From the 1940s on, the Bay of Quinte was subjected to increased point-source phosphorus loading. This and other stresses (exploitation, fish species invasion, and perhaps climate fluctuation) altered the productivity of the ecosystem and community structure. In the late 1970s, point-source loading was greatly reduced. At the same time, walleye (Stizostedion vitreum vitreurn) produced the largest year class ever recorded and surged to abundance greater than those seen in the 1950s before its collapse. Also, the dominant invader white perch (Morone americana), which had flourished in the 1960s and 1970s, was almost eliminated by a low-temperature kill. Studies of the bay ecosystem before and after these events indicated that while nutrients and algae responded to changes in nutrient loadings, changes in secondary producers hinged on the changes in the fish community. Use of Borgmann's biomass size spectrum model provides a focus for the interpretation of nutrient- and predator-related changes in biomass distribution. Deviations of the biomass spectrum are interpreted. The biomass and production spectra are linked to both phosphorus concentrations, observed and predicted, and to fishery exploitation rates. The results suggest that while nutrients control the overall biomass of the ecosystem and the slope of the biomass spectrum, trophic interactions and perturbations (e.g. exploitation) determine the degree to which a smooth biomass spectrum is achieved.

Accuracy of time estimation by 112 preschool and elementary-school children was investigated, using a reproduction method in which Ss attempted to reproduce the time required for a hidden toy to travel along a path to a designated point. In Experiment I, half the Ss were given feedback information after each reproduction trial, providing both the direction and magnitude of error, whereas a second group simply made reproductions of the time interval after each presentation of the standard interval. All age groups in both feedback and no-feedback conditions showed initial tendencies to underestimate the standard, but all groups except the preschool Ss in the no-feedback condition showed increasingly accurate estimations over trials. Average relative estimation errors on the last trial block (trials 13–15) were 5 and 1% of the standard for the preschool and elementary-school children in the feedback conditions, and were 34 and 1% for these age groups in the no-feedback condition. Results of Experiment II, carried out with preschool Ss only, indicated that these Ss were unable to learn accurate time estimations without informational feedback between test trials: motivating praise in the absence of error-correcting information resulted in decreasing accuracy over trials.