While NDVI has been used for only a few years in viticulture, it has been more than 20 years since cereal farmers adopted NDVI index to monitor nitrogen fertilization. One of the reasons why viticulture industry has been slow in using such technology has to do with how difficult it is to analyze NDVI images taken from discontinuous crops (ie. crop with a tree-like structure) compared to continuous crops (see previous article).
Vineyard in champagne; Côte des bars (Source)
However, if we can remove the contribution of between rows area to the overall picture definition, then the NDVI image becomes actionable. By focusing on the chlorophyll content of each planted row, it is possible to adjust fertilizer inputs according to the vigor areas delineated by the map. However it remains strongly recommended to support mapping analysis with plant based measurements obtained directly in the vineyards such as leaf or petiolar nutritional content. Plant based measurements will support the winegrower to better diagnose what causes such spatial heterogeneities. As we discussed previously, an NDVI map is useful to reveal vineyard spatial variability. However, it does not explain what is the origin of spatial heterogeneities unless your perform ground observations (see previous article on sampling).
An NDVI map obtained before harvest, usually at veraison, can be a useful tool for winter fertilization, pruning and inter-row management decisions.
For fertilization decisions, NDVI map allows winegrower to apply different amounts of fertilizers according to vigor zones spatial distribution. With the right hardware, these maps can even be integrated into GPS consoles and from there it is possible to guide the spread of fertilizer inputs locally. Fortunately, even without such sophisticated equipment attached to a tractor, it is possible to target fertilization input differentially either by carving your vineyard block onto homogeneous zones or by speeding up or slowing down tractor speed locally.
For pruning decisions, if an area appears to show a reduced growth by veraison, it is possible to prune more severely during the winter. By leaving less growing points on the vine for next season, you will promote a greater individual shoot growth rate which eventually helps the vine to achieve a better balance between leaf area development and fruit production.
For inter row management decision, planting a competitive cover crop on a very vigorous block section will be helpful to slow down leaf area growth; while it will be relevant to sow legumes on block section showing nitrogen deficiencies.
Burgundy vineyards (Source)
An NDVI map at the beginning of the season (May or June) is also relevant for vineyard management decisions. Since shoot growth rate is maximal between the fruit stages of “split flower” and “pea- size”, an early season map is useful to discriminate vineyard areas with abnormal functioning. The analysis of an early map can help you diagnose areas where plant mineral assimilation (like nitrogen) are disrupted or suboptimal. From there, it becomes possible to apply foliar sprays between the “pea-size” and “veraison” stages, as the vine is still being receptive. The work published by Schreiner (2016) has shown that vine assimilates more than 25% of its mineral needs between these two phenological stages.
All these cultural practices, once adapted and localized, aims at reducing or even erasing the spatial variability within the block. When uniform areas are too fragmented, vineyard management becomes extremely complex for the wine grower, which can taint the final quality of the block production.
By uniformizing the block, you make the work of the winegrower easier and consequently you reduce management cost. In addition, regular NDVI pictures, taken at key stages of each season will measure the success of winegrower actions from one year to the next.
At Fruition Sciences, we acknowledge uniqueness, natural abilities and potential growth for plants and people. While respecting tradition, we provide winemakers with a highly integrated, terroir and vintage specific, data driven web application.