Imported from human-avatar/skills-for-agriculture (skills/s4ag-biodiversity/SKILL.md). Install upstream with npx skills add human-avatar/skills-for-agriculture --skill s4ag-biodiversity. Copyright stays with the author.

Biodiversity

Functional biodiversity is not just an ecological good — it is a practical farm asset. Pollinators, beneficial insects, birds, and soil invertebrates directly underpin farm productivity. Your goal is to build habitat that supports ecological function across the farm: pest control, pollination, soil biology, and resilience. The most cost-effective entry point is almost always the farm boundary — and the biggest return often comes from doing less, not more.

How this skill works: Each sub-tool pauses at a Checkpoint to confirm the assumptions it is about to build on before producing output. A recommendation built on a wrong assumption wastes time and money — confirm the checkpoint before acting. Each sub-tool ends with Next steps — the skills worth running once you have acted on this one.

Expert Lineage

The thinkers whose frameworks underpin this skill — and what they specifically discovered that changes how you farm.

E.O. Wilson — Biodiversity Science and the Area-Species Relationship Wilson established that species diversity is directly proportional to habitat area and connectivity — the island biogeography principle. His practical finding for farmers: small, isolated habitat patches support far fewer species than connected patches of the same total area. A network of hedgerows, strips, and corridors is worth more than equivalent isolated blocks. His concept of the biodiversity hotspot shapes how this skill prioritises habitat investment: connect first, expand second.

Dave Goulson — Pollinator Biology and Habitat Requirements Goulson's research at the University of Sussex established that bumblebee populations require a minimum foraging range of 250–1,500 metres, and that floral resource gaps of more than 6–8 weeks in the flowering calendar cause colony collapse regardless of overall habitat quality. His actionable finding: it is not enough to create habitat — you must ensure a continuous floral sequence from March to October. A wildflower strip that flowers only in June solves nothing. Sequence is more important than total flower cover.

Aldo Leopold — Land Ethic and Edge Ecology Leopold identified the "edge effect" — the ecological principle that species diversity peaks at the boundary between two habitat types. Farm edges (hedgerow-field, woodland-pasture, stream-meadow) are the highest-diversity zones on any agricultural landscape. His practical implication: the ecological value of farm hedgerows, margins, and buffers is not incidental to the farm — it is the productive core of its ecological function.

George Monbiot — Trophic Cascades and Rewilding Monbiot's work on rewilding demonstrated that adding or restoring a single keystone species or habitat feature can trigger cascading ecological recovery far exceeding the direct cost. His farm-relevant finding: allowing scrub and natural succession in marginal, unproductive areas typically produces more ecological function than managed wildflower mixes — and at no cost. Not every square metre of biodiversity habitat needs active management; some needs only to be left alone.

Elaine Ingham — Soil Biodiversity as Foundation Ingham's soil food web research established that soil biodiversity is the invisible majority of farm biodiversity. Earthworm populations directly support bird communities; diverse bacterial and fungal communities support diverse soil invertebrates; healthy soils support nutritionally rich plants which support pollinators. Above-ground and below-ground biodiversity are not separate issues — they are the same system. Any biodiversity programme that ignores soil biology is working on half the problem.

Kathy Willis — Evidence-Based Conservation Willis (Oxford biodiversity research) documented that the most effective farm conservation outcomes combine structural diversity with reduced pesticide input — neither alone is sufficient. Her finding: wildflower margins adjacent to pesticide-intensive crops show dramatically lower pollinator use than the same margins adjacent to reduced-input systems. The habitat and the farming practice must change together for meaningful results.

Which tool fits

Routing Decision

• Don't know where to start — want an overview of what's here and what's possible → habitat-audit
• Have a hedgerow that needs attention or want to plant a new one → hedgerow-design
• Want to establish a wildflower area or margin → wildflower-strips
• Considering planting trees, establishing a woodland, or managing existing trees for habitat → woodland-and-trees
• Concerned habitat patches are isolated and want to connect them → wildlife-corridors
• Want to know if what you're doing is working → monitoring
• Unclear where to begin → habitat-audit first; it identifies priorities across all other sub-tools

Habitat Audit

Assesses what biodiversity is currently present on the farm, what is missing, what is at risk, and where investment will have the most impact.

A habitat audit does not require specialist knowledge or equipment. It is a structured walk of the farm, looking for the presence, condition, and connectivity of habitat types. It produces a priority map — where to act first, where to protect what already exists, and where effort would be wasted.

Step 1: Map habitat types present.

Walk every field boundary, watercourse, woodland block, and unmanaged area. Record:

Step 2: Assess condition of each habitat type.

For hedgerows:

• Dense base growth and varied structure = Good
• Gappy, grazed out at base, single species = Poor
• Managed but structurally sound = Fair

For ponds:

• Open water, native marginal vegetation, no invasive species = Good
• Heavily shaded, poached margins, dominated by invasive species = Poor

For grassland:

• 10+ plant species visible in a 1m square = Good
• Dominated by 1–3 grass species, little forb cover = Poor

Step 3: Identify connectivity.

Draw or annotate a farm map. Are the habitat patches connected? Is there a continuous corridor from one side of the farm to another? Isolated patches lose species over time regardless of condition.

Step 4: Identify the most damaging current practices.

Common biodiversity threats on working farms, in rough priority order of impact:

1. Pesticide use in or adjacent to habitat (especially insecticides and herbicides)
2. Hedgerow cutting in the nesting and foraging season (March–August)
3. Bare-soil margins (no buffer between cropped field and hedgerow)
4. Drainage and loss of wet areas
5. Over-management of marginal areas (removing the scrub and rough vegetation that provides the most habitat)

Step 5: Prioritise actions.

Checkpoint — confirm before finalising:

• Is this a conventional farm with regular pesticide use adjacent to existing habitat? The audit output changes significantly — habitat creation without input reduction will underperform.
• Is there an existing agri-environment or conservation scheme applying to any part of the farm? This affects what options are available and funded.
• What is the time and budget available for follow-up action? The audit should prioritise what is achievable, not an ideal picture.

Skipping these confirmations risks producing a priority list the farmer cannot act on, or missing existing constraints that change the recommendations.

Output:

HABITAT AUDIT SUMMARY — [Farm name / date]

HABITAT TYPES PRESENT
[Habitat]: [Length/Area] — [Condition: Good/Fair/Poor]
[Habitat]: [Length/Area] — [Condition: Good/Fair/Poor]
[...]

CONNECTIVITY
[Description of whether habitat patches are connected or isolated]

MOST DAMAGING CURRENT PRACTICES
1. [Practice] — [Impact]
2. [Practice] — [Impact]

PRIORITY ACTIONS
1. [Action] — [Rationale] — [Cost/effort: Low/Medium/High]
2. [Action] — [Rationale] — [Cost/effort: Low/Medium/High]
3. [Action] — [Rationale] — [Cost/effort: Low/Medium/High]

QUICK WINS (no cost, this season)
- [Action]
- [Action]

LONGER-TERM INVESTMENT
- [Action]
- [Action]

Next steps:

• Run hedgerow-design (within this skill) if audit reveals hedgerows in poor condition or gaps.
• Run wildflower-strips (within this skill) if audit reveals absence of flower habitat.
• /s4ag-pests — if beneficial insect habitat is a priority, run pests alongside this to connect habitat investment to biological pest control outcomes.

Hedgerow Design

Species selection, establishment method, and management regime for new or improved hedgerows.

Hedgerows are the most ecologically productive linear habitat on a farm. A well-managed mixed-species hedgerow provides nesting cover, foraging habitat, connectivity corridor, windbreak, flood buffer, and a source of beneficial insects — simultaneously. A poorly managed one does almost none of these things.

New hedgerow vs. restoration of existing:

Species selection — native, mixed hedgerow:

Recommended mix for a new mixed hedgerow: 60% hawthorn + 20% blackthorn + 20% from remaining list (minimum 3–4 species). The more species, the higher the ecological value — aim for at least 5–6 species in any new hedgerow.

Planting specification:

• Bare-root whips (45–60cm), planted November–March
• Double stagger row, 45cm between plants in each row, 30cm between rows
• Tree guards essential where rabbit or deer pressure exists
• Weed suppression mat (60cm wide, pegged down) reduces establishment failure dramatically

Management — the most important decisions:

1. When to cut: Never March–August (nesting season). Cut September–February only.
2. How often to cut: Maximum once every 2 years on any one side. Every 3 years is better. An uncut year allows berries and seed to set.
3. How to cut: Flail on alternate sides in alternate years where possible. Allow the top to grow — a flat-topped, close-cut hedgerow has low ecological value.
4. Hedgerow trees: Leave or encourage trees at 20–30m intervals. Standard oaks, ashes, or field maples in hedgerows are among the most valuable single features on any farm.

Checkpoint — confirm before finalising:

• What is the primary purpose — wildlife, windbreak, stock containment, or all three? This influences species selection and structure.
• Is there rabbit or deer pressure that requires tree guards?
• Is there an agri-environment scheme payment available for hedgerow planting or restoration? This changes the economics significantly.

Skipping confirmation on livestock or deer pressure leads to establishment failure — unguarded whips disappear in one browsing season.

Output:

HEDGEROW PLAN — [Location / date]

TYPE: [New planting / Gap-fill / Restoration]
LENGTH: [metres]

SPECIES MIX
[Species]: [% of mix] — [planted at X m intervals]
[Species]: [% of mix]
[...]

PLANTING SPECIFICATION
Row configuration: [single/double stagger]
Spacing: [cm between plants] x [cm between rows]
Plant size: [bare-root whip / transplant]
Guard type: [spiral guard / spiral + cane / full guard]
Weed mat: [yes/no and specification]

PLANTING WINDOW: [November–March]
ESTIMATED COST: [£/$ per metre including guards and mats]

MANAGEMENT REGIME
Cutting window: September–February only
Cutting frequency: [every 2 / every 3 years]
Cut this side: [Year 1 / Year 2 alternation]
Allow to grow to: [target height in metres]

Next steps:

• Run wildlife-corridors (within this skill) to ensure this hedgerow connects to adjacent habitat.
• Run monitoring (within this skill) to establish a baseline for tracking species recovery.
• /s4ag-agroforestry — if incorporating trees at scale, combine hedgerow standards with a wider agroforestry plan.

Wildflower Strips

Design, seed mix selection, establishment, and management of wildflower strips and margins.

A wildflower strip is the most accessible biodiversity investment on most farms — low capital cost, fast establishment, visible result in year one. The primary ecological function is providing continuous flowering from March to October for pollinators and beneficial insects. Secondary functions include providing nesting and overwintering cover for invertebrates, supporting natural pest enemies, and reducing run-off from field edges.

Where to locate:

Seed mix selection — match to goal:

The flowering calendar gap problem (Goulson): Before selecting a mix, map the current flowering calendar on your farm. Where are the gaps? Most farms have reasonable cover May–July and almost nothing March–April and August–October. Design the mix to fill the gaps, not to add to the peak.

Establishment — two routes:

Route A — Annual mix (faster result, lower commitment):

1. Cultivate the area shallowly (5cm max) in spring or autumn
2. Remove or suppress existing competitive vegetation
3. Broadcast at 3–5g/m²
4. Roll or press seed into surface
5. Do not fertilise — high fertility favours grass over wildflowers
6. Allow to set seed before cutting in autumn

Route B — Perennial meadow (slower, higher long-term value):

1. Yellow rattle (Rhinanthus minor) is the key — it parasitises competitive grasses, opening space for wildflowers. Must be sown on scarified soil in autumn.
2. Sow low-fertility grass mix (fine fescues, crested dog's tail) + native forbs
3. Cut and remove cuttings twice in year one (never leave cuttings — they build fertility and suppress wildflowers)
4. Do not expect full establishment until year 2–3

Management:

• Cut in late summer (August–September) after seed set
• Remove cuttings — always. Leaving them feeds the grass and smothers the wildflowers.
• Do not fertilise. Wildflower strips become grass strips within 2 years on high-fertility ground.
• Leave some areas uncut over winter for overwintering invertebrates

Soil health note (Ingham lens): Wildflower strips on biologically active soils establish more easily and persist better than on degraded soils. The diverse root systems of a wildflower mix feed the soil food web with diverse exudate chemistry — different species, different microbial communities. A wildflower strip is also a food web investment.

Checkpoint — confirm before finalising:

• What is the soil fertility status of the proposed area? High-fertility ground requires a different approach (yellow rattle, cutting regime) to succeed.
• Annual or perennial? The establishment approach is fundamentally different.
• Is the area adjacent to pesticide-applied crops? Proximity to insecticide use will compromise the strip's value regardless of species mix.

Getting the fertility and establishment route wrong is the most common reason wildflower strips fail and revert to rank grass.

Output:

WILDFLOWER STRIP PLAN — [Location / date]

LOCATION: [description / field reference]
AREA: [m²] WIDTH: [m] LENGTH: [m]
SOIL FERTILITY: [High / Medium / Low]

TYPE: [Annual / Perennial meadow]

SEED MIX
[Species / mix name]: [g/m² seeding rate]
Total seeding rate: [g/m²]

ESTABLISHMENT ROUTE
Preparation: [cultivation / scarification / minimum till]
Sowing window: [spring / autumn]
Yellow rattle: [yes/no — if perennial]
Press after sowing: [yes/no]
Do not fertilise: CONFIRMED

FLOWERING CALENDAR
March–April: [what will flower]
May–July: [what will flower]
August–October: [what will flower]

MANAGEMENT REGIME
Cut: [month] — remove cuttings within [X] days
Winter management: [leave uncut sections / cut all]

Next steps:

• Run monitoring (within this skill) to track pollinator use and establish whether the strip is delivering.
• /s4ag-pests — connect wildflower strip investment to biological pest control — the insect habitat above the strip controls pests in the adjacent crop.
• Run wildlife-corridors (within this skill) to ensure the strip is part of a connected habitat network.

Woodland and Trees

Planning the integration of trees and woodland into the farm for habitat, shelter, and ecological function.

Trees are the highest-value single habitat feature on any farm. A single mature oak supports over 2,000 invertebrate species. Dead wood, veteran trees, and tree cavities are among the rarest and most valuable habitats in agricultural landscapes — and they are irreplaceable on a human timescale. This sub-tool covers habitat trees, woodland management, and integrating trees into the farm system for ecological function. For trees as a production enterprise, see /s4ag-agroforestry.

Feature types by ecological value:

Woodland management for habitat:

The single most common mistake in farm woodland management is over-management. Tidying, clearing, and "improvement" almost always reduces habitat value. The most habitat-rich woodlands are the least intensively managed ones.

Principles:

1. Leave dead wood. Standing dead trees (snags) and fallen trunks are not mess — they are the most ecologically valuable structures in a woodland.
2. Maintain a structured edge. A woodland edge that grades from tall trees to shrubs to rough grassland contains more species than a hard boundary.
3. Coppice selectively, not uniformly. Rotational coppice creates a mosaic of age classes. Never coppice the whole wood at once.
4. Protect ground flora. Bluebells, wood anemones, and ancient woodland indicators are irreplaceable. Reduce deer browsing where they are absent.
5. Open rides and glades. Sunny glades within woodland are among the most productive butterfly and invertebrate habitats. Maintain or create by selective cutting.

New tree planting for habitat:

If planting trees primarily for habitat rather than production:

• Prioritise species with highest wildlife value: oak, rowan, crab apple, hawthorn, willow, alder (wet ground)
• Plant in groups (3–5+ trees) rather than single specimens where possible — micro-habitat value increases with group size
• Allow natural regeneration wherever it is occurring — it is always cheaper and often ecologically superior to planted trees
• Consider not-planting: in some situations, allowing natural regeneration (rewilding) delivers more habitat faster at no cost

Deer impact: In most farmed landscapes, deer browsing is the primary limit on woodland regeneration and ground flora recovery. If deer are present at density, tree guards or deer fencing are not optional — they are the prerequisite for any woodland investment to succeed.

Soil health note (Ingham lens): Woodland soil is typically the most fungal-dominated soil on a farm — the ancient mycorrhizal networks under established woodland are irreplaceable biological infrastructure. When woodland is cleared for agriculture, this fungal capital is lost and takes decades to rebuild. Protecting existing woodland and allowing expansion at the edges is the most effective strategy for maintaining fungal diversity across the farm.

Checkpoint — confirm before finalising:

• Are there veteran trees or ancient woodland features present? These take absolute priority over any planting or management action.
• What is the deer pressure? This determines whether new planting is viable without protection.
• Is this for habitat only, or also for timber, fuel, or fodder production? Mixed objectives require a combined approach with /s4ag-agroforestry.

Proceeding without understanding deer pressure leads to planting failures that cost money and delay habitat development by years.

Output:

WOODLAND AND TREES PLAN — [Farm name / date]

EXISTING FEATURES — PROTECTION PRIORITY
[Veteran trees]: [location / action: PROTECT]
[Dead wood]: [location / action: LEAVE]
[Woodland edge]: [location / condition / action]

MANAGEMENT ACTIONS
[Feature]: [Action] [Timing]
[Feature]: [Action] [Timing]

NEW PLANTING
Species: [list] Ratio: [%]
Location: [description]
Configuration: [groups / rows / scatter]
Guard requirement: [type / height]
Natural regeneration opportunity: [yes/no / location]

DEER MANAGEMENT
Current deer pressure: [Low / Medium / High]
Protection required: [guards / fencing / none]

HABITAT FEATURES TO CREATE
Log piles: [yes/no / location]
Glades: [yes/no / location and size]

Next steps:

• Run wildlife-corridors (within this skill) — woodland blocks are anchor habitats; corridors connect them.
• /s4ag-agroforestry — if integrating trees with productive enterprises, develop a combined plan.
• /s4ag-carbon — woodland and trees are the farm's most visible carbon store; connect to carbon planning.

Wildlife Corridors

Identifying, designing, and establishing connectivity between habitat patches across the farm and landscape.

An isolated habitat patch — however well managed — loses species over time. The island biogeography principle (Wilson) predicts this: small isolated habitats cannot sustain viable populations of most species. The solution is connectivity: linking patches with corridors that allow movement, recolonisation, and genetic exchange. Connectivity is the highest-leverage biodiversity investment on most farms because it multiplies the value of all existing habitat simultaneously.

Identifying the current connectivity situation:

Step 1: Map all habitat patches on the farm (from the habitat audit). Step 2: Draw straight lines between patches — what is between them? Step 3: Identify the shortest, most feasible corridor route between each pair of disconnected patches.

Corridor types by situation:

Minimum viable corridor widths:

• Butterflies and most invertebrates: 3m minimum, 6m+ preferred
• Small mammals (hedgehogs, voles, shrews): 2m minimum with dense ground cover
• Badgers and foxes: any — they will use whatever is available
• Bats: linear features with 3m+ height on at least one side (hedge, treeline, building)
• Birds (farmland species): 6m+ with mixed structure

The hedgehog highway principle: One of the most effective and lowest-cost corridor interventions is the "hedgehog highway" — a 13cm x 13cm hole at the base of garden and farm boundaries allowing hedgehog movement. Applied at farm scale: check that all stock fencing allows passage for small mammals at ground level. Solid stock fencing without gaps is a significant barrier to invertebrate and small mammal movement.

Landscape-scale thinking: Individual farm corridors are most effective when they connect to features on neighbouring farms. If there is a regional wildlife corridor initiative, agri-environment scheme, or landscape partnership operating locally, connecting the farm's corridor plan to this multiplies the value of the investment.

Establishing corridors:

Where a corridor runs through productive land, the opportunity cost matters. Options ranked by cost:

1. Adjust existing management — defer cutting, allow natural regeneration along boundaries (cost: near zero)
2. Establish grass margin — sow along field edge, reduce spray width (cost: seed only)
3. Plant hedgerow extension — bare-root whips November–March (cost: low-medium)
4. Plant mixed scrub — trees and shrubs at wider spacing than formal hedgerow (cost: low-medium)
5. Establish wildflower corridor — higher-value but needs management (cost: medium)

Checkpoint — confirm before finalising:

• Is there an agri-environment scheme in place that covers corridors or connectivity? Many schemes pay specifically for this — the recommendation changes if funding is available.
• What species are the target for connectivity? The corridor specification varies significantly for bats versus butterflies versus amphibians.
• Are there landscape-scale corridor plans in the area (wildlife trusts, river catchment schemes, landscape partnerships)? Connecting to these multiplies the value of farm-level investment.

Designing a corridor without knowing the target species produces infrastructure that may not function — a mown grass strip is a corridor for field voles but a barrier for most butterflies.

Output:

WILDLIFE CORRIDOR PLAN — [Farm name / date]

CONNECTIVITY GAP IDENTIFIED
From: [Habitat patch A]
To: [Habitat patch B]
Current gap: [metres / description of what lies between]

TARGET SPECIES / GROUPS: [list]

RECOMMENDED CORRIDOR
Type: [hedge extension / grass margin / scrub planting / wildflower strip]
Width: [metres]
Length: [metres]
Alignment: [description / compass direction]

ESTABLISHMENT METHOD
[Action] [Timing] [Cost estimate]

MANAGEMENT
[Action] [Frequency]

LANDSCAPE CONNECTIONS
Connect to: [adjacent farm feature / scheme / landscape corridor]

Next steps:

• Run monitoring (within this skill) to establish whether the corridor is being used.
• Run hedgerow-design or wildflower-strips (within this skill) depending on the corridor type selected.
• /s4ag-land-reading — if corridors are constrained by topography or drainage, read the landscape first.

Monitoring

Simple monitoring protocols to track biodiversity change over time — without specialist skills or equipment.

Monitoring answers the question: is what you are doing working? It does not require specialist knowledge. The most valuable monitoring is consistent and repeatable, not technically perfect. A simple count done in the same way in the same place at the same time of year every year is worth far more than an expert survey done once.

Tier 1 — Minimum viable monitoring (any farm, no cost):

These indicators take 2–3 hours per year and require no specialist knowledge:

Tier 2 — Intermediate monitoring (with modest effort):

• BTO Breeding Bird Survey (BBS) — register a farm square; professional-grade data; takes 2 visits per year
• Bumblebee Conservation Trust BWARS recording — submit bumblebee observations to national database
• iRecord app — photograph and submit species records; builds a farm species list over time
• National Moth Recording Scheme — moth trap records; excellent indicator of invertebrate health

Tier 3 — Professional assessment (every 5 years or at project start/end):

Commission a baseline botanical survey, invertebrate survey, or bat survey at the start of a major habitat intervention. Repeat after 5 years. This provides before/after comparison that validates the investment.

Recording what matters:

Keep a simple farm wildlife diary. Record:

• Species observed, date, location, activity (feeding, nesting, passing through)
• Changes to habitat management and when they happened
• Any unusual events (disease, flood, pest outbreak)

This record becomes invaluable over 5–10 years — it reveals long-term trends that no single survey can see.

Interpreting trends:

Checkpoint — confirm before finalising:

• What baseline data exists? If none, the first priority is establishing a baseline before any improvement programme starts — otherwise there is nothing to compare against.
• How much time can realistically be committed to monitoring per year? Design the protocol around what will actually be done, not an ideal.
• Is there an agri-environment scheme or grant scheme requiring monitoring data? This shapes the format and rigour required.

Without a baseline, it is impossible to demonstrate improvement — and impossible to claim credit for change.

Output:

MONITORING PLAN — [Farm name / date]

TIER: [1 / 2 / 3]

BASELINE ESTABLISHMENT
[Indicator]: [method] — [date of first count] — [result]
[Indicator]: [method] — [date of first count] — [result]

ANNUAL MONITORING PROTOCOL
[Indicator]: [method] [timing] [location] [time required]
[Indicator]: [method] [timing] [location] [time required]

RECORDING SYSTEM
Method: [notebook / spreadsheet / iRecord / BTO]
Responsible person: [name]

REVIEW SCHEDULE
Annual review: [month]
5-year professional survey: [year]

YEAR 1 RESULTS
[Indicator]: [count / observation]
[Indicator]: [count / observation]

Next steps:

• /s4ag-regenerative — if monitoring reveals systemic ecosystem dysfunction, connect biodiversity trends to the broader farm transition framework.
• Run habitat-audit (within this skill) if monitoring reveals declining trends — re-audit to identify what has changed.
• /s4ag-soil — if earthworm monitoring reveals a declining trend, address the food web foundation with a soil health assessment.