Historic, Archive Document

Do not assume content reflects current
scientific knowledge, policies, or practices.

t

*

' m

SEND TO STACKS

f

FORESTRY

RESEARCH

what’s new in the west

U.S. Department of Agriculture Forest Service

OCTOBER 1977

What’s Inside

Fuelbreak

Guidelines _ 1

Classifying

Montana’s Forests _ 5

Snow Management _ 9

,cn

cd

— - \

no

Tin

*rca

J

<c:

— Douglas-fir Tussock
Moth Research _ 14

—New Publications _ 18

.r. _ A.\ .. . .

a note to you

Forestry Research: What's New in the West, is a report on the
work of the USDA Forest Service’s four Forest and Range Experiment
Stations in the West. These research centers, and the States included
in their areas of study are: Rocky Mountain (North Dakota, South
Dakota, Nebraska, Kansas, Colorado, Arizona, New Mexico, and part
of Wyoming, Oklahoma, and Texas); Intermountain (Montana, Idaho,
Utah, Nevada, and part of Wyoming); Pacific Northwest (Alaska,
Oregon, and Washington); and Pacific Southwest (California, Hawaii,
and the Pacific Basin).

on the cover

Snow Scientist Ron Tabler checks a Weather Monitoring Sta¬
tion, developed by the USDA Forest Service, near Wyoming’s Inter¬
state 80. Weather data is relayed to the Wyoming Highway Depart¬
ment for use in making decisions for regulating traffic. See “Snow
Management’’ on page 9.

our addresses

Single copies of most of the publications mentioned in this issue
are available free of charge. When writing to research Stations, please
include your complete mailing address (with ZIP) and request
publications by author, title, and number (if one is given).

For INT publications write:

Intermountain Forest and
Range Experiment Station
507 25th Street
Ogden, Utah 84401

For PNW publications write:

Pacific Northwest Forest and
Range Experiment Station
Post Office Box 3141
Portland, Oregon 97208

For PSW publications write:

Pacific Southwest Forest and
Range Experiment Station
Post Office Box 245
Berkeley, California 94701

For RM publications write:

Rocky Mountain Forest and
Range Experiment Station
240 West Prospect Street
Fort Collins, Colorado 80521

If you are planning to move, please notify us as much in advance
as possible. Send your old address, your new address, and the address
label from the back cover to Forestry Research: What's New in the
West, 240 West Prospect Street, Fort Collins, Colorado 80521.

When reprinting articles, please credit USDA Forest Service.
Mention of commercial products in this issue is for information only —
no endorsement by the U.S. Department of Agriculture is implied.

Researchers

develop

fuelbreak

guidelines

More than 2,000 miles of fuelbreaks wind
along ridgetops, roads, and canyon
bottoms in California wildlands. The
fuelbreaks are usually at least 200 feet across
— wide enough to provide a break, or opening,
in heavy, dense forest or brushland vegetation.
The grasses or shrubs growing in the fuel-
breaks are “light fuels’’ — plants that are less
of a hazard to firefighters than the thick, flam¬
mable vegetation growing beyond the fuel-
break borders.

Strategically located, fuelbreaks serve as
ready-made fire control lines. They are access
routes for firefighters and equipment, and rela¬
tively safe places to control a fire — either by
fighting it as it burns towards the fuelbreak, or
by starting a backfire out to meet the wildfire.

Fuelbreaks have other uses. They can
serve as routes for vacationers, and, depending
upon the plants growing in the fuelbreak, may
provide food for wildlife or livestock.

A fuelbreak' s usefulness during a wildfire may depend upon such factors as the fuelbreak location, and
wind and fuel condition.

\

1

A large brushland disk, pulled by a tractor, will up¬
root, cut up, and mulch brush.

Over a 17-year period, researchers at the
Pacific Southwest Station’s Forest Fire Lab¬
oratory in Riverside, California, have con¬
ducted studies on almost every aspect of fuel-
break development and management. Repre¬
sentatives of the National Forests of Cali¬
fornia, California Department of Forestry, Los
Angeles County Fire Department, Los Angeles
State and County Arboretum, and other Fed¬
eral, State, and County organizations have
been cooperators in this research. Studies have
covered the uses of fuelbreaks, their cost, de¬
sign, construction, revegetation, and mainte¬
nance. Most of the research took place in
southern California brushfields, but some dealt
with fuelbreaks in conifer forests — mainly
ponderosa pine.

The results of these studies are not limited
just to California. Instead, they have useful im¬
plications for land managers in areas where the
climate, terrain, and vegetation are similar to
California’s. Here are some highlights of this
research:

The best way to plan fuelbreaks is on a
large-scale basis. Complete fuelbreak systems
— interconnected networks of individual fuel-
breaks — should be mapped out, even if the
money needed to construct the breaks isn’t
available at the time the plans are made. If
they are well designed, fuelbreaks will not only
meet the needs of firefighters, but will, within a
few years after they are constructed, look like
natural changes in the vegetation.

Clearing techniques

There are a variety of techniques that can
be used to construct fuelbreaks. Brush grow¬
ing in the path of a planned fuelbreak can be
cleared by hand, by bulldozers or other ma¬
chinery, by prescribed fire, or a combination of
these. All of these methods have been tested at
sites throughout California. Each has advan¬
tages and limitations. Hand clearing is the
most expensive, and is slow, but may be the
only available method where fuelbreak sites are
too steep or too rocky for bulldozers or other
machinery to navigate.

Machine clearing is faster and cheaper, but
must be done carefully, so that erosion and
sedimentation, an inevitable result of machine
operations, will be minimal. In California wild¬
lands, the most efficient machines to use are
either a tractor with a bulldozer blade, or a
tractor with heavy, offset disks. Of the two, the
bulldozer is more versatile — it can be used on
slopes up to 55 percent, while the disk is ineffi¬
cient on slopes of more than 30 percent.

The cheapest way to dispose of brush de¬
bris is to burn it. In a bulldozing operation,
brush can be crushed for later broadcast burn¬
ing, or pushed into piles for later burning.

Two alternatives to burning — chipping
debris or burying it — have been tried, but
both are more expensive than burning.

After disking, burning is sometimes unnec¬
essary. In most chamise stands, for example,
two passes of the disk will be enough to chop
and partially bury most of the vegetation.

Clearing brush by burning usually costs
less than bulldozing and generally creates less
soil disturbance. The disadvantages of pre¬
scribed burning include the possibility of es¬
cape and the associated liability, the smoke it
produces, (although researchers claim the
smoke is more of a nuisance than a pollutant),
and the possibility of some undesirable side
effects on the soil, such as burning of organic
matter, destruction of soil organisms, and crea¬
tion of a non-wettable layer in the surface soil.

Repeated, low-intensity prescribed fires
are recommended for clearing fuelbreaks in
pine forests. They can efficiently remove
needles, cones, dead branches, and other fuel
from the forest floor, without harming the
overstory trees.

2

Chemical treatment

In brushfields, the studies have shown
that the vegetation needs to be treated before
burning, for best results. The brush should
either be crushed, or be treated with chemicals
that will dry it out (desiccants). If the brush
fuels aren’t treated, the prescribed fire is more
likely to be uneven, and some patches of brush
may not be burned.

Brittle-stemmed, mature brush can be
crushed effectively by a bulldozer, an anchor
chain pulled by two tractors working in con¬
cert, by a ball and chain pulled by one tractor,
or by tractor-mounted rollers or drags. (Agri¬
cultural Handbook No. 487, “Mechanical
Methods of Chaparral Modification,” has de¬
tails; copies are available from the Pacific
Southwest Station.)

Young, flexible brush should be treated
with desiccants. Two classes of desiccants —
contact herbicides and systemic herbicides —
have been studied. The contact herbicides
quickly dry out leaves and small twigs; the sys-
temics are slower acting and attack the plant
internally. The chemicals PCP (pentachloro-
phenol) or DNBP (dinitro secondary butyl
phenol) in diesel oil are the most cost-effective
of the government-approved contact herbi¬
cides. Systemics such as 2,4-D or 2,4,5-T, or
2,4-D plus the herbicide picloram, cost less
than the contact herbicides to apply, and usu¬
ally result in a much more complete plant kill.

Applying herbicides — whether from the
air, or from the ground by tractor-mounted or
hand-carried boom sprayers — is the most ef¬
fective way to prevent regrowth of unwanted
plants. Regrowth is by far the biggest obstacle
to successful establishment of fuelbreaks in
California brushfields. This is because some
brush species can resprout within 10 days after
burning, and can be 8 to 10 inches tall a month
later. For this reason, researchers recommend
that herbicides to control brush regrowth be
applied no later than 2 years after the initial
clearing, and be followed by one or two more
applications in the next few years. The most
useful chemical for controlling regrowth is 2,4-
D.

The fuelbreak cover, planted or seeded in
place of the original vegetation, must meet sev¬

eral requirements. The fuelbreak plants should
be slower to burn than the plants they have re¬
placed; they should also help stabilize the soil,
establish readily, reproduce well, and require
little maintenance., Further, the plants should
be low-growing and should be low in fuel vol¬
ume — the weight of the new plants, plus the
litter they produce, should be no more than 2
tons per acre, dry weight.

At the Forest Fire Laboratory, more than
50 species of native and foreign grasses and
shrubs that showed promise of having most, or
all, of these characteristics, were evaluated.
Grasses were included in the research, for sev¬
eral reasons. Although firespread in grass is
rapid, its behavior is predictable. Fire in
chaparral is less predictable, and control is
much more difficult. Furthermore, firefighters
can use fast-moving grass fires to their advan¬
tage, by quickly “firing out” control lines in
grass-covered fuelbreaks. Finally, grass adds
only 2 to 3 tons of fuel per acre, while brush can
add from 5 to 40 tons, per acre.

Certain conditions are required for aerial application
of herbicides — winds should be under 5 miles per
hour, the air temperature should be below 70 degrees
F., and the humidity should be greater than 40
percent.

3

Creeping sage (foreground) was among the most successful shrubs in field tests of low-growing, slow-burn¬
ing plants.

Fuelbreak cover

Perennial grasses are recommended over
annuals, because perennials have a deeper root
system, stay green further into the fire season,
and are generally more stable. Of the peren¬
nials studied, intermediate and pubescent
wheatgrasses are the most highly recom¬
mended.

Of the low-growing, woody shrubs avail¬
able, bearmat is the best naturally occurring
ground cover for fuelbreaks in conifer forests
of the west-side Sierra Nevada. For brushland
fuelbreaks, creeping sage is an excellent cover.
It grows on a wider variety of chaparral sites
than any of the other low-growing brushland
shrubs tested. Like bearmat, creeping sage
produces low, mat-like foliage that successfully
crowds out herbaceous plants. Creeping sage
can be established by seed (if the seeds are
treated in gibberellic acid to break dormancy),
or by cuttings or transplants.

Descanso rockrose is also highly recom¬
mended, because it is relatively drought-tol¬
erant and is well-adapted to a wide range of
sites. Although it can grow as high as 2 feet, it
is a relatively low shrub, compared to other
chaparral species. Mature rockrose is sus¬
ceptible to fire when its moisture content drops
in the late summer and fall; but, because of its

relatively low height and low fuel volume, it
burns with less intensity and less heat than
most chaparral species.

Two tall shrubs — fourwing saltbush and
allscale saltbush — are also recommended as
fuelbreak cover. Fourwing saltbush is drought-
tolerant, relatively cold-tolerant, and provides
excellent browse. It can be established either
from direct seeding or potted transplants, but
should be enclosed during its first year, to keep
out foraging rabbits. Allscale saltbush is also
drought-tolerant and relatively cold-tolerant.
It is a good browse for livestock and deer, and
is easy to establish.

Fuelbreaks are not the ultimate answer to
wildland fire management problems, but their
value to firefighters has been proven through¬
out the State. Details about the fuelbreak
studies described here are presented in two
new publications — “Fuelbreaks and Other
Fuel Modification for Wildland Fire Control,”
USD A Agricultural Handbook No. 499-FR12,
by Lisle R. Green, and “Low-volume and Slow-
burning Vegetation for Planting on Clearings
in California Chaparral,” Research Paper PSW-
124-FR12, by Eamor C. Nord and Lisle R.
Green. Copies of each are available from the
Pacific Southwest Station.

By Marcia Wood, Pacific Southwest Station

4

Classifying

Forest site classifications based on habi¬
tat types are now used by land man¬
agers and researchers in more than 20
western areas.

“Forest Habitat Types of Montana,” a
general technical report published by the Inter¬
mountain Station, describes the first such
classification based on a large-scale reconnais¬
sance study. More than 1,500 stands in 10 Na¬
tional Forests were sampled to develop this
habitat type classification for Montana’s ex¬
tensive forests.

The study was led by Robert Pfister,
Project Leader of the Station’s Forest Ecosys¬
tems research work unit, headquartered at
Missoula. The Northern Region of the Forest
Service provided major financial assistance,
and the University of Montana’s Forestry
School and Botany Department provided ad¬
ministrative and technical support.

The Montana classification involved a pro¬
gressive series of analyses by Pfister and his
associates. Development included 4 years of
field testing by land managers and researchers.

Montanas

Habitat typing as an approach to forest
ecosystem classification was originally de¬
veloped for northern Idaho and eastern Wash¬
ington by Professor Rexford Daubenmire of
Washington State University. Daubenmire’s
system proved valuable, and it served as a
model for classification work throughout west¬
ern forests.

Habitat types are based on the potential
climax tree and undergrowth vegetation for
given sites. The theory is that climax vegeta¬
tion, developing over long periods of time with¬
out disturbance, most clearly reflects environ¬
ment and site characteristics.

Pfister explains that successional trends
toward climax vegetation can usually be identi¬
fied even in rather young stands. Thus a given
habitat type includes all land areas potentially
capable of producing similar plant communi¬
ties when vegetation reaches climax, even
though the present vegetation may be domi¬
nated by early successional species.

To represent Montana’s diverse forest en¬
vironments, Pfister and his team have defined
64 habitat types. However, only a few are
found in specific drainages or mountain ranges,
and they occur in a generally consistent and
predictable local pattern related to topog¬
raphy.

A “key to habitat types,” similar to the
familiar keys for identifying plants, permits
Montana foresters to identify types after mini¬
mal training. Numerous shortcourses have
been conducted to help land managers effec¬
tively apply habitat typing concepts.

5

Status of forest habitat type classifications in west¬
ern North America. Numerals refer to year in which
classifications were completed. Letters refer to
curren t studies.

6

The Classification Framework

Habitat types provide a simple, orderly
classification system. The first part of each
habitat type name is based on the climax tree
species, which is usually the most shade-tol¬
erant tree adapted to the site. This level of
classification is called the series, encompassing
all habitat types having the same dominant
tree at climax. The second part of the habitat
type name is based on the dominant or charac¬
teristic undergrowth species in the climax com¬
munity type.

A third level, the phase, is used when nec¬
essary to recognize subdivisions of a habitat
type — for example, the Agropyron spicatum
(bluebunch wheatgrass) phase represents the
warmest, driest part of the PseudotsugaJCala-
magrostis (Douglas-fir/pinegrass) habitat type.

Applying Habitat Types

Some current applications of habitat
typing are:

HABITAT TYPE MAPPING - Habitat
type maps can serve as a fundamental contri¬
bution to land-use planning. Maps at scales of
2 inches per mile have been prepared for sev¬
eral planning units, and also for some entire
National Forests. Maps in greater detail have
been prepared for project planning and re¬
search studies.

TIMBER MANAGEMENT - The
Montana report provides timber productivity
data for most habitat types. It also shows dis¬
tributions and roles of tree species by habitat
type. This helps silviculturists determine
which species to plant or to favor in thinning or
partial cutting. Habitat types are used in
guidelines for seed collecting, selecting sites
for planting nursery stock, categorizing sites in
genetic tree-improvement programs, compar¬
ing regeneration success, and evaluating
cutting methods.

RANGE AND WILDLIFE MANAGE¬
MENT — The report provides information on
the frequency and abundance of all species —
including important forage plants — in mature
stands representing each habitat type. It also

summarizes observations of use by domestic
stock and wildlife in each type.

WATERSHED MANAGEMENT -
Certain habitat types and phases are identified
with seepage, subsurface water, or wet sites
that require special considerations, particu¬
larly in specifying timber harvesting methods
and roadbuilding. The Montana report gives
climatic data for many types.

RECREATION MANAGEMENT -
Habitat types have been used to assess suit¬
ability of sites for recreational use, evaluate im¬
pacts of use on plant communities and sites,
and predict recovery rates following disturb¬
ance.

FOREST PROTECTION - Habitat types
can be used in categorizing fuel buildup, devel¬
oping fuel management plans, and evaluating
the natural role of fire. They can also be used in
assessing susceptibility of vegetation to in¬
sects and diseases.

Estimating coverage of undergrowth species in a
plot.

7

Pseudotsuga menziesii/Calamagrostis rubescens (Douglas- fir/pine grass) is a habitat type that is widespread
in Montana and adjacent western States and Canadian provinces.

NATURAL AREA PRESERVATION -
Habitat type information can help to insure
that the environmental spectrum is adequately
represented in a natural area system.

Pfister cautions that use of habitat types is
not a panacea for land management decision¬
making or research interpretation. In many
cases, however, they can provide a common
frame of reference complemented by informa¬
tion on current vegetation, soils, and other site
characteristics. The Montana classification is a
major step toward the goal of developing and
applying classification systems for manage¬
ment of forest ecosystems.

Pfister points out that further work must
be done to fully realize the potential benefits of
habitat types in land management. Continued
support is needed to complete regional classifi¬
cations. Cooperative research-management
programs could use habitat type classifications
to summarize and communicate existing

knowledge from many disciplines. Classifica¬
tions, or models, of successional vegetation
within the habitat type framework are needed
to serve many management purposes.

Authors of “Forest Habitat Types of Mon¬
tana” (INT-GTR 34-FR12) are Pfister; Bernard
L. Kovalchik, timber management planner on
the Helena National Forest, Montana; Stephen
F. Arno, associate plant ecologist of the Forest
Ecosystems unit; and Richard C. Presby, for¬
est biotic planning specialist on the Idaho Pan¬
handle National Forests.

Copies of the report are available from the
Intermountain Station.

Readers needing additional information
should contact Robert Pfister in Missoula,
Montana at (406) 329-3533, FTS-585-3533.

— By Delpha Noble, Intermountain

Station

8

Snow

management

Snowfall in the mountain West is often a
real blessing. Mountain and plains com¬
munities and farms depend on the snow¬
melt for survival. Sometimes, however, snow
can become a curse. Explosive development
and increasing recreational activities have in¬
tensified two natural hazards of increasing im¬
portance — blowing snow and snow ava¬
lanches. Blowing snow creates visibility
problems along transportation routes and
often causes road closures — avalanches trap
travelers and demolish valuable property.

Avalanches are of increasing concern to
land managers and planners throughout the
West. Avalanche accidents have been on the
rise in recent years, due to the escalating devel¬
opment of ski resorts, private residences, roads
and highways, and a general increase in out¬
door recreational activities. Recorded ava¬
lanche accidents in recent years have taken an
annual average of 12 lives and caused over Va
million dollars in property damage and loss.

Various government and private groups
throughout the country are involved in ava¬
lanche research. One is the Mountain Snow
Management and Avalanche Hazard Evalua¬
tion Project located at the Rocky Mountain
Station in Fort Collins, Colorado. Project
Leader Mario Martinelli, Jr., says the mission
of his unit is to find ways to minimize the
hazard of avalanches to life and property.

Avalanches, such as this, are common in the moun¬
tains of the western United States. Most go unno¬
ticed by man.

9

One problem of increasing concern is the
development of second-home, resort, and re¬
tirement communities near or in avalanche
paths and runout zones — the bottom boun¬
dary of an avalanche path. This concern is com¬
pounded by the general lack of a systematic ap¬
proach toward land use planning in hazardous
areas.

Station scientists have developed a system
which combines early studies by European
scientists with recent findings to produce for¬
mulas for predicting how far an avalanche will
travel, its mass, and the impact force it will
carry. These formulas, and direction for their
use, will be of interest to government officials
and private developers who have the responsi¬
bility of zoning and recreation planning in
mountain areas.

Avalanche Warning System

In addition to zoning hazards, other
problems result from winter recreation. In the
backcountry, avalanche control measures, as
might be used in developed ski areas or along
highways, are nonexistant. Cross country
skiers, winter hikers, and mountaineers are
prime victims of avalanches in these areas.

In 1973, a joint venture between the Na¬
tional Weather Service and the Rocky Moun¬
tain Station resulted in an “Avalanche Warn¬
ing System’’ to help evaluate current hazard
conditions in Colorado’s mountains, and to
warn mountain residents and travelers when
avalanche danger is high. Most avalanche ac¬
tivity is directly related to precipitation, wind
speed, and other weather factors.

The Forest Service has established nu¬
merous weather reporting stations in critical
avalanche areas. Information is collected by
on-the-spot observers or by instruments that
relay observations to automatic recording sta¬
tions. These data, combined with snow condi¬
tions and weather reports from other govern¬
ment and private cooperators throughout the
State, help researchers develop an up-to-date
evaluation of potential avalanche activity in
Colorado’s mountains. Bulletins are issued to
local media for broadcast to the public when¬
ever avalanche danger exists.

Increased mountain development has placed some
structures in or near avalanche paths. This condo¬
minium is going up directly in an area noted for ava¬
lanches. Note the dead and torn trees as a result of
avalanche activity.

The “ Avalanche Warning System" was designed to
warn recreationists, such as these cross country
skiers, of existing avalanche conditions before they
enter into the backcountry.

10

Similar avalanche warning systems have
been, or are being, organized in other western
states.

Rocky Mountain Station researchers are
also trying to improve their ability to predict
when a given snowpack might release from a
slope. Dick Sommerfeld, research geologist
with the project, has placed recording sensors
in the snow at Berthoud Pass, Colorado, an area
notorious for large and frequent avalanches.
Results from this study are still being eval¬
uated. Indications are that prior to an ava¬
lanche, the snow emits certain sounds due to
the shifting and collapse of the internal snow
structure. It is hoped that the analysis of such
noise emissions from a snowpack will improve
the accuracy of avalanche forecasts in heavily
used mountain areas.

Blowing snow

Snow management problems are not
unique to the mountains. A newly constructed
77-mile section of Interstate 80, between
Laramie and Walcott in southeastern Wyom¬
ing, was opened to traffic in October of 1970.
Keeping it open in winter became an immediate
challenge. Wyoming’s high plains are charac¬
terized by strong winds and low, sparse vegeta¬
tion. Snow transported by these winds resulted
in large drifts and ice accumulating on the
highway and poor visibility for motorists.

The cost of removing snow to keep 1-80
open during the winter proved excessive. The
Wyoming Highway Department (WHD), with
hopes of finding answers to this problem, con¬
tacted the Rocky Mountain Station’s Snow
Management Research Project at Laramie. For
several years researchers at the lab had been
experimenting with snow fences to control
blowing snow. This research resulted in new de¬
sign techniques to make snow fences more effi¬
cient and economical than earlier methods.

Using these techniques, Ron Tabler, leader
for the Laramie project, and Bob Jairell, re¬
search technician, designed a system of snow
fences for the most critical locations along the
highway.

During the next two years the WHD
erected over 60,000 feet of 12-foot-high fences
along 1-80. By the winter of 1973-74, a total of

17.06 miles of the highway were protected at a
total cost of $1.5 million.

The fences made a dramatic difference.
Most of the blowing snow was trapped before it
reached the highway. Plowing was virtually
eliminated along the fence sites. In addition,
visibility in the lee of the fences was greatly im¬
proved during ground blizzards. A 1975 cost
analysis study showed the fences cut snow re¬
moval costs in half, an annual savings of
$80,000, based on the 1975 dollar value. At this
rate, savings in snow removal alone will pay for
the fences in about 15 years. In addition, there
was a 15 percent reduction in accidents and a
30 percent reduction in damage costs.

This 12-foot tall fence is designed to withstand over
100 mph winds and remain in service for about 25
years.

Although this snow fence system has been
of great benefit to all involved, blowing snow
still occurs along 1-80, and at times severe con¬
ditions reduce visibility to the point that the
highway must be closed altogether.

The Laramie and Fort Collins projects
worked together to develop a “visual range
monitoring system.’’ Part of this system is an
electronic snow particle counter that measures
the size and number of particles transported by
wind at a given point and time. With this in¬
strument, and an anemometer, it was possible
to develop a technique for estimating the visi¬
bility motorists might experience during blow¬
ing snow conditions.

11

Snow is removed from a trouble spot prior to fence installation (L). With fences, the same spot is clear, even
at times of peak accumulation.

The WHD has put this research to good
use in developing a “road condition monitoring
system” that gathers information from strate¬
gically located weather instruments and relays
it to the WHD. These data are recorded con¬
tinuously on chart recorders and analyzed by
computer to help make traffic regulation deci¬

sions. Warnings to motorists can then be dis¬
played on variable message electric signs. This
unique system warns motorists of icy roads,
blowing snow conditions, safe driving speeds,
driving time to the next town, and alternate
routes of travel should conditions warrant clo¬
sure of this section of 1-80.

R. A. Schmidt, research hydrologist, checks particle counter (L) that relays blowing snow data to a WHD com¬
puter for visibility analysis. This information is then made available to motorists, on variable message signs.

12

Experience gained from the research and
application of new fencing techniques along
Wyoming’s 1-80 is being used in other states
for help in managing blowing snow.

Other accomplishments

Snow research at the Rocky Mountain Sta¬
tion has played an important role in other
areas, too. For example:

— The two projects recently helped a pipe¬
line company assess the feasibility of using
fences to trap blowing snow for a road base,
thereby keeping equipment off the fragile
Alaskan tundra.

— New methods for highway design have
been developed to help keep them free from
drifting snow.

— New techniques have been applied to trap
blowing snow and improve the availability of
water for livestock and irrigation — a com¬
modity of increasing importance in the
drought-stricken West.

— Snow fences have been used in several
western ski areas to trap and hold blowing
snow on ski slopes, and to reduce avalanches
by keeping snow off hazardous slopes.

Continued research, along with coopera¬
tion from private and public organizations, will
help man better cope with snow management
problems and use snow to his benefit.

Readers requesting additional information
on blowing snow research at the Rocky Moun¬
tain Station should contact Ron Tabler at
Laramie, Wyoming, (307) 742-6621, FTS
operator — 328-1110.

For additional information on avalanche re¬
search, call Mario Martinelli, Jr., at Fort
Collins, Colorado, (303) 482-7332, FTS - 323-
1207.

Publications available from the Rocky
Mountain Station describing these and related
studies in greater detail include:

Judson, Arthur, and Bernard J. Erickson.
1973. Predicting Avalanche Intensity from
Weather Data: A Statistical Analysis. USDA
Forest Service Research Paper RM-112-FR12,

12 p.

Judson, Arthur. 1975. Avalanche Warn¬
ings: Content and Dissemination. USDA For¬
est Service Research Note RM-291-FR12, 8 p.

Judson, Arthur. 1976. Colorado’s Ava¬
lanche Warning Program. From Weatherwise,
Vol. 29, No. 6, December 1976, 10 p.

Leaf, Charles F., and M. Martinelli, Jr.
1977. Avalanche Dynamics: Engineering Ap¬
plications for Land Use Planning. USDA
Forest Service Research Paper RM-183-FR12,
51 p.

Martinelli, M., Jr. 1974. Snow Avalanche
Sites: Their Identification and Evaluation.
USDA Agriculture Information Bulletin 360-
FR12, 27 p.

McNair, Donald, and Frank Wolfe, Jr.
1977. An Acoustic Emissions Monitoring Sys¬
tem for Avalanche Snowpacks. USDA Forest
Service Research Note. RM-340-FR12, 4 p.

Perla, Ronald I., and M. Martinelli, Jr.
1976. Avalanche Handbook. USDA Forest
Service Agriculture Handbook 489, 238 p.
Available from the Superintendent of Docu¬
ments, U.S. Government Printing Office,
Washington D.C. 20402, Price $3.95. Request
Stock No. 001-000-03466-9/Catalog No.
Al. 76:489.

Tabler, Ronald D. 1973. Snow Fences Im¬
prove Highway Safety. Pub. Works 104(8):74-
75.

Tabler Ronald D. 1973. New Snow Fence
Design Controls Drifts, Improves Visibility,
Reduces Road Ice. From the Annual Transpor¬
tation Engineering Conference (Denver, Colo¬
rado, February 1973) Proceedings 46:16-27.

Tabler, Ronald D. 1974. New Engineering
Criteria for Snow Fence Systems. From
Transp. Res. Rec. 506, p. 65-78. NAS-NRC,
Washington, D.C.

Tabler, Ronald D. 1975. Predicting Profiles
of Snowdrifts in Topographic Catchments.
From the Western Snow Conference Proceed¬
ings (Coronado, California, April, 1975) Pro¬
ceedings 43, p. 87-97.

Tabler, Ronald D. 1975. Estimating the
Transport and Evaporation of Blowing Snow.
From the Snow Management on the Great
Plains Symposium (Bismarck, North Dakota,
July, 1975) Great Plains Agricultural Council
Publication 73, 186 p.

—By Rick Fletcher, Rocky
Mountain Station

13

Scientists create a
model tussock moth

outbreak

When the next outbreak of the Douglas-fir
tussock moth rolls around, forest man¬
agers in the western United States will
have a sophisticated new tool — a compu¬
terized crystal ball — to help them make deci¬
sions about what, if anything, should be done
to control the moth.

When the most recent of the periodic out¬
breaks of this defoliating insect occurred in the
western United States in 1971-74, professional
foresters had no systematic basis for reconcil¬
ing the differing predictions from experts
about the probable duration of the outbreak
and the amount of damage it would cause. Be¬
cause of heavy defoliation in some areas, and
land managers’ concern about additional dam¬
age, thousands of acres of western conifer for¬
ests were sprayed with DDT.

The computerized crystal ball is really
three computer models that will be used to in¬
tegrate knowledge about the Douglas-fir tus¬
sock moth and forecast the effects of an out¬
break on tree stands and watersheds. The
models are one of the products of the expanded
Douglas-fir Tussock Moth Research and Devel¬
opment Program, established in late 1974 to
help find new answers to the tussock moth
problem.

Three principal models — the Stand Prog¬
nosis Model, the Stand-Outbreak Model, and

the Socio-economic Model — will be ready for
testing by the time the research and develop¬
ment program concludes in September 1978.
The Stand Prognosis Model is actually a series
of regional models structured to describe host-
type timber stands throughout the range of the
tussock moth. These models will process ac¬
cumulated information on the progressive con¬
dition of stands, assuming “normal” growth
rates and planned timber harvest. The Stand
Outbreak Model will project the interaction be¬
tween stand characteristics at the start of an
outbreak, behavior of the insect during the out¬
break, and methods for control. The Socio¬
economic model deals with several kinds of
costs, for example, the cost of artificial control
methods if those are employed, and the costs of
the outbreak to timber, water, wildlife, and rec¬
reation values.

A model is a formula for predicting what is
likely to happen — given certain circumstances
— on the basis of what has happened in the
past. It is a substitute for an expert who might
keep all relevant knowledge constantly in mind
and apply it in the correct sequence to forecast
a result. For example, an ideal tussock moth
outbreak expert would be able to look at a
sample of moth larvae, appraise the condition
of the timber stand the sample came from, and
predict the progress of the outbreak and the
amount of damage over the duration of the out¬
break. He would not only have comprehensive
knowledge about the insect and the impact of
past outbreaks on timber stands, but he could
also make the lightning fast calculations of a
computer.

14

Two sides of a cross section from a sample tree will
be cut away in preparation for gathering growth
data for the Normal Stand Model for a timber stand.

Data on natural tree growth is an example of
information collected in the field and laboratory
to form the basis for models.

The ideal tussock moth expert exists — not
as one person but as a team of scientists and
mathematicians skilled in the use of com¬
puters. Entomologists who have studied the
insect’s structure, physiology, and behavior for
many years are pooling the knowledge they
have accumulated in the laboratory and forest.
A skilled modeler is working with the ento¬
mologists to condense their knowledge into its
essential relationships and produce a mathe¬
matical structure that duplicates the processes
of a real outbreak. Scott Overton, professor of
forest biometrics at Oregon State University in
Corvallis, Oregon, is the mathematician-
modeler for the Stand-Outbreak Model. He and
several colleagues are structuring the equa¬
tions of the model and translating them into
mathematical instructions so the thousands of
calculations required to project all possible out¬
comes of a real outbreak can be made by the
computer and printed out for the decision¬
maker. Overton describes modeling as “placing
a mathematical structure on knowledge.”

Robert Monserud, an associate of model-maker
Albert Stage, Intermountain Station, Moscow,
Idaho, and Research Technician Jonalea Tonn
examine a groove cut into the rectangular slab to
make the growth rings stand out sharply.

The grooved slab is placed on an optical scanner
which can be operated to read the width of the
growth rings and feed the information to a key
puncher. Each ring width is recorded on a card.
Together the cards tell the complete growth history
of the sample tree.

15

The essential process in the model, as in a
real outbreak, is the amount of defoliation the
insect will cause under given conditions. When
an outbreak is recognized, probably from a
spring survey of larval hatch, the forest man¬
ager will furnish certain information concern¬
ing the species mixture of the affected stand,
habitat community type, age or size of trees,
foliage composition, and the larval counts.
Given this information, the forest manager can
request a computer run to predict the esti¬
mated damage.

The working model

The following information from the Stand-
Outbreak Model illustrates the way models
work. Equations translate the development
and feeding behavior of the insect into the
amount of defoliation the insect will cause in
each life stage, and during each year of the
course of the outbreak. For example, one equa¬
tion represents the effect of a given number of
larvae on a typical tree branch. This equation
takes into account the amount of food each
larva must eat to pass through its develop¬
mental stages and the amount it destroys but
does not consume. Newly hatched larvae feed
on the underside of new needles, often taking a
single bite from each needle — enough to put
the needle out of business but not doing very
much to fill up the growing larvae. The amount
of foliage damaged varies with the species of
tree and the age of the larvae. Newly hatched
larvae require new needles for food, but as they
get older they can feed on older needles. How¬
ever, this relationship varies with the geo¬
graphic region and the density of the larval
population. These few facts suggest the enor¬
mous detail and variations that are built into
the equations of the Stand-Outbreak Model.

Estimates of the amount of top kill, tree
mortality, and growth loss are based on pro¬
jected patterns of defoliation in the affected
stand if natural controls take their course.
These natural controls include birds which eat
both larvae and pupae, and other organisms —

such as insect predators, parasites, and dis¬
eases. These factors presumably have con¬
trolled moth populations for thousands of
years — but not before some defoliation has oc¬
curred, and occasionally significant damage, as
in 1971-74.

The effects on the outbreak of direct
controls, such as aerial sprays of natural micro¬
bial materials and chemicals, are included by
specifying the insect mortality expected from
available control programs. In cases where na¬
tural controls are affected by direct control
methods, these factors are also specified.

The Stand-Outbreak Model is constructed
to forecast the effects over the duration of a
typical outbreak, plus three post-outbreak
years. “It is important to realize that the model
cannot predict outbreaks,” says Overton. “It
can only forecast the course of an identified
outbreak.”

The condition of the affected stand at the
end of the outbreak can then be incorporated
into the Normal Stand Model. Some of these
models have already been built to describe
stands in western National Forests. They are
based on prognosis methods developed by
Albert Stage of the Intermountain Station,
who is in charge of this phase of modeling.

The third major model is the Socio¬
economic Model, which is being developed
under the leadership of Gerard Schreuder, pro¬
fessor of forest economics, and Director of the
College of Forest Resources at the University
of Washington in Seattle. This model will
contain equations to project the costs of all
possible outcomes of an outbreak and ways of
dealing with it. Among these costs are the ef¬
fects of defoliation on timber growth and mor¬
tality, fish, wildlife, water, and recreation, as
well as the costs for purchase and application
of control materials.

The final product from the three models
will be a series of estimates of the results and
costs of each course of action that might be
taken regarding an outbreak.

A sound tool

Entomologist Bob Campbell of the Pacific
Northwest Station’s Corvallis Laboratory

16

serves the Tussock Moth Program as leader of
the Integration Working Group. This group is
responsible for putting both old and newly-de¬
veloped information into a pest management
system. Regarding the models, he says, “Al¬
though they are imperfect in some details, they
form a conceptually sound tool which is flexible
and based on data from the forest manager’s
local area. The models are based on more than
20 years of basic research on both the pest and
its impact on forests.’’

When the next outbreak occurs the family
of models will be available for trial. They will be
operated by the Methods Application Group at
Davis, California. The purpose of this group,
part of the Forest Service’s Insect and Disease
Management Staff, is to develop, test, and
make available methods and technology for
dealing with insect and disease management
problems in western forests.

When the next outbreak threatens, the for¬
est manager will be able to ask for projections,
based on the larval sample and characteristics
of his stand, to forecast the outcome of all the
options he wishes to consider. These forecasts
will list the losses the manager can expect in
timber mortality and growth over a rotation, as
well as losses to recreation, wildlife, fish habi¬
tat, and water quality. Perhaps the cost of all
available artificial controls will be greater than
the anticipated loss in forest values. The forest
manager may even discover that the projected
timber losses are about the same as what he
plans to harvest and decide to adjust his
harvest plan and save the cost of control.

The computer print-outs will not make
decisions. The consequences of the options will
still have to be weighed, but, the models and
the vast store of information they represent
will provide sound biological and economic
bases for making decisions.

As additional knowledge is discovered by
scientists after the expanded 4-year program
concludes, succeeding generations of models
will update and refine the modeling process.
However, an operational model will remain
available at Davis.

Some day researchers may discover what
triggers a population increase of the Douglas-
fir tussock moth every 7 or 8 years, and out¬
breaks may be prevented. Until that time,

scientists, speaking through the family of com¬
puter models, will provide coordinated, scien¬
tific projections on which forest managers can
base their decisions.

Information about these models may be
obtained from Ken Wright, manager of the
Douglas-fir Tussock Moth Program, at the Pa¬
cific Northwest Station — (503) 234-3361, FTS-
429-4918.

— By Dorothy Bergstrom, Pacific
Northwest Station

Aquatic habitat
study begins

What influence does livestock grazing
along the high mountain meadow streams of
the Idaho Batholith have on fish and aquatic
habitat?

William S. Platts, fisheries biologist at the
Intermountain Station’s Boise laboratory, is in
charge of new studies that will provide reliable
answers to the question. The research will be
conducted within the Boise and Sawtooth Na¬
tional Forests on 10 sites characteristic of
Idaho Batholith meadow ecosystems.

Platts and his coworkers will evaluate
stream channels and banks and streamside en¬
vironment during the grazing season. They will
document natural changes in streamflow and
identify how different grazing systems affect
fish populations and the aquatic and stream-
side vegetation. They will also determine if fish
cover and bank stability can be maintained or
improved with alternative grazing systems.

17

Publications

Aspen management

Quaking aspen is an important wild¬
life food in the Southwest. It is available
during most of the year, and provides
needed protein for deer and elk when other
plants are unavailable or are low in nu¬
trient value. Although some clones pro¬
duce enormous numbers of tiny seeds,
aspen reproduction in the Southwest is
mostly by sprouting from roots.

Ninety-three percent of the existing
aspen stands in the Southwest are mature
or overmature, and do not have large
numbers of sprouts. A recent report dis¬
cusses factors that affect the regeneration
and management of the tree. Aspen does
not usually form large stands in the
Southwest; small aspen patches are gener¬
ally associated with larger stands of coni¬
fers. The report suggests that because of
this characteristic, management of aspen
for deer and elk should be aimed at provid¬
ing large amounts of leaves and twigs for
food. Nearby conifers will provide needed
cover. The report states that unless
stands are regenerated by cutting or burn¬
ing to enhance sprout growth, aspen acre¬
age in Arizona and New Mexico will grad¬
ually decline.

Copies of “Managing Aspen for Wild¬
life in the Southwest,’’ General Technical
Report RM-37-FR12, by David R. Patton
and John R. Jones, are available upon re¬
quest from the Rocky Mountain Station.

Producing trees and elk

Trees and elk are renewable resources
which occupy many of the same areas
throughout the western United States.
Evidence shows that both can be pro¬
duced in the same area. Information about
the effects of human activity on elk be¬
havior can help land managers make full
utilization of the land without losing the
wildlife resource.

A recent report describes the results
of a 4-year study of elk behavior in relation
to timber harvesting and forest road
traffic. Radiotelemetry, time-lapse pho¬
tography, and counts of elk tracks on
roads were used to assess elk behavior in
relation to timber harvest activities on the
Medicine Bow Range in south-central
Wyoming. Results show that elk preferred
to stay at least 800 meters (V2 mile) from
people engaged in timber harvest or clean¬
up operations. Elk move back to the har¬
vested areas soon after human activity,
especially that beyond 400 meters (Va mile)
from the road. Most road crossings by elk
were made where desirable feeding sites,
such as clearcuts and natural openings
occurred near the road.

More information on this study can be
found in “Elk Behavior in Relation to
Timber Harvest Operations and Traffic on
the Medicine Bow Range in South-Central
Wyoming’’, by A. Lorin Ward, a reprint
from the Elk — Logging — Roads Sym¬
posium Proceedings, held December, 1976
at the University of Idaho at Moscow,
Idaho. Copies are available from the
Rocky Mountain Station.

18

Dwarf mistletoe spread
studied

Dwarf mistletoe that is infecting
otherwise-healthy young red or white firs
may spread so slowly that it will not ser¬
iously affect the trees until their height
growth slows down at maturity, a Cali¬
fornia study shows. Plant Pathologists
Robert F. Scharpf of the Pacific South¬
west Station, Berkeley, and John R. Par-
meter, Jr., of the University of California,
Berkeley, conducted the study to learn
how fast dwarf mistletoe builds up and
spreads through the crowns of true firs
that are not exposed to infected overstory
trees. They reasoned that this rate of
spread, in relation to tree height growth,
could be used to predict future levels of
infection and impact on tree vigor.

For the experiment, they located 26
young firs that were out of the range of
any overstory source of mistletoe infec¬
tion, and inoculated them with dwarf mis¬
tletoe seeds. These trees ranged from 5 to
32 feet in height, and were located on sites
in the southern Cascades and central
Sierra Nevada.

The study results, based on examina¬
tions of the trees during a 12- to 15-year
period following the inoculation, showed
that dwarf mistletoe spread from the
original site of infection to new infection
sites further up in the crown at an average
rate of only 3 inches per year, or less. All
of the test trees grew in height several
times faster than the parasite spread.

The researchers say these results sug¬
gest that dwarf mistletoe will have little
impact on young-growth firs, growing on
good sites free from exposure to infected
overstory trees, if the young-growth
stands are “managed for optimum height
growth consistent with other management
considerations.”

Details are in Research Paper PSW-
122-FR12, “Population Buildup and Verti¬
cal Spread of Dwarf Mistletoe on Young
Red and White Firs in California.” Copies
are available from the Pacific Southwest
Station.

New soil core sampler

Collecting soil samples from steep,
rocky sites has always been a tough
problem — too many rocks make digging
impractical, and the terrain may prevent
the use of heavy, power-driven equipment.

M. F. Jurgensen, Michigan Tech¬
nological University, Houghton; M. J.
Larsen, Forest Products Laboratory,
Madison, Wisconsin; and A. E. Harvey,
Intermountain Station, have developed a
device that solves many of the problems —
a portable, hand-operated soil core sam¬
pler. It provides a 4- by 11.8-inch soil core
and can be built by most machine shops.

Called the “Thumper,” the sampler
has two basic units — a soil-coring cyl¬
inder, and a combination cylinder driver
and soil core extractor. Two people can
operate the device, and it is designed so
that it can be taken apart and carried in a
backpack.

Harvey, a plant pathologist at the
Intermountain Station’s Forestry
Sciences Laboratory, Missoula, needed
over 2,000 soil samples for a study on the
environmental effects of timber harvest¬
ing on steep, rocky sites in western Mon¬
tana. “It was a matter of trial and error,”
Harvey says, “but we finally got a
sampler that worked.”

The sampler is described in “A Soil
Sampler for Steep, Rocky Sites,” INT-RN-
217-FR12, written by Jurgensen, Larsen,
and Harvey. Copies are available from the
Intermountain Station.

19

How long will they stand?

Before a forest fire is “dead out,” land
managers begin planning for the future of
standing snags left in the aftermath. Pulp-
wood, house logs, firewood, and wildlife
habitat are among the alternative uses.
One of the most pressing questions is,
“How much time do we have before the
snags go down?”

As part of a study of postfire vegetal
succession on the Sleeping Child Burn,
Bitterroot National Forest, Montana, L.
Jack Lyon, Intermountain Station wildlife
biologist, examined snag attrition. He
periodically counted the number of stand¬
ing lodgepole pine snags on eight tran¬
sects within the 28,000-acre burn. Initial
data were taken one year after the fire in
1962; counts were repeated in 1963, 1966,
1969, 1971, and 1976.

Following two years with little wind-
throw, snags on this burn fell at an annual
rate of 13.4 percent. At current rates,
most of the snags will fall within the next
40 years. A few larger snags, less sus¬
ceptible to windthrow, will stand indefi¬
nitely.

Details of Lyon’s study are available
in “Attrition of Lodgepole Pine Snags on
the Sleeping Child Burn, Montana,” INT-
RN-219-FR12. Write to the Intermoun¬
tain Station for a copy.

Habitat requirements for
deer and elk

Guidelines to help land managers and
resource planners evaluate the conse¬
quences for deer and elk of timber man¬
agement activities in the Blue Mountains
of eastern Oregon and Washington are

outlined in a recent publication from the
Pacific Northwest Station. The essential
integrating factor in considering habitat
for Rocky Mountain elk and Rocky Moun¬
tain mule deer in timber management
planning is the ratio of cover to forage
areas. The optimum ratio is estimated as
40-percent cover to 60-percent forage
areas, although the sizes of cover patches
and forage areas, and their location in rela¬
tion to each other, must meet certain re¬
quirements.

A series of tips for enhancing deer and
elk habitat, while producing timber, are of¬
fered. The publication concerns habitat
management in the Blue Mountains, but
the principles of management apply to
other parts of the West.

The paper is the result of a coopera¬
tive effort by the State Departments of
game in Oregon, Washington, and Idaho,
the U.S. Bureau of Land Management,
and the U.S. Forest Service and was pre¬
sented at the Elk-Logging-Roads Sympo¬
sium held at Moscow, Idaho, December
1975. Reprints of “Relationships of Rocky
Mountain Elk and Rocky Mountain Mule
Deer Habitat to Timber Management in
the Blue Mountains of Oregon and Wash¬
ington,” by Hugh Black, Richard J. Scher-
zinger, and Jack Ward Thomas are avail¬
able from the Pacific Northwest Station.

Graphs by computer

Graphs drawn by a plotter attach¬
ment to a computer can be used directly in
publications, under a program used at the
Pacific Northwest Station. Computer-
drawn graphs have two principal advan¬
tages over hand-drawn graphs: they are
more accurate and take only minutes to
prepare instead of hours. By following
requirements for submitting data, a scien¬
tist can dispense with the chore of plotting

20

his own graphs and also save many hours
for the illustrator.

A program which produces graphs of
good quality that require very little addi¬
tional preparation before publication was
developed by Mensurationist Don
DeMars in 1973 for use with the Wang 700
Series Calculator and has been used to
produce graphs for many publications
since that time. DeMars and Illustrator
Delbert Thompson expect that in the fu¬
ture most manuscript graphs will be
drawn by computer.

The format for plotting data can be
used with all types of computers that have
plotting capability. In addition to produc¬
ing graphs for publications, the system de¬
veloped by DeMars can assist forest man¬
agers and scientists in their daily work by
spotting errors in collecting data and
plotting equations.

The requirements for submitting data
have been outlined in “Computer Plotting
of Graphs for Reports,” Research Note
PNW-289, by Donald J. DeMars and Del¬
bert E. Thompson. The publication is di¬
vided into three main sections: 1) plotting
data requirements regardless of the type
of graph used; 2) graph types and their in¬
dividual requirements, and 3) special
plotting techniques. Copies are available
from the Pacific Northwest Station.

Water makes firelines

A simple technique for installing fire¬
lines before prescribed burning of light
rangeland fuels utilizes plain water and
standard firefighting-equipment. Pickup
pumpers are used to apply water through
two 4-foot wands attached to a Y. Back¬
pack pumps are also used to apply water.
The process is rapid, effective, and does
not permanently disturb the landscape.
Costs are low.

Three variations of the technique are
described in a recent publication, “Wetline
Technique for Prescribed Burning Fire¬
lines in Rangeland,” Research Note PNW-
292, by Robert E. Martin, Stuart E. Cole¬
man, and Arlen H. Johnson. Copies are
available from the Pacific Northwest Sta¬
tion.

You’ll find lots of good reading in the
January issue. Feature articles will cover:
Forest Genetics; Habitat Needs of Wood¬
peckers; Forest Insect Research; and
more.

If you know of someone who would be
interested in this publication, he or she
can be added to the mailing list by filling
out the coupon below and mailing it to us.

i - -t

Please add my name to the mail¬
ing list for Forestry Research:
What's New in the West:

ZIP

Mail to: Forestry Research:

What's New in the
West

U.S. Dep. of Agriculture
F orest Service
240 W. Prospect St.

Fort Collins, CO 80521

21

q. U. S. GOVERNMENT PRINTING OFFICE 1977 - 780-676/71 Reg. 8

~0

O

3

Qj

'"+■

^ o

o' $
o o'
^ti S'

S' DO
Oj ^

r — ^3

r> CD

xCb Co

DO

FORESTRY RESEARCH: Whafs Hew in the West POSTAGE AND FEES PAID

U.S. Department of Agriculture Forest Service U.S. DEPARTMENT OF AGRICULTURE

240 West Prospect Street AGR - 101

Fort Collins, Colorado 8052 1