Cumulative impacts of a gravel road and climate change in an ice-wedge polygon landscape, Prudhoe Bay, Alaska

Environmental impact assessments for new Arctic infrastructure do not adequately consider the likely long-term cumulative effects of climate change and infrastructure to landforms and vegetation in areas with ice-rich permafrost. This is due in part to lack of long-term environmental studies that monitor changes after infrastructure is built. This case study examines long-term (1949–2020) climate- and road-related changes in a network of ice-wedge polygons, Prudhoe Bay Oilfield, Alaska. We studied four trajectories of change along a heavily traveled road and a relatively remote site. During 20 years prior to oilfield development, the climate and landscapes changed very little. During 50 years after development, climate-related changes included increased numbers thermokarst ponds, changes to ice-wedge-polygon morphology, snow distribution, thaw depths, dominant vegetation types, and shrub abundance. Road dust strongly affected plant-community structure and composition, particularly small forbs, mosses, and lichens. Flooding increased permafrost degradation, polygon center-trough elevation contrasts, and vegetation productivity. It was not possible to isolate infrastructure impacts from climate impacts, but the combined datasets provide unique insights into the rate and extent of ecological disturbances associated with infrastructure-affected landscapes under decades of climate warming. We conclude with recommendations for future cumulative impact assessments in areas with ice-rich permafrost.

Better cumulative impact assessment guidelines are needed for new Arctic infrastructure in regions with ice-rich permafrost There is a scarcity of long-term environmental studies that have monitored changes after infrastructure was built. • Focus on the often-ignored indirect impacts that follow the direct impact (footprint) of new roads. U.S. definition of cumulative impacts "...the impact on the environment which results from the incremental impact of the action when added to other past, present and reasonably foreseeable future actions, regardless of what agency (federal or non-federal) or person undertakes such other action. Cumulative impacts can result from individually minor but collectively significant actions taking place over a period of time." (Council on Environmental Quality, 1997).
• CIs are difficult to predict because of the complexity of ecological interactions, the scarcity of environmental baseline data, and the difficulty in defining the spatial and temporal boundaries for meaningful assessments (Clark 1994).
• Objective rules for conducting CIAs are also generally lacking (Jones 2016).
• CIAs are generally not integrated into comprehensive regional planning processes, so it is difficult to follow through on generated recommendations (NRC 2003).
•Nonetheless, CIAs are often mandated for large projects despite these difficulties because of the potential large long-term consequences that would likely occur in the absence of a thorough consideration of cumulative impacts

Site description
The Alaska North Slope oilfields, major infrastructure, and location of study areas

Comparison of plots sampled in 2014 and 1970s
Loss of diagnostic species: Diagnostic species: Species with high fidelity to a given vegetation unit, i.e., they are regularly found in the given vegetation unit and are not regularly found in other units. Defined by the phi value:

Comparison of Colleen site vegetation patterns in 2013 and 1970s
Mapping

Variation with distance from road
Coleen transects T1 & T2: Thaw depth, snow depth, water depth, patterned ground features

Comparison of key site factors along JS and CS transects:
Center-trough elevation difference Vegetation-type distribution

Summary of of the four trajectories of change
Trajectory A: Pre-road Jorgenson and Colleen sites (1949-1968)

Drivers of change
Climate: • Climate data from Barrow (1949 and1968) • Mean annual air temperature: -12.6 ± 1.2 ˚C; a slight cooling trend Other: • Natural landscape-evolution processes (e.g., annual frost heave, successional processes related to pond expansion and drainage, small annual input of eolian dust from the Sagavanirktok River).

Trajectory A cumulative impacts
• Little detectable change in thermokarst ponds (l< 1% change at Jorgenson and Colleen sites), landforms, or vegetation patterns.

Summary of of the four trajectories of change
Trajectory C: Colleen Site T1 side of road Climate change and road dust (1969-present)

Trajectory C cumulative impacts:
Climate change • Similar changes in pond area, microtopography, snow, thaw layers as the Jorgenson Site. • Greater cover of deciduous shrubs near roads is likely due to a combination of warmer climate and disturbed soils. Dust • Large dust impacts to vegetation included much less cover of evergreen shrubs, forbs, mosses, and lichens compared to undisturbed vegetation types sampled in the 1970s. Loss of many diagnostic species, particularly lichens in undisturbed lichen-rich moist tundra.
• Accumulation of dust in troughs near roads reduced centertrough elevation contrasts and thermokarst compared to the JS.

Flooding
• Cumulative landscape effects of flooding on the T2 side compared to the other trajectories include the thickest active layers in polygon centers, and deepest thaw depths, snow depths, and water depths in polygon troughs. • The leaf area index (LAI) was overall 34% higher in vegetation plots on the flooded side of the road compared to the T1 side and 2.1 times higher in polygon centers on the T2 side compared to T1 centers. The higher productivity is likely due to wetter early-summer soil moisture regimes, deeper thaw, higher rates of organic-matter decomposition, more nutrients from the dust, and high inputs of feces and decayed organic matter from the waterbirds that persistently graze the area. • Enhanced productivity and erosion of mineral material into the troughs is adding to the deeper litter layers and helping to protect ice wedges from further thaw (Kanevskiy et al. 2017(Kanevskiy et al. , 2022.

Conclusions
Climate change impacts • The explosive growth in the numbers and size of small icewedge thermokarst ponds during the past 30+ years has transformed local microtopography, drainage patterns, vegetation, and ecosystem processes in ice-rich permafrost areas with near-surface ice-wedges.
• The increase in erect-shrub cover since the 1970s near roads and at over 200 m from the roads is likely a consequence of a combination the warming climate and road-related disturbances. This increase of shrubs in cold coastal landscapes is occurring more slowly than in inland areas.

Dust impacts
• Road dust has changed the distribution of common plant growth forms and the occurrence of many diagnostic species, especially small forbs, mosses, and lichens.
• Many dust-related impacts have logarithmic relationships to distance from the road and now evident in areas that are over 200-m from the road, especially on the downwind side of the road.
• Road dust has complex relationship with snow drifts, affecting ground temperatures, active-layer thickness, the timing of snow melt and green-up, and ice-wedge degradation near roads.