Browsing College of Natural Science and Mathematics (CNSM) by Subject "Alaska, Gulf of"
Now showing items 1-2 of 2
Assessing seasonal trends in harbor seal (Phoca vitulina) diet using stable isotope analysis along the whiskerHarbor seals (Phoca vitulina) that use tidewater glaciers in the summers to molt, pup, and breed have declined in recent years, which could be attributed to reduced prey availability caused by regime shifts in the coastal marine environment. Recording seasonal dietary trends of harbor seals in different years could improve our ability to test if regime shifts caused these declines. However, such analysis has not been possible, because wild harbor seals are difficult to recapture. Stable isotope analysis of serial sections of growing whiskers (mystacial vibrissae) can be used as a tool to assess diet over different seasons, but uncertainty about whisker growth status and shed dates have prevented accurate estimates of stable isotope deposition date in the past. In Chapter 1, I characterized harbor seal whisker morphology to improve estimates of stable isotope deposition date. First, I measured 567 whiskers collected from wild harbor seals in the Gulf of Alaska from 2003 to 2012. Measurements included the length of a smooth root section (SRS), the length of the bumpy section, and the distance between each bump (inter-bump length; IBL). I found that the SRS was longer for spring-collected whiskers than fall-collected whiskers and matched the length of fully-grown, shed whiskers. These results suggest that the SRS can be used to differentiate whisker shed and growth status, and can be used to determine the sequence of whisker shedding by cohort in summer-captured seals. I also found that the mean IBL was correlated with whisker length and provides a proxy for whisker growth rate. I compared stable carbon isotope ratios along the three longest whiskers from 10 harbor seals and found that intra-individual patterns of whisker stable carbon isotope ratios became more synchronous when expressed by deposition date rather than by position along the whisker. In Chapter 1, I proposed a method to improve deposition date estimates by applying individually adjusted growth rates and better estimates of shed date to wild harbor seal whiskers. In Chapter 2, I analyzed stable isotope ratios from serial sections of whiskers of 32 harbor seals from a population that uses tidewater glacial habitats in southeast Alaska. I used a mixed-effects repeated-measures model to determine the characteristics that influence stable isotope ratios over time. Mean stable carbon and nitrogen isotope ratios differed significantly among size classes (p < 0.005), with no effect of sex. Seals were then grouped by size to describe isotopic differences between different demographic groups using Standard Ellipse Corrected Area (SEAc). Larger seals (>1.4 m) exhibited a broader isotopic niche in the fall, winter, and spring relative to smaller seals (< 1.4 m), but had a similar niche width in the summer. These results suggest that seals using tidewater glacial habitat share common prey base in the summer, while larger seals diversify their diets throughout the rest of the year. Overall, the results of this thesis suggest whisker morphometric characteristics can be used to improve the ability to make longitudinal inferences using serial sections of the whiskers, which reveal differences in prey utilization by size class in harbor seals that merit further study.
Effect of filling methods on the forecasting of time series with missing valuesThe Gulf of Alaska Mooring (GAK1) monitoring data set is an irregular time series of temperature and salinity at various depths in the Gulf of Alaska. One approach to analyzing data from an irregular time series is to regularize the series by imputing or filling in missing values. In this project we investigated and compared four methods (denoted as APPROX, SPLINE, LOCF and OMIT) of doing this. Simulation was used to evaluate the performance of each filling method on parameter estimation and forecasting precision for an Autoregressive Integrated Moving Average (ARIMA) model. Simulations showed differences among the four methods in terms of forecast precision and parameter estimate bias. These differences depended on the true values of model parameters as well as on the percentage of data missing. Among the four methods used in this project, the method OMIT performed the best and SPLINE performed the worst. We also illustrate the application of the four methods to forecasting the Gulf of Alaska Mooring (GAK1) monitoring time series, and discuss the results in this project.