Winter-terminated cover crop strategies for vegetable farms in northern New England

Objective 1: On-Farm Cover Crop Variety Trial

Seasonal precipitation and temperature were recorded with a Davis Instrument Vantage Pro2 weather station, equipped with a WeatherLink data logger at Borderview Research Farm in Alburgh, VT (Table 2 and 3).

Table 2. Seasonal weather data collected in Alburgh, VT, 2017.

Based on weather data from a Davis Instruments Vantage Pro2 with WeatherLink data logger. Alburgh precipitation data from August-October was provided by the NOAA data for Highgate, VT. Historical averages are for 30 years of NOAA data (1981-2010) from Burlington, VT.

 Table 3. Seasonal weather data collected in Alburgh, VT, 2017-2018

Based on weather data from a Davis Instruments Vantage Pro2 with WeatherLink data logger. Alburgh precipitation data from August-October was provided by the NOAA data for Highgate, VT. Historical averages are for 30 years of NOAA data (1981-2010) from Burlington, VT.

In 2017, August was cooler and wetter than historical averages while September and October were unseasonably hot and dry. The winter months of November thru January were cold and dry. The early months of 2018 experienced a lot of variation. February was unseasonably warm, March was typical, April was unseasonably cold and wet, and May was warm and dry. The months of July and August were warmer and dryer than historical averages.

Results from Fairfax Location

Table 4. Cover crop mix yield and quality, Fairfax, VT, 2017-18.                

*Treatments marked with an asterisk did not perform statistically worse than the top performing treatment (p=0.10) shown in bold.

NS – There was no statistical difference between treatments in a particular column (p=0.10).

NA – Statistical analysis was not performed as only one treatment had living biomass to measure in the spring.

Both treatments 13 (oats, radish, clover) and 15 (winter rye and vetch) performed comparably in the fall for yield and percent soil cover (Table 4). Treatment 15 had a greater nitrogen concentration, which was to be expected as it contained a legume. In the spring, the cover crop that contained winter rye and vetch provided ample soil cover and biomass.

Table 5. Soil active carbon and wet aggregate stability, Fairfax, VT, 3-May 2018. 

The top performing treatment (p=0.10) shown in bold.

NS – There was no statistical difference between treatments in a particular column (p=0.10).

Surprisingly, the control had a greater amount of active carbon in the soil compared to treatment 15 (winter rye and vetch) (Table 5). Soil active carbon was measured prior to termination of the cover crop. Hence, it is possible that the measurement of active carbon would have been higher had the cover crop been amended into the soil before the samples were taken. Ample dead and decomposing weed biomass were present in the control plots and may have contributed to active carbon.

Table 6. Soil NO3-N within the different cover crop treatments, Fairfax, VT, 2017-18. 

The top performing treatment (p=0.10) shown in bold.

NS – There was no statistical difference between treatments in a particular column (p=0.10).

Clearly, the cover crops helped to scavenge excess nitrogen in the soil. In the fall of 2017, the control plots had the highest level of soil nitrates compared to those with a cover crop. For all treatments, soil NO₃-N peaked between 6-Jun and 19-Jun (Figure 1, Table 6). Interestingly, the rye and vetch treatment peaked the highest for soil NO₃-N out of all the treatments. This occurred on the 6-Jun sampling and was significantly higher than the other treatments. This indicates that the N from the rye/vetch cover crop took approximately 40 days to begin mineralizing and releasing NO₃-N. The high levels of NO₃-N in the control show excess nutrients that would likely be lost into the environment. The value of having a cover crop is to scavenge excess nutrients and keep them from leaving the system through leaching or other pathways.

Results from the Grand Isle Location

Table 7. Cover crop mix yield and quality, Grand Isle, VT, 2017-18.          

*Treatments marked with an asterisk did not perform statistically worse than the top performing treatment (p=0.10) shown in bold.

NA – Statistical analysis was not performed as only one treatment had living biomass to measure in the spring.

In the fall, treatments 11 (radish) and 13 (oats, clover, radish) were the best performers for yield and ground cover (Table 7). Treatments 11, 13, and 14 (winter rye) had comparable nitrogen concentrations, which may reflect the strong ability for winter rye to absorb available nitrogen. Biomass in the control plots were weeds but still provided adequate ground cover to protect the soil from erosion. Treatment 14 (winter rye) was the best performer for percent soil cover in the spring, which is not surprising since it overwinters.

Table 8. Soil active carbon and wet aggregate stability, Grand Isle, VT, 10-May 2018. 

The top performing treatment (p=0.10) shown in bold.

NS – There was no statistical difference between treatments in a particular column (p=0.10).

Treatment 11 (radish) outperformed treatment 14 (winter rye) for wet aggregate stability (Table 8). Soil aggregates are formed when biological activity in the soil causes soil particles and organic matter to become glued together. The more glue the more stable a soil aggregate may become. The stronger the aggregate the more resistant it is to being degraded when disturbed by rain or mechanical action. Higher aggregate stability can improve soil drainage and other biological properties. The radish growing in the fall likely improved biological activity and helped to build soil aggregates. Once the radish died from cold winter temperature, microbial activity to decompose the root may have further enhanced the aggregate stability.

Table 9. Soil NO3-N within the different cover crop treatments, Grand Isle, VT, 2017-18. 

*Treatments marked with an asterisk did not perform statistically worse than the top performing treatment (p=0.10) shown in bold.

NS – There was no statistical difference between treatments in a particular column (p=0.10).

There were low levels of soil nitrate in the fall of 2017. Interestingly, control soil nitrate levels were higher indicating that the cover crops were utilizing excess soil nitrate. Throughout the 2018 season, the differing cover crop treatments performed comparably for soil NO₃-N concentrations (Figure 2, Table 9). Given the drought conditions that were observed across the growing season, it is possible that cover crop decomposition was slow due to lack of moisture and above average temperatures.

Table 10. Sweet corn yield and quality, Grand Isle, VT, 30-Aug 2018.        

 NS – There was no statistical difference between treatments in a particular column (p=0.10).

 The sweet corn grown following the cover crop did not show any significant differences for yield or quality, between cover crop treatments (Table 10).

 Results from Borderview Farm

 Table 11. Cover crop mix yield and quality, Alburgh, VT, 2017-18.       

*Treatments marked with an asterisk did not perform statistically worse than the top performing treatment (p=0.10) shown in bold.

At Borderview Farm, the best performers for fall yield were treatments 3 (winter rye, clover, turnip), 7 (oats, clover, turnip), and 8 (ryegrass, turnip) (Table 11). All treatments had adequate soil cover and would have been protected from erosion. In the spring, two of the treatments consisting of overwintering varieties, treatments 14 (winter rye) and 15 (winter rye and vetch), had the best yields. These two treatments were among the top performers for spring ground cover. Interestingly, the control and two winterkilled treatments (treatment 10 – oats and 13 – oats, radish, crimson clover) also were top performers for ground cover, which may indicate that the dead plant materials were effective as a spring cover.

Table 12. Soil active carbon and wet aggregate stability, Alburgh, VT, 10-May 2018. 

NS – There was no statistical difference between treatments in a particular column (p=0.10).

Among the treatments evaluated for active carbon and aggregate stability, there were no significant differences (Table 12).

Table 13. Soil NO3-N within the different cover crop treatments, Alburgh, VT, 2017-18. 

*Treatments marked with an asterisk did not perform statistically worse than the top performing treatment (p=0.10) shown in bold.

NS – There was no statistical difference between treatments in a particular column (p=0.10).

 When most treatments were close to peak soil NO3-N for the season, on 21-Jun, treatments 7 (oats, clover, turnip) and 17 (hairy vetch) were top performers (Figure 3, Table 13). This is not surprising since both of those treatments contained nitrogen fixing varieties.

 Table 14. Sweet corn yield and quality Alburgh, VT, 2018. 

NS – There was no statistical difference between treatments in a particular column (p=0.10).

 No significant differences were seen in the sweet corn grown following the cover crop treatments (Table 14).

DISCUSSION

At both Pomykala Farm and Borderview Farm, there was no measurable impact on the subsequent cash crop that would indicate differences between the cover crop treatments. However, it is interesting to note of when peak soil NO₃-N generally was at each farm. For River Berry Farm, peak soil NO₃-N was between 6-Jun and 19-Jun, approximately 75 days after the field was prepped and planted with strawberries. This was earlier than the other farms and may be influenced by River Berry’s light soil, which would have warmed faster than the soils at the other two farms. Also, regular irrigation at River Berry Farm likely helped cover crops decompose more quickly. At Pomykala Farm, peak was on 16-Aug, which was approximately 45 days after field prep and planting. This was fairly late in the season and likely influenced by the extremely hot and dry conditions experienced during the growing season. At Borderview Farm, peak was from 21-Jun to 19-Jul, which was 30-60 days after field prep and planting of the sweet corn.  

It is important to consider the effect of soil texture and seasonal differences on soil NO₃-N availability from cover cropping. Also, using a winterkilled cover crop variety may provide the benefit of not having to manage terminating the crop in the spring, when timing of this may be difficult due to wet, spring conditions. Cover cropping decisions will likely be based on the demands and goals within each operation and field management considerations.

Objective 2: Cover Crop Planting Date Trial

Seasonal precipitation and temperature were recorded with a Davis Instrument Vantage Pro2 weather station, equipped with a WeatherLink data logger at Borderview Research Farm in Alburgh, VT (Table 15).

Table 15. Seasonal weather data collected in Alburgh, VT, 2017-2018.

Based on weather data from a Davis Instruments Vantage Pro2 with WeatherLink data logger. Alburgh precipitation data from August-October was provided by the NOAA data for Highgate, VT. Historical averages are for 30 years of NOAA data (1981-2010) from Burlington, VT.

 In 2017, August was cooler and wetter than historical averages while September and October were unseasonably hot and dry. The winter months of November thru January were cold and dry. The early months of 2018 experienced a lot of variation. February was unseasonably warm, March was fairly typical, April was unseasonably cold and wet, and May was warm and dry. Overall, between August and May there were a total of 1705 growing degree days, which is 608 more than historical averages.  

Results by planting date

Table 16. Cover crop yield and coverage for each planting date across all mixtures, Alburgh, VT, 2017-2018.

*Treatments marked with an asterisk were not statistically different compared to the top performing treatment (p=0.10) shown in bold.

Unsurprisingly, the 24-Aug planting generated the most biomass in the fall (Table 16). However, the 19-Sep planting generated the most biomass and percent cover in the spring, although the 24-Aug planting provided a comparable amount of cover. These observations are not surprising, as we would expect some species such as annual ryegrass and oats to produce significant quantities of biomass if planted in late summer. The amount of biomass would decline with later planting dates. However, winter grains must be planted a bit later to establish just prior to winter. Planting winter grains to early may result in high quantities of biomass that can smoother the plant crown over the winter.  Other research has shown that planting by late September produces the best yields. This study supports this evidence as highest spring cover crop biomass was obtained from the 19-Sep planting date.

Results by cover crop treatment

Table 17. Cover crop yield and soil coverage for each cover crop treatment, Alburgh, VT, 2017-2018. 

Treatments marked with an asterisk were not statistically different compared to the top performing treatment (p=0.10) shown in bold.

Across all plantings, the top performer for fall yield was treatment 7 (oats, clover, turnip), while treatment 3 (winter rye, clover, turnip), 4 (triticale, clover, turnip), and 11 (radish) performed comparably (Table 17). Top performers for the spring yield included overwintering varieties that had the advantage of being able to grow again in the spring. For the spring, the top performers were treatments 6 (triticale and rape), 14 (winter rye), and 15 (winter rye and vetch). Out of the 4 top performing treatments for spring percent cover, 3 of them had an overwintering variety in their mix.

Results for each planting date

Table 18. Cover crop yield and coverage for the 24-Aug 2017 planting, Alburgh, VT, 2017-2018. 

*Treatments marked with an asterisk were not statistically different compared to the top performing treatment (p=0.10) shown in bold.

From the 24-Aug planting, treatments 3 (winter rye, clover, turnip), 4 (triticale, clover, turnip), 7 (oats, clover, turnip), 8 (ryegrass, turnip), 9 (ryegrass, radish), and 16 (ryegrass) were the top performers for yield in the fall (Table 18). Interestingly, none of the top performers for fall yield were top performers for spring yield. Two out of 6 successful fall yielding treatments included overwintering varieties, however, they also included turnip which would not survive the winter and generate more biomass in the spring.

In the spring, treatments 14 (winter rye) and 15 (rye and vetch) were top performers. This makes sense considering that winter rye and vetch will over winter and continue to grow in the spring. Another interesting result is that in the spring, top performers for percent cover included overwintering treatments, yet also included the control. It is possible that the weed seeds, naturally found in the soil, were able to grow well enough to provide comparable ground cover.

Table 19. Cover crop yield and coverage for the 11-Sep 2017 planting, Alburgh, VT, 2017-2018. 

*Treatments marked with an asterisk were not statistically different compared to the top performing treatment (p=0.10) shown in bold.

The top performers from the 11-Sep planting for fall yield were treatments 2 (triticale, radish, clover), 4 (triticale, clover, turnip), 5 (oats, radish), 7 (oats, clover, turnip), 10 (oats), and 11 (radish) (Table 19). The treatments generally performed well for percent cover in the fall. In the spring, unsurprisingly, overwintering varieties were more successful with generating biomass and top performers included treatments 6 (triticale, rape), 14 (winter rye), 15 (winter rye and vetch), and 17 (hairy vetch). Surprisingly, the control had a comparably high yield for the spring, which was likely due to the weed seed bank coming out of dormancy. 

 Table 20. Cover crop yield and coverage for the 19-Sep 2017 planting, Alburgh, VT, 2017-2018. 

*Treatments marked with an asterisk were not statistically different compared to the top performing treatment (p=0.10) shown in bold.

The top performers from the 19-Sep planting for fall yield included treatments 2 (triticale, radish, clover), 3 (winter rye, clover, turnip), 4 (triticale, clover, turnip), 5 (oats, radish), 6 (triticale, rape), 9 (ryegrass, radish), 11 (radish), 13 (oats, radish, clover), and 14 (winter rye) (Table 20). Top performers for the spring yield included overwintering varieties. Specifically, top performers were treatments 3 (winter rye, clover, turnip), 4 (triticale, clover, turnip), 6 (triticale, rape), and 15 (rye and vetch).

Table 21. Cover crop yield and coverage for the 27-Sep 2017 planting, Alburgh, VT, 2017-2018. 

 *Treatments marked with an asterisk were not statistically different compared to the top performing treatment (p=0.10) shown in bold.

From the 27-Sep planting, the highest fall yielding performer was treatment 11 (radish) (Table 21). The top performers for spring yield including overwintering varieties, representing treatments 2 (triticale, radish, clover, pea), 4 (triticale, clover, turnip), 6 (triticale, rape), and 14 (winter rye).

Results from the planting date by variety interaction

The interaction of planting date and variety were significant for fall yield (p=0.0007), fall percent cover (p=0.0019), spring yield (p<0.0001), and spring percent cover (p<0.0001), meaning that cover crop mixture performance differed by planting date. (Figure 4, 5, 6, 7). It was expected that winter terminated cover crops such as oats, annual ryegrass, and tillage radish would produce more biomass and soil coverage if planted in late summer compared to fall. Cover crop mixtures with winter terminated and surviving species tended to experience less variability in cover and yield across planting dates.

The trend within each planting date was that the top performers for yield in the spring included an overwintering variety, which is unsurprising since these varieties will be able to grow more biomass in the spring after winter dormancy. However, it is interesting to note that in the 24-Aug, 19-Sep, and 27-Sep plantings top performers for spring percent cover included treatments that did not overwinter. This indicates that these mixes were still able to provide the benefit of ground coverage, consisting of dead plant materials from the fall. Using a winterkilled cover crop variety may also provide the benefit of not having to manage terminating the crop in the spring, when timing of this may be difficult due to wet, spring conditions. Also, it was clear that planting earlier in the fall would result in great fall biomass yields. Cover cropping decisions will have to be made based off of the demands within each operation and field management considerations.