Tomato Variety Trials for Flavor, Quality and Agronomic Performance, to Increase High-value Direct Marketing Opportunities for Farmers and On-farm Trialing Capacity

Tomato Variety Trials for Flavor, Quality and Agronomic Performance, to Increase High-value Direct Marketing Opportunities for Farmers and On-farm Trialing Capacity

Summary

This project uses participatory research methods to evaluate tomato varieties for agronomic traits, disease resistance, flavor and quality for local and regional markets in the NCR. We conducted variety trials in organic conditions at the West Madison Agricultural Research Station, comparing hoop house and open field production, and at the Rock County Extension Farm, using organic practices on non-certified land. We also had five participating farmers trial subsets of the tomato varieties on their farms. We assessed quality and flavor by using a combination of field crew evaluations and Madison area chef evaluations, in addition to several public taste tests at field days and other events. We have worked on improving methodology for on-farm trials to make them more relevant for farmers. This project will create a variety trialing and quality evaluation network that will also be used for other crops.

Objectives/Performance Targets

• Conduct organic and low-input IPM field trials investigating selection of tomato varieties for optimal economic and environmental sustainability on two research stations and on 6 participating farms.
• Characterize varieties with acidity, Brix and nutritional measurements, and identify potential correlations with quality and flavor.
• Evaluate flavor for each variety with a panel of chefs currently sourcing local products.
• Provide data for breeders, researchers, farmers and seed companies to identify both promising varieties and traits needing improvement for the NCR
• Develop a network of farmers, extension specialists and chefs interested in participatory research.
• Create an online database that will be easy for farmers to both access and contribute to, following the model of successful citizen science projects and participatory plant breeding trial methodology

Accomplishments/Milestones

Conduct organic and low-input IPM field trials investigating selection of tomato varieties for optimal economic and environmental sustainability on two research stations and on 6 participating farms.

In 2015 we conducted two on-station variety trials, each of which included 20 varieties of tomatoes, chosen based on data from preliminary trials in 2014. The trials were at the West Madison Agricultural Research Station, and the Rock County Extension Farm in Janesville, WI.  At West Madison, we compared hoop house and open field production of all varieties, on certified organic land. At the Rock County Farm, we tested varieties in open field production, using organic practices on non-certified land. On-station trials were conducted as randomized complete block designs with the same check varieties (Big Beef and Pruden’s Purple) as in the 2014 screening trials. Six on-farm satellite variety trials planted a subset of 7 varieties: Big Beef, Pruden’s Purple, Caiman, Bear Creek, Defiant, Plum Regal and 45L23. Unfortunately one farm had to drop out of the trial when pesticide from a neighboring farm killed all the plants. Two of the remaining five farms included a comparison of field and hoop house production of tomatoes. The other three tested the varieties in open-field conditions. On-farm trials were planted with a replicated check (Big Beef) as an augmented experimental design.

Tomatoes were harvested weekly from mid July to mid October at West Madison, and three times from mid-September to early October at Rock County. The Rock county trials did not ripen until mid-September. At each harvest, tomatoes were sorted into “marketable” and “unmarketable” categories. Any fruits damaged by splitting, disease, insects, rodents or weather were considered unmarketable. Natural cat-facing on heirlooms or small dry cracks due to rapid growth were still considered marketable. Total weight and number of marketable tomatoes were recorded for each plot, as were total weight of unmarketable tomatoes and causes of damage. Perfectly ripe tomatoes were put aside for flavor evaluation and flavor components analysis. Disease ratings were taken three times during the season in each location.

The greatest difference was between tomatoes grown in the open-field and those in the hoophouse, with variety being a significant but lower in magnitude effect. In the 2014 preliminary trials, marketable yield was 20% greater in the hoop house, and in 2015, marketable yield was 45% higher in the hoop house, primarily due to greater disease in the field, particularly bacterial speck. Disease was measured as area under the disease progress curve, and was twice as high in the field in 2014 and 7 times as high in the field in 2015 compared to the hoop house. Varieties showed some differences in susceptibility to septoria and early blight, the two major foliar diseases, but these differences were much smaller than the effect of the management system (hoop house vs. open field).

Characterize varieties with acidity, Brix and nutritional measurements, and identify potential correlations with quality and flavor.

Brix and titrateable acidity was evaluated on all tomato plots from West Madison Agricultural Research Station twice during the growing season, and separately on each sample used for flavor evaluation. Brix levels were tested using a digital refractometer. Citric acid content by volume was measured using an automated titrator. pH was also measured for each tomato juice sample, though it was of lesser interest than titrateable acidity, since the latter measures acidity by volume, thereby giving a more accurate indication of how acid is perceived as a flavor.

Titrateable acidity did not differ significantly between hoophouse and field production, while brix was higher on average in the hoop house. Varietal differences were significant, and contributed more to total variation than management. The ratio of citric acid:˚Brix was somewhat correlated with perceived flavor intensity, and could potentially be used as a screening tool to help choose a subset of varieties most likely to have good flavor for further evaluation by a panel.

Samples of tomato from each sample used for brix and titrateable acidity have been processed and frozen for nutritional analysis.

Evaluate flavor for each variety with a panel of chefs currently sourcing local products.

We conducted two flavor evaluations with five chefs at each evaluation. Tomato varieties were first evaluated by members of the field crew at West Madison after a simple training exercise to allow the recognition and scoring of different flavor attributes. These included sweetness, acidity, saltiness, bitterness and umami. Color and texture were also rated by crew panelists. A mixed model analysis of variance was used to analyze data from this tasting, to determine which factors were most important in differentiating varieties. Using the variety means from the ANOVA, principal component analysis (PCA) was used to visualize the relationship among varieties using their entire quality profile. Both the estimated variety means and the PCA evaluation of variety similarity were used to select a subset of varieties to be evaluated by participating project chefs for flavor.

Chefs participated in a sensory evaluation exercise known as “Projective Mapping.” This is a similarity-based method of descriptive analysis particularly well suited to culinary professionals rather than professional sensory evaluators (Frost et al. 2015). There is a quantitative component based on the distance samples are placed from each other and a qualitative descriptive component. Chefs independently taste each sample and place the samples on the mapping sheet according to their perception of similarity and dissimilarity. Once placed, chefs write the words they associated with a variety’s flavor and texture directly on the map.

This allowed the chefs to evaluate only the most important attributes of each species and intuitively compare varieties rather than rating each on a 5-point scale, which many chefs find fairly arbitrary and uninformative (Healy Thesis 2016). After completing the mapping exercise, chefs were instructed to evaluate the varieties again, this time from a hedonic perspective. Questions like “would you by this for your restaurant?” “how would you use this” and “what is the best trait of this variety?” helped assess the chef’s preferences for certain varieties, and ideas about how favored varieties might be useful in a culinary setting. Analysis of the chef tasting data is done using Multiple Factor Analysis (MFA). Individual chef maps are scaled so that each chef contributes the same amount to the total variance in the dataset. A consensus map of varieties is then produced is then interpreted using the comments and preference ratings given by the chefs.

We also conducted several public taste tests, at field days and open-house events at the West Madison Research Station, at the Wausau Farmers Market and Stoney Acres’ farm Pizza Night, at the Fondy Farmers Market and Alice’s garden in Milwaukee and at the Centro Hispano Fiesta Hispana in Madison.

Provide data for breeders, researchers, farmers and seed companies to identify both promising varieties and traits needing improvement for the NCR

Data from these trials was summarized and distributed to contributing breeders, farmers, and chefs by email, and to the public online. Participating breeders were contacted to follow up on any questions and to request feedback on reports. All breeders felt that the reports were useful and are interested in testing other breeding lines and varieties in the trials in coming years. We identified several promising breeding lines and varieties that have good performance in the upper Midwest for productivity and flavor. We are planning to publish an extension report on these results in the coming year.

Data from these trials is available in summary form online at dawson.horticulture.wisc.edu and is being analyzed for publication.

Develop a network of farmers, extension specialists and chefs interested in participatory research.

We have expanded our network of farms, chefs and researchers involved in the trials for this coming year. We had five participating farms in 2015, and 15 farms have expressed interest in hosting a trial for 2016. We also are working with UW Extension in Spooner, Superior and Appleton to expand trialing in those areas, in addition to our current work with extension in the Milwaukee Area and Rock County.  

All of the participating chefs from 2015 are intending to continue, and additional chefs have attended planning meetings in February. We have also identified chefs in Milwaukee, Wausau, Spooner and Superior who have an interest in evaluating flavor in collaboration with participating farmers in those areas. We are working to develop methods that facilitate the involvement of more farmers and chefs in areas farther away from Madison, where it is more difficult for project personnel to attend evaluations or regularly visit on-farm trials.

Create an online database that will be easy for farmers to both access and contribute to, following the model of successful citizen science projects and participatory plant breeding trial methodology

This will be completed in 2016-17, we have completed background research on databases and are working on implementation.

Impacts and Contributions/Outcomes

The most significant impact so far is the development of an on-farm participatory trialing network which farmers would like to participate in. We have more farmers asking to host trials in 2016 than in 2015, and have also worked to develop a method of participatory trialing that is accessible and takes advantage of the particular expertise of the farmers and the researchers.

In our pilot year of 2014, we distributed data sheets that were fairly simple, which asked farmers to rate each variety on a scale of 1-5 for early vigor, disease resistance, earliness (for harvest), yield, marketable yield, and marketability. Of the 10 farms involved in the on-farm pilot project, which included tomatoes and other crops, we got one data sheet back. So we met with the farmers involved to work out how to make the data collection more manageable and more relevant. We asked the farmers what kind of information they want in order to make decisions about growing a variety and what they would like to know from other farmers about varieties in the trial. Farmers were most interested in marketability and quality, and while productivity was important, as long as varieties produced reasonably well it wasn’t the key trait they were looking for. We came up with a list of questions they would ask other farmers about varieties, and this became our data sheet for the 2015 year.

This datasheet included: Would you grow this again? How marketable is it? What did you think of the flavor? Strongest point; Major flaws; Best/Worst Variety (choose one best and one worst variety for each crop); General Notes (any other observations you have that don’t fit in the previous categories)

We asked them to include notes on productivity and disease susceptibility in the strongest points/ major flaws category, and really focused on capturing their expertise as farmers. Farmers have developed, over many years of experience, an ability to assess varieties holistically and we want to bring that expertise to breeders so that they are able to focus on the most important elements for direct market growers. Both the breeders and the farmers felt that this assessment was more useful, combined with more detailed data from the research station trials on the relative performance of different varieties.

I think we need to recognize that the goal of on-farm research should not be to turn farmers into scientists. One of the reasons the farmers gave for not filling out the data sheet in 2014 was that they were worried about not doing it correctly because they had no formal training in research or plant pathology. We had not considered that they might think there was only one “correct” way to fill out the data sheets and had unconsciously structured them in a way that is familiar to researchers but not to farmers.   We certainly did not intend to make the farmers uncomfortable, or insecure about their knowledge, or to imply that the way researchers look at information is somehow better than how farmers assess varieties.

The change to a farmer-developed datasheet has had an enormous effect on the contributions from farmers. All but one of the ten farmers involved in the participatory trials this year (tomatoes and other crops) returned completed datasheets, and we have expanded the number of farms interested in participating. While we went into participatory research with the goal of combining farmer and researcher expertise on varieties, we did not realize at the beginning how important it was to ask farmers about varieties in the way that they think about varieties, rather than the way researchers think about varieties. I think this change will help us get better information, that is truly complementary to the research station trials, and will make the trial information much more relevant to farmers.

We presented preliminary information on this project as a poster at three conferences this winter: the Wisconsin Local Food Network, Jan 14-15 2016 in Sheboygan, WI; the Organic Agriculture Research Symposium, Jan 20 2016 in Asilomar, CA, and the MOSES Organic Farming Conference, Feb 26-27 2016 in LaCrosse, WI. Kitt Healy presented project results as part of her Masters defense seminar Jan 29, 2016 in Madison, WI.  Kitt Healy and Julie Dawson also presented the project at the Organic Seed Growers Conference, Feb 5 2016 in Portland, OR, and at the University of Minnesota Horticulture Seminar Series, Feb 24 2016 in St. Paul, MN.

Preliminary data was distributed to all participants and is posted online with information about the on-farm trials. Information on high tunnel production, disease resistance and management, variety adaptation and flavor will be published in an extension bulletin, and a scientific publication later this year. This data was included in the Master’s Thesis of Kitt Healy, who was partly funded by this project and who successfully defended her thesis in Jan, 2016

Collaborators: