Nearshore marine bird surveys: Data synthesis, analysis and recommendations for sampling frequency and intensity to detect population trends. Exxon Valdez Oil Spill Long-Term Monitoring Program (Gulf Watch Alaska) Final Report (Exxon Valdez Oil Spill Trustee Council Project 12120114-F)
Coletti H and Wilson TL. 2018. Nearshore marine bird surveys: Data synthesis, analysis and recommendations for sampling frequency and intensity to detect population trends. Exxon Valdez Oil Spill Long-Term Monitoring Program (Gulf Watch Alaska) Final Report (Exxon Valdez Oil Spill Trustee Council Project 12120114-F). Exxon Valdez Oil Spill Trustee Council. Anchorage, Alaska
Sustainability of long-term monitoring programs requires optimization of sampling intensity and efforts to minimize costs while concurrently having sufficient power to detect a trend. While there has been critical thought in the past regarding these issues, we attempted to use currently available analytical methods to allow for use of existing data in simulations to estimate number of samples and sample frequency required to detect a specified trend as well as to examine factors contributing to variation, such as imperfect detection. We were unable to complete the full suite of analyses originally proposed, based on findings from preliminary analyses. Based on results from the Synthesis Report, we concluded that effort varied for each transect sampled and this would need to be modeled. This could include time-on-transect or actual length travelled during a single transect survey. There also was high model-selection uncertainty (all models had nearly the same AIC – Akaike information criterion). This indicated that there was still some un-modeled heterogeneity and this could be improved by calculating more appropriate habitat covariates, e.g., assigning habitat classes that take into account both shoreline type and bathymetry. Sample size also was an issue in the preliminary analysis. Although we had five replicate encounter histories for each transect, there were large uncertainties associated with estimates. Essentially, the limited number of transects did not capture the level of heterogeneity in the existing data. Despite this, the current sampling protocol represented the maximum effort that can be expended on surveys given logistical constraints. Because we cannot add more transects at this point, further discussion and analysis may have us: 1) reducing the scope of the monitoring program by focusing our efforts on specific habitats; 2) increasing the number of transects sampled; 3) changing the spatial grain of sampling (sample unit size); and 4) considering more complex model structures in a fully Bayesian framework. The optimal course of action will depend on refinement of monitoring objectives. Any discussion of objectives should address the following: spatial extent of analysis, spatial grain of analysis, target species, hypothesized population drivers, and feasible courses of action (e.g. management or conservation) if change is detected.
- Type
- Published Report
- Authors
- Coletti, Heather; Wilson, Tammy
- Date of Issue
- 2018-07
- Publisher
- Exxon Valdez Oil Spill Trustee Council
- Units
- KATM , KEFJ , SWAN
- Keywords
- Detection, encounter histories, Gulf of Alaska, Kenai fjords, marine bird, occupancy, population trend, Prince William Sound, sample size, transect length
- Subjects
- Ecological Framework: Biological Integrity | Focal Species or Communities | Marine Communities , Ecological Framework: Biological Integrity | Focal Species or Communities | Birds