Document: GAI-3-65-33

Detecting change across broad scales: An inferential design for intertidal monitoring at Glacier Bay National Park, Alaska.

IRVINE, G.V.*

USGS-BRD, Anchorage, AK USA ¹

Abstract:
The detection of change in populations and communities is the goal of many monitoring programs. However, the ability to detect change becomes more complex if broad physical scales and co-occurring spatial variation are involved. Managers of expansive national parks, such as Glacier Bay National Park and Preserve in Alaska, are faced with daunting challenges imposed by the tremendous physical scale of the environments they must manage. Additionally, at Glacier Bay NP&P, gradients set up by the retreating glaciers are correlated with "longitudinal" patterns in species diversity of intertidal communities. A major challenge in designing and implementing a monitoring program is how to balance the ability to sample across broad physical scales and make inference to the entire managed area versus the sampling intensity necessary to detect changes occurring within biological communities. A three-staged inferential sampling plan was devised that included: 1) aerial surveys of a subset of coastal segments; 2) low intensity sampling of a large number (25) of sites of a selected habitat type; and 3) high intensity sampling of a few (6) sites of a selected habitat type. An adaptive management approach was adopted, where results of the first phase of sampling informed the decision-making leading to the next phase, etc. Aerial surveys characterized the frequency of different habitat types, and also revealed an unexpected predominance of cobble/boulder habitat. Power analyses of data from the first year's sampling indicated that sampling a large number of sites, even at low intensity, provided a high probability of detecting change in the predominant species (Fucus, mussels, barnacles). The ability to detect change in mussels is contrasted with the gradient in species diversity. The inferential design of the sampling allows managers to extrapolate the results to similar habitat throughout the bay.

Keywords: monitoring, intertidal, inferential design, power analyses, mussels, Fucus, barnacles, Alaska

This abstract is being presented at: 11:00 AM in session:
Oral Session #56: Metapopulation Analysis.