Journal of

Ecotechnology
Vol.2 No.2 December 2006

Water Environment Research Center

www.ntut.edu.tw/~wwwwec/

Water Resources Agency, Ministry of Economic Affairs, ROC

University of Virginia, USA

National Taipei University of Technology, ROC

Water Environment Research Center

 Content
Preface Shaw L. Yu Ι
News Brief
Ecological Engineering in Taiwan－ An Assessment of Progress
台灣生態工程評估 Shin-Hwei Lin 1
Ecotechnology at work－A success story at Guandu Nature Park in Taipei
生態工程－台北關渡自然公園 4
Educational and Public Outreach Activities－ Visit to the Fei-tsuei Reservoir Watershed
水庫集水區教育宣導活動－翡翠水庫參觀 6
2006 International Symposium on Ecological Engineering
2006 生態工程國際研討會 8
“Climate Change, Taiwan Cares” Festival and Parade
“氣候變遷，台灣在意＂反全球暖化大遊行 Mitchell Tsai 9
Technical Paper
Bridging the Gap Between BMP Effectiveness and Receiving Water Quality Protection
T. Andrew Earles, Jane K. Clary, Rachel Pittinger, David B. Mehan 10
Tools and Processes for Implementing a Road Ecology Approach to Transportation and the
Environment
Alison M. Berry, Katie Benouar, Sandra L. Jacobson, Carson Poe, and Victor Claassen 31
Modeling the Relationships Between Benthos and Stream Environmental Variables
Lilian D. Minja (Corresponding author), Shaw L. Yu, Howard Epstein 48

Publisher Tsu-Tian Lee

Circulation Dep. Water Environment Research Center, NTUT.

Chief editor Shaw-Lei Yu, UVa, USA

Editorial consultants Jen-Yang Lin, NTUT, ROC.

Cheng-Dao Hsieh, WRA, ROC.

Executive Chi-Chang Liu, Wei-Tsung Liao

Telephone No. +886-2-2771-2171 ext 2664

FAX +886-2-87719190

Website http://www.ntut.edu.tw/~wwwwec
 Preface
J. of Ecotechnology, 2006 (2)

Preface

This issue marks the end of the second year of existence of the Journal of Ecotechnology
(JET). The publication of the JET is pert of the research activities of the Water Environment
Research Center (WERC) of the National Taipei University of Technology (NTUT). WERC was
established in 1999, with major funding from the Water Resources Bureau (predecessor of the
current Water Resources Agency or WRA), and collaboration from NTUT and the University of
Virginia (UVA). Since its inception, WERC has made ecological engineering one of the focal
points of its research and outreach, as well as international collaborative activities.

Launched in 2005, JET was designed to provide a platform for professionals to obtain and
dissimilate information on ecological engineering. Presently the Journal contains a news brief
section and a technical paper section. The eventual goal is for the Journal to become a full-fledged
scientific journal and attain SCI recognition.

The mission of WERC, with JET as one of the implementation tools, is described in a broader
sense by the American Ecological Engineering Society’s (AEES) mission statement:

To promote the development of sustainable ecosystems that integrate human society with its
natural environment for the benefit of both by fostering education and outreach, extending
professional development and associations, raising public awareness, and encouraging original
research.

In the past few years, the application of ecological engineering methods to water and
environmental projects has received much attention in Taiwan. The government essentially requires
all water resources engineering projects be considered for potential ecological engineering
applications. Leading the way is the Water Resources Agency, which funded several large projects
that included ecological engineering method applications. Similar studies have been funded by the
Council of Agriculture and the Environmental Protection Administration. To promote a dialogue
among local and international ecological engineering professionals, the Public Construction
Commission sponsored a hugely successful major international conference on ecological
engineering, organized by NTUT, in Taipei in 2004. The second such conference was just held in
Taipei earlier this month.

The successful application of ecological engineering in water/environmental projects requires a

close coordination between ecologists and engineers. In Taiwan, there has been increasingly more

I
 Preface
J. of Ecotechnology, 2006 (2)

such coordination in recent years, thanks to the great efforts exercised by governmental agencies
and academic groups. It is our sincere hope that the trend will continue and that the principles of
ecological engineering will be considered in all engineering applications.

Shaw L. Yu
November 19, 2006
Reston, Virginia

II
 News Brief
J. of Ecotechnology（2）2

Ecological Engineering in Taiwan

－ An Assessment of Progress

Shin-Hwei Lin1

Taiwan has begun to promote the application of ecological engineering since 1991.
Governmental agencies as well as academic groups have been aggressively developing and
searching for engineering technologies that can satisfy ecological requirements. Among
governmental agencies, the Soil & Water Conservation Bureau of the Council of Agriculture (COA)
has implemented many ecological engineering projects, including stream habitat improvement,
fish-passage design, landscaping and vegetative cover for villages, detention pond and debris flow
damaged area ecological engineering design, and establishment of reference and description
materials, etc. The Forest Bureau of COA has also supported R & D studies on ecological
engineering methods applied to forest roads and on fish-passage and wooden structures in lumber
areas.

The Water Resources Agency (WRA) of the Ministry of Economic Affairs has sponsored a
substantial number of studies on the application ecological engineering methods in water resources
engineering and has incorporated these methods in many water projects, such as non-traditional
levee design and restoration works in flooded areas. At the Department of Transportation, the
Highway Bureau has assessed the application of landscaping and ecological corridor design in road
projects. The Council of Public Works has held ecological engineering fairs; international
conferences on ecological engineering, and established a category of ecological engineering for its
public works gold medal competition, which is designed to honor those associated with good
ecological engineering applications, and to enhance the enthusiasm of participation from
governmental agencies and common citizens.

In the academic arena, much information on environmental – ecological indices, biological

indices, and composite environmental indices have been generated through research studies,
conferences, and journal publications. This information provides a quantitative basis for comparing
before and after the implementation of ecological engineering projects, which is needed in the

1
Professor, Department of Soil & Water Conservation, National Chung-Hsing University, Taichung, Taiwan, ROC

1
 News Brief
J. of Ecotechnology（2）2

evaluation of the cost-effectiveness of such projects.

There have been successes as well as failures in the application of ecological engineering
methods in Taiwan. Failures usually are associated with engineering design problems, location of a
project and improper construction materials. In the future, we need to assess the success and failure
cases in order to find ways of improving ecological engineering applications. Regional design and
specification guidelines need to be established. A set of ecological engineering design principles
and implementation protocols, which differ from the traditional civil and construction guidelines
and protocols, should be developed. By compiling all the information on ecological engineering
applications in Taiwan and abroad, we can establish a “localized” framework for ecological
engineering applications in Taiwan.

(Translated and edited by Shaw L. Yu)

2
 News Brief
J. of Ecotechnology（2）2

台灣自 1991 年起開始推行生態工程，政府機關與學術團體積極研發與探索合乎生態需求

之工程技術，在政府機關方面，行政院農業委員會水土保持局辦理多項生態工程計畫，內容
包含野溪棲地改善、魚道設計、農村景觀植被、滯洪池與土石流災區生態工程設計、參考圖
說資料建置等項目；行政院農業委員會林務局亦委託進林道生態工法、林班地魚道與木構造
物之研發；經濟部水利署進行非規格化堤防與與易淹水地區整治技術之研發；交通部公路總
局進行景觀與生態廊道設計與道路應用案例評估；行政院公共工程委員會推動生態工程博覽
會、國際研討會與公共工程金質獎(首次增列生態工程類)選拔，獎勵優良生態工程之執行單
位與相關人員，提高政府部門與民間參與之熱忱。學術團體方面，許多關於環境生態指標、
生物指標及環境綜合指數之研究陸續進行研討會與論文之期刊發表，可提供生態工程改善前
後環境品質改變之量化數據，以利工程成效評估。

生態工程之推行至今，有成功也有失敗之案例，失敗案例是工程設計問題、設置地點錯
誤或是資材使用年限到期? 未來仍需要將成功與失敗實例進行探討與改進，進行不同區域環
境提出制式化之規範，且對於生態工程設計監造需另建立有一套異於一般土木或傳統工程之
標準作業流程，將本土化生態工程資訊整合，以擴大台灣生態工程之格局。

3
 News Brief
J. of Ecotechnology（2）2

Ecotechnology at work－
A success story at Guandu Nature Park in Taipei

I
n northern Taiwan at the junction of Damchui and Jilong Rivers there is a heavenly place for
migrating birds that it’s called the “Guandu Nature Park”. Guandu is a major stopover site for
migrating birds, especially waterfowls and shorebirds, as well as an important wintering and
breeding ground for many species. Two hundred twenty-nine species of birds have been recorded
at Guandu so far, qualifying this wetland complex as an Important Bird Area (IBA) recognized by
BirdLife International.

In 1983 the government created the "Guandu Waterbird Refuge", and the Tourism Bureau
classified Guandu as a major tourist attraction. The Nature Park covers 57 hectares divided into the
Main Area, Core Reserve Area, Outdoor Observational Areas, and the Sustainable Management
Area, as shown in Figure 1. Figure 2 depicts an overview of the Park.

Figure 1. Park Map(http://www.gd-park.org.tw/en/e1.htm)

On December 1, 2001, administration of Guandu Nature Park was handed over to the Wild
Bird Society of Taipei, a Non-Governmental Organization (NGO) with thirty years of experience in
conservation and environmental education. Guandu Nature Park thus became the first protected
area in Taiwan to be managed by an NGO. As a non-profit organization, the Wild Bird Society of

4
 News Brief
J. of Ecotechnology（2）2

Taipei channels 100% of the income generated by Guandu Nature Park back into the management
of the park. With this unprecedented action, the Wild Bird Society of Taipei aims to combine the
resources and energy of both governmental and public interests and recreate a healthy wetland
ecosystem for the purposes of both conservation and environmental education (see Figure 3).

Figure 2. Overview of Park Figure 3. Field class for kids

(http://www.gd-park.org.tw/en/e1.htm) (http://www.gd-park.org.tw/en/e1.htm )

Reference:

http://www.gd-park.org.tw/en/e1.htm

5
 News Brief
J. of Ecotechnology（2）2

Educational and Public Outreach

Activities
－Visit to the Fei-tsuei Reservoir Watershed

Date: Septerber/23th/2006

Destination: Fei-tsuei Reservoir and the Jr-tan Water Treatment Plant, Taipei.

Sponsored by: Water Resources Agency, Ministry of Economic Affairs

Water Environment Research Center, NTUT

Hosts: Bureau of Fei-tsuei Reservoir Administration,

Taipei City Government Taipei Water Department

Participants: Graduate students, Seniors majoring in environment, civil, hydraulic engineering,

and interested citizens.

Table1. Activities schedule

Time Activity Location

08：00～08：30 Assemble NTUT
10：00～10：30 Introduce Fei-tsuei Reservoir Administration of Fei-tsuei Reservoir
10：30～11：30 Tour of dam Fei-tsuei Reservoir
11：30～13：00 Lunch Jr-tan Water Treatment Plant
Introduce Jr-tan Water
13：30～14：00 Jr-tan Water Treatment Plant
Treatment Plant
Visit to Jr-tan Water
14：00～15：00 Jr-tan Water Treatment Plant
Treatment Plant
15：00～16：00 Conclusion NTUT

6
 News Brief
J. of Ecotechnology（2）2

Figure 1. Educational tour to the Fei-tsuei Reservoir

Figure 2. Educational tour of the Jr-tan Water Treatment Plant

7
 News Brief
J. of Ecotechnology（2）2

2006 International Symposium on

Ecological Engineering

Date:11/9th ~10th/2006 （Thursday～Friday）

Sponsored by: Public Construction Commission, Executive Yuan

Organized by: The Ecological Engineering Research Center, National Taiwan University (NTU),
Taiwan

Co-Sponsored by: Department of Bioenvironmental Systems Engineering, NTU, Taiwan

Chinese Institute of Civil and Hydraulic Engineering

Taiwan Geotechnical Society

Chinese Society of Agricultural Engineers

Genesis Group, Taiwan

Description:

The conference was the second of such meeting sponsored by the Public Construction
Commission. The first one, in 2004, was organized by NTUT and was a huge success.

Reference:

The objective of the 2006 conference was to provide a forum for invited scholars and experts,
who are professionals in ecological engineering and related fields, from Europe, America and Asia
to report on their experiences and recent advances in ecological engineering, especially in the areas
of stream restoration and road construction ecology. The exchange between experts from abroad
and Taiwan will be helpful in the integration of ecological principles with engineering practices in
Taiwan. The conference was very successful. The invited papers are published in a book. A
Conference Proceeding CD is available.

Reference:
http://203.64.157.65/2006isee/news.asp

8
 News Brief
J. of Ecotechnology（2）

“Climate Change, Taiwan Cares”

Festival and Parade

Mitchell Tsai1

On November 4, 2006 people from 45 countries simultaneously demanded action from global
leaders to stop climate change. In Taiwan more than 500 people gathered together at Daan Forest
Park in Taipei for the “Climate Change, Taiwan Cares” Festival and Parade
(http://blog.yam.com/climatechange). The festival featured skits, musical performances, informa-
tion booths, and the sale of environmentally-friendly goods. Parade participants carried umbrellas
as well as dressed as bicyclists, animals, and trees promoting environmentally friendly lifestyles.

The event was held by a coalition of more than 11 Taiwanese environmental groups, including
Taiwan Environmental Action Network, Society of Wilderness, Taiwan Green Party, Homemaker’s
Association, Green Citizen Action Alliance, Wild at Heart Legal Defense Association, and Taiwan
Environmental Protection Union. The groups demanded that Taiwan implement Kyoto Protocol
requirements stipulating that CO2 emissions be brought to 5% below year-1990 levels by
2012-2016. The groups also want the government to set long-term targets for CO2 emissions at
50% below 2005 levels by 2050.

Despite the island’s political and diplomatic isolation, the country is increasing its participa-
tion in international climate change instruments. Taiwan Environmental Action Network recently
joined Climate Action Network (http://www.climatenetwork.org), a coalition of over 365 interna-
tional non-governmental organizations working collectively on climate change issue. This will give
an opportunity for Taiwanese environmental groups to exchange information and work collectively
with other NGOs around the world to address climate change.

1
Taiwan Environmental Action Network-Campaign Coordinator

9
 Technical Paper
Bridging the Gap Between BMP Effectiveness and Receiving Water Quality Protection

Bridging the Gap Between BMP Effectiveness and Receiving Water Quality
Protection

T. Andrew Earles1, Jane K. Clary2, Rachel Pittinger3, and David B. Mehan4

1
Water and Civil Engineer, Environmental Scientist, Ph.D., P.E., D.WRE., Project Engineer and Wetlands
Scientist, respectively. Wright Water Engineers, Inc., 2490 West 26th Avenue, Suite 100A, Denver, Colo-
rado, USA 80211. Phone: (303) 480-1700, Fax: (303) 480 1020, Corresponding Author’s email: aear-
[email protected]
2
Water and Civil Engineer, Environmental Scientist, Project Engineer and Wetlands Scientist, respectively.
Wright Water Engineers, Inc., Denver, Colorado, USA.
3
Water and Civil Engineer, Environmental Scientist, P.E., Project Engineer and Wetlands Scientist, respec-
tively. Wright Water Engineers, Inc., Denver, Colorado, USA.
4
Water and Civil Engineer, Environmental Scientist, P.W.S., Project Engineer and Wetlands Scientist, re-
spectively. Wright Water Engineers, Inc., Denver, Colorado, USA.

Abstract

Since the early 1990s when Phase I of the National Pollutant Discharge Elimination System
(NPDES) Stormwater Regulations went into effect in the United States, significant research has
been conducted to evaluate the effectiveness of stormwater best management practices (BMPs).
This type of research has proliferated since the adoption of Phase II NPDES Stormwater Regula-
tions in 2003 since many smaller municipalities, unregulated under Phase I, are now required to
have NPDES permits for stormwater discharges and face the challenges of selecting BMPs that are
effective for management of non-point source (NPS) pollution and protection of receiving water
quality.

Significant progress has been made in assessing BMP effectiveness and understanding the un-
derlying physical, chemical and biological processes that remove or transform pollutants in storm-
water. Engineers now have reasonable expectations of BMP pollutant removal and effluent water
quality given a targeted parameter (often total suspended solids [TSS]), anticipated BMP loading
from a defined tributary area, BMP design criteria and/or other factors; however, determining the
effects of multiple BMPs on the chemistry, biology and physical characteristics of receiving waters
at the development- or watershed-scale is less well documented.

This paper presents an approach to evaluating the effects of BMPs on receiving water quality
by examining the expansion of the Copper Mountain Base Area in Summit County Colorado, USA.

10
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.10-30

The Copper Mountain Base Area expansion project provides an example of how careful planning,
implementation and monitoring of water quality protection strategies consisting of structural and
non-structural BMPs can be effective at protecting receiving waters from the effects of development.
Water quality and biological monitoring data were collected before, during and after construction
for West Tenmile Creek, which runs through the heart of the base area expansion project. The re-
sults demonstrated that the creek has maintained excellent water quality, habitat quality and aquatic
life, and that BMPs for the Copper Mountain Base Area expansion project have been effective at
protecting receiving water quality.

Keywords: NPDES; Stormwater; BMPs; TSS; Water Quality Protection

11
 Technical Paper
Bridging the Gap Between BMP Effectiveness and Receiving Water Quality Protection

BMP 效率與水質保護關係之建立

自 1990 年代初，美國推動全國污染排放消除系統（NPDES）中有關雨水第一梯次之管
制辦法以來，已經有相當多針對最佳管理措施（BMP）去污效率探討之研究。尤其從 2003
年第二梯次管制辦法實施後，因為人口十萬以上的城市皆列入管制範圍，而需要申請雨水排
放許可證，並對非點源污染及 BMP 之使用必須執行，所以自 2003 年以來，全國有更多對 BMP
效率評估之研究。

目前，經由對 BMP 效率多年以來之研究成果。一般而言，對 BMP 處理雨水逕流帶來污

染物之機制，如物理、化學或生物性等，已有相當之瞭解，BMP 對某些指定之污染物（通常
含總懸浮固體物，TSS）的去除率，以及相關之設計準則等也有甚多的文獻報導。相對而言，
對於以一集水區整體為考量，區內 BMP 之使用對水質改善之評估之文獻資料則甚少。

本文即針對集水區內整體性 BMP 之使用，對區內水質改善之效果做一探討。選擇科羅拉

多州丹佛市附近之 West Tenmile Creek 做一實例之研究。該流域有一銅山開發區，本研究在該
開發行為前、中以及完成後作水質監測及分析，結果證明該流域在全面推動 BMP 以後，對
West Tenmile Creek 之水質及棲地保護有極優良的效果。

關鍵詞：污染排放消除系統、雨水逕流、最佳管理作業、總懸浮固體物、水質保護

12
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.10-30

Introduction and Background WERF and others provide a wealth information

on BMP performance and design; however, a
Since the early 1990s when Phase I of the
significant knowledge gap still exists in deter-
National Pollutant Discharge Elimination Sys-
mining how BMP performance relates to re-
tem (NPDES) Stormwater Regulations went
ceiving water quality (WERF 2006). Data
into effect in the United States, significant re-
now exist that can provide engineers with a
search has been conducted to evaluate the ef-
reasonable expectation of BMP pollutant re-
fectiveness of stormwater best management
moval and effluent water quality given a tar-
practices (BMPs). This type of research has
geted parameter (often TSS), anticipated BMP
proliferated since the adoption of Phase II
loading from a defined tributary area, BMP
NPDES Stormwater Regulations in 2003 since
design criteria (design rainfall, water quality
many smaller municipalities, unregulated under
capture volume, release structure details, etc.)
Phase I, are now required to have NPDES per-
and other factors. However, determining the
mits for stormwater discharges and face the
effects of multiple BMPs at the development or
challenges of selecting BMPs that are effective
watershed-scale on the chemistry, biology and
for management of non-point source (NPS)
physical characteristics of receiving waters is
pollution and protection of receiving water
less well documented. Determining how mul-
quality. Internationally, awareness of the ef-
tiple, distributed BMPs work on a watershed
fects of NPS on receiving water quality has in-
scale to influence receiving water quality is an
creased dramatically, and researchers in the
emerging research topic in water resources en-
United States, Asia, Europe and other popula-
gineering (Wu et al. 2006, WERF 2006).
tion centers around the world have focused on
determining how BMPs can effectively provide In the late 1990s, Intrawest, the owner of
treatment for pollutants in stormwater runoff. Copper Mountain Resort, began a large-scale
expansion of the ski resort base area, including
Significant progress has been made in as-
extensive construction of commercial, residen-
sessing BMP effectiveness and understanding
tial and parking facilities and work on the
the underlying physical, chemical and biologi-
lower slopes of the ski mountain to improve
cal processes that remove or transform pollut-
skier access and lift facilities. West Tenmile
ants stormwater (Water Environment Research
Creek, an excellent trout fishery characterized
Foundation [WERF] 2005). Resources in-
by exceptional water quality and a healthy
cluding the International Stormwater BMP Da-
community of macroinvertebrates, runs through
tabase (www.bmpdatabase.org) (Urban Water
the Copper Mountain Base Area (Base Area).
Resources Research Council et al. 2001) and
Additionally, the alluvium of West Tenmile
research efforts by organizations including
Creek serves as a public water supply source,
American Society of Civil Engineers (ASCE),
with municipal wells located less than two

13
 Technical Paper
Bridging the Gap Between BMP Effectiveness and Receiving Water Quality Protection

miles downstream of the project site. The

level of water quality protection demanded by
pristine receiving waters and challenging envi-
ronmental settings with steep slopes, erodible
soils, intense short-duration rainfall character-
istics calls for regulatory requirements and
voluntary practices that go beyond what is
typically required for development (Earles et al.
2000), and, consequently, the proposed Base
Area expansion generated a high level of public
and regulatory interest and scrutiny. Figure 1
shows the general location of Copper Mountain
and the West Tenmile Creek watershed. Fig- Figure 2. Copper Mountain Base Area Expan-
ure 2 shows the core of the Base Area expan- sion Core
sion project and West Lake (a central water
Water quality protection efforts for the
feature and intermediate receiving water) and
Base Area expansion began with a thorough
illustrates the proximity of the project to West
assessment of existing water quality data for
Tenmile Creek.
West Tenmile Creek, characterization of exist-
ing and proposed land uses in the West Tenmile
Creek watershed including Copper Mountain
and other tributary areas, geomorphic and bio-
logical characterization of West Tenmile Creek
and review of water quality criteria and other
environmental regulations. Based on this in-
formation and past experiences dealing with
water quality in high mountain settings with
sensitive receiving waters, Wright Water Engi-
neers, Inc. (WWE) worked with Intrawest and
local regulators to develop a Water Quality
Protection Strategies plan (WWE 1998) for the
proposed expansion. This plan emphasized
Figure 1. General Location Map and West multiple layers of structural and non-structural
Tenmile Creek Watershed BMPs and recognized the need for a multidis-
ciplinary approach to water quality protection
including input from engineers, biologists,

14
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.10-30

planners, and others with expertise related to tecting West Tenmile Creek.
BMPs and receiving waters (Figurski et al.
2004). Concurrent with development of water There were multiple projects associated
quality protection strategies for the Base Area with Base Area expansion between 1999 and
expansion, WWE and Intrawest began data 2005, including (1) The Village buildings, four
collection to document baseline water quality multi-unit residential buildings on a 12-acre
and biological quality of West Tenmile Creek. site with 3-acres of underground parking; (2)
Periodic grab samples were collected at loca- Passage Point, a multi-unit residential devel-
tions upstream, in the middle of and down- opment across West Tenmile Creek from The
stream from the Base Area. Macroinverte- Village, including underground parking; (3)
brate surveys were also conducted to examine The Lake District, retail shops and a boardwalk
the character and diversity of benthic organ- along the boundary of West Lake; and (4)
isms and to evaluate habitat quality. Water Trail’s End, the Cirque and Lewis Ranch, resi-
quality sampling and bioassessments were dential construction projects ranging from
completed voluntarily by Intrawest and were multi-unit condominiums to townhomes to
not a regulatory requirement; however, the large lot custom homes. The Village, Passage
project team recognized the importance of Point and The Lake District are clustered
baseline data for evaluating the overall effec- around West Lake, shown in Figure 2, and West
tiveness of water quality protection strategies Lake was a common thread with respect to the
for the development. water quality management strategy for all of
these projects. The Village used West Lake for
Construction and Challenges construction dewatering and stormwater treat-
ment, and Passage Point took advantage of the
The Base Area expansion at Copper capacity of West Lake for construction dewa-
Mountain consisted of multiple projects that tering treatment. The Lake District (and West
were constructed between 1999 and 2005. At Lake Restoration associated with the project)
the time of publication of this paper, Intrawest posed one of the greatest challenges since it
is in the planning stages of a second phase of was necessary to dewater West Lake itself for
the Base Area expansion. The effectiveness work on footings for The Lake District and re-
of the water quality protection strategies for the contouring of the lake. The following sec-
phase of the development described in this pa- tions describe some of challenges faced during
per has provided regulators and the public with construction, a period of particular vulnerabil-
a good level of comfort that a plan consisting of ity for stream biological integrity and water
structural and non-structural BMPs accompa- quality due to exposed soils and extensive con-
nied by water quality and biological monitoring struction activity.
of stream health can be highly effective at pro-

15
 Technical Paper
Bridging the Gap Between BMP Effectiveness and Receiving Water Quality Protection

The Village subsurface flows from melting of natu-

ral and man-made snow.
Initial site investigations of The Village 4. Construction site runoff had to be man-
project site, as well as prior design and con- aged concurrently with dewatering dis-
struction experience in the vicinity, indicated charges, and it was not feasible to iso-
that groundwater would be encountered when late all of the dewatering flows from
excavating and constructing the parking and Passage Point and The Village from the
building foundations for The Village and Pas- runoff waters of The Village and
sage Point. As a result, permanent de- water- on-mountain areas.
ing systems would be required for these build-
ings. Planning indicated that groundwater Construction of The Village began in
management during construction would be early May 1999. By mid-May 1999, the
challenging for multiple reasons: excavation had progressed to a depth at
which groundwater was encountered and
1. West Tenmile Creek, with an average construction dewatering commenced.
annual discharge of approximately WWE, on behalf of Intrawest and the
21,000 acre-feet, was located immedi- general contractor, obtained a Colorado
ately north of the construction zone, and Department of Public Health and Envi-
deep excavation occurred within 40 ronment (CDPHE) construction dewater-
horizontal feet of the channel. ing discharge permit that specified two
2. The West Tenmile Creek alluvium discharge locations, shown in Figure 2:
transmits large quantities of water dur-
ing the summer construction season • Discharge 001 to West Tenmile Creek
(which corresponds to the snowmelt and via a bank-side detention/filtration sys-
thunderstorm seasons). A compounding tem.
factor was that the historic West Ten- • Discharge 002 to West Tenmile Creek
mile Creek channel (now a preferential via West Lake, also a primary storm-
flow path for groundwater) passed di- water outfall.
rectly through the parking garage loca-
tion. The general dewatering strategy was to
3. The magnitude of the groundwater in- direct “clean” dewatering discharges to a
flows to the construction zone was fur- bank-side detention facility and to direct sedi-
ther increased because of the site’s loca- ment-laden dewatering discharges to West Lake
tion at the base of the ski mountain, for extended detention and sedimentation prior
which produced substantial surface and to discharge. As a condition of the CDPHE
construction dewatering permit, sample collec-

16
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.10-30

tion for water quality analyses for TSS, total water quality swale and sedimentation pond.
dissolved solids (TDS), and total phosphorus These preventative measures were imple-
(TP) was required on a weekly basis. While mented to reduce sediment entering the ex-
reporting was required for all monitored pa- cavation area and West Lake. Since dewa-
rameters, the only numeric standard for gaug- tering flows and stormwater were both
ing compliance was TSS. The permit specified routed to West Lake for treatment, it was
that the TSS concentration in the discharges critical to assure that sediment carried to
should not exceed a concentration of 45 milli- West Lake in runoff was minimized. West
grams per liter (mg/L) for a weekly average Lake was specified in the dewatering dis-
and a concentration of 30 mg/L for a monthly charge permit as a treatment measure prior
average. to discharge to the receiving water (West
Tenmile Creek), and, as a result, permit
A variety of measures were identified and compliance was assessed by sampling at the
implemented to manage stormwater and con- outflow from West Lake. While the con-
struction dewatering discharges and to increase struction stormwater permit required the
the sedimentation effectiveness of West Lake. implementation and maintenance of storm-
BMPs for the Village included: water BMPs, no numeric water quality
standards were specified in the stormwater
1. Alum dosage of West Lake and construc-
permit. The construction de- watering per-
tion dewatering discharge flows were used
mit, however, specified numeric standards
as measures to enhance sedimentation. The
for TSS. Since dewatering discharges and
pH of West Lake and the discharge from
runoff were combined in West Lake and
West Lake to West Tenmile Creek were
since compliance was gauged at the outflow
monitored on a daily basis to assure that the
from West Lake, the mixture of runoff and
alum polymer addition did not create toxic
dewatering water (rather than just the de-
effects on aquatic life. These pH measure-
watering discharge) was required to meet
ments indicated that there were no signifi-
the numeric standard for TSS specified in
cant pH changes resulting from the alum
the dewatering permit.
polymer addition. The alum polymer was
3. The well point placement strategy for con-
quite effective for the low-temperature/low
struction dewatering was designed to inter-
-alkalinity groundwater encountered in this
cept clean groundwater before reaching the
mountain setting.
area of disturbance. Dewatering plans were
2. Extensive erosion and sediment control
developed to isolate well points from the
measures were implemented upgradient of
excavation and activities of heavy machin-
the excavation at the toe of the ski moun-
ery that disturb sediments and impair the
tain. Measures included construction of a
quality of dewatering discharges.

17
 Technical Paper
Bridging the Gap Between BMP Effectiveness and Receiving Water Quality Protection

4. Inflows to West Lake from West Tenmile that they would be on hand if needed. The con-
Creek (a snowmaking diversion for resort tingency involved pumping the dewatering
operations) were effectively shut off, discharge across West Tenmile Creek to West
thereby increasing residence time for Lake for treatment. The contractor purchased
stormwater and dewatering discharges pass- additional alum polymer for use if necessary.
ing through the lake.
5. Stabilization measures were implemented It was fortunate that these contingency
on the banks of West Lake to minimize ero- plans were developed since the on-site pond
sion of banks during storm events. Meas- was not effective enough at removing sus-
ures included installation of silt fence with pended sediments to meet permit requirements.
a rigorous inspection regimen and applica- When the on-site pond proved ineffective, a
tion of a tackifier. treatment system analogous to the 1999 Village
6. A diversion system was created, allowing system was implemented. This included direct-
for diversion of dewatering discharges to ing “dirty” water to West Lake and “clean”
West Tenmile Creek (if “clean”) or to West water to a bankside filtration /detention facility
Lake (if “dirty”) via manipulation of in-line using a diversion. Alum polymer treatment was
valves. set up at the point of discharge to West Lake
(good mixing at the point of discharge). This
Passage Point system proved effective again during the sum-
mer 2000.
With knowledge of the challenges en-
countered at The Village in 1999 and similari- The Lake District and West Lake
ties between The Village and Passage Point in Rehabilitation
terms of stormwater and construction dewater-
ing management, WWE and Intrawest carefully West Lake served the project well during
planned the treatment strategy for Passage 1999 and 2000, but with dewatering of West
Point in anticipation of similar sediment chal- Lake required for The Lake District and West
lenges. Flows from the Passage Point excava- Lake Restoration, an alternate strategy was
tion were expected to be significantly less than needed for 2001. Since The Lake District was
from The Village, so initially, an on-site treat- the final stage of The Village and associated
ment pond for dewatering discharges was projects, space for treatment was scarce.
planned to avoid having to convey dewatering
WWE worked with Intrawest to develop a
discharges across West Tenmile Creek to West
treatment strategy using a pond-within-a-pond
Lake for treatment. Based on the experiences
approach. Two ponds, one for stormwater run-
with The Village in 1999, a contingency plan
off and one for construction dewater discharges,
was developed and supplies were gathered so
were planned to be constructed on the bottom

18
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.10-30

of West Lake and positioned to allow dewater- around the lake to draw in the cleanest water.
ing in areas where footings needed to be poured These techniques proved highly effective at
for The Lake District, or where work on the providing water quality protection while the
banks of West Lake was necessary. Water from ponds were constructed in West Lake.
the permanent foundation drain system for The
Village, which flows to West Lake, was re- West Tenmile Creek Setback Enhancement
routed to discharge to the stormwater pond to
During the construction of the Trail’s End
maximize residence time in the dewatering
portion of the base area development at Copper
pond. A polymer for enhanced sedimentation
Mountain (north of the Village Core), a devel-
was obtained by the contractor as a contin-
opment constraint was maintaining good access
gency.
for the heavily used bike path on the south side
Construction of the ponds within West of the development running along West Ten-
Lake presented a logistical challenge since mile Creek and heading up to Vail Pass. To
West Lake had to be dewatered to create the construct a suitable alignment, several areas of
ponds. During construction of the ponds in encroachment into the wetland and riparian
West Lake an alternate dewatering treatment area buffer along West Tenmile Creek were
method was needed. West Lake was first drawn necessary. Important purposes of wetland and
down via gravity as far as possible (not regu- riparian buffers are to provide water quality
lated by dewatering permit). For pumped de- protection and wildlife and habitat benefits.
watering discharges (regulated by the dewater- To offset buffer encroachments and to actually
ing permit), treatment was provided via filtra- improve the function of the buffer, Intrawest
tion. Dewatering discharges from West Lake developed a Wetland Setback Encroachment
were pumped into a “dewatering bag” con- Mitigation and Enhancement Plan (WWE
structed of a biodegradable geotextile fabric. 2003). This plan identified portions of the
Flow rates were kept as low as practical to buffer that were in relatively poor condition
minimize the loading on the filter bags, and prior to the start of the project and targeted
dewatering discharges were drawn from near them for enhancement. Enhancement meas-
the surface of the lake to draw out water with ures involved planting wetland plants including
the least sediment in it. Bags were replaced willows and various shrubs and planting transi-
approximately monthly, or as soon as elevated tional and upland areas within the buffer with
sediment levels (evidenced by slight discolora- shrubs and native grasses. Intrawest consulted
tion) were observed in the discharge. The with the United States Army Corps of Engi-
pump intake was attached to a small paddleboat, neers and the Colorado Division of Wildlife on
tethered to the banks with rope. This set-up al- this project. Follow-up monitoring of en-
lowed the contractor to move the intake point

19
 Technical Paper
Bridging the Gap Between BMP Effectiveness and Receiving Water Quality Protection

hancement areas has demonstrated that the Habitat quality was evaluated following
project has been a success. the habitat assessment procedure in USEPA
guidance (USEPA 1998). Habitat quality was
MEASURING SUCCESS—ASSESSMENTS assessed by completing a “habitat assessment
OF STREAM BIOLOGY AND field data sheet⎯high gradient streams” at each
WATER QUALITY site. This entailed rating habitat quality in
terms of ten parameters that address flow con-
Biological Monitoring
ditions, in-stream habitat quality, and riparian
zone conditions. Parameters were rated on a
Biological monitoring of West Tenmile
scale from 0 to 20, with 20 being “optimal.”
Creek was conducted prior to the start of con-
The total habitat score possible was 200.
struction in 1999, during the construction phase
However, the total score was not as important
of The Village and Passage Point in 2000 and
as the relative differences between scores.
most recently in 2005. Four locations were
selected for biological monitoring: one site up-
Habitat Assessment
stream of Copper Mountain, two sites within
the resort Base Area (upstream and downstream West Tenmile Creek has a moderate gra-
of a primary tributary, Wheeler Gulch), and one dient that becomes flatter in the study area.
site downstream of con- struction activities. In-stream and riparian habitat are similar at
The approximate locations of these sites along sites where the creek channel has not been al-
with water quality sampling locations are tered, primarily upstream of the Base Area de-
shown on Figure 3. velopment. The creek channel at the more
upstream sites (WTC-1 and WTC-2) was ar-
During each biological survey, WWE col-
mored by large cobble and boulders through a
lected sample of benthic macroinvertebrates
riparian corridor dominated largely by willows.
and performed a habitat assessment. Benthic
In-stream habitat at these sites consisted prin-
macroinvertebrates were sampled using stan-
cipally of riffle habitat, with pockets of slower
dard procedures (USEPA 1998), and habitat
water near the shoreline and behind larger
types were sampled in proportion to their oc-
boulders. Substrate was relatively clean,
currence over an approximately 50-meter reach
though a thin layer of brown sediment was ob-
at each site. Samples were preserved in the
served in slower areas along the shoreline. A
field, and a 100-organism sub-sample was cre-
substantial amount of coarse particulate organic
ated from each sample. Organisms were iden-
matter in the form of relatively recent leaf fall
tified to the genus level or the lowest taxo-
existed at these sites.
nomic level possible. The taxonomic guides
used were Merritt and Cummings (1988) and
Ward and Kondratieff (1992).

20
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.10-30

One of the sites monitored was directly WTC-4 was located on the golf course,
downstream of The Village (WTC-3) and downstream of the Base Area development.
downstream of West Lake. The creek down- The stream gradient was flatter at this location.
stream of WTC-3 is a lower gradient riffle/run This site included more shallow pool habitat.
armored by large cobbles and boulders. A Shallow riffle habitat was sampled at the up-
thin layer of fine sediment was observed in stream portion of this site. Substrate at this
slow water along the shore. The creek chan- site was smaller than the upstream sites, and
nel has been altered in this reach and the ripar- consisted largely of cobble.
ian zone encroached by a road and base area
landscaping. Table 1 shows the habitat assessment
scores for both the May 2000 and 2005 bio-
assessments.

Table 1. Habitat Assessment Scores for Bioassessment Sites⎯West Tenmile Creek

WTC-1 WTC-1 WTC-2 WTC-2 WTC-3 WTC-3 WTC-4 WTC-4
Metric
5/2000 9/2005 5/2000 9/2005 5/2000 9/2005 5/2000 9/2005

Epifaunal sub-
20 10 20 20 20 20 20 20
strate/available cover

Embeddedness 19 19 19 19 18 19 13 19

Velocity/depth regime 13 14 13 13 13 13 17 17

Sediment deposition 20 19 19 19 19 19 13 17

Channel flow status 19 18 20 19 18 17 20 12

Channel alteration 20 20 20 20 15 13 18 18

Frequency of riffles 20 20 20 20 20 20 20 20

Bank stability 20 20 18 18 18 19 18 19

Vegetative protection 20 20 19 19 10 12 14 15

Riparian vegetative
17 17 12 13 8 8 10 10
zone

TOTAL SCORE 188 187 180 180 159 160 163 167

Habitat ratings were similar for both bio- Benthic Organisms

assessments. Overall, habitat quality was
good to excellent at all the sampling sites. All The results of the metrics that describe the
sites had good available substrate and channel nature of the benthic community are shown in
stability. Table 2 for both the May 2000 and the 2005
bioassessments.

21
 Technical Paper
Bridging the Gap Between BMP Effectiveness and Receiving Water Quality Protection

Table 2. Tenmile Creek Bioassessment Summary of Benthic Sampling Results

WTC-1 WTC-1 WTC-2 WTC-2 WTC-3 WTC-3 WTC-4 WTC-4
Metric
5/2000 9/2005 5/2000 9/2005 5/2000 9/2005 5/2000 9/2005

Taxa richness 22 15 18 11 15 12 11 19
EPT Index 16 9 12 4 10 5 8 12
Percent contribution of
19 29 35 33 64 33 45 36
dominant taxon
Scraper-filtering col-
0.4 0.6 0.6 0.4 0.25 1.1 0 0.9
lector ratio
EPT-Chironomidae
3.9 31.0 25.5 9.0 30.7 0.9 * 3.4
ratio
Modified HBI 3.4 4.0 2.9 4.0 3.1 4.7 2.4 3.7
Notes: EPT= Ephemeroptera-Plecoptera-Trichoptera; HBI= Hilsenhoff Biotic Index

*No chironomids found

caddisflies, all of which are relatively more

In general, a healthy benthic community sensitive, were found at WTC-4.
was found at each of the sampling sites typical • The trend for the EPT index, which is the
of those found in similar creeks in the state. number of species in the mayfly, caddisfly,
The community was dominated by mayflies, and stonefly orders, followed a similar
stoneflies, and caddisflies at each site. How- trend to taxa richness.
ever, the characteristics of the benthic commu- • The percent contribution of the dominant
nity varied between the sampling sites in terms taxon was relatively constant at all sites.
of the dominant species present and the pollu- This metric was slightly lower at WTC-1,
tion tolerance of the organisms. which indicates a healthier benthic commu-
nity. The dominant species at sites
The following observations were made
WTC-1, WTC-2, and WTC-4 were either
regarding differences between the sampling
mayflies (family Baetidae) or riffle beetles
sites in 2005:
(family Elmidae). The dominant species
at WTC-3 were chironomids.
• The greatest number of species, referred to
as taxa richness, was found at WTC-1 and • The EPT-chironomidae ratio, which is the
number of mayflies, stoneflies, and caddis-
WTC-4. The higher number of species
flies (relatively sensitive organisms) to the
found at WTC-4 may be due to the greater
number of chironomids (more tolerant or-
variety of habitat types available at the site.
ganisms), varied greatly between the sites.
Several species of mayflies, stoneflies and
This ratio was the highest (indicating the

22
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.10-30

healthiest community) at WTC-1 and low- • The percent contribution of the dominant
est at WTC-3, where chironomids were the taxon was generally lower in 2005, indicat-
most common organism. ing a more balanced community structure.
• The Hilsenhoff Biotic Index (HBI) was There is also less variation in this metric in
relatively similar between the sites. This 2005.
metric was the lowest (had the most sensi-
tive community) at WTC-4 and was highest • The scrapering-filtering collector ratio was
at WTC-3 where the benthic community relatively similar between both sampling
was more tolerant. times. This suggests a relatively stable
• The ratio of scraper to filtering collectors is food base in the creek.
a measure of the food base available for
• Large differences were found in the
benthic organisms. This ratio was rela-
EPT-chironomidae ratio between the sam-
tively constant between sites and was high-
pling dates. A relatively large number of
est at WTC-3.
chironomids were found at WTC-1 and
Although the same methodology was used WTC-3 in 2000. Chironomids favor soft
for the 2000 and 2005 bioassments, results sediment, which may have been more
from the two dates are relatively different. prevalent in the previous sampling.
These differences may be due to several factors,
• Values of the HBI were higher, indicating a
the most important of which is the different
more tolerant benthic community, at all
time of year of the sampling. Samples were
sites in 2005. This indicates that the
collected in the spring of 2000 versus the fall of
community became more tolerant between
2005. Past studies have shown relatively
these two dates.
large changes in the benthic community on a
seasonal basis. Another factor to consider is
• Riffle beetles from the family Elmidae were
that historic drought conditions occurred be- relatively rare in 2000, but comprised a sig-
tween the sampling times. This may have re- nificant proportion of the community at the
sulted in relatively low flows, increased water sites in 2005. No explanation is readily
temperatures, and other changes that could apparent for this finding.
have affected the benthic community. The
following observations are made with regard to Discussion and Evaluation of Biological
results from the two sampling efforts. Monitoring Results

• Taxa richness was higher at most of the The benthic organism sampling of 2005
sites in 2000. The same was found with showed differences between sample locations
the EPT index. and differences relative to the previous sam-

23
 Technical Paper
Bridging the Gap Between BMP Effectiveness and Receiving Water Quality Protection

pling in 2000. During both sampling efforts, a both types of habitat were sampled, a mix of
healthy benthic community, dominated largely both sensitive and more tolerant species was
by mayflies, stoneflies and caddisflies, was found. This is consistent with results from
found at all sites. The spring versus fall sam- other bioassessments in similar settings
pling, and occurrence of significant drought throughout Colorado.
conditions between the sampling times, may
explain the differences between sampling While the results of the second bio- as-
events. sessment raise some questions regarding the
nature of the benthic community in West Ten-
In the 2000 sampling results, the benthic mile Creek, neither sampling event shows any
community health was better correlated with significant impairment of the benthic commu-
habitat quality. In 2002, the most robust ben- nity. The presence of some more tolerant
thic community was generally found at WTC-1, species may be due to the relatively small
which also had the highest habitat score, and amount of fine sediment in slow water at the
the health of the community declined with de- sampling sites. The presence of very sensitive
creasing habitat quality. Also, there was more species at each site, including mayflies from
variation in the metric values in 2000. the families Ameletidae and Ephemerellidae,
stoneflies from the family Leuctridae, and cad-
Another finding that is not readily ex- disflies from Rhyacophillidae, indicate that ex-
plained is the prevalence of riffle beetles from cellent water quality conditions exist in West
the family Elmidae in 2005. These beetles Tenmile Creek in the study area.
only occurred in relatively low numbers in
2000, but were the first or second most domi- Water Quality Monitoring
nant species in 2005. Several species of
Elmidae beetles live in Colorado. All occur in From May through August 2006, Wright
riffle habitat and are moderately sensitive. No Water Engineers, Inc. (WWE) staff collected
conditions were identified that would explain monthly synoptic samples from West Tenmile
their increased dominance. Creek at five locations ranging from above the
Copper Mountain Resort to just above the con-
A relatively fine film of sediment was ob- fluence with Tenmile Creek. The purpose of
served in slow water areas at each site in 2005. this sampling effort was to evaluate water qual-
Sediment typically encourages midges (from ity conditions for the stream through the Base
the family Chironomidae) and other more tol- Area and to determine if the construction ac-
erant species; however, adjacent cobble and tivities and developmet from 1999 to 2005 and
boulder habitat in flowing water was clean, existing resort conditions appeared to be having
encouraging more sensitive species. Because any adverse impacts on water quality. Similar

24
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.10-30

data were analyzed when the Water Quality a lower detection limit for phosphorus
Protection Strategies plan was developed prior analyses in July and the laboratory was
to the start of construction in 1999, and data at able to report “J” qualified values for
that time indicated excellent water quality and the last two sampling events that reflect
compliance with water quality standards estab- estimated values below the reportable
lished for West Tenmile Creek by the Colorado detection limit. No “J” values were
Water Quality Control Commission (CWQCC). reported for August 31 and “J” qualified
values of 0.02 or 0.03 mg/L were re-
Water quality sampling locations are ported for total phosphorus during the
shown on Figure 3. The stations labeled on August 1 sample event at all sample lo-
Figure 3 as “Above I-70,” “Below Union,” cations, including upstream of the resort.
“Above Wheeler” and Above Confluence One “J” value was reported for dis-
roughly corresponded to biological monitoring solved phosphorus upstream of I-70 on
locations WT-1, WT-2, WT-3 and WT-4, re- August 2, 2006, but in no other samples.
spectively. Table 3 contains a summary of the These data indicate that the resort does
sampling results. Key results of water quality not appear to be increasing the phos-
monitoring included: phorus loading to West Tenmile Creek.
• Dissolved oxygen (DO) averaged 7.6
• Water quality can be characterized as
mg/L, with no measurement below
excellent. During all site visits, flows
CWQCC stream standards.
were clear, with no significant sediment
• The average pH was 8 and remained
loading evident. No decrease in water
consistent (standard deviation = 0.2)
quality occurred from above Copper
and met CWQCC stream standards
Mountain Resort to below the resort.
through all sampling events, with no
• Nutrients, including ammonia, nitrate,
upstream to downstream trend.
nitrite, dissolved phosphorus and total
• Total suspended solids (TSS) were be-
phosphorus were consistently below
low the detection limit of 5 mg/L at all
detection limits, with the exception of
locations and for all sample events, with
one nitrate sample that was slightly
the exception of one low detected value
above detection limits, but well below
of 6.5 mg/L above Wheeler Gulch in
the stream standard. WWE requested
May.

25
 Technical Paper
Bridging the Gap Between BMP Effectiveness and Receiving Water Quality Protection

Figure 3. Water Quality Sampling Locations and Approximate Locations of Biological Surveys

The original sampling plan for metals in- the August 1, 2006 sample.. (For this reason,
cluded analyses for metals with assigned September samples were not analyzed for met-
CWQCC stream standards at the confluence als.) Of the 12 metals for which analyses
location during the May and September sam- were conducted, all were below detection limits
pling events. All metals were analyzed in the at all sample locations in August, with the ex-
dissolved form, with the exception of total ception of copper, iron and manganese, which
mercury, total recoverable iron and total recov- had a few detectable concentrations. On Au-
erable arsenic. Metal samples were collected gust 1, copper concentrations were above the
in May at this location and planned for collec- stream standard upstream of Copper Mountain
tion in September. However, the laboratory and below Union Creek but were below detec-
inadvertently analyzed metals at all locations in tion limits at all other locations. Manganese

26
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.10-30

was detected at a few locations, but was well Area expansion demonstrate that receiving wa-
below the stream standard. Total recoverable ter quality of West Tenmile Creek has been
iron was reported at the detection limit and well protected.
below the stream standard at the confluence
monitoring location in May.

These samples were collected during dry

weather flow conditions, but were collected
over a time period that reflects the spring run-
off, followed by gradually decreasing flow
conditions in the later summer. Relative to the
annual hydrograph, the four sampling events
would be considered to be a high flow period.

Conclusions

The multi-tiered program developed in the

Water Quality Protection Strategies plan for the
Base Area development at Copper Mountain
provides an example of a successful BMP
-based approach to water quality protection for
development, both during construction and op-
erational phases. A unique aspect of this pro-
ject has been the extensive documentation of
water quality and biological characteristics of
West Tenmile Creek. While many evaluations
of BMP effectiveness focus on BMP inflows
and outflows and provide useful information on
the processes that remove pollutants in BMPs,
the monitoring data collected at Copper Moun-
tain provides a broader perspective and demon-
strates how a development-wide system of
structural and non-structural BMPs can be ef-
fective at protecting water quality. The water
quality and biological data collected before
during and after the construction of the Base

27
 Technical Paper
Bridging the Gap Between BMP Effectiveness and Receiving Water Quality Protection

Table 3. West Tenmile Creek 2006 Water Quality Data

Conduc- Hard- P, As, Fe,

Location Date NH3-N P, Diss. DO NO3-N NO2-N pH Temp TSS Cd Cr Cu Pb Mn Hg, Tot. Se
tivity ness Tot. Trec. Trec.

Units mg/L us/cm mg/L mg/L mg/L mg/L mg/L SU mg/L C mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L
Above 70 10-May-06 <0.8 NA <0.1 68 <0.056 <0.076 8.1 <0.1 <5
Above 70 21-Jun-06 <0.8 130 <0.1 7.3 53 <0.056 <0.076 7.6 <0.1 9.2 <5
0.02
Above 70 02-Aug-06 <0.8 192 0.02J 7.2 79 <0.056 <0.076 7.7 9.7 <5 <0.01 <0.01 0.0190 <0.073 0.0098 <0.1
J
Above 70 31-Aug-06 <0.8 222 <0.1 7.5 99 <0.056 <0.076 8.1 <0.1 7.6 <5
Above Wheeler 10-May-06 <0.8 NA <0.1 NA 74 <0.056 <0.076 8.1 <0.1 NA 6.5
Above Wheeler 21-Jun-06 <0.8 131 <0.1 7.7 53 <0.056 <0.076 7.9 <0.1 9.9 <5
0.02
Above Wheeler 02-Aug-06 <0.8 131 <0.1 8.0 82 <0.056 <0.076 7.9 10.0 <5 <0.01 <0.01 <0.005 <0.073 <0.005 <0.1
J
Above Wheeler 31-Aug-06 <0.8 222 <0.1 8.3 97 <0.056 <0.076 8.1 <0.1 7.3 <5
Below Union 10-May-06 <0.8 NA <0.1 NA 68 <0.056 <0.076 8.1 <0.1 NA <5
Below Union 21-Jun-06 <0.8 128 <0.1 7.7 53 <0.056 <0.076 7.6 <0.1 9.7 <5
0.02
Below Union 02-Aug-06 <0.8 192 <0.1 7.3 79 <0.056 <0.076 7.9 10.0 <5 <0.01 <0.01 0.0098 <0.073 0.0066 <0.1
J
Below Union 31-Aug-06 <0.8 221 <0.1 7.6 97 <0.056 <0.076 8.1 <0.1 8.0 <5
Below Wheeler 10-May-06 <0.8 NA <0.1 NA 74 <0.056 <0.076 8.2 <0.1 NA <5
Below Wheeler 21-Jun-06 <0.8 149 <0.1 7.5 60 <0.056 <0.076 7.9 <0.1 9.5 <5
0.02
Below Wheeler 02-Aug-06 <0.8 198 <0.1 7.6 83 <0.056 <0.076 8.1 10.0 <5 <0.01 <0.01 <0.005 <0.073 <0.005 <0.1
J
Below Wheeler 31-Aug-06 <0.8 221 <0.1 8.1 98 <0.056 <0.076 8.2 <0.1 7.6 <5
Confluence 10-May-06 <0.8 NA <0.1 NA 74 0.059 <0.076 8.2 <0.1 NA <5 <0.05 <0.01 <0.01 <0.005 0.07 <0.073 0.0066 <0.0001 <0.1
Confluence 21-Jun-06 <0.8 137 <0.1 7.1 57 <0.056 <0.076 7.8 <0.1 10.2 <5
0.03
Confluence 02-Aug-06 <0.8 178 <0.1 7.2 82 <0.056 <0.076 8.0 10.0 <5 <0.01 <0.01 <0.005 <0.073 <0.005 <0.1
J
Confluence 31-Aug-06 <0.8 223 <0.1 7.6 98 <0.056 <0.076 8.2 <0.1 8.7 <5
<0.0001
Average <0.8* 178 <0.1* 7.6 76 <0.056* <0.076* 8.0 <0.1* 9.2 <5 <0.05 <0.01* <0.01* 0.0048 0.07 <0.073* 0.0038 <0.1*
*
<0.0001
Minimum <0.8* 128 <0.1* 7.1 53 <0.056* <0.076* 7.6 <0.1* 7.3 <5 <0.01* <0.01* <0.005 NC <0.073* <0.005 <0.1*
0.03 *
<0.0001
Maximum <0.8* 223 0.02J 8.3 99 0.059 <0.076* 8.2 10.2 6.5 <0.01* <0.01* 0.019 NC <0.073* 0.0098 <0.1*
J *
Std. Dev. 40 0.3 16 0.2 1.0
J = Value reported below Practical Quantification Limit at Request of WWE
<__*=all values below detection limits

28
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.10-30

base. Prepared in cooperation with the

References USEPA Office of Water.
[6] USEPA 1998. Revision to Rapid Bio-
[1] Earles, T.A., J.E. Jones, and W.F. Lorenz
assessment Protocols for Use in Streams
2000. Case Studies of Local Strategies
and Rivers⎯Benthic Macroinvertebrates
for Control of Non-Point Source Pollution
and Fish. January 1998. Final draft.
in Colorado (USA). In Advances in Ur-
[7] USEPA 1989. Rapid Bioassessment Pro-
ban Stormwater and Agricultural Runoff
tocols for Use in Streams and Riv-
Source Controls, Marsalek, J., Watt, E.,
ers⎯Benthic Macroinvertebrates and Fish.
Zemen, E. and Sieker, H. (eds.). NATO
EPA/444/4-89-001.
Science Series: IV Earth and Environ-
[8] Ward, J.V. and B.C. Kondratieff 1992.
mental Sciences—Vol. 6. Kluwer Aca-
An Illustrated Guide to the Mountain
demic Publishers: Boston.
Stream Insects of Colorado. Univer-
[2] Jones, J.E. and T.A. Earles 2000. Best
sity Press of Colorado.
Management Practices for Development
[9] WERF 2006. Linking Stormwater BMP
Projects in the Rocky Mountains. Proc.
Systems Performance to Receiving Water
Inaugural Session of the Water Congress
Protection to Improve BMP Selection and
of the ASCE Environmental and Water
Design. Request for Preproposals, RFPP
Resources Institute. Orlando, Florida.
06-SW-1. WERF: Alexandria, Virginia.
Washington, D.C.: ASCE. May 20-24.
[10] WERF 2005. Critical Assessment of
[3] Figurski, M.J., Roesner, M.J. and T.A.
Stormwater Treatment and Control Selec-
Earles 2004. Multipurpose Facilities are
tion Issues. Final Report 02-SW-1.
Improved Through Multidisciplinary In-
WERF: Alexandria, Virginia and IWA
put. Water Resources Impact, Volume 6,
Publishing: London.
No. 5, pp.13-16. American Water Re-
[11] WWE 1998. Copper Mountain Base Area
sources Association.
Water Quality Protection Strategies.
[4] Merrit, R.W. and K.W. Cummings 1989.
Prepared for Copper Mountain Resort
An Introduction to the Aquatic Insects of
Development Group, Intrawest U.S.
North America—Second Edition. Kend-
Holdings, Inc., Copper Mountain, Colo-
all/Hunt.
rado.
[5] Urban Water Resources Research Council
[12] WWE 2003. Wetland Setback En-
of ASCE, WWE, Denver Urban Drainage
croachment Mitigation and Enhancement
and Flood Control District and URS
Plan. Prepared for Copper Mountain
Greiner Wood Ward Clyde 2001. User’s
Resort Development Group, Intrawest U.S.
Guide National Stormwater BMP Data-
Holdings, Inc., Copper Mountain, Colo-
rado.

29
 Technical Paper
Bridging the Gap Between BMP Effectiveness and Receiving Water Quality Protection

[13] Wu. J, Yu, S.L. and R. Zou 2006. A Water

Quality-Based Approach for Watershed
Wide BMP Strategies. Journal of the
American Water Resources Association,
Volume 42, No. 5, pp. 1193-1204.

30
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.31-47

Tools and Processes for Implementing a Road Ecology Approach to

Transportation and the Environment

Alison M. Berry1, Katie Benouar2, Sandra L. Jacobson3, Carson Poe4, and Victor Claassen5

1
Alison M. Berry, Professor of Plant Sciences and Director, Road Ecology Center, John Muir Institute of the
Environment, University of California at Davis.
2
Katie Benouar, Senior Research Coordinator, Road Ecology Center, John Muir Institute of the Environment,
University of California at Davis.
3
Sandra L. Jacobson, Wildlife Biologist, USDA Forest Service, Pacific Southwest Research Station, Redwood
Sciences Laboratory, 1700 Bayview Drive, Arcata, CA. [email protected], 707-825-2985.
4
Carson Poe, Planning and Policy Division, John A. Volpe National Transportation Systems Center, United
States Department of Transportation, Cambridge, Massachusetts.
5
Victor Claassen, Assistant Researcher, Land, Air and Water Resources Department, University of Califor-
nia at Davis.

Abstract

Road systems are an essential infrastructure need for economic and social wellbeing. As road
networks expand and alter the landscape, ecological processes are affected, in turn affecting core
resources such as water quality, air quality, wildlife habitat, fisheries, wetlands and vegetation.
Awareness of the effects of roads on the environment has grown, but research is still needed to
achieve a scientific understanding of the underlying processes and to apply this understanding to the
design of sustainable transportation systems.

The science of road ecology is being developed to address the interface between roads and
natural systems. Scientific research will provide the foundation for development of a road ecology
approach to transportation systems. The road ecology approach will include tools and processes that
incorporate ecological considerations in transportation planning, design and project delivery.
Road ecologists also aim to develop ways that transportation agencies and resource protection
agencies can collaborate to lessen the impacts of roads on the environment.

Road ecology is an emerging field, both in transportation planning and engineering on the one
hand, and in academic research. Several major publications worth noting have appeared in recent
years. The keystone book in this area is Road Ecology (Forman et al. 2003), which provides an ex-
cellent foundation and overview of the key concepts of road ecology. A report addressing potential
assessment tools and approaches, sponsored by the U.S. National Academy of Sciences (Assessing

31
 Technical Paper
Tools and Processes for Implementing a Road Ecology Approach to Transportation and the Environment

and Managing the Ecological Impacts of Paved Roads, 2005) is available from National Academy
Press (http://www.nap.edu/catalog/11535.html . Most recently, Davenport and Davenport (2006)
bring together a series of papers on transportation impacts on ecosystems, ranging from marine
travel to highway transportation. Several chapters deal with wildlife impacts, and include practical
solutions.

The intent of this review is to summarize some recent tools and processes that may be useful
and adaptable for a range of transportation planning situations. The topics include:

I、 Ecological approaches to management of roadside soils and vegetation;

II、 Identifying and mitigating road impacts on wildlife;
III、 Eco-Logical, an integrative planning tool for professionals;
IV、 University Centers for environment and transportation; and
V、 ICOET, the International Conference on Ecology and Transportation.

Keywords: Road ecology; Transportation

32
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.31-47

簡論推行道路生態規劃的一些工具及方法

道路系統為社會經濟民生福祉不可或缺設施之ㄧ，在道路網絡不斷擴充情況下，沿路之
景觀及生態將會受影響，而對重要資源，如水、空氣、野生動物棲地、漁產、濕地及植物等
引起衝擊。目前道路會對環境產生影響已被認同，但有關對何種影響細節之瞭解，以及如何
將瞭解後得到之資訊應用到設計規劃一永續性之交通系統等方面之研究則仍有待加強。

道路生態領域現在仍正在發展中，其內涵是道路與自然系統介面之探討，而目的則是建
立已含生態考量之交通系統規劃之基礎並提供考量生態的一些工具及方法。道路生態之另一
目標為促進交通及資源部門之合作來減輕道路對環境引起之衝擊。

目前，在交通規劃、工程及學術研究方面，道路生態之考量還僅是在起步階段。近幾年
來，有一些重要的文獻值得參考：

Š Forman 等（2003）所著的書「道路生態」，對道路生態之重要基礎及觀念有詳
盡之描述。
Š 美國國家科學院 2005 年的一本報告「鋪面道路對生態衝擊之評估及管理」，
（http://www.nap.edu/catalog/11535.html.）介紹了一些評估工具及管理方法。
Š Davenport and Davenport（2006）發表了一系列之論文，內容主要為交通（航運、
公路等）對生態系統之衝擊，包括對野生動物之影響及實務的解決方法等。

本文之目的為討論在交通規劃中考量到路生態的一些工具及方法，內容包括：

I、 保護生態之路邊土壤及植物的管理
II、 道路對野生動物衝擊之鑑定及減輕方法
III、 Eco-Logical，進行實務工作時整體性規劃工具
IV、 大學中環境及通研究中心簡介
V、 ICOET，國際生態及交通研討會

關鍵詞：道路生態、交通規劃

33
 Technical Paper
Tools and Processes for Implementing a Road Ecology Approach to Transportation and the Environment

Part I. Regenerating functional soils to substrates. Also, in order for implementation to

support native plants for slope
occur, the treatments specified must be cost ef-
stabilization
fective to purchase, transport and apply to sites
Soil erosion along roads is a severe prob- that are often remote, and difficult to work
lem in many parts of the world, particularly (steep, unstable, rocky or clayey).
where roads intersect with sloping topography,
and where environmental disturbances such as The guidelines, as currently developed,
heavy rainfall or fire can lead to excessive run- include:
off and destabilization. To identify and remedi-
1) identification of a reference site;
ate the causes of non-point source erosion on
2) evaluation of infiltration capacity;
drastically disturbed slopes (meaning all topsoil
3) assessment of plant available water;
and biological material removed by excavation,
4) knowledge of soil organic matter
landslide or burial), soil scientists at the Soil
pools and mineralization rates of C
and Revegetation Laboratory (University of
and N;
California, Davis) have developed a set of
5) non-N nutrients and soil chemical
guidelines that focus on soil stabilization as a
characteristics (pH, CEC, EC);
starting point. Thus, if the soil or substrate
6) soil biological activity (symbiotic
conditions are made adequate, many locally
microbes, saprophyte activity);
occurring plant species will colonize and stabi-
7) surface erosion stabilization; and
lize the slopes, bringing their associated toler-
8) needs for site specific plant materi-
ances for local climatic conditions, grazing
als.
pressure and disease. This is an alternative to
approaches that emphasize single-species 1. Reference sites are selected that represent
plantings that may be adapted to some condi- the potential outcomes within 3 to 5 or 10
tions, but that can become either weedy invad- years after construction. The purpose is to
ers in other local habitats, or that do not survive envision what outcomes are acceptable to
during periodic weather extremes. the stakeholders and to make sure that
everyone is agreeing on the same result.
Each of the target soil characteristics must
The reference site also serves as an exam-
be measured in ways that are relevant to wild-
ple of a soil / plant system that works well
lands conditions (as opposed to agricultural
in the local environment and can be used
conditions) and that have known or adaptable
to develop target soil characteristics.
treatments that can be constructed by heavy
Typically, a “disturbed-but-revegetated”
equipment. While many soils tests have been
site is preferable to an “undisturbed” site,
developed for horticultural, agricultural or for-
since less disturbed sites often have soil
estry uses, these may not be useful in degraded

34
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.31-47

materials that have developed or accumu- 3. Threshold levels for plant available water
lated for hundreds to thousands of years, is estimated from evapotranspiration data
making an unrealistic target for a short and plant water use data of native grasses
term revegetation project. Plant commu- and shrubs. The amount of soil needed for
nity density and composition are often adequate plant moisture varies widely, but
evaluated on the revegetated reference and is typically between 0.5 m and 1.5 m, de-
the impacted (barren) sites, but actual pending on the rock content of the soil,
measurement of soil conditions is made on which can dilute the water holding capac-
both sites through the remaining soil ity. It is not necessary to till to this depth,
evaluation steps. if the underlying geology is already frac-
tured, which is common in many rock
2. Infiltration rates for impacted sites are types.
often reduced because the fine mineral
particles often settle and pack tightly or 4. Soil organic matter is important for gen-
form crusts due to the low organic matter erating soil aggregates, as mentioned in
content. This reduces downward infilt- ra- the infiltration section, but also for pro-
tion of rainfall, causing more water to viding a long-term supply of nitrogen (N)
flow over the surface of the site. Our ap- for plant growth. Fertilizer applications of
proach to this situation is to set a target nutrients, especially of N, tend to be
condition for infiltration that is adequate highly available for a short period, en-
to prevent overland flow at the site. This couraging weedy growth. Shortly there-
can be done by copying the characteristics after, available N declines and the plant
of the revegetated reference site, or by community thins and erosion returns.
getting a rate from storm intensity data for Amending the site with large loads of
various storm return frequencies. When composted materials (50 to 100 Mg or
the site is tilled or ripped with heavy more) can supply a long-lasting but slowly
equipment, the soil particles often re-settle available supply of N for regeneration of
back into a close packed configuration the plant community. Note that regenera-
within a season or so, returning the site to tion of the community on a previously
an erosive condition. Incorporation of barren site requires not only N for the
coarse woody material (forest thinning plant biomass, but for soil organic matter,
slash, unscreened yard waste compost, tub secondary woody tissue, a woody duff
ground (not chipped) wood) helps keep layer, and for all the soil biota that will
the soil macropores open for the several colonize the site. This may require many
years needed to start regeneration of soil hundreds of kg N/ha to be available for
structure. incorporation into living or dead biomass.

35
 Technical Paper
Tools and Processes for Implementing a Road Ecology Approach to Transportation and the Environment

5. Non-N nutrient evaluation (P, K, S, Ca, from running onto the site from upslope,
Mg, micronutrients) and soil chemical undercutting the toe of slopes, and making
characteristics (pH, CEC, EC) are gener- sure that the vegetative community is vi-
ally easily measured by existing soil fertil- brant and dense enough to take on the task
ity tests. Target values, however, are set of protecting the soil. Mulch effects are
for wildlands systems rather than for commonly known for intercepting rain
commercial agricultural systems. When drop splash impact, but thermal protection
this information is not available, we again (from winter cold and from summer heat)
use the revegetated reference site to indi- are also important, as is reduction in
cate what is adequate for the plants grow- evapotranspiration from the soil during
ing in this environment. dry seasons.

6. Soil biological activity is divided into 8. The final aspect that has become important
saprophytic groups, which are generally as we take on progressively more harsh
easily stimulated by organic matter addi- and atypical sites is the issue of selecting
tions, and to symbiotic microbes, includ- site adapted plant materials. For valley
ing N-fixing symbionts and mycorrhizal locations and moderate climates, many
fungi. These goals are very site specific horticultural accessions are available, but
and cannot easily be generalized except to as the site conditions become more ex-
say that we emphasize the use of locally treme, the need is greater to have plants
collected inoculants, which have been ex- that can tolerate these conditions. For
posed to local climatic conditions for an these reasons, we try to identify conditions
extended period of time and are hopefully that will require special plant selection, so
adapted to these conditions. that these seed sources can be collected
and propagated in preparation for the site
7. Surface stabilization involves either tem- treatment.
porary erosion control, long term erosion
control or surface mulch effects. Tem- Some horticultural and agricultural plants
porary erosion control is a well-developed can become weedy and invasive if planted in
commercial activity, making available highly disturbed sites such as roadsides.
many products and approaches. We note There are current efforts to develop appropriate
that if infiltration is regenerated, much less practices for restoration and maintenance of
emphasis can be placed on surface erosion native vegetation along roadsides, as described
control efforts, since overland flow is in Harper-Lore and Wilson (2000). To insure
predominantly eliminated. Long term the establishment of site-adapted native geno-
erosion control involves controlling water types, restoration may involve a phase of

36
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.31-47

propagating seeds collected from the restora- tract and protect wildlife as they cross, remov-
tion site, and then placement of the resulting ing vegetation in some areas to discourage
plants or the next generation of seeds. Since wildlife from approaching the road, and study-
vegetation provides habitat for wildlife, road- ing the width of medians and how that affects
side plantings can have positive effects on the likelihood of wildlife to cross roads.
wildlife. At the same time, precautions need to
be taken to prevent creating a wildlife “sink” Traffic volume impacts to terrestrial and
due to road mortality. More study in this area is aquatic species
needed.
A major difference between road effects to
Part II. Identifying and mitigating road terrestrial and aquatic organisms is traffic vol-
impacts on wildlife ume. Traffic volume, considered daily or annu-
ally when the species of interest are active, is
Road networks have been built without
predictive of the impact a road will have on
consideration of wildlife movement networks
terrestrial species (Van Langevelde and Jaarsma
and as a result, often disrupt the movement of
2005). As traffic volume increases, roads affect
various species and negatively impact the
wildlife species in different ways (Müller and
health of the ecosystems that support their sur-
Berthoud 1995, Jacobson in prep.). At very low
vival. From the perspective of wildlife biolo-
traffic volume, most animals except very slow
gists, road networks pose a threat to the health
species are minimally affected, but these slow
and survival of wildlife by placing barriers
species can be wiped out if the impact contin-
across wildlife movement corridors, introduc-
ues over time. At moderate traffic volume (be-
ing noise, toxic run-off and death from crashes.
tween 2000-8000 ADT), mortality is heaviest
From the perspective of transportation agencies,
for most species, and habitat permeability is
wildlife attempting to cross roads is a safety
moderately impacted. Again, slow species will
hazard to the traveling public. Transportation
continue to experience heavy mortality as well
agencies are also increasingly seeking ways to
as barrier effects. At high traffic volumes,
balance their mission of providing roads with
above 8000 ADT, mortality is reduced because
good stewardship of the environment.
most animals are intimidated from attempting
To address these concerns about wildlife to cross, but the highway becomes a virtually
and highway interactions, a variety of tech- complete barrier to animal movement. Only
niques and tools are being developed. These those species with great motivation to cross
include approaches directed at wildlife and how will continue to attempt to cross at very high
they behave such as building wildlife crossing traffic volumes. Examples include migratory
structures over or under roads, enhancing the deer herds, or species whose behavior is more
vegetation around the crossing structures to at- instinctual than intelligent such as frogs and

37
 Technical Paper
Tools and Processes for Implementing a Road Ecology Approach to Transportation and the Environment

turtles. Focusing traffic volume on few roads allows

mitigation to be more efficiently focused as
Impacts to animals from highways vary by well. Generally, roads with moderate traffic
species. The groups of species most at risk are volume in high quality habitat are the highest
slow species of all taxonomic categories, such priority for urgent and best mitigation measures
as turtles, frogs, badgers, and gallinaceous birds. such as wildlife passages and barrier fencing.
This is because it takes them longer to cross a Very high volume roads may be able to avoid
road and exposes them to risk of mortality fencing unless migrant species are present, but
longer. Species whose defense mechanisms are need wildlife passages to mitigate the impacts
incompatible with vehicles are at great risk as of loss of habitat connectivity.
well. These include species such as snakes that
coil in the presence of danger, small animals Wildlife crossings
that depend on cryptic coloration and mo-
tionlessness to avoid predation, and species that The most effective mitigation measure
have circuitous travels such as rabbits. Other currently known for both mortality and habitat
categories are those that must travel across fragmentation impacts is a combination of
highways to meet key survival needs, such as wildlife crossing structures of the appropriate
winter range, or daily water or food. design for the species impacted, barrier fencing
Wide-ranging mobile species, including most to divert them from the road into the structure,
species of carnivores, are at risk because their and escape structures to allow any clever ani-
mobility places them in frequent contact with mals to leave the highway if they find their way
roads. onto it (Forman et al. 2003, Clevenger et al.
2001b). Further, a mixture of large crossings
Traffic calming to address road impacts to for large mammals and more frequent smaller
terrestrial species structures for smaller animals will reduce mor-
tality and increase the permeability of the
Because traffic volume is the best predic- highway for most terrestrial species (Clevenger
tor of impacts to terrestrial species, any meas- and Waltho 2000, Clevenger et al. 2001a).
ures taken to reduce traffic volume will benefit
them (Jaarsma and Willems 2002). This in- Collaboration among all agencies and
cludes transit options, traffic calming mecha- non-governmental organizations is key to the
nisms and installing wildlife passages on roads success of maintaining aquatic and terrestrial
with predicted increases in traffic volume on species health. A road does not affect only
existing roads. A useful traffic calming ap- those species that wander onto its surface, but
proach is to focus high traffic volume on one rather it can affect species far beyond it. This
road instead of dispersing it among many roads. suggests that landowners distant from the road

38
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.31-47

will benefit from involvement in transportation sage over time will save money in the long run.
planning, and the public in general will be able Two websites which provide some guidance on
to maintain its wildlife and fisheries resources. wildlife and aquatics passage are the Wildlife
Crossings Toolkit at http://www. wildlifecross-
The training course, Innovative Ap- ings.info and the Fish Xing website at
proaches to Wildlife and Highway Interactions http://stream.fs.fed.us/fishxing /beta.html.
was based on the premise that many disciplines
will need to be integrated into transportation Designing water conveyance structures as
planning to solve some of the major challenges wildlife and fish passages
associated with highway impacts to wildlife. A
key component of the short course is to include In areas with high rainfall, roads will have
the skills and experience of both engineers and numerous water conveyance structures. Be-
biologists, who tend to view challenges differ- cause many animal species use drainages as
ently and have different methods of problem travel corridors, water conveyance structures
solving. Networking is also a useful component are often useful for wildlife passages. Although
of these training sessions, because often pro- water goes through these structures, it cannot
fessionals are unaware of other agencies’ or- be assumed they will function as passages for
ganizational structures or how to contact the fish or other aquatic organisms. Water convey-
appropriate person within another agency. ance structures such as bridges and culverts can
be ideal crossing structures for both aquatic and
Additional wildlife crossing guidance terrestrial species if they are constructed with
the needs of the target species in mind.
Designing crossing structures is most effi-
ciently done when considering all the species, The easiest way to determine how to make
aquatic and terrestrial, in a geographic area. a water conveyance structure function for wild-
Some aquatic organism passage techniques are life and fish is to simulate a natural stream
detrimental to wildlife passage, and can be so crossing. “Stream simulation” means that the
expensive as to preclude subsequent modifica- natural gradient, width and substrate are main-
tion to allow for wildlife use. This may happen tained so that the stream is not fighting with the
when a crossing is considered impactive to structure at any flow levels (Clarkin et al. 2005).
fisheries but the traffic volume is not impactive When stream crossings are constructed with
to wildlife at the time of planning and construc- stream simulation in mind, aquatic species will
tion, but later the road evolves into a higher not have to negotiate culverts that are pitched at
traffic volume or fun- ctional class that will af- too great an angle or that have perched outlets.
fect wildlife. Crossing types designed to be Stream simulation does not address the height
functional for both aquatics and terrestrial pas- of a structure except as it relates to flood events,

39
 Technical Paper
Tools and Processes for Implementing a Road Ecology Approach to Transportation and the Environment

and usually streams are better simulated as Part III. Eco-Logical: An Ecosystem Ap-
proach to Developing Infrastruc-
wide crossings rather than as high crossings.
ture Projects
Thus, some structures that serve aquatic species
very well may not be suitable for wildlife spe- Over the last several decades, the under-
cies because they do not have adequate head- standing of how infrastructure impacts habitat
room. Wide structures are better than tall and and ecosystems has grown. Consequently,
narrow structures for most wildlife species as awareness of how to better avoid, minimize,
long as the headroom is adequate (Gordon and and mitigate impacts has also matured. A new
Anderson 2003, Foster and Humphrey 1995). A guide has been developed to help agencies form
key feature of good stream simulation is that partnerships and translate this awareness into
dry areas are maintained along the side of the action. Articulating a vision of cost-effective
stream except in flood events, and water is at infrastructure development that contributes to
approximately the same depth as the natural ecosystem conservation, EcoLogical: An Eco-
stream (that is, not too shallow and spread out). system Approach to Developing Infrastructure
This allows terrestrial animals, even those very Projects offers a non-prescriptive approach to
capable of swimming, to travel on dry land in- making infrastructure more sensitive to wildlife
stead of wading through water, and aquatic and ecosystems through greater interagency
species adequate depth to swim. cooperative conservation.

Many existing water conveyance struc- Central to Eco-Logical, which was devel-
tures can be ‘retrofitted’ to serve aquatic and oped by a steering team of representatives from
terrestrial passage. Sometimes this may be as eight Federal agencies and the Departments of
simple as to install a wide shelf (0.5 m) along Transportation from four States, is the “ecosys-
the edge so that there is a dry surface for terres- tem approach,” a method for sustaining or re-
trial species to walk. Another example of a storing ecological systems and their functions
useful retrofit is to install sound-moderating and values. An ecosystem approach is a process
surface material in the ceilings of structures to for the comprehensive management of land,
make the sound inside the structures more water, and biotic and abiotic resources that eq-
natural, and therefore less intimidating, to ani- uitably promotes conservation and sustainable
mals (Jacobson, in prep.). Retrofitting existing use. The goaldriven ecosystem approach in
structures is an emerging field; so many new Eco-Logical is based on a collaboratively de-
designs will be developed in the coming years. veloped vision of desired future conditions that
integrates ecological, economic, and social fac-
tors. It is applied within a geographic frame-
work defined primarily by ecological, rather
than political or other boundaries.

40
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.31-47

To help achieve the approach, the guide essary – mitigation are acted upon be-
offers the conceptual groundwork for integrated fore critical ecological resources are
planning, a process by which agencies and their lost.
partners:
Eco-Logical provides a process for the
• Combine planning efforts – Manage- collection, sharing, analysis, and presentation of
ment plans that agencies and partners data in agencies’ plans. A key product of this
have developed individually are impor- process is a regional ecosystem framework
tant sources of information in the inte- (REF). A REF can help identify ecologically
grated planning process. Some examples significant areas, potentially impacted resources,
of plans include: long-range transporta- regions to avoid, and mitigation opportunities
tion plans; watershed plans; species re- before new projects are initiated. With this un-
covery plans; resource management derstanding, agencies can work together to
plans; state wildlife action plans, forest more accurately identify the areas in most need
management plans; Special Area Man- of protection, better predict and assess cumula-
agement Plans; coastal management tive resource impacts, and streamline infra-
plans; and community growth plans. structure development through increased pre-
• Understand where planned infrastruc- dictability. Although there is no standard for
ture projects and resources (natural, creating a REF, Eco-Logical recommends that a
historic, and cultural) will interact – REF consist of an “overlay” of maps of agen-
Maps from infrastructure and conserva- cies individual plans, accompanied by descrip-
tion plans can be overlaid to determine tions of conservation goals in the defined re-
the projects and resources that “link” gion(s).
agencies. An overlay of maps can show
how planned projects and objectives Eco-Logical also supports the considera-
might cumulatively impact a region’s tion of ecosystem-based mitigation where ap-
resources, and identify areas where col- propriate. Ecosystem-based mitigation is the
laboration on project location and de- process of restoring, creating, enhancing, and
sign may address multiple and diverse preserving habitat and other ecosystem features
goals within the community and the in areas where environmental needs and the
ecosystem. potential environmental contributions are
• Define ecological resources of highest greatest. It is implemented in conjunction with,
concern – Identification of ecological- or in advance of, infrastructure projects. Eco-
priority areas can help ensure that op- system-based mitigation extends existing com-
portunities to protect them through pensatory mitigation options by offering a way
avoidance, minimization, and – if nec- to evaluate alternatives for off-site mitigation

41
 Technical Paper
Tools and Processes for Implementing a Road Ecology Approach to Transportation and the Environment

and/or out-of-kind mitigation in the ecologi- sign infrastructure in ways that mini-
cally most important areas as defined by inter- mize habitat fragmentation and protect
agency partners and the public. It is a poten- larger scale, multi-resource ecosystems.
tially enhanced approach to crediting mitigation • Efficient project development –
that builds on existing approaches. Integrating Uncertainty during project development
this new concept with lessons learned from imposes a high cost on agencies and
previous experience can allow agencies to partners, in both time and money. An
capitalize on opportunities for substantial habi- ecosystem approach fosters cost-effec-
tat connectivity and wildlife conservation while tive environmental solutions that can be
developing needed infrastructure. incorporated early in the planning and
design of infrastructure projects.
Moving beyond the customary pro- • Increased transparency – Infrastruc-
ject-by-project approach to meeting infrastruc- ture projects developed with an ecosys-
ture needs and toward the ecosystem approach tem approach provide opportunities for,
defined in Eco-Logical can result in a range of and encourage, public and stakeholder
benefits, including: involvement at all key stages of plan-
ning and development.
• Safer, improved infrastructure – All
agencies and stakeholders contribute to Eco-Logical is available on the Internet at
the delivery of infrastructure. The col- www.environment.fhwa.dot.gov/ecological/eco
lective abilities and knowledge shared in _index.asp, and in hardcopy by contacting
the ecosystem approach should allow a Carol Adkins at (202) 366-2054 or carol.adkins
more balanced understanding of eco- @fhwa.dot.gov.
logical and social concerns.
• Improved watershed and ecosystem Tools for professionals
health – Integrating the preventive, di-
agnostic, and prognostic aspects of eco- An important aspect of road ecology is the
system management should lead to need to bring together interdisciplinary groups
greater understanding of the relation- of scientists and professionals to develop the
ships between ecological issues and science from several disciplines to work to-
human activities. gether in new ways and develop innovative ap-
• Increased connectivity and conserva- proaches to the road and natural systems inter-
tion – Since the ecosystem approach face. This leads to development of interagency
takes a broad view that encompasses the tools and processes that help professionals from
interaction of human and natural sys- various federal, state and local agencies work
tems, it can help agencies plan and de- together despite differing missions and per-

42
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.31-47

spectives. historical resources. Context sensitive solu-

tions is changing the approach to project deliv-
The Federal Highway Administration ery by training engineers and planners in this
(FHWA), the agency responsible for transporta- approach. The science of road ecology can be
tion planning and project development nation- an important input to this process. More infor-
wide, has developed many products that sup- mation about context sensitive solutions is can
port environmental stewardship in transporta- be found at http://www.fhwa.dot.gov/csd/index.
tion provision. There is an extensive website of cfm
their products that includes guidance for wet-
lands mitigation, context sensitive solutions Integrated planning approaches
and integrated planning best practices. By
funding research and development of tools and Transportation planning is evolving to-
best practices, the federal agency is promoting ward a multi-disciplinary, multi-agency prac-
understanding of the impacts of roads on the tice that includes consideration of road ecology,
natural environment and awareness among sharing of environmental data among agencies,
transportation engineers and planners that this setting up processes where coordination can
is a consideration to be incorporated in their occur and integrating the various plans in a
work. The FHWA webpage for environmental given region and ecosystem. Plans may include
streamlining and stewardship is a wealth of in- the long-range transportation infrastructure
formation and can be found at: http://www.en- plan, habitat conservation plans, local land use
vironment.fhwa.dot.gov/strmlng/ndex.asp plans for accommodating population growth
and economic development plans.
Context sensitive solutions
This approach moves the focus from the
Context sensitive solutions is an important project-level to the landscape level so that all
initiative of the FHWA. It is a process for de- the natural systems and the transportation net-
veloping transportation projects within the en- work can be studied as an integrated whole.
vironmental and social context of the sur- This can help agencies collaborate at the re-
rounding natural environment and the commu- gional level and consider a number of planned
nity. Road projects are undertaken with the in- projects together rather than piecemeal. This
volvement of the communities surrounding the can allow more innovative and long-term solu-
project and inclusion of the ecological data of tions. It also can allow planners to put together
the area. Project design is based on the full larger areas for mitigating environmental im-
consideration of community input and envi- pacts of transportation projects and create
ronmental conditions such as water, soils, ero- higher quality habitat.
sion, wildlife and aquatic habitat, cultural and
Some areas where collaboration and inte-

43
 Technical Paper
Tools and Processes for Implementing a Road Ecology Approach to Transportation and the Environment

gration of planning processes include: re- series bringing expert panels on transportation
gional integrated watershed management plans and the environment to professionals via
that enable small water agencies in an area to web-cast. CTE also organizes the International
work together to manage an entire watershed conference on the Environment and Transpor-
more effectively and protect it from growth, tation (ICOET), every two years which brings
state wildlife action plans that identify wildlife international experts together to present the
species for protection and measures each state latest research and best practices to profession-
will take for protection and conservation plans als in transportation and the environment. Op-
for preserving large land areas from urbaniza- portunities for distance learning through CTE
tion. One good example is the State of Flor- can be found at http://www.itre.ncsu.edu/CTE
ida Department of Transportation Efficient /index.asp.
Transportation Decision Making Process
(ETDM) that provides a large GIS database of Western Institute of Technology (WTI),
environmental resources and land use data for Montana State University, Bozeman, Montana,
use in a collaborative transportation planning conducts research on a wide range of transpor-
process that also streamlines the environmental tation topics, including road ecology. The focus
permitting required for delivery of projects. in road ecology research is on wildlife crossing
and animal-vehicle interactions. WTI can be
Part IV. Road Ecology Research and found at http://www.coe.montana.edu/wti/.
Education
The UC Davis Road Ecology Center
University Centers provide focus for
(REC), University of California, Davis, Cali-
collaboration of scientists
fornia, fosters a synergistic approach to solving
In the United States, there are three uni- road ecology problems, facilitating collabora-
versity centers with a focus on transportation tive projects among scientists, policy makers,
and the environment. The university center is a transportation planners and engineers. The
model and a key driving force in establishing REC has been successful in elevating aware-
road ecology as a focus for research and policy ness of key issues among academic researchers
development. The three centers and their spe- and professionals through research and educa-
cific areas of activity are listed below. tional workshops. The Center also works with
federal, state and local government agencies to
Center for Transportation and the Envi- identify scientific and applied research needed
ronment (CTE), North Carolina State Univer- by policy makers and professionals in trans-
sity, Raleigh, North Carolina, focuses on edu- portation and the environment. This enables
cation and training in transportation and the the identification of high-priority research di-
environment. CTE hosts an excellent speaker rections. The REC can be found at http://road

44
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.31-47

ecology.ucdavis.edu/. In addition to university academic educa-

tion, the Road Ecology Center, as well as CTE
Interdisciplinary road ecology research and WTI, provide education for transport- ation
needs assessments professionals who put road ecology science
into practice to address problems in the deliv-
Some of the challenges to successfully in-
ery of transportation projects. Technology
tegrating the disciplines of ecology and trans-
transfer workshops are organized to engage the
portation lie in lack of scientific basis for poli-
professional community in developing the sci-
cies and practices that help transportation
ence of road ecology as well as applying it in
agencies meet goals of environmental protec-
their practice. Summaries and presentations
tion. The Road Ecology Center has developed a
from REC workshops are posted on the Cen-
research needs assessment workshop approach
ter’s website at: http://roadecology.ucdavis.edu/
to identifying focus areas for agencies respon-
events.html
sible for transportation and resource protection.
The workshop provides a small-group forum to Road Ecology Center Visiting Scholars
discuss common challenges and barriers to en- Visiting scholars at UC Davis work on projects
vironmental protection in transportation project of high international interest and relevance.
delivery. This interactive approach to re- The REC continues to expand partnerships
search needs assessment enables the university among Pacific Rim universities and transport-
to respond effectively to current, real-world ation officials.
needs of those agencies responsible for envi-
ronmental protection and providing transporta- Part V. International Conference on Ecol-
ogy and Transportation
tion infrastructure in a way that could not oth-
erwise occur. The International Conference on Ecology
and Transportation, or ICOET, is an interna-
Road ecology education
tional gathering of scholars and professionals to
The first-ever graduate course in road share the latest research findings that integrate
ecology is planned for Spring 2007, to include ecological awareness and best practices with
experiential learning opportunities that allow transportation projects, to create sustainable
students from ecology, engineering and social transportation systems. The ICOET meetings
sciences to work together with agencies on are held every two years and are organized by
real-world problems in transportation and the Center for Transportation and the Environ-
ecology, again emphasizing both the inter- dis- ment (North Carolina State University). The
ciplinary nature and the interactive aspects of next meeting will be held in May 2007 in Little
road ecology research. Rock, Arkansas. Proceedings of the past con-
ferences are found on the ICOET website at

45
 Technical Paper
Tools and Processes for Implementing a Road Ecology Approach to Transportation and the Environment

http://www.icoet.net/index.asp.

Conclusion

Addressing road networks and natural

systems in an integrated fashion is a complex
contemporary issue. Solutions are urgently
needed in the face of rapid economic growth
and urbanization. The science of road ecology
is a promising new multi-disciplinary field that
is being developed to address the range of en-
gineering, planning and ecological solutions to
problems such as fragmentation of wildlife
habitat, soil erosion and sedimentation caused
by roads, roadside plant communities and
re-vegetation with native species, wetlands
restoration, and storm water run-off and water
quality impacts. Scientific research in turn
leads to innovative tools for practitioners in
transportation departments and resources pro-
tection agencies. Such scientific and cost-ef-
fectiveness analyses are needed by pol-
icy-makers to achieve sustainable transporta-
tion systems.

46
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.31-47

References ence and Solutions. Island Press. Washington

DC.
[1] Clarkin, K., and A. Conner, M.J. Furniss, B. [7] Foster, M.L. and S.R. Humphrey. 1995. Use
Gubernick, M. Love, K. Moynan, S. Wil- of highway underpasses by Florida panthers
son-Musser. 2005. National inventory and and other wildlife. Wildlife Society Bulletin
assessment procedure for identifying barriers 23(1):95-100.
to aquatic organism passage at road-stream [8] Gordon, K. and S.H. Anderson. 2003. Mule
crossings. USDA Forest Service, National deer use of underpasses in western and
Technology and Development Program, 444. southeastern Wyoming. In: Irwin C.L.,
E. Bonita Avenue, San Dimas, CA 91773. Garrett P., McDermott K.P., eds, Proceedings
Available online at of the International Conference on Ecology
http://stream.fs.fed.us/fishxing/publications/P and Transportation Center for Transporta-
DFs/NIAP.pdf tion and the Environment, North Carolina
[2] Clevenger, A.P. and B. Chruszcz, K. Gunson. State University, Raleigh NC.
2001a. Drainage culverts as habitat linkages [9] Harper-Lore, and M. Wilson (eds.). 2000.
and factors affecting passage by mammals. Roadside Use of Native Plants. Washington,
Journal of Applied Ecol- D.C., Island Press.
ogy 38 (6), 1340-1349. [10] Jaarsma, C.F. and G.P.A. Willems. 2002.
[3] Clevenger, A.P. and B. Chruszcz, K.E. Reducing habitat fragmentation by minor ru-
Gunson. 2001b. Highway mitigation fencing ral roads through traffic calming. Landscape
reduces wildlife-vehicle collisions. Wildlife and Urban Planning. 58:125-135.
Society Bulletin 29(22): 646-653. [11] Müller, S. and G. Berthoud. 1995. Fauna and
[4] Clevenger, A.P. and N. Waltho. 2000. Factors traffic safety: Manual for Civil Engineers.
influencing the effectiveness of wildlife un- LAVOC, Ecole Polytechnique Federale de
derpasses in Banff National Park, Alberta, Lausanne, Switzerland.
Canada. Conservation Biology 14(1): 47-56. [12] Van Langevelde, F. and C.F. Jaarsma. 2005.
[5] Davenport, John, and J.L. Davenport (eds.). Using traffic flow theory to model traffic
2006. The Ecology of Transportation: Man- mortality in mammals. Landscape Ecology
aging Mobility for the Environment. The 19(8): 895-907.
Netherlands, Springer.
[6] Forman, R.T.T. and D. Sperling, J.A. Bisson-
ette, A.P. Clevenger, C.D. Cutshall, V.H.
Dale, L. Fahrig, R. France, C.R. Goldman, K.
Heanue, J.A. Jones, F.J. Swanson, T. Turren-
tine, T.C. Winter. 2003. Road Ecology: Sci-

47
 Technical Paper
Modeling the Relationships Between Benthos and Stream Environmental Variables

Modeling the Relationships Between Benthos and Stream

Environmental Variables

Lilian D. Minja1 (Corresponding author), Shaw L. Yu2, Howard Epstein3

1
Department of Environmental Engineering, Faculty of Lands and Environmental Engineering, University
College of Lands and Architectural Studies, P. O. Box 35176, Dar es Salaam, Tanzania. Email:
[email protected]
2
Department of Civil Engineering, University of Virginia, P. O. Box 400742 , Charlottesville, VA 22904-4742.
Email: [email protected]
3
Department of Environmental Science, University of Virginia, Email: [email protected]

Abstract

This study investigated the quantitative relationships between benthic macroinvertebrates and
stream environmental variables using generalized additive models. The benthic data used in the
study were collected by the Virginia Department of Environmental Quality from 1994 through 2001.
The benthic macroinvertebrate assemblages were represented by five benthic metrics namely taxo-
nomic richness, proportion of the two most dominant taxa, proportion of sensitive taxa, proportion
of tolerant taxon, and proportion of shredders. Twenty environmental variables, of which eight
were physicochemical and twelve were habitat related, represented stream environmental conditions.
The study findings suggest a sharp decline in taxonomic richness at total suspended solids concen-
tration of 9 mg/L to 17 mg/L. At flow rates above 85 cfs, taxonomic richness decreases with in-
creasing flow rate. As the dissolved oxygen concentration increases, the proportion of sensitive taxa
also increases, while the proportion of tolerant taxon and the two most dominant taxa decreases.
None of the physicochemical variables showed a significant effect on the proportion of shredders.
Also, the variations of pH within the pH range observed in the case study streams (pH 6 to pH 10)
had no significant effects on the responses of the benthic metrics. For the habitat variables, taxo-
nomic richness and sensitive taxa positively correlated with channel flow status and substrate, re-
spectively. Tolerant taxon negatively correlated with substrate.

Keywords: Stream Environmental; Benthos; macroinvertebrate;

48
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.48-65

水底生物與河流環境因子相關模式之探討

本研究對水底大型無脊椎動物（Macroinvertebrates 或 MI）與河流環境因子間量化性之
關係作一探討。因 MI 常用為河川生態狀況之重要指標，故此種量化性之關係，可用來做改
善環境因子（如流量及水質參數）之根據。

本研究所用之數據為維吉尼亞州環境品質局於 1994 年至 2001 年所收集之資料，MI 族群

之分類指標為物種之數量及多樣性、數量最多之二個物種、最敏感物種之比例，以及最抗污
染物種之比例。考慮之溪流環境因子共有 20 個，其中 8 個為物理及化學方面之因子，其他
12 個則與棲息地有關。

研究結果顯示在溪流中懸浮固體（TSS）濃度從 9 mg/l 升高到 17 mg/l 時，MI 物種之數

量及多樣性有顯著的減少，流量到 85 CFS 以上時亦有相同之情形。在溶氧量增加時，最敏感
物種之比例亦隨之增加，而最抗污物種及數量最多之二個物種之比例則相對減少。另外，所
用之數據之 pH 值在 6 至 9 之間，對 MI 似無任何影響。在棲地因子方面，MI 族群數量及多
樣性，以及最敏感物種之比例，都與流量及河床介質有正面之相關關係，而最抗污物種比例
則與河床介質呈負面之相關關係。

關鍵詞：溪流環境，水底生物，大型無脊椎動物，統計模式分析

49
 Technical Paper
Modeling the Relationships Between Benthos and Stream Environmental Variables

Introduction flow rates and the natural water chemistry.

Furthermore, in some cases a selected reference
Over the past decades, water resources
site is a minimally-impaired site or least-dis-
managers and the public have embraced the use
turbed site due to lack of a non-impaired stream
of benthic metrics to supplement physico-
in the area (Plotnikoff and Wiseman, 2001). A
chemical analysis in monitoring and assessment
minimally-disturbed site is a site that has ex-
of stream environmental conditions. Changes in
perienced very little historical human activity
benthic metrics are predictable when environ-
that alters stream integrity (Wiseman, 2003). A
mental equilibrium is upset due to environ-
least-disturbed site is a site that has been de-
mental stressors such as habitat degradation or
graded by previous human activities, but exhib-
pollution (Barbour et al., 1999). Generally, as-
its some level of recovery (Wiseman, 2003). It
sessment of benthos in an impaired stream
could be argued that the minimally- or
reach is conducted and the information ob-
least-disturbed sites do not provide adequate
tained is used to compute benthic metrics. Also,
representation of biotic integrity in an undis-
water quality analyses are carried out, along
turbed site.
with evaluation of physical stream habitat fea-
tures. The results of the assessment are com- In an attempt to provide an alternative to
pared to benthic metrics, water quality and use of reference sites, this study carried out
habitat condition at a reference site. The refer- generalized additive modeling (GAM), which
ence site needs to have characteristics (location, is described by Hastie and Tibshirani (1990), to
elevation, geology and hydrology) similar to investigate the quantitative relationships be-
the waterbody being evaluated, and should tween benthic macroinvertebrates and stream
support a viable, diverse benthic macroinverte- environmental variables. The environmental
brate community (Barbour et al., 1999). Limi- variables investigated comprised eight phys-
tations in use of reference sites to identify and icochemical variables and twelve habitat vari-
evaluate the level of impairment in a target ables. The benthic macroinvertebrate assem-
stream have been pointed out in the literature. blages were represented by five benthic metrics
Gergel et al. (2002) have observed that identi- namely; taxonomic richness, proportion of
fication of reference sites can be a challenge, the two most dominant taxa, proportion of
especially for large rivers. Walker et al. (2002) sensitive taxa (Ephemeroptera, Plecoptera and
point out that it is virtually impossible to select Trichoptera), proportion of tolerant taxon
a comparable reference site that exactly (Chironomidae), and proportion of shredders.
matches the characteristics of the target stream. According to Barbour et al. (1999), taxonomic
They especially note that reference sites are richness decreases with a decrease in water
often different from the target stream in terms quality and/or habitat degradation as the less
of watershed sizes, physical characteristics, tolerant species are eliminated. The proportion

50
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.48-65

of two most dominant taxa is a tolerance metric servations instead of 2624. The VDEQ stores
that measures the dominance of the two most these data in a data management system called
abundant taxa. A community with a high pro- Ecological Data Application System (EDAS)
portion of dominant taxon indicates that the (TetraTech, 1999), developed for use with Mi-
community is under the influence of environ- crosoft Access®. The EDAS database was used
mental stress (Plafkin et al., 1989). The metric to obtain information on the benthic metrics
referred to as proportion of sensitive taxa that were selected to illustrate changes in ben-
represents aquatic insects that are intolerant of thic macroinvertebrate assemblages along en-
pollution and other environmental stressors. vironmental gradients.
Thus, the metric usually decreases with in-
creasing environmental stress. The proportion
of tolerant taxon metric increases with increas-
ing pollution and environmental stress. Shred-
ders are sensitive to riparian zone impacts;
therefore the proportion of shredders increases
as the riparian vegetation zone width increases.

Methods Data sources and acquisition

methods Figure 1. Location of benthic sampling sites in
Virginia, U.S.A.
The benthic data used in this study were
collected by the Virginia Department of Envi- The physicochemical variables that were
ronmental Quality (VDEQ) from 164 benthic investigated are total suspended solids (TSS),
sampling sites (Figure. 1) in 1st through 4th total phosphorus (TP), nitrate (NO3), dissolved
order streams located throughout western Vir- oxygen (DO), water temperature, electrical
ginia, from 1994 through 2001. The benthic conductivity, pH, and flow rate. Data on in-
sampling sites span five ecoregions namely the stantaneous measurements of DO, water tem-
North Piedmont, the Blue Ridge Mountains, the perature, electrical conductivity, and pH were
Central Appalachian, the Piedmont, and the available in the EDAS database (TetraTech,
Central Appalachian Ridges and Valley. The 1999). These measurements were taken at the
VDEQ used the U.S. EPA rapid bioassessment time when benthic sampling was carried out.
protocols (Barbour et al., 1999) for these bio- Data for TSS, NO3, and TP were obtained from
logical stream surveys, and sampling was car- the VDEQ website of surface water monitoring
ried out twice per year (spring and autumn). program. These data were collected around the
However, not all benthic sampling sites were time period that the benthic sampling was con-
sampled every year from 1994 through 2001. ducted.
Therefore, the benthic data consist of 839 ob-

51
 Technical Paper
Modeling the Relationships Between Benthos and Stream Environmental Variables

Some benthic sampling sites are located in gauged sites to the benthic site. The gauged site
gauged streams. Therefore, stream flow data for that was closest to the benthic sampling site
these sampling sites were obtained from the was selected for generating the synthetic flow
National Water Information System (NWIS) series.
online database of U.S. Geological Survey
(USGS). Synthetic flow series were generated Habitat assessment data for each benthic
for the benthic sampling sites that had no sampling site were provided in the EDAS da-
stream flow data using the Drainage-area-ratio tabase (TetraTech, 1999). These data were ob-
(DAR) method (Hirsch, 1979). The DAR tained through visual assessment of habitat
method relies on the assumption that the ratio features at the time of benthic sampling.
of flows at the outlet of two nearby watersheds Twelve features of physical habitat were rated
is equal to the ratio of the drainage areas of the on a scale of 1 to 20, in which 20 corresponds
two watersheds. The equation for estimating to highest quality of the habitat feature being
the synthetic flow series is expressed as (Equa- assessed. The habitat features assessed were: 1)
tion 1): bank stability, 2) bank vegetative protection, 3)
channel alteration due to man-induced activities,
⎛A ⎞ 4) channel flow status, 5) embeddedness of
y i = ⎜ y ⎟ xi [1]
⎝ Ax ⎠ stream, 6) epifaunal substrate, 7) grazing or
other bank disruptive pressure, 8) in-stream
where yi is the estimated ith flow at cover, 9) frequency of riffles, 10) riparian
the ungauged site, Ay is the drainage area at the vegetation zone width, 11) sediment deposition
ungauged site, Ax is the drainage area at the in stream, and 12) velocity-depth regimes of
gauged site, and xi is the ith observed flow at stream.
the gauged site. The procedure for generating
the synthetic flow series for a target benthic Statistical analysis
sampling site involved identification of a
gauged site that is close to the benthic sampling The Kruskal-Wallis test, a non-parametric
site. This was achieved by entering location ANOVA, was used to examine the variation in
information (latitude, longitude and name of benthic macroinvertebrate assemblages in order
county) of the target benthic sampling site into to determine if it was appropriate to analyze the
NWIS database, which then generated all benthic data collectively. The test was specifi-
gauged sites close to the target site. The loca- cally carried out for the following reasons: 1)
tion of the benthic sampling site and all the ad- Since the benthic data used in this study were
jacent gauged sites were then mapped using collected during the spring and autumn seasons,
ArcView version 8.3 (ESRI, 2002), which al- the observed variability in the benthic macro-
lowed visual assessment of proximity of the invertebrate community structure could have

52
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.48-65

been due to seasonal variations and not an- model-driven. That is, in contrast to some ana-
thropogenic activities such as pollution or lytical procedures (e.g., ordination and linear
stream habitat degradation; 2) The fact that the regression models), generalized additive mod-
benthic sampling sites are located across four eling does not make a priori assumptions about
different stream orders, the observed variation underlying relationships, thus allowing the data
in benthic macroinvertebrate assemblages may to determine the fit of the model instead of the
be caused by natural variation of stream condi- model determining the acceptability of the data.
tions as it flows from head- waters to down- Because of the non-parametric nature of GAM,
stream reaches; 3) Since the benthic sampling there are no associated regression equations. In
sites span five ecoregions, there is a possibility order to improve the distributional characteris-
that the observed variability in the diversity of tics of regression residuals of GAM, the taxo-
the organisms is caused by geographic classifi- nomic richness metric was logarithmic trans-
cation. formed while the proportional abundance met-
rics were arcsine-square root transformed. Dis-
Principal component analysis (PCA) was tribution of total suspended solids data was
performed on the eight physicochemical vari- strongly skewed therefore this variable was log
ables and twelve habitat variables. The analy- transformed. The five benthic metrics were re-
sis was performed due to the fact that it is gressed against the physicochemical and habitat
common for environmental variables to be sig- variables to reveal the relationships between
nificantly correlated. Therefore, there was a the metrics and the predictor variables. The
chance that some environmental (predictor) significance of a predictor variable in a gener-
variables would be redundant if all twenty alized additive model was assessed using
variables were included in the development of model deviance. The deviance for a fitted
the benthos-environment relationships. Ac- model is defined as
cordingly, the predictor variables were reduced
to a smaller number without losing too much D = -2 ln [Lc / Ls] [2]
information. This approach of dimension re-
duction is a common practice in quantitative where Lc is the likelihood of the current
ecology (Legendre and Legendre, 1998). model, and Ls is the likelihood of the saturated
model. A saturated model uses the observed
Development of benthos-environmental data as fitted values and therefore ascribes all
variables relationships of the variation in data. Generally, deviance
plays the role of the residual sum of squares for
Generalized additive modeling (Hastie and generalized models, and is used to compare
Tibshirani, 1990) is a non-parametric regres- models (Hastie and Tibshirani, 1990). For each
sion technique, which is data-driven rather than smoothing function in a given model, a χ2-test

53
 Technical Paper
Modeling the Relationships Between Benthos and Stream Environmental Variables

was performed to compare the deviance be- vertebrate assemblages between spring and au-
tween a model with a predictor variable xj and tumn seasons, among the five ecoregions, and
the model without the predictor variable. A among the four stream orders. Benthic
predictor variable xj was deemed to have a re- macroinvertebrate assemblages did not vary
gression effect if p < 0.05 (Xiang, 2001). The significantly (p>0.05) among the five ecore-
partial residual diagnostic plots of the general- gions. Similarly, there were no significant dif-
ized additive models of benthic metrics versus ferences (p>0.05) in benthic macroinvertebrate
environmental variables were examined to de- assemblages between spring and autumn sea-
termine how responses of the benthic metrics sons. Significant differences (p<0.01) among
change along gradients of environmental vari- benthic macroinvertebrate assemblages were
ables. indicated among the four stream orders in
which the benthic sampling sites were located.
Due to a lack of sufficient field data for Therefore, it was deemed inappropriate to ana-
evaluating the developed models of ben- lyze collectively the benthic data collected
thos-environment relationships, the robustness from different stream orders because of the
of the models was evaluated using the bootstrap confounding effects of variations in stream or-
method (Efron and Tibshirani, 1991). The der. The benthic data were then divided into
bootstrap algorithm was run and created 100 subsets, in which each subset comprised of data
bootstrap samples. Then generalized additive collected from one stream order. Subsequent
modeling was performed on the 100 bootstrap statistical analyses were restricted to data col-
samples to examine if the benthos- environ- lected from one stream order, which served as a
ment trends in these bootstrap samples were typical case. In this case, the third order
similar to the trends that had been developed streams were selected, and the data pertinent to
using field data. this stream order consisted of 280 observations
from 52 sampling sites.
Results Statistical analysis

Table 1 shows the results of Kruskal-Wal-

lis test on the variability of benthic macroin-
Table 1. Results of Kruskal-Wallis Test: p-values on the variability of benthic macroinvertebrate
assemblages between spring and autumn seasons, among the five ecoregions, and among
four stream orders.
Taxonomic Proportion of two Proportion of sen- Proportion of Proportion of

richness most dominant taxa sitive taxa tolerant taxon shredders

Seasons 0.41 0.29 0.08 0.36 0.77

Ecoregions 0.20 0.09 0.28 0.41 0.09

54
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.48-65

Order 0.00 0.01 0.00 0.01 0.00

For the physicochemical variables, results

of PCA revealed that the first four principal For the twelve habitat variables, results of
components sufficiently represented the vari- PCA indicated that the first three principal
ability in the physicochemical data. These four components sufficiently represented variability
components were retained according to the in the habitat features. These three components
broken-stick cut-off criterion (Legendre and explained approximately 68.4% of the variation
Legendre, 1998). That is, the individual vari- in the habitat variables. Table 3 shows the
ances of the first four principal components loadings of the components.
exceeded those of their broken-stick model
counterparts. The four components accounted
for approximately 83% of the total variation in
the physicochemical variables. The first prin- Table 3. Rotated principal components of
cipal component accounted for 34% of the habitat variables
variation in the data set, the second component Component

accounted for 22% of the remaining variation, 1 2 3

and the third and fourth components accounted Epifaunal substrate (benthic
0.91
for 16% and 11% of the remaining variation, macroinvertebrate)
respectively. Table 2 shows how the variables Riffle frequency of stream 0.84
loaded on the principal components. Embeddedness of stream 0.83
In-stream cover 0.79 -0.31
Table 2. Rotated principal components of Velocity-depth regimes of stream 0.78
physicochemical variables Sediment deposition in stream 0.68 0.45

Component Riparian vegetation zone width 0.88

Variable
1 2 3 4 Bank vegetative protection 0.75

TSS 0.96 Channel alteration due to man-

0.75
NO3 0.95 induced activities

TP 0.78 Bank stability 0.42 0.61 0.36

DO -0.87 Grazing or other bank disruptive

0.59
Temperature 0.86 pressure

Flow rate 0.93 Channel flow status 0.30 0.95

Conductivity 0.37 -0.71 0.32 Note: Only loadings above 0.3 are displayed
PH 0.97
Variables that had the greatest amount of
Note: Only loadings above 0.3 are displayed.
axis loading in PCA (i.e., TSS, DO, flow rate,

55
 Technical Paper
Modeling the Relationships Between Benthos and Stream Environmental Variables

and pH for physiochemical variables, and epi- Spearman’s correlation coefficients of the se-
faunal substrate, riparian vegetation zone width lected environmental variables are shown in
and channel flow status for habitat variables) Table 4. The small correlation coefficients
were selected to represent suites of covarying (r<0.25) indicate that the variables are inde-
variables in the subsequent statistical analyses. pendent.

Table 4. Correlation table (r-values) of the selected environmental variables

TSS DO pH Flow rate Substrate *RVZW **CFS

TSS 1.00 -0.01 0.06 0.07 -0.19 0.02 0.09

DO 1.00 0.07 0.04 -0.07 -0.08 0.01

PH 1.00 -0.16 0.11 -0.17 -0.05

Flow rate 1.00 -0.13 0.06 0.20

Substrate 1.00 0.17 -0.09

* RVZW 1.00 -0.05

** CFS 1.00

*RVZW = riparian vegetation zone width; **CFS = channel flow status

Table 5. Physicochemical and habitat variables that have significant association with benthic met-
rics. Blank spaces indicate that the variable is not significant at p < 0.05
Physicochemical variables Habitat variables
Metric
TSS DO pH Flow rate Substrate * RVZW **CFS

Taxonomic richness < 0.001 < 0.05 < 0.01

Proportion of two-
< 0.01
most dominant taxa

Proportion of sensi-
< 0.001 < 0.001
tive taxa

Proportion of tolerant
< 0.05 < 0.01
taxon

Proportion of shred-
< 0.05
ders

*RVZW = riparian vegetation zone width; **CFS = channel flow status

56
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.48-65

Benthos-environmental variables relation-

ships

The results of generalized additive mod-

eling indicated that not all benthic metrics are
significantly associated with the environmental
variables (Table 5).

2-B
Figures 2A and 2B show the gradient of
taxonomic richness versus TSS and flow rate,
Figure. 2. Partial residual values and GAM fit
respectively. Figure 2A shows that taxonomic
for (2-A) taxonomic richness versus
richness is negatively correlated at high con-
log(TSS+1), (2-B) taxonomic rich-
centrations of TSS. The partial residual values
ness versus flow rate. The 95% con-
in Fig. 2A seem to line up along log(TSS+1) =
fidence limits are shown as dotted
0.6. This is because the minimum detection
lines.
limit of the method that was used to analyze
TSS is 3 mg/L. Therefore, all observations be- Figure 3 shows changes in the proportions
low this TSS concentrations were reported as 3 of sensitive taxa, the two most dominant taxa
mg/L. Figure 2B shows that taxonomic richness and tolerant taxon along the gradient of DO
positively correlates with flow rates less than concentration. The proportion of sensitive taxa
25 cfs, then becomes uniform over flow rates of increased as DO concentration increased (Fig-
25 cfs to approximately 85 cfs. At flow rates ure. 3A), while the proportion of tolerant taxon
above 85 cfs, taxonomic richness decreases (Figure. 3B) and the proportion of the two most
with increasing flow rate. dominant taxa (Figure. 3C) decreased with in-
creasing DO concentrations from 7 mg/L to 10
mg/L, then leveled off. It is noted that al-
though Figures 3B and 3C show similar trends,
different types of taxa account for the two met-
rics represented in the figures. Fig. 3B repre-
sents the proportion of the two most dominant
taxa in which the taxa often varied from one
observation to another, while Fig. 3C represents
the proportion of tolerant taxon, which was
2-A
sometimes accounted for in the two most
abundant taxa. The results of generalized addi-

57
 Technical Paper
Modeling the Relationships Between Benthos and Stream Environmental Variables

tive modeling indicated that the observed range shown as dotted lines.
pH range in the case study streams (i.e., pH 6 to
10) did not cause significant changes in the Figures 4 and 5 show the response of ben-
benthic metrics. Also, none of the physico- thic metrics to changes in habitat features. It is
chemical variables had a significant effect on notable that partial residual values in these fig-
the proportion of shredders. ures are lined up in perfect vertical lines. This
is because the scores that describe the quality
of habitat features are discrete (rather than con-
tinuous) variables. The proportion of sensitive
taxa positively correlated with substrate (Figure.
4A), while the proportion of tolerant taxon
negatively correlated with substrate (Figure.
4B). Taxonomic richness increased with in-
creasing score of channel flow status (Figure.
3-A 5A).

3-B
4-A

3-C
4-B
Figure 3. Partial residual values and GAM fits for
(3-A) sensitive taxa, (3-B) dominant taxa, Figure 4. Partial residual values and GAM fits
and (3-C) tolerant taxon versus dissolved for (4-A) sensitive taxa and (4-B)
oxygen. The 95% con- fidence limits are tolerant taxon versus epifaunal sub-

58
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.48-65

strate. The 95% confidence limits are Figures 6A to 6C are typical and repre-
shown as dotted lines. sentative bootstrap curves generated in the
evaluation of the developed benthosenviron-
The proportion of shredders increased ment relationships. The bootstrap curves show
with increasing score of riparian vegetation trends of benthos environment relationships
zone width (Figure. 5B), except for scores of similar to those originally developed, indicative
approximately 9 to 12, there was a decrease in of the robustness of the developed models.
the proportion of shredders. Changes in these
three habitat variables had no significant effect
on the proportion of the two most dominant
taxa.

6-A

5-A

6-B

5-B

Figure 5. Partial residual values and GAM fit

(5-A) for taxonomic richness versus
channel flow status. (5-B) for shred-
ders versus riparian vegetation zone
width. The 95% confidence limits are
shown as dotted lines.
6-C

59
 Technical Paper
Modeling the Relationships Between Benthos and Stream Environmental Variables

Figure 6. Typical examples of bootstrap GAM can be interpreted that the first component re-
curves for taxonomic richness versus flects a water quality process that involves nu-
(6-A) log(TSS+1) and (6-B) flow rate. trients and sediment interrelations. The second
(6-C) sensitive taxa versus dissolved component suggests an oxygen exchange proc-
oxygen concentration. Only 15 of the ess in streams, and its dependence on water
100 bootstrap curves are shown. temperature as indicated by the negative corre-
lation of DO with water temperature. The third
Discussion Variability in benthos component involves the relationship between
assemblages flow and electrical conductivity. An increase in
flow tends to dilute the dissolved salts with a
The observed variability in benthic
resultant reduction in electrical conductivity.
macroinvertebrate assemblages among the four
The fourth component suggests acidity proper-
stream orders was probably caused by the same
ties of streams. With respect to habitat vari-
factors as those suggested by Vannote et al.
ables, the first component mainly corresponds
(1980). That is, the variation in composition
to habitat features in the stream channel. The
and structure of benthic macroinvertebrate as-
second component represents habitat features
semblages along the longitudinal gradient of a
and activities associated with the surrounding
river system is linked to shifts in organic matter
environment of a stream channel. The third ro-
supply. Thus, the morphological-behavioral
tated component corresponded to features asso-
adaptations of invertebrates reflect shifts in
ciated with flow processes in the streams.
types and locations of food sources within the
stream size, and downstream communities are Benthos-environment relationships
fashioned to capitalize on upstream organic
matter processing inefficiencies (Vannote et al., Previous studies have observed that sus-
1980). The observed low classification strength pended sediments have deleterious effects on
for the ecoregions supports previous studies benthos (Wohl and Carline, 1996; Wood and
conducted in the region (Waite et al., 2000; Armitage, 1999). However, these studies have
Yuan and Norton, 2003). not indicated TSS concentrations that cause
adverse impacts on benthos. On the other hand,
Selection of environmental variables findings of this study suggest that there is a
sharp decline of taxonomic richness at TSS
The rotation of components in PCA al-
concentrations of 9 mg/L to 17 mg/L. It is rec-
lowed better identification of groups of co-
ognized that the variation in the concentration
varying environmental variables. The compo-
of suspended sediment and their deposition, is a
nents reflect the underlying processes that have
result of natural variability of stream flow,
created the correlation among the variables
from high to low flows. For instance, high
(Tabachnick and Fidell, 1996). Therefore, it

60
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.48-65

flows in streams associated with runoff may The observed DO concentrations in the
cause dilution of suspended particulates thereby case study streams (i.e., 7 mg/L to 14 mg/L) are
reducing TSS concentration in a stream. Con- at a high end of the spectrum, even for fresh
versely, high flows can be associated with an clean waterbodies. This may be a reflection of
increased TSS concentration because of erosion very good re-aeration mechanisms such as tur-
and sediment delivery in a stream. Also, high bulence caused by rapids and horizontal direc-
flows may cause scouring of the streambed, tion changes, in addition to the streams being
with a consequent increase in TSS concentra- relatively shallow. The observed decrease in
tion. Then again, high flows can widen stream the proportion of the two most dominant taxa
banks, and ultimately reduce flow velocity, and the proportion of tolerant taxon at DO lev-
which in turn allows settling of sediment. De- els above 7 mg/L can be explained by implica-
spite this complex association between TSS tions of changes in benthic community diver-
and flow rate, it should be noted that this study sity. It can be argued that high DO concentra-
shows that the decline in taxonomic richness tions cause an increase in the diversity of ben-
along the TSS gradient is independent of its thic macroinvertebrate community, and in turn
decline along the gradient of flow rate. This is this reduces the likelihood of a few taxa domi-
evident from the lack of significant correlation nating the community. That is, at higher DO
between TSS and flow rate. levels the dominant taxa and tolerant taxon lost
lose their competitive advantage. In general, an
The effects of stream flow on benthos are important question here is “Do high dissolved
described in Scheidegger and Bain (1995) and oxygen levels have direct adverse impacts on
Poff et al. (1997), but the estimates of flow any taxa, or the impacts are due to competition
rates that are either too high or too low for borne out of other environmental conditions?”
benthos are not indicated. Findings from this This is in view of the fact that when the DO
study indicate that the benthos in the case study levels are high, conditions are favorable for all
streams thrive at flow rates of 25 cfs to 85 cfs aquatic fauna. Therefore, the resulting condi-
(Figure 2B). Nonetheless, it is recognized that tions may enable other non-dominant taxa gain
velocity is a more direct measure of effects of competitive advantage due to their other attrib-
discharge on benthos than flow rate. For exam- utes. Since there are no observations at low DO
ple, a stream may have a high flow rate but low concentrations, the study could not determine
velocity if its cross-sectional area is large. As the effects of low DO on benthos. Dauer et al.
such, the stream may not exhibit significant (1992) observed that characteristics of macro-
effects on benthos. On the other hand, a stream invertebrate communities in waters having low
with low flow rate but high velocity may cause DO (<2 mg/L), include lower species diversity.
significant impacts on benthos as the fast cur- The observed insignificant effects of DO on
rent dislodges the organisms and their habitats. taxonomic richness and the proportion of

61
 Technical Paper
Modeling the Relationships Between Benthos and Stream Environmental Variables

shredders is probably because these two ben- Figure 5A shows taxanomic richness in-
thic metrics include several species of benthos, creases as the channel flow status increases.
which thrive under different levels of DO con- When water covers much of the streambed, the
centration. Thus, the adverse response of some amount of suitable substrate for benthos in-
individual taxons to changes in DO concentra- creases because riffles, cobble, logs, and snags
tion tends to be offset by positive responses of are not exposed (Barbour et al., 1999). Conse-
other taxa, resulting in no noticeable change in quently, taxonomic richness increases due to
the overall metric response to changes in DO the increased areas of good habitat (Barbour et
concentration. al., 1999). Shredders are particularly good in-
dicators of toxic effects when the toxicants (e.g.,
Figure 4A shows that there is a steep in- pesticides, herbicides) involved are readily ad-
crease in the proportion of sensitive taxa be- sorbed to the coarse particulate organic matter
tween substrate scores of 8 and 12, and also (CPOM) such as leaf litter, and either affect
between scores of approximately 15 and 20. No microbial communities colonizing the CPOM
change is observed between scores of about 12 or the shredders directly (Mandaville, 2002).
and 15. The leveling off of proportion of sensi- Logically, as the riparian vegetation zone width
tive taxa between scores of about 12 and 15 increases, more removal of pollutants from
was probably caused by other factors such as runoff takes place due to more retention time
predation, which were not quantified in this attributed to longer flow distance. Specifically,
study. For example an increase in the propor- the longer flow distance allows more time for
tion of sensitive taxa might have triggered pre- processes such as microstraining and plant up-
dation by other organisms, consequently inhib- take of toxicants to take place. However, based
iting the increase in sensitive taxa population. on the findings from this study, it is not clear
However, due to the favorable conditions, the why these processes exhibited significant ef-
sensitive taxa were able to transcend the im- fectiveness only at scores higher than 12 (Fig-
pacts of predation and continue to increase as is ure. 5B). The observed trend at scores less than
seen at scores above approximately 15. The 12 was probably caused by other factors that
proportion of tolerant taxon showed the ex- influenced correlation of the proportion of
pected trend at increasing scores of substrate shredders with scores of riparian vegetation
(Figure. 4B). That is, a wide variety and/or zone width. For instance the riparian vegetation
abundance of substrate in a stream increased may not be able to trap or uptake toxicants in
habitat diversity, and this in turn increased spe- the subsurface flow. Therefore in case streams
cies diversity which might have resulted in a are located in highly permeable soils, toxicants
decline of the tolerant taxon. can reach the streams through the subsurface
flow. Consequently, the toxicants impact the
shredders. Therefore, in such cases the scores

62
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.48-65

of riparian vegetation zone width might not thos-environment relationships have been de-
positively correlate with the proportion of rived using data collected from these streams.
shredders as expected. However, the developed approach to quantify
benthos-environment relationships can be used
Conclusion
for other stream orders or regions.
Using real-world data, this study has
Acknowledgement
demonstrated that the relationships between
benthos and environmental variables can be This paper is based upon work supported
quantified. The study has identified several at- under a Fulbright Scholarship from the Institute
tributes in benthic macroinvertebrate assem- of International Education (IIE), U.S.A., and an
blages that respond to changes in physico- award from Delta Kappa Gamma Society In-
chemical conditions and habitat features. It has ternational (DKGSI), U.S.A., to the senior au-
been observed that benthos vary in their re- thor Any opinions, findings, or recommenda-
sponse to environmental variables. Taxonomic tions expressed in this publication are those of
richness decreased as TSS and flow rate in- the authors and do not reflect the views of IIE
creased. Likewise, the proportion of the two or DKGSI. We thank Alex Barron from Vir-
most dominant taxa and proportion of tolerant ginia Department of Environmental Quality for
taxon decreased with increasing DO concentra- providing the bio-assessment data.
tion. On the contrary, the proportion of sensi-
tive taxa increased as the concentration of DO
increased. In regard to habitat variables, it was
observed that taxonomic richness positively
correlated with channel flow status. Also, the
proportion of sensitive taxa positively corre-
lated with substrate, while the proportion of
tolerant taxon negatively correlated with sub-
strate. Understanding this association between
benthos and environmental variables is critical
for diagnosing correlates of benthos impair-
ment. This is important mainly because at pre-
sent the diagnosis process depends on reference
sites despite the acknowledged limitation of
these sites. The study results are mainly ap-
plicable to third order streams in the case study
area since the developed models of ben-

63
 Technical Paper
Modeling the Relationships Between Benthos and Stream Environmental Variables

References Available at
http://lakes.chebucto.org/H-1/tolerance.pdf
[1] Barbour, M.T., Gerristsen, J., Snyder, B.D.,
[9] Plafkin, J.L., Barbour, M.T., Porter, K.D.,
and Stribling, J.B. 1999. Rapid Bioassess-
Gross, S.K. and Hughes, R.M. 1989. Rapid
ment Protocols for Use in Streams and
Bioassessment Protocols for Use in Streams
Wadeable Rivers: Periphyton, Benthic
and Rivers: Benthic Macroinvertebrates and
Macroinvertebrates and Fish, 2nd Edition.
Fish. U.S. EPA, EPA/440/4-89-001. Wash-
EPA 841-B-99-002. U.S. Environmental
ington, DC, 187 p.
Protection Agency; Office of Water; Wash-
[10] Plotnikoff, R.W. and Wiseman, C. 2001.
ington, D.C.
Benthic Macroinvertebrate Biological Moni-
[2] Dauer, D.M., Rodi Jr, A.J. and Ranasinghe,
toring Protocols for Rivers and Streams.
J.A. 1992. Effects of low dissolved oxygen
Washington State Department of Ecology,
events on the macrobenthos of the lower
Olympia, WA. Publication No. 01-03-028.
chesapeake bay. Estuaries, 15: 384-391.
Available at
[3] Environmental Systems Research Institute
http://www.ecy.wa.gov/biblio/0103028.html.
(ESRI Inc.). 2002. ArcView Software Ver-
[11] Poff, N.L., Allan, J.D. Bain, M.B., Karr, J.R.,
sion 8.3.
Prestegaard, K.L., Richter, B.D., Sparks, R.E.
[4] Gergel, S.E., Turner, M.G., Miller, J.R., Me-
and Stromberg, J.C. 1997. The natural flow
lack, J.M. and Stanley, E.H. 2002. Landscape
regime: A paradigm for river conservation
indicators of human impacts to riverine sys-
and restoration. BioScience, 47 (11):
tems. Aquatic Sciences, 64 (2): 118-128.
769-784.
[5] Hastie, T.J. and Tibshirani, R.J. 1990. Gener-
[12] Scheidegger, K.J. and Bain, M.B. 1995. Lar-
alized Additive Models. New York: Chapman
val fish in natural regulated rivers: assem-
and Hall.
blage composition and microhabitat use.
[6] Hirsch, R.M. 1979. An evaluation of some
Copeia, 1995: 125-135. In: Poff, N.L., Allan,
record reconstruction techniques. Water Re-
J.D. Bain, M.B., Karr, J.R., Prestegaard, K.L.,
sources Research, 15 (6): 1781-1790.
Richter, B.D., Sparks, R.E. and Stromberg,
[7] Legendre, P. and Legendre, L. 1998. Nu-
J.C. 1997. The natural flow regime: A para-
merical Ecology, 2nd English Edition. Am-
digm for river conservation and restoration.
sterdam, The Netherlands. Elsevier Science
BioScience, 47 (11): 769-784.
B.V.
[13] Tabachnick, B.C. and Fidell, L.S. 1996. Us-
[8] Mandaville, S. M. 2002. Benthic Macroin-
ing Multivariate Statistics. 3rd Edition.
vertebrates in Freshwaters - Taxa Tolerance
HarperCollins College Publishers, New York
Values, Metrics, and Protocols. Soil and Wa-
[14] TetraTech. 1999. Ecological Data Applica-
ter Conservation Society of Metro Halifax.
tion System (EDAS).

64
 Technical Paper
J. of Ecotechnology, 2006 (2) : pp.48-65

[15] Vannote, R.L., Minshall, G.W., Cummins, [22] Xiang, D. 2001. Fitting Generalized Additive
K.W., Sedell, J.R. and Cushing, C.E. 1980. Models with the GAM Procedure. SAS Insti-
The river Continuum Concept. Canadian tute Inc., SUGI Proceedings.
Journal of Fisheries and Aquatic Sciences 37: [23] Yuan, L.L. and Norton, S.B. 2003. Compar-
130-137. ing responses of macroinvertebrate metrics to
[16] Virginia Department of Environmental Qual- increasing stress. J. N. Am. Benthol. Soc., 22
ity (VDEQ). 2003. The Quality of Virginia (2): 308-322.
Non-Tidal Streams: First Year Report. Rich- [24]
mond, VA. VDEQ Technical Bulletin
WQA/2002-001.
[17] Waite, I.R., Herlihy, A.T., Larsen, D.P. and
Klemm, D.J. 2000. Comparing strengths of
geographic and non-geographic classifica-
tions of stream benthic macroinvertebrates in
the Mid-Atlantic Highlands, USA. Journal
of the North American Benthological Society,
19 (3): 429-441.
[18] Walker, J.L., Porter, K.R. and Younos, T.
2002. Monitoring Needs to Meet Benthic
TMDL Requirements. National Monitoring
Conference, Madison, Wisconsin, USA.
World Meeting Number 000 6280.
[19] Wiseman, C.D. 2003. Multi-Metric Index
Development for Biological Monitoring in
Washington State Streams. Washington
State Department of Ecology, Olympia, WA.
Publication No. 03-03-035.
[20] Wohl, N.E. and Carline, R.F. 1996. Relations
among riparian grazing, sediment loads,
macroinvertebrates, and fishes in three Cen-
tral Pennsylvania streams. Can. J. Fish. Aquat.
Sci., 53 (Suppl. 1): 260-266.
[21] Wood, P.J. and Armitage, P.D. 1997. Bio-
logical effects of fine sediment in the lotic
environment. Environmental Management, 21
(2): 203-217.

65
 Manuscripts Submission Guidelines
Journal Content
The Journal of Ecotechnology will publish (1) academic papers in watershed
management or ecological engineering, (2) description of local or international
meetings or activities, (3) case studies, (4) technical reports, and (5) review articles.
Fields of interest include, but not limited to, hydrology, water and soil conservation,
ecology, vegetation engineering, environmental engineering, and civil and hydraulic
engineering.
1. Language: All manuscripts must be submitted in English. The manuscript title,
authors, affiliation, and abstract should also include a Chinese translation.
2. Format: Manuscripts shall be typed in Times New Roman, 12-point font, single
column format. Please submit both a Document (.DOC) file and a Portable
Document Format (.PDF) file for editing purposes.
3. Figures: Please do NOT reduce the resolution of illustrations. If illustrations
were not included in the text files, please send them separately with at least 300
dpi resolution.
Manuscripts Submission
1. Please submit your manuscripts by E-mail if the file size is less than 10 MB.
E-mail to: [email protected]
2. Please submit your manuscript stored in CD-ROM disks by mail if the file size is
larger than 10 MB.
Mail to: Mr. Chi-Chang Liu
Water Environment Research Center,
National Taipei University of Technology
No. 1, Sec. 3, Chung-hsiao E. Rd. Taipei, Taiwan, R.O.C. 10608
Note
The editors of the J. of Ecotechnology have the rights to edit manuscripts
submitted for publishing. If authors do not wish their manuscripts to be edited, please
specify. Copyright on all published articles is held by the J. of Ecotechnology.
Manuscripts will not be returned whether the articles are published or not, unless
requested by the authors at the time of submission.

Contact us: www.cc.ntut.edu.tw/~wwwwec/

 稿約
稿件內容
廣邀各界撰寫有關水庫集水區、生態工法之學術論文，以及國內外地方或國際性、學術或工
程界活動報導、案例介紹均可，領域包含水保、水文、生態、植生、環工、土木、水利等相關領
域，凡符合此主題之學術論文、技術報告、理念介紹、文獻回顧等論述，皆歡迎來稿。

來稿須知
1. 本期刊投稿撰寫一律以全英文稿為限，惟論文題目與作者名、機關名、摘要均請同時並列
中英文。
2. 稿件一律橫寫，以 MS WORD 格式之電腦文書軟體編輯，全文統一字體大小 12，字型 Times
New Roman，單欄格式，各標題並請盡量勿套用特殊格式，以利編輯小組調整。
3. 投稿請一併附上 Document (.DOC) 檔案以及 Portable Document Format (.PDF) 檔案，以利
本期刊編輯小組編輯。
4. 圖表若非附於文章內，請另附清晰之圖片檔案(300 dpi 以上)，連同光碟寄至本期刊編輯小
組。

投稿方式（二擇一）
1. 小於 10MB 之檔案，可以 E-mail 寄至 [email protected] 將光碟檔案寄至 10608 台
北市忠孝東路三段一號
台北科技大學土木館三樓 水環境研究中心 劉吉倉先生 收

刊載說明
本期刊對來稿均具有修改縮編權利，如不允許修改者，請於稿件中一併註明。稿件經刊載後，
本期刊可基於學術交流等非營利目的再版、製作光碟、或放置於相關網頁上。
來稿均由編輯委員審查。稿件不論刊載與否，均不退件。如需退件者，請於投稿時註明。

來函照登
本人張俊斌於 2005 年發表於 Ecotechnology 期刊第 1 卷第 2 期第 11 頁至 24 頁之「自然災害
對集水區的景觀生態變遷之研究」文章，其中結果摘錄與部分圖表部份，未經農委會計畫主持人
（應用遙測影像及景觀生態計畫方法於集水區土地利用變遷之研究）－林裕彬先生之同意就直接
引用，在此鄭重向其至上十二萬分之歉意，並衷心感謝能得到林裕彬先生之諒解與不計較。
特此聲明與感謝
立書聲明人：張俊斌
中華民國九十五年八月三十日

歡迎您的意見：www.cc.ntut.edu.tw/~wwwwec/