Measuring, Reporting, and Verification of Forest Restoration

1.1 Quantifying restoration success
Irrespective of the nature of the restoration project, restoration objectives should be explicit, measurable, and time limited. At the outset, and prior to restoration, a baseline survey describes the current biotic and abiotic elements of the site, as well as external threats. The baseline is key to understanding the capacity of the site to achieve desirable outcomes and an appropriate reference ecosystem. A reference ecosystem represents a non-degraded version of the site, complete with its flora, fauna, and other biota, abiotic elements, functions, processes, and successional states that might have existed had degradation not occurred (Gann et al. 2019). Reference sites may be rare in regions that have few protected areas, and agencies need to be aware that assumptions in their selection may imply lower standards than would have been considered historically (i.e. referred to as shifting baselines). Based on information from multiple reference sites, a reference model predicts the expected ecosystem condition that the restoration site could achieve in the future, adjusted to accommodate changed or predicted environmental conditions.

Typically, the first condition restoration funding requires stakeholders to demonstrate is that the intervention has changed the measure of the ecosystem such that it is appreciably different to locations where the management interventions did not take place. Evidence actions have moved outcomes beyond the range of business-as-usual scenarios is the minimum standard necessary for any project to be deemed successful and the criteria that qualifies projects to claim additionality. If this is not the case and the site is still similar to unrestored sites, then either this indicator is not an effective measure of the kinds of change brought by the restoration, or it is too soon for a project to claim successful impact for that particular indicator. We refer to the expected values of an indicator in a standard, un-restored, system as the counterfactual model.
If additionality has been achieved, then the second condition for describing the quality of restoration success is based on its similarity to states describes by the reference model. Ecosystems naturally display variation, and recovery following restoration does not follow a precise orderly process, and therefore the expectations of both the counterfactual and reference models will include a range of values that reflect the observed variation in measured outcomes.

As stated, restoration objectives should be explicit, measurable and time limited. For this reason, restoration objectives should be selected from among properties that can be reliably observed and modelled at reference ecosystems. However, many ecosystems, not just forests, may take a long time to fully recover and the rate at which particular properties will approach the reference state is not necessarily linear. Therefore, within SUPERB we adopted a broader view of reference states to include sites that had been recovering for different periods of time in order to characterise the trajectory of succession, and the condition a project should have achieved after a given timeframe.

Projects should therefore claim they have achieved additional ecological outcomes only when it is highly unlikely indicator values fall within the range of expectations for a counterfactual, and should only claim they have successfully delivered restoration outcomes when the indicators confidently fall within the range of the reference model. While forest restoration may take many forms, additionality for some indicators will in practice be a trivial exercise. For example, afforestation projects that aim to diversity tree species will immediately have had an impact beyond the expectations of traditional monoculture forestry if multiple species are planted. However, to be considered high quality, the diversity of trees that persist and grow would be expected to continue to match that of other similar restoration projects. Ecosystem restoration cannot be defined by a single variable, and therefore it is important to recognise that multiple indicators will be required to ensure success is robustly defined. A project may therefore achieve rapid success for some indicators that diverge quickly (Fig. 1 Indicator A), and simultaneously still within the range of counterfactual conditions for other indicators that diverge later as the habitat matures (Fig. 1 Indicator B).