Measuring, Reporting, and Verification of Forest Restoration

Ecosystems are multidimensional

Ecosystems are composed of both living (biotic) and non-living (abiotic) components that interact in complex ways. Although explicit consideration of the entire suite of components and interactions in an ecosystem is impossible, the reference model should be developed with the aspiration of including as many components and interactions as feasible and at minimum should include indicators for each type of ecosystem attribute. Holistic approaches to restoration will track progress across as many sub-attributes as possible. For example, the Society for Ecological Restoration recommend selection of attributes that evidence six categories:

Absence of threats: e.g. direct degradation due to invasive species, pests, pathogens are minimal or effectively absent.
Physical conditions: e.g. physical and chemical conditions of soil, water, and topography required to sustain the ecosystem are present.
Species composition: e.g. species characteristic of communities observed in the appropriate reference ecosystem are present.
Structural diversity: e.g. differences in forest canopy height and age structure, or complex faunal trophic levels including predators and parasites present.
Ecosystem function: e.g. appropriate levels of growth and productivity, nutrient cycling, decomposition, and rates of natural disturbance are present.
External exchanges: e.g. the site is connected to other habitats in the landscape providing positive abiotic and/or biotic flows. 

Selecting appropriate response variables to address restoration objectives requires an understanding of the ecosystem, but frequently several good response variables may be identified. Selection can then be based on cost, ease, accuracy, timing, frequency, subjectivity of measurement, and duration for expected response. Yet whenever possible response variables that can be empirically quantified (rather than subjectively determined) should be chosen to enable objective interpretation.

The rate of restoration progress is multidimensional

A key to understanding why multiple ecological objectives are informative is that at a particular point in space and time, measures of environmental properties and biological indicators integrate processes over variable periods of time preceding collection. Generally, organism size and life span determine the spatial and temporal scales of their responses to environmental conditions, and therefore different biotic groups (e.g. microbe, invertebrate, vertebrate) can be used reflect different ecological processes. Measurements can therefore target a range of organisms, not only to demonstrate conservation outcomes, but to identify the nature of underlying functional changes operating at different times (Figure 2). Not all measures need to be collected with the same intensity, and thus as our understanding of ecosystem responses improves, we will better understand how to adapt our monitoring to also acknowledge the effects of disturbances events versus longer-term regional trends.




Figure 2. The spatial and temporal scales over which environmental conditions change, and how these vary by taxonomic groups (adapted from Dafforn et al. 2016).

As practitioners, there is great value in the ability to routinely demonstrate the continual progression of ecological restoration, however slow it may appear to our eyes. Humans are innately biased toward optimism; underestimating the likelihood of failure, or the time in which we should expect success. Reliable reference models are critical to define properties that indicate the transition of the ecosystem along the restoration trajectory and are therefore suitable indicators of success; but also because a reference trajectory will temper expectations of all stakeholders at the start of projects, highlighting when in the succession of the restoration trajectory different indicators could be realistically expected. Monitoring a variety of response indicators at different levels of organisation can reflect the establishment of an ecosystem processes. Sometimes the change in the environmental conditions may be measurable as soon as the work is complete (e.g. physical alteration to standing canopy structure), whereas in others the effects must “trickle-down” to finer levels of organisation (i.e., component/habitat and species) that may take years.

Monitoring Inputs vs. Monitoring Outcomes

It is important to draw a contrast between restoration projects that deliver well-defined measures of impact, and those that simply report project inputs. Inputs are the resources invested, usually counted as staff, equipment, consumables, and money. Similarly, outputs of these investments might be recorded as material or legal products, such as the total numbers of hectares restored, or kilometres of fencing, or numbers of public engagement events. Inputs are rarely directly related to ecological outcomes, and often the relationships are indirect and weak, but assumed to be positive on a theoretical basis. By contrast, measures of impacts are the gains in the ecological content or state of restoration projects. Inputs have dominated past environmental assessment because historically inputs were much simpler to document relative to biodiversity, but their imprecise meaning has potentially wasted resources and obscured progress for practitioners (Pressey et al. 2021). A variety of technologies now exist to collect biodiversity information and environmental properties at scale, and measures of impact should therefore be used to claims of additionality and restoration success (Section 1.1).

Naturally forest restoration is a slow process, and during early phases of a project practitioners often need to recover at least some of their costs when there is little evidence of ecological benefit. At this stage the most relevant indicators of progress are those confirming actions invested in delivering habitat interventions as well as high quality restoration practice e.g., the total number of seedlings planted, seed provenance, local community engagement, governmental administration (link to the D7.4 report). Inevitably the indicators of success for forest restoration in the short to medium term must rely on output measures e.g., survival rate of planted seedlings (expressed in %), browsing damage rate (%), restored area (ha). Nevertheless, while such metrics may establish responsible project delivery, they do not guarantee ecological objectives will be secured.