Introduction
Biological invasions cause global changes due to their impacts on ecosystems and biodiversity (Pyšek et al., Reference Pyšek, Hulme, Simberloff, Bacher, Blackburn, Carlton, Dawson, Essl, Foxcroft, Genovesi, Jeschke, Kühn, Liebhold, Mandrak, Meyerson, Pauchard, Pergl, Roy, Seebens, van Kleunen, Vilà, Wingfield and Richardson2020), being one of the main threats to species extinctions (Mollot et al., Reference Mollot, Pantel and Romanuk2017; Spatz et al., Reference Spatz, Jones, Bonnaud, Kappes, Holmes, Guzmán, Young and VanderWerf2023; Su et al., Reference Su, Cui, Man, Li, Huang, Chen and Zhao2023). International trade, transport and tourism have contributed to an exponential increase in invasive species worldwide in recent decades (Seebens et al., Reference Seebens, Blackburn, Dyer, Genovesi, Hulme, Jeschke, Pagad, Pyšek, van Kleunen, Winter, Ansong, Arianoutsou, Bacher, Blasius, Brockerhoff, Brundu, Capinha, Causton, Celesti-Grapow, Dawson, Dullinger, Economo, Fuentes, Guénard, Jäger, Kartesz, Kenis, Kühn, Lenzner, Liebhold, Mosena, Moser, Nentwig, Nishino, Pearman, Pergl, Rabitsch, Rojas-Sandoval, Roques, Rorke, Rossinelli, Roy, Scalera, Schindler, Štajerová, Tokarska-Guzik, Walker, Ward, Yamanaka and Essl2018; Sun et al., Reference Sun, Koski, Wickham, Baranchikov and Bushley2024). Invasive species must present adaptive responses to different selective pressures faced in the environments where they are introduced in order to ensure their survival and colonisation of new habitats (Schäfer et al., Reference Schäfer, Berger, Rohner, Kjaersgaard, Bauerfeind, Guillaume, Fox and Blanckenhorn2018). Studies on rapid evolutionary changes in invasive species are of great interest because they can identify phenotypic characteristics which favour dispersal and successful colonisation of new habitats (Rejmánek and Richardson, Reference Rejmánek and Richardson1996).
Evolutionary studies with invasive insect species have widely used wing morphology (Huey et al., Reference Huey, Gilchrist, Carlson, Berrigan and Serra2000; Gilchrist et al., Reference Gilchrist, Huey, Balanyà, Pascual, Serra and Noor2004; Loh et al., Reference Loh, David, Debat and Bitner-Mathé2008; Laparie et al., Reference Laparie, Vernon, Cozic, Frenot, Renault and Debat2016). This is an attractive structure for this kind of investigation, since wings are of wide importance in a variety of aspects of an insect's life, such as reproduction, territorial display, foraging, defence mechanisms, thermal regulation and aerodynamics (Bettsand and Wootton, Reference Betts and Wootton1988; Wootton, Reference Wootton1992; Berwaerts et al., Reference Berwaerts, Aerts and van Dyck2006; Pass, Reference Pass2018). Changes in environmental conditions are capable of promoting quantitative variations in physiology and morphology of insect wings. Reducing the frequency of wingbeats at high temperatures is an adaptive strategy in bees which favours thermoregulation (Glass et al., Reference Glass, Burnett, Combes, Weisman, Helbling and Harrison2024). Several studies on the Drosophilidae family have demonstrated that abiotic parameters (such as temperature) influence the wing size of invasive species, with flies developing larger wings as an adaptive response to occupying environments with lower temperatures (Karan et al., Reference Karan, Moreteau and David1999; Huey et al., Reference Huey, Gilchrist, Carlson, Berrigan and Serra2000; Gilchrist et al., Reference Gilchrist, Huey, Balanyà, Pascual, Serra and Noor2004; Gilchrist and Huey, Reference Gilchrist and Huey2004; Loh et al., Reference Loh, David, Debat and Bitner-Mathé2008).
The Asian Drosophila nasuta (Diptera: Drosophilidae) fly is an invasive species in Brazil with a notable ability to expand its geographic distribution in this area of introduction (Martins et al., Reference Martins, Santos, Santos, Araújo, Neves, Garcia and Montes2023). The first record of this species in Brazil occurred in the Cerrado (savanna), approximately 10 years ago (Leão et al., Reference Leão, Roque, Deus and Tidon2017). Since then, D. nasuta has spread throughout different Brazilian biomes, such as the Atlantic Forest (Vilela and Goñi, Reference Vilela and Goñi2015; Silva et al., Reference Silva, Schmitz, Medeiros, Rohde, Montes and Garcia2020), the Caatinga (xeric shrubland) (Montes et al., Reference Montes, Neves, Ferreira, Santos, Quintas, Manetta, Oliveira and Garcia2021), the Amazon Forest (Medeiros et al., Reference Medeiros, Monteiro, Caçador, Pereira, Praxedes, Martins, Montes and Garcia2022) and the Pantanal wetlands (Martins et al., Reference Martins, Santos, Santos, Araújo, Neves, Garcia and Montes2023). The species currently occupies more than half of Brazil's area (Martins et al., Reference Martins, Santos, Santos, Araújo, Neves, Garcia and Montes2023). Genetic studies in Brazil indicate population structuring of D. nasuta (Santos et al., Reference Santos, Neves, Oliveira, Ribeiro, Faria-Júnior, Montes and Garcia2021), reflecting its evolutionary potential in this territory. Some possible effects of ecosystem imbalances caused by D. nasuta in Brazil have been observed in the north of the Atlantic Forest, where a reduction in the abundance of native drosophilids was reported after this species arrived (Oliveira, Reference Oliveira2021). Furthermore, D. nasuta shows a preference for preserved environments compared to anthropised areas (Silva et al., Reference Silva, Schmitz, Medeiros, Rohde, Montes and Garcia2020), which represents a threat to the biodiversity of invaded territories.
Ecological niche modelling data projected the geographic expansion of D. nasuta in different invasion areas, especially in conservation units in Central and South America. These data also indicated that rainfall and temperature parameters are mainly responsible for limiting the global distribution of this species (Garcia et al., Reference Garcia, Silva, Neves and Montes2022).
In this work, significant differences in the wing size of Brazilian D. nasuta populations obtained in Amazon Forest, Caatinga, Cerrado and Atlantic Forest areas were evaluated. The influence of abiotic factors known to be important for the geographic distribution of D. nasuta, such as rainfall and temperature, were tested to understand the geographical pattern of the observed morphological variation.
Materials and methods
Drosophila nasuta sampling locations and capture method
Drosophilids were collected in Brazil in areas within the Amazon Forest, Caatinga, Cerrado and Atlantic Forest biomes (fig. 1). Sampling was always performed during periods of greater rainfall in the areas investigated between 2019 and 2021 in order to remove the morphological variation associated with seasonality (Przybylska et al., Reference Przybylska, de Brito and Tidon2016) (table 1).
On the left, map of Brazil with an indication of its biomes. On the right, partial enlargement of the map, indicating the sampling locations of Drosophila nasuta.
Drosophila nasuta biomes and sampling locations in Brazil with data on geographic coordinates, sampling dates and climate characterisation (temperature and rainfall)
The codes for the locations are the same as those used in fig. 1.
a Climatempo (2024a) Climatologia histórica de Altamira, Pará. https://www.climatempo.com.br/climatologia/228/altamira-br (Accessed 20 January 2024).
b Climatempo (2024b) Climatologia histórica de Manaus, Amazonas. https://www.climatempo.com.br/climatologia/25/manaus-am (Accessed 20 January 2024).
c Climatempo (2024c) Climatologia histórica de Caruaru, Pernambuco. https://www.climatempo.com.br/climatologia/764/caruaru-pe (Accessed 20 January 2024).
d Climatempo (2024d) Climatologia histórica de Belo Jardim, Pernambuco. https://www.climatempo.com.br/climatologia/2179/belojardim-pe (Accessed 20 January 2024).
e Climatempo (2024e) Climatologia histórica de Brasília, Distrito Federal. https://www.climatempo.com.br/climatologia/61/brasilia-df (Accessed 20 January 2024).
f Climatempo (2024f) Climatologia histórica de Igarassu, Pernambuco. https://www.climatempo.com.br/climatologia/1256/igarassu-pe (Accessed 20 January 2024).
g Climatempo (2024g) Climatologia histórica de Itatiaia, Rio de Janeiro. https://www.climatempo.com.br/climatologia/303/itatiaia-rj (Accessed 20 January 2024).
The biomes studied are areas where D. nasuta has been recorded in greater abundance in South America (Leão et al., Reference Leão, Roque, Deus and Tidon2017; Silva et al., Reference Silva, Schmitz, Medeiros, Rohde, Montes and Garcia2020; Montes et al., Reference Montes, Neves, Ferreira, Santos, Quintas, Manetta, Oliveira and Garcia2021; Martins et al., Reference Martins, Santos, Santos, Araújo, Neves, Garcia and Montes2023), representing a wide territory of its distribution on this continent (Martins et al., Reference Martins, Santos, Santos, Araújo, Neves, Garcia and Montes2023). The Amazon is the largest tropical rainforest in the world, covering nine countries in South America, with 59% of its area in Brazil (IBGE, 2019). It is one of the biomes with the greatest biodiversity on the planet (Guayasamin et al., Reference Guayasamin, Ribas, Carnaval, Carrillo, Hoorn, Lohmann, Riff, Ulloa and Albert2024). The Caatinga is the largest and most diverse seasonally dry tropical forest in the world, occurring exclusively in Brazil, where it occupies around 10% of its territory (Silva et al., Reference Silva, Leal and Tabarelli2017; IBGE, 2019). The Cerrado is the most biodiverse savanna on the planet, extending across three countries (Brazil, Paraguay and Bolivia), but mainly found in Brazil where it occupies 24% of its territory (Walter et al., Reference Walter, Carvalho, Ribeiro, Sano, de Almeida and Ribeiro2008; IBGE, 2019). This biome is recognised as one of the hotspots for conservation (Mittermeier et al., Reference Mittermeier, Turner, Larsen, Brooks, Gascon, Zachos and Habel2011). The Atlantic Forest occupies approximately 13% of the Brazilian territory. It is mainly distributed along its coast, also extending to part of Argentina and Paraguay (IBGE, 2019). It is one of the richest humid tropical forests in the world, featuring many endemic and endangered species, and is one of the world's hotspots (Tabarelli et al., Reference Tabarelli, Pinto, Silva, Hirota and Bede2005).
Drosophilids were sampled in each of the eight investigated locations (table 1), using ten traps made from plastic bottles containing banana bait (Tidon and Sene, Reference Tidon and Sene1988). The traps were suspended 1.5 m from the ground and distributed randomly (at a minimum distance of 30 m between them) and 50 m away from the edges of the forest fragments where they remained exposed for three consecutive days. The captured drosophilids were stored in 70% ethanol and the D. nasuta specimens were identified according to Vilela and Goñi (Reference Vilela and Goñi2015) by their light body colour, the presence of a longitudinal brown stripe in the middle dorsal area of the pleura, a silvery and whitish fringe in the head region when viewed from the front, a row of cuneiform setae on the anteroventral side of the femur on the forelegs, wings with a costal index of about 3.1, and male terminalia characteristics.
Morphometric and statistical analyses
The D. nasuta individuals collected were separated by sex and geographic origin. Separation by sex was performed by analysing the flies' terminalia, with males being distinguished from females by the presence of an aedeagus and hypandrium and the absence of an ovipositor. Flies were discarded if they had torn or wrinkled wings. A total of 30 male individuals from each population were dissected with a 70% ethanol solution. The right-side wing of each individual was removed by squeezing the wing joint with tweezers and pulling the wing away from the body, using a pair of tweezers to hold the body in place. Only the right-side wings were used to avoid fluctuating asymmetry variations.
The dissected wings were placed on microscope slides with the ventral side facing down and covered with a 1:1 solution of absolute ethanol and glycerol. Slides were covered with coverslips and any air bubbles were gently removed by pressing the coverslip with forceps. The wings were digitally photographed on an Instrutherm MBB-200 microscope at 40 × magnification.
Next, 11 measurements were taken from the digitised images on each wing from reference points at the junction or termination of the venations, following the parameters of Bitner-Mathé and Klaczko (Reference Bitner-Mathé and Klaczko1999) (fig. 2). Measurements were performed using the tpsDIG program (Rohlf, Reference Rohlf2016). The wings of all specimens were mounted, photographed and measured by the same person in order to minimise possible errors in morphometric analyses, in accordance with the recommendations of Fox et al. (Reference Fox, Veneracion and Blois2020).
Right wing of a male Drosophila nasuta with indications of the measurements that were taken from six reference points: OA, OB, OE, AB, AE, BC, BD, BE, CD, CE and DE.
Arithmetic means and standard deviations were obtained for each of the 11 wing measurements for samples from different geographic locations. Analysis of variance (ANOVA) was performed with the Tukey a posteriori test to observe possible differences in wing measurements between locations. The wing measurements of individuals from locations which did not show statistical differences in previous tests were grouped. The established groups were analysed using a linear discriminant function. Pearson's correlation test was performed between wing measurements and abiotic factors (rainfall and maximum and minimum temperatures). All of these analyses were carried out using the PAST version 4.3 program (Hammer et al., Reference Hammer, Harper and Ryan2001) and a significance level of P < 0.05 or P < 0.001 was used in the statistical tests.
Results
The lowest averages for the 11 wing measurements evaluated in 240 D. nasuta individuals were observed for the populations of the Amazon Forest (Altamira and Manaus) and the north of the Atlantic Forest (Igarassu). The Caatinga (Caruaru and Belo Jardim), the Cerrado (sensu strictu and Gallery Forest) and the south of the Atlantic Forest (Itatiaia) populations presented the highest averages for these measurements (table 2).
Arithmetic means (in mm) and standard deviations for the different reference point measurements of the right wings of Drosophila nasuta males from different biomes and locations in Brazil
The reference points for wing measurements are illustrated in fig. 2.
The two groups observed by analysing the mean D. nasuta wing measurements were also verified by ANOVA and the subsequent Tukey's test (table 3, Supplementary table 1). One of the groups was formed by populations from the Amazon Forest (Altamira and Manaus) and the north of the Atlantic Forest (Igarassu), with no significant difference between these samples. Another group brought together populations from the Caatinga (Caruaru and Belo Jardim), the Cerrado (sensu strictu and Gallery Forest) and the south of Atlantic Forest (Itatiaia), also without significant differences between the wing measurements of these populations. Comparisons of wing measurements between the populations of these two groups showed significant differences (P < 0.001) (table 3).
Tukey's test (P < 0.001) for measurements of the right wings of Drosophila nasuta males collected in different locations in Brazilian biomes
The groupings formed in the previous analyses were evaluated using a linear discriminant function, which confirmed the existence of these two distinct groups. In this analysis, 91.67% of individuals were correctly identified in their corresponding groups by the cross-validation test (table 4).
Classification by discriminant function analysis followed by cross-validation for the two groups obtained by analysis of means and ANOVA/Tukey's test, based on measurements of the right wings of Drosophila nasuta males collected in different Brazilian locations and biomes
Group 1 = populations from the Amazon Forest (Altamira and Manaus) and the north of the Atlantic Forest (Igarassu); and Group 2 = populations from the Caatinga (Caruaru and Belo Jardim), the Cerrado (sensu strictu and Gallery Forest) and the south of the Atlantic Forest (Itatiaia).
The group of individuals with the smallest wings coincided with the locations with the highest maximum and minimum temperatures detected. The group with the largest wings corresponded to the areas with the lowest maximum and minimum temperatures (table 1). A high negative and significant correlation was observed between maximum temperatures and four of the 11 wing measurements investigated, as well as for all wing measurements and minimum temperatures. No significant correlation regarding rainfall was observed with any of the wing measurements analysed (table 5).
Pearson's correlation between measurements of the right wings of Drosophila nasuta males collected in different biomes in Brazil and abiotic factors (rainfall, maximum and minimum temperatures)
Discussion
The Asian D. nasuta fly invaded Brazil approximately 10 years ago (Leão et al., Reference Leão, Roque, Deus and Tidon2017). The species has already expanded over an area of 4.6 million km2 in this short period, which corresponds to 55% of the Brazilian territory (Martins et al., Reference Martins, Santos, Santos, Araújo, Neves, Garcia and Montes2023). Data from 11 wing measurements taken on 240 individuals of this species from different Brazilian biomes in the present study revealed statistically significant differences between the geographic samples.
The variations in wing sizes observed herein resulted in forming two groups of D. nasuta. Other invasive drosophilids in the Neotropical region also showed significant differences in wing morphology in different areas of introduction. Loh and Bitner-Mathé (Reference Loh and Bitner-Mathé2005) detected variations in the wing size and shape of the African Zaprionus indianus fly in areas recently invaded by the species in Brazil. Some authors have observed significant differences in the morphometry of drosophilid wings in comparison with invaded areas, and in comparing these areas with locations where the species are native; for example, in studies conducted with Z. indianus (David et al., Reference David, Araripe, Bitner-Mathé, Capy, Goñi, Klaczko, Legout, Martins, Voudibio, Yassin and Moreteau2006; Yassin et al., Reference Yassin, David and Bitner-Mathé2009) and D. suzukii (Fraimout et al., Reference Fraimout, Jacquemart, Villarroel, Aponte, Decamps, Herrel, Cornette and Debat2018). Taken together, our results and those of these investigations reveal the capacity for morphological differentiation in the wings of invasive drosophilids in introduced areas.
Drosophila nasuta individuals with larger wings were observed in locations with colder temperature extremes (Caatinga, Cerrado and south of Atlantic Forest) and those with smaller wings occurred in locations with higher minimum and maximum temperature extremes (Amazon Forest and north of the Atlantic Forest). Changes in environmental temperature conditions are recognised to promote quantitative variations in drosophilid wing morphology. As pointed out by our results, other studies have shown that invasive drosophilids have larger wings in areas with lower temperatures in places of introduction. For example, this has been observed for the European species D. subobscura in invaded areas in North and South America (Huey et al., Reference Huey, Gilchrist, Carlson, Berrigan and Serra2000; Gilchrist et al., Reference Gilchrist, Huey, Balanyà, Pascual, Serra and Noor2004; Gilchrist and Huey, Reference Gilchrist and Huey2004) and for the African species Z. Indianus in invaded areas in India (Karan et al., Reference Karan, Moreteau and David1999) and South America (Loh et al., Reference Loh, David, Debat and Bitner-Mathé2008). These authors deemed that changes in the wing size of invasive drosophilids in response to temperature variations were associated with an adaptive process.
Our results revealed a high negative correlation between wing measurements and maximum temperatures, and especially for minimum temperatures. Fraimout et al. (Reference Fraimout, Jacquemart, Villarroel, Aponte, Decamps, Herrel, Cornette and Debat2018) tested the influence of different temperatures (16, 22 and 28°C) in a laboratory on the wing morphology of the Asian species D. suzukii from samples collected in its area of origin in Japan and in two invasion areas, France and the United States. As observed in the present study for D. nasuta, the extreme minimum temperature most influenced the wing size of D. suzukii, resulting in individuals with larger wings compared to those at temperatures of 22 and 28°C (which did not present significant differences in wing morphology between them). The importance of minimum temperatures for the occurrence of D. nasuta has been highlighted by Garcia et al. (Reference Garcia, Silva, Neves and Montes2022) in an ecological niche modelling study. These authors revealed that cold temperatures explain 21% of the global geographic distribution model of this species. Thus, colder minimum temperatures seem to influence the wing morphology of different drosophilid species, and at the same time, account for the geographic distribution capacity of D. nasuta.
Why were the largest wings of D. nasuta observed in individuals occupying locations with the lowest minimum temperatures? This probably occurs because large-winged insects are more effective at absorbing heat, making this trait advantageous in areas with more extreme cold conditions where obtaining and retaining heat are critical for survival (Heinrich, Reference Heinrich1974; Douglas, Reference Douglas1981). Thus, the phenotypic variation found in the wings of D. nasuta individuals could be the result of an adaptive process related to temperature. Laboratory experiments may confirm this result by cultivating geographic samples of this species at different temperatures.
The present study is a pioneer in describing a morphological variation pattern in the wing size of D. nasuta, sampling individuals from a large part of the geographic distribution of this recent invasive species in Brazil. This condition reveals the adaptive potential of D. nasuta in introduced areas.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0007485324000580
Acknowledgements
This work was supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, process number: 425274/2018-7).
Competing interests
None.
