Southeastern Naturalist
577
D.E. Kroes and C.W. Bason
22001155 SOUTHEASTERN NATURALIST 1V4o(3l.) :1547,7 N–5o9. 53
Sediment-trapping by Beaver Ponds in Streams of the
Mid-Atlantic Piedmont and Coastal Plain, USA
Daniel E. Kroes1, 2,* and Christopher W. Bason3
Abstract - The effect of beaver ponds on sediment deposition is undocumented in the
Piedmont and Coastal Plain of Virginia and North Carolina. We used 3 methods to examine
sedimentation: 1) depth-integrated base-flow sampling, 2) repeat channel-surveys,
and 3) sediment-accumulation pads. During base flow, Piedmont ponds exported sediment
and Coastal Plain ponds had little or no effect on downstream suspended-sediment
concentration. Most ponds accumulated sediment within the channel until dam breaching.
Ponds inundating the floodplain trapped more sediment. Ponds of varying configuration
trapped sediment differently. Mean floodplain accretion rates in these beaver ponds
(2002-2003: 20 mm/yr 2003-2005: 15 mm/yr) greatly exceeded the mean deposition rate
of similar unimpounded streams in these areas. Intact Piedmont ponds trapped 11 m3/yr on
the floodplain and 77 m3/yr in the channel. Intact Coastal Plain ponds trapped 107 m3/yr
on the floodplain and 8 m3/yr in the channel.
Introduction
In the US, many animals that function as ecosystem engineers have been reduced
to functional extinction. One such ecosystem engineer, Castor canadensis
(Kuhl) (Beaver) was trapped and hunted to local extinction, or to such low numbers
that it was no longer found in most low-gradient streams of the eastern and
central US (Johnston and Chance 1974). Eradication of Beaver populations has led
to widespread beaver-dam failures resulting in systemic stream-change (Polvi and
Wohl 2012, Walter and Merritts 2008). Streams that were once composed of a series
of ponds (Morgan 1868) became free-flowing and exported large volumes of stored
sediment. On some streams, a portion of this sediment was trapped in mill ponds,
many of which have failed or are failing because of disrepair (Merritts et al. 2004).
Estimates of pre-colonial Beaver densities range from 2.2 to 74 Beaver/km2
(Bailey 1927, Hodgden and Hunt 1955), depending on habitat quality and stream
density. These population estimates could equate to <1 to >10 beaver ponds/km2,
assuming a drainage density of 2.1 km/km2 for the Coastal Plain and 4.9 km/km2
for the Piedmont (Calvo-Alvarado and Gregory 1997). Beaver-pond densities in
some Colorado mountain valleys currently range from 5 to 46 ponds/stream km
(Ringelman 1992). Johnston and Naiman (1990) estimated 13% of their Minnesota
study site to be Beaver impounded, with greater than 60% of those ponds exhibiting
decadal stability.
1US Geological Survey, 3535 South Sherwood Forest Boulevard, Suite 120, Baton Rouge,
LA 70816. 2US Geological Survey, 430 National Center, Reston, VA 20192. 3Center
for the Inland Bays, 39375 Inlet Road, Rehoboth, DE 19971. *Corresponding author -
dkroes@usgs.gov.
Manuscript Editor: Jennifer Rehage
Southeastern Naturalist
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
578
In the southeastern US, Beaver populations and ponds have increased dramatically
since the reintroduction and subsequent protection of 70 Beavers to
Virginia (VA) and North Carolina (NC) in the 1930s (Newbill and Parkhurst 2000,
Woodward and Hazel 1991). Human resistance to a real or perceived loss of land
and timber due to Beaver activity may maintain a low beaver pond to stream-km
ratio (McKinstry and Anderson 1999). Butler (1991) estimated Beaver density
to be less than 1 Beaver/20 km2 in the southeastern US, with concentrations of 1
Beaver/100 km2 in VA and NC. Despite this low density of beaver ponds relative
to the pre-European densities, Beaver are considered a nuisance species and may
be hunted year round (NCWRC 2010, VDGIF 2013). In areas where Beaver come
into conflict with humans, dam destruction and eradication of Beaver colonies occur
frequently.
It is important for us to understand the role of Beavers in stream/floodplain
systems, especially because their ponds trap sediment. However, the sedimenttrapping
potential remains understudied in the Piedmont and Coastal Plain of the
mid-Atlantic US. When higher-velocity water and sediment enter an area of ponded
water, some of the bedload and suspended sediment may be deposited (Kroes and
Kraemer 2013). This process results in increased sediment-storage rates within the
pond. Sediment deposition and channel processes in beaver ponds are well-studied
in high-gradient stream areas. In Germany, John and Klein (2004) observed the
greatest deposition amounts to be within the channel of beaver ponds (mean = 120
mm/yr), with decreased deposition in shallower areas (mean = 40 mm/yr). Butler
and Malanson (1995) observed an average of 21 mm/yr deposition in Montana
beaver ponds. Generally, ponds in these studied areas trapped more sediment than
was eroded from the downstream bed and banks (Bigler et al. 2001, Gurnell 1998,
Meentemeyer and Butler 1999).
In comparison to the focal areas of previous beaver pond sediment-deposition
studies, the Piedmont and Coastal Plain have lower stream gradients, resulting in
beaver ponds with greater ratios of surface area to volume than in higher-gradient
areas. These high ratios typically result in long, wide ponds with longer watertransit
times where suspended fine sediment has a better chance of settling before
exiting the pond. Longer settling times are important for sediment deposition in
streams where fine-grained sediment may be the only sediment type present. Further,
methods commonly used to determine sediment deposition in beaver ponds
in other physiographic regions have limited utility on the Coastal Plain as a result
of the sediment properties. Typically, low-order stream-bed load material in the
Coastal Plain would be considered fine sediment in a mountain stream, commonly
with median particle diameters of less than 0.5 mm (Kroes and Kraemer 2013). Floodplain
deposition typically has a grain size of less than 0.063 mm, primarily silt and clay (Kroes
and Hupp 2010). Low-order Piedmont streams in this area carry a bed load of medium
pebbles and finer particles, with a median diameter less than 10 mm, and if floodplains
are depositional, the deposition is fine sand, silt, and clay (less than 0.125 mm diameter)
(Hupp et al. 2013). All of the sediment is fine material, thus, it is very difficult to
use sediment grain-size changes to differentiate beaver-pond deposition from unSoutheastern
Naturalist
579
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
impounded deposition as is commonly done in areas with coarse sediments (Bigler
et al. 2001). Streams in the mid-Atlantic Coastal Plain commonly have organic
floodplain deposition that may have organic content values of 80% without beaverpond
activity (Kroes and Hupp 2010, Noe and Hupp 2005).
Currently, sediment-deposition rates on unimpounded floodplains along
Piedmont streams average 4 mm/yr, with many exhibiting reaches of erosional
floodplains (Allmendinger et al. 2005, Schenk and Hupp 2009, Schenk et al. 2013).
Average deposition in low-order southeastern US Coastal Plain streams ranges
from 1.6 to 5.4 mm/yr (Craft and Casey 2000, Hupp 2000, Kroes and Hupp 2010).
In this study, we quantified the effect of beaver ponds on sediment transport in
low-order streams along the mid-Atlantic coast where climate, agriculture, geomorphology,
and Beaver control differred from previously-studied areas. Secondly,
we aimed to determine the timing and events that influenced the storage and export
of sediment from beaver-pond complexes.
Methods
Field-site description
In order to determine Beaver effects on sediment deposition, we selected 3 areas
for our study: (1) Fairfax County, VA, (2) Westmoreland County, VA, and (3) Pitt
County, NC (Fig. 1A). These areas have different land forms and sediment sources
and, hypothetically, should store sediment at different rates than previously studied
areas. Fairfax County, VA (hereafter Fairfax) is within the Piedmont physiographic
province (Hunt 1967), and has a land-surface form of moderate-relief table lands
with 20–50% of the area on gentle slopes (Hammond 1964). This region has undergone
intensive housing and commercial development (suburbanization) resulting
in increased upland erosion (Booth and Bledsoe 2009). Development has increased
the percentage of impermeable surfaces resulting in increased runoff and, subsequently,
accelerated stream-channel erosion (Booth and Reinelt 1994, Hupp et al.
2013, Schenk et al. 2013). Study streams in this area were forested by Fraxinus spp.
(ash), Acer rubrum L., (Red Maple), and Quercus spp. (oak).
Westmoreland County, VA (hereafter Westmoreland) is within the Coastal
Plain physiographic province (Hunt 1967) and has a land-surface form of irregular
plains; 20–50% of the area slopes gently (Hammond 1964). Uplands in this area are
primarily forested or agricultural. During the study period, active head-cut erosion
and gullying were occurring on the slopes between some highlands and stream bottoms.
Many of the streams in this area were modified by a series of active beaver
ponds. Study streams in this area were forested by ash, Red Maple, and oak.
Pitt County, NC (hereafter Pitt) is within the Coastal Plain physiographic province
(Hunt 1967) and has a land-surface form of flat plains (Hammond 1964).
Uplands in this area are primarily agricultural or forested. Sources of sediment in
Pitt County are erosion of field and roadside ditches, erosion of agricultural land,
increasing levels of development, and past stream channelization. In a similar
Coastal Plain area, Gellis et al. (2009) used radioisotopes to determine that a large
portion of the entrained sediment comes from agricultural ditch beds and banks.
Southeastern Naturalist
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
580
Study streams in this area were forested by ash, Red Maple, Nyssa biflora (Walter)
(Tupelo), and Taxodium distichum L. (Richard) (Bald Cypress).
Site selection
All selected ponds had naturally vegetated buffer zones that minimized overland
sediment-transport from uplands directly into the ponds. We tried to choose sites
Figure. 1. (A) Study sites and (B) typical pond features with water-sampling locations.
Southeastern Naturalist
581
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
that had a single-thread stream input and, where possible, we chose new ponds
(trees present and little open water). We did not include ponds where the primary
water source appeared to be groundwater discharge within the pond. Although we
made every attempt to select ponds in parks and in remote locations with the lowest
chances of being destroyed, the Beaver is considered a nuisance species in the mid-
Atlantic States of the US, and many ponds were destroyed prior to the completion
of the study. There were no apparent significant changes in the landscape surrounding
the ponds or the watersheds during the study.
We defined a pond site from the downstream confluence of dam drainages to the
upstream channel-bed elevation that would normally be higher than the dam elevation
at the time of the initial site survey (Fig. 1B). We selected a total of 14 pond
sites for the study. We identified 5 sites in Fairfax (Table 1) through complaints of
Beaver activity to Fairfax County Park Authority (Fairfax, VA) and conducted field
reconnaissance to verify the presence of beaver ponds. Fairfax sites were located
on 1st–3rd- order streams (Strahler 1957), and were composed of a single pond,
inundating disjunct serial ponds, and channel disjunct serial ponds (Fig. 2 A, B)
of up to 6 ponds (5–80 m wide, 120–470 m long). We used topographic maps and
field reconnaissance to select 5 sites in Westmoreland (Table 1). These ponds were
located on streams that ranged from 1st–3rd order and were either treeless or had
standing dead wood. The Westmoreland sites were composed of single ponds (43–
83 m wide, 70–175 m long). We selected 4 sites in Pitt (Table 1) using 1998 digital
orthophotography, infrared aerial photography (Townsend and Butler 1996), and
field reconnaissance of sites sampled by Bason (2004). Pitt sites were on streams
that were 2nd order and were composed of single ponds (39–110 m wide, 210–410
m in long).
Table 1. Site names and preexisting conditions. F = Fairfax, W = Westmoreland, P = Pitt, U = unchannelized,
C = channelized, UF = unchannelized filled, HDS = high density suburban, F/S = forest
suburban mix, S= suburban, F/A = forest–agriculture mix, A = agriculture, and fp = floodplain.
Ponded
Stream Watershed Channel floodplain
Site Area order Channel cover % slope area (m2) area (m2)
Fryingpan Brook F 1 U HDS 0.29 320 540
South Run upstream F 1 C F/S 0.61 1490 0
Johnny Moore F 2 C S 0.74 1910 4800
South Run downstream F 3 U S 0.24 1070 0
Horsepen Run F 1 C S 0.59 3260 25,100
Canal Swamp upstream W 2 UF F/A 1.03 190 7500
Canal Swamp downstream W 3 UF F/A 0.34 130 8400
Tributary to Fox W 1 U F 0.49 160 5290
Bundy Swamp W 2 U F/A 0.95 250 3910
Fox Hall Swamp W 2 U F 0.33 180 4400
Tower Swamp P 2 C F 0.17 420 560
Howell Swamp P 2 U A 0.29 580 15,950
Bynum Mill P 2 U F 0.14 460 22,400
Juniper Branch P 2 C A 0.15 2870 16,000
Southeastern Naturalist
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
582
Description of pond types
In the process of conducting this study, it became evident that there were limited
(Hair et al. 1978, Pullen 1971), if any, physical descriptions of the many configurations
of beaver ponds. Thus, we describe the 4 basic beaver-pond types here (Fig. 2A):
Figure 2. A schematic of (A) pond type, (B) pond setting (in or out of series), (C) channel
type, (D) dam conditions, and (E) channel-fill conditions that a ffect sediment storage.
Southeastern Naturalist
583
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
(1) Inundating ponds occur where the beaver pond inundates the floodplain
from the dam to the upstream extent of the pond. The pond has no break in
continuity and stream-channel edges are not exposed at normal pool levels.
(2) Channel ponds occur when Beavers create a dam within the channel, but
at normal pool levels the floodplain is not inundated.
(3) Discontiguous ponds have the same dam layout as inundating ponds,
but the incoming channel is separated from the majority of the pond either
because the banks are exposed due to high sediment deposition, dam
construction along reaches of side-ditched streams, or as spoil banks from
channelization. Side-ditching is a common type of channelization where a
ditch is dug along the side of a floodplain instead of in the existing channel.
In this study, cuts between the ditch and the historical channel appeared to
have been dug by Beaver reverting flow to the natural floodplain.
(4) Floodplain ponds occur where dams are built in sloughs or in backswamp
areas (Townsend and Butler 1996).
Ponds are either pioneer or serial (Fig. 2 B). Pioneer ponds occur when Beavers
initially move up a stream (Hilfiker 1991) and create a single pond with no upstream
ponds. Serial ponds occur either in disjunct or in stair-step formation. Ponds
also occur along natural (Fig. 2C) and modified stream reaches. There are 3 major
types of channel modification that affect the stability and shape of beaver ponds:
(1) channelization, where the channel is straightened along the natural stream bed;
(2) side-ditched, where a ditch is dug at the edge of the floodplain and streamflow
is captured by the ditch; and (3) deepened/widened, where the natural channel is
deepened or widened either by manual excavation or by erosion of the banks and
bed (incision).
Sediment release occurs when Beaver dams are breached (Fig. 2D). The condition
of the dam may also significantly affect sediment transport (Woo and Waddington
1990). Additionally, during the period of ponding, the stream channel may fill with
sediment and organic debris, affecting sediment storage (Fig. 2E). If a channel fills
during ponding, the site is often referred to a beaver meadow (Polvi and Wohl 2012),
although the term meadow does not include all filled channel ponds.
Sediment measurement
We employed 3 methods of measuring sediment dynamics in beaver ponds:
(1) depth-integrated base-flow sampling (hereafter, base-flow samples), (2) repeat
surveys within the channel, and (3) depositional surfaces (pads) on the floodplain.
We collected base-flow samples to determine the instantaneous effect of beaver
ponds on suspended solids. We utilized repeat surveys and pads to examine variation
in sediment storage at each site.
We collected depth-integrated samples in downstream to upstream order using
the multiple-vertical method during base-flow conditions in the channel downstream
of the confluence of drainages (Edwards and Glysson 1999). We filtered
known volumes of samples using a 0.7-μm ash-free, glass-fiber filter, dried the
collected solids and filters for 24 h at 100 °C, and weighed them to determine the
suspended-sediment concentration (mg/l). We ashed the dried samples at 400 °C
Southeastern Naturalist
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
584
for 16 h and weighed them to determine volatile organic loss on ignition (Nelson
and Summers 1996). We calculated the suspended-sediment concentration (SSC)
reductions/increases from paired upstream and downstream samples.
We surveyed stream channels during site selection (2002) using a Topcon
RL-HA rotating laser (Topcon, Livermore, CA) with an accuracy of 1 mm/25
m. Surveying consisted of establishing a benchmark, measuring channel dimensions,
and determining bed elevation relative to the benchmark. We measured
pond dimensions and surveyed the beaver-dam elevation and width. We surveyed
the ponded channel at 5-m intervals from below the dam to the distance above
the pond where the channel bed exceeded pool elevation. We determined channel
gradient by surveying the thalweg (the deepest point of the channel that is a
longitudinally continuous feature, not a scour hole) 20 m downstream of the pond
and upstream of the pool elevation, and reported it as channel percent-slope between
these points. We resurveyed the ponds in 2003 and 2005 and reported the
difference in channel-bed elevation as a change in volume (channel width x length
x average-elevation change). We reported changes in channel elevation of less than 1 cm
as being below detection limits and did not include beaver dams and lodge structures
as depositional volumes. The channel of Howell Swamp (Pitt County) was
excluded from channel surveys because the water in the channel exceeded 2 m in
depth and was not boat accessible.
In 2002, we placed pads on floodplains and in the ponds to determine sediment-
deposition rates outside of the channel (hereafter referred to as floodplain
deposition). We used feldspar-clay pads in locations that were normally dry (Bauman
et al. 1984, Kleiss et al. 1989). In locations with standing water, we anchored
4-mm-thick plastic sheets or ~0.3 m2 steel pads to the floodplain surface. We placed
3–15 pads on the floodplain depending on pond size and shape to collect at the
upstream, middle, or downstream portions of the pond. We measured clay pads by
removing plugs of material and measuring the deposition above the marker horizon.
We measured deposition on the plastic and steel sheets by inserting a sharp
knife with a blunted tip into the deposited material until it came into contact with
the sheet and measured the depth of insertion; 3 measurements of the depositedsediment
thickness were made per pad. We measured depth of sediment deposition
in 2003 and 2005. We reported accumulations of less than 1 mm as being below detection
limits. Sediment deposition is reported as a vertical rate (mm/yr) and volume (m3/
yr) (area of pond x average deposition). We determined sediment texture in the field
(Thein 1979).
We employed multiple regressions and hierarchical cluster analysis using the
Minkowsky method to analyze the data and determine meaningful groupings of
depositional data in relation to physiographic area, channel gradient, presence/absence
of channelization, beaver-dam removal, and watershed condition (suburban,
forested, agriculture). We compared sediment storage across the 3 regions using an
ANOVA (SPSS v16.0).
Southeastern Naturalist
585
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
Table 2. Base-flow depth-integrated samples for suspended-sediment concentration (SSC) and particulate
organic matter (POM) upstream (up) and downstream (down) of sites. Percentage difference
is given for paired (upstream and downstream) samples. F = Fairfax, W = Westmoreland, and P = Pitt.
Standard errors are in parentheses. For sites: (u) = upstream, (d) = downstream.
SSC POM
Up Down Avg % Up Down Avg %
Site Area n (mg/L) (mg/L) change (mg/L) (mg/L) change
Johnny Moore F 9 6.8 (1.4) 8.3 (1.3) 26 3.5 (0.9) 5.0 (0.6) 44
South Run (u) F 10 20.0 (3.3) 24.0 (4.9) 21 10.3 (2.9) 13.0 (4.6) 27
South Run (d) F 5 20.0 (7.1) 20.0 (8.6) -13 7.6 (1.0) 7.5 (0.7) 0.1
Horsepen Run F 5 15.0 (4.3) 20.0 (3.7) 40 6.6 (0.7) 7.5 (0.9) 13
Fryingpan Brook F 5 29.0 (4.9) 46.0 (3.7) 59 7.8 (1.3) 10.0 (2.0) 28
Farifax avg 27 22
Canal Swamp (u) W 5 15.0 (6.4) 22.0 (14.0) 46 4.1 (2.1) 4.0 (2.0) -3
Canal Swamp (d) W 4 15.0 (7.2) 19.0 (16.0) 27 4.3 (1.3) 2.9 (0.7) -33
Bundy Swamp W 5 19.0 (7.1) 14.0 (7.0) -26 8.1 (6.1) 5.7 (2.3 -30
Fox Hall Swamp W 5 39.0 (20.0) 15.0 (5.8) -61 5.9 (1.9) 8.3 (6.6) 41
Westmoreland avg -4 -6
Tower Swamp P 7 7.1 (3.6) 10.0 (5.9) 40 3.0 (3.3) 4.9 (4.2) 63
Howell Swamp P 6 5.5 (2.6) 4.3 (1.6) -22 3.1 (1.2) 3.5 (1.5) 13
Bynum Mill P 5 2.3 (0.4) 1.5 (0.5) -35 1.6 (0.3) 1.3 (0.3) -19
Juniper Branch P 8 5.6 (1.6) 5.1 (1.4) -9 1.6 (1.5) 1.2 (1.6) -25
Pitt avg -7 8
Total average 5 7
Results
All areas
Base-flow concentrations of particulate organic matter (POM) increased
through the beaver ponds in Fairfax and decreased in Westmoreland and Pitt
(Table 2). Sediment storage in the channel was the most variable on the Piedmont
(Fairfax) and least variable on the Coastal Plain (Pitt) (Table 3). Breached beaver
dams released sediments previously stored in-channel (Fig. 2D). Regions varied
in sediment storage (ANOVA: P = 0.001; Fig. 3). Sediment storage on the floodplain
was greatest in Fairfax along inundating ponds and lowest in Westmoreland
(Table 3). Hierarchical cluster analysis of channel storage grouped the Pitt sites
together with 1 Westmoreland site and 1 Fairfax site. The Westmoreland stair-step
serial sites grouped together, and most of the Fairfax sites did not group. None of
the measured parameters (gradient, ponded area, stream order) correlated with
deposition volumes (R2 = 0.04, 0.18, 0.01), SSC (R2 = 0.02, -0.01, -0.01), or POM
(R2 = 0.05, 0.12, 0.22).
We identified 3 channel-storage groupings. The first group was unbreached
dam and open channel. Channel accumulation of sediment occurred in this group
during the first year of study. The second group was unbreached, filled. These
sites lost negligible amounts of sediment from the channel during the first year.
In filled-channel beaver-ponds, the greatest deposit of sediment was within the
Southeastern Naturalist
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
586
channel. When the channel was filled, sediment deposition was evenly distributed
between the channel and floodplain. The third group was breached; several dams
were breached during the study and lost considerable amounts of unconsolidated
sediment (161–718 m³) from the channel.
Piedmont, Fairfax County, VA
Suspended-sediment concentrations in base-flow samples for the 5 Fairfax
sites ranged from 2.3 mg/l to 48 mg/l. During the study period, average base-flow
SSC increased by 27% from upstream to downstream. POM ranged from 1.8 mg/l
to 21 mg/l, constituting an average 40% of base-flow SSC. There was an average
POM increase of 22% from upstream to downstream (Table 2). In-channel sediment-
storage volume at Fairfax ranged from -718 m³/yr to 131 m³/yr. Sediment
losses occurred at breached dams, and accumulation occurred in unbreached
ponds (Fig. 2D). On average, the 5 sites lost 200 m³/yr of sediment from the
channel. If the 3 sites breached by human activity are removed from the analysis,
the 2 remaining sites stored an average of 77 m³/yr. All beaver ponds were
destroyed by 2005, but channel blockages remained at 2 sites. In 2005, surveys
at those ponds indicated a loss of 15 m³/yr in one and no measurable difference at
the other (Table 3, Fig. 4).
Floodplain deposition at the Fairfax sites ranged from 11 m³/yr to 533 m³/
yr (less than 1 mm/yr to 111 mm/yr). The lowest volume occurred at a small inundating
pond (Fig. 2A) with the lowest channel gradient of the 5 sites. The greatest
deposition volume occurred at the Johnny Moore site, which was a large inundating
pond with the steepest channel gradient. No measurable deposition occurred
Figure 3. Hierarchical cluster analysis of in-channel sediment storage. Pitt and Westmoreland
sites grouped based on area. Fairfax sites did not group. P = Pitt, F = Fairfax, W =
Westmoreland, I = dam intact, B = dam breached. Howell Swamp (Pitt) was excluded from
channel storage analyses because it exceeded wadeable depth.
Southeastern Naturalist
587
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
Table 3. Site names, dam conditions, and depositional rates for 2002–2003 and 2003–2005. F =
Fairfax, W = Westmoreland, P = Pitt, I = dam intact, A = abandoned, BH = breached by humans, Pb
= partial blockage, BD = below detection limits (resurveyed elevational change less than 1 cm, floodplain
deposition less than 1 mm), Cl = clay, St = silt, O = organics, Sd = sand, and Fg = fine gravel. (u) = upstream;
(d) = downstream. Howell Swamp (Pitt) was excluded from channel storage analyses because it exceeded
wadeable depth.
Channel Avg. floodplain Floodplain
deposition deposition deposition Deposited
Dam (m³/yr) (mm/yr) (m³/yr) sediment
Site Area 2003 2005 02–03 03–05 02–03 03–05 02–03 03–05 texture
Fryingpan Brook F I BH 23 BD 21 6 11 3 Cl, St, O
South Run (u) F I BH 131 - BD - BD - Cl, St, Sd
Johnny Moore F BH Pb -278 -15 111 33 533 290 Sd, Fg
South Run (d) F BH BH -161 - BD - BD - -
Horsepen Run F BH BH -718 - 15 7 377 184 Cl, St, Sd
Fairfax avg -200 7 26 15 184 115
Canal Swamp (u) W I A -16 -5 - - - - -
Canal Swamp (d) W I I -6 BD - - - - -
Trib to Fox W I - 5 - 2 - 11 - O, St
Bundy Swamp W I A 21 21 9 19 37 74 O, St, Cl
Fox Hall Swamp W I A BD BD 17 6 75 27 O
Westmoreland avg 1 4 9 13 41 53
Tower Swamp P I I 18 BD 2 7 2 10 O, St
Howell Swamp P I - - - 7 - 116 - O, St, Cl
Bynum Mill P I I 23 BD 2 10 39 219 O, St
Juniper Brook P I I 34 175 29 15 466 197 Sd, O
Pitt avg 26 58 10 11 156 142
Figure 4 (following page). Repeat longitudinal surveys of channel-bed elevations at 5 m
intervals relative to benchmarks: 2002, 2003, and 2005 from Fairfax (A–E), Westmoreland
(F–J), and Pitt (K–M). Howell Swamp (Pitt) was excluded from channel storage analyses
because it exceeded wadeable depth.
on the pads at the 2 channel ponds (Fig. 2A) where water did not inundate the
floodplain except during storm events. Average deposition for Fairfax was 184
m³/yr. The average floodplain deposition for inundating ponds in Fairfax was 307
m³/yr. Deposited sediment texture varied by watershed and location in the pond
ranging from fine gravel to silty clay. Measurements in 2005 indicated greatly reduced
storage rates (Table 3, Fig. 5).
Coastal Plain, Westmoreland County, VA
Base-flow samples collected for 4 Westmoreland sites had SSC ranging from 2.2
mg/l to 92 mg/l. Beaver ponds reduced the average SSC by 3.5% from upstream to
downstream. POM ranged from less than 1 mg/l to 29 mg/l, constituting an average 37%
of base-flow SSC. There was an average POM decrease of 26% from upstream to
downstream. Repeated channel surveys showed channel-sediment volume changes
of -16 m³/yr to 21 m³/yr. Losses occurred where channels were filled prior to initial
Southeastern Naturalist
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
588
Southeastern Naturalist
589
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
site surveying (Fig. 2E). Accumulation occurred in ponds where the channels were
open (Fig. 2E). On average, the 5 sites accumulated 1 m³/yr of sediment in the channel
by 2003 and increased to 4 m³/yr during 2003–2005 (Table 3, Fig. 5). The smallest
change in channel storage occurred in the stable, serial beaver ponds (Fig. 2B).
We observed floodplain deposition ranging from 11 m³/yr to 75 m³/yr (2–19
mm/yr) at 3 sites in Westmoreland. The greatest deposition occurred at the lowest
channel-slope site. The average floodplain-deposition volume for the area was 41
m³/yr and increased to 53 m³/yr during the period 2003–2005. Many ponds had been
Figure 5. Beaver pond configurations and floodplain deposition or erosion rates (mm/yr) at
each pad location in ponds during 2002–2003: Fairfax (A–E), Westmoreland (F–H), and Pitt
(I–L), and 2003–2005: Fairfax (M–O), Westmoreland (P, Q) and Pitt (R–T). (--) indicates a
lost pad. Upstream pads at Johnny Moore were upstream of the influence of the pond. Only
ponds with pad deposition or erosion data are shown. Positive rates indicate accretion; negative
rates indicate erosion.
Southeastern Naturalist
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
590
abandoned, with partial dam breakage leaving partial channel blockages. Floodplain
deposition rates increased between 2003 and 2005, and deposited sediment
particle size ranged from silt to organics (Table 3, Fig. 5).
Coastal Plain, Pitt County, NC
Base-flow samples collected for the 4 Pitt sites had SSC ranging from less than 1 mg/l
to 44 mg/l. During base flow, average SSC decreased by 7% through the ponds.
Discontiguous (Fig. 2A) and channel ponds exported sediment. POM ranged from
less than 1 mg/l to 29 mg/l, constituting an average 80% of base-flow SSC, and increased
8% from upstream to downstream. Repeated channel surveys (excluding Howell
Swamp) showed the greatest average channel-sediment storage (18–34 m³/yr),
with all ponds exhibiting sediment storage in the channel. On average, the 3 sites
gained 26 m³/yr of sediment in the channel; this rate doubled between 2003 and
2005 (Table 3, Fig. 5). Floodplain deposition ranged from 0.4 m³/yr to 466 m³/yr
(2–29 mm/yr). The greatest deposition volume occurred at a discontiguous pond located
at the confluence of a channelized reach with the natural channel. The lowest
deposition-accumulation volume occurred at a side-ditched, channel pond (Tables
2, 3). The average floodplain-deposition rate for Pitt was 156 m³/yr; there was a
decrease to 142 m³/yr between 2003 and 2005. Deposited sediment was primarily
sand at high-deposition sites and silt/clay at low-deposition sites (Table 3; Fig. 5).
Discussion
This study investigated the role of beaver ponds in sediment storage on the
Coastal Plain and Piedmont of the mid-Atlantic, and detected differences in sediment
storage between the Piedmont and Coastal Plain. Sediment storage in beaver
ponds appears to be influenced by several factors including shape, position in relation
to other ponds, channel condition, and frequency of dam destruction. Beaver
ponds that were not destroyed during the study period stored large volumes of
sediment. Beaver ponds were not static; pond repositioning and reconfiguration
occurred during the study. Often, new dams were built within pioneer ponds and
sometimes dams were built that inundated older dams.
Mean floodplain accretion rates in these beaver ponds (2002-2003: 20 mm/yr
2003-2005: 15 mm/yr) greatly exceeded the mean deposition rate of similar unimpounded
streams. Previous studies of unimpounded low-order stream floodplains
in these areas have documented mean sediment accretion rates up to 5.4 mm/yr
and shown evidence indicating those in the Piedmont sometimes exhibited erosion
(Allmendinger et al. 2005, Craft and Casey 2000, Hupp 2000, Kroes and Hupp
2010, Schenk and Hupp 2009, Schenk et al. 2013). Accretion rates in the Piedmont
beaver ponds ranged from below to within the range of published rates for
beaver impoundments, but accretion rates in the ponds on the Coastal Plain were
consistently less than most previously reported rates (Bigler et al. 2001, Butler
and Malanson 1995, John and Klein 2004). Despite the inequity in accretion rates,
the volume of deposition was similar between Piedmont and Coastal Plain beaver
ponds as a result of the typically much greater area of the Coastal Plain ponds. Inundation
ponds were the most effective at trapping sediments. Channel ponds were
Southeastern Naturalist
591
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
ineffective at increasing sediment trapping on the floodplain. The dams that formed
channel ponds did not appear to slow water velocities enough to induce sediment
deposition on the floodplain during flood events and frequently breached or blewout
due to the dynamic hydrology of the stream. It seems probably that inundation
ponds have greater volume and slower flow resulting in increased settling of suspended
sediment.
Low sediment-storage rates in channel ponds may also be attributed to Beaver
activity, with bioturbation increasing with greater confinement. Beaver activities
associated with the creation and maintenance of dams and the excavation of bank
burrows and slides release fine particles into the water column and can be a significant
source of sediment (Meentemeyer et al. 1998). However, at our sites burrows
were common only along channel and discontiguous ponds (D.E. Kroes, Unpubl.
data). Bioturbation of fine sediment is especially relevant when the primary sediments
are clay and silt.
In the Coastal Plain, POM concentration decreased through most of the ponds
during base flow. In Fairfax, POM concentration increased through most sites. In
shallow meadows, this carbon accumulation is supplemented by dense herbaceous
growth and substantial root mass in a shade-free environment (Wohl 2013). As beaver
ponds age, they accumulate and bury POM and large organic debris, possibly
becoming important carbon repositories.
Sediment-storage patterns appear to be different between the channel and the
floodplain. The amount of in-channel sediment deposition appears to be related
to the amount of bedload and whether the pond is pioneer or serial. We identified
3 channel-storage groups: (1) unbreached dam and open channel, (2) unbreached,
filled, and (3) breached. Breached ponds showed considerable loss of stored
sediment. These lost sediments may have been stored in the channel for the life of
the pond (days to decades) until breaching, and this phenomenon demonstrates the
dynamic nature of in-channel storage. Where a pond was abandoned after pond maturity
as in Westmoreland, deposition rates increased as bioturbation was decreased
and woody debris stabilized channel sediment. Apparently, where no large wood
was present in the channel, complete sediment washout occurred; however, we did
not measure large wood features.
In the Piedmont (Fairfax), individual watershed occurrences of Beaver eradication,
or dam-busting overwhelmed the regional similarity in regard to channel
storage. In contrast, the highest stream-gradient area (Westmoreland) and lowestgradient
site (Pitt) had strong similarity in channel storage between sites and areas.
One cause of the dissimilarity in Fairfax storage rates may be the relative scarcity
of ponds in combination with the active, frequent destruction of beaver dams. In a
serial pond setting, much of the in-channel sediment from the breached pond would
be trapped in downstream ponds. However, Beaver eradication would eventually
result in the total failure of the pond series.
Floodplain-sediment storage in beaver ponds is longer-term than channel storage
(Westbrook et al. 2011), especially in low gradient Coastal Plain systems,
where deposited sediments may be stored for centuries or millennia (Meade et al.
1990). Ponds with the slowest velocities and longest turnover-time should then be
Southeastern Naturalist
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
592
the most effective at trapping fine-grained sediments and associated phosphorus
(Noe et al. 2007).
Conclusions
Beaver ponds have great potential to trap sediments. When beaver ponds are
present on channelized streams, they often hydraulically reconnect the floodplain
with the stream, thus, restoring natural function to the floodplain. The drainage
density of the VA and NC Piedmont is 4.91 km/km2 and the drainage density of the
Coastal Plain is 2.14 km/km2 (Calvo-Alvarado and Gregory 1997). If the deposition
for intact beaver ponds were applied to 1 pond/stream km on Piedmont VA and NC
floodplains, it would result in 22 million m3/yr of deposition and reduce previously
reported rates of stream incision and erosion of post-colonial floodplain deposits.
On the Coastal Plain, 1 pond/km would result in 19 million m3/yr of deposition in
VA and NC.
Beaver ponds in suburban Fairfax County were destroyed more rapidly than in
the other areas despite protection by Park regulations (FCPA 2006). Conflicts with
humans occur as Beaver populations increase, ponds form, subsequent tree damage
occurs, or drainages become blocked (Marjorie Pless, Fairfax County Parks
Authority, Fairfax,VA, pers. comm.). Although the motivations for Beaver removal
are varied, the end result is increased sediment yield from watersheds relative to
yields when impoundments are present. Beaver ponds may be an underutilized asset
in reducing the sediment and associated nutrient-delivery rate to the eutrophic
estuaries of the mid-Atlantic US.
Our study was designed to investigate the effect of beaver ponds on sediment
deposition in relation to physiographic settings. It appears however that the pond
and dam conditions as well as their disturbance frequency overwhelm the signal
of the differing gradients, sediment loads, and land use. Future research should be
focused on ponds and channels of differing configurations to determine the specific
effects of each pond and channel type. Further investigation should be conducted
on sediment trapping in beaver ponds on the higher sediment-load streams of the
Great Plains and Central Lowlands.
Acknowledgments
This research was funded by the US Geological Survey’s Chesapeake Bay and National
Research Programs. Thanks to Joshua Elwell for his help in the field despite his
unreasonable fear of leaches. Thanks to Fairfax County Parks, Fairfax County Stormwater
Management, all landowners, and land managers who allowed us access to their lands. 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
Allmendinger, N.E., J.E. Pizzuto, N. Potter Jr., T.E. Johnson, and W.C.Hession. 2005. The
influence of riparian vegetation on stream width, eastern Pennsylvania. US Geological
Society of America Bulletin 117:229–243.
Southeastern Naturalist
593
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
Bailey, V. 1927. Beaver habitats and experiments in Beaver culture. USDA Technical Bulletin
Number 21.
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.
Baumann, R.H., J.W. Day, and C.A. Miller. 1984. Mississippi deltaic wetland survival:
Sedimentation versus coastal submergence. Science 224:1093–1095.
Bigler, W., D.R. Butler, and R.W. Dixon. 2001. Beaver-pond sequence morphology and
sedimentation in northwestern Montana. Physical Geography 22:531–540.
Booth, D.B. and B.P. Bledsoe. 2009. Streams and urbanization. Pp. 93–123, In L.A. Baker
(Ed.).The Water Environment of Cities. Springer US, New York, NY. 375 pp.
Booth, D.B., and L.W. Reinelt. 1994. Consequences of urbanization on aquatic systems:
Measured effects, degradation thresholds, and corrective strategies. Pp. 545–550. In
Steering Committee (Eds.). Proceedings Watershed 93: A National Conference on Watershed
Management, 12–14 March 1993, Alexandria, VA. EPA publication 840-R-94-
002. US Government Printing Office, Washington, DC. 914 pp.
Butler, D.R. 1991. The reintroduction of the Beaver into the south. Southeastern Geographer
1:39–43.
Butler, D.R., and G.P. Malanson. 1995. Sedimentation rates and patterns in beaver ponds in
a mountain environment. Geomorphology 13:255–269.
Calvo-Alvarado, J.C., and J.D. Gregory. 1997. Predicting mean annual runoff and suspended-
sediment yield in rural watersheds in North Carolina. Water Resources Research
Institute Report Number. 307, University of North Carolina, Chapel Hill, NC.
Craft, C.B., and W.P. Casey. 2000. Sediment and nutrient accumulation in floodplain and
depressional wetlands of Georgia, USA. Wetlands 20:323–332.
Edwards, T.K., and G.D. Glysson. 1999. Field Methods for Measurement of Fluvial Sediment:
Techniques of Water-Resources Investigations of the USGS. Book 3: Applications
of Hydraulics. US Geological Survey, Reston, VA. 97 pp.
Fairfax County Park Authority (FCPA). 2006. Park authority regulations. Available online
at http://www.fairfaxcounty.gov/parks/parkpolicy/app7regs.pdf. Accessed 7 July 2015.
Gellis, A.C., C.R. Hupp, M.J. Pavich, J.M. Landwehr, W.S.L. Banks, B.E. Hubbard, M.J.
Langland, J.C. Ritchie, and J.M. Reuter. 2009. Sources, transport, and storage of sediment
at selected sites in the Chesapeake Bay Watershed. Scientific Investigations Report
2008-5186. US Geological Survey, Reston, VA. 95 pp.
Gurnell, A.M. 1998. The hydrogeomorphological effects of beaver dam-building activity.
Progress in Physical Geography 22:167–189.
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.) Strategies for Protection and Management
of Floodplain Wetlands and other Riparian Ecosystems: Proceedings of the Symposium,
11–13 December 1978, Callaway Gardens, GA USDA GTR-WO-12. US Department of
Agriculture Forest Service, Washington, DC. 410 pp.
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.
Hilfiker, E.L. 1991. Beavers, Water, Wildlife, and History. Windswept Press, Interlaken,
NY. 198 pp.
Hodgdon, K.W., and J.H. Hunt. 1955. Beaver management in Maine. Game Division Bulletin
No. 3, State of Maine Department of Inland Fisheries and Game, Augusta, ME.
102 pp.
Southeastern Naturalist
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
594
Hunt, C.B. 1967. Physiography of the United States. W.H. Freeman and Company, San
Francisco, CA. 480 pp.
Hupp, C.R. 2000. Hydrology, geomorphology and vegetation of Coastal Plain rivers in the
southeastern USA. Hydrological Processes 14:2991–3010.
Hupp, C.R., G.B. Noe, E.R. Schenk, and A.J. Benthem. 2013. Recent and historic sediment
dynamics along Difficult Run, a suburban Virginia Piedmont stream. Geomorphology
181:156–169.
John, S., and A. Klein. 2004. Hydrogeomorphic effects of beaver dams on floodplain
morphology: Avulsion processes and sediment fluxes in upland valley floors (Spessart,
Germany). Quaternaire 15:219–231.
Johnston, C.A., and D.H. Chance. 1974. Pre-settlement overharvest of upper Columbia
River Beaver populations. Canadian Journal of Zoology 52:1519–1521.
Johnston, C.A., and R.J. Naiman. 1990. The use of geographical information system to
analyze long-term landscape alteration by Beaver. Landscape Ecology 4:5–19.
Kleiss, B.A., E.E. Morris, J.F. Nix, and J.W. Barko. 1989. Modification of riverine water
quality by an adjacent bottomland hardwood wetland. Pp. 429–438, In D.W. Fisk (Ed.).
Proceedings of Wetlands: Concerns and Successes. Tampa, FL, 17–22 September1989.
TPS 89-3. American Water Resources Association, Bethesda, MD. 568 pp.
Kroes, D.E., and C.R. Hupp. 2010. The effect of channelization on floodplain-sediment deposition
and subsidence along the Pocomoke River, Maryland. Journal of the American
Water Resources Association 46:686–699.
Kroes, D.E., and T.F. Kraemer. 2013. Human-induced stream-channel abandonment/capture
and filling of floodplain channels within the Atchafalaya River Basin, Louisiana.
Geomorphology 201:148–156.
McKinstry, M.C., and S.H. Anderson. 1999. Attitudes of private and public land managers
in Wyoming, USA, toward Beaver. Environmental Management 23:95–101.
Meade, R.H., T.R. Yuzyk, and T.J. Day. 1990. Movement and storage of sediment in rivers
of the United States and Canada. Pp. 255–280, In M.G. Wolman (Ed.). Surface
Water Hydrology. The Geology of North America, Volume 0–1. Geological Society of
America, Boulder, CO. 382 pp.
Meentemeyer, R.K., and D.R. Butler. 1999. Hydrogeomorphic effects of beaver dams in
Glacier National Park, Montana. Physical Geography 20:436–446.
Meentemeyer, R.K., J.B. Vogler, and D.R. Butler. 1998. The geomorphic influences of burrowing
Beavers on streambanks, Bolin Creek, North Carolina. Zeitschrift für Geomorphologie
42:453–468.
Merritts, D., R. Walter, C. Lippincott, and S. Siddiqui. 2004. High suspended-sediment
yields of the Conestoga River watershed to the Susquehanna River and Chesapeake Bay
are the result of ubiquitous post-settlement mill dams. Fall Meeting Supplement, H51F-
06. Eos Transactions AGU 85:47.
Morgan, L.H. 1868. The American Beaver and His Works. J.B. Lippincott and Company,
Philadelphia, PA. 330 pp.
Nelson, D.W., and L.E. Sommers. 1996. Total carbon, organic carbon, and organic matter.
Pp. 1002–1005, In D.L. Sparks (Ed.). Methods of Soil Analysis, Part 3: Chemical
Methods-SSSA Book Series 5, SSSA, Madison, WI. 1358 pp.
Newbill, C.B., and J. Parkhurst. 2000. Managing wildlife damage: Beavers (Castor canadensis)
Virginia Cooperative Extension Publication Number 420-202, Blacksburg, VA.
Noe, G.B., and C.R. Hupp. 2005. Carbon, nitrogen, and phosphorus accumulation in floodplains
of Atlantic coastal plain rivers, USA. Ecological Applications 15:1178–1190.
Southeastern Naturalist
595
D.E. Kroes and C.W. Bason
2015 Vol. 14, No. 3
Noe, G.B., J. Harvey, and J. Saiers. 2007. Characterization of suspended particles in Everglades
wetlands. Limnology and Oceanography 52:1166–1178.
North Carolina Wildlife Resources Commission (NCWRC). 2010. Beaver management in
North Carolina. Available online at http://216.27.39.101/Wildlife_Species_Con/Beaver/
Documents/Beaver_Mgt_in_NC.pdf. Accessed 12 November 2013.
Polvi L.E., and E. Wohl. 2012. The Beaver meadow complex revisited: The role of Beavers
in post-glacial floodplain development. Earth Surface Processes and Landforms
37:332–346.
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.
Ringelman, J.K. 1992. Ecology of montane wetlands. Fish and Wildlife Leaflet 133.3.6
Waterfowl Management Handbook, US Fish and Wildlife Service, US Department of
Interior, Washington, DC.
Schenk, E.R., and C.R. Hupp. 2009. Legacy effects of colonial millponds on floodplain
sedimentation, bank erosion, and channel morphology, mid-Atlantic, USA. Journal of
the American Water Resources Association 45:597–606.
Schenk, E.R., C.R. Hupp, A. Gellis, and G. Noe. 2013. Developing a new stream metric for
comparing stream function using a bank–floodplain sediment budget: A case study of
three Piedmont streams. Earth Surface Processes and Landforms 38:771–784.
Strahler, A.N. 1957. Quantitative analysis of watershed geomorphology. American Geophysical
Union Transactions 38:913–920.
Thein, S.S. 1979. A flow diagram for teaching texture-by-feel analysis. Journal of Agronomic
Education 40:54–55.
Townsend, P.A., and D.R. Butler. 1996. Patterns of landscape use by Beaver on the lower
Roanoke River floodplain, North Carolina. Physical Geography 17: 253–269.
US Climate Data. 2013. Average annual temperature and precipitation. Available online at
http://www.usclimatedata.com/climate.php?location=USNC0281 (for Greenville, NC),
http://www.usclimatedata.com/climate.php?location=USVA0474 (for Matthews, VA),
and http://www.usclimatedata.com/climate.php?location=USVA0735 (for Sterling,
VA). Accessed 20 November 2013.
Virginia Department of Inland Game and Fisheries (VDGIF). 2013. Furbearer trapping
seasons. Available online at http://www.dgif.virginia.gov/hunting/regulations/furbearertrapping.
asp. Accessed 12 November 2013.
Walter, R.C., and D.J. Merritts. 2008. Natural streams and the legacy of water-powered
mills. Science 319:299–304.
Westbrook, C.J., D.J. Cooper, and B.W. Baker. 2011. Beaver-assisted river-valley formation.
River Research and Applications 27:247–256.
Wohl, E. 2013. Landscape-scale carbon storage associated with beaver dams. Geophysical
Research Letters 40:3631–3636.
Woo, M.K., and J.M. Waddington. 1990. Effects of beaver dams on subarctic wetland hydrology.
Arctic 43:223–230.
Woodward, D., and R. Hazel. 1991. Beavers in North Carolina: Ecology, utilization, and
management. Cooperative Extension Service Publication No. AG-434, North Carolina
State University, Raleigh, NC.