British Columbia Mine Reclamation Symposium

Benthic invertebrate community recovery in a remediated stream Batchelar, Katharina; Stecko, Pierre; Hughes, Colleen 2019

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


59367-Batchelar_K_et_al_Benthic_invertebrate.pdf [ 1.94MB ]
JSON: 59367-1.0391927.json
JSON-LD: 59367-1.0391927-ld.json
RDF/XML (Pretty): 59367-1.0391927-rdf.xml
RDF/JSON: 59367-1.0391927-rdf.json
Turtle: 59367-1.0391927-turtle.txt
N-Triples: 59367-1.0391927-rdf-ntriples.txt
Original Record: 59367-1.0391927-source.json
Full Text

Full Text

BENTHIC INVERTEBRATE COMMUNITY RECOVERY IN A REMEDIATED STREAM   Katharina Batchelar, M.Sc., R.P.Bio1 Pierre Stecko, M.Sc, EP, R.P.Bio1 Colleen Hughes, EP2   1 Minnow Environmental Inc. 204-1006 Fort Street Victoria, B.C.  V8V 3K4  2 Mount Polley Mining Corporation PO Box 12 Likely, B.C.  V0L 1N0    ABSTRACT  A foundational failure of the perimeter embankment of the Mount Polley Mine Tailings Storage Facility (TSF) on August 4th 2014 resulted in a breach that released approximately 25 million cubic meters of debris (water and solids that consisted of tailings, construction materials, and scoured sediment and soil).  The debris flowed along the entire length of Hazeltine Creek, scouring native materials and depositing TSF material in the creek channel.  A first phase of channel remediation in Hazeltine Creek was completed in 2015, and a second phase which incorporated natural geomorphological and fish habitat values was completed in 2016 and 2017.  Benthic invertebrate community monitoring was completed most recently in these remediated areas in 2017 and 2018.  This monitoring documented rapid recolonization on the basis of organism densities that were similar to pre-breach in some areas, including colonization by sensitive EPT taxa.  A number of differences relative to pre-breach remained, indicating that the community was still recovering from physical disturbances associated with the remediation work (including the influences of ongoing work on downstream areas).  These included lower organism density in some areas, lower taxon richness, and a community that was dominated by early successional or tolerant taxon groups.  KEY WORDS  Recolonization, stream remediation, breach   INTRODUCTION  Failure of a glacio-lacustrine unit underlying the perimeter embankment of the Mount Polley Tailings Storage Facility (TSF) resulted in a breach on August 4th 2014.  The TSF breach released approximately 25 million cubic meters of debris (water and solids that consisted of tailings, construction materials, and scoured sediment and soil), which flowed along the length of Hazeltine Creek (Figure 1).  The Mount Polley   Mining Corporation (MPMC) initiated channel reconstruction along the length of Hazeltine Creek in 2014, and implemented the remediation work in two phases (Bronsro et al., 2016).  The first phase involved the removal of tailings and soft material, and construction of an erosion resistant foundational channel with primary meander patterns.  This work was completed for the entirety of Hazeltine Creek in May 2015 (Bronsro et al., 2016).  The second phase involved the construction of natural geomorphological and fish habitat values (secondary and tertiary meander patterns, pools, riffles, instream and overstream cover, and riparian planting; Ogilvie at al., 2018).  This was completed in two areas of upper Hazeltine Creek in September of 2016 (Reach 1) and 2017 (Reach 2; Figure 2) and is currently ongoing in Hazeltine Creek downstream of Reach 2.  This second phase of remediation in upper Hazeltine Creek was in progress prior to and during the completion of benthic invertebrate community monitoring in lower Hazeltine Creek in 2017 (which was downstream of the remediation works).  Benthic invertebrate community monitoring is one component of the ongoing assessment of Hazeltine Creek recovery, and was initiated in 2015 following the first phase of channel remediation works.  Monitoring continued in lower creek areas in 2017 and in upper creek areas in 2018; areas which were remediated most recently in 2015 and 2016/2017, respectively (Figure 2).  In order to evaluate recovery of the benthic invertebrate community relative to pre-breach conditions (2007; Weech and Stecko 2009), remediated monitoring locations were selected to be within the vicinity of pre-breach locations (Figure 2).  Due to differences in sample collection techniques in 2015 compared to 2017 and 2018, this paper is focused only on benthic invertebrate community monitoring completed in 2017 and 2018.  The fundamental objectives of this work were to characterize the recovery of the benthic invertebrate community in these remediated areas in order to support the evaluation of potential effects to aquatic ecosystem health (e.g., potential effects to higher trophic levels such as fish and fish habitat quality).  METHODS Sampling        Benthic invertebrates were sampled at one area of lower Hazeltine Creek in 2017 (HAC-D) and at two areas of upper Hazeltine Creek in 2018 (HAC-R1 and HAC-U; Figure 2) using sampling methods consistent with technical guidance (BCMOE 2013).  Sampling methods were also consistent with those used prior to the breach (UHC and LHC; 2007) to allow comparability to pre-breach conditions (Weech and Stecko 2009).  Samples were collected using a 30.5 x 30.5 cm (1’ x 1’ square; 0.093 m2) Surber sampler equipped with 250 µm mesh.  Five stations were sampled in each area and were established at a minimum distance of approximately three bankfull widths from adjacent stations.  Sampling locations were within the vicinity of pre-breach monitoring locations, and were selected to ensure that water velocity and substrate characteristics were comparable among sub-samples, stations, and sampling areas to minimize natural influences on community variability.    At each Surber sampling station, one composite sample of three sub-samples (i.e., approximately 0.279 m2 total creek surface area) was collected.  Each sub-sample was collected by carefully inserting the base of the Surber sampler into undisturbed substrate.  Cobble and gravel contained within the sampler was carefully washed, allowing the current to carry dislodged organisms into the collection net.  The samplerwas then moved to the next sub-sampling location and the procedure repeated.  Following collection of the third sub-sample using the procedure outlined above, material and organisms retained in the collection net were carefully transferred to internally and externally labelled 1 or 2 L wide mouth plastic jars using a wash bottle while working over a plastic tub to avoid any potential loss of organisms.  Each jar received an internal and external label with the station number, area identifier, project number, date, and the initials of the field personnel.  Each sample was preserved with buffered formalin solution to achieve a nominal concentration of 10%, and stored at room temperature until shipment to the laboratory.  Supporting information collected at each sampling station included GPS (Global Positioning System) coordinates, sampling depth, substrate size (20 randomly selected substrate pieces measured per station, using methods adapted from Environment Canada 2012a), water velocity, station photographs, and notes of presence of aquatic vegetation and other physical observations that would aid interpretation.  Field meter measurements of temperature, specific conductance, dissolved oxygen and pH (using a YSI EXO™ handheld portable field meter equipped with YSI EXO2™ Sonde) were also recorded at each station.    Analytical Methods  Taxonomic identification of the benthic invertebrate community samples was completed by Cordillera Consulting in Summerland BC, using standard methods that incorporate QA/QC measures (e.g., Environment Canada 2012b; 2014).  Upon receipt, each sample was processed as follows: Rose Bengal dye was added, the sample was washed, remaining organic material was removed, and the sample was preserved in 70% ethanol.  Samples were sieved and processed in the laboratory using 250 µm mesh.   The number of invertebrates in each sample was estimated, and if greater than 600, a subsample was taken using a Marchant Box (Marchant 1989) as described by Environment Canada (2014).  Samples (or sub-samples) were then counted to a minimum of 300 and sorted into family/order using a low power stereo microscope.  If the 300 organisms had not been reached by the 50th cell of the Marchant Box, the entire sample was sorted.  Sorted organisms were then identified to the lowest practicable level (LPL; typically genus or species) by qualified taxonomists using standard keys, by comparing to an externally-verified reference collection, and using effort lists compiled by CABIN (Canadian Aquatic Biomonitoring Network), SAFIT (Southwest Association of Freshwater Invertebrate Taxonomists), and PNAMP (Pacific Northwest Aquatic Monitoring Partnership).  Laboratory QA/QC included an assessment of sub-sampling error, sorting efficiency, and taxonomic quality control on at least 10% of the samples (Environment Canada 2012a, 2014).   Data Interpretation  Benthic invertebrate communities were evaluated using metrics of total organism density (organisms per m²), taxonomic richness, Simpson’s Index of Diversity, Simpson’s Index of Evenness (calculated as in Smith and Wilson 1996; Environment Canada 2012b), and densities and relative densities (proportions) of dominant/indicator taxa.  Dominant/indicator taxon groups were defined as those groups representing more than 5% of total organism abundance within each set of areas being compared, or any taxon groups considered to be important indicators of environmental stress.  Further analysis of benthic community structure using the Bray Curtis Index of Dissimilarity or multivariate analyses (e.g., Correspondence Analysis) were not employed because differences in benthic invertebrate community structure relative to pre-breach were identifiable without use of these techniques.    Data were evaluated using a before-after study design.  Specifically, results from upper and lower areas of Hazeltine Creek sampled in 2017 and 2018 were compared to similarly located areas sampled prior to the breach (i.e., HAC-R1 and HAC-U vs. UHC; HAC-D vs. LHC).  A before-after-control-impact (BACI) study design was not used to evaluate benthic community recovery because recovery was expected to be in early stages due to the recent disturbance.  As the community becomes more similar to pre-breach conditions, a BACI design will be an important tool for data interpretation.    Comparison to pre-breach conditions represents a key contrast for evaluating benthic community recovery in these remediated areas.  However, challenges exist in comparing recent results to these historic data.  This is due to the use of different taxonomic laboratories for the analyses, and the large time span (>10 years) over which general improvements in laboratory protocols may have occurred.  One major difference was that aquatic mites and spring-tails were identified to a more specific taxonomic level by the laboratory used in 2017/2018 than by the laboratory used in 2007.  To maintain comparability with pre-breach data from 2007, aquatic mite and spring-tail data from 2017/2018 were collapsed to a less-specific taxonomic level (Class) prior to calculation of benthic community endpoints.  Benthic community metrics were calculated for each sampling area and year at the LPL of taxonomy following exclusion of Porifera, Nemata, Platyhelminthes, Ostracoda, Copepoda, Cladocera, and any terrestrial drop-ins (taxa excluded from laboratory counts).  Metrics were then summarized by area and year by calculating mean, median, minimum, maximum, and standard deviation.  Benthic community metrics for upper Hazeltine Creek in 2018 (HAC-U and HAC-R1; remediated) were statistically contrasted against pre-breach results (UHC; 2007) using analysis of variance (ANOVA).  For these analyses, assumptions of normality and homogeneity of variance were tested (α = 0.05), and data were transformed to satisfy assumptions if necessary.  The transformation with the highest resulting p-value for normality (Shapiro Wilk’s test) was selected for analysis as long as the assumption of homogeneity of variances was also met.  For density endpoints, logarithmic transformations were preferred (when they met the model assumptions) over no transformation or other transformations.  When the model assumptions could not be met, the Kruskal-Wallis test was used.  If the overall results of ANOVA or the Kruskal-Wallis test were significant, pairwise comparisons among the three groups were evaluated using Tukey’s honestly significant differences method for ANOVA or Dunn’s test (with a Bonferroni adjustment to the significance level) for the Kruskal-Wallis test.  Benthic community metrics for lower Hazeltine Creek in 2017 (HAC-D; remediated) were statistically contrasted against pre-breach results (LHC; 2007) by two sample t-tests.  Data were transformed prior to analysis to meet test assumptions as described for ANOVA.  When the assumption of normality was not met, a Mann-Whitney test was used.  When the assumption of homogeneity of variances was not met, a two-sample t-test for unequal variances was used.   Statistical contrasts were conducted using R (R Core Team, 2017), and verified using Minitab® v17 (Minitab 2016).  An effect on the benthic invertebrate community among areas was defined as statistically significant using an α of 0.1 (Environment Canada 2012a).  Interpretation of benthic invertebrate community metrics was enhanced by inspection of raw data and taxonomic proportions to detect patterns of ecologically relevant differences between the remediated and pre-breach areas.    RESULTS AND DISCUSSION  The benthic invertebrate community of lower Hazeltine Creek differed substantially in 2017 from pre-breach conditions in terms of lower mean organism density (4,534 organisms/m2 in 2017, 32,599 organisms/m2 in 2007) and lower mean richness (33 taxa in 2017, 41 taxa in 2007, Figure 3, Table 1).  These differences were not unexpected given the large physical disturbance to this area due to the breach, ongoing upstream remediation efforts (causing disturbance downstream), and the limited time frame (approximately two and a half years) since remediation of this area.  .  Despite this, Simpson’s Indices of Evenness and Diversity were similar to pre-breach conditions (Figure 3).  Differences in benthic invertebrate community composition relative to pre-breach were evident, with higher proportions of Trichoptera (caddisfly larvae), Diptera (non-Chironomidae winged insect larvae), and Oligochaetes (aquatic worms; primarily the Tubificid worms Enchytraeus and Nais), and lower proportions of Ephemeroptera (mayfly larvae) and Arachnida (aquatic mites) in 2017 (Figure 4, Table 1).  Despite these clear differences in taxonomic composition following remediation, the proportion of EPT taxa (mayflies, stoneflies and caddisflies) which are considered to be sensitive (e.g., Lenat 1988, Herman and Nejadhashemi 2015) was similar in 2017 compared to pre-breach.  This similarity was driven by higher proportions of Plecoptera and Trichoptera in 2017 than pre-breach (Table 1), and was indicative of ongoing community recovery in this remediated area.  The benthic invertebrate community of upper Hazeltine Creek in 2018 showed rapid recovery of organism density at HAC-U to pre-breach conditions only 1 year after the second remediation phase was completed (39,326 organisms/m2 in 2018, 39,228 organisms/m2 in 2007).  At HAC-R1, total organism density remained significantly lower than pre-breach two years after remediation (19,747 organisms/m2 in 2018), and was lower than at HAC-U (Figure 5, Table 2).  This difference between the remediated areas was primarily attributable to a higher prevalence of Chironomidae, Diptera (excluding Chironomidae), and Oligochaeta at the more recently remediated area (remediated in 2017; HAC-U; Figure 2).  The community of both remediated areas clearly differed from pre-breach conditions in terms of significantly lower mean taxon richness (20 and 19 taxa at HAC-R1 and HAC-U in 2018, respectively; 42 taxa in 2007; Table 2), and lower Simpson’s Diversity (Figure 5, Table 2).  Differences in community composition between the two upper Hazeltine Creek remediated areas and pre-breach conditions were clearly indicative of a community structure in early recovery.  Specifically, the communities at both remediated areas were more dominated by Chironomidae (Figure 6, Table 2), which are considered to be tolerant of environmental stress (Mandaville 2002) and are generally expected to increase in relative proportion with increasing environmental disturbance (Barbour 1999).  The combined proportion of sensitive EPT taxa was significantly lower at both remediated areas than pre-breach, driven by lower proportions of Ephemeroptera and Plecoptera (Figure 6, Table 2).  Combined, these differences relative to pre-breach conditions are indicative of physical disturbance, which is consistent with the limited timeframe (1 and 2 years [full seasonal cycles] at HAC-U and HAC-R1, respectively) since conclusion of the second phase of remediation in these areas.       SUMMARY AND CONCLUSION  The Mount Polley TSF breach and subsequent channel remediation works in Hazeltine Creek had a direct physical impact on the benthic invertebrate community.  Following initial remediation efforts, the benthic invertebrate community in lower Hazeltine Creek differed from pre-breach conditions on the basis of substantially lower organism density, lower taxon richness, and community composition.  This is consistent with the large disturbance in this area approximately 2 and a half years prior to monitoring, and upstream remediation efforts (which caused disturbance at this downstream area) both prior to and during monitoring of this area.  The community in this remediated area had higher proportions of tolerant taxa groups (Diptera and Oligochaetes) than pre-breach conditions, but a similar proportion of sensitive EPT taxa.  This similarity to pre-breach of sensitive taxa prevalence is evidence of recovery, and suggests no chemical barrier to recolonization of these sensitive taxa.  Following the second phase of remediation works, the benthic community in upper Hazeltine Creek showed rapid recovery of organism density to pre-breach conditions at the more recently remediated monitoring area (1 year [including a full seasonal cycle] following remediation).  Community composition at both upper creek areas was indicative of a community in early recovery, as evidenced by lower taxon richness, lower Simpson’s Diversity, higher dominance of tolerant taxon (Chironomidae), and lower relative proportion of sensitive EPT taxa than pre-breach conditions.  These differences in community composition are not unexpected given the recent large disturbance to these upper creek areas one to two years prior to monitoring.  Overall, monitoring indicated rapid recovery of the benthic invertebrate community in areas representing both the first and second phases of remediation, with production relating to fish habitat quality (organism density) equal to pre-breach in one area.  The community composition in lower Hazeltine Creek has improved more than the upper creek areas, which is reflective of a longer period of physical stability in the lower creek areas where the second phase of remediation has not been completed.  Despite this difference, taxa that represent important food for fish were present at pre-breach proportions at all monitoring areas.  REFERENCES  Barbour, M.T., J. Gerritsen, B.D. Snyder, and J.B. Stribling. 1999. Rapid Bioassessment Protocols for Use in Streams and Wadeable Rivers: Periphyton, Benthic Macroinvertebrates, Fish, Second Edition,  EPA 841-B-99-002. United States Environmental Protection Agency, Office of Water, Washington, DC. 337 pp. BCMOE (British Columbia Ministry of Environment). 2013. British Columbia Field Sampling Manual For Continuous Monitoring and the Collection of Air, Air-Emission, Water, Wastewater, Soil, Sediment, and Biological Samples, 2013 Edition. BCMOE. 363 pp. Bronsro, A., J. Ogilvie, L. Nikl., and M. Adams. 2016. River Rehabilitation Following a Tailings Dam Embankment Breach and Debris Flow. Proceedings Tailings and Mine Waste. Keystone, Colorado, USA, October 2-5. Environment Canada. 2012a. CABIN (Canadian Aquatic Biomonitoring Network) Field Manual: Wadeable Streams. Environment Canada, 57 pp. Environment Canada. 2012b. Metal Mining Guidance Document for Aquatic Environmental Effects Monitoring, Report EEM/2002/1. Environment Canada. 550 pp. Environment Canada. 2014. CABIN (Canadian Aquatic Biomonitoring Network) Laboratory Methods: Processing, Taxonomy, and Quality Control of Benthic Macroinvertebrate Samples. Environment Canada. 36 pp. Herman, M.R. and A.P. Nejadhashemi. 2015. A Review of Macroinvertebrate and Fish-based Stream Health Indices. Ecohydrology & Hydrobiology. 15: 53–67 Lenat, D.R. 1988. Water Quality Assessment of Streams Using a Qualitative Collection Method for Benthic Macroinvertebrates. Journal of the North American Benthological Society. 7: 222-233. Mandaville, S.M. 2002. Benthic Macroinvertebrates in Freshwaters – Taxa Tolerance Values, Metrics, and Protocols, Project H-1. Soil & Water Conservation Society of Metro Halifax. 128 pp. Marchant, R. 1989. A Subsampler for Samples of Benthic Invertebrates, Bulletin of the Australian Society for Limnology. 12: 49-52.   Minitab 17 Statistical Software. 2016. State College, PA: Minitab, Inc. ( Ogilvie, J. and D. Carter. 2017a. Hazeltine Creek: Reach 1 Record Drawings and Habitat Rating.  Prepared for Mount Polley Mining Corporation by Golder Associates Ltd., 23 March 2017. 18 pp. Ogilvie, J. and D. Carter. 2017b, Hazeltine Creek Reach 2 Record Drawings and Habitat Rating, Prepared for Mount Polley Mining Corporation by Golder Associates Ltd., 18 December 2017. 21 pp. Ogilvie, J., H. Topps, and L. Nikl. 2018. An innovative approach to monitoring the physical stability of constructed fish habitat using drones. British Columbia Technical and Research Committee on Reclamation Symposium. Williams Lake, BC. R Core Team. 2017. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. Smith, B. and J.B. Wilson. 1996. A Consumer’s Guide to Evenness Indices, Oikos. 76: 70-82. Weech, S. and P. Stecko. 2009. Mount Polley Mine Aquatic Environmental Characterization – 2007. Prepared for Mount Polley Mining Corporation by Minnow Environmental Inc.  October 2009. 312 pp.    


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items