Two private companies enforce 41% of England's biodiversity covenants

England's mandatory Biodiversity Net Gain regime turned two in February 2026. The Biodiversity Gain Sites Register, hosted by Natural England, now covers 245 sites across 8,196 hectares, projecting 34,837 net biodiversity units over 30 years (live register data via Bristol Tree Forum's mirror, accessed 16-17 May 2026).

Direct query of the official register reveals a structural fact not previously written about in trade press or operator marketing: two private companies, RSK Biocensus Limited and Harry Ferguson Holdings Limited, hold the Responsible Body role — the legal enforcement role for conservation covenants — on 100 of 245 sites, or approximately 41% of all registered habitat banks. RSK Biocensus alone holds 75 (30.6%). Harry Ferguson Holdings holds 25 (10.2%). Local Planning Authorities, charities, and other smaller bodies hold the remainder.

This document presents that finding alongside counterparty profiles of eight habitat bank operators of materially different scale. It is a desk-based, primary-source analysis of who actually sells, owns, and polices BNG units in England — using only public data from Companies House, the Natural England register, and operator disclosures.

A note on timing: the register has grown from 19 sites in October 2024 to 245 today, roughly doubling every 8-10 months. Industry pipeline disclosures suggest 250-400 additional sites in development, and the May 2026 extension of BNG to Nationally Significant Infrastructure Projects is expected to drive a further wave. The structural facts below are a snapshot. The proportion of sites with private Responsible Bodies will change. The operator landscape will broaden. This is v0.1; updates will be versioned and dated.

This document attempts to make visible what isn't visible at the point of purchase: who is selling the unit, how they are funded, what governs their 30-year obligation, and what risks are documented in the public record. It is a counterparty profile, not a biodiversity-outcome rating. The regime is too young for outcome data to exist.

A finding the register surfaces immediately

Before profiling individual operators, one finding from the public register: two private companies hold the Responsible Body role on approximately 41% of all sites registered with Natural England's Biodiversity Gain Sites Register.

Verified by direct query of the official register (environment.data.gov.uk/biodiversity-net-gain) on 17 May 2026:

• RSK Biocensus Limited (part of the RSK Group): 75 sites as Responsible Body
• Harry Ferguson Holdings Limited (Isle of Wight): 25 sites as Responsible Body
• Combined: 100 of 245 registered sites = 40.8%

Methodology note. Counts verified directly against the official register by name search on 17-18 May 2026 (screenshots available on request). The register's "Local Planning Authority or responsible body" field is unstructured text, with company names appearing in several capitalisations (e.g. "RSK Biocensus Limited", "RSK Biocensus Ltd", "RSK Biocensus"). The 75 and 25 figures cover all visible variants returned by the search, treated as the same legal entity in each case. A single-word search for "RSK" returns the same 75 sites as "RSK Biocensus" — no additional RSK Group entities (including the separately-designated RSK ADAS Limited) currently hold sites on the register, ruling out under-counting through name variation. Total registered sites (245) is taken from Bristol Tree Forum's nightly mirror of the official register and may differ from the live register by a small daily-update margin. The combined 41% figure should be considered accurate to approximately ±2-3 percentage points. It is also a snapshot: the register grew from 19 sites in October 2024 to 245 today, and the proportion of private-Responsible-Body sites may change as more LPA-led sites are registered.

The role of a Responsible Body under the Environment Act is to hold and enforce conservation covenants — analogous to the role of a trustee in a long-duration legal arrangement. The Responsible Body, not the habitat bank operator, is the legal counterparty for enforcement against the landowner over the 30-year obligation.

This concentration is not visible at the point of purchase. Developers buying biodiversity units do not choose who polices the covenant — that choice was made by the operator when the site was registered. Buyers inherit it. They may not know which company holds the enforcement role on the unit they have just purchased until they read the legal agreement carefully.

RSK Biocensus' visible role-holdings include the largest single site on the register (Wendling Beck Environment Project, 537 hectares), Nattergal's Boothby Wildland Phase 2 and Harold's Park Wildland sites, and many third-party-operated habitat banks (Belmont, Iford, Dingley Park, Ardingly, Witchampton among others, all visible on the public register).

Other Responsible Bodies on the register include Local Planning Authorities (the most common category overall, fragmented across dozens of councils), Restoration Soil Limited, Plantlife, The Land Trust, The Land Restoration Trust, The Lifescape Project, and several Wildlife Trusts. Defra has designated other potential private Responsible Bodies — including RSK ADAS Limited and The Environment Bank Limited itself — but as of 17 May 2026 those entities hold zero sites on the register. They are dormant designations, not currently active enforcers. Whether that changes — particularly whether RSK ADAS begins onboarding sites alongside RSK Biocensus, pushing the RSK Group concentration even higher — is worth tracking.

Summary table

Profiles

1. The Environment Bank Limited

The market's largest operator by stated unit volume. Customers include Aldi, National Grid, and Everton FC's stadium transport developer. Operates a leasehold model: 30-year leases with farmers, manages habitat creation in-house. Per Environment Bank's own April 2026 statement to Farmers Weekly, approximately 9,000 units generated with around 1,000 allocated to development sites.

Refinancing horizon is the watch-item. Solvency in 2026 is not in question. The risk is what happens when the Gresham facility matures and the natural-capital credit market has to absorb a refinancing event, potentially during a less benign cycle.

Governance. In-house ecology team. Lease model means land risk sits with the farmer; operational and refinancing risk sit with Environment Bank.

2. Nattergal Limited

Multiple rewilding sites registered with Natural England: Boothby Wildland Phase 1 (64.6 ha) and Phase 2 (492.9 ha, registered November 2025), High Fen Wildland (113 ha), and Harold's Park Wildland (197 ha). Together over 860 ha currently on the register, with Nattergal stating over 4,000 biodiversity units generated. Founder connections to Knepp Castle (UK's flagship rewilding estate) and natural-capital finance circles. Research partnerships with Cambridge and UKCEH.

Responsible Body diversification. The Nattergal sites use three different Responsible Bodies: RSK Biocensus (Boothby Phase 2, Harold's Park), The Lifescape Project (High Fen), and South Kesteven LPA (Boothby Phase 1). This is unusually diversified compared to peers and reduces single-counterparty risk on covenant enforcement.

Counterparty profile. Equity-funded structure is lower default risk than debt (no maturity wall, no covenant trips), but operational viability depends on unit sales meeting projections plus growing carbon/water/voluntary credit revenue. Receives Defra Landscape Recovery Scheme baseline funding (~£275k), giving runway while markets mature.

Watch. Multiple emerging natural-capital revenue lines (BNG, carbon, water, ecotourism, Landscape Recovery). Diversification is a strength; dependency on multiple unproven markets is a weakness. The rewilding model is genuinely different from habitat-creation orthodoxy and ecological outcomes will diverge accordingly.

3. Lopemede Farm

Single 27.24-hectare site in the Thame Valley, registered as BGS-150424001 with Buckinghamshire LPA as Responsible Body. Per the live register, 14 allocations have already been made to development schemes from this site as of May 2026 — high utilisation for a site of this size. High-distinctiveness grassland and floodplain wetland mosaic. Inventory regionally constrained to Buckinghamshire/Oxfordshire and immediately neighbouring LPAs.

What it can't do. Scale. A developer needing thousands of units cannot use Lopemede. Habitat diversity is limited. The model is hard to replicate — every Lopemede equivalent needs a farmer–LPA–charity tripartite assembly that takes years.

4. BioGains Ltd

First operator on the Natural England register (Exeter site, early 2024). Now multi-site. Customers include Barratt Redrow. Co-founders Jon Garner (CIEEM ecologist; MD of GE Consulting) and Andy Rowe (chartered RTPI planner; former PLC housebuilder board director).

Counterparty profile. PLC-backed via Origin gives stronger balance sheet support than typical mid-sized operators. The structural conflict is a disclosure-and-process issue, not a financial-stability issue.

5. Integrated Land Management Ltd (ILM)

15 habitat sites across the Midlands and southern England. Multi-LPA coverage (Wealden, North Northants, East Suffolk, St Albans/Three Rivers among others). Habitat diversity is genuine — traditional orchards, wood-pasture, native hedgerows, lowland meadow, neutral grassland. Run as a family business by the Paske family with director Jonathan Thompson.

Capital and balance sheet. Small private company filing total-exemption accounts (turnover below the audit threshold). No charges visible on Companies House search results. The business model is "farmer-direct": landowners retain ownership and receive index-linked income; ILM manages.

Counterparty considerations. The combination of small balance sheet plus 30-year obligations across 15 sites is the textbook case where buyers should ask about reserve funding, succession planning, and what happens if a key director becomes unavailable. None of these are red flags on their own; together they're the diligence questions a serious buyer asks.

6. Maydencroft Limited

Family-run group with multiple operating brands: landscape services, woodland management (per Maydencroft's website, 10,000+ hectares of woodland sustainably managed), tree-pest specialism (Oak Processionary Moth), recycled-plastic boardwalks, conservation grazing (Longhorn cattle, native sheep), events at Maydencroft Manor. BNG is one product line within a broader estate-services business that has been operating since 2007.

Counterparty profile. A long-established group with non-BNG revenue streams cross-subsidising habitat creation. The diversified parent reduces single-product-line risk. The trade-off: BNG is one priority among many, not the company's defining mission, which can cut either way for buyers — operational resilience yes, ecological focus less obvious.

Watch. Group structure means buyer due diligence should clarify which legal entity is signing the conservation covenant or S106 agreement, and what intra-group dependencies exist.

7. Kent Wildlife Trust Group

One of the regional Wildlife Trusts within The Wildlife Trusts federation. Operates a planning team that engages with Kent County Council and LPAs, and delivers BNG plans. Conservation covenant or S106 secured. Notably, the Trust publicly states they apply due diligence to developers before agreeing to work with them — a buyer-screen rare among commercial operators.

Watch. Capacity. Wildlife Trusts are conservation charities, not commercial habitat factories. Total federation-wide BNG capacity is finite. Pricing tends to be higher than commercial operators on like-for-like habitats because profits are reinvested in conservation rather than driving down unit cost.

8. BNG Habitat Banks Limited

Included here as the textbook "ambiguous name" case. There are several companies trading under variations of "Habitat Bank" or "BNG" naming: The Habitat Bank, UK Habitat Bank, BNG Habitat Banks Ltd, Legacy Habitat Banks, and others. This is one of them. Limited public profile; a small private company in Congleton.

Counterparty profile. Not enough public signal to assess in depth. This is itself a signal: small operators with limited public disclosure require greater buyer-side due diligence, particularly around the conservation covenant terms, reserve funding, and what happens to the obligation if the company is dissolved.

What the comparison reveals

At the point of unit purchase, all eight operators look identical to the developer's planning condition: registered, legally agreed, monitored. The metric output is the same.

The visible differences sit in four dimensions:

Capital structure. Institutional debt (Environment Bank), equity (Nattergal), corporate-PLC investment (BioGains), small private (ILM), established group (Maydencroft), charity (Kent Wildlife Trust), family freehold (Lopemede). Each has a different default mode and different 30-year persistence profile.

Governance. In-house management vs charity-delegated vs vertically-integrated. The Lopemede / Kent Wildlife Trust / Buckinghamshire model with independent charity oversight is structurally stronger than self-monitored commercial models — at the cost of scale.

Related-party structure. Most acute at BioGains, where the parent group is also invested in ecology consultancies that advise on unit purchase. Buyers using a consultancy from the Origin Environmental portfolio should request related-party disclosure.

Scale vs depth trade-off. Environment Bank can supply 9,000 units; Lopemede cannot. Lopemede has multi-generational freehold security and charity oversight; Environment Bank has refinancing risk. These are not better-or-worse comparisons — they are different products for different buyers.

What failure looks like, and what's detectable

The regime depends on 30-year obligations. Some failure is inevitable across a portfolio of 245 sites over three decades. The honest question is not whether failures happen, but which are detectable in time to act on, and which only become visible after the damage is done.

Six failure modes are visible in industry discussion and the published regulatory framework:

1. Slow ecological underperformance. Habitat is created but predicted targets are not met. Wildflower meadow ends up with three indicator species rather than twelve. The covenant is technically intact; the ecological output is not. Typically becomes visible at condition-assessment intervals — five years and beyond. Not detectable in advance from public data. The metric measures predicted unit value, not delivered unit value.

2. Management failure or abandonment. Operator stops doing required management: grazing, mowing, scrub control. Habitat reverts. Detectable on satellite imagery within two to three years through vegetation-structure change. This is one of the failure modes a desk-based monitoring system can catch.

3. Operator insolvency. Conservation covenant survives — it runs with the land — but operational capacity vanishes. Enforcement falls to the Responsible Body. UK insolvency rates for small private companies run roughly 3-5% per year. Across 245 sites over 30 years, this is not an unusual event. Detectable in real time from Companies House.

4. Landowner default or land sale. Land changes hands. New owner inherits the covenant but enforcement requires legal action by the Responsible Body. Typically surfaces around inheritance, sale, or insolvency events. Detectable via Land Registry monitoring if instrumented.

5. Catastrophic event. Fire, flood, disease outbreak. Wipes out habitat. Whether remediation funding is ring-fenced depends on the specific Habitat Management and Monitoring Plan and on operator resilience. Detectable immediately; recovery trajectory takes years.

6. Regulatory drift. Metric updates, exemption changes, NSIP rules, small-site policy shifts. Doesn't destroy physical habitat but re-rates economic value. Detectable from Defra publication channels.

Three of these — operator insolvency, management abandonment, regulatory drift — are detectable from public data feeds (Companies House, satellite imagery, Defra publications) with lead times measured in months rather than years. That is where structured monitoring of the kind sketched in this document becomes operationally useful. The other three failure modes need physical inspection, monitoring reports filed to LPAs, or post-hoc condition assessments — sources not currently aggregated and published, and beyond the scope of public-data analysis today.

A reasonable working assumption from private industry conversations is that ten to twenty percent of currently-registered units will not deliver as predicted over their 30-year lifetimes. That number is not authoritative — no published figure exists — but it reflects how people inside the regime privately talk about delivery risk. The product question is not which sites fail. It is which operators have the structural resilience to absorb failure across their portfolios, and which fail catastrophically.

A note on unit quality

This document does not include pricing per operator, and does not assess ecological unit quality. Both omissions are deliberate.

Pricing is opaque by market design. Operators submit transaction prices to brokers and industry aggregators (notably Biodiversity Units UK's quarterly Pricing Report) only on condition of anonymity. Aggregate pricing data exists; operator-level pricing data does not exist in any public source. Including specific prices would mean repeating trade-press numbers without provenance.

Unit quality is formally defined by the Statutory Biodiversity Metric (distinctiveness, condition, strategic significance, connectivity, time-to-target). Two operators selling units with identical metric outputs may deliver very different ecological outcomes. Verifying that distinction requires species-level monitoring data, soil chemistry, hydrology, and on-the-ground ecological assessment — none of which is in the public register, and none of which is in scope for this document. Buyers requiring substantive quality diligence currently rely on chartered ecologists carrying out site visits. That is the existing standard and this document does not substitute for it.

Methodology and sources

Data sources. All free, all public. Companies House (find-and-update.company-information.service.gov.uk) for incorporation, accounts status, charges, persons of significant control. Natural England Biodiversity Gain Sites Register (environment.data.gov.uk/biodiversity-net-gain), accessed via Bristol Tree Forum's live mirror (bgs.bristoltrees.space) for aggregate counts and structured site-level data. Operator websites. Local Government Association BNG library (local.gov.uk/pas). Trade press (Farmers Weekly, ENDS Report). Charity Commission for charity operators.

Deliberate exclusions. No satellite imagery in v0 (planned for v1). No subjective ecological quality judgments. No letter grades (premature given absence of outcome data). No price commentary (out of scope; tracked separately by Biodiversity Units UK).

Data tiers. Every claim in this document sits in one of three tiers:

Tier A — primary-source verified. All Companies House data (incorporations, charges, statuses); all Natural England register figures (RSK Biocensus 75 sites, Harry Ferguson Holdings 25 sites, total 245 sites, totals for hectares and units). For Responsible Body counts, screenshots of the query results from environment.data.gov.uk dated 17-18 May 2026 are available on request.

Tier B — operator-stated, attributed in text. Unit-generated counts ("~9,000 per Environment Bank"; "4,000+ per Nattergal"), hectares managed by parent entities ("per Maydencroft"), and similar self-reported scale figures. These are reported as the operator states them and labelled as such.

Tier C — public derived data via third-party mirror. Aggregate register statistics (245 sites, 8,196 ha, 34,837 net units) are taken from Bristol Tree Forum's nightly mirror of the official register because the official Natural England interface supports per-query search only, not aggregate views. The underlying data is from environment.data.gov.uk; Bristol Tree Forum republishes and aggregates it. Aggregate figures should be considered approximate ± a small daily-update margin.

Tier D — analytical inference, explicitly labelled. Phrases such as "refinancing risk," "structural strength," "consistent with a refinancing," "capacity-limited," and any 30-year persistence judgment are analytical interpretation of the underlying facts, not facts themselves. They are not statements about any operator's actual financial condition or future behaviour.

Conflicts of interest. The author has no financial relationship with any operator covered, is not paid to publish, and has not given any operator pre-publication review.

Right of reply. Any operator covered may respond. Corrections to fact will be incorporated. Disagreements with interpretation will be appended verbatim, attributed to the operator.

When this analysis is useful — and when it isn't

The conservation covenant under BNG runs with the land, not the operator. If an operator fails, the legal obligation to manage the habitat survives — passing to the landowner, with enforcement falling to the Responsible Body. In that strict legal sense, no operator failure ever leaves a buyer's planning condition unmet.

What does not survive operator failure is the delivery. The management work stops. The funding stream stops. Enforcement is slow and expensive. The land sits in legal limbo while habitat degrades. The unit purchaser has been paying for "30 years of actively managed habitat creation" and ends up with "30 years of land that is legally supposed to be managed but isn't." Those are very different products even though the contract appears intact.

This distinction shapes when counterparty analysis is most useful:

Most useful — before commitment. At the point of unit purchase, financing, or insurance underwrite, the buyer can still choose between operators or price the risk. A developer comparing two operators for the same site can pick the more resilient one. A lender can decline to finance or adjust covenants. An insurer can adjust premiums. The information is decision-actionable.

Less useful — after commitment. Once units are allocated to a development, the planning condition is discharged. The buyer has no contractual route to switch operators mid-covenant. Counterparty information becomes an input to ESG disclosure and risk reporting, but cannot drive a remedy.

The implication: operator counterparty diligence is fundamentally a transaction-time tool. Most valuable to whoever is making the decision before the commitment closes. That includes developers, their planning consultants, lenders, insurers, and any fund taking operator-level exposure.

What this is for

For a developer or their planning consultant: a structured way to look beyond the sales page before allocating units to a development.

For a lender or insurer underwriting nature finance: a starting point for counterparty diligence in an asset class without established credit comps.

For a fund evaluating operator equity exposure: a public-record sanity check before deeper diligence.

For a planning ecologist at an LPA: context on the operator behind a Biodiversity Gain Plan application.

For a journalist or researcher: a base layer of operator-level fact in a market where most analysis is aggregate.