This is the fifth in a recent series of papers on the poorly known western North American pebblesnail genus Fluminicola (Caenogastropoda, Lithoglyphidae). Herein we clarify the taxonomic status of the currently undescribed pebblesnail fauna in the upper Klamath River drainage (UKL) based on morphologic evidence, and mitochondrial DNA sequence data from 58 UKL collection localities. We describe one new species (F. klamathensis) from eight UKL localities which is differentiated by mtDNA sequences and unique penial morphology, and document range extensions to the UKL for three species from closely proximal drainages (F. fresti, F. modoci, F. multifarius). Fluminicola fresti was found at a single locality along the western edge of upper Klamath Lake. Fluminicola modoci and F. multifarius are widely distributed in the UKL; both species exhibit marked morphologic variation yet are relatively little differentiated genetically in this basin.

Caenogastropoda , freshwater, mitochondrial DNA, morphology, Pacific Northwest, systematics, Truncatelloidea

This is the fifth in a recent series of papers on the freshwater pebblesnails (Lithoglyphidae: Fluminicola) of the Pacific Northwest, USA. The previous contributions in this series, which treated the faunas in the Rogue–Umpqua (Hershler et al. 2017), upper Sacramento River (Hershler et al. 2007), and Snake River watersheds (Hershler and Liu 2012; Liu et al. 2013), increased the number of Fluminicola species from 9 to 27 and also documented large range extensions for F. coloradoense Morrison, 1940 and F. multifarius Hershler, Liu, Frest & Johannes, 2007.

The pebblesnails in the upper Klamath River drainage (UKL), California–Oregon, have been little studied historically and are currently unassigned to species (Hershler and Frest 1996). In various contract reports documenting their extensive field surveys of UKL freshwater mollusks, Frest and Johannes (1998, 2000, 2004, 2005, and other references cited therein) recognized 24 purportedly undescribed, narrowly ranging pebblesnail species from this watershed (based largely on shells and body pigmentation) and gave them provisional scientific (e.g., “Fluminicola n. sp. 1”) and colloquial names. Although the UKL pebblesnails have subsequently become a focus of conservation attention–e.g, four of the putative novelties recognized by Frest and Johannes were incorporated into the Northwest Forest Plan as “survey and management species” (USDA and USDI 1994; also see Frest and Johannes 1999) and three of these were petitioned (unsuccessfully) for addition to the Federal List of Threatened and Endangered Species (Curry et al. 2008; USFWS 2012)–there have been no recently published studies of this fauna aside from a molecular phylogenetic analysis which delineated a close relationship between specimens from the Link River (the outlet of Upper Klamath lake) and F. modoci Hannibal, 1912 from the Goose Lake basin (Hershler et al. 2007). Herein we utilize both molecular (mitochondrial DNA sequences) and morphological data to delimit the UKL pebblesnail species. The former has proved very useful in previous taxonomic studies of pebblesnails, enabling delineation of both morphologically cryptic, and morphologically variable species (Hershler et al. 2007, 2017; Liu et al. 2013).

For this project we sequenced specimens from 58 UKL localities that were sampled in August 2012 and May and September 2013. Collections were made at three localities (Harriman Springs, Wood River south spring source, spring brook below Schoolhouse Meadow) on more than one occasion in an effort to increase sample sizes. Specimens were collected by hand or with a small sieve and preserved in 90% (non-denatured) ethanol in the field. Portions of several samples were relaxed with menthol crystals, fixed in dilute formalin, and preserved in 70% ethanol for anatomical study. Vouchers were deposited in the Smithsonian Institution’s National Museum of Natural History (USNM) collection.

Some of the collections contained multiple shell morphotypes which were sorted and analyzed separately, yielding a total of 80 samples (UKL12–UKL91). Cytochrome c oxidase subunit I (COI) and cytochrome B (cytB) sequences were obtained from 283 and 259 UKL specimens, respectively. Genomic DNA was extracted from entire snails using a CTAB protocol (Bucklin 1992); each specimen was analyzed for mtDNA individually. LCO1490 and HCO2198 (Folmer et al. 1994) were used to amplify a 709 base pair (bp) fragment of COI; cytB427F (5’TGA GGK GCN ACT GTT ATT ACT AA3’) and cytB1049R (5’GTG AAA ACT TGS CCR ATT TGC TC3’) were used to amplify a 644 bp fragment of the cytB gene. The cytB427F and cytB1049R primers were designed based on conserved regions of cytB in an alignment using previously published sequences from Oncomelania hupensis (Gredler) (NC13073) and Potamopyrgus antipodarum (Gray) (GQ996433). Amplification conditions and sequencing of amplified polymerase chain reaction product methods were those of Liu et al. (2013). Sequences were determined for both strands and then edited and aligned using Sequencher™ version 5.4.1 (Gene Codes Corporation, Ann Arbor, MI). In order to generate easily readable topologies, one example of each unique UKL haplotype was used in the phylogenetic analyses, which were performed separately for the COI and cytB datasets. The analyses of the COI dataset also included the previously published UKL haplotypes (from a single collection locality), and sequences from 14 regional Fluminicola species and representatives of two other North American lithogyphid genera (Somatogyrus, Taylorconcha). Trees were rooted with Pristinicola hemphilli (Pilsbry) (Hydrobiidae). The cytB dataset also included sequences from 13 Fluminicola species (a cytB sequence is not available for F. gustafsoni Hershler & Liu). Given that cytB sequences are not available for other North American lithoglyphid genera, basally positioned F. virens was used to root the trees based on this dataset (Hershler et al. 2007; Hershler and Liu 2012). Sample codes, locality and voucher details, and GenBank accession numbers for the sequences used in the molecular phylogenetic analyses are in Suppl. material 1.

Genetic distances were calculated using MEGA7 (Kumar et al. 2016), with standard errors estimated by 1,000 bootstrap replications with pairwise deletion of missing data. MrModeltest v. 2.3 (Nylander 2004) selected the GTR + I + G model parameters as the best fit for both the COI and cytB datasets (using the Akaike Information Criterion). Phylogenetic analyses were performed using maximum parsimony (MP), maximum likelihood (ML), and Bayesian inference (BI) methods; trees were also generated using a distance method. The MP and ML analyses were performed using PAUP* v. 4.0b10 (Swofford 2003) and the BI analyses were conducted using MrBayes v. 3.2.6 (Ronquist et al. 2012). The MP analyses were conducted with equal weighting, using the heuristic search option with tree bisection reconnection branch-swapping and 100 random additions. Nodal support was evaluated by 10,000 bootstrap replicates. The ML analyses were performed using the GTR + I + G model. The optimized parameter values for COI were base frequencies of A = 0.3089, T = 0.3856, C = 0.1684, G = 0.1371; shape of gamma distribution = 1.1801; proportion of invariant sites = 0.5691. The optimized parameter values for cytB were base frequencies of A = 0.3146, T = 0.3671, C = 0.1917, G = 0.1268; shape of gamma distribution = 0.7706; proportion of invariant sites = 0.3818. A GTR distance-based neighbor-joining (NJ) tree was used as the initial topology for branch-swapping. Nodal support was evaluated by 1,000 bootstrap pseudo-replicates. For the BI analyses Metropolis-coupled Markov chain Monte Carlo simulations were run with four chains (using the model selected by MrModeltest) for 5,000,000 generations. Markov chains were sampled at intervals of 100 generations to obtain 50,000 sample points. We used the default settings for the priors on topologies and the GTR + I + G model parameters. At the end of the analyses, the average standard deviations of split frequencies were 0.005763 (COI dataset) and 0.004997 (cytB dataset) and the potential scale reduction factor (PSRF) was 1, indicating that the runs had reached convergence. The sampled trees with branch lengths were used to generate 50% majority rule consensus trees, with the first 25% of the samples removed to ensure that the chain sampled a stationary portion. For the distance analyses, HKY distances were used to generate neighbor-joining (NJ) trees (Saitou and Nei 1987).

We also studied pertinent specimens in the USNM collection, including UKL material collected by Frest and Johannes that was acquired during the planning stage of this project. The total number of shell whorls (WH) was counted for each specimen; and the height and width of the entire shell (SH, SW), body whorl (HBW, WBW), and aperture (AH, AW) were measured from camera lucida outline drawings (Hershler 1989). Photographs of alcohol preserved specimens (that had been anaesthetized with menthol crystals prior to fixation) were taken using a Coolpix 990 mounted on an Olympus SZX12 dissecting microscope. Other methods of morphological study were routine (Hershler et al. 2007). Shell descriptive statistics were generated using Systat for Windows 11.01 (SSI 2004).

Ninety-four (94) COI and 96 cytB haplotypes were detected in the analyzed UKL specimens (Suppl. materials 2, 3, respectively). The molecular phylogenetic and distance analyses of the two datasets generated closely similar trees in which the UKL haplotypes were resolved into four clades. Three of the clades were strongly supported (>95% bootstrap or posterior probability) in all analyses, while the fourth (clade A) was strongly supported only in the cytBBI analysis. The BI topology based on the COI sequences is shown in Figure 1 and the geographic distributions of the four clades are shown in Figure 2. Pairs of the lineages were sympatric at eight localities and three of the lineages co-occurred at one site (Wood River south headsprings).

Figure 1. 

Bayesian tree based on the COI dataset. The four clades (A–D) containing UKL haplotypes are color coded as in Figure 2. Posterior probabilities for nodes are shown when ≥ 95%. Specimen codes are from Suppl. materials 1–3.

Figure 2. 

Map of southwest Oregon and northwest California showing the distribution of mtDNA clades A–D with color codes matching those in Figure 1.

Clade C is composed of pebblesnails from eight UKL localities. This lineage is well differentiated genetically from currently recognized Fluminicola species (7.6–17.2% for COI, 6.4–20.3% for cytB) and is further distinguished by unique penial morphology; we describe it as a new species below. The phylogenetic relationships of this new species were not well resolved.

Clades A, B and D contain both UKL pebblesnails and currently recognized species from other regional drainages (Figs 1, 2). Clade A contains F. multifarius and a large number of UKL populations varying in shell size and shape (Fig. 3). Although most of the UKL pebblesnails in this clade have subglobose to narrowly conical shells conforming to F. multifarius, populations in the Jenny Creek drainage often contain additional forms that fall outside of the range of variation previously reported for this species (e.g., Fig. 3F, shell neritiform; Fig. 3H, shell having a distinct swelling on the inner apertural lip). Clade B contains F. modoci and UKL pebblesnails that well conform to this species aside from having a much larger range in shell size among populations (Fig. 4A–G). Clade D is composed of F. fresti Hershler, Liu & Hubbart, 2017 and pebblesnails from Harriman Springs that closely resemble this species in their shells (Fig. 4H).

Figure 3. 

Shells of UKLF. multifarius A USNM 1144951 B USNM 1144587 C USNM 1190091 D, E USNM 1020970 F, G USNM 1145066 H USNM 1144903 I USNM 1190128 J USNM 1190104 K USNM 1144463. Scale bar: 1.0 mm.

Figure 4. 

Shells of UKLF. modoci (A–G) and F. fresti (H) A USNM 1144454 B USNM 1144520 C USNM 1144942 D USNM 1144407 E USNM 1144966 F USNM 1190096 G USNM 1144565 H USNM 1144900. Scale bar: 1.0 mm.

The sequence divergence between the UKL pebblesnails and the extra-limital species in clades A, B, and D is <4% for both genes (Table 1), which is relatively small compared to that among currently recognized Fluminicola species, which ranges from 1.7–18.7% for COI and 1.4–25.7% for cytB, with 93% (301/325) and 96% (288/300) of the pairwise comparisons >4%, respectively. The morphologically diverse UKL pebblesnails in both clades A and B are also relatively little differentiated genetically (clade A, mean sequence divergence 2.2% for COI and 2.9% for cytB; clade B, 0.4% for both genes). (Note that clade B was shallowly structured in all the trees while clade A was somewhat structured only in the BI topologies.) We also found that the pronounced phenotypic variation within populations (belonging to clade A) in the Jenny Creek drainage is not accompanied by substantial genetic divergence. For example, specimens having neritiform, simple conical, and conical shells with a swelling on the inner apertural lip that were collected in sympatry at the Fall Creek locality (UKL 50–52) differ by only 0.6–0.7% for COI, and 0%–0.5% for cytB. Based on this body of evidence we recognize clades A, B and D as single species corresponding to F. multifarius, F. modoci, and F. fresti.

Our findings, based on both morphologic and genetic (mtDNA sequences) evidence, have shown that contrary to previous assertions in the grey literature (Frest and Johannes 1998, 2000, 2004, 2005), the UKL does not have a large, highly endemic fauna of undescribed pebblesnails, but instead contains only four species, one of which is new and three of which were previously described from other regional drainages. Although we only surveyed 58 of the >200 UKL pebblesnail localities reported by Frest and Johannes, we sampled at least one locality for each of the putatively new species that they recognized, and we sampled numerous localities in the Jenny Creek watershed where much of the phenotypic diversity of UKL pebblesnails is concentrated. Thus, we are confident that we have well delineated the taxonomic diversity of UKL pebblesnails. Additional studies will be needed to further delineate the distributions of the four species in the UKL and to determine whether these pebblesnails range into the lower reach of the large Klamath River watershed.

Our study also provides evidence of striking morphologic variation in F. modoci and F. multifarius similar to what has been observed in various marine caenogastropod lineages (e.g., Littorina; Reid 1996). Further investigations utilizing rapidly evolving nuclear markers such as microsatellites should provide additional insight into the apparent decoupling of morphologic and genetic variation in these two species and, together with ecological studies, help tease apart the underlying mechanisms for the sympatric occurrence of multiple F. multifarius morphotypes at various localities in the Jenny Creek area.

We thank Edward Johannes for contributing specimens and locality data that helped enable this study. Missy Anderson, Corbin Bradford, Jim Chambers, Kelli Van Norman, Darci Rivers-Pankratz, Rob Robinson, Jen Smith, Roger Smith, Terry Smith, and Kelli Van Norman provided field assistance. We also thank Missy Anderson for logistical assistance relating to fieldwork, and Kelli Van Norman for her continued support and encouragement during the long course of this project. Freya Goetz prepared the anatomical drawings and figures, Karolyn Darrow assisted with shell photography, and Yolanda Villacampa prepared the scanning electron micrographs and measured shells. This project was supported by an award from the Bureau of Land Management, Oregon State Office (L11AC20325, modification no. 1) and funding from the Center of Biological Diversity, and the Xerces Society, Inc. We also thank Art Bogan, Chuck Lydeard, and editor Thierry Backeljau for their comments on the submitted manuscript.