Field Sites and Sampling

Cyprinella lutrensis, C. venusta, and P. vigilax were collected from three streams in eastern Texas that support populations of F. askewi: Sabine River near Highway 14, Smith County; Neches River near Highway 294, Anderson County; and Lake Fork Creek near Highway 80, Wood County (Fig. 1). We collected fishes from the Sabine and Neches rivers on eight different days between May and October of 2014 (Table 1) based on times of maximum glochidial infestation in these rivers reported by Marshall (2014). We collected fishes from Lake Fork Creek on a single date (August 4, 2014) to increase sample sizes of target fishes when high flow prevented sampling on the Sabine and Neches rivers. Fishes were collected from each site over a 150-m reach near mussel beds using a 7.5-m bag seine. Electrofishing was not used to avoid mortality or stress to fish that may cause release of encysted glochidia. We attempted to collect fishes of varying sizes (3–7 cm length) for each species. Water temperature, pH, and conductivity were measured using a YSI multi-probe meter (YSI Incorporated, Yellow Springs, OH, USA) for each sampling event.

Figure 1.

Texas collection locations for Red Shiners (Cyprinella lutrensis), Blacktail Shiners (Cyprinella venusta), and Bullhead Minnows (Pimephales vigilax).

Laboratory Housing of Fishes

Fishes collected from the field were brought back to the Department of Biology Aquatic Ecology Laboratory at the University of Texas at Tyler. Fish were then placed in 3-L tanks in an AHAB unit (Pentair Aquatic Eco-Systems, Inc., Apopka, FL, USA) in groups of up to seven smaller individuals or two to three larger individuals and separated by species, collection date, and collection site. Multiple individuals of the same species and origin were housed together to decrease stress in these shoaling species and because our system was limited to 20 tanks. We monitored water temperature, pH, and conductivity every other day with a multi-probe YSI meter, and we adjusted these conditions to be similar to river sites where the fish were collected.

Table 1.

Infestation of Red Shiners (Cyprinella lutrensis), Blacktail Shiners (Cyprinella venusta), and Bullhead Minnows (Pimephales vigilax) by mussel glochidia at three eastern Texas collection sites (Sabine = SBN, Neches = NCHS, Lake Fork Creek = LKFRC). (n) refers to the number of fishes examined on each date. Number of glochidia is divided into those that excysted as juveniles (Juv.), those that were sloughed prior to metamorphosis into juveniles (Gloch.), and those that remained encysted at the end of the experiment (Encysted).

Juvenile Collection

Juvenile collection devices consisting of 3.5-cm-long polyvinyl chloride pipe segments with 118-µm mesh netting on one end were placed on each tank; these permitted water exiting the tanks to flow through while retaining glochidia and juvenile mussels (Barnhart 2006). This mesh size is smaller than the minimum size of F. askewi glochidia (128 µm; Howells et al. 1996). We removed and inspected juvenile collectors every other day for the first 2 wk of the trial and sporadically until termination of the trial. We examined material retained on the netting under an Olympus SZ dissection microscope (Olympus Corporation of the Americas, Center Valley, PA, USA) and counted all glochidia and juveniles. Juvenile mussels were distinguished from glochidia based on the presence of internal tissue development and movement, such as protrusion of the foot from the shell (Howells et al. 1996). We calculated overall infestation intensity ([number of juveniles + number of sloughed or encysted glochidia]/number of fish), juvenile production (number of juveniles produced/number of fish), and metamorphosis success (number of juveniles/[number of juveniles + number of sloughed glochidia]) for each potential host species across all trials. We collected subsamples of at least 10 juveniles for genetic identification from each tank on each day that tanks were inspected. Each individual was placed in a separate 1.5-mL centrifuge tube with 95% ethanol and stored at -20°C.

Duration of each trial ranged from 3 to 6 wk. If juvenile production ceased or if fishes did not produce any glochidia or juveniles for 3 wk, we terminated the trial and euthanized all fishes in that tank. This 3-wk termination criterion was based on past observations of the authors, as well as observations that unionid glochidia tend to excyst between a few days and several weeks following encystment (e.g., Haag and Warren 1997), with this process expedited in warmer regions (Watters and O'Dee 2000). Euthanized fish were then examined for glochidial encystment on their gills and fins.

DNA Sequencing and Identification

Genomic DNA was extracted from individual juveniles using a Chelex double-stranded DNA extraction protocol (Casquet et al. 2011). We modified the protocol of Casquet et al. (2011) by adding 50 µL of a 1:15 solution of proteinase K and 10% Chelex 100 resin instead of the recommended 150 µL; this was done to avoid diluting the small amounts of genomic DNA extracted from juvenile mussel tissue. Extracted DNA was stored at -20°C until use in PCRs. The primers Leu-uurF and LoGlyR were used to amplify mitochondrial (mtDNA) NADH dehydrogenase (ND1) gene (Serb et al. 2003). PCR reactions used for amplification of the ND1 gene consisted of 20 µL: 6.7 µL purified H₂O, 0.1 µL TopTaq, 2.0 µL PCR buffer (Qiagen Sciences Inc, Germantown, MD, USA), 0.4 µL dNTPs, 2.0 µL 10X Coral Load (Qiagen), 4.0 µL Q-Solution, 1.0 µL of each primer, 0.4 µL bovine serum albumin, and 2.4 µL of DNA (∼150 ng/µL). An extra 10% of the PCR reaction was created to provide a negative control with each PCR. An Eppendorf Mastercycler gradient thermal cycler (Eppendorf, Hamburg, Germany) with a heated lid was used to amplify the reactions. The reaction settings for amplification of double-stranded DNA were as follows: 94°C for 5 min; 30 cycles of 94°C for 45 s, 54°C for 60 s, and 72°C for 60 s; followed by a final extension of 72°C for 5 min. Gel electrophoresis was used to test the quality of amplification. The successfully amplified PCR products were purified using and E.Z.N.A. cycle pure kit (Omega Bio-tek, Norcross, GA, USA) following the protocol with an additional 30 µL of purified water for resuspension. Purified DNA was concentrated at 17–20 ng/µL with a 260/280 ratio around 1.8 to 2.0 as recommended by Eurofins MWG Operon where reactions were shipped to for sequencing using BigDye Terminator v 3.1 Cycle Sequencing kits (Applied Biosystems, Foster City, CA, USA). Sequences were edited with the Sequencher 5.2.4 program (Gene Codes Corporation, Ann Arbor, MI, USA) and then initially compared with unionid sequences available on the National Center for Biotechnology Information database ( http://www.ncbi.nlm.nih.gov). The edited sequences were also cross-referenced with an adult molecular key that provides sequences for all the 37 unionid species that occur in eastern Texas (Marshall 2014). The tissue samples from mussels used to create the molecular key included adult mussels collected from the same sampling sites we used on the Sabine River and Neches River. ClustalX2.0.11 (Conway Institute UCD, Dublin, Ireland) was used to generate an alignment file of the juvenile sequences with the adult sequences of the molecular key. The alignment file from ClustalX2.0.11 was then uploaded into Mesquite (version 2.75, Mesquite Project Team,  http://mesquiteproject.org) to provide ocular observation of the alignment with the sequences of the molecular key.

Infestation on Wild-caught Fish

A total of 114 C. lutrensis, 87 C. venusta, and 46 P. vigilax were collected during the study (Table 1). Pimephales vigilax had the highest glochidial infestation intensity (average = 14.3/ fish), but only two juveniles were produced in a single trial from the Sabine River (overall juvenile production = 0.04 juveniles/fish; metamorphosis success = 0.3%). No glochidia were found encysted on the gills of deceased P. vigilax at the end of our trials. Cyprinella lutrensis had a lower glochidial infestation intensity (7.9/fish), but it had the highest rate of juvenile production (2.1/fish) and moderate metamorphosis success (29.4%). In addition, 73 glochidia were encysted on the gills of deceased fish at the end of our trials. Cyprinella venusta had the lowest infestation intensity (2.5/fish) and the second highest juvenile production (0.8/fish), but it had the highest metamorphosis success (46.3%). Sixty-seven glochidia were found encysted on deceased fish at the end of our trials. For all three fish species, glochidial infestation was observed from late May to early June until July or early August, and no fishes were infested in October.

Molecular Identification of Glochidia and Juvenile Mussels

DNA was extracted from a total of 127 juveniles, which consisted of 86 juveniles from C. lutrensis, 39 juveniles from C. venusta, and the two juveniles from P. vigilax. Of these, DNA from eight juveniles from C. lutrensis and seven juveniles from C. venusta was succesfully amplified, sequenced, and identified. These juveniles included at least one individual from each fish species from all three sampling sites. We were unsuccessful in amplyfing and sequencing DNA from juveniles collected from P. vigilax.

Fourteen of our 15 sequences represented a single haplotype (GenBank accession number KY442832) that was 100% identical to both a National Center for Biotechnology Information sequence from F. askewi and one generated by Marshall (2014) for Triangle Pigtoes (Fusconaia lananensis) and F. askewi. Only one sequence represented a haplotype (GenBank accession number KY442833) not previously detected in eastern Texas, but this sequence was consistent with F. lananensis and F. askewi, and it differed from the other haplotype we detected by only a single base pair difference and was over 99% identical to that haplotype.