Ecological Intensification of Oregon Hazelnut Orchards: Restoring Native
Plant Communities in Shared Ecosystems
by
Marissa Lane-Massee
A thesis accepted and approved in partial fulfillment of the
Requirements for the degree of
Master of Science
in Environmental Studies
Thesis Committee:
Lauren Hallett, Chair
Lucas Silva, Member
Lauren Ponisio, Member
University of Oregon
Spring 2024
2
© 2024 Marissa Lane-Massee
This work is openly licensed via CC BY 4.0
3
THESIS ABSTRACT
Marissa Lane-Massee
Master of Science in Environmental Studies
Ecological Intensification of Oregon Hazelnut Orchards: Restoring Native Plant Communities in Shared
Ecosystems
The rapidly expanding Oregon hazelnut industry offers a unique opportunity for restoring
ecosystem services to private lands that were historically oak-prairie-dominated habitats. With typical
orchard management consisting of bare-soil orchard floors, ecological intensification through the use of
native conservation cover may directly benefit farmers and their operations, saving time and money spent
on land management. With the hazelnut industry currently investing resources into young orchards, soil
management with cover crops has become a contentious area of research. Looking towards the future,
understanding how cover crops can be tailored towards an expanding and aging Oregon hazelnut industry
is imperative. Here, I study the feasibility of large-scale native conservation cover implementation in a
mature orchard, with measurements of compatibility with orchard management practices and desirable
ecosystem services that farmers can directly utilize. My results show that native conservation cover can
successfully suppress orchard weeds, align with important pest management timeframes, facilitate
hazelnut pickup during wet harvest years, reduce chemical and mechanical inputs, and while not having a
significant effect on soil moisture, significantly reducing soil temperature during summer months. This
study demonstrates the feasibility and compatibility of native conservation cover to be used in
commercial hazelnut systems, and the capacity at which native conservation cover directly benefits the
farmer and agroecosystem alike.
4
ACKNOWLEDGMENTS
I would like to acknowledge my funding sources for this project – USDA Western SARE
GW23-251, USDA NIFA A1451, NSF RaMP 1939511, and the University of Oregon’s Department of
Environmental Studies – for helping purchase materials, conduct research, supplying funds for tuition and
pay, and providing a platform for developing relationships with hazelnut farmers, researchers, and
community members.
I would also like to acknowledge my lab PI, Lauren Hallett, for supporting me throughout the
entire process of my undergraduate and post baccalaureate program, masters program, and going into the
future. Lauren has been enormously influential on my personal and career development, and I cannot
thank her enough for the support and mentorship she has provided me. She is inspiring, tenacious, astute,
and so incredibly kind. It has been an amazing opportunity and an honor to be in the Hallett Lab.
I would like to acknowledge the Hallett Lab as a whole for providing feedback and support for
every milestone of my undergraduate and masters thesis. Without your support, I would not be here! I
would also like to specifically thank Alejandro Brambila, Ari Brown, Calvin Penkauskas, Carmen
Watkins, Paul Reed, Jazz Dhillon, Eliza Hernandez, Ethan Torres, Jake Swanson, Jasmin Albert, Lina
Aoyama Batas, Rachael Dennis, Steven Haring, and Ellie Ingraham. Ellie Ingraham provided a great deal
of help with the set up, labor, and support of the past years project. Thank you Ellie!
I would also like to acknowledge my community partners, hazelnut farmers, and researchers who
have supported me throughout this process. Thank you to Amy Bartow and Ian Silvernail at the Corvallis
Plant Materials Center, Betsey Miller at Oregon State University, Brian and Heather Cottings, Emily
Huckstead at University of Oregon, Ethan Schra, Hazelnut Growers of Oregon, Jason Kelley with
Asperatus Consulting, Judith Massee, Lauren Ponisio at University of Oregon, Lucas Silva at University
of Oregon, Lynda Boyer at Heritage Seedlings and Liners, Lynnea Lane and Paul Massee, Oregon
Organic Hazelnut Collective, Patrick Kingston and Jason at Helena Agri-Enterprises, Peter McGhee at
5
Pacific Biocontrol, Raina Wickstrom, Rebecca Sweet at Buzz Cover Crop Seeds, Salmon Safe, Tom Kaye
with the Institute for Applied Ecology, Vân Trần, and Wayne Steenson.
I would also like to make a very special thank you to my family who have tirelessly loved me and
supported my research in every imaginable way. From letting us set up the initial pilot project on the farm
back in 2019 (and expanding the project to twelve acres as of last year), driving with me to complete tasks
I need to do for research, helping me collect data in the pouring rain, double checking my math, to
inspiring me to try out new ways to achieve goals and carry out tasks. Thank you so much, I can’t thank
you enough!
6
TABLE OF CONTENTS
I. LIST OF FIGURES………………………………………………………………8
II. LIST OF TABLES………………………………………………………………..9
III. INTRODUCTION………………………………………………………………..10
IV. HYPOTHESES & EXPECTED RESULTS…………………………………….. 12
Community Composition & Phenology………………………………………………………… 12
Harvest Nut Entrapment………………………………………………………………………… 13
Input Treatments………………………………………………………………………………… 13
Soil Moisture & Temperature Monitoring………………………………………………………. 14
V. METHODS……………………………………………………………………….15
Study Area……………………………………………………………………………………….. 15
Study Design…………………………………………………………………………………….. 15
Site Preparation & Management………………………………………………………………… 16
Cover Crop Selection……………………………………………………………………………. 17
Sampling Design………………………………………………………………………………… 17
Community Composition & Weed Exclusion…………………………………………... 17
Phenology………………………………………………………………………………..18
Harvest Nut Entrapment…………………………………………………………………18
Input Treatments…………………………………………………………………………18
Soil Moisture & Temperature……………………………………………………………19
VI. DATA ORGANIZATION & ANALYSIS………………………………………..20
VII. RESULTS……………………………………………………………………….. 21
Community Composition & Weed Exclusion……………………………………………………21
Phenology………………………………………………………………………………………...22
Harvest Nut Entrapment………………………………………………………………………….24
Input Treatments………………………………………………………………………………….25
Soil Moisture & Temperature Monitoring………………………………………………………. 27
VIII. DISCUSSION…………………………………………………………………… 30
Community Composition & Weed Exclusion……………………………………………………30
Phenology……………………………………………………………………………………….. 31
Harvest Nut Entrapment………………………………………………………………………… 33
Input Treatments………………………………………………………………………………… 33
Soil Moisture & Temperature Monitoring………………………………………………………. 35
Other Considerations……………………………………………………………………………. 37
IX. CONCLUSION…………………………………………………………………..39
X. CONFLICT OF INTEREST STATEMENT……………………………………..40
XI. APPENDIX………………………………………………………………………41
XII. LITERATURE CITED…………………………………………………………...47
7
I. LIST OF FIGURES
I. Mean Species Cover (%) by Cover Type………………………………………………...21
II. Phenologic Stages of Individual Species……………………………………………….. 23
III. Crop Loss by Cover Type and Year...................................................................................24
IV. Orchard Floor Treatments by Cover Type……………………………………………….25
V. Soil Water Content (mm) at Increasing Soil Depths……………………………………. 27
VI. Temperature (℃) at Increasing Soil Profile Depths……………………………………. 29
VII. Site Design……………………………………………………………………………… 41
VIII. Cover Crop Filtering Methodology (Physical)…………………………………………..42
IX. Cover Crop Filtering Methodology (Temporal)................................................................43
X. Target Species Cover in Community Plots………………………………………………44
XI. Crop Loss by Cover Type and Implement……………………………………………….45
XII. Conventional vs Conservation Cover Management Cycle………………………………47
8
II. LIST OF TABLES
I. Forecasted Operational Cost for Cover Type Management…………………………….26
II. Mean Soil Moisture Values at Various Depths………………………………………… 28
III. Mean Soil Temperature Values at Various Depths…………………………………….. 29
IV. Selected Seed Mix Species and Associated Target Traits………………………………43
V. Individual Species’ and Weed Plot P-Values by Year…………………………………..44
9
III. INTRODUCTION
Increasing plant cover in intensified agricultural systems is important for soil health, below- and
above-ground productivity, and the overall integrity of an ecosystem (Davis et al., 2023; Cárceles
Rodríguez et al., 2022; Wittwer et al., 2017). This is true for many nut orchards in the western U.S., as
many of these systems utilize bare-ground management to facilitate mechanical nut-pickup during
harvest. In Oregon hazelnut systems, this bare-ground management is typically achieved through various
quantities of herbicide application, flailing (mowing), and scraping (leveling, dragging, or floating)
treatments. This prolonged management style can result in water contamination, degraded soils (e.g.,
erosion, soil compaction, reduced biological activity), and loss of understory biodiversity, which can
ultimately lead to reduced ecosystem services (e.g., water infiltration, nutrient cycling) in orchards and
the surrounding landscape (Babcock & Kim, 2020).
Eliminating this need for a bare-soil management regime comes with three steps: (1) reducing the
weed seed bank established in orchards, (2) reducing mechanical and chemical inputs, and in place, (3)
filling this empty understory space with desirable, low-maintenance, compatible vegetation communities
that can complement an array of management strategies used by the majority of Oregon hazelnut farmers.
My goal is to harness my knowledge of the system – both historically and present – to develop cover
crops that are compatible with day-to-day orchard operations while also providing long-term
sustainability and enhanced ecosystem services benefiting the farmer, orchard, and surrounding
agroecosystem.
Maintaining cover in bare-ground managed orchards is important for soil health, but finding
solutions that meet the needs of orchard management can be difficult. There are a plethora of cover crop
types available on the market, though there are pros and cons to implementing any type in the given
system. Many farmers worry about the effects cover crops will have on orchard weeds, soil moisture, nut
pickup, and the cost to manage (Haring et al., in review). Native conservation cover (NCC),
self-sustaining perennial cover crops that are composed of native plant species, may be a good solution
10
for overcoming the precedent of bare-ground management in orchard systems and address these concerns.
Native plants may be especially utilitarian in Mediterranean climates where herbaceous vegetation grows
in the winter, spring, and early summer, but senesces in the late summer when harvest preparation is
needed. In such ecosystems, like western U.S. prairies, native species are adapted to disturbance, and our
past work has shown that these species are a viable alternative to many cover crop types currently used.
These native species can flourish under farm operations and may enhance certain ecosystem services
desirable to hazelnut farmers (Brambila et al., in review). If this type of cover crop can be a viable
alternative in commercial-scale hazelnut orchards, it could be worth investing in for many farmers.
Species selection is a key component to refining any type of cover mix for nut orchards. There are
key criteria for a species to be viable– they must be adapted to shade conditions on the farm, they must
not interfere with farm operations (such as nut pickup and pest management), they must not harm trees
(such as competing for water). NCC may fill these gaps due to native species’ adaptations to the
bioregion. But ideally, these NCC mixes must also enhance on-farm operations – for example, by
suppressing weeds and moderating soil temperatures. Suppressing weeds may be a very compelling
service offered by these mixes, as an effective conservation cover mix would both reduce mechanical and
chemical inputs on the farm, helping the farmer reduce costs associated with labor and expendable inputs
like herbicides and diesel, while also facilitating nut pickup during harvest.
Species mix development is a tandem consideration when understanding their combined effect on
orchard health and management. For example, many studies in ecology show that diverse plant
communities better resist weed invasion by filling all available ecological niches (Halassy et al., 2023).
These mixes with greater diversity may be better suited for suppressing weeds, however, these mixtures
may also have the capacity to better access all resources, potentially reducing soil water content for
hazelnut growth and crop production. Whether this affects water available to the trees in an open
question, however, as trees are tapping into a much deeper soil profile. Here I describe key questions and
my approach:
11
IV. HYPOTHESES & EXPECTED RESULTS
Community Composition & Phenology
My first research objective was to develop a self-sustaining, NCC seed mix for a mature hazelnut
orchard and evaluate compatibility with orchard management goals. These orchard management goals
included compatibility with pesticide treatments and competition with agricultural weeds. Due to cover
composition affecting multiple factors in weed exclusion and orchard management function, I tested the
individual species that the seed mix was composed of and the community cover mix as a whole, as
interspecific competition for resources vs intraspecific competition may result in different varying levels
of performance and desirability for each management category. Community mixes may inhibit weed
survival and reproduction more than individual species alone due to diversified niche occupation and the
associated invasion resistance (Halassy et al., 2023). I hypothesize that (H1.1) NCC communities could
suppress weeds effectively enough to see a decline in plot-level weed cover between year one and year
two. I also expected, due to diversified niche occupation, diversified growth characteristics, and
diversified phenologies, (H1.2) cover community plots would suppress weeds greater than individual
species plots. Because each species has various resource acquisition and diversified niches, I hypothesized
that (H2.1) certain individual species may suppress weeds better than others. (H2.2) In the first year, I
expected to find the annual species, that have similar growth characteristics and phenology to weeds and
would suppress weeds better than other individual species. Whereas in year two, I expected that the
perennial species would suppress weeds more after a year to establish and develop. Because selected
cover species must dually suppress weeds and match key points in the orchard management timeline, it is
important to evaluate points of interference. If the cover is flowering at these given points, an ecological
trap can occur (Hale & Swearer, 2016) Thus, (H3.1) I expected that species that are in vegetative stages
(or flower in a condensed time period) would match orchard management timelines greater in accordance
to sucker spraying. I expected that species that could produce viable seed prior to (H3.2) filbertworm
12
moth spraying (which requires flailing beforehand to avoid an ecological trap for pollinators and other
beneficial insects) would successfully reseed for the following year.
Harvest Nut Entrapment
My second research objective was to ensure that native cover did not entrap nuts in plant residue
or plant regrowth during harvest. Crop pickup is very important to the farmer and it is imperative to
design a cover crop that will not interfere with this process. Selecting species based on a variety of
factors, including basal rosette shape, leaf structure, and regrowth potential is important when designing
seed mixes. My overarching question was whether NCC mixes can be designed, based on vegetative
characteristics, to be compatible with standard hazelnut orchard harvesting techniques. When assessing
this, it is also important to consider differences in hazelnut pickup and vegetative regrowth in dry vs wet
harvest years. (H4.1) I hypothesized that dry harvest years will have equal nut pickup comparing cover
plots to control plots and (H4.2) wet harvest years to have slightly less nut pickup in cover plots than in
control plots due to increased plant growth from pre-harvest rains.
Input Treatments
My third research objective was to understand if NCC could successfully reduce on-farm
mechanical and chemical inputs compared to conventional orchard management practices (i.e.
bare-ground). To evaluate this, I selected three management strategies commonly used by hazelnut
farmers to measure general input quantities in the form of treatments. Reduction in these three inputs
generally means less time, money, labor, diesel, and equipment wear over time. Reduction in these inputs
can contribute to a healthier ecosystem and a more sustainable land management regime. I asked if
conservation cover can cumulatively reduce the need for broadcast herbicide applications, flailing passes,
and scraping passes over a two-year period. Broadcast herbicide applications could be reduced by having
a significant amount of conservation cover take the place of weeds, flailing could be reduced by needing
to flail weeds less, and scraping could be reduced by the lack of rut creation from herbicide application
13
during spring. (H5.1) I hypothesized that NCC would reduce the need for broadcast herbicide sprays
compared to conventional orchard management practices. (H5.2) I also hypothesized that NCC would
reduce the cumulative need for flailing passes compared to conventional orchard management practices.
Lastly, I expected that (H5.3) NCC would fully reduce the need for scraping passes compared to
conventional orchard management practices. To compare these input reductions over time, I elaborated on
an expected input cost for implementing and managing each cover type. I hypothesized that (H6.1)
conservation cover would reduce overall ground management costs compared to conventional
management, but (H6.2) is costlier to implement than conventional management.
Soil Moisture & Temperature Monitoring
My fourth research objective was to evaluate NCC’s effects on soil water content and soil
temperature by measuring soil moisture and temperature at multiple depths, with and without cover. My
overarching questions were to ask whether conservation cover significantly reduces soil water content,
and what depths are most affected by cover drawdown, and if conservation cover reduces overall soil
temperature, and at which depths are most affected by cover shading. Soil moisture and temperature are
important factors for regulating biological and chemical soil processes on various scales (Niu et al., 2024;
Dai et al., 2020). Having ample moisture available to trees is important for crop production (Tombesi &
Rosati, 1997) and temperature is an important factor influencing soil water content and plant growth
(ScienceDirect, 2016). Though it would be hard to directly test whether cover competed with hazelnut
trees for plant-available water, some inferences could be made based on assumptions at varying depths. I
hypothesized that (H7.1) cover communities would have a negligible to slightly reduced effect on soil
moisture compared with conventionally managed soils in the summer. Soil temperature would likely
decrease as depth increases due to insulation from the soil profile rather than plants. As depth increases,
soil temperature increasingly lags behind daily and seasonal above ground temperatures, though the
topmost profile is still generally influenced by plant cover (ScienceDirect, 2016). (H7.2) I expect that
conservation cover would have a slight positive effect on soil temperature at every depth.
14
V. METHODS
Study Area
This study was conducted in the Willamette Valley on a hazelnut farm near Gervais, Oregon. The
orchard consists largely of older-stand (~60 years) Corylus avellana var. ‘Barcelona’ (hazelnut) trees
intermixed with C. avellana var. ‘Jefferson’ and C. avellana var. ‘McDonald’ planted on a 6.1m x 6.1m
grid. Over the course of the study, the orchard went through various levels of canopy structure
transformation due to tree replacement and pruning of Anisogramma anomala (Eastern filbert blight).
Special carefulness and consideration weren taken for plot preservation, though greater disturbance was
seen in year two compared to year one of the study. Compared to other orchards in the region, this orchard
experienced higher magnitudes of winter and spring disturbance due to this prolonged replacement and
pruning of Eastern filbert blight.
Over the course of the two years that the study took place, temperatures at the study site varied
from -6°C to 40°C and the site received 1,002.03 mm of precipitation in 2022 and 1,037.08 mm in 2023
(Weatherlink, Davis Instruments). The soil type of the study site is Chehalis silty clay loam.
Study Design
Following the hazelnut harvest in 2021, a four-acre block of plots was established in the
northwest corner of the orchard. Plots were established in a checker block pattern, with each plot being
separated on every side by a ‘management row’ for ease of access to orchard equipment. Management
rows were managed in accordance with conventional orchard management practices. Out of eighty 6.1m x
6.1m meter plots, twenty plots consisted of a ‘community cover’ seed mix blend, twenty control plots,
and forty ‘individual species’ plots, with five plots of the eight individual species from the seed mix. Plots
were randomized inside of the four-acre block and only established if the plot had 3/4 hazelnut trees of
mature status. Each plot corner was marked by a hazelnut tree (appendix figure 1).
15
Community cover plots were used to assess community composition, weed suppression, soil
moisture / temperature, and nut entanglement data. Control plots were used to assess soil moisture /
temperature and nut entanglement data. Individual seed mix plots were used to assess phenology, and
weed suppression. The entire NCC block of plots was used to assess chemical and mechanical input data.
A control block was set up adjacent to the large block of plots to be used to assess chemical and
mechanical input data. The control block and control plots were managed in accordance to conventional
orchard management practices, with exception to spraying for Cydia latiferreana (filbertworm moth) in
control plots, which management was instead accomplished through Isomate pheromone dispensers
(Miller et al., 2018).
Site Preparation & Management
After the hazelnut harvest in 2021, I prepared the site for the experiment by removing any leaves
from the area with a Flory sweeper. I then broadcast sprayed glyphosate over the four-acre block with a
6% concentration to create a baseline vegetation load, utilizing recommended practices from
oak-savanna/woodland restoration ecology methodology (Boyer, 2013). I waited fourteen days for the
vegetation to die off and the herbicides to break down before seeding. After removing any additional
fallen leaves using a leaf blower, the seeds were spread at a rate of 10g/37.16 m2 (being the size of one
plot) using a hand-held broadcast seed spreader, with the seed ratios being calculated by seed weight and
mature plant size (Xerces Society and Cruz, 2020).
Mechanical treatments started in January and consisted of winter shredding (mulching pruned
branches and winter leaf cover), summer flailing, and scraping (if the farmer determined it was a needed
process). After these processes took place, a final flailing pass was performed to crush up any blank nuts
that had fallen prior to the majority of nut drop. A month later, after most nuts had fallen, the project sites
were harvested using the farmer’s personal harvesting equipment (supplement 12).
After harvest, each cover type was rested during the winter, letting the NCC regrow from roots /
germinate, and letting the weeds regrow in the conventionally managed plots. In year two, the cover crops
16
filled into the management rows through mechanical dispersal as management rows were deemed
unnecessary for workflow.
Cover Crop Selection
Individual species were selected for their functional compatibility traits to hazelnut life-cycle
(supplemental figures 2 and 3, supplemental table 4), disturbance tolerance, competitiveness, plant height,
price, and commercial availability. Species were actively removed from selection if they had an extended
flowering period (to prevent life-cycle mismatch with hazelnut management), had very robust basal
leaves or were bunching (to prevent harvest entanglement), or did not meet the required habitat
characteristics for growth in hazelnut orchards (i.e. disturbance tolerant, shade tolerant, soil type adaptive,
etc.). The perennial (P) species selected were Camassia leichtlinii, Geum macrophyllum, Lomatium
dissectum, and Sidalcea campestris. The biennial (B) species selected were Phacelia nemoralis and
Prunella vulgaris. The annual (A) species selected were Collinsia grandiflora and Collomia grandiflora.
All species selected were local biotypes purchased from Heritage Seedlings & Liners in Salem, Oregon.
Sampling Design
Community Composition and Weed Exclusion
Vegetation cover was surveyed multiple times throughout the spring and summer months,
averaging one survey every two weeks. Data was collected by conducting ocular cover percentage
estimates across entire plots. Mean cover was evaluated in community cover and individual species plots
for target species, weeds (in general), litter cover, and bare ground. During data analysis, peak vegetation
data (when plants no longer grow fuller) from June was used to compare 2022 and 2023 results of both
community cover and individual species community composition. Target species cover in community
cover plots is the mean sum of every seeded plant species. The mean target cover in individual species
plots is also the mean sum of that particular plant species. The mean weed cover is a sum of all species
cover that was not seeded into the plot, including native and non-native species.
17
Phenology
Phenology was assessed by visiting each individual species plot and determining what
percentage of the population was at which given life stage: germination, vegetation, flowering, seed
development, seed maturation, and senescence/death/dormancy. Phenology data was surveyed multiple
times throughout the spring and summer months, averaging one survey every two weeks.
Harvest Nut Entrapment
In 2022, nut entrapment was surveyed in ten community composition plots and ten control plots.
A 1m2 quadrat was set in the middle of the plot where the harvester collected nuts from two hours prior.
Inside each quadrat, quality nuts entangled in vegetation or plugged into the soil were collected. Blank
and shrivel nuts, nuts attached to husks, and older nuts from previous harvests were removed due to their
likelihood of being ejected from the harvester based on weight or surface area. In 2023, the same methods
were used, though the survey was expanded to include a sweeper survey. In this survey, a 1m2quadrat was
laid in the middle of the tree row, on the North side, where the sweeper paddles extend to, and nuts were
collected based on the criteria listed above. In 2023, harvester data was also collected, but expanded to
include twenty seed mix plots and 20 control plots.
Input Treatments
Mechanical and chemical treatment inputs were tracked in accordance with how many iterations
were performed over the course of the year to the given cover type. Herbicide sprays were accounted for
if either cover type received a broadcast herbicide spray. Spot-spraying weeds or hazelnut tree suckers did
not count as a broadcast herbicide application. Flailing and scraping passes were accounted for if the
cover type received an entire pass, one pass being defined on an axis (i.e. North to South, up and down
the row). All three treatment types were determined effective by the land managers, ensuring that the
minimum cumulative treatments in each treatment type met the overall expectations for harvest
preparation regardless of cover type. Economic data was elaborated and based on the cost to perform each
18
activity, accounting for input costs of herbicides and labor performing each activity. Herbicide cost was
based on the average concentration of glyphosate used to spray each cover type (3 quarts per acre) and
current market value (~$8/quart). Labor cost was assessed using current the minimum wage for rural
areas in counties in the Willamette Valley ($14.20/hr) (Oregon Bureau of Labor and Industries, n.d.). Fuel
usage was not accounted for.
Soil Moisture & Temperature
Twelve semi-permanent soil moisture sensors (Sentek Drill & Drop) were installed in six
community composition plots and six control plots. The top of each probe was set 5cm below the soil
surface to prevent probe destruction from mechanical treatments. Each probe measured temperature and
soil moisture at the following depths: 5cm, 15cm, 25cm, 35cm, 45cm, and 55cm, respectively adding 5cm
to each depth to account for probe set depth. The probes automatically recorded temperature and moisture
data points every 30 minutes from mid-May until flailing in mid-July. Data was stored internally, with
data uploads to the server site occurring every two weeks.
19
VI. DATA ORGANIZATION & ANALYSIS
I used ‘R’ version 4.3.3 to process and manipulate all raw data, visualize all figures, and calculate
all statistics. To test Native Conservation Cover’s ability to perform in the orchard system, I fit a linear
mixed model with cover type as an explanatory variable. I compared data from each category of the
questions section: community composition, phenology, soil moisture and temperature, nut entrapment,
and input treatments. Each category is a respective response variable aside from input treatment
evaluation, which was not fit to a model.
For the community composition data set, I created two models to evaluate success of target cover
compared to weed cover performance. First, I evaluated the abundance of target species (individual target
species, community mix target cover, and control) as the explanatory variables. Secondly, I compared the
abundance of weed cover with the same explanatory variables (individual target species, community mix
target cover, and control). This design showed whether certain target species or the community mix
performed better than the other, and which target species or community mix excluded weeds more than
the others. When comparing target cover to weed cover for both individual species and the community
mix, I calculated statistics using Welch’s two-sample unpaired t-tests. When calculating the mean change
of target cover or weed cover year-to-year, I used Welch’s two-sample paired t-tests.
To test nut entrapment, I compared conventional management (bare-ground) to NCC. I calculated
these statistics using a Welch’s two-sample unpaired t-tests to compare mean nut loss across cover type by
year. I did not compare year-to-year data by cover type due to environmental differences.
To test soil moisture and temperature, I calculated my statistics using a representative interval of
the last ten days of data collection (7/7/23 - 7/17/23). I compared conventional management to
conservation cover for both data sets, using a two-way ANOVA and then a post-hoc analysis with a Tukey
HSD.
Statistics were not calculated for phenology and input treatments due to the nature of the data
sets.
20
VI. RESULTS
Community Composition and Weed Exclusion
Figure 1 - Mean percent cover of all plots types comparing target cover to weed cover during 2022 and
2023. The top left panel represents mean conservation cover across community plots. Every other panel
represents individual species plots comparing target cover to weed cover. Bare ground cover and litter
cover were observed but not included in the figure. A table of p-values can be found in the supplement
(table 5).
Conservation cover in 2022 averaged 82.35% across all twenty community plots, whereas weed
cover averaged 6.05%. In 2023, conservation cover increased to a mean of 88.10% (p = 0.003) and weed
cover decreased to a mean of 2.35% (p = <0.0001). These findings support my hypothesis (H1.1),
displaying that NCC communities can significantly suppress weeds year to year.
Cover in single-species plots varied from year to year, with a general trend (excluding L.
dissectum (P), P. nemoralis (B), and S. campestris (P)) of decreasing conservation cover and increasing
weed cover from 2022 to 2023. Mean L. dissectum cover significantly increased from 2022 to 2023 and
mean P. nemoralis (B) and S. campestris (P) cover or weed cover did not significantly change from year
21
to year. C. leichtlinii (P) and S. campestris (P) were planted as bulbs and plugs, respectively, thus they
should not be evaluated for weed competition in the same light. These findings support my hypothesis
(H1.2), displaying that community plots would suppress weeds greater than all other individual species
plots year to year. The only contender to as good of cover maintenance and weed suppression is likely
seen in P. nemoralis (B) plots, where the general trends are similar. This supports my (H2.1) hypotheses,
showing that some species, like G. macrophyllum (P) and P. nemoralis (B), may suppress weeds better
than other species, like Collinsia grandiflora (A) and P. vulgaris (B), when seeded by themselves. I
hypothesized that (H2.2) annual species would suppress weeds better than perennials in year one, and
perennials would suppress weeds better in year two, though there is no distinct trend across life habit, and
is entirely species reliant.
Phenology
Key management points for each phenologic stage were flowering and sucker spraying, and seed
maturation and filbertworm spraying. Management slightly differs between years based on orchard
phenology, but events are generally completed within two weeks each year. Results for this component
were evaluated categorically. While sucker management was more flexible and largely dependent on
farmer discretion (figure 2, blue triangle), filbertworm management was more of a hard deadline (figure
2, red circle), leaving little room for seed maturation to take place on a flexible timescale. In each year,
peak flowering phenology happened in between sucker management timepoints and filbertworm
management happened roughly about the same time that seed maturation for many species occurred.
Flowering phenology for all species categories was largely the same but is most likely this way due to
seed selection and filtering criteria. This supports my hypothesis (H3.1) that species that flower in a
condensed time period would be best suited for sucker spraying timeline match. In 2022, not as many
species went to seed, but seed maturation for annual species was effectively completed prior to
filbertworm management, ensuring that the farmer could flail prior to spraying (figure 2, gray dash).
22
Figure 2 - Phenological stages of individual species across both 2022 and 2023, broken up into three
primary stages: vegetation, flowering, and seed maturation. Individual plant species are colored by
annuals (magenta), biennials (yellow), and perennials (teal). Colored shapes indicate time of month in
which a management activity took place, with sucker management as a blue triangle and filbertworm
management as a red circle. The flailing deadline (gray dash) is also included to indicate when cover was
mowed down prior to filbertworm management. Characteristics such as germination, first flower, last
flower, and dormancy/death/senescence were evaluated but excluded from this visualization.
In 2023, annual species effectively went to seed well before this time point, perennial species cut it close
but completed seed maturation, and biennial species went to seed last. Some
biennial species may not be fit for this rigorous timeline, though given preliminary cover data from 2024,
it seems these species did not need to fully mature before being flailed down. This supports my
hypothesis (H3.2) suggesting that species that are able to go to seed prior to filbertworm management will
have success in reseeding for the following year.
23
Harvest Nut Entrapment
Figure 3 - Crop loss, counted by nuts leftover after harvest per square meter, by cover type and year.
Conventional management is in tan and conservation cover is in green. Data represented is collected in
the harvester row. Sweeper data can be found in supplement figure 7.
Nut entrapment data was collected in the fall of 2022 and 2023, with differing harvest conditions
each year. Crop loss in dry years (2022) was comparable, with no significant difference between cover
types. In a wet harvest year (2023), crop loss in conservation cover was significantly less than crop loss
under conventional management, with a mean reduction in crop loss of 1.5 nuts per m2 (p = 0.0204). This
supports my hypothesis that (H4.1) dry harvest years would have equal nut pickup comparing cover plots
to control plots, as results from nutpickup between the comparison groups were insignificant. I also
hypothesized that (H4.2) wet harvest years would have slightly less nut pickup in cover plots than in
control plots due to increased plant growth from pre-harvest rains. These findings do not support my
hypothesis, but instead shows that conservation cover can facilitate nut pickup in wet harvest years.
24
Input Treatments
Figure 4a- Mechanical and chemical orchard floor treatments by cover type. Treatments were measured
in the amount of time each activity was performed by cover type in both 2022 and 2023. In tan is
conventional management and in green is conservation cover.
In both years of data collection, conservation cover reduced the overall number of herbicide
applications, flail passes, and scraping passes. Herbicide applications were reduced by one in 2022 and
two in 2023. In 2022 and 2023, flailing passes were reduced by three in conservation cover compared to
conventional management. In both years, scraping in the conservation cover block was also reduced by
two passes compared to the conventional management block, with conservation cover negating the overall
need to scrape altogether.
I translated this data to a forecasted economic analysis comparing conservation cover and
conventional management side-by-side, evaluating cost to implement and manage a ten-acre block over a
five year period. For year one, the cumulative total and yearly total for each cover type is the same, with
conventional management estimated at $707.10 and conservation cover at $15,673. This high value for
conservation cover is due to the price of the seed mix, with an estimated input cost of $1,500 / acre. From
25
years two through five, input costs remain the same for conventional management, accruing an additional
$707.10 each year to a cumulative total of $3,535.5 at the end of the five-year period. Input costs for
conservation cover are forecasted to reduce in years two through five, accruing an additional $68.10 each
year to a cumulative cost of $15,945.4 at the end of the five-year period. This data supports my
hypotheses (H6.1) suggesting that conservation cover will reduce overall management input costs, but
(H6.2) be costlier to implement.
Year 1 Year 2 Year 3 Year 4 Year 5
CM CC CM CC CM CC CM CC CM CC
Seed Cost $0 $15,000 $0 $0 $0 $0 $0 $0 $0 $0
Herbicide $56.80 $28.40 $56.80 $0 $56.80 $0 $56.80 $0 $56.80 $0Labor
Herbicide $480 $240 $480 $0 $480 $0 $480 $0 $480 $0Cost
Flailing $113.50 $68.10 $113.50 $68.10 $113.50 $68.10 $113.50 $68.10 $113.50Labor $68.10
Scraping $56.80 $0 $56.80 $0 $56.80 $0 $56.80Labor $0
$56.80 $0
Yearly Total $707.10 $336.50 $707.10 $68.10 $707.10 $68.10 $707.10 $68.10 $707.10 $68.10
Cumulative $707.10 $15,673 $1,141.2 $15,741.1 $2,121.3
Total $15,809.2
$2,828.4 $15,877.3 $3,535.5 $15,945.4
Table 4b - Forecasted operational cost for cover type management over a five-year period. Conventional
management (CM) is displayed in tan columns and conservation cover (CC) is displayed in green
columns. Cost estimate accounts for seed cost, broadcast herbicide application labor, flailing labor,
scraping labor, and the cost of general herbicide inputs on a ten-acre scale. Unaccounted for are fuel
costs and the potential for price inflation for each activity.
26
Soil Moisture & Temperature Monitoring
Figure 5a - 2023 mean soil water content (mm) data at different depths in the soil profile starting at 5cm
and progressing in increments of 10cm. Conservation cover is displayed in green and conventional
management is displayed in tan.
Mean soil water content progressively decreased as months went by, with a greater difference in
soil water content between cover types occurring at shallower depths in the soil profile. At any given
depth, soil water content was significantly impacted by cover type (p = 0.000988). This data supports my
hypothesis (H7.1) that cover communities would have a negligible to slightly reduced effect on soil
moisture compared with conventionally managed soils in the summer. Depth also impacted soil moisture
significantly (p = 0.000603), though there is no significant interaction between the two (p = 0.635432)
(table 5b).
Temperature in the soil profile steadily increased over time regardless of cover type, with more
variance happening at shallower depths compared to deeper depths of the soil profile. At any given depth,
soil temperature was significantly impacted by cover, with conservation cover plots lowering soil
27
temperature overall (p = 8.41e-11). As depth increased, mean temperature differences between cover
types also decreased significantly (p = 2.90e-08), though there is not a significant interaction between
cover type and depth (p = 0.953). This data does not support my hypothesis (H7.2) that conservation
cover would slightly cool soil temperatures at every depth. Conservation cover significantly decreased
soil temperature at every depth at every temporal period represented.
Cover Type + Depth
(cm) Difference Lower / Upper P-Value
NC / CM: 5cm -2.9049805 -11.23 / 5.42 0.9839281
NC / CM: 15cm -1.9982243 -10.32 / 6.33 0.9993291
NC / CM: 25cm -2.1842963 -10.51 / 6.14 0.9984949
NC / CM: 35cm -3.2012400 -11.53 / 5.12 0.9672628
NC / CM: 45cm -3.2870498 -11.61 / 5.04 0.9606539
NC / CM: 55cm -7.3890852 -15.71 / 0.94 0.1227823
Table 5b - Soil moisture statistics comparing differences in soil moisture content at various depths across
cover type. Data is pulled from the representative interval of declining soil moisture, from July 7th, 2023
to July 17th, 2023. Native Conservation Cover is displayed first as NC and conventional management is
displayed second as CM.
28
Figure 6a - 2023 mean soil temperature data (°C) at differing depths in the soil profile starting at 5cm
and progressing in increments of 10cm. Conservation cover is displayed in green and conventional
management is displayed in tan.
Cover Type + Depth
(cm) Difference Lower / Upper P-Value
NC / CM: 5cm -2.0445 -3.86 / -0.22 0.0168042
NC / CM: 15cm -2.0465 -3.86 / -0.22 0.0166350
NC / CM: 25cm -2.1425 -3.96 / -0.32 0.0101595
NC / CM: 35cm -2.0513 -3.87 / -0.23 0.0162326
NC / CM: 45cm -1.7666 -3.58 / 0.05 0.0637745
NC / CM: 55cm -1.5033 -3.32 / 0.31 0.1897940
Table 6b - Soil temperature statistics comparing differences in soil temperature (°C) at various depths
across cover type. Data is pulled from the representative interval of declining soil moisture, from July 7th,
2023 to July 17th, 2023. Native Conservation Cover is displayed first as NC and conventional
management is displayed second as CM.
29
VII. DISCUSSION
Community Composition & Weed Exclusion
The NCC community selected for this study effectively demonstrated an ability to suppress weeds over a
two year period (figure 1). Inside of these community plots, Collomia grandiflora cover dominated in
both years, with varying levels of cover from all of the other species (supplementary figure 6). In both
community cover plots and individual species plots, Collomia grandiflora suppressed weeds effectively,
though mean Collomia grandiflora cover in individual species plots declined in year two (figure 1). Other
individual species that competed well with weeds, in both years, were Collinsia grandiflora, G.
macrophyllum, P. nemoralis, and P. vulgaris (figure 1). Collinsia grandiflora may have had an individual
advantage due to its early germination start, which usually takes place in the early fall, and like Collomia
grandiflora, is an annual that tends to be somewhat weedy. G. macrophyllum and P. nemoralis may have
had individual advantages due to their basal leaves forming rosette patterns, which compete well for
space. With P. vulgaris acting as a biennial “boom or bust” species in this system, this species may have
solely done well in both years due to this nature.
It is likely that this native plant community can suppress weeds much greater than individual
native plant species alone (figure 1) due to diversified niche occupation (Holmes et al., 2017), though
additional evidence suggests that biomass production plays a key role in weed suppression (MacLaren et
al., 2019). Many of the species selected for the tested seed mix had multiple traits desirable in competition
against weeds, having characteristics such as rhizomes, tap roots, prolonged seed dormancy and
associated germination strategies, seed appendages and coatings, etc. (Emery 2021; Russell 2011;
Corvallis Plant Materials Center, n.d.).
While both annual species and most of the perennial species did well in competition with weeds
in individual species plots, some decline in individual species plot mean cover may be due to mechanical
seed dispersal (from flailing, sweeping, and harvesting) to outside of the plot area. Notably, the spread of
Collomia grandiflora, P. nemoralis, and P. vulgaris was wide, with the largest span of distance between a
30
plot and outside area to be observed at 12m, with 6m spread being regularly observed. This may be due to
seed shape and weight, as all three species have relatively heavy, rounded seeds that may have the ability
to disperse better via equipment ejection during harvest (LaForgia (in preparation)). Spread of G.
macrophyllum was also observed widely, and this may be due to seed shape and weight, and/or due to
seed burrs. The ‘weedy’ quality of the highest performing species is very desirable in an orchard setting
as competition with weeds will likely be an ever persisting challenge due to the weed seed bank in
orchards, annual disturbance patterns (Brambila et al., 2022), and sink-source dynamics from neighboring
lands (Petit et al., 2012). Overall, diverse mixes of native plants that display persistent colonization habits
and have the ability to spread easily should be used in conservation cover mixes to achieve maximum
weed suppression and mean cover, as they compete better against weeds compared to individual species.
Phenology
When considering individual species to select for NCC seed mixes, it is important to consider the
phenology of each species in accordance with the mechanical and chemical treatments that are required
on-farm to prepare for a quality hazelnut harvest (supplementary figure 3). Two key management steps
that take place in order to satisfy a quality harvest are sucker management and filbertworm management,
which both often take form in a pesticide application. With regards to the phenology data, it is important
to note the time of management activities in relation to the associated stage of cover crop development.
For sucker spraying, it is important to avoid the flowering time period as much as possible so floral
resources for pollinators are not compromised. For this phenology stage, most individual species fall
within a safe window between sucker spraying (figure 2). G. macrophyllum and P. vulgaris first flowers
come close to this window in differing years, but would likely not inhibit a safe sucker spraying window.
Given that many farmers spray their orchard suckers differently, this inherently has pros and cons to how
timing of sucker spray happens and whether or not farmers feel comfortable with the seed mix phenology.
If farmers have the correct equipment for spot spraying suckers, and are more comfortable taking sucker
spraying at a slower and more careful pace, suckers may be able to be sprayed when cover crops are
31
coming into peak flower or exiting peak flower, as most species do not grow directly at the base of trees
(within a 0.5 m radius of the trunk) due to the increased shade, canopy drip during rain events, and lesser
amount of available rooting space. Some farmers strip spray the tree line to burn back suckers and kill
vegetation in the tree line, which if also using NCC, would be an important time period to avoid for
flowering as this could end up directly coating flowering plants with herbicides, potentially creating an
ecological trap for pollinators (Hale & Swearer, 2016). Increasing research into non-suckering trees,
diversified seed mix phenologies (so peak bloom does not happen at once), diversified seed mix
pollination strategies (wind pollinated, etc.), and new sucker management technologies will help in this
regard.
For filbert worm management, most farmers spray one or more insecticides to mitigate life-cycle
completion for the filbertworm (Wiman et al., 2014). In regards to phenology, and if a farmer is using
insecticides to manage filbertworms, it is important that species produce viable seed prior to spraying for
filbertworm (~early to mid-July) so the cover can be flailed down in time to spray for filbertworm but
avoid creating an ecological trap for pollinators and beneficial species (figure 2). Every individual species
selected for the seed mix develops its seeds to either the hard-dough or mature stage (Young & Young,
1986) by the time flailing needs to happen in preparation for filbertworm insecticide application. Species
that have seeds reaching maturation before flailing are C. leichtlinii, Collinsia grandiflora, Collomia
grandiflora, G. macrophyllum, and partially S. campestris. Species that have seeds reaching hard-dough
stage prior to flailing are P. vulgaris, P. nemoralis, and partially S. campestris. Seeds only reaching
hard-dough stage by the flailing timeframe likely mature on the ground after being separated from the
desiccated plant material (Young & Young, 1986). L. dissectum did not flower in 2022 or 2023, but will
likely reach a hard-dough stage by the time flailing occurs as well.
In this study, the plot area was not sprayed with any insecticides, but instead used isomate
filbertworm rings (Miller et al., 2018) to hang in the trees to prevent filbertworm mating from occurring. I
managed filbertworm this way in order to minimize any possible disruption to insect life present in the
cover cropped area. The ismoate filbertworm rings effectively deterred filbertworm mating inside of the
32
cover cropped area, with threshold traps recording zero filbertworm catches over the two years of the
study, compared to zero in the control in 2022 and one in the control in 2023. The enhanced trophic-web
in the cover cropped area may have also increased filbertworm predation, combining in effect with the
isomate rings to disrupt mating efforts. Some species included in the seed mix may have acted as host
plants for parasitoid wasps (Russell, 2015; Stephens et al., 2006) that prey on filbertworm, especially
members of the Apiaceae (L. dissectum), Boraginaceae (P. nemoralis), and Rosaceae (G. macrophyllum).
Harvest Nut Entrapment
One of the biggest factors playing into the use of cover crops in Oregon hazelnut orchards is the
entrapment of hazelnuts in cover crops when mechanically harvesting. With conventional orchard
management, the orchard floor is typically bare from weeds and debris, allowing the harvester to pass
through each row of trees to pick up the nuts easily. With conservation cover matching the phenology of
the hazelnuts rather well (figure 2), conservation cover can be flailed down just in time for mechanical
harvest (supplementary figure 3, supplementary figure 8). Mechanical harvest consists of sweeping the
nuts into windrows and then picking them up with the mechanical harvester. Through analysis of the data,
it is determined that the most amount of nuts are typically lost in the sweeping portion of the harvesting
process (supplementary figure 7). Though, this is significantly reduced when sweeping over conservation
cover. The harvester loses less nuts overall, but crop loss is also significantly reduced by conservation
cover in wet harvest years. Cover type has no significant impact on crop loss during dry harvest years
(figure 3), and this is likely due to conservation cover and conventional management cover types creating
similar orchard floor textures. In wet harvest years, crop loss may be significantly less in conservation
cover plots due to the improved root structure in the topsoil (Hudek et al., 2021), preventing nuts from
being plugged in the soil as seen in conventionally managed plots. In 2023 (wet harvest), some species in
the conservation cover mix regrew from their roots prior to harvest , though this seemed to have a
negligible impact on crop loss given these findings. Given this information, basal leaf structure is a very
33
important factor when designing NCC seed mixes, as certain perennials with fine leaf patterns or
bunching basal leaves may inhibit hazelnut pick-up (ex. Achillea millefolium) (Brambila et al., (in
review)).
Input Treatments
Harvest preparation typically happens over the duration of the spring and summer, and usually
consists of herbicide application, flailing, and scraping the orchard floor to get a smooth, vegetation-free
surface for mechanical harvest in the fall (supplementary figure 2, supplementary figure 3, supplementary
figure 8). Conventional orchard floor preparation usually consists of 1-2 broadcast herbicide applications
a year (or one burndown application mixed with a preemergent), 4-6 flail passes, and 1-3 scraping passes,
depending on the condition of the orchard floor due to debris load and ruts, and the desired floor texture
of the farmer. The earliest time frame for flailing is during initial soil drydown in the late spring, and the
latest time frame for flailing can be condensed right during blank drop in the late summer (Olsen &
Peachy, 2013). With such a large allowable time frame for harvest preparation to occur, it is important that
seed maturation has occurred prior to flailing so that the cover crop can self-seed for the following year.
Given that native seeds are so expensive (table 4b), self-seeding is the most economical way to prepare
for next year’s cover crop and hazelnut harvest. Building up a native seed bank in an orchard will help
suppress weeds long-term, support an ecological community that facilitates many ecosystem services, and
can help keep costs low year to year. One of the most acclaimed reasons for using cover crops is the lure
of reducing chemical and mechanical inputs in orchards, reducing the costs for labor, time, and
expendable products like diesel, pesticides and surfactants, flail teeth, and grease. Based on the findings
from this study, it can be concluded that conservation cover can reduce the need for broadcast herbicide
sprays after initial cover establishment. This initial broadcast application to reduce the weed seed bank
should not be dismissed, though, as it is imperative to the success and establishment of these native
plants. In this study, both mechanical inputs (flailing and scraping) were also reduced (figure 4a). Flailing
was reduced in part by reducing the need to mow down weeds, whereas scraping was fully reduced by the
34
lack of need to scrape out ruts in the soil from traversing equipment. Scraping may be reduced in part due
to less cumulative rutt-creating activity happening in the plot area from flailing and herbicide spraying,
and in part due to the conservation cover root structure preventing soil displacement (Hudek et al., 2021;
Balkcom et al., n.d.).
Conservation cover management also largely reduces overall orchard floor management costs
year to year compared to conventional orchard floor management (table 4b). While these preliminary
results show a great opportunity for economic sustainability, seed prices solely make conservation cover
implementation out of economic reach for many farmers. Generally, native seed prices can range
anywhere from $90 - $350 per pound, meaning that a NCC seed mix alone can outprice conventional
orchard floor management on a ten acre scale over a five year period. In order for these mixes to be
adopted by the majority of hazelnut farmers, price will need to be reduced. If a cost share program could
be implemented, this cover type may be very valuable for farmers for long term management.
A downfall of this dataset is that labor costs are largely based on tractor speed. Regardless of
tractor speed, the ratio between flailing passes and labor used should largely be the same. Herbicide
expense data is largely based on market price for pesticides, which fluctuates year to year. Fuel estimates
were not calculated into this analysis due to variations between tractor type, implement type, and fuel
consumption from farm to farm. Overall, this data set fits well within the range expressed amongst
Willamette Valley hazelnut farmers (Murray et al., 2022), though should be investigated further to parse
out variations in year-to-year and farm-to-farm data.
Soil Moisture & Temperature
Cover crops are notoriously criticized for their water consumption, as they tend to use more water
in perennial systems than perennial systems without cover (Kaspar 2006). Because studies are limited to
evaluating conventional cover crops and not native cover, it is important to understand what water usage
looks like at differing depths inside orchards and if cover actively competes for water with hazelnuts.
Given the findings from this study, NCC may have little effect on soil water content (table 5b). As depth
35
increased, soil water content also increased (figure 5a), but still without significant difference between
cover types. In terms of this data set, this was not statistically significant, though the number of probes
were limited to six per treatment, limiting the power of these statistics. A lot of literature discusses the
minimal water usage of native plants, referencing the adaptations to climate and environment these
species have developed over time (McMahan, 2018). For future research, it would be beneficial to have
side by side comparison of water usage in native cover to non-native cover to see if these adaptations
truly come into play in agricultural systems. In the context of understanding competition for water, this
study is unable to answer that question directly, but may give some insight to how this could be
approached in the future. With hazelnut rooting patterns consisting of roots mostly developing in the top
60cm of soil, depending on the soil profile and water table of an individual orchard, hazelnuts can have
roots reaching depths beyond the majority of the cover crop root system, reaching depths of 600 - 1000
cm (Olsen, 2014). Given these findings, competition for water might happen in the upper soil profile in
which juvenile trees are rooted, but likely does not see much direct competition in the lower soil profile
(>35cm) where mature trees have more roots and the conservation cover has less roots. Many of the
species in the seed mix have very small, fine roots, but some, like S. campestris and L. dissectum may
have roots reaching depths of 15 - 35cm (University of Washington, 2024). Roots may have impacted the
results seen in this study as some soil moisture probes had readings consisting of large errors that
rebounded quickly. NCC roots may have encapsulated probe nodes and removed plant-available water
from around the probes quicker than in the rest of the soil profile due to the increased pore space around
the probes from insertion. These large errors were removed, though the impacted probes consistently read
lower than non-affected probes.
In perennial systems, cover crops are also used to prevent evapotranspiration by shading the
ground (Blanco-Canqui et al., 2015). In younger hazelnut orchards, it is especially important to cover the
orchard floor to prevent problems associated with heated soils, like limited soil microbial activity, which
can lead to limited decomposition, nitrogen mineralization, root growth, etc. (ScienceDirect, 2016.). In
more mature orchards with a denser canopy, the orchard floor is decently shaded, though a cooler soil
36
temperature throughout the summer months can still improve conditions for biological activity, enhancing
all of these listed factors. In this study, even given the mature status and shading of the orchard floor,
temperature in all profile segments measured had significantly decreased temperatures (table 6b) and
stabilized amplitude fluctuations in soil when under NCC (figure 6a). In other hazelnut orchards with less
canopy cover, this temperature difference may be larger (Brambila et al., (in review)). If conservation
cover can stabilize soil temperature throughout the winter months as well, hazelnut root growth can
continue to be active when temperatures reach 4.5 ℃ or higher (Olsen, 2014).
Other Considerations
Some limitations to this study may be in the factors that couldn’t have been unaccounted for, such
as seed dispersal, mechanical seed scarification, pollinator-plant relationships, etc., but are hard to
determine and may well be worth researching in the future. Throughout the study, it became apparent that
seed dispersal in hazelnut systems likely happens largely due to mechanical treatments (flailing, scraping,
sweeping, harvesting, etc.). These treatments can spread any type of seed and could be used to their
advantage and/or disadvantage in this system. With these mechanical treatments, some may act to
unintentionally treat seed coats (scarification, etc.) to give them a germination advantage or accidentally
terminate the seeds. Species in the selected seed mix, like S. campestris, germinate better with
scarification (Corvallis Plant Materials Center USDA, n.d.) whereas other seeds in the seed mix do not
require any treatment (Emery, 2021). This could be changing the persistence of conservation cover and
weed cover and should be evaluated in the future to determine species’ seed responses to these
mechanical treatments. Another factor that could influence success of individual target cover and weed
competition is certain plant growth preferences, like the preference for growing in a monoculture or
polyculture. C. leichtlinii species are known for their performance in monocultures from their adaptations
to indigenous cultivation over time (Carney et al., 2021), so competition with weeds may not be
demonstrated when not tightly packed. In the hazelnut orchard context, C. leichtlinii did not compete well
with weeds in a sparse monoculture, but had an easier time in the polyculture (community plots). L.
37
dissectum did well as a monoculture, growing from year one to year two significantly, but did not do as
well in the polyculture. Collomia grandiflora did not do well in monocultures, significantly declining
from year one to year two, but dominated polyculture cover increasing from year one to year two (figure
2a), almost creating a monoculture in the polyculture itself. These phenomena should be more thoroughly
investigated as they could be very interesting to study and have a large effect on which plants are selected
for conservation cover seed mixes.
Another outside factor that may have implications on the success of NCC is habitat type and
connectivity to wildland habitat. These qualities may have an effect on wildflower seed development and
year-to-year cover performance, especially if a plant species has a specialist pollinator(s) that may not
have suitable habitat in close proximity to the hazelnut orchard. Another factor in pollination and seed
production success may be in the management of the cover itself. Mulching cover crops annually may
contribute to insect decline in intensified agricultural settings as habitat is significantly altered and
reduced every year (Ganser et al., 2019). Other studies indicate that even in these ecosystems, pollinator
populations increase over time (Schmied et al., 2022), thus more research should be done for this specific
system.
With this study being limited to the native forbs that are commercially available, it is important to
note the wide array of species that may be compatible with the management and function of these
orchards that are not yet commercially available. In the Willamette Valley, native seed production is
steadily growing, but restoration efforts far outpace the quantity and availability for practitioners,
researchers, tribes, and the public (International Network for Seed-Based Restoration, 2023). Many more
native species have desirable functional traits that would be optimal for use in differing types of orchards
for light/shade tolerance, differing management practices, soil quality, etc., but do not exist in bulk
quantities. Going forward, investing more research into which species are compatible with the majority of
orchards may be able to improve native seed production goals if NCC becomes widely adopted.
38
VIII. CONCLUSION
Native conservation cover (NCC) may be a great cover crop option for Oregon hazelnut farmers
to adopt as the practice aligns with key orchard management goals studied. From this research,
conservation cover communities demonstrated effective weed suppression capabilities and alignment with
key orchard management activities, like sucker spraying, filbertworm management, and the ability to
increase hazelnut pickup during wet harvests. Conservation cover also demonstrated the ability to reduce
chemical and mechanical inputs, likely saving farmers input costs, time, and energy spent controlling
orchard weeds and maintaining smooth ground for harvest. While NCC may reduce soil available water at
the soil surface and down to 35cm consistently, it did not significantly impact soil water content at these
depths or lower than this. In orchard floor areas covered by NCC, temperature was significantly reduced
in the summer months while also stabilizing temperature fluctuations over time. Given the ecosystem
services that this type of cover has the capability to provide orchard systems and farmers alike, broader
seed mix design and scaled-up implementation could further enhance services targeted towards farms, the
surrounding communities, adjacent wildland habitats, and the greater bioregion as a whole. With the
current price of native seed, a cost share program would likely be needed in order for affordability by the
majority of farmers. Given the economic services that NCC has the capability to provide, further
investigation and investment into these practices could help agricultural sustainability efforts, stimulate
the economy, and reduce time and money spent on pest and orchard floor management.
39
IX. CONFLICT OF INTEREST STATEMENT
I would like to acknowledge that this study site is located on my family’s hazelnut farm.
40
X. APPENDIX
Supplemental Figure 1 - Site Design. On the left is the ‘checkerboard’ pattern represented from an aerial
view and on the right is the cover type key.
41
Supplemental Figure 2 - Cover crop filtering methodology (physical). Species selection criteria was
based on plant (light availability, flailing, and scraping) and agricultural filters (soil moisture, weed
suppression, and harvest compatibility). A list of compatible cover crops were then selected through the
filtering criteria. Figure adapted from Brambila et al. (in review).
Supplemental Figure 3 - Cover crop filtering methodology (temporal) – orchard management activities
timeframes. This figure shows the basis for the temporal cover crop pool species filter. Along the bottom
are keys indicating key time points or management frames to align cover crop species pool with.
42
Species Plant Plant Listed Listed Target Traits Demonstrated
Code Growth Height Flowering
Habit Phenology
Camassia CAML Perennial 20 - 130 March - Disturbance Disturbance tolerant,
leichtlinii spp. cm July tolerant, shade semi-shade tolerant,
suksdorfii tolerant, bulb weed intolerant
forming
Collinsia COLI Annual 6 - 40 April - Semi-shade Semi-shade tolerant,
grandiflora cm August tolerant, short in phenology match,
stature short in stature
Collomia COLO Annual 10 - 90 April - Disturbance Disturbance tolerant,
grandiflora cm July tolerant, semi-shade semi-shade tolerant,
tolerant, drought drought tolerant,
tolerant weedy, tall
Geum GEUM Perennial 20 - 110 April - Disturbance Disturbance resistant,
macrophyllum cm August resistant, shade shade tolerant,
tolerant, rhizomatous,
rhizomatous spreading
Lomatium LODI Perennial 30 - 200 March - Disturbance Disturbance resistant,
dissectum var. cm July resistant, shade very shade tolerant,
dissectum tolerant, drought drought tolerant, does
tolerant well in monocrops,
flowering stalks very
tall
Phacelia PNEM Biennial 20 - 120 May - Disturbance Disturbance resistant,
nemoralis ssp. cm August resistant, shade very shade tolerant,
oregonensis tolerant, drought drought tolerant,
tolerant spreading
Prunella PRUN Biennial 10 - 60 May - Disturbance Disturbance resistant,
vulgaris cm September resistant, shade semi-shade tolerant,
tolerant, short in short in stature
stature
Sidalcea SIDA Perennial 50 - 200 May - Disturbance Disturbance resistant,
malviflora cm August resistant, semi-shade tolerant,
campestris semi-shade tolerant, persistent, very tall,
rhizomatous rhizomatous
Supplemental Table 4 - Selected seed mix species and associated target traits
Sources: Oregon Flora, Washington Native Plant Society, CalFlora
43
Species 2022 target vs 2023 target vs 2022 vs 2023 2022 vs 2023
weeds cover weeds cover target cover weeds cover
C. leichtlinii <0.0001* (-) <0.0001* (-) <0.0001* (-) 0.3811
Collinsia <0.0001* (+) <0.0001* (+) 0.0004* (-) 0.4177
grandiflora
Collomia <0.0001* (+) 0.0178* (+) 0.0262* (-) 0.0192* (+)
grandiflora
G. macrophyllum <0.0001* (+) 0.0022* (+) 0.1497 0.1052
L. dissectum 0.0018* (-) 0.0164* (+) 0.0003* (+) 0.0157* (-)
P. nemoralis 0.0003* (+) <0.0001* (+) 0.8559 0.3910
P. vulgaris <0.0001* (+) 0.0359* (+) 0.0116* (-) 0.4346
S. campestris <0.0001* (-) 0.0033* (-) 0.3046 0.3508
Supplemental Table 5 - Individual species’ and weed plot p-values by year, comparing mean
conservation cover (by species) to general weed cover per plot and significant changes by cover type.
Mean conservation cover and mean weed cover was also compared against itself, Denoted by the p-value
is a (+/-) indicating whether the change in target cover was significantly positive or negative compared to
mean weed cover, or whether the change was significantly positive or negative comparing cover type
across years. All p-values were calculated with a confidence interval of 95%, with designation being
acknowledged with the symbol: *.
Supplemental
Figure 6 -
Community plot
composition by
species type in
both trial years.
44
Supplemental Figure 7 - 2023 data comparing implement type and cover type crop loss. Harvester crop
loss is reduced in conservation cover, compared to conventional management, by a mean of 1.5 nuts per
m2, with a p-value of 0.0204. Sweeper crop loss is also reduced by conservation cover, compared to
conventional management, by a mean of 4.85 nuts per m2 , with a p-value of 0.0003. In both cases,
conservation cover significantly reduces crop loss compared to conventional orchard floor management.
45
Supplemental Figure 8 - Conventional vs Conservation Cover Management Cycle in Oregon Hazelnut
Systems. On the left in tan is the standard, bare-ground, conventional management style used in the
majority of Oregon hazelnut orchards. In early spring, weeds are sprayed, followed by flailing in late
spring-early summer. After flailing, scraping takes place in mid-summer to create a smooth orchard floor
in late summer-early fall. The ground in then flailed again and harvest proceeds after nutfall in fall.
Leaves fall on the ground in winter and the cycle then continues. Conservation cover management is
identical from late summer-early fall through winter, with spring and summer being replaced by growing
cover. Flailing is shifted to mid-summer to achieve a bare-ground orchard floor for nut pickup in the fall.
46
XI. LITERATURE CITED
Adams , Mark, et al. “Crops, Nitrogen, Water: Are Legumes Friend, Foe, or Misunderstood Ally?”
ScienceDirect, CellPress, June 2018, doi.org/10.1016/j.tplants.2018.02.009.
Babcock, J., and J. Kim. 2020. Willamette Basin Hazelnut Resource Stewardship Assessment Report.
https://orsolutions.org/. Oregon Solutions.
Balkcom, K., H. Schomberg, W. Reeves, A. Clark, L. Baumhardt, H. Collins, J. Delgado, S. Duiker, T.
Kaspar, and J. Mitchell. (n.d.). Managing Cover Crops in Conservation Tillage Systems.
Sustainable Agriculture Research and Education Program.
https://www.sare.org/publications/managing-cover-crops-profitably/managing-cover-crops-in-con
servation-tillage-systems/.
Boyer, L. 2013. NATIVE WILLAMETTE VALLEY PRAIRIE AND OAK HABITAT RESTORATION .
Heritage Seedlings.
Brambila, A., P. B. Reed, S. D. Bridgham, B. A. Roy, B. R. Johnson, L. Pfeifer‐Meister, and L. M.
Hallett. 2022. Disturbance: a double‐edged sword for restoration in a changing climate.
Restoration ecology 31.
Brambila, A., Brown, A., Lane-Massee, M., Hallett, L., Feasibility of Native Cover Crops in Hazelnut
Orchards of Varying Ages in the Willamette Valley, Oregon. Under review for the Journal of
Applied Ecology.
CaliforniaAgNet. 2024, April. Hazelnut Grower Keeps Nuts Out of the Mud by Cover Cropping.
https://www.youtube.com/watch?v=kykZRdxyt8Y.
Cárceles Rodríguez, B., V. H. Durán-Zuazo, M. Soriano Rodríguez, I. F. García-Tejero, B. Gálvez Ruiz,
and S. Cuadros Tavira. 2022. Conservation Agriculture as a Sustainable System for Soil Health:
A Review. Soil Systems 6:87.
CalFlora. 2024. Calflora - Information on wild California plants. https://www.calflora.org/.
Carney, M., S. Tushingham, T. McLaughlin, and J. d’Alpoim Guedes. 2021. Harvesting strategies as
evidence for 4000 years of camas (Camassia quamash) management in the North American
Columbia Plateau. Royal Society Open Science 8:202213.
Cavaleri, M. A., and L. Sack. 2010. Comparative water use of native and invasive plants at multiple
scales: a global meta-analysis. Ecology 91:2705–2715.
Clackamas Soil Water Conservation District. 2023, June 8. Cover Crop Field Day.
Clapp, J. 2021. Explaining Growing Glyphosate Use: The Political Economy of Herbicide-Dependent
Agriculture. Global Environmental Change 67:102239.
Clark, A. 2015. Cover Crops for Sustainable Crop Rotations. Sustainable Agriculture Research and
Education Outreach.
https://www.sare.org/resources/cover-crops/#:~:text=No%2Dtill%20farming%20or%20other.
47
Corvallis Plant Materials Center USDA. (n.d.). Native Seed Production Manual for the Pacific Northwest.
United States Department of Agriculture.
Dai, Z., M. Yu, H. Chen, H. Zhao, Y.-L. Huang, W. Su, F. Xia, S. X. Chang, P. C. Brookes, R. A.
Dahlgren, and J. Xu. 2020. Elevated temperature shifts soil N cycling from microbial
immobilization to enhanced mineralization, nitrification and denitrification across global
terrestrial ecosystems. Global Change Biology 26:5267–5276.
Davis, A. G., D. R. Huggins, and J. P. Reganold. 2023. Linking soil health and ecological resilience to
achieve agricultural sustainability. Frontiers in Ecology and the Environment 21:131–139.
Demir, Z., and D. Isik. 2020. Using Cover Crops to Improve Soil Quality and Hazelnut Yield . Fresenius
Environmental Bulletin 29:1974–1987.
Emery, D. E. 2021. Seed Propagation of Native California Plants. Santa Barbara Botanic Garden.
Eubanks II, W. 2014, June. A Rotten System: Subsidizing Environmental Degradation and Poor Public
Health with Our Nation’s Tax Dollars. Department of Justice .
https://www.justice.gov/sites/default/files/atr/legacy/2014/06/13/AGW-00064-d.docx.
Ganser, D., E. Knop, and M. Albrecht. 2019. Sown wildflower strips as overwintering habitat for
arthropods: Effective measure or ecological trap? Agriculture, Ecosystems & Environment
275:123–131.
Halassy, M., Péter Batáry, Anikó Csecserits, Katalin Török, and Orsolya Valkó. 2023. Meta-analysis
identifies native priority as a mechanism that supports the restoration of invasion-resistant plant
communities. Communications biology 6.
Haring, S., Aoyama, L., Lane-Massee, M., Ponisio, L., and Hallett, L. (in preparation). Agricultural
sustainability requires multidimensional solutions that address environmental and financial
benefits in the Oregon hazelnut industry.
Holmes, A., A. Thompson, and S. Wortman. 2017. Species-Specific Contributions to Productivity and
Weed Suppression in Cover Crop Mixtures. Agronomy Journal 109:2808–2819.
HOOPER, D. U., D. E. BIGNELL, V. K. BROWN, L. BRUSSARD, J. MARK DANGERFIELD, D. H.
WALL, D. A. WARDLE, D. C. COLEMAN, K. E. GILLER, P. LAVELLE, W. H. VAN DER
PUTTEN, P. C. DE RUITER, J. RUSEK, W. L. SILVER, J. M. TIEDJE, and V. WOLTERS.
2000. Interactions between Aboveground and Belowground Biodiversity in Terrestrial
Ecosystems: Patterns, Mechanisms, and Feedbacks. BioScience 50:1049.
Hudek, C., C. Putinica, W. Otten, and S. De Baets. 2021. Functional root trait‐based classification of
cover crops to improve soil physical properties. European Journal of Soil Science.
Ingels, C., and K. Klonsky History. 1998. Historical and Current Uses Tradeoffs in Cover Cropping
Potential Benefits. University of California, Davis.
International Network for Seed-Based Restoration. 2023, August. Native Seeds: Supplying Restoration.
McKenna Asakawa. https://www.youtube.com/watch?v=QqBlfTOt7Jo.
48
Kaspar, T., Kladivko, E., Singer, J., Morse, S., Mutch, D. 2006. POTENTIAL AND LIMITATIONS OF
COVER CROPS, LIVING MULCHES, AND PERENNIALS TO REDUCE NUTRIENT
LOSSES TO WATER SOURCES FROM AGRICULTURAL FIELDS. Research Agronomist,
USDA-ARS.
Kremen, C. 2020. Ecological intensification and diversification approaches to maintain biodiversity,
ecosystem services and food production in a changing world. Emerging Topics in Life Sciences
4:229–240.
LaForgia, M., L. Hallett, M. Borton, S. Harrison, P.-O. Cheptou, and J. Gremer. (in preparation). Diverse
seed traits predict a trade-off between spatial and temporal dispersal.
MacLaren, C., P. Swanepoel, J. Bennett, J. Wright, and K. Dehnen-Schmutz. 2019. Cover Crop Biomass
Production Is More Important than Diversity for Weed Suppression. Crop Science 59:733–748.
McMahan, L. R. 2018, June 14. Incorporating Pacific Northwest native plants into your water-wise
landscapes. Oregon State University Extension Service.
https://extension.oregonstate.edu/gardening/techniques/incorporating-pacific-northwest-native-pl
ants-your-water-wise-landscapes.
Miller, B., D. Dalton, L. Brewer, and V. Walton. 2018, May 1. Sustainable Hazelnut Production:
Filbertworm Control by Mating Disruption. Oregon State University Extension Service .
https://extension.oregonstate.edu/catalog/pub/em-9198-sustainable-hazelnut-production-filbertwo
rm-control-mating-disruption.
Murray, K., I. Sandlin, N. Wiman, P. Ellsworth, P. Jepson, H.-K. Luh, and C. Hedstrom. 2022, October
14. Measuring the Economic Impact of Pests and Pest Management on Oregon Hazelnuts. Oregon
State University Extension Service.
https://extension.oregonstate.edu/catalog/pub/em-9370-measuring-economic-impact-pests-pest-m
anagement-oregon-hazelnuts#costs-of-chemical-management-by-pest-species.
Niu, Y., Y. Li, M. Lou, Z. Cheng, R. Ma, H. Guo, J. Zhou, H. Jia, L. Fan, and T. Wang. 2024. Microbial
transformation mechanisms of particulate organic carbon to mineral-associated organic carbon at
the chemical molecular level: Highlighting the effects of ambient temperature and soil moisture.
Soil biology and biochemistry/Soil biology & biochemistry:109454–109454.
Olsen, J. 2014, December 18. Growing Hazelnuts in the Pacific Northwest: Orchard Site Selection.
Oregon State University Extension Service.
https://extension.oregonstate.edu/catalog/pub/em-9076-growing-hazelnuts-pacific-northwest-orch
ard-site-selection#:~:text=Although%20most%20of%20a%20hazelnut.
Olson, J. 2014, December 18. Growing Hazelnuts in the Pacific Northwest: Orchard Design. Oregon State
University Extension Service.
https://extension.oregonstate.edu/catalog/pub/em-9077-growing-hazelnuts-pacific-northwest-orch
ard-design.
Olsen, J., and E. Peachey. 2013. Growing Hazelnuts in the Pacific Northwest Orchard Floor Management
Weed Management.
Oregon Bureau of Labor and Industries. (n.d.). BOLI : Oregon Minimum Wage : For Workers : State of
Oregon. State of Oregon. https://www.oregon.gov/boli/workers/Pages/minimum-wage.aspx.
49
Oregon State University Department of Botany and Plant Pathology. 2024. OregonFlora. Symbiota.
https://oregonflora.org/.
Penkauskas, C., A. Brambila, D. Donahue, T. Larson, B. Miller, and L. M. Hallett. 2021. Hogs and
hazelnuts: adaptively managing pest spillover in the agricultural-wildland matrix. Agroforestry
Systems.
Petit, S., A. Alignier, N. Colbach, Alexandre Joannon, Didier Le Cœur, and C. Thenail. 2012. Weed
dispersal by farming at various spatial scales. A review. Agronomy for sustainable development
33:205–217.
Rey Benayas, J. M., and J. M. Bullock. 2012. Restoration of Biodiversity and Ecosystem Services on
Agricultural Land. Ecosystems 15:883–899.
Russell, M. 2015. A meta-analysis of physiological and behavioral responses of parasitoid wasps to
flowers of individual plant species. Biological Control 82:96–103.
Russell, M. 2011. Dormancy and Germination Pre-treatments in Willamette Valley Native Plants.
Northwest Science 85:389–402.
Scavo, A., S. Fontanazza, A. Restuccia, G. R. Pesce, C. Abbate, and G. Mauromicale. 2022. The role of
cover crops in improving soil fertility and plant nutritional status in temperate climates. A review.
Agronomy for Sustainable Development 42.
ScienceDirect. (2016). Soil Temperature - An Overview .
https://www.sciencedirect.com/topics/agricultural-and-biological-sciences/soil-temperature#:~:tex
t=Temperature%20governs%20evaporation%20and%20aeration.
Shelef, O., P. J. Weisberg, and F. D. Provenza. 2017. The Value of Native Plants and Local Production in
an Era of Global Agriculture. Frontiers in Plant Science 8.
Stephens, C. J., N. A. Schellhorn, G. M. Wood, and A. D. Austin. 2006. Parasitic wasp assemblages
associated with native and weedy plant species in an agricultural landscape. Australian Journal of
Entomology 45:176–184.
Sweet, R., & Schreiner, R.P. (2010). Alleyway Cover Crops Have Little Influence on Pinot noir
Grapevines (Vitis vinifera L.) in Two Western Oregon Vineyards. American Journal of Enology
and Viticulture.
Shackelford, G., P. R. Steward, T. G. Benton, W. E. Kunin, S. G. Potts, J. C. Biesmeijer, and S. M. Sait.
2013. Comparison of pollinators and natural enemies: a meta-analysis of landscape and local
effects on abundance and richness in crops. Biological Reviews 88:1002–1021.
Schmied, H., L. Getrost, O. Diestelhorst, G. Maaßen, and L. Gerhard. 2022. Between perfect habitat and
ecological trap: even wildflower strips mulched annually increase pollinating insect numbers in
intensively used agricultural landscapes. Journal of Insect Conservation.
Tombesi, A., and A. Rosati. 1997. HAZELNUT RESPONSE TO WATER LEVELS IN RELATION TO
PRODUCTIVE CYCLE. ISHS. https://www.actahort.org/books/445/445_36.htm.
50
United States Fish & Wildlife Service. 2017. Willamette Valley Conservation Study - Strategic Habitat
Conservation in Oregon’s Willamette Valley. United States Department of the Interior.
University of Washington. 2024. Consortium of Pacific Northwest Herbaria. University of Washington
Herbarium Burke Museum of Natural History and Culture.
https://www.pnwherbaria.org/index.php.
Washington Native Plant Society. 2023. WNPS Home. https://www.wnps.org/.
West Coast Nut. 2020, July 10. Cover Crops Catching on in Hazelnuts. West Coast Nut.
https://www.wcngg.com/2020/07/10/cover-crops-catching-on-in-hazelnuts/.
Wiman, N. 2019, January 15. Final report for OW16-028 - SARE Grant Management System. Western
SARE. https://projects.sare.org/project-reports/ow16-028/
Wiman, N., J. Pscheidt, and M. Moretti. 2014, December 18. 2023 Hazelnut Pest Management Guide for
the Willamette Valley. Oregon State University Extension Service.
https://extension.oregonstate.edu/catalog/pub/em-8328-2023-hazelnut-pest-management-guide-wi
llamette-valley.
Wittwer, R. A., B. Dorn, W. Jossi, and M. G. A. van der Heijden. 2017. Cover crops support ecological
intensification of arable cropping systems. Scientific Reports 7.
Young, J. A., and C. G. Young. 2009. Collecting, Processing and Germinating Seeds of Wildland Plants.
Timber Press.
51