Can rainfall explain the variability of sugarbeet root and sugar yield across North Dakota and Minnesota?
Sugarbeet production regions of North Dakota and Minnesota contribute more than half of the production nationally. Across this region, sugarbeet yield and sugar content vary widely from north to south. The study described in this article used statistical analyses of five years (2013–2017) of county-wise monthly rainfall data to interpret the spatial variation of sugarbeet yield and quality.
Rainfall is the most prominent driver of crop production to both dryland–rainfed and limited-irrigation crop production in the U.S. Great Plains (Lobell et al., 2007; Nielsen, 2018). Leng and Huang (2017) determined that a one-unit increase of precipitation lead to a 5.9% increase in corn yield (area-weighted county mean) in the contiguous United States. However, yield and rainfall relationship is not always positive. In California, Lobell et al. (2007) found that high precipitation was associated with a reduction in yields due to reduced pollination and increased disease pressure in wetter years. Teasdale and Cavigelli (2017) reported that late-season precipitation had the highest correlation with corn and soybean yields at Beltsville, MD. Understanding the rainfall distribution pattern within a production region is beneficial for establishing local recommendations and interpretation of results by researchers, growers, and stakeholders.
Sugarbeet production regions of North Dakota and Minnesota contribute more than half of the production nationally. Across this region, sugarbeet yield and sugar content vary widely from north to south. Counties in the southern part of the region have more variability and a higher average yield than northern counties, indicating environmental conditions are seldom uniform across the region (Campbell, 1995). Commonly, sugarbeet is planted in early May, and it is harvested at the end of September. In the Red River Valley of North Dakota, sugarbeet used around 16.6 to 22.8 inches of water, which is comparatively higher than other crops grown in this area (Bauder & Ennen, 1981). Six weeks after germination, sugarbeet root begins an accelerated growth and continues to accumulate dry matter linearly throughout the growing season (Theurer, 1979). Sucrose content also increases linearly during the growing season with the most rapid rate of accumulation during June followed by a slightly decreased rate in July, remaining relatively constant thereafter until harvest (Theurer, 1979). The photosynthate allocated to the root is partitioned between growth and sucrose storage, and this partitioning is influenced by rainfall during the growing season (Wyse, 1979).
The study described in this article used statistical analyses of five years (2013–2017) of county-wise monthly rainfall data to interpret the spatial variation of sugarbeet yield and quality. The specific research questions investigated were:
- How much is the county-level variability in sugarbeet root yield and sugar content in the region?
- Rainfall in which months are associated with root yield and which are for sugar content? Are they similar across the region?
- What extent of sugarbeet root yield and sugar content can be explained by monthly rainfall data in this region?
Estimation of Rainfall Contribution to Sugarbeet Root Yield and Sugar Content
County-level monthly rainfall data of the growing season (May-September) for five years (2013–2017) were obtained from the North Dakota Agriculture Weather Network (http://ndawn.ndsu.nodak.edu) and the Minnesota Department of Natural Resources website (www.dnr.state.mn.us/climate/index.html). If there was no weather station in a county, the nearest weather station data were used. County-level sugarbeet root yield and sugar content data of major sugarbeet-producing counties of North Dakota and Minnesota were collected from three factory districts: American Crystal Sugar Company, Southern Minnesota Beet Sugar Cooperative, and Minn-Dak (personal communication with agronomists).
A series of steps and computations were involved in quantifying the variability in the relationship between rainfall and sugarbeet production. Exploratory data analyses were conducted, and a five-year average of all variables was calculated using SAS 9.4 (SAS Inc., Cary, NC). ArcGIS10.5 was used to prepare the map for distribution patterns of county-level five-year average root yield, sugar content, and monthly rainfall. Rainfall, root yield, and sugar percent data were detrended using linear regression over time, and residuals were used for the correlation and regression analyses. Pearson correlation coefficients were calculated to relate the residuals of monthly rainfall as well as root yield and sugar content at the county level, using SAS 9.4. To identify the best predictor variable, multiple linear regression with a forward selection of variables was conducted between yield residuals (root yield and sugar content) against monthly rainfall residuals at the county level.
Sugarbeet Production in North Dakota and Minnesota
Dominant sugarbeet production areas are spread across 30 counties in Minnesota (MN) and six counties in North Dakota (ND) (Figure 1). Based on a five-year (2013–2017) average, sugarbeet production covers 428,000 acres in MN and 200,000 acres in ND (USDA-NASS, 2019). Polk county within MN and Pembina within ND of the Red River Valley had the highest sugarbeet area of 90,000 and 55,000 acres, respectively. Five-year county average root yield varied between 20.9 tons/ac at Cottonwood, MN to 29.5 tons/ac at Norman, MN (Figure 2). In southern MN, the top three counties in root yield are Meeker (29.1 tons/ac), Nicollet (28.7 tons/ac), and Redwood (28.1 tons/ac). In the Red River Valley of ND and MN, Norman (29.5 tons/ac), Stevens (29.4 tons/ac), and Becker (28.9 tons/ac) of MN are the top three in root yield. The Red River Valley of ND and MN had higher sugar content than southern MN (Figure 2). Sugar content ranged between 14.3% at Cottonwood, MN to 17.6% at Grand Forks, MN. Higher values of sugar content are concentrated in the Red River Valley of ND and MN compared with southern MN.
Monthly Rainfall Distribution
County-wise monthly rainfall for the 2013–2017 growing seasons is presented in Figure 2. The top three average monthly rainfall amounts were observed at Watonwan (5.05 inches), Cottonwood (5.04 inches), and Brown (4.87 inches) counties of MN for May, McLeod (7.10 inches), Sibley (6.11 inches), and Brown (6.10 inches) counties of MN for June, Douglas (4.55 inches), Becker (4.30 inches), and Nicollet (4.18 inches) counties of MN for July, Kandiyohi (6.76 inches), Yellow Medicine (6.07 inches), and Chippewa (6.04 inches) counties of MN for August, and Otter Tail (4.19 inches), Nicollet (3.81 inches), and Becker (3.41 inches) counties of MN for September. The total growing season rainfall ranged between 12.2 (Cass, ND) and 23.6 inches (McLeod, MN).
Association between Monthly Rainfall and Sugarbeet Yield and Sugar Content
County wise and overall, correlation coefficients between monthly rainfall and sugarbeet root yield and sugar content are presented in Table 1. Overall, root yield had a negative association with rainfall in June and August and a positive association with rainfall in July and September. Sugar content had a negative association with June, August, and September and a positive association with May. May rainfall had a positive association with root yield only for Traverse county, MN and with sugar content for Becker county, MN. June rainfall had a negative association with root yield for Cottonwood, Lyon, Polk, Swift, and Yellow Medicine counties of MN and Cass county of ND. June rainfall also had a negative association with sugar content for Grant county of MN. July rainfall had a positive association with root yield for Cottonwood, Kandiyohi, Lac qui Parle, Nicollet, Norman, and Swift counties of MN and Richland and Trail counties of ND; it had a negative association with root yield only for Lyon county of MN. July rainfall had a positive association with sugar content only for Cottonwood, MN and negative association for Nicollet, Stearns, Stevens, and Yellow Medicine counties of MN and Walsh county of ND. August rainfall had a negative relationship with root yield for Douglas and Redwood counties of MN and Grand Forks county of ND but a positive relationship with root yield for Marshall county of MN. August rainfall had a negative relationship with sugar content for Lyon and Pope counties of MN and a positive relationship for Brown county, MN. September rainfall had a positive relationship with root yield for Stevens and Watonwan counties of MN and Walsh county of ND, and September rainfall and root yield had a negative association for Otter Tail, MN. September rainfall had a negative association with sugar content for Cottonwood, Lac qui Parle, Redwood, and Travers counties of MN and Cass county of ND; the only positive relationship was observed for Clay county, MN.
Table 1. Pearson correlation coefficient (r) between county-wise and overall data residuals of sugarbeet root yield and sugar with residuals of monthly rainfall from May to September for the 2013–2017 growing seasons
| Root yield (tons/ac) | Sugar (%) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| State | County | May | June | July | Aug. | Sep. | May | June | July | Aug. | Sep. |
| MN | Becker | 0.04 | –0.81 | 0.31 | 0.03 | 0.52 | 0.88 | –0.18 | –0.51 | 0.70 | –0.82 |
| MN | Brown | –0.09 | –0.53 | 0.26 | 0.29 | 0.12 | 0.94* | 0.36 | 0.86 | 0.98 | 0.80 |
| MN | Chippewa | –0.23 | –0.44 | 0.61 | –0.85 | 0.47 | 0.70 | –0.01 | –0.81 | –0.14 | –0.32 |
| MN | Clay | –0.15 | 0.34 | 0.53 | –0.46 | 0.50 | 0.01 | 0.08 | 0.90 | –0.87 | 0.90 |
| MN | Cottonwood | 0.17 | –0.99 | 0.94 | –0.71 | –0.74 | 0.53 | –0.68 | 0.90 | –0.21 | –0.94 |
| MN | Douglas | –0.50 | 0.13 | 0.62 | –0.92 | 0.31 | 0.86 | –0.20 | –0.60 | 0.88 | –0.25 |
| MN | Grant | 0.44 | –0.75 | –0.04 | 0.20 | 0.48 | –0.65 | –0.90 | –0.67 | –0.38 | –0.62 |
| MN | Kandiyohi | –0.59 | –0.70 | 0.99 | 0.32 | 0.39 | –0.12 | –0.45 | 0.88 | 0.62 | –0.19 |
| MN | Kittson | –0.19 | –0.92 | 0.79 | 0.85 | 0.86 | 0.75 | 0.18 | –0.69 | 0.03 | –0.76 |
| MN | Lac Qui Parle | 0.32 | –0.52 | 0.95 | 0.71 | 0.61 | 0.10 | 0.30 | –0.56 | –0.18 | –0.93 |
| MN | Lyon | –0.22 | –0.90 | –0.89 | –0.85 | 0.01 | –0.08 | –0.49 | –0.68 | –0.93 | 0.41 |
| MN | Mahnomen | 0.23 | –0.74 | 0.20 | –0.38 | 0.42 | 0.03 | –0.27 | 0.60 | –0.41 | 0.84 |
| MN | Marshall | 0.09 | 0.43 | –0.66 | 0.91 | –0.74 | –0.36 | –0.62 | –0.46 | –0.65 | –0.79 |
| MN | Mc Leod | 0.40 | –0.54 | 0.51 | 0.42 | 0.17 | –0.35 | –0.62 | –0.46 | –0.66 | –0.80 |
| MN | Meeker | 0.28 | –0.78 | 0.70 | 0.76 | 0.35 | 0.61 | 0.39 | –0.35 | –0.63 | –0.86 |
| MN | Nicollet | –0.03 | –0.52 | 0.93 | –0.14 | –0.54 | 0.41 | 0.24 | –0.98 | –0.15 | 0.39 |
| MN | Norman | 0.59 | –0.68 | 0.99 | 0.77 | 0.32 | 0.05 | 0.12 | –0.83 | –0.77 | –0.71 |
| MN | Otter Tail | –0.45 | –0.82 | 0.88 | –0.59 | –0.93 | 0.57 | –0.82 | 0.32 | 0.04 | –0.71 |
| MN | Polk | –0.05 | –0.97 | 0.09 | 0.20 | –0.13 | 0.31 | –0.58 | –0.21 | –0.24 | –0.77 |
| MN | Pope | 0.84 | –0.74 | 0.25 | –0.08 | 0.43 | 0.51 | –0.37 | –0.76 | –0.93 | –0.62 |
| MN | Redwood | –0.13 | –0.34 | 0.29 | –0.96 | 0.44 | 0.57 | 0.80 | 0.22 | –0.01 | –0.92 |
| MN | Renville | –0.03 | –0.65 | 0.79 | –0.85 | –0.13 | 0.84 | –0.11 | –0.73 | 0.76 | 0.19 |
| MN | Sibley | 0.03 | –0.96 | 0.58 | –0.69 | 0.44 | 0.80 | 0.50 | –0.52 | –0.19 | –0.44 |
| MN | Stearns | –0.77 | 0.27 | –0.75 | –0.52 | –0.60 | –0.72 | –0.35 | –0.95 | –0.75 | –0.87 |
| MN | Stevens | –0.13 | –0.55 | –0.71 | –0.39 | 0.93 | 0.30 | 0.32 | –0.91 | –0.05 | –0.87 |
| MN | Swift | 0.28 | –0.97 | 0.94 | –0.73 | 0.58 | 0.85 | –0.10 | 0.24 | 0.43 | –0.55 |
| MN | Traverse | 0.96 | –0.87 | 0.59 | –0.03 | 0.45 | –0.13 | –0.12 | –0.26 | –0.27 | –0.94 |
| MN | Watonwan | –0.54 | –0.39 | 0.45 | 0.18 | 0.94 | 0.71 | 0.40 | –0.64 | –0.46 | –0.63 |
| MN | Wilkin | 0.86 | –0.79 | 0.39 | –0.17 | 0.09 | –0.19 | 0.45 | 0.08 | –0.31 | –0.53 |
| ND | Yellow Medicine | 0.27 | –0.95 | 0.09 | 0.53 | 0.01 | –0.82 | 0.47 | –0.94 | –0.50 | –0.48 |
| ND | Cass | –0.58 | –0.98 | 0.37 | –0.38 | –0.03 | 0.70 | –0.34 | –0.46 | –0.04 | –0.97 |
| ND | Grand Forks | 0.61 | –0.85 | –0.21 | –0.99 | –0.01 | 0.40 | –0.40 | –0.56 | –0.78 | –0.71 |
| ND | Pembina | 0.30 | –0.87 | 0.72 | –0.14 | 0.32 | 0.01 | –0.78 | –0.23 | 0.08 | –0.55 |
| ND | Richland | –0.06 | –0.59 | 0.96 | –0.64 | –0.18 | 0.69 | –0.06 | 0.55 | –0.03 | –0.84 |
| ND | Traill | –0.12 | –0.66 | 0.93 | –0.34 | 0.44 | 0.51 | 0.90 | –0.76 | –0.49 | –0.49 |
| ND | Walsh | 0.39 | –0.72 | 0.21 | –0.81 | 0.90 | –0.29 | –0.22 | –0.96 | –0.51 | –0.32 |
| Overall | –0.02 | –0.59 | 0.40 | –0.18 | 0.20 | 0.22 | –0.17 | –0.11 | –0.19 | –0.38 | |
Forward selection of the regression equation for June rainfall explained 34% of the root yield variation across the sugarbeet production region of ND and MN; subsequently adding July and May rainfall improved the explanation to 40 and 44% (Table 2). May and June rainfall had negative associations, but July had a positive association. Sugar content is mostly explained by September rainfall; subsequent selection of June and August rainfall do not improve the prediction much.
Table 2. Linear regression coefficient with forward selection of monthly rainfall residuals to fit sugarbeet root yield and sugar content residuals of North Dakota and Minnesota during the 2013–2017 growing seasons
| Root yield | Sugar (%) | |||||||
|---|---|---|---|---|---|---|---|---|
| R2 | Variable entered | Parameter estimate | Pr > F | R2 | Variable entered | Parameter estimate | Pr > F | |
| Step 1 | 0.34 | June | –0.96 | <0.001 | 0.14 | September | –0.23 | <0.001 |
| Step 2 | 0.40 | June | –0.84 | <0.001 | 0.18 | June | –0.08 | 0.001 |
| July | 0.37 | <0.001 | September | –0.23 | <0.001 | |||
| Step 3 | 0.44 | May | –0.44 | 0.001 | 0.19 | June | –0.07 | <0.001 |
| June | –0.94 | <0.001 | August | –0.06 | 0.001 | |||
| July | 0.36 | <0.001 | September | –0.23 | 0.075 | |||
Discussion
Yield prediction is very important for optimizing processing campaigns of sugarbeet factory districts, and this study provides insights on the impact of monthly rainfall on root yield and sugar content. It is evident that yield and sugar response to monthly rainfall varied across counties. Counties vary in size and degree of environmental heterogeneity. The wide variability of county yields within individual years indicated the lack of uniformity in growing conditions throughout the region (Campbell, 1995).
Only June and September rainfall can explain the 34% and 14% variability of sugarbeet root yield across 36 counties of MN and ND. Kukal and Irmak (2018) found that climate explained more than 20% of the crop yield variability of a considerable number of counties in the Great Plains, and precipitation alone explained yield variability in 18% of counties, representing 32% of the regional production. In our results, some counties did not show any association with rainfall, indicating that other factors such as air temperature, soil organic matter, or management practices like subsurface drainage play a dominant role in the production. Kenter et al. (2006) found that rainfall only slightly affected leaf growth and did not show any significant influence on the taproot growth rate.
Sugarbeet root yield showed a negative association with June rainfall. Sugarbeet was quite tolerant of mid- to late-season water deficit and plant moisture stress (Carter et al., 1980); but excess rainfall can damage sugarbeet with an increase in root diseases—Aphanomyces, Fusarium, Rhizoctonia, Rhizophus, or Pythium. Also, excess moisture also accelerates the loss of nitrogen through leaching and denitrification. The negative association of late-season rainfall (September/October) on sugarbeet sucrose content in the Northern Great Plains was also reported by Akeson (1981). Besides, southern MN counties receive considerable manure application due to increased livestock operations, mainly poultry and swine, leading to mineralization of nitrogen (Lamb et al., 2016). Integrating both late-season rainfall and N mineralization reduces the sugar content of southern MN.
Conclusion
Rainfall in June and September could be critical for the prediction of sugarbeet root yield and sugar content, respectively. Out of 36 counties of ND and MN, only a few counties showed significant associations with monthly rainfall, indicating other factors need to be identified. Subsurface drainage has been installed at an increasing rate across the region to reduce the water damage. Impact of rainfall on county-wise yield and sugar content will depend on the relative adoption rate of subsurface drainage. Future research studies might separate this sugarbeet production region into two groups, counties responding to rainfall and not responding to rainfall; and subsequently determine climatic, soil, or management variables relevant for both groups.
Dig deeper
Akeson, W.R. (1981). Relationship of climate and sucrose content of sugarbeet roots. Journal of the American Society of Sugar Beet Technologists, 21(1), 27–40
Bauder, J.W., & Ennen, M.J. (1981). Water use of field crops in eastern North Dakota. North Dakota Farm Research, 39, 1, 3–5.
Campbell, L.G. (1995). Long-term yield patterns of sugarbeet in Minnesota and eastern North Dakota. Journal of Sugar Beet Research, 32, 1, 9–22.
Carter, J.N., Jensen, M.E., & Traveller, D.J. (1980). Effect of mid- to late-season water stress on sugarbeet growth and yield. Agronomy Journal, 72, 806–816.
Kenter, C., Hoffmann, C.M., & Marlander, B. (2006). Effects of weather variables on sugar beet yield development (Beta vulgaris L.). European Journal of Agronomy, 24, 62–69. https://doi.org/10.1016/j.eja.2005.05.001
Kukal, M.S., & Irmak, S. (2018). Climate-driven crop yield and yield variability and climate change impacts on the US Great Plains agricultural production. Scientific Reports, 8, 3450. https://doi.org/10.1038/s41598-018-21848-2
Lamb, J.A., Bredehoeft, M.W., Radamacher, J., VanOs, N., Dunsmore, C., & Bloomquist, M. (2016). Swine manure application management in a sugarbeet rotation. Sugarbeet Research and Extension Reports, 46, 59–66.
Leng, G.Y., & Huang, M.Y. (2017). Crop yield response to climate change varies with crop spatial distribution pattern. Scientific Reports, 7, 1463. https://doi.org/10.1038/s41598-017-01599-2
Lobell, D.B., Cahill, K.N., & Field, C.B. (2007). Historical effects of temperature and precipitation on California crop yields. Climatic Change, 81, 187–203. https://doi.org/10.1007/s10584-006-9141-3
Nielsen, D.C. (2018). Influence of latitude on the US Great Plains east-west precipitation gradient. Agricultural & Environmental Letters, 3. https://doi.org/10.2134/ael2017.11.0040
Teasdale, J.R., & Cavigelli, M.A. (2017). Meteorological fluctuations define long-term crop yield patterns in conventional and organic production systems. Scientific Reports, 7, 688. https://doi.org/10.1038/s41598-017-00775-8
Theurer, J.D. (1979). Growth patterns in sugarbeet production. Journal of the American Society of Sugar Beet Technologists, 20, 343–367.
Wyse, R.E. (1979). Parameters controlling sucrose content and yield of sugarbeet roots. Journal of the American Society of Sugar Beet Technologists, 20, 368–385.
USDA-NASS. (2019). Crop production 2018 summary. Retrieved from https://nass.usda.gov/Publications/Todays_Reports/reports/cropan19.pdf
Text © . The authors. CC BY-NC-ND 4.0. Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.
