C‑DRONE
Aerial view of a green mountain valley and its stream

C-DRONE GUIDE · 7 SEPTEMBER 2026

AOP chestnut orchards by drone: gall wasp, multispectral mapping and Cévennes terraces, price

The Châtaigne d'Ardèche AOP alone accounts for close to 60 % of French chestnut production, grown on an orchard that climbs the slopes of the Cévennes on hand-built dry-stone terraces, in places up to 900 m above sea level. From 2007 the industry went through one of the most serious health crises in its history: the arrival of the chestnut gall wasp, an insect from China that biological control eventually got the upper hand on — though it still needs watching. On terrain a tractor cannot follow and an orchard too vast for an exhaustive walking survey, a multispectral drone flight offers a monitoring and inventory tool the Cévennes chestnut orchard has never had before. Here is what it genuinely documents, and what it does not replace.

Published on 7 September 2026, reviewed on 11 September 2026 — regulations in force as of September 2026.

An AOP orchard like no other, on terrain of its own

Recognised as a controlled designation of origin in 2006, then registered as a protected designation of origin by the Official Journal of the European Communities on 14 January 2014, the Châtaigne d'Ardèche AOP covers an area of 197 municipalities — 188 in Ardèche, plus a handful of neighbouring municipalities in the Gard and Drôme départements. The exploited orchard covers roughly 6,000 ha, for an annual output of around 5,000 t of chestnuts under the designation — close to 60 % of French production — grown by more than 600 producers.

What sets this orchard apart from most French fruit-growing land is its terrain. In the Cévennes valleys, chestnut trees are planted on dry-stone terraces — locally called "bancels" or "faïsses" — hand-built to hold the soil on slopes that nothing else would have made farmable, in places up to 900 m above sea level. The decline of farm mechanisation and the gradual abandonment of part of these plots, documented by the Cévennes National Park, have weakened some of these terraces over the decades. In practice, that means a good share of the AOP orchard is out of reach for any motorised ground survey: a tractor, a quad bike, or even a simple regular walk-through run into the slope, the dry-stone walls and the tangled canopy. That is exactly the terrain where a drone flight changes things: it covers in a single outing what a walking survey would take days to cover, without depending on the state of access tracks.

The chestnut gall wasp: a crisis receding, not a crisis closed

First reported in France in 2007 and durably established around 2010, the chestnut gall wasp (Dryocosmus kuriphilus), a small insect native to China, quickly became the most feared pest in French chestnut growing. The female lays eggs inside the buds; the following spring, the larvae trigger the formation of galls on young shoots and leaves, which reduce the leaf area available for photosynthesis and, in turn, fruit production — not killing the tree outright, but durably weakening it if the infestation persists.

The response took the form of a biological control programme now documented as one of the field's success stories: the introduction of Torymus sinensis, a tiny parasitoid wasp specific to the gall wasp and also native to China. After first releases in Italy from 2005, France began its own experimental releases from 2011, under the joint steering of INRAE and a committee bringing together the chestnut industry's main players, ahead of large-scale deployment across every producing region. The outcome, published by INRAE under the title "Le crépuscule du cynips du châtaignier" ("The chestnut gall wasp's twilight"), is clear: gall wasp populations are declining steadily and durably wherever the parasitoid has established itself.

That decline does not remove the need for monitoring: Torymus sinensis has not established itself perfectly evenly from one valley to the next, and a residual outbreak or a localised reinfestation remains possible on isolated plots or ones the parasitoid reached late. That is where a regular flight earns its place: spotting, across an orchard spanning several thousand hectares, the areas where the canopy still shows gall damage, without waiting for a grower to notice a localised yield drop on their own.

What a multispectral flight documents on the canopy

Research has already validated, on chestnut trees, what drone imagery can add to this kind of monitoring. A Portuguese team led by Luís Pádua, with Pedro Marques, Lino Martins, André Sousa, Emanuel Peres and Joaquim J. Sousa, published a random-forest classification method applied to drone multispectral imagery in 2020 in the journal Remote Sensing: at the scale of each individual tree crown, the algorithm first distinguishes the presence or absence of a phytosanitary issue from vegetation indices and spectral bands, then identifies whether the cause is biotic (pest, disease) or abiotic (water stress, deficiency), on chestnut plots actually affected by several phytosanitary issues, including the gall wasp (see the study on Google Scholar). Repeated season after season, this kind of monitoring lets a grower compare the same tree crown's health over time, rather than a single-point diagnosis.

Part of the same team — Pedro Marques, Luís Pádua, Telmo Adão, Jonáš Hruška, Emanuel Peres, André Sousa and Joaquim J. Sousa — had already published, in 2019 and also in Remote Sensing, a method for individual chestnut-tree detection from RGB and near-infrared imagery combined with a canopy digital surface model, reaching over 95 % segmentation accuracy and specifically designed for multi-temporal monitoring able to spot missing or newly planted trees (see the study on Google Scholar). Applied to a Cévennes orchard, that individual tree count serves a very concrete purpose: documenting, terrace by terrace, how a replanting effort is taking after removing trees too weakened to keep, or giving an AOP defence body or a regional forestry advisory body (CRPF) an up-to-date inventory of an orchard nobody has the practical means to recount on foot every year.

What a drone does not do: it does not replace the official phytosanitary diagnosis of a suspected outbreak — still a matter for a technician or an approved laboratory, working from a sample — nor the entomological monitoring of how well Torymus sinensis has established itself, which follows its own dedicated field protocol. It delivers a spotting and inventory dataset, leaving the industry to direct its field visits to wherever the map shows a change.

Method and 2026 prices

This kind of mission suits an AOP defence body or an interprofessional committee wanting to objectively assess orchard health across a sector, a chestnut cooperative monitoring its members, a forestry manager or regional advisory body (CRPF) supporting a replanting effort, or an individual grower whose holding is too large for an exhaustive walking survey. Sensor choice follows the same logic as for other perennial crops: multispectral for crown-level health spotting, high-resolution RGB for counting and individual tree inventory — our guide to choosing between a multispectral and an RGB camera details that trade-off.

Ranges observed in France in 2026, excl. VAT:

ServiceRange (excl. VAT, 2026)
Multispectral health-spotting flight, one terrace sector (up to 15 ha)€400 to €800
Health monitoring, 2 passes over the season (before/after leaf-out)€900 to €1,800 per sector
Individual tree count and inventory, high-resolution RGB orthophoto€600 to €1,400 depending on area and terrain
Post-replanting survival monitoring, across several seasonsquote-based, by number of passes
Grouped mission for an AOP body or cooperative, several holdingsquote-based, by number of sectors and area

These amounts cover the flight, index or orthophoto processing and map delivery; they exclude both the official phytosanitary diagnosis of a suspected outbreak (an approved laboratory or the plant health body, FREDON) and entomological monitoring of how well Torymus sinensis has established. On site, expect the same constraints as any mission in a steep-sided valley: sometimes degraded GNSS signal, terrain-driven wind, and care around power lines crossing some sectors of the orchard — our guide to forest health monitoring by multispectral drone details a related method on a different species. For an AOP defence body, a cooperative, a CRPF or a grower, request a quote stating the area, the terrain (slope, access) and the priority goal — health spotting or tree inventory.

Request a free quote

Also worth reading