Diversifying Northwestern fields and palates

Objective 1: Establish a regional baseline for production and end-use quality of millet and buckwheat varieties currently grown in Idaho and Washington by determining their flavor, food functionality, and nutritional quality to bolster marketability via appropriate end-use recommendations (Year 1).

In Year 1, we will assess regional millet and buckwheat production and end-use to establish baseline information on agronomic and end-use performance and to inform improvement strategies. We will map commercial millet and buckwheat production and gather quantitative data (acreage, variety, yield, end-use). We will collect representative grain samples from commercial growers and analyze them as outlined in Objective 3. We will conduct producer interviews at the beginning and end of the project (Table 1) to document producer needs and enable evaluation and impact assessment.   

In the following research and education objectives, we will show how we will assess flavor, food functionality, and nutritional quality data to identify optimal end uses for each crop and variety grown. We will work with our producer-collaborators and stakeholders to develop reliable food processing methods to bring these crops and products to market (Figure 2) (Objective 3). For example, a producer-collaborator-grown millet crop would be preliminarily evaluated, milled, and then used in products developed with a collaborating stakeholder and Dr. Ganjyal’s lab. Expert sensory panels, consisting of research team members and stakeholder-collaborators, will be used to provide focused feedback that informs the optimization of product formulations (Objective 3). Panelists will complete a questionnaire (Table 2). Later in the project, consumer acceptability and preferences will be crowd-sourced at outreach events using rapid response surveys at booths with taste-testing activities (Table 2) (Objective 5.4).

Objective 2: Conduct researcher- and farmer-led millet and buckwheat field trials to identify new varieties for the region that improve upon baseline (Years 1-2).

The project aims to select new millet and buckwheat varieties that improve upon baseline agronomics, flavor, food functionality, and nutritional quality. Varieties will be assessed with a mother and baby trial (MBT) design that captures regional adaptability (Figure 3). Most commonly used in participatory plant breeding, the MBT approach links researcher-led “mother” trials, where all crop factors are evaluated in a replicated design, to farmer-led “baby” trials (not replicated on farm but replicated across farms) to capture diverse management practices and environmental conditions ^[21].

MBT locations (Figure 4) will span temperature and precipitation regimes in Western WA/Northern OR (buckwheat) and Eastern WA/Idaho Panhandle/Northern OR (millet). Buckwheat sites will have varied planting times as well as precipitation and temperature regimes. The buckwheat mother trial site will be located at the 78^th St. Heritage Farm in Vancouver, WA, typified by a Willamette Valley regional climate with slightly higher in-season temperatures and lower precipitation than most confirmed buckwheat baby trial sites located in more maritime regions of the NW. The millet mother trial will occur at the Spillman Research Farm in Pullman, WA; the site receives 21-in. mean annual precipitation (MAP). Baby trials will be located within 260 miles of Pullman in WA, ID, and OR, with MAP ranging from 11-in to 24-in. We will identify additional baby trial hosts in cooperation with Shepherd’s Grain, a group with extensive farmer networks in the Inland Northwest.

One mother trial will be planted for each crop species (Years 1-2); a researcher-led randomized and replicated variety trial of millet in Eastern WA (7 varieties) and buckwheat in Southwest WA (6 varieties) will each. The mother trials will include varieties identified in Objective 1, lesser known commercially available varieties, and breeding populations developed in Washington state. We will determine key traits of interest for each crop with the producer interview (Table 2). During the proposal development, growers identified weed suppression, ground cover, and a short growing season as important traits for buckwheat and millet in their rotation. Evaluated traits will include percent emergence, days to canopy closure, disease incidence/severity, days to maturity, plant height, and yield.

A sub-set of varieties will be grown across a range of environments in non-replicated, farmer-led, on-farm baby trials (Years 1-2; 1-4 varieties, plus a control variety) per crop to determine variety performance across environments and capture farmers’ perceptions and acceptance of new varieties. The control variety will be replicated at each baby trial. The host farmer will determine the size of a baby trial and the number of entries at each trial. Replication of varieties across farms will increase statistical power for assessing variety differences across locations. We will analyze data among baby trials if more than three growers are assigned more than one of the same variety or if growers elect to replicate within their farms. We will encourage participating farmers to plant field-scale experimental plots (> 0.25 acres) to effectively assess variety performance in their system. Large plot sizes will capture the heterogeneity across a field, which helps improve inference when replication is otherwise lacking. Host farmers will provide qualitative data (e.g., vigor, overall acceptance, production challenges) and quantitative data (e.g., emergence, canopy development, yield).

By using both replicated trials (“mother”) and non-replicated trials (“baby”), the MBT experimental design allows flexibility in analysis when using a mixed-effects model. We can: (1) determine the correlation between variety performance on-farm (baby trial) and variety performance on-station (mother trial); (2) measure variety reliability using the probability it will perform above or below the mean; and (3) conduct pairwise comparisons of varieties. If a variety is not replicated across farms and is therefore not well represented across the region, fewer inferences can be made ^[22]. However, these data remain critical to the project because they inform future variety selection decisions by farmers. Additionally, the qualitative data provided by farmers will be included in our publications, with permission.

Seed produced in MBTs will be used to develop product formulations and for nutritional quality and food functionality assessments (Objective 3). The remaining baby trial seed will be incorporated into products by stakeholder collaborators, used for bird seed by Global Harvest Foods (Spokane, WA), and used for on-farm livestock feed.

Objective 3: Characterize new millet and buckwheat varieties for flavor, food functionality, and nutritional and malt quality to determine end-use suitability while also developing optimized product formulations for focaccia mix, pancake mix, and children’s breakfast bar (Years 1-3).

Dr. Murphy’s lab will perform nutritional characterization of millet and buckwheat samples from MBTs. Mineral content will be measured with traditional analytical methods. Data will be used to develop calibrations for high-throughput analysis via near-infrared and x-ray fluorescence spectroscopy. An innovative flatbed scanner phenotyping platform will measure seed size, shape, color, and thousand seed weight.   

Physicochemical properties of flours are important for identifying how the flours perform in different food applications. The chemical compositions and functionalities of millet and buckwheat from breeding trials in Objectives 1 and 2 will be measured in Dr. Ganjyal’s laboratory. The chemical compositions to be evaluated include starch, protein, fiber, crude fat, and ash content. These will be determined using standard methods. Flour functional properties to be evaluated include thermal, pasting (rheological), and hydration properties^[23,24]. A Differential Scanning Calorimeter will perform thermal analyses. Pasting properties will be analyzed with a Micro-Visco amylograph^[24]. The water absorption index and water solubility index will be measured according to published standard methods^[23]. All data from the above physicochemical analyses will be used to categorize the millet and buckwheat varieties for processing conditions and potential end-use applications (food and malting/brewing). (Figure 2). Using this approach, Dr. Murphy and Dr. Ganjyal have successfully categorized diverse varieties of quinoa into groups based on the most appropriate end use, such as noodles, pasta, sauces, soups, and different baking products, depending on physicochemical analyses^[23].

We will select millet and buckwheat varieties based on their chemical composition and functionalities to develop three food product formulations: focaccia mix, pancake mix, and children’s breakfast bar. Formulations will be optimized to maintain or improve flavor, texture, and nutritional quality. WSU Malt Quality Lab (Pullman, WA) will provide consultation and support malting, malt quality analysis, and interpretation of malt quality with sensory data provided through stakeholder-expert panels.

The varieties of millet and buckwheat will be selected and tested in the following specific food applications: a) focaccia mix, b) pancake mix, and c) children’s breakfast bar.

Focaccia mix – Focaccia mix formulations will be tested. Wheat flour + millet flour blends (0, 25, 50, 75, and 100%), as well as wheat flour + buckwheat flour blends (0,25,50,75,100%) will be used as the base formulation for the focaccia mix. A commercial wheat-based focaccia will be used as control.

Pancake mix – The product formulation will be initiated from a commercial pancake mix. The wheat + buckwheat flour blends (0, 25, 50, 75, and 100%) and the wheat + millet flour blends (0, 12.5, 25, 50, and 100%) will be evaluated and compared with the 100% wheat pancake mix as controls. 

Children’s breakfast bar – The product formulation will be initiated from a commercial breakfast bar, and up to 40% of buckwheat and millet blend will be used in the formulations. Formulations will be optimized to have millet and buckwheat as leading ingredients and to maintain taste and texture without compromising nutritional value

An expert sensory panel and the research team will evaluate select products from each of the above-mentioned categories for consumer acceptability (Figure 2). The expert sensory panel will include stakeholders who already carry the identified three products, are interested in branching out into those products, or produce other products (e.g., noodles, beer) but will provide informative feedback on the varieties tested. Our stakeholder-collaborators will also use the products as the initial formulations in development of commercial products. For example, a buckwheat variety identified as a high performing flour in quick bread mixes will be 1) sourced from a baby trial (Hidden River Farm); 2) dehulled, kept as whole groats, or milled; and 3) shipped to a stakeholder-collaborator (Snacktivist). Snacktivist will use the buckwheat in the pancake formulation provided by the research team but will also be encouraged to experiment beyond the provided formulation (Figure 3). Snacktivist will complete a questionnaire (Table 2) designed to provide the research team critical information on end-use suitability of the variety and performance of the product formulation. Several of our stakeholder letters of support identify additional members on the expert sensory panel.

Objective 4: Conduct a market assessment for regionally produced and processed millet and buckwheat with consumer surveys, school district questionnaires, and stakeholder interviews.

We will work with WSU’s Division of Governmental Studies and Services (DGSS) to assess marketability for millet and buckwheat products. Over the three-year project period, we will: 1) conduct consumer surveys; 2) provide school districts comprehensive questionnaires; and 3) interview stakeholders to better understand pricing and feasibility of product incorporation (Table 2).

Because the majority of consumer groups relative to our project are not easily reached via random sampling, nonprobability methodology (convenience sampling) will be used. The increased response numbers obtained through nonprobability survey methods will provide important information regarding the local market for these products while avoiding the substantial costs of probability surveys. The survey will be promoted among partners such as the Culinary Breeding Network and Organic Seed Alliance and will take advantage of mailing lists, other listservs, newsletters, social media, and websites. The targeted audience will also be reached via our Field to Flavor Showcase events.

School districts are a key institutional buyer. We will develop a comprehensive questionnaire to capture school district requirements such as price ranges, sourcing constraints, product fit, food safety certifications, volumes handled, and volume periodicity. We will email a link to a web-based survey to all 294 school districts in Washington State and to school districts in Idaho and Oregon familiar with farm-to-school projects.

We will conduct a series of in-person interviews with businesses/stakeholders (including those on the expert sensory panel), to gain a better understanding of the potential they see for the grain varieties they tested, feasibility of incorporating millet/buckwheat into their products, and to assist in guiding price estimation. The target audience for this outreach will be largely informed by responses to the surveys conducted earlier in the study. Survey data will be entered, cleaned, and analyzed by the Division of Governmental Studies and Services, as will data obtained through the interviews.