Hidden diversity in waterfall environments: The genus Acrorbis (Gastropoda: Planorbidae) from the Upper-Paraná Atlantic Forest

High-energy freshwater environments such as rapids and waterfalls in the Upper-Paraná Atlantic Forest are home to highly endemic minute freshwater snails of the genus Acrorbis. Only one species, Acrorbis petricola, is currently included within this genus, whose geographical distribution is restricted to three known populations, one in Brazil and the other two in Argentina. Because of habitat specificity and limited geographical distribution, the species is considered vulnerable in Argentina and endangered in Brazil. In this work, we identify five new populations of A. petricola in southern Upper-Paraná Atlantic Forest, exclusively found on waterfalls from the Misiones Province, Argentina. Based on these populations and on specimens of one of the two historical populations from the Misiones Province, we explored the morphological features of shells and reproductive system of specimens from each location and provide the first molecular data on the species. We used DNA sequences from cytochrome c oxidase subunit I (COI) and 16S-rRNA genes to investigate the molecular diversity, genetic distances and genealogical relationships among populations. We verified the existence of intra- and interpopulation morphological variability, with the greatest variation being found in spire, spiral sculpture, penis sheath, flagella, prostatic diverticula and bursa copulatrix. We found interpopulation genetic diversity, with no intrapopulation variation, and identified six geographically structured genetic lineages with maximum genetic distances of up to 2.3%. Different combinations of morphological characters with the same genetic background within each locality were observed. The finding of new populations genetically differentiated not only broadens the known distribution of the species, but also illustrates that waterfall environments in the Atlantic Forest harbour a hidden diversity of Acrorbis that still remains to be discovered. This scenario suggests a complex evolutionary history that needs to be unveiled and taken into account for future development of conservation strategies in this endemic genus.


Introduction
Waterfalls and river rapids are natural laboratories for evolutionary research as they usually host a highly endemic flora and fauna, especially adapted to the constant humid conditions and the force of water, including several molluscan species [1][2][3][4]. Such high-energy freshwater environments are abundant in the Atlantic Forest, which extends across southeastern Brazil, eastern Paraguay and northeastern Argentina [5,6].
The Upper-Paraná Atlantic Forest, the southwestern and largest ecoregion of the Atlantic Forest, constitutes one of the "hottest hotspots" of biodiversity and endemism worldwide [5,7]. Despite ongoing landscape transformation, the Misiones Province in Argentina currently contains the largest remaining tract of this ecoregion [6,8], where numerous streams and rivers still maintain complete native species assemblages, including a high species richness of freshwater gastropods [9][10][11]. Nonetheless, in spite of being one of the most explored areas at malacological level, the molluscan biodiversity inhabiting waterfalls and rapids in the Upper-Paraná Atlantic Forest remains largely unknown [1,[12][13][14].
Snails of the genus Acrorbis Odnher, 1937 (Gastropoda: Planorbidae) represent a clear example of endemic micromolluscs occurring in poorly studied high-energy freshwater habitats of the Upper-Paraná Atlantic Forest, for which currently available knowledge is scarce and fragmentary. Odhner [15] introduced the generic name Acrorbis to accommodate Acrorbis petricola Odhner, 1937, an atypical planorbid described from specimens collected in rocky banks of the Ariranha River, Nova Teutônia, Santa Catarina State, Brazil [15][16][17][18][19][20]. Under the name Acrorbis odhneri Hylton Scott, 1960, a new species was added to the genus from Argentina based on specimens collected in Salto Encantado, a waterfall located in the Salto Encantado Provincial Park, Aristóbulo del Valle municipality, Misiones Province [21][22][23][24]. Subsequent studies based on shell, anatomy and radular morphology of the species from Salto Encantado, verified that Argentinian specimens were indistinguishable from the Brazilian congeners, and consequently A. odhneri was regarded as a junior synonym of A. petricola [22]. A new comparative study in the 1990s, based on the finding of Acrorbis individuals in the Iguazú National Park, Misiones Province, Argentina, showed some variability in shell morphology and male terminal genitalia in relation to previous descriptions [24]. Nonetheless, based on the Iguazú specimens and on a reexamination of the type-series of A. odhneri, that study also suggested that both nominal species were conspecific under the name A. petricola [24].
To date, Acrorbis is considered to be monotypic [9][10][11]14,22,24,25] and is known only from a geographical area restricted to three localities in northeastern Argentina and southern Brazil: Salto Encantado Provincial Park (Salto Encantado waterfall) and Iguazú National Park (Salto Arrechea, Salto Dos Hermanas and Salto Rivadavia waterfalls), both in the Misiones Province, and Nova Teutônia, Santa Catarina State, its type locality in Brazil [14,[24][25][26]. The species has always been found associated with high-energy freshwater environments in a landscape dominated by subtropical forest. The specimens seem to have restricted microhabitat preferences, such as the water-air interface on the surface of rocky substrates covered by moss and epilithic algae in the area kept wet by the water [1,22,24,26]. Owing to its limited distribution and niche specificity, the species is recognized as vulnerable in Argentina and endangered in Brazil [11,27]. However, no conservation strategies have been adopted for this species yet.
In this study, we surveyed waterfall environments in central and southern Misiones Province, Argentina, and identified new Acrorbis populations occurring at the Upper-Paraná Atlantic Forest. The objectives of this paper are: (i) to confirm the taxonomic identity of the populations using traditional morphological criteria based on shell morphology and morphological features of the reproductive system; (ii) to investigate the morphological variability among localities; and (iii) to explore the genetic background of the populations and to provide the first molecular data for the species. In this process we gained insights into a hidden diversity of Acrorbis in waterfall environments of the Atlantic Forest.

Snail collecting and preservation
Living snails were collected by hand in six waterfall environments in central and southern Misiones Province during mollusc surveys conducted between 2017 and 2018. The waterfall known as Salto Encantado in Salto Encantado Provincial Park was specifically surveyed as it represents the first historical record for Acrorbis petricola in the Misiones Province [21]. Permission for collection was granted by Ministerio de Ecología y Recursos Naturales Renovables de la Provincia de Misiones (Disp. No. 027/2018). Specimens were mostly collected among mosses and epilithic algae directly on the wet walls of the waterfalls, or on nearby rocks at the base of the waterfalls, in the area sprayed by water.
Living specimens were relaxed in water with menthol crystals for 30-60 min and immersed in hot water (80˚C) for 30-60 sec. Soft parts were then separated from the shell and preserved in 96% ethanol for anatomical and molecular studies. Shells were firstly cleaned with a fine paint brush in a 10% sodium hypochlorite solution for 10 min to remove vegetation and other encrustations from external shell surface, and then washed with distilled water. Subsequently, shells were cleaned in isopropanol using an ultrasonic cleaner for 5-10 sec, and then air dried [28].
Voucher specimens were deposited in the malacological collection of the Instituto de Biología Subtropical (IBS-Ma), CONICET-Universidad Nacional de Misiones (UNaM), Misiones Province, Argentina. In addition, we obtained ethanol preserved tissues of museum specimens of A. petricola from Nova Teutônia, Brazil, and Iguazú National Park, Argentina, housed in the malacological collection at the Museo de La Plata (MLP-Ma; La Plata, Argentina). These samples were included herein in order to explore the genetic background of the other historical records of A. petricola in the Atlantic Forest, including its type locality.

Morphological examination
Digital images in dorsal, ventral, and lateral views were obtained for ten specimens from each of the new localities with a Canon 6D camera equipped with a Nikon BD Plan 10X objective. Five shell measurements were taken in that material (Fig 1), namely the length (L = maximum dimension in the basal plane), width (W = maximum dimension perpendicular to L in the basal plane), height (H = maximum dimension perpendicular to the basal plane), aperture length (AL = maximum dimension of the aperture parallel to W, measured in the apertural plane), and aperture width (AW = maximum dimension of the aperture perpendicular to AL). Number of whorls (NW) was determined according to Diver [29]. Shells were measured using the software ImageJ 1.49 [30]. Shell measurements were normalized by logarithmic transformation and subjected to a principal component analysis (PCA) using PAST 3.25 [31] in order to explore the morphological variation among the different populations.
DNA extraction, PCR amplification, and DNA sequencing DNA was extracted from 1 mm 3 samples of foot muscle of ethanol-preserved individuals with an ADN PuriPrep-T kit (INBIO-Highway, Tandil, Argentina) following the manufacturer's instructions. We selected 30 samples belonging to A. petricola (5 specimens per waterfall environment, including those anatomically dissected specimens) and the outgroup species Antillorbis nordestensis (Lucena, 1954) and Biomphalaria peregrina (d'Orbigny, 1835). Collection information and GenBank accession numbers for the samples analysed are presented in Table 1.
Partial sequences of the mitochondrial 16S-rRNA (hereafter 16S) and cytochrome c oxidase subunit I (COI) genes were amplified using the primers 16SF-104 (5 0 -GAC TGT GCT AAG GTA GCA TAA T-3 0 ) and 16SR-472 (5 0 -TCG TAG TCC AAC ATC GAG GTC A-3 0 ) for 16S [32], and the primers LCO1490 (5 0 -GGT CAA CAA ATC ATA AAG ATA TTG G-3 0 ) and HCO2198 (5 0 -TAA ACT TCA GGG TGA CCA AAA AAT CA-3 0 ) for COI [33]. PCR amplifications of the 16S and COI genes were conducted as in Rumi et al. [34] and Vogler et al. [12], respectively, and run on a T18 thermocycler (Ivema Desarrollos). Success of PCR reactions was verified by agarose gel electrophoresis. Successfully amplified products were purified with an AccuPrep PCR Purification Kit (Bioneer, Daejeon, Korea). Owing to the co-amplification of nonspecific fragments, some PCR products were purified from 1.5% (w/v) agarose gel using an ADN PuriPrep-GP kit (INBIO-Highway, Tandil, Argentina). After purification, each gene was then directly cycle-sequenced in both directions (Macrogen Inc., Seoul, South Korea). Newly generated sequences and chromatograms were visualized and trimmed to remove the primers with Chromas Lite 2.6.5 (Technelysium Pty Ltd). The forward and reverse strands for each sequence were assembled by means of the BioEdit 7.2.5 software [35], with ambiguities checked and corrected manually. For the museum material of A. petricola, the repeated attempts to amplify COI and 16S loci were unsuccessful; consequently, that material was not included in further analyses.

Sequence data and phylogenetic analyses
Multiple sequence alignment of 16S gene was performed using MAFFT 7 through the MAFFT web-server (https://mafft.cbrc.jp/alignment/server/) [36] with the G-INS-I algorithm and optimized by visual inspection. The COI sequences were aligned using Clustal X 2.1 [37], and translated into amino acids to check for stop codons and frameshift mutations in ORF Finder (https://www.ncbi.nlm.nih.gov/orffinder/). For 16S, a putative secondary structure model was generated following the template proposed by Lydeard et al. [38] and contrasted with structural models for other molluscs available at the Comparative RNA Website [39]. This model was used to examine nucleotide substitutions in relation to conserved sequence motifs, alignment and secondary structure of domains IV and V of the 16S gene predicted for other molluscs. The number of haplotypes (h), number of segregating sites (S), as well as nucleotide (π) and haplotype (H d ) diversities per marker were computed by means of the DnaSP 6.12.01 program [40]. Nucleotide composition of haplotypes was analysed in BioEdit 7.2.5. Genetic distances were analysed in MEGA X software [41] using the number of differences (p).
Phylogenetic trees were constructed using maximum likelihood (ML), and Bayesian inference (BI). Acrorbis phylogenetic datasets were analysed separately as 16S (279 bp), COI (655 bp) and concatenated (16S + COI, 934 bp). The ML analysis was conducted with PhyML 3.0 [42] via the ATGC bioinformatics platform (http://www.atgc-montpellier.fr/) with the Nearest-Neighbor Interchange branch swapping algorithm. Optimal models of nucleotide substitution were selected using the SMS program [43] by means of the Akaike Information Criterion: GTR+G model for all datasets. The statistical support of the nodes was evaluated using 1,000 bootstrap replicates [44]. For the concatenated dataset, we also conducted a ML analysis partitioned by gene with RAxML 8.2.12 [45] via the CIPRES portal (https://www.phylo.org/) [46], inferring with RAxML the appropriate parameters of the GTR+G model for each partition [47]. Nodal support was assessed using RAxML rapid bootstrapping with 1,000 bootstrap replicates. The BI was performed in MrBayes 3.2.6 [48] with the parameters from the best model (GTR+G for COI and concatenated datasets, and GTR+I for 16S sequences) as identified in jModelTest 2.1.7 [49] under the corrected Akaike Information Criterion. Two runs were performed simultaneously with four Markov chains for 1 million generations. Trees were sampled every 100 generations, and posterior probabilities were obtained after discarding the first 1,001 samples of each run as burn-in. In addition, a median-joining network was constructed for each individual locus and the concatenated dataset using Network 5.0.0.3 [50], in order to visualise relationships among haplotypes.

New localities records
Living specimens of Acrorbis were recorded in the six waterfall environments surveyed in central and southern Misiones Province, five of these being new localities for the genus (Figs 2 and 3 and Table 1).

Morphological examination
The specimens from the six waterfalls surveyed exhibited the morphological characters of shell and reproductive system defined for Acrorbis petricola as described by Odhner [15], Paraense & Deslandes [19] and Paraense [20]. The following distinctive characters are emphasized: Shell: pseudodextral, small, helicoid; spire ranging from low and nearly depressed to elevated, with convex whorls increasing rapidly in diameter, up to 2 ⅞ in number in the largest specimens; suture deep (Fig 4 and Tables 2 and S1). Shell from pale amber to deep brown, apex deep brown, somewhat reddish in several specimens. Shell surface with oblique striae, some specimens sculptured with spiral lines from faint to coarse ones superimposed onto oblique striae giving them a somewhat reticulated appearance (e.g. Fig 4B and 4E). Apex laterally displaced, not the highest point on spire in several specimens. Last whorl strongly widening. Aperture wide, oblique, D-shaped to subcircular; lips simple, sharp. Most of specimens widely umbilicated with umbilicus partially covered by small basal reflection of inner lip, several specimens with umbilicus fully covered ( Fig 4A). Average measurements of ten adult shells and the conchological variability recorded per waterfall environment are presented in Tables 2 and 3, respectively. In the PCA, the first two principal components explained cumulatively 91.81% of the total variation (PC1: 84.80%; PC2: 7.01%; S2 Table). There was no clear separation among the Acrorbis populations in the morphometric space, as all populations overlapped with at least one other (Fig 5).
Reproductive system: the complete genital system is shown in Fig 6. The following distinctive characters are emphasized: ovotestis multilobed with a single row of up to 10 simple unbranched diverticula. Seminal vesicle well-developed, convoluted, at the middle portion of hermaphroditic duct (Fig 6C). Prostate composed of long, slender, unbranched, finger-shaped diverticula; 6-12 diverticula in the main series converging into spermiduct in a fan-like arrangement, 3-4 diverticula protruding into distal spermiduct from opposite direction ( Fig  6E). Spermiduct short, highly sinuous. Vas deferens long, inserting at proximal portion of penis sheath. Penis sheath distinctly narrower than praeputium with two flagella at proximal end. Flagella slender, finger-shaped, from ⅓ to about same length as penis sheath; several specimens with flagella unequal in length, asymmetrical and/or widened towards insertion into penis sheath (Fig 6F-6K). Penis unarmed, slender, acicular, longer than the penis sheath ( Fig 6D). Praeputium with some tendency to intussusception, separated from penis sheath by a diaphragm. Oviduct short, highly sinuous. Nidamental gland proximally expanded, with numerous small digitiform protuberances at the proximal end. Vagina short, tubular. Bursa copulatrix rounded to pear-shaped. Morphological variability observed between the locations examined and that described for the type locality is presented in Table 4.

Sequence data and phylogenetic analyses
Both mitochondrial genes were successfully amplified in Acrorbis specimens from most of the localities, except for the museum material from Puerto Iguazú and Nova Teutônia where amplification of the 16S and COI markers was not possible. Partial COI sequences were of 646 bp in length for all individuals, and partial 16S sequences ranged between 257 and 261 bp. No genetic variation was detected within each waterfall environment. A single unique haplotype per locality per mitochondrial marker was recognised, resulting in a total of six haplotypes per locus. Mean genetic diversity recorded for both markers was high (COI: H d = 0.862 ± 0.018, π = 0.01050 ± 0.00080; 16S: H d = 0.862 ± 0.018, π = 0.01208 ± 0.00103). Sequence alignments showed a total of 21 and 14 variable positions for the COI and 16S genes, respectively (S3 and S4 Tables). Nucleotide composition of both genes is given in S5 Table. For COI haplotypes stop codons were absent, and an ORF = +2 was identified. The secondary structure of the 16S gene  Table 1.
https://doi.org/10.1371/journal.pone.0220027.g003 was conserved among the different localities, with polymorphic sites being recorded mostly in unpaired regions or involving alternative basepairing in stems without affecting structural motifs (Fig 7). Genetic distances among COI and 16S haplotypes are summarised in Table 5.
Phylogenetic analyses recovered all Acrorbis petricola sequences as a well-supported monophyletic group (Figs 8 and 9). In all tree topologies, six phylogenetic lineages for the species were recognized, each one corresponding to a waterfall environment. However, the relationships among these lineages differed between the markers, and could not be resolved through the analysis of the individual regions nor from the concatenated dataset (Figs 8 and 9). Similarly, the organization of haplotype networks differed depending on the marker examined (Fig 10).

Discussion
In this study, we explored the morphology and genetic background of one the two historical populations of A. petricola in the Misiones Province, Argentina. In addition, we report for the first time five new populations of the species in southern Upper-Paraná Atlantic Forest, exclusively found on waterfalls from the Misiones Province, which were also morphologically and molecularly characterized. The new records increase the known distribution of Acrorbis to both the west and south of the Upper-Paraná Atlantic Forest, with the westernmost record located in Salto Krysiuk and the southernmost record located in Salto Paca. It is worth to mention that the occurrence of the genus Acrorbis was reported in highly oxygenated freshwater habitats near the Yacyretá-Apipé rapids in the High Paraná River at the border area between Argentina and Paraguay in the 1980s [10,51]. Such an occurrence, at about 150 km to the west from Salto Krysiuk, would imply a wider distribution of the genus westward. Nonetheless, the shells collected by Rumi [51] supporting that record are unavailable and probably lost. In addition, the Yacyretá-Apipé rapids have disappeared because of damming and the filling up of  the Yacyretá Reservoir in the early 1990s [10,12], thus hampering a reliable characterization of that record that remains enigmatic to us. Morphologically, the specimens of the six populations from the Misiones Province showed intra-and interpopulation conchological variability, mostly in relation to spire and spiral  sculpture, which falls within the variation documented for A. petricola in its type locality [19,20]. For the historical population from Salto Encantado, we found specimens that fit the conchological description of Hylton Scott [21], who refers specimens without spiral sculpture and with an umbilicus partially covered by a brief reflection of the inner lip. However, for that location we also observed specimens with a marked spiral sculpture and a fully covered  Acrorbis snails from the Upper-Paraná Atlantic Forest umbilicus, indicating a greater conchological variability than previously reported. The shells of specimens from Salto Capioví, Salto Chávez, Salto Teodoro Cuenca and Salto Paca were similar to those recorded in Salto Encantado; in contrast, Salto Krysiuk exhibited the most extreme shell configurations in which depressed shells with low spires were observed, resembling those described for the Iguazú National Park [24]. Nonetheless, PCA exhibited no clear geographic separation among the populations in the morphometric space. The general morphology of the reproductive system of the specimens from the six records examined here is in good agreement with previous studies of Paraense & Deslandes [19] and Paraense [20] for A. petricola. The penis sheath, flagella, prostatic diverticula and bursa copulatrix were identified to be the structures that exhibited the greatest intra-and interpopulation variability. Paraense [20] has shown that the length proportions of organs with muscular tissue such as the penis sheath and praeputium are highly variable in Planorbidae, since beyond phenotypic variation some variability exists as a result of distension of each organ at the time of fixation. Consequently, morphological differentiation based on these structures should be made with caution. In all the dissected material we have found a pair of slender and finger-like flagella showing a great variability in length within and among localities. Although a single flagellum was referred for A. petricola in the original description [15], based on material from the same locality Hubendick [18] and Paraense & Deslandes [19] confirmed the existence of two flagella in the male genitalia of the individuals from Nova Teutônia, as recorded here for the Argentinian populations. In other Planorbidae species, the length of flagella is generally useful to discriminate between species (e.g. Drepanotrema Fischer & Crosse, 1880), although   Table 5. p-distances between Acrorbis petricola haplotypes based on the COI (below the diagonal) and 16S (above the diagonal) genes. differences found here can be ascribed to intraspecific variation, as the morphology and proportion of flagella are within the intrapopulation variability previously described for specimens from Nova Teutônia [18][19][20]. In relation to prostatic diverticula, although it was possible to determine the number of diverticula for the six localities examined, it is likely that such a number will vary as more specimens from each locality be examined; thus, the count presented in this study for each waterfall should be regarded as purely indicative and not definitive. In addition, we documented differences in the shape (from spherical to pyriform) and size (from small to large) of the bursa copulatrix within and among localities. This requires further study to evaluate whether there is a relationship between the shape and size of this organ with the reproductive status of the individuals. Despite this, it is worth noting here that the largest bursa copulatrices were observed in specimens from Salto Paca, which presented the largest shells and also exhibited the longest flagella among the material examined. Although most of the A. petricola populations may be indistinguishable based on morphological features alone, our study revealed the existence of interpopulation genetic diversity within the species' limited distribution range, with absence of intrapopulation genetic variation. The six examined locations were genetically distinct with the presence of a single and unique haplotype per locality, which allows each of them to be recognized as a geographic group. Pairwise genetic distances among the six locations showed similar divergence levels for both markers, ranging from 0.6% to 2.0% for COI, and from 0.4% to 2.3% for 16S. Upon considering the historical population of Salto Encantado as a reference for comparison, Salto Capioví was shown to be the location with the smallest genetic distance for the COI gene, but was the one with the greatest divergence for the 16S gene. Conversely, Salto Chávez exhibited the greatest distance for COI and the smallest divergence for 16S, thus suggesting that both markers are probably evolving at different rates among some of the populations, a finding in need of further investigation. Beyond this, the highest divergences found are within the range of intraspecific variability usually reported in Planorbidae (e.g. [34,52,53]). Based on these values, and supported by morphological evidence, all examined specimens are considered here as belonging to A. petricola. Nevertheless, it is important to note that DNA sequences are not yet available for the species from its typical locality in Nova Teutônia. Although we attempted to examine the genetic background of that location based on historical material, we had no success in recover DNA probably as a consequence of tissue samples age [54]. Thus, further studies focusing on the genetic divergences and evolutionary affinities of Brazilian specimens are required in order to confirm that all populations belong to the single species A. petricola. On the contrary, the junior synonym A. odhneri would need to be revalidated for the Argentinian populations.
For both mitochondrial markers, genetic diversity indices showed high values of mean haplotype and nucleotide diversities. When interpreted in light of the spatial distribution of genetic variation, these findings suggest that despite the short geographical distance between some localities, populations have been genetically isolated for a large amount of time with limited gene flow, a condition attributable to stable, geographically subdivided populations having a long evolutionary history [55,56]. By considering that all A. petricola populations in the Misiones Province were found in waterfall environments, such a pattern seems fully consistent with vicariance events. These could be related to climatic and tectonic factors associated with the geological history and geomorphological evolution of the fluvial systems of Misiones during the Cenozoic, where the hydrographic network was completely different from the current one [57,58]. Similar evidence of microevolutionary differentiation at small spatial scale has already been documented for other South American endemic freshwater snails (e.g. [59,60]). Further research based on a greater number of individuals and loci per locality is required to gain insights into the evolutionary history, and the spatio-temporal framework of the diversification of Acrorbis in the Upper-Paraná Atlantic Forest.
In summary, this study not only broadens the known distribution of the species, but also illustrates a hidden diversity of Acrorbis that still remains to be discovered in waterfall environments of the Atlantic Forest, which include genetically differentiated populations. The genealogical relationships among haplotypes depict a complex evolutionary history that needs to be unveiled and taken into account for future development of conservation strategies in this endemic genus. Nonetheless, the identification of six geographically isolated evolutionary lineages in the Misiones Province appear to warrant their future recognition as evolutionarily significant units (ESUs) for conservation prioritisation [61][62][63]. The Misiones Province has one of the largest systems of protected natural areas in Argentina, comprising different categories ranging from the strictest protected areas to the sustainable-use ones [64]. The information provided here is expected to contribute to the development of future conservation strategies aimed at preserving the gene pool of Acrorbis based on the territorial network of protected areas of Misiones. To date, only Salto Encantado and Salto Capioví populations are included within protected areas, in a provincial and municipal park, respectively. In this context, the finding of Acrorbis populations in the waterfalls known as Salto Chávez, Salto Teodoro Cuenca, Salto Krysiuk and Salto Paca, widely used as ecotourism resources, could serve as the basis for the environmental protection of such areas, as they harbour a unique genetic diversity of A. petricola, which is documented here for the first time.
Supporting information S1