Phytoplankton dynamics in relation to turbidity and other environmental factors in Lake Diefenbaker

Climate warming (i.e., drought, flooding, rising temperature) and the increased demand for water from reservoirs promote algal blooms in reservoirs. Lake Diefenbaker (LD) is a poorly studied reservoir in the Canadian Prairies that is susceptible to episodic algal blooms, especially in summer and fall following calm periods. These booms are of concern to residents. LD has experienced variable flow from its major tributary, the South Saskatchewan River (SSR), which receives its water from the melting of snow-pack in the Rocky Mountains (Alberta). It is expected that as the Canadian Prairies continue to undergo climate change, lakes located in this region are predicted to experience more frequent extreme events. Water withdrawal from the reservoir is an additional problem and plans are underway to increase water abstraction from LD for irrigation. But our knowledge remains unclear about how climate variability (i.e., extreme hydro-meteorological events such as drought and flooding and associated turbidity) due to climate warming and human activities will influence the phytoplankton dynamics in LD. Therefore, reconstructed historic turbidity and chlorophyll-a levels from Landsat-images, and long-term, extensive field observations were used to investigate how algal biomass and composition with turbidity will respond under different hydro-meteorological and limnological conditions in LD. \nThe SSR is naturally turbid with clay-sized particles with inherently low settling velocity. However, long-term records of turbidity are not available to fully understand the factors associated with it in Lake Diefenbaker. Therefore, I reconstructed historic turbidity levels using Landsat-imagery on LD and examined the factors that were associated with it (Chapter 2). Estimated turbidity decreased spatially and temporally which is likely associated with the decline in the SSR flow and the settling of suspended sediments. Tributary flow and wind speed explained 64 %, 54 % and 69 % of the variability in estimated turbidity levels in the riverine zone, in the transition zone and in the entire reservoir, respectively. The relationship between estimated turbidity and wind speed (WS) may be associated with the re-suspension of bottom sediment at the upper reach of LD. I observed high turbidity in 2002 that exceeded other estimates of turbidity. Because a prolonged drought preceded 2002, the high estimated turbidity may be related to an increase in sediment loads from the SSR flow and an increase in shoreline erosion from a rise in LD’s water level, this is novel information that would have been impossible to obtain without Landsat-imagery. \nHydrological drought as characterized by low river inflow rate, and in turn, low water level (WL) and greater water residence time (WRT), may support greater algal biomass and blooms in lakes. To test the hypothesis that periods of lower inflows are related to increased algal biomass in LD, I examined the association between summer chlorophyll-a and hydrologic variables. A 31-year period of hydrologic data and estimated chlorophyll-a (i.e., derived from Landsat imagery) was examined (Chapter 3). Summer estimated chlorophyll-a increased as WL decreased in late spring/early summer (r2adj = 0.30, P = 0.00081). This inverse relationship may be a function of improved light conditions and internal nutrient loading. Summer estimated chlorophyll-a had a weak, inverse relationship with late spring/early inflow rate (r2adj = 0.12, P = 0.0315) that was likely driven by an extreme high flow event in 2011. Summer estimated chlorophyll-a was not significantly related to WRT (r2adj = 0.06, P = 0.1), which may be due to the limited variability in WRT and the size of LD.\nReduced water levels in reservoirs due to drought, or water abstraction, or both are associated with high cyanobacterial abundance in reservoirs. Because Lake Diefenbaker has experienced reduced water levels associated with drought and plans are underway to increase water abstraction from LD for irrigation, I used a nine-year dataset that comprised a drought year (1984), four consecutive high flow years (2011 to 2014), and four subsequent low flow years (2015 to 2018) to investigate how these changes could affect the major phytoplankton groups and cyanobacterial community (Chapter 4). Diatoms and cryptomonads dominated the algal community under low and high flow years. Diatoms were associated with greater mixing in late spring and fall, whereas the cryptomonads were related to the high nutrients from spring flow. Cyanobacteria dominated the phytoplankton (79.3% of the total biomass) under drought, which may be associated with thermocline deepening and subsequent internal loading of nutrients as observed in other reservoirs. Microcystis, a potential bloom-forming and toxin-producing genus, was dominant during the drought and correlated with reduced WL, increased air temperature, and moderate WS. Although its biomass was low, another potential bloom-forming and toxin-producing genus, Aphanizomenon, was present in low and high flow years. Aphanizomenon was correlated with decreased inflow rate and increased particulate carbon to particulate phosphorus ratios, which may be related to improved light conditions and to their ability to cope with P limitation. \nIn conclusion, these results highlight that LD is vulnerable to an increase in potential toxic cyanobacteria species during lower water levels. It is expected that climate variability will lead to more frequent extreme events (i.e., drought) that will affect reservoirs around the globe, including those in the Canadian Prairies, due to climate warming. Moreover, water abstraction from reservoirs to meet the needs of the growing human population poses an additional challenge. Such reduction in water level due to drought and the associated increase in water abstraction in the SSR basin will cause further deterioration of water quality (algal blooms) in LD.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.