Southeastern Naturalist C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 584 2017 SOUTHEASTERN NATURALIST 16(4):584–602 The Effect of Beaver Ponds on Water Quality in Rural Coastal Plain Streams Christopher W. Bason1,3,*, Daniel E. Kroes2, and Mark M. Brinson3,† Abstract - We compared water-quality effects of 13 beaver ponds on adjacent free-flowing control reaches in the Coastal Plain of rural North Carolina. We measured concentrations of nitrate, ammonium, soluble reactive phosphorus (SRP), and suspended sediment (SS) upstream and downstream of paired ponds and control reaches. Nitrate and SS concentrations decreased, ammonium concentrations increased, and SRP concentrations were unaffected downstream of the ponds and relative to the control reaches. The pond effect on nitrate concentration was a reduction of 112 ± 55 μg-N/L (19%) compared to a control-reach–influenced reduction of 28 ± 17 μg-N/L. The pond effect on ammonium concentration was an increase of 9.47 ± 10.9 μg-N/L (59%) compared to the control-reach–influenced reduction of 1.49 ± 1.37 μg-N/L. The pond effect on SS concentration was a decrease of 3.41 ± 1.68 mg/L (40%) compared to a control-reach–influenced increase of 0.56 ± 0.27 mg/L. Ponds on lower-order streams reduced nitrate concentrations by greater amounts compared to those in higher-order streams. Older ponds reduced SS concentrations by greater amounts compared to younger ponds. The findings of this study indicate that beaver ponds provide water-quality benefits to rural Coastal Plain streams by reducing concentrations of nitrate and suspended sediment. Introduction Land-use practices and associated processes in agriculturally influenced rural landscapes affect surface-water quality by contributing high levels of nutrients and sediments to streams and estuaries. In the Coastal Plain of the southeastern US, these practices and processes include fertilizer application, concentrated animal operations, soil erosion, atmospheric deposition of nitrogen, use of domestic septic systems, and extensive stream channelization (Correll et al. 1992, Duda 1982, Glasgow and Burkholder 2000, Jordan et al. 1997, Mallin et al. 1997, Omernik 1976, Paerl 1995, Yarbro et al. 1984). Nutrient and sediment inputs can contribute to ecosystem degradation and inability of waters to support their designated uses (Carpenter et al. 1998, USEPA 2002). Advances in water-quality management practices are being incorporated on agricultural lands (Correll 1997, Osmond et al. 2002), but expression of eutrophic conditions is expected to worsen in most south Atlantic estuaries of the US (Bricker et al. 2007). Although inputs of both nitrogen and phosphorous are high, estuaries in North Carolina are typically nitrogen limited (Harned and Davenport 1990, Harned et al. 1995). Phosphorous concentrations are 1Delaware Center for the Inland Bays, 39375 Inlet Road, Rehoboth Beach, DE, 19971. 2US Geological Survey, Lower Mississippi-Gulf Water Science Center, Baton Rouge, LA, 70817. 3Department of Biology, East Carolina University, Greenville, NC 27858. †Deceased. *Corresponding author - chrisbason@inlandbays.org. Manuscript Editor: Robert Krenz Southeastern Naturalist 585 C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 naturally high in the waters of North Carolina due to geologic sources (Spruill et al. 1998), and they are often augmented by inputs from croplands and effluent from sewage-treatment facilities. Castor canadensis (Kuhl) (Beaver) populations have reestablished throughout much of the southeastern US (Arner and Hepp 1989, Butler 1991, Naiman 1988). Reintroductions of Beaver to North Carolina began in 1939 (Woodward and Hazel 1991). By 1983, populations were present in 80 of 100 counties and affected a minimum of 35,858 ha of bottomland (Woodward et al. 1985). Population growth to such nuisance proportions was driven by abundant food resources, few natural predators, and a weak fur market. Beaver construct dams across stream channels and floodplains (Hair et al. 1978, Pullen 1971). The ponds that form behind the dams decrease water velocities, expand water–sediment interactions, and open forest canopies, thus increasing aquatic primary production, facilitating sedimentation, and altering biogeochemical transformations (Naiman et al. 1988). Beaver ponds have been shown to affect water quality by reducing concentrations of nitrate and suspended sediments in various regions of North America (Cirmo and Driscoll 1993, Correll et al. 2000, Devito et al. 1989, Klotz 2010, Kroes and Bason 2015, Maret et al. 1987, Margolis et al. 2001, Naiman et al. 1988). However, we are unaware of published research documenting the effects of beaver ponds on stream-water quality in watersheds with intensive agricultural land-use typical of the southeaster n Coastal Plain. The purpose of this study was to explore variability in effects among different beaver ponds on water quality in rural streams of the North Carolina Coastal Plain. We used a paired study design to compare the differences in concentrations of nitrate, ammonium, soluble reactive phosphorus (SRP), and suspended sediments (SS) between paired upstream and downstream sampling locations in 13 Beaverimpounded stream reaches and 13 upstream, free-flowing control reaches during fall and winter, 2002–2003. We focused on the landscape scale to allow for inference from a group of varied ponds, rather than concentrating our efforts on the detailed biogeochemical dynamics of a single site using a mass balance approach, as undertaken in the Maryland Coastal Plain by Correll et al. (2000). This approach forgoes the resolution necessary to determine influences of season and streamflow on the effects of 1 pond in exchange for the capacity to explore potentially influential spatial and temporal factors, such as drainage-basin size and age, across multiple ponds. Study Area The inner Coastal Plain of North Carolina is characterized by low elevations and a muted relief that results in a poorly drained landscape (Hammond 1964). Streams in that region typically have low gradients (less than 1% slope) that contribute to more frequent (~5 times/year), and longer-lasting overbank flow than in higher-gradient settings (Sweet and Geratz 2003). Floodplains without Beaver activity typically support bottomland hardwood forests (Shafale and Weakley 1990). Stream-water quality in the study area is typical of southeastern Coastal Plain blackwater streams Southeastern Naturalist C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 586 (Smock and Gilinsky 1992), with moderately acidic to near neutral pH (min–max = 4.94–6.74, median = 6.08) and generally low specific conductance (min–max = 63–225 μS/cm, median = 98 μS/cm), but often with high levels of nutrients depending on watershed land-use and condition of the channel and riparian area (Yarbro et al. 1984). The climate is warm temperate with an average annual temperature of 16.5 ºC and average annual precipitation of 126 cm (US Climate Data 2013). The study area received 38 cm of precipitation during the study period (5 cm above average) and the temperature was 5.5 ºC (1.1 ºC) below average (NOAA 2017a, b). High summertime evapotranspiration can cause 1st- and 2nd-order streams (Strahler 1957) to become intermittent and large portions of beaver ponds to become dry. Low wintertime temperatures can cause some ponds to periodically freeze, although we did not observe this during our study . Land use in the study area is dominated by agriculture and silviculture, and confined animal feeding operations (CAFOs) are also common. These areas were largely converted from deciduous forest with subsequent stream channelization, establishment of field and roadside ditches, and development of roads and residential areas (Cashin et al. 1992, Rheinhardt et al. 1999). Drainage efforts began as early as the 1700s in North Carolina, and more than 40% of production cropland in North Carolina requires drainage (Thomas et al. 1995). Field ditches typically connect to 1st- to 4th-order channelized streams. Channelization practices ended in the 1970s following challenges by environmental groups (Coffey 1982). These practices have affected most headwater streams and associated riparian areas on the Coastal Plain. Such impairment contributes to high nutrient and sediment loads in receiving streams, and ultimately, estuaries (Mallin et al. 1993). Beaver activity in the study area was extensive. Beaver dams across low-gradient stream channels and floodplains often created serial stair-step ponds (Kroes and Bason 2015) that extended for several kilometers. Field assessments and study of orthophotography led us to the conservative estimate that over one quarter of 2nd- to 5th-order stream length within the study area was impounded by Beaver. Associated flood damage to timber and roads was substantial (Butler 1991, Woodward et al. 1985) and led to concerted population control ef forts. Methods Site selection We studied orthophotography from 1998 to identify 60 potential site pairs within an area of ~2500 km2 (Fig. 1). We selected 13 pairs of sites based on several criteria: (1) site pairs occurred on 2nd- to 4th-order streams with a vegetated buffer of at least 10 m in width on each side of the channel, (2) control reaches were free-flowing and had no visible evidence of recent, sustained ponding from Beaver activity, and (3) pond and control-reach pairs had comparable length, gradient, and channelization status. We avoided road crossings, hydrologic inputs such as field ditches, and sites in close proximity to CAFOs. At one site (#35), we discovered a small hydrologic input with a nitrogen-rich discharge after the study began and Southeastern Naturalist 587 C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 eliminated this site from nitrogen analyses. In all cases, the study dam was the first dam occurring downstream of the control reach. Site factors During the study period, dams were typically maintained so that pond-water levels were at the top of the dam (full) or within centimeters of full. Some of these Figure 1. (a) Study area on the inner Coastal Plain (ICP) of North Carolina with site locations numbered. Inset shows approximate location relative to the Piedmont (PM) to the west and the outer Coastal Plain (OCP) to the east. (b) Sampling locations at a typical study site; PD = pond downstream, CD = channel downstream, CU = channel upstream. Southeastern Naturalist C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 588 ponds were previously observed to experience summer drawdowns as precipitation and groundwater inputs were reduced. All study sites had floodplains composed of loam soils including loam, fine sandy loam, and sandy loam of the Bibb, Johnston, Muckalee, and Rains complexes (NRCS 2016). We evaluated site pairs for stream order, drainage-basin area, gradient, pond area, dam height, dam-maintenance condition, channelization status, CAFO presence in the watershed, and pond age. We estimated stream order and drainage-basin area upstream from each study dam from digital US Geological Survey (USGS) topographic maps (1:24,000) in ArcView GIS v3.2 (ESRI 1999). We recorded dam height as the mean of 3 measurements taken from the top of the effective dam to the downstream floodplain surface, not as the height of the dam in the stream channel. We classified dams as unmaintained if we observed no evidence of upkeep, but they were still intact and impounding water. We used GPS-derived point data of site perimeters in ArcView to calculate pond-surface area at dam-full level. Longitudinal floodplain gradient (% slope) was derived from pond length at dam-full condition and dam height. We calculated an estimate of damfull pond volume as one half of the product of mean pond length, width, and dam height. We estimated pond age as the number of years since evidence of ponding was first detected in annual aerial photography plus 1 year to compensate for tree-canopy thinning required for ponding to be detected. We determined presence or absence of CAFOs from aerial photography. We considered sites channelized if we observed spoil banks from channel excavation. Water-quality sampling and analysis We used a paired-study design to compare differences in nutrient and sediment concentrations between upstream and downstream sampling locations in ponds and in upstream, free-flowing control reaches. We established 3 sampling locations within the hydrologically dominant channel at each site pair: pond downstream (PD) was the compilation of major outflows ~2 m downstream of the study dam, control downstream (CD) was just upstream of the backwater influence of the pond, and control upstream (CU) was at a distance upstream from CD similar to the distance from PD to CD. Renewed dam construction during the study raised water levels at some sites and required that we move CD slightly upst ream (less than 15 m). We collected grab samples on 4 to 6 dates per site pair from late October 2002 to February 2003 while water flowed in streams and over dams. We did not collect samples during warmer months because of highly intermittent flow. Analysis of USGS National Water Quality Assessment Program samples taken near these sites from January 2002 through December 2003 in the Neuse River at Kinston, NC (station # 02089500), showed samples from October 2002 from February 2003 to have a mean filtered nitrate + nitrite concentration of 505 μg/l compared to 438 μg/l for all 2002– 2003. Filtered samples had a mean SRP concentration of 23 μg/l, compared to 36 μg/l for all 2002–2003 sampling dates (USGS 2015). These concentrations indicate that nitrogen concentrations sampled during our study were slightly elevated compared to the rest of the year, whereas orthophosphate concentrations were seasonally low. We sampled sites at a range of wadeable water levels to incorporate variation in water residence times and to better approximate the range of biogeochemical conditions in Southeastern Naturalist 589 C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 stream systems during the study period. We stored our samples on ice before filtering them in the laboratory within 24 h of collection. We measured SS concentrations by filtering samples through pre-ashed, pre-weighed glass microfiber filters (1.5-μm effective-pore size) within 24 h of collection, drying at 105 °C for 24 h, and reweighing (APHA et al. 1995). Samples used in determining dissolved inorganic nutrient concentrations were filtered as above and stored frozen until analysis. We determined soluble reactive phosphorus with the ascorbic acid method (APHA et al. 1995). We modified Solorzano’s (1969) phenolhypochlorite method to determine ammonium: sample and reagent volumes were halved and color developed for at least 2 h before measuring absorbance. We employed the cadmium-reduction method (APHA et al. 1995) to determine total oxidized nitrogen (NO3 -+NO2 --N), which we report as nitrate. Data analysis We determined reach effects on sediment and nutrient concentrations as differences between concentrations at downstream (CD) and upstream (CU) locations for the control reach and, for the pond reach, as differences between concentrations at CD and locations below the dam (PD): Control-reach Influence on Concentration = (CD - CU) Pond-reach Influence on Concentration = (PD - CD). This paired-sampling structure allowed determination of a single response variable to estimate the effect of ponds adjusted for their control reaches: Effect of pond on concentration adjusted for control = (PD - CD) - (CD - CU) Proportional change (+, -) in concentration = (Effect/CD)(100) We averaged influences for each water-quality variable by site. We employed paired t-tests to determine significant differences between influence of control reaches (CD - CU) and influence of ponds (PD - CD). Relationships between site factors and the effect of ponds on concentrations as adjusted for their controls were tested using Pearson’s correlation (r). We log10-transformed SS data to meet test assumptions of normality. We also log10-transformed pond age to reduce the undue influence of an outlier (16 y) on relationships. We examined associations among site characteristics with Pearson’s correlation coefficients or Kendall’s nonparametric coefficient of concordance (τ), as appropriate for the data distributions. We employed t-tests (one-tailed, unequal variance) to compare differences between nutrient concentrations of control streams in watersheds with and without CAFOs and simple linear regressions to examine relationships between incoming concentrations and reductions. We conducted all analyses in SPSS v. 11 (SPSS Inc, Chicago, IL 2001). Exceptionally high nitrate concentrations prevented inclusion of site 20 in parametric statistical analyses of nitrate; all other water-quality variables at site 20 were statistically analyzed. We excluded site 35 from nitrate and ammonium analyses altogether due to nitrogen-rich discharge from an agricultural tributary that we Southeastern Naturalist C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 590 discovered after the study began and whose influence confounded upstream–downstream relationships (2–5 times other study reach concentrations). Results Site factors Drainage-basin areas varied from ~1.67 km2 to 40.4 km2 (Table 1). Drainage areas were inversely correlated to stream gradient, i.e., ponds with smaller drainage areas had steeper gradients (r = -0.71, P = 0.004), and smaller pond surface-to-volume ratios (r = 0.61, P = 0.014). Beaver dam height averaged 0.47 m, and 9 dams (69%) were actively maintained during the study. We were unable to determine the age of site 43 from aerial photographs and excluded it from age-related comparisons. The mean pond age of the other ponds was 4.9 years; the single 16-year-old pond was an outlier. Two sites (20 and 52) were channelized and another site (56) was just downstream from a channelized reach. CAFOs were present in 42% of study-site watersheds. Nutrient concentrations did not differ significantly between streams in watersheds with and without CAFOs (nitrate: t = 1.23, P = 0.12; ammonium: t = 0.00 P = 0.36; SRP: t = 0.6, P = 0.28). Nitrate Mean stream-nitrate concentrations spanned an order of magnitude from 151 μg N/L at site 28 to 5114 μgN/L at site 20, and were significantly higher in channelized streams (τ = 0.58, P = 0.009). Beaver ponds conveyed lower concentrations of nitrate than controls (t = -3.7, P = 0.002, omitting site 35). Mean nitrate concentrations showed greater reductions through the ponds (141 ± 48 μg N/L), than through the control reaches (28 ± 17μgN/L; Fig. 2a). The effect of 11 of 12 ponds was to Table 1. Characteristics of 13 Beaver ponds arranged by drainage-basin area. Dam maint. = whether the dam was maintained. Chan. = was there channelization. CAFO indicates the presence or absence of confined animal feeding operations within the watershed of the site (A = abandoned). n = 4–6, depending on site. Drainage Pond Dam Pond Stream area Gradient area height volume Dam Site order (km2) (%) (ha) (m) (m3) maint. Chan CAFO Age (y) 60 3 1.67 0.44 0.37 0.62 1150 Y N N 2 17 2 3.29 0.19 1.95 0.68 6350 Y N Y 7 35 2 3.94 0.52 0.17 0.42 392 N N Y 6 20 2 4.20 0.41 0.19 0.41 327 Y Y Y 7 52 3 4.54 0.20 0.60 0.28 735 Y Y N 1 32 2 4.71 0.27 1.00 0.93 4100 Y N N 5 50 2 8.47 0.21 1.97 0.55 6220 Y N N 6 56 3 18.67 0.13 0.97 0.27 1020 N N A 2 39 3 19.81 0.12 3.01 0.41 5410 N N Y 2 24 3 24.53 0.10 4.84 0.46 11,300 Y N N 3 36 4 28.93 0.10 3.53 0.46 7030 Y N Y 16 43 4 31.34 0.10 3.00 0.38 4890 Y N Y - 28 4 40.35 0.14 0.56 0.19 475 N N N 2 Southeastern Naturalist 591 C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 decrease average concentrations. Reductions varied widely from 3 μgN/L to 753 μgN/L (Fig. 3a). Site 43 was the only pond that exhibited an increasing effect on nitrate concentrations. High variation within site types may be attributed to small sample size and inclusion of at least 1 high-discharge sample per site. Ponds that had smaller drainage areas had higher stream gradients; both of these characteristics were related to greater reductions in nitrate concentrations (r = -0.75, P = 0.007 and r = 0.72, P = 0.012, respectively; Table 2). Greater concentrations of nitrate entering control streams correlated to greater reductions by ponds on concentrations (r = 0.48, P < 0.001). Figure 2. Distributions of control reach influence (CD–CU) and pond reach influence (PD–CD) on concentrations of (a) nitrate, (b) ammonium, (c) soluble reactive phosphorus (SRP), and (d) suspended sediment (SS). Dotted lines are means, solid lines are medians, the bottom and top of each box are 25th and 75th percentiles, whiskers are the 10th and 90th percentiles, and solid dots are outliers. Site 20 was omitted from (a) due to exceptionally high nitrate concentrations. Table 2. Pearson correlation coefficients between site factors and the effect of impoundments on water-quality variables adjusted for control reaches. Dam maint. = dam maintenance; chan. = channelization.. *P < 0.05, **P < 0.01 and † indicates data were log 10 transformed Stream Drainage Surface Dam Damfull Dam Factor n order area Gradient area height volume maint. Chan. Age† Nitrate 11 0.43 0.72* -0.75** 0.49 -0.20 0.33 -0.18 -0.33 0.51 Ammonium 12 -0.12 -0.29 0.29 -0.41 -0.16 -0.42 0.41 0.42 -0.12 SRP 13 -0.38 -0.08 0.48 -0.37 -0.10 -0.34 -0.29 0.13 0.10 SS† 13 -0.08 -0.04 -0.29 -0.12 -0.03 -0.06 -0.23 0.13 -0.71** Southeastern Naturalist C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 592 Ammonium Mean ammonium concentrations spanned an order of magnitude (10.9–209 μg N/L) and were significantly greater in streams associated with channelization (τ = 0.62, P = 0.006). Beaver ponds conveyed significantly higher concentrations of ammonium than control reaches (t = 3.007, P = 0.012; n = 12). On average, Figure 3. Effect of beaver ponds on concentrations of (a) nitrate and ammonium, (b) soluble reactive phosphorus (SRP), and (c) suspended sediment (SS) as adjusted for their free-flowing control reaches, by site. Bars are mean (+ 1 SE). Sites arranged by increasing drainage area from left to right. Data were omitted for site 35 in (a) as explained in Methods. Southeastern Naturalist 593 C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 ammonium concentrations increased through the ponds (7.97 ± 11.7 μgN/L) and decreased through the control reaches (1.49 ± 1.37 μgN/L; Fig. 2b). The effect of 9 of 12 ponds was to convey higher concentrations of ammonium (Fig. 3a). We found no significant relationships between site factors and the effect of ponds (Table 2). Higher concentrations of ammonium entering control streams was weakly correlated to a greater effect of ponds on concentrations ( r = 0.28, P = 0.03). Soluble reactive phosphorus The variation of mean SRP concentrations in streams (3.0–46 μg P/L) was narrower than for ammonium or nitrate, and the effect of ponds on SRP concentration was not significant (t = -0.31, P = 0.76; n = 13). Ponds and control reaches exhibited similar influences (1.15 μg P/L vs -0.82 μg P/L; Fig. 2c). Nine of 13 ponds had a reducing effect on SRP concentrations, but about half of the sample-set differences in concentrations fluctuated around zero (Fig. 3b). No significant relationships were found between site factors and effect (Table 2). Higher concentrations of SRP entering control streams correlated with greater pond effects on outgoing concentrations (R2 = 0.68, P < 0.001). Suspended sediment The variation of mean SS concentrations entering ponds was over an order of magnitude (1.3–19 mg/L). Ponds were better at trapping SS than the control reaches (log10 transformed data, t = -4.75, P < 0.001; n = 13). Beaver ponds reduced SS concentrations by 2.84 ± 1.48 mg/L. The control reaches acted as a source of sediment (0.56 ± 0.27 mg/L; Fig. 2d). Eleven of 13 ponds had an effect of decreasing SS concentrations, but 3 sites fluctuated around 0 (mg/L) difference. Mean reductions ranged from 0.1 mg/L to 15 mg/L (Fig. 3c). The effect of ponds on SS concentrations showed a significant negative relationship with pond age (log10 transformed data: r = -0.71, P = 0.009; Table 2). The relationship between concentrations of SS entering the control reach with the effect of the ponds was strong (r = 0.99, P less than 0.001) with high SS concentrations resulting in greater sedimen t trapping. Discussion The study was comprised of paired, adjacent reaches with and without effects of beaver ponds to control for variation due to factors other than pond effect. This restriction made it difficult to locate site pairs that lacked confounding hydrologic inputs. After the study began, we discovered that 1 site (35) had a nitrogen-rich input; thus, we eliminated it from nitrogen analyses. We could have employed a mass-balance approach that measured all flows and nutrient loadings of a single pond to overcome this problem, but that would not have met our study objectives. Our objectives were to explore the variation in effects of beaver ponds on water quality among a range of ponds and to examine the potential influence of factors, such as watershed size and stream channelization, on those effects (Table 1). We undertook fall and winter sampling because we assumed that stream-water nutrient and suspended solids concentrations would be higher due to reduced vegetative and Southeastern Naturalist C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 594 microbial assimilation and entrapment. In addition, extended lapses in flow from Beaver impoundments due to high rates of evapotranspiration precluded sampling from many sites during summer. We observed high variability in water-quality parameter concentrations among the sites. Ponds conveyed significantly lower nitrate concentrations (-19%) and significantly higher ammonium concentrations (59%) relative to free-flowing stream reaches (Fig. 2). A number of nitrogen transformations within beaver ponds may be responsible for the effects we observed. Denitrification has been suggested as the primary means of nitrogen removal for many riverine and riparian wetlands (Brinson et al. 1984, Ettema et al. 1999, Klotz 2010, Willems et al. 1997), and was likely responsible for the majority of nitrate concentration decreases in this study, although we did not measure this directly. Beaver ponds provide excellent conditions for denitrification because of the extended duration of water contact with pond sediments rich in organic matter and the elevated nitrate concentrations of agriculturally influenced stream flows entering ponds. In a study of 3 Rhode Island beaver ponds, Lazar et al. (2015) found that denitrification dominated nitrogen transformations and was negatively correlated with the dissolved-oxygen saturation of pond water. Rates of assimilation of nitrogen into the organic nitrogen pool of these ponds were lower than denitrification followed by rates of ammonium generation; neither rate correlated with pond characteristics. The variation in nitrate is of primary concern for water quality within beaver ponds. Increases in ammonium concentrations below ponds may have been due to the liberation of ammonium from anoxic pond sediments, mineralization, or from dissimilatory nitrate reduction to ammonium. The fact that we did not measure oxygen levels in the pond sediments limits our ability to speculate on the importance of these potential transformations. We found that 2 sites conveyed exceptional concentrations of nitrate or ammonium. Pond 43 conveyed nitrate concentrations 9.3% higher than its control reach, and samples from site 52 showed remarkably high ammonium concentrations on several sampling dates. It is not clear why these sites had elevated concentrations of nitrogen relative to other sites, and we could not determine a specific source. It is possible that CAFOs or undetected conveyances of nitrogen-rich water such as points of focused groundwater discharge in the watershed of site 43 could have contributed, although this condition was not observed in all other sites with CAFOs in the watershed. The scale of the study did not allow determination of hydrographs for the sites or how the effects of ponds on water quality varied between rising and falling stream stage. Although we sampled each site at least once during elevated flow, we did not determine if flow was rising or falling, which could influence nutrient and sediment concentrations in streams (Lowrance and Leonard 1988, McDiffet et al. 1989). It is notable that the collective ponds in this study reduced SRP concentrations and SS concentrations to a greater degree at the highest flows compared to the lowest flows (Wilcoxon signed-rank test: n = 13; SRP: z = -2.27, P = 0.023; SS: z = -1.85, P = 0.064). There was no significant effect on SRP concentrations in this study, and only 3 site-pairs demonstrated a notable effect on concentrations: 2 site-pairs Southeastern Naturalist 595 C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 reduced concentrations and 1 increased concentrations (Fig. 3b). Other studies have found inconsistent effects of beaver ponds on phosphorus with some ponds augmenting and others reducing concentrations of SRP (Correll et al. 2000, Devito and Dillon 1993, Klotz 1998, Maret et al. 1987). Higher SRP concentrations discharged from ponds in New York were related to anoxic conditions during periods of ice cover, a relatively rare phenomenon in the southeastern Coastal Plain (C.W. Bason, pers. observ.). Beaver ponds in this study acted as sediment traps, decreasing concentrations of SS in waters entering ponds by 40% (3.41 mg/L) relative to the control-reach influence (Fig. 2). The mechanism by which this decrease is accomplished is the reduction in the velocity of water entering ponds. Beaver ponds in Maryland, Virginia, North Carolina, and Wyoming caused decreases in stream SS concentrations for at least part of the year (Correll et al. 2000, Kroes and Bason 2015, Maret et al. 1987). The Maryland stream conveyed total SS concentrations of 67.6 mg/L prior to creation of a beaver pond and 49.3 mg/L after construction (Correll et al. 2000). Decreases in SS concentrations in Virginia and North Carolina ponds varied from -1.5 mg/L to -17 mg/L (Kroes and Bason 2015). Despite the regional differences in sediment characteristics and concentrations, the SS proportional concentration reductions were similar between these studies. During high flows, deposition of sediment-associated nutrients in beaver ponds could be an important mechanism for reducing total nutrient concentrations, as suggested by Maret et al. (1987). Likewise, Correll et al. (2000) suggested that organic phosphorus and organic nitrogen concentrations were reduced via sediment deposition. Relative to larger ponds, the faster water-transit times in small ponds may act to decrease total SS concentration by deposition of sands and coarse silts while simultaneously increasing suspended fine silt and clay concentrations via bioturbation. Interestingly, the 3 ponds exhibiting increases in SRP in our study (sites 20, 28, and 35) also had the smallest pond volumes and likely had higher bioturbation by volume. Although we did not determine particle sizes of incoming and outgoing SS, bioturbation of very fine sediment could have resulted in increases of SRP because phosphorus is disproportionately adsorbed to clay. Concentrations of SRP released from beaver ponds can also be influenced by the differences in phosphorus concentrations between pond sediments and the overlying water column (Klotz 1998). Sediments either act as sources or sinks of P to the water column depending on the concentration gradient at the sediment–water interface. We did not calculate sorption capacity in this study, but assumed that pond sediments had higher sorption capacity than free flowing reaches due to their finer grain size. Reductions in SRP concentrations attributable to ponds in this study tended to be largest during the highest stream-flows and when SRP concentrations upstream of the pond were high. This result suggests that the pond sediments had high sorption capacity to reduce the elevated in solution P entering the ponds. Sorbed P would likely be released during subsequent periods of low flow when we would expect incoming SRP concentrations to be lower and pond sediments to be in a more reduced state. Southeastern Naturalist C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 596 Channel gradient and drainage area were the most influential site characteristics on the reduction of nitrate concentrations in ponds. Both upstream position and steeper gradients co-occur with higher nitrate concentrations because lower-order streams on the NC Coastal Plain are surrounded by, and are in closer proximity to, agricultural lands (Spruill et al. 1998). Where nitrate concentrations are higher, the potential for greater removal of nitrogen via denitrification is also higher (Hanson et al. 1994, Spieles and Mitsch 2000, Willems et al. 1997). Increasing drainage area can also decrease water-retention time, reducing the capacity of wetlands to improve water quality (Woltemade 2000). Estimates of retention time for a subset of this study’s ponds indicated that sites with small drainage areas had average waterresidence times over 4-fold longer than sites with large drainage areas. Our results suggest that, like the smaller streams on which they are created (see Peterson et al. 2001), beaver ponds on smaller streams play a greater role in controlling nitrogen concentrations in waters travelling to downstream estuaries. Stream channelization alters biogeochemical functions by increasing and sustaining groundwater-derived baseflow and decoupling channels from their floodplain wetlands (Yarbro et al. 1984). Relative to an unchannelized stream, a channelized stream typically has less heterogeneity that results from straightening and removal of snags, low levels of particulate carbon in bed sediments, and shorter hydraulic retention time. The net effect is lower retention of nitrate, total organic nitrogen, particulate phosphorus, and sediments (Noe and Hupp 2005, Yarbro et al. 1984). Two of 3 channelized sites in this study (sites 20 and 56) were also located in close proximity to active or abandoned CAFOs. Higher nitrate concentrations entering our pond sites were not reduced to a greater degree than in ponds with lower incoming concentrations. However, the greatest single reduction in nitrate occurred with high incoming concentrations at 1 site. Beaver were found to hydraulically reconnect floodplain wetlands with channelized streams in sites of this study and those of Kroes and Bason (2015). In this study, recoupling of these floodplains in the presence of a beaver pond had the ef fect of reducing nitrate concentrations. Older ponds in this study reduced suspended sediment (SS) concentrations to a greater degree than younger ponds. As ponds age, trees of floodplain forests die, large, downed wood accumulates, and closed tree-canopies are replaced by dense herbaceous vegetation. These structural changes reduce water velocities and create complex microtopography in channels and on pond bottoms, thus facilitating sedimentation in streams and floodplains (Harmon et al. 1986, Sand-Jensen 1998). The process of sediment deposition changes and may be reduced as old ponds fill in with sediments (Kroes and Bason 2015). However, we did not observe this change in our study area perhaps due to generally low SS concentrations of Coastal Plain streams and the recolonizing Beaver population. The lack of a significant relationship between pond age and effects on nitrate suggests that the mere presence of a beaver pond was important for nitrate reductions. Inferences about the effects of pond age on study variables must be made in context with the recolonizing population dynamics of Beaver in this region. Our study design required free-flowing reaches upstream of ponds; thus, site selection Southeastern Naturalist 597 C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 was biased towards the most upstream point of Beaver recolonization within a stream corridor, and consequently, younger ponds. Our results capture conditions at 1 point in time during what is expected to be a cycle of colonization and recolonization of Beaver populations as they respond to resource availability and geomorphology over years to centuries (Johnston 2015, Johnston and Naiman 1990). The effects of Beaver on water quality may change dynamically over space and time as populations and ponds age and are vacated due to declining resources and trapping. However, it seems important that studies over time and regions of North America have reported reductions in nitrate and sediment concentrations due to beaver ponds. In more-developed landscapes such as the agricultural coastal plain, human control of Beaver populations is likely to have a greater influence on the water-quality effects of Beaver than the less-developed areas where research on population dynamics has previously focused. In addition to improving water quality, beaver ponds also diversify plant and animal communities (Bason 2004, Feldmann 1995, Metts et al. 2001, Snodgrass and Meffe 1998), provide valuable habitat (Menzel et al. 2000, Otis and Edwards 1999), and offer renewable harvest of Beaver and waterfowl (Hair et al. 1978). In contrast to these ecological services, they create economic losses by flooding timber, pastures, roads, and agricultural fields, and can facilitate invasive species establishment (Perkins and Wilson 2005). Beaver populations are often managed in response to individual nuisance complaints, and in some instances, are excluded from potential habitat entirely. This form of control can reinforce the concept of Beaver as problem animals (Halley and Rosell 2002) instead of as a keystone species and ecosystem engineers in lotic ecosystems and riparian areas. Trade-offs between ecosystem services and negative economic impacts should be considered when managing Beaver populations. However, our results suggest that best management of Beaver populations may be an option deserving of more consideration as a water-quality improvement measure. Conclusions Our results indicate that beaver ponds may ameliorate some agricultural effects on water quality of rural Coastal Plain streams by reducing nitrate concentrations (19%) and sediment concentrations (40%). Ponds on lower-order streams reduced nitrate concentrations by greater amounts. Older ponds reduced sediment concentrations by greater amounts. In order to utilize the water-quality effects of beaver ponds, resource agencies and landowners could consider (1) allowing expansion of Beaver populations into impaired waters and, where appropriate, headwater streams, (2) measured control of newly established populations to prevent rapid population growth that can result in negative attitudes towards Beaver, (3) costshare programs for trapping and water-level management structures and, 4) where Beaver must be removed, leaving unmaintained dams to improve wa ter quality. Nutrient runoff models like SPARROW (USGS 2016) could aid in identification of watersheds that would most benefit from greater abundance of beaver ponds. The basin-wide effects of beaver ponds have not been evaluated at a scale appropriate Southeastern Naturalist C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 598 for quantifying a reduction in nutrient loading on estuarine water quality. However, our conservative estimate that over ¼ of the 2nd- to 5th-order stream length within the study area was impounded by Beaver suggests that non-point-source nitrogen and sediment loading to North Carolina’s eutrophic and typically nitrogen-limited estuaries was lower due to the presence of beaver ponds. Acknowledgments Unfortunately, Mark Brinson passed away prior to finalization of this paper. Wherever he may be, we wish him the best. His friendship and guidance were essential to accomplishing this research. We thank Chantal Bouchard, Bob Christian, Dennis Whigham, Paul Vos, and David Beamer for their support, guidance, and field assistance. Although research described in this article has been funded in part by the US Environmental Protection Agency through cooperative agreement R-82868401 via a contract to East Carolina University from Pennsylvania State University, it has not been subjected to the Agency’s required peer and policy review, and therefore, does not necessarily reflect the views of the Agency; no official endorsement should be inferred. Any use of trade, product, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the US Government. Literature Cited American Public Health Association (APHA), American Water Works Association, and Water Environment Federation. 1995. Standard methods for the examination of water and wastewater. Washington, DC. 1100 pp. Arner, D.H., and G.R. Hepp. 1989. Beaver pond wetlands: A southern perspective. Pp. 117–128, In L.M. Smith, R.L. Pederson, and R.M. Kaminski (Eds.). Habitat Management for Migrating and Wintering Waterfowl in North America. Texas Tech University Press Lubbock, TX. 574 pp. Bason, C.W. 2004. Effects of beaver impoundments on stream-water quality and floodplain vegetation in the inner coastal plain of North Carolina. M.Sc. Thesis. East Carolina University, Greenville, NC. 147 pp. Bricker, S., B. Longstaff, W. Dennison, A. Jones, K. Boicourt, C. Wicks, and J. Woerner. 2007. Effects of Nutrient Enrichment in the Nation’s Estuaries: A Decade of Change. National Centers for Coastal Ocean Science, Silver Spring, MD. 32 pp. Brinson, M.M., H.D. Bradshaw, and E.S. Kane. 1984. Nutrient assimilative capacity of an alluvial floodplain swamp. Journal of Applied Ecology 21:1041–1057. Butler, D.R. 1991. The reintroduction of the Beaver to the South. Southeastern Geographer 31:39–43. Carpenter, S.R., N.F. Caraco, D.L. Correll, R.W. Howarth, A.N. Sharpley, and V.H. Smith. 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecological Applications 8:559–569. Cashin, G.E., J.R. Dorney, and C.J. Richardson. 1992. Wetland alteration trends on the North Carolina Coastal Plain. Wetlands 12:63–71. Cirmo, C.P., and C.T. Driscoll. 1993. Beaver pond biogeochemistry: Acid-neutralizing capacity-generation in a headwater wetland. Wetlands 13:277–292. Coffey, A. 1982. Stream improvement: The Chicod Creek episode. Journal of Soil and Water Conservation 37:80–82. Southeastern Naturalist 599 C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 Correll, D.L. 1997. Buffer zones and water-quality protection: General principles. Pp. 7–20, In N. Haycock, T. Burt, K. Goulding, and G. Pinay (Eds.). Buffer Zones: Their Processes and Potential in Water Protection. The Proceedings of the International Conference on Buffer Zones, September 1996. Quest Environmental Harpenden, Hertfordshire, UK. 326 pp. Correll, D.L., T.E. Jordan, and D.E. Weller. 1992. Nutrient flux in a landscape: Effects of coastal land- use and terrestrial community mosaic on nutrient transport to coastal waters. Estuaries 15:431–442. Correll, D.L., T.E. Jordan, and D.E. Weller. 2000. Beaver-pond biogeochemical effects in the Maryland Coastal Plain. Biogeochemistry 49:217–239. Devito, K.J., and P.J. Dillon. 1993. Importance of runoff and winter anoxia to the P and N dynamics of a beaver pond. Canadian Journal of Fisheries and Aquatic Sciences 50:2222–2232. Devito, K.J., P.J. Dillon, and B.D. Lazerte. 1989. Phosphorus and nitrogen retention in 5 Precambrian-shield wetlands. Biogeochemistry 8:185–204. Duda, A.M. 1982. Municipal point-source and agricultural non-point–source contributions to coastal eutrophication. Water Resources Bulletin 18:397–407. Environmental Systems research Institute (ESRI) 1999. ArcView GIS v3.2. Redlands, CA. Ettema, C.H., R. Lowrance, and D.C. Coleman. 1999. Riparian-soil response to surface nitrogen input: Temporal changes in denitrification, labile and microbial C and N pools, and bacterial and fungal respiration. Soil Biology and Biochemi stry 31:1609–1624. Feldmann, A. 1995. The effects of Beaver (Castor canadensis) impoundment on plant diversity and community composition in the Coastal Plain of South Carolina. M.Sc. Thesis. University of Georgia, Athens, GA. Glasgow, H.B., Jr., and J.M. Burkholder. 2000. Water-quality trends and management implications from a 5-year study of a eutrophic estuary. Ecological Applications 10:1024–1046. Hair, J.D., G.T. Hepp, L.M. Luckett, K.P. Reese, and D.K. Woodward. 1978. Beaver pond ecosystems and their relationships to multi-use natural resource management. Pp. 80–92, In R.R. Johnson, and J.F. McCormick (Eds.). Strategy for Protection and Management of Floodplain Wetlands and Other Riparian Ecosystems. US Forest Service, General Technical Report. WO-12. 217 pp. Washington, DC. Halley, D.J., and F. Rosell. 2002. The Beaver’s reconquest of Eurasia: Status, population development, and management of a conservation success. Mammal Review 32:153–178. Hammond, E.H. 1964. Classes of land-surface form in the 48 states, USA. Annals of the Association of American Geographers 54(1):map supplement. Hanson, G.C., P.M. Groffman, and A.J. Gold. 1994. Denitrification in riparian wetlands receiving high and low groundwater nitrate inputs. Journal of Environmental Quality 23:917–922. Harmon, M.E., J.F. Franklin, F.J. Swanson, P. Sollins, S.V. Gregory, J.D. Lattin, N.H. Anderson, S.P. Cline, N.G. Aumen, J.R. Sedell, G.W. Lienkaemper, K. Cromack, and K.W. Cummins. 1986. Ecology of coarse woody debris in temperate ecosystems. Advances in Ecological Research 15:133–302. Harned, D.A., and M.S. Davenport. 1990. Water-quality trends and basin activities and characteristics for the Albemarle–Pamlico estuarine system, North Carolina and Virginia. US Geological Survey Open-File Report 90-398. Reston, VA. 164 pp. Harned, D.A., G. McMahon, T.B. Spruill, and M.D. Woodside. 1995. Water-quality assessment of the Albemarle–Pamlico Drainage Basin, North Carolina and Virginia: Characterization of suspended sediment, nutrients, and pesticides. US Geological Survey Open-File Report 95-191. Reston, VA. 131 pp. Southeastern Naturalist C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 600 Johnston, C.A. 2015. Fate of 150-year-old Beaver ponds in the Laurentian Great Lakes Region. Wetlands 35:1013–1019. Johnston, C.A., and R.J. Naiman. 1990. Aquatic-patch creation in relation to Beaver population trends. Ecology 71:1617–1621. Jordan, T.E., D.L. Correll, and D.E. Weller. 1997. Effects of agriculture on discharges of nutrients from coastal plain watersheds of Chesapeake Bay. Journal of Environmental Quality 26:836–848. Klotz, R.L. 1998. Influence of Beaver ponds on the phosphorus concentration of stream water. Canadian Journal of Fisheries and Aquatic Sciences 55:1228–1235. Klotz, R.L. 2010. Reduction of high nitrate concentrations in a central New York state stream impounded by Beaver. Northeastern Naturalist 17:349–356. Kroes, D.E., and C.W. Bason. 2015. Sediment-trapping by Beaver ponds in streams of the Mid-Atlantic Piedmont and Coastal Plain, USA. Southeastern Natu ralist 14:577–595. Lazar, J.G., K. Addy, A.J. Gold, P.M. Groffman, R.A. McKinney, and D.Q. Kellogg. 2015. Beaver ponds: Resurgent nitrogen sinks for rural watersheds in the northeastern United States. Journal of Environmental Quality 44(5):1684–93. Lowrance, R., and R. A. Leonard. 1988. Streamflow-nutrient dynamics on Coastal Plain watersheds. Journal of Environmental Quality 17:734–740. Mallin, M.A., H.W. Paerl, J. Rudek, and P.W. Bates. 1993. Regulation of estuarine primary- production by watershed rainfall and river flow. Marine Ecology Progress Series 93:199–203. Mallin, M.R., J.M. Burkholder, M.R. McIver, G.C. Shank, H.B. Glasgow Jr., B.W. Touchette, and J. Springer. 1997. Comparative effects of poultry- and swine-waste lagoon spills on the quality of receiving streamwaters. Journal of Environmental Quality 26:1622–1631. Maret, T.J., M. Parker, and T.E. Fannin. 1987. The effect of Beaver ponds on the nonpointsource water-quality of a stream in southwestern Wyoming. Water Research 21:263–268. Margolis, B.E., M.S. Castro, and R.L. Raesly. 2001. The impact of Beaver impoundments on the water chemistry of 2 Appalachian streams. Canadian Journal of Fisheries and Aquatic Sciences 58:2271–2283. McDiffet, W.F., A.W. Beidler, T.F. Dominick, and K.D. McCrea. 1989. Nutrient concentration– stream-discharge relationships during storm events in a 1st-order stream. Hydrobiologia 179:97–102. Menzel, M.A., C.T. Carter, W.M. Ford, and B.R. Chapman. 2000. Tree-roost characteristics of subadult and female adult Evening Bats (Nycticeius humeralis) in the upper Coastal Plain of South Carolina. American Midland Naturalist 145:112–119. Metts, B.S., J.D. Lanham, and K.R. Russell. 2001. Evaluation of herpetofaunal communities on upland streams and Beaver-impounded streams in the upper piedmont of South Carolina. American Midland Naturalist 145:54–65. Naiman, R.J., C.A. Johnston, and J.C. Kelly. 1988. Alteration of North American streams by Beaver. Bioscience 38:753–762. Natural Resource Conservation Service (NRCS). 2016. Web soil survey. Available online at: http://websoilsurvey.nrcs.usda.gov/app/WebSoilSurvey.aspx. Accessed 14 October 2016. National Oceanic and Atmospheric Administration (NOAA). 2017a. Climate at a glance— US time series: Precipitation. National Centers for Environmental Information. Available online at: http://www.ncdc.noaa.gov/cag/. Accessed 7 February 2017. NOAA. 2017b. Climate at a glance—US time series: Aveage temperature. National Centers for Environmental Information. . Available online at: http://www.ncdc.noaa.gov/cag/. Accessed 7 February 2017. Southeastern Naturalist 601 C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 Noe, G.B., and C.R. Hupp. 2005. Carbon, nitrogen, and phosphorus accumulations in floodplains of Atlantic Coastal Plain rivers, USA. Ecological Applications 15:1178–1190. Omernik, J.M. 1976. The influence of land use on stream-nutrient levels. Ecological research series. US Environmental Protection Agency, Office of Research, Corvallis, OR. 106 pp. Osmond, D.L., J.W. Gilliam, and R.O. Evans. 2002. Riparian buffers and controlled drainage to reduce agricultural nonpoint-source pollution. North Carolina State University, Raleigh, NC. 57 pp. Otis, D.L., and N.T. Edwards. 1999. Avian communities and habitat relationships in South Carolina Piedmont Beaver ponds. The American Midland Naturalist 141(1):158–171. Paerl, H.W. 1995. Coastal eutrophication in relation to atmospheric nitrogen deposition: Current perspectives. Ophelia 41:237–259. Perkins, T.E., and M.V. Wilson. 2005. The impacts of Phalaris arundinacea (Reed Canarygrass) invasion on wetland-plant richness in the Oregon Coast Range, USA depend on Beavers. Biological Conservation 124:291–295. Peterson, B.J., W.M. Wollheim, P.J. Mulholland, J.R. Webster, J.L. Meyer, J.L. Tank, E. Marti, W.B. Bowden, H.M. Valett, A.E. Hershey, W.H. McDowell, W.K. Dodds, S.K. Hamilton, S. Gregory, and D.D. Morrall. 2001. Control of nitrogen export from watersheds by headwater streams. Science 292:86–90. Pullen, T.M. 1971. Some effects of Beaver, Castor canadensis, and beaver-pond management on the ecology and utilization of fish populations along warm-water streams in Georgia and South Carolina. Ph.D. Dissertation. University of Georgia, Athens, GA. 84 pp. Rheinhardt, R.D., M.C. Rheinhardt, M.M. Brinson, and K.E. Faser Jr. 1999. Application of reference data for assessing and restoring headwater ecosystems. Restoration Ecology 7:241–251. Sand-Jensen, K. 1998. Influence of submerged macrophytes in lowland streams. Freshwater Biology 39:663–679. Shafale, M.P., and A.S. Weakley. 1990. Classification of the natural communities of North Carolina. 3rd approximation. North Carolina Natural Heritage Program, Division of Parks and Recreation, NC Department of Environmental Health and Natural Resources, Raleigh, NC. 325 pp. Smock, L.A., and E. Gilinsky. 1992. Coastal-plain blackwater streams. Pp. 271–313, In C.T. Hackney, M. Adams, and W. Martin (Eds.). Biodiversity of the Southeastern United States. John Wiley and Sons, Inc. New York, NY. 779 pp. Snodgrass, J.W., and G.W. Meffe. 1998. Influence of Beavers on stream-fish assemblages: Effects of pond age and watershed position. Ecology 79:928–942. Solorzano, L. 1969. Determination of ammonia in natural waters by the phenolhypocholrite method. Limnology and Oceanography 14:799–801. Spieles, D.J., and W.J. Mitsch. 2000. The effects of season and hydrologic and chemical loading on nitrate retention in constructed wetlands: A comparison of low- and highnutrient riverine systems. Ecological Engineering 14:77–91. Spruill, T.B., D.A. Harned, P.M. Ruhl, J.L. Eimers, G. McMahon, K.E. Smith, D.R. Galeone, and M.D. Woodside. 1998. Water-quality assessment of the Albemarle–Pamlico drainage basin, North Carolina, and Virginia, 1992–95. United States Geological Survey, Circular 1157, Raleigh, NC. 36 pp. Strahler, A.N. 1957. Quantitative analysis of watershed geomorphology. American Geophysical Union Transactions 38:913–920. Southeastern Naturalist C.W. Bason, D.E. Kroes, and M.M. Brinson 2017 Vol. 16, No. 4 602 Sweet, W.V., and J.W. Geratz. 2003. Bankfull hydraulic geometry relationships and recurrence intervals for North Carolina’s Coastal Plain. Journal of the American Water Resources Association 39:861–871. Thomas, D.L., C.D. Perry, R.O. Evans, F.T. Izuno, K.C. Stone, and J. Wendell Gilliam. 1995. Agricultural drainage effects on water quality in southeastern US. Journal of Irrigation and Drainage Engineering 121:277–282. US Climate Data. 2013. Average annual temperature and precipitation. Available online at http://www.usclimatedata.com/climate.php?location=USNC0281. Accessed 20 November 2013. US Environmental Protection Agency (USEPA). 2002. 2000 National water-quality inventory. Office of Water, Washington, DC. 207 pp. US Geological Survey (USGS). 2015. Water-quality samples for the nation, USGS 02089500 Neuse River at Kinston, NC. Available online at http://nwis.waterdata.usgs. gov/usa/nwis/qwdata. Accessed 27 August 2015. USGS. 2016. SPARROW surface water-quality modeling. Available online at http://water. usgs.gov/nawqa/sparrow/. Accessed 18 May 2016. Willems, H.P.L., M.D. Rotelli, D.F. Berry, E.P. Smith, R.B. Reneau Jr., and S. Mostaghim. 1997. Nitrate removal in riparian wetland soils: Effects of flow rate, temperature, nitrate concentration, and soil depth. Water Research 31:841–849. Woltemade, C.J. 2000. Ability of restored wetlands to reduce nitrogen and phosphorus concentrations in agricultural drainage water. Journal of Soil and Water Conservation 55:303–308. Woodward, D.K., and R.B. Hazel. 1991. Beavers in North Carolina: Ecology, utilization, and management. North Carolina Cooperative Extension Service, R aleigh, NC. 20 pp. Woodward, D.K., R.B. Hazel, and B.P. Gaffney. 1985. Economic and environmental impacts of Beavers in North Carolina. Pp. 89–96, In P.T. Bromley (Ed.). Proceedings of the 2nd eastern wildlife-damage control conference. Raleigh, NC. 281 pp . Yarbro, L.A., E.J. Kuenzler, P.J. Mulholland, and R.P. Sniffen. 1984. Effects of stream channelization on exports of nitrogen and phosphorus from North Carolina Coastal Plain watersheds. Environmental Management 8:151–160.