> For the complete documentation index, see [llms.txt](https://sexytrees.savimbo.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://sexytrees.savimbo.com/biodiversity-credits/calculation.md).

# Biodiversity calculation

<mark style="color:$danger;">This is a</mark> [<mark style="color:$danger;">bricolage</mark>](/foundations/bricolage.md) <mark style="color:$danger;">protocol; which means we are building it in public! Some pages are under construction — but check back soon its updating fast!</mark>&#x20;

Generating Interoperable Biodiversity Units (IBUs) in an uplift scenario requires an increment-sensitive, absolute measure of ecosystem integrity that can be tracked over time, affordably, and without expert personnel on-site. No currently published protocol meets this requirement to our satisfaction.&#x20;

Accurate assessments of absolute integrity require increment-sensitive scales anchored in well-defined reference conditions. Our previous protocol has been successful in certifying, crediting, and [OTC-trading](https://carbon-pulse.com/432328) with [offtake agreements](https://carbon-pulse.com/450718/). But because the ISBM relies on indicator-species presence or absence (an inherently binary scale at the higher end of integrity), it cannot achieve this level of granularity but is limited to conservation scenarios.&#x20;

To our knowledge, no existing certifiable protocol currently offers absolute uplift measurements. So we set out to develop one, starting with the question of which available tools can reliably monitor which scale-based metrics.&#x20;

### There are a lot of metrics

In nature credits, raw data compiles into metrics. Metrics are analyzed by methods (incorporating statistical controls) that are then abstracted into units for interoperability and trading.&#x20;

We think its eaiser to start with units and work backwards, narrowing the field as you go, but you can work either direction. The point is you don't need all the metrics available; you need the right metrics for your argument, and those metrics need to produce an interoperable unit.

Carbon has >170 methods for its measurement. Biodiversity will likely have a lot more because the life in an ecosystem is much more complex than its inorganic components.&#x20;

Here is a nice summary we like from [BioInt](https://www.bioint.fr/biodiversitymeasurementapproaches), our go-to deep-nerds on Western biodiversity science ([BioInt](https://www.bioint.fr/biodiversitymeasurementapproaches), 2025).&#x20;

<figure><img src="/files/7LgvU5HnPuZuXjHWrlYU" alt=""><figcaption><p>Overview of biodiversity measurement approaches from <a href="https://www.bioint.fr/biodiversitymeasurementapproaches">BioInt,</a> V4 2025. Used with permission. Citations for the graphic on BioInt site. </p></figcaption></figure>

Biodiversity is measured at two scales. Alpha diversity captures what's at a single site — species richness (which species) and evenness (how individuals are distributed across them). Beta diversity captures how composition changes between sites or over time. Both presence/absence and abundance feed into alpha diversity, but they're not interchangeable in the field. [ISBM](https://isbm.savimbo.com/executive-summary) (the conservation protocol) uses presence/absence only.  In that case abundance counts require more training, more time, and more cost per sample, which would have priced communities out of running the protocol themselves. The trade-off is real: a presence/absence protocol is cheaper to run and more inclusive, but coarser than a protocol that counts individuals. Credits issued under it should be priced and described accordingly, and we expect this distinction to eventually show up in market prices.

For this protocol, an uplift protocol, we had to pick a method that would capture more granular changes, predominantly in tropical forests. That narrows the scope considerably, because we pick the metric based on the argument we are trying to make.

We knew the indices we were looking for. To show biodiversity recovery in degraded tropical forests begin with increases in bird activity and diversity  (within 4 years: [Roels et al. 2019](https://doi.org/10.5751/ACE-01330-140109)) and arthropod diversity (first 5-8 years: ([Cole et al. 2016](https://doi.org/10.1002/ece3.2220); [Pinto et al. 2025](https://doi.org/10.1016/j.foreco.2025.122749)). Further, incorporating multiple taxonomic groups improves the ability of biodiversity metrics to distinguish primary forests from agroforestry systems ([Kessler and Bromet 2013](https://doi.org/10.1890/08-1074.1)), and ideally, we have both abundance and species richness of habitat specialists serve as informative indicators ([Scales and Marsden 2008](https://doi.org/10.1017/S0376892908004840)).&#x20;

### The current state of the science

We reviewed each available tool class systematically before arriving at our proposed approach. On extensive investigation, it's clear that claims to measure biodiversity restoration are scientifically problematic.&#x20;

We do not consider this a deterrent to the attempt, but begin with awareness of the true limitations in the field.  &#x20;

**Indicator species observation (ISBM).** The existing certified biodiversity crediting protocol — Savimbo's [Indicator Species Biodiversity Methodology (ISBM)](https://isbm.savimbo.com/) — uses the presence or absence of indicator species as its metric. This binary scale cannot detect gradual restoration trajectories. It is therefore restricted to conservation scenarios where ecosystems are already at or near full integrity, and is not applicable to agroforestry uplift.

**Arthropod trapping**. While arthropod trapping is a reliable academic methodology, to base market-wide scale and grassroots adoption it's impractical. As a tool its invasive, requires significant taxonomic expertise for identification, and involves substantial permitting in most jurisdictions — making it unsuitable for community-led, low-cost monitoring at scale.

**Acoustic monitoring.** Acoustic approaches avoid the invasiveness of trapping and can capture early bird diversity signals ([Roels et al. 2019](https://doi.org/10.5751/ACE-01330-140109)). However, insect species cannot yet be identified acoustically ([van Klink et al. 2024](https://doi.org/10.1098/rstb.2023.0101)), acoustic biodiversity indices perform poorly and inconsistently across ecosystems ([Alcocer et al. 2022](https://doi.org/10.1111/brv.12890)), and automated bird detection still fails in tropical biomes ([Funosas et al. 2025](https://doi.org/10.21203/rs.3.rs-7832874/v1)). Acoustic data thus cannot provide the multi-taxon signal needed to anchor an absolute integrity curve.

**Visual point counts**. Visual surveys rely on trained expert observers and produce limited verifiable documentary evidence for third-party audits. They are also highly sensitive to observer effort and skill, undermining comparability across sites and time.

**Remote sensing (NDVI, LiDAR, satellite)**. Vegetation indices and canopy structure metrics derived from drone or satellite imagery are valuable for carbon assessments (see [Carbon](/carbon-credits.md)) but do not capture faunal community composition or below-canopy biodiversity. A dense canopy can coexist with severely degraded faunal assemblages. Remote sensing is therefore orthogonal to biodiversity measurement as defined by the IBU — it measures a different dimension and cannot substitute for it.

The conclusion from this review is that credible, scalable, multi-taxon, increment-sensitive biodiversity monitoring in smallholder agroforestry systems requires a tool that is non-invasive, deployable without taxonomic expertise, capable of detecting multiple species groups simultaneously, and feasible under the logistical and economic constraints of remote tropical settings. The team has concluded that, at present, only environmental DNA (eDNA) metabarcoding ([Lacoursière-Roussel 2019](https://doi.org/10.1111/jfb.14177)) satisfies these criteria.

### eDNA as a measurement approach

Environmental DNA refers to genetic material shed by organisms into their surrounding environment — water, soil, or air — and collected non-invasively for analysis ([Power et al. 2023](https://doi.org/10.1002/edn3.497)). eDNA detection has been validated as suitable for Indigenous- and smallholder-led monitoring due to its non-invasiveness, adaptability to remote field settings, and ability to detect multiple species simultaneously ([Bélisle et al. 2026](https://doi.org/10.1111/1365-2664.70253)). eDNA surveys targeting terrestrial organisms are recognized as promising tools for agricultural systems, though their implementation in these contexts remains at an early stage ([Kestel et al. 2022](https://doi.org/10.1016/j.scitotenv.2022.157556)).

Serious limitations of this approach include equity, scientific practicality, and ethics.&#x20;

* **Equity risks** include affordability, lack of skills to interpret, store, or access data, or laboratory services. Communities are inherently reliant on ethical partners for this technology, and are therefore at a disadvantage in exercising their sovereignty.
* **Scientific challenges** are significant. Spatial and temporal resolution is scale-dependent and variable across ecosystems, as DNA degradation rates and dispersion patterns differ by geography and season ([Kestel et al. 2022](https://doi.org/10.1016/j.scitotenv.2022.157556)). Taxonomic identification is generally less accurate in tropical regions due to gaps in local biodiversity reference databases ([Chimeno et al. 2023](https://doi.org/10.1371/journal.pone.0290173)). Random error in eDNA samples can affect data integrity and must be acknowledged ([Lahoz-Monfort et al. 2016](https://doi.org/10.1111/1755-0998.12486)). These constraints are addressed in part through the Water Bucket Protocol design (see [WBP](/biodiversity-credits/water-bucket-protocol.md)) and through our ongoing work on reference level establishment (see [Ongoing work](#ongoing-development)).
* **Ethical risk** requires explicit acknowledgment: eDNA sampling has the potential to [inadvertently collect human DNA](/appendices/edna-case-study.md). Consent frameworks, constraints on analysis scope, and data storage governance fall under the principles of [Indigenous Data Sovereignty (IDSov)](/foundations/data-sovereignty.md) and are addressed in our concurrent ethics protocols and public materials, which should be consulted alongside this methodology (see Practice Guide [eDNA consents](/practice-guide/edna/ethical-consents.md)).

But the technology is improving too rapidly to ignore. The emergence of portable sequencing technologies such as Oxford Nanopore Technology opens the possibility of in-field data generation — from collection through to analysis — without requiring sample export, which substantially reduces logistical barriers and IDSov risks in remote settings ([Sánchez-Vendizú et al. 2025](https://doi.org/10.1038/s41597-025-05697-z)). The #SexyTrees protocol is designed to be compatible with in-field sequencing as this technology matures.

### Pragmatic first steps for eDNA

We don't need to name every species to measure biodiversity uplift. Metabarcoding works as a community fingerprint — tracking shifts in richness and evenness over time without resolving every sequence to a Latin binomial.&#x20;

A statistically significant increase between baseline and follow-up sampling is a valid, if crude, metric of ecological improvement. This falls reasonably aligned with the ISBM's successful strategy of good-enough logic, and rigorous logical parameters ([ex-post](/foundations/orthogonal-stacking.md#nature-was-complex-now-its-also-chaotic), ecosystem-compared-to-itself), and reduces dependence on tropical reference databases, which remain patchy. We acknowledge that building those databases still matters for genetic accuracy ([Di Capua et al. 2024](https://doi.org/10.1038/s41598-024-69520-2)), but academic science is not the market's science and has fundamentally different aims.

### Theoretical eDNA measurement

The #SexyTrees biodiversity crediting pathway therefore proceeds in two stages.&#x20;

* **Stage 1 data collection** (current pilots), projects establish eDNA baselines and plant #SexyTrees, generating ex-post tree survival data eligible for outcomes-based payments (see [Trees](/tree-credits.md)).&#x20;
* **Stage 2 per-project normalized curve** (in development), once reference levels are established and the IBU conversion pathway is certified, accumulated eDNA time-series data from Stage 1 plots will be eligible for retroactive IBU issuance for the uplift biodiversity dimension.

This ex-post approach is explicitly chosen over ex-ante projection. Our position is that IBU issuance is always tied to measured outcomes, not modeled predictions — consistent with both the logic of [complexity science](/foundations/orthogonal-stacking.md#nature-was-complex-now-its-also-chaotic) and the integrity requirements of commodities exchange trading.

### Stage 1 — Data collection

#### Figure H. Rudimentary diagram of Water Bucket Protocol (WPB) bucket placement.&#x20;

<figure><img src="/files/zRmOiBYU0fRXVDO1odPW" alt="Aerial schematic of a ~1-hectare agroforestry plot with four eDNA sampling buckets arranged in a diamond around the plot&#x27;s centroid, each ~10 m from its neighbours. Two light-coloured buckets (yellow, sky blue) target pollinators; two dark terrain-coloured buckets (brown, forest green) target wood-dwelling insects. A 50 m scale bar is shown; bucket symbols are enlarged for legibility."><figcaption><p>Aerial layout of the Water Bucket Protocol (WBP). Four bleach-sterilised buckets are placed around the centroid of an agroforestry plot (averaging ~1 ha)</p></figcaption></figure>

### Stage 2 — Per-project normalized curve

#### Figure Ga. Rudimentary diagram of uplift biodiversity crediting methodology, with a normalized curve generated from using eDNA from the Water Bucket Protocol (WPB).

<figure><img src="/files/kP7leEtkRH3svdYdWyII" alt="Bar-stack graph showing biodiversity integrity increasing linearly from degraded land at year zero to near-primary-forest levels by year four across hectares of SexyTrees-compliant reforestation, measured by eDNA methodology under Savimbo&#x27;s Water Bucket Protocol. Each stack represents one hectare&#x27;s measured biodiversity integrity score."><figcaption><p>Unrealistic (should be 60-80yr uplift) diagram showing eDNA linear regression method for eDNA absolute integrity sampling. </p></figcaption></figure>

This is an illustrative image of linear regression with a random selection of 1-ha plots in a project area used to generate a normalized curve.&#x20;

But the 4-year recovery to primary forest is rhetorically powerful but scientifically naive. Real eDNA-measured biodiversity recovery on tropical agroforestry typically doesn't hit primary-forest values at year 4 — published trajectories using traditional plot-based botanical inventories usually show 60–80 years to full recovery on most metrics, with some functional traits never converging ([Rozendaal et al. 2019](https://doi.org/10.1126/sciadv.aau3114); [Poorter et al. 2021](https://doi.org/10.1126/science.abh3629)). (Which is why principle #1 remains protecting primary forests first.)

Here is a more realistic (if still wildly optimistic), visual timeline for our plots.

#### Figure Gb. *Slightly* more realistic diagram of uplift biodiversity crediting linear regression.

<figure><img src="/files/AMbwx87vOOZk7F2XD37q" alt=""><figcaption></figcaption></figure>

### Ongoing development

Additional items under active development or negotiation include:

* **Integrity curve normalization**: Establishing the species richness and community composition signals that correspond to defined integrity increments in tropical agroforestry recovery trajectories, enabling translation of eDNA metabarcoding outputs into IBU-compatible integrity scores.
* **Reference database gaps:** Tropical eDNA identification depends on local species sequence databases that are currently incomplete. The protocol is being co-developed with sequencing partners and regional biodiversity data initiatives to fill the most critical gaps for the Putumayo and broader LAC agroforestry context.
* **In-field sequencing integration:** Piloting of Oxford Nanopore-based sequencing in the field to evaluate whether a closed data pipeline (collection → analysis → credit claim, without export) is operationally feasible and IDSov-compliant.
* **Third-party certification pipeline:** Three certifiers have expressed willingness to adopt the protocol into their biodiversity crediting programs upon successful validation. Formal certification requirements are being aligned in parallel with protocol development.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://sexytrees.savimbo.com/biodiversity-credits/calculation.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
