Knowledge about the diversity of Syrphidae in the Colombian Amazon region is still incipient, with only one study recording the expansion of the geographic distribution of Cepa apeca Thompson, 2007; a species considered rare (Parada-Marín et al. 2021), as well as the description of Copestylum enriquei Montoya, Parada-Marín and Ramos-Pastrana, 2021, an endemic Amazonian species (Montoya et al. 2012; Parada-Marín et al. 2021; Montoya et al. 2021) which was exclusively collected together with Quichuana angustiventris (Macquart, 1855) in the dense secondary forest during HIR.
Our study records 72 species (Table 1; Fig. 2) for Reserva Natural y Ecoturística La Avispa, corresponding to 20% of the Colombian Syrphidae fauna (Montoya et al. 2012; Montoya 2016; Parada-Marín et al. 2021; Montoya et al. 2021; Montoya et al. 2022; Montoya and Wolff 2023), including, 16 species recorded for the first time in the country, increasing the Syrphidae fauna of Colombia to 384 species, including species that extend their distribution range to the Amazon region as well as other potentially new to science (Table 1).
Although VSRTs have been used mainly in studies of butterfly diversity and ecology (Villarreal et al. 2004), obtaining good results in different vegetation covers (Brown and Freitas 2002; Freitas et al. 2014; Casas-Pinilla et al. 2017; Álvarez et al. 2021), these traps have also been employed in studies of Diptera, managing to be very effective in Synanthropic studies of Calliphoridae (Montoya et al. 2009; Pinilla et al. 2012), Muscidae (Ramos-Pastrana et al. 2022) and Sarcophagidae (Pinilla et al. 2012; Yepes-Gaurisas et al. 2013). Although studies on Syrphidae diversity mainly employ Malaise traps (de Souza et al. 2014; Sommaggio and Burgio 2014) or entomological nets (Rotheray et al. 2007; Ángel et al. 2021; Montoya et al. 2021), in our study, VSRTs presented high effectiveness in capturing Syrphidae, being decomposing fish the highest efficiency bait, compared to shrimp and fermented fruit (Table 1; Fig. 3), constituting a potential alternative method for rapid inventories of flower flies in the Neotropics.
The three habitats were dominated by Copestylum species (Aquatic Saprophagous) (Fig. 4), a New World endemic genus that harbor more than 400 described species (Thompson 1981; Rotheray et al. 2007; Ricarte et al. 2015; Montoya et al. 2022), of which 71 have been recorded in Colombia (Restrepo and Carrejo 2009; Montoya 2016; Montoya et al. 2021, 2022). In our study, 12 species of this taxon are recorded for the first time in the country (Table 1), highlighting the strategic importance of the studied reserve, which harbors the endemic Amazonian species such as Copestylum enriquei, exclusively collected in the dense secondary forest during HIR (Fig. 11), suggesting the relevance of VSRTs as a complementary tool for studying Syrphids diversity.
Salpingogaster nigra (Terrestrial Zoophagous) reached its highest abundance (n = 39) in the agroforestry system during HIR (Fig. 4c), which could be because the species is considered the natural enemy and main biological controller of the spittlebug, Aeneolamia varia (Hemiptera: Cercopidae), which can prey up to 17–40 nymphs during its larval stage and reaching up three generations per life cycle (Guppy 1913; Sotelo and Cardona 2001; Bustillo 2011; Espitia et al. 2022) in sugarcane crops (Páez et al. 1985; Lastra et al. 2007), crops that was also dominant in the AFS. In particular, its prevalence during HIR could be related to the higher availability of spittlebugs during rainy seasons, which according to Veríssimo et al. (2018) tends to decrease considerably during the dry season.
Dioprosopa clavata (Fabricius, 1794) was exclusive to agroforestry systems during HIR, a habitat dominated by sugarcane (Table 1; Fig. 7b). This species is widely distributed in tropical and subtropical areas of America (Kassebeer 2000; Lillo et al. 2021), being listed as a natural enemy of numerous aphid species of agricultural importance, considered also as biological control of Aeneolamia Fennah, 1949 (Hemiptera: Cercopidae), one of the main pest in sugarcane crops (Rojo et al. 2003; Arcaya et al. 2018; Lillo et al. 2021; Ferrer and Salas 2024).
In the interpolation and extrapolation curves, the highest completeness, concerning the sample coverage estimator was obtained by the dense secondary forest (98%) (Fig. 5a-c), possibly due to the constant rainfall during HIR, which causes an increase in water reservoirs (Phytotelmatas), with the availability of decomposing organic matter, which are optimal conditions for the development of saprophagous larvae of Eristalinae (Copestylum, Palpada and Quichuana) (Howarth 2000; Rotheray et al. 2007), subfamily that presented the highest richness during the rainy season (n = 1,242, 47 species). The forest edge obtained a completeness of 92.9%, the same value obtained by the agroforestry system, attributable to the prevalence of some species that prefer opened areas, with high light incidence (de Souza et al. 2014). In Montoya et al. (2021), grassland habitats, characterized by large open areas, obtained the lowest number of genera (23) and species (59), therefore, this habitat obtained the lowest completeness concerning the other sampled habitats. Despite this, grassland was the second most abundant habitat (n = 571), representing 20% of the collected material. However, the authors state that anthropization, plant homogeneity and extreme climatic conditions could impact diversity, with a prevalence of generalist and widely distributed species.
The agroforestry system obtained the same richness as the dense secondary forest (44 species), however, the less completeness (92.9%) (Fig. 5a-c; Table 2). This agrees with Gittings et al. (2006), who emphasize the importance of open spaces for the maintenance of syrphid diversity in forest plantations, where about 80% of the species recorded in their study were associated with such areas. All values obtained for the sample coverage estimator in LIR were below 71% (only in AFS and FE), while those for extrapolation completeness during HIR, averaged 77% and 68% in LIR (Fig. 5g-i; Table 2), which may be because the increase in temperature produced a decrease in water reservoirs, dries the decomposing organic material and limits the survivor of populations of other key insects, critical for the development of some species of Syrphinae (Terrestrial Zoophages), limiting the persistence of specialist functional groups (Gittings et al. 2006; Meyer et al. 2009; Montoya et al. 2021).
The Hill numbers suggested a greater richness during HIR (Table 3). Ari et al. (2021) indicate that high rainfall promotes more abundant and diverse insect assemblages, due to environmental conditions, which considerably impact plant heterogeneity in conserved habitats, contributing to the stability and increasing diversity, meanwhile, the dry season and the plant homogeneity that characterizes disturbed habitats causes a decrease in insect biodiversity. Likewise, Djellab et al. (2019) state that seasonality has a considerable impact on the Syrphids composition, as it influences food resources and availability of developmental microhabitats, directly intervening in population dynamics. The high diversity obtained during HIR may be attributable to the increase in water reservoirs formed in phytotelmata, providing ideal conditions for the development of functional groups such as the Aquatic Saprophagous and Xylosaprophagous (Howarth 2000; Wolff et al. 2023), the first one being particularly diverse and abundant in our study.
Sommaggio (1999) reported that the diversity of Syrphidae is determined by the availability of food resources for their immature stages. In our study, we were able to identify the presence of four larval trophic categories, in which richness at genera and species level, as well as abundance was skewed (Fig. 6), with Terrestrial Zoophagous having the highest genera number (8), being the second most speciose group (18) (Table 1; Fig. 6a). This group was equally diverse in the study of Montoya et al. (2021) in the Paramos Santa Inés Belmira and Sonsón complexes, recording 22 genera and 84 species.
Concerning the species and individuals collected, Aquatic Saprophagous presented significant differences compared to the other larval trophic categories, harboring 71% of the species and 90% of the specimens collected (Table 1; Fig. 6b-c). These results coincide with those reported by Montoya et al. (2021), who obtained higher abundance and richness in the genera Copestylum, Palpada and Quichuana, a pattern that is consistent with the observed in our study, in which these genera had also a good representation in conserved forest.
In terms of diversity, the dense secondary forest (14 exclusive species) and the agroforestry system (12 exclusive species) had the same richness (44 species) and exclusive genera (1) (Fig. 7a). These results differ from those reported by de Souza et al. (2014) in Brazil using Malaise traps, where the forest edge obtained the highest richness (98 species), where more than half species were exclusive, compared to the more conserved forested areas. However, it should be considered that many Syrphidae species are associated with heterogeneous forests, which seem to offer a large number of microhabitats in European habitats (Branquart and Hemptinne 2000; Meyer et al. 2009).
Argentinomyia was exclusive to the dense secondary forest (Fig. 7c), a group that has been extensively studied in pristine high Andes Forest and Paramo ecosystems, where it harbors high diversity and endemism in highlands (Montoya et al. 2012; Montoya and Wolff 2020; 2023), despite, some species occur in the lowlands closed to the Amazonian region (Montoya and Wolff 2023). This is the first report of the genus in the Colombian Amazon region.
Betadiversity was dominated by species turnover between and within habitats (Fig. 8). Results that coincided with the dissimilarity obtained in the nMDS carried out to compare the species distribution using the Bray-Curtis and Jaccard indices (Fig. 9). The nMDS conducted to compare species composition and baits, yielded that decomposing fish was the preferred bait (Fig. 10) by Copestylum species (most diverse group) and Ornidia, which have been related to decomposing carcasses (Martins et al. 2010; Ramos-Pastrana et al. 2018; Ramos-Pastrana et al. 2019). While, nMDS also suggests that discomposed shrimp and fermented fruit overlap, which could be a consequence of shared species.
The greatest dissimilarity between habitats was observed between dense secondary forest and agroforestry systems (0.64) (Fig. 8; Supplementary Material S1), attributable to the low number of shared species (3). Also, by the plant species that characterize each habitat and the incidence of light, the latter being very important in open areas (agroforestry system and forest edge), since species occurring in these areas generally do not occur in forests (Gitttings et al. 2006). This agrees with the beta diversity results in the pairwise comparison between forest edge and agroforestry system (0.5) (Fig. 8; Supplementary Material S1), which were the least dissimilar, as they are the habitats with the highest number of shared species (9), with turn over as the dominant component of beta diversity (Fig. 8; Supplementary Material S1).
Irven et al. (1999) state that the seasonal availability of pollen is an important phenological factor, an argument supported by Djellab et al. (2019), who suggest that the availability of food resources during the seasons affects the abundance and population dynamics of flower flies. During our study, greater activity of Syrphids was evidenced during HIR (Fig. 11), suggesting that plant heterogeneity, availability of food resources and phytotelmata as main microhabitat are the important phenological factors involved in the composition of flower flies in each habitat and climatic season.
Van Someren-Rydon traps as an alternative method to assess Syrphidae diversity in rapid sampling inventories
The present is the first study of flower fly diversity in Colombia using exclusively Van Someren-Rydon traps, which showed high sampling integrity. However, the absence of genera and species typical of the Amazonian fauna (Miranda 2017a) evidences the need to use complementary collection methods such as rearing immature, as well as the use of Malaise traps and entomological netting, which allows to cover more groups and make a more compressive diversity estimation. Despite this, the finding of new records and critical functional groups suggest that the VSRTs constitutes a potential alternative for rapid inventories of flower flies in the Neotropics.
It is noteworthy that in this study the preferred and most effective bait in the collection of Syrphidae was decomposing fish, contributing considerably to the new record of species in Colombia (16) and the extension of the range of distribution in the Colombian Amazon (31).