Somewhere between acceptable and sustainable: When do impacts to resources become too large in protected areas?

Utilization of marine and terrestrial protected areas is fundamentally important for their acceptance and success. Yet even appropriate uses can negatively impact resources requiring managers to make decisions as to when the impacts become unacceptably large. These decisions can be difficult because the level at which impacts occur may be far below the level at which resource persistence is threatened. In Glacier Bay National Park, Alaska, managers must make a recurring decision regarding the number of cruise ships that are allowed to enter the park each year. Cruise ships bring >95% of all visitors to the park but have been involved in several lethal collisions (ship strikes) with humpback whales. Using an individual-based simulation model, we demonstrate that increasing the annual ship volume from current to maximum allowable levels would have negligible impacts on population growth of whales. Over the next 30 years the median number of collisions would likely increase from 3 (95% CI: 0-7) to 4 (1-8) or, worst case scenario, from 5 (0-7) to 8 (3−13), while median annual growth rates would, at most, shift from 4.4% (3.7%-5.2%) to 4.2% (3.5%-4.9%), depending upon assumptions. By comparison, a median of 67 (50-82) ship strikes would need to occur over the next 30 years to threaten the persistence of whales. Confronted with an impact level that is far below what would threaten the conservation of whales, managers are tasked with the decision of placing values on 2 million additional visitors for every additional dead whale. We argue that decision-making related to use-impact trade-offs for protected areas could be more consistent and effective if site-values are defined explicitly, clearly communicated among stakeholders, and linked to biological metrics. Protected areas managers can then utilize monitoring programs to evaluate management effectiveness when the objective is conserving both resources and values.