INTRODUCTION

Bioassessments that characterize and quantify the impacts of eutrophication are prerequisites to monitoring the efficacy of management practices designed to improve the integrity of nutrient-impaired waters (Smucker and Vis 2009). Algae are a major component of the trophic base of most shallow lotic systems and assemblage composition may reflect periphyton habitat quality (Stancheva and Sheath 2016). The composition of algae assemblages of the majority of streams in the Interior Plateau Level III Ecoregion that extends from southern Indiana and Ohio to northern Alabama is unknown (Grimmett and Lebkuecher 2017). This lack of knowledge limits the ability of watershed managers to monitor changes of habitat quality. This work documents the composition of algal assemblages as they are used to assess the effects of water quality in the upper and middle reaches of the Harpeth River in Middle Tennessee.

Common methods to evaluate the impact of nutrient enrichment of shallow lotic systems include measurements of the concentrations of benthic chlorophyll a (chl a), ash-free dry mass of benthic organics (AFDM), nutrient concentrations in water, and organism composition. The use of concentration of chl a as an indicator of trophic state is complicated by the influence of many abiotic and biotic characters including irradiance, temperature, water velocity, herbivory, volume of submerged wood, and time period between spates of high discharge (Lebkuecher et al. 1998, Kurle and Cardinale 2011). Ash-free dry mass of benthic organics may be influenced by the same factors that influence the concentration of chl a. In addition, AFDM is affected by organic inputs that differ by season, streambank characteristics, and allochthonous inputs (Lebkuecher et al. 2015). Chemical analyses do not indicate ecological condition and may not accurately reflect water quality (Andrus et al. 2013). Pulses of pollution may be missed during sampling and water with a high concentration of biomass may have low nutrient concentrations due to high nutrient demand (Dodds 2006). Periphyton composition is often a more accurate indicator of trophic state relative to biomass and nutrient concentrations partially because of rapid, yet sustained changes of composition as a result of changes of habitat quality.

Bioassessments using algal assemblages are most useful for management practices because they reflect a wide range of anthropogenic impacts (Rimet 2012). Several states including Kentucky, Montana, Oklahoma, and Texas use evaluations of diatom composition as a standard protocol to monitor changes of water quality (Stevenson et al. 2008, Szczepocka and Szule 2009). The objectives of this study were to document the richness, composition, and biomass of algae in the Harpeth River at two sites upstream and at two sites downstream of the Franklin Wastewater Treatment Facility to assess the impact of nutrient enrichment on the algal assemblages.

METHODS

Sampling Site Locations and Dates

Four sites were sampled in the Harpeth River on September 30, 2017 (Figure 1): two sites upstream and two sites downstream of the Franklin Wastewater Treatment Facility in downtown Franklin (Appendix 1). The uppermost site (Site 1) is located approximately 17 km upstream of Franklin in a rural, agricultural region. Site 2 is located approximately 4 km upstream of downtown Franklin in a densely populated neighborhood. Effluent from the Franklin Wastewater Treatment Facility enters the river at river near downtown Franklin approximately 5 km downstream of Site 2 and approximately 5 km upstream of Site 3. Site 3 is located approximately 6 km downstream of downtown Franklin in a densely populated neighborhood. Site 4 is located approximately 16 km downstream of Franklin in an area with a mix of neighborhoods and agriculture.

Figure 1.

Location of Harpeth River Watershed (black area) in Tennessee, location of the Franklin Wastewater Treatment Facility (square), and locations of the four sampling sites (circles). Sites 1 and 2 are upstream and Sites 3 and 4 are downstream of the Franklin Wastewater Treatment Facility.

Laboratory Methods

Periphyton pigment concentrations and ash-free dry mass

Methods to determine periphyton pigment concentrations and AFDM followed those described in Lebkuecher et al. (2015) and Grimmett and Lebkuecher (2017). One cobble was placed in a glass pan containing 0.1 L of 90% acetone and the periphyton was removed with a single-edged razor blade and scrub brush. Ten mL aliquots of periphyton suspended in acetone were placed in a mortar, ground with a pinch of sand and a pestle for 2 min., and filtered through Whatman no. 1 filter-paper circles. Optical density (OD) of the supernatant was determined at 664 nm to determine the concentration of chl a, then at 665 nm following acidification with 0.1 N HCl to determine the concentration of pheophytin a. Optical density determinations occurred within minutes of acetone extraction and in very low light to ensure pheophytin was not degraded. Concentrations of chl a corrected for pheophytin a were calculated using the equations of Lorenzen (1967). The chl a to pheophytin a ratio was indicated as the ratio of OD₆₆₄ to OD₆₆₅ (Baird et al. 2017).

Ash-free dry mass of organic benthic matter was determined following evaporation of the acetone at 25°C. During the drying process, periphyton was transferred to beakers, then to the same crucibles in which weights were measured to ensure no loss of organic matter. Ash-free dry mass of organic benthic matter and inorganic sediment weights were determined as described by Baird et al. (2017). Values for AFDM were increased by the proportion of the periphyton removed to determine pigment concentrations. The surface area of cobble from which periphyton was removed was calculated by covering the upper surface of cobble with aluminum foil, weighing the foil, and extrapolating weight to surface area (Hauer and Lamberti 2006).

Statistical methods

Means for concentrations of chl a, optical density (OD) ratios of OD₆₆₄ to OD₆₆₅ of pigment extracts, and AFDM were compared using Tukey-Kramer honestly significant difference tests preceded by analysis of variance tests. The data were homoscedastic and normal. Assay means were considered significantly different if they differed at the experiment-wise error rate of α = 0.05.

Identification of soft-bodied algal taxa

Large filamentous algae were cut with scissors such that well-mixed aliquots of filamentous-algae sections and microalgae in the sample could be obtained. Wet mounts on a ruled microscope slide (NeoSci, Nashua, New Hampshire) with a 16-mm² grid divided into eight 2-mm² squares were used to determine percent composition as described by Woelkerling et al. (1976) and Schoen (1988). Macroalgae were not separated from microalgae. Soft algae within a 2-mm² square were observed at 100×, 400×, and 1000× magnification and identified to the lowest taxon possible. Taxa were recorded as units. A unit was considered one cell of unicellular taxa, one colony of colonial taxa, and each 10 μm-length of filamentous taxa as recommended by Stevenson and Bahls (2006). Taxa were enumerated until at least 800 units counted, or for samples with very little soft-bodied algae relative to diatoms, until at least 20 wet mounts were observed. Wet mounts from samples collected by multi-habitat sampling were searched using 100×, 400×, and 1000× magnification until no new taxa were observed in at least five consecutive wet mounts. Taxa identified from multi-habitat samples were recorded as present and thus were not compared to taxa sampled from cobbles to determine relative abundance. Primary taxonomic references used to identify soft-algae taxa included, Cocke (1967), Prescott (1982), Whitford and Schumacher (1984), Anagnostidis and Komárek (1988), and John et al. (2011). The percent of overall soft-bodied algal units and overall diatom units at each site was estimated by counting the number of soft-bodied algae units and diatom units in 2-mm squares of the ruled microscope slide until at least 800 units were counted.

Identification of diatom taxa

Frustule preparation for permanent mounts followed the methods of Carr et al. (1986). Organic debris and intracellular material were completely removed by placing concentrated frustules in 2.5% sodium hypochlorite for 1 h. Aliquots of cleaned frustules (50 μL) were pipetted onto glass cover slips, dried at 50°C, and mounted on glass microscope slides with Permount mounting medium. Valve ornamentations were clearly visible using 1000× magnification. All valves in the field of view at 1000× magnification were identified and tallied until a minimum of 200 valves from each stream site were identified, the minimum number required to calculate the pollution tolerance index of diatom assemblages (KDOW 2008). Primary taxonomic references used to identify diatom taxa included Patrick and Reimer (1966, 1975), Krammer and Lange-Bertalot (1986, 1988, 1991a, 1991b), and Ponader and Potapova (2007). The permanent mounts are maintained in the Austin Peay State University Herbarium in Clarksville, Tennessee.

Calculation of metrics and indices

Shannon diversity index (H′) and evenness (J) of soft-algae and diatom assemblages were calculated to provide information about assemblage structures by the equations of Shannon and Weaver (1949):

where P_i = abundance of species i, and S = richness (number of taxa). Pairwise percent similarities of diatom and soft-algae assemblages associated with cobble were calculated as the sum of the smaller of the two percent-composition values for all taxa common to two sites (Whittaker and Fairbanks 1958). A percent-similarity value of 0 indicates that the two assemblages compared do not have a single taxon in common. A percent-similarity value of 100 indicates that the two assemblages compared have the exact same percent composition of taxa.

The pollution tolerance index (PTI) for diatom assemblages (KDOW 2008) was calculated as:

where n_j = number of individuals of taxon j, t_j = eutrophication-tolerance value (one to four; KDOW 2008) of taxon j, and N = total number of individuals assigned a eutrophication-tolerance value and tallied to calculate the index. The PTI ranges from one (all taxa very tolerant to eutrophic conditions) to four (all taxa very intolerant of eutrophic conditions). A value for the PTI < 2.6 indicates an assemblage is impaired by eutrophic conditions as determined by comparisons of PTI values of stream sites with different trophic states in Middle Tennessee (Lebkuecher et al. 2011).

The organic pollution index (OPI) is the percentage of diatoms tolerant of organic pollution listed in Kelly (1998). Organic pollution index values of >20 imply organic pollution impacts the composition of diatom assemblages and values >40 imply the habitat is severely impaired by excessive concentration of organic matter (Kelly 1998). The siltation index (SI) is the sum of Navicula, Nitzschia, and Surirella, and the taxa formerly identified as Navicula, Nitzschia, and Surirella divided by the total number of diatoms and multiplied by 100 (Bahls 1993). Motile diatoms other than Navicula sensu lato, Nitzschia senu lato, and Surirella are not used to calculate the SI (Bahls 1993). The SI values range from 0 to 100. High SI values signify that sediments impact the structure of diatom assemblages. Belton et al. (2005) suggested that SI values near 40 indicate an impact of sediments on diatom assemblages.

The algae trophic index of soft-algae assemblages (ATI) was calculated as:

where: n_j = number of taxon units j sampled at a site, ti_j = trophic-indicator value for taxon j, and N = total number of taxon units at the sampling site used to calculate the index (Grimmett and Lebkuecher 2017). The trophic-indicator values are the abundance-weighted average of the concentration of benthic chl a, listed in Grimmett and Lebkuecher (2017). Taxa not identified to species were excluded from index calculations and included less than 10% of the algae observed. The algae trophic index of soft-algae assemblages values for stream reaches studied by Grimmett and Lebkuecher (2017) were very similar to and significantly correlated with concentrations of benthic chl a (mg^.m⁻²). Therefore, ATI values correspond with Dodds et al. (1998) suggested classification of temperate streams with chl a concentrations <20 mg^.m⁻² as oligotrophic and concentrations >70 mg^.m⁻² as eutrophic.

Concentrations of total phosphorus and total nitrogen of water samples

Nutrient concentrations of water samples collected approximately 5 cm below the surface were determined using a Lachat QuickChem 8500 Flow Injection Analyzer (Lachat Instruments, Loveland, Colorado). Concentrations of total phosphorus (TP) were determined using the persulfate-digestion and the ascorbic-acid method (Baird et al. 2017). Concentrations of total nitrogen (TN) were determined by the persulfate-digestion and cadmium-reduction method (Baird et al. 2017). Concentrations of TP and TN were determined because Dodds (2003) suggested concentrations of TP and TN are stronger indicators of trophic state relative to the concentrations of dissolved forms of N and P. In addition, concentrations of TP and TN are used to designate the trophic state of streams by Dodds et al. (1998).

RESULTS

Nutrient Concentrations and Periphyton Biomass

Water samples from all four sites contained eutrophic concentrations of TP and ranged from 310 μg^.L⁻¹ at the uppermost site to 1035 μg^.L⁻¹ at the site immediately downstream of the wastewater treatment facility (Table 1). Concentrations of total nitrogen (TN) were in or near the range considered mesotrophic for streams and ranged from 687 μg^.L⁻¹ at the uppermost site to 1434 μg^.L⁻¹ at site three immediately downstream of the wastewater treatment facility. The TN:TP ratios were lowest at the two sites downstream of the Franklin Wastewater Treatment Facility as a result of the very high concentrations of TP at these sites.

Table 1.

Concentrations of total phosphorus (TP) and total nitrogen (TN) of water samples and periphyton characteristics at sites sampled in the Harpeth River. Means ± standard error for concentrations of chlorophyll a, optical density (OD) ratios of OD₆₆₄ to OD₆₆₅ of pigment extracts, and ash-free dry mass of benthic organic matter represent four replicates and are not significantly different at the experiment-wise error rate of α = 0.05. Sites 1 and 2 are upstream and Sites 3 and 4 are downstream of the Franklin Wastewater Treatment Facility.

Concentrations of benthic chlorophyll chl a (Table 1) were all >103 mg^.m⁻² and did not differ significantly. None of the optical density (OD) ratios of OD₆₆₄ to OD₆₆₅ of the pigment extracts were <1.5, a threshold value used to indicate the algae were in poor health or senescent (Baird et al. 2017). The concentrations of AFDM were all were >10 g^.m⁻² and did not differ significantly.

Composition of Algal Assemblages

We identified 186 taxa of algae: 92 taxa of soft-bodied algae (Appendix 3) and 94 taxa of diatoms (Appendix 4). The most abundant soft-bodied algae sampled (Table 2) was the filamentous Rhodophyta Audouinella hermannii (Roth) Duby (16.0%) followed by the filamentous cyanobacterium Leptolyngbya foveolarum (Mont.) Anagn. and Komárek (11.4%). The most abundant diatom sampled (Table 3) was Achnanthidium rivulare Potapova and Ponander (10.4%) followed by Navicula minima Grun. (7.6%).

Table 2.

Most abundant soft-bodied algal taxa at sites sampled in the Harpeth River. Numbers in parentheses represent percent composition. Sites 1 and 2 are upstream and Sites 3 and 4 are downstream of the Franklin Wastewater Treatment Facility.

Table 3.

Most abundant diatom taxa at sites sampled in the Harpeth River. Numbers in parentheses represent percent composition. Sites 1 and 2 are upstream and Sites 3 and 4 are downstream of the Franklin Wastewater Treatment Facility.

Percent composition of algae groups such as diatoms, cyanobacteria, and Chlorophyta differed dramatically between sites (Table 4). Cyanobacteria were the most abundant at the uppermost site and diatoms were most abundant at the lowermost site. Both the soft-bodied algal and diatom assemblages were distinct across sites (Table 5). The most similar assemblages of soft-bodied algae and diatoms were the assemblages at the two uppermost sites, the two sites with the lowest and most similar concentrations of TP of water samples (Table 1). The most dissimilar assemblages of soft-bodied algae and diatoms were those at the sites with the greatest differences in concentrations of TP, the uppermost site and the site immediately downstream of the wastewater facility the sites.

Table 4.

Percent composition of algae groups at sites sampled in the Harpeth River. Sites 1 and 2 are upstream and Sites 3 and 4 are downstream of the Franklin Wastewater Treatment Facility.

Table 5.

Percent similarity of soft-bodied algae and diatom assemblages between sites sampled in the Harpeth River. Sites 1 and 2 are upstream and Sites 3 and 4 are downstream of the Franklin Wastewater Treatment Facility.

Metrics and Indices for Algae Assemblages

Richness of soft-bodied algal taxa increased downstream (Table 6). The similar values for the Shannon diversity index for soft-algae assemblages among sites were due partially to similar evenness. The highest value for the algae trophic index (ATI) for the assemblage of soft-bodied algae at site three indicates that this assemblage was most impacted by eutrophication.

Table 6.

Metrics and indices for soft-bodied algal assemblages at sites sampled in the Harpeth River. Sites 1 and 2 are upstream and Sites 3 and 4 are downstream of the Franklin Wastewater Treatment Facility.

Greater values for the Shannon diversity index for the diatom assemblages (Table 7) relative to the soft-algae assemblages at the same sites is due largely to greater the evenness of the diatom assemblages. Values for the pollution tolerance index for diatom assemblages (PTI) at the all the sites are <2.6, which indicate eutrophic conditions. Values for the organic pollution index (OPI) > 20 and the siltation index (SI) > 40 for the diatom assemblages at Sites 3 and 4 were above the threshold values used to designate assemblages as impacted by organic pollution and siltation, respectively.

Table 7.

Metrics and indices for diatom assemblages at sites sampled in the Harpeth River. Sites 1 and 2 are upstream and Sites 3 and 4 are downstream of the Franklin Wastewater Treatment Facility.

DISCUSSION

The high concentrations of benthic chl a at all four sites indicate eutrophic conditions based on Dodds et al. (1998) classification of temperate streams with chl a concentrations >70 mg^.m⁻² as eutrophic. The concentrations of AFDM at all four sites also indicate eutrophic conditions. All sites had AFDM values >10 g^.m⁻², a threshold value indicative of eutrophic environments based on earlier studies (O'Brien and Wehr 2010, Lebkuecher et al. 2015, Grimmett and Lebkuecher 2017). The eutrophic concentrations of TP at the sites upstream of the Franklin Wastewater Treatment Facility likely reflect the heavy anthropogenic activities in the watershed and naturally high concentrations of P in the limestone bedrock (TDEC 2005). Concentrations of approximately 180 μg TP^.L⁻¹ are suggested to be a more realistic expectation of moderate levels of TP of surface waters in the Nashville Basin (TDEC 2005) as opposed to the boundary value of <75 μg TP^.L⁻¹ suggested by Dodds et al. (1998) to designate lotic systems as mesotrophic.

The TN:TP ratios at all four sites are well below TN:TP ratios hypothesized to indicate potential N limitation of algae growth. For example, N may be considered limiting when the TN:TP ratio is <10, and by P when the TN:TP ratio is >17 (Dodds 2003). The low TN:TP ratios may have precluded the potential for significant differences of concentrations of benthic chl a and AFDM among sites upstream and downstream of the wastewater treatment facility. Earlier studies of determinations of chl a from four cobbles per stream site were adequate to demonstrate significant increases of chl a at sites immediately downstream of wastewater treatment plants just outside the Nashville Basin (Lebkuecher et al. 2015, Grimmett and Lebkuecher 2017). However, we acknowledge that four cobbles per sampling site is a relatively small number to estimate chl-a concentration given the spatial heterogeneity of small rivers and may have contributed to the absence of significant differences. In addition, environmental factors such as canopy angle and river morphological characteristics influence periphyton biomass.

High values for the algae trophic index (ATI) for soft-bodied algal assemblages and low values for the pollution tolerance index for diatom assemblages at the sites downstream of the Franklin Wastewater Treatment Facility indicate that additional nutrient enrichment of the Harpeth River as the river flows through Franklin impacts the structure of these assemblages. These index values demonstrate that the assemblages of soft-bodied algae and diatoms sampled downstream of the wastewater treatment facility have a greater relative abundance of taxa tolerant of eutrophic conditions. The higher values for the ATI at the lower three sites were due largely to the high abundances of Audouinella hermannii. The highest value for the ATI at site three was due largely to the high abundance of Cladophora glomerata (L.) Kütz. Audouinella hermannii and C. glomerata are assigned high trophic-indicator values for the ATI which indicate these taxa are abundant at eutrophic sites in Middle Tennessee (Grimmett and Lebkuecher 2017).

The high value for the pollution tolerance index (PTI) for diatom assemblages for the assemblage at the uppermost site is due largely from the abundance of Achnanthidium taxa (52%) and Cymbella affinis (16.5%). Both of these taxa are assigned eutrophication-tolerance values of three or four, given they are common in oligotrophic and mesotrophic lotic systems (KDOW 2008). The lowest PTI value for the assemblage at the site immediately below the wastewater treatment facility is due largely from the low abundance of Achnanthidium (8.3%; Appendix 4) and the greatest abundance of Navicula minima (11.0%). Navicula minima is an indicator of poor quality water and is assigned a eutrophication-tolerance value of 1.0 (KDOW 2008).

The PTI values ≤2.6 for the diatom assemblages of the Harpeth River sites are similar to those of other stream sites in predominately agricultural and urban regions impaired by nutrient enrichment in Middle Tennessee. The PTI values for stream sites in Middle Tennessee considered some of the most nutrient impaired range from 2.3 to 2.0, whereas PTI values for stream sites in Middle Tennessee considered reference sites with good water quality range from 2.8 to 3.0 (Lebkuecher et al. 2011, Lebkuecher et al. 2015, Grimmett and Lebkuecher 2017). The organic pollution index (OPI) values >20 for the diatom assemblages at the sites downstream of downtown Franklin suggest these assemblages may be impacted by organic pollution, yet are well below the threshold value of 40, which indicates severe impairment. Values well above 40 are common in reaches with very high concentrations of organics, such as Elk Fork Creek in the Red River Watershed in northern Middle Tennessee (Lebkuecher et al. 2011) and Jones Creek downstream of the Jones Creek Wastewater Treatment Plant near Dickson, Tennessee (Grimmett and Lebkuecher 2017).

The apparent impact on the diatom assemblages by siltation at the two lowermost sites sampled in the Harpeth River is typical for middle and lower reaches of lotic systems in agricultural regions. For example, diatom assemblages in two streams in morphologically similar watersheds in New Jersey with 1% and 28% agriculture had mean siltation index values of 18 ± 7 and 43 ± 4, respectively (Belton et al. 2005). Siltation index values are most informative when comparing values from stream sites with similar regional conditions given that soil erodibility and land use affects the composition of diatom assemblages. Siltation index values of six stream sites in the Red River Watershed in northern Middle Tennessee, which has highly erodible soils and where >60 % of the land is used for agriculture, were 54 at the watershed's reference site and 78 at the site most impacted by siltation (Lebkuecher et al. 2011).

We do not know the reason why the uppermost site was dominated by cyanobacteria and the lowermost site was dominated by diatoms. Lebkuecher et al. (2015) and Grimmett and Lebkuecher (2017) found no correlation between trophic state and abundances of diatoms verses soft algae in Middle Tennessee streams. The lower abundance cyanobacteria and the greater abundance of Chlorophyta at the sites downstream of downtown Franklin sites relative to the sites upstream are consistent with earlier studies that demonstrated cyanobacteria biomass and diversity were lower at river sites with higher concentrations of soluble reactive phosphorus of water samples in central Spain (Perona et al. 1998, Douterelo et al. 2004). Similar lower abundances of cyanobacteria relative to Chlorophyta at sites with higher concentrations of nutrients occur in other Middle Tennessee streams (Grimmett and Lebkuecher 2017).

The greater number of the same taxa among the diatom assemblages relative to the soft-algae assemblages at the sites studied in the Harpeth River is consistent with earlier studies of several Middle Tennessee streams. A study by Lebkuecher et al. (2015) of three mesotrophic sites and one hypereutrophic site in Sulphur Fork Creek in Middle Tennessee demonstrated that the similarity of diatom assemblages between sites sampled in August was far more consistent, ranging from 44% to 66%, relative to the similarity of soft-bodied algal assemblages, which ranged from 16% to 57%. In a different study of eight sites in eight streams in Middle Tennessee, percent similarity for diatom assemblages between sites sampled in August ranged from 25% to 67% and for soft-bodied algal assemblages, from 2% to 39% (Grimmett and Lebkuecher 2017).

Noteworthy and unexpected taxa identified in the Harpeth River by this study include Chilomonas sp., a nonphotosynthetic cryptomonad, Chroomonas sp., a photosynthetic cryptomonad, and Paulinella chromatophora Lauterborn, a filose thecamoeba. Cryptomonads are much more common in oligotrophic lentic water, although two sp. of Cryptomonas were documented in Middle Tennessee streams with relatively good quality water (Lebkuecher et al. 2015, Grimmett and Lebkuecher 2017). Paulinella chromatophora was not known to occur in the southeastern United States, possesses primitive, cyanobacteria–like plastids, and is frequently referenced as evidence for the endosymbiotic origin of plastids (Graham et al. 2016). Genetic uniqueness of P. chromatophora plastids suggests that all plastids were not acquired from a single primary endosymbiotic event and thus implies that the Archaeplastida supergroup may not be monophyletic (Nowack et al. 2008).

CONCLUSIONS

This study documents the composition of soft-algal and diatom assemblages needed to help monitor the integrity of periphyton in the upper and middle reaches of the Harpeth River. High concentrations of TP of water, benthic chl a, and AFDM at sites upstream of Franklin are likely a result of extensive anthropogenic activities in the watershed and naturally high concentrations of P in the limestone bedrock. Additional eutrophication of river sites downstream of the Franklin Wastewater Treatment Facility is indicated by increased concentrations of TP of water samples, increased values for the algae trophic index for soft-algal assemblages, and decreased values for the pollution tolerance index for diatom assemblages. The results indicate that additional degradation of water quality as the Harpeth River flows through Franklin alters the composition of algal assemblages.