Acute ammonia toxicity of Texas unionid mussels Open Access

Abstract

Introduction

Freshwater mussels of the family Unionidae are widely distributed across the world. Some of the greatest diversity occurs in the United States, which contains more than 290 species from 54 genera (Haag, 2012; Lopes-Lima et al., 2017; Williams et al., 2017). This family is also one of the most imperiled taxa, with almost 70% of species on the North American continent classified as extinct, endangered, threatened, or near-threatened (Lopes-Lima et al., 2017; Williams et al., 1993). Texas has 52 recognized species, including nine that are endemic to the state (Dascher et al., 2018; Freshwater Mollusks Conservation Society (FMCS), 2023; Moulpied et al., 2022). The taxa are in decline throughout the state, with 12 species that have been listed or proposed for listing under the U.S. Fish and Wildlife Service (USFWS) Endangered Species Act (USFWS, 2018, 2022, 2023a, 2023b, 2024) and another five that have been designated as state threatened (Texas Parks and Wildlife Department [TPWD], 2020). Water quality degradation from environmental pollutants, including ammonia, has been implicated as one of the largest threats to mussel populations globally (Haag, 2012; Lopes-Lima et al., 2017; Strayer et al., 2004; Williams et al., 1993) and to imperiled species in Texas (USFWS, 2018, 2022, 2023a, 2023b, 2024).

Ammonia is an important nutrient source in plants but can be toxic to aquatic organisms in higher concentrations (Shi et al., 2020; Yu et al., 2022). Anthropogenic sources include municipal or industrial wastewater effluent discharges and runoff from fertilized agricultural fields (U. S. Environmental Protection Agency [USEPA], 2013). In water, total ammonia nitrogen (TAN) consists of unionized ammonia (NH₃) and ionized ammonium (⁠$NH4+$⁠). Certain environmental conditions, such as higher pH and higher temperature, increase the proportion of the more toxic unionized ammonia (Emerson et al., 1975). Ammonia nitrogen can damage gills, reduce oxygen transport, and cause nervous system dysfunction (Xu et al., 2021).

The USEPA has developed aquatic life ambient water quality criterion for ammonia in freshwater to guide protections for wildlife (USEPA, 2013). Published studies on TAN lethal toxicity from 69 different aquatic species in the United States were evaluated by the USEPA to identify protective environmental criterion that dynamically scale with pH and temperature. Forty-five of the 50 U.S. states including Texas have not voluntarily adopted USEPA recommended ammonia criterion into state specific water quality standards 5 years after publication of the report (Virginia Department of Environmental Quality [VDEQ], 2018).

Past acute ammonia toxicity studies have largely focused on species native to temperate environments (Kwok et al., 2007; USEPA, 2013). Previous studies have found significant differences in ammonia sensitivity between aquatic species sourced from tropical and temperate climates, but confounding effects from thermal stress and water chemistry have been implicated over physiological differences between the species (Henry et al., 2017; Mooney et al., 2019; Wang & Leung, 2015). Some regional declines of Texas freshwater mussels, where the majority of the state’s rivers and freshwater mussel populations occur in the subtropical region, have been attributed to ammonia pollution (Horne & McIntosh, 1979; Nobles & Zhang, 2015). The life history of freshwater mussels includes three life stages with variable susceptibility to environmental contamination effects. Early glochidia and juvenile life stages have been shown to be particularly vulnerable to the effects of ammonia nitrogen exposure (Augspurger et al., 2003; Salerno et al., 2020; Wang et al., 2007a; 2007b). Freshwater mussels comprise the twelve most sensitive species in the most recently published revision of the USEPA aquatic life ambient water quality criterion for ammonia in freshwater and were a primary driver of water quality guidance for this pollutant (USEPA, 2013). Acute ammonia toxicity tolerances for the majority of unionid mussel species have not been documented in published literature.

The goal of this study was to experimentally test the ammonia tolerances of glochidia and juvenile life stages from federally endangered (Cyclonias necki, Fusconaia mitchelli, and Lampsilis bergmanni), state threatened (Fusconaia askewi, Lampsilis satura), and common species (Lampsilis hydiana, Lampsilis teres, and Potamilus purpuratus) collected throughout the state of Texas. Differences in sensitivity for mussels collected in the humid subtropical rivers of Texas were evaluated against previously published toxicity data from unionids native to higher latitudes of the United States. The species-specific data from this study will inform future conservation management for tested species and assess the suitability of current federal ammonia criterion to freshwater mussels of Texas.

Materials and methods

Collection of brooding mussels

East Texas species

Three species of gravid mussels were collected from the Sabine River near Hawkins, Texas (Fusconaia askewi, Lampsilis satura, and Potamilus purpuratus) in spring and fall of 2022. Broodstock were transferred to the USFWS San Marcos Aquatic Resources Center in San Marcos, Texas. Mussels collected in spring were transported in a cooler with Sabine River water, and mussels collected in the fall were wrapped in damp paper towels in a cooler prior to transportation to reduce premature expelling of glochidia. Brooding mussels were held in flow-through well water with high total hardness (372 mg calcium carbonate/L) at 20 °C and fed a diet of cultured and concentrated phytoplankton (Shellfish and Nanno; Reed Mariculture). Short-term brooders (F. askewi) were kept in isolated tanks with 100-μ mesh to prevent loss of expelled conglutinates and long-term brooders (L. satura and P. purpuratus) were held in a communal tank until glochidia could be extracted.

Central Texas species

Four species of gravid mussels were collected from the Guadalupe River in Central Texas (Cyclonaias necki, Fusconaia mitchelli, Lampsilis bergmanni, and Lampsilis hydiana) in spring and fall of 2022. Lampsilis bergmanni and C. necki were collected from the Upper Guadalupe near Kerrville, TX, L. hydiana were collected from the Middle Guadalupe near Seguin, TX, and F. mitchelli were collected from the Lower Guadalupe near Gonzales, TX. Mussels were transported in individual containers with river water to collect any conglutinates aborted during transfer. Mussels were held in a recirculating water system with water from the Guadalupe River with regular water changes. Short-term brooders (C. necki and F. mitchelli) were kept in separate tanks in a recirculating system with 100-μ mesh over the effluent drains to prevent loss of expelled glochidia, and long-term brooders (L. bergmanni and L. hydiana) were kept in a communal tank until glochidia were harvested. Glochidia collected in effluent drain meshes from short-term brooders were retained for testing.

Lampsilis teres were collected from the Yegua Creek tributary of the Brazos River downstream of Somerville Lake east of Somerville, TX. These mussels were transported in ambient water from the Brazos River and held in source water in a communal tank with aeration and regular water changes until glochidia were extracted and used in the trial.

Glochidia testing

Only glochidia with greater than 90% viability were used in the trials and mixed with glochidia from other individuals of the same species. Glochidia were acclimated to dilution water (ASTM International, 2022) and study temperature of 20 °C prior to trials by providing 50% water changes over 2 hr. Approximately 500 to 1000 glochidia were transferred to each 250-ml test beaker containing 100 ml TAN treatment solution. Beakers rested in a water bath at 20 ± 2 °C inside an incubator (Power Scientific diurnal plant growth chambers, VWR incubator) with a 16:8-hr light: dark photoperiod. Nominal concentrations of ammonia for glochidia were 0, 1, 2, 4, 8, 16, and 32 mg/L TAN (Figure 1). Glochidia viability from each test beaker was measured after 6, 24, and, in some cases, 48 hr of exposure using a subsample of approximately 100 glochidia per beaker. Trials consisted of three to four replicates (only Fusconaia mitchelli) for each testing concentration and were considered valid only if > 90% of average replicated control groups survived following adjustment for initial viability as referenced in the ASTM International (2022) guidelines. Ammonia concentrations were tested and recorded from stock solutions at the beginning of each glochidia trial and from exposure beakers of pooled replicates at the endpoint of each trial (Table A1, Appendix A). The final ammonia exposure concentrations were not recorded for F. mitchelli.

Figure 1.

Flow chart of the experimental design to determine the total ammonia nitrogen (TAN) lethal concentrations for freshwater mussel glochidia and juveniles. Glochidia were extracted from brooding female mussels and either used in glochidia testing or used for host fish infestation to obtain juvenile mussels for testing. Replicated testing chamber concentrations ranged from 0–32 mg of TAN/L and were incubated at 20 ±2 °C. Survival was determined at indicated intervals of 6 (glochidia only), 24, 48, or 96 (juvenile only) hr. All procedures followed ASTM International (2022) guidelines.

Open in new tabDownload slide

Ammonia test solutions were prepared in 1,000-ml volumetric flasks by adding anhydrous ammonium chloride (NH₄Cl) reagent to reconstituted water. The 32 TAN mg/L and 16 TAN mg/L stock solutions were created by respectively adding 0.1222 g ammonium chloride, and 0.0611 g ammonium chloride to the flasks, and the 8, 4, 2, and 1 mg/L TAN stocks were created by serially diluting the 32 mg/L TAN stock solution with reconstituted moderately hard water (ASTM International, 2023). The pH of 32, 16, and 8 mg/L stock testing solutions was adjusted to a range of 7.5 to 8.0 S.U by dropwise additions of 6 N sodium hydroxide to counteract the lowered pH of testing solutions resulting from the addition of ammonium chloride. Water quality measurements were taken prior to and at the conclusion of each trial. For water quality measurements, solutions were pooled from the same test concentrations. Ammonia, hardness, and alkalinity concentrations were measured for 0, 4, and 32 mg/L TAN test solutions. Temperature, pH, dissolved oxygen, and conductivity were measured for all test solutions. Dissolved oxygen (mg/L), pH, and temperature (°C) were measured using a YSI Professional Plus probe. Conductivity (µS/cm) was measured using an EcoSense 1030A probe. Hardness (mg CaCO₃/L) alkalinity, and ammonia (TAN mg/L) were measured using a Hach DR 3900 spectrophotometer and Hach TNTplus methods (Hach Company, 2018b; 2020). Alkalinity was determined using a Hach digital titration method (Hach Company, 2018a). Quality control samples were analyzed for TAN following instrument setup. Calibration blanks were recovered at concentrations below the method detection limit of < 0.015 mg/L. Calibration verification spikes and laboratory duplicate spikes were tested at TAN concentrations of 0.015, 0.25, and 1.0 mg/L resulting in percentage of recovery values ranging from 93.3% to 120% and relative percent differences ranging from 0.06% to 1.11%.

Juvenile mussel testing

Three brooding females with viability > 90% were pooled and used to infest their corresponding host fish, except for L. bergmanni, where glochidia were pooled from two contributing females and used for infestation. Bluegill (Lepomis macrochirus) and green sunfish (Lepomis cyanellus) were used for all Lampsilis species and freshwater drum (Aplodinotus grunniens) were used for Potamilus species (Ford & Oliver, 2015). Infested fish were held in recirculating tanks where effluent was captured by mesh sieves. Well water with measured total hardness of 372 mg/L was used in the culture systems and temperatures were maintained between 20 and 25 °C. Juveniles were collected daily from the sieves during peak drop-off, sorted from particulates, and placed into 800-ml beakers with gentle aeration, where they were acclimated to the dilution water (ASTM International, 2022) and study temperature of 20 °C prior to trials by providing 50% water changes over 2 hr. Juveniles with observed valve or foot movement were entered into trials within 48 hr of being collected.

Juvenile toxicity exposure tests were conducted in 50-ml glass beakers containing 30 ml exposure solution with five juveniles per beaker. Reconstituted moderately hard water (ASTM International, 2023) was used to dilute the ammonia solution. Nominal concentrations of ammonia for juveniles were 0, 2, 4, 8, 16, and 32 mg/L TAN, where each treatment concentration had three replicates (Figure 1). The lethal concentrations resulting in 50% mortality (LC50) and 5% mortality (LC05) were determined by exposing juvenile mussels to the ammonia solution for 96 hr with survival checks at 24, 48, and 96 hr across the concentrations. Juvenile survival was assessed at the end of the 96 hr and mussels were considered to be alive if valve or foot movement was observed within 5 min of observation under a dissection scope. Trials consisted of three replications at each testing concentration and were only considered valid if the calculated average of replicated control group survival was > 90% through completion of the trial (ASTM International, 2022). Additionally, test solutions were renewed during the 48-hr assessment by replacing approximately 80% of the solution volume with fresh ammonia solution to maintain target concentrations. Ammonia concentrations were tested for stock solutions at the beginning of each juvenile trial and from exposure beakers of pooled replicates after 48 hr and at the endpoint of each trial, but the initial stock concentration was not recorded for two trials of L. bergmanni and two trials of L. satura. (Table A1, Appendix A). Mussels were not fed from juvenile transformation through the duration of trials. Water quality measurement procedures were consistent with the glochidia trials with an additional measurement performed during 48-hr survival checks. Juvenile trials were also carried out in incubators (see Glochidia testing) to provide consistent temperatures and photoperiods.

Water quality

Measured water quality parameters, including pH, hardness, and alkalinity measurements, did not vary more than 10% throughout the duration of independent trials. In the individual treatments collected from pooled exposure beakers and measured at the beginning and ending of all trials, temperature ranged from 17.0 to 20.8 °C, pH ranged from 7.6 to 8.4, conductivity ranged from 222 to 644 mg/L, alkalinity ranged from 37 to 135 mg/L, hardness ranged from 68 to 99 mg calcium carbonate/L, and dissolved oxygen ranged from 4.4 to 8.0 mg/L. Dissolved oxygen concentrations never fell below the 4.0 mg/L maintenance threshold identified for glochidia and juvenile acute test acceptability (ASTM International, 2022). The range of pH measurements exceeded the targeted 7.4 to 7.8 of moderately hard water (ASTM International, 2023), but similar tests have produced acceptable results with pH in this range (Wang et al., 2007b; 2016). The variability in water quality measurements could be due to differences in glochidia and juvenile holding water, biological degradation during testing, inconsistencies in test chamber volumes, thermal variability, and ammonia gas venting. The average water quality measurements from treatments collected at the starting and ending point of each test were assessed for all trials and presented in Table A2 (Appendix A).

Data analysis

The arithmetical averages of LC50 and LC05 results were calculated for species with more than one successful trial and the coefficient of variation (CV) was calculated from average replicated trial results for each species at each life history stage to evaluate inter- and intraspecies variability. The overlap of 95% CIs from LC50 and LC05 results were evaluated. A one-way analysis of variance (ANOVA) and Tukey’s honestly significant difference test were used to detect significant differences in LC50 and LC05 between species. Binomial regression models, lethal concentrations, and associated test statistics were calculated with the “ecotox” package version 1.4.4 (Hlina et al., 2021), whereas ANOVA and Tukey’s honestly significant difference tests were calculated with the “stats” package (version 4.4.1) in R version 4.2.3 (R Core Team, 2020).

Results

Acute glochidia tests

The average glochidia viability of the control was > 90% in 14 of 16 glochidia trials following 24-hr of exposure and 13 of 16 trials following 48 hr of survival (ASTM International, 2022). A P. purpuratus trial with average control viability of 88% and an L. satura trial with average control viability of 63% were also included in the analysis and denoted with a footnote in Table 1. Glochidia of F. mitchelli did not experience 50% mortality after 24 hr of exposure at the highest testing concentration and the 24-hr LC50 for this species was reported as greater than the calculated value. Control survival for F. mitchelli fell to 73% after 48 hr of exposure. Although trials for all species except C. necki were evaluated after 48-hr exposures, these results had less consistent control survivals and are of uncertain ecological significance (ASTM International, 2022; Bringolf et al., 2012; Pillow, 2009; USEPA, 2013). Glochidia 24-hr TAN lethal concentrations varied both within and between the eight species tested. Intraspecies variation of 24-hr LC50s (CV: 9%–18%) and LC05s (CV: 6%–20%) was smaller than between species LC50s (CV: 1%–78%) and LC05s (CV: 1%–55%, Table 1). There were significant differences between species in 24-hr LC50s (F_{7, 8} = 15.39, p < 0.001) and LC05s (F_{7, 8} = 51.12, p < 0.001; Figure 2, Table 1). Lampsilis teres was the most sensitive species tested with the lowest average LC50 (32.5 mg/L) and LC05 (5.6 mg/L). Fusconaia askewii had the highest 24-hr LC50 (88.8 mg/L, Figure 2), and L. bergmanni had the highest average LC05 (51.3 mg/L, Figure 2). Lethal concentrations after 48 hr of exposure had greater variability than 24-hr results both within (CV: 6%–35%) and between species (CV: 7%–84%). Increased variability was also observed in 48-hr LC05s within (CV: 4%–95%) and between (CV: 4%–140%) species. There were also differences between species in 48-hr LC50s (F_{6, 8} = 14.8, p < 0.001) and LC05s (F_{6, 8} = 17.14, p < 0.001; Table 2). Most species shared intermediate TAN sensitivity with significant overlaps of 95% CIs around the mean, but relationships were not consistent between LC50 and LC05 results (Figure 2).

Figure 2.

Total ammonia nitrogen (TAN, A) lethal concentration 50% (LC50) and (B) LC05 (24 hr) for glochidia of eight species of freshwater mussels. Error bars represent 95% confidence intervals (CIs) around the LC50s or LC05s. All results were normalized to a pH of 7 and 20 °C. For species tested more than once (n = 2 for L. bergmanni, n = 3 for L. satura and P. purpuratus, n = 4 for L. hydiana) the mean LC50s and LC05s and 95% CIs were computed.

Open in new tabDownload slide

Acute juveniles tests

Three 96-hr juvenile trials were completed for each of four species of unionid mussels. The average control viabilities for 11 of the 12 trials met the 90% survival acceptance guidelines after 96 hr of exposure (ASTM International, 2022). The results from one L. bergmanni trial were also included with an average control viability of 87%.

Juvenile 96-hr TAN lethal concentrations exhibited larger variation than glochidia tests both within and between the four species tested. Similar to glochidia trials, intraspecies differences in LC50s (CV: 6%–30%) and LC05s (CV: 13%–37%) were smaller than between species LC50s (CV: 1%–51%) and LC05s (LC05 CV: 24%–75%, Table 3). There were significant differences between species in LC50s (F_{3, 8}) = 10.1, p = 0.004), and LC05s (F_3,8) =13.85, p = 0.002 (Figure 3, Table 3). Juveniles of L. satura were the most sensitive species tested with an average 96-hr LC50 of 20.2 mg/L and LC05 of 5.9 mg/L. In contrast to their high glochidia sensitivity, L. hydiana juveniles were the least sensitive species tested, with an average LC50 of 43.1 mg/L and LC05 of 19.1 mg/L.

Figure 3.

Mean total ammonia nitrogen (TAN; A) lethal concentration 50% (LC50) and (B) LC05 (96 hr) for juveniles of four species of freshwater mussels. Error bars represent SE (three trials per species). All results were normalized to a pH of 7 and 20 °C. Different letters indicate significant differences in lethal concentrations identified by Tukey’s honestly significant difference test (p < 0.05).

Open in new tabDownload slide

Average juvenile LC50s of 30.7 mg/L (18.1–45.1 mg/L) and LC05s of 11.7 mg/L (4.4–21.6 mg/L) were lower than average glochidia LC50s of 59.9 mg/L (31.4–82.7 mg/L) and LC05s of 28.4 mg/L (12.6–53.6 mg/L), and statistically significant differences in LC50s (F_{1, 22} = 26.29, p < 0.001) and LC05s (F_{1, 22} = 15.94, p < 0.001) were observed for L. bergmanni, L. hydiana, L. satura, and P. purpuratus, which were tested during both life stages for this study.

Discussion

This is the first study to assess acute ammonia toxicity lethal concentrations from early life stages of freshwater mussels in Texas, none of which had been tested previously, and to the best of our knowledge, our results include the first lethal concentrations assessed for a representative of the Cyclonaias genus. The range of LC50s observed in the lethal concentration trials of both early life stages tested were comparable with the range of values and variability in previously published results (Miao et al., 2010; Mummert et al., 2003; Newton & Bartsch, 2007; Wang et al., 2007a; Wang et al., 2007b; Wang et al., 2008; Wang et al., 2016, Figure 4).

Figure 4.

Total ammonia nitrogen (TAN) species mean acute values (SMAVs) based on values reported by U.S. Environmental Protection Agency, (2013, grey bars) and computed as geometric means of 50% lethal concentrations (LC50s) obtained in this study (white bars).

Open in new tabDownload slide

The LC05 was calculated and reported for this study, because models of toxicant induced reductions in population growth have predicted that changes as small as 5% make extinction twice as likely for some aquatic organisms (Snell & Serra, 2000). It is also consistent with previously published acute thermal tolerance 5% lethal temperatures in early life stages of the federally endangered species (C. necki, F. mitchelli, and L. bergmannni) tested in this study (Khan et al., 2019) that are limited to one or two populations and contained within a single river basin.

Our finding that juveniles were more sensitive than glochidia (LC50: 2.1-fold; LC05s: 2.6-fold) was consistent with previous studies that found a greater than twofold higher ammonia sensitivity of juveniles compared with glochidia (Raimondo et al., 2016), although the average 48-hr TAN LC50s for glochidia found in (Wang et al., 2007b) were actually lower (8.6 mg/L) than that of the juveniles (14.6 mg/L). Juvenile mussels are often burrowed in the sediment, but previous research has shown that this behavior offers minimal protection from acute exposures to ammonia (Miao et al., 2010; Wang et al., 2011). The most likely explanation for the decreased lethal tolerance in juveniles was the extended 4-day duration of juvenile trials compared with the 24 hr prescribed exposure for glochidia (ASTM International, 2022). An ANOVA comparison to the 24-hr endpoints from glochidia trials (59.9 mg/L) identified no statistically significant differences (F_{1, 22} = 1.828, p = 0.19) between the 24-hr LC50s (51.3 mg/L) of newly transformed juveniles from the same species.

Intraspecies repeated measure variability in the glochidia trials (measured as CV of LC50 values of the same species, which was 8% to 18% in this study) were somewhat lower compared with previously documented acute freshwater mussel trials (25%–27%, Wang et al., 2007b). Similarly, for juveniles trials the measured CV of LC50s of the same species in our study (6%–30%) was lower than intralaboratory variability for 96-hr juvenile mussel trials for L. silliquoidea (42%, Wang et al., 2008). The intraspecies variability from these trials was likely less pronounced than previous studies due to consistent environmental testing conditions, with many trials conducted simultaneously within the same incubator with pooled glochidia extracted from the same adult mussels.

The failure to meet ASTM International’s prescribed 90% endpoint control survivals for two of the 16 glochidia trials undertaken during this study was potentially due to several confounding causes. Brooding mussels were held for extended periods in well water with different water quality characteristics than the reconstituted moderately hard water used for glochidia and juvenile toxicity trials. The average pH of 7.2 of holding water was lower than the approximate equilibrium pH of 7.6 to 8.4 used for testing. The total hardness (372 mg CaCO3/L), and total alkalinity (260 mg/L) of holding water was also more than two times the concentration of the water used during toxicity trials (Table A2, Appendix A). These differences in water chemistry between well water and testing solutions may have caused stress-induced mortalities in glochidia controls as they acclimated to new conditions (ASTM International, 2022). The measured TAN concentration in the pooled control beakers at the endpoint of the L. satura trial with unacceptable survival measured 0.07 mg/L. This elevated ammonia concentration in the control beaker may indicate that cross-contamination from proximity to concentrated ammonia treatments during the incubation period contributed to glochidia mortalities.

For glochidia, the average and range of acute 24-hr TAN LC50s from the eight species of genera in this study (62.3 mg/L, range: 31.4–> 113 gm/L) was similar to previously published results (72.8 mg/L, range: 17.8–> 160) from eight mussel species in seven genera (Wang et al., 2007b; USEPA, 2013). For newly transformed juveniles, the range of 96-hr juvenile TAN LC50 in this study (18.1–45.1 mg/L) from four tested species in two genera was also comparable with previously reported data (13.6–94.1 mg/L) from six species in three genera (Wang et al., 2007b; USEPA, 2013).

Congeners of the species tested in this study had lower, similar, or higher LC50s in previous studies. For example, L. satura tested here had a mean 96-hr juvenile LC50 of 21.1 mg/L (18.1–21.3 mg/L), which was less than half of the previously reported LC50 of 50.6 mg/L (47.2–54.1 mg/L) for the morphologically and genetically similar Lampsilis cardium (Inoue et al., 2020; Newton & Bartsch, 2007; USEPA, 2013; Williams et al., 2008). For L. bergmanni, average TAN LC50s of both glochidia (75.5 mg/L; range: 70.5–80.6 mg/L) and juveniles (27.4 mg/L (range: 21.9–39.3 mg/L) were slightly lower than those of L. siliquoidea, which is one of its closest genetic relatives (Inoue et al., 2020; glochidia: 93.3 mg/L, 49.7–> 160 mg/L; juveniles: 40.1 mg/L range: 24.3–94.1 mg/L; Wang et al., 2007b; Wang et al., 2008; USEPA, 2013). The only Potamilus species with available published data was P. ohiensis that had a 24-hr glochidia LC50 of > 109 mg/L, which is larger compared with our results (68.6–82.7 mg/L) for P. purpuratus (Wang et al., 2007b; USEPA, 2013). A 6-hr glochidia LC50 of 47.7 mg/L from Fusconaia masoni was the only published data available from this genus, which was less than half of the 24-hr glochidia LC50 of 88.8 mg/L for F. askewi and lower than >69.8 mg/L for F. mitchelli in this study.

Three of the species tested here were short-term brooders (tachytictic, i.e., C. necki, F. askewi, and F. mitchelli) that also expel their glochidia encased in protective conglutinate membranes. The removal of conglutinate casings may have reduced the toxicological resistance of glochidia, as conglutinates have been shown to protect viable glochidia from toxins for at least 4 days prior to uptake by a fish host (Bringolf et al., 2012; Gillis et al., 2008, Pillow, 2009). Protective effects of conglutinate membranes were not assessed in this study.

Regulatory guidance

The USEPA has published multiple iterations of ambient water quality criteria (AWQC) for ammonia under the Clean Water Act since 1985 and received a significant update in 2013 to include 24-hr exposure tests from glochidia of freshwater mussels. The AWQC serve as recommendations to states for setting water quality standards. These values are derived from genus mean acute values (GMAV) from published data of 69 aquatic species ranked by sensitivity. The current AWQC includes multiple life stages of 16 unionid mussels from 11 genera. The pH and temperature–normalized acute criterion magnitude were designed to be protective of 95% of aquatic genera in waters of the United States, so long as the 1-hr average TAN concentration is not exceeded more than once every 3 years. The AWQC also advises the consideration of site-specific ammonia criterion in waters with known populations of federally listed species (USEPA, 1985, 1999, 2013). Mean 24-hr glochidia LC50s and 95% CIs from all eight species of Texas mussels tested were greater than federal criterion, with the exception of the L. teres LC50 of 32.5 mg/L (95% CI, 23.3–50.4) which fell between the final acute value of 33.5 mg/L and the acute criterion magnitude of 17.0 mg/L at pH 7 and 20 °C.

The inclusion of the results from this study would change future calculations of USEPA GMAVs for several unionid genera. The LC50 of (43.8 mg/L) for the federally endangered Cyclonaias necki in this study is the first ammonia toxicity tolerance data available for the Cyclonaias genus in literature. This genus was more sensitive than the Lampsilis genus and would displace that genus on the AWQC as the 5th most sensitive overall if it were to be included in the AWQC. The geometric mean of the LC50s (35.7 mg/L) for all life stages of the four Lampsilis species and single Potamilus species (47.1 mg/L) tested in this study are lower than the current USEPA–calculated GMAV for their respective genera (46.6 and 109 mg/L), whereas the LC50s for the two Fusconaia species in this study (88.8 and > 69.8 mg/L) were greater than the current GMAV (47.4 mg/L) for that genus.

The Texas Commission on Environmental Quality (TCEQ) is responsible for the implementation of state-wide surface water quality standards; however, a universal state-wide acute or chronic numerical criterion for ammonia nitrogen in the surface waters of Texas has not been implemented to date. Ammonia toxicity is addressed on a site-specific basis by implementation of chronic whole-effluent toxicity testing of aquatic organisms in permitted discharges (TCEQ, 2010, 2022). Researchers have recently documented the feasibility of incorporating juvenile unionid mussels into whole-effluent toxicity tests as indicators of aquatic health, which could potentially be used in future testing guidance (Wang et al., 2021).

Conclusion

This study found that glochidia and juvenile life stages of freshwater unionid mussels collected in Texas had similar ammonia tolerances to species found throughout the United States under controlled laboratory conditions. Testing conditions were likely not representative of water quality conditions that mussels will experience in the wild, and it is important to recognize that unionids living in subtropical regions like Texas will undoubtedly be exposed to higher temperatures than in northern latitudes, which can lead to increased ammonia toxicity. A greater sensitivity to ammonia in the juvenile life stage was also consistent with previously tested species, although this may stem from differences in prescribed test durations between life stages. The lethal tolerance values from the eight species tested support the appropriateness of the 2013 USEPA AWQC for the protection of aquatic invertebrates, as the LC50 for all but one species was greater than the final acute value. Low-level LC05 ammonia tolerance results for the federally listed C. necki, F. mitchelli, and L. bergmanni will inform the development of water quality guidance and site-specific conservation measures to ensure the future health of these species, but acute sensitivity variabilities observed in these species highlight the need for additional species-specific chronic toxicity data to inform longer term conservation management decisions for these at-risk populations. Although chronic ammonia toxicity was not the focus of this study, the demonstrated variations in ammonia sensitivity among endangered and threatened species tested in this study could be used to tailor species-specific conservation management and policy decisions and additional chronic toxicity data for these at risk species would also help to inform these decisions.

Supplementary material

Supplementary material is available online at Environmental Toxicology and Chemistry.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Author contributions

Lee J. Gudgell (Formal analysis, Investigation, Methodology, Writing—original draft, Writing—review & editing), Somerley J. Swarm (Investigation, Methodology, Writing—original draft, Writing—review & editing), Ericah D. Beason (Investigation, Methodology, Writing—review & editing), Tara L. Lanzer (Investigation, Writing—review & editing), Clinton R. Robertson (Conceptualization, Writing—review & editing), Astrid N. Schwalb (Conceptualization, Funding acquisition, Methodology, Project administration, Supervision, Writing—original draft, Writing—review & editing)

Funding

This study was funded by the Texas Parks and Wildlife Department and U.S. Fish and Wildlife through State Wildlife Grant (TX T-237-R-1, F21AF03051-00) awarded to A.N.S. Additional personnel and resources were contributed by the Guadalupe-Blanco River Authority and the U.S. Fish and Wildlife Service San Marcos Aquatic Resources Center.

Conflicts of interest

The authors declare no conflicts of interest.

Ethical statement

Fish were held and handled by U.S. Fish and Wildlife Service at their facility in San Marcos and followed internal animal welfare guidelines.

Disclaimer

The views presented herein are those of the authors and do not necessarily reflect those of the U.S. Fish and Wildlife Service.

Acknowledgments

We would like to acknowledge C. Norris (Guadalupe-Blanco River Authority) and R. Gibson (U.S. Fish and Wildlife Service [USFWS]) for providing supervisory support and technical expertise. Thanks to A. Seagroves Ruppel (USFWS) for her technical assistance in launching this project. Thanks to K. Phelps (TXST) for her assistance in the lab. Thanks to M. Johnson (USFWS) and C. Smith (UT Austin) for their assistance in brood collection.

References