Sweetpotatoes: Testing traits for increased market value and sustainable production for direct market farms in the maritime northwest

On-station trials

On station trials were held at WSU Mount Vernon NWREC (48°26’28”N, 122°23’38”W). The region has a cool Mediterranean climate, and during the summer growing season (June through September) the average (1970 to 2023) air temperature is 15.7°C (average minimum 10.1°C, average maximum 21.5°C) and precipitation is 135.7 mm (AgWeatherNet, 2023).

For all on-station trials, air temperature, relative humidity (RH), solar radiation, and rainfall data during the cropping season were collected from the WSU AgWeatherNet station located approximately 140 m from the experimental field plot. Soil temperature and moisture were recorded at 15-minute intervals with sensors (TEROS 11; Meter Environment, Pullman, WA) installed 10 cm deep in the center of the middle plot in replicate 2. 

1. Wireworm Resistant Cultivars Trial

A replicated variety trial was carried out in a certified organic field in 2023 and 2024. It will be repeated in 2025. The experiment had a randomized complete block design with 5 sweetpotato varieties and 4 breeding lines as the treatments and four replications. Varieties Bayou Belle, Beauregard, Covington, Monaco, and Orleans were from the North Carolina State University sweetpotato breeding program (J. Schultheis, Horticultural Science; G.C. Yencho and K. Pecota, Sweetpotato Breeding Program, Raleigh, NC). Breeding lines PI 666141 (released as cv. Cascade), USDA-04-136, USDA-04-284, and USDA-04-791 were from the USDA ARS sweetpotato breeding program (P.A. Wadl, Research Geneticist, Charleston, SC). ‘Covington,’ which is susceptible to wireworm, was the control treatment. Each plot included 15 plants in a single row with 25.4 cm spacing between plants on a raised bed covered with black polyethylene (PE) mulch (1 mm, Filmtech, Allentown, PA). Plots were separated by 1.5 m in the bed to facilitate data collection, and beds were 1.8 m center-to-center. An organic fertilizer (8-2-4, Stutzman Environmental Products, Inc., Canby, OR) was applied over the center of the row at the rate of 112 kg‧ha^-1 of nitrogen and incorporated as the beds were formed. Drip irrigation tape (Rivulis Irrigation T-Tape, Kibbutz Gvat, Jezreel Valley, Israel) with 20 cm emitter spacing was installed at the time PE mulch was laid. Slips were produced from sweetpotato roots grown and stored at WSU NWREC. Planting was May 31 in 2023.

Plant establishment/growth. Plant establishment was measured as the total number of live plants per plot 2 and 4 weeks after transplanting. Plants with any visible amount of green tissue were counted as alive. Length of the longest vine was measured from the soil level to the base of the growing point for the center six plants in each plot 2 weeks after transplanting (June 14, 2023), in mid-season (August 9, 2023), and at harvest (September 27, 2023). Canopy cover was measured at the same time as vine length and included the five northernmost plants of the center six plants. Percent canopy cover was measured using Image-J (Rasband, 2023). Photos were taken from 1 m above the surface of the raised bed.

Storage root yield and wireworm damage. Harvest was October 3 in 2023, before the soil temperature reached below 13℃ (Brandenberger et al., 2014). The center 2 m of each plot was hand harvested for yield and wireworm damage assessment. For yield assessment, sweetpotatoes were sorted into categories using U.S. standards for grades of sweetpotatoes (USDA, Agricultural Marketing Service, 2005): Jumbo (>22.9 cm length and >8.9 cm diameter), U.S. No. 1 (7.6‒22.9 cm length and 4.4-8.9 cm diameter), U.S. No. 2 (≥3.8 cm diameter), and cull (storage roots of any size with off-shapes and/or damages from disease, rodents, bruises, or other means). Sweetpotato roots that were not at least 2.5 cm in diameter and 7.6 cm in length were discarded. The number and weight of harvested sweetpotatoes were recorded for each cultivar. Average number and marketable yield of sweetpotatoes per plant were also calculated for each treatment. Root damage by wireworms was rated using a severity index based on the number of feeding scars (no scars = 0, 1-5 scars = 1, 6-10 scars = 2, >10 scars = 4) (Schalk et al., 1993), and the overall wireworm damage severity per plot was calculated as a weighted average score between 0 and 4. Only wireworm scars greater than 1 mm in diameter were counted.

Data analysis. All data were analyzed using statistical analysis software (R version 4.3.2, R Studio, Vienna, Austria). The assumptions of normality and homogeneity of variance were tested using the Shapiro-Wilk Test (α=0.05) and Levene’s Test (α=0.05), respectively. Means were separated using Tukey’s HSD Test (α=0.05).

2. In-Row Plant Spacing Trial

This experiment was added in response to grower interest in spacing efficiency in the field. The experiment had a split-plot design with four main plot treatments and three subplot treatments, and four replications. The main plot treatment was sweetpotato varieties and breeding lines, and subplot treatment was plant spacing. Main plot treatments included the USDA ARS breeding line Cascade (PI 666141) and cvs. Monaco, Beauregard and Covington. The subplot treatment included three plant spacings: 8 in, 10 in, and 12 in. Each subplot included 10 plants in a single row on a raised bed covered with PE mulch. Plots were separated by 1.5 m, and beds were 1.8 m center-to-center. Urea (46-0-0) was applied at the rate of 112 kg‧ha^-1 of nitrogen and incorporated as the beds were formed. Drip irrigation tape was installed at the time mulch was laid. Planting was May 31 in 2023.  

Plant establishment/growth. Plant establishment was measured as the total number of live plants per plot 2 and 4 weeks after transplanting. Plants with any amount of green tissue were counted as alive. Length of the longest vine was measured from the soil level to the base of the growing point for the center six plants in each plot 2 weeks after transplanting (June 14, 2023), in mid-season (August 10, 2023), and at harvest (October 2, 2023). Percent canopy cover was measured using Image-J to analyze photos taken from 1 m above the surface of the raised bed. Canopy cover was measured for the same center plants in each plot at the same time as vine length was measured. For plots with 10 inch spacing, five plants were included in canopy cover analysis. For plots with 8 and 12 inch spacing, photos were taken from the same 1 m height so they included the same area of analysis, which corresponded to six and four plants, respectively.  

Storage root yield. Harvest was October 2 in 2023, before the soil temperature reached below 13℃ (Brandenberger et al., 2014). The center 2 m of each plot was hand harvested for yield and wireworm damage assessment. For yield assessment in all trials, sweetpotatoes were sorted into categories using U.S. standards for grades of sweetpotatoes (USDA, Agricultural Marketing Service, 2005): Jumbo (>22.9 cm length and >8.9 cm diameter), U.S. No. 1 (7.6‒22.9 cm length and 4.4-8.9 cm diameter), U.S. No. 2 (≥3.8 cm diameter), and cull (storage roots of any size with off-shapes and/or damages from disease, rodents, bruises, or other means). Sweetpotato roots that were not at least 2.5 cm in diameter and 7.6 cm in length were discarded. The number and weight of harvested sweetpotatoes were recorded for each variety. Average number and marketable yield of sweetpotatoes per plant were also calculated for each treatment. Root damage by wireworm was rated using a severity index based on the number of feeding scars (no scars = 0, 1-5 scars = 1, 6-10 scars = 2, >10 scars = 4) (Schalk et al., 1993), and percentage of wireworm damaged roots (by weight and number) was calculated. Only wireworm scars greater than 1 mm in diameter were counted.   

3. Dual-Cropping for Greens and Roots Trial

This experiment was added in response to grower interest in greens production. The experiment had a split plot design with three main plot treatments and four subplot treatments, and four replications. The main plot treatment was sweetpotato breeding lines, and the subplot treatment was greens harvesting schedule. Main plot treatments included the breeding lines USDA-04-284, USDA-04-791, and USDA-04-136 obtained from USDA ARS (P.A. Wadl, Research Geneticist, Charleston, SC). These breeding lines are being evaluated in a separate study for wireworm resistance in western Washington, so additional information on how the breeding lines respond to greens harvesting will also be useful to the breeder for potential variety release. The subplot treatment included four greens harvesting schedules: no harvesting (control), early harvesting, late harvesting, and continuous harvesting (Table 1). Each subplot included 10 plants in a single row with 10-inch spacing between plants on a raised bed covered with polyethylene (PE) mulch. Plots were separated by 1.5 m, and beds were 1.8 m center-to-center. Based on soil test results, urea (46-0-0) was applied at the rate of 112 kg‧ha^-1 of nitrogen in 2023 and incorporated as the beds were formed. Drip irrigation tape was installed under the mulch. Planting was June 8 in 2023.  

Table 1. Sweetpotato greens harvesting schedule; greens were harvested for all treatments 16 weeks after transplanting, which was 1-2 days before storage root harvest.  

Plant establishment/growth. Plant establishment was measured as the total number of live plants per plot 2 and 4 weeks after transplanting. Plants with any amount of green tissue were counted as alive. Length of the longest vine was measured from the soil level to the base of the growing point for the center six plants in each plot 2 weeks after transplanting (June 21, 2023), in mid-season (August 15, 2023), and a few weeks before harvest (September 25, 2023). Percent canopy cover was measured using Image-J to analyze photos taken from 1 m above the surface of the raised bed. Canopy cover was measured at the same time as vine length was measured and included five of the center six plants. 

Greens harvest and yield. At each harvest date, the terminal 15 cm (measured from the base of the growing point) was cut from every vine that was at least 15 cm long on each plant. For subsequent harvests, vines were only re-harvested if there was at least 15 cm of new growth. Vine branches were harvested in addition to harvesting the main vine if the branches were at least 15 cm long from the branching point. All plants in each plot were harvested, but only the vines from the center eight plants were collected, counted, and weighed. On the final greens harvest date, which was 1-2 days before root harvest, the terminal 15 cm of vines were removed and counted. Then all remaining vines were cut and collected. The total vines from each plant (the terminal 15 cm and the remaining vines) were placed in a paper bag, and fresh weight was recorded within 1 hour of harvest. Bags containing the vines were dried in an oven at 60°C for up to 120 hours, until completely dry. In 2023, greens were harvested on August 2-4 (8 WAT), August 16-21 (10 WAT), August 30-31 (12 WAT), September 13-14 (14 WAT), and September 28-October 9 (16 WAT).  

Storage root yield. Harvest was October 13 in 2023, before the soil temperature reached below 13 ℃ (Brandenberger et al., 2014). The center six plants in each subplot were harvested by hand. Yield and wireworm damage were assessed for the center six plants in each subplot. Sweetpotatoes were sorted into categories using U.S. standards for grades of sweetpotatoes (USDA, Agricultural Marketing Service, 2005): Jumbo (>22.9 cm length and >8.9 cm diameter), U.S. No. 1 (7.6‒22.9 cm length and 4.4-8.9 cm diameter), U.S. No. 2 (≥3.8 cm diameter), and cull (storage roots of any size with off-shapes and/or damages from disease, rodents, bruises, or other means). Sweetpotato roots that were not at least 2.5 cm in diameter and 7.6 cm in length were discarded. The number and weight of harvested sweetpotatoes in each category were recorded for each subplot. Average number and marketable yield of sweetpotatoes per plant were also calculated for each treatment. Root damage by wireworms was rated using a severity index based on the number of feeding scars (no scars = 0, 1-5 scars = 1, 6-10 scars = 2, >10 scars = 4) (Schalk et al., 1993), and percentage of wireworm damaged roots (by weight and number) was calculated. Only wireworm scars greater than 1 mm in diameter were counted.  

On-farm trials

Two replicated and 12 observational field experiments were carried out on farms in Clallam, Jefferson, and Kitsap Counties in northwest Washington in 2023. In 2024, trials were expanded to farms in Mason, King, Skagit, Snohomish, Whatcom and Thurston Counties and included 2 replicated and 23 observational experiments. Trials will be repeated in 2025 with additional recruited participants. In 2023, participating farms were provided sweetpotato roots so that they could produce their own slips to plant. In 2024, we enlisted a local nursery to produce slips, because slip production proved to be a barrier to full trial participation in the previous year. Along with roots or slips, participating farms were provided with a draft slip production guide that was created as a part of preliminary trials (Shrestha and Miles, 2023). The on-farm replicated experiments were a randomized complete block design, and treatments included cv. Cascade (formerly PI 666141), a USDA ARS wireworm-resistant variety, and two commercial varieties, Beauregard and Covington (control treatment). The replicated experiments included three replications and 10 plants per plot with 12 inch spacing between plants. Most of the observational experiments included the varieties Cascade, Covington, and Beauregard. However, due to challenges with on-farm slip production in 2023, some observational experiments included Mahon Yam or Monaco in place of one or more of Cascade, Beauregard, and Covington. Mahon Yam slips were not produced on farms but were provided by a local grower who has had a good experience growing the variety locally and purchased the slips online (Johnny’s Selected Seeds, Winslow, Maine). In 2024 farmers were given slips of the three main varieties, Cascade, Beauregard and Covington, and other commercial varieties that were left over from on-station trials including Bayou Belle, Orleans and Monaco (for greens production only). Tables 2a and 2b provide a summary of cultivars planted at each farm and warming techniques used in 2023 and 2024. Observational experiments included a single plot of each cultivar with a minimum of 10 plants per plot and 8 to 12 inch spacing between plants (Table 3 a and b). 

Table 2a. 2023 On-farm trial locations, cultivars, mulch.  

PE = black polyethylene mulch  

Table 2b. 2024 On-farm trial locations, cultivars, mulch and other warming methods

Table 3a. 2023 on-farm trial slip production and transplanting methods.  

Table 3b. 2024 on-farm trial slip production and transplanting methods. 

na= most slips were produced centrally and distributed to farmers in 2024.

We provided growers with draft publication, Growing Sweetpotatoes in Western Washington, that was first developed during preliminary WSU NWREC trials and then updated in 2024 based on this work (Shrestha and Miles, 2022b), and we recorded how each grower planted and maintained the crop (Tables 2-4). Each grower used black polyethylene (PE) mulch or their preferred alternative (i.e. black silage tarp, weed fabric, biodegradable black film, etc.) to grow sweetpotatoes in their farm experiments (Table 2a and b). Some growers implemented additional or alternative warming methods in 2024 such as planting in high tunnels, covering rows with floating row cover, or covering greens with slitted plastic hoops (Table 2b). Due to western Washington’s cooler temperatures, farmers were interested whether planting pre-rooted slips would reduce the loss of 2-3 weeks of active growing time that slips experience when they are planted directly. In both years, some growers chose to experiment with planting pre-rooted slips (3 in 2023 and 10 in 2024). A handful of growers decided to plant and harvest later in the season to avoid the risk of frost damage in June. Data for the on-farm replicated trials was collected by WSU Extension personnel on the Olympic Peninsula. Farmers hosting an observational experiment collected their own data using a short questionnaire that was provided at the time of planting in 2023. Farmers received assistance with data collection from WSU Extension personnel in 2024 in order to obtain better data. During farm visits, the research assistant helped with weed control, recorded data about growing practices, checked-in with growers to assess needs and was present to collect harvest data on the majority of observational farms.  

Table 4a. 2023 on-farm trial field practices and harvest dates.  

WAT = weeks after transplanting 

¹The first 2-6 weeks after transplanting is the critical weed free period for sweetpotato (Nedunzhiyan et al., 1998; Seem et al., 2003; Wadl et al., 2023).  

Table 4b. 2024 on-farm trial field practices and harvest dates*. 

*Different data is displayed between 2023 and 2024 (Tables 4a and b). This will be consolidated and rectified in the final report. 

Weather data. Air temperature, relative humidity (RH), solar radiation, and rainfall data during the cropping season were collected from the WSU AgWeatherNet stations located within 10 miles of each on-farm trial. Soil temperature and moisture were recorded at 15-minute intervals with sensors (TEROS 11; Meter Environment, Pullman, WA) installed 10 cm deep in the center of the middle plot in replicate 2 of each replicated on-farm experiment.   

Plant survival and vigor. Slip survival rate was recorded as a percentage of plants alive 2-4 weeks after transplanting. Plant vigor was rated at harvest using rating system based on photos that showed examples of plants that were not vigorous, slightly vigorous, moderately vigorous, or very vigorous. Plant vigor ratings were coded as 0 (plant was dead), 1 (not vigorous), 2 (slightly vigorous), 3 (moderately vigorous), or 4 (very vigorous) for data analysis. For replicated experiments, plant vigor was recorded for each plant selected for harvest (3 plants per plot). For observational experiments, overall plant vigor was recorded the entire experiment.  

Weed Control. Weed control was rated by WSU Extension personnel during mid-season farm visits as none, insufficient, adequate, or excellent (Table 4). 

Storage root yield and wireworm damage. Harvest dates are shown in Table 4. Farmers were asked to harvest before the soil temperature reached below 13℃ (Brandenberger et al., 2014). In the replicated on-farm trials, 3 central plants from each plot were hand harvested for yield and wireworm damage assessment. If one or more of the original central plants was missing, different plants from the same plot were harvested instead. In the on-farm observational experiments, 1 to 3 representative plant(s) of each cultivar were harvested for yield and wireworm damage assessment. Sweetpotato roots less than 1 inch in diameter and 3 inches in length were discarded. The remaining total root weight was recorded for the 3 representative plants in each plot, and the number of harvested sweetpotatoes in each size category was recorded. The average number and marketable yield of sweetpotatoes per plant were calculated for each treatment. Categories were U.S. standards for grades of sweetpotatoes (USDA, Agricultural Marketing Service, 2005): Jumbo (>22.9 cm length and >8.9 cm diameter), U.S. No. 1 (7.6‒22.9 cm length and 4.4-8.9 cm diameter), U.S. No. 2 (≥3.8 cm diameter), and cull (storage roots of any size with off-shapes and/or damages from disease, rodents, bruises, or other means). Root damage by wireworm was rated using a severity index based on the number of feeding scars (no scars = 0, 1-5 scars = 1, 6-10 scars = 2, >10 scars = 4) (Schalk et al., 1993), and a wireworm severity index from 0 to 4 was calculated as a weighted average for each plot.  

Data analysis. All data were analyzed using statistical analysis software (R version 4.3.2, R Studio, Vienna, Austria). A cumulative link mixed effects model (clmm) was used to determine if weed control, wildlife interference, and growing degree days were significant predictors of plant vigor at harvest. Plant vigor at harvest was used instead of root yield due to missing data and inconsistencies with how root yield was measured and reported by farmers. Growing degree days (GDD) was calculated using air temperature data from the AgWeatherNet station closest to each farm. GDD was calculated both with a base temperature of 10 ⁰C and 15.5 ⁰C (Duque et al., 2022). GDD was used to account for planting date, harvest date, and temperatures throughout the growing season all in one metric. Weed control was rated for the first 2-6 weeks after transplanting as 0 = none, 1 = insufficient, 2 = adequate, 3 = excellent. Wildlife interference was rated to account for significant damage by rabbits or deer as 0 = no major wildlife interference, 1 = major wildlife interference. There were 20 farms in total with complete data sets in 2024.