Kevin P. Timoney

A dyin delta? A case study of a wetland paradigm

The Peace-Athabasca Delta, Canada, a RAMSAR Wetland of International Significance, has since the 1970s been viewed as an ecosystem in decline. The putative reason for its decline has been the regulation of the Peace River by the W.A.C. Bennett Dam in British Columbia. The dying delta paradigm has its origin in a short-term study that coincided with a transient drawdown of the open drainage lakes in the delta. The paradigm has circumscribed all subsequent studies of the delta in its a priori assumption that all detected changes are due to the dam and are negative. As a result, the growth of scientific understanding has been stifled. Factors that contribute to the lack of ecological understanding unclude problems of data quality, quantity, and scale, ecological complexity, media marketing, failure to consult or analyze older/historical datasets, over-reliance on gray literature, too few wetland ecologists, and too little interdisciplinary thinking. Factors that may be involved in the recent changes, or lack of changes, include climatic variation and change. normal wetland dynamism, stochasticity, flow regulation, weirs, dredging, avulsions and their prevention, influxes of weeds and contaminants, delta evolution, and cultural change. The assessment of health in a delta is problematic since deltas are naturally stressed, dynamic ecosystems. Indicators of anthropogenic stress, such as declines in diversity and abundance, changes in biomass and primary production, or retrogressive succession may be difficult to apply in deltas. Long-term datasets are required that allow differentiation of normal from anthropogenic changes. Critique of the ‘dying delta’ view reveals little scientific support. The often-stated decline in flood frequency seems to have no statistical basis. The ecological health of the Peace-Athabasca Delta was assessed based on 26 attributes; 18 attributes indicated health, 3 indicated stress/disease, and 5 attributes were either neutral or required study. When compared to other major deltas in North America, the Peace-Athabasca Delta stands out as a paragon of ecosystem health. This study presents a cautionary lesson in the power of unchallenged paradigms in shaping scientific and popular opinion. A new paradigm views the delta as predominantly healthy, driven by large-scale natural processes, complex, and dynamically varying.

Vegetation development of boreal riparian plant communities after flooding, fire, and logging, Peace River, Canada

Abstract In this study we compare and contrast vegetation development following natural and logging disturbances in a major boreal river valley. Permanent sample plots and releves were established and sampled for vegetation and landscape attributes in June and July of 1993 and 1994 in the Peace River Lowlands, Wood Buffalo National Park, Canada. In the Peace River Lowlands, primary succession is a flood-origin process. Secondary succession may be either autogenic through gap dynamics mediated by nursery logs, buried wood, and suckering, or allogenic, following fire or logging. Flood origin accounts for 72% and fire origin for 29% of the undisturbed forests. From 1951–1995, 24% of the forest land burned, yielding a fire return interval of 186 years. Forest successional trajectories are set soon after flood, logging, or fire, with little evidence of gradual replacement of one forest type by another. Vegetation composition and relative species abundance are strongly correlated with living moss depth, moss-lichen total cover, total tree cover, herb cover, and canopy height. Species with high indicator value are Hylocomium splendens, Picea glauca, Pyrola chlorantha, Equisetum pratense , and Epilobium angustifolium . Strong correlations exist between white spruce tree density and canopy height, total tree cover and canopy height, total tree cover and basal area per hectare, basal area and canopy height, and between canopy height and surface age. Clearcuts are initially dominated by rose-raspberry followed by balsam poplar (with lesser amounts of Alaska birch and aspen). After logging, temporal changes in composition and dominance occur more rapidly than during natural succession. There is no evidence of post-logging convergence toward the original white spruce and mixedwood forests; a long-term deciduous disclimax is predicted. Vegetation associations, successional pathways, landscape relationships, and ecological benchmarks are identified.

Failure of natural regeneration after clearcut logging in Wood Buffalo National Park, Canada

Abstract Logging of boreal riparian old-growth white spruce took place in Wood Buffalo National Park from 1951 to 1991. In this study we focus on one of the areas, Timber Berth 408, logged from 1966 to 1991. Regeneration surveys of white spruce were conducted in clearcuts and undisturbed vegetation during June 1994. The age of a clearcut is unrelated to either white spruce stocking or density over a 25 year period. Post-logging natural regeneration in clearcuts has failed: median stocking (% frequency in 10 m2 plots) and density (stems ha−1) of seedlings are 18.2% and 300 ha−1, and of transgressives are 14.3% and 200 ha−1. Median stocking and density of spruce (including residual growth) in the cutover areas are 36.4% and 848 ha−1. In order to achieve the level of stocking found prior to logging, ≈93% of the logged area, or 9300 ha, would require planting. There is a clear relationship between spruce stocking rates and densities in clearcuts and the distance to the nearest white spruce seed tree. White spruce seed tree distance is important in determining the ability of a clearcut to regenerate naturally. Failure of post-logging natural regeneration in Wood Buffalo National Park is attributable primarily to two factors: (1) large size of the clearcuts, placing most of any area logged beyond the effective dispersal distance of white spruce seed; (2) destruction of advance growth and lack of residual growth. Clearcutting and site preparation are shown to degrade boreal riparian ecosystem structure and function. Recommendations are provided whose goals are to maintain biodiversity and healthy ecological structure and function in boreal riparian ecosystems.

Spring ice-jam flooding of the Peace-Athabasca Delta : Evidence of a climatic oscillation

A historic record of spring ice-jam floods of the Peace-Athabasca Delta was analyzed for the years 1826–1995. The temporal pattern of flooding is non-random. The likelihood of a flood following a flood, or a non-flood following a non-flood, is greater than expected by chance. Probability analysis of flood occurrence reveals that the period 1860–1880 was a time of unusually few floods, and the period 1915–1950 was a time of unusually frequent floods. The long-term flood frequency is 1 flood in 6.25 years. Changes in flood frequency over the record reveal a pattern of oscillation described by a sine-based model that is correlated with the long-term (Gleissberg) cycle of solar activity. Monte Carlo simulation was used to test a Bennett Dam Model and a Cyclic Model. The Bennett Dam Model is unlikely to have generated the observed flood history (p=0.04). The observed flood history shows a better fit to the Cyclic Model (p=0.65). No correlations between floods and ENSO cold or warm events was detected. The most recent wet period began about 1900 and ended in the early 1960s prior to completion of the W. A. C. Bennett Dam in British Columbia. As independent corroboration of climatically-driven changes in flood frequency we present three additional lines of evidence. The pattern of annual muskrat returns (95 year record) reveals both 10 year cycles and long-term patterns that agree well with the observed flood cycle. The annual area burned in Wood Buffalo National Park is inversely related to flood occurrence. Incised channels and dendritic drainage patterns in the bed of Lake Mamawi provide probable evidence of a previous dry period in the delta. Climatic change or oscillation likely underlies the drying trend observed in recent decades in the Peace-Athabasca Delta.

Old-growth white spruce and balsam poplar forests of the Peace River Lowlands, Wood Buffalo National Park, Canada : development, structure, and diversity

Abstract Thirty-seven permanent plots were established and sampled during 1993 and 1994 in Timber Berth 408, Peace River Lowlands, of Wood Buffalo National Park. Sites were sampled for vegetation, bird, physical, and structural attributes. The transition from mature to old-growth conditions occurs at different times in balsam poplar and white spruce forests. In lowland balsam poplar forests, old-growth attributes begin to appear after a stand age of ≈ 80 years, and in lowland white spruce, after ≈ 160 years. Both forest types may persist in excess of 300 years. Attributes shared by, and unique to, both lowland balsam poplar and white spruce old-growth are described. The development of log attributes over time follows opposing trends in flood and fire-origin forests. The preferred rooting medium of white spruce shifts from mineral soil early in succession to decaying wood under pre-existing canopies. Log and snag structure, decay, and dynamics are discussed in relation to time, bryophyte and lichen communities. Canopy height and roughness, total tree cover, and snag and tree animal cavities are described; indicator communities of birds are identified. We conclude with a summary of attributes characteristic of boreal riverine old-growth forests.

Landscape Cover Change in the Peace-Athabasca Delta, 1927–2001

This study examined landscape change in the Peace-Athabasca Delta, northern Alberta, Canada. The proportion of the landscape covered by ten habitat types was determined for five vintages of air photos (1945–2001) from 24 randomly chosen study areas. To test for reproducibility, three vintages were analyzed by two independent teams, neither of which knew the dates or locations of air photos. Their results were highly correlated. Comparison with another air photo study focused in the southeastern delta revealed good agreement. Analysis of oblique air photos extended the reconstruction back to 1927. Of four cover types (water, marshes, willows, forests), only water differed significantly in area between vintages. Cover trends for the most general of types (wet communities and dry communities) indicated drying from 1927 to 2001. This trend may be due to multidecadal geomorphic evolution and climate change. Currently, the changes are statistically non-significant, but nevertheless may inform about the long-term future of the ecosystem. Large spatial and temporal variation in landscape cover is characteristic of the system. The types and ranges of change in the delta’s vegetation post-Bennett Dam (1968) do not appear unusual relative to pre-Bennett Dam change. A multi-decadal perspective is necessary to encompass normal oscillations in abundance. Key to understanding change in the delta is to think at multiple scales and to remain aware that trends or patterns are scale-dependent. The delta may not be predictable on the meso-scales relevant to society and management.

Boreal diatom ponds: A rare wetland associated with nesting whooping cranes

This paper documents and characterizes a rare form of boreal wetland associated with the nests of whooping cranes, an endangered species. Diatom ponds are found in wetlands in association with bulrush marshes. They are shallow (<50 cm deep) and vary in size from 10 to >1000 m in diameter, may evaporate down to diatom muck by late summer, are strongly influenced by dissolution of gypsum, and are circumneutral to alkaline and high in sulphates. Aquatic macrophytes are few. Pond waters are clear, and the predominant primary producers are a benthic diatom community that gives the ponds a characteristic yellow color (viewed from the air). As the diatom ponds dry, they change color from yellow to pink (when water table is at the surface) to cream (due to a dried diatom and sulphate crust). Diatomaceous earth or sedimentary peat underlies the ponds, which exist in a dynamic relationship with bulrush marshes, wet meadows, fens, and bogs. In the U.S. wetland classification system, diatom ponds fit most nearly within the palustrine, unconsolidated bottom, aquatic bed type. In the Canadian wetland classification system, the diatom ponds might fit in the marsh/shallow open water complex, with a new distinction at the type level. The association between nesting cranes and diatom ponds may be due to a combination of factors such as long sight lines for detection of predators, the proximity of bulrush (their favored nesting material), and use of the ponds for feeding.

HOW LONG MUST NORTHERN SAXICOLOUS LICHENS BE IMMERSED TO FORM A WATERBODY TRIMLINE

Saxicolous lichen trimlines are relatively level and distinct transition zones found on bedrocklined shores. They occur as a result of disturbance to the rock lichen community, typically due to high water events. Trimlines can inform about previous high water conditions, and in conjunction with lichenometry, can be dated. In wetlands which fluctuate in water level, dated trimlines may be useful in reconstruction of the timing and magnitude of former high water. We assessed seven common northern saxicolous lichen species over a period of 270 days to determine the period of immersion required for mortality. Overall mortality rates increased after about 30 days of immersion. At 90 days, mean mortality was about 68%; at 180 days, mean mortality was about 91%; total mortality of the immersed lichens was observed at day 270. There were large interspecific differences in susceptibility to immersion.Phaeophyscia sciastra was the least susceptible to immersion andXanthoparmelia somloënsis was the most susceptible to immersion. A prolonged period of high water (months, rather than days) may be required to form a trimline in a northern delta. A northern saxicolous trimline might remain visually distinct for several decades in the absence of disturbance.

Lichen Trimlines in Northern Alberta: Establishment, Growth Rates, and Historic Water Levels

Abstract This study examines water-formed saxicolous lichen trimlines on bedrock outcrops in the Peace-Athabasca Delta, northern Alberta, Canada. We determined the lichen species growing on rock above and below lichen trimlines; described the trimlines with regard to height above water, location, elevation above sea level, and physical factors; and dated the trimlines through lichenometry. Trimlines were found at open, semi-restricted, and perched basins at a mean elevation of 210.85 m asl; 14 of 16 trimline elevations were between 210.26 and 210.92. Mean trimline height above current water was 155 cm (range 40 to 230 cm). The all-lichen species diameter growth rate mean was 0.416 mm yr−1. Physcia caesia grew at 0.331 mm yr−1 and Xanthoria elegans grew at 0.441 mm yr−1. Many of the trimlines may have established after drawdown of high water that existed during the mid- to late-1970s. Based on cemetery data, establishment lag time is about six to 10 years. Most trimlines date from the period 1976 to 1990. The early 1980s appears to be a key time of basin drawdown and trimline establishment. Site and species influences on establishment lag and growth rates, and the complex hydrology, may be reflected in the range of trimline dates. The mean 95% confidence interval range in establishment dates was 6.2 years. In a hydrologically complex ecosystem such as a delta, lichenometric dating may prove a useful tool to reconstruct water level changes.

Rates of vegetation change in the Peace-Athabasca delta

Landscape changes were examined in the Peace-Athabasca Delta, northern Alberta, Canada for the period 1945 to 2001 in 24 randomly chosen study areas. Proportions of the landscape covered by four community types (open water, marshes and meadows, shrub communities, and forests) were determined on airphotos from five periods. Progressive, retrogressive, and oscillatory changes were observed. Variation in cover was about 4–10 times greater than net change. Median rates of both increase and decrease were about 0.2% to 0.6% of the landscape per year. There was evidence of temporal stability in some areas, while in other areas, a dynamic mosaic existed. Local or short-term changes were often large (> 1% per year), but averaged over the entire delta, or over time, net changes were small. Net changes were the lowest in forests and the highest in marshes. Succession between open water and marshes and between marshes and willows were common. Forest showed little evidence of transition to other types. Oscillatory changes in open water and marshes were typical. Open water, marshes, and willows changed at similar rates and more rapidly than did forest. Succession from open water and marsh communities toward willow communities and forests is to be expected at centennial scales. In a naturally oscillating system, trends appear or disappear as temporal and spatial scales change. Short duration studies, or those of limited spatial extent, may provide inaccurate estimates of vegetation change in a large dynamic delta. Interactions between drivers such as isostatic rebound, sedimentation, and avulsions mean that the Peace-Athabasca Delta will remain dynamic.