Medical and Veterinary Entomology (1991) 5 , 403-41 1
Sandfly distribution and abundance in a tropical rain forest J A N E M E M M O T T Department of Pure and Applied Biology, University of Leeds Abstract. 1. The distribution patterns of sandflies (Diptera: Psychodidae) upon tree buttresses were studied in tropical rain forest at Finca la Selva in the Caribbean lowlands of Costa Rica. 2. Four species of sandfly, Lutzornyia shannoni Dyar, L. trapidoi F. & H., L.ylephiletor F. & H. and L.vespertilionis F. & H. comprising 97% of those caught, used tree buttresses as diurnal resting sites. Their distribution on the buttresses was aggregated. 3. During the dry season tree species had no significant effect upon the distribution of the sandflies. However, during the wet season the distributions of two of the species, L. trapidoi and L.ylephiletor, were significantly affected by the species of tree; it is suggested that some species of tree may provide greater protection from rainfall than others. 4. L. vespertilionis was restricted to a single buttress on each positive tree. Distribution of this species is evidently determined by the distribution of its host animal, the bat (Emballonuridae). Female flies feed upon the bat’s blood and male flies may be attracted to the bat as it provides a source of female sandflies. It is suggested, therefore, that tree buttresses serve as sandfly swarming sites. 5. Within a large buttress the sandflies are not randomly distributed but are aggregated in particular areas. Within these aggregations, the sandflies are vertically zoned upon the buttress with a shift in species composition with height. Two hypothesis were suggested to account for this distribution pattern: a response to an environmental gradient or an interaction between the four species of fly.
Key words. Lutzomyia, sandfly, tree buttress, aggregation, swarming, tropical rain forest, Costa Rica. Introduction
Sandflies (Diptera: Psychodidae: Phlebotominae) are best known for their bites and their role as vectors of the zoonotic disease leishmaniasis. A great deal is known about sandflies from the perspective of leishmaniasis transmission, but some facets of their basic ecology suffer from Correspondence: Dr Jane Memmott, Department of Entomology, The Natural History Museum, Cromwell Road, London SW7 SBD.
neglect. For example, during daylight hours sandflies can be found in a variety of resting sites. In tropical rain forest these sites comprise hollow trees, animal burrows, low shrubs and tree buttresses, each with a different spectrum of sandfly species (Christensen et al., 1983). Tree buttresses are the most accessible of these diurnal resting sites and harbour great numbers of flies. The starting point of this investigation was the field observation that there is considerable variation in the use of trees as resting sites by sandflies. 403
404
Jane Menirnotr
Three aspects of the distribution of sandfly populations have come under consideration in the past. Firstly the distribution of sandfly species upon diurnal resting sites. Secondly the distribution of nocturnal flying heights (Williams, 1968; Porter & Defoliart, 1981; Arias & de Frietas, 1982), one of the factors which determines the height at which sandflies bite and thereby transmit leishmaniasis. The third aspect concerns the vertical stratification of sandfly species on tree buttress resting sites (Chaniotis et al., 1972). N o suggestions were found in the literature as to why a zonation of species is a feature of sandfly aggregations on tree buttresses. The aim of previous studies of sandfly diurnal resting sites had largely been to form a basis for making better, more efficient and selective capture of vector species (Chaniotis et a!., 1972). Little effort has been made to understand the reasons why sandfly species use specific diurnal resting sites. In contrast to previous sandfly studies, this one aims to explore and interpret the patterns in the distribution and abundance of adult sandflies found on tree buttresses in a tropical rain forest. This is approached by systematically examining the distribution of flies at four levels: the overall distribution of flies o n trees in the rain forest, the effect of tree species upon the flies distribution, the distribution of flies on a single tree and the distribution of flies on individual tree buttresses. Hypotheses will be suggested and discussed once patterns are established. A better understanding of sandfly ecology may be of use when considering how best to control leishmaniasis in tropical rain forest environments.
Methods
The study was conducted over a 3-year period at Finca la Selva. an Organisation for Tropical Field Studies field station located in the Caribbean lowlands of Costa Rica. Fieldwork was undertaken during three field seasons: February to July 1986, March to August 1987 and April to August 1988. Holdridge eta/. (1971) describe the site in detail with reference to its vegetation, soil types, abiotic factors and history. Sandflies are most abundant on the basal (lower) 2 m of a buttress (Chaniotis eta/., 1972) and mechanical aspirators were used to collect
sandflies from this area. The flies were cleared and mounted in Berlese medium. Diagnostic features of mounted specimens are described in Murillo & Zeledon (1985). Four aspects of the distribution of sandflies in a rain forest were examined. (1) Overall distribution. To gain an overview of the pattern of sandfly distribution in a rain forest, sandflies were collected from sixty-eight emergent or canopy level trees. All the trees sampled had a diameter at breast height greater than 700cm. Five species of tree were sampled: Hieronyma oblongata (twenty-one trees), Dipteryx panamemis (nineteen trees), Lecythis costaricensis (nineteen trees), Ceiba perztanrira (six trees) and Sloaniu species (three trees). All these trees except Lecythis have large buttresses. Lecythis with its non-existent to very small buttresses was used as a means of testing whether sandflies are choosing large buttresses or large trees as diurnal resting sites. Sandflies were collected from all sixty-eight trees once during the wet season and once during the dry season for the first year of the study. ( 2 ) The effect of tree species on the distribution of sandfies. One factor which could influence the distribution of sandflies is an affinity for a particular species of tree. To test for this possibility, Hieronyma, Dipteryx and Lecythis were tested for evidence of preferential use by sandflies. Ceiba and Sloatiia were not included in this test due to their small sample size. ( 3 ) Distribution upon a single tree. Field observations indicated that the various buttresses of a tree were not used t o an equal extent by sandflies. To determine how the flies distribute themselves upon a tree, sandflies were collected weekly from the buttresses of a Ceiba pentarzdra over a period of 14 weeks. (4) Distribution within a buttress. A single buttress of a tree can be a considerable size, in some cases several metres in both height and width. Two methods were used to investigate the distribution of sandflies upon individual buttresses. Firstly, a grid of l m squares was chalked onto two buttresses and the flies collected separately from each square. Secondly, sampling transects were used to investigate the detailed stratification upon the bottom 2 m of the buttress. Transects, 1m in width and 2 m in height, were chalked onto seventeen buttresses. The distribution of sandflies with respect t o height was investigated by dividing the transect
Sandfly distribution and abundance
into five horizontal zones: 0-25 cm, 26-50cm, 51-75 cm, 76-100cm and 101-200cm, the uppermost being widest as the flies were relatively scarce in this area of the buttress.
Results (1) Overall distribution of Lutzornyia spp.
Sandflies are aggregated in their distribution on trees during both the wet and the dry season (Fig. 1). Many trees have some sandflies upon their buttresses, but only a few of the trees harbour large numbers of flies. The variance is considerably larger than the mean, indicating that the flies are aggregated rather than randomly distributed (Table 1). Four species of sandfly were commonly found upon tree buttresses at La Selva: Lutzornyiu shannoni Dyar 1929, Lutzornyia trupidoi Fairchild & Hertig 1952, Lutzornyia ylephiletor Fairchild & Hertig 1952 and Lutzornyia vespertilionis Fairchild &
405
Table 1. The variance:mean ratios of the observed frequency distribution of sandflies on trees during the wet and dry season. ~~
Wet season Dry season Mean no. of flies per tree 33.44 Variance 5243.20 Variancehean 156.90 SE of ratio 2/(n - 1) 0.173 t = (obs - exp)/SE 906.90 Probability PiO.001
10.77 208.80 19.40 0.173 106.30 P
Sandfly distribution and abundance in a tropical rain forest J A N E M E M M O T T Department of Pure and Applied Biology, University of Leeds Abstract. 1. The distribution patterns of sandflies (Diptera: Psychodidae) upon tree buttresses were studied in tropical rain forest at Finca la Selva in the Caribbean lowlands of Costa Rica. 2. Four species of sandfly, Lutzornyia shannoni Dyar, L. trapidoi F. & H., L.ylephiletor F. & H. and L.vespertilionis F. & H. comprising 97% of those caught, used tree buttresses as diurnal resting sites. Their distribution on the buttresses was aggregated. 3. During the dry season tree species had no significant effect upon the distribution of the sandflies. However, during the wet season the distributions of two of the species, L. trapidoi and L.ylephiletor, were significantly affected by the species of tree; it is suggested that some species of tree may provide greater protection from rainfall than others. 4. L. vespertilionis was restricted to a single buttress on each positive tree. Distribution of this species is evidently determined by the distribution of its host animal, the bat (Emballonuridae). Female flies feed upon the bat’s blood and male flies may be attracted to the bat as it provides a source of female sandflies. It is suggested, therefore, that tree buttresses serve as sandfly swarming sites. 5. Within a large buttress the sandflies are not randomly distributed but are aggregated in particular areas. Within these aggregations, the sandflies are vertically zoned upon the buttress with a shift in species composition with height. Two hypothesis were suggested to account for this distribution pattern: a response to an environmental gradient or an interaction between the four species of fly.
Key words. Lutzomyia, sandfly, tree buttress, aggregation, swarming, tropical rain forest, Costa Rica. Introduction
Sandflies (Diptera: Psychodidae: Phlebotominae) are best known for their bites and their role as vectors of the zoonotic disease leishmaniasis. A great deal is known about sandflies from the perspective of leishmaniasis transmission, but some facets of their basic ecology suffer from Correspondence: Dr Jane Memmott, Department of Entomology, The Natural History Museum, Cromwell Road, London SW7 SBD.
neglect. For example, during daylight hours sandflies can be found in a variety of resting sites. In tropical rain forest these sites comprise hollow trees, animal burrows, low shrubs and tree buttresses, each with a different spectrum of sandfly species (Christensen et al., 1983). Tree buttresses are the most accessible of these diurnal resting sites and harbour great numbers of flies. The starting point of this investigation was the field observation that there is considerable variation in the use of trees as resting sites by sandflies. 403
404
Jane Menirnotr
Three aspects of the distribution of sandfly populations have come under consideration in the past. Firstly the distribution of sandfly species upon diurnal resting sites. Secondly the distribution of nocturnal flying heights (Williams, 1968; Porter & Defoliart, 1981; Arias & de Frietas, 1982), one of the factors which determines the height at which sandflies bite and thereby transmit leishmaniasis. The third aspect concerns the vertical stratification of sandfly species on tree buttress resting sites (Chaniotis et al., 1972). N o suggestions were found in the literature as to why a zonation of species is a feature of sandfly aggregations on tree buttresses. The aim of previous studies of sandfly diurnal resting sites had largely been to form a basis for making better, more efficient and selective capture of vector species (Chaniotis et a!., 1972). Little effort has been made to understand the reasons why sandfly species use specific diurnal resting sites. In contrast to previous sandfly studies, this one aims to explore and interpret the patterns in the distribution and abundance of adult sandflies found on tree buttresses in a tropical rain forest. This is approached by systematically examining the distribution of flies at four levels: the overall distribution of flies o n trees in the rain forest, the effect of tree species upon the flies distribution, the distribution of flies on a single tree and the distribution of flies on individual tree buttresses. Hypotheses will be suggested and discussed once patterns are established. A better understanding of sandfly ecology may be of use when considering how best to control leishmaniasis in tropical rain forest environments.
Methods
The study was conducted over a 3-year period at Finca la Selva. an Organisation for Tropical Field Studies field station located in the Caribbean lowlands of Costa Rica. Fieldwork was undertaken during three field seasons: February to July 1986, March to August 1987 and April to August 1988. Holdridge eta/. (1971) describe the site in detail with reference to its vegetation, soil types, abiotic factors and history. Sandflies are most abundant on the basal (lower) 2 m of a buttress (Chaniotis eta/., 1972) and mechanical aspirators were used to collect
sandflies from this area. The flies were cleared and mounted in Berlese medium. Diagnostic features of mounted specimens are described in Murillo & Zeledon (1985). Four aspects of the distribution of sandflies in a rain forest were examined. (1) Overall distribution. To gain an overview of the pattern of sandfly distribution in a rain forest, sandflies were collected from sixty-eight emergent or canopy level trees. All the trees sampled had a diameter at breast height greater than 700cm. Five species of tree were sampled: Hieronyma oblongata (twenty-one trees), Dipteryx panamemis (nineteen trees), Lecythis costaricensis (nineteen trees), Ceiba perztanrira (six trees) and Sloaniu species (three trees). All these trees except Lecythis have large buttresses. Lecythis with its non-existent to very small buttresses was used as a means of testing whether sandflies are choosing large buttresses or large trees as diurnal resting sites. Sandflies were collected from all sixty-eight trees once during the wet season and once during the dry season for the first year of the study. ( 2 ) The effect of tree species on the distribution of sandfies. One factor which could influence the distribution of sandflies is an affinity for a particular species of tree. To test for this possibility, Hieronyma, Dipteryx and Lecythis were tested for evidence of preferential use by sandflies. Ceiba and Sloatiia were not included in this test due to their small sample size. ( 3 ) Distribution upon a single tree. Field observations indicated that the various buttresses of a tree were not used t o an equal extent by sandflies. To determine how the flies distribute themselves upon a tree, sandflies were collected weekly from the buttresses of a Ceiba pentarzdra over a period of 14 weeks. (4) Distribution within a buttress. A single buttress of a tree can be a considerable size, in some cases several metres in both height and width. Two methods were used to investigate the distribution of sandflies upon individual buttresses. Firstly, a grid of l m squares was chalked onto two buttresses and the flies collected separately from each square. Secondly, sampling transects were used to investigate the detailed stratification upon the bottom 2 m of the buttress. Transects, 1m in width and 2 m in height, were chalked onto seventeen buttresses. The distribution of sandflies with respect t o height was investigated by dividing the transect
Sandfly distribution and abundance
into five horizontal zones: 0-25 cm, 26-50cm, 51-75 cm, 76-100cm and 101-200cm, the uppermost being widest as the flies were relatively scarce in this area of the buttress.
Results (1) Overall distribution of Lutzornyia spp.
Sandflies are aggregated in their distribution on trees during both the wet and the dry season (Fig. 1). Many trees have some sandflies upon their buttresses, but only a few of the trees harbour large numbers of flies. The variance is considerably larger than the mean, indicating that the flies are aggregated rather than randomly distributed (Table 1). Four species of sandfly were commonly found upon tree buttresses at La Selva: Lutzornyiu shannoni Dyar 1929, Lutzornyia trupidoi Fairchild & Hertig 1952, Lutzornyia ylephiletor Fairchild & Hertig 1952 and Lutzornyia vespertilionis Fairchild &
405
Table 1. The variance:mean ratios of the observed frequency distribution of sandflies on trees during the wet and dry season. ~~
Wet season Dry season Mean no. of flies per tree 33.44 Variance 5243.20 Variancehean 156.90 SE of ratio 2/(n - 1) 0.173 t = (obs - exp)/SE 906.90 Probability PiO.001
10.77 208.80 19.40 0.173 106.30 P
