Three technologies the Institute did not invent, and one chain linking them that has not been assembled before. Each panel states what is established in the literature, and what remains untested on this site.
The two biological tracks are kept apart by one boundary: invasive biomass GHRI removes would be transferred to a separate company to process, while endemic propagation is GHRI's alone, under a binding commitment not to commercialise it.
Lignin is the polymer that makes wood woody, and almost nothing digests it. White-rot fungi do — they secrete oxidative enzymes that dismantle it from the outside. Cut cane becomes substrate, the fungus does the conversion, and what it leaves behind is a feed and a soil amendment.
Growing mushrooms on crop residue is a global industry of more than 40 million tonnes a year. Oyster mushrooms are roughly a fifth of it.
White-rot Agaricomycetes are the only organisms known to substantially decay lignin. Laccase, manganese peroxidase and lignin peroxidase oxidise it non-specifically, from the outside. Floudas et al. 2012 · Science 336:1715
42.8 Mt world production (2020). Biological efficiency 25–55 % on supplemented straw substrates; 2.52 % on unsupplemented walnut shell — woody feedstock sits at the low end without nitrogen supplementation. FAOSTAT via Okuda 2022 · Sözbir 2015 · Miśkiewicz 2025
Spent substrate is an established ruminant feed at 10–15 % DM inclusion. Above 20 %, intake and feed-to-gain both worsen. It inherits the toxicology of whatever it was grown on. Xu et al. 2024 · Front Microbiol 15:1425218
No published trial has grown oyster mushrooms on Rubus niveus cane. And cultivated oyster mushrooms have escaped elsewhere and halved native fungal diversity where they spread — on an island that is a biosecurity question before it is a production question.
Pleurotus citrinopileatus escaped cultivation in North America; invaded elms carried roughly half the fungal richness of uninvaded ones. Current guidance recommends prohibiting outdoor cultivation of non-native fungi. Current Biology 2025 · Stallman & Allen 2026
Design intent is a locally isolated white-rot strain, not an imported one. ABG has no published position on cultivated fungi; this is to be settled with the regulator, not assumed.
Nearest analogue is a single unrefereed thesis: woody forest invaders performed comparably to oak. Lignin content, particle size, supplementation regime and pasteurisation schedule for R. niveus cane are all undetermined. Kaszynski 2017 · MSc thesis, UW–Madison
GHRI removes invasive biomass and would transfer it. A separate company would process it; GHRI would not operate that company. Removal targets are set by the restoration plan and by the regulator, not by processor demand.
The boundary is coordinates, not wire. A collar sounds a rising tone as the animal nears the line; only if it keeps going does a brief pulse follow. The animal is not learning a map — it learns that the tone predicts the pulse, and turns at the sound. Once trained, the pulse mostly stops.
Cattle learn the cue in about three weeks. In a burned sagebrush trial, virtually fenced cattle took under 3 % of the forage in the recovering area; unfenced cattle took nearly 70 %.
Stimulus: 82 dB rising tone at 1 m, then 0.2 J at 3 kV for 1.0 s. Across 13 studies cattle learned the association in a mean of 20.8 days, with the pulse-to-tone ratio falling significantly over time. Wilms et al. 2024 · J Anim Sci 102:skae108
Exclusion from regenerating Eucalyptus saplings held for 99.8 % of the trial; 74.5 % of 4,378 cues were audio-only. Burned sagebrush: <3 % vs ~70 % forage use against unfenced controls. Campbell 2020 · Animals 10:1069 · Boyd 2022 · Rangel Ecol Manag 81:55
Manufacturer reports 200,000+ collars across six countries. Legal in NO, UK, IE, ES, US, SE; prohibited in DE, AT, CH, FI. Deployment figures are company-reported and not independently audited.
Published trials hold animals at paddock scale. Collar position error on a moving animal is around 11 m, worse under canopy and on rough ground. Holding cattle out of a 10–15 m planting group is smaller than the error itself — that is a trial to run, not a specification to publish.
Mobile accuracy error 10.8 ± 6.8 m, against 4.6 ± 3.2 m static. 2.9× worse at high terrain ruggedness; closed canopy degrades both precision and accuracy. Versluijs et al. 2024 · Anim Biotelem 12:33
Published guidance recommends a 5–10 ha minimum cell and a ~30 m buffer between a virtual line and any feature that matters. A 10–15 m radius zone is 0.03–0.07 ha — two orders of magnitude below that floor. Versluijs 2024 · Univ. Arizona Ext. AZ2089
No peer-reviewed study validates exclusion at this radius, and none applies virtual fencing to invasive-species control. Both are extensions of the technology, not established uses of it.
A plant cell keeps the instructions for the whole plant. A fragment of shoot on sterile medium, pushed toward shoots by cytokinin, makes more shoots instead of growing on as one. Divide and repeat — the multiplication is exponential across cycles, not additive.
A mature industry, 1.5–2 billion plants a year, and the accepted route for the roughly one third of critically endangered plants whose seed cannot be banked. Cylindrocline lorencei, an island-endemic daisy extinct in the wild on Mauritius, was brought back in vitro and returned to the field in 2024.
9 shoots per explant per 21-day cycle on semi-solid medium, 24 in temporary immersion; 1,496 ± 110 after six cycles, with no polymorphism detected across 11 ISSR primers. Uma et al. 2021 · Sci Rep 11:20371
The honest comparator is a threatened wild tree, not banana: 10.2 ± 0.22 shoots per 4-week cycle, 93 % rooting, 70 % acclimatisation survival. Dang et al. 2011 · AoB PLANTS plr012
36 % of critically endangered species produce recalcitrant seed unsuitable for conventional banking. BGCI treats tissue culture as a required route for these exceptional species. Wyse et al. 2018 · Nature Plants 4:848
There is no published in vitro protocol for any Scalesia. The one attempt, in 2006, lost its Scalesia cultures to contamination and browning. Protocols are species-specific and have to be developed. That gap is the work — not a footnote to it.
High contamination and lethal browning across all three disinfection treatments for Scalesia affinis; S. affinis appears in none of the paper's conclusions, and no follow-up has been published in twenty years. Izquierdo 2006 · Lyonia 9(2):57
Only 13.7 % of timber tree species have any micropropagation study; 6.4 % of wet-tropical species. There is no citable figure for how long protocol development takes — do not publish one. Ley-López et al. 2026 · Biodivers Conserv 35:156
Somaclonal variation rises with subculture number, and micropropagation multiplies genotypes rather than creating them. Restoration stock needs many founders modestly multiplied, not one line mass-produced. Majumder 2025 · CPC best practice: ≥50 founders