Molecular identification and distribution of insect larvae in the Lower Danube River

Aquatic insects represent a large and diverse group derived from various terrestrial ancestors that re-inhabited aquatic environments, comprising about 76,000 species adapted to all types of freshwater habitats (Samways & Deacon 2021). Although most insects are characterized by a subaquatic life cycle, few species spend their entire life submerged in waters. Most aquatic insects undergo an aquatic immature stage followed by a terrestrial adult (e.g. Odonata, Ephemeroptera, Trichoptera, Megaloptera), being considered semiaquatic, associated only with aquatic and semiaquatic vegetation (Bouchard 2004). These organisms are an important component of aquatic (and sometimes terrestrial) food webs, playing an essential role in nutrient cycling processes (Balian et al. 2008). Most of the larval stages are active and passive filter or deposit feeders, representing an intermediate step between microorganisms and fish in the food chain and constituting an important food source for the latter (Rosenberg & Resh 1993). Due to their large biomass, high reproduction rates with short-lived generations and rapid colonization, insect larvae are successfully used in assessing the structure and function of aquatic ecosystems as important tools in ecology (Hershey & Lamberti 2001), genetics (Sivaramakrishnan et al. 2014) and evolutionary studies (Kristensen 1981; Mazzuccoa et al. 2015). Aquatic ecosystems are exposed to multiple anthropogenic pressures that significantly affect their health condition (Buczyńska and & Buczyński 2019; Gómez-Baggethun et al. 2019; Mureșan et al. 2019; Teacă et al. 2019; Begun et al. 2020) due to habitat change, pollution (Davidson et al. 2013) and soil erosion (Broadman 2013). Being highly sensitive to pollution, aquatic insects are frequently used as bioindicators to assess the impact of environmental stressors on lentic and lotic water systems and the water quality (Nasirian & Irvine 2017), as any change in their composition or density indicates a change in the water quality (Varma & Pratap 2006). Compared to fish and plankton, they have a higher ability to tolerate pollution-induced environmental stress (da Rocha et al. 2010; Andem et al. 2012).

Taxonomic assignment of macroinvertebrates at the species level has proven challenging when using a morphology-based identification technique (Barrett & Hebert 2005). Moreover, identification of early life stages of aquatic insects (larvae and nymphs) based on morphological characters is particularly difficult since most of the identification keys are available only for adults. During the last decades, complementary molecular techniques were used based on DNA analysis for accurate identification and uncertainties linked to the morphological approach (Navajas & Fenton 2000). To date, DNA barcoding based on amplification of the mitochondrial cytochrome C oxidase subunit I gene region has been the most frequently used technique for discriminating invertebrate species (Hebert et al. 2003). In addition to morphological identification, which in most cases could only be performed at the family or genus level, DNA barcoding has allowed identification of specimens up to the species level (Baird et al. 2012).

The distribution and abundance of most aquatic insect communities are considered to be affected by several factors such as current velocity, temperature, altitude, season, total suspended solids, availability of food, and the like (Crisci-Bispo et al. 2007). In general, these organisms require sufficient amounts of dissolved oxygen to survive, and this parameter has the potential to limit aquatic insect diversity (Kamsia et al. 2007). Studies have reported that their dispersal is strongly related to water depth (Hasmi et al. 2021), and the occurrence of certain species is correlated with the presence of vegetation (Crisci-Bispo et al. 2007). Furthermore, the nature of the substrate plays an important role in determining the species distribution and the structure of their communities (Ciutti et al. 2004).

Over the last decades, the distribution, ecology and diversity of aquatic insect species populating the Danube River have been the subject of several studies (Graf et al. 2006; Tubić et al. 2013; Farkas et al. 2014; Krno et al. 2018; Navara et al. 2020), including the Danube sector that crosses the Romanian territory (Chiriac 2004; Pavel et al. 2018). Meanwhile, a very limited number of studies have used molecular identification by DNA barcoding based on COI gene fragments of insects collected from terrestrial environments in Romania, not counting a survey of 180 butterflies (Dincă et. al. 2011), and no identification of aquatic insects using genetic methods has been carried out so far.

In this context, the current study aimed to identify insect larvae belonging to seven different species of the following orders: Odonata – Erythromma viridulum (Charpentier 1840) and Gomphus flavipes (Charpentier 1825), Trichoptera – Triaenodes bicolor (Curtis 1834) and Hydropsyche bulgaromanorum (Malicky 1977), Ephemeroptera – Heptagenia flava (Rostock, 1878), Lepidoptera – Acentria ephemerella (Denis & Schiffermüller, 1775) and Megaloptera – Sialis morio (Klingstedt, 1932), using DNA COI barcoding, and to investigate their distribution and quantitative occurrence in the Lower Sector of the Danube River in relation to different environmental parameters such as substrate type, water depth and oxygen concentration. To our knowledge, this study presents the first molecular identification and characterization of the targeted species from the aquatic environment in Romania.

Due to their sensitivity to several abiotic and biotic factors, immature stages of aquatic insects are particularly useful for monitoring the water quality in freshwater ecosystems (Arimoro & Ikomi 2008; Barman & Gupta 2015). As such, our study addressed the identification and distribution of several biological indicators in the Danube section that crosses the Romanian territory to highlight their presence at different depths and in different substrates.

In Romania, the first detailed taxonomic, biological, and ecological studies of Odonata were performed by Cîrdei & Bulimar (1969), and by Bulimar (1973, 1976, 1993). Odonata larvae were used as bioindicators of anthropogenic disturbances, as shown for E. viridulum (Buczyńska & Buczyński 2019). Therefore, the results of a survey performed in two areas of the El-Kebir-East hydrosystem with different degrees of human impact showed that the species was only found at less disturbed sites characterized by good water quality with high concentrations of dissolved oxygen (Benchalel et al. 2017), as in our case. As a cosmopolitan species, it is widespread in most European countries (de Jong. et al. 2014). It was previously reported in the Romanian reaches of the Danube near the city of Galați (Şerban & Cristescu 2013), Timiş (Manci 2016), and Sf. Gheorghe (de Vries et al. 2017). Furthermore, small populations were reported from the Danube Delta (Mahmudia, Murighiol and Caraorman areas; de Vries et al. 2017). Our current study revealed its presence in only two of the Romanian Danube sectors, Bechet and Cetate. Although previous studies (Carchini et al. 2003; Benchalel et al. 2013) identified this species at water depths < 1m, the current study revealed its presence at greater depths (4–8.2 m). Similar to our results, Heidemann & Seidenbusch (2002) found this species in areas with aquatic vegetation, represented mainly by species from the genera Ceratophyllum and Myriophyllum.

G. flavipes, an indicator species of less polluted sections of large rivers, listed in Annex IV of the Habitats Directive, is currently no longer considered a threatened species in Europe and its populations are expanding (Kalkman et al. 2010). The species was reported from the Danube Delta, Banat, Galați and Suceava counties (Cîrdei & Bulimar 1965; Bulánková et al. 2013), on the banks of the Danube, and the access pier to the meander of the Danube Zătuni-Cotul Pisicii (Şerban & Cristescu 2013). This species showed the highest prevalence among the European species of the genus Gomphus, spreading from France to eastern Siberia in the Sava, Tisza, Danube and Morava rivers (AnđAnđus 1992; Rajkov & Šćiban 2012), with very abundant populations identified in the Small Danube and the Danube Delta (Şibl et al. 2003; Bulánková et al. 2013). In contrast, only one individual of this species was detected in the Slovak section of the Danube during a 20-year survey (Dolny & Bulánková 2006). Large and stable populations of G. flavipes were encountered along the entire Romanian section of the Danube, in areas characterized by a silty substrate, within a wide range of water depths, from shallow (< 3 m) to 30 m deep waters (Pavel et al. 2019). However, our survey revealed the presence of this species at only two sites, at km 626 and 762.5 at a depth of 3.7 m and 4.8 m, respectively, in the Corabia and Pisculeţ sections. It was shown that both G. flavipes and E. viridulum prefer areas with reduced hydrological connectivity and dynamics and high rates of sedimentation, dominated by sandy substrates (Schultz et al. 2003; Bulánková et al. 2013). Our results were consistent with these findings, as samples were collected from a similar substrate.

Another cosmopolitan species recorded during our survey was T. bicolor. The species was recorded in the Danube Delta area (Ciubuc 2004), in the German and Austrian sections of the Danube River (Moog et al. 1994; Graf et al. 2006), and in the Prut River (Munjiu 2014). The current study revealed the presence of this species in the Bechet and Pisculet sections, as well as in the Sulina branch. Furthermore, our data confirmed the presence of larvae in sandy and gravelly sandy substrates, as previously suggested by van den Brink et al. (2013). These larvae feed on aquatic plants and build cases from macrophytes tissues, developing abundant populations in areas rich in aquatic vegetation (Graf et al. 2008), suggesting their preference for areas dominated by submerged vegetation, as confirmed by our survey. This Trichoptera species was reported as a marker for oxygen-rich waters of unpolluted environments (Nijboer & Goedhart 2006). While a recent study by Buczyńska et al. (2018) reported its presence at depths of up to 2 m in waters characterized by low oxygen concentrations, our report indicates that the depth for T. bicolor in the Danube River is up to 6.3 m.

Larvae of the caddisfly H. bulgaromanorum were found during our investigation in all analyzed Romanian sections of the Danube River, including the Sulina branch. It is a widespread species in Europe (Malicky 1984), colonizing the Austrian, German, Slovak and Serbian sections of the Danube (Moog et al. 1994; Elexová & Nemethová 2003; Paunovic et al. 2007), inhabiting areas with a hard benthic substrate containing stones and gravel (Pîrvu et al. 2012). Our data indicating the presence of this species in several types of substrates (sand, mud, gravel) have expanded the knowledge of the types of aquatic habitats of this insect’s larvae. Moreover, although previous studies considered the distribution of H. bulgaromanorum as correlated with the stream velocity (van den Brink et al. 2013), being restricted to the lotic habitat (eupotamon), Ciubuc (2004) found this species in the Danube Delta (including the Razelm Sinoe Lagoon System).

Furthermore, the reported eurybathic character of this species (Savić et al. 2013) appeared to be consistent with our findings that identified an abundant population in both shallow (< 5 m) and deep (> 10 m) Danube locations, suggesting that depth variation is not a limiting factor for its distribution.

So far, the eurybiontic mayfly H. flava has been reported along the Bulgarian, Serbian and Austrian Danube sections (Russev 1992; Moog et al. 1994; Vidinova & Russev 1997; Petrovic et al. 2015). In Romania, this species was observed in the Vişeu and Tejeajen rivers, establishing abundant populations in unpolluted environments (Curtean-Bănăduc 2009; Sava 2016–2017). However, Vidinova & Russev (1997) found larvae in heavily polluted areas with low oxygen concentrations, as well as living in fragments of decaying vegetation, under stones and gravel. The preference for such living substrates was also confirmed in our study, showing high abundance of larvae in substrates characterized by a significant amount of gravel. During an investigation conducted at 46 lotic slow-flowing study sites located in the Pannonian lowland ecoregion in Croatia, a single individual of H. flava was identified in fine sediments at a depth of 1 m (Vilenica et al. 2020). Our findings along the Pisculeţ, Bechet and Corabia sections of the Danube showed its presence at different depths, ranging from 3.2 m to 8.2 m.

The herbivorous caterpillar A. ephemerella is one of the few insect species that can live underwater throughout its entire life cycle (Miller & Bargsten 2016). It is commonly found in lakes and coastal waters throughout Europe (Berg 1941; Gross et al. 2001), including the Austrian part of the Danube (Moog et al. 1994). In our study, this caterpillar was found at only two sites in the Bechet section. Similar to our results, where A. ephemerella was found in areas with macrophytes and a significant amount of organic matter, the study conducted by Gross et al. (2002) revealed that this species occurs in high abundance, feeding also on many submerged macrophytes, especially in areas with Potamogeton sp., Ceratophyllum demersum (Linnaeus, 1753) and Myriophyllum spicatum (Linnaeus 1753). This aquatic moth is an important herbivore insect feeding on aquatic macrophytes (Gross et al. 2002). Therefore, biomonitoring of this species has demonstrated its potential as a biological control agent limiting the growth of the Eurasian watermilfoil M. spicatum (Johnson et al. 1998; Newman 2004).

The alderfly S. morio is also a widespread species in Europe, previously found in Norway, Sweden, Finland, Slovakia, countries of former Yugoslavia, and Romania (Aspöck et al. 1980; Meinander 1996; Ábrahám & Kovács 1999; Tierno del Figueroa et al. 2002). S. morio larvae were found in both lentic and lotic ecosystems (Kaiser et al. 1977; Aspöck et al. 1980). Even though several studies (Aspöck et al. 1980; Ábrahám & Kovács 1999) highlighted the eurybiontic character of this species, the tolerance to hostile environmental conditions and the vertical distribution ranging from lowlands to higher altitudes, Yakovlev (2009) indicated that the species is mostly associated with lowland lakes. This could explain the absence of this alderfly in most of the study areas, being found only at one site in the Sulina branch (Meander) characterized by a muddy substrate and slow water flow.

The dissolved oxygen content in water appeared to be one of the most important environmental factors for the survival, development, and reproduction of aquatic insects, which generally require high concentrations of dissolved oxygen to complete their life cycle (Hossain et al. 2015; Prommi & Payakka 2015). In our study, relatively high concentrations of dissolved oxygen were recorded in all analyzed Danube areas. Despite this, only H. bulgaromanorum was encountered at most sites, showing a wide distribution along the Romanian sections of the river. This could be attributed to its high ecological plasticity, as indicated by both previously published results and the current study.

Given the importance of these organisms as biological indicators, their accurate taxonomic identification is essential. Thereby, DNA barcoding was used during our survey as the main method for taxonomic identification of species, being a successfully applied technique in biodiversity assessment studies (Cordero et al. 2017). The morphological investigation of larvae failed to identify the taxonomic characters of S. morio. As such, DNA barcoding proved to be effective in resolving this intricacy, given that, in addition to this species, two others were reported in Romania, S. fuliginosa (Pictet 1836) and S. lutaria (Linnaeus 1758) (Aspöck et al. 2001; Devetak et al. 2016). Moreover, Ábrahám & Kovács (1999) showed difficulties during taxonomic identification based on morphological characters due to high degree of phenotypic similarities between S. morio and S. lutaria. Several studies (Doskocil et al. 2008) focused on the species-level resolution for insects, showing that conventional taxonomic methods can be ambiguous and may lead to inaccurate results. In this regard, Jalalizand et al. (2012) reported challenges in identifying aphids, while Talbalachi & Shaikevich (2010) and Chan et al. (2014) in identifying Diptera species based on morphology. Nevertheless, the COI gene fragment amplification for identification of aquatic insect species was used as a sensible tool in recent studies (Molina et al. 2017; Pramual et al. 2020). For most of the currently examined species, only a limited number of nucleotide sequences are available in the reference databases for DNA barcodes (Benson et al. 2007; Ratnasingham & Hebert 2007). Also, few studies focused on the molecular identification and characterization of these taxa. In some of these studies, E. viridulum and other damselflies and dragonflies populating the water bodies of Vienna were identified using DNA barcoding and the eDNA approach (Fisher et al. 2018). Studies on the distribution of Odonata and Trichoptera species in Italian environments using molecular methods also focused on E. viridulum and G. flavipes (Galimberti et al. 2020) and H. bulgaromanorum (Geraci et al. 2011; Kučinić et al. 2018), respectively.

The current data, advancing the knowledge about the spatial distribution and abundance of Odonata, Trichoptera, Ephemeroptera, Lepidoptera and Megaloptera larvae in the lower Danube River, provide a novel insight into their role as freshwater biological indicators. Aquatic insects are among the most sensitive organisms to changes in environmental conditions, therefore further research should be undertaken to assess their ecological role in relation to various environmental parameters that may affect their distribution.