The three figures are near-equivalents rather than exact ones, and the gap is small but worth knowing. Cutting 10% of a forest landscape every ten years turns the whole of it over in a century, so the 90-year average rotation offered as the equivalent is slightly more permissive than the cap itself: a 90-year rotation removes 11.1% of the area per decade [11]. The set-aside route resolves differently. Withhold a fifth of the forest, hold the landscape total at 10% per decade, and the remaining 80% can be cut on an 80-year rotation, because 0.8 of the area divided by 0.1 per decade is eight decades [12]. Reserving land does not buy extra harvest on top of the 90-year figure. It buys shorter rotations on less ground.
The design is a space-for-time comparison. Counts from Finland's national bird monitoring in the south and centre of the country, 2006 to 2020, were set against European maps of accumulated forest disturbance [3], so landscapes that have already absorbed heavy clearcutting stand in for the future of landscapes that have not. That is the practical way to reach a cumulative effect that single-harvest studies cannot see, and the gap in the literature was exactly that landscape-scale accumulation [10]. It also means nobody assigned clearcutting at random, and the report does not state what the analysis adjusted for [14].
Rémi Duflot reports a direction rather than a magnitude: common forest species and old-growth specialists were both less abundant where clearcutting had accumulated over 20 years [4]. How much less abundant, and across how many survey routes, the report does not say [14]. The 20-year accumulation window is also longer than the 15 calendar years of bird data described [15], which fits if the disturbance maps reach back further than the counts, though that is not spelled out. The limit itself is defined generously in one respect: it is the level below which old-growth-associated populations can stay stable or recover from earlier declines [6], not merely hold at their current depleted level.
What travels well here is the procedure. Pair a monitoring time series with a disturbance map, find the accumulation above which specialist abundance stops recovering, and express it as a share of landscape per decade; Mikko Mönkkönen says the method suits use worldwide [8]. The pass mark it produced is Finnish, from Finnish birds, in a 15-year window, and a scheme that writes 10% per decade into a standard adopts that provenance along with the number.
