Development, Assessment and Application of Benthic Algal Biomonitoring Protocols for Canadian Waters
Thomas, Kathryn Elizabeth
MetadataShow full item record
Stressors such as residential and industrial development and climate warming are escalating in North America, which increases stress to aquatic ecosystems. In the face of this, monitoring biologists must continually improve protocols for long-term monitoring programs in order to adequately characterize changes in biological communities. To address this need, this thesis has developed, applied, and assessed benthic algal biomonitoring protocols in lakes and rivers. In the Muskoka-Haliburton area of Ontario, benthic algal protocols were developed to assess effects of differences in shoreline development. In the South Nahanni River watershed, Northwest Territories, benthic algal biomonitoring protocols were developed to assess effects of two mining companies on rivers in an otherwise pristine ecosystem. In the Muskoka-Haliburton area I developed and evaluated bioassessment protocols based on benthic algae growing in the littoral zone of lakes to track effects of shoreline development. To do this, I sampled a suite of study sites (n = 28 in 2006, n = 29 in 2007) spanning a gradient of shoreline development (e.g., intact forests, cottages, marinas). The protocols were modified from protocols developed for rivers (Biggs and Kilroy, 2000), and five levels of assessment were completed for each site that differed in the amount of time, resources and expertise required. Level 1 comprised visual assessments of benthic algal cover. Level 2 involved biomass estimates (ash-free dry mass and chlorophyll-a). Level 3 included coarse-level taxonomic enumeration of benthic algal community composition (i.e., to major algal classes). Level 4 included quantification of pigment concentrations using High-Performance Liquid Chromatography (HPLC). Level 5 involved high-taxonomic resolution enumeration of diatom community composition (to species and sub-species levels). Uni- and multivariate analyses were used to assess relations between shoreline development, water chemistry and benthic algal metrics. Results of this study showed that Level 5 (diatom community composition) best discriminated among shoreline development categories and, despite the higher technical skill and time required, was recommended for use as the most promising metric for Precambrian Shield lake nearshore biomonitoring with benthic algae. Photosynthetic pigment concentration (Level 4) showed modest potential as a biomonitoring tool, but further development is required for their use in monitoring protocols. The South Nahanni River watershed is remote with good water quality. However, activities conducted by two mining operations within the watershed potentially threaten the water quality and ecological integrity of downstream sites. Here, I conducted three studies. The first study examined physical and chemical conditions at river sites unaffected by human activities and how the conditions related to three algal metrics (benthic algal community composition, diatom community composition, and photosynthetic pigment concentration). To do this, I sampled 44 reference sites (i.e., unaltered by human activities such as mining or other infrastructure) from across the South Nahanni River watershed in 2008 and 18 sites in 2009 (12 repeated from 2008, 6 new). Multivariate analyses were utilized to assess patterns of variation in physical and chemical data and their relation to benthic algal community composition. Results showed that physical and chemical conditions differed distinctly between two ecoregions within the Nahanni Watershed (Selwyn Mountain and Nahanni-Hyland ecoregions). Patterns of variation in the benthic algal metrics corresponded well with gradients of physical and chemical variables. Diatom community composition discriminated best between the two ecoregions. Photosynthetic pigment concentration only discriminated between the ecoregions in 2009, showing some promise as a biological monitoring tool. The second study examined the extent that algal pigment versus taxonomic descriptors of algal community structure varied due to the Cantung mine along the Flat River in the South Nahanni River watershed in order to evaluate the use of photosynthetic pigment concentration as a biomonitoring approach. To do this, I sampled 4 sites upstream and 6 sites adjacent to and downstream of the Cantung mine site and compared relations of water physico-chemical conditions with photosynthetic pigment concentration and taxonomic-based benthic algal community composition at the study sites. Patterns evident in ordinations by PCA and RDA identified that photosynthetic pigment concentrations varied along Flat River and were related to variance in physical and chemical variables. My analyses showed that there were substantial and often statistically significant differences in photosynthetic pigment concentration at non-exposed sites located upstream of the mine versus exposed sites located adjacent to and downstream of the mine. Photosynthetic pigment concentrations were more strongly and consistently associated with physical and chemical conditions than the taxonomy-based data, suggesting pigment analysis is effective for detecting environmental degradation. Additionally, cost comparisons showed that the base analytical cost for in-house analysis of pigment was low ($66.48/sample) and generally lower than traditional taxonomy-based assessments, making it a cost-effective alternative for biomonitoring protocols. In the third study, I developed Reference Condition Approach (RCA) models based on benthic algae for the South Nahanni River watershed. To do this, I sampled a suite of reference sites across the watershed in 2008 (n = 44) and 2009 (n = 18; 12 resampled from 2008 and n = 6 new) and test sites (potentially affected) downstream of two mining companies (n=20 in 2008 and n = 17 in 2009). The BEAST (BEnthic Assessment of SedimenT) model was used to develop the benthic algal RCA models for each of the three benthic algal metrics. All reference sites (unaffected by mining activities) from 2008 and 2009 were grouped into biologically similar assemblages. Only physical and chemical variables unaffected by mining activities were used in developing the RCA model. The biological assemblages at test sites were compared to their predicted reference assemblage using non-metric Multimetric Dimensional Scaling Analysis (MDS) and assessed for impairment. Three probability ellipses were used to create four categories of impairment: Category 1: ≤ 90% (reference condition), Category 2: 90 – 99% (possibly stressed), Category 3: 99 – 99.9% (stressed), Category 4: ≥ 99.9% (severely stressed). Patterns of downstream impairment were assessed and zones of influence were identified for each algal metric in each year. Assessments downstream showed that the RCA models identified reasonably consistent ‘zones’ of stress downstream of Cantung mine along Flat River. However, changes in photosynthetic pigment concentrations were more prominent compared to the other two metrics. Along Prairie Creek, only photosynthetic pigment concentrations identified sites outside of the reference condition directly downstream of the Prairie Creek mine. My results show that benthic algal RCA models (specifically photosynthetic pigment concentration models) show promise as biological monitoring tools, but should be tested in other ecosystems to assess the widespread utility of the method. I developed, applied and assessed benthic algal community compositions for Canadian lakes and rivers. I assessed a variety of algal metrics in different ecosystems and associated with differing stressors, and found that photosynthetic pigments were the most sensitive metric to differences in physical and chemical conditions downstream of the two mines. Conversely, diatoms were the most responsive metric to differences between ecoregions, and similarly to differences in shoreline development categories in Muskoka lakes. Photosynthetic pigment concentrations can be influenced differently by stressors such as light compared to other metrics or biological traits. Indeed, I found that differences in pigment concentrations were often associated with differences in turbidity and thus, light may play an important role in pigment concentration in biological assessments. Despite this, pigments and the RCA approach show promise as a biomonitoring tool for detection of impairment, and should be further tested and refined based on studies in other watersheds.