees, chosen to represent the phylogeny of the group (Fig. 1).
Modified from character 21 of PY-DANIEL (1982: 298); 1 of EVANS & ADLER (2000: 1471).Character 38 of GIL-AZEVEDO & MAIA-HERZOG (2007: 49).The derived character state (2) is a synapomorphy for Lutzsimulium.In the discussion of his character 21, Py-Daniel (1982), suggested that there are small spicules on the internal surface of the spermatheca in Lutzsimulium (except L. simplicicolor).It is however impossible to observe clearly with a light microscope.Therefore the presence/absence of these spicules was not used in this analysis.
19. Ratio between height and width of the gonostylus: (0) the height more than twice the base; (1) height less than twice the width, but distinctly greater (Figs 13,28,37 and 48); (2) the height and width with similar lengths.ci = 0.66.
Character 25 of PY-DANIEL (1982: 299).The presence of teeth on the distal area of the median sclerite is an autapomorphy of L. simplicicolor.
The derived character state (0) is an autapomorphy of L. simplicicolor.
The derived state ( 1) is a synapomorphy for Lutzsimulium.Character 05 of MIRANDA-ESQUIVEL & COSCARÓN (2001).The body tegument with hairs is an autapomorphy of L. hirticosta.PY-DANIEL (1982) (character 53, p. 300) described membranous projections on the cuticle.However, I was not able to observe those projections.Also, Py-Daniel (1982) (character 54, p. 300) proposed that there are ventral tubercles in all Lutzsimulium (except L. simplicicolor), but I was not able to see those tubercles in any species.

Discussion
Even though I have tried to use all characters of PY-DANIEL (1982) in my analysis, some characters were found to be inappropriate for cladistics, as follows: 1) character 9 (p.298): PY-DANIEL proposed that L. simplicicolor does not have setae on the Sc vein, but a few setae (three) are present on the Sc vein of L. simplicicolor; 2) character 15 (p.298): this character, pertaining the shape of the katepisternum, has several intermediate conditions, making it difficult to distinguish between the states proposed by the author; 3) character 39 (p.299): the presence of two transversal grooves on the dorsum of the pupal abdomen (present in L. simplicicolor) is a confusion the author made with the sutures between the tergites and pleurites, present in all Simuliidae.
The cladistic analysis under equal weights resulted in two most parsimonious trees with 92 steps, CI = 0.65 and RI = 0.71.The difference between the two trees pertains the sister group of Lutzsimulium, which can be either Araucnephia iberaensis or Simulium + Paraustrosimulium.The tree that shows A. iberaensis as the sister-group of Lutzsimulium was chosed for character optimization and discussion.This tree (Fig. 1) does not show autapomorphic characters.
The analysis of PY-DANIEL'S (1982) data set resulted in three trees (Fig. 2).All trees support the monophyly of Lutzsimulium + Kempfsimulium, with seven synapomorphies.Also, analysis of his data resulted in ten autapomorphies for Kempfsimulium, which PY-DANIEL (1982) used to support his new genus.Consequently, even under the data set of PY-DANIEL (1982), the creation of a new monotypic genus is not heavily justified.Given that Lutzsimulium (sensu lato) is well supported and easily recognizable, there is no reason to maintain a monotypic genus.I propose to include Kempfsimulium as a junior synonym of Lutzsimulium.
In a phylogenetic analysis of the Simuliinae genera from the Southern Hemisphere (GIL-AZEVEDO & MAIA-HERZOG 2007) and here, A. iberaensis was recovered as the sister group of Lutzsimulium, and did not cluster with Araucnephia.However, this sister-group relationship is not only weakly supported, but relies on synapomorphies that are not easily recognizable: gonostylus height less than twice the wegth, but distinctly greater (character 19, state 1); facial and dorsal-scutum tri-chomes apex coiled (respectively: character 27, state 0; character 29, state 0).For these reasons I have chosen not to transfer A. iberaensis to Lutzsimulium.The problem is that A. iberaensis, the African Paracnephia, Procnephia, Araucnephioides and Araucnephia montana conserve many primitive features of the Simuliini.The position of those groups is uncertain and further investigation is necessary to clarify them.
What makes Lutzsimulium more enigmatic is that, in spite of presenting basal features of Simuliini, it shares many character states with Simulium (GIL-AZEVEDO & MAIA-HERZOG 2007), a genus of Simuliini considered to be highly derived.These features are: median arms of furcasternum (adult) with projections (character 7, state 1); basal cell (vein m-cu) absent (character 10, state 1); larval antennomere 3 with as long as or longer than 1+2 (character 43, state 1); mandibular serrations with less than seven teeth (character 44, state 1); hypostomal teeth reduced -next to the hypostomal plate, which partially covers the teeth (character 45, state 1); paralateral teeth reduced (character 46, state 1); postgenal cleft deep (character 48, state 1).As the monophyly of the clade Simulium -2 (sensu GIL-AZEVEDO & MAIA-HERZOG 2007) is very strong, it is here hypothesized that those characters have evolved independently in both genera.Nevertheless, the occurence of so many homoplasies (seven at least) between Lutzsimulium and Simulium needs further investigation.Based on phylogenetic and biogeographic data, two distinct groups can be recognized in Lutzsimulium.One group has three species (L.hirticosta, L. pernigrum, and L. flavopubescens), and is supported by five synapomorphies: paraproct (female terminalia) with a distinctly tail-like expansion (character 12, state 1); ventral plate (male terminalia) in dorsal view with medial portion narrower than the basal portion (character 21, state 0); and endoparamere reduced or absent (character 24, state 1); sexual dimorphism:pupal frontal trichomes absent in females and present in males (character 28, state 1); larval articles 1 and 2 unpigmented, contrasting with the dark antennomere 3 (character 41, state 1).Inside this group, L. pernigrum and L. flavopubescens are close species, a hypothesis corroborated by the following synapomorphies: larva with deep postgenal cleft subsquare (character 49, state 2), and with irregular border (character 50, state 1).The other group is composed by a single highly autapomorphic species, L. simplicicolor, with five non-homoplasious autapomorphies: adult female with clypeus acute apex distinctly developed (character 3, state 1); paraproct (female terminalia) with digit-like projection on the anterior and basal area (character 13, state 1); median sclerite of male terminalia without laterally pointed apex (character 23, state 0) and with distal area with teeth (character 22, state 1); and hypostoma of larva with 1+1 paralateral teeth (character 47, state 0).The fact that this species has many unique features probably led PY-DANIEL (1982) to place it in a monospecific genus, which was reassigned to subgenus status by (CROSSKEY & HOWARD 1997).I think that a monotypic subgenus is redundant, especially for a genus with only four species.
Distribution and biology.The genus occurs in the eastern side of South America and seems to be strongly correlated with forest and other humid areas.The distribution area is disjoint, ranging from northeastern and southeastern Brazil to northeastern Argentina, and also in the south and east borders of Amazonia .This disjoint distribution may be caused by the presence of the South American semi-arid diagonal, which ranges from northeastern Brazil to the northwestern Argentina.
ZOOLOGIA 27 (5): 761-788, October, 2010 Simuliidae are viewed as a pest most of the time but, like other insects, they also suffer a risk of extinction, mainly caused by the destruction of their habitat.Three species of Lutzsimulium need some attention because of their restricted habitats.L. flavopubescens was only found in a restricted area in Itatiaia mountain range, and probably is endemic of that region.Luckily, that area is inside one of the most important Brazilian National Parks, the first to be created in Brazil.Lutzsimulium pernigrum is restricted to altitudinal relict areas, which need to be preserved, and L. hirticosta is also related to areas of low anthropic impact.The biology and ecology of those species need more studies because maybe they could serve as useful bioindicators of the state of preservation of areas.
Keys to the species of Lutzsimulium Adult 1.General coloration of thorax yellow.Female -clypeus acute apex distinctly developed (Fig. 42); spermatheca with netlike structure present only in the duct (Fig. 47); paraproct without a distinctly tail-like expansion and with a digit-like projection in the anterior and basal area (Fig. 45).Maleventral plate in dorsal view with medial portion wider than basal (Fig. 49); endoparamere developed ..... L. simplicicolor 1'.General coloration of thorax black or light brown.Femaleclypeus acute apex normal (Fig. 3); spermatheca with netlike structure present in spermatheca neck and duct (Fig. 12); paraproct with a distinctly tail-like expansion and without a digit-like projection in the anterior and basal area (Figs 10, 27 and 36  Description.Female.General body color black.Body length 2.4-2.7 mm (n = 2); mesothorax length 1.1-1.4mm, mesothorax width 0.8-0.9mm (n = 4); wing length 2.4 mm, wing width 0.9 mm (n = 1).Head dichoptic with black eyes, with reddish edge; nudiocular area small and rounded (Fig. 3).Frons, clypeus and occiput black with silver pruinosity and many long copper bristles; frons higher than wide and with a medial suture (Fig. 3), clypeus as wide as long and with acute apex (Fig. 3).Antennae with silver pruinosity; scape and pedicel light brown, flagellomeres black.Mouthparts dark; palp and labium black with silver pruinosity, labrum reddish.Mandible with about 35 teeth on internal side, without teeth on external side.Maxilla with about 14 teeth each side; palp with the last and the penultimate palpomeres approximately equal; Lutz's organ length about 1/3 of basal article length.Cibarium unarmed, with reduced and sclerotized cornua (Fig. 4).Scutum and postpronotal lobe entirely dark brown, with silver pruinosity and many copper bristles uniformly distributed.Scutellum dark brown with silver pruinosity and many long black bristles.Postnotum and pleural region black with silver pruinosity.Medium arms of furcasternum with projections (Fig. 5).Chaetotaxy of wing veins, C and R 1 with hair-like and spinelike setae; Sc with an only row of hair-like setae; basal portion of R and Rs with hair-like setae; vein Rs unbranched and basal cell absent (Fig. 6).Halteres entirely black.The coloration pattern of the legs is not obvious.Hind leg with calcipala present (basitarsus) and a shallow pedisulcus (Fig. 7).All legs with many long copper bristles.All leg claws bearing small triangular subbasal tooth (Fig. 8).Abdomen entirely black with many long copper bristles.Hypogynial valve distally straight; inner margins slender, sclerotized and subparallel, diverging distally (Fig. 9).Cerci hemispheric, with width more than twice of the height; paraproct square and small, with anterior area high and posterior area narrow resembling a tail and with setae only in anterior side; length of paraproct and of cerci similar sized (Fig. 10).Genital fork stem as long as lateral plates; lateral plates short and broad, with a triangle form, and posterior edge straight and sclerotized; stem entirely pigmented and lateral plates unpigmented; lateral plates forming a oval space (Fig. 11).Spermatheca spherical and entirely pigmented; surface of the spermathecal duct and insertion area of the spermatheca with net-like structure (Fig. 12).Male.General body color black.Body length 2.9-3.2mm (n = 2); mesothorax length 1.1-1.2mm, mesothorax width 0.9 mm (n = 2); wing length 1.9-2.7 mm, wing width 0.8-1.2mm (n = 3).Head holoptic with black reddish eyes.Frons, clypeus, mouthparts, occiput and antennae colorations as in female, except for the dark red scape of the antenna.Lutz's organ length about 1/6 of basal article length.Thorax coloration entirely black, with silver pruinosity and many copper bristles uniformly distributed; furcasternum, wing veins, halteres coloration, and legs as in female.Abdomen as in female.Gonocoxite small, length almost the same of the gonostylus; gonocoxite with trapezium shape, so that the right side is smaller than the left and the greater width and the greater height sub-equal (Fig. 13).Gonostylus canine-like, with height less than twice the base, but distinctly greater than base; apex with two developed spinules (apical + sub-apical) (Fig. 13).Ventral plate hemispheric with sclerotized basal arms, medium area high, lateral margins convex and concave pointed apex (Figs 14 and 14).Ventral plate and gonostylus similar in size.Median sclerite elongated, flattened and hourglass shape; with enlarged base, middle with a constriction, and lateral pointed and sclerotized apex, with a membrane between the points (Fig. 16).Paramere and dorsal plate absent.Pupa.Length 3.2-4.6mm (n = 10); gill length 1.0-1.7 mm (n = 10).Gill with 12 gill filaments branching in a dichotomous way (arborescent) and directed forward (Fig. 17); the basal stalk immediately giving rise to two main trunk each with six filaments [2 + (2 + 2)] + [2 + (2 +2)] (Fig. 17); all filaments with equal length and width; surface of gill filaments smooth; apex of all filaments digit-like.Head and thorax surface light cracked with few small rounded tubercles with a random distribution; head with 2+2 frontal trichomes only in male (none in female), 1+1 facial trichomes; thorax with 6+6 dorsalscutum trichomes; all the trichomes simple and hair-like; trichomes of the dorsal-scutum region and facial trichomes with coiled apex.Abdomen well sclerotized, majority of plates sclerotized and with small rounded tubercles and setae; pleurites of the abdomen reduced to striated areas with a small plate that carries the hook.Abdominal chaetotaxy (Fig. 18), tergites III-IV each with 4+4 anteriorly directed hooks long and thin (similar to the ventral hooks); tergites VI-IX each with a transverse row of small comb spines on anterior margin; tergite IX with 1+1 developed and slightly curved terminal spines; segment IX with many simple, long and curve (cane-like) terminal setae.Sternite IV without hooks; sternite V with 2+2 anteriorly directed simple hooks; sternites VI-VII with 1+1 anteriorly directed simple hooks.Pleurites VI-VII with 1+1 anteriorly directed simple hooks.Cocoon (Fig. 19) light brown, with shape not defined, ill-formed and asymmetrical; anterior and ventral rim not defined; cocoon often covering all the body and gill, and often with debris.
Remarks.This species was described by LUTZ (1910) based in some specimens collected in Itatiaia, next to the Agulhas Negras mountain range (about 2,400 m).All the discussion in the literature on this species is based on the type material, which consists of few pupae mounted in slides and three poorly preserved females.The male was hitherto unknown and the larva described by Lutz in the original description has been later shown to belong to L. hirticosta (PY-DANIEL 1982).LANE & PORTO (1939) identified a female collected in Campos do Jordão (FSP-USP 1,222) as L. flavopubescens.This female, mounted in slide by Py-Daniel, has been damaged by postal handling and according to him this specimen could be referred to any Lutzsimulium species, but, probably not to L. flavopubescens due to small differences (PY-DANIEL 1982).So, Campos do Jordão is a dubious record, and until now L. flavopubescens is endemic to Itatiaia.This species was the last to be transferred to Lutzsimulium (PY-DANIEL 1982), probably because of the little available information.Almost one century after the original description, this species was rediscovered in Itatiaia, and abundant material was collected, so it is possible to provide here descriptions of male and larva, and more complete redescriptions of female and pupa.
Biological and distributional notes.Larvae and pupae were found in small, temporary stream which was located in an open area in highlands (2,400m).The immatures were found grouped, with the cocoons connected, on the mud in water at 4°C.
flavopubescens.Thorax and abdomen covered by small setae.Anal sclerite X-shaped, ventral and dorsal arms similar sized (Fig. 33); rectal organ with three simple lobes; anal ring with about 64 rows of 12-13 hooks each.
Examined material.MLP: ARGENTINA, Misiones, 15 km S.E.Deseado, 1 female with pupal pelt (slide), 20/VIII/1972, S. Coscarón leg.BRAZIL, RJ, Angra dos Reis (km 58), 1 female with pupal pelt (slide), 29/VIII/1979, S. Coscarón leg. 1 male with pupal pelt (slide) (MLP). 1 larva (slide).LUTZ-LSO: BRAZIL, RJ, Petrópolis, 6 females and 4 males (pinned) #12,525, 12,528, 12,529, 12,530, 12,531, 12, Biological and distributional notes.This species has a wide distribution, stretching from northeastern Brazil to northwestern Argentina.But its occurrence is punctual and always few numbers of specimens are collected, suggesting a restricted habitat preference associated with variables still unknown.There seems to be no relation between the occurrence of L. hirticosta and the altitude, since the species was found at different heights.On the other hand, there is an association of this species with small creeks within rainforest areas with low anthropic impact.
Biological and distributional notes.This species has a wide distribution, from Southeastern Region of Brazil to northwestern Argentina.But its occurrence is restricted, and there is a strong relation between the presence of the species and the altitude/latitude.In southeastern Brazil (between 22°10'-22°44'S range) the altitudes recorded for the species are around 2,000m, in southern Brazil (between 29°21'-29°40'S range) the altitudes recorded are about 1,000m, and in Corrientes, Argentina (about 28°06'S) the altitude of the single record is at sea level.But this last record is doubtful, because only one pupa was found among a lot of A. iberaensis, and more collecting is necessary in the locality to confirm it.Anyway, this species is probably related to relicts of Araucaria Forests (Araucaria angustifolia), but more studies are necessary to confirm it.Like L. hirticosta, L. pernigrum is associated with small creeks with low anthropic impact, but unlike the first, it is possible to collect large numbers of L. pernigrum specimens.
Male.General body orange.Body length 3.4-4.1 mm (n = 3); mesothorax length 1.0-1.2mm, mesothorax width 0.9-1.0mm (n = 5); wing length 2.3-2.6 mm, wing width 1.2-1.3mm (n = 2).Head holoptic with black reddish eyes.Frons, clypeus, mouthparts, occiput and antennae colorations as in female.Lutz's organ length about 1/4 of basal article length.Thorax coloration, furcasternum, wing veins, halteres coloration, and legs as in female.Abdomen as in female.Gonocoxite large, length longer than the gonostylus; gonocoxite with trapezium shape, that the right side is smaller than the left, and the greater width and the greater height sub-equal; gonocoxite of this species more rounded than the other species of Lutzsimulium (Fig. 48).Gonostylus as L. flavopubescens; gonostylus apex with two developed spinules (apical + sub-apical) (Fig. 48).Ventral plate hemispheric with sclerotized basal arms, medium area high, lateral margins concave in the middle and distinctly pointed in the distal area, and concave pointed apex; the ventral plate is very faintly (Fig. 49).Ventral plate longer than gonostylus.Median sclerite elongated and flattened; rectangle shape with rounded lateral sides (Fig. 50).Paramere present and developed, with small spines in the apex; dorsal plate absent.
Pupa.Length 3.0-3.8mm (n = 10); gill length 1.1-1.3mm (n = 10).Gill with 12 gill filaments branching in a dichotomous way (arborescent) and directed forward (Fig. 51); the basal stalk immediately giving rise to two main trunk each with about six (Fig. 51); all filaments with equal length and width; surface of gill filaments smooth; apex of all filaments digit-like; filaments shorter and thicker than the filaments of the others Lutzsimulium (Fig. 51).Head surface light cracked and thorax surface smooth, both without tubercles; head and thorax chaetotaxy as L. flavopubescens, except by the absence of sexual dimorphism, with 2+2 frontal trichomes in male and female.Abdomen as L. flavopubescens.Abdominal chaetotaxy as L. flavopubescens, except sternite IV with 2+2 hooks (Fig. 52).Cocoon as L. flavopubescens, except cocoon covering abdomen and part of thorax (Fig. 53).
But before that, the species identity was nebulous and much confusion occurred in the literature, such as misidentification of pupae (FAIRCHILD 1940;FLOCH & ABONNENC 1946;CERQUEIRA 1967), and very doubtful records of females (NEIVA & PENNA 1916;LACEY 1981).PY-DANIEL (1979, 1981) proposed the transfer of this species to Lutzsimulium without explanation.But one year later he decided (with Nunes de Mello) to create Kempfsimulium to place the species, providing an exhaustive study about it, description of male, pupa and larva, and redescription of female, and a comparison of his new genus with other genera (PY-DANIEL 1982).
Biological and distributional notes.The species is found in the southern and eastern borders of Brazilian Amazonia.The immatures occur in Amazonian streams inside the forest (Igarapés) (PY-DANIEL 1982).Females were collected biting horses, cows, monkeys, and humans (PY-DANIEL 1982).
The species was originally considered a human pest, because bites were a problem during the construction of the railway Madeira-Mamoré (LUTZ 1910).But, in almost one century, only one study was carried out on the anthropophilylic habits of the species (LACEY & CHARLWOOD 1980), In spite of the fact that it considered anthropophilic, and is present in a region next to the Amazonian focus of onchocerciasis (SHELLEY et al. 1997), and given the possibility of expansion of the disease (SHELLEY et al. 2000), very little is known about the biology of this species (PY-DANIEL 1982).

Figure 1 .
Figure 1.One of two trees obtained with maximum parsimony, unweighted characters, chosen to represent the phylogenetic relationships of the Lutzsimulium (81 steps, consistency index = 0.65 and retention idex = 0.71, not considering the autapomorphic characters).Only unambiguous apomorphies are represented in the diagram; circles represents apomorphies (white = homoplastic; black = nonhomoplastic); the numbers above the circles represent characters and the numbers below refer to character states.Branch support in the boxes: decay index (above) and bootstrap (below).

Figure 2 .
Figure 2. The three most parsimonious trees obtained from the analysis of Py-Daniel's (1982) data set.