Campsurus truncatus Ulmer , 1920 ( Polymitarcyidae ) : an Ephemeroptera in eutrophic waters Campsurus truncatus Ulmer , 1920 ( Polymitarcyidae ) : um Ephemeroptera em águas eutrofizadas

Aim: The order Ephemeroptera, together with Trichoptera and Plecoptera, has been considered as sensitive in environmental quality assessment. However, the species Campsurus truncatus Ulmer 1920 (Polymitarcyidae) has been sampled in anthropically altered sites. In this paper, data on the occurrence of that species in the water quality monitoring network of the Environmental Agency of São Paulo State (CETESB) were gathered to review its position in the environmental gradient and in the quality assessment of rivers and reservoirs. Methods: For this, a compilation of the occurrences and densities of Campsurus obtained in rivers and reservoirs of São Paulo state between 2001 and 2018, with grab samplers, was carried. Environmental factors, the Trophic State Index (TSI) and the Benthic Community Index (BCI) were related to the species densities to define its environmental requirements. Results: The sublittoral zone of reservoirs was the most common habitat to C. truncatus, occurring more frequently at mesotrophic sites with fair ecological quality. Higher mean densities were recorded at eutrophic sites with good ecological quality conditions. The statistical analysis did not show linear correlation between C. truncatus densities and any of the analyzed environmental variable such as grain size, organic matter content, DO and light penetration, although there was a tendency for higher species densities occur in muddy, dark sediments and with intermediate levels of organic matter. As described in literature the presence of aquatic plants can actually benefit that population once the highest densities were observed in sites with many macrophytes, as Salto Grande reservoir, but it may be confusing the relationship between population density and environmental factors. Conclusions: The high occurrence and densities of C. truncatus obtained in eutrophic environments suggest that the species should not be considered as a sensitive taxon in qualitative indexes, such as the BCI for rivers and the sublittoral region of reservoirs. Moreover, its exclusion from the richness of sensitive groups had low impact on the diagnoses already performed at the CETESB monitoring network.

greater depths by other species of benthic fauna and microbial fauna (Leal et al., 2003(Leal et al., , 2005. As a biological indicator of pollution, the species is recognized as relatively sensitive (Henriques-de-Oliveira et al., 2007). The order, Ephemeroptera, and also Plecoptera and Trichoptera, are considered sensitive to the environmental changes promoted by anthropic activities and, together (EPT), have produced a series of metrics based on both their densities and their occurrences, especially applicable to the diagnosis of small lotic environments, such as streams. Even so, the Richness of sensitive group (S sens ) metric, which includes the EPT families, was incorporated in the diagnosis of larger rivers and the sublittoral reservoir region in the Benthic Community Index (BCI) adopted by the Environmental Agency of São Paulo State (CETESB, 2018), in spite of its naturally lower richness and density in these types of environments.
The water quality monitoring network of CETESB evaluates the ecological quality of water bodies in the São Paulo State, through the incorporation of biomonitoring with aquatic macroinvertebrates since 2002. Until 2018, 280 assessments were performed, comprising 49 rivers and 38 reservoirs, with diagnoses for the sublittoral and profundal regions, from the 22 Water Resource Management Units (WRMU) of the state. In these surveys, the species Campsurus truncatus was often observed colonizing degraded environments, making questionable its classification as sensitive, as well as its incorporation in the S sens metric in the BCI multimetric index. This work evaluates the concordance between species occurrence and its sensibility and its application

Introduction
The Polymitarcyidae family of aquatic insects of the Ephemeroptera order is widespread, occurring at Holartic, Ethiopian-Oriental and Neotropical regions (Domínguez et al., 2006). In South America, the family is represented by three genera: Campsurus, Asthenopus and Tortopus. Of these, Campsurus is the genus with the highest number of species, 28 (Molineri et al., 2015). Their nymphs burrow U-shapped tunnels in bottom sediments of lentic habitats (Domínguez et al., 2006). The species Campsurus truncatus Ulmer 1920 occurs in Brazil, Bolivia, Colombia and Peru (Molineri et al., 2015). At Águas Claras reservoir (State of MG/Brazil), specimens of C. truncatus reached 2.4 cm in length and its population, with high secundary production (3.2 g.m -2 .year -1 ), was classified as multivoltine, with the aquatic phase lasting between 16 and 24 weeks (Lima & Pamplin, 2017). The potential high secondary production of the species is an important issue for the energy flow and nutrient cycling for the aquatic system, as they are preferential food item for some fish species (Meschiatti, 1995). In addition, it has been reported that populations of the genus play a significant role in the sediment bioturbation (Leal et al., 2003;Figueiredo-Barros et al., 2007), with potential to alter the carbon cycle (Leal et al., 2007) that, in impacted environments, can lead to the resuspension of contaminants and nutrients accumulated in the sediments to the water column. The burial behavior also modifies the substrate by altering its composition and increasing its oxygenation, which would favor colonization at Campsurus em amostras obtidas em rios e reservatórios do estado de São Paulo no período de 2001 a 2018, com pegadores. Fatores ambientais e os índices de estado trófico (IET) e da comunidade bentônica (ICB) foram comparados com a ocorrência e densidade desta espécie na tentativa de definir suas preferências ambientais. Resultados: A região sublitoral de reservatórios foi o habitat preferencial da espécie, sendo que C. truncatus ocorreu com maior frequência em ambientes mesotróficos e de qualidade ecológica regular, mas apresentou maiores densidades médias em ambientes eutróficos e de qualidade ecológica boa. As análises estatísticas não evidenciaram relação linear de suas densidades com nenhum dos fatores abióticos analisados, como a granulometria, conteúdo de matéria orgânica, OD e penetração de luz, embora tenha havido uma tendência para a ocorrência das maiores densidades em sedimentos lodosos, com menos de 1% de penetração de luz e com teores intermediários de matéria orgânica. Como descrito na literatura, a presença de plantas aquáticas pode de fato beneficiar a população de Campsurus truncatus, uma vez que os as densidades mais elevadas foram observadas em ambientes com muitas macrófitas, como o res. Salto Grande, mas é um fator que pode estar confundindo a relação populacional com as variáveis ambientais. Conclusões: A frequência de ocorrência e as elevadas densidades obtidas em ambientes eutrofizados sugerem que a espécie não deve ser considerada forma sensível em índices qualitativos, como a riqueza de espécies sensíveis utilizada no cálculo do ICB para rios e região sublitoral de reservatórios. Além disso, sua exclusão na riqueza de grupos sensíveis surtiu baixo impacto sobre os diagnósticos já realizados pela rede de monitoramento da CETESB.
in the diagnosis of the ecological quality of rivers and reservoirs.

Material and Methods
Triplicate samples were obtained in the winter/dry period in each habitat type, using Ponar or van Veen grab, in depositional banks of rivers and sublittoral zone of reservoirs, and Ekman-Birge grab, in profundal zone of reservoirs. Taxonomic identification to family-level and counting of the organisms were made in the laboratory under stereoscopic microscope. The identification of the species was performed by Dr C. Molineri (Molineri et al., 2015). Water transparency, chlorophyll-a and total phosphorus of surface water, dissolved oxygen in the bottom water, as well as grain size and organic matter content in the sediments were obtained on the same dates. Table 1 resumes the methods used for each variable.
The trophic state of the sites was evaluated by a modified Trophic Sate Index (TSI) (Lamparelli, 2004) that consider chlorophyll-a and total phosphorus data. To evaluate the benthic community integrity was applied the Benthic Community Index (BCI), a multimetric index that incorporates diversity, richness, sensitivity and tolerance measurements. The methods of calculations of both indices are described in CETESB (2018). Table 2 presents the potential diagnoses for the two indices and their representatives white and grey patterns. The depth to 1% light penetration, sufficient for the primary production, was obtained by multiplying the transparency result by factor 2.709 (Esteves, 2011). Campsurus densities were log (x + 1) transformed and related by Pearson correlation test with abiotic variables and TSI and BCI gradients. All statistical calculations were performed by the Past software (Hammer et al., 2001).

Results
The Polymitarcyidae Campsurus truncatus was observed in only 19 of the 84 environments investigated between 2002 and 2018, located in nine of 22 Water Resource Management Units (WRMU) of the state that comprise the basins of the Tiete (WRMU 6, 5, 10, 13 and 16), Mogi-Guaçu (WRMU 9), Paranapanema (WRMU 14), Paraíba do Sul (WRMU 2) and Juqueriquerê (WRMU 3) rivers ( Figure 1).   Although it occurred in the three habitat types sampled in the monitoring network, it was more frequent in reservoir sublittoral regions (Table 3), where it also reached the highest mean densities (Figure 2).
While the species occurs along almost the entire trophic gradient, except in the hypereutrophic condition, its frequency of occurrence ( Figure 3) shows a bell shaped distribution being more common in sites with some degree of organic enrichment (meso and eutrophic state), reaching higher mean values of density at eutrophic sites ( Figure 4). As the relationship was not linear, no significant correlation was found between the Campsurus density and the TSI values (r = -0.05; p = 0.7695) or classes (r = -0.15; p = 0.4029).
In the ecological quality gradient stated by the benthic community index (BCI), the species only occurred in the quality range between excellent and fair. Although more commonly found in fair quality    environments ( Figure 5), the mean density was higher in good quality ( Figure 6). Since the species was absent in the bad and very bad categories, the relationship between Campsurus densities and the BCI result was falsely linear, because it included only part of the quality gradient, and the correlation was significantly negative (r = -0.37 ; p = 0.03). No significant linear correlation was found between the densities of Campsurus truncatus and the abiotic data (Table 4) of sand% (r = -0.16, p = 0.3682), silt% (r = -0.03, p = 0.8446), clay% (r = 0.21, p = 0.2454), silt% + clay% (r = 0.16, p = 0.3654), light (r = 0.09, p = 0.999) and dissolved oxygen in deep water (r = 0.02, p = 0.9109). Even so, the XY graphs (Figure 7) show that there is a tendency for higher species densities in muddy, dark sediments with intermediate levels of organic matter.

Discussion
Most sampled in the sublittoral region of reservoirs, in this study Campsurus truncatus did not occur only under conditions of hypertrophy in the trophic gradient. Although most frequently observed in mesotrophic environments, it was in an eutrophic and macrophyte-rich environment that it reached its highest densities. No significant linear relationship was observed between densities and abiotic factors that could describe possible environmental requirements for the species, but a trend was observed for the occurrence of higher densities in muddy environments, with less than 1% light penetration and with intermediate content of organic matter.
Because of its burial habit, it was expected that the population of Campsurus truncatus would be Figure 7. XY plots relating Campsurus trucatus densities to abiotic data. For light, 1 = less than 1% of light penetration; 2 = more than 1% of light penetration. favored by fine sediments, similar to that observed by Leal & Esteves (2000) in Lake Batata, where the substrate alteration promoted by clay was favorable to the population of C. notatus. On the other hand, in a Uruguayan lagoon system, Clemente et al. (2018) obtained a negative correlation between the density of C. violaceus and the percentage of fine fractions (clay and silt) in one of the study years. The absence of a positive linear correlation between the density of the species and the fine grain size in our study may indicate that multiple factors act together to control its population.
The organic matter in the sediments, as well as the possibility of the presence of trophic chain components associated with the primary production, would be factors enriching the potential food source for the species, but no correlation was observed between the densities of Campsurus and these two factors, although the higher density values occurred in organic substrate (VS> 10%). In these cases, there was not enough light for photosynthesis to occur to sustain a grazing food chain. Although Clemente et al. (2018) obtained a positive correlation between the densities of C. violaceus and organic matter, our data showed a tendency for the highest densities of C. truncatus occur under intermediate values, corroborating the response of that parameter to the trophic gradient measured by the TSI.
Being a taxon belonging to the EPT group, it would be expected that Campsurus truncatus would be sensitive regarding the availability of dissolved oxygen. However, not only the dissolved oxygen concentrations did not have a significant linear and positive effect on the population, but the highest densities occurred under low concentrations (3.5 and 2.1 mg.L -1 ) of this variable, already considered restrictive for insect components of the benthic fauna (Heliövaara & Väisänem, 1993). Takeda et al. (2003) observed that the oxygen contained in the plant tissues of aquatic plants can supply individuals of the genus Campsurus in a situation of hypoxia, when the authors observed the migration of nymphs of the genus towards the lower portion of macrophytes. That adaptive behavior can explain its independence from the dissolved oxygen concentrations of the environment. Many of the sites where C. truncatus occurred presented aquatic plants, even as a result of the process of their enrichment. In addition to supplying oxygen for some taxa, macrophyte clumps represent habitat and food for the macroinvertebrates and their presence in the environment may have interfered in the expected relationship between the density of C. truncatus and the environmental factors tested. The presence of macrophytes diversifies habitats and can positively influence the richness of the community. Perhaps for this reason the highest mean densities of Campsurus were observed in environments considered good for the development of biota, but already eutrophic.
Populations of the genus Campsurus have been reported in domestic sewage receiving environments (Henriques-de-Oliveira et al., 2007), already eutrophic (Clemente et al., 2018) and impacted by mining residues (Leal & Esteves, 2000). Although not occurring in extreme situations of poor ecological quality, as other authors have also observed (Henriques- de-Oliveira et al., 2007), we suggest that the species Campsurus truncatus be disregarded as a sensitive taxon, especially in qualitative metrics, as they occur more frequently in environments with intermediate degrees of degradation. In a score system for rivers and reservoirs, as in the biotic qualitative or quantitative indexes commonly used in streams, the species would receive an intermediate score. The impact of the removal of the species from the calculation of richness of sensitive groups (S sens ) on the diagnoses of BCI expanded its occurrence to environments classified as poor concerning ecological quality. However it changed the diagnosis one quality class worst in only four cases (1.4% of total diagnoses): in the sublittoral zones of Taiaçupeba (2003), Billings (2008) and Guarapiranga (2006) reservoirs, which were poor, and in Claro river -UGRHI 3 (2013), which was regular. This low impact is due to the fact that the BCI is a multimetric index and the richness of sensitive groups is only one of its four components for reservoir's sublittoral and rivers assessment.