Jarrod O. Miller, Assistant Professor and Extension Specialist, Agronomy, firstname.lastname@example.org; Amy L. Shober, Professor and Extension Specialist, Nutrient Management and Environmental Quality, email@example.com, Jamie Taraila, Graduate Research Assistant;
As part of a Northeastern SARE graduate student grant, we used a drone to predict the nitrogen (N) that may be present in cover crops prior to burndown. We flew fields in Laurel, Georgetown, and Harbeson with a readily available consumer drone (Phantom4) equipped with a standard (RGB) camera. Each of the fields were flown prior to cover crop burndown (late April to late May) resulting in 25-100 images per field that had to be stitched together into one image (Figure 1a). The camera captured different wavelengths of light (i.e., red, green, blue) that were reflected by plants which were transformed into the Visible Atmospherically Resistant Index (VARI). This allowed us to estimate plant biomass by comparing VARI values to cover crop biomass that was collected in the field. We collected 10 samples per field, which were dried, weighted, and then analyzed for N content by the UD soil testing lab.
When VARI measurements were compared to cover crop biomass and total N (lbs/acre) we saw a strong relationship in all three fields (Figure 2). Yet, each field had a unique relationship (e.g., different slope or magnitude) due to variability in field characteristics, time of sampling, and type of cover crop. These linear relationships were used to convert VARI to cover crop N content (lbs/acre), as shown in Figure 1b.
Relationships between cover crop weight (biomass) were strong, but improved when converted to lbs N per acre, which indicates how both leaf area and N content (green) may influence the VARI measurements. Cover crop N estimated by VARI do not necessarily represent the actual plant N available from the cover crop; the carbon to nitrogen ratio (C:N) of the cover crop biomass also needs to be assessed to determine the potential for N mineralization to the cash crop. Lower C:N (<20) are desired if you want to assume a net relates of N to the following crop.
In the case of the rye cover crops planted in Harbeson (Figure 3) and Georgetown (Figure 4), it is more likely that a portion of the cover crop N will not be immediately available to a newly planted cash crop. The timing of burndown also has an effect on this, where the earlier rye termination in Georgetown resulted in a C:N of 14, while the late May burndown in Harbeson has a C:N of 25. So where 20-80 lbs of N is predicted in some parts of the Harbeson field (Figure 3), it is not as likely to be available compared to the Georgetown field (Figure 4). While we would expect the rye/clover mix in Laurel to provide more N, the late burndown and greater rye biomass resulted in a C:N of 30. Therefore, adding clover may not necessarily result in greater N where rye survival and growth remain higher.
Observations of field variability may be more useful than actual N content. While our estimates (based on VARI and biomass N) suggest that cover crops could provide 70 to 130 lbs N/acre to the next crop, imagery from all three fields show how uneven the distribution of that total N could be. While we estimated that cover crop biomass in most of the Laurel field (Figure 1b) has > 40lbs N/ac across the field, there are pockets where biomass N content is lower. Biomass N distribution is even more variable in the Harbeson field (Figure 3), where most biomass is < 10lbs N/ac with a few patches around 16-20 lbs N/ac. In Georgetown (Figure 4), where winter soil moisture killed many portions of the cover crop, there is a strong contrast between zero cover crop N and 20-200 lbs N/ac.
Mapping cover crops with a standard consumer drone and comparing tissue N samples may provide valuable for farmers in our region, particularly if cover crops were planted with the goal of providing early season N to cash crops. A similar exercise can also be done with satellite imagery instead of a drone, but collection of tissue samples from each field is imperative to account for variability in cover crops biomass, C:N ratios, and field conditions. Based on this project, each field would need its own analyses.
Sponsored by Northeast SARE