Description of Lutzomyia (pifanomyia) Robusta N. Sp. (diptera, Psychodidae, Phlebotominae) from Peruvian Equadorean Interandean Areas

Description of Lutzomyia robusta, n. sp. (Diptera, Psychodidae, Phlebotominae) from interandean areas of Peru and Equador. Lutzomyia robusta, n. sp., probable vector of human bartonellosis and cutaneous leishmaniasis, is described and illustrated. This species presents strong affinity with L serrana (Damasceno & Arouck, 1949) but they can be distinguished by variance analysis of four male characteristics and only one female characteristic. In the variance analysis, populations of L. serrana, of Amazonian areas of Brazil, Peru and Bolivia, the coast of Equador and other areas of Brazil were studied. The synonymy of Lutzomyia guayasi (Rodriguez) and L serrana was corroborated. Captures of sand flies carried out in the Region Nor-Oriental del Marañón, Peru, between 5° 3' SL-78° 53' WL and 6° 5'30" SL-78° 30'29" WL, in the provinces of San Ignacio (interandean valley of Chinchipe), Jaén (interandean valley of Chamaya and Chinchipe) and Utcubamba (right bank of River Marañón), during the period 1987-1992, motivated by the occurrence of an outbreak of human bar-tonellosis, showed the presence of five new species: two, of the subgenus Helcocyrtomyia Barretto (Galati & Cáceres 3 , 1994) and three, of the sub-genus Pifanomyia Ortiz & Scorza; two of wich latter have already been described (Galati et al 4 ,1995) and the third of wich, L. robusta, n. sp., is the object of the present study. This species, in both sexes, is very close to L. serrana (Damasceno & Arouck, 1949), the differences between them being more quantitative than qualitative; hence the two species have been confused. In the mentioned areas, in intradomiciliary captures, with light CDC trap, L. robusta, n. sp. showed the greatest density and in the peridomi-cile, with human bait, was one of the most prevalent ; this behaviour suggests its involvement in the transmission of the parasitosis. Using the same capture techniques this new species was the most frequent in the Equador area, near the border of Peru, in the Amazonian province of Zamora-Chinchipe, where cases of human bartonellosis were also identificated and which is considered a cutaneous leishmaniasis endemic zone (Le Pont et al. 5 , 1994). Rodriguez 6 (1956) describes the male and female of Phlebotomus (Brumptomyia) guayasi from the coast of Equador, which has been regarded as identical with L. serrana by Fairchild & Hertig 2 (1961). Because of the great similarity between the two species and the probable implication of this new taxon in the transmission of the above …


Captures of sand flies carried out in the Region
Nor-Oriental del Marañón, Peru, between 5° 3' SL -78° 53' WL and 6° 5'30" SL -78° 30'29" WL, in the provinces of San Ignacio (interandean valley of Chinchipe), Jaén (interandean valley of Chamaya and Chinchipe) and Utcubamba (right bank of River Marañón), during the period 1987 -1992, motivated by the occurrence of an outbreak of human bartonellosis, showed the presence of five new species: two, of the subgenus Helcocyrtomyia Barretto (Galati & Cáceres 3 , 1994) and three, of the subgenus Pifanomyia Ortiz & Scorza; two of wich latter have already been described (Galati et al 4 ,1995)  and the third of wich, L. robusta, n. sp., is the object of the present study.This species, in both sexes, is very close to L. serrana (Damasceno & Arouck, 1949), the differences between them being more quantitative than qualitative; hence the two species have been confused.
In the mentioned areas, in intradomiciliary captures, with light CDC trap, L. robusta, n. sp.
showed the greatest density and in the peridomicile, with human bait, was one of the most prevalent; this behaviour suggests its involvement in the transmission of the parasitosis.Using the same capture techniques this new species was the most frequent in the Equador area, near the border of Peru, in the Amazonian province of Zamora-Chinchipe, where cases of human bartonellosis were also identificated and which is considered a cutaneous leishmaniasis endemic zone (Le Pont et al. 5 , 1994).
Rodriguez 6 (1956) describes the male and female of Phlebotomus (Brumptomyia) guayasi from the coast of Equador, which has been regarded as identical with L. serrana by Fairchild & Hertig 2 (1961).
Because of the great similarity between the two species and the probable implication of this new taxon in the transmission of the above mentioned parasitoses, it is necessary to clarify its taxonomic status.With this purpose in view, some morpholog-ical characteristics were submitted to variance analysis, by means of a comparison of populations of Lutzomyia robusta, n. sp. of the interandean areas with those of the Lutzomyia serrana of the areatype (Amazonian Brazilian region), as well as with the Peruvian and Bolivian Amazonian areas and of other areas of Brazil and of the coast of Equador.

Material and Method
For the variance analysis, 8 male characteristics were observed: length: flagellomere I, labrum-epipharynx, hind tibia, R 5 , genital ductus; width of the wing and paramere and ratio between length and width of paramere apical area (setal area) of two populations of Lutzomyia robusta, n.sp. of the interandean areas: Peru-Region Nor-Oriental (NO) and Equador -Zamora-Chinchipe Region (ZC) and of nine populations of L. serrana (except for hind tibia of the areas: Amazonian: Peru Loreto/Madre de Dios (PA), Bolivia -Pando (BO) and Brazil -Pará/Maranhão (PM), Rondônia (RO), Acre (AC), Mato Grosso (MT); Atlantic -Brazil -Bahia/Espírito Santo/Rio de Janeiro (AT) and inland areas of Brazil Minas Gerais (MG) and the coast of Equador (EC).For the hind tibia, the following areas were not included: BO, AC, RO, because in the specimens examined it was missing.For females, 7 characteristics were observed: length: flagellomere I, labrumepipharynx, R 5 , fore, mid and hind tibiae and wing width, in the populations of L. robusta of NO, ZC and only four populations of L. serrana: BO, MG, MT, EC.
Sclerotized characteristics, the least subject to measurement variations, were selected for observation only.
The program used was prepared by J. L. F. dos Santos, professor of Department of Epidemiology, of the Faculdade de Saúde Pública, of the Universidade de São Paulo.The confidence interval was calculated by Gabriel's formula (Sokal &  Rolhff 8 , 1981), which is: where: gi is the average values of each characteristic; m the critical value (alpha = 0.05) obtained from table 21 of "Studentized range" Rolhf & Sokal 7 , 1981); Sy i the standard error; k* the n° of places of the origin and the v degrees of freedom within of the treatment.

Results
The greatest values and the overlapping of the all confidence intervals (CI) can be observed with regard to all eight characteristics submitted to variance analysis for the males of the interandean populations ZC and NO (Fig. 27).
The overlapping of the CI occurs with regard to the populations of the Amazonian region: PA, BO, AC, PM, RO, MT and those from the Atlantic area (AT) and the coast of Equador (EC).
The MG population presents the CI values closest to those of ZC and NO, with overlapping of the labrum-epipharynx length and wing width of the males.
By means of the R 5 length CI three population groups can be distinguished: the first covering the Amazonian and Atlantic areas and of the coast of Equador; the second MG and the third of the interandean region.
Besides R 5 length, the two interandean populations are distinguished from the others by their hind tibia, paramere width and ratio between length and width of the apical part of the paramere.

Comments
Based on the eight characteristics of the males and seven of the females submitted to variance analysis, the interandean populations are similar.On the other hand, it is possible to distinguish them from all the populations of L. serrana studied by means four characteristics of the males: lengths of R 5 and hind tibia, width of the paramere and ratio between length and width of the apical area of the paramere (Fig. 26) and only one of the females: length of R 5 .
It is possible distinguish the males of the two interandean populations of the L. serrana from the area-type (Amazonian) by means of the length of the AIII, labrum-epipharinx and by the wing width also.
The Mato Grosso (MT) population was included among those of the Amazonian area because the specimens were from the north of the State, which has the same ecological characteristics, as also the specimens from Pará and Maranhão (PM) which were studied together because they belong to areas characterised by similar vegetal cover.
The females of the L. robusta and L. serrana are not distinct, except as regards the R 5 length.
The population of L. serrana from Minas Gerais (MG) presents the confidence interval values nearest to those of L. robusta and at times, these are different from those of the Amazonian and Atlantic areas; hence it needs to be better evaluated, in view of possible genie flux interruption among the three L. serrana populations.
Based on this data, we denominate the new species of L. robusta, encountered in the Nor-Oriental del Marañón region of Peru and Zamora-Chinchipe, Equador, because of the larger size of the characteristics analysed, as compared to those of the L. serrana.
Since it was impossible to distinguished between the confidence intervals of the coastal population of the of Equador and those of the other L. serrana populations, as regards the various characteristics submitted to variance analysis, the synonymy of P. guayasi Rodriguez, as determined by Fairchild & Hertig 2 (1961), is corroborated.
Taxonomic Discussion.The serrana series of the subgenus Pifanomyia, according to Galati et al. 4 (1995) is characterized, in both sexes, by presenting long palpomere V, rosette sensilla on AV (except for series evansi) and the absence of the rosette sensilae on AXII and AXIII.Males with one ascoid on AXIV and AXV; style with subapical seta and three spines (the inferior external spine was lost) with internal spine situated on basal region of the structure; coxite with basal tuff constituted up to nine setae implanted directly on the surface, and lateral lobe with round top, without differentiated setae.The females present setae on tergite VIII; the spermathecae are striated and vesiculous, with terminal annulus.The species with these characteristics are: L. odax (Fairchild & Hertig, 1961), L. oresbia (Fairchild & Hertig, 1961), L. orestes (Fairchild & Trapido, 1950) The females of L. robusta, n. sp. in can be distinguished from the those of L. oresbia because they present spermathecae with large terminal annulus, its length is about 1/2 of those of the striated part.The separation of the other species was not possible, except by R 5 length, studied in the variance analysis, with regard to the L. serrana.
The association between the sexes of L. robusta, sp.n. was based on body coloration and by the simultaneous capture of males and females in the absence of other species of Pifanomyia, series serrana.

Conclusions
It was possible to distinguish L. robusta, n. sp.from the various L. serrana populations on the basis of four male characteristics among the eight submitted to variance analysis and only one of the female characteristics among the seven analysed.
The males of L. robusta, can be distinguished from those of the L. serrana by the appearance of the paramere, that in the former presents strong dorsal concavity preceding the setal area and is straigther in the latter.
The females of L. robusta are distinct those of the L. serrana as regard R 5 length only.
The L. serrana population from Minas Gerais presents male labrum-epipharynx and R 5 length and wing width confidence intervals distinct from those of the populations of the Amazonia and Atlantic areas.Thus, further more studies are necessary for its taxonomic characterization.
The L. serrana population from the coast of Equador is indistinguishable from the other population of the Amazonian areas.Thus, the synoymy of P. guayasi with L. serrana is corroborated.