The forest that was planted wrong

Two million trees were planted on a hillside. Eight years later about a tenth were alive, and the soil was in worse condition than before.

The species was the problem, and so was the counting. A fast-growing non-native conifer was chosen because it establishes quickly and because the programme's target was trees planted, a number that can be verified in the week of planting. Survival was not measured, because measuring it requires coming back. The conifer acidified the soil, shaded out the ground cover that had been holding it, and died in the second drought.

Restoration ecologists have a blunt rule: a planting target without a survival target is a target for digging holes. The useful figure is canopy cover after five or ten years, which is slow, expensive and inconvenient for anyone needing a result before the next budget. It is also the only number that corresponds to anything the forest does — holding soil, storing carbon, sheltering species.

The second lesson concerns what was already there. The hillside had a thin scrub that looked like neglect and was functioning as the first stage of natural succession. Clearing it to plant was the single most damaging act of the programme. In several studied sites, fencing the land against grazing and leaving it alone produced more woodland after fifteen years than planting did, at a twentieth of the cost.

None of this argues against planting trees. It argues against measuring the wrong thing, which is a different and more common failure. A programme that reports survival will plant fewer trees and grow more forest.