Ecology of the vegetation of the Saskatchewan River delta

The vegetation of the Cumberland Marshes, a 2,760 km2 portion of the Saskatchewan River Delta in east-central Saskatchewan, was examined with respect to vegetation types, site characteristics, environmental control of vegetational variation, and soil and peat forming processes. As a means of obtaining an overview of the landscape and its major components, a physiognomic-floristic reconnaissance map of the vegetation was initially prepared through interpretation of aerial photographs and 190 ground checks. Field study was subsequently confined to a representative block of 660 km2. Vegetational and environmental measurements were obtained from 98 stand-samples that included all common vegetation types. These data were classified by the association analysis technique into 10 groups, distinct on the basis of species presence and absence, which fell into five environmental categories: aquatic, fen, wooded fen, bog, and alluvial stream levee. Lists of major species and average environmental attributes were prepared for all groups and typical community structures described. To examine vegetational and environmental relationships, the stand-samples of bog, fen, and wooded stream levee vegetation were further subjected to ordination by principal component analysis. Moisture regime and nutrient status were found to be major environmental gradients that interact in controlling overall vegetational composition. Relative site maturity and man-made disturbance are additional important gradients causing diversity of forest communities on alluvial levees. The floristic composition of bog communities is primarily determined by a complex gradient of relative ombrotrophy which incorporates decreasing pH values and nutrient content, and increasing organic matter content and water-holding capacity through the accumulation of Sphagnum moss. Periodic measurements of water-table depth and soil temperature along representative field-transects during three consecutive growing seasons showed that frost within the peat layer holds melt water at the surface of fens during the early part of the growing season. In sites with immobile ground water-tables, ice lenses persist until early July and in Sphagnum peat until at least early September. The presence of a continuous frost layer within the Sphagnum substrate of the black spruce bogs, during most of the growing season, isolates these communities from the surrounding minerotrophic fens and thus is instrumental in maintaining their ombrotrophic character. Wide floating mats which surround the shallow lakes of the study area, move up and down with fluctuating lake water levels, and thus possess a fairly stable moisture regime. Macroscopic examination of peat cores from representative field-transects showed peat deposits of open basins with moving sub-surface waters to be less than 1.5 m thick and derived from fen and aquatic plants. Floating mats consist of buoyant, fibric fen peat. The growth of Phragmites communis is now restricted to the outer edge of the floating mats, but the presence of old culms and rhizomes at lower levels of the peat, further landward, suggests that the species initiated the encroachment of the organic mat into the lake basins but was later replaced by sedges as the main peat-forming plants. Closed basins with impeded drainage, typically have peat deposits more than 3 m thick. Basin-filling through seral replacement of aquatic communities by fens and, eventually, by Sphagnum bogs is evident from the stratigraphy of the examined cores. The spatial and dynamic relationships of vegetation and environment in the Cumberland Marshes are visualized as primarily dependent on the original landform and drainage pattern and, secondarily, on the modifying influence of peat accumulation on moisture and nutrient regime. Man-made disturbance has also initiated some adjustments in the vegetation. Compared to the rapid, cyclic changes evident in the wetland vegetation of the neighbouring aspen grove and grassland regions, the vegetation changes in the study area are slower, less fluctuating, and essentially unidirectional