Quick check: do Wiltshire long barrows avoid woodland?

Date: 9 Sep 2026 (Europe/London) · n barrows: 129

1. Ancient Woodland (Natural England AWI) — preferred historical proxy

AWI = woodland continuously present since ~1600 AD. County cover ≈ 3.65% (127.24 km² of 3484.78 km²).

Metric	Long barrows	County random (n=2000)	County area fraction
Inside AWI	1 (0.78%)	3.7%	3.65%
≤100 m of AWI	3 (2.33%)	7.4%	—
≤250 m of AWI	7 (5.43%)	13.5%	—
≤500 m of AWI	10 (7.75%)	22.1%	—
Median distance	2621 m	1261 m	—

Only 1 / 129 barrow is inside AWI: West Woods (SU 15683 65628). Barrows are ~3–4× less often in/near AWI than random county points; median distance to AWI is about double the random baseline.

2. Modern woodland proxy (OSM `natural=wood` + `landuse=forest`)

Crowd-sourced; incomplete; includes plantations & small copses. County cover ≈ 7.28%.

Metric	Long barrows	County random	County area
Inside OSM woodland	14 (10.85%)	7.1%	7.28%
≤100 m	37 (28.68%)	32.95%	—
≤250 m	79 (61.24%)	63.3%	—
Median distance	195 m	176 m	—

No avoidance of modern OSM woodland — rates are similar to (or slightly above) the random baseline. Small copses/plantations sit near many chalk monuments.

Verdict for Tim

Partial map pattern, weak Neolithic claim.

Yes: barrows strongly under-sit surviving Ancient Woodland vs a Wiltshire-wide random baseline.
No: they do not avoid modern OSM woodland generally.
Best reading: long barrows concentrate on open chalk (Wessex Downs / Plain fringes). Surviving AWI is concentrated on clay vales & bigger woodland blocks (e.g. Savernake / West Woods). So “empty chalk pockets ≈ not woodland” is mostly chalk vs clay geography, not a recoverable Neolithic “avoid trees” rule from AWI alone.
Little overlap with your 52 hillforts is still consistent with different landform/function niches — woodland is only one weak correlate.

Go / no-go: NO-GO on “long barrows avoid woodland” as a Neolithic behavioural hypothesis from this quick check. GO only as a descriptive modern/early-modern pattern: “scheduled long barrows sit away from surviving AWI patches because they sit on open chalk.”

Caveats

AWI ≠ Neolithic woodland (1600 AD continuity).
Survival / visibility bias: monuments in woodland are harder to detect and schedule.
County-wide random baseline ignores chalk-only null models; a chalk-downs-only randomisation would shrink the AWI effect.
OSM modern layer is not NFI/CORINE — directionally useful only.
No pollen / mollusc / charcoal reconstruction here.

Absolute paths: /workspace/map-overlays/woodland-check/

wiltshire_long_barrows_woodland.png
summary.html
stats_summary.json
barrows_woodland_distances.geojson
ancient_woodland_wiltshire_clipped.geojson

Quick check: do Wiltshire long barrows avoid woodland?

1. Ancient Woodland (Natural England AWI) — preferred historical proxy

2. Modern woodland proxy (OSM natural=wood + landuse=forest)

Verdict for Tim

Caveats

2. Modern woodland proxy (OSM `natural=wood` + `landuse=forest`)