In recent decades we have witnessed a true technological revolution, where countless innovative technologies applicable to agriculture have appeared. Hand in hand with this revolution, the amount of available information has increased exponentially, which is a double-edged sword that can overwhelm even experts.
Figure 1. Drone in cherry orchard. Source: Avium Team.
Nowadays, through remote sensing techniques, it is possible to obtain information from a distance using sensors mounted on remote devices. The most popular devices are satellites and unmanned aerial vehicles (UAVs). The images obtained through remote sensing are composed of pixels, which represent numerical values of reflectance on a surface in different bands, that is, each of them represents a range within the electromagnetic spectrum.
Figure 1. Electromagnetic spectrum. Source: Adapted from Chuevieco 2010, by Torrealba 2013.
In cherry cultivation, different works and in different seasons evaluated by the Avium Team indicate that the vegetation index is not necessarily going to be correlated with a particular variable. In addition, it has been shown that NDVI and SAVI obtained from satellite images are not capable of predicting cherry phenology (Von Bennewitz et al., 2018). On the other hand, according to Blanco et al., 2020, the NDVI index is not related to the water potential measured with a Scholander pump in pre-harvest, but it is related to the post-harvest with a coefficient of statistical determination r2 of 0.6.
Avium has been working hard on the incorporation of these new technologies applied to cherry cultivation with promising results. Conventional management is usually implemented in a homogeneous manner for an entire surface, based on an average that is not necessarily representative of the same, completely forgetting a component that is key in precision agriculture and a solid foundation for agriculture 4.0: spatial variability.
Figure 2. Avium team on the ground programming drone flight plan in cherry orchards.
The visualization of variables in maps with striking colors is a key pillar in understanding spatial variability and can be explaining underlying problems that can take years to be identified or misdiagnosed without these types of tools. In the same context, the use of multispectral images allows the calculation of indices for each pixel associated with a specific location and its colorimetric representation based on a classification system, which is also debatable. To analyze spatial variability, it is not entirely correct to compare two variety/rootstock combinations within the same histogram to be classified, since logically the less vigorous combination will have lower values of a vegetation index, altering the map's visual.
In literature there is a large number of multispectral indices which are the result of mathematical operations between bands of the electromagnetic spectrum. The equipment Tech Avium, using the logic of spatial variability representation, compared the spatial patterns visualized from different multispectral indices, identifying that the spatial distribution is very similar in cherry cultivation regardless of the index used. In this first instance, the objective was to simplify and/or limit the indices to be used to facilitate the interpretation of results, give them a correct classification and define a reduced number of categories per index.
Figure 3. Spatial distribution of different vegetation indices in a Santina/Colt cherry orchard, KGB management system, San Vicente de Tagua Tagua, Chile. A) Chlorophyll; B) NDRE; C) NDVI; D) Nitrogen; E) Leaf moisture. Source: Copeval-Agrodreams.
Given the large amount of information and the similarity of results between indices, Avium proposes using the NDVI (Normalized Difference Vegetation Index) in cherry cultivation, which corresponds to the relationship between the reflectance of the near-infrared and red bands (Rouse et al., 1974).
NDVI= NIR-REDNIR+RED
Where, NIR: near infrared band; Red: red band.
In the case of cherry cultivation, it is not correct to assume that the NDVI vegetation index is exclusively associated with vigor. In fact, the definition of vigor is the characteristic of annual sprout development and not the size of the plants. You can perfectly have a large, voluminous plant, but not vigorous, that is, without or with little annual sprout development. It is not enough to just view the image from the comfort of an office; it is essential to complement the image capture with the reality of each particular case. Within the evaluations of this type of tool carried out by the Avium team, it has been seen that in an area a lower NDVI index can be associated with different problems, such as lower vegetative expression, irrigation problems, pest and/or disease outbreaks, differences in soil texture, weed populations, among others.
Figure 4. Areas with low NDVI index in cherry trees. A) Santina/MaxMa 14 foci of wood diseases. B) Lapins/Colt area with differences in soil texture and plants with Chondrostereum purpureum. C) Santina/Colt plants subjected to water stress, short irrigation lines. Source: Avium-Copeval-Agrodreams.
Another important analysis to take into account is to comprehensively observe spatial and temporal variability, as this allows for a more precise understanding of potential management zones. During the 2022-2023 season, Avium, within the evaluation of these new technologies applied to cherry cultivation, compared spatial patterns on different dates, specifically in pre-harvest, early post-harvest and late post-harvest periods, demonstrating that the patterns identified by the same type of classification are maintained over time, but not the numerical value of the same index.
Figure 5. Identification of management zones in Santina/Colt cherry orchards in KGB conduction system, San Vicente de Tagua Tagua, Chile. Source: Avium-Copeval-Agrodreams.
The correct use of this type of technological tools not only has a technical impact, but also an economic one. According to Figure 5, approximately 50% of the surface from an agronomic management point of view might not be a candidate for summer pruning, and this task could even be incorrect in this situation, since plant material would be removed, affecting the normal development of the plants in summer.
Among the most interesting indices evaluated on the ground by Avium, beyond the visible and near infrared spectrum, we find the thermal spectrum, which has great potential to identify different problems in real time, before seeing them expressed in the vegetation indices mentioned above. In Figure 6, temperature differences can be observed towards the northeast edge of the plot caused by excessive tension in the irrigation lines that caused them to become constricted, preventing the passage of water through them.
Figure 6. Using multispectral and hyperspectral images. A) NDVI index; B) thermal index; C: constricted irrigation lines. Source: Avium-UPL-Skyquest
The incorporation of this type of technology, relatively new in the agricultural world in recent decades, is truly a challenge. The excess of information from different sources and the lack of knowledge of how these technologies work can lead users to make errors in interpretation. For this reason, it is essential to accompany the results with an analytical, critical look from a technical point of view, verifying in the field the origin or origins of the spatial variability expressed in the images that allow decisions to be made, which in most cases will have an impact on the sustainability of the agricultural system, strongly on the economic and environmental pillars.
Bridging the gap in knowledge of technological tools applicable to agriculture is a topic to be addressed in the industry. It is necessary to dispel myths and propose strategies that are applicable not only from a technical point of view, but also from an operational one. The number of site-specific managements proposed according to the identified variability must be feasible and profitable from a technical-operational point of view, in order to achieve excellence in cherry production.
Special thanks:
Manuel Baumann, CEO of SkyQuest, has been the driving force behind Avium's vision for using multispectral images.
Agrodreams-Copeval Program. Which enabled the development of analytics and technical foundations for indicators in cherry cultivation.
Literature
von Bennewitz, E., Cazanga-Solar, R., & Carrasco-Benavides, M. (2018). Studying phenological stages of cherry (Prunus avium L.) using field observations and satellite-derived vegetation indexes.
Blanco, V., Blaya-Ros, PJ, Castillo, C., Soto-Vallés, F., Torres-Sánchez, R., & Domingo, R. (2020). Potential of UAS-based remote sensing for estimating tree water status and yield in sweet cherry trees. Remote Sensing, 12(15), 2359.
Chuvieco, E. 2010. Environmental remote sensing, the observation of the earth from space. New updated edition. Ariel Ciencia. Barcelona. pp: 591.
Rouse, J.; Haas, R.; Deering, D.; Schell, J. and Harlan, J. 1974. Monitoring the vernal advancement and retrogradation (green wave effect) of natural vegetation. NASA/GSFC Type III Final Report. 371 p.
