Mosquito species occurrence in association with landscape composition in green urban areas

Ocorrência de espécies de mosquitos associada à composição de paisagem em áreas verdes urbanas

Abstract

Aedes aegypti prefers densely populated habitats, but has been shown to explore less anthropogenic environments. We investigated composition of the abundance of mosquitoes in forested areas and assessed relationships between species occurrences and different types of land use and land cover at three spatial scales (100m, 500m and 1000m). Mosquitoes were collected from October 2012 to March 2013 using oviposition traps. We collected 4,179 mosquitoes in total including at least 10 species. Aedes albopictus and Limatus durhami were eudominant species, representing 90% of all collected individuals. We found intraspecific differences in response to land use and land cover, and species response patterns were similar at all spatial scales. Ae. albopictus relative abundance was associated with urbanized areas, while Li. durhami, Haemagogus leucocelaenus and Toxorhynchites sp., abundances were associated with native forest. Aedes aegypti were found in five of the eight areas studied, including in an Atlantic forest fragment at a considerable distance from the forest edge (370 m). Aedes aegypti occurrence was not influenced by type of land use or land cover.

Keywords:
Aedes aegypti; mosquito; oviposition trap; urbanization; landscape

Resumo

Aedes aegypti tem como habitat preferencial áreas densamente povoadas, mas que tem se mostrado capaz de explorar ambientes menos antropizados. O objetivo desse trabalho foi investigar a abundância de mosquitos em áreas florestadas e avaliar a relação dos diferentes tipos de uso e cobertura do solo, em três escalas de paisagem (buffer de 100m, 500m e 1000m). Os mosquitos foram coletados entre outubro de 2012 a março de 2013 utilizando armadilhas de oviposição. Foram coletados 4,179 culicídeos, de pelo menos 10 espécies, destas, Aedes albopictus e Limatus durhami foram eudominantes, representando 90% do total de indivíduos coletados. As espécies de culicídeos respondem de maneira diversa aos tipos de uso e cobertura do solo, indicando que as afinidades mais evidentes se mantém nas três escalas avaliadas. A abundância relativa de Ae. albopictus se mostrou relacionada às áreas mais urbanizadas, enquanto a de Li. durhami, Haemagogus leucocelaenus e Toxorhynchites sp., à presença da mata nativa. Aedes aegypti esteve presente em cinco das oito áreas estudadas, incluindo um fragmento de Mata Atlântica a uma distância considerável da borda (370m). A ocorrência de Aedes aegypti nas áreas analisadas não demonstrou ser influenciada pelo tipo de uso e cobertura do solo.

Palavras-chave:
Aedes aegypti; mosquito; armadilha oviposição; urbanização; paisagem

1. Introduction

Urbanization can alter the environment in ways that impact mosquito species diversity by reducing abundances or even causing local extinction of some species. A decline in population numbers of a particular species results in an empty ecological niche that may become available for invasion by other species. These new species may or may not be functionally equivalent and can potentially be harmful, possibly resulting in outbreaks of infectious diseases. Some mosquito species have high genetic and ecological plasticity and, driven by anthropogenic pressure, easily adapt to new environmental conditions. This is particularly true in urban areas neighboring green areas, which often contain an abundance of artificial containers for oviposition and ample blood sources (i.e., humans and pets), potentially leading to changes in species natural habitats (Lopes, 1997).

Among the Culicids that benefit from urban sprawl, Aedes aegypti is notorious due to its role as a vector of arboviruses such as Zika, dengue, and chikungunya, causing a strong impact on human health. This is an anthropophilic and synanthropic species, well known for exploring and occupying a diversity of microhabitats in intensely urbanized environments. Aedes aegypti habitat preferences are generally well understood, and this knowledge is fundamental for the development of effective control measures. Although this species exhibits strong preference for urban environments immature forms have been found in forested areas (Lourenço-de-Oliveira et al., 2003; Medeiros et al., 2009; Zequi et al., 2005), environments that are traditionally not covered by municipal vector control programs.

Understanding how landscape changes influence Culicidae species composition and abundance may help us to predict spatial distributions of target species under urbanization, potentially providing important insights into urban planning and prevention of disease epidemics. We investigated Aedes aegypti occurrence in forest areas in Porto Alegre, Brazil. We specifically evaluated potential links between types of land use and land cover (LULC) and container species composition and abundance at three spatial scales.

2. Material and Methods

Mosquitoes were collected from eight forested areas in Porto Alegre, Rio Grande do Sul, Brazil. The collection areas are distributed along an urbanization gradient (i.e., from municipal center to more rural areas) (Figure 1 and Table 1). Study areas were classified as urban fragments, green areas, and remnant forest fragments according to Maciel and Barbosa (2015). Green areas (4) included the Parque Farroupilha (PF; 22J, 478925.00, 6677169), the Parque Marinha do Brasil (PM; 22J, 477655.89, 6675151.10), the Parque Knijnik (PK; 22J 480451.90, 6669719.02) and the Jardim Botânico (JB; 22J 482989.01, 6675651.87); these areas have a park-like structure formed by groves of scattered trees, continuous canopy, and field. Grupo Silk (GS; 22J 489947.02, 6679643.06) is a partially preserved urban, native forest fragment surrounded by densely populated area. Morro Santana (MS; 487932.01, 6673829.11) and Morro São Pedro (SP; 22J, 490116.00, 6661360.00) are Atlantic forest remnants in a permanent preservation area in granitic hills, the latter being the largest native forest fragment in Porto Alegre. The Reserva Biológica do Lami José Lutzenberger (RL; 22J, 490759.04, 6655009.08) is a municipal biological reserve of native Restinga forest fragment.

Figure 1
Map of the study area (highlighting the green areas) evaluated in Porto Alegre, RS, Brazil between October 2012 and March 2013.

Table 1
Original land use and land cover (LULC) classes as described by Hasenack (2008), and reconstructed (grouped) classes used for analyses. See Methods for more details.

Original Classes	Reconstructed Classes
Permanent Cropland	Woodland
Woody Vegetation
Shrubland
Degraded Forest Land
Wetland	Grassland
Temporary Cropland
Pasture
Native Grasslands
Residential	Urban
Isolated Houses
Rural Housing
Buildings
Buildings and Houses
Pavilions
Spontaneous Occupation
Rocky Outcrops	Inhospitable
Pavement
Bare Soil
Water
Native Forest Land	Native Forest

Oviposition traps were constructed from 2 L, plastic soda bottles with side openings (15 cm × 8.5 cm) cut at the label level. Trap exteriors were spray-painted with matte black paint (Chemicolor®). A 15 × 5 cm strip of hardboard (Eucatex®) was placed inside each bottle, partially immersed in 600 mL of distilled water. We installed three traps in each of the eight areas at ground level, spaced one meter away from each other in a triangular layout.

Containers were sampled quarterly from October 2012 to March 2013. Oviposition traps were left in the field throughout the experimental period. During collections the content of the traps was filtered, larvae and pupae were removed, and the water was returned to the trap. The water volume was replenished to the original 600 mL if necessary. Fourth-instar larvae were sacrificed immediately after collection and stored for later identification. First to third instar larvae were raised in the laboratory in 135 mL plastic containers with distilled water. Rearing containers were covered with voile fabric, and larvae were fed fish food (Tatramin®) until reaching the forth instar. Pupae were kept in distilled water until they emerged. Adult and fourth-instar larval identification was carried out using taxonomic keys by Lane (1953) and Forattini (2002). Genera were abbreviated according to Reinert (2009).

Landscape analyses were performed using LULC maps from the Environmental Assessment of Porto Alegre (Hasenack, 2008), which is based on QuickBird® imaging. Briefly, for each sample area we calculated the LULC class percentages at three different spatial scales, defined by circular buffers around each triangular trap arrangement with radii measuring 100 m, 500 m, and 1000 m. Culicids typically do not disperse further than 1000 m from the breeding site, which informed our choices for experimental buffer radius. The twenty original LULC classes were grouped into five classes considered relevant for this study (see Table 1). The class “water” was reclassified as “inhospitable”, as the sampled Culicidae did not use these bodies of “water” sensuHasenack (2008) for oviposition. GIS analyses were performed using ArcView 3.2 software (ESRI).

We followed the classification proposed by Cardoso et al. (2011) to determine the degree of constancy of species: ‘constant’ species were present in more than 50% of the samples, ‘accessory’ species were present in 25 to 50% of the samples, and ‘incidental’ species appeared in less than 25% of the samples. Dominance (Dom.) was defined as the relative abundance of each species in the sample, where: eudominant ≥ 10%, dominant 5 ≤ 10%, subdominant 2 ≤ 5%, recessive 1 ≤ 2% and rare <1% (Cardoso et al., 2011).

We used redundancy analysis (RDA) to explore the association between the composition and abundance of container-inhabiting Culicidae species and landscape structure at three different spatial scales. Abundance data were Hellinger-transformed (Peres-Neto et al., 2006; Legendre and Legendre, 2012). Analyses were performed using R software (R Core Team, 2012).

3. Results

We collected a total of 4,179 mosquitoes belonging to six genera, including at least ten species. Aedes albopictus (Skuse, 1894) and Limatus durhami (Theobald, 1901) had the highest relative abundances and occurred in all surveyed areas (Table 2). Ae. albopictus was the most dominant species, and was constant in all areas. Limatus durhami was incidental at PM, accessory at PK and PF, and constant in the remaining areas. Haemagogus leucocelaenus (Dyar and Shannon, 1924) was constant at SP, and Toxorhynchites sp. was constant at PK and SP. Culex quinquefasciatus and Sabethes albiprivus (Theobald, 1903) were considered incidental; only one individual was collected for the latter.

Table 2
Numbers of collected specimens, dominance (Dom.), species constancy (A-C), and relative abundance (%) of Culicids captured with oviposition traps in eight green areas in Porto Alegre, RS, Brazil; (A) = Incidental; (B) = Acessory; (C) = Constant.

	GS	PF	PM	JB	MS	SP	RL	PK	Total	Dom.	%	Acr.
Aedes albopictus	643 (C)	470 (C)	649 (C)	223(C)	44 (C)	96 (C)	259 (C)	105 (C)	2,489	EUD	59.56	Ae.alb.
Limatus durhami	73 (C)	4 (A)	25 (B)	398 (C)	254 (C)	106 (C)	401 (C)	8 (A)	1,269	EUD	30.37	Lm.dur.
Aedes aegypti	8 (B)	35 (C)	28 (A)	5 (A)	17 (A)				93	SUB	2.23	Ae.aeg.
Aedes sp.						2 (A)			2	RAR	0.05	Ae.sp.
Aedes fluviatilis			20 (B)	6 (A)		2 (A)			28	RAR	0.67	Ae.flu.
Culex dolosus				10 (A)	26 (A)			69 (A)	105	SUB	2.51	Cx.dol.
Culex quinquefasciatus			34 (A)						34	RAR	0.81	Cx.qui.
Haemagogus leucoceleanus				3 (A)	15 (B)	84 (C)			102	SUB	2.44	Hg.leu.
Toxorhynchites sp.	13 (B)	1 (A)	2 (A)	3 (A)	7 (B)	12 (C)	5 (B)	13 (C)	56	REC	1.34	Tox.sp.
Sabethes albiprivus					1 (A)				1	RAR	0.02	Sa.alb.
									4,179		100

Ae. aegypti was classified as subdominant and was found in five of the eight areas, including three green areas (PM, PF e JB), the GS urban fragment, and the permanent forest fragment (MS). The highest Ae. aegypti abundances were in PF and PM with 35 and 28 specimens, respectively. Ae. aegypti was constant in PF, accessory in GS and PM, and incidental in JB and MS.

The RDA for the 100 m scale showed that the LULC variables explained 76% of the total variation in species abundances (Figure 2). The first and second axes accounted for 55% and 13% of the variation, respectively. The RDA for the 500 m scale showed that 79% of the total variation in species abundances was explained by landscape variables, with 50% being explained by the first axis and 21% by the second (Figure 2). Finally, analysis at the 1000 m scale showed that LULC explained 73% of the total variation in species abundances, 45% by axis 1 and 21% by axis 2 (Figure 2).

Figure 2
Ordination diagram from redundancy analysis (RDA) of Culicid abundance and LULC classes at three spatial scales (100m, 500m and 1000m) around each oviposition trap site.

The first axis explained the majority of the variation in species abundances, and represents the landscape structure gradient. This gradient ranges from more preserved environments composed of complex natural structures and high vegetation cover, to urban settings with less vegetation cover, less shade, and higher rates of human movement. The pattern of associations between species and habitats holds fairly constant at all three spatial scales. GS, PM and PF appear at the right side of the RDA diagram and were associated with the variables URB, INO and NAT. MS, SP, JB and RL were situated on the left side (along the first axis), and were associated with NAT, as expected. In terms of the species’ responses to environmental variables, Li. durhami (correlation with axis 1: -0.61; -0.58; -0.53) and Hg. leucocelaenus (correlation with axis 1: -0.27; -0.33; -0.36) appear to be associated with more forested environments. In contrast, Ae. albopictus appears near URB (correlation with axis 1: -0.47; -0.42; -0.39) at all three spatial scales. Culex dolosus is strongly correlated with the second axis, in association with MS and PK.

This study documents Sa. albiprivus and Li. durhami for the first time in Porto Alegre. Sa. albiprivus was previously found in three cities in Rio Grande do Sul: Derrubadas, Garruchos and Santo Antônio (Cardoso et al., 2005), while Li. durhami has been recorded in Nova Bassano, Nova Petrópolis (Cardoso et al., 2005) and Maquiné (Cardoso et al., 2011).

Aedes albopictus was dominant in green urban areas. This species is known to be well adapted to transition zones between urban and non-urban/forest habitats. According to Forattini (2002), such habits result from being able to utilize breeding sites and blood sources from both environments. The high incidence of Ae. albopictus in our study was thus expected (Lourenço de Oliveira et al., 2004; Medeiros et al., 2009).

4. Discussion

Our analyses indicated an association of Li. durhami and Hg. leucocelaenus with non-urban/forest habitats, with Li. durhami occurrence being strongly associated with native forest. This contradicts Lopes (1997), who compared the two species and characterized Li. durhami as “strongly prone to domiciliation” based on its ability to use artificial containers for breeding. Following this logic, Hg. leucocelaenus can be considered better adapted to non-urban habitats, preferring to oviposit in tree holes and bamboo nodes (Forattini, 2002). Our choice of artificial container favored oviposition by Li. durhami, possibly influencing our results and leading to categorization of the species as “wild” compared to Hg. leucocelaenus. Capturing Sa. albiprivus with oviposition traps is unusual, as this species typically does not utilize artificial containers; it prefers to oviposit in phytotelmata such as tree hollows, bamboo internodes, bromeliads, and leaf axils (Forattini, 2002).

Toxorhynchites sp. was found in all sample areas, including those with higher urban influence. They frequently co-occurred with Ae. aegypti, Ae. albopictus, and Hg. leucocelaenus. Species belonging to this genus exhibit predatory behavior in larval stages (Forattini, 2002; Albeny et al., 2010), and have been evaluated for their potential role as biological control agents (Collins and Blackwell, 2000). However, because we did not record Toxorhynchites larval instars during the collections, we were not able to assess any negative effects on mosquito vector populations.

While Ae. aegypt’s proclivity to urban environments is well understood, our analyses did not reveal specific associations of the species with anthropogenic environments. Furthermore, in the current study it was less prone to urban habitats than Ae. albopictus, known for its affinity to peri-urban habitats. This discrepancy implies that factors related to LULC cannot fully explain Ae. aegypti occurrence in green areas of the city, and that perhaps other factors are equally or more influential.

The average maximum flight distance for female mosquitoes is around 100 m (Forattini, 2002). At the PF, JB, PM, and GS sites where the oviposition traps were installed within 100 m of populated areas, species distributions may be explained by the presence of nearby sources of human blood. At the MS site, however, traps were placed more than 370 m away from human-occupied areas, and human movement and activity within the study site was uncommon. According to Reiter et al. (1995), Ae. aegypti dispersal is driven by search for oviposition sites. The methodology employed by the city of Porto Alegre’s vector control administration, which includes breeding site suppression, could be causing Ae. aegypti to disperse to neighboring green areas in search of oviposition sites. Forattini (2002), Lourenço de Oliveira et al. (2004) and Reiter et al. (1995) also highlighted the inefficiency of such vector control strategies, as it leads to species dispersion rather than elimination.

The availability of artificial breeding sites such as trash and containers used in religious rituals, especially around forests, can also contribute to the presence of Ae. aegypti in forest patches and fragments (e.g. the MS area). According to Lourenço de Oliveira et al. (2004), forest remnants can serve as refuge for this species, harboring the necessary conditions for re-colonization of other areas in the absence of proper population control by authorities. The explicit preference for human blood as a food source and artificial containers as breeding sites can also explain habitat expansion of Ae. aegypti to forested areas adjacent to urban centers. The occurrence of Ae. aegypti at the MS site is thus likely not indicative of late adaptation to natural or native areas, but rather another aspect of its well-known opportunistic behavior in terms of seeking habitat for reproduction and shelter.

In conclusion, we observed Ae. aegypti breeding in areas not considered characteristic for the species, far from the reach of city vector control operations. We also found that mosquito species respond differently to types of land use and land cover, and that they show affinities to LULC classes that are similar at the three spatial scales evaluated. Ae. albopictus was dominant in the study. Occurrence of Hg. leucocelaenus, an important regional vector of jungle yellow fever (Cardoso et al., 2010), was associated with native forest habitat, but was also found in areas with frequent human traffic such as the Botanical Garden (JB). Knowledge from this study can contribute to urban planning and development of future strategies to limit the transmission of arboviruses in Porto Alegre.

Acknowledgements

We are grateful to Dr. Fernanda Mello for access to the LACEN/RS entomological reference collection, and to Dra. Márcia Bicudo de Paula at USP for help with species identifications. This project was funded through CAPES scholarships to F.R.G. Montagner.

• (With 2 figures)

References

• ALBENY, D.S., ROSA, C.S., ATAÍDE, L.M.S. and VILELA, E.F., 2010. Primeiro registro do gênero Toxorhynchites (Theobald) (Diptera, Culicidae) em Mata Atlântica, Viçosa, MG. Revista Ceres, vol. 57, no. 2, pp. 182-184. http://dx.doi.org/10.1590/S0034-737X2010000200007
  » http://dx.doi.org/10.1590/S0034-737X2010000200007
• CARDOSO, J.C., ALMEIDA, M.A., SANTOS, E., FONSECA, D.F., SALLUM, M.A., NOLL, C.A., MONTEIRO, H.A., CRUZ, A.C., CARVALHO, V.L., PINTO, E.V., CASTRO, F.C., NUNES NETO, J.P., SEGURA, M.N. and VASCONCELOS, P.F., 2010. Yellow Fever Virus in Haemagogus leucocelaenus and Aedes serratus mosquitoes, Southern Brazil, 2008. Emerging Infectious Diseases, vol. 16, no. 12, pp. 1918-1924. PMid:21122222. http://dx.doi.org/10.3201/eid1612.100608
  » http://dx.doi.org/10.3201/eid1612.100608
• CARDOSO, J.C., CORSEUIL, E. and BARATA, J.M.S., 2005. Culicinae (Diptera, Culicidae) ocorrentes no Estado do Rio Grande do Sul, Brasil. Brasileira de Entomologia, vol. 49, no. 2, pp. 275-287. http://dx.doi.org/10.1590/S0085-56262005000200013
  » http://dx.doi.org/10.1590/S0085-56262005000200013
• CARDOSO, J.C., PAULA, M.B., FERNANDES, A., SANTOS, E., ALMEIDA, M.A.B., FONSECA, D.F. and SALLUM, M.A.M., 2011. Ecological aspects of mosquitoes (Diptera: Culicidae) in an Atlantic forest area on the north coast of Rio Grande do Sul State, Brazil. Journal of Vector Ecology, vol. 36, no. 1, pp. 175-186. PMid:21635656. http://dx.doi.org/10.1111/j.1948-7134.2011.00155.x
  » http://dx.doi.org/10.1111/j.1948-7134.2011.00155.x
• COLLINS, L.E. and BLACKWELL, A., 2000. The biology of Toxorhynchites mosquitoes and their potential as biocontrol agents. Biocontrol News and Information, vol. 21, no. 4, pp. 105N-116N.
• FORATTINI, O.P., 2002. Culicidologia médica: identificação, biologia, epidemiologia São Paulo: EDUSP, vol. 2, 860 p.
• HASENACK, H., 2008. Diagnóstico ambiental de Porto Alegre: geologia, solos, drenagem, vegetação, ocupação e paisagem Porto Alegre: Secretaria Municipal do Meio Ambiente. 88 p.
• LANE, J., 1953. Neotropical Culicidae São Paulo: Universidade de São Paulo. 548 p.
• LEGENDRE, P. and LEGENDRE, L., 2012. Numerical ecology 2nd ed. Amsterdam: Elsevier, vol. 24, 106 p.
• LOPES, J., 1997. Ecologia de mosquitos (Diptera:Culicidae) em criadouros naturais e artificiais de área rural do Norte do Estado de Paraná, Brasil. V. Coleta de larvas em recipientes artificiais instalados em mata ciliar. Revista de Saude Publica, vol. 31, no. 4, pp. 370-377. PMid:9595766. http://dx.doi.org/10.1590/S0034-89101997000400006
  » http://dx.doi.org/10.1590/S0034-89101997000400006
• LOURENÇO DE OLIVEIRA, R., CASTRO, M.G., BRAKS, M.A.H. and LOUNIBOS, L.P., 2004. The invasion of urban forest by dengue vectors in Rio de Janeiro. Journal of Vector Ecology, vol. 29, no. 1, pp. 94-100. http://dx.doi.org/10.1603/033.046.0505 PMid:15266746.
  » http://dx.doi.org/10.1603/033.046.0505
• LOURENÇO DE OLIVEIRA, R., VAZEILLE, M., FILIPPIS, A.M.B. and FAILLOUX, A.B., 2003. Aedes albopictus from Brazil and southern United States: genetic variation and vector competence for dengue and yellow fever viruses. The American Journal of Tropical Medicine and Hygiene, vol. 69, no. 1, pp. 105-114. PMid:12932107.
• MACIEL, T.T. and BARBOSA, B.C., 2015. Áreas verdes urbanas: história, conceitos e importância ecológica. CES Revista, vol. 29, no. 1, pp. 30-42.
• MEDEIROS, A.S., MARCONDES, C.B., DE AZEVEDO, P.R., JERÔNIMO, S.M., SILVA, V.P. and XIMENES, M.F., 2009. Seasonal variation of potential flavivirus vectors in an urban biological reserve in Northeastern Brazil. Journal of Medical Entomology, vol. 46, no. 6, pp. 1450-1457. PMid:19960696. http://dx.doi.org/10.1603/033.046.0630
  » http://dx.doi.org/10.1603/033.046.0630
• PERES-NETO, P.R., LEGENDRE, P., DRAY, S. and BORCARD, D., 2006. Variation partitioning of species data matrices: estimation and comparison of fractions. Ecology, vol. 87, no. 10, pp. 2614-2625. PMid:17089669. http://dx.doi.org/10.1890/0012-9658(2006)87[2614:VPOSDM]2.0.CO;2
  » http://dx.doi.org/10.1890/0012-9658(2006)87[2614:VPOSDM]2.0.CO;2
• R CORE TEAM, 2012 [viewed 30 March 2016]. R: a language and environment for statistical computing [online]. Viena: R Foundation for Statistical Computing. Available from: http://www.R-project.org
  » http://www.R-project.org
• REINERT, J.F., 2009. List of abbreviations for currently valid generic-level taxa in family Culicidae (Diptera). European Mosquito Bulletin, vol. 27, pp. 68-76.
• REITER, P., AMADOR, M.A., ANDERSON, R.A. and CLARK, G.G., 1995. Short report: Dispersal of Aedes aegypti in an urban area after blood feeding as demonstrated by rubidium-marked eggs. The American Journal of Tropical Medicine and Hygiene, vol. 52, no. 2, pp. 177-179. PMid:7872449. http://dx.doi.org/10.4269/ajtmh.1995.52.177
  » http://dx.doi.org/10.4269/ajtmh.1995.52.177
• ZEQUI, A.C., LOPES, J. and MEDRI, I.M., 2005. Imaturos de Culicidae (Diptera) encontrados em recipientes instalados em mata residual no município de Londrina, Paraná, Brasil. Revista Brasileira de Zoologia, vol. 22, no. 3, pp. 656-661. http://dx.doi.org/10.1590/S0101-81752005000300021
  » http://dx.doi.org/10.1590/S0101-81752005000300021

Publication Dates

• Publication in this collection
  27 July 2017
• Date of issue
  May-Aug 2018

History

• Received
  31 Mar 2016
• Accepted
  12 Oct 2016