Assessing Cover Crop Biomass Using Aerial Imagery: Lessons Learned During the UNL-NRCS Soil Health Initiative
The Nebraska On-Farm Research Network does not endorse the use of the products tested in on-farm research trials. While treatments are replicated within trials and may be replicated across multiple sites under various conditions, your individual results may vary.
Why is Cover Crop Biomass Important?
Cover crop biomass production is important and related to benefits such as weed and pest suppression, erosion control, and nitrogen (N) retention and supply. For example, weed suppression can occur through the physical interference with weed seed emergence and growth, competition with weeds for water, nutrients and sunlight, and possible allelopathic compounds. Atmospheric N fixed by legume cover crops can be used by cash crops following decomposition of cover crop residue, which contributes to crop production through nutrient cycling. Cover crops reduce soil erosion by providing physical cover while growing and after termination with residue left in the field. The processes leading to these benefits, be it erosion control, N fixation or weed suppression, are all related to cover crop biomass production.
Aerial Imagery (Remote Sensing) Can Be Used to Estimate Cover Crop Biomass
Spatial variation in cover crop growth within a field is rarely measured by farmers or agronomists given the time and efforts associated with sampling. Crop canopy sensors and aerial imagery have been used to track cover crop adoption on a regional scale, as well as estimate N content from cover crop above-ground biomass (Seifert et al., 2019; White et al., 2019).
Normalized difference vegetation index (NDVI) is a metric commonly applied from aerial imagery of plant canopy spectral reflectance. This value represents the ratio of the red light to near-infrared wavelengths and is calculated as NDVI = (NIR-Red)/(NIR+Red). The combination of surface reflectance at two or more wavelengths is designed to highlight a particular property of the vegetation. The NDVI is correlated with both plant biomass (e.g. leaf area) and chlorophyll content (e.g. plant photosynthetic activity), with higher NDVI values representing more chlorophyll and biomass.
The NDVI and other vegetation indices are extensively used to evaluate vegetative cover for applications in site-specific crop management practices such as irrigation scheduling, nutrient management, and pest control. Similarly, NDVI sensors can also be used to estimate cover crop biomass. In turn, this may also have implications on planting and termination management, as well as site-specific decisions based on in-field variability of cover crop growth.
Aerial Imagery Can Capture Cover Crop Biomass Differences Resulting From Different Cover Crop Practices
The Nebraska Soil Health Initiative (SHI) is a partnership for education and on-farm assessment launched in 2016 connecting the Nebraska Extension On-Farm Research Network and the USDA-NRCS. In this article we report preliminary cover crop biomass data collected in 2019 and its relationship to remotely-sensed information. To answer the question, “Can aerial imagery be used to estimate cover crop biomass across several fields and cover crop management treatments?”, cover crop biomass not winter terminated was collected in spring 2019 before chemical termination in four fields (Table 1).
|County||Cover crop management comparisons||Cover crop planting date||Cover crop Termination date||Cover crop mixture composition|
|Greeley||Cover crop and no cover crop||11/17/2018||6/1/2019||cereal rye 50 lb/ac, forage collards 1 lb/ac, ‘Purple top’ turnip 1 lb/ac, rapeseed 1 lb/ac, kale 1 lb/ac|
|Howard||Cover crop and no cover crop||9/21/2018||5/14/2019||‘Barkant’ turnip 0.8 lb/ac, African cabbage 0.6 lb/ac, ‘Impact’ forage collard 0.5 lb/ac, ’Dwarf Essex’ rapeseed 1.6 lb/ac, eco-till radish 1.1 lb/ac, ’Peredovick’ sunflower 1.0 lb/ac, ’Finch’ safflower 1.0 lb/ac, VNS hairy vetch 1.6 lb/ac, ’Viceroy’ lentil 1 lb/ac, cereal rye 33 lb/ac|
|Knox||Grazed and non-grazed cover crop||8/12/2018||6/28/2019||pearl millet 2.5 lbs/acre, japanese millet 5 lbs/ac, spring oats 10 lbs/ac, winter triticale 10 lb/ac, soybean 10 lbs/ac, mungbean 5 lb/ac, sunnhemp 1.29 lb/ac, common vetch 3 lb/ac, bmr dwarf sorghum 4 lbs/ac|
|Stanton||Single species and multi-species mixture||7/27/2018||5/16/2019||Single species: cereal rye 50 lb/ac
Multi species: cereal rye 30 lb/ac, red clover 3 lb/ac, rapeseed/canola 2 lb/ac, hairy vetch 6 lb/ac
Aerial imagery was obtained from a commercial aerial imagery provider (TerrAvion, San Leandro, CA) with a spatial resolution of 6 to 8 inches. The NDVI values were correlated to a wide range of cover crop treatments, including monocultures and mixtures of grasses, legumes and brassicas (Table 1).
We observed that NDVI values for cover crop biomass were related to management factors such as cover crop planting date, seeding rate, species in the mixture, and grazing component, across the four different farms. Key observations/findings included:
- Cover crops that were not grazed in the fall had greater spring biomass and NDVI values than the grazed treatments in Knox County (Figure 1-A).
- Higher biomass and NDVI values were observed in a single species cover crop compared with a cover crop mixture in Stanton County. Biomass samples from both treatments were collected in May 2019. The cover crop mixture in the spring was dominated by cereal rye, whereas the other species winterkilled (Figure 1-B).
- The Howard County cover crop mixture was seeded in September, whereas the Greeley County cover crop mixture was seeded in November. As such, the Howard County site had almost 70 more days of growth than in Greeley County. This resulted in limited biomass production and lower NDVI values at the Greeley County site. The Howard County cover crop accumulated greater biomass prior to the onset of cold weather (Figures 2-A and 2-B).
- The very low biomass production in Greeley County (less than 500 lb/ac) that resulted in lower NDVI can be attributed to a low brassica seeding rate (1 lb/ac) with the resulting growth being primarily brassica. Brassicas winterkill in Nebraska environment and the November 17th planting date was past the optimum planting date for brassicas (Figure 2-B).
- At the Howard County site, variation in cover crop biomass production result in heterogenous NDVI patterns across the field (Figure 3). Aerial imagery, possibly paired with end of season yield information, might be used to generate biomass production conditions maps for the better cover crop management.
The data presented here demonstrates the value and use of high-resolution multispectral images to assess variation in cover crop biomass production and management practices. Although NDVI and biomass values might not align perfectly and the response might not always be linear, the relationship described here provide us some helpful insights on cover crop management considerations that affect biomass production such as planting time, seeding rate and method (broadcast or drill), species in the mixture (single or multi-species), herbicide program for cover crop termination, and grazing opportunities. Cover crop growth was variable as a result of spatial and temporal availability of water and nutrients, similar to what we observe in cash crop yields. Using aerial imagery, a non-destructive and easy-to apply method, we are able to gain insight into cover crop biomass production across an entire field, which would not be possible with traditional, boots-on-the-ground biomass sampling.
The authors would like to thank the Natural Resources Conservation Service (USDA-NRCS) and Robert B. Daugherty Water for Food Global Institute (DWFI) for funding support.
Seifert, C. A., Azzari, G., & Lobell, D. B. (2019). Corrigendum: Satellite detection of cover crops and their effects on crop yield in the Midwestern United States (2018 Environ. Res. Let. 13 064033). Environmental Research Letters, 14(3), 039501.
White, C. M., Bradley, B., Finney, D. M., & Kaye, J. P. (2019). Predicting Cover Crop Nitrogen Content with a Handheld Normalized Difference Vegetation Index Meter. Agricultural & Environmental Letters, 4(1).