Expanding Local Markets through Evaluating Sensory Characteristics and Agronomic Performance of Flint Corn Varieties

The overall objective of our project is to use flint corn field trials and a sensory directed product development approach to expand the flint corn market in the Northeast, while sharing the approach and results to the benefit of other grain growers in the United States.

Flint Corn Variety Screening Trial

The availability of high-quality flint corn varieties in the northeast region is variable and relatively low compared to dent corns. After several years of trialing as many varieties as we could find, it became clear that many of the varieties that are available on the market are either poorly adapted to our region’s growing season, do not possess the agronomic characteristics needed in commercial production, or are too expensive for a commercially viable business. Therefore, we decided to broaden the scope of our varietal evaluations to screen flint corn germplasm for basic agronomic traits. Sixty flint corn accessions were obtained from the USDA National Plant Germplasm System. Approximately 100 seeds of each accession were seeded in three rows planted 30” apart in Alburgh, VT on 21-May 2021. Each accession was therefore planted at approximately 26,000 seeds ac^-1. Each plot was evaluated for plant population, number of barren plants, plant height, ear height, lodging severity, ear disease severity, corn yield, and corn test weight when the corn reached black layer. For accessions for which there was sufficient material, samples were processed in order to evaluate the sensory characteristics of the corn. Sensory evaluation is presented below. 

In addition, 2 farm collaborators worked with the UVM team to identify a corn variety that had desirable qualities for the value-add corn market. After connecting each grower with a potential value-add corn market, a corn variety was selected and grown for the new collaboration. One farm was located in NY and one in Vermont. One farm grew Wapsie Valley and the other Early Riser on 10 acres per farm. UVM Extension helped to access the variety and provide technical assistance throughout the growing season. Both were able to produce saleable corn from the trial and sold to the value-add corn market. 

Are the production practices (i.e. populations) for flint corn different than dent corn?

Impact of Seeding Rate on Heirloom Corn Yield

In 2021 and 2022, two grain corn varieties, one flint and one heirloom dent, were each seeded at six different seeding rates at Borderview Research Farm in Alburgh, Vermont (Table 1). The trial design was a randomized complete block with split plots and four replications. Main plots were the varieties while sub-plots were seeding rates ranging from 20,000 to 30,000 seeds ac^-1. Plots were evaluated for populations, lodging, grain yield, grain moisture, and grain test weight. The soil type at the Alburgh location is a Covington silty clay loam. The seedbed was prepared with a Pottinger TerraDisc. The previous crop was corn silage. Prior to planting, plots were fertilized with 19-19-19 at a rate 300 lbs ac^-1 on 6-Apr. Plots were planted on 24-May with a 4-row cone planter with John Deere row units fitted with Almaco seed distribution units (Nevada, IA). Liquid starter fertilizer (9-18-9) was applied at planting at a rate of 5 gal ac^-1. Plots were 20’ long and consisted of four rows of corn 30” apart. Populations were counted in each plot after emergence at prior to harvest. An application of Acuron was made following planting at a rate of 3 qt ac^-1 to control weeds. In early July plots were top-dressed with 400 lbs ac^-1 24-12-18.

Table 1.  Treatment and trial management information.

In 2021, on 8-Oct and 5-Nov the Cascade Ruby Gold and Wapsie Valley plots were harvested respectively. In 2022, only Wapsie Valley was harvested on 26-Oct. Corn populations and the number of barren plants, plants that did not form an ear, were counted. Plots were also visually assessed for lodging severity on a scale from 0 (no lodging) to 5 (completely lodged). Corn was picked by hand and fed through an Almaco SPC50 plot combine. The corn from each plot was weighed and the moisture and test weight measured using a Dickey John Mini-GAC Plus moisture and test weight meter. Yield data and stand characteristics were analyzed using mixed model analysis using the mixed procedure of SAS (SAS Institute, 1999).  Replications within trials were treated as random effects, and hybrids were treated as fixed. Hybrid mean comparisons were made using the Least Significant Difference (LSD) procedure when the F-test was considered significant (p<0.10).  

Sensory

Fall 2020:                                Preparation for the 2021 growing season

In addition to the identification of the corn varieties to be used in this project, the descriptive sensory analysis piece was kicked-off and cornmeal sample preparation and sensory method development for its evaluation was initiated. One of the aims of this project is to develop new and effective ways to prepare cornmeal samples on which objective descriptive sensory analysis can be conducted to generate data that can be successfully correlated with both practices at the farms and sensory performance of final products such as corn tortillas and chips. We plan to adapt the standard approach used by many large international food companies to conduct sensory screening on grains such as corn. Typically, a 10% aqueous solution of each cornmeal sample is produced and submitted for objective descriptive sensory analysis.

The timeline and task updates are as follows:

August 2020:                          Project kick-off and logistics planning

A virtual meeting was held with the UVM Extension Northwest Crops and Soils Program team and staff from the All Souls Tortilleria to discuss the award and review overall project activities and timing, including the fall 2020 and winter 2021 preparations for the 2021 growing season.

January - March 2021:                 Sensory Orientation and Training

Early in January 2021, we collected corn tortilla chip samples produced by All Souls Tortilleria in Burlington, Vermont, to use in a sensory orientation of the trained descriptive sensory analysis panel at the University of Vermont Extension. Due to Covid-19 restrictions, individual samples were distributed to the UVM Extension trained tasters, and representatives from All Souls Tortilleria, and a virtual sensory orientation taste session was facilitated by Roy Desrochers on January 8, 2021. Participants used standard sensory methods to assess the headspace aroma of each sample bag and then record both flavor and texture data.

A modification of the Flavor Profile Method (FPM) of Sensory Analysis called Total Intensity of Aroma (TIA) was used for the bag headspace aroma analysis. FPM and TIA are standard sensory methods well defined and accepted through the American Society for Testing and Materials (ASTM), specifically committee ASTM E-18 which develops standard sensory methods.   (Reference: ASTM ML 13)

The flavor and texture of the samples was assessed using another modification of FPM call Profile Attribute Analysis (PAA) which is another standard ASTM method. These sensory methods will be used by the UVM Extension descriptive sensory analysis panel in the fall of 2021 to evaluate flint corn meal samples at the farm level, as well as flint corn tortillas and tortilla chips produced by All Souls Tortilleria using the trial flint corn samples.

In addition,  a virtual hands-on flint and dent corn sensory demonstration was included as a component of the 2021 Northern Grain Growers Association Annual Conference. 

November - December 2021:              Harvest, Physical Testing, and Sensory Directed Product Development

The flint corn trials were harvested in October & November of 2021. Prior to harvest, plots were assessed for populations, number of ears fully formed, number of plants forming ears, lodging presence and severity, ear orientation and husk cover. Plots were hand harvested and yield, average ear length and number of kernel rows were recorded. A subsample of ears were shelled, and harvest grain moisture and test weight measured using a Dickey-John Mini-GAC Plus meter. Ears from each variety were fully dried in a grain bin, shelled, and a subsample submitted to the UVM Cereal Grain Testing Lab for a basic corn nutrient analysis via NIR (Near Infrared Reflectance) procedures and Deoxynivalenol (DON) content.

In 2022, cornmeal samples were prepared for each of the 3 flint corn varieties and assessed by the UVM Extension descriptive sensory analysis panel using the sample preparation and analysis methods developed in the first quarter of 2021. We then selected samples of flint corn from the variety trials, based on agronomic performance and DON concentrations, and delivered them to All Souls Tortilleria for processing into food products. The samples were nixtamalized, a process of cooking in a hot water bath with culinary lime to remove the pericarp and alter the proteins to create a malleable dough. The trained panel conducted descriptive sensory analysis on a small portion the whole kernel, nixtamalized corn, known as hominy. Remaining hominy was wet milled into a masa dough and rolled into tortillas.  The tortillas were baked into whole tortillas and cut and fried into tortilla chips. Objective descriptive sensory analysis was conducted on both tortilla and tortilla chips samples. The trained sensory panel generated texture profiles for all the finished corn product samples. Aroma, flavor, and texture results for the end-products were correlated with all other farming, flint corn variety, and flint corn sensory data, to connect these variables to end-product performance, which is interpreted using the Flavor Leadership Criteria. (FLC) The FLC are a powerful set of sensory criteria, developed by ADL, to identify products with market leading potential. These criteria include early identifiable flavor, amplitude, mouthfeel, off-notes, and aftertaste.

January - September 2022:               Sensory Directed Product Development

We screened a range of flint corn samples using a standard approach of wetting the ground grains (10% with spring water) and evaluating the aroma of each sample using modified flavor profile Total Intensity of Aroma (TIA). The data generated using of this sensory screening approach did not yield differences that would suggest it would be a good way to screen samples for sensory quality moving forward. Consequently, we decided to use a different unique approach by making corn grits with the samples using a standard recipe and evaluating them using a different modified Flavor Profile method called Profile Attribute Analysis (PAA).

The standard recipe for corn grits we developed was:

• Ingredients:
  • 590 ml of clean, flavor free, spring water
  • 3 grams of table salt
  • 85 grams of corn grits
• Instructions:
  • Bring water and salt to a boil in a clean 2-qt saucepan
  • Reduce heat to medium and gradually stream in corn grits, whisking constantly
  • Reduce heat to medium-low, maintaining a slow simmer
  • Cook grits for 35 minutes, stirring every few minutes to prevent sticking, or scorching, on the bottom of the pan
  • Remove from heat and serve warm

We originally planned to use a modification of the Flavor Profile Method called Total Intensity of Aroma (TIA) alone as a screening method. However, this approach did not give us as much useful information as we had hoped, so we elected to use a more detailed modification of the Flavor Profile Method called PAA to assess the corn grit samples. PAA is a powerful method that uses a ballot containing predetermined aroma, flavor, and texture attributes. These attributes are selected based on previous sensory findings on similar samples.

The PAA ballot that we used for the corn grit sample sensory screening included the following attributes (Figure 1):

• TIA: The Total Intensity of Aroma by smelling the sample using small repeated sniff, recording the overall intensity and supporting descriptive terms.
• Corn description: The intensity of each of the following corn aroma types in flavor;
  • Raw corn
  • Cooked corn
  • Canned corn
  • Creamed corn
  • Sweet corn
• Other grain description: The intensity of each of the following other grain aroma types in flavor;
  • Starchy
  • Cooked grain
  • “Cream of Wheat” character
  • Flour
• Paper/cardboard/woody: The intensity of any woody, cardboard, or papery characteristics
• Sulfidy: The intensity of sulfide related aromas including:
  • Vegetables, such as peas
  • Rubbery, or any odd smelling sulfide
• Brothy/meaty: The intensity of any meat, or brothy, characteristics
• Other: The intensity of any other flavor characteristics detected but not measured by one of the other predetermined PAA attributes
• Bitter: The intensity of bitter flavor detected on the tongue

Intensity was measured using the standard seven-point intensity scale that was developed by Arthur D. Little as part of the original Flavor Profile Method. The numbers reflect the change to whole numbers typically used when conducting PAA:

1 = None (original Flavor Profile value = 0)

2 = Very Slight (original Flavor Profile value = ½ )

3 = Slight (original Flavor Profile value = 1)

4 = Slight-to-Moderate (original Flavor Profile value = 1 ½ )

5 = Moderate (original Flavor Profile value = 2)

6 = Moderate-to-Strong (original Flavor Profile value = 2 ½ )

7 = Strong (original Flavor Profile value = 3)

In addition to aroma and flavor, we used two texture attributes on the corn grit PAA ballot:

• Particulate 1: The overall number of particulates perceived ranging from none to many
• Particulate 2: The size of the particulates detected ranging from very small to large

The final PAA ballot used to assess the corn grit samples contained 6 samples and all aroma, flavor, and texture attributes. (see attached Figure 1)

 We included six varieties of flint corn in the corn grit screening test:

1. Comstock Family
2. Johnny Cake
3. King Philip
4. Salzer’s White
5. Longfellow
6. Rotor Tess

Although we did not see big differences, we did see differences and decided to select four varieties to use to produce both corn tortillas and corn chips. These varieties included Comstock Family, Johnny Cake, King Phillip, and Salzer’s White. The corn grit PAA data for these varieties included the panel average scores for each attribute that was measured. (Table 2)

Next, we provided corn samples of each of these four varieties of flint corn to All Souls Tortilleria, who produce both corn tortillas, and corn chips, using their standard recipe and processing. The samples were submitted to the UVM trained sensory DSA panel for evaluation.

The frozen tortilla samples were wrapped in a section of clean, flavor free, paper towel and heated in a microwave for approximately 20 seconds. They were then presented to the UVM DSA panel warm for evaluation.

The UVM DSA panel used a different PAA ballot to assess the tortilla samples. (Figure 2)

The corn tortilla PAA aroma, flavor, and texture attributes included:

• TIA: The Total Intensity of Aroma by smelling the sample using small repeated sniff, recording the overall intensity and supporting descriptive terms.
• Balance: An integrative measure of the harmony of flavor components measured from unblended to blended
• Fullness: An integrative measure of the complexity of flavor components measured from thin to full body
• Toasted Corn: The intensity of toasted corn flavor
• Other Corn: The intensity of corn flavor, other than toasted, that is detected
• Other grain: The intensity of any grains detected other than corn
• Green, grassy: The intensity of green or grassy flavor detected in the sample
• Sweet: The intensity of sweet taste detected on the tongue
• Sour: The intensity of sour taste detected on the tongue
• Salty: The intensity of salty taste detected on the tongue
• Mouthfeel: The overall combined intensity of mouthfeel detected including dry, oily, and irritation
• Others: The combined intensity of any other flavor notes detected in the samples
• Aftertaste: The intensity of flavor left in the mouth one minute after the last taste

Intensity was again measured using the standard seven-point intensity scale that was developed by Arthur D. Little as part of the original Flavor Profile Method.

In addition to aroma and flavor, we included four texture attributes on the corn tortilla PAA ballot:

• Hardness: The perception of hardness during the first bite, measured from soft to hard
• Crumbly: The perception of how much the sample crumbles upon the first bite, measured from not crumbly to crumbly
• Grain size: The perception of the size of the grains detected during tasting, measured from small to large
• Moistness: The perception of moistness of the sample, measured from not moist to moist

The corn tortilla PAA data indicated differences between the samples. (Table 3, note: a 0.5 difference on a trained DSA panel average for any attribute is considered meaningful and is highlighted in the attached tables)

The findings derived from these tables and analysis of the data will be discussed in a subsequent section.

Next, corn chip samples produced using the same four flint corn varieties were presented to the UVM DSA panel at room temperature for assessment.

The UVM DSA panel used the attached PAA ballot (Figure 3) to assess the tortilla samples.

These corn chip PAA aroma, flavor, and texture attributes included:

• TIA: The Total Intensity of Aroma by smelling the sample using small repeated sniffs while recording the overall intensity and supporting descriptive terms.
• Balance: An integrative measure of the harmony of flavor components measured from unblended to blended
• Fullness: An integrative measure of the complexity of flavor components measured from thin to full body
• Toasted Corn: The intensity of toasted corn flavor
• Other Corn: The intensity of corn flavor, other than toasted, that is detected
• Other grain: The intensity of any grains detected other than corn
• Fresh Fried Oil: The intensity of fresh fried oil flavor detected in the sample
• Oxidized, or Rancid, Oil: The intensity of oxidized, or rancid, oil flavor detected in the sample
• Sweet: The intensity of sweet taste detected on the tongue
• Sour: The intensity of sour taste detected on the tongue
• Salty: The intensity of salty taste detected on the tongue
• Oily or Greasy Mouthfeel: The intensity of oily or greasy mouthfeel detected in the sample
• Dry Mouthfeel: The intensity of dry mouthfeel detected in the sample
• Astringent Mouthfeel: The intensity of astringent (puckering) mouthfeel detected in the product
• Others: The combined intensity of any other flavor notes detected in the samples
• Aftertaste: The intensity of flavor left in the mouth one minute after the last taste

Intensity was again measured using the standard seven-point intensity scale that was developed by Arthur D. Little as part of the original Flavor Profile Method.

In addition to aroma and flavor, we included five texture attributes on the corn chip PAA ballot:

• Hardness: The perception of hardness during the first bite, measured from soft to hard
• Crispiness: The intensity of crispy texture measured from not crispy to strong crispy
• Crumbly: The perception of how much the sample crumbles upon the first bite, measured from not crumbly to crumbly
• Grain size: The perception of the size of the grains detected during tasting, measured from small to large
• Oily or greasy: The perception of oiliness, or greasiness, of the sample, measured from not oily or greasy to very oily or greasy

The corn chip PAA data indicated differences between the samples: (see attached Table 4 note: a 0.5 difference on a trained

DSA panel average for any attribute is considered meaningful and is highlighted in the prior tables. 

The findings derived from these tables and analysis of the data will also be discussed in a subsequent section.