Amicus Curiae Brief — Decker v. Northwest Environmental Defense Center

Nos. 11-338 and 11-347

IN THE

Supreme Court of the United States

DouG DECKER, in his official capacity as Oregon State

Forester, et al.,

Petitioners,

We

NORTHWEST ENVIRONMENTAL DEFENSE CENTER,

Respondent.

GEORGIA-PACIFIC WEST, INC., et al.,

Petitioners,

Vv.

NORTHWEST ENVIRONMENTAL DEFENSE CENTER,

Respondent.

On Writs of Certiorari to the United States Court of

Appeals for the Ninth Circuit

BRIEF OF DR. KEVIN BOSTON AS AMICUS

CURIAE IN SUPPORT OF RESPONDENT

SHAUN A. GOHO

COUNSEL OF RECORD

EMMETT ENVIRONMENTAL LAW & POLICY CLINIC

HARVARD LAW SCHOOL

6 Everett St., Suite 4119

Cambridge, MA 02138

617-496-2058

sgoho@law.harvard.edu

Counsel for Amicus Curiae

i

TABLE OF CONTENTS

EE I shisiccenniccsmuidsitdabbinsiibainepanensenee i

ee Re Bee i ickctnctiennscinntscdssencssencascnnes ii

INTEREST OF THE AMICUS CURIAE.................... 1

INTRODUCTION AND SUMMARY OF

IIIT’ ‘ichiicincestshternbinnindandeidandiasicidibimsaiinstabiiadiiiiaoucedanaulibidid 2

I ITITTED cinitcnnidicicrittichriesinisiahestinntshaiiniiieasiinniaitindiiniiaaidaabiteatiiiabiiiibes 3

I. Large-Scale, Modern Forestry is an Industrial

SE TTEIITN, sicehcsipsitinecieitniadeoesshdeteiakceihebatebiistiiithdaiininiadiilibetiedinaiaiaiael 3

Il. Active Logging Roads Have Significant Impacts

I, iis ciinititicceeeitentinininiedianndiabiatedateid 6

Fh ION, i cciniccscscictisnieniniieinbinednmmniidienies 8

PE III, cicusntscnindcadscnanichninmnsiontidne 10

i III insist rpntneinncinsinmnpainndbiiaeiitianiioinenuits 11

Ill. Creating a Logging Road is a Complex, Multi-

step Undertaking, with Implications for

Sediment Pollution at Each Step. ...................... 12

A. Road System Planning. ...................cseccseeesees 13

Be I isi inte sccciitaleibnnniniipdieidbichnnaealaanainindaibeide 15

ee I sinc dnechesnesdeenisioamneiidaabenemestiliia 21

EE, TCO SET 23

i: EINER 15. scsnrccsnipisinisesnesieabehasmanieippeieiintt 25

RT sciicincsinitininnstnindanciininncsnidintanaainiicinianstaseeiaiel 27

TABLE OF AUTHORITIES

Page(s)

STATUTES

Be ee Tt TIE vinpiccusdpossesnusniuneesiiveimeddpedanmtabimaienatenes 3

REGULATIONS

De. AGE. Th. GRD GE) vescccsccccscoscscsvscccosscocsens 2

OTHER AUTHORITIES

Robert E. Bilby, Kathleen Sullivan & Stanley H.

Duncan, The Generation and Fate of Road-Surface

Sediment in Forested Watersheds in Southwestern

Washington, 35 FOREST SCI. 453 (1989).................. g

Kevin Boston, Marvin Pyles & Andrea Bord,

Compaction of Forest Roads in Northwestern

Oregon—Room for Improvement, 19 INT’LJ. OF

FOREST ENGINEERING 24 (2008). ...............2.006 15, 22

STEVE CONWAY, LOGGING PRACTICES: PRINCIPLES OF

TIMBER HARVESTING SYSTEMS (1976) .............22-.-006 5

Jacky Croke & Simon Mockler, Gully Initiation and

Road-to-Stream Linkage in a Forested Catchment,

Southeastern Australia, 26 EARTH SURFACE

PROCESSES & LANDFORMS 205 (2000).................... 14

EPA, Guidance Specifying Management Measures for

Sources of Nonpoint Pollution in Coastal Waters

Pail iciiiieddbisencnnaindddatdeninsiaiibidadinasidibiedaniiimsantipudiaeaiind 6

Randy B. Foltz, Sediment Reduction from the Use of

Lowered Tire Pressures, in CENTRAL TIRE

INFLATION SYSTEMS: MANAGING THE VEHICLE TO

SURFACE 47-52 (Soc’y of Automotive

Taney SUMETITII: ‘ocicichaciccensdiniestidcld dettsiiceabiaiastieidiclenpiiianianditininiiasiainea 25

ili

Randy B. Foltz, Traffic and No-Traffic on an

Aggregate Surfaced Road: Sediment Production

Differences (1996) (paper presented at the FAO

Seminar on Environmentally Sound Forest Roads),

available at

http://forest.moscowfsl.wsu.edu/engr/library/Foltz/

I nis sceiaeicccriatdelncaiatinindeduimiiamias 23

Randy B. Foltz & William J. Elliot, Effect of Lowered

Tire Pressures on Road Erosion, 1589 TRANSP.

RESEARCH RECORD 19 (1997) .............2-2ccceeeeeeeees 9, 25

Randy B. Foltz, Gary L. Evans & Mark Truebe,

Relationship of Forest Road Aggregate Test

Properties to Sediment Production, in WATERSHED

MANAGEMENT & OPERATIONS MANAGEMENT 2000

(Marshal Flug et al. eds. 2000) ............................. 23

Randy B. Foltz & Mark Truebe, Locally Available

Aggregate and Sediment Production, 1819B

TRANSP. RESEARCH RECORD 185 (2003)............. 9, 23

Pieter J.B. Fransen, Chris J. Phillips & Barry D.

Fahey, Forest Road Erosion in New Zealand:

Overview, 26 EARTH SURFACE PROCESSES AND

ee Be IED vtitictirincintntnnianicndtietininies 24, 26

GREAT LAKES ENVTL. CTR., NATIONAL LEVEL

ASSESSMENT OF WATER QUALITY IMPAIRMENTS

RELATED TO FOREST ROADS AND THEIR PREVENTION

BY BEST MANAGEMENT PRACTICES (2008).......... 6,11

LOREN D. KELLOGG, PETE BETTINGER & DON STUDIER,

TERMINOLOGY OF GROUND-BASED MECHANIZED

LOGGING IN THE PACIFIC NORTHWEST (1993),

available at

iv

http://ir.ibrary.oregonstate.edu/xmluybitstream/h

I iiiiicdigrtiicigniindndnatinniiinminennan 4

Charles H. Luce & Thomas A. Black, Spatial and

Temporal Patterns in Erosion from Forest Roads,

in LAND USE AND WATERSHEDS: HUMAN INFLUENCE

ON HYDROLOGY AND GEOMORPHOLOGY IN URBAN

AND FOREST AREAS 165 (Mark S. Wigmosta &

Stephen J. Burges eds., 2001) .............ccccececeeeeeeeees 20

Charles H. Luce & Beverley C. Wemple, Jntroduction

to Special Issue on Hydrologic and Geomorphic

Effects of Forest Roads, 26 EARTH SURFACE

PROCESSES AND LANDFORMS 111 (2001)................ 26

Mary Ann Madej, Erosion and Sediment Delivery

Following Removal of Forest Roads, 26

EARTH SURFACE PROCESSES AND LANDFORMS 175

-__. SERESES REP roe Peers eh mea here Or Te RO EOE 21

Walter F. Megahan, Monte Wilson & Stephen B.

Monsen, Sediment Production from Granitic

Cutslopes on Forest Roads in Idaho, USA, 26

EARTH SURFACE PROCESSES & LANDFORMS 153

IIIT i isiciesisisdinicbintepicietiieseieatatiranianamaaidipdineniesadeidihdagaleiiaticidimeaed 26

KEITH MILLS, OR. DEP’T OF FORESTRY, FOREST ROADS,

DRAINAGE AND SEDIMENT DELIVERY IN THE KILCHIS

RIVER WATERSHED (1997), available at

www.oregon.gov/ODF/privateforests/docs/kilchis.p

AER DRE RL ite tek ly Re SORE Ree A Nee eS 8, 9, 10, 11

Jeffry Moll, Ronald Copstead & David Kim

Johansen, U.S. Forest Service, Traveled Way

ey ST IE sicediniiciiteniniciniitebcnndiniaicguditimecaores 17

NATIONAL COUNCIL FOR AIR AND STREAM

IMPROVEMENT, FOREST ROADS AND AQUATIC

Vv

ECOSYSTEMS: A REVIEW OF CAUSES, EFFECTS, AND

MANAGEMENT PRACTICES (2003)....14, 16, 20, 21, 26

Or. Dep’t of Forestry, Forest Practices Technical

Note No. 9, Wet Weather Road Use (2003),

avatlable at

http://www.oregon.gov/odf/privateforests/docs/wets

easonroadusefptechnote9. pdf.................... 22, 23, 24

Or. Dep’t of Forestry, State Forests Program, Forest

SN a caam niin 16

Or. Dep’t of Forestry, Technical Report # 17, Wet

Season Road Use Monitoring Project: Final Report

(2003), available at

http://www.oregon gov/odf/privateforests/docs/road

Bradley T. Piehl, Robert L. Beschta & Marvin R.

Pyles, Ditch-Relief Culverts and Low-Volume

Forest Roads in the Oregon Coast Range,

62 NORTHWEST SCI. 91 (1988).................0.0c0ee0 19, 26

Leslie M. Reid & Thomas Dunne, Sediment

Production from Forest Road Surfaces, 20 WATER

RESOURCES RES. 1753 (1984) .........0c0sesseceecececseeseeeeee 9

Gary J. Sheridan & Philip J. Noske, A Quantitative

Study of Sediment Delivery and Stream Pollution

from Different Forest Road Types,

21 HYDROLOGICAL PROCESSES 387 (2006) ............. 24

Matthew Thompson et al., Intelligent Deployment of

Forest Road Graders, 18 INTLJ. OF FOREST

IED OP CINE ccciccccccccecnsccccsddesdoncetncesesncse 26

U.S. FOREST SERVICE, ROCKY MOUNTAIN RESEARCH

STATION, LEGACY ROADS AND TRAILS MONITORING

PROJECT UPDATE 2012, available at

vi

http://www.fs.fed.us/GRAIP/downloads/

case_studies/2012LegacyRoadsMonitoringProjectU

IIIIICIITTIT sins sccsiiutsdetsnedahenchesisieladidelinemindapiidbanitbdiaictiabiiadandebualiiek 7

Virginia Tech University, Dep’t of Forestry,

Harvesting Process,

http://web1.cnre.vt.edu/harvestingsystems/Harvest

ees eineeniciatnimenaaniienie 4

Beverley C. Wemple, Frederick J. Swanson & Julia

A. Jones, Forest Roads and Geomorphic Process

Interactions, Cascade Region, Oregon, 26

EARTH SURFACE PROCESSES AND LANDFORMS 191

IIT nssisiehticinsniaiadencinitnninbileseainpsitinad idiuleedanpbeiapiaaennniatidinudiadseiiadisiitl 21

Wildlands CPR, Road Reclamation: Measuring

Success 6 (2012) (summarizing Forest Service

results), available at

http://www.wildlandscpr.org/files/GRAIP%20Repor

t%20Wildlands%20CPR_O0.pdf .............. cece eee eee 12

1

INTEREST OF THE AMICUS CURIAE!

Dr. Kevin Boston is an associate professor in the

Forest Engineering, Resources, and Management

Department at Oregon State University, where he

teaches courses in Forest Road Engineering and

Forest Road System Management. He is engaged in

active research on the physical properties of various

materials used in logging road construction, and has

authored or co-authored numerous publications re-

lating to forest road management. Dr. Boston has a

Bachelor of Science in Forestry from Humboldt State

University and a Masters of Forestry and a PhD in

Forest Engineering from Oregon State University,

and is a Registered Professional Forester in Califor-

nia and Professional Engineer, registered in Oregon.

Prior to working at Oregon State University, Dr.

Boston worked as an assistant professor at the Uni-

versity of Georgia and as a lecturer at the New Zeal-

and School of Forestry at the University of Canter-

bury. He has spent many years working in the tim-

ber industry in New Zealand and the western United

States.

Dr. Boston believes that the principles of logging

road engineering and the processes of sediment gen-

eration and delivery from logging roads have not

1 All counsel of record have consented to the filing of this brief.

Petitioners Georgia-Pacific West, Inc., et al., Petitioners Decker,

et al., and Respondent have all filed blanket consents. No

counsel for any party authored this brief either in whole or in

part. No persons other than amicus or its counsel has made

any monetary contribution to the preparation or submission of

this brief. Dr. Boston submits this brief on his own behalf and

not on behalf of his employer Oregon State University.

2

been adequately explained in the other briefs before

the Court. A sound understanding of these concepts

is necessary to inform the Court’s decision about

whether sediment generated from the movement of

heavy machinery and the hauling of timber on log-

ging roads and conveyed to waters of the United

States via engineered drainage systems constitutes a

point source discharge of industrial stormwater.

INTRODUCTION AND

SUMMARY OF ARGUMENT

The main purpose of this brief is to explain rele-

vant facts surrounding the science and engineering of

logging roads and their drainage systems.? These

facts demonstrate that the primary purpose of these

roads is to facilitate industrial logging operations;

that the pipes, ditches, and channels associated with

these roads are designed to prevent road washout

and failure that would disable logging operations;

and that these drainage structures can be a signifi-

cant source of water pollution. 2JA 90. In short, the

sediment pollution caused by the creation and use of

logging roads is not natural in origin and the choices

made in the planning, design, construction, and use

of logging roads can determine the volume of sedi-

° In accord with the usage adopted by Respondent, this brief

uses the term “logging roads” to refer to forest roads that are

designed and constructed for heavy vehicle traffic and used for

active timber cutting and hauling operations. The distinction

between active logging roads and other forest roads is a com-

mon one, recognized, for example, in the Oregon regulations’

definition “active roads,” which “are roads currently being used

or maintained for the purpose of removing commercial forest

products.” Or. Admin. R. 629-600-0100(3).

3

ment pollution that reaches rivers and streams. Asa

result, when this sediment pollution is directly dis-

charged into waters of the United States through

man-made ditches, pipes, culverts, and channels, it

should be subject to the National Pollutant Dis-

charge Elimination System (NPDES) permitting

program under section 402 of the Clean Water Act,

33 U.S.C. § 1342.

ARGUMENT

I. Large-Scale, Modern Forestry is an Indus-

trial Activity.

The active logging roads that are the subject of

this litigation are integral parts of large-scale, com-

plex, industrial operations. Although these logging

roads may eventually be used for other purposes—

such as recreation and fire suppression—their pri-

mary purpose is to enable the harvesting and extrac-

tion of timber and other forest products. This pur-

pose determines the decisions made at each stage of

the process of planning, designing, constructing,

operating, and maintaining these roads.

Logging roads are built to safely and efficiently

bear the heavy vehicle traffic necessary for modern

industrial forestry. This traffic involves both moving

the huge machines used in cutting and processing

trees into the forest and removing timber from the

forest on massive logging trucks. Roads can also be

the site of some of the timber processing, as tree

limbs are removed and stems are cut into logs for

transport to the mill.

Paradoxically, these heavy-duty logging roads—

built to safely bear massive equipment—are also

4

extremely vulnerable. Constructed of dirt and built

along steep and often unstable slopes, logging roads

are susceptible to washout and failure. In order to

protect these roads, which are indispensible to the

logging process, logging companies design and install

complex networks of drainage systems. These pipes,

ditches, and channels are intended to fulfill an es-

sential job: to preserve the integrity of the road by

carrying water away from its surface.

The basic steps involved in getting timber “from

the ‘stump to mill” are roadbuilding, felling, extrac-

tion, processing, loading, and trucking.* Each step in

this industrial process is typically carried out by a

specialized piece of heavy machinery and/or by indi-

viduals with a specialized skill. As described in

detail below,‘ the first step, foundational to all oth-

ers, is roadbuilding, which involves clearing the road

with bulldozers, smoothing the surface with graders,

and installing drainage structures. Once the road is

built, the “felling” (cutting down) of trees is often

performed by mechanical harvesters—tracked or

wheeled machines that can weigh more than 30 tons.

The operator of one of these machines, who sits in a

3 Virginia Tech University, Dep't of Forestry, Harvesting

Process,

http://web1.cnre.vt.edu/harvestingsystems/HarvestingProcess.h

tm (last visited October 18, 2012). For illustrations and de-

scriptions of the types of machinery described below, see LOREN

D. KELLOGG, PETE BETTINGER & DON STUDIER, TERMINOLOGY OF

GROUND-BASED MECHANIZED LOGGING IN THE PACIFIC NORTH-

WEST (1993), available at

http:/Ar.library.oregonstate.edu/xmluibitstream/handle/1957/7

615/RC1.pdf.

4 See infra Part III.

5

cab, uses a huge claw-like arm to grab the entire

tree, cut it off from the stump, and stack it with

others on the ground. On steep ground or areas with

large trees, this task may be accomplished by highly-

skilled timber fallers using chainsaws. Next, the

tree is “extracted” (moved) to a landing or roadside.

To carry out this step, the tree is either carried on a

machine called a forwarder or dragged by a different

machine called a skidder. On steep slopes, machines

similar to cranes, called yarders, are used to haul

logs on a system of cables. Typically, at the landing

or roadside, another machine called a delimber or

people with chainsaws are used to cut off the

branches and the stem is manufactured into the

preferred log lengths. A log loader is then used to lift

the logs onto a truck for transport to the mill.

Given the size and sophistication of the machi-

nery and the skill of the individuals involved at each

step of this process, it is not surprising that a classic

text on logging practices describes a modern indus-

trial forestry operation as “a loose grouping of mobile

factories. The manufacturing process taking place in

those factories is not unlike the process taking place

in other extractive industries such as coal mining

and oil production.”5

All of this massive machinery is delivered into

the heart of a forest by means of logging roads. The

logging trucks—with a total weight between approx-

imately 80,000 and 105,000 pounds depending on

state regulations—then use the roads to carry the

5’ STEVE CONWAY, LOGGING PRACTICES: PRINCIPLES OF TIMBER

HARVESTING SYSTEMS 48 (1976).

6

timber to the mill for processing. As described in the

following section, the cumulative impacts of this

intensive use of the forest in support of industrial

logging operations is significant water pollution.

II. Active Logging Roads Have Significant

Impacts on Water Quality.

Logging roads are significant sources of pollu-

tion—the most important source of forestry-related

sediment pollution. 2JA 119, 128. In particular, the

Environmental Protection Agency (EPA) has con-

cluded that “up to 90% of the total sediment produc-

tion from forestry operations” comes from logging

roads. Furthermore, a report commissioned by EPA

concluded that “forestry-related sediment is a lead-

ing source of water quality impairment to rivers and

streams nationwide.”? The magnitude of these im-

pacts is directly related to the decisions that are

made at each stage of the development and use of

logging roads.* As a result, these impacts are also

the direct result of logging roads’ function in the

overall industrial logging operation.

6 EPA, Guidance Specifying Management Measures for Sources

of Nonpoint Pollution in Coastal Waters 27 (1993).

* GREAT LAKES ENVTL. CTR., NATIONAL LEVEL ASSESSMENT OF

WATER QUALITY IMPAIRMENTS RELATED TO FOREST ROADS AND

THEIR PREVENTION BY BEST MANAGEMENT PRACTICES 2 (2008)

{hereinafter NATIONAL LEVEL ASSESSMENT] (citing EPA’s 2000

and 2002 National Water Quality Inventories).

6 Or. Dep't of Forestry, Technical Report # 17, Wet Season Road

Use Monitoring Project: Final Report 24 (2003), available at

http://www.oregon.gov/odf/privateforests/docs/roaduse.pdf.

7

Not all of the sediment generated by logging

roads enters streams. Indeed, in a properly planned,

designed, constructed, and maintained logging road,

virtually all of the sediment will be discharged over-

land. 2JA 78. As a practical matter, however, the

amount of sedimentation actually delivered to

streams from logging roads can be staggering.

An imaginative exercise based on the actual re-

sults of removing abandoned logging roads from the

landscape may help sharpen the problem. In 2008,

in recognition of the impact of logging roads on water

quality, Congress created the Legacy Roads and

Trails Remediation Program. Recently, the U.S.

Forest Service’s Rocky Mountain Research Station

completed research measuring the benefits of this

program.’ The research group chose a notably large

measurement unit: metric tons per year. The re-

search revealed that rehabilitating just 42 miles of

abandoned roads prevented delivery of 200 metric

tons per year of sediment into streams—the equiva-

lent of a 1-ton pickup truck dumping a full load of

dirt and aggregate into a stream, 200 times.

Active logging roads contribute to water pollution

through several mechanisms. Specifically, whether a

logging road causes stream sedimentation depends

on three factors: generation, mobility, and connectiv-

ity. That is, whether and how much sediment is

delivered from logging roads to streams depends on

9 U.S. FOREST SERVICE, ROCKY MOUNTAIN RESEARCH STATION,

LEGACY ROADS AND TRAILS MONITORING PROJECT UPDATE 2012,

available at

http://www.fs.fed.us/GRAIP/downloads/case_studies/2012Legac

yRoadsMonitoringProject Update. pdf.

8

(1) the amount of sediment that is generated and

thus made available for delivery; (2) the mobility of

the sediment (in other words, the degree to which it

is able to be transported by the available water}; and

(3) the hydrological connection between roads and

streams that allows the sediment generated and

mobilized to enter the streams.

A. Sediment Generation.

Sediment is generated at the road site in a num-

ber of ways. First, a massive amount of soil is made

available during the excavation of a road.!° This soil

is loosened by being torn from its anchor of vegeta-

tion. The cut-slopes are thus exposed to potentially

damaging rainfall, often without any of the vegeta-

tive protection remaining. This excavated material,

already loose, is made more vulnerable to erosion

when it is used to make fillslopes (on the slope below

a road), which can be placed at a steeper angle than

native hillslopes and thus are prone to mass wasting

events (landslides).!1! Such landslides frequently

produce sedimentation when they reach streams,

either through road-surface flows and obstruction of

drainage (likely outcomes for mass wasting on cut-

slopes, which are upslope from roads and thus fre-

quently land on the road surface), or through debris

flows (a likely outcome for fillslopes, which are

19 See KEITH MILLS, OR. DEP’T OF FORESTRY, FOREST ROADS,

DRAINAGE AND SEDIMENT DELIVERY IN THE KILCHIS RIVER WA-

TERSHED 2 (1997), available at

www.oregon.gov/ODF/privateforests/docs/kilchis.pdf [hereinaf-

ter MILLS REPORT].

1) Id.

9

downslope of roads and can be carried wholesale to a

stream during a storm).!2

Second, sediment can be generated from the rock

and dirt, or “aggregate,” used to surface the road.

Although aggregate is used to allow for year-round

use of a forest road, it also reduces the erosive poten-

tial when compared to dirt roads. Additionally, when

logging trucks pass over aggregate, they can crush it

into even more fine sediment.!3 Studies have found

that tire crushing alone may yield from 10 to up to

500 tons of sediment per kilometer of road.1!4

Third, the creation of ruts in the surface of the

roads from repeated use can disable the road’s ability

to shed water as it becomes concentrated in the

road’s surface. This water can accelerate the erosion

of the road surface that can lead to an increase in

sedimentation of forest roads. As a result, a rutted

road can produce two to four times as much sediment

as a freshly graded road.!5

12 Jd. at 5.

13 Randy B. Foltz & Mark Truebe, Locally Available Aggregate

and Sediment Production, 1819B TRANSP. RESEARCH RECORD

185, 190 (2003).

14 Robert E. Bilby, Kathleen Sullivan & Stanley H. Duncan, The

Generation and Fate of Road-Surface Sediment in Forested

Watersheds in Southwestern Washington, 35 FOREST SCI. 453,

459 (1989); Leslie M. Reid & Thomas Dunne, Sediment Produc-

tion from Forest Road Surfaces, 20 WATER RESOURCES REs.

1753, 1759 (1984).

15 Randy B. Foltz & William J. Elliot, Effect of Lowered Tire

Pressures on Road Erosion, 1589 TRANSP. RESEARCH RECORD

19, 19 (1997).

10

Finally, additional sediment is generated from

ditches, which are generally surfaced not with ag-

gregate but with more erodible native dirt. 2JA 77.

B. Sediment Transport.

The movement of water over the road transports

the sediment loosened by road building and road use.

For example, even a light rain falling on roads can

mobilize fine sediment, direct it into the road-side

ditch, and ultimately carry the sediment to streams.

Rain can also saturate cut and fill slopes, causing

landslides that results in materials entering

streams. Stormwater rushing across a roaded land-

scape increases the likelihood of these sources of

pollution entering streams.

Streamwater can also carry away sediment from

logging roads. When a logging road crosses a stream,

a pipe or “culvert” is commonly installed to carry the

water under the road.'6© However, during a storm,

flows may be heavier than the culvert can bear.!?

The flow may be diverted from the stream channel

and into the roadside ditch or may create new chan-

nels near stream banks, where the soil is more erod-

ible than in the established channel.'® Or the flow

may carry off soil that has been loosened as part of

roadbuilding, such as that on fillslopes that even-

tually reaches the stream.!9

16 MILLS REPORT, supra note 10, at 2.

Td.

18 Td.

19 Id.

11

C. Connectivity.

Sediment may be produced by roadbuilding and

use and may be transported across the landscape by

water, but it only damages stream life if it reaches

streams—a question of connectivity. Unfortunately,

there are numerous possible points of connectivity

between roads and streams. Not surprisingly, mass

movement events like landslides often have momen-

tum sufficient to deliver sediment directly to

streams. More significant to this litigation, many

logging roads discharge directly to rivers and

streams through purposefully-designed systems of

ditches. Studies in the Pacific Northwest have found

that between 25% and 75% of road drainage points

discharge directly to streams.”° 2JA 122, 128.

This statistic offers both good news and bad news

for stream health. The bad news is that direct drai-

nage of sediment-laden runoff from logging roads to

streams is real—and widespread. 2JA 129 (conclud-

ing that, despite “forest practices rules, which re-

quire filtering of muddy runoff through the forest

floor,” in practice “roads are designed and main-

tained for efficient delivery of water to channels”).

The good news is such direct drainage is not inevita-

ble. After all, when viewed in reverse, the statistic

reveals that 25% to 75% of road drainage points do

not discharge directly to streams. Sediment-laden

water from these drains is reincorporated into the

landscape before reaching streams—usually through

dispersion onto the forest floor where it infiltrates

into the soil matrix. Stream-friendly drains that

20 Jd.; NATIONAL LEVEL ASSESSMENT, supra note 7, at 43-44.

12

avoid direct discharges into streams are thus possi-

ble, even common.

There is further good news for streams: intensive

analyses of watersheds in Oregon and Idaho by the

U.S. Forest Service’s Rocky Mountain Research Sta-

tion has revealed that 90% of the sediment that

reaches streams is deposited by a mere 7% of drain

points.?! The implication is clear: stream health can

be dramatically improved by regulating the small

fraction of drains that act as problematic point

sources.

III. Creating a Logging Road is a Complex, Mul-

ti-step Undertaking, with Implications for

Sediment Pollution at Each Step.

The planning, design, construction, use, and

maintenance of logging roads is a complicated

process, requiring significant expertise and study.

The presence of a road represents a significant ma-

nipulation of the forested landscape; the road is en-

gineered to achieve a particular purpose. Key issues

in this process include the selection of a road surface

shape and the design and placement of drainage

structures to transport water away from the surface

of the road as efficiently as possible. By moving the

water in this way, the goal of forest road engineers is

to maintain the usefulness of the road as a conduit

for logging equipment and timber extraction, as well

21 Wildlands CPR, Road Reclamation: Measuring Success 6

2012) (summarizing Forest Service results), available at

http://www. wildlandscpr.org/files/GRAIP%20Report%20Wildlan

ds%20CPR_0.pdf.

13

as to reduce the impact of the road on the environ-

ment.

Successful road design requires an understand-

ing of forest hydrology and forest erosion processes in

addition to knowledge of civil, mechanical, and in-

dustrial engineering. Scientists and engineers in the

discipline of forest engineering produce many papers

each year, reporting research on topics such as the

ability of surface aggregates to sustain mechanical

and chemical breakdown, the causes of fillslope fail-

ures, the efficiency of travel routes taken by graders,

and the choice and distribution of logging machinery.

Even this listing—necessarily only a tiny sample of

forest engineering literature—helps illuminate the

complexity of the decision making involved in creat-

ing a logging road.

A. Road System Planning.

Before logging can begin, the necessary road

network must be in place. Depending on the loca-

tion, this process can involve the construction of new

roads, the rehabilitation of old roads, or some of

each. To arrive at the appropriate road network

design, the landscape must be evaluated, and loca-

tion alternatives must be generated and contrasted.

The feasibility of any particular logging project is

dependent on the possibility of designing the neces-

sary road network. Forest landowners will frequent-

ly plan access routes when designing timber sales,

even mandating in contracts that contractors or

subcontractors use specified roads and/or provide

notice of their proposed hauling routes. C.A. ER 47

at 98-99.

14

It is essential that location decisions be made

carefully: even the best design and construction deci-

sions are often not enough to overcome a poor loca-

tion. The challenge here is the same that runs

through most decisions about forest roads— it lies in

reconciling the need for safe, efficient roads on the

one hand with the unpredictability and variability of

a natural place on the other.

As a basic matter, sediment delivery to waters of

the United States can be reduced by simply locating

roads away from streams.?2 In fact, where the topo-

graphy allows it, the link between logging roads and

streams can be nearly severed by locating roads on

ridgetops, which allows road runoff to be dispersed

and reenter the soil.23 Thus, an effective method is

the severing of the connection between roads and

streams. There are numerous other ways, however,

to reduce sedimentation at the road system planning

stage. For example, avoiding unstable, landslide-

prone slopes not only improves safety but also reduc-

es sedimentation.?4

Although engineered fills (i.e. material with a

known set of properties that will perform in well-

22 See NATIONAL COUNCIL FOR AIR AND STREAM IMPROVEMENT,

FOREST ROADS AND AQUATIC ECOSYSTEMS: A REVIEW OF CAUSES,

EFFECTS, AND MANAGEMENT PRACTICES 9 (2003) [hereinafter

NCASI WHITE PAPER].

24 See Jacky Croke & Simon Mockler, Gully Initiation and

Road-to-Stream Linkage in a Forested Catchment, Southeastern

Australia, 26 EARTH SURFACE PROCESSES & LANDFORMS 208,

216 (2000).

24 See NCASI WHITE PAPER, supra note 22, at 9.

15

understood manner) are used in highway construc-

tions, logging roads are constructed from the mate-

rials found in place. Thus there is a high variability

in the engineering soil properties that are encoun-

tered during the construction of logging roads. Some

may be wet or weak soils while others will be much

stronger. The location of the road has the goal to

place the road to avoid the weaker soils or locate on

the stronger soils. However, limitations on grade

and road alignment may result in the road located on

less desirable construction materials.?®

B. Road Design.

Once the location for a logging road has been se-

lected, the road must be designed. The basic ele-

ments of a logging road, and the terminology used to

describe them, are illustrated in Figure 1.

25 See Kevin Boston, Marvin Pyles & Andrea Bord, Compaction

of Forest Roads in Northwestern Oregon—Room for Improve-

ment, 19 INT’LJ. OF FOREST ENGINEERING 24 (2008).

16

Natural

(Sround

> Top of Cut

Cutslope

res » Surfacing Subgrade/Roadbed

Fillslope

Road Prism

Figure 1: Logging Road Terminology [adapted from

Or. Dep’t of Forestry, State Forests Program, Forest

Roads Manual (2000)}.

The main adversary here is water. Water that is

not quickly dispersed can cause surface erosion and

rutting. These conditions hinder the safe passage of

logging trucks and other heavy equipment.

Landslides can also lead to road failures.26 Thus

road engineers must design roads that shed water

before it can erode the road and disable the drainage

system. 2JA 81, 97.

26 NCASI WHITE PAPER, supra note 22, at 5, 17. As Industry

Petitioners acknowledge, “{a]s a practical matter, forest roads

cannot be built or maintained without stormwater drainage

systems: without them, the roads would wash out.” Industry

Br. at 34.

17

They accomplish this task through several

means. The first is choice of road shape. A perfectly

flat road will retain too much water, so the road

engineer will generally design the road to be in-

sloped, out-sloped, or crowned (see Figure 2).

— tee

Out-Sloped Road a

Figure 2: Logging Road Surface Shapes [adapted

from Jeffry Moll, Ronald Copstead & David Kim

Johansen, U.S. Forest Service, Traveled Way Surface

Shape (1997)}.

These surface shapes are designed to “encourage

shedding of water from the surface before it [can]

18

cause unacceptable surface erosion” or ruts.27 On in-

sloped roads the surface is angled so that surface

runoff is diverted toward the cutbank (the uphill

side), where it can be collected in a purposely-

designed drainage ditch. When rain falls, water

flows off the road and into the ditch, then downslope

in the ditch until it is emptied either onto the forest

floor via a ditch-relief culvert or cross-drain, or emp-

tied directly into a stream at a road-stream crossing.

Currently, the best practice is to disconnect these

ditches and culverts from the streams, but that is not

physically possible in all circumstances. While an in-

sloped surface is commonly used, it likely provides

the least protection of the three road surface shapes

against chronic sediment pollution discharges to

waterways, as it collects the largest portion of the

water from the road surface.

Out-sloped roads divert surface runoff down the

hill toward the exposed edge of the roadway, thereby

avoiding the concentration and channelization of

runoff. Because of safety concerns, out-sloped roads

are generally appropriate only for roads with gentler

road grades. 2JA 81, 97.

A crowned road combines some elements of each

of the other two designs, including the use of a cut-

bank drainage ditch. Crowned roads are the most

common road shape in the areas at issue in this case.

2JA 81. Since the distance the water travels is re-

duced in a crowned road, this road type has the abili-

2" Jeffry Moll, Ronald Copstead & David Kim Johansen, U'S.

Forest Service, Traveled Way Surface Shape 1 (1997).

19

ty to shed water the fastest and is often used in

areas with significant rainfall.

After settling on the road surface shape, a road

engineer must next decide on the drainage system to

use. As indicated above, these two decisions are

interrelated—the choice of an in-sloped or crowned

road generally necessitates the use of a drainage

ditch located on the cutbank side of the road.

Ditches are built during road construction. They are

typically 3-5 feet wide and 1-3 feet deep. Other drai-

nage devices used in connection with logging roads

include culverts, waterbars (diagonal berms made by

cutting and filling across the road and usually asso-

ciated with roads that are temporarily or seasonally

closed), and dips (gentle grade reversals of the road

surface). 2JA 83.

The entire purpose of these structures is to con-

vey water from the road surface. The design objec-

tive of these drainage systems is to collect surface

runoff (and intercepted subsurface flow from the

cutbank) and then divert it via cross-drain structures

to the forest floor. The majority of logging road drai-

nage systems do just that. 2JA 122, 128. Some-

times, however, the water is discharged into a river

or stream—those discharges, and only those dis-

charges, are at issue in this case.

As with road system planning, many aspects of

the road design decision affect sedimentation. First,

poor drainage choices can be a major cause of slope

failures and subsequent delivery of sediment to

streams through mass wasting events (landslides).?®

28 See Bradley T. Piehl, Robert L. Beschta & Marvin R. Pyles,

20

As a basic precaution, the risk of landslides, and the

sedimentation that follows, can be reduced by taking

care not to direct drainage onto landslide-prone

slopes. 2JA 75. Even if the road has been con-

structed across an extended area of high landslide

risk, cross-drains can still be used to redirect water

to gentler slopes where water is more likely to be

dispersed safely. 2JA 77.

Second, the ditches and other aspects of the

drainage networks themselves also produce sedi-

ment. 2JA 77. Ditches generally have finer, more

erodible soil material than stream beds.*9 2JA 77.

However, the sedimentation impact of drainage

networks can be reduced, for example by carefully

designing the spacing between the ditch-relief cul-

verts to manage water velocity.2°° 2JA 123. Logging

companies can reduce sedimentation by creating a

network that slows water by, for example, frequently

discharging the ditches into the forest to reduce the

volume and velocity of the water in ditch. Slower-

moving water detaches less soil matrix and results in

less sediment production.?!

Ditch-Relief Culverts and Low-Volume Forest Roads in the

Oregon Coast Range, 62 NORTHWEST SCI. 91, 91 (1988).

2° Charles H. Luce & Thomas A. Black, Spatial and Temporal

Patterns in Erosion from Forest Roads, in LAND USE AND Wa~-

TERSHEDS: HUMAN INFLUENCE ON HYDROLOGY AND GEOMOR-

PHOLOGY IN URBAN AND FOREST AREAS 165, (Mark S. Wigmosta

& Stephen J. Burges eds., 2001).

40 See NCASI WHITE PAPER, supra note 22, at 9.

31 See id.

21

C. Road Construction.

At the construction stage, logging companies im-

plement the design parameters chosen above and

create a road that will support the logging project.

In general, construction involves clearing the

roadbed of brush, timber, and stumps; filling holes or

dips; compacting the subgrades; and choosing and

placing aggregate.*?

As in the other stages of creating a logging road,

the choice of construction practices strongly affects

the generation of sediment pollution. As a founda-

tional matter, construction itself can be a significant

source of sedimentation. In fact, studies have shown

that the highest rates of sediment production occur

in the years immediately following road construc-

tion.33 Sources of sedimentation from construction

include sidecast materials—uncompacted material

from excavation that is disposed on the fill side of the

road. This material has low strength and is subject

to gully formation or landslides when it becomes

saturated.*4

Choices made at the construction stage affect not

only the amount of sedimentation that is immediate-

ly created, but also the amount of sediment that will

32 See id. at 7.

33 Beverley C. Wemple, Frederick J. Swanson & Julia A. Jones,

Forest Roads and Geomorphic Process Interactions, Cascade

Region, Oregon, 26 EARTH SURFACE PROCESSES AND LANDFORMS

191, 191 (2001).

34 Mary Ann Madej, Erosion and Sediment Delivery Following

Removal of Forest Roads, 26 EARTH SURFACE PROCESSES AND

LANDFORMS 175, 175 (2001).

22

be generated over the life of a road. For example,

one method to improve forest road performance is to

compact the subgrade. The subgrade is the native,

base soul of a road, to which aggregate and other

types of surfacing are added.*> Compaction of the

subgrade increases the strength of the road.5°¢ A

well-compacted road can better resist catastrophic

road failures, but its primary purpose is that is can

resist rut formation. If the subsurface is not ade-

quately compacted—particularly if it is composed of

fine, loose soil—a portion of the aggregate can sink

into the subsurface, leading to the creation of ruts.37

Additionally, the quality of the aggregate

(crushed rock) used to surface the road significantly

influences the quantity of fine sediment that is sub-

sequently generated by the road. Native-surfaced

(dirt) roads are generally the most erosive, meaning

that they produce the most sediment pollution. Log-

ging roads are usually surfaced with aggregate to

allow year-round hauling. This process involves

trucking in many tons of rock from a quarry. These

materials are placed on the road surface and then

compacted and smoothed by road building machi-

nery.

Lower quality aggregates, which are weaker,

create substantially more sedimentation because

35 Or. Dep’t of Forestry, Forest Practices Technical Note No. 9,

Wet Weather Road Use 8 (2003), available at

http://www.oregon.gov/odf/privateforests/docs/wetseasonroadus

efptechnote9.pdf [hereinafter Wet Weather Road Use}.

36 Jd.; Boston, Pyles & Bord, supra note 25, at 24.

37 See Boston, Pyles & Bord, supra note 25, at 24.

23

they are more vulnerable to mechanical breakdown

from truck tires or to chemical breakdown from

weather. Studies indicate that roads surfaced with

high quality aggregate can produce one-quarter to

one-twentieth the fine sediment produced by roads

surfaced with poor quality aggregate.*®

In addition, it is important to use an aggregate

with the right mix, or distribution, of coarse and fine

material: although some amount of fine sediment in

the aggregate mix is desirable to facilitate compac-

tion and road stability, too much is undesirable as it

is easily washed off the road surface during rain

events.39

D. Road Operation.

The use of a logging road for logging is, of course,

its purpose. To sustain the road for the duration of a

project, a logging company must balance considera-

tions of traffic volume and truck weight—which can

damage the road—with the need to timely transport

heavy equipment into—and logs out of—the forest.

36 See Wet Weather Road Use, supra note 35, at 2; Randy B.

Foltz, Traffic and No-Traffic on an Aggregate Surfaced Road:

Sediment Production Differences (1996) (paper presented at the

FAO Seminar on Environmentally Sound Forest Roads), avail-

able at

http://forest.moscowfsl.wsu.edu/engr/library/Foltz/Foltz1996f/19

96f.pdf.

39 Foltz & Truebe, supra note 13, at 188; Randy B. Foltz, Gary

L. Evans & Mark Truebe, Relationship of Forest Road Aggre-

gate Test Properties to Sediment Production, in WATERSHED

MANAGEMENT & OPERATIONS MANAGEMENT 2000 (Marshal Flug

et al. eds. 2000).

24

Perhaps most significantly, a vast amount of se-

dimentation is caused by the movement of logging

trucks and other heavy equipment over the road.

The heavy truck traffic associated with industrial

forestry increases sediment pollution in two specific

ways. First, the weight of the tires passing over the

road crushes the aggregate into fine sediment.

Second. “vibrations from heavy traffic can move fine

material up through the aggregate to the top of the

road.”*° The fine sediment generated by these two

mechanisms remains at the surface of the road until

it is washed away by the next significant rainfall.

In one study, truck traffic levels alone explained

97% of the variation in annual sediment yield.4!

Another study demonstrated that twenty passes of a

truck can produce as much sediment as is generated

in a year by the road surface alone.42 The Oregon

Department of Forestry has itself concluded that

“[wjet season road use can be the most significant

forest practice-associated source of chronic turbidity

and fine sediment in streams.”43

This level of sediment pollution is not inevitable,

however. One method of decreasing sediment pro-

duction is to decrease the tire pressure of the logging

40 Wet Weather Road Use, supra note 35, at 4.

*? Gary J. Sheridan & Philip J. Noske, A Quantitative Study of

Sediment Delivery and Stream Pollution from Different Forest

Road Types, 21 HYDROLOGICAL PROCESSES 387, 394 (2006).

'‘- See Pieter J.B. Fransen, Chris J. Phillips & Barry D. Fahey,

Forest Road Erosion in New Zealand: Overview, 26 EARTH

SURFACE PROCESSES AND LANDFORMS 165, 169 (2001).

*s Wet Weather Road Use, supra note 335, at 2.

25

trucks and other heavy vehicles as they pass over the

logging road. Some logging trucks are manufactured

with “central tire inflation,” an in-cab system for

adjusting tire pressures. The use of such a system to

reduce tire pressures on heavy-haul vehicles when

traveling over vulnerable aggregate has been shown

to reduce road surface sediment production an aver-

age of 80% over a three-year study.44 Even manually

reducing the tire pressure in unmodified trucks re-

sulted in a 45% reduction in sediment production.*®

Lowering logging truck tire pressures decreases

sedimentation in two ways. Not only does it reduce

the mechanical breakdown of the aggregate, but it

also minimizes the creation of ruts.46 A rutted road

produces twice to four times as much sediment as a

freshly graded road.‘47

E. Road Maintenance.

After construction, the road must be periodically

maintained to allow for continued use, with special

care given to the smoothness of its surface, mainten-

ance of the road shape, and the functionality of its

drainage structures. Maintenance activities include

surface grading and gravel supplementation; ditch

cleaning; cleaning and replacement of damaged relief

44 Randy B. Foltz, Sediment Reduction from the Use of Lowered

Tire Pressures, in CENTRAL TIRE INFLATION SYSTEMS: MANAG-

ING THE VEHICLE TO SURFACE 47-52 (Soc’y of Automotive Eng’rs

1995).

45 Id.

46 Id.

4° Foltz & Elliot, supra note 15, at 19.

26

culverts and stream crossing culverts; and grading of

rolling dips.

Maintenance choices, like choices at other steps

in the process of building a logging road, can signifi-

cantly reduce the rate of sedimentation. For exam-

ple, overly frequent grading of the road surface can

increase sediment generation.49 Meanwhile, failure

to clean plugged ditches contributes significantly to

road failures.5° In one study, blocked culverts ac-

counted for 45 per cent of landslides caused by log-

ging roads.5! Blocked ditches and culverts can also

cause water to travel on the landscape in other un-

natural and uncontrolled ways—for example, by

creating gullies along virgin slopes.5? Erosion treat-

ments such as spreading grass seed on the slope

below roads can also be an effective way to reduce

sedimentation.®

46 See NCASI WHITE PAPER, supra note 22, at 10.

4° See Matthew Thompson et al., /ntelligent Deployment of

Forest Road Graders, 18 INT’LJ. OF FOREST ENGINEERING 15, 15

(2007).

50 See Piehl, Beschta & Pyles, supra note 28.

51 See Fransen, Phillips & Fahey, supra note 42, at 170.

52 Charles H. Luce & Beverley C. Wemple, Jntroduction to

Special Issue on Hydrologic and Geomorphic Effects of Forest

Roads, 26 EARTH SURFACE PROCESSES AND LANDFORMS 111,

112-13 (2001).

53 See Walter F. Megahan, Monte Wilson & Stephen B. Monsen,

Sediment Production from Granitic Cutslopes on Forest Roads

in Idaho, USA, 26 EARTH SURFACE PROCESSES & LANDFORMS

153, 161 (2001).

27

In sum, the runoff that flows from logging roads

into streams is not natural runoff—it is runoff whose

timing, pathways, and sediment content has been

modified by the presence, composition, and use of

logging roads. Indeed, much of the sediment from

logging roads is not soil from natural hill slopes—

instead, it is material that has been exposed by the

road construction process or aggregate that has been

placed on the road surface to aid the logging opera-

tions and has then been ground up into fine particles

by the repeated passage back and forth of heavy

trucks and other equipment used in logging opera-

tions. Under these circumstances, when sediment-

laden water from logging roads is carried by ditches

and other man-made drainage structures and dis-

charged into rivers or streams, it is best understood

to be a discharge from a point source, generated by

industrial activity, and not mere natural runoff.

CONCLUSION -

The discipline of forest engineering teaches that

discharges of water pollution to streams from logging

roads is not inevitable. Indeed, scientists and engi-

neers have demonstrated that discrete choices in the

planning, design, construction, and use of logging

roads can provide substantial protection to streams.

These choices, detailed above, are both limited in

number and achievable. They include locating roads

on strong soils and away from streams; designing

road shape and drainage networks to promote shed-

ding of water from the road; compacting the subsur-

face to reduce rutting; selecting quality aggregates

that resist weathering and crushing; using vehicles

and tire adjustment systems that reduce pressure on

28

the road; and maintaining the road by, for example,

clearing drainage structures to prevent clogging and

seeding slopes to reduce erosion.

The function of the NPDES permitting program

in this context is to stimulate sources to adopt these

important practices.*4 NPDES permitting need not

be burdensome. As explained by Respondent, the

EPA Administrator has a variety of mechanisms,

including general permitting and variances in re-

sponse to economic hardship, to minimize the admin-

istrative burden of the permitting requirement.

Respondent Br. at 54-56. For example, individual

sources may opt in to a general permit that sets out

conditions applicable to a particular geographic area

merely by providing notice to the agency. Such gen-

54 Amici supporting Petitioners suggest that the impacts of

stormwater runoff from logging roads are best dealt with by

Best Management Practices (BMPs). See generally Amicus Br.

of the Soc’y of Am. Foresters. While there are some benefits to

BMPs, this argument misses the point that the only legal

difference between a point source and a non-point source is the

manner in which pollution enters waters of the United States.

There can be no question that the discharges from ditches and

other human-created drainage structures at issue in this case

are point-source discharges. Indeed, petitioners and their amici

do not dispute that logging roads use pipes, ditches, and chan-

nels to collect and discharge stormwater. See, e.g., Industry Br.

at 1, 34; State Br. at 2, 44; Amicus Br. of the Soc’y of Am. Fore-

sters at 20: Amicus Br. of Ass'n of Or. Counties at 19-32 (dis-

cussing the numbers of culverts and ditches associated with

logging roads in Oregon). Additionally, in many states BMPs

are voluntary and do not provide the certainty of protection

that would be accorded by conditions included in an enforceable

permit. See generally Amicus Br. of Northwest Envtl. Advo-

cates.

29

eral permitting minimizes both the burden on the

agency and on sources.

In short, permitting of direct discharges from

logging roads is compelled by the CWA, is necessary

to achieve the goals of the Act, and—thanks to gen-

eral permitting—places no unacceptable burdens on

EPA or sources. The judgment below should be af-

firmed.

Respectfully submitted,

SHAUN A. GOHO

Counsel of Record

Emmett Environmental

Law & Policy Clinic

Harvard Law School

6 Everett St., Suite 4119

Cambridge, MA 02138

(617) 496-2058

sgoho@law.harvard.edu

Counsel for Amicus Cu-

riae Dr. Kevin Boston

OCTOBER 23, 2012