Content uploaded by Joe Berg
Author content
All content in this area was uploaded by Joe Berg on Jan 28, 2022
Content may be subject to copyright.
December 2020 ECOLOGICAL RESTORATION 38:4 • 257
Color version of this article is available through online subscription
at: http://er.uwpress.org
Ecological Restoration Vol. 38, No. 4, 2020
ISSN 1522-4740 E-ISSN 1543-4079
©2020 by the Board of Regents of the University of Wisconsin System.
DESIGN APPROACHES TO ECOLOGICAL RESTORATION
Stream Restoration Using Engineered
Wood Structures Harvested from On-site:
The Past and Future of Streams
Joe Berg, Doug Streaker and Chris Streb
ABSTRACT
Modifications of our watersheds, starting with beaver extirpation and continuing through current land development
practices (whether for agriculture, residential, commercial, or institutional uses), have negatively impacted our stream
resources. Changed watershed hydrology has led to stream channel enlargement and isolation from stream valley and
floodplain resources, with a cascade of resource degradation. Stream restoration—specifically restoration using engi-
neered wood structures to restore the stream valley and floodplain connection—can sustainably reverse the cascade of
degradation and regenerate natural processes that will continue to improve the system’s ecological function over time.
Wood has been used to enhance fish habitat in streams for decades, but it has largely served to augment the use of rock
in stream restoration projects. Stream restoration using wood is gaining interest and acceptance following the publication
of manuals such as “Stream Restoration Using Large Wood” (Bureau of Reclamation and U.S. Army Corps of Engineers
Research and Development Center, 2016) and “Low-Tech Process Based Restoration Manual” (Wheaton etal. 2019).
I
n a degraded coastal plain stream in Maryland, earth and
wood harvested from the project site were used to create
in-channel structures to reverse channel enlargement and
reconnect the stream to its valley and oodplain resources.
Following construction, ground and surface water eleva-
tions increased, restoring wetland hydrology in the stream
valley. In addition, peak stormwater discharge decreased as
the time of concentration increased (increased hydrograph
duration). Slower ow velocities and increased contact with
oodplain vegetation benetted both water quality and
aquatic habitat. is restoration approach is analogous to
that used by beaver to create beaver dams. Compared to
other stream restoration techniques, this approach greatly
reduces both greenhouse gas emissions (GHGs) and con-
struction costs associated with the restoration, because it
uses materials harvested on-site.
Stream channel degradation can be observed throughout
the urbanized Mid-Atlantic. In the Chesapeake Bay water-
shed, stream restoration is a best management practice for
reducing total nitrogen, phosphorous, and sediment loads.
Jurisdictions with Municipal Separate Storm Sewer System
(MS4) permits have signicant investments in stream
restoration projects to both halt channel degradation and
satisfy the waste load allocation associated with their MS4
permit. A comprehensive approach to stream restoration
is to reverse the hydrological impacts on stream networks
resulting from patterns of land development. While it is not
practical to restore watersheds to pervious forested cover
to truly restore the watershed hydrology, it is possible to
have incrementally positive eects on watershed hydrology
when restoring streams and oodplains. Techniques that
use locally sourced materials, minimize energy inputs, and
leverage feedback loops to hasten the system’s ability to self-
repair are needed to lower the costs, minimize disturbance,
and increase the sustainability of ecological restoration.
To leverage natural feedback loops that contribute
to self-repair, it is useful to review the hydrologic and
hydraulic changes within the landscape that lead to stream
and ecosystem degradation. Landscape development has
greatly reduced the ability for precipitation to inltrate
into the ground, recharge aquifers, and slowly move from
higher elevations through a natural network of streams and
wetlands to receiving waterways. Impervious cover and
its associated network of pipes radically alters the water
balance of a landscape. With infrastructure designed to
convey water downstream as quickly as possible, a cascade
of resource degradation ensues.
e increased volume of surface water runo with
heightened ow velocity leads to stream channel erosion
258 • December 2020 ECOLOGICAL RESTORATION 38:4
Figure 1. General Location of Project Area.
December 2020 ECOLOGICAL RESTORATION 38:4 • 259
and incision. In turn, groundwater elevations drop and
oodplains are drained by incised stream channels, leading
to the loss of spring seeps and perennial stream ow. As
stream channels enlarge, ood ows of increasing magni-
tude are contained within the channel banks, establishing
a reinforcing, degenerating feedback loop of more chan-
nel widening and incision and loss of aquatic habitat.
Enlarged stream channels also drain riparian and ood-
plain wetlands, leading to the loss of vernal pool habitats
and associated amphibian species, simplication of forest
community composition, and reduced diversity of associ-
ated ora and fauna.
e most common form of stream restoration in use
today is natural channel design (NCD), an approach orga-
nized and proselytized by Dave Rosgen (Rosgen 1997).
NCD has been and continues to be taught to and adopted
by many stream restoration practitioners. While this
approach is a dramatic improvement over previous stream
stabilization techniques (e.g., concrete trapezoid, rip-rap,
and gabion basket-lined channels), stream restoration
methodologies are not so well rened that improvements
are not possible. e NCD approach (as implemented by
many practitioners) stabilizes an enlarged channel and con-
veys water and sediment through a stream reach designed
not to erode or aggrade. is approach has little to no eect
on attenuating the hydrologic impacts from watershed
development. Consequently, when the restored reach ends,
and it must, the developed watershed hydrology exits the
restored reach, and enters the downstream, unrestored
channel. e unrestored channel may continue to degrade.
NCD projects localize benets associated with restora-
tion by stabilizing reaches and reducing the load of eroded
sediment to receiving waters. Do these localized benets
oset the costs associated with construction from both
an ecological and economic perspective? To maximize a
return on investment, a principle of ecological engineering
is to leverage feedback loops that lead to emergent proper-
ties and self-repair (Streb 2001). Moreover, benets associ-
ated with a restoration should strive to expand beyond the
project limits by attenuating the impacts of the developed
watershed.
Mitigating watershed impacts within a stream restora-
tion can be achieved by storing more runo along the ow
path through a combination of three tactics: increasing
ow path roughness, storing more water in the incised
channel, and increasing the connection with the ood-
plain so all appreciable runo events result in oodplain
inundation. ese three elements increase the area and
volume of aquatic habitat, reduce in-stream ow velocity
and channel erosion, reduce peak discharges, increase time
of concentration, raise groundwater elevation, restore wet-
land hydrology, restore oodplain functions, and improve
the diversity of the forest community and associated ora
and fauna.
Landscape Setting
e project described in this paper is in the coastal plain
physiographic province in Anne Arundel County, Mary-
land, within the South River watershed, downstream of
the Bacon Ridge Natural Area, an Anne Arundel County
Park (Figure 1). e restoration occurred within a 62ha
(154 acre) property owned by the Elks Club of America.
e property is predominately forest, and the majority of
work occurred along a forested oodplain that is used for
passive recreation.
Rather than the relatively level lands so common in the
mid-Atlantic coastal plain province, this area has a steep,
well-dissected, and highly erodible landscape. e streams
on the property were actively eroding from both bed inci-
sion and cross-section/bank widening (Figure 2). Historic
land use and stormwater runo from upstream residential
developments and roadways contributed to ongoing channel
degradation. e active incision and bank erosion through-
out the property resulted in the reduction of oodplain
function through decreased connectivity to stream ow
and associated overbank events along the main stem and
main tributary. It also drained groundwater (through stream
bank groundwater interception), which degraded oodplain
wetlands. e apparent high rates of bank and bed erosion
also reduced instream complexity and channel bed form
diversity, resulting in a mobile wedge of sediment along the
channel bottom. Sediment generated from stream bank ero-
sion and bed scour within was being exported downstream
and deposited in the larger volume and cross-section of the
receiving South River, degrading that resource.
e stream valley slope in the project area ranges from
a low of ~1.2% on the mainstem to more than 20% for the
1st order tributaries. e drainage area to the project is
2104ha (5,700 acre) or 23km (8.9mi). e land cover is
a mix of forest, large lot residential, and institutional, with
an impervious area of 4 to 6%, but with historic and cur-
rent pervious land impacts larger than the impervious area.
e project area included approximately 32ha (80ac) of
forested oodplain, ranging from poorly drained forested
wetlands along the valley side slopes to better drained,
degraded forested wetlands in the vicinity of the incised
stream channel.
Design Goals
e design goals for this project were to 1)modify the
hydraulics of the stream channel and valley to optimize
oodplain reconnection of storm ows; 2)improve geo-
morphic conditions of the stream channel to ensure long-
term bed and bank stability; 3)detain and slow stormwater
ows throughout the full width of the valley bottom, assist-
ing with the improvement in physiochemical functions and
water quality; and 4)create and enhance the ecological
functions of existing and historic non-tidal wetlands and
stream habitats and functions.
260 • December 2020 ECOLOGICAL RESTORATION 38:4
Figure 2. Representative pre-restoration images.
December 2020 ECOLOGICAL RESTORATION 38:4 • 261
Restoration of the two main perennial channels (main
stem and main tributary) was accomplished using a low-
impact approach that included engineered log structures
designed to raise baseow stage to just below the currently
abandoned oodplain surface, maintaining a baseow
water surface elevation that is just below the oodplain
elevation, restoring the stream and oodplain function. To
protect the integrity of this sensitive bottomland habitat,
valley-wide log sill structures were constructed to span
the oodplain and connect to instream engineered log
structures. e sill structures also limit the potential for
the development of channel avulsions and knickpoints. Use
of on-site wood for structures reduced the need to import
construction and lowered the risk of signicant impacts
related to construction access.
e design approach used for this project is a modi-
cation of the baseow channel design approach, which
involves building a stream channel that is sized to convey
the normal spring stream baseow or discharge. is is
accomplished by creating rie grade control cross sections
with a width and depth that are sized to convey the normal
spring discharge. Any appreciable increase in discharge
results in a stage increase which raises the water surface in
the stream above the oodplain elevation and the surplus
ow enters the oodplain.
is baseow channel design was selected as the pre-
ferred approach to achieve the project goals. e channel
cross section of the engineered log structures was sized to
convey the “normal” base ows in a channel with a high
surface area to volume ratio, a physical relationship asso-
ciated with eective material processing. During frequent
storm events that produce increased ows, the increased
water surface elevation spills out of the baseow channel
and onto the adjacent broad, forested oodplain. is
results in a loss of energy due to a slower, broad, shallow
ow through a complex forest with a comparable reduc-
tion in the channel adjustment and sediment entrainment.
Channel stability was enhanced by reducing ow depth and
in-channel shear stress via reconnection of frequent storm
ows to the full oodplain width, and the use of repetitive
grade control structures along the active channel owpath
to maintain the targeted bed elevation. e engineered log
structures were used at locations along the stream designed
in series to maintain a relatively continuous water surface
elevations with minimal plunging ows to support aquatic
species movement. Structure placement was selected to
minimize disturbance and promote continued oodplain
connectivity. Structures are located so the top elevation
“ties” into existing top of bank elevation with minimal
disturbance.
e project accomplished this oodplain connection
by constructing approximately 60 rie grade controls of
soil and wood harvested on-site. Each rie grade control
functions as a plug in the oversized channel, resulting in
the oversized channel lling with stream ow until the
water surface rises above the rie grade control invert and
continues downstream. is technique also minimizes the
groundwater drainage eect associated with the oversized
channel. is takes just a few minutes once the rie grade
control is installed and the ‘pump around’ is turned o.
Over time, bed load, suspended sediments, and vegetation
will “ll in” the oversized channel sections between the
rie grade control structures.
A log structure is engineered with specic dimensions
and top elevation to prevent channel degradation and
promote the development of specic hydraulic criteria as
described above. e location and relative spacing of each
structure are primarily based on the oodplain slope, so
that each ‘ties’ into the existing top of bank with minimal
disturbance of the oodplain. Other factors in alignment
location include avoidance of existing trees and use of exist-
ing “leaning” trees and side channel conuence locations
to support oodplain reconnection and stream channel
complexity.
Engineered log structures were constructed from tar-
geted trees within the project area that were harvested near
the location of installation. All elements of the felled tree,
including the root wad, trunk, and smaller branch material,
were used in the construction of the engineered log struc-
tures. For each engineered log structure, an estimated three
to seven trees were needed, depending on the channel size.
Trunk material was used as foundation logs crossing
the channel and keyed into the streambank to provide
rigid support for the blockage. Vertical, inverted root wad
keyed into the channel bottom provided further stability
and increased blockage due to their physical complexity.
At each structure, vertical posts were used to pin the key
logs and root wads together and provide further blockage
elements and attachment points for the brush mattress. A
brush mattress core consisting of smaller branch material
was packed into voids between key logs, root wads and
vertical posts. Soil material generated from top of bank
grading of adjacent pool areas was added to the brush
mattress core, which was installed in lis. e upstream
and downstream side of each structure was blocked using
brush mattress and excavated soil to further choke the
structure and encourage water surface increases over top
of each structure and to minimize plunging ows between
structures (Figure 3).
Long-term protection of the reconnected oodplain
was provided primarily by maintaining the existing ood-
plain vegetation. is was accomplished by minimizing
disturbed areas and using appropriately sized smaller,
low ground pressure construction equipment. In any dis-
turbed areas, permanent seeding and plant installation
was included to reestablish plants intended to develop
thick root density, provide good ground cover, and protect
slopes from soil erosion during out of bank ow events. To
provide additional, redundant protection along the ood-
plain surface, oodplain log sills were installed in locations
262 • December 2020 ECOLOGICAL RESTORATION 38:4
Figure 3. Representative engineered wood structures.
December 2020 ECOLOGICAL RESTORATION 38:4 • 263
where there was a higher risk of concentrated ow paths
forming. Sills consist of buried logs (minimum 25cm [10" ]
DBH) ush with or just above the local oodplain surface
to protect against scour and provide oodplain roughness.
Supplemental live posts installed along the sills pinned logs
to the oodplain surface to minimize oatation. ese
structures protect against the formation of headcuts along
secondary owpaths which we anticipate will develop into
an anabranching channel network over time.
Discussion
e most innovative element of this design approach is the
use of materials harvested from the project area. Normally,
a stream restoration project of this magnitude would have
required approximately 12,000m (15,000 CY) of stone be
imported and placed, requiring at least 2000 truck trips to
and through the project area, with signicantly more dis-
turbance in the stream valley, trac on the roads around
the project site, and fuel consumed, and at a signicantly
higher cost in dollars and carbon. Instead, the project
required a small construction crew with two excavators,
~20% of the standing stock of the forest to be harvested,
and buried in a saturated soil environment, and limited
channel grading and disturbance. One growing season
aer construction, the forested oodplain condition is such
that it is dicult to see signs that hundreds of trees were
harvested from the site (Figure 4). e earth and wood
rie structures appear as natural landscape features, unlike
many of the rock design elements that have no parallels in
the coastal plain. Within two weeks of construction com-
pletion and demobilization, anadromous yellow perch were
running upstream and spawning in the restored stream.
is restoration project is also interesting in that in
addition to improving aquatic and wetland resources, the
restored reach is now delivering ecosystem functions and
services it had not provided in decades due to its pre-resto-
ration degraded conditions. e reconnection of the stream
to its oodplain restored lost oodplain functions known
to be important, including water quality enhancement
and ood attenuation. Prior to its restoration, this stream
and its oodplain were in a conservation easement, but
that status conferred little benet to the resource because
the watershed hydrology responsible for its degradation
Figure 4. Representative before (top panel) and after (bottom panel) restoration conditions.
264 • December 2020 ECOLOGICAL RESTORATION 38:4
continued to adversely impact the resource. While plac-
ing resources in a conservation easement may protect the
resource from specic development threats, such ease-
ments have little eect on the slow, chronic degradation
initiated by o-site land use.
All stream restoration projects in the mid-Atlantic
region have a requirement for ve years (or more) of
post-construction monitoring as a condition of Army
Corps of Engineers and State permits. Monitoring typi-
cally includes inspections of the placed structures to be
sure they are working, evaluation of the plant community
to be sure installed plant material is thriving and undesir-
able invasive species aren’t taking over, as well as a variety
of project-specic monitoring conditions (e.g., aquatic
life abundance and diversity, increasing indices of biotic
integrity), all of which must be regularly documented in
monitoring reports submitted annually over the course of
the monitoring period.
is project approach has broad applicability. Aer all,
the beaver used this type of structure and construction
technique to control streams and stream valleys across all
of North America. is design and construction approach
can be adapted to a variety of stream valley conditions, if
the project areas are in forest cover and there is condence
that they will remain in forest cover. Obviously, areas which
have been cleared of forest cover would not be good can-
didate sites for this type of restoration.
Acknowledgments
anks to Amy Nelson for her help in reviewing and editing
this paper, as well as all the Biohabitats sta that worked on this
project and made it possible. Also, the sta at Greenvest and EQR
deserve much of the credit for making this project possible as a
result of their funding and building this project (respectively).
References
Bureau of Reclamation and U.S. Army Corps of Engineers Research
and Development Center 2016. National Large Wood Manual:
Assessment, Planning, Design, and Maintenance of Large Wood
in Fluvial Ecosystems: Restoring Process, Function, and Structure.
Pollock, M.M., G. Lewallen, K. Woodru, C.E. Jordan and J.M.
Castro. 2015. e Beaver Restoration Guidebook: Working with
Beaver to Restore Streams, Wetlands, and Floodplains. Version
1.0. Portland, OR: United States Fish and Wildlife Service.
Pollock, M.M., T.J. Beechie, J.M. Wheaton, C.E. Jordan, N. Bouwes,
N. Weber and C. Volk. 2014. Using beaver dams to restore incised
stream ecosystems. Bioscience 64:279–290.
Rosgen, D.L. 1997. A Geomorphological Approach to Restoration of
Incised Rivers. In S.S.Y. Wang, E.J. Langendoen and F.D. Shields
(eds.), Proceedings of the Conference on Management of Land-
scapes Disturbed by Channel Incision. Oxford, MS: University
of Mississippi.
Streb, C.A. 2001. Woody Debris Jams: Exploring the Principles of
Ecological Engineering and Self-Design to Restore Streams. MS
esis. University of Maryland, College Park, MD.
Wheaton J.M., S.N. Bennett, N. Bouwes, J.D. Maestas and S.M.
Shahverdian (eds). 2019. Low-Tech Process-Based Restora-
tion of Riverscapes: Design Manual. Version 1.0. Logan, UT:
Utah State University Restoration Consortium. doi:10.13140/
RG.2.2.19590.63049/2.
Joe Berg (corresponding author), Biohabitats Inc,
2081 Clipper Park Road, Baltimore, MD 21211,
jberg@biohabitats.com.
Doug Streaker, Biohabitats Inc, Baltimore, MD.
Chris Streb, Biohabitats Inc, Baltimore, MD.
Eupatoriadelphus maculatus. USDA-NRCS PLANTS Database. Britton, N.L. and A. Brown. 1913. An Illustrated Flora of
the Northern United States, Canada and the British Possessions. New York, NY: Charles Scribner’s Sons.
December 2020 ECOLOGICAL RESTORATION 38:4 • 265
Ecological Restoration Vol. 38, No. 4, 2020
ISSN 1522-4740 E-ISSN 1543-4079
©2020 by the Board of Regents of the University of Wisconsin System.
ECOLOGICAL DESIGN COMMENTARY
From Dead Wood to Living Landscape
Stacy Passmore
While “dynamic landscapes” are popular concepts in
academic settings, in reality most designers and
clients are uncomfortable working with living and chang-
ing matter in the landscape, gravitating to concrete and
stone that can be xed with minimal weathering. As the
plant material in our landscapes grows, moves, and changes
more predictably based on known biology and environ-
mental factors, the idea that we might install an engineered
structure designed for disturbance and decay is dicult to
comprehend and more dicult to justify, but increasingly
important. Ecologists Berg, Streaker, and Streb with Bio-
habitats describe such a system built in a Maryland forest,
using techniques that are gaining favor in application:
locally harvested timber-made structures that mimic the
dam building of beavers and natural tree-fall in the forest
by creating hydrological barriers across the landscape. In
recent years, much interest has developed around the engi-
neering work of beavers and their role in the restoration of
hydrological landscapes. Among the many brilliant scien-
tists and writers contributing to this work, two stand out,
1)Glynnis Hood, a Canadian ecologist who documents
the recovery of the North American beaver in her book
“e Beaver Manifesto” (2011) and their potential role in
mitigating and climate change, and 2)Ben Goldfarb, who
proles the “Beaver Believer” movement (2019), interview-
ing a diverse range of advocates, scientists and land manag-
ers working to return the beaver to their historic habitat.
I have also studied beaver landscapes across the Western
Slope of the Rocky Mountains (Passmore 2019), visiting
around 30 beaver wetlands in New Mexico, Colorado, Utah
and Idaho, to learn from them about bringing water back
to arid landscapes.
Like beaver dams, these wood structures designed by
Biohabitats support increased water retention, ecosystem
enhancements and broader oodplain health by creating
micro-conditions where blockages can occur. By mimick-
ing what would have been a messier, more complex forest
oor condition pre-development, they are engaging with
the aliveness of the forest and the wood itself. In this, they
connect an engineered solution directly to the character
and materiality of this coastal forested oodplain, where a
traditional rock dam or weir would be foreign. Although
the wood structures will surely transform, break, and rot,
it is this lightness, soness, and weakness of wood that
makes it so eective. Sourced from the immediate site, it
can be transported easily and fabricated and installed using
low-tech and cost-eective methods. is work could be
done in conjunction with forest thinning initiatives or
maintenance programs that address broader forest health.
Where I work in the West, similar beaver dam analog sys-
tems (“post-assisted log structures”) have been essential
to post-re landscapes to mitigate erosion and the loss of
sediments and nutrients (Shahverdian et al. 2019).
Biohabitat’s approach to using raw timber will be most
eective in forested contexts, and therefore may need
to be adapted depending on the surrounding landscape.
Beaver also build with what is available nearby; along
Western streams I observed dams built with willow (Salix
spp.), cottonwood/aspen (Populus spp.) and even sagebrush
(Artemisia tridentata). Conifer species such as spruce or
pine are available, but less palatable to beaver. ese west-
ern woods tend to be soer, more porous and pliant; they
can achieve strength in numbers but may decay quickly.
Biohabitats’ structures, if built using these materials, would
likely require more maintenance or have a shorter life span.
If regularly maintained, like beaver dams, I would expect
that this system would last for decades, but ultimately
depends on how “ashy” or prone the watershed is to large
destructive ash ood events.
Biohabitats’ system in its context is remarkable because it
also has the potential to regenerate independently, though
chaotically. As the forest continues to shed branches and
detritus, new wood will feed the dams as it is caught by the
structures. Over time the landscape will take on a life of its
own, a truly living system. Do we accept this uncertainty,
develop models for prediction, or create a stewardship
program to control the outcome? It is the unpredictability
and maintenance required that may make it dicult to
convince a municipality or private client to pay for this type
of structure. Our technical drawings are geared towards
static design elements with minimal follow up; site fea-
tures are expected to be built and planted exactly as they
are drawn. To be successful over time, Biohabitats’ project
must be accompanied with a plan for the tending and care
of the structure, just as a beaver continuously adds waddle
and daub to their dam. We can also expect Biohabitat’s
project to generate dramatic geomorphological changes;
266 • December 2020 ECOLOGICAL RESTORATION 38:4
the authors mention the potential for the emergence of an
anabranching stream network, and this may be the most
radical and valuable outcome of all, a new braided stream
that weaves through the landscape with complexity and
variation, reforming the land as it ows. Yet this also has the
potential to develop land use conicts, depending on the
site. A successful example of a landscape designed for ux,
Swiss landscape architect Georges Decombes’ project “e
Renaturation of the River Aire” (Descombes et al. 2018)
creates a similar condition, controlled within the limits of
a canalway, his design built sandy diamonds that morph
and move drawing new lines as waters ow.
From the perspective of performance and sustainabil-
ity, the Maryland stream restoration is beautifully low-
impact—yet almost too subtle—for these practices to
become more understood and accepted they must be active
spaces that are seen and maintained—people must engage
with them. is might include educational programs where
community members visit the forest to work like the bea-
vers, adding wood and playing with the structures. e
timber posts and horizontal log structures could be exag-
gerated in size to be more visible when water levels are
high. Contemporary practices in nature playscape design,
or even Victorian stumperies that incorporate deadwood
and tree stumps, may also oer clues for integrating resto-
ration practices that have both environmental and human
attributes. Biohabitats shows us that our future requires
that we think like non-human species such as beavers
who work like stewards in the landscape, engaged with
continual processes of growing ecosystems, knowing that
landscapes are not xed in place.
References
Goldfarb, B. 2018. Eager: e Surprising, Secret Life of Beavers and
Why ey Matter.White River Junction, VT: Chelsea Green
Publishing.
Hood, G. 2011. e Beaver Manifesto. Victoria, Canada: Rocky
Mountain Books.
Passmore, S. July 2019. Landscape with Beavers. Places Journal. doi.
org/10.22269/190709.
Shahverdian, S., S.N. Bennett, and J.M. Wheaton. 2018. Baugh Creek
Post-Fire Emergency Restoration: Prioritizing and Planning
Post-Fire Restoration in the Baugh Creek Watershed. Anabranch
Solutions, Technical Report. doi.org/10.13140/RG. 2. 2. 17537.
71528.
Descombes, J., C. Van Cauwenberghe, V. Correnti , F. Gerber (eds).
2018. AIRE: e River and its Double. Zurich, Switzerland: Park
Books.
Stacy Passmore is the Harvard University Graduate School
of Design’s 2018 Charles Eliot Traveling Fellow and currently
works with Civitas, Inc., 1200 Bannock Street, Denver, CO
80204, laurastacypassmore@gmail.com.
Broken trees. Warren, G.F. 1913. Elements of Agriculture. London, UK:
MacMillan and Co., Ltd. The Florida Center for Instructional Technology,
fcit.usf.edu.
December 2020 ECOLOGICAL RESTORATION 38:4 • 267
Ecological Restoration Vol. 38, No. 4, 2020
ISSN 1522-4740 E-ISSN 1543-4079
©2020 by the Board of Regents of the University of Wisconsin System.
ECOLOGICAL DESIGN COMMENTARY
Applying Design Lessons from
an Ecosystem Engineer
Ken P. Yocom
As a species, Castor canadensis, otherwise known as
the North American beaver, has had a complex and
oen controversial relationship with people. Highly favored
for their pelts during colonial expansion in the U.S. and
Canada and seen as pests, a hindrance to settlement, during
the 19 and 20 centuries, they were eradicated from many
parts of the continent. eir remaining populations were
isolated and by the turn of the 21 century only about 10%
of their pre-colonialization population estimates remained
(Wilson and Reeder 2005).
However, since the mid 20th century research has shown
the importance of C.canadensis as a keystone species in
riverine and coastal environments (Boswell etal. 1988).
is recognition coupled with a decreasing lack of public
support for programs that trap and kill or relocate individu-
als has contributed to a re-establishment and resurgence
of beaver populations in areas that would, for many, seem
unlikely. For example, in 2007, a beaver was sighted in
New York City’s waterways, the rst in more than 200 years
(O’Connor 2007). In Seattle, beaver have re-established
their presence in nearly 90% of suitable habitat within the
city (Dittenbrenner et. al. 2018).
An industrious and persistent “ecosystem engineer”, the
activities of the beaver are now recognized to provide many
ecological benets and ecosystem services. By impound-
ing water and sediment through the construction of dams
and weirs in uvial systems, beaver can directly improve
degraded hydrologic regimes and geomorphology while
enhancing habitat for native plant and animal communi-
ties. In regions where beaver have the potential to rees-
tablish, environmental designers have begun to actively
plan and design to facilitate such opportunities (Bailey
et. al. 2018). Yet, in other areas where beaver populations
have been entirely eradicated designers are emulating the
structures of beaver to improve riverine systems and mimic
historical ecological conditions within the region. e
project described by Joe Berg, Doug Streaker, and Chris
Streb of Biohabitats in the South River watershed of Anne
Arundel County, Maryland serves as case in point.
Grounded in current best practices for stream rehabilita-
tion, the project utilized a scaold system of grade control
structures to establish rie-pool sequences that would
maintain baseows and discharge into the oodplains
during high ow events. ey specically utilize materials
such as trees, rock, and woody shrubs sourced from the
site, as the beaver would have done, to reduce costs and
construction impacts. e use of large woody debris (LWD)
in stream rehabilitation is common practice serving to not
only enhance and diversify localized conditions, but further
mitigate and improve hydrological impacts of the system
as a whole by forcing ows to slow, spread, and soak into
the oodplains. ough some consider the use of LWD
to be unsightly initially, when used in combination with
strategies such as bank and oodplain plantings and live
staking of water-thriving, fast growing woody vegetation,
they improve conditions for the future growth of native
riparian vegetation. While past practices to control ows
using concrete and boulder-lined channels are still used
and provide benet to protect infrastructure such as road
crossings and water and waste pipelines, such strategies are
now understood as detrimental the larger system, channel-
izing ows and leading to channel degradation. rough
proposed monitoring and adaptive management practices
the scaold of initial bed structures incorporated into this
project will eventually break down, to be replaced by the
trees and plant communities that are just now establishing.
is project, and many others like it around the country,
are furthering the potential for environmental designers
working in degraded streams, rivers, and other habitat types
to draw understanding from species like C. canadensis to
establish design management practices that are regionally
grounded, ecologically responsive, and adaptive to future
changes in conditions.
References
Bailey, D.R., B.J. Dittbrenner and K.P. Yocom. 2018. Reintegrating
the North American beaver (Castor canadensis) in the urban
landscape. Water 6:e1323.
Boswell, G.P., N.F. Britton and N.R. Franks. 1988. Habitat fragmen-
tation, percolation theory and the conservation of a keystone
species. Proceedings of the Royal Society B: Biological Sciences,
265:1921–1925.
268 • December 2020 ECOLOGICAL RESTORATION 38:4
Dittbrenner, B.J., M.M. Pollock, J.W. Schilling, J.W. Olden, J.D. Lawler
and C.E. Torgersen. 2018. Modeling intrinsic potential for beaver
(Castor canadensis) habitat to inform restoration and climate
change adaptation. PLoS One, 13:e0192358.
O’Connor, A. 2007. Aer 200 years, a beaver is back in New York
City. e New York Times. February 23. www.nytimes.com/ 2007/
02/ 23/ nyregion/23beaver.html.
Wilson, D.E. and D.M. Reeder. (Eds.) 2005. Mammal Species of the
World: A Taxonomic and Geographic Reference. 3rd ed. Balti-
more, MD: Johns Hopkins University Press.
Ken P. Yocom, Department of Landscape Architecture, Uni-
versity of Washington, 348 Gould Hall, Box 355734 Seattle,
WA 98195, kyocom@uw.edu.
Beaver. Hooker, W. 1882. Natural History. New York, NY: Harper & Brothers. The Florida Center for Instructional
Technology, fcit.usf.edu.
