American Journal of Plant Sciences, 2011, 2, 744-752
doi:10.4236/ajps.2011.26089 Published Online December 2011 (http://www.SciRP.org/journal/ajps)
Plant Community Development in a Dryland
CREP in Northeastern Oregon
John D. Williams1*, Heidi M. Hartman2, Lori M. Spencer3, James O. Loiland4
1USDA Agricultural Research Service, Columbia Plateau Conservation Research Center, Pendleton, USA; 2Umatilla Soil & Water
Conservation District, Pendleton, USA; 3Department Environmental Science and Regional Planning, Washington State University
Tri-Cities, Richland, USA; 4District Conservationist (USDA-NRCS), Dillingham Field Office, Dillingham, USA.
Received August 10th, 2011; revised September 13th, 2011; accepted October 9th, 2011.
ABSTRACT
Riparian areas in dryland crop regions of the Intermountain Pacific Northwest have largely been converted to cropland
or pasture during the last 140 years. Some formerly cultivated floodplains have become difficult to farm; enrollment o f
these lands into conservation programs provides the opportunity to use them as wildlife habitat and as buffer areas
near streams. Our objective was to evaluate plant community development on an USDA Conservation Reserve Enhan-
cement Program site in northeastern Oregon from when the plant was planted in 1999 through 2008. The researc h was
designed as a descriptive study. We established permanent line-transects to quantify vegetation establishment and
changes in species composition through time. We collected data in 2000-2001 and 2007-2008. Vegetation cover in
2000-2001 was 100%, dominated by tall wheatgrass. Living plant material cover decreased from 98% in 2000-2001 to
33% in 2007 and 68% in 2008; dead plant residue significantly increased and tall wheatgrass cover decreased. Native
species we re pre sen t in sim ila r pe rcen tag es fro m 2000 to 2008, although there was a shift from target to nontarget spe-
cies. The 1999 seeding can be judged a success because of the ground cover provided and the establishment of one tar-
get species, tall wheatgrass. However, the increased ratio of dead to living plant material and shift to non-target annual
weed species suggests that more active management (i.e., fire, grazing, or mowing) of the tall wheatgrass stand is
needed to maintain its productivity and/or a healthy mix of multiple species.
Keywords: Conservation Buffers, Riparian Management, Weed Control, Plant Community
1. Introduction
Although riparian areas comprise less than 2% of the
land area in the arid and semiarid western United States,
they contribute disproportionately to physical and bio-
logical processes. They serve as pathways for the flow of
energy, matter, and organisms through the landscape, act-
ing as ecotones between the terrestrial and aquatic zones
and corridors across regions [1]. Riparian vegetation
plays important roles in trapping soil eroded from up-
lands and removing nutrients from surface and soil water
[2,3], stream morphological dynamics [4,5], and aquatic,
avian, and large game habitat requirements [6-8].
Since the late 1800s, dryland small grain production
has been practiced on nearly all the arable land of the
inland Pacific Norwest [9]. Before widespread motorized
mechanization of farming practices in the 1940s, the bo-
ttomland of second and higher order streams in this
region were used extensively for grazing livestock, par-
ticularly draft horses, mules, and oxen. Beginning in the
1940s, much of this bottomland was converted to small
grain production, resulting in the elimination of natural
stream channels and riparian areas and the disruption of
flood plains.
Infrastructure (roads and railroads) maintenance requi-
rements, the need for farm operation efficiency, and go-
vernment incentives led to the channelization of many of
the streams in this region. Channelization creates steep
banks unprotected by vegetation cover or consolidating
root structure. Deep, channelized storm flow saturates
unprotected stream banks, creating positive pore pres-
sures that cause bank failure when the storm flow re-
cedes [10], and concentrates energy to transport soils
eroded from uplands, stream banks and bottoms to depo-
sition areas. Whereas the goal of stream channelization is
to drain soil water more efficiently, the effect is to dis-
connect the hydrologic flux between stream channel and

Plant Community Development in a Dryland CREP in Northeastern Oregon745
adjoining land. In forest or rangeland situations, this chan-
ge in hydrology facilitates the establishment of opportun-
istic weed species. In croplands, the rapid draining of soil
water short-circuits chemical and biochemical processes
that would occur if the water were resident longer. For
example; if water is stored in a floodplain from 2 to 10
days, nitrate concentrations would be reduced through
denitrification [11].
Functioning riparian areas are necessary to create mul-
tifunctional production systems as described by Jordan et
al. [12]. Until the late 1990s, efforts to restore or reha-
bilitate riparian areas occurred primarily in forests and
rangelands on public lands through the efforts of USDA-
Forest Service, USDI-Bureau of Land Management, and
USDI-National Park Service, with smaller scale private
land projects sponsored by nongovernmental organiza-
tions such as The Nature Conservancy. The introduction
in 1998 by the USDA of the continuous Conservation
Reserve Enhancement Program (CREP) provided crop
producers with the opportunity to reestablish some of the
structure and function of former riparian areas [3]. CREP
is a voluntary land retirement program intended to help
agricultural producers protect environmentally sensitive
land, decrease soil erosion, restore wildlife habitat, and
safeguard ground and surface water. Where this program
applies to lands bordering waterways, the stream must
provide current or historical habitat for threatened or
endangered fish species and must not be located above a
permanent barrier to fish passage. The program also ap-
plies to any area with a completed agricultural water qua-
lity management area plan, as well as reservation and
tribal trust lands. Eligible practices include planting and
maintaining riparian forest buffers, filter strips, wetland
restoration, fencing, off-site watering and others. Con-
tracts are generally 10 to 15 yr in duration. CREPs are
designed by the USDA-Natural Resources Conservation
Service (NRCS) and funded through the USDA-Farm
Service Agency (FSA). Technically, CREP projects are
not considered ecological restoration, because native and
non-native species are used [13,14], but are known in-
stead as rehabilitated production systems (RPS) [3]. The
first CREP project in northeastern Oregon was estab-
lished in 1999.
After initial post-implementation evaluation by NRCS
of the sediment filtering effectiveness of the seeded
grasses and establishment success of shrubs, CREP pro-
jects are visited periodically to assure compliance with
contractual agreements. Few systematic plant community,
soil sampling, or soil erosion studies are conducted in
RPS to evaluate project success in terms of structure and
function, although numerous studies on larger ecological
issues have used data gathered from these sites [3]. For
example, much of the research has been conducted as ha-
bitat evaluations of Conservation Reserve Program (CRP),
e.g., [15,16]. Systematic studies of individual CREP pro-
jects, particularly in the Pacific Northwest, are lacking,
although a statewide evaluation was recently conducted
in the state of Washington [17]. The objective of this
research was to conduct a multi-year vegetation observa-
tional study to describe plant community development in
the Gerking Flat CREP.
2. Materials and Methods
2.1. Study Site
Gerking Flat is located on Gerking Creek, approximately
on 26 km northeast of Pendleton, Oregon (45˚49'41"N,
118˚32'49"W). Gerking Creek is an intermittent tribu-
tary of Wildhorse Creek, the major northern tributary of
the Umatilla River, draining rainfed agricultural lands.
Gerking Creek enters the project site as an incised chan-
nel, broadens into a multi-channel stream in the mid-
section of the project, and then once again becomes an
incised channel in the lower one-third of the site. The
flood plain was only 2 - 4 m wide in the incised seg-
ments. The channel length through the project was 2.2
km (Figure 1), and extended 50 m on either side of the
channel, encompassing a total project area of 44.5 ha
(Figure 2).
In the early 1960s, Gerking Flat was converted to
small grain production in an annual spring barley system
from pasture. After flooding in December 1964, Gerking
Creek was straightened to improve surface runoff effi-
ciency, soil drainage for early-season field access, effi-
cient operation of farm machinery, and to reduce the im-
pact of salt accumulation on crop production. In subse-
quent years, drain tiles were installed to further assist
drainage of wet areas. However, barley yields were poor
because of excess soil water, salinity, and high soil pH.
Flooding and channel migration in 1996 and 1997, com-
bined with poor crop yields, led the landowners to con-
clude that barley production was no longer economically
feasible, and in 1999, they enrolled a portion of Gerking
Flat in a 15 yr CREP contract to retire unproductive land,
contribute to soil and water conservation, and create bird
habitat.
Planting in the project area was divided into three
zones based on then-current topography: 1) active stream
channel, 2) floodplain, and 3) upland. A mix of willows,
other trees, and shrubs were to be planted in zones 1 and
2 (Table 1), and grasses, forbs and shrubs in zone 3 (Ta-
ble 2). In this article, we describe our results in terms of
target/nontarget-native/nonnative species, where target
species are those that were planted and nontarget are vo-
lunteer or invasive (Baer et al. 2009). The project plan

Plant Community Development in a Dryland CREP in Northeastern Oregon 746
Figure 1. Location of Gerking Creek within Oregon.
Figure 2. Aerial photograph taken in 2008 of Gerking Flat
CREP, channels, and vegetation transects.
called for a density of 500 live trees per hectare. Based
on NRCS and Pheasants Forever planting lists, a total of
200 cottonwoods, 500 willows, and 875 shrubs of vari-
ous species were planted (Table 1). Trees and shrubs
were planted from 0.9 to 4.3 m or an average of 2.6 m on
each side of the channel in April, 2000, and January,
2002. Shrubs and trees were hand watered once after
planting.
Table 1. Tree and shrub species planted in 1999 in the Gerk-
ing Flat CREP (source: USDA- Natural Resources C onserva-
tion Service and Pheasants Forever).
Species No. Planted
Black cottonwood (Populus trichocarpa) 200
Willow (Salix exi q ua , Sa li x spp.) 500
Woods rose (Rosa woods ii) 75
Nootka rose (Rosa nutka na) 75
Snowberry (Symphor ic arpos al bus ) 100
American plum (Pr un us am erican a) 75
Chokecherry (Prunus virginiana) 100
Elderberry (Sambucus race mo sa) 50
Golden current (Ribes aureum) 50
Buckbrush (Ceanothus cuneatus) 50
Western white clematis (C lematis ligustici folia.) 25
Table 2. Target grass and forb species planted in 1999 in
the Gerking Flat CREP.
Species Percent of mix
Nonnative
Meadow foxtail (Alopecurus pratensis) 13
Alfalfa (M edicago sa tiva ) 10
Yellow sweetclover (Melilotus of ficin al is ) 8
Tall fescue (Festuca ar u ndin acea) 22
Tall wheatgrass (Thinopyrum ponticum) 11
Native species
Basin wildrye (Elymus cinerus) 6
Streambank wheatgrass (Elymus lanceolatus) 4
Western wheatgrass (Agropyron sm ithii) 5
Alkali sacaton (Sporobol us a ir o ides) 21
Prior to seeding, weeds were controlled with herbicides
and disking. The grass/forb mix was seeded in the spring
of 1999 at 22.4 kg/ha, with a seed weight ratio of 36:100
native to nonnative species. Post-seeding weed control
was conducted on the entire site by mowing in the spring
of 2000 and 2001. In 2002, the eastern, upstream quarter
of the site was disked lightly to control weeds and elimi-
nate the heavy mat of dead vegetation that had developed.
The landowner, in consultation with NRCS personnel,
concluded that burning some portion of the site annually
would not damage the established grasses, and would
result in less open ground for nontarget species invasion.
Subsequently, spring burns were conducted before bird
nesting season on approximately one-quarter of the site
each year from 2002 to 2005. Exact records of burn dates
and specific areas burned were not kept. Alleyways were
also randomly cut through the site each year to provide

Plant Community Development in a Dryland CREP in Northeastern Oregon747

upland game bird hunting access.
Meteorological records dating from 1931 at the USDA
Columbia Plateau Conservation Research Center and Ore-
gon State University Columbia Basin Agricultural Re-
search Center, 14 km south of the research site, show
minimum and maximum air temperatures of −34˚C and
46˚C, with a 71 yr average mean annual temperature of
11˚C. Annual frost-free days range from 135 to 170 [18].
Approximately 70% of precipitation occurs between No-
vember and April, with annual precipitation averaging
422 mm. Snow cover is transient, with accumulated snow
subject to rapid melting by frequent marine warm fronts
from the Pacific Ocean. Soils are Hermiston silt loams
(coarse-silty, mixed, mesic Cumulic Haploxeroll) and
Pedigo silt loam (coarse-silty, mixed, superactive, mesic
Cumulic Haploxerolls), formed in silty alluvium from
loess and ash on flood plains and low terraces. These
soils overlay the fractured Miocene basalt layers of the
Columbia Plateau, and slopes range from 0 to 3% [18].
2.2. Monitoring and Sampling Procedures
We established permanent line-point transects [19] ex-
tending 50 m perpendicularly east and west from the ac-
tive channel edge to quantify plant cover by species (Fig-
ure 2). Transects were established at regular intervals of
100 m throughout the project area using survey grade
GPS equipment. Each 50 m transect was stratified by
distance from the channel edge in 5 m increments, and
five points were randomly sampled within each of these
increments. Cover, by species, litter, or bare ground was
recorded at each point. Plant species were referred to as
target-native, target-nonnative, nontarget-native, and non-
target-nonnative, where target species were those that had
been planted under the CREP program [3]. We sampled
in May and June in 2000, 2001, and 2008, and in July
and August in 2007. Individual trees and shrubs in the
project area were counted in August, 2008; willow spe-
cies were identified separately from this census.
2.3. Experimental Design and Statistical
Procedures
The study reported here is descriptive. Data were gra-
phically and statistically analyzed to determine if there
were differences in species composition and vegetation
cover among years. We calculated Simpson’s diversity
index (N2),
 
2
N
NN 11,
ii
nn (1)
where n is the total number of organisms of a particular
species and N is the total number of organisms of all
species, and the Simpson equitability index (E),
2
ENS, (2)
where S = the total number of species identified at time
of sampling. These values were calculated, by year, for
the entire site [20]. Analysis of cover and species com-
position was conducted using a mixed-model, repeated
measures ANOVA GLIMMIX procedure to model the
response using a binomial distribution, and least square
means separation tests where significant main effects and
interaction terms were found [21].
3. Results and Discussion
3.1. Research Site Conditions
Weather conditions were warmer and drier than normal
from 2001 through 2008, relative to the previous 70 yr.
Mean annual temperatures were higher than normal in all
years except 2004 and 2005 and total annual precipita-
tion was below the long-term average in 2001, 2002,
2003, 2005, and 2008.
Construction of a building and pasture establishment
between 2001 and 2007 eliminated parts of two transects
(16 & 17). We were, however, able to conduct a census
of trees and shrubs planted within 10 m of the channel in
these areas.
3.2. Ground Cover
Total ground cover was nearly 100% during all years of
sampling, with means and standard errors of 99.8% ±
0.1%, 99.9% ± 0.1%, 97.1% ± 0.5%, and 97.3% ± 0.8%
for 2000, 2001, 2007, and 2008. However, bare soil in-
creased significantly from 0% in 2001 to 1.7% ± 0.5% in
2007 (P ≤ 0.05) (Figure 3), and there was a significant
increase in detritus (dead and down material) from 2001
to 2007 (Figure 4). The increase in bare soil and detritus
and decrease in living plant material from 98% in 2000-
2001 to 33% in 2007 and return to 68% in 2008 corre-
sponded with a reduction in tall wheatgrass and increase
in other nontarget species (Figure 5). After the tall wheat-
grass matured, largely as a monoculture in the first two
years, it began to die. In 2007 we found large areas of
standing dead tall wheatgrass 2 m in height, without any
other vegetation in the understory, where annual nonna-
tive species established in 2008.
The total contribution by native species to plant cover
was <10% during any year (Figure 6). Six nontarget-
native species contributed 3% and 16 nontarget-nonna-
tive species contributed 16% of the plant cover by 2008
(Table 3).
3.3. Plant Species Diversity and Community
Development
We identified 32 individual species of grasses and forbs
in the permanent transects over the course of the study,
of which 23 were nontarget species (Table 3). Not all

Plant Community Development in a Dryland CREP in Northeastern Oregon 748
Figure 3. Changes in bare soil demonstrating increased
plant community complexity from 2000 through 2008. Figure 4. Ground cover contribution by detritus from 2000 through 2008.
Figure 5. Composition by individual species at Gerking Flat from 2000 through 2008.
species were present every year, with 15 in 2000, 18 in
2001, 18 in 2007, and 21 in 2008. Across the site N2 was
1.6 in 2000 and 1.5 in 2001, increasing to 5.3 in 2007
with a shift down to 4.4 in 2008. This change in N2 indi-
cates an increase in species diversity. However, an in-
crease in diversity alone is not sufficient grounds to
judge a rehabilitation project a success. We will discuss
this concept further in this article. No shrubs or trees
were found in any of the line transects during any sample
year.
Despite a substantial die-off, tall wheatgrass remained
the dominant species in 2008 (Figure 5). The other tar-
geted species, native and introduced, contributed ≤10%
to total cover and only basin wildrye contributed >1%
cover in 2008. Of the remaining four target-nonnative
species, none contributed more than 2% cover during any

Plant Community Development in a Dryland CREP in Northeastern Oregon749
Figure 6. Nati ve species contri bution to pla nt co ver.
Table 3. Nontarget species identified in Gerking Flat CREP.
Species
Nonnative
Wild oat (Avena fatua)
Kocia, Mexican fireweed, mock cypress, Mirabel (Kochia scoparia
Ripgut brome (Bromus diandrus)
Downey brome (Bromus tectorum)
Canada thistle (Cirs i um arve nse)
Hairy fleabane, wavy-leaf fleabane, flax-leaf fleabane,
asthmaweed (Conyza bonariensis)
Prickly lettuce (Lactuca serriola)
Oxeye daisy (Chrysanthemum leucanthemum)
Catnip (Nepeta cataria)
Scotch thistle, Scotch cottonthistle (Onopordum acanthium)
Rabbitfoot polypogon, annual rabbitsfoot grass
(Poly pogon mo nspeliensis)
Tumble mustard (Sisym brium altissimum)
Pineapple weed, disc mayweed (Matriciaca ri a matric arioides)
Lambsquarters (Chenopodium album)
Fiddleneck Tarweed (Amsinckia lycopsoid es)
Cutleaf Nightshade (Solanum triflorum)
Black Mustard (Brassica nigra)
Native
Willowweed; dwarf fireweed (Epilobium latifolium )
[Canadian] Horseweed (Conyza canadensis)
False quackgrass (Elymus × pseudorepens)
Field horsetail (Equisetum arvense)
Hardstem bulrush (Sch oenop lectus acutus var. acutus)
Common cattail, broadleaf cattail (Typha latifolia L.)
American speedwell (Veronica americana)
year of sampling and none were present in 2008. Only
two of the four target-native species, basin wildrye and
alkali sacaton remained present in 2008, contributing 5%
and 2% cover, respectively.
Of the nontarget species identified, 16 were nonnative
and 7 were native. Four of the nontarget native species
were early colonizers of moist, primary successional
sites (field horsetail, hardstem bulrush, common cattail,
and American speedwell). Although considered weed
species in croplands, all might be expected to contribute
to native riparian habitat. Two nontarget natives, Cana-
dian horseweed, and dwarf fireweed or willowweed,
were found sparsely distributed throughout the site. False
quackgrass, a nontarget native found in 2001, was not
found in 2007 or 2008. The progressive shift from tar-
get-nonnative toward nontarget-nonnative dominance at
the site from 2000 to 2008 resulted from increases in
wild oats, kochia, Canada thistle, and tumble mustard.
Low species diversity and evenness values (on Gerk-
ing Flat N = 32, N2 ≤ 5.3, and corresponding E ≤ 0.29)
are commonly found in rehabilitated agricultural systems
[3]. Generally, these values are reported with a reference
value from local native remnant plant communities as a
measure of project success. For our purposes, and be-
cause we lack a native comparison site, we rely on the
general definition of the indices to conclude that the
Gerking Flat RPS has low species diversity and evenness.
The relevance of these values to the functioning of plant
communities is the center of an ongoing debate [22],
although it is important from a managerial perspective to
understand that fully functioning ecosystems generally
score higher in both indices.
3.4. Shi ft f r om Targ et t o No nt ar get Spe ci es
Target-nonnative species comprised significantly more
of the vegetation composition than all other classes (P ≤
0.05), during 2000 and 2001, but nontarget species in-
creased substantially in 2007 and 2008 (Figure 7). This
increase was predominately by Eurasian annuals typi-
cally found in disturbed semiarid and arid landscapes
(Figure 5). Nontarget-native species can alter the in-
tended development trajectory of an RPS [3]. These spe-
cies are exceptionally well adapted for invading dis-
turbed areas where soil conditions and lack of seed
sources reduce competition from native plants. Once est-
ablished, these communities of invasive species tend to
persist unless there is substantial management interven-
tion [3]. Management options in CREP are limited to
light soil surface disturbance (disking), mowing, burning,
and limited herbicide use. The producer managing Gerk-
ing Flat mowed the site after the first year to control seed
production from nonnative annuals. An attempt at disk-
ing part of the site was judged counterproductive, and the

Plant Community Development in a Dryland CREP in Northeastern Oregon 750
Figure 7. Vegetation composition by target-nontarget native-
nonnative species.
producer opted for burning some portion of the site each
spring thereafter. Burning eliminated accumulating dead
material, but apparently had no effect on slowing the
increase in nontarget-nonnative species. In Mississippi,
Greenfield et al. [16] were able to improve bobwhite
habitat by disking or burning a 10 yr old CRP field, but
the improvement was short-lived and plant community
composition was unaffected. Disking, mowing, and burn-
ing are effective if applied at the appropriate plant phe-
nological stage, i.e. before seed set. But such timing is
likely to interfere with critical wildlife use, such as nest-
ing. Grazing has been proposed as a means of upland
weed suppression (e.g. [23]), and can contribute to nu-
trient cycling advantageous to target species [3]. How-
ever, grazing is restricted in most CREP contracts, and
many of the CREP projects on the Columbia Plateau are
managed by single commodity producers without the
animal or managerial resources needed for grazing.
3.5. Tree and Shrub Development
Establishment of willows can be considered successful.
Of the 500 willows originally planted, we were able to
identify 279 individual plants in 2008. These were lo-
cated in the lower half of the project where most of the
tree and shrub planting took place (Mr. Bud Schmidtgall,
landowner, personal communication). The survival of
other trees and shrubs appears to have been low, and com-
pletely unsuccessful with respect to cottonwood, Nookai
rose, American plum, choke cherry, elderberry, golden
current, buckbrush, and western clematis. We recorded
one nontarget-native species, red osier dogwood. Com-
plete results of the tree and shrub census appear in Table
4.
3.6. Projec t Eva luat ion
The project on Gerking Flat met the basic objective of
Table 4. Trees and shr ubs f ound in z ones 1 a nd 2 (descri be d
in text) in a September, 2008, census.
Species Number
Willows (Salix spp.) 279
Pacific willow (Salix lucida ssp. lasiandra)
Scouler willow (Salix scouleriana)
Coyote willow (Salix exigua)
Dusky willow (Salix melanopsis)
Booth’s willow (Salix boothii)
Woods rose (Rosa woods ii) 15
Snowberry (Symphor ic arpos al bus ) 15
Red osier dogwood (Cornus sericea) 5
Black cottonwood (Populus trichocarpa) 4
providing ground cover, >90% in 2008, to conserve soil.
A continuous canopy of willows now covers the channel
in the lower one-third of the project. Increased root and
stem biomass slows erosion, both within and outside the
stream channels, and traps soil eroded from surrounding
fields and borrow ditches [2,24-26]. An early concern
with the planting of trees and shrubs in channels was
they would be too effective and cause the channel to mi-
grate, especially in areas where the original channel was
some distance, and lower, than the constructed channel.
Although more sinuosity in a channelized stream might
ultimately be a desirable objective, in the early phases of
the project channel movement could have jeopardized
establishment of newly planted vegetation.
Anecdotally, there seemed to be substantially more
raptors and upland game birds at the site than in sur-
rounding fields or adjacent to Gerking Creek above and
below the project. This was expected, as avian communi-
ties increase with increasing plant community complex-
ity, including variety in plant size and life form, and ac-
cumulation of detritus [15]. Before the project was begun,
the site was a mono-cropped agricultural field, and in
2000 and 2001 the site supported a monoculture of the
targeted-nonnative tall wheatgrass. By 2008, we observed
patchiness in the distribution of vegetation, with distinct
areas of tall wheatgrass, nontarget-nonnative species (pri-
marily annual or biannual forbs), and accumulations of
detritus in various states of decomposition throughout the
site.
4. Conclusions
The Gerking Flat CREP project planted in 1999 has ful-
filled the primary objectives of establishing a plant com-
munity with sufficient cover to trap sediment from off-
site, reduce erosion onsite, and provide cover and habitat
for upland game birds. In the first 2 yr after planting,
2000 and 2001, the stand was dominated by a target-

Plant Community Development in a Dryland CREP in Northeastern Oregon751
nonnative species, tall wheatgrass, with low overall di-
versity and evenness. After 6 and 7 yr, diversity values
and evenness values both increased incrementally. Com-
plexity of the site increased with an increase of detritus
in various stages of decomposition, and patchiness with-
in the community that was observed, but not captured by
the sampling regime. Nontarget-nonnative species in-
creased most at the site, which suggests that the current
spring burning regime will have to be supplemented by
other weed control measures to prevent conversion of the
site to an annual nonnative plant community. Possible
alternatives include well-timed applications of herbicide
or intensive grazing management, and are decisions that
must be taken in the context of landowner willingness to
keep the land enrolled in this program. Although the
immediate objectives of this project were met, establish-
ment and development of plant communities dominated
by nontarget-nonnative species create a seed source for
infestation of surrounding crop land, potentially creating
an economic drain on producers and ill will toward such
projects. CREP projects are finite and depend on the
competitiveness of program payments with alternate land
uses and landowner satisfaction with project develop-
ment and outcome. Arguably, participants will be chal-
lenged by economic pressures to return these sites to
production [27], even with successful establishment of
healthy stands of targeted species. However, emerging
resource concerns, such as downstream water quality
issues and ecosystem service markets, including the car-
bon sequestration potential of restored CREP sites, may
help counter such pressures. A more extensive evaluation
of plant communities, hydrologic response, and other
resource values in CREP projects should be undertaken,
and lessons learned throughout the Pacific Northwest
compiled to aide in future rehabilitation efforts.
5. Acknowledgments
We wish to thank Bud Schmidtgall, Athena, Oregon for
his cooperation in allowing us to conduct the sampling
and publish a manuscript based on that information.
Dave Robertson, Daryl Haasch, Felicity Dye, and Stefka
Waite aided in collecting and summarizing the data. In
addition to participation in this project by USDA-NRCS
and -FSA, resources were contributed by the Umatilla
Soil and Water Conservation District, Confederated
Tribes of the Umatilla Reservation, Oregon Department
of Forestry, Potlach Corporation, and Pheasants Forever.
REFERENCES
[1] R. T. T. Forman, “Land Mosaics: The Ecology of Land-
scapes and Regions,” Cambridge University Press, Cam-
bridge/New York, 1995.
[2] R. A. Pearce, M. J. Trlica, W. C. Leininger, D. E. Mergen
and G. W. Frasier, “Sediment Movement through Ripar-
ian Vegetation under Simulated Rainfall and Overland
Flow,” Journal of Range Management, Vol. 51, No. 3,
1998, pp. 301-308. doi:10.2307/4003415
[3] S. G. Baer, D. M. Engle, J. M. H. Knops, K. A. Lange-
land, B. D. Maxwell, F. D. Menalled and A. J. Symstad,
“Vulnerability of Rehabilitated Agricultural Production
Systems to Invasion by Nontarget Plant Species,” Envi-
ronmental Management, Vol. 43, No. 2, 2008, pp. 189-
196. doi:10. 10 07/ s00 267 -0 08- 91 67- 6
[4] S. J. Bennett, T. Pirim and B. D. Barkdoll, “Using Simu-
lated Emergent Vegetation to Alter Stream Flow Direc-
tion within a Straight Experimental Channel,” Geomor-
phology, Vol. 44, No. 1-2, 2002, pp. 115-126.
doi:10.1016/S0169-555X(01)00148-9
[5] D. Corenblit, E. Tabacchi, J. Steiger and A. M. Gurnell,
“Reciprocal Interactions and Adjustments between Flu-
vial Landforms and Vegetation Dynamics in River Cor-
ridors: A Review of Complementary Approaches,” Earth-
Science Reviews, Vol. 84, No. 1-2, 2007, pp. 56-86.
doi:10.1016/j.earscirev.2007.05.004
[6] J. B. Kauffman and W. C. Krueger, “Livestock Impacts
on Riparian Ecosystems and Streamside Management Im-
plications: A Review,” Journal of Range Management,
Vol. 37, No. 5, 1984, pp. 430-438. doi:10.2307/3899631
[7] F. L. Knopf, R. R. Johnson, T. Rich, F. B. Samson and R.
C. Szaro, “Conservation of Riparian Ecosystems in the
United States,” Wilson Bulletin, Vol. 100, No. 2, 1988, pp.
272-284.
[8] S. V. Gregory, F. J. Swanson, W. A. McKee and K. W.
Cummins, “An Ecosystem Perspective of Riparian Zones,”
BioScience, Vol. 41, No. 8, 1991, pp. 540-551.
doi:10.2307/1311607
[9] A. C. McGregor, “Counting Sheep: From Open Range to
Agribusiness on the Columbia Plateau,” University of
Washington Press, Seattle, Washington, 1982.
[10] A. Simon and A. J. C. Collison, “Pore-Water Pressure
Effects on the Detachment of Cohesive Streambeds:
Seepage Forces and Matric Suction,” Earth Surface
Process e s and Landfor m s , Vol. 26, No. 13, 2001, pp. 1421-
1442. doi:10.1002/esp.287
[11] D. W. Rassam, C. S. Fellows, R. D. Hayr, H. Hunter and
P. Bloesch, “The Hydrology of Riparian Buffer Zones;
Two Case Studies in an Ephemeral and a Perennial
Stream,” Journal of Hydrology, Vol. 325, No. 1-4, 2006,
pp. 308-324. doi:10.1016/j.jhydrol.2005.10.023
[12] N. Jordan, G. Boody, W. Broussard, J. D. Glover, D.
Keeney, B. H. McCown, G. McIsaac, M. Muller, H.
Murray, J. Neal, C. Pansing, R. E. Turner, K. Warner and
D. Wyse, “Environment: Sustainable Development of the
Agricultural Bio-Economy,” Science, Vol. 316, No. 5831,
2007, pp. 1570-1571. doi:10.1126/science.1141700
[13] A. D. Bradshaw, “Underlying Principles of Restoration,”
Canadian Journal of Aquatic Science, Vol. 53, Supple-
ment 1, 1996, pp. 3-9.
[14] J. B. Kauffman, R. L. Beschta, N. Otting and D. Lytjen,
“An Ecological Perspective of Riparian and Stream Res-

Plant Community Development in a Dryland CREP in Northeastern Oregon 752
toration in the Western United States,” US Fisheries, Vol.
22, No. 5, 1997, pp. 12-24.
doi:10.1577/1548-8446(1997)022<0012:AEPORA>2.0.C
O;2
[15] K. F. Millenbah, S. R. Winterstein, H. Campa III, L. T.
Furrow and R. B. Minnis, “Effects of Conservation Re-
serve Program Field Age on Avian Relative Abundance,
Diversity, and Productivity,” Wilson Bulletin, Vol. 108,
No. 4, 1996, pp. 760-770.
[16] K. C. Greenfield, M. J. Chamberlain, J. Bruger, L. Wes
and E. W. Kurzejeski, “Effects of Burning and Discing
Conservation Reserve Program Fields to Improve Habitat
Quality for Northern Bobwhite (Colinus Virginianus),”
American Midland Naturalist, Vol. 149, No. 2, 2003, pp.
344-353.
doi:10.1674/0003-0031(2003)149[0344:EOBADC]2.0.C
O;2
[17] C. J. Smith, “Evaluation of CREP Riparian Buffers in
Washington State. Prepared for Washington State Con-
servation Commission,” Washington State Conservation
Commis s i on Board Membe r s , Washington, 2006, p. 61.
[18] D. R. Johnson and A. J. Makinson, “Soil Survey of Uma-
tilla County area, Oregon,” USDA—Soil Conservation Ser-
vice, US Government Printing Office, Washington DC,
1988.
[19] C. D. Bonham, “Measurements for Terrestrial Vegeta-
tion,” John Wiley & Sons, New York, 1989.
[20] R. H. Green, “Sampling Design and Statistical Methods
for Environmental Biologists,” John Wiley & Sons, New
York, 1979.
[21] SAS, SAS/STAT 9.2 User’s Guide. SAS Institute Inc.,
Cary, NC, 2008.
[22] A. R. Ives and S. R. Carpenter, “Stability and Diversity of
Ecosystems,” Science, Vol. 317, No. 5834, 2007, pp. 58-
62. doi:10.1126/science.1133258
[23] J. McIver and L. Starr, “Restoration of Degraded Lands
in the Interior Columbia River Basin: Passive vs. Active
Approaches,” Forest Ecology and Management, Vol. 153,
No. 1-3, 2001, pp. 15-28.
doi:10.1016/S0378-1127(01)00451-0
[24] R. D. Harmel, C. T. Haan and R. Dutnell, “Bank Erosion
and Riparian Vegetation Influences: Upper Illinoies River,
Oklahoma,” Transactions of the American Society of Ag-
ricultural Engineers , Vol. 42, No. 5, 1999, pp. 1321-1330.
[25] K. Gran and C. Paola, “Riparian Vegetation Controls on
Braided Stream Dynamics,” Water Resource Research,
Vol. 37, No. 12, 2001, pp. 3275-3283.
doi:10.1029/2000WR000203
[26] Z. O. Toledo and J. B. Kauffman, “Root Biomass in Re-
lation to Channel Morphology of Headwater Streams,”
Journal of the American Water Resources Association,
Vol. 37, No. 6, 2001, pp. 1653-1663.
doi:10.1111/j.1752-1688.2001.tb03667.x
[27] T. L. Napier, “Grain Scarcity: A New Era for Conserva-
tion Policies and Programs,” Journal of Soil and Water
Conservat i on, Vol. 64, No. 1, 2009, pp. 7A-10A.
doi:10.2489/jswc.64.1.7A