No description available.

Nous avons étudié la végétation de la région du mont Mégantic au moyen de la méthode phytosociologique de Zurich - Montpellier (Braun - Blanquet, 1951). La prospection sur le terrain s'est soldée par l'acquisition de 94 relevés de végétation, chacun complété par une étude pédologique et une étude dendrométrique. Ces informations nous ont permis de définir 13 associations, 10 sous-associations et 9 variantes. Il faut souligner la diversité de nos groupements en raison d'une physiographie particulière du mont Mégantic lui-même, dont l'altitude varie de 1 350 à 3 625 pieds. Une carte des dépôts de surface et une carte phytosociologique complètent notre travail.

No description available.

No description available.

Nous avons étudié la végétation de la région du mont Mégantic au moyen de la méthode phytosociologique de Zurich - Montpellier (Braun - Blanquet, 1951). La prospection sur le terrain s'est soldée par l'acquisition de 94 relevés de végétation, chacun complété par une étude pédologique et une étude dendrométrique. Ces informations nous ont permis de définir 13 associations, 10 sous-associations et 9 variantes. Il faut souligner la diversité de nos groupements en raison d'une physiographie particulière du mont Mégantic lui-même, dont l'altitude varie de 1 350 à 3 625 pieds. Une carte des dépôts de surface et une carte phytosociologique complètent notre travail.

I investigated the impacts of a catastrophic ice storm on the old-growth, hardwood forests of Mont St. Hilaire, Quebec. The mass of litter resulting from the ice storm of January 1998 was estimated using equations relating the basal diameter of fallen branches with branch mass for each of the ten major species. The ice storm of January 1998 produced 19.9 metric tonnes or 33.6 m3 of woody-litter per hectare. These losses of woody biomass are approximately 20 times greater than what is expected in a normal year and correspond to between 7--10% of the total above-ground biomass of the prestorm forest. This level of litter production positions the ice storm of 1998 as the most severe ice storm on record and amongst the most powerful forms of climatic disturbance experienced in forested ecosystems. / I also investigated differences in the magnitude and nature of the biomass losses sustained by each study species. While the magnitude of biomass lost by the study species was not related to either wood strength or stiffness, the nature of the biomass lost was. All species primarily lost branches less than 5 cm in diameter, but it was the relatively few branches greater than this diameter that accounted for the majority of downed biomass. Smaller branches were lost in relation to differences in species-specific mechanical properties, while larger branches appear to be lost in response to weakening by decay and other age-dependent factors. The ecological and evolutionary implications of these results emphasise the need for an analysis of the interplay between mechanical properties and canopy architecture in determining overall susceptibility to ice damage.

I investigated the effect of defoliation by the gypsy moth, Lymantria dispar L. (Lepidoptera: Lymantriidae) on the radial growth of tree species that differ in their acceptability as hosts for gypsy moth larvae. Annual growth rings were measured from 1950 to 1992 on increment cores taken from three species: red oak (Quercus rubra L.), a preferred species, sugar maple (Acer saccharum Marsh.), an acceptable species, and white ash (Fraxinus americana L.), an avoided species. The number of gypsy moth larvae on these same individual trees had already been recorded from 1979 to 1992. To remove the potentially-confounding effects of climate on radial growth, I developed regression models of growth on climate in a pre-gypsy moth period (1950 to 1975), and then used these models to predict growth in a post-gypsy moth period (1978 to 1992). The residuals from these growth models were then examined with respect to gypsy moth numbers. (Abstract shortened by UMI.)

The dynamics in Fagus-Acer forest have been frequently investigated, a particular interest being the replacement patterns and coexistence of the two dominant species, Fagus grandifolia Ehrh. and Acer saccharum Marsh. This thesis examines whether the community pattern and dynamics in a Fagus-Acer forest at Mont. St. Hilaire, Quebec, are consistent with predictions made by the disturbance hypothesis. This conceptual model explains the mechanisms underlying the coexistence and replacement patterns of the two dominant species. While the hypothesis suggests disturbance frequency and interspecific differences in growth under variable light conditions to be the key factors that determine the dynamics in Fagus-Acer forests, I demonstrate that edaphic factors and regeneration of Fagus by root sprouts can also play a significant role. Based on the findings, I propose an alternative model for the coexistence and replacement patterns in Fagus-Acer forest to better account for the influence of edaphic factors and of Fagus root sprouts.

A new sampling method that decouples spatial and environmental correlations was developed and applied to a temperate forest understory. Data were used to contrast niche theory with neutral theory, and only showed support for niche theory. A spatial and environmental partitioning analysis indicated that the effects of dispersal are primarily important within the spatial extent of environments suitable for a given species. The same sampling data were used to test correlates of non-native species invasion at a species level and as a group. The distributions of non-native plant species are also better explained by the niche model; however, non-native species do not appear to negatively impact native species, nor to be negatively impacted by native species. Together, these results suggest that the forest understory is strongly niche-structured, but likely not saturated. Diversity in this forest appears to be primarily determined by regional processes, and only secondarily by local species interactions.

The validity of niche-assembly rules in explaining community structure is revisited by testing for ecological separation among ferns. An intensive, fine-scale survey of fern abundance and environmental variation was done in 1-ha of old-growth forest. Three methods of detecting niche segregation (canonical correspondence analysis, detrended correspondence analysis and GIS mapping) suggested that most fern species at the site are distributed according to distinct environmental preferences. The most important gradients separating fern species are first, the amount of soil moisture and second, soil nitrate concentration. Contrary to other findings, pH had little influence on controlling fern distribution. Spatial autocorrelation, detected by partialled ordinations, obscured the presence of niche partitioning. As well, sampling grain changed the apparent location of some species on environmental gradients and their ecological similarity to other species. Finer-scaled environmental heterogeneity or dispersal-mediated processes may account for the unexplained variation in fern species abundance of this site.