2020 ESA Annual Meeting (August 3 - 6)

PS 55 Abstract - Simulation models account for uncertainty and reveal impacts of future development on Alaskan wildlife

Timothy Fullman1, Ben Sullender2, Matthew Cameron3 and Kyle Joly3, (1)The Wilderness Society, Anchorage, AK, (2)Audubon Alaska, Anchorage, AK, (3)National Park Service, Fairbanks, AK
Background/Question/Methods

Wildlife management often involves tradeoffs between protecting species and allowing human activities and development. Ideally, these decisions are guided by scientific studies that quantify the impacts of proposed actions on the environment. However, critical information to assess impacts of proposed activities may be lacking, such as certainty in where actions will take place, which may hinder a robust impact assessment. In Alaska, uncertainty often challenges the early stages of decision-making, such as determining which areas will be available for petroleum or mineral leasing prior to submission of specific development proposals. To address this gap, we use Monte Carlo simulation modeling of future development to predict potential impacts of various development scenarios while accounting for uncertainty. We apply our model to oil and gas leasing under a revised Integrated Activity Plan (IAP) for the National Petroleum Reserve – Alaska (NPR-A). For each alternative in the IAP Draft Environmental Impact Statement (DEIS), oil production pads and roads were randomly simulated across the available area in proportion to estimated oil availability. We assessed changes to habitat values for two caribou herds and nine bird species based on prior studies of responses to development, repeating the process 100 times for each alternative.

Results/Conclusions

Model results indicated significant differences across the alternatives. Habitat loss was lowest for the Teshekpuk Caribou Herd under Alternative A – the no-action alternative – and greatest under alternatives C and D. The eight shorebird species analyzed consistently showed the lowest levels of habitat loss under alternatives A and B, and the greatest under Alternative D. Alternative C varied by species but typically featured intermediate levels of habitat loss. Molting brant showed no impact under alternatives A and B, minimal impact under Alternative C and greatest impact under Alternative D, though these results were affected by a spatially-limited sampling area. Brant molt in other parts of the NPR-A, which may increase impacts under all alternatives. Simulated infrastructure levels suggested that infrastructure footprints predicted by the IAP DEIS greatly underestimate what may occur over time. Our results demonstrated that simulation models can account for uncertainty while exploring potential effects of future development. While any single development simulation is unlikely to reflect the actual future footprint, examining many possible iterations reveals a range of potential impacts. We used this tool to compare leasing alternatives for the NPR-A IAP, but it is equally applicable for other management decisions with spatially explicit alternatives for development restrictions.