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

ON-STATION TRIALS

1. Wireworm Resistant Cultivar Trial

Environmental conditions. Across all three years, the soil temperature in the raised bed covered with black plastic mulch was on average 4.1 °C warmer than the air temperature. In general, air and soil temperatures throughout the growing season were 0.2–0.6 °C lower in 2024 than in 2023 and 2025. 

Plant survival. While plant survival differed by year at 2 WAT (P < 0.001), 4 WAT (P = 0.004), and 6 WAT (P = 0.001), plant survival was high for all cultivars in all years (all ≥ 90%).

Longest vine length. According to the International Potato Center’s sweetpotato descriptors, sweetpotatoes can be classified into four plant types based on longest vine length at the end of the growing season: erect <75 cm, semi-erect 75–150 cm, spreading 151–250 cm, and extremely spreading >250 cm (Huamén 1991). Based on these classifications and the 3-year average longest vine lengths at 16 WAT, cv. Monaco has erect growth habit, breeding lines USDA-04-136 and USDA-04-284 are semi-erect, Beauregard, Cascade, Covington, Orleans, and USDA-04-791 are spreading, and Bayou Belle is extremely spreading. 

Canopy cover. Canopy cover differed by year (P < 0.001) and cultivar (P ≤ 0.028) on all three measurement dates, except for cultivar at 16 WAT in 2023 (P = 0.091). At 9 WAT, average canopy cover across cultivars was 92%, 88%, and 76% in 2023, 2024, and 2025, respectively, and the range for all three years was 33–96%. At 16 WAT, average canopy cover across cultivars was 90%, 87%, and 96% in 2023, 2024, and 2025, respectively, and the range for all three years was 79–97%. Differences in canopy cover between cultivars were greatest at 9 WAT, and all three years was lowest for USDA-04-136 (66% on average), and for USDA-04-791 and cv. Covington in two out of three years (78% and 85% on average across all years, respectively). All other cultivars tended to have similarly high canopy cover at 9 WAT (90% on average). Canopy cover decreased slightly between 9 and 16 WAT for most cultivars in 2023 and 2024, likely due to shading of nearby leaves causing the plants to shed leaves (Ravi and Saravanan 2012). However, in 2025, canopy cover increased for all cultivars between 9 and 16 WAT, indicating vegetative growth continued after 9 WAT in 2025. 

Storage root yield. Storage root yield differed by year (P ≤ 0.001) and cultivar (P ≤ 0.048) for all grades, except for jumbo in 2023 (P = 0.075; in 2024 and 2025 there were no jumbo roots produced) and US no. 2 in 2025 (P = 0.127). Total yield averaged 31.2, 16.9, and 18.4 t‧ha^-1 across all cultivars in 2023, 2024, and 2025, respectively, while marketable yield (total yield minus cull; wireworm damage was not considered) averaged 25.3, 11.2, and 14.4 t‧ha^-1, respectively. The highest total yield was produced by cv. Bayou Belle and Orleans in 2023 (45.3 t‧ha^-1 on average), Beauregard and Bayou Belle in 2024 (26.8 t‧ha^-1 on average), and Bayou Belle in 2025 (32.2 t‧ha^-1). The highest marketable yield was produced by cv. Bayou Belle and Orleans in 2023 (37.8 t‧ha^-1 on average), Beauregard, Orleans, Bayou Belle, and Covington in 2024 (17.6 t‧ha^-1 on average), and Bayou Belle and Covington in 2025 (24.5 t‧ha^-1 on average). US no. 1 yield across all cultivars averaged 19.7, 8.4, and 11.8 t‧ha^-1 in 2023, 2024, and 2025, respectively. The highest yield of US no. 1 roots was produced by cv. Bayou Belle in 2023 (34.3 t‧ha^-1), Beauregard and Orleans in 2024 (16.2 t‧ha^-1 on average), and Bayou Belle, Beauregard, Cascade, Covington, Orleans, Monaco, and USDA-04-284 in 2025 (14.2 t‧ha^-1 on average). 

Average root weight (kg) differed by year (P < 0.001) and cultivar (P < 0.001). Average root weight across all cultivars averaged 0.23, 0.18, and 0.20 kg in 2023, 2024, and 2025, respectively. Average root weight was highest for cv. Beauregard in all three years (0.29 kg on average), Orleans and Covington in two of three years (0.25 kg on average across all three years), and Bayou Belle and USDA-04-791 in 2023 (0.20 kg on average across all three years). Average root weight was lowest for cv. Monaco, Cascade, and USDA-04-284 in all three years (0.15 kg on average), and USDA-04-136 and USDA-04-791 in two of three years (0.18 kg on average across all three years). 

Wireworm damage. WDS severity index differed by year (P < 0.001) and cultivar (P < 0.001). WDS severity index across all cultivars averaged 2.33, 1.98, and 2.46 for 2023, 2024, and 2025, respectively. WDS severity index was highest (indicating susceptibility to wireworm) for cv. Beauregard, Covington, and Orleans in all three years (3.26 on average), and USDA-04-284 in 2025 (2.72 on average across all three years). WDS severity index was lowest (indicating resistance) for USDA-04-136 and USDA-04-791 in all three years (0.94 on average), and cv. Cascade in 2024 (1.51 on average across all three years). Cv. Cascade had significantly lower average WDS severity index (1.51) than the commercial standard cv. Beauregard (3.19) and Covington (3.17) in all three years. Cv. Bayou Belle (2.19) and Monaco (2.21) had significantly lower average WDS severity index than Beauregard and Covington in all three years, though higher than Cascade in two of three years.

Sweetpotato shows great promise as a new crop in northwest Washington and yield potential can be equal to or greater than the national average for some cultivars. However, wireworm resistant sweetpotato cultivars and advanced breeding lines have lower yield than susceptible cultivars. Farmers can successfully grow wireworm-susceptible cv. Beauregard, Covington, and Orleans in fields with no wireworm pressure, Bayou Belle in fields with low pressure, and Cascade in fields with high pressure. Future research should continue to evaluate new sweetpotato cultivars and advanced breeding lines for wireworm resistance and higher yield in northwest Washington, especially for organic growers who lack other control options.

2. In-Row Plant Spacing Trial

We found that 20 or 25 cm in-row spacing is optimal for sweetpotato grown in black plastic mulch in northwest Washington, due to overall increases in production of US no. 1, marketable, and total yield per hectare. In a year with favorable environmental conditions, average root size and total root weight per plant increases with wider in-row spacing though root number within each grade remains the same; thus, slightly larger roots result in higher yield per plant. These differences in per-plant yield are offset by increased plant density, resulting in higher yield per hectare at narrower in-row spacing. In a cooler year, though root yield per plant also increased with wider in-row spacing, no differences were observed between spacings in yield per hectare.

Farmers may prefer 25 cm spacing over 20 cm spacing to reduce slip costs and transplanting labor per hectare. For example, farmers trialing sweetpotatoes in northwest Washington have reported organic slip prices of around $200 per 1000 slips. This corresponds to slip costs of $5,400, $4,300, and $3,600 per hectare for 20, 25, and 30 cm in-row spacing, respectively. Farmers in northwest Washington have reported selling sweetpotatoes to local direct, wholesale, and processing markets for $4.40 to $11.00 per kg ($2.00 to $5.00 per lb), with $6.60 per kg ($3.00 per lb) being the most common price. Using the marketable root yields in this study, gross returns at a price of $6.60 per kg would be $81,800, $81,800, and $71,300 per hectare for 20, 25, and 30 cm in-row spacing, respectively. Subtracting the slip costs, this results in a gross income (not including other costs such as labor) of $76,400, $77,100, and $67,700 per hectare for 20, 25, and 30 cm in-row spacing, respectively. When growing sweetpotatoes in black plastic mulch, weeds are primarily present in the planting holes, so a slightly wider in-row spacing may also reduce weeding labor needed per hectare. Thus, we recommend an in-row spacing of 25 cm (10 inches) for sweetpotato in northwest Washington. 

3. Dual-Cropping for Greens and Roots Trial

Based on requests from producers, we added a trial at NWREC to measure the impact of harvesting sweetpotato greens on root yield. Sweetpotato greens are commonly consumed in West Africa, the American Southeast and other locations where sweetpotatoes are widely grown. Sweetpotato greens are high in many nutrients, including carotene, calcium, and iron (Ishiguro et al. 2004) and can be cooked and consumed like other greens such as spinach and chard. 

Greens yield. Sweetpotato greens yield (fresh weight) differed by year (P ≤ 0.004), and overall was greater in 2023 than in 2024 (Table 3). Total greens yield was greater in the early and continuous harvest treatments in 2023 (4.85 t‧ha-1 on average) and lowest in the late and mid-late harvest treatment (3.59 t‧ha-1 on average) (P = 0.002), but there were no differences in 2024.

Root yield. Root yield differed by year for all grades (P < 0.001) except US no. 1 (P = 0.270), and was greater in 2023 for all grades except jumbo (total yield 12.1 t‧ha-1, marketable yield 7.3 t‧ha-1) than in 2024 (total yield 6.4 t‧ha-1, marketable yield 5.3 t‧ha-1). There were differences due to greens harvest treatment for US no. 1 and marketable yield both years (P  ≤ 0.012), and both were greater for the late harvest treatment (US no. 1 yield 7.9 t‧ha-1, marketable yield 10.6 t‧ha-1, averaged across years) than for the other greens harvest treatments (US no. 1 yield 3.2 t‧ha-1, marketable yield 4.8 t‧ha-1, averaged across years and treatments). In 2023, US no. 2 and total yield were highest for the late harvest treatment (US no. 2 yield 4.2 t‧ha-1, total yield 19.6 t‧ha-1), intermediate for the mid-late harvest treatment (US no. 2 yield 2.9 t‧ha-1, total yield 11.9 t‧ha-1), and lowest for the early and continuous harvest treatments (US no. 2 yield 2.1 t‧ha-1, total yield 8.5 t‧ha-1 on average) (P < 0.001); in 2024 there were no differences. In both years, US no. 1, marketable, and total yield were lowest for USDA-04-136, and were highest for USDA-04-284 in 2023 but for USDA-04-791 in 2024 (P < 0.001).

Economic returns. In both years, based on prices common in this region for greens ($3.00 per bunch) and roots ($6.60 per kg, $3.00 per lb), total gross economic returns in 2023 were higher for the late harvest treatment ($139,000‧ha-1) than the early and mid-late harvest treatments ($110,000‧ha-1 on average) (P = 0.025), while in 2024 there were no differences ($83,000‧ha-1 on average across all treatments) (P = 0.332). In 2023, the late and continuous harvest treatments produced similar total economic returns, suggesting that the reduction in root yield in the continuous harvest treatment was offset by the increase in greens production.

Growers might consider harvesting vine tips from sweetpotato plants a few days before storage root harvest to produce sweetpotato greens with no impact on storage root yield. At the end of the season when the plants are harvested, the entire vines can be used as feed for livestock. 

4. Growing Degree Day Trial

Roots increase in bulk more slowly in western Washington due to the climate, resulting in a high number of fingerling roots, but all root grades except for culls are sellable for direct market growers. When considering that marketable roots in western Washington include US no. 1, US no. 2, jumbo, and fingerling, this study found an average marketable yield at 18 WAT of 23,505 kg^.ha^-1 and 14,971 kg^.ha^-1 for cv. Covington and Cascade with PE mulch, respectively, and 6,484 kg^.ha^-1 and 6,120 kg^.ha^-1 for Covington and Cascade without mulch, respectively.  PE mulch was essential for high yield, but as some growers in western Washington are resistant to single-use plastic, alternative soil warming methods for sweetpotato should be assessed.

In this study, few sweetpotato storage roots were formed at 8 WAT and 10 WAT. At 12 WAT, there was marketable yield for all treatments, suggesting that at least a minimum of ~556 air GDD with 10 °C T_b (~133 air GDD with 15.5 °C T_b) are needed to produce a marketable sweetpotato crop. Further GDD accumulation beyond this minimum is needed for adequate marketable yield. For example, in this study, 5,491 kg^.ha^-1 was attained at 692 air GDD at 16 WAT with 10 °C T_b in 2024 and 5,127 kg^.ha^-1 was attained at 696 air GDD at 14 WAT with 10 °C T_b in 2025. Further research is needed to determine the GDD baseline to attain adequate yield based on the economics of sweetpotato in western Washington.

Marketable and total yield increased at each subsequent harvest date with additional GDD accumulation (P < 0.001). In this study, the greatest total and marketable yield were at 18 WAT, with very few jumbo roots produced (5% of total yield, overall). This suggests that if temperature conditions had been conducive for extending harvest, additional yield would likely have been achieved at a later harvest date.

Most studies report GDD for sweetpotato using 15.5 °C T_b (Duque et al. 2022; Stoddard and Weir 2002; Villordon et al. 2009a; Villordon et al. 2009b; Villordon et al. 2010a; Villordon et al. 2010b; Wees et al. 2015; Wees et al. 2016; Weir and Stoddard 2001), and only a few studies use 10 °C T_b (Duque et al. 2022; Rao et al. 2023; Wees et al. 2015). In this study, GDD were calculated for both 15.5 °C T_b and 10 °C T_b, and our results indicate 10 °C T_b is better suited for sweetpotato GDD calculations for western Washington as growth continued between 10 and 15 °C.

ON-FARM TRIALS

1. Wireworm Resistant Cultivar Trials

Across both farms over the three years of trials, there was no statistically significant difference in yield between varieties. This was likely due complications caused by crop damage and the learning process experienced by farmers who are growing this crop for the first time. Cultivar data from the on-station trials is more reliable for side-by-side assessment of cultivar performance. 

On farm no. 1, Cascade exhibited the highest average yield per plant in 2023 but was the lowest yielding variety in 2024 and 2025. Overall sweetpotato yield was highest in 2024 (similar to data from the on-station trials). The improvement in yield between 2023 and 2024 was likely due improvements in crop management and a frost event that occurred 2 weeks after planting in 2023. Improvements in crop management included more consistent watering, better weed control early in the season, and higher soil temperature due to plastic mulch being more tightly tucked around the beds. This farm reported being able to cure and market majority of their harvest from 2024 and 2025 via a virtual farm stand, local farmer’s markets and grocery stores.  

The second farm only participated for two years. Heavy deer grazing on vines in first two months after transplanting in both 2023 and 2024 severely stunted root development. In the second season, slitted plastic low tunnels were set up over the plants after initial grazing damage. In addition to protecting sweetpotato greens from further damage, the added warmth and protection from wind helped plants start to recover from earlier damage. Had the tunnels been placed over the crop sooner after damage occurred, we believe the harvest could have recovered enough to produce a saleable crop. 

These two farms experienced little wireworm damage on sweetpotato roots at harvest. At farm no. 1, the most severe damage was only 1.5 on a scale from 0 to 4 on the wireworm severity index in the first trial year. By 2024 and 2025, wireworm severity was averaging around 0.2 on the scale. While wireworm damage was minimal in the replicated trials, damage caused by voles at farm no. 1 increased over the three growing seasons. Interestingly, a farmer in western Oregon that had tried to grow sweetpotatoes for several years reported stopping after they felt that the crop increased vole populations on their farm. More investigation into vole management is warranted. Farm number 2 ultimately dropped out of the trial because deer pressure could not be resolved. This highlighted a need to identify additional pest management tools for growers in our sweetpotato growing guide publications. 

2. Unrooted vs. Rooted Slips Trial

Western Washington’s short cooler growing season led to interest in using pre-rooting slips to maximize active growing time. Unrooted slips typically experience a 2-3 week delay between planting and initiation of new vine growth in western Washington. In all three years, some growers experiment with pre-rooting the slips they received (3 in 2023, 8 in 2024, 19 in 2025). Replicated trials were held in 2024 and 2025. Data did not show a significant difference in average root yield between unrooted and rooted slips in 2024 (p-value > 0.05), however 2025 harvest data did show a difference, (p < 0.001). Growers reported that rooted slips experienced less shock and initiated new vine growth sooner after planting than unrooted slips. Due to evidence that pre-rooted slips were more prone to producing irregularly shaped sweetpotatoes than unrooted slips, we trialed different pre-rooting methods in the greenhouse to minimize root binding and found that rooting many slips in a larger container (as opposed to individual pots/cells) in potting soil for only 1-2 weeks before planting was time and space efficient and avoided roots getting over grown before planting. In addition to pre-rooting slips, a handful of growers preferred planting and harvest later in the season to avoid the risk of frost damage often experienced in June and/or early October. More research on pre-rooting is warranted.

3. Observational Cultivar Trial

Participation in observational on-farm trials increased each year of the project. Eleven farms participated in 2023 and by 2025 there were 34 participants from 15 counties. A subset of 20 farms that provided the most complete data in 2024 and 2025 was analyzed using a multiple linear regression to test whether soil type, soil mulch, wildlife interference, or wireworm damage were significant predictors of average sweetpotato root yield (lbs). The overall regression was statistically significant (R2 = 0.7, F (8, 45) = 12.83, p < 0.0001). Both plastic mulch (β = 1.89, p < 0.0001) and alternative soil warming mulch (β = 2.24, p < 0.0001) significantly predicted higher root yields. This result mirrored the on-station trials that indicated that use of black PE mulch to warm soil improves yields in the region. More importantly for small farms, this model suggests that recycled black silage tarp or weed fabric, are viable alternatives to single use PE mulch. Growers who did not use plastic mulch or warming alternatives, such as high tunnels or caterpillar tunnels, reported poor yields over the three years of trials. 

The level of interest from farmers and gardeners in western Washington in this project, and the unique requirements required for growing sweetpotatoes successfully, indicated that availability of locally focused growing guides and locally available slips from sweetpotato varieties that are demonstrated to produce well in the region were essential tools for producers to consider attempting this new crop. Farm visits allowed researchers to collect more consistent and complete data, assess grower needs, provide technical assistance, and assist with early weed management, which was found to be a significant barrier for successful sweetpotato production in the first year of the trial. The digital survey format used to collect data in 2025 was more accessible and led to better response.

Variety trials: While none of the five varieties grown on farms exhibited a statistically difference in yield over the three years of study, 2025 questionnaire results indicated that Bayou Belle was rated slightly more often by growers the highest yielding variety. Questionnaire results didn’t indicate a consistently favored cultivar among growers. This was likely due to two factors- farmers reported cultivar preferences based on diverse factors such as flavor, skin color, yield, and storage life, and most trial participants were new to growing sweetpotatoes and may not have had enough growing time/experience to identify favorites. Farms greater than 1 acre (considered large for this project) often reported most dissatisfaction in Cascade yield. They reported the shape and size of cascade roots were not as marketable as the other cultivars. 

Wireworm damage: Incidences of wireworm damage was low among participants and didn’t yield significant data. The handful of growers that reported higher wireworm presence in their fields did not produce well due to other factors such as not using soil warming mulch, deer grazing, or cold summer temperatures. These complicating factors made it difficult to assess resistance of different sweetpotato cultivars to wireworms in an on-farm trial environment, supporting the usefulness of partnering with a research facility. 

Wildlife interference: A one-way ANOVA was conducted to examine the effects of wildlife interference on plant root yield (lbs) for the subset of 20 farms that provided the most complete data. There was a statistically significant difference between the levels of wildlife interference on root yield (F(2, 264) = 17.13, p <0.0001). Tukey’s HSD post hoc tests were carried out. The mean root yield for plants that experienced moderate wildlife interference was significantly higher than the average root yield for plants that experienced severe wildlife interference as well as plants with no interference from wildlife (p <0.0001). It is likely that root yield on average was higher for plants with moderate wildlife impacts than no interference, because of the large amount of vole damage observed in the 2024 and 2025 season. Farms that did experience losses from voles were often also using plastic or landscape fabric. The use of soil warming mulch results in larger roots, and the mulch itself provides voles protection from predators, therefore creating conditions that may be more attractive to rodents. Many of the farms reporting no wildlife activity were also growers that chose not to use plastic mulch, suggesting their root harvest was lower due to other variables and conditions.  

Voles were the most common wildlife pest damaging sweetpotato roots. Over 35% of growers reported rodent damage in 2024 and 2025. The second most common source of wildlife impact was deer. Several farms experienced deer grazing that started in the early weeks after planting each year. This led to at least 3 growers dropping out of the trial completely due to the inability to manage deer damage. Several growers who had experienced damage from deer or rabbits in 2023 used floating row cover successfully as a wildlife deterrent to protect greens in subsequent seasons. Both floating row covers and slitted plastic low tunnels were observed to reduce grazing damage from deer and rabbits in these trials. We updated the sweetpotato growing guides to include more specific information from growers on protecting sweetpotatoes from deer and rabbits and added more information about voles.  

Soil Type: Sweetpotatoes grow best in well drained and fertile loam, sandy-loam or clayey-loam soils (Mukhopadhyay et al. 2011). In our 2024 trials, more than 70% of the participating farms had soil with high clay content. Heavy clay soil restricts the development of sweetpotato roots, causing shape irregularity that leads to lower yields and breakage during harvest (Mukhopadhyay et al. 2011). Clay soils are often cooler and hold more water, leading to physical deformities such as corky root when harvest was delayed too late in the fall after rain had fallen. In 2025, we started providing more information about growing sweetpotatoes in clay soil, such as planting in raised beds, amending soil with compost, and adjusting harvest times to fall before significant rain fall. Only 11 out of the 34 farmers in 2025 reported planting in soil with high clay content. We suspect that some growers dropped out because of difficulties with clay soil.

In our first year of study, farmers in the observational and replicated trials encountered problems with processes that are specific to sweetpotatoes: slip production and curing. Both processes require heat and humidity conditions not required for any other crops that are grown locally. Slip production problems were remedied in 2024 by producing slips centrally and distributing them to farms instead of asking farmers to produce their own. Farmers continued to struggle with the curing process after harvest in 2024. Only properly cured sweetpotatoes can be stored successfully. Ideal curing conditions require maintaining a temperature of 80 - 90 °F and 80 - 90% relative humidity for 1 week. At lower but adequate temperatures, the curing process can still be successful but takes longer. Some participating farmers attempted to cure sweetpotatoes in unheated greenhouses, a method that is successful for other crops. However, sweetpotato harvest occurs in the fall when nighttime temperatures may fall below 50°F, and farmers who attempted to cure them in an unheated greenhouse experienced complete or near complete loss of the crop. Temperatures below 50°F, even sporadically, causes cold damage to roots that manifests later as a shortened storage life. We offered curing and storage educational workshops in 2025 as well as direct technical assistance. These tools greatly improve curing success in 2025.

Over the three years of trials, we incorporated the feedback and experiences of participating growers in a series of “Growing Sweetpotatoes in Western Washington" fact sheets included in this report.