Periphytic diatoms from an oligotrophic lentic system, Piraquara I reservoir, Paraná state, Brazil

Knowledge of biodiversity in oligotrophic aquatic ecosystems is fundamental to plan conservation strategies for protected areas. This study assessed the diatom diversity from an urban reservoir with oligotrophic conditions. The Piraquara I reservoir is located in an Environmental Protection Area and is responsible for the public supply of Curitiba city and the metropolitan region. Samples were collected seasonally between October 2007 and August 2008. Periphytic samples were obtained by removing the biofilm attached to Polygonum hydropiperoides stems and to glass slides. The taxonomic study resulted in the identification of 87 diatom taxa. The most representative genera regarding the species richness were Pinnularia (15 species) and Eunotia (14 species). Five species were registered for the first time in Brazil and seven in the State of Paraná. Taxonomic and ecological comments of the species registered are provided.


Introduction
Diatoms are considered one of the most representative groups of microalgae due to the number of described species.Approximately 12,000 species of diatoms have been described hitherto (Guiry 2012) although the estimated number of species is thought to be at least 30,000 (Mann e Vanormeligen 2013).Diatoms are excellent bioindicators of water quality, being sensible to environmental pollution and a number of anthropogenic pressures, such as the increase in nutrient concentrations, acidification, land use, and the presence of toxic chemicals dissolved in water (Leskinen & Hällfors 1990, Lowe & Pan 1996, Lobo et al. 2002, Rimet et al. 2015).
Diatoms release mucilage through raphe, rimoportulae and apical pore fields, facilitating the frustule adhesion to different types of substrates.This strategy contributes to diatoms representativeness in the periphytic community (Round et al. 1990, Lowe 1996).The species composition of periphytic assemblages can be influenced by the substrate micro-topography (Murdock andDodds 2007, Souza &Ferragut 2012).Inventories using substrates with different surface types usually hold a great algal diversity (Ács et al. 2000).Therefore, diatom inventories should be carried out in a great number of different environments and using diverse substrates to maximize the recovery of new species, extend the distribution of known species and better define their ecological preferences, which increase the reliability of environmental diagnostics.
Man-made lentic environments with oligotrophic conditions are rare in urban areas and are examples of environmental health.Unfortunately, human activities in the surrounding watershed have been accelerating the processes of eutrophication, affecting the water quality and the biodiversity, leading to the loss of important ecological functions (Tundisi 2003, Torrisi et al. 2010).Due to the undesirable impact of these activities on the water physical and chemical conditions, efforts have been generally focused on studying eutrophic environments rather than on preserving oligotrophic water bodies.For this reason, there is usually a gap in the knowledge of the algal diversity of protected areas, which creates exceptional opportunities to study the remaining oligotrophic environments (Kociolek & Stoermer 2009).
Recent studies in oligotrophic Brazilian reservoirs contributed to the description of new diatom species in the genera Kurtkrammeria and Encyonema (Marquardt et al. 2016, Marquardt et al. 2017) and extended the geographic distribution of the already known species (Canani et al. 2011).Paleolimnological studies with diatoms are important in providing a further understanding of biodiversity, detecting the response of the assemblages over time.Generally, a decrease in the diatom diversity during the eutrophication process can be observed when comparing environments that were oligotrophic but are currently eutrophised (Faustino et al. 2016, Wengrat et al. 2017).
Previous diatom inventories were carried out in urban reservoirs of the Iguaçu river basin along a trophic gradient ranging from mesotrophic (Passaúna, Bertolli et al. 2010 andPiraquara II, Marra et al. 2016), eutrophic (Iraí, Silva et al. 2010) to hypereutrophic (Itaqui, Faria et al. 2010).The Piraquara I reservoir, which was surveyed in the present study, is connected to Piraquara II reservoir.The macrophyte Polygonum hydropiperoides Michaux and glass slides were used as substrates to assess the periphytic diatoms in these reservoirs.
The present study aimed at assessing the composition of periphytic diatom assemblages from the oligotrophic reservoir Piraquara I. We provided measures, illustrations, taxonomical comments and ecological data based on scientific literature, in order to contribute to the knowledge of Brazilian diatoms from oligotrophic environments.

Material and Methods
The Piraquara I Reservoir (25°30'24.16''S, 49°1'29.4''W) is a man-made public water supply of the Iguaçu River basin located in an Environmental Protection Area which is in a transition zone between the dense and mixed ombrophilous forest.The Cayuguava river was dammed to construct the reservoir in 1979 (Figure 1) (Guimarães 2008).The reservoir has a 3.3 km 2 of flooded area, 7 meters in depth, a flow rate of 600 l/s, and a water residence time of 438 days (Júnior et al. 2005).
The physical and chemical data were provided by the Paraná Environmental Institute (IAP) (unpublished data) and by the water company of Paraná State (SANEPAR) (unpublished data).The Trophic State Index (TSI) was calculated according to Lamparelli (2004) (Zorzal-Almeida et al. 2017).
Samplings were carried out seasonally (i.e. in spring, summer, autumn, and winter) from October 2007 to August 2008, at two sampling stations.Periphytic samples were obtained from natural and artificial substrates.As natural substrates, we used stems of the macrophyte Polygonum hydropiperoides Michaux that were collected near the reservoir margin.The periphyton was obtained by scraping the stems with a razor blade wrapped in foil.Glass slides (7.5 cm x 2.5 cm) were used as artificial and inert substrates fixed to woody frames that were left submerged in the water column for 30 days.The artificial substrates were placed at about 80 meters far from the margin of the reservoir to prevent being eventually trapped in the macrophytes by the wind action.
Substrate fragments were fixed in Transeau's solution (6:3:1) (Bicudo & Menezes 2017).The samples were treated by oxidation according to the method of Simonsen (1974) modified by Moreira-Filho & Valente-Moreira (1981).The material was air-dried onto glass slides and mounted in Naphrax ® resin.Observations, measurements, and diatom photomicrographs were performed at 1000× magnification in an Olympus BX40 microscope equipped with a DP71 Olympus camera.Diatom identification was performed up to the lowest level of taxonomic hierarchy based on recent and classic literature (e.g.Lange-Bertalot 1993, 1999, 2001, Lange-Bertalot & Metzeltin 1996, Metzeltin & Lange-Bertalot 1998, 2002, 2005, 2007, Bahls 2015).For each taxon, we included information on morphometry, remarks on the autoecology (when available in literature) as well as on the occurrence in the samples (see also table 2).Taxonomic comments were provided only for poorly known species in Brazil.We considered that a species is recorded for the first time in Brazil and in the state of Paraná if it was not previously recorded in a published article.The studied materials were deposited at the herbarium of the State University of Paraná (UPCB) under the numbers 63371 to 63374 (glass slides) and 63375 to 63378 (macrophyte P. hydropiperoides).

Results and Discussion
The Piraquara I reservoir was classified as oligotrophic by the TSI index, characterized by low concentrations of phosphorus and nitrogen, high transparency of the water column (up to 3.2 meters), and a slightly acid to nearly neutral pH.Summer showed the highest values of temperature and accumulated rainfall occurred in the summer sample (27°C and 326.9 mm, respectively).The physical and chemical water parameters are shown in Table 1.
We identified 88 infrageneric taxa that belong to 35 genera.Pinnularia Ehrenberg (15 spp.) and Eunotia Ehrenberg (14 spp.) were the most representative diatoms in both substrates.The macrophyte P. hydropiperoides showed the highest richness (77) and the highest number of exclusive species (35) that represented 45.4% of the total substrate richness, whereas the artificial substrate presented 52 species and 10 exclusive taxa, which corresponded to 19.2% of the total substrate richness.Pinnularia and Eunotia (15 and 14 species, respectively) presented a higher species richness in Piraquara I reservoir.The presence of these genera was favored by the slight acidity, low conductivity and water oligotrophy conditions of the studied reservoir (Round et al. 1990, Metzeltin & Lange-Bertalot 1998, Spaulding & Edlund 2009, Costa et al. 2017).Eunotia was also the most representative genus in a eutrophic reservoir (Silva et al. 2010) located near to the Piraquara I.However, only three Eunotia species (E.naegelii, E. minor and E. subarcuatoides), which have a wide tolerance to trophic gradients (Van Dam et al. 1994, Silva et al. 2010, Costa et al. 2017) and they are able to occur under oligotrophic and eutrophic conditions, were common to both reservoirs.
The genera Navicula, Gomphonema, and Nitzschia, which are commonly registered in nutrient-rich environments (Goldsborough & Robinson 1996), showed a high species number in both substrates at the mesotrophic and hypereutrophic reservoirs located nearby to Piraquara I (Bertolii et al. 2010, Faria et al. 2010, Marra et al. 2016).
The richest assemblage was obtained from Polygonum hydropiperoides (natural substrate), 77 species, and can be explained by the glabrous or pubescent petiole of this macrophyte (Melo 2008), which offers a greater architectural complexity than the smooth and uniform surface of glass slides.The roughness difference between substrates usually does not contribute to an increase in algal biomass, but it is an important factor in selecting the species composition of the periphyton (Burkholder 1996, Souza & Ferragut 2012).
The diatom species richness found in the oligotrophic Piraquara I reservoir was lower than in the mesotrophic reservoirs Piraquara II and Passaúna (135 and 106 taxa, respectively), the eutrophic reservoir Iraí (96 taxa) and the hypereutrophic reservoir Itaqui (124 taxa) (Bertolli et al. 2010, Silva et al. 2010, Faria 2010, Marra et al. 2016).The Piraquara II is a mesotrophic reservoir located downstream from Piraquara I.Both reservoirs shared the presence of 31 diatom species, of which 11 were recorded from oligotrophic to mesotrophic environments and 20 species had a wide trophic tolerance.The diatom assemblage had more species that are tolerant in Piraquara II than in Piraquara I.Although eutrophication might lead to a loss in diatom diversity (Wengrat et al. 2017), the species richness in shallow lakes might be also influenced by the presence, abundance and diversity of macrophytes (Sayer et al. 1999, Bicudo et al. 2007).Piraquara II and Itaqui had more diatom richness than the other reservoirs, which can be related to the fact that in the former reservoir were sampled three different species of macrophytes (Marra et al. 2016), and that the latter reservoir was almost totally covered by the floating macrophyte Pistia stratiotes L. (Faria et al. 2010, Faria et al. 2013).
Ecology: this species is fairly common in alpha-mesosaprobic and eutrophic conditions (Cantonati et al. 2017, van Dam et al. 1994).In contrast to previous studies, it was common in our samples, indicating a tolerance to oligotrophic waters, poor in nutrients, with slightly acid to circumneutral pH (6-6.85), and low conductivity (24-24.5 μS cm-1 ).
Although it is cosmopolitan and often present in periphyton samples from lakes, rivers, and streams of temperate regions (Delgado et al. 2015), it is rarely found in tropical regions (Silva et al. 2010, Tremarin et al. 2009 The specimens analyzed in the present study are shorter, with a lower striae density, and the absence of a central area.Hofmann et al. (2013) referred specimens of 30-100 µm long, 2-3 µm wide and with a striae density of 17-20 in 10 µm, and the often absence of a central area.
Ecology: considered as indifferent to tolerant to the trophic state of the environment (van Dam et al. 1994, Hofmann 1994).In our samples, it occurred in circumneutral pH (6.75), low conductivity (24.5 μS cm-1 ) and oligotrophic conditions, only in summer (temperature of 27°C and accumulated rainfall 326.9 mm).
Family  (Costa et al. 2017, Reichardt 1995).Ecology: previously registered for mesotrophic environments, with a preference for slightly acidic waters (pH 6.6) with medium values of conductivity (71 µS cm -1 ) (Costa et al. 2017).In our samples, it occurred only in winter (temperature 16°C) and expanded the occurrence to oligotrophic environments with slightly acid pH (6), and low conductivity (24.5 μS cm -1 ).This is the first record for Paraná state.
Occurrence in samples: natural substrates (UPCB 63375).Ecology: previously registered for moody pools with sphagnum moss (Krasske 1948), and from samples of surface sediments collected in environments with acid pH and poor in nutrients (Costa et al. 2017).
Ecology: considered as cosmopolitan (Krammer 2000); it is rare in waters with neutral to alkaline pH and in moderate to high conductivity (Cremer et al. 2004, Noga et al. 2014).It was previously registered for streams with acid pH and low conductivity, in Central Brazil (França et al. 2017).In our study, it was recorded from samples of periphyton associated with aquatic macrophytes, in temperatures of 16°C, acid pH (6), low conductivity (24.5 μS cm -1 ), and oligotrophic conditions.This is the first record for Paraná state.
Occurrence in samples: artificial substrates (UPCB 63771).Reichardt by the presence of subcapitate apices, narrower valves and a higher striae density.The striae are difficult to distinguish under the light microscope and the areolae are inconspicuous (Reichardt 1995, Lange-Bertalot & Metzeltin 1996).In this species, the proximal ends of the raphe are straight and not inflated and teardrop-shaped as in S. gracilior (Cantonati et al. 2017).

Figures: 4R-4T
Length: 16-20 µm; width: 4.5 -5 µm; striae: 16-18 in 10 µm; fibulae: 7 in 10 µm.Nitzschia semirobusta differs from Nitzschia amphibia Grunow by the presence of longer fibulae and from Nitzschia amphibioides Hustedt by having a greater striae density.Denticula kuetzingii Grunow is distinguished from Nitzschia semirobusta by the presence of fibulae that are extended completely and with similar thickness from margin to margin in the first species (Lange-Bertalot 1993), while in the latter species the fibulae decrease in thickness towards the margin.Ecology: Due to the taxonomic problems and misidentification, the ecological preference of N. semirobusta is unclear.Therefore, information on its occurrence and ecological preferences is of high relevance for delimiting the ecological requirements and tolerances of N. semirobusta.It was frequently registered for oligo-to mesotrophic reservoirs in São Paulo state (Bartozek et al. 2018).In our samples, it occurred in slightly acid to circumneutral pH (6-6.85),low conductivity (22.5-24.5 μS cm -1 ) and oligotrophic conditions.This is the first record for Paraná state.

Conclusions
The inventory performed at the Piraquara I reservoir brings an important contribution to the diatom diversity from oligotrophic environments.We registered the first occurrence of five diatom species in Brazil and seven in the state of Paraná, contributing to the geographic

Table 1 .
Physical and chemical characteristics of the Piraquara I reservoir registered seasonally between October 2007 and August 2008.

Table 2 .
Diatom species and seasonal occurrence on artificial and natural substrates in Piraquara I reservoir.