The Qu'Appelle Valley drainage system provides water to a third of the population of the Canadian Great Plains, yet is plagued by poor water quality, excess plant growth, and periodic fish kills. Fossil algae (diatoms, pigments) and invertebrates (chironomids) in Pasqua Lake were analyzed by variance partitioning analysis (VPA) to determine the relative importance of climate, resource use, and urbanization as controls of aquatic community composition . From fossil analyses, we identified three distinct biological assemblages in Pasqua Lake. Prior to agriculture (ca. 1776-1890), the lake was naturally eutrophic with abundant cyanobacterial carotenoids (myxoxanthophyll, aphanizophyll), eutrophic diatoms (Stephanodiscus niagarae, Aulacoseira granulata, Fragilaria capucina/bidens), and anoxia-tolerant chironomids (Chironomus). Principal components (PCA) and dissimilarity analyses demonstrated that diatom and chironomid communities did not vary significantly (P Ͼ 0.05) before European settlement. Communities changed rapidly during early land settlement (ca. 1890-1930) before forming a distinct assemblage ca. 1930-1960 characterized by elevated algal biomass (inferred as ␤-carotene), nuisance cyanobacteria, eutrophic Stephanodiscus hantzschii, and low abundance of deep-water zoobenthos. Recent fossil assemblages (1977)(1978)(1979)(1980)(1981)(1982)(1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992)(1993)(1994) were variable and indicated water quality had not improved despite 3-fold reduction in phosphorus from sewage. Comparison of fossil community change and continuous annual records of 83 environmental variables (1890-1994) using VPA captured 71-97% of variance in fossil composition using only 10-14 significant factors. Resource use (cropland area, livestock biomass) and urbanization (nitrogen in sewage) were stronger determinants of algal and chironomid community change than were climatic factors (temperature, evaporation, river discharge). Landscape analysis of inferred changes in past algal abundance (as ␤-carotene; ca. 1780-1994) indicated that urban impacts declined with distance from point sources and suggested that management strategies will vary with lake position within the catchment.

Analysis of diatoms, algal pigments, and chironomids in sediment cores from two otherwise similar prairie reservoirs demonstrated that differences in reservoir formation (river valley impoundment versus lake inundation) and hydrological regime (variable versus stable water level) resulted in distinct patterns of aquatic community change. Lake Diefenbaker, a 500-km2 reservoir created by damming the South Saskatchewan River in 1968, experiences water level fluctuations of 6 m·year-1. In contrast, impoundment of Buffalo Pound Lake in 1952 flooded a natural lake, raised mean water levels ~2.0 m, and reduced water level fluctuations from ~3 to <1 m·year-1. Comparison of fossil records showed that reservoir formation did not inevitably lead to eutrophication. Lake Diefenbaker exhibited typical reservoir ontogeny with three trophic periods, including an initial ~4-year period of eutrophy, a decade of mesotrophy, and a gradual shift to modern productive conditions. Planktonic taxa dominated diatom communities at all times, whereas benthic chironomid and algal remains were rare. In contrast, pigment analyses suggested that phytoplankton standing crops declined after impoundment in Buffalo Pound Lake but that chironomid and macrophyte populations expanded. Such site specificity in trophic development appears to result from differences in the extent of inundation (500 versus ~5 km2) and the magnitude of subsequent water level fluctuations (6.3 versus <1 m).

1. Palaeolimnological analyses of fossil diatoms and pigments were conducted in four lakes of the Qu'Appelle Valley, Saskatchewan, Canada, to quantify the effect of upstream depositional basins on lake response to urban and agricultural human activities. Pasqua, Echo, Mission and Katepwa lakes exhibit similar modern limnological characteristics, lie sequentially downstream from urban point sources of growth‐limiting nitrogen (N), yet drain similarly large areas of farmland (38–40 × 103 km2). 2. Analyses indicated that all lakes were naturally productive, contained eutrophic diatoms (i.e. Stephanodiscus niagarae, S. hantzchii, S. parvus and Aulacoseira granulata), and supported blooms of colonial (as myxoxanthophyll) and potentially toxic N‐fixing cyanobacteria (aphanizophyll), even prior to the onset of European settlement (ca. 1890) and urban development (ca. 1930). 3. The onset of agricultural practices ca. 1890 had only modest effects on algal communities in the Qu'Appelle lakes, with subtle increases in eutrophic diatom species (Pasqua, Mission and Katepwa lakes) and 25–50% increases in pigment‐inferred algal abundance (Echo, Mission and Katepwa lakes). 4. Despite naturally high production, total algal abundance (β‐carotene) in upstream Pasqua Lake increased by more than 350% after intense urbanization beginning ca. 1930, while eutrophic diatoms became more common and cyanobacteria populations increased ten‐fold. Principal components analysis (PCA) explained 64% of diatom variance, and identified three eras corresponding to baseline, pre‐agricultural communities (1776–1890), an era of high production (ca. 1925–1960) and recent variable community composition following tertiary treatment of urban sewage (ca. 1977–1990). 5. Analyses of three downstream lakes demonstrated that urban impacts following 1930 remained evident in fossil profiles of β‐carotene and myxoxanthophyll, but that large blooms of N‐fixing cyanobacteria were restricted to the past 25 years at downstream Mission and Katepwa lakes. Similarly, PCA showed that fossil diatom assemblages exhibited little directional variation until the 1970s. 6. Together, these analyses support the hypothesis that upstream lakes were effective at reducing the impacts of point‐source urban nutrients on downstream lakes. In contrast, diffuse agricultural activities had only limited impacts on water quality and these were less well ameliorated by upstream basins.

Diatom valves and chrysophyte scales were analyzed in sediment cores to assess historical (post-1860) and recent (since 1970) trends in lake water pH, total [Al], and [Ni] in Lumsden, George, and Acid lakes (Killarney Provincial Park, Ontario). Although acidification in Lumsden and Acid lakes started in the 1930s, rapid acidification only occurred since the 1960s. As lake water pH declined, inferred [Al] greatly increased causing the extirpation of the sport fisheries. In nearby George Lake, acidification started in the 1920s, and inferred lake water pH declined about 0.5 of a pH unit by 1960. The fish community in George Lake survived where the inferred acidification was less severe than Acid and Lumsden lakes. As in previous paleolimnological studies, chrysophytes generally indicate earlier and more marked acidification trajectories when compared with trends inferred from diatoms, probably because the vernal blooming and euplanktonic chrysophytes are tracking spring pH depressions. Although all three lakes have acidified and metal concentrations have increased during this century, our study provides evidence that these lakes, located as far as 60 km away from Sudbury, are showing signs of biological and chemical recovery as a result of recent reductions in SO2 emissions from the Sudbury smelters.