Rotifers of Bahia State , Brazil : News records and limitations to studies

Rotifera are microscopic euthelic metazoans (50-2000 μm), with approximately 2030 described species represented by two classes: Pararotatoria and Eurotatoria (Segers, 2002). The former includes only the order Seisonacea, while Eurotatoria comprises the subclasses Monogononta and Bdelloidea. Seisonacea is the least representative group, consisting of four epizootic marine species of branchiae crustaceans, Seison africanus, S. nebaliae, Paraseison annulatus, and P. kisfaludyi (Ricci et al., 1993; Sørensen et al., 2005; Leasi et al., 2012). Bdelloidea and Monogononta are the best known and most diversified, with approximately 2000 species between them (Monogononta ~1600, and Bdelloidea ~360) (Segers, 2007; Serra et al., 2019). Rotifers are cosmopolitan organisms, although largely restricted to continental waters, and constitute important components of the zooplankton in lakes, rivers, and reservoirs (Serafim et al., 2003; Tundisi and Matsumura-Tunsdisi, 2008); they are considered opportunistic, having well-developed adaptive abilities that allow them to quickly colonize a wide variety of habitats under favorable conditions (Sharma and Sharma, 2005; Bonecker et al., 2009). Rotifers make up part of the fundamental food chains of continental waters, occupying the ecological niche of small filterers. They are considered indicators of specific ecological conditions and can be used in evaluating the trophic states of bodies of water (Bërzinš and Pejler, 1989; Duggan et al., 2001). Some species develop well in highly eutrophic sites, while others are more sensitive to organic and chemical residues (Edmondson and Litt, 1982). Keratella cochlearis, Polyarthra vulgaris, Brachionus urceolaris, and Pompholyx sulcata, for example, are known to indicate eutrophic conditions due to their saprobic valences (Dorak, 2019). Abstract A first checklist of Rotifera species in freshwater environments in Bahia State, in northeastern Brazil, is provided. The list includes sampling data from 26 aquatic environments (lotic and lentic) undertaken from 2010 to 2016. One hundred and fifty-five species were recorded, with 68 new records for the state. The family Brachionidae and Lecanidae were the most representative (54.8%). The greatest richness was recorded in the Colônia River (57 species). Those results reflect the low numbers of studies previously undertaken in the region, indicating more research needs to be focused on Rotifera biodiversity in Bahia, the fifth largest state in Brazil (567,295 km2) with large numbers of freshwater bodies.


Introduction
Rotifera are microscopic euthelic metazoans (50-2000 μm), with approximately 2030 described species represented by two classes: Pararotatoria and Eurotatoria (Segers, 2002). The former includes only the order Seisonacea, while Eurotatoria comprises the subclasses Monogononta and Bdelloidea. Seisonacea is the least representative group, consisting of four epizootic marine species of branchiae crustaceans, Seison africanus, S. nebaliae, Paraseison annulatus, and P. kisfaludyi (Ricci et al., 1993;Sørensen et al., 2005;Leasi et al., 2012). Bdelloidea and Monogononta are the best known and most diversified, with approximately 2000 species between them (Monogononta ~1600, and Bdelloidea ~360) (Segers, A complete inventory of species and reliable quantitative data are required to accurately describe the structures and functioning of zooplankton communities (Karjalainen et al., 1996) -important groups in terms of our knowledge of global biodiversity, and relevant to all macroecological analyses (Tittensor et al., 2010). A total of 625 Rotifera species are known to Brazil (representing 84 genera), including 103 species of Lecanidae and 72 of Brachionidae (the most representative families) (Garraffoni and Lourenço, 2012).
It is estimated that about 14% of the Rotifera species recorded globally can be found in Brazil (Lewinsohn and Prado, 2002), although the true biodiversity of that group in continental aquatic ecosystems in that country is still only poorly understood and difficult to estimate (Agostinho et al., 2005). While a number of studies focusing on the Rotifera community in the continental waters of Bahia State have been published (Neumann- Nogueira-Paranhos, 1987, Crispim andWatanabe, 2000;Souza et al., 2004;Oliveira et al., 2015;Araújo andNogueira, 2016, Santos et al., 2019), most were reports, monographs, dissertations, and theses. Bahia is the largest state in northeastern Brazil (covering approximately 564,732 km 2 ) (IBGE, 2017) and detailed Rotifera inventories will surely identify significant species richness there.
We undertook a taxonomic inventory of the planktonic Rotifera inhabiting continental aquatic environments in Bahia State (the first such survey for the region), citing new occurrences and discussing the limitations of studies of those microorganisms.

Material and Methods
The list of species presented here was based on a literature search for published studies concerning the Rotifera of Bahia State (Neumann- Leitão and Nogueira-Paranhos, 1987;Crispim and Watanabe, 2000;Souza et al., 2004;Oliveira et al., 2015;Araújo and Nogueira, 2016;Santos et al., 2019). Sampling was also undertaken at 23 sites in lentic and lotic environments, during the period between 2010 and 2016, including the southern portion of the state (near the cities of Vitória da Conquista and Mucugê) (Table 1) (Figure 1). Three sites were sampled in the Lagoa das Bateias lake, three sites in the reservoir at the Fazenda Beija-flor farm, four sites in the Cumbuca River, four sites in the Piabinha River, and nine sites in the Cachoeira River basin.
Published articles concerning field collections of Rotifera generally employed sample volumes of 25 to 100 L of water and horizontal drags for 10 minutes, except Santos et al. (2019), who collected 100 L of water using a graduated bucket. The samples were collected using plankton nets with 50 μm, 65 μm, and 68 μm meshes, and the captured specimens were fixed with 4% formaldehyde; in some cases they were neutralized with sodium tetraborate, and some with a saturated sugar solution or precipitated calcium carbonate.
The surveys we conducted between 2010 and 2016 used plankton net nets (20 μm mesh) and the specimens were fixed in 4% formaldehyde buffered with Hexamethylenetetramine. The samples were obtained using horizontal subsurface drags with a 30X70 cm conical plankton net for five minutes per collecting site, covering a drag distance of approximately 10 m 2 (except in the CRB, where 400 L of water (using a graduated bucket) per point was filtered.
The samples were processed in Sedgewick-Rafter type chambers and viewed using an optical microscope; the individual Rotifers found were separated on slides with glycerin for better visualization and manipulation. When necessary, a 75% hypochlorite solution was used for trophos extraction, and rose bengal staining was used to better visualize the specimens. An Olympus CX31 microscope with a coupled digital camera was used to photograph the specimens.
Some forms of Rotifera, principally the Bdelloidea group, and representatives of the families of Dicranophoridae and Notommatidae, must be alive for accurate taxonomic identification, as the fixation solution causes contraction into their lorica. The samples analyzed from 2010 to 2016 were fixed in formaldehyde, however, and thus only useful for studying the Monogononta group. Segers (2007) was consulted to elucidate questionable nomenclature for the species; those not classified are indicated in the list of species occurrences. Only the lowest possible taxonomic level was considered in the compilation of data presented in the published literature (e.g., Polyarthra sp. de Souza et al. (2004) was not included in the list).
The frequency of occurrence (FO%) of the species was determined considering the number of samples in which they occurred in relation to the total number of samples, with the species being classified as constant (present in more than 80% of samples), frequent (present in between 50 to 80% of the samples), common (from 20 to 50%), or rare (< 20%) (Dajoz, 1983).
The Jackknife2 richness estimator was calculated based on the presence and absence data of species using EstimateS 9.1 software. We generated an accumulation curve to verify that the analyzed samples were sufficient to estimate Rotifera richness using asymptotic behavior analysis (Santos et al., 2019). The average percentage of richness extrapolation was calculated based on Heck Junior et al. (1975).
The specimens collected between 2010 and 2016 were deposited in the limnological collection of the Federal University of Bahia -Anísio Teixeira Campus and in the Laboratory of Plankton Ecology at the State University of Santa Cruz (UESC).
Analyses of the numbers of species in the sampled areas showed that the Reservoir of Pedra do Cavalo (70 species) was the most species rich, followed by Colônia river (57 species), the rivers Cachoeira, Salgado, and Lagoa das Bateias (47 species, each), and the reservoir at Fazenda Beija-flor (22 species); the least representative areas were: the tilapia cultivation tank, the São Francisco River (municipalities of Sobradinho and Rodelas), and the Salgado River (each 10 species), followed by the Sobradinho reservoir (nine species) ( Figure 5).
The rarefaction curve evidence that the Rotifera samples taken in Bahia were not sufficient to define an asymptotic trend ( Figure 6). The average percentage of wealth extrapolation calculations indicated the rotifer data was able to assess approximately 67% of the species.

Discussion
The only checklist available of the Rotifera of northeastern Brazil was prepared by Melo and Almeida (2007) for Pernambuco State, and listed 64 planktonic species for 19 freshwater environments. That total is lower than described here for Bahia. Souza-Soares et al. (2011) studied 250 aquatic environments in São Paulo State and recorded 277 species, highlighting the potential for significant increases in species richness if additional efforts were undertaken in Bahia.
Brachionidae and Lecanidae, the most representative families in our study, are highly diversified in the tropics (Segers, 1995), and considered the main representatives of Rotifera in tropical freshwater environments in South America (Rocha et al., 1995;Aoyagui and Bonecker, 2004;Melo and Almeida, 2007). Among the genera registered in Bahia State, Brachionus (considered endemic to Australia and South America) (Dumont, 1983) and Lecane stand out; the species included in those genera have been identified as bioindicators of eutrophic environments (Sládecek, 1983;Pontin and Langley, 1993). Lecane is a predominantly subtropical or warm-water genus, with numerous regional and local endemics, but also comprises numerous Holarctic, Palaeotropical, Australasian, New World, and Old World species (Segers 2008). They are predominantly nonplanktonic (Borges and Pedrozo, 2009) and associated with aquatic macrophytes (Duggan et al., 2001;Kuczyn'ska-Kippen, 2009), although they are frequently recorded as plankton.
Although the Rotifera richness in Bahia can appear high when compared to other studies with more intense sampling efforts (Souza-Soares et al., 2011), the accumulation curve for Rotifera species has not yet reached its asymptote, indicating that much more sampling needs to be done. Heck Junior et al. (1975) noted, however, that inventories where 50 to 75% of the species represent common taxa can be considered satisfactory. Due to differences in sampling efforts, it was not possible to infer Rotifera diversity patterns in the available published studies. Reservoir systems, natural lakes, and rivers are all distinct environments, with reservoirs generally being more complex than natural lakes due to their interactions with entire river basins, and the inflow from many tributaries (Straskraba and Tundisi, 1999). Aquatic communities will therefore vary among different bodies of water, with their compositions and dynamics being influenced by surrounding environmental conditions and dependent on a variety of factors -whether local or regional. Biological, physical, and chemical factors, as well as interactions between them, all play important roles in the selection of predominant species (Casanova et al., 2009).
In the present study, the Cachoeira River Basin and Lagoa das Bateias were observed to have the highest nutrient concentrations, mainly due to the inflow of organic sewage, quite different from the Cumbuca and Piabinha rivers in the  Table 1.      the other hand, support smaller numbers of plankton species but larger populations of each (Maitland, 1990). Therefore, to fully understand Rotifera diversity patterns in Bahia State, taxonomic and ecological studies will need to be combined. The shortage of studies in Bahia, together with the lack of Rotifera specialist in northeastern Brazil have resulted in artificially lower established richness of that group as Sempre Vivas Municipal Park. High densities of Brachionus angularis angularis have been found to be associated with high nutrient concentrations (Branco and Senna, 1996;Slàdec �ek, 1983), and that species was encountered in three rivers in the Cachoeira river basin (Table 2).
phenotypic plasticity (Segers and De Smet, 2008), and factors such as temperature and predation that can modify their morphological characteristics and make identifications much more difficult (Gilbert, 2011). As such, the complex morphological details of small metazoans and, in some cases, their polymorphic cycles, tend to inflate the numbers of species (Finlay et al., 1996).
Among the representatives of Rotifera (Monogononta, Bdelloidea, and Seisonacea), taxonomic difficulties are most evident among the Bdelloidea, one of the most sustained clades of ancient asexuals (Butlin, 2002). They reproduce only by parthenogenesis, as compared to the states Mato Grosso do Sul (Roche and Silva, 2017) and São Paulo (Souza-Soares et al., 2011). Nonetheless, 155 species were recorded here, including 68 new occurrences, and most freshwater environments in Bahia have not yet even been sampled. Many news sites will need to be surveyed and many aspects of the ecology and physiology of that group investigated (which will also contribute to our knowledge of New World biodiversity).

Limitations for taxonomic studies of Rotifera
The taxonomy of the Rotifera can be extremely complex due to wide morphological variations observed within the group, cyclomorphosis, their capacity for     Table 1. no males or signs of meiosis have ever been observed (Welch et al., 2004), and demonstrate high morphological variability. Furthermore, they retract during the process of chemical fixation, so that live organisms are necessary for the identification of the group. The use of anesthetics such as marcaine (0.5%) (Fontaneto et al., 2008) and carbonated water (De Carli et al., 2017) are recommended, although those methods are not always efficient. Orstan (2015) recently proposed a new method for fixing bdelloids: the hot glutaraldehyde method, which preserves those organisms in superior conditions, suitable for taxonomic identification and internal anatomical studies.
The accuracies and reliabilities of data sets in which rotifer taxonomies are not fully resolved and that contain cryptic species are inevitably low (Malekzadeh-Viayeh et al., 2014). Among Rotifers, the species complex Brachionus plicatilis Müller, 1786 (Monogononta, Brachionidae) is a classic example of high diversity that remained hidden and unclear using only the tools of morphological taxonomy. Research involving a range of genetic techniques, however, has improved our knowledge of the diversity of those species complexes (Gomez et al., 2002).
Among Rotifera, mainly the loricate organisms, taxonomic identifications depend on analyzes of the ultrastructure of the mastax (a modified pharynx composed of muscle sets and trophos) (Obertegger et al., 2006). The use of scanning electron microscopy (SEM) allows a more detailed visualization of the ultrastructure of the trophos that cannot be obtained using light microscopy. A thorough examination of trophos using light and scanning electron microscopy should be part of all Rotifera taxonomic studies (De Smet, 1998), and their morphological taxonomy should be aligned with ecological and molecular taxonomic studies to increase the precision of the diagnoses (Roche and Silva, 2017).
The use of valid nomenclature, excluding synonyms, is also a recurrent problem, and recent research has addressed synonyms referred to as valid species or variant names of the subspecies (Souza-Soares et al., 2011). Synonyms are used in many Rotifera species based on morphological characters, such as in the genus Brachionus where Brachionus angularis orientalis (Sudzuki 1989) = B. angularis (Gosse 1851) (Segers, 2007