The Evidence Layer for Scientific AI

 Quick Explanation Copied

 Long Explanation

What Wood Is, Structurally

Wood is secondary xylem: a hierarchical natural composite of cellulose microfibrils wound around hollow cells, embedded in lignin and hemicellulose. Its two most predictive micro-structural parameters are basic density and microfibril angle (MFA) — the angle cellulose fibrils make with the cell axis. In nine-year Eucalyptus in Sumatra, MFA ranged only 9.2–11.0° across three species, yet together with basic density (0.41–0.46 g/cm³) it predicted stiffness (MOE) and strength (MOR) better than either parameter alone Eucalyptus wood property survey.

Note: reported E. urophylla MOE/MOR values in the source table contain formatting ambiguities; E. grandis and E. pellita values shown are unambiguous Data caveat for Eucalyptus table.

Wood Changes With Age: Juvenile vs Mature

The same tree makes two different woods. Near the pith, juvenile wood has shorter fibers, larger MFA, and lower density and strength; properties then stabilize toward the cambium. In Japanese yellow-poplar (Liriodendron tulipifera), all ten measured properties changed logarithmically near the pith and stabilized after cambial age ~9 years (~14 cm diameter) — notably later than peak growth increment (CAI at 4.9 yr; MAI at 7.4 yr), implying a growth–quality tradeoff Xylem maturation timing study. A 10-species Chinese study (70 trees) generalized the juvenile/mature pattern across conifers and hardwoods Ten-species juvenile/mature comparison.

Measuring and Breeding Wood Quality

Because destructively testing timber is slow, near-infrared spectroscopy calibrations now predict properties from increment cores: across 3,323 cores from 14 pine species in 5 countries, R² was 0.77 (air-dry density), 0.82 (MFA), and 0.901 (stiffness) Global NIR wood-property calibrations. Genomic selection adds prediction across generations: in ~6,000 phenotyped, ~2,500 genotyped Norway spruce, pedigree-based (ABLUP) and marker-based (GBLUP) models showed comparable predictive ability for wood properties Spruce genomic prediction validation.

Environment vs Genetics — and Blind Spots

For Victorian Douglas fir (~55 yr, 24 trees), green-timber density (372–432 kg/m³) and strength differed significantly by locality, but after air-drying locality effects vanished; strength grouped as S6/SD6, weaker than mature North American timber Victorian Douglas fir properties. Growth form also matters: foxtailing Caribbean pines in Fiji were taller and straighter with only minor wood-density differences vs normals Fiji foxtail pine wood study.

Confidence and caveats: Key uncertainties include (1) single-site designs — the Eucalyptus and Liriodendron studies cannot separate genotype from environment; (2) small samples (nine Liriodendron trees, 24 Douglas fir); (3) possible provenance effects misattributed to site in the Douglas fir study; (4) unverified legacy extractions. Findings would be revised if multi-site trials showed consistent negative growth–quality correlations or if NIR/genomic calibrations failed on independent validation sets. What is missing: molecular mechanisms linking MFA variation to cell-wall biosynthesis, and long-term data on property stability.