PAEC: Panel-based assessment of ecosystem condition of Arctic tundra

Research notes

The state of Norwegian ecosystems is under the magnifying glass in the new national System for Assessment of Ecological Condition. In 2021, the first operational assessment of Low and High Arctic tundra was undertaken. This article summarises the findings and their implications.

By: Åshild Ønvik Pedersen, Ingrid MG Paulsen, Eva Fuglei, Jesper Mosbacher and Virve Ravolainen // Norwegian Polar Institute. Jane Uhd Jepsen // Norwegian Institute for Nature Research. Rolf Anker Ims // UiT The Arctic University of Norway

Foto av landskap med en elv som renner gjennom en dal

The assessment, done by a scientific panel, concludes that while the main structural and functional attributes of the Arctic tundra ecosystems are still maintained, extensive incipient ecosystem changes from the ongoing rapid temperature increases will over time threaten the integrity of tundra ecosystems.

Foto av åpent landskap
Varanger. Photo: Jakob Iglhaut

Mandated by the Norwegian Ministry of Climate and Environment, the System for Assessment of Ecological Condition is set to determine whether and to what extent the condition of the nation’s terrestrial and marine ecosystems deviates from reference conditions. The reference condition is defined as “an intact ecosystem state”, which implies that the structural and functional characteristics of the ecosystem are under limited influence from human pressures. The system further defines a climatic baseline as “a climate as described for the climatic normal period 1961–1990”.

Scientists from the FRAM-based Climate-ecological Observatory for Arctic tundra (COAT) contributed in 2020 both to methodological developments of the system and, in 2021, by conducting the first operational assessments of the ecological condition of Low and High Arctic tundra in Finnmark and Svalbard.

Foto av åpent landskap
Sassendalen, High Arctic Svalbard. Photo: C Jaspers

First PAEC for Norwegian Arctic tundra

Last year’s assessment of High Arctic tundra in Svalbard and Low Arctic tundra in Finnmark is based on analyses of 34 long-term datasets and 66 indicators. Sixteen of these indicators are shared between the two sub-ecosystems; 26 are unique to Low Arctic tundra, and eight are unique to High Arctic tundra ecosystems. The key conclusions from the assessment are the following:

The Arctic tundra is more threatened by climate change than any other terrestrial biome on Earth and this is also evident from the assessment of ecological state of the Norwegian sector of the biome. Relative to the climatic baseline, the abiotic compartments of the ecosystems have undergone significant changes, such as generally increasing surface and permafrost temperatures, longer growing season, and shorter snow-covered season. In the Low Arctic, bioclimatic subzone D, where July temperatures are below 9°C, has essentially vanished.

However, the biotic implications of these abiotic changes are still mostly limited, and are mainly seen in ecosystem characteristics and indicators with strong causal links to climate. Some biotic components are presently on significant change trajectories, especially in the Low Arctic. This should be considered a warning of more extensive, incipient ecosystem changes. Of the two sub-ecosystems assessed, the Low Arctic tundra in Finnmark shows more pronounced and consistent deviations in biotic characteristics than the High Arctic tundra in Svalbard. These deviations are expressed as an increasing borealisation and a loss of Low Arctic endemic species (Arctic fox and snowy owl) and bird communities.

Foto av landskap med en elv som renner gjennom en dal
Komagdalen, Low Arctic Finnmark. Photo: RA Ims


The System for Assessment of Ecological Condition is intended to inform knowledge-based management of both terrestrial and marine ecosystems in Norway. How the PAEC protocol is suited for this task is discussed in a separate report (see Further reading).

Ecosystem-based management requires distinguishing between manageable and non-manageable stressors, and natural variation. PAEC does this through a focus on phenomena, which are qualitative expectations of directional change in indicators as a result of relevant drivers of change. Statistical ecosystem modelling can strengthen the validity of such phenomena, but only if long-term, ecosystem-based monitoring data are available. Generation of such data and models linked to tundra phenomena are key objectives of COAT (see previous article).

Bilde av 4 ulike kart satt sammen
Map of the climatic delineation of each of the Low Arctic bioclimatic subzones (D and E), as well as the northern-boreal zone (F) in the climatic reference period (1961–1990) and subsequent decades (1991–2000, 2001–2010, 2011–2019). The subzones are defined based on average July temperature, and subzone D here contains all areas belonging to subzone D or colder (e.g. subzone D ≤ 9°C, E = 9–12°C, F ≥ 12°C. The bioclimatic subzones set living conditions for distinctly different biological communities; e.g. subzone F forest communities. Modified from Pedersen et al (2021) Norwegian Polar Institute Report 153

Arctic climate change poses a substantial challenge because it represents — on the ecosystem level — an overriding, non-manageable driver. For tundra ecosystems that are rapidly leaving their defining bio-climatic envelopes, it will be necessary to develop management strategies based on what is realistic and desirable to achieve. Such strategies should be based on dynamic concepts of managing and modifying trajectories of change, rather than attempting to preserve past ecosystem states.

Through the formulation of phenomena and assessment of evidence for ecosystem change trajectories, the PAEC framework can be used as a vehicle for selecting realistic targets and interventions for ecosystem-based management, and then serve as a tool for assessing the efficiency of such management.


COAT logo

COAT is a long-term, ecosystem-based and adaptive observation system. It aims to unravel how climate change impacts Arctic tundra food webs, and to enable prudent science-based management. The baseline of the approach is conceptual models. These models guide the monitoring design that may include management actions. They also guide the structure of statistical models that estimate impacts and derive predictions based on the monitoring data. The approach accommodates three adaptive loops driven by (1) new knowledge, (2) needs of stakeholders and (3) new technologies/methods.

COAT takes a food web approach on three grounds:

  1. Climate change impacts on tundra ecosystems are often mediated and enhanced by trophic interactions.
  2. Humans affect ecosystems by their involvement in food webs as harvesters and managers of ecosystems services and conservation targets.
  3. Food web ecology has a strong scientific tradition in terms well-established general theories and development of new tactical models.

COAT focuses on Low and High Arctic tundra. The consortium contains five Fram Centre institutions (MET, NINA, NPI, UiT and UNIS).


Thanks to the Norwegian Environment Agency for funding PAEC and the scientific panel for Arctic tundra: H Böhner, KA Bråthen, D Ehrich, J-A Henden, K Isaksen, S Jakobsson, J Madsen, E Soininen, A Stien, I Tombre, T Tveraa, OE Tveito, OPL Vindstad, NG Yoccoz, and E Øseth.

Further reading

Pedersen et al (2021) Norwegian Polar Institute Report 153

Pedersen et al (2021) Norwegian Polar Institute Brief Report 056

Panel-based Assessment of Ecosystem Condition (PAEC)


COAT researchers, in collaboration with colleagues from the Institute of Marine Research, developed and tested the Panel-based Assessment of Ecosystem Condition. The PAEC protocol hierarchy (blue boxes) assesses, step by step: 1) the knowledge base, 2) the condition of individual indicators, 3) the condition of each of seven ecosystem characteristics, and 4) the condition of the ecosystem as a whole. Individual indicators – both abiotic (e.g. snow cover duration, winter melt days, mean July temperature) and biotic (abundance of e.g. lemming, snowy owl, and pink-footed goose; abundance and reproduction of e.g. Arctic fox and reindeer) – are assessed in terms of the extent to which expected changes (phenomena) are supported by evidence of observed changes (based on statistical analysis of the underlying data).

Figur over arktisk tundra økosystem

The formulation of phenomena is central in PAEC. Each phenomenon specifies causal links between human-induced drivers of change and indicators of ecosystem function and structure. The causal links are expressed as qualitative predictions of change in ecological indicators and what significance such changes would have for the ecosystem. Central to PAEC is also a structured evaluation of key sources of uncertainty that impinge on the assessment, including data coverage in time and space for each of the indicators, and how well the set of indicators covers each of the complex ecosystem characteristics being assessed.