How to Map a Food Forest You Already Have (Tree by Tree, With a Phone)

Ask the internet how to map a food forest and you get planning software: tools that help you design layered plantings, check shade and spacing, and project growth twenty years ahead. Those tools are good at what they do. But thousands of people are standing in food forests they did not design, inherited, or planted years ago without a map, and their question is the reverse one: how do I document what is actually here? A food forest is the hardest orchard to inventory precisely because it is not rows of one cultivar. It is dozens of species at different ages, in clumps, at layered heights, often planted by someone else. This article is a field method for mapping exactly that, using the phone in your pocket.

Why map a planting that exists only in your head

The case for a map is really a case for per-plant memory. Once a planting passes a few dozen plants, the questions you actually ask stop being answerable from memory: which of the twelve avocado seedlings turned out worth grafting, which banana clump produced in the dry season last year, where the essay-worthy graft is on that multi-variety mango, what died over winter and left a gap. A map turns those from archaeology into lookup. It also transfers: a documented food forest survives a change of owner, a volunteer crew, or your own twelve-month absence, because the knowledge lives on a per-plant page rather than in one person's head.

What the existing tools do (and do not do)

It is worth being precise about the current landscape, because it explains why food forests so often go unmapped. The design tools, like AgroForest Designer and LiveStacks, are built for the planning phase: AgroForest Designer maps your site and helps you review spacing, growth and indicative shade before planting, with a free first project; LiveStacks focuses on plant research, guild-building and property maps, and markets itself to permaculture design course students and instructors. At the other end of the spectrum, Falling Fruit crowdsources publicly accessible foraging locations, and small farm apps like Fruit Forest App let you pin trees, add species and notes, schedule tasks and organize zones. What none of these do is take a walk around an existing, unstructured planting and turn your photos into individually identified plants on a map. Design tools assume a designed site; pin-by-pin apps assume you already know what every plant is and where it stands. The un-designed, half-documented food forest falls between them.

The method: photograph, cluster, identify, enrich

The method has four steps, and the first two cost nothing but discipline.

1. 1Turn on GPS photo capture and walk the site slowly, photographing each plant from two or three angles: one shot framing the whole plant, one close on a leaf or bark or fruit for identification, one on the base of the trunk. The geotag on each photo is what lets software tell plants apart later, so do not skip it.
2. 2Cluster the photos into plants. Photos taken within a metre or two of the same trunk belong to the same plant, so location plus judgment collapses a few hundred frames into a clean list of individuals. This is the step that makes the method scale: you are not maintaining a spreadsheet while you walk, you are shooting, and the structure happens afterwards.
3. 3Identify what you photographed. For known species this is typing a name. For the plants you inherited or cannot name, an identification tool closes the gap (CanopyTwin's free plant identifier works on any phone browser with no account). Guess honestly: "likely mangosteen, unconfirmed" is a usable record; a confident wrong name is worse than none.
4. 4Enrich each plant with the fields you will actually use: species and cultivar if known, layer (canopy, understory, shrub, ground), planting year or age estimate, rootstock and graft info where relevant, and condition. Then let the record grow: every future visit, photo, count or note attaches to the same plant.

Worked example: the clump problem

The one genuinely tricky decision in a food forest is what counts as a mappable unit. Rows of apples make it obvious; layered polycultures do not. A worked example: a 0.4 hectare tropical food forest with roughly 30 species contains a banana patch of maybe 40 pseudostems, a clump of three surf cherry trees grown from one seed ball, and a jackfruit with five saplings volunteered beneath it. Mapping every pseudostem would be madness. The workable rule: map a unit when you would ever want to ask a question about it as a unit. The banana patch becomes one entry (a banana mat), possibly with a photo count of pseudostems; the three surf cherries become one entry only if you will never treat them differently, otherwise three; the five volunteer jackfruit saplings are five entries if you intend to select and graft one, which means you will. In practice a mixed planting of a few hundred visible plants tends to collapse to somewhere between 80 and 150 tracked units, which is a very manageable inventory. Write your rule down before the walk, because the worst time to invent a convention is mid-survey.

What to record per plant

• Identity: species, cultivar, and layer. Mark uncertain identifications honestly.
• Age and origin: planting year, seedling or grafted, rootstock, who planted it if you know.
• Location: GPS position captured automatically by the photo, corrected by eye where canopy deflects the fix (under canopy, phone GPS can be off by several metres, so drop the pin on the trunk when the map lets you).
• Condition and history: any treatments, pruning, or damage, and dated photos so the plant's story is auditable later.
• Productivity: fruiting season, counts from photo surveys, and any taste or performance notes that make the difference between keeping and culling a seedling.

The test of a good record is a stranger question. Hand your map to a visitor: can they find the abiu, tell it is on its own roots, and see it set fruit last October? If yes, the map is done. If they must ask you, keep enriching.

Keeping the map alive

A map made once decays as fast as the planting grows. The fix is cheap: re-photograph. A survey walk repeated a few times a year refreshes every plant's photo history, catches new volunteers and dead plants, and starts building the time series that makes records interesting. Per-plant fruit and flower counts from those photos (which CanopyTwin's AI produces from survey photos) turn "it fruited okay this year" into a number you can compare across plants and seasons, which is the raw material for the only real food forest question: which of these thirty species actually earn their space here.

CanopyTwin exists because we hit this exact problem on our own mixed tropical plantings. You walk the site photographing plants with your phone, the GPS in each photo clusters the frames into individually tracked plants on a satellite map, and every plant gets a permanent page: photos over time, species from the 85-species tropical and rare-fruit database, notes, and counts. Unknown plants can be identified as you go with the free plant identifier, and from the Orchard plan up everything exports as CSV if you want the data elsewhere. The Grower plan covers 50 plants at $12 per month with a 30-day free trial and no credit card up front; the demo farms on the site are real mixed orchards you can browse to see the end result before you start your own survey.