Abstract

EPIDEMIOLOGY

Classification of immature stage habitats of Culicidae (Diptera) collected in Córdoba, Argentina

Walter R Almirón⁺ + Researcher of Consejo Nacional de Investigaciones Científicas y Técnicas de la República Argentina (CONICET). ; Mireya E Brewer

Centro de Investigaciones Entomológicas de Córdoba, Facultad de Ciencias Exactas, Físicas y Naturales, Universidad Nacional de Córdoba, Av. V. Sarsfield 299 (5000) Córdoba, Argentina

ABSTRACT

In order to classify mosquito immature stage habitats, samples were taken in 42 localities of Córdoba Province, Argentina, representing the phytogeographic regions of Chaco, Espinal and Pampa. Immature stage habitats were described and classified according to the following criteria: natural or artificial; size; location related to light and neighboring houses; vegetation; water: permanence, movement, turbidity and pH. Four groups of species were associated based on the habitat similarity by means of cluster analysis: Aedes albifasciatus, Culex saltanensis, Cx. mollis, Cx. brethesi, Psorophora ciliata, Anopheles albitarsis, and Uranotaenia lowii (Group A); Cx. acharistus, Cx. quinquefasciatus, Cx. bidens, Cx. dolosus, Cx. maxi and Cx. apicinus (Group B); Cx. coronator, Cx. chidesteri, Mansonia titillans and Ps. ferox (Group C); Ae. fluviatilis and Ae. milleri (Group D). The principal component analysis (ordination method) pointed out that the different types of habitats, their nature (natural or artificial), plant species, water movement and depth are the main characters explaining the observed variation among the mosquito species. The distribution of mosquito species by phytogeographic region did not affect the species groups, since species belonging to different groups were collected in the same region.

Key words:Aedes - Anopheles - Culex - Mansonia - Psorophora - Uranotaenia - immature stage habitats - classification - Argentina

Large outbreaks of equine encephalitis have been documented for the temperate zone of Argentina, caused by western equine (WEE) and eastern equine encephalitis (EEE) virus (Mitchell et al. 1985, 1987). Other arboviruses of potential medical importance in Argentina include St. Louis encephalitis (SLE) virus, Rocio virus and Venezuelan equine encephalitis (VEE) virus (Sabattini et al. 1985). At present, only Aedes albifasciatus from Córdoba Province (Argentina) has been incriminated as an experimental vector of WEE virus (Avilés et al. 1990).

The ecology of arboviruses remains far obscure due to the few studies carried out in Argentina. To determine the host preferences, feeding patterns were studied for mosquitoes collected in Chaco, Santa Fe and Río Negro Provinces (Mitchell et al. 1985, 1987) and in Córdoba Province (Almirón & Brewer 1995a). The discovery of eggs, larvae and pupae of several mosquito species in natural field environments during the autumn-winter period in Córdoba Province suggested that immatures continue to develop throughout these seasons (Almirón & Brewer 1994). In addition, studies on seasonal distribution of immature stages and females were conducted in Córdoba Province (Almirón & Brewer 1995b). The species collected as females and/or larvae were, from the most to the least abundant, Cx. quinquefasciatus, Cx. apicinus, Cx. dolosus, Cx. saltanensis, Cx. bidens, Cx. brethesi, Cx. maxi, Ae. albifasciatus, Cx. spinosus, Cx. acharistus and Cx. chidesteri. Ae. albifasciatus was mainly captured in summer and fall while Culex species were collected throughout the year, although peaks of mosquito abundance were from September to March.

The World Health Organization has recommended that public health personnel concentrate on vector control, including biological control and environmental management (Rejmankova et al. 1991). The knowledge of relationships between habitats, environmental factors and occurrence of mosquito larvae is essential for an efficient application of mosquito control methods. In a coastal lowland farm in Rio de Janeiro (Brazil), Oliveira et al. (1986) collected several species of Aedes, Anopheles, Culex, Limatus, Mansonia, Phonio-myia, Psorophora, Uranotaenia and Wyeomyia, describing their immature stage habitats. The majority of species preferred to breed on the ground, specially natural sites, but they also developed in urban areas. Some species were more abundant in temporary breeding places while others occurred usually in permanent ones. Some species were collected in natural recipients, others in bromeliads, in tree-holes, in plant leaf axils, and in artificial containers. Field surveys of mosquito larval sites carried out on the Pacific coast of southern Chiapas, Mexico, found several quantified environmental factors to be promising as predictors of anopheline larval occurrence (Rejmankova et al. 1991). One of the most important factors was the vegetation that favors the remote sensing and geographic information system technologies to monitor and predict larval densities (Savage et al. 1990, Rejmankova et al. 1991, 1992, Rodriguez et al. 1993).

No research on this subject has been conducted in Argentina. Del Ponte and Blaksley (1945/1948), Prosen et al. (1960), Bachmann and Casal (1962) reported the type of breeding sites where they collected immature stages in Buenos Aires and Córdoba Provinces, and the associated species, providing no information on the habitat features. As part of the mosquito fauna studies of Córdoba Province, we carried out field surveys of possible mosquito breeding sites in order to describe them and to classify the mosquito species according to the similarity of their habitats.

MATERIALS AND METHODS

Study site - Córdoba is a mediterranean province located between 29°29'-35°1'LS and 32°54'-61°46'LW. The climate in this province is temperate (Capitanelli 1979). Freezing temperatures are recorded from April through September, July being the month with the majority of days (6.5) with frost. The mean temperatures are 17°C in the SE and 20°C in the NE. Mean annual rainfall in the area is between 380-900 mm. The principal rainy season is in March, with a secondary one from October through December. Phytogeographically, Córdoba Province is divided into different regions (Luti 1979). The Chaco region is divided into the western and eastern woodlands, and shows the greatest physiognomical variety in the province. The annual precipitation is between 550-600 mm. In the Espinal region grassland and woodland are mixed, with about 800 mm of annual precipitation. The Pampa region is a wide flat grassland, sub-humid in the east with 900 mm of annual rainfall, and semi-arid in the southwest with 500 mm.

Collections - Eggs, larvae and pupae were sampled in 42 localities which included the three phytogeographic regions (Table I). A total of 252 samples were collected and analyzed.

Immature stage habitats were classified using the following criteria: natural or artificial; size; type; location in relation to light (sunlight, partial shade and complete shade) and houses; breeding depth; water permanence (permanent, semipermanent and temporary); water movement (stagnant or in motion), although the speed was not measured; turbidity was estimated visually either as turbid or clear (turbid samples were those where the white dipper background could not be seen); pH (papers MN-Macherey-Nagel D-5160) and vegetation (present or absent).

Identification - Immature stages were identified using species descriptions and keys by Lane (1953), Forattini (1962, 1965a,b) and Darsie (1985). Identification was based on fourth instar larvae and/or adults, so eggs, first to third instar larvae and pupae collected were reared to fourth instars or adults for identification. The specimens were deposited in the mosquito collection of Centro de Investigaciones Entomológicas de Córdoba, Facultad de Ciencias Exactas, Físicas y Naturales, Universidad Nacional de Córdoba.

Vegetation sampling - A botanical survey was conducted to collect and identify the plant species in the habitats that were visited. Plants were sent to the Botanical Museum of the National University of Córdoba for identification. Plant species appearing after the preliminary survey was completed, or whose identification was in question were collected, and preserved for subsequent identification.

Data analysis - Data were analyzed according to the following steps (Crisci & López Armengol 1983): operative taxonomic units (OTUs) were chosen; a basic matrix of data (BMD) was developed; the similarity was calculated for each pair of OTUs; a similarity matrix (SM) was developed; and mosquito species were grouped.

Since the purpose was to group species with similar breeding features, the operative taxonomic units (OTUs) chosen were the species collected (Table I). The recorded characteristics were analyzed to identify common patterns of immature stage habitats where the different species were collected. Both quantitative (size, distance from breeding sites to the nearest house, pH, and depth) and qualitative (natural or artificial, type, shade, water permanence, water movement, turbidity, and vegetation) characters were codified as 1/0 (= presence/absence). Quantitative data were subdivided into intervals (see depth, for instance, in Table III), each one being considered as a character.

A BMD was developed based on the codified data, consisting of 19 rows for the mosquito species and 86 columns for all the characters. Values within the cells represented 1 or 0 if each character was recorded for each species or not respectively. This table was used for analysis to calculate the similarity for all possible pairs of OTUs, and is not presented here.

The similarity for all possible combinations of species pairs (OTUs) was calculated using the Dice or SD association coefficient (Crisci & López Armengol 1983). The coefficients obtained were used to develop the SM (Table II). OTUs were grouped by similarity using cluster analysis. The UPGMA (unweighted pair group method using arithmetic averages) was used as a linkage technique, representing the groups by a phenogram (Sneath & Sokal 1973).

The principal component analysis (ordination method) was used to examine the relationships among all the OTUs based on their similarity. The SM was used to represent all the characters by a reduced number of principal components according to a linear model, where each component contains part of the total variability of the characters (Crisci & López Armengol 1983). The NTSYS program (Rohlf 1988) was used for this analysis.

RESULTS AND DISCUSSION

Four groups of species can be distinguished in the phenogram (Fig.) according to the association coefficients obtained (Table II): Ae. albifasciatus, Cx. saltanensis, Cx. mollis, Cx. brethesi, Ps. ciliata, An. albitarsis and Ur. lowii (Group A), Cx. acharistus, Cx. quinquefasciatus, Cx. bidens, Cx. dolosus, Cx. maxi and Cx. apicinus (Group B), Cx. coronator, Cx. chidesteri, Ma. titillans and Ps. ferox (Group C), Ae. fluviatilis and Ae. milleri (Group D). Cx. coronator and Cx. chidesteri showed the highest association coefficient (0.785), being grouped together (Group C). The lowest coefficient (0.160) was for Cx. mollis and Ae. milleri (Table I), so they were included in the group A and D respectively (Fig.). Species of group B are clearly separated from the other groups. Species of groups A, B and C were collected from two or three phytogeographic regions (Table I), so their geographical distribution did not affect the formation of the groups.

The Group A species were mainly collected in natural habitats, though Ae. albifasciatus, Cx. saltanensis and An. albitarsis were also found in artificial containers (Table IV). Ground pools ranging in size from small to large represent the principal habitat of this group. Aquatic plants were either present or absent. Other characteristics included: either shade or sunlight; either close to far away from houses; either clear or turbid water; a range of pH from 6.4-8.0; shallow; stagnant water; and either temporary or permanent water source. Ae. albifasciatus-Cx. saltanensis and Cx. brethesi-Ps. ciliata form the two clusters that share the greatest number of habitat features (Fig.).

The species of group B were exclusively in the genus Culex. They were found both in natural and artificial habitats and in a great variety of small to large places. Aquatic vegetation was an important feature, although it was not found in every site. Habitats were located in either shade or sunlight, both close to and far away from houses. Temporary and permanent habitats showed clear to turbid water, pH (6.4->8), and depths 0.03-1.8 m. All species were collected in stream margins except Cx. bidens, those species being found in slowly-moving water in the presence of aquatic plants. This group shows higher uniformity of habitat features compared with groups A and C (Table III). Two nuclei can be recognized: Cx. acharistus-Cx. quinquefasciatus and Cx. bidens-Cx. dolosus. The differences between the nuclei are found in breeding type, vegetation and degree of illumination. Cx. maxi is closely related to the second nucleus sharing more types of habitats, while Cx. apicinus is clearly separated from the other species (Fig.). Cx. apicinus and Cx. quinquefasciatus were found more frequently (67 and 71% respectively) in artificial containers (Table III). The type of habitat, size and location (closely related to human environment) distinguish Cx. apicinus from the other species.

The Group C species were collected only in natural habitats, medium to large in size, mainly ground pools, with or without aquatic plants, partially illuminated, relatively close to houses; in clear to turbid water, pH between 7-7.6, 0.1-0.2 m depth, always stagnant; and in temporary to permanent places. The Culex species in Group C form a nucleus which displays the highest similarity coefficient (0.785) (Fig.).

Larvae of Ae. fluviatilis and Ae. milleri (Group D) were only found in small artificial containers, without vegetation, in partial to complete shade, with clear, stagnant water, pH 6.7-6.8, 0.2 m depth, and in temporary habitats located in a house.

Ae. albifasciatus, Cx. acharistus, Cx. apicinus, Cx. bidens, Cx. brethesi, Cx. dolosus, Cx. maxi and Cx. quinquefasciatus were often collected together and with other species (Table IV); they were also found as unique species from a single sample. The most frequent associations (based on 252 samples) were Cx. acharistus-Cx. dolosus (11.9%), followed by Cx. apicinus-Cx. quinquefasciatus, and Cx. dolosus-Cx. quinquefasciatus (2.77% for both combinations), and Cx. bidens-Cx. maxi (2.38%). All these species belong to group B. Except for Cx. bidens, the other species of this group were found at least once all together in the same habitat. The least frequent associations were Ae. albifasciatus-Ps. ciliata, Ae. fluviatilis-Cx. acharistus-Cx. apicinus-Cx. dolosus-Cx. quinque-fasciatus, An. albitarsis-Cx. dolosus-Cx. maxi-Cx. saltanensis, and Cx. bidens-Cx. brethesi-Ur. lowii, found only once in these combinations. To find species of different genera in the same habitat suggests a similarity of habitat requirements.

According to the principal component analysis, the first three components explain 52.4% of the variation observed among the different mosquito species. The type of breeding place, aquatic plants, habitat nature (artificial or natural), and water movement contributed the most to explain the variation in the first component. Water depth is the most important character in the second component (Table III). Type of breeding place and vegetation are the most important in the third component. Relationships between vegetation and immature stages were reported by several authors (Savage et al. 1990, Rejmankova et al. 1991, 1992, Rodriguez et al. 1993), who investigated spatial and seasonal variations on anopheline larval densities and their plant associations in Mexico, finding that larval abundance was related to the presence of certain types of vegetation.

Ae. albifasciatus has also been found in swamps (Del Ponte & Blaksley 1945/1948), in hypersalty water (NaCl 31.7-49.8 g/l, Na₂SO₄ 11.32-19.44 g/l, CaCO₃ 7.5-12.8 g/l) with the aquatic plant Atriplex (Bachmann & Casal 1962), and in habitats at 2,300 m above sea level (Forattini 1965a), displaying considerable diversity in habitats. Our data on Cx. saltanensis agree with those reported by Oliveira et al. (1986) who found immature stages in partially shaded natural and artificial habitats, with clear to turbid water.

Very little is knwon about the habitat of the immature stages of Cx. brethesi. Bachmann and Casal (1962) found it breeding in hypersalty water (NaCL 2.5 g/l, CaCo₃ 730 mg/l). Ps. ciliata larvae were also found in Bromeliaceae feeding on mosquito larvae of Aedes, Ae. taeniorhynchus, Culex, Ps. (Janthinosoma), Ps. ferox and Ps. confinnis (Prosen et al. 1962/1963, Forattini 1965a).

Cx. mollis was collected in permanent natural ground pools in our study. However, it can also be found in temporary habitats, Bromeliaceae, drinking places, tanks, barrels, cans, fountains, banana tree leaf axils and small pools, associated with Cx. coronator and some species of Aedes (Forattini 1965a, Clark-Gil & Darsie 1983).

An. albitarsis can breed in several habitats such as lagoons, dikes, drains, swamps, floodlands, stream margins, in stagnant or slow-moving water, with abundant or scarce debris, in fresh or salty water, and associated with one or more of the following: An. albimanus, An. argyritarsis, An. darlingi, An. noroestensis, An. pseudopuncti-pennis, An. strodei, and An. triannulatus (Deane et al. 1948, Forattini 1962).

Ur. lowii was collected in the margins of lakes, lagoons, in very sunny places, usually with aquatic plants, and associated with An. albimanus and Cx. coronator (Carpenter & La Casse 1955, Clark-Gil & Darsie 1983). Oliveira et al. (1986) found larvae in a small plastic receptacle. We also collected this species in shaded habitats, without vegetation.

Cx. acharistus and Cx. quinquefasciatus are closely related species (similarity coefficient = 0.765) (Fig.), though acharistus was mainly collected in natural habitats (76%) and quinquefasciatus in artificial containers (71%) (Table III). No information was found on Cx. acharistus to compare with our results. Our data on quinquefasciatus agree with those by Dyar (1922), Del Ponte and Blaksley (1945/1948), Carpenter and La Casse (1955), Prosen et al. (1960), Forattini (1965a), Ishii and Sohn (1987), Kulkarni and Naik (1989); these authors also found this species in barrels, dikes, drains, sewages, and less frequently in crabholes, bamboo internodes and Bromeliaceae.

Cx. dolosus and Cx. bidens are members of a nucleus. The first species was collected in artificial containers and stream margins whereas Cx. bidens preferred ground habitats with aquatic plants. This observation is in concordance with that reported by Oliveira et al. (1986).

According to Forattini (1965a) Cx. apicinus can breed in highlands, in small stream bed pools. In our study, this species was mainly collected in artificial receptacles (67%), domestic or peridomestic, including domiciliary water tanks. Based on these data this species can be considered domestic as well, at least in these regions of Argentina.

Cx. coronator and Cx. chidesteri were collected in ground pools and swamps. However, the first species has also been found in artificial containers, drains, hoof-prints, tree holes, Bromeliaceae, fruit shells, in clear to polluted water, shaded to sunny places, and associated with some species of Aedes, Anopheles, Culex and Ps. confinnis (Carpenter & La Casse 1955, Forattini 1965a, Clark-Gil & Darsie 1983). Larvae of Cx. chidesteri were also collected in lagoons, ponds, cement pools, streams, breeding habitats with or without plants, and associated with An. argyritarsis and An. pseudopunctipennis (Forattini 1965a, Clark-Gil & Darsie 1983, Oliveira et al. 1986).

Ma. titillans is close to the Cx. coronator-Cx. chidesteri nucleus because they were all collected in large breeding habitats. Ma. titillans was found associated with Pistia sp., and in addition it can also be found in ground pools with Araceae, in dikes, and in stream margins attached to Eicchornia, Ipomea and grass roots (Prosen et al. 1960, Forattini 1965b).

Although in the present study Ps. ferox was collected only in a small ground pool without vegetation, it can also be found in swamps with abundant aquatic plants, shady places or breeding habitats where the water flow is slow (Forattini 1965a).

According to Anduze (1941), Forattini and Rabello (1960), and Forattini (1965a), Ae. fluviatilis breeds in artificial containers (barrels, metallic containers, flower vases in cemeteries) and tends to be domestic. Our findings agree with these data. In addition, we collected this species in natural breeding habitats consisting of fresh water contained in rock holes, and sunny places in stream margins in common with Dyar's (1922) findings. It has been found associated with Ae. aegypti in artificial receptacles (Anduze 1941). Ae. milleri was collected in an artificial container, but larvae of this species can be also found in fresh slow-moving water (Prosen et al. 1960) and in rock holes (Forattini 1965a).

The analysis of habitat features in the present study grouped the mosquito species into four clusters, two of them included species from different genera, suggesting a similarity of habitat requirements. Based on this supposition, finding one species at one habitat it is possible to predict what other species could be found in such habitat.

ACKNOWLEDGEMENTS

To Dr David Gorla and Marta Sabattini for their suggestions; to Dr Rosa Subils for identifying the aquatic plants.

Received 13 March 1995

Accepted 28 August 1995

This work was supported in part by CONICOR (Consejo de Investigaciones Científicas y Tecnológicas de Córdoba).

Publication Dates

History