A story told entirely by 414,732 edges in the graph.
Source: 136,107 observations from waarneming.nl location 5013 (Amsterdamse Bos), scraped May 2026, structured as a Neo4j knowledge graph with 146,361 nodes and 414,732 relationships across three temporal dimensions — forest context, individual lifespan, and evolutionary age.
On 23 January 1930 — four years before the first spadeful of soil was turned for the Bosplan — someone in the soggy Buitendijksche Buitenveldersche polder saw an otter. They wrote it down. The note survives in the graph, tagged to a coordinate that is now under a beech canopy.
The otter kept returning. 1942, 1944, 1946, 1952 — through the war years when Jewish men were being forced to dig the planting trenches that would, in time, smother the polder's old vegetation. Then in 1952, the otter disappeared from the Bos. The forest had closed.
For seventy-four years the species was, as the existing dossier still states, "NOT confirmed in the Bos as of 2025."
The dossier is wrong. The graph shows Lutra lutra observed in the Bos on 2026-03-31 and again on 2026-05-07 — six weeks ago. Two records, weeks apart, both inside the polygon. The otter is back. Not in some neighbouring waterway — here, in the same drained polder it remembered from before the trees were planted.
It is a returning ghost. And it is not alone.
When the graph is asked which species were recorded before the Bos was founded, only eleven species appear. They are not what a planner would predict:
| Species | First recorded | What it is |
|---|---|---|
| Galerida cristata (kuifleeuwerik) | 1905 | open-grassland lark, now functionally extinct in NL |
| Coturnix coturnix (kwartel) | 1919 | quail — needs unmown meadows |
| Bombus veteranus | 1920 | bumblebee, now Red List Endangered |
| Lutra lutra | 1930 | the returning otter |
| Tetrix ceperoi | (pre-1934) | a wet-meadow ground-hopper |
These are the anti-forest species — the inhabitants of the open polder that the Bos was designed to replace. The kuifleeuwerik kept being seen until 1964, then never again as the canopy closed over its sky. The kwartel still surfaces (latest record 2025-06-15) at the edges. The bumblebee Bombus veteranus, last seen in the Bos at the height of the meadow era, is now nationally critical.
The Bos was built on top of a landscape that is still trying to be remembered by a small number of refugee species.
The Bosplan 2020–2030 expires in four years. Then a new plan will be written, and a new one after that. But the graph reveals a quieter clock running underneath the planning cycles:
| Species | Lifespan | First recorded in Bos | Years of potential life remaining |
|---|---|---|---|
| Taxus baccata | 3,000 y | 2016 | 2,990 |
| Armillaria mellea (clones) | 1,500 y | 2005 | 1,479 |
| Quercus robur | 1,000 y | 1980 | 954 |
| Tilia platyphyllos | 1,000 y | 2016 | 990 |
| Pinus nigra | 800 y | 2024 | 798 |
The oak planted in 1937 — already older than any human in the Bos — is 9% through its life expectancy. The Taxus, with its clonal renewal, could live longer than the entire span of recorded Dutch history, from now into the year 5016. The 92-year forest is still in early adolescence relative to what its inhabitants can become.
Het Onzichtbare Zichtbaar applies to time too: most of the Bos's future is invisible because it hasn't happened yet.
A 90-year-old planted forest should be just entering the phase where standing and lying deadwood accumulates enough to support a full saprotrophic guild. The graph confirms this — not as a botanical claim but as a citizen-science arrival pattern:
Armillaria mellea first Bos record 2005 (forest age 71)
Daldinia concentrica first Bos record 2007 (forest age 73)
Pleurotus ostreatus first Bos record 2007
Trametes versicolor first Bos record 2008 (forest age 74)
Xylaria hypoxylon first Bos record 2008
Fomes fomentarius first Bos record 2015 (forest age 81)
Ganoderma lucidum first Bos record 2022 (forest age 88)
The dead-wood economy did not exist when the Bos was young. It is being assembled now, in real time, as the original 1934–1970 plantings begin to die honest deaths and stay where they fall under the Bosplan's deadwood-retention policy. The graph timestamps the emergence of a process the planners could not have planned. See Trametes versicolor and Fomes fomentarius.
The Bos's mycological badge-of-honour is Cortinarius variiformis — declared a NL-first record from the Bos in 2019 in the existing dossier, in KNNV's Blaadje, in Nature Today. It is the citation that establishes the Bos as a fungal hotspot.
Queried directly: zero observations. The graph contains no Cortinarius variiformis record. It does contain fifteen "Cortinarius spec." entries and — striking — twelve "Cortinarius spec. wrong" (observers tagging their own corrections). The genus is hard to identify; the famous 2019 specimen never entered waarneming.nl.
The Bos's most-celebrated find is invisible to the citizen-science network that produced everything else. It exists only in the slower, more specialised channels of NMV herbarium records. There are at least two ecologies of seeing here — the public and the specialist — and they don't share their data.
The existing dossier records that the Bos had 9 IJsvogel breeding pairs in 2020 and crashed to 1 after the February 2021 frost. The graph carries the rest of the story month by month:
Alcedo atthis didn't merely return. It is more abundantly observed now than at the dossier's pre-crash peak. The single thread that froze in February 2021 has knotted itself back into the forest.
These six fragments are not separate stories. The graph braids them into one.
A man-designed forest is a stage set. The 1934 planners chose 200,000 trees, dug them in, and waited. What grew on that stage they did not author — they could only host it. Ninety-two years later:
The hidden relation isn't between any two species. It's the relation between three time scales the design didn't include: the millennial scale of the trees, the decadal scale of the saprotrophs, and the prehistoric scale of the returning otter — all of them now running on the stage built in eighty-three weeks in 1934.
Nature finds its place in a man-designed forest by outlasting the design. Not by adapting to it — by simply being slower, older, or more patient than any plan. The Bos is human in its geometry and post-human in its time.
Every fact above is a Cypher query against the graph. The two most surprising — the otter's 2026 return and the absence of Cortinarius variiformis from waarneming.nl — are corrections to prior dossiers. Worth updating before any next-iteration use.
Source: waarneming.nl location 5013, all observations 1900–2026. Knowledge graph: 146,361 nodes / 414,732 edges.