Abstract

This paper presents an updated analysis of trends in the abundance of harbour porpoise, Phocoena phocoena, in central and northern California, for the period 1986-95. The most recent survey effort (1995) was comparable to previous years, and regional patterns of density were similar to those found on past surveys, with densities lowest south of Monterey Bay, intermediate from Monterey Bay to the Russian River and highest off northern California. An analysis of covariance model was constructed to test for a trend in abundance while accounting for the effects of sea state, cloud cover and area. The results are qualitatively similar to those obtained for the 1986-93 time series, but encounter rates were higher in 1995, and the estimated rate of decline over the entire time period changed from 9.4% to 5.9% per year. The decreasing trend is no longer significant at a = 0.10 (p = 0.149). A power analysis based on Monte Carlo simulations revealed that power remains low to detect trends of less than 10% per year. Possible effects of oceanographic conditions, as measured by the September average sea surface temperature anomaly (SSTa), on porpoise abundance are investigated using two different techniques. correlation tests indicate an inverse relationship between SSTa and relative porpoise abundance for the eight survey years. The correlation is greatest when considering the change between survey years (decreases in relative abundance and increases in SSTa), rather than the individual values of relative abundance and SSTa. An alternate, Poisson-based generalised additive model (GAM) of porpoise sighting rates in relation to area, sea state, cloud cover, year and SSTa indicates a significant, non-linear effect of sea surface temperature on porpoise sighting rates, with no significant year effect once SSTa is included. These results suggest that harbour porpoise may exhibit interannual movement in and out of the study area in relation to changing oceanographic conditions.