The behavior of eight species of slug (Dergceras reticulatum, D. caruanae, Arion atar, A. subfuscus, A. circumscriptus, A. hortensis, Limax saximus, and Ariolimax columbianus) and the snail, Cepaea nemoralis, was studied in relation to weather, shelter, food, and competition. Experiments that depended on artificial shelters quickly showed how the distribution and abundance of slug populations could vary within large outdoor enclosures. If undisturbed by aggressive individuals, the molluscs selected shelters closest to their food. All species were capable of returning to shelters by homing on an odor they deposited there, and by following slime trails. During hot, dry weather the animals usually returned to the same shelter repeatedly, but they were less likely to do so in the wet, cool weather of spring and fall. Adult L. maximus, D. caruanae, and A. subfuscus became highly aggressive during the summer months. Their attacks caused smaller conspecifics and other non-aggressive species to avoid shelters they occupied. Slugs were not so aggressive in spring and fall, and not at all in winter. Three hundred and fifty eight hourly observations of molluscan activity and weather were made on 21 nights from May until October, 1976. Factors causally important to the activity patterns of the molluscs were determined by controlled laboratory experiments. These factors were included in a multiple correlation-regression analysis of daily and seasonal activity patterns in relation to weather. The analysis was also performed for each species using weather data from the previous hour's observation. Equations incorporating lag-weather explained slightly more variability than did those that used concurrent weather. The best r² values obtained for the subterranean species, A. hortensis and A. circumscriptus were 0.2269 and 0.5533 respectively. For the other species studied, r² values ranged from 0.7283 to 0.8966. Factors included in the regression equations, in descending order of importance, were: time of day (circadian rhythm), surface temperature, light intensity, photoperiod, time of sunset, temperature gradient (shelter to ground surface), wind speed, moon phase, atmospheric moisture, changes in light intensity, barometric pressure, shelter temperature (acclimation), changes in barometric pressure, and temperature changes, age and hydration were also shown to be key factors in other experiments. A model incorporating weather thresholds estimated from field data explained 83.06% of the variability in the activity of Limax maximus over the season. The values predicted from the model did not differ significantly from those actually observed in the field (Kolmogorov-Smirnov test, p>0.50). Seasonal changes in the strength of the homing response, activity patterns, and the aggressiveness of the three species noted above were closely related to one another through their mutual association with weather. / Land and Food Systems, Faculty of / Graduate