Autonomous Agents for Regenerative Ecologies Lab. How can autonomous agents contribute to or scaffold regeneration of degraded land?

• mineral content
• sunlight
• atmosphere
• latent seed stock? (typically desert flora can go dormant for extensive periods, but this is man made desert so how vital is the latent seed stock?)

• rain
• water retention
• organic matter in soil
• soil retention (erosion)
• microclimates
• economy
• infrastructure (electricity, etc)

• condensation (water harvesting, shadow) > out: pioneering vegetation
• ditches or 'contour ridges' (increase geodiversity) > out: water retention, erosion, microclimates, hydrologic corridors, ‘housing’ for ground dwelling fauna)
• biodegradable 'waterboxx' to grow trees (Pinus silvestris)
• mucuva: seeding an area with biodiverse mix of local species (increases tree diversity from 160 to over 2000 species per hectare initially and over 5000 after a decade)
• refugia (at large scale) vegetation pockets in the most favourable areas that serve as basecamps to grow from
• assisted migration (zoochory) > out: dramatically speeds up regeneration
• assisted evolution?? (like Cobalt-60 exposure)

• transplantation ecology (disperse topsoil samples) out: dramatically speeds up regeneration
• structural complexity enhancement (jump towards late-successional diversity) knock over some of the trees and expose roots, strip bark from some of the trees to create snags (standing dead trees) breaking up canopy cover, make piles of branches, or somehow bring in heavy longs from elsewhere > out: creates habitat for insects, birds, mammals, amphibians that mitigate disease, pests and renewed space for pioneering vegetation and the associated wildlife
• atmospheric water cycle: canopy cover starts to work as an atmospheric moisture pump (near coast or on mountain side)
• establish forestfire corridors

• primary trophic layer > pioneering vegetation / foundation species > annuals and algae > in: water / out: mine and accumulate available nutrients, create biomass, water retention, microclimates
• secondary trophic layer > insects, small herbivores > in: ‘accommodation’ / out: pollination, trophic
• third trophic layer > birds, carnivores > in: canopy or shrubbery / out: pest mitigation, zoochory (seed dispersal, microbe dispersal, nutrient dispersal)

• foundation species: a strong role in structuring a community (grass, trees, kelp, ants)
• indicator species (as a way of analysing system health) lichen
• keystone species: high regulatory impact in relation to its abundance (often predators) wolf, wild boar,
• nuisance species: that tend to discourage human occupation: leech, hogweed, mosquito
• hyper-acumulators (in case of toxins in the ecosystem) tuberous plants

• first stage: algae and annuals (herbs)
• second stage: shrubs and first trees (contour hedge rows to maximise impact)
• third stage: tree cover, nitrogen fixing trees: root deep, self seed, grow quickly, grow easily from cuttings (automated tree nursery)

(Essentially we're making a ‘vegetation forecast’)

• shade giving
• high water retention
• deep rooted
• fruit bearing (to attract more life)
• high characteristic return time: rate at which a population recovers from disturbance
• high jump dispersal: can cover long distances over inhospitable habitat in one generation (typically: seeds, insects, birds)
• high speciation (allopatric)

• low cryptic diversity (genepool)
• low tipping point
• low buffering capacity

• Albacete region in Spain (google maps)

see also: environmental_machine_learning