This research was conducted at the University of Saskatchewan’s Horticultural Field Research Station and North Management Area in Saskatoon on a Dark Brown Chernozem (Sutherland Association), encompassing five site-years. For three years (2021-2023) the experiment was established on a clay loam soil (characterized by pH of 7.4, 4.9% organic matter, bulk density of 1.18 g cm^{- 3} in the 0–15 cm depth). A sandy loam site was included in 2022 (0–15 cm soil pH of 7.3, 3.7% organic matter, bulk density of 1.24 g cm^-3) and also in 2023 (0–15 cm soil pH of 7.7, 3.8% organic matter, bulk density of 1.31 g cm^-3). Prior to planting, soil total extractable N (in the 0–60 cm depth) was 77.5, 92.5, and 110.9 kg N ha^-1 at the clay loam site in 2021, 2022, and 2023, respectively; and 84.4 and 100.6 kg N ha^-1 at sandy loam site in 2022 and 2023 ( Supplementary Table 1 ). In each year, the experiment was established on an area, that in the year prior, grew barley or was otherwise fallow and had not previously grown potatoes in the past decade. Each year, the soil was prepared for production by discing the top 10 cm of the entire field area in early spring, and a pre-season herbicide [Eptam^® (PCT) and Sencor^® (metribuzin)] was applied.

To compare six different potato cultivars under a range of N fertilizer applications, 3 by 3 m plots were established and arranged in a split-plot design with cultivar as the main effect and fertilizer rate as the sub-effect. Potato cultivars selected were fresh market or processing types that are commonly grown for seed potato production on the Prairies, selected in consultation with the Saskatchewan Seed Potato Growers Association (SSPGA). All seed was certified and sourced from growers in the SSPGA. Cultivars included Clearwater Russet (russet skin, mid-season maturity), Dark Red Norland (red skin, early-season maturity), Milva (yellow skin, mid-season maturity), Poppy (red skin, mid-season maturity), Russet Burbank (russet skin, late-season maturity) and Sangre (red skin, mid-season maturity). All six potato cultivars are suitable for the fresh market, although some are dually suitable for the processing market (the Russets). Additionally, all six cultivars are regionally representative of the seed potato market. For the purposes of this research, we are considering potato production for the fresh market. For a target yield of 44 Mg ha^-1 (Government of Manitoba, 2022), soil tests conducted through AgVise Laboratories recommended average applications of 22 kg N ha^-1 at the clay loam sites and 152 kg N ha^-1 at the sandy loam sites. For our experiment, fertilizer treatments were established for 0, 10, 20, 30 and 40 kg N ha^-1 at the clay loam site (2021-2023), and 0, 50, 100, 150, and 200 kg N ha^-1 at the sandy loam sites (2022-2023). This represented a range of N application starting at zero and including rates below, similar to, and above the recommended rate for each site. Granular urea (46-0-0) was the source of N and was broadcast incorporated into the top 10 cm of each plot prior to planting. Other nutrients such as phosphorus (triple super phosphate 0-45-0) and/or potassium (soluble potash 0-0-60) fertilizer were broadcast incorporated alongside N fertilizer as needed based on soil-test recommendations. At the 2022 sandy loam site, 20 kg P ha^-1 and 48 kg K ha^-1 was applied. No additional fertilizer was applied in other site-years.

The plots were seeded May 13, 2021, May 12, 2022, and May 9, 2023, at the clay loam site, and May 17, 2022, and May 15, 2023, at the sandy loam site. Potato seed pieces were placed 10 cm deep with 75cm between-row spacing and 38 cm in-row spacing; each plot encompassed four rows. Potato plots were mechanically hilled after emergence in mid-June each year. Irrigation was applied using a travelling gun system (Micro Rain 58, Southern Irrigation, Saskatoon, SK) as needed based on soil moisture observations and the weather forecast.

Weather data was collected from the Saskatchewan Research Council Saskatoon station located at the sandy loam site (52.09227°N 106.36288°W). This weather data is representative for all sites as the station is within 3 km of all sites. The average temperature and total precipitation were monitored during the growing season of each study year (May-August). All three years of the trial were drier and warmer than the historical average ( Table 1 ). Growing degree days were calculated using a T_min of 7°C and a T_max of 30°C. The sites accumulated a total degree days of 724.3, 907.6, 961.8 at the clay loam sites (2021-2023) and 907.6 and 916.8 at the sandy loam sites in 2022 and 2023, respectively. The clay loam sites received total precipitation (rainfall + irrigation) of 198.9 mm (2021), 219.7 mm (2022), and 273.7 mm (2023). The sandy loam sites received total precipitation of 236.4 mm (2022) and 278.9 mm (2023) ( Table 1 ).

In-season N was monitored with Soil Plant Analysis Development (SPAD) readings collected every week starting after hilling (SPAD-502+ chlorophyll meter, Konic Minolta, Japan). Two readings were recorded from each plot, on the apical leaflet of the 4^th youngest leaf (Li et al., 2012). Potatoes were monitored for insects, especially the Colorado potato beetle, and insecticides [deltramethrin (applied as Decis^®) or chlorantraniliprole (Coragen^®)] were applied when needed. Plots were hand-weeded in-season to control weeds. Two weeks prior to harvest, plants were top-killed via herbicide [diquat (Reglone^®)] application.

Plots were harvested mid-September each year (Sept 14, 2021, Sept 12, 2022, and Sept 11, 2023, at the clay loam site, and Sept 13, 2022 and Sept 12, 2023 at the sandy loam site). From each plot, tubers from one 3 m center row were harvested, collected into mesh bags and immediately weighed to record fresh weight. Tubers from each bag were sorted into Grade 1, Grade 2, Undersize or Cull categories based on size, shape, and firmness (Canadian Food Inspection Agency, 2021). Total yield was calculated as the sum of the weight of tubers in each category, and yield is reported as tuber fresh weight in megagrams/metric ton (Mg ha^-1). A sub-sample of ten tubers were checked for internal defects and a sub-set of five tubers were collected, cut using a fry cutter and dried at 60°C for three weeks until no weight change was recorded to determine dry matter. Dry tubers were ground using a Thomas Wiley Laboratory Mill with a 1mm mesh screen. To determine %N, the dried and ground tuber samples were analysed by LECO (CN 628 series combustion analyser, LECO Corporation, MI, USA).

Soil samples were collected prior to planting as a composite sample of 6–8 cores (3.5 to 5 cm diameter) randomly spaced throughout the entire field at 0-15, 15-30, and 30–60 cm depths. Samples were shipped to AgVise or A&L Laboratories for soil characterization. Bulk density measurements were collected from each site using the core method (Hao et al., 2008). After potato harvest, soil samples (in 15 cm increments from 0–60 cm depth) were collected to determine soil nutrient levels separately for each potato cultivar. Soil samples were kept in a cooler while working in the field, sieved to 2 mm, and kept frozen until analysis. All samples were analyzed for extractable N (NO₃-N and NH₄-N) at the University of Saskatchewan following a KCl extraction procedure (Keeney and Nelson, 1982). Inorganic N was determined with the extract analyzed colorimetrically using a Technicon AutoAnalyzer III.

Two NUE metrics were investigated, N Balance Intensity (NBI) and tuber N uptake efficiency (NUpE). The NBI conceptualizes the difference between N removed as yield relative to the N added as fertilizer, indicating whether N outputs are balanced by N inputs, or not. Based on this balance, NBI indicates if an accumulation or a decline in soil N may be expected over a growing season. Calculation of NBI was determined as the difference between fertilizer N applied and N removed as yield (Equation 1)

where Yn denotes N contained in the yield (tuber N) (kg N ha^-1); Fn denotes the amount of fertilizer N applied (kg N ha^-1). Positive NBI values indicate that tubers removed a quantity of N equivalent to the N fertilizer applied in addition to removing more N from the indigenous soil N pool. Negative values indicate the tubers did not remove the equivalent quantity of the N fertilizer applied, rather they left a portion of fertilizer N in the soil as excess. Values closer to zero indicate a balance between N fertilizer input and tuber N removal from the system (Congreves et al., 2021).

Tuber NUpE was calculated as an index relating tuber N at harvest to the relative availability of soil N at planting (Equation 2)

where Yn denotes N contained in the yield (tuber N) (kg N ha^-1); Fn denotes the amount of fertilizer N applied (kg N ha^-1) and Sn represents the inorganic N levels (kg N ha^-1) in the spring prior to seeding. Tuber NUpE was calculated using the extractable soil N levels at pre-plant from 0–60 cm.

Statistical analyses were done using R Statistical Software (v4.2.2; R Core Team, 2022) within RStudio (version 2023.12.0; RStudio Team, 2023). R is an open-source language and environment for statistical computing, while RStudio provides an interface to facilitate visualization and analysis. The site-years were analysed separately because of the unique growing-season conditions and soil characteristics in each case; our main objective was to determine cultivar differences in each case. To test the null hypothesis that treatment groups have identical mean values, analyses of variance (ANOVA) tests were performed. Models were fit using the “agricolae” package in RStudio that considered the split plot design (de Mendiburu and Yaseen, 2020) and relies on least squares estimation. Replicate was used as the blocking term (random factor), cultivar was the main effect, and fertilizer rate was the sub-effect (both fixed factors). The model was run for various response variables including yield, tuber N, and NUE indices. Shapiro-Wilks test (p>0.05) and visual tests (qq plots) were used to assess the residuals for normality. Levene’s test and residual plots were used to assess homogeneity of variance. Residuals were normally distributed and the data met the assumption of homogeneity of variance in most cases, with the exception of the 2022 sandy loam site. Data for that site-year was either log or square-root transformed with a shift and was back transformed for tables and figures. Probability values were evaluated at a 95% confidence interval (p<0.05). Two-way interactions between the fixed effects (cultivar and fertilizer) were only significant in one case, discussed in more detail in the results and discussion. To investigate significant differences between levels of the same factor (i.e. different cultivars or fertilizer rates), post-tests (multiple means comparison) were performed using the least significant difference (LSD) test via the “agricolae” package with a 95% confidence interval (de Mendiburu and Yaseen, 2020). Linear regressions (model II regressions) were done for each site-year separately to analyze the relationship between total yield and NUE using base R (R Core Team, 2022).

Cultivar significantly influenced total yield in all site-years, whereas the two-way interaction of cultivar and fertilizer rate did not ( Table 2 ). The highest yielding cultivars varied by site and year, with Sangre and Milva having the greatest yield at the clay loam site in 2021, Poppy in 2022, and Dark Red Norland, Poppy and Sangre in 2023. At the sandy loam site, Sangre, Poppy and Dark Red Norland had top yields in 2022, and Dark Red Norland and Milva in 2023. Sangre and Dark Red Norland averaged 15.8% and 15.2% higher yields than Russet Burbank across the years. Clearwater Russet had consistently lower yields in all site-years. Similar yield differences among cultivars were also observed at another (adjacent) experiment, but one that focused on P fertilizer and P use efficiency instead of NUE (Carruthers and Congreves, 2025). Given Saskatchewan’s short growing season, potatoes were harvested in September regardless of cultivar. While staggering cultivar harvest (as would normally be done by a grower) might have reduced the gap between cultivar yields or tuber N, the cool fall climate would likely have limited any substantial benefits from extended growth periods. Regardless, yield differences between cultivars of similar market class and maturity emerged, demonstrating variation in yield. To place our yield results in context of expected potato yield in Canada, the average total tuber yields across all cultivars were 36 Mg ha^-1, 33.6 Mg ha^-1, and 51.8 Mg ha^-1 (clay loam 2021, 2022, and 2023, respectively), and 17.8 Mg ha^-1 and 43.3 Mg ha^-1 (sandy loam 2022 and 2023) ( Table 2 ). The yields produced in our study are comparable to the national average of 36.9 Mg ha^-1 (Agriculture and Agri-Food Canada, 2021), with the exception of the 2022 sandy loam site-year which had significantly lower yield likely due to moisture stress and a compounding effect of drought in previous years ( Table 1 ).

Cultivar influenced tuber N in four of five site-years, while the two-way interaction of cultivar and fertilizer rate was negligible ( Table 3 ). Only at the clay loam site in 2021 was there no significant differences between cultivars in terms of tuber N content. In other years at the clay loam site, the cultivars with the highest tuber N were Dark Red Norland and Poppy (2022), and Dark Red Norland and Sangre (2023) ( Table 3 ). At the sandy loam site in 2022, Clearwater Russet and Milva had significantly lower tuber N relative to other cultivars which yielded similarly, possibly suggesting the ability of these cultivars to take up N was impacted more by water availability relative to other cultivars. Overall, Dark Red Norland, Sangre, and Poppy had consistently high tuber N content throughout this study, with tuber N content 8.18-44.45 kg N ha^-1 greater than other cultivars. As long as nitrate accumulation in tubers is not excessive (to the point of potential health impacts), then these cultivars may provide an agronomic and environmental benefit by removing more N from the field as harvested tubers, thereby leaving less residual soil N in the field, vulnerable to losses over the subsequent winter and spring. In the current study, average tuber N content was 115 kg N ha^-1, 134 kg N ha^-1, and 140 kg N ha^-1 for clay loam sites in 2021, 2022, and 2023, respectively. At the sandy loam site, average tuber N was 78.1 kg N ha^-1 in 2022 and 135 kg N ha^-1 in 2023 ( Table 3 ). These tuber N values are in line with other reported findings (Haase et al., 2007; Rens et al., 2016; Waddell et al., 1999; Zebarth et al., 2012).

Cultivar influenced NBI at four of the five site-years ( Table 4 ). NBI was not influenced by the two-way interaction of cultivar and fertilizer rate, except for the 2022 sandy loam site-year ( Table 4 ). In this case, the interaction was caused by minor differences in cultivar performance across fertilizer rate, but the overall trend was decreasing NBI with increasing fertilizer rate, thus the interaction was not explored further. All cultivars had positive NBI values in all site-years, except for the 2022 sandy loam site (the site-year with the driest conditions). Positive NBI values from most years indicates that tubers removed a quantity of N equivalent to the amount of N fertilizer applied in addition to acquiring more N from the soil. However, the negative NBI values at the 2022 sandy loam site indicates that cultivars left more fertilizer N behind in the soil than put towards yield - all except Dark Red Norland. Overall, Dark Red Norland resulted in the greatest NBI values, up to 1.5–5 times greater than the other cultivars, depending on the site-year ( Table 4 ), demonstrating potential for better NUE across different years and different environmental conditions. There was sufficient variation in NUE between cultivars of the same market class and therefore further exploration of these differences could allow for cultivar selection as a sustainable management tool.

Tuber NUpE was influenced by cultivar at four of the five site-years ( Table 5 ). Tuber NUpE ranged from 61-240% in the clay loam site-years, and from 7.82-227% in the sandy loam site-years. Such high values indicate a significant amount of N was mineralized from the soil over the course of the growing season, enabling the tubers to take up more than 100% of the quantity of N that was present at planting. Lower tuber NUpE values at the sandy loam site were caused by low tuber N in 2022 due to drought that limited plant N uptake paired with the higher N fertilizer rates at this site. There was no cultivar difference in NUpE at the clay loam site in 2021, but Sangre numerically recovered the most N. Dark Red Norland, Poppy and Sangre had the highest NUpE in this study, outperforming the other cultivars by 22.5-33.2%.

Significant variation in NUE driven by cultivar has been widely reported (Cohan et al., 2018; Getahun et al., 2022, 2020; Jones et al., 2021; Ospina et al., 2014). In the current study, the improved NUE observed in Dark Red Norland, Sangre and Poppy was primarily driven by yield differences among cultivars - specifically the ability of those cultivars to produce high yields, which in turn increased N contents of tubers. The relationship between NUE and total yield was assessed using linear regression ( Figure 1 ). Despite yield differences across site-years (driven by environmental factors), there was a significant (p<0.001) positive linear trend between average total yield and NUE metrics. Given that soil N was not limiting, NUE was primarily driven by yield differences among cultivars and site-year yield production. When investigating the relationship between yield and NBI at clay loam site-years, R² was found to be 0.862, 0.865, and 0.812 in 2021, 2022, and 2023, respectively. Similarly, the relationship between yield and NUpE showed R² of 0.866, 0.865, and 0.811 for those same years, indicating a consistent and strong relationship between yield and NUE on clay loam soils. At the sandy loam site, a strong relationship was observed between yield and NBI in 2023 (R² = 0.865), which was comparable to the clay loam site-years, while the NUpE relationship was slightly weaker (R² = 0.794). However, the 2022 sandy loam site displayed the weakest relationship for both metrics (R² = 0.681 for NBI and R² = 0.753 for NUpE). This suggests that at the 2022 sandy loam site, environmental factors other than N availability (like water limitation or heat stress) may have had a stronger influence on yield. This could have resulted in a weaker relationship between yield and NUE relative to other site-years. The low yields at the sandy loam site in 2022 translated into lower tuber N removal, which may also explain why the regression for that site-year is distinct from other years. Despite differences in site-years, a few consistent trends emerged: Clearwater Russet produced consistently low yield and NUE, while Dark Red Norland, Sangre, and Poppy had high yields and NUE ( Figure 1 ). These findings highlight a win-win situation for producers: selecting high yielding cultivars also leads to greater NUE. This creates both economic and environmental benefits with increased yields generating greater profits, and lower fertilizer requirements reducing N lost to the environment.

Looking to the N translocation patterns of the cultivars, it appears that in addition to yield potential, cultivars may have different mechanisms driving increased NUE. Dark Red Norland had a different pattern of N translocation over the growing season relative to other cultivars, shown by SPAD values over time ( Figure 2 ). Due to its early-season maturity, Dark Red Norland relocated N from aboveground biomass to tubers earlier in the season relative to other cultivars, which could explain the higher NUE, especially under the driest conditions at the sandy loam site in 2022. Thus, earlier-maturing cultivars may have stronger performance under short growing seasons or harsh environmental conditions due to their more advanced stage when drought stress occurs. Some advantages of earlier cultivars were uncovered in a greenhouse study that found higher photosynthetic capacity, starch accumulation, and final yield (Mokrani et al., 2018). However, other research has found late-maturing cultivars to have higher NUE than early cultivars (Getahun et al., 2020). Climate plays a substantial role in these findings, as later-maturing cultivars may achieve greater yield and NUE, but only if the length of the growing season allows for it.

Poppy and Sangre (both mid-season maturity) generally maintained the highest SPAD values across the growing season. The N content in leaves did not plummet as sharply as Dark Red Norland, suggesting a different mechanism for increased NUE compared to Dark Red Norland. One explanation for this could be the dark leaf color of Poppy and Sangre observed in the field, suggesting these cultivars might have an increased capacity for N in vegetative tissues during the growing season, compared to other cultivars. In theory, higher N content in leaves could increase total N translocated to tuber later in the growing season. However, more research would need to be done to link improvements in NUE to N translocation during the growing season. Quantitative trait loci (QTL) studies have found multi-trait genomic regions with associations with potato NUE on chromosomes III, V, and VI (Getahun et al., 2022), but this research also highlighted the importance of growing season conditions on NUE. Further research to determine specific phenotypic traits linked to NUE and their interaction with environment is critical.

The cultivar differences observed in our experiment may not necessarily indicate a superiority in specifically acquiring or utilizing N but instead may reflect superiority in other aspects (photosynthetic efficiency, water use efficiency, early-season vigor, early maturity, etc.) which then translate into using more N, when soil N is readily available for uptake. It is possible that earlier-maturing cultivars are better able to move more N to tubers before harvest than later-maturing cultivars. Fertilizer considerations are also important, as high applications or split applications can promote extended vegetative growth which can delay N translocation to tubers. Overall, Dark Red Norland and Sangre consistently performed equal to or better than other cultivars across yield and NUE metrics. Hence, the ability of a plant to produce high yielding tubers also corresponds with greater NUE. In other research, the development of “Alturas” - a late-maturing russet that requires 40% less fertilizer than Russet Burbank, showed potential for improvements in NUE through cultivar (Novy et al., 2003). Our findings demonstrate sufficient variation to prompt further investigation into NUE mechanisms in potato, and that cultivar selection could allow producers to reduce fertilizer rates and improve the N balance of potato production systems. Healthy and vigorous plants that produce high biomass tubers not only represent agronomic yield benefits, but also present environment benefits in the form of NUE.

Fertilizer rate did not significantly impact total yield or tuber N content in any site-year ( Tables 2 , 3 ). This highlights the potential to reduce or omit N application without yield penalties, and our findings align with prior studies. Research in Manitoba investigated potato yields at 80% of the recommended N rate (ranged from 119–168 kg N ha^-1) and found that fertilizer recommendations over-estimated crop N needs and could be reduced (Gao et al., 2018). In evaluating 15 site-years across Canada, there was no yield response at N rates above 60 kg N ha^-1 for four of those site-years, and no yield response above 120 kg N ha^-1 for another five site-years (Nyiraneza et al., 2021). Research shows that typical N fertilizer recommendations for potato are excessively high, and could be reduced without yield reductions in many cases (Gao et al., 2018; Nyiraneza et al., 2021).

Excessive fertilizer rates are known to reduce the NUE of cropping systems (Ruža et al., 2013). Our findings confirm this, with both indicators of NUE: NBI and NUpE, being influenced by fertilizer rate alone, rather than by interactions of cultivar and fertilizer rate ( Tables 4 , 5 ). Consistently, the 0 kg N ha^-1 treatment led to the highest NBI values, while increases in fertilizer generally led to incrementally lower values ( Table 3 ). Likewise, tuber NUpE generally decreased with greater N applications ( Table 5 ). These results suggest lower fertilizer rates (0–10 kg N ha^-1 at the clay loam site, or 0–50 kg N ha^-1 at the sandy loam sites) would be sufficient for potato production, while minimizing residual N buildup that could increase risk of N loss and input costs for producers.

The lack of yield response to fertilizer points to the importance of soil N mineralization during the growing season. Prior to planting, total extractable N (0–60 cm) at the clay loam site was 77.5, 92.5, 110.9 kg N ha^-1 for 2021, 2022, and 2023, respectively. At the sandy loam site, pre-plant total extractable N (0–60 cm) was 84.4 and 100.6 kg N ha^-1 for 2022 and 2023. These levels are below the typical crop N demand for potato production, which ranges between 125-225 kg N ha^-1 on the Canadian Prairies (Gao et al., 2018; Government of Saskatchewan). However, the lack of yield response to fertilizer is likely explained by significant amounts of soil N mineralized during the growing season.

Growing season N mineralization was estimated by subtracting the pre-plant soil N levels (0–60 cm) from the amount of N in the soil (0–60 cm) plus the plant N contained in the unfertilized controls after harvest. At the clay loam site, estimated soil N mineralization was 55, 89, and 304 kg N ha^-1 for 2021, 2022, and 2023, respectively. At the sandy loam site, soil N mineralization varied, with only 17 kg N ha^-1 in 2022 compared to 154 kg N ha^-1 in 2023. This demonstrates a moderate to substantial contribution to crop N needs depending on the site-year. Moisture was likely a driving factor behind mineralization, with dry conditions at the 2022 sandy loam site-year resulting in the lowest mineralization, yields and tuber N uptake. In comparison, improved moisture conditions in other years (especially 2023) resulted in significantly higher mineralization, yields and tuber N uptake. Thus, environmental factors have a critical role in facilitating soil N mineralization and subsequent crop N uptake. It is important to note that total rainfall plus irrigation during the growing season at the sites was still generally low (182–280 mm) in comparison to typical potato production systems that target 25 mm per week (Government of Saskatchewan). Therefore, the moisture provided in other potato fields could be sufficient to promote significant growing season N mineralization.

This points to the importance of determining mineralization potential of soils, to avoid over-application of fertilizers on sites that may be able to support high yields without additional N. Therefore, accounting for mineralizable soil N is a critical link that many producers may be missing. However, research still needs to determine a standardized protocol to estimate soil potentially mineralizable N (Clark et al., 2020; Ros et al., 2011), let alone to determine an accurate projection. Large differences in results are found between various methodologies and the length of incubations can vary significantly (7–210 days) (Ros et al., 2011). Advancement in this area could reduce the trade-offs which include processing time, cost, and accuracy (Clark et al., 2020). Fast and effective mineralizable N testing would further enhance the toolbox producers can utilize to determine appropriate fertilizer rates to optimize input costs and NUE.

This research confirms that there is sufficient cultivar variation in potato to create differences in yield and NUE. Cultivar selection can therefore be a sustainable management tool, assuming future research will work to identify top performing cultivars in different market or maturity classes. While we identified cultivar-based differences, there was a strong relationship between yield and NUE, and the implications from this are positive for producers: selecting a high yielding cultivar also confers improved NUE. The impact of environment and climate is one limitation, as growing season can restrict later-maturing cultivars from reaching their full potential; however cultivars should be selected with the climate in mind. Additionally, there is more risk than reward to high fertilizer rates, especially if growing an improved NUE cultivar or on soils with repeated historical application of N and high mineralization potential that could satisfy potato N requirements. It is clear improvements in soil testing are needed to give producers a quick decision tool that allows for reductions in fertilizer rates. Using these strategies together will improve sustainability of potato production.