Three An. squamifemur specimens were collected on the 5^th of September 2005 at Arca de Israel (Pando Department, Federico Roman Province, Municipality of Nuevo Manoa (Nueva Esperanza)) a small community located in the extreme north-east of Bolivia, on the banks of the river Madera, on the border with Brazil (Fig. 1). This small settlement was founded in the year 2000, but is at present the most populated of the area (≈250–300 inhabitants grouped in ≈70 houses). The area is characteristic of the humid Amazonian forest (thermicity index 710–740 °C, ombrothermic index 4.5–6.5 mm °C⁻¹ and dry season ombrothermic index 0.5–2.3 mm °C⁻¹), at a mean elevation of ≈120 m [5]. Two CDC light-traps were placed inside and outside a house (S10°15′40″S, 65°19′11″W, altitude 117 m) located outside of the community center, on the road to Nueva Esperanza, between 18:00 h and 6:00 h next morning. The site was surrounded by the Amazonian forest. One An. squamifemur female was captured inside the house along with two An. oswaldoi, while outside, two An. squamifemur females were collected along with four An. darlingi, four An. oswaldoi and one An. triannulatus. The presence of the characteristic tuff of long, erect black scales on the distal third of hind femurs permitted the unambiguous classification of the specimens as An. squamifemur.

The monotypic Subgenus Lophopodomyia was established in 1937 by Antunes with An. squamifemur as the type species [7]. At present, along with An. squamifemur, five other species belong to this strictly Neotropical Subgenus [1]: An. gilesi (Peryassú 1908); An. gomezdelatorrei Levi-Castillo, 1955; An. oiketorakras Osorno-Mesa, 1947; An. pseudotibiamaculatus Galvão and Barreto, 1941; and An. vargasi Gabaldón, Cova-García and López, 1941. From the Lophopodomyia Subgenus, only An. gilesi has been previously recorded from Bolivia [8,9] (Table 1).

Adults of the six species of this Subgenus are morphologically very close to those of the Anopheles Subgenus and are generally best distinguished by examining the male genitalia [10,11], except for An. squamifemur for which a special distinguishing mark exists: both sexes show a characteristic tuff of long, erect black scales on slightly less than distal third of hind femurs that permits an easy and undoubted identification. In the Neotropical region, An. squamifemur is the only Anopheles species exhibiting such scale tuffs on its hindlegs. Other Anopheles species are known to exhibit tuffs of scales on their legs such as all the species of the Lophoscelomyia Series (Subgenus Anopheles) from the Hindu-Asiatic region and the Malaysian species An. wellingtonianus Alcock, 1912 (Subgenus Anopheles, Series Anopheles).

The description of An. squamifemur female was first published by Antunes [7], then the male was described [12] and later the larva, pupa and egg [13]. All stages of An. squamifemur can easily be distinguished from the other species of the Lophopodomyia Subgenus [13]. However, using usual identification keys [14,15], the 4^th stage larvae would be misidentified as An. (Anopheles) eiseni which is also present in Bolivia (Table 1). However, An. squamifemur larvae can be easily differentiated by the presence of a palmate tuft hair on the first abdominal segment (seta 1-I) and a generally shorter size as compared to eiseni [13].

An. squamifemur is also known from Brazil (States of Para and Amazonas, close to the Venezuelan border [12,16] and State of Rondonia, frontier with Bolivia [17,18]), Colombia [7] from where comes the holotype, Costa Rica [19], French Guiana [20], Panama [21], Peru [22], Suriname [23] and Venezuela [24].

Lophopodomyia mosquitoes are sylvatic species, biting man and animals at twilight and their aquatic stages are passed in shaded side pools of streams rich in organic matter [25]. An. squamifemur follows these rules: it has always been captured in tropical humid forest areas [[18,20,21,26], among others]. It is, for example, widespread and locally common in the Atlantic coastal forests of the Isthmus of Panama [13]. Little is known of the bionomics of An. squamifemur. Immature stages have rarely been collected and the only types of breeding sites known at present are from Venezuela [27] and Peru [28]. They were encountered in small shaded areas in the banks of small or large rivers, where the stream is slow and permits the accumulation of natural floating debris and vegetation (Eichhornia crassiipes y Pistia stratiotes). Such an area may also be the consequence of a fallen tree trunk (for example) which enables the accumulation of organic debris and the slowing down of the stream. Few An. squamifemur were also collected in ponds [27]. Adults were also collected in Panama by horse-bait traps [13,29] and light traps [21,26], in a forest area of French Guiana [20] and at the borders of forests using animal baits at twilight [30]. An. squamifemur may bite humans [13], but nothing is known on its importance as a Plasmodium vector. It is probably limited, due to generally low densities of this mosquito in areas where human populations are also scarce.

In Bolivia, the surroundings of the collection site were similar to the few descriptions of other adult collecting sites elsewhere [18,20,21,26,30]. The Bolivian collecting site was situated in a characteristic Amazonian semi-evergreen seasonal forest (Fig. 1) and it is likely that An. squamifemur may extend its distribution range in Bolivia throughout the extension of this biogeographic region in that country.

These two species belong to the Anopheles subgenus (Series Arribálzagia) and are very closely related to An. mediopunctatus. An. costai has been recently resurrected from synonymy with An. mediopunctatus [31], while An. forattinii has been recently described [32]. These three species are morphologically similar and therefore were misidentified as An. mediopunctatus for several years in South America, as a consequence of identifications based only on adult female characters which could not discriminate [31]. An. costai and An. forattinii females are undistinguishable, and apparently undistinguishable from An. mediopunctatus except by the shape of the cercus [31,32]. Differences amongst the three species are more marked in male genitalia, pupae and larvae. In Bolivia, where the usual keys of Gorham et al. [9,15] or Stojanovich et al. [14] are routinely used to identify Anopheles species, An. costai and An. forattinii, absent from these keys, were de facto confused with An. mediopunctatus. The correct identification of these two species in Bolivia was recently carried out by mounting larvae collected during the “Bolivian Vectors Mapping Project” and using the differentiating characters from [31,32]. Both species were first identified in the Cochabamba Department, Chapare Province, Municipality of Villa Tunari: An. forattini from the Moleto locality and An. costai from the Villa 14 de septiembre locality (Fig. 1). However, only a few larvae from these two localities were classified undoubtedly as costai or forattinii. Other several tens of larvae examined from five other localities of the Chapare Region (Aroma, Moleto, San Mateo Bajo, San Rafael, Villa 14 de septiembre) and two localities from the Tarija Department in the south of Bolivia (Sachapera in the Municipality of Yacuiba (Gran Chaco Province), and Valle Dorado in the Municipality of Padcaya (Arce Province)) (Fig. 1) presented a mixture of characters that prohibited a clear classification. They shared characters of both species, even amongst the most discriminating ones (i.e., the antenna length/width, setae 3C, 4C, 9C, 12C, 14C, 7P, 13P, 2S and 1-II) [31]. To sum up the determination results, some larvae shared more characters with forattini (in particular in the south of Bolivia), while others more with costai. However, none were very close to mediopunctatus s.s. although sharing some of its characters (on setae 8P for example). If costai, forattini and mediopunctatus are really separated species and not mutant forms, some larvae from Bolivia may represent something else characterized by a mixture of (at least) the two former species. Genetic analysis (in prep.) of Bolivian specimens will likely give some cues to clear this situation. Larval sites were shaded pools (>40 cm deep and <10 m2) in the margin of small rivers, in forest situations, at low altitude (200–300 m) with cold clear water (with or without floating debris and algae).

Adults An. mediopunctatus s.l. formerly identified from the Bolivian collections came from various sites of Bolivia, at low altitude: (1) in the Cochabamba Department, Chapare Province, localities of Muyurina, Ambrosia, San Mateo alto, San Mateo Bajo and Aroma; Carrasco Province, localities of Puerto Villaroel, Bulo Bulo and Sajta; (2) Santa Cruz Department, Ichilo Province, localities of Buena Vista and Huaytu, (3) Pando Department, Manuripi Province, locality of Puerto Rico; (4) Beni Department, Vaca Diez Province, localities of Riberalta and Guayaramerin, and (5) Tarija Department, Gran Chaco Province, localities of Carapari, Estacion Caiza and Sachapera; Arce Province, localities of La Goma and Valle Dorado (Fig. 1). These adults were examined using the discriminating character of the cercus (elliptical for An. costai and An. forattinii, and triangular for An. mediopunctatus) [31,32]. Although it was impossible to classify them as An. costai or An. forattinii because of the exact morphological similarity of the two species for this character, they were not An. mediopunctatus s.s. It seems that An. mediopunctatus s.s. is at present only known from southeastern Brazil [32], and because none of the specimen examined from Bolivia was identified as An. mediopunctatus s.s., this species was removed from the list of Bolivian Anopheles (Table 1). However, in Bolivia, more research is needed to clarify the taxonomic status of An. costai, An. forattini and the closely related forms encountered.

An. costai is also known from Argentina, Brazil, Colombia, French Guiana, Trinidad, Tobago, Panama, Peru, Surinam [31], and Venezuela [33] and might be present in all other countries where An. mediopunctatus was formerly (mis)identified: British Guiana, Colombia, Ecuador, Mexico [31]. An. forattinii is also known from Amazonian Brazil, Colombia, French Guiana, Peru, Surinam [32] and Venezuela [33].

Little is known about the biology of the two mentioned species, but in Bolivia their distribution range is wide, from the north (Amazonian region) to the south, but always in areas of low altitude (<600 m), in ecoregions such as Southwest Amazonia and in the lower parts of the Yungas and the Tucumano-Bolivian Forest (Fig. 1). An. mediopunctatus s.l. can be involved in malaria transmission [31]. However, in Bolivia, because of the rarity and low densities of An. costai and An. forattinii and the low human densities where these species have been found, their role as a malaria vector might be limited.

This species was captured during surveys we carried out in 2002–2003 in Guayaramerín, a city of ≈35000 inhabitants on the banks of the Mamoré River, a river ≈1 km wide, on the border with Brazil (Fig. 1). Guayaramerín is facing its “sister” Brazilian city (Guajara-Mirin) of similar size, and from which An. deaneorum was first described and included in the albitarsis complex [34]. Adults An. deaneorum captured in Guayaramerín were distinguished from the closely related species An. albitarsis s.s. also present in the samples by a paler general external appearance for An. deaneorum, and the insertion of postlateral tufts of black scales appearing from the third tergite for An. albitarsis and from the fourth tergite for An. deaneorum [34]. The presence of An. deaneorum in Guayaramerín was also reported shortly before us in 2002 [35], but it has been impossible to examine specimens from these collections and therefore the presence of the species could not be confirmed at that time.

An. deaneorum is one of the five species of the albitarsis complex, along with albitarsis s.s., albitarsis sp. B, albitarsis sp. E and An. marajoara [36–38]. In Bolivia, it is likely that An. deaneorum has been confused for several years with An. albitarsis s.s. because of their very close morphological appearance. In other countries, species of the albitarsis complex are known to transmit Plasmodium sp., but further research is needed to ascertain the malaria vector status of each of these cryptic species and in particular the role of the three (An. deaneorum, An. albitarsis s.s. and An. marajoara) that have been identified in Bolivia (Table 1). In the past, An. albitarsis has also been divided in two closely related species: An. albitarsis and An. allopha [39]. Some specimens labelled An. allopha were found in the collection of pinned adults at Escuela Técnica de Salud Boliviano–Japonés de Cooperación Andina (Cochabamba). Although the two species might be morphologically distinguished [39], An. allopha is now in a nomen dubium [2,40]. However, Bolivian specimens matched the description of both “species” (An. allopha and An. albitarsis) as described by Faran and Linthicum [39].

The history of An. trinkae in Bolivia was made possible by the examination of apocrypha literature, mainly internal reports of Dr. Lien (a former collaborating scientist) to the Bolivian Ministry of Health from 1980 to 1985, and held in the Entomology Laboratory at Escuela Tecnica de Salud Boliviano–Japonés de Cooperación Andina (Cochabamba, Bolivia) [41]. An. trinkae has been described in 1979 and its presence in Bolivia was at that time hypothesized [42]. Its real identification in Bolivia was confirmed by Lien in 1981 [43]. Unfortunately this discovery went unnoticed because of the unavailability of the local Bolivian journal where it was published. Lien, who at that time worked in Bolivia with one of us (RR), first identified An. trinkae from San Mateo Bajo (S 17.0165, W 65.4378, Altitude 351 m), a small community in the Chapare Province of the Cochabamba Department, by a careful examination of larvae and adults reared in the laboratory.

An. trinkae belongs to the Albimanus Section of the Nyssorhynchus Blanchard Subgenus of Anopheles. In this Section, it is classified in the Oswaldoi Subgroup of the Oswaldoi Group of the Oswaldoi Series [44]. The Subgroup comprises eleven species [1] of which two: An. rangeli Gabaldón, Cova García and López, 1940 and An. nuneztovari Gabaldón, 1940 (or more precisely the nuneztovari complex) are morphologically very close to An. trinkae. As such, the adult females of An. trinkae, An. nuneztovari and An. rangeli can sometimes be confused due to intraspecific variability and the paucity of reliable differentiating characters [39]. As a consequence, before 1979, An. trinkae was misidentified as An. nuneztovari or An. rangeli. In Bolivia, where the three species are present (Table 1), the latter misidentification was the rule. Since the identification of An. trinkae in Bolivia, its presence has been confirmed by molecular tools [45,46]. In Bolivia, it has a wide distribution range from the Andes (tropical areas) in the Departments of La Paz (Yungas region), Cochabamba (Chapare region), Santa-Cruz and Tarija, to the Amazonian region in the Departments of Pando and Beni (to a lesser extent) (Fig. 1). This species is also known to occur along the eastern slope of the Andes in Colombia [42], Ecuador [42] and Peru [47]. It is known as a poor malaria (P. vivax) transmitter [39,44,47]. In Bolivia, it is suspected to be a malaria vector in some localities of the Chapare region of the Cochabamba Department, along with An. species C [48]. However, because of its feeding behavior (mostly zoophilic) and its low susceptibility to Plasmodium infection, its role as an efficient (and locally important) vector in Bolivia still remains to be clarified.