We assessed how litter piles (i.e., terrestrial leaves and twigs) in the river mouth act as a refuge for the marine amphipod (Anisogammarus pugettensis) from predation. We compared the survival rates of A. pugettensis under predation pressure from flounder (Pleuronectes schrenki) using different organic debris particle size classes and different amounts of debris (batch sizes). Particle size classes were <10, 10-50, and 50<mm, which were supplied in 0-, 5-, or 50-g batches. After 24 h, A. pugettensis survival was higher in larger batches of debris regardless of particle size. In addition, we conducted feeding experiments using kelp and terrestrial leaves as food resources for A. pugettensis in June and October 2005 to estimate the usage of these food resources. Although A. pugettensis that were fed only terrestrial leaves exhibited negative growth rates, carbon stable isotope analysis in both experimental seasons indicated that terrestrial leaves were part of the A. pugettensis diet. From these experiments, we concluded that terrestrial leaves are occasionally used as a food resource by A. pugettensis, although their principal role is as a refuge.

In order to improve the accuracy of growth rates prediction in larval and juvenile Japanese sand lance Ammodytes personatus, growth analysis in body length using otolith microstructure was carried out. Because the widths of the rings near the center of otoliths are very narrow, we used a scanning electron microscope together with a biological microscope. In laboratory reared larvae, growth trajectories of body length back-calculated by the biological intercept method corresponded with those measured directly by periodical sampling. This result supports that application of this method to the growth trajectories of natural A. personatus is adequate. Upon application of the method to the fish caught by commercial boat seine fishery on the opening days of fishing seasons in 1997, 2000, 2001, and 2004, the growth in early life history of the A. personatus showed logistic-like trajectories. There was a strong correlation between growth rate at the 37th day after hatching (0.64–0.88 mm day^-1) and mid-layer water temperature in Osaka Bay during early-February. Hatching dates estimated by counting otolith daily growth increments agreed with those estimated by field investigations on mature adult fish in each year.

The deepsea smelt Glossanodon semifasciatus is one of the most important fishes for the demersal fishery in the Pacific coastal waters of central and southern Japan. In the present study, the spawning season in Tosa Bay was estimated to be from January to April (spring spawning group), and also at a lower frequency in October to November (autumn spawning group) based on gonad somatic index changes and stages of the ovaries of adult fish landed by commercial fisheries in Tosa Bay. Furthermore, the abundances of 0-year-old fish of the two seasonal spawning groups were investigated by a trawl survey. The spring spawning group was much more abundant than the autumn spawning group. The growth of 0-year-old fish was traced by the modes of the histogram of the specimens by measuring the Scale Length (SCL) of fish caught by a monthly trawl survey using a research vessel from April 2002 to March 2005. The smallest specimen sampled was about 19 mm SCL in mid April 2002 at the depth of 150 m. Although the histograms consisted of polymodal classes and at least three or more growth lines were traced, the main mode of SCL of the spring spawning group from 2002 to 2003 was 25 mm in April, 50 mm in May, 75 mm in June, 90 mm in August, 105 mm in October, 110 mm in November and about 140 mm in March in the next year. The growth rate of the 0-year-old fish in Tosa Bay is higher than that estimated for the sub-population in the Sea of Japan, possibly due to the higher temperature of the nursery ground of the Tosa Bay. The recruitment index of the spring spawned group fluctuated between years however reasons for the fluctuations require further investigation.