停止砍伐森林 · 修复森林 Stopper la déforestation · Restaurer les forêts
Referee's note · 裁判说明 · Note de l'arbitre
Two teams. One ball. A scoreboard that has been lying to us politely for thirty years — because a hectare is not a hectare, and a year is not a year.
很多年来我们被一个定式带着走：只要有人种树，人越多就越好。但比分不是这样算的。同样写作“公顷”，一边是三百年，一边是三天。
Team B's 5.2 Mha/yr includes plantations. A three-year-old eucalyptus row is counted like a 300-year-old forest. Carbon, water and species are not. 种植园也算“森林”。
A chainsaw scores in an afternoon. A seedling needs 10–20 years of care before it is a tree, and often 20–100 years before it repays the carbon that was lost. 砍一天，种二十年。
“Net loss −4.7” sounds gentle. It is 10 Mha destroyed and 5.2 Mha started elsewhere — different places, different biomes, different centuries. 净值掩盖了真实损失。
Team B's shots take twenty years to land. Team A scores in an afternoon.
B队的球，要二十年才落进筐里。A队一个下午就能得分。
« Les tirs de l'équipe B mettent vingt ans à retomber. L'équipe A marque en un après-midi. »
The regional scoreboard · 分区域比分 · Le tableau par région
Globally Team A leads. But the game is not global — it is eight or nine local games with wildly different scores, and two of them are being played where no camera is pointed.
Un seul championnat, neuf matchs très différents.
GFW/WRI 2025 · INPE PRODES
GFW/WRI 2025
GFW/WRI · Forest Declaration Assessment
GFW/WRI · FAO FRA
FAO FRA 2020 · Trase
Forest Survey of India · FAO FRA
European Commission · WWF 2021
GFW/WRI · GWIS · FAO FRA
FAO FRA 2020 · AFR100
Finance, 2010 → 2050 · 资金对比 · L'argent sur le terrain
Rule for this section: if a flow of money causes clearing with no clear compensation, it plays for A. If it protects or regrows forest, it plays for B. Bars show the order of magnitude, not a precise ratio — the figures beside them are the citable part.
结论先说：绝对数字更糟了，方向第一次清楚了。2010年没有任何规则说钱不能砍树；2025年有了。
350,000,000 ha restoration goal (Bonn Challenge, by 2030) × US$1,000–3,000 per ha establishment + 5–10 years of care (published project ranges) = US$0.35 – 1.05 trillion total ÷ 10 years = US$35 – 105 bn / year ← restoration alone + protecting standing forest (payments & opportunity cost) + land tenure recognition & enforcement + monitoring, fire response, community wages ────────────────────────────────────────────── ≈ US$460 bn / year published estimate of total forest-goal need vs ≈ US$2.2 bn / year dedicated international public forest finance ⇒ ≈ 0.5 % of the need — while ≈ US$7 tn / year still flows nature-negativeThe middle block is our own back-of-envelope arithmetic on published unit costs; the US$460 bn, US$2.2 bn and US$7 tn figures come from the Forest Declaration Assessment and UNEP (see sources). Please argue with the numbers — that is the point of the roundtable.
Human activity, sorted · 双方阵容：人类行为分队 · Les effectifs
Every player here is a human behaviour, not a force of nature. Tags say which way each one is moving: surging, braking, or unwatched — the third being the category that decides games.
Around half of all deforestation. Soy, palm, cocoa, coffee, maize. surging
Roughly a further 38%. Braking in the Brazilian Amazon, accelerating in Bolivia, Paraguay and the Cerrado. mixed
In 2024, fire — not the chainsaw — became the leading cause of tropical primary forest loss for the first time on record. surging · new star
Degradation removes structure and carbon without ever registering as “deforestation”. In tropical moist forests it rivals outright clearing in area. unwatched
Small footprint, enormous leverage: a road is a permanent invitation to the frontier. unwatched
Record Amazon droughts; patches of forest flipping from carbon sink to source. A player we recruited by accident. biggest hidden risk
Plantations counted as forest; net figures published instead of gross. A scores points in the spreadsheet, not only in the field. unwatched
Nature-negative public and private flows (UNEP, 2023), including ≈US$540–700 bn/yr of largely unconditional agricultural support and US$300 bn+ of credit to forest-risk commodity companies since 2016.
Let it come back. Cheapest per tonne of carbon, fastest to biodiversity, and almost entirely unfunded because you cannot photograph a groundbreaking. underplayed superstar
At least a third of the world's remaining intact forest sits on Indigenous peoples' lands, with consistently lower loss rates. Best defence on the court; worst paid. strong · underpaid
Amazon Soy Moratorium, Indonesian peat moratorium. Proof that a rule can move a curve within one season. holding
The EU Deforestation Regulation — the first attempt to make an importer legally responsible for a distant hectare. Repeatedly delayed. lagging
REDD+, LEAF, the Amazon Fund, and now a facility to pay per hectare kept. Growing — while the integrity of forest carbon credits is under serious scientific challenge. growing · contested
China, India, AFR100, Bonn Challenge: 210+ Mha pledged of a 350 Mha goal. Huge numbers, thin survival data, wrong-biome risk on grasslands and savannas. big numbers · weak evidence
Funding usually ends at year 3. Trees need twenty. This is the hole in B's roster, and nobody is scouting for it. the real risk
Dedicated international public forest finance — against an estimated need of ≈US$460 bn/yr. Even counting all nature-positive flows (≈US$200 bn/yr), B is playing four-on-five with borrowed shoes.
Five plays that decide the decade · 五大关键 · Les cinq priorités
Not five wishes. Five places where a decision taken before 2035 changes the score in 2050.
Make farm support and bank credit conditional on deforestation-free supply. This single lever is bigger than every restoration budget combined.
先关掉“砍树也能拿钱”的水龙头。2024 changed the game: fire beat agriculture. Fire crews, early warning and drought-resilient landscapes are forest policy now.
火，已经是A队的头号球员。Keeping is cheaper than replanting by an order of magnitude. Per-hectare payments for intact forest, with honest measurement, beat offset arithmetic.
保住一棵老树，胜过种十棵小树。Recognised Indigenous and community tenure is the cheapest proven protection on the board. Fund it directly, not through five intermediaries.
把土地权利还给守护森林的人。Report trees alive, not trees planted. Default to natural regeneration. Budget maintenance, monitoring and wages for two decades — or don't start.
种下去只是开始，第3到第20年才是胜负。The wood wide web · 森林的万维网 · Le réseau souterrain
The Canadian forest ecologist you are remembering is almost certainly Suzanne Simard, of the University of British Columbia — her 1997 Nature paper on carbon moving between Douglas fir and paper birch through fungal networks is the reason a journal headline coined the phrase “the wood wide web”. Her book is Finding the Mother Tree (2021).
你说的《大树之歌》可能是两本书记混了：加拿大科学家 Suzanne Simard 的《寻找母树》(Finding the Mother Tree) 讲的是菌根网络与“母树”；而《树木之歌》(The Songs of Trees) 是美国生物学家 David Haskell 写的。地下网络这条线，属于前者。
So: do trees pass carbon and nutrients to each other? Measurably, yes — in experiments, isotope-labelled carbon moves through mycorrhizal fungi between plants. Is there a maternal tree deliberately provisioning her children across a forest? That is where science and story part company, and honesty here matters more than wonder.
If forced to pick: Team B — but scouting Team A, because B only ever wins by taking A's money off the table. Planting more trees while the subsidies keep flowing is running faster on a treadmill.
Why C, though. A and B are not two industries; they are two columns of the same balance sheet, and almost always the same actors — the same bank, the same ministry, the same shopper. A merged Team C has one rule: no hectare enters the books without its ledger. Cleared, compensated, and kept alive to year twenty, or it doesn't count. Merging is not a truce. It is finally making both sides show up in the same accounts.
为什么合并成 C 队：A 和 B 从来不是两个行业，而是同一张资产负债表的两栏，往往还是同一批人 —— 同一家银行、同一个部门、同一个消费者。C 队只有一条规则：每一公顷都要带着自己的账进来 —— 砍了多少、补了多少、活到第二十年没有。合并不是停战，而是让两边终于出现在同一本账上。
Pourquoi l'équipe C : A et B ne sont pas deux industries, mais deux colonnes du même bilan — souvent les mêmes acteurs. Une seule règle : aucun hectare n'entre dans les comptes sans son registre. Coupé, compensé, et vivant à la vingtième année.
It looks like fun — a match programme, two teams, a scoreboard. It is not a joke. Keep attention: this is deep dive 01, and the numbers here are meant to be attacked, corrected and updated in public.
这看着像玩笑，但不是。这只是第一次深潜 —— 数据欢迎被质疑、更正、持续更新。下一次带上啤酒，我们把账算细。
Every number above lives in one of these · 资料来源 · Sources