nena masthead
NENA Home Staff & Editors For Readers For Authors

Effects of Phragmites Management on the Ecology of a Wetland
Amy Krzton-Presson, Brett Davis, Kirk Raper, Katlyn Hitz, Christopher Mecklin, and Howard Whiteman

Northeastern Naturalist, Volume 25, Issue 3 (2018): 418–436

Full-text pdf (Accessible only to subscribers. To subscribe click here.)

 

Access Journal Content

Open access browsing of table of contents and abstract pages. Full text pdfs available for download for subscribers.



Current Issue: Vol. 30 (3)
NENA 30(3)

Check out NENA's latest Monograph:

Monograph 22
NENA monograph 22

All Regular Issues

Monographs

Special Issues

 

submit

 

subscribe

 

JSTOR logoClarivate logoWeb of science logoBioOne logo EbscoHOST logoProQuest logo

Northeastern Naturalist 418 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 22001188 NORTHEASTERN NATURALIST 2V5(o3l). :2451,8 N–4o3. 63 Effects of Phragmites Management on the Ecology of a Wetland Amy Krzton-Presson1,*, Brett Davis1, Kirk Raper1, Katlyn Hitz1, Christopher Mecklin1,2, and Howard Whiteman1 Abstract - Effective management of wetlands often involves the suppression of invasive species, and understanding the ecological consequences of management efforts is an important goal. A non-native strain of Phragmites australis (Common Reed; hereafter Phragmites) is aggressively invading North American wetlands, and the species has been shown to alter hydrology and impact aquatic organisms. Clear Creek Wildlife Management Area, in Kentucky, is heavily impacted by Phragmites invasion. Portions of the Wildlife Management Area with Phragmites were treated with herbicide to restore wetland flora, and in this study, we evaluated the effects of both the presence and management of Phragmites on wildlife populations and ecosystem function. We selected 3 locations (treated Phragmites, untreated Phragmites, and Phragmites-free) and surveyed each for water chemistry, wildlife populations, and stable-isotope signatures over a 2-y period. Water chemistry varied with the presence or absence of Phragmites, suggesting differences in nutrient cycling. Fish diversity did not differ among sites, but individual species varied in distribution and abundance between the Phragmites sites and the Phragmites-free site. Turtles showed significant differences in both diversity and body size based on the presence or absence of Phragmites, but not herbicide treatment. We detected no significant differences in frog diversity across treatments. We recorded 8 Kentucky Species of Greatest Conservation Need, but there were few differences in the distribution of these species across sites. Stable-isotope analysis revealed variation in food-web structure based on the presence of Phragmites. These results indicate that herbicides had little effect on fish and herpetofaunal communities in the short term, but potentially significant ecological changes may occur if Phragmites were eradicated. Our conclusions highlight the importance of monitoring habitat restoration to guide future management. A holistic, ecosystem-level approach is necessary to understand the impacts of both invasive species and their management. Introduction Invasive species have been severely impacting habitats and communities worldwide for decades (Gratton and Denno 2006, Simberloff 1996). Wetland ecosystems are particularly susceptible to invasion because they accumulate water, nutrients, and sediment from disturbances to both terrestrial and aquatic systems, providing ample resources for invasive plants to thrive (Zedler and Kercher 2004). One of the best-known and aggressively managed invasive wetland plants in North America is Phragmites australis (Cav.) Trin. ex Steud. (Common Reed; hereafter Phragmites) (Rogalski and Skelly 2012). 1Watershed Studies Institute and Department of Biological Sciences, Murray State University, Murray, KY 42071. 2Department of Mathematics and Statistics, Murray State University, Murray, KY 42071. *Corresponding author - amy.krztonpresson@gmail.com. Manuscript Editor: Robert Bertin Northeastern Naturalist Vol. 25, No. 3 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 419 Phragmites is a coarse, perennial, wetland grass with several haplotypes including 1 native to North America. However, a Eurasian haplotype is rapidly invading wetlands of North America via high seed-production and vegetative propagation, often forming dense monocultures (Ailstock et al. 2001, Saltonstall 2002). This invasive strain of Phragmites can replace more diverse floral communities, as well as alter structural habitat complexity and wetland functioning. Communities consisting of monocultural stands of Phragmites are often considered poor wildlife habitat with low faunal diversity (Paxton 2007, Roman et al. 1984). Invasive Phragmites alters hydrologic and chemical cycles in wetlands where it becomes monotypic, replacing the more diverse native flora; thus, changing the structure, and also the function of the wetland (Meyerson et al. 2000). For example, Armstrong and Armstrong (1988) found that Phragmites limits the cycling of phosphorus and other nutrients by increasing rhizosphere oxidation, which binds these elements in the sediments. The documented and deleterious ecological effects associated with Phragmites invasion warrant the control and eradication of this plant to restore affected wetlands to their pre-invasion state. Options for Phragmites reduction and/or removal include chemical, mechanical, and biological controls (Ailstock et al. 2001, Haslam 1971). Regardless of the method used, the destruction of a plant that has such a large presence in a community is likely to have an impact on the surrounding fauna by simply changing community composition and habitat structure. Phragmites is considered one of Kentucky’s most noxious invasive wetland species, with prevalent populations throughout the western part of the state and along the Ohio River (Mahala 2008). Invasive Phragmites has been particularly successful in establishing a dominant monoculture at the Clear Creek Wildlife Management Area (CCWMA), a property in Hopkins County, KY, managed by Kentucky Department of Fish and Wildlife Resources (KDFWR). In an effort to increase access to this public area and begin restoration of this wetland ecosystem, the KDFWR and Ducks Unlimited collaborated to manage 121 ha of Phragmites on CCWMA. Our goal was to study the effects of herbicide management on the fish and herpetofaunal communities and develop recommendations for future management of Phragmites. Due to limited management funding, the CCWMA was the only Phragmites-invaded wetland in the area being managed at this scale. Thus, we were unable to replicate this study with additional sites. Although our results must be interpreted cautiously, we thought it prudent to take advantage of the unique situation provided and document observations through the early stages of the management effort. In this study, we focused on 2 questions: (1) What effect does Phragmites and its management have on fish, reptile, and amphibian populations and Species of Greatest Conservation Need (SGCN)? and (2) What impact does Phragmites and its management efforts have on the ecosystem as a whole? Our efforts were aimed at monitoring the effects of Phragmites, as well as herbicide treatment of Phragmites, on fish and herpetofaunal diversity, the presence of SGCN, water quality, and ecosystem function. Previous research suggested that Phragmites has a negative effect on wetland fauna, and that vegetation removal results in an alteration of the fish Northeastern Naturalist 420 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 Vol. 25, No. 3 and herpetofaunal communities (Able et al. 2003, Fell et al. 2003, Meyer 2003). Therefore, we hypothesized that areas impacted by Phragmites represented lowerquality habitat for SGCN and would have reduced diversity and altered community structure. We also hypothesized that Phragmites management would increase species richness and abundance in the fish and herpetofaunal communities when compared with unmanaged Phragmites-dominant areas. Finally, we hypothesized that the high growth rate of Phragmites would impact the entire ecosystem by lowering nutrient availability. Field-Site Description Clear Creek is a low-gradient, 5th-order stream (~51 km long) that drains 15,661 ha, of which 347 ha make up the CCWMA (Fig. 1). Since the 1850s, timber removal, a lack of effective soil-conservation techniques in agriculture, and intensive coal mining operations have caused extreme sedimentation, hydrologic alterations, and acid-mine drainage (AMD) in Clear Creek (Hill 1983, Leuthart 1975, Neichter 1972). Prior to these disturbances, Clear Creek had defined banks surrounded by bottomland hardwood forests; however, it is now a broad, permanently inundated wetland (Hill 1983, Leuthart 1975, Neichter 1972). Before reclamation, abandoned mine sites leached AMD into Clear Creek until at least the early 1980s (Hill 1983), with pH measurements in the main channel recorded as low as 2.8 (Niechter 1972). Concentrated sulfate, heavy-metal loads, and acidic water conditions rendered the creek channel depauperate of all but the most tolerant wildlife (Hill 1983). No fish species were observed at all in Clear Creek during research from 1969 to 1974 (Leuthart 1975). Phragmites was not present in plant community descriptions from previous research at Clear Creek, although Hill (1983) noted a dense Phragmites stand about 48 km north near Smith Mills, KY. Thus, Phragmites likely invaded Clear Creek within the last 30 y and has since come to dominate the system. Concurrent genetic analysis of Phragmites specimens from Clear Creek WMA revealed this population to be the exotic, more invasive genotype (Croteau et al. 2013). At the time of this study, the Phragmites in CCWMA had established dense stands throughout the inundated wetland, except for areas where the water deepened, typically near the stream thalweg. There was shallow, slow-moving water inundating the dead Phragmites mulch within the Phragmites stands. The rest of the wetland community was comprised primarily of native plants, including Lemna minor L. (Duckweed), Ceratophyllum demersum L. (Coontail), Peltandra virginica (L.) Schott (Arrow Arum), Potamogeton spp. (pondweeds), and Typha sp. (cattail). As part of this study, we surveyed an adjacent Phragmites-free wetland as a comparison to Clear Creek (Fig. 1). Weir Creek is a 3rd-order stream and tributary of Clear Creek. It drains an area of ~6133 ha and is 18 km long. The predominant plant species included Nuphar advena (Ait.) Ait. f. in Ait. & Ait. f. (Immigrant Pond-lily), Duckweed, Cattail, and Coontail. Weir Creek was characterized by low-gradient, tea-colored water, and a loose bottom consisting of detritus, vegetation, large woody debris, and silt. Northeastern Naturalist Vol. 25, No. 3 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 421 Figure 1. Clear Creek Wildlife Management Area (CCWMA; outlined area that includes treated sites) and associated study sites. Sampling locations are represented by filled circles. Dead Phragmites, the result of herbicide application, is apparent in the treated area. Phragmites-free reference sites located on Weir Creek located in upper left, outside of CCWMA. Map created by Jane Benson, Mid-American Remote-sensing Center (MARC), Murray State University. Northeastern Naturalist 422 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 Vol. 25, No. 3 On 22 August 2009, KDFWR and Ducks Unlimited carried out a chemical treatment of Phragmites on a section of the CCWMA. They undertook an aerial application of a glyphosate herbicide (AquaMaster®) via helicopter on ~121 ha of the 347-ha site at a rate of 15.3 L/ha (E. Williams, KDFWR, Drakesboro, KY, pers. comm.). We collected data in 2 Phragmites-dominated sites on the CCWMA: 1 treated with herbicide, and 1 untreated. The Weir Creek wetland served as our Phragmites-free reference site (Fig. 1). Methods We conducted all statistical analyses in the software R, version 2.10 (https:// cran.r-project.org/). Water chemistry To evaluate water quality, we collected 1–3 replicate water samples from each site every month. We analyzed these samples for nitrogen (nitrate + nitrite), soluble reactive phosphorus (SRP), total phosphorus (TP), and total nitrogen (TN) using a Lachat Quick Chem 8000 Flow Injection Analyzer at the Hancock Biological Station (HBS), Murray, KY. We filtered dissolved organic carbon (DOC) water samples through Whatman GF/F glass-microfiber filters with a nominal pore size of 0.7 mm. We recorded turbidity, temperature, pH, conductivity, and dissolved oxygen (DO) during each sampling event using a YSI 6600 multi-parameter sonde (YSI, Inc., Yellow Springs, OH) and YSI 650MDS hand-held data display system (YSI). We employed MANOVA to compare overall water-quality across sites and individual ANOVAs for each parameter. Faunal sampling We sampled fish at 6 locations per site at least every 3 weeks from September 2009 until August 2010, except during duck-hunting season (October 2009–March 2010). We collected fish by seining and electrofishing according to the standardized methods of the Kentucky index of biotic integrity (KIBI; Kentucky Department of Fish and Wildlife Resources 2008); identifications were to the species level. We conducted ANOVA to evaluate differences in KIBI and Shannon diversity among sampling locations and dates. We used the JACCARD index to compare fish community similarity and analyzed the data via a permutational MANOVA to test for significance among sample locations and dates. We used hoop traps to sample turtles at the same 6 locations per site throughout the summers of 2009, 2010, and early summer 2011. We checked traps daily and removed them from the water at the end of a trap period to prevent mortality. Information we recorded for captured turtles included sex, carapace length, carapace width, carapace height, plastron length, and mass. We marked each adult turtle with a unique code composed of holes drilled in marginal scales on the carapace (Dustman 2010, Gibbons 1990). We used turtle-capture data to analyze movement patterns, size–density relationships, and species diversity. We employed the Shannon diversity index and an ADONIS analysis of similarity to compare turtle Northeastern Naturalist Vol. 25, No. 3 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 423 diversity between study sites, followed by a Tukey’s HSD test to identify which sites differed significantly in diversity. Excluding juveniles, we compared turtle measurements across sites using MANOVA. We trapped Siren intermedia (Goin) (Western Lesser Siren, hereafter Siren), an SGCN within Kentucky, at each site in the spring, fall, and summer. We set, baited with sardines, and checked daily 6 modified trashcan traps (Luhring and Jennison 2008) at each site (n = 18). We opened and upended traps after each trap period to allow captured organisms to escape and prevent mortality. Snoutto- vent length, total length, and mass were measured for each captured Siren. At each site, we set and baited with dry dog food 24 minnow traps. We identified to the lowest taxon possible and released all other organisms (snakes, frogs, tadpoles, fish, and invertebrates) captured in the minnow and trashcan traps. All SGCN that were captured or casually observed during fish and herpetofaunal sampling were also recorded. We utilized a Song Meter1® (Wildlife Acoustics Inc., Maynard, MA) automated recording device (ARD) that recorded ambient sound for 2 min every hour to estimate relative densities of frog and toad species. One ARD was placed at each site and data were regularly downloaded. We identified species using the songs of breeding males. We estimated semi-quantitative densities using the North American Amphibian Monitoring Program (NAAMP) ranking system (Bridges and Dorcas 2000, Royle and Link 2005). We employed a Gower’s measure of dissimilarity to statistically measure the level of dissimilarity of both species richness and ordinal densities, and performed an ADONIS analysis to identify significant dissimilarities in the Gower’s results. Stable-isotope analysis We collected samples for stable isotope analysis from the wetland substrate on 24–28 May 2010 and 6–10 June 2011. We also collected samples from dominant primary producers (Phragmites, Duckweed, cattail, Immigrant Pond-lily, Zygnematales [green algae]), invertebrates (Belostoma [hemipterans], Dytiscus [predaceous diving beetles], Ranatra [water scorpions], Odonata [dragonflies and damselflies], Decapoda [crustaceans], Notonecta [backswimmers], and Gastropoda [molluscs]), turtles (Trachemys scripta elegans [Wied-Neuwied] [Red-eared Slider] Chelydra serpentina L. [Common Snapping Turtle]), frogs (Rana and Pseudacris), and fish (Lepomis [sunfish], Gambusia [mosquitofish], Noturus [madtoms], Esox [pike], and Aphredoderus [perch]). We placed on ice and stored samples in either Whirl-Paks® or scintillation vials until they could be transferred to a freezer. We dried samples at 55 °C for 72 h and then ground them to a fine powder using a ceramic mortar and pestle. Small organisms and small tissue-samples of the same species were combined to provide an adequate mass as a composite sample. We placed 0.05–3.00 mg of each sample in an 8 x 5 mm Elemental Microanalysis Ltd. (Devon, UK) tin capsule. The sample capsules were analyzed in a Finnigan Delta Plus XP mass spectrometer (Thermo Fisher Scientific, Waltham, MA). Northeastern Naturalist 424 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 Vol. 25, No. 3 Results Water chemistry Five of the 11 water-quality parameters varied significantly among sites. DOC, SRP, TP, and TN were significantly higher at the Phragmites-free reference site than at either Phragmites-dominated site (all F2,38 > 8.0, all P < 0.01). In contrast, conductivity was significantly lower at the reference site than the other 2 sites (F2,38 = 76.3, P < 0.01). The treated Phragmites site never differed significantly from the untreated Phragmites site. DOC, TP, turbidity, ORP, TN, and conductivity showed significant seasonal variation (F1,38 ≥ 4.7, P < 0.05), but only conductivity exhibited a significant interaction between site and month (Fig. 2). This interaction inversely correlated with discharge data at the 2 Phragmites-dominated sites, while conductivity at the Phragmites-free reference site remained stable throughout rain events and a drought (Fig.2). Nitrogen, temperature, pH, and DO did not differ significantly across treatments or over time (all P > 0.06). Fish Over the course of this study, we captured and recorded a total of 26 species of fish representing 14 families and 2525 individuals. Fifteen fish species were each represented by more than 15 total individuals captured. Of these well-represented species, 12 had similar capture numbers at the Phragmites-dominated sites, but were substantially different at the Phragmites-free reference (χ2 = 5.4, P = 0.02, df = 1; Fig. 3). The permutational MANOVA of the Jaccard similarity index Figure 2. Specific conductivity in microsiemens/cm over time at each of the 3 sites: treated Phragmites (diamonds), Phragmites-free (squares), and untreated Phragmites (triangles). Northeastern Naturalist Vol. 25, No. 3 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 425 Figure 3. Total number of individuals captured across 25 fish species per sample site. See Appendix 1 for authorities and common names of fish species. Northeastern Naturalist 426 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 Vol. 25, No. 3 showed significant dissimilarity between the 2 Phragmites-dominated sites and the reference site (both Phragmites sites F1,12 > 2.4, both P < 0.001), but not between the treated and untreated sites (F1,12 = 0.77, P = 0.75). Shannon’s diversity index showed no differences across sites, including when controlling for time (both F < 0.5, both P > 0.61). Of particular interest among the fish samples was the distribution of Erimyzon sucetta (Lake Chubsucker). The Lake Chubsucker is threatened in the state of Kentucky, is an SGCN, and was found at all 3 sampling locations. We captured a total of 477 Lake Chubsuckers during this study, the second-highest abundance of any species; the majority of these individuals (57%) were collected at the untreated Phragmites-dominated site. Turtles Over 2 field seasons, we trapped each site for 288 trap days, resulting in 661 turtles captured, consisting of 328 Common Snapping Turtles, 330 Red-eared Sliders, 2 Sternotherus oderatus (Latreille in Sonnini & Latreille) (Musk Turtle), and 1 Chrysemys picta margenata Agassiz (Midland Painted Turtle). The ADONIS analysis showed a significant difference in turtle biodiversity across sites (F2,106 = 3.2, P = 0.01). The subsequent Tukey’s test showed diversity was significantly lower at the Phragmites-free reference site than either of the 2 Phragmites-dominated sites (P < 0.02). The 2 Phragmites-dominated sites did not differ significantly (P > 0.7). MANOVA comparisons of overall body sizes showed significant differences among both Common Snapping Turtles and Red-eared Sliders. Turtles of both species were significantly larger at the Phragmites-free reference site than either of the 2 sites with Phragmites (both F > 3.4, both P < 0.01); the Phragmites-dominated treated site did not differ significantly from the untreated Phragmites site (both F > 1.6, both P > 0.1). Sirens and other SGCN We monitored trash-can traps at each site for ~654 days over the 2-y period for Siren. Trapping resulted in 7 Sirens captured within the Phragmites sites, only 1 of which was at the treated site. In addition to Sirens, we occasionally captured Nerodia rhombifer Hallowell (Diamondback Water Snake), Nerodia erythrogaster neglecta (Conant) (Copperbelly Water Snake), and Agkistrodon piscivorus leucostoma (Troost) (Western Cottonmouth) in these traps. We captured or observed Sirens and Western Cottonmouths only at the Phragmites-dominated sites; the 6 other recorded SGCN were observed at all 3 sites. Anurans The ARDs recorded for 10 months at each site, which provided over 360 h of sound for analysis. We identified 12 anuran species across the 3 sites including 3 SGCN: Hyla avivoca Viosca (Bird-voiced Tree Frog), Lithobates areolata circulosa (Rice and Davis) (Crawfish Frog), and the Lithobates sphenocephala Cope (Southern Leopard Frog). Gower’s measure of dissimilarity showed no trends, Northeastern Naturalist Vol. 25, No. 3 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 427 and the subsequent ADONIS analysis showed no significant differences in anuran biodiversity across sites (F2,21 = 0.54, P > 0.05). Stable-isotope analysis Only 3 organisms showed variation between sites in δ 15N (Fig. 4). One fish (Esox) and 1 invertebrate (Dytiscus) displayed an increase of 2.5‰ and 4‰ δ 15N, respectively, in the Phragmites-free reference site as compared to both Phragmites sites (Fig. 4). Another fish (Noturus) had a heavier δ 15N in both Phragmites sites than in the reference site by 2.5‰. The majority of the observed variation in stable isotopes was in the δ 13C signatures. Noturus exhibited enriched levels in its δ 13C signature at both Phragmites sites in comparison to the reference site’s more negative signature. Notonecta, Odonata, Ranatra, and Trachemys showed the opposite result, with heavier δ 13C levels at the reference site than both Phragmites sites. This shift led to reduced variation in δ 13C in the Phragmites-free reference site (most organisms between -32‰ to -24‰) when compared to both Phragmites (-35‰ to -25‰). We observed a difference in the δ13C signatures of Phragmites between the treated and untreated Phragmites sites. The treated site showed a lighter mean δ 13C value (-29‰) as compared to the untreated (-26‰) site. Discussion Overall, we detected no differences between the treated and untreated Phragmites- dominated sites, with exception to the δ 13C signature of the Phragmites itself. All statistically significant differences found were between the Phragmites-free site and the Phragmites-dominated sites. These dissimilarities are complex due to a broad range of dynamic environmental variables and should therefore be interpreted cautiously. Water chemistry All significant differences in the water-chemistry analysis paralleled the presence or absence of Phragmites, suggesting that this plant may be directly impacting nutrient cycling in Clear Creek. The land use surrounding all 3 sites is similar— historical agricultural activity and coal mining adjacent to every site (Davis 2011, Hill 1983, Neichter 1972). The differences across sites included the major nutrients involved in aquatic ecosystems. An increase in DOC can reflect the faunal community that the ecosystem can support. Sources of DOC include excretion of waste materials by animals, cell breakdown, and microbial decomposition (Lampert and Sommer 2007). An increase in DOC, as well as other more limiting nutrients, such as phosphorus and nitrogen, can indicate an increased trophic state of an aquatic habitat and even increased species diversity (Lampert and Sommer 2007). Our data contradict this trend, with the Phragmites-free reference site having higher nutrient availability and lower faunal diversity. Reactive phosphorus (SRP) is typically the most limiting nutrient in a wetland (Cole 1994). More available phosphorus can benefit algal communities at the base Northeastern Naturalist 428 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 Vol. 25, No. 3 Figure 4. Stableisotope analysis for each site. Nuphar was sampled at the Phragmites-free site in the absence of Phragmites. Northeastern Naturalist Vol. 25, No. 3 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 429 of the aquatic food web. Phragmites has higher net primary-productivity and plant biomass than the native plants it displaces (Ehrenfeld 2003), and thus could have sequestered much of the phosphorus at the Phragmites-dominated sites, limiting available phosphorus for other autotrophs. At the treated Phragmitesdominated site, we observed a decrease in living Phragmites shoots. If Phragmites management had any effect on nutrients such as SRP, it was not detected during this short-term study. However, nutrient cycling in a stream ecosystem has a longitudinal (downstream) element (Newbold 1992); unmanaged Phragmites existed upstream and may have impacted the influx of nutrients to the ar ea. Conductivity varied significantly over time at both Phragmites sites, while levels at the Phragmites-free site were much more stable. This finding suggests that the flow of minerals into Weir Creek was much more constant and stable than that entering Clear Creek. High mineral-content in watersheds is often a result of the erosion of rocks and soils (Golterman 1975). Although the historical coal mining in Hopkins County increased erosion in both streams, Clear Creek was more severely impacted than Weir Creek (Neichter 1972). This situation could explain the higher and more-variable conductivity levels found at the Clear Creek sites in the current study (Fig. 2). Additionally, Clear Creek is a higher-order stream with more exposure to potential drainage sources across a larger watershed. Drainage from strip mines is often characterized by a low pH and increased heavy metals (Robb and Robinson 1995). Phragmites has been shown to increase sediment deposition (Rooth and Stevenson 2000), which may have affected the transport of conductive dissolved solids in Clear Creek. The pH levels during this study were relatively neutral for all 3 sites. Considering the historical documentation of very acidic conditions in Clear Creek (Niechter 1972) and research findings that indicate that Phragmites has the capacity to increase the pH of runoff from mines by adding alkalinity through gas exchange in the substrate (Robb and Robinson 1995), it is plausible that Phragmites has stabilized the pH at CCWMA. Other water-chemistry parameters showed seasonal effects, but no variation across sites. Alternatively, the differences seen in conductivity across sites could be sitespecific characteristics, independent of the presence of Phragmites. For example, drainage from strip mines that increased conductivity may have degraded Clear Creek, making it difficult for native plants to survive, thereby reducing competition for Phragmites and making it easier for this species to invade CCWMA. Peverly et al. (1995) documented Phragmites thriving and expanding in high concentrations of metal leachate from an adjacent strip mine, with its roots acting as filters to absorb metals while the rhizosphere released oxygen to form metal precipitates. If this is the case at our sites, Clear Creek may be more difficult to restore because Phragmites is providing an ecological buffer for strip-mine drainage. Even after the invasive plant is removed, native species may have difficulty reestablishing in a degraded habitat. Further monitoring would provide insight into this question. Faunal population effects Our results do not support the hypothesis that the CCWMA fish community was affected by Phragmites management. The lack of difference between the treated and Northeastern Naturalist 430 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 Vol. 25, No. 3 untreated Phragmites sites could be attributed to the structurally similar habitat that still exists at these 2 locations. Although the treatment of Phragmites resulted in the elimination of green plant tissue in adult individuals, we observed that throughout this study the dead plant stems remained standing and the amount of open-water habitat remained essentially the same as before the herbicide application. The short duration of the present study may not reflect long-term changes in the fish community that could result from continued Phragmites management. For example, the fish community at the Phragmites-free reference site was made up of a significantly different assemblage than those at either of the Phragmites locations. This finding supports the hypothesis that the native plant community at the Phragmites-free site provides habitat for a different fish community than those sites dominated by Phragmites. As the remaining Phragmites stems decompose and active management of Phragmites presumably continues, open-water habitat and native species may yet increase at the CCWMA, with the caveats described above. Phragmites presence, but not management, also affected turtle species. We captured more large turtles of the 2 most-dominant species at the Phragmites-free site than at either of the 2 Phragmites sites. There were large numbers of native Immigrant Pond-lily and Duckweed at the non-Phragmites site, both of which are commonly consumed by Red-eared Sliders (Gibbons 1990). Although Common Snapping Turtles are much more carnivorous than Sliders, both species are omnivorous. Our observation that large turtles of both species were more common at the Phragmites-free reference site indicates habitat selection according to age and size. Invasion by an exotic plant such as Phragmites lowers plant diversity (Warren et al. 2001), and Phragmites may have outcompeted more palatable plants at our 2 Phragmites sites. Our fish-community results revealed no significant differences in overall prey availability for turtles among sites. Thus, all 3 areas had adequate food for younger and smaller turtles that typically have a more protein-based diet, suggesting that larger turtles may have been more concentrated in the non-Phragmites site because of the vegetation differences. The Phragmites-free reference site had lower turtle diversity than either of the 2 Phragmites sites, primarily because of low species evenness. The 2 Phragmites sites did not differ significantly from each other. The ratio of Common Snapping Turtles to Red-eared Sliders was much higher in the Phragmites-free site, as opposed to either the treated or untreated Phragmites sites (~4:1 at Phragmites-free versus ~1:2 for both Phragmites sites). These results suggest Phragmites has species-specific effects on turtle distribution and population size. It is possible that turtle species differ in habitat selection, either preferring or avoiding Phragmitesdominant habitats. Alternatively, the species that are more aggressive, such as Common Snapping Turtles, may outcompete those that are less aggressive, such as Red-eared Sliders, for access to the high-quality resources at the Phragmites-free locality. Further research will be required to evaluate these disparate hypotheses. We found no significant difference in anuran diversity based on either species richness or ordinal densities using the NAAMP ranking. This result supports the null hypothesis that anurans are not affected by the presence of Phragmites and may Northeastern Naturalist Vol. 25, No. 3 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 431 therefore show little reaction to Phragmites management. Although Phragmites does not grow in deep water, there was standing water throughout the stands of Phragmites in Clear Creek during our study. Our results suggest that Phragmites provides adequate habitat for the frog species of this area. This result is congruent with the findings of Mazerolle et al. (2013). We confirmed the presence of Sirens at both the treated and untreated Phragmites sites. Although Sirens were not captured at the Phragmites-free site, we found a Farancia abacura (Holbrook) (Western Mud Snake), a Siren-specialist predator (Petranka 1998) and an SGCN, nearby, suggesting both species are present in that area. Of the 10 aquatic reptile and amphibian SGCN that are listed as confirmed in Hopkins County, KY (Kentucky Department of Fish and Wildlife Resources 2010), we observed 8 at or directly adjacent to the study sites. The presence of these species suggests that Clear Creek has high-quality habitat for herpetofaunal SGCN, despite the invasion of Phragmites. Some of these species, particularly the Siren, were rare and might benefit from the successful management of this invasive plant. Other SGCN at our sites may have been affected by anthropogenic disturbance. For example, the Phragmites-free reference site is heavily trafficked by the landowner, and the treated Phragmites site is open to public use, and we observed relatively few Western Cottonmouth snakes at both sites. Based on personal communications with individuals in the local community, the Western Cottonmouth is likely affected by human-induced mortality at these 2 sites. Frequent observations of this species at the untreated Phragmites site suggest the Phragmites is not responsible for lower numbers of Western Cottonmouths at the treated and reference site. Ecosystem effects The stable-isotope results provide insight into the food-web dynamics at each of the 3 study sites. The stable-isotope signatures of the organisms at the treated and untreated Phragmites sites resemble one another overall. However, the signatures of the organisms at the Phragmites-free site show a slightly different food web, particularly in δ13C levels (Fig. 4). Emergent plants that grow in littoral zones in aquatic systems tend to have a less-negative carbon signature than do algae and organisms that feed on algae (Post 2002). The lack of Phragmites at the reference site may have allowed herbivores to feed on the more available native macrophytes within this community, such as Immigrant Pond-lily and Duckweed. In a similar study, Gratton and Denno (2006) found the food web at their control treatment (Spartina [cordgrass]-dominated wetland) to be trophically linked to the dominant macrophyte, while Phragmites-invaded areas had isotopic signatures linking the food web to algal resources. This result corresponds to the greater range of δ13C values at both Phragmites-dominated sites. For example, 3 invertebrates (backswimmers, damselflies/dragonflies, and water scoprions) and 1 turtle (Red-eared Slider) showed a heavier δ 13C signature at both Phragmites sites as compared to the reference site. If the reference site has more palatable macrophytes, then consumers will not need to rely so much on algal resources. However, the isotopic signature of backswimmers contradicts this hypothesis Northeastern Naturalist 432 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 Vol. 25, No. 3 with a lighter δ13C signature at the Phragmites-free site, suggesting that it consumes more algae or more organisms that eat algae there. One potential effect of management efforts can be observed in the δ13C signature of Phragmites. The live Phragmites analyzed from the treated site exhibited a more negative signature than that from the untreated Phragmites site by approximately 4‰. Despite no significant differences in water-sample nutrient levels between the Phragmites sites, the δ13C signatures potentially reflect a difference in the microbial community in the substrate. Wang et al. (2017) found a negative correlation between microbial biomass and the δ13C levels of the associated plant community. It is possible that either an increase in decomposing Phragmites mulch due to herbicide application or direct effects of the herbicide itself might have impacted the microbial community and thereby may have subsequently altered nutrient availability to the remaining live Phragmites plants. The nutritional state of plants can impact isotopic signatures (O’Leary 1981). Neither site had atypical δ13C signatures for Phragmites nor high variation among sample replicates (~0.7‰ for both), indicating the higher signature at the treated site is unlikely due to temporal variation. Two predators (predaceous diving beetles and pike) existed at an increased trophic level (mean δ15N value) at the Phragmites-free reference site as compared to either Phragmites site, suggesting the reference site supported a longer food chain. In a bottom-up–controlled ecosystem, a longer food chain is supported by increased biomass in the preceding trophic levels (Campbell and Reece 2002). An increase in biomass on all trophic levels means there would be an increase in faunal excretory waste and cellular decomposition, which corresponds to increases in DOC (Lampert and Sommer 2007) as found at our Phragmites-free reference site. Again, this finding was not supported by the nitrogen signature of madtoms, which showed an enriched δ15N in both Phragmites sites (+~3‰) as compared to the reference site. Neither the carbon nor the nitrogen signature of madtoms corresponds to the more positive trends seen in other organisms in the Phragmites-free site, but this is not entirely unexpected. Other studies regarding Phragmites’ effect on fish have shown mixed results (Able and Hagan 2000, Davis 2011, Fell et al. 2003), and it is not surprising that species react differently to resource alterations. Conclusion This study demonstrates the complex and variable nature of invasive species treatment in a wetland ecosystem. Plant invasions generally affect all levels and processes within an ecosystem; thus, multi-disciplinary and cross-taxa monitoring research is important for understanding the ecological implications of invasive species management (Blossey 1999). This interdisciplinary approach is reflected in the management goals and monitoring strategies at Clear Creek WMA. Replicating this study in conjunction with continued management efforts could tease out nuances in how different populations of flora and fauna react to invasive plant ma nagement. The ability of Phragmites to quickly dominate marsh-plant communities (Roman et al. 1984) makes possible the alteration of wetlands on an ecosystem level. Our results illustrate the complexities of wetland dynamics and fluxes, some of which originate from the bottom of the food web and manifest in top predators. The Northeastern Naturalist Vol. 25, No. 3 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 433 successful management of Phragmites and any subsequent effects on the wetland community may not be apparent without long-term monitoring in conjunction with continued management. Successful habitat restoration in this context can be defined as an attempt to re-establish communities to a state that is both physically and functionally similar to the pre-invasion habitat (Palmer et al. 1997). Complementing management efforts with thorough biological monitoring allows for a deeper understanding of how this invasive plant and its management can affect wetlands, and thus provides guidance for future restoration efforts. Acknowledgments This research was funded by a State Wildlife Grant from the Kentucky Department of Fish and Wildlife Resources. We thank the Hancock Biological Station for logistical support and B. Kik for access to the Clear Creek Wildlife Management Area and for lodging. We are grateful to E. Dustman, Dr. G. Kipphut, T. Johnson, T. Flynt, K. Johnston, G. Harris, J. Benson, S. Roach, T. Anderson, and others for their valuable guidance and assistance. Finally, we thank 2 anonymous reviewers for their helpful suggestions on an earlier version of this manuscript. Literature Cited Able, K.W., and S.M. Hagan. 2000. Effects of Common Reed (Phragmites australis) invasion on marsh-surface macrofauna: Response of fishes and decapod crustaceans. Estuaries 23:633–646. Able, K.W., S.M. Hagan, and S.A. Brown. 2003. Mechanisms of marsh-habitat alteration due to Phragmites: Response of young-of-the-year Mummichog (Fundulus heteroclitus) to treatment for Phragmites removal. Estuaries 26:484–494. Ailstock, M.S., C.M. Norman, and P.J. Bushman. 2001. Common Reed, Phragmites australis: Control and effects upon biodiversity in freshwater nontidal wetlands. Restoration Ecology 9:49–59. Armstrong, J., and W. Armstrong 1988. Phragmites australis: A preliminary study of soiloxidizing sites and internal gas-transport pathways. New Phytologist 108:373−382. Blossey, B. 1999. Before, during, and after: The need for long-term monitoring in invasive plant-species management. Biological Invasions 1:301−311. Bridges, A., and M. Dorcas. 2000. Temporal variation in anuran calling behavior: Implications for surveys and monitoring programs. Copeia 2:587−592. Campbell, N.A., and J.B. Reece. 2002. Biology, 6th Edition. Benjamin Cummings, New York, NY. 1175 pp. Cole, G.A. 1994. Textbook of Limnology, 4th Edition. Waveland Press, Long Grove, IL. 412 pp. Croteau, E., A. Krzton-Presson, and H. Whiteman. 2013. Identification of non-native Phragmites at Clear Creek Wildlife Management Area, Kentucky, using genetic techniques. Castanea 78:323−328. Davis, B.D. 2011. Fish community response to Phragmites treatment at Clear Creek Wildlife Management Area, Hopkins County, Kentucky. M.Sc. Thesis. Murray State University, Murray, KY. Dustman, E. 2010. Effects of roads on turtle life history. M.Sc. Thesis. Murray State University, Murray, KY. Northeastern Naturalist 434 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 Vol. 25, No. 3 Ehrenfeld, J.G. 2003. Effects of exotic plant invasions on soil-nutrient cycling processes. Ecosystems 6:503−523. Fell, P.E., S.R. Warren, J.K. Light, R.L. Rawson Jr., and S.M. Fairley. 2003. Comparison of fish and macroinvertebrate use of Typha angustifolia, Phragmites australis, and treated Phragmites marshes along the Lower Connecticut River. Estuaries 26:534−551. Gibbons, J.W. 1990. Life History and Ecology of the Slider Turtle. Smithsonian Institution Press, Washington, DC. 368 pp. Golterman, H.L. 1975. Physiological Limnology: An Approach to the Physiology of Lake Ecosystems. Elsevier Scientific Publishing Company, New York, NY. 489 pp. Gratton, C., and R. Denno. 2006. Arthropod food-web restoration following removal of an invasive wetland plant. Ecological Applications 16:622−631. Haslam, S.M. 1971. Community regulation in Phragmites communis Trin: II. Mixed stands. Journal of Ecology 59:75−88. Hill, P.L. 1983. Wetland-stream ecosystems of the Western Kentucky Coalfield: Environmental disturbance and the shaping of aquatic community structure. Ph.D. Dissertation. University of Louisville, Louisville, KY. Kentucky Department of Fish and Wildlife Resources (KDFWR). 2010. Kentucky’s Comprehensive Wildlife Conservation Strategy. Frankfort, KY. Lampert, W., and U. Sommer. 2007. Limnoecology: The Ecology of Lakes and Streams. Oxford University Press, Oxford, UK. 489 pp. Leuthart, C.A. 1975. Reclamation of orphan strip-mined land in southern Illinois and western Kentucky. Ph.D. Dissertation. University of Louisville, Louisville, KY. Luhring, T., and C. Jennison. 2008. A new stratified aquatic sampling technique for aquatic vertebrates. Journal of Freshwater Ecology 23:445−450. Mahala, M. 2008. Kentucky aquatic nuisance-species management plan. Kentucky Department of Fish and Wildlife Resources Technical Publication. Frankfort, KY. Mazerolle, M.J., A. Perez, and J. Brisson. 2014. Common Reed (Phragmites australis) invasion and amphibian distribution in freshwater wetlands. Wetlands Ecology and Management 22:325−340. Meyer, S.W. 2003. Comparative use of Phragmites australis and other habitats by birds, amphibians, and small mammals at Long Point, Ontario. M.Sc. Thesis. University of Western Ontario, London, ON, Canada. Meyerson, L.A., K. Saltonstall, L. Windham, E. Kiviat, and S. Findlay. 2000. A comparison of Phragmites australis in freshwater and brackish-marsh environments in North America. Wetland Ecology and Management 8:89−103. Neichter, P.L. 1972. Water-quality survey of the Weirs Creek levee system, Kentucky. M.Sc. Thesis. University of Louisville, Louisville, KY. Newbold, J.D. 1992. Cycles and spirals of nutrients. Pp. 379−408, In P. Calow (Ed.). Rivers Handbook. Volume 1. Blackwell Science, Cambridge, MA. 544 pp. O’Leary, M.H. 1981. Carbon-isotope fractionation in plants. Phytochemistry 20:553−567. Palmer, M.A., R.F. Ambrose, and N.L. Poff. 1997. Ecological theory and community restoration ecology. Restoration Ecology 5:291−300. Paxton, B.J. 2007. Potential impact of Common Reed expansion on threatened high-marsh bird communities on the seaside: Breeding-bird surveys of selected high-marsh patches. Center for Conservation Biology Technical Report Series 07−03. College of William and Mary, Williamsburg, VA. Petranka, J.W. 1998. Salamanders of the United States and Canada. Smithson Institution Press, Washington, DC. 592 pp. Northeastern Naturalist Vol. 25, No. 3 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 435 Peverly, J.H., J.M. Surface, and T. Wang. 1995. Growth and trace-metal absorption by Phragmites australis in wetlands constructed for landfill leachate treatment. Ecological Engineering 5:21−35. Post, D.M. 2002. Using stable isotopes to estimate trophic position: Models, methods, and assumptions. Ecology 83:703−718. Robb, G.A., and J.D. Robinson. 1995. Acid drainage from mines. Geographical Journal 161:47−54. Rogalski, M.A., and D.K. Skelly. 2012. Positive effects of nonnative invasive Phragmites australis on larval bullfrogs. PLoS One 7:e44420. Roman, C., W. Niering, and R. Warren. 1984. Salt-marsh vegetation change due to tidal restriction. Environmental Management 8:141−150. Rooth, J.E., and J.C. Stevenson. 2000. Sediment deposition patterns in Phragmites australis communities: Implications for coastal areas threatened by rising sea-level. Wetland Ecology Management 8:173−183. Royle, J., and W. Link. 2005. A general class of multinomial-mixture models for anuran calling-survey data. Ecological Society of America 86:2505−2512. Saltonstall, K. 2002. Cryptic invasion by a non-native genotype of the Common Reed, Phragmites australis, into North America. Proceedings National Academy of Science 99(4):2445−2449. Simberloff, D. 1996. Impacts of introduced species in the United States. Consequences 2:13−22. Wang, B., Y. Wu, and D. Chen. 2017. Stable isotopes of carbon and nitrogen help to predict the belowground communities at a regional scale. Scientific Repo rts 7:72–76. Warren, R.S., P.E. Fell, J.L. Grimsby, E.L. Buck, G.C. Rilling, and R.A. Fertik. 2001. Rates, patterns, and impacts of Phragmites australis expansion and effects of experimental Phragmites control on vegetation, macroinvertebrates, and fish within tidelands of the Lower Connecticut River. Estuaries 24:90−107. Zedler, J.B., and S. Kercher. 2004. Causes and consequences of invasive plants in wetlands: Opportunities, opportunists, and outcomes. Critical Reviews in Plant Sciences 23:431−452. Northeastern Naturalist 436 A. Krzton-Presson, B. Davis, K. Raper, K. Hitz, C. Mecklin, and H. Whiteman 2018 Vol. 25, No. 3 Appendix 1. The following are the scientific names, authorities, and common names of fish species captured during this study at CCWMA as seen in Figure 3. Scientific name Authority Common name Amia calva L. Bowfin Lepisosteus oculatus Winchell Spotted Gar Lepisosteus osseus (L.) Long-nose Gar Esox americanus Gmelin Grass Pickerel Dorosoma cepedianum (Lesueur) Eastern Gizzard Shad Erimyzon sucetta (Lacepède) Lake Chubsucker Cyprinus carpio L. Common Carp Notemigonus crysoleucas (Mitchill) Golden Shiner Labidesthes sicculus (Cope) Brook Silverside Ameiurus melas (Rafinesque) Black Bullhead Ameiurus natalis (Lesueur) Yellow Bullhead Noturus gyrinus (Mitchell) Tadpole Madtom Fundulus olivaceus (Storer) Blackspotted Topminnow Gambusia affinis (Baird & Girard) Mosquito Fish Aphredoderus sayanus (Gilliams) Pirate Perch Centrarchus macropterus (Lacepède) Flier Lepomis cyanellus Rafinesque Green Sunfish Lepomis gulosus (Cuvier) Warmouth Lepomis macrochirus Rafinesque Bluegill Lepomis megalotis (Rafinesque) Longear Sunfish Micropterus salmoides (Lacepède) Largemouth Bass Pomoxis annularis Rafinesque White Crappie Pomoxis nigromaculatus (Lesueur) Black Crappie Elassoma zonatum Jordan Banded Pygmy Sunfish Etheostoma squamiceps Jordan Spottail Darter