Multi-Stem Tree

The dividing line is where the union happens relative to breast height (1.3m / 4.5 ft). If the stems separate below that point, it's a multi-stem tree and you record it as such — the tree itself doesn't have a single DBH, because there's no single stem to measure at that height. If the stems separate above breast height, it's a single-stem tree with codominant stems higher up — you take one DBH reading on the shared lower trunk as normal, and note the codominant structure (and any included bark at the union, a known structural weak point) as a defect rather than a separate stem measurement.

Measure DBH on each stem independently, at breast height on that stem (not on the shared root flare below the union). Record the readings together, largest to smallest by convention, as a list — "24, 18, 12" cm rather than 54cm combined or 18cm averaged. If a survey wants condition or notes at the stem level too (one stem has a cavity, the others don't), record those per stem rather than for the tree as a whole — a single condition rating for the whole tree can hide a real defect on one stem.

Where a calculation needs a single figure for the whole tree — most notably a root protection area, which is root-based rather than stem-based — use the aggregate diameter method (square root of the sum of each stem diameter squared), not a simple average or sum. That's a separate step from recording the raw measurements, and it should happen after the field data is captured, not instead of it.

See the tree inventory guide for how multi-stem trees fit into a broader survey, or LogLog's multi-stem editor — a dedicated marker mode adds and removes stems on a tree, each with its own DBH, condition, and notes, so you're not squeezing five readings into one text field by hand.