In clonal cherry orchards, every tree shares the same genetic background. Yet walking the rows reveals clear differences in structure and fruit load between neighbouring plants. Quantifying that variability at block scale remains a practical challenge, and averages systematically hide it.
Digital phenotyping aims to make this variability measurable and traceable at the individual plant level.
Phenoria is a smartphone-based tool we are developing for digital phenotyping of fruit trees, to extract plant information from a single photo per plant.
From that image, an AI model detects and counts individual fruits, segments the canopy structure, and quantifies variables such as foliar coverage, canopy density, foliage maturity, vertical foliage and fruit distribution, leaf-to-fruit ratio, etc.
Figure 1. Phenoria pipeline: from smartphone image capture to canopy structure segmentation and fruit counting.
Working in a commercial cherry orchard near harvest, we compared the structural phenotype of trees in the top 20% of fruit load (755 fruits/plant on average) against those in the lowest 20% (137 fruits/plant on average). The patterns diverged clearly.
High-load trees shared open canopies, lower foliar coverage, more mature foliage, and lower leaf-to-fruit ratio: fruit exposed, architecture tilted toward reproduction.
Low-load trees shared dense canopies, abundant active shoot growth, and much higher leaf-to-fruit ratio: fruit hidden, architecture tilted toward vegetative growth.
Figure 2. Comparison of structural phenotypes between trees in the top 20% and bottom 20% of fruit load within the same orchard block.
Mapping these phenotypes spatially within the block revealed distinct sectors concentrating high-load and low-load trees.
Interestingly, in the low-load sectors the trees appeared most vigorous from drone imagery, consistent with the dense canopies and active vegetative growth that Phenoria detected at ground level.
Figure 3. Drone-based map of the orchard block showing the spatial distribution of high- and low-load phenotypes.
A single smartphone photo cannot explain why those sectors diverge. The drivers often lie in soil, irrigation, or management history. But it can point precisely to where that explanation should be sought.
Image source: Tomás del Campo Smith
Tomás del Campo Smith
Phenoria
24 Apr 2026
Paclife Conference 2026 focuses on cherries, postharvest, automation and global markets, highlighting how data, innovation and integrated logistics are crucial drivers to ensure quality, competitiveness and value across the Chilean and international export supply chain.
11 Mar 2024
Cherry Growers Australia Inc demonstrates the innovative pergola training system. The system promotes early fruit production, optimal fruit quality and a host of other benefits, potentially achieving impressive yields of up to 300 kg per day.
02 Nov 2026
Phenoria uses a single smartphone image and AI to assess cherry tree canopies and fruit load. Comparing trees averaging 755 versus 137 fruits, the study shows how digital phenotyping reveals within-orchard variability and helps identify areas for further investigation.
02 Nov 2026
Claudia Soler examines the future of Chilean cherries after two difficult seasons, focusing on premium quality, changing consumer demand in China, export diversification into the US and Europe, stronger logistics and strategies to build a more resilient industry by 2035.
