We apply the tree-ring technique of crossdating to generate highly accurate age data and evaluate error in annual growth increment (annual growth zone) counts for long-lived Pacific geoduck ( Panopea abrupta ) in the Tree Nob Islands, northern British Columbia, Canada. Crossdating is the most fundamental procedure of tree-ring analysis and is based on the tendency of environmental variability to synchronize the growth of all individuals at a given site. By crossmatching these synchronous growth “bar codes”, all growth increments can be correctly identified and assigned the correct calendar year, including the innermost year of recruitment. In this analysis, a total of 432 geoduck individuals were aged using crossdating methods as well as annual growth increment counts. The entire crossdating process was completed using visual techniques, requiring no additional equipment beyond a microscope or microprojector. When compared with crossdated ages, growth increment counts consistently underaged Pacific geoduck, particularly in the oldest individuals. These inaccuracies obscured major recruitment pulses and underestimated the rarity of strong recruitment events. To date, crossdating has been used to develop growth chronologies in a variety of marine and freshwater bivalve and fish species, but no study has demonstrated how the technique can be used to dramatically and economically improve accuracy in age data.

The hunting performance of the common merganser (Mergus merganser) was evaluated in relation to prey density and merganser flock size by stocking three enclosed sections of a natural stream with known densities of juvenile coho salmon (Oncorhynchus kisutch). Two size classes of coho were stocked: smolt averaging 43 g at densities of 0.02–0.65/m2, fry averaging 2 g at 0.08–1.6/m2 and various mixtures of smolt and fry at a combined density of 0.65/m2. The stream enclosures differed in the amount of cover available to fish. Mergansers were less successful at capturing coho smolt or fry in the enclosures wth cover from undercut banks. Smolt exposed to mergansers earlier that day were less vulnerable than smolt with no previous exposure. The feeding success of individual mergansers was not significantly affected by flock size for flocks of 25 birds or less. A smoothly asymptotic functional response (type II) was observed under all experimental conditions. Coho smolt were selected over coho fry. It is concluded that a merganser's average daily food requirement (ca. 400 g) can be satisfied at smolt densities of 0.02–0.30/m2 depending on the availability of cover for smolt and their previous exposure to mergansers.

Geoduck (Panopea abrupta) stocks are perceived as stable and their fisheries as sustainable, but this may reflect a mismatch between slow-paced dynamics (maximum recorded age 168 years) and short-term perception. Management is based on biological reference points, whose appropriateness as a means to ensure sustainability is limited by a sedentary lifestyle and long-term trends in productivity. Analysis of age frequency distributions for 1979–1983, postharvest recovery rates measured in Washington in tracts pulse-fished during the 1980s and 1990s, and age frequency distributions compiled in British Columbia during the 1990s consistently suggest that recruitment declined for decades (long before the onset of the fishery), reaching a minimum around 1975, and rebounded afterwards. In such scenario, reliance on biological-reference-point-based harvest rules without timely feedback could accelerate population declines, eventually driving an apparently sustainable fishery to collapse. The merits of approaches that rely on monitoring and feedback using data-driven decision rules are stressed. Transition from a biological-reference-point-based strategy to one based on monitoring and feedback will demand a shift in research focus to the design of practical monitoring programs and the evaluation of management procedures by means of simulations. For geoducks and other long-lived organisms, monitoring should integrate data informative at different temporal scales.

In this paper we develop a multivariate analysis of length and age at maturity that includes the univariate maturity model of Schnute and Richards (1990, Can. J. Fish. Aquat. Sci. 47: 24–40) as a special case. In addition, we address the problem of drawing meaningful conclusions from large data sets oriented to fish maturity, and we present statistical tests of such conclusions. We illustrate our approach with comparisons among male and female lingcod (Ophiodon elongatus) from three stocks along the coast of British Columbia, Canada. From the univariate analysis, we demonstrate that male lingcod mature at a smaller size than female lingcod, and that for each sex, size at maturity increases with latitude. From the multivariate analysis, we determine that length and age together provide a better prediction of lingcod maturation than either variate considered alone. The multivariate model is applicable to any situation for which one or more positive variates is asymptotically related to a probability measure in the range (0, 1).