Planting Methods

Planting methods

Canopi is the only platform that models how planting method affects tree survival at the site level. The choice between manual planting and drone seeding isn’t just a logistics decision — it changes the probability of success.

Why method matters

A seedling planted by hand into prepared soil has a fundamentally different start in life than a seed pod dropped from 50 meters by a drone. The seed-soil contact, initial rooting depth, microsite selection, and competition exposure all differ. These differences compound over time, especially in challenging conditions.

Canopi captures this by including planting method as a core model feature. The model learns from thousands of observations how method choice interacts with site conditions — soil type, slope, moisture availability, competition — to affect survival outcomes.

Manual planting

ID: manual

Hand-planted seedlings installed by field crews, typically with mechanical or manual site preparation. Benefits include careful microsite selection by experienced planters, good root-soil contact, and the option for site prep (scalping, mounding) that reduces competition.

Manual planting generally shows higher survival rates in the model, particularly on challenging sites — steep slopes, rocky soils, and high-competition areas where careful placement makes the biggest difference.

Drone seeding

ID: drone_seeded

Aerial dispersal of seed pods or seedling capsules via drone. Benefits include speed, cost-efficiency at scale, and access to terrain too steep or remote for field crews.

Drone seeding survival predictions are generally lower than manual, reflecting reduced seed-soil contact and less microsite optimization. However, the differential varies significantly by site. On accessible, moderate-slope sites with good soil, the methods may converge. On steep, rocky, or heavily vegetated sites, the gap widens.

Using the compare endpoint

The /v1/predict/compare endpoint is designed specifically for this decision. Send a site and species, and get both methods’ predictions with an explicit differential:

curl -X POST https://api.canopitech.ai/v1/predict/compare \
  -H "X-API-Key: cnpi_live_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{"latitude": 44.5, "longitude": -122.0, "species_code": "PSME"}'

The method_differential in the response tells you exactly how much survival probability changes between methods. If the differential is small at your target site, drone seeding may be the economically rational choice despite slightly lower survival. If the differential is large, manual planting may justify the higher per-tree cost.