UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A comparative assessment of stormwater runoff from a coastal and interior log yard Fikart, Alena 2002

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

Item Metadata

Download

Media
831-ubc_2002-0399.pdf [ 6.32MB ]
Metadata
JSON: 831-1.0090410.json
JSON-LD: 831-1.0090410-ld.json
RDF/XML (Pretty): 831-1.0090410-rdf.xml
RDF/JSON: 831-1.0090410-rdf.json
Turtle: 831-1.0090410-turtle.txt
N-Triples: 831-1.0090410-rdf-ntriples.txt
Original Record: 831-1.0090410-source.json
Full Text
831-1.0090410-fulltext.txt
Citation
831-1.0090410.ris

Full Text

A C O M P A R A T I V E ASSESSMENT O F S T O R M W A T E R RUNOFF F R O M A C O A S T A L AND INTERIOR L O G Y A R D  by Alena Fikart B.Sc, Simon Fraser University, 1995 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES Resource Management and Environmental Studies  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA September 2002  © Alena Fikart 2002  In  presenting  degree freely  at  this  the  available  copying  of  department publication  of  in  partial  fulfilment  University  of  British  Columbia,  for  this or  thesis  reference  thesis by  this  for  his thesis  and  study.  scholarly  or  her  for  of I  financial  gain  shall  that  agree  may  representatives.  requirements  agree  I further  purposes  the  It not  that  be  the  Library  by  understood be  an  allowed  advanced  shall  permission  granted  is  for  for  the  without  it  extensive  head  that  make  of  my  copying  or  my  written  permission.  Department  of  few,re  T h e U n i v e r s i t y o f British Vancouver, Canada  Date  DE-6  (2/88)  / 7 f ^ f i ^ ^ Columbia  faj  ^r?u//^7/7i^/  d  /  J T ^ y ^  Abstract Stormwater runoff from log yards in different BC regions can affect aquatic habitats to varying degrees given differences in weather, water quality and tree species. The objective of this thesis was to compare runoff quality and total runoff loadings from a coastal and interior log yard. Chemical analyses, toxicity tests and treatments were conducted. Data were compared to criteria, statistically compared between sites and assessed for seasonal trends. Export coefficients (ECs) were compared between sites. Relationships between toxicological and chemical variables were explored statistically. Runoff toxicity was similar between sites and fairly low. LC50s for 48-hour  Ceriodaphnia  dubia  tests ranged from 32.95 to > 100 and 58.70 to > 100 for coastal and interior runoff, respectively. Microtox ®.5 minute EC50s ranged from 27.12 to > 100 for coastal runoff and 22.22 to > 100 for interior runoff. Several metals and dehydroabietic acid (DHA) exceeded criteria in runoff from both sites. Biochemical oxygen demand, alkalinity, pH and metals were significantly higher (p < 0.05) at the interior site. Sodium and conductivity were higher at the coastal site. No seasonal differences in runoff quality were observed. Therefore, acute effects would occur during periods of high runoff, during autumn at the coastal site and late winter at the interior site. The unpaved interior site generated less runoff per square meter due to ground infiltration. ECs were comparable to the paved coastal site. Exceptions to this include tannins and lignin (11 fold higher at the coastal site) and DHA (9 fold higher at the interior site). C. dubia toxicity was partially associated with TSS for both sites. Tannins and lignins were correlated (r =0.91) with 2  C.  dubia  toxicity for coastal runoff. Tannin and lignin concentrations ranged from 45 to 263 mg/L and 43 to 75 mg/L in coastal and interior samples, respectively. Since results suggest that TSS is partially responsible for toxicity, and since contaminants are often bound to TSS, source control and treatment options for TSS should be implemented. ii  Table of Contents Abstract  ii  Table of Contents  Hi  List of Tables  vii  List of Figures  x  Acknowledgements  1  xi  Introduction and Literature Review  1  1.1  Introduction  1  1.2  Literature Review  2  1.2.1 Forest Harvest and Operations in BC  2  1.2.2 Wood Chemistry and Leachate Toxicity  3  1.2.3  Characteristics of Log Yard Runoff.  11  1.2.4 Management Policies and Guidelines for Log Storage Sites  17  1.2.5  20  1.3  2  Stormwater Runoff Treatment Study Objectives  21  Materials and Methods  2.1  24  Site Descriptions  24  2.1.1  Interior Site  24  2.1.2  Coastal Site  26 iii  2.2  Field Sampling  29  2.2.1  Sample Location  29  2.2.2  Sample Collection  29  2.2.3  Flow Rate Measurem ents  31  2.2.4  Field Measurements  33  Laboratory Analyses  33  2.3.1  Chemical Analyses  33  2.3.2  Toxicity Testing  42  2.3  2.4  Data Analyses  49  2.4.1  Evaluation of Log Yard Stormwater Runoff  49  2.4.2  Statistical Comparisons Between Sites  49  2.4.3  Seasonal Trends  50  2.4.4  Calculation of Loadings for Coastal and Interior Sites  51  2.4.5  Relationships between Toxicological and Chemical Variables  57  3  Results  3.1  61  Evaluation of Log Yard Stormwater Runoff.  61  3.1.1  Runoff Chemistry  61  3.1.2  Runoff Toxicity  75  3.2  Statistical Comparisons Between Sites  81  3.3  Seasonal Trends  81  3.4  Calculation of Loadings for Coastal and Interior Sites  85  3.4.1  Site Precipitation  85 iv  3.4.2  Data Evaluation  86  3.4.3  Loadings  87  3.5  Relationships between Toxicological and Chemical Variables  3.5.1  Spearman's Rank Correlation Analyses  3.5.2  Linear Regression  3.5.3  Principle Components and Regression Analyses  4  89 89 •  90 94  Discussion  99  4.1  Comparison of Water Quality between Sites  99  4.2  Relationships between Microtox® and Ceriodaphnia  4.3  Comparisons to Other Log Yard Runoff Studies  102  4.4  Relationships between Toxicological and Chemical Variables  102  dubia toxicity  101  4.4.1  Sample Treatments  102  4.4.2  Regression Analyses  106  4.4.3  Criteria  108  4.4.4  Weight of Evidence  108  4.5  Factors affecting Log Yard Stormwater Toxicity  109  4.6  Effects from Log Yard Stormwater  112  4.6.1  Acute Effects  112  4.6.2  Contaminant Loadings  113  4.7  Conclusions and Recommendations  115  v  5  References  118  Appendix A: Data Tables  127  Appendix B: Water Quality Trend Figures  131  Appendix C: Spearman's Rank Correlations  136  vi  List of Tables Table 1.1: Relative toxicity of leachate from six different tree species (adapted from Moore 1992)  7  Table 1.2: Toxic effects of sitka spruce and western hemlock bark extracts on pink salmon fry, larval and adult pink shrimp, and larval dungeness crabs in seawater (modified from Buchanan et al. 1976)  8  Table 1.3: Toxic effects of western hemlock, sitka spruce, western redcedar and yellowcedar wood extracts to pink salmon fry in fresh and saltwater (Pease 1974)  9  Table 1.4: Summary of chemical and toxicological characteristics of runoff from log handling areas  13  Table 1.5: Summary of compounds implicated in toxicity of wood and woodwaste leachate  14  Table 1.6: Removal effectiveness (%) of solids and trace metals from stormwater runoff by filtration/infiltration techniques (modified from Macdonald et al. 1997)  21  Table 2.1: Year 2001 log inventory for the coastal and interior site  25  Table 2.2: List of metals analyzed and associated ICP-OES detection limits  42  Table 2.3: Volume of EDTA added for C. dubia sample treatments  47  Table 3.1: Summary of collected stomwater runoff samples from a coastal and interior log yard  62  Table 3.2: Summary of general water quality characteristics from coastal and interior log yard stormwater runoff  63  Table 3.3: Iron and manganese interference in tannin and lignin analytical method  67  Table 3.4: Summary of metal concentrations in coastal and interior log yard stormwater runoff.  69  Table 3.5: Summary of metal criteria (or guideline) exceedances in coastal and interior log yard stormwater runoff  72  Table 3.6: Summary of C. dubia and Microtox toxicity data from coastal and interior log yard stormwater runoff  76  Table 3.7: Treatment effects (TSS removal and EDTA chelation) on log yard stormwater runoff toxicity to C. dubia  77  Table 3.8: Results of TSS additions from log yard stormwater runoff (Sample N-07/19) to control water and centrifuge control results for C. dubia toxicity tests vii  78  Table 3.9: Summary of total to soluble COD, tannins and lignins and DHA in log yard stormwater runoff.  79  Table 3.10: Summary of contaminant concentrations from wet and dry weather coastal log yard stormwater runoff (n=4 for dry weather and n=6 for wet weather)  83  Table 3.11: Simple linear coefficients of determination for total sample flow vs. C. dubia toxicity from combined and coastal log yard stormwater runoff  85  Table 3.12: Significant simple linear coefficients of determination (p < 0.05) between total flow and select water quality parameters from combined and coastal log yard stormwater runoff.  85  Table 3.13: Summary of 2001 precipitation for the coastal and interior log yard sites (data from Environment Canada Atmospheric Environment Service)  86  Table 3.14: Summary of Sample Mean Concentrations (SMC) for contaminants in coastal and interior log yard stormwater runoff  87  Table 3.15: Summary of total and seasonal runoff volumes of log yard stormwater from the coastal and interior sites  88  Table 3.16: Summary of select stormwater runoff export coefficients for contaminants from the coastal and interior log yard sites  89  Table 3.17: Significant simple linear coefficients of determination (p <0.05 and r > 0.60) for contaminants vs. C. dubia toxicity in log yard stormwater runoff  92  Table 3.18: Summary of significant (p < 0.05) pH and hardness additions to coastal and combined data  94  Table 3.19: Rotated principal component (PC) loadings between PCs and C. dubia sample toxicity for combined and coastal log yard stormwater runoff  95  Table 3.20: Simple linear coefficients of determination for principal component scores vs. C. dubia toxicity for combined and coastal log yard stormwater runoff  96  Table 4.1: Comparison of coastal and interior log yard stormwater runoff data to runoff data from log handling areas found for other studies  103  Table 4.2: Summary table of significant linear coefficients of determination ( p < 0.05 and r > 0.80) for contaminants vs. C. dubia toxicity in log yard stormwater runoff. 2  107  Table 4.3: Spearman's rank correlation coefficients between coastal log yard stormwater runoff contaminants highly correlated (r > than 0.80) with C. dubia toxicity 2  108  Table 4.4:Comparison of export coefficients from urban runoff studies to log yard runoff export coefficients (kg/year/ha)  114 Vlll  Table A . l : General water quality characteristics in interior log yard stormwater runoff  127  Table A.2: Metals concentrations in interior log yard stormwater runoff  128  Table A.3: General water quality characteristics in coastal log yard runoff  129  Table A.4: Metals concentrations in coastal log yard stormwater runoff  130  Table C . l : Spearman's rank correlations for the combined log yard stormwater runoff contaminants and flow data  136  Table C.2: Spearman's rank correlations for the coastal log yard stormwater runoff contaminants and flow data  139  ix  List of Figures Figure 2.1: Schematic of the interior log yard  26  Figure 2.2: Schematic of the coastal log yard  28  Figure 3.1: Comparison of mean general water quality characteristics from coastal and interior log yard stormwater runoff (error bars are standard deviations)  64  Figure 3.2: Comparison of mean metal concentrations from coastal and interior log yard stormwater runoff (error bars are standard deviations)  71  Figure 3.3: Freshwater criteria comparisons for copper, manganese, nickel and zinc for coastal and interior log yard stormwater runoff (circles are coastal samples, triangles are interior samples)  73  Figure 3.4: Comparison of mean general ion concentrations from coastal and interior log yard stormwater runoff (error bars are standard deviations)  74  Figure 3.5: Presence of outliers in the correlation between toxicity and DHA dominated Principal Component 2 (Comp. 2) of the combined PCA. Toxicity and DHA concentrations decrease along the y and x axes  97  Figure 3.6: Presence of outliers in the correlation between toxicity and DHA dominated (negative relationship) Principal Component 2 (Comp. 2) of the coastal PCA. Toxicity and DHA concentrations decrease along the y and x axes  98  Figure B . l : Trends for general water quality characteristics in interior and coastal log yard stormwater runoff (circles are interior runoff samples, squares are coastal runoff samples)  131  Figure B.2: Trends for metal concentrations in interior and coastal log yard stormwater runoff (circles are interior runoff samples, squares are coastal runoff samples)  133  Figure B.3: Trends for general ion concentrations in interior and coastal log yard stormwater runoff (circles are interior runoff samples, squares are coastal runoff samples)  134  Figure B.4: Trends for C. dubia and Microtox® toxicity in interior and coastal log yard stormwater runoff (circles are interior runoff samples, squares are coastal runoff samples)  135  x  Acknowledgements / would like to thank my supervisory committee: Ken Hall, Sheldon Duff and Les Lavkulich for their continued support throughout this process. Funding for this work was provided by the BC Hydro Water and Wastewater Centre. I would also like to thank EVS Environment Consultants (EVS) for allowing me to use their toxicology laboratory, specifically Howard Bailey (Senior Ecotoxicologist)  and James Elphick  (laboratory manager) for their support, guidance and general interest. Thanks are also due to the sawmill managers who allowed access to their sites for field sampling and responded to requests for information, Coast River Environmental Services Ltd. for provision of sampling equipment and to Valerie Lemay of the UBC Department of Forestry for her valuable statistical help.  Special thanks go also to Julie Beer, Lea Elliot and Nick Page for their valuable contributions to my general sanity and this thesis during discussions over breakfast. Thanks also to the following graduate students in the UBC pulp and paper laboratory for their help in the technical aspects of laboratory analysis and their contribution to the good humour of the laboratory: Steve Helle, Preston Hoy, Norman Woo, Christine Woodhouse and Mike Zenaitis.  Last, but not least, I'd like to thank my husband, Edward Gregr, whose support (financial, emotional and technical) was invaluable and whose kindness and encouragement continued throughout.  xi  1  Introduction and Literature Review  1.1 Introduction Forestry is one of British Columbia's (BC) major industries, with exports in 2000 accounting for approximately half of the total provincial goods exported to international markets (BC Ministry of Forests 2001). In 2000, the total harvest in BC was 77.4 million cubic meters (volume of all products billed), the majority of which were softwoods (coniferous trees). Recently, the storage and processing operations of the forest industry have come under scrutiny by environmental agencies such as Environment and the Department of Fisheries and Oceans (DFO) due to the potential for release of toxic leachate from woodwaste residues and logs. Once trees are harvested, they are often processed in large volumes in log yards or dryland sorts either separate from, or adjacent to processing operations such as sawmills or shingle mills. When logs are stored in open areas, naturally occurring toxic chemicals found in the wood and bark of logs may dissolve or leach due to continued exposure to precipitation (Samis et al. 1999). In addition to leachate generated from the exposed logs, stormwater runoff from log yards can include metals from equipment or galvanized roofing (Good 1993, Bailey et al. 1999) as well as oil and grease from on-site machinery (National Council of the Paper Industry for Air and Stream Improvement Inc. (NCASI) 1992). As log yard areas are either paved or highly compacted due to the use of heavy machinery on site, log yard runoff is likely to flow off the site during precipitation events (rainfall or snowmelt) and enter nearby aquatic systems. Depending on a number of factors such as runoff toxicity and volume, and characteristics of the aquatic system such as dilution capacity, pH and hardness, the runoff may be detrimental to aquatic organisms such as fish and invertebrates. Since the effect of log yard runoff on aquatic organisms is subject to a number of environmental  1  factors, log yards located in different regions of BC may affect aquatic habitats to a greater or lesser extent. Log yards located in coastal regions are subject to more rainfall (and thus more leachate) than log yards situated in interior regions of the province. In some areas, spring snowmelt of a heavy snowpack can cause a flush of log yard runoff to enter nearby aquatic systems. Additionally, log yards in different regions of the province often process logs from different tree species that may affect the quality and toxicity of the log leachate generated (Moore 1992, Buchanan et al. 1976, Pease 1974).  1.2 Literature Review 1.2.1  Forest Harvest and Operations in BC  Approximately 2.0% of the province's total log harvest is exported as raw logs, with the majority of exports being processed wood products. Lumber and pulp and paper products account for a large majority of forest products exported, at 49 and 35%, respectively, with the remaining products including plywood, shingles and shakes, poles, chips, logs and millwork (Council of Forest Industries (COFI) 2001). In 2000, the total harvest in BC was 77.4 million cubic meters (volume of all products billed), the majority of which were softwoods (coniferous trees). Lodgepole pine (Pinus contorta) was the most dominant species, accounting for 28% of the total harvest. Picea sp. (spruce) was the next most harvested species, accounting for approximately 16%. Hemlock (Tsuga sp.) and Douglas-fir {Pseudotsuga menziessi) each accounted for 14% of the total BC harvest (BC Ministry of Forests 2001). The majority (approximately two thirds) of softwood production occurs in interior BC, with the remaining one third of log production occurring on the coast. In 2000, the main softwood species harvested in coastal BC in order of volume included hemlock (Tsuga sp.), Douglas-fir, western  2  redcedar (Thuja plicata) and true firs (Abies sp.). The main softwood species harvested in the interior of BC in order of volume include lodgepole pine, spruce (Picea sp.), true firs and Douglas-fir. Other harvested softwood species in BC include the yellow-cedar or cypress (Chamaecyparis nootkatensis), larch (Larix sp.), white pine (Pinus monticola), whitebark pine (Pinus albicaulus), ponderosa pine (Pinus ponderosa) and yew (Taxus brevifolia), but these  account for only 4% of the total harvest (BC Ministry of Forests 2001). The majority of hardwood harvesting occurs in the interior and includes aspen (Populas tremuloides), Cottonwood (Populas balsamifera), birch (Betula papyrifera), alder (Alnus rubra)  and maple (Acer macrophyllum). BC hardwood species (deciduous trees) make up only 3-4% of the total harvest (BC Ministry of Forests 2001).  1.2.2  Wood Chemistry and Leachate Toxicity  The chemical composition of wood is known to vary among tree species due to the presence of natural compounds that are often unique to each species. Additionally, the chemical composition varies by age, the part of the tree and its geographic location. Although many of the same constituents are found in different parts of the tree (i.e., bark and wood) the proportions are quite different. For example, wood extractives such as resins and condensed tannins, as well as lignins are found in considerably higher concentrations in the bark. Additionally, an aliphatic acid called suberin is only present in the tree bark. The presence of these compounds makes bark more resistant than wood to decay or insect attack (Samis et al. 1999). As a result, the leachate generated from different types and parts of wood may differ in toxicity and likely in the toxic agents.  1.2.2.1 Chemical Composition  In general, compounds present in the wood and bark of all tree species include carbohydrates,  3  lignins and wood extractives. Wood extractives are soluble in neutral organic solvents or water and can be divided into three groups, namely phenolic compounds, terpenoids and aliphatic compounds (Sjdstrom 1993, Samis et al. 1999, McDougall 1996).  Carbohydrates Carbohydrates of softwood and hardwood species include polysaccharides (cellulose, hemicellulose) and monosaccharides (simple water soluble wood sugars). All carbohydrates are considered to be biodegradable (Samis et al. 1999) because they can be broken down into shorter fragments and wood sugars, however, cellulose is more resistant to degradation. Cellulose molecules (made of polysaccharides) bundled together into cellulose fibers create the wood skeleton. They make up approximately 40-45% of the dry weight of softwood and are insoluble in most solvents. Hemicellulose surrounds the cellulose skeleton and acts as a matrix of supporting material in the cell wall. In softwoods, approximately 20% of the wood is hemicellulose (McDougall 1996, Sjdstrom 1993).  Lignins Lignins are phenolic substances composed of polymers of phenylpropane units. Like hemicellulose, lignin surrounds cellulose and bonds wood cells together in a rigid structure, is relatively resistant to biodegradation, and is not readily solubilized (Samis et al. 1999, Sjostrom 1993). Softwoods generally have a higher lignin content (26-32% dry wt) than hardwoods (2025%) dry wt) (McDougall 1996). The highest lignin content occurs in the sapwood of the western redcedar and hemlock species (Samis et al. 1999).  Wood Extractives Extractives are a broad category of substances and include phenolics, terpenoids and aliphatic acids. Phenolic extractives are mainly located in the heartwood and bark and provide trees with 4  microbiocidal and insecticidal properties to protect them from decay (Samis et al. 1999, McDougall 1996). These include tannins (hydrolyzable and condensed), lignans, stilbenes, flavenoids and tropolones. Tannins are mainly found in the bark, foliage and roots of softwoods, and in lesser amounts in heartwood. Hydrolyzable tannins are readily hydrolyzed by acids to phenol carboxylic acids and simple wood sugars or polyhydroxy alcohols. Condensed tannins undergo progressive polymerization in the presence of heat and acid to yield mainly water insoluble complexes. Tannins found in the heartwood only occur in the condensed form. Lignans have fungicidal, insecticidal and anti-oxidant properties. Tropolones are acidic, water soluble compounds found only in the heartwood of the western redcedar and species of the cypress family. The decay resistance of the heartwood of these species is due mainly to their presence (Samis etal. 1999). Terpenoids are water insoluble compounds that exist primarily in pine, spruce, larch and Douglas-fir and are largely absent from hardwoods. They provide a defense against wood boring insects (McDougall 1996). There are more terpenoids in pine species than in other species. Up to 5% of softwood (dry weight) consists of terpenoids, which can be subdivided into volatile and non-volatile fractions. Non-volatile terpenoids consist mainly of resin acids (diterpenoids) found in resin canals of trees (Samis et al. 1999). After a tree is harvested, resin acids can become more water soluble due to oxidation (McDougall 1996). Volatile terpenoids include monoterpenoids and other volatile oils. These are found in foliage and resin canals of all softwoods and account for the characteristic odour of fresh wood (Samis et al. 1999). Between 1 and 5% of the dry weight of wood consists of aliphatic acids. Aliphatic acids (or fatty acids) are mainly water insoluble, long-chained saturated and unsaturated fatty acids that are a source of stored energy for trees (Samis et al. 1999). Fatty acids are mostly concentrated in seed tissues (e.g. cones and fruit) but they are also found in the wood resins of all softwood and 5  hardwood species (Samis et al. 1999, McDougall 1996).  Other Chemical Constituents Wood also contains many other constituents present in smaller quantities. These include aliphatic alcohols and sterols, such as juvabione and dehydrojuvabione that are formed from the esterification of fatty acids (Samis et al. 1999). 1.2.2.2 Leachate Toxicity  Wood leachate can be generated upon exposure of wood to water through precipitation, snowmelt and flooding. The toxicity of the resulting leachate depends largely on the tree species. It is, however, difficult to assess the relative toxicity of leachate generated by different tree species, as toxicity is also dependant on the test organism and the method of leachate generation. Several studies have attempted to rank tree species in order of leachate toxicity to different organisms. Four studies summarized below examine the relative toxicity of leachate generated from trees commonly harvested on the west coast of BC. Moore (1992) conducted a study that looked at the relative toxicity of six different tree species commonly found on the west coast: western redcedar; yellow-cedar; sitka spruce; Douglas-fir; western hemlock; and black cottonwood. Tests included three different organisms and were as follows: (1) luminescent bacteria Microtox®; (2) 96-hr LC50 rainbow trout; and (3) 48-hr LC50 Daphnia  magna.  Leachate from the different  species was generated by soaking woodwaste in deionized water for 28 days. As expected, each species exhibited different toxic responses to different leachate types (Table 1.1). Black cottonwood leachate was the most toxic to both Microtox® and rainbow trout; Douglas-fir leachate was the most toxic to Daphnia  magna.  Sitka spruce leachate was the least toxic to all  three organisms. Sitka spruce and western redcedar leachate were found to have the highest  6  concentration of resin acids, however, there was no correlation between toxicity and total resin acid concentrations. Table 1.1: Relative toxicity of leachate from six different tree species (adapted from Moore 1992). Microtox® Toxicity Most Toxic  Least T o x i c  Rainbow Trout  Daphnia magna  5 minute EC50  15 m i n u t e E C 5 0  96-hr LC50  48-hr LC50  black c o t t o n w o o d  black c o t t o n w o o d  black c o t t o n w o o d  Douglas-fir  western hemlock  Douglas-fir  yellow-cedar  yellow-cedar  redcedar  western hemlock  Douglas-fir  black c o t t o n w o o d  Douglas-fir  redcedar  western hemlock  western hemlock  yellow-cedar  yellow-cedar  redcedar  redcedar  sitka s p r u c e  sitka s p r u c e  sitka s p r u c e  sitka s p r u c e  Buchanan et al. (1976) studied the toxic effects of sitka spruce and western hemlock bark extracts and particulates on pink salmon fry {Oncorhynchus gorbuscha), larval and adult pink shrimp (Pandalus borealis) and larval dungeness crabs {Cancer magister). Cessation of swimming and death were the criteria of toxic effect. Bark samples were ground with a mortar and pestle and homogenized with seawater. Subsequently, a portion of the homogenized solution was filtered and test solutions - both filtered and unfiltered - were prepared by serial dilution with filtered seawater. Toxicity of log bark extracts (fdtered samples) was tested for all four organisms. Toxicity of extracts plus particulates (unfiltered samples) was tested for larval and adult pink shrimp as well as larval dungeness crab. Sitka spruce extracts were found to be more toxic than hemlock extracts on all invertebrates tested. However, hemlock extracts were more toxic to pink salmon fry. In general, salmon fry were the most sensitive organisms to the extract solutions (Table 1.2). Note that only results from the bark extracts (fdtered samples) are reported, as complete data for the extracts plus particulates (unfiltered samples) were not presented by Buchanan et al. (1976). However, fdtered 7  spruce extracts were less toxic than unfiltered spruce extracts. Pease (1974) conducted a study to determine the toxicity of western hemlock, sitka spruce, western redcedar and yellow-cedar log leachates to pink salmon (Oncorhynchus gorbuscha) fry. Freeze-dried extracts were prepared from cross sections of a log cut into thin wood shavings and soaked in distilled water for four days. The leachate was filtered to remove solids and freeze dried. The freeze-dried extracts were subsequently re-dissolved at various test concentrations in fresh or saltwater for toxicity tests. Results indicate that leachate was more toxic in freshwater tests than those conducted in saltwater. Sitka spruce and western redcedar extracts were most toxic to pink salmon fry in freshwater, while yellow-cedar was most toxic in saltwater (Table 1.3). Table 1.2: Toxic effects of sitka spruce and western hemlock bark extracts on pink salmon fry, larval and adult pink shrimp, and larval dungeness crabs in seawater (modified from Buchanan et al. 1976). C o n c e n t r a t i o n of bark e x t r a c t s ( m g / L ) Test Organism  < 1 ) < 2 )  Test  Western Hemlock  Sitka Spruce  Pink s a l m o n f r y  96-h L C 5 0  56  100-120  C r a b larvae  96-h L C 5 0  >1000  530  S h r i m p larvae  96-h L C 5 0  >1000  415  S h r i m p adult  96-h L C 5 0  >1000  205  C o n c e n t r a t i o n s w e r e e s t i m a t e d using total bark w e i g h t u s e d t o c r e a t e e x t r a c t s . T e s t serial dilutions w e r e p r e p a r e d with filtered bark s a m p l e s h o m o g e n i z e d with s e a w a t e r (bark extracts).  8  Table 1.3: Toxic effects of western hemlock, sitka spruce, western redcedar and wood extracts to pink salmon fry in fresh and saltwater (Pease 1974).  yellow-cedar  C o n c e n t r a t i o n of bark e x t r a c t s ( m g / L ) (D(2) Test F r e s h w a t e r (96-hr threshold t o x i c i t y )  <3)  S a l t w a t e r (96-hr t h r e s h o l d toxicity) ( 1  '  ( 2 >  ( 3 )  Sitka S p r u c e  Western redcedar  Yellow-cedar  Western hemlock  25-40  35-45  >50  75-90  >200  >200  150-200  >200  C o n c e n t r a t i o n s w e r e e s t i m a t e d by m e a s u r i n g c o n c e n t r a t i o n s of total h y d r o x y l a t e d a r o m a t i c c o m p o u n d s in l e a c h a t e T e s t serial dilutions w e r e p r e p a r e d with f r e e z e - d r i e d e x t r a c t s p r e p a r e d f r o m s o a k i n g w o o d s h a v i n g s in distilled w a t e r a n d s u b s e q u e n t l y filtered. 9 6 - h r t h r e s h o l d toxicity v a l u e s w e r e c a l c u l a t e d f r o m l_T50s a n d a r e r e f e r r e d t o a s incipient L C 5 0 s .  Schermer and Phipps (1976) report the results of 96-hr LC50 rainbow trout tests conducted with Douglas-fir, western redcedar and western hemlock extracts. Douglas-fir and hemlock extracts were leached from a mixture of approximately 50/50 bark and wood chips (hogfuel), western redcedar was derived from shredded wood of mostly bark or sapwood. The woodwaste was placed in plastic containers (static "generators" with drains for sampling) and covered with water. To determine the relative rates of leaching under these static conditions, and the time required for a leachate to approach a maximum concentration, generators with the three types of woodwaste were monitored for 247 hours. After two hours, chemical oxygen demand (COD) and tannin concentrations in western redcedar were 75 to 85% of the corresponding values at 227 hours. Leaching rates for Douglas-fir and hemlock were much slower, although Douglas-fir leachate ultimately reached a higher COD (but not tannin) concentration. The maximum concentration leachate for each species was subsequently used for fish toxicity tests. Results indicated that the order of toxicity was fir (LC50s: 5.5-30.6%), cedar (LC50s: 20 - 45.6%) and hemlock (LC50s: 19 - 100%). LC50 values for cedar were thought to be low due the lack of heartwood, and thus tropolones, present in the sample extracts. Other studies have examined the relative toxicity of leachate from tree species commonly  9  harvested in the interior of BC or other parts of Canada (or foreign species of the same genus found in Canada). Borga et al. (1996) separated timber with respect to species and sprinkled with freshwater (either oligotrophic [nutrient poor] or eutrophic [nutrient rich] water) for 18 weeks. They found that Norway spruce (Picea abies) wastewater was more toxic than that of Scotch pine (Pinus sylvestris L.) as measured by Microtox® (5 and 15 minute tests), particularly when sprinkled with oligotrophic water. Although Microtox® toxicity was not thought to be a function of resin acid concentration, resin acid loads were larger from spruce wastewater thanfrompine. Unlike the results obtained by Borga et al. (1996), results from a study conducted by the Alberta Forest Products Association (AFPA 1999a) indicate that pine leachate is more toxic than spruce leachate. The AFPA conducted a study characterizing surface runoff from log yard sites in Alberta. Their results indicate that sites that process a large proportion of pine (Pinus sp.) can generate log yard runoff that is toxic to aquatic organisms (96-hr rainbow trout, Microtox®). They also found that runoff from sites containing predominately spruce (Picea sp.) species did not generate toxic responses in aquatic organisms. 96-h rainbow trout LC50s for sites storing some proportion of pine (40% or less of total log volume) ranged from 32 to 92%, total runoff, while LC50s for sites storing predominantly spruce (85% or more) were >100%. O'Connor et al. (1992) conducted a study that included representatives of tree species commonly employed in Canada from coastal (western hemlock) and interior (white pine, balsam fir [Abies balsamea],  black spruce and aspen) areas. They simulated a series of mechanical pulping  effluents that were analyzed for wood extractives and tested for acute and chronic toxicity to Ceriodaphnia C. affinis  affinis  and acute toxicity to the fathead minnow (Pimephales promelas). For both  and the fathead minnow, the acute toxicity of the simulated mechanical pulping  effluents followed the order of white pine > balsam fir > western hemlock > black spruce > aspen. 10  Taylor et al. (1996) conducted a study on the toxicity of aspen wood leachate, derived by leaching fresh aspen wood chips for 35 days. Solids were removed from the extract by filtration. The median lethal concentration to rainbow trout (96-hr LC50) was 1-1.8% of undiluted leachate, while toxicity was marginally less to Daphnia magna (48-hr acute) around 1.7-3.4%. Leachate was also strongly inhibitory to algal growth and Microtox® bacteria. In general, it is difficult to assess the relative toxicity of different wood species due to different methodologies in obtaining extracts (i.e., different portions of the tree used) and different organisms used in testing.  1.2.3 1.2.3.1  Characteristics of Log Yard Runoff Chemistry and  Toxicity  There have been relatively few studies conducted on the characteristics of runoff from log yards (NCASI 1992, Taylor 1994, AFPA 1999a, Bailey et al. 1999 Zenaitis and Duff 2002). Log yard stormwater runoff is generally composed of leachate from log piles and wood waste residue, as well as contaminants introduced by machinery, equipment and roofing. Log yard runoff is often characterized by high chemical oxygen demand (COD), biochemical oxygen demand (BOD), total suspended solids (TSS) and total organic carbon (TOC), as well as contaminants such as metals and oil and grease. Often runoff has a low pH (3 to 6.5) due to the degradation of organic compounds to organic acids and carbon dioxide, which reacts with water to produce carbonic acids (Samis et al. 1999). Agitation of liquid containing wood leachate can result in the formation of foam consisting of constituents such as resin and fatty acids and lignin. Surface slicks on stagnant pools of wood leachate are associated with terpenes (Samis et al. 1999, NCASI 1992). Naturally-occurring wood compounds often react with metals. For example, the chelation of tannins with dissolved iron and other metals causes the formation of complexes (Thomas 1977). The chelation of tropolones with dissolved iron causes the formation of red11  coloured complexes. Many surface leachate streams (such as log yard runoff streams), tend to hold most of their colour in the sediments, while the liquid portion is fairly clear. This can happen as a result of precipitation and settlement of coloured complexes (Sweet and Fetrow 1975, Samis et al. 1999). Several studies have documented the chemical and toxicological characteristics of log yard or woodwaste runoff (Table 1.4). The former BC Ministry of Environment, Lands and Parks (now the BC Ministry of Water, Air and Land) conducted a study in 1994 (Taylor 1994) on the leachate generation of an aspen (Populus  tremuloides)  logpile in Northern BC. Leachate from the  aspen logpile was characterized by dark colour, acidic pH, elevated conductivity, high BOD, total organic carbon and low dissolved oxygen. The greatest acute toxicity from the aspen woodpile was found to occur in the spring and after light and moderate showers. Heavy rains tended to remove more soluble material in total, however, the leachate was more dilute and therefore less toxic. Other studies summarized in Table 1.4 include: a program characterizing the surface water runoff from Albertan log yards processing various proportions of spruce, pine and aspen (AFPA 1999a); a literature review and case study of log yard runoff by NCASI (1992); an investigation of the causes of toxicity in stormwater runoff from sawmills (Bailey et al. 1999); and the characterization of log yard runoff for ozone treatment trials (Zenaitis and Duff 2002). McDougall (1996) summarized the results of stormwater assessment conducted by Weyerhaeuser Canada Ltd. (Weyerhaeuser) at four industrial log handling and sawmill sites between 1989 and the early 1990s. McDougall's personal communication with Weyerhaeuser indicated that the stormwater quality varied from toxic to non-toxic, and could vary with the season. Based on the monitoring data, it was found that water quality parameters such as metals, resin and fatty acids and other organics, TSS, and BOD can exceed water quality objectives.  12  o CD Q  1^  8"  CNI  O  _l  co o  CO  i  CT)  LO  CT) o CT) .c  LO  CD  CO <  T3  o CD i CD  o  00  CN LO  c 0 O  oo  CT> o CT) o CO CT) C O CT) 1^-  CNI  < 0.  O V  UO CN  o o o v  CN V  CO  O  CM I  4 CO  CD  o LO LO CN I  CN  o  CO  CO  CO  00  o  o o  o  00  CO  O  o o  V  05  o  LO  CO CO  o co 00 LO co • CO  1^  06 ^ .  o o  o  CM  00  O O  /\  I  CM  LO CN  1^ O  (35  CT) CT)  00  I  o  0  O O  LO I CO  cb  CD  CQ  CO  CN O O CM  o  o o  CN  00  O CD  CD  LO  CT) oo  CN  T3  o co I  CN CO  CT)o O  00  cb o  1^  LO  CNI  CN  CO CM  c 0 N  CD E,  CD CD Q.  CO  O LO  O LU  CO  =J <D  CO  •g o < c CO CD  CD CO  !Q CD O i_  D3  E O  c  CD  .c CD Q  _  c Q O  CL  CO  o  ZJ C  O O o  X CL  CO CO  o  O  o _l I JZ 1  c  co  LO  ©  C O C O E E o  0  -C  ^  o LO  rout  0 0 E  < X Q •g o <  CT)  o LO O _i i_  .c 00 CD  agn  CD  I— S X •2 o ^ o "5 c o= Q. CD — < o 'CD "5 P cc 2 Q  1.2.3.2  Potential  Toxicants in Log yard  Runoff  Due to the limited information available specifically on toxicants in stormwater runoff from log yards, the following summary of potential contaminants of concern includes information from a variety of studies of wood and woodwaste runoff toxicity (Table 1.5). Table 1.5: Summary of compounds implicated in toxicity of wood and woodwaste leachate. Type of Wood Pine {Pinus sp.), spruce {Picea true firs {Abies sp.)  sp.),  Contaminants Implicated in Toxicity  Source  Resin and Fatty Acids  BC Research 1975 BC Research 1977 Leach and Thakore 1976  Western redcedar {Thuja  Balsam fir {Abies  balsamia)  Western hemlock {Tsuga Aspen {Populus  plicata)  Heartwood Tropolones  Peters ef al. 1976  Foliage (volatile) Terpenes  Peters ef al. 1976  Lignans  Peters ef al. 1976  Dehydrojuvabione  O'Connor etal.  1992  Lignans  O'Connor etal.  1992  heterophylla)  tremuloides)  Total Phenols  Not specified  Tannins  Not specified  Tropolone and Hinokitiol  Not specified  Tannins and Lignins  Not specified  Zinc  AFPA 1999 Field etal.  1988  Inamori ef al. 1991 Bailey etal.  1999  Bailey ef al. 1999  Volatile Terpenes Peters et al. (1976) found that foliage terpenes of the western redcedar were toxic to coho salmon fry, having a 96-h LC50 of 2.7 mg/L.  Resin Acids (non-volatile terpenes) and Fatty Acids BC Research found that resin and fatty acids accounted for 90% of the toxicity in hydraulic debarker effluent generated by spruce, pine and fir species (BC Research 1977) and 60 to 90% of the toxicity to rainbow trout from mechanical pulping effluents from spruce and lodgepole pine (BC Research 1975). Leach and Thakore (1976) linked 7 resin acids (dehydroabietic, palustric, abietic, isopimaric, pimaric/sandaracopimaric and neoabietic) to 60-90% of overall acute toxicity 14  of mechanical pulping effluents generated from spruce, pine and fir species. However, Peng and Roberts (2000) found that of the most common resin acids, dehydroabietic acid (DHA) was the most soluble but least toxic to Daphnia  magna.  Total Phenols AFPA (1999a) noted that toxic responses of rainbow trout (96-hr LC50) and Microtox® marine bacteria were observed when total phenol concentrations reached 0.5 mg/L. They also noted that the generation of phenols appeared to be associated with sites that store aspen tremuloides)  in their log yards rather than pine or spruce. However, Taylor et  (Populus al.  (1996) found  that although aspen leachate contains elevated concentrations of phenolic compounds, experimental data suggest that phenols were not the chief toxicant.  Tannins and Lignins Bailey et al. (1999) found that tannin and lignin concentrations correlated strongly with the toxicity of stormwater runoff from log yard areas to rainbow trout. They also found that a pH of 8.5 ameliorated the toxic effects of tannins and lignins, particularly at lower concentrations. Field  et al. (1988) found that much of hydraulic debarker effluent toxicity was attributable to tannins. Tropolones Peters et al. (1976) found that tropolones were primarily responsible for the toxicity of western redcedar leachate to fish (coho salmon fry, steelhead trout and prickly sculpins) but that toxicity was eliminated with chelation by iron. Inamori et al. (1991) found that 50 mg/L of tropolone and hinokitiol (beta-thujaplicin) extracts showed strong inhibitory activity on the growth (root length) of eight food plants examined.  Dehydrojuvabione O' Connor et al. (1992) found that the nonpolar organic extractive dehydrojuvabione was likely 15  primarily responsible for the chronic toxicity of simulated mechanical pulping effluent from balsam fir (Abies balsamed) to Ceriodaphnia affinis.  Others Bailey et al. (1999) found that zinc concentrations correlated significantly with sawmill runoff toxicity to rainbow trout. Peters et al. (1976) found lignan extractives from western redcedar were mildly toxic to coho salmon fry and alevins, and less toxic to mayfly nymphs. Lignans were also implicated as potential toxicants in simulated mechanical pulping effluent from western hemlock to Ceriodaphnia affinis (O'Connor et al. 1992).  1.2.3.3 Environmental Effects of Log Yard Runoff  More studies have been conducted on the effects of log storage in rivers, lakes, saltwater bays and estuaries (Hall et al. 1988, Pease 1974, Sedell and Duval 1985, Conlan and Ellis 1979) than on effects from land-based storage operations. Hatfield Consultants Ltd. (1996) wrote a fisheries sustainability review of the Strait of Georgia, which includes a review of log handling impacts. Information from this review relevant to runoff generated from land-based storage facilities is included below, along with information from other sources. Physical and chemical impacts are associated with log storage and handling. Many of the physical impacts are due to the accumulation of wood and bark debris (Sedell and Duvall 1985) in aquatic receiving environments. Habitat alteration can occur as a result of wood debris deposition on bottom sediments. Sedell and Duvall (1985) noted the formation of hydrogen sulphide in anaerobic sediments smothered in bark debris. Suspension feeding organisms, which rely on access to the water column, are negatively affected by wood residue (Pease 1974, Conlan and Ellis 1979). In marine environments, thick deposits create much reduced benthic communities with a few sulphide tolerant organisms such as bacteria, protozoa and nematodes  16  dominating in the sulphide rich sediments (Jackson 1986). Wood residue deposited in streams, along lake margins or in marine foreshore areas may clog gravel and destroy areas important for fish production (e.g. spawning areas). Deposition of wood residue and leachate can also cause fungal and bacterial growth, which can affect fish and fish habitat (Samis et al. 1999). Impacts of wood debris may not always be negative. Stanhope and Levings (1985) concluded that intertidal estuarine wood litter supported sufficient amphipod prey populations for rearing salmonids. Additionally, the toxicity of leachates in seawater is thought to be negligible due to precipitation in combination with chloride ions (Sedell and Duvall 1985).  1.2.4  Management Policies and Guidelines for Log Storage Sites  The Alberta Environmental Protection Agency (AEP, McDougall 1996) conducted a review of policies and guidelines adopted for log storage facilities in other jurisdictions. A summary of this review is presented below, along with information obtained from other sources. In 1990, the United Stated Environmental Protection Agency (USEPA) enacted stormwater regulations requiring all industrial facilities to apply for and obtain National Pollutant Discharge Elimination System (NPDES) permits for any stormwater flowing directly into surface waters or into municipal separate sewer systems. Stormwater discharges from log storage areas are included under these regulations. Priority pollutants from log yard stormwater identified by the USEPA include bark and wood debris, TSS and wood extractives. In addition to these regulations, the EPA requires industrial facilities to develop storm water pollution prevention as a means to ensure pollutants are controlled as required by the Clean Water Act. As part of the pollution prevention plans, Best Management Practices (BMPs) must be developed by each facility. Certain states are designated as "permitting authority states", which means they can issue their 17  own permits. For example, the Washington State Department of Ecology (WDOE) has issued permits that are more stringent than the USEPA requirements. The WDOE requires that log yards eventually comply with groundwater quality and sediment management standards. In May 1995, the Washington State Department of Ecology published Best Management Practices to Prevent Stormwater Pollution at Log Yard (WDOE 1995). Source control BMPs recommended by (WDOE 1995) specifically for high activity log yard areas include: •  Paving the area if soil is not contaminated;  •  Sloping all high activity paved and rock areas to prevent erosion and minimize the formation of leachate; and  •  Provide slopes sufficient to convey contaminated stormwater to appropriate pollution control systems.  Paving the area is recommended to facilitate cleaning of debris, to reduce generation of solid waste and to increase efficiency of log yard operations (WDOE 1995). Canadian regulations differ between provinces. For example, the New Brunswick Department of Environment has no specific regulations or guidelines for log yards in New Brunswick. Ontario has required storm water dischargers, including pulp and paper and saw mills that have log storage areas, to conduct storm water control studies and prepare a report once every three years. In Alberta, industrial stormwater management is assessed on a site-specific basis, with limits established for certain sectors under the Alberta  Environmental  Protection  and  Enhancement  Act. Historically, stormwater runoff was not considered high risk to the environment. However, largely in response to concerns regarding toxic leachate from aspen log woodpiles in BC, the AEP published a document pertaining to the assessment of log yard runoff in Alberta (McDougall 1996). The intent of the report was to provide recommendations to minimize the 18  impact of log storage sites on the environment. In response to these concerns, AFP A initiated a research program in 1996 to investigate the effects and characteristics of log yard runoff (AFP A 1999a). They developed a decision tree to aid regulators and operators in evaluating risk from specific log yards (AFPA 1999b) In BC, regulating the storage, use and disposal of waste is mainly within provincial jurisdiction. The provincial government has the authority to develop legislation concerning land use and pollution control. The BC Waste Management Act provides the authority for lead agencies to require permits and approval for waste management. However, federal government agencies such as DFO and EC are responsible for the administration and enforcement of the federal Fisheries Act,  which includes pollution prevention and habitat protection provisions (Liu et al.  1996). Under federal legislation (Fisheries Act, Subsection 35(1)), it is illegal to "carry on any work or undertaking that results in the harmful alteration, disruption or destruction of fish habitat" (Liu et al. 1996). Currently in BC, a Waste Management Branch (WMB) permit (now the Environmental Protection Division of the BC Ministry of Water, Air and Land), issued under the provincial Waste Management Act, is not required for the discharge of stormwater from log yards. However, as part of an effort to address fisheries related impacts from wood leachate to the aquatic environment in the Pacific and Yukon region, Environment Canada and DFO have jointly published guidelines on storage, use and disposal of wood residue (Liu et al. 1996). A companion document to these guidelines has also been published by DFO (Samis et al. 1999) which summarizes sources and toxicity of leachate generated from wood residues. The document also provides general guidelines for the reduction of leachate generation and runoff control. Despite these general guidelines for wood waste storage, no operational guidelines or BMPs have been established specifically for log storage facilities in BC. Some practices, such as continuous cleaning of log storage sites, are conducted by some log yards as a risk management 19  measure. This helps reduce the amount of debris and fine particles potentially harmful to fish and fish habitat in the log yard runoff  1.2.5  Stormwater Runoff Treatment  Settling basins or other filtration/infiltration type systems are recommended to remove suspended solids as well as other contaminants from urban runoff (WDOE 1995, ASCE 1992). Although a detailed discussion of stormwater treatment is beyond the scope of this thesis, some of these systems may be appropriate for treatment of log yard stormwater runoff. Filtration systems include concrete, earthen or lined basin, tank or similar containment facilities that provide stormwater detention time to allow for settling and removal of floating and suspended solids. Limitations of filtration systems include: (1) episodic high flow events may need attenuation; and (2) soluble and emulsifiable compounds (such as resin and fatty acids) may not be removed. Additionally, filtration systems can become easily clogged with fine particulates (WDOE 1995). Infiltration systems, where the majority of run-off is percolated through the ground include ponds, trenches, vaults, porous pavements and concrete grids (WDOE 1995, Gibb et al. 1999). This option must be carefully considered with respect to impact on ground water quality. Other important considerations are the type of soil used in the system; very fine soils reduce infiltrative capacity, while very coarse soils often have low removal of dissolved pollutants. On the positive side, infiltration systems are applicable to small drainage areas (WDOE 1995). Biofiltration systems make use of grasses or other types of vegetation to filter, adsorb and uptake stormwater soluble and insoluble pollutants (WDOE 1995). This category includes the constructed wetland. Contaminant removal mechanisms in wetlands include gravity settling of particulates, filtration of solids by root mats and soils, adsorption to soil particles, chemical 20  transformation and uptake and conversion to less harmful forms by plants and bacteria. Constructed wetland variants for stormwater management include: (1) the shallow wetland; (2) the wet pond followed by a shallow wetland; (3) the shallow marsh with an extended detention live storage; and (4) the pocket wetland where the water pool intersects the groundwater table. Wetlands are practical for new developments, and suitable for small on-site facilities and large regional facilities (Gibb et al. 1999). Table 1.6 (modified from Macdonald et al. 1997) summarizes the removal effectiveness of solids and trace metals from stormwater runoff by various filtration/infiltration type systems. Table 1.6: Removal effectiveness (%) of solids and trace metals from stormwater runoff by filtration/infdtration  techniques (modifiedfrom  Macdonald  et al. 1997).  Percent Removal Treatment System  Solids  Wet Detention Basin  60-100  Organic Carbon  Lead  57-71  9-95  24  50-70  75-90  77  Grassed Swale  80  45  1.3 Study  40  (Schueler and Helfrich 1989)  30-60  (Stahre and Urbonas 1990)  95,55  96,88  (Yousef et al. 1986)  57  73  51  (Schueler 1997)  60  80  60  (Horner 1988)  0-20  (Schueler 1987)  20-100 78  (Athayde etal. 1983)  77, 50  0-40  Wetlands  5-71  (Schueler 1987)  73-74  Wet Pond  Reference (Schueler 1987)  60-100  Multiple Wet Pond  Vegetated Filters  Zinc  60-80  5-91 Dry Detention Basin  Copper  (Gibb etal. 1991) 28  39  63  54  (Schueler 1997)  Objectives  The impacts of stormwater runoff on the receiving environment have been classified as acute and cumulative (Harremoes 1988). Acute effects are short-term effects typically resulting from a 21  single event with a duration measured in hours, such as bacteriological pollution or dissolved oxygen (DO) depletion. Cumulative effects can be characterized by a gradual build-up of pollutant in the receiving environment. In the case of log yard stormwater, the evaluation of pollution discharged with runoff can be based on immediate environmental effects (contaminant concentrations, toxicity) or the effects of accumulated contaminants (contaminant loadings) in the environment over time. The objective of this thesis was to conduct a comparative assessment of stormwater runoff from a coastal and interior log yard with both a short term (acute) and long term (contaminant loadings) perspective by: •  Evaluating and comparing the potential for acute environmental effects from log yard stormwater runoff (chemistry and toxicity) between the two sites;  •  Evaluating and comparing the contaminant loadings from log yard stormwater runoff between the two sites; and  •  Identifying potential toxic constituents of the stormwater runoff for each site.  Thesis objectives were based on the assumption that regional differences in precipitation and the type of trees processed on-site are relevant to the concentration, volume and toxicity of log yard stormwater runoff generated; and that these differences are relevant to the management of runoff from log yards in different regions of BC. The two study sites were chosen such that: (1) sites were located in different biogeoclimatic zones with different precipitation regimes; (2) different wood species were processed on site; (3) no anti-microbial treatment was applied to logs or lumber on the property; and (4) practical considerations. Both of the study sites chosen were sawmills with attached log yards. The 22  interior site is situated in the northern interior portion of BC in the Sub-Boreal Spruce (SBS) biogeoclimatic zone, the lumber processed at the site reflects the dominant tree species in this biogeoclimatic zone and includes white spruce, subalpine fir and lodgepole pine. The coastal site is located in the southern portion of the Coastal Western Hemlock (CWH) biogeoclimatic zone, lumber processed is largely western hemlock and true firs. The CWH zone is considered to be the wettest zone in BC with approximately 15% of the total precipitation occurring as snow in the southern regions of the zone. The SBS has less precipitation than the CWH with a greater proportion (approximately 25-50%) occurring as snow (Meidinger and Pojar 1991). Practical considerations in choosing the study sites included permission of the site managers to sample runoff and publish results, site accessibility, and existence of appropriate sampling points.  23  2  Materials and Methods  Chemical analyses, toxicity tests and limited treatability studies were conducted on log yard stormwater runoff from a coastal and interior log yard site. To examine the potential for acute effects, data were evaluated and compared to criteria. Additionally, data were statistically compared between sites and assessed for seasonal trends. In order to evaluate contaminant loadings to the receiving environment, export coefficients (area standardized loadings) were calculated and compared between sites. Possible relationships between toxicological and chemical variables were explored statistically.  2.1 Site Descriptions The two sawmill operations participating in this study have requested that their identities remain confidential and will hereafter be referred to as the interior and coastal sites. Note that both mills experienced temporary shut-downs in production in 2001. The coastal site has since been permanently shut down. 2.1.1 Interior Site 2.1.1.1 General Operations and Site Description  The interior operation is located on the shore of a river in the northern interior portion of BC in the Sub-Boreal Spruce (SBS) Biogeoclimatic Zone. The lumber processed at the site reflects the dominant tree species in the SBS and includes white spruce, subalpine fir and lodgepole pine. The volume of logs stored on site for 2001 is summarized in Table 2.1. No anti-microbial treatment is applied to the logs or lumber on the property. Two log yards lie adjacent to the main sawmill buildings, one north and one east of the main sawmill operations (Figure 2.1). No portions of the log yard areas are paved. Parking lots are located at the north-western end of the property. Three ditches drain the property (northern,  24  central and southern), flowing eastward through the eastern log yard and a riparian buffer area before discharging into a nearby river. The northern ditch drains a portion of both the eastern and northern log yards, as well as the parking lots. The central ditch runs through the eastern log yard and receives stormwater from both the log yard and the upland sawmill area. The southern ditch drains the southern portion of the eastern log yard (Figure 2.1). Table 2.1: Year 2001 log inventory for the coastal and interior site  Coastal Log Inventory (m)  Interior Log Inventory (m)  January  3121  172,112.6  February  3247  175,891.8  March  3414  83,397.77  April  2027  71,023.4  May  2073  109,174.7  June  2288  135,243.1  July  2288  107,601.6  August  2760  76,347.79  September  2759  37,620.7  October  2505  60,264.79  November  2505  194,566.3  December  2505  169,546.6  Month  3  25  3  Riparian Areas  Figure 2.1: Schematic of the interior log yard 2.1.1.2  Climate  The SBS climate is continental and characterized by cold snowy winters and relatively warm, moist, short summers. The mean annual temperature of the SBS ranges from 1.7 to 5°C. Average temperature is below 0°C for 4 to 5 months of the year and above 10°C for 2 to 5 months. Mean annual precipitation is moderate, ranging from 439-1588 mm, of which approximately 25-50% is snow. The driest month of the year is generally April, with 15 to 37 mm of precipitation. The wettest month is August (50 - 279 mm) due to numerous thunderstorms that occur in the summer months (Meidinger and Pojar 1991). 2.1.2  Coastal Site  2.1.2.1 General Operations and Site Description  This coastal sawmill operation is located in southern coastal BC, in the Coastal Western Hemlock (CWH) zone. Lumber processed on site is approximately 50% fir (true firs) and 50% 26  western hemlock. On average, approximately 2,600 cubic metres of logs were stored on site each month. Logs are also stored in an adjacent saltwater bay. No anti-microbial treatment is applied to the logs or lumber on the property. Buildings on site include the administrative and sawmill buildings, located roughly in the center of the property (Figure 2.2). A parking lot is located immediately northeast of the administrative buildings. The log yard area, including a crane for moving logs between the bay and the property, is located in the southeastern portion of the site. Approximately 70% of the log yard area is paved. Lumber is generally stored in the south western portion of the property. Surface runoff is controlled through two main drainage areas on the property. Stormwater on the southeastern side flows towards the ocean, where it drains the log yard area and the southern portion of the sawmill buildings. It is diverted into a settling pond via a barrier of strawbales. Stormwater from the settling pond is pumped upland (northwest) into a gravel /soil field. This area of the property is largely paved, with the exception of a debris fdled area under the crane and areas on the southern edge of the property. Stormwater from the northeastern and northwestern parts of the property flows towards the ocean in an open culvert on the northeastern edge of the property. It drains a portion of the sawmill, administrative buildings and parking lots. Since the southwestern area of the property is sloped towards this culvert, stormwater from this area is drained via the open culvert. These areas are entirely paved (Figure 2.2).  2.1.2.2  Climate  The CWH is the wettest biogeoclimatic zone in BC. It has a cool, mesofhermal climate with cool summers and mild winters. The mean annual temperature is approximately 8°C and ranges from 5.2 to 10.5°C among the CWH subzones. The mean monthly temperature is above 10°C for 4 to 27  6 months of the year. Mean annual precipitation is 2228 mm and ranges from 990 to 4387 mm. Less than 15% of this occurs as snowfall in the south. The driest and wettest months of the year are July and December, with mean annual precipitation ranging from 17 to 15.1 mm and 146 to 625 mm, respectively (Meidinger and Pojar 1991).  Paved L o g Y a r d Unpaved L o g Y a r d Sampling Point  Figure  2.2:  Schematic  of the coastal  log  yard.  28  2.2 2.2.1  Field Sampling Sample Location  Sampling locations were chosen to maximize contaminant inputs from log yard areas and minimize those from other areas of the sawmill operations. At the interior site, one sample was collected from the northern ditch and two samples were collected from the southern ditch. Samples were collected approximately 10 metres downstream from the beginning of the ditch (Figure 2.1). At the coastal site, samples were collected just prior to the straw bale opening to the settling pond (Figure 2.2).  2.2.2  Sample Collection  Ideally, water quality samples for runoff assessment should be collected throughout an entire storm event. The USEPA (Athayde et al. 1983) defines a measure of a storm's contaminant loading by the event mean concentration (EMC). The EMC is a flow weighted average for the entire storm obtained by sampling throughout an entire storm event. However, such intensive sampling is often unpractical due to limited resources and the duration of storm events, particularly on the West Coast. Partial event sampling of urban runoff has been conducted by the City of Bellevue (Bellevue Utilities 1995). The measure of the storm's contaminant loading was referred to as the sampled mean concentration (SMC) to distinguish it from the EMC calculated from monitoring an entire storm event. The results from this study show that although in some cases, samples were not representative of the entire storm, the SMC approach can provide useful information in a cost-effective manner (Macdonald et al. 1997). For this study, the SMC approach was used. Flow rated composite sampling (constant time and flow proportional volume) was conducted manually for each storm event over a period of approximately 3 hours, based on methods summarized by the USEPA (1992). Composite samples are intended to reflect the average wastewater characteristic during the time of 29  monitoring and provide a single integrated wastewater characteristic profde. Sampling was conducted by Environmental Dynamics Inc. at the interior site. Sampling at the time of storm initiation was attempted to catch first flush events. However, due to travel logistics and sawmill operating hours, sampling started after storm initiation for most storm events. The dry period prior to the storm was obtained from Environment Canada Atmospheric Services and recorded. Sampling was conducted as follows. Every 20 minutes, stormwater flow rate was measured and a grab sample collected into a 5-L clean high density polyethylene (HDPE) plastic pail. Field sample measurements (pH, conductivity, temperature, dissolved oxygen) were taken immediately upon sample collection and total rainfall at the time of the grab sample was recorded from a plastic rain gauge installed at the beginning of each sampling event. The 5-L plastic pail was covered with a lid to prevent dilution with rainwater. In total, nine grab samples were collected over a period of 160 minutes. After collection of nine 5-L grab samples, the aliquot of each grab sample to be included in a 10L flow proportioned composite sample was calculated according to the following equation:  Aliquot  rr 7  i  Volume  \mL)  T  \  Grab  Sample  Flow  = Total  Event  Flow  .  n  n  n  n  x 10,000  r  (1)  mL  The aliquot volume from each grab sample was measured using a 500 mL graduated cylinder and placed into a 20-L HDPE plastic pre-cleaned jerry can. Once aliquots from each grab sample were measured and combined, the jerry can was gently shaken to combine sample water. The 10L composite was then distributed into the following containers: •  1-L HDPE plastic collapsible containers for Ceriodaphnia  •  1-L HDPE plastic sample bottles for total suspended solids (TSS), biochemical oxygen  30  dubia  toxicity testing;  demand (BOD), tannins and lignins, and Microtox® testing; •  125 mL HDPE plastic sample bottle for chemical oxygen demand (COD) analysis;  •  1-L amber glass bottle for dehydroabietic acid (DHA) analysis; and,  •  250 mL HDPE plastic sample bottle for total metals analysis.  Care was taken to ensure that sample foam was included in the sample when distributing the composite samples to the appropriate sample bottles since foam from woodwaste runoff often contains resin and fatty acids. In some cases, this meant not filling the sample bottle right to the top. Samples for COD analysis were preserved immediately with 2 mL sulphuric acid, samples for total metals analysis were preserved with 4 mL nitric acid. Sample bottles were placed immediately into cooler with icepacks. Separate sampling gear was used at the two sites. With the exception of sample bottles for total metals analysis, sample bottles were re-used over several sampling events. Bottles were cleaned prior to use by three rinses with tap water, followed by three with distilled water. Additionally, bottles were rinsed with site water in the field. All sampling equipment was scrubbed and cleaned with regular tap water, then rinsed with site water in the field.  2.2.3 Flow Rate Measurements 2.2.3.1  Interior  Flow at the interior site was measured using the velocity-based stopwatch and floating object method (USEPA 1992). Width and depth measurements were taken over a 1.5 m straight section of the ditch (float track). A floating object was dropped into the current and timed to determine the travel time taken to traverse 1.5 m. Flow rate was calculated according the following equation: 31  Flow  Rate (m /s)= 3  Area  (m )x 2  Velocity  (m/s)  (2)  Where: Area = water depth (m) x width of flow (m); and Velocity: length of float track (m)/float travel time (s).  2.2.3.2  Coastal  Stormwater runoff from the southern drainage was directed into the settling pond via a straw bale barrier with a small opening, approximately 30 cm wide and 60 cm long. Flow was channeled through this opening before dispersing into the settling pond. During heavier rains, a pool would form upstream of the entrance of the channel. For the first two samples, a temporary channel was constructed using strawbales and plywood and the floating object/stopwatch method was employed. However, due to the limited distance over which flow width and depth were relatively constant, the floating object/stopwatch method employed at the interior site could not be used with accuracy at the coastal site. Therefore, for the remaining eight samples, a 20 cm (8 inch) cutthroat flume with a maximum capacity of 3.79 cubic metres was used to measure flow rate. In order to determine flow rate through the flume, water depth (stage) was measured at the entrance to the flume using a scale etched on the side of the flume. Flow rate was then calculated using conversion tables provided by the flume manufactures (Baski Inc., Denver, Colorado) As the site is subject to use by heavy machinery, the cutthroat flume could not be permanently installed at the site, but was installed in the opening between the strawbales upon arrival. During occasions in which a pool of water had formed at the mouth of the straw bale opening, sampling was delayed for approximately 10 minutes to allow flow to stabilize. 32  2.2.4  Field Measurements  All field measurements (pH, conductivity, DO and temperature) at the coastal site were collected with a YSI Model 600 SN Multiprobe meter (Yellow Springs Instrument Company, Yellow Springs, OH, USA) with the exception of the sample collected on October 24, 2001 (S-10/24). On this date, the YSI meter was not functioning and several other meters had to be used. Dissolved oxygen (DO) and temperature were measured in the field with a YSI meter (Model 57). pH and conductivity were measured in the laboratory a few hours after sampling. pH was measured with an Orion pH meter Model 266 (Orion Research, Beverly, MA, USA) and conductivity with a YSI meter (Model 135A). Field measurements at the interior site were conducted with a YSI meter (Model 57) for DO and temperature and an Oakton Model TDS-TESTR 3 for conductivity. Unfortunately, a pH meter was not available for use and field pH had to be estimated with pH strips. All meters were calibrated prior to use according to equipment specific instructions and using appropriate pH buffer and conductivity solutions.  2.3 Laboratory Analyses 2.3.1  Chemical Analyses  It is difficult to characterize all the compounds present in log yard runoff due to the high number of wood compounds and other potential contaminants present in stormwater. Therefore, the parameters chosen for analysis in this study are based on parameters identified from other sources (AFPA 1999a, Bailey et al. 1999, NCASI 1992, Samis et al. 1999, Taylor 1994) and are thought to be at least partially responsible for runoff toxicity. The following chemical analyses were conducted in the UBC Pulp and Paper Center: Tannins and Lignins; COD; BOD ; TSS; and DHA. 5  33  Alkalinity analyses were conducted at the EVS Environment Consultants (EVS) laboratory. Total metals analyses were conducted by Analytical Laboratory Services (ALS). Hardness was calculated from calcium and magnesium concentrations provided by ALS.  2.3.1.1  Tannins and Lignins  Tannin and lignin analyses were conducted according to standard methods (American Public Health Association [APHA] 1998). Tannins and lignins contain aromatic hydroxyl groups that react with folin phenol reagent (tungstophosphoric and molybdophosphoric acids) to form a blue colour suitable for estimation of concentrations up to 9 mg/L (APHA 1998). Carbonate-tartrate solution for the analysis was prepared by dissolving 200 g sodium carbonate (Na C03) and 12 g sodium tartrate (Na2C H 06 x 2H 0) in 1-L of distilled water in an 2  4  4  2  Erlenmeyer flask. Undiluted sample was shaken to prevent settling of solids and diluted by 1:25 to 1:50, depending on sample strength, with room temperature distilled water. A 50 mL volume was placed in a clean Erlenmeyer flask. One mL commercially-prepared folin phenol reagent and 10 mL carbonate-tartrate reagent were added in rapid succession and the flask was gently swirled. Flask contents were poured into three pre-cleaned HACH vials and capped. Thirty minutes was allowed for colour development, at which point the vials were rinsed with distilled water and wiped with tissue. A reagent blank was simultaneously prepared as above using 50 mL distilled water instead of sample. Absorbance readings were measured with a HACH direct reading spectrophotometer (Model DR/2000) at a wavelength of 700 nm. The spectrophotometer was set to zero with both distilled water and the reagent blank prior to taking sample readings. Absorbance readings were converted into concentrations (mg/L) using a calibration curve 34  equation. Calibration curves were plotted from series of standard solutions (10, 7.5, 5, 2.5, 1.25 mg/L) prepared from 100 mg/L stock solution of tannic acid. Folin phenol reagent and carbonate-tartrate solution were added and absorbance readings were measured against a simultaneously prepared reagent blank as described above. Sample concentrations were then multiplied by the sample dilution volume to achieve tannin and lignin concentrations in undiluted sample. New calibration curves were prepared for each new batch of carbonate-tartrate solution. Unfortunately, the analysis is not specific for tannins and lignins, since the reaction will occur with other organic and inorganic reducing substances such as iron (II) and manganese (II) (APHA 1998). Since iron and manganese are ubiquitous in the environment (and present in the runoff samples), an experiment was conducted to ensure that minimal interference was occurring from these two metals. A series of solutions (10, 7.5, 5, 2.5, 1.25 mg/L) was prepared from 100 mg/L stock solutions of ferrous chloride (FeCL) and manganese chloride (MnCl ) and distilled 2  water. Distilled water was used to dilute the stock solution as runoff samples were diluted with distilled water by 1:25 to 1:50 prior to analysis. Subsequently, the sample solution was more similar to distilled water than to 100% sample. Folin phenol reagent and carbonate-tartrate solution were added and absorbance readings were measured against a simultaneously prepared reagent blank as described above. Absorbance readings were converted into concentrations to determine the potential interference of these substances in the tannins and lignins analysis.  2.3.1.2 Chemical Oxygen Demand (COD)  COD is a measure of the organic matter content of a sample (in oxygen equivalents) that is susceptible to oxidation by strong chemical oxidants. However, some compounds, such as volatile straight-chain aliphatic compounds, are not oxidized to any appreciable extent (APHA 1998). 35  COD analyses were conducted within 48 hours of sampling according to the (APHA 1998) Closed Reflux Colourimetric Method. Undiluted sample was shaken to prevent settling of solids and diluted (between 1:2 or 1:5 fold) with room temperature distilled water. A 2.5 mL volume was placed in three pre-cleaned HACH vials. Digestion reagent (1.5 mL of K C r 0 7 , H2SO4, 2  2  HgS0 ) and sulphuric acid reagent (3.5 mL of H S0 , Ag S0 ) were added to each vial. Vials 4  2  4  2  4  were capped and inverted several times and placed in a pre-heated COD Reactor and heated for 120 minutes at 150°C. Reagent blanks were simultaneously prepared as above using 2.5 mL distilled water in each of three HACH vials. After digestion, vials were cooled to room temperature, rinsed with distilled water and wiped with a tissue. Absorbance readings were measured with a HACH direct reading spectrophotometer (Model DR/2000) at a wavelength of 600 nm. The spectrophotometer was set to zero with both distilled water and the reagent blank prior to taking sample readings. Absorbance readings were converted into concentrations (mg/L) with a calibration curve equation. Calibration curves were calculated from series of six standard solutions prepared from potassium hydrogen phthalate solution with COD equivalents from 20 to 900 mg 0 /L. Digestion 2  and sulphuric acid reagents were added, samples were digested and absorbance readings were measured against a simultaneously prepared reagent blank as described above. The calibration curve was constructed by plotting absorbance measurements vs. known solution concentrations. Sample concentrations were multiplied by the sample dilution volume to achieve COD concentrations in undiluted sample. New calibration curves were prepared for each new batch of reagents.  2.3.1.3 Biochemical Oxygen Demand (BOD)  The 5 day BOD determination is a test in which standardized laboratory procedures are used to  36  determine the relative oxygen requirements of wastewater, effluents and polluted waters. The test measures the oxygen used during an incubation period of five days for the biochemical oxidation of organic material and some inorganic material such as sulphides and ferrous iron (APHA 1998) Analyses were conducted according to standard procedures (APHA 1998) within 24 hours of sampling for coastal samples and upon receipt of interior samples, approximately 48 hours after sampling. Fresh dilution water was prepared for each analysis by adding 10 mL of phosphate buffer solution, magnesium sulfate solution, calcium chloride solution and ferric chloride solution to 10-L of distilled water. Dilution water was saturated with DO by shaking the partially filled container. Sludge was obtained from the recycle activated sludge (RAS) line of activated sludge treatments from two coastal pulp and paper operations. The RAS was diluted 1:3 with distilled water prior to use as a seed source for the BOD test. A series of sample dilutions (ranging from 1:10 to 1:500) were prepared in triplicate using fresh dilution water and placed in 60 mL incubation bottles. To each incubation bottle, 0.5 mL of seed was added and initial DO was measured with a YSI DO meter (Model 59). The bottles were sealed with stoppers and caps. Controls were simultaneously prepared in the same manner. However, initial DO was measured prior to adding 0.5, 1.0 and 1.5 mL seed to the three dilution water controls. Bottles were incubated for five days after test initiation at 20°C. After 5 days, the final DO of each incubation bottle was measured using the DO meter. To calculate BOD concentrations, a calibration curve was prepared by plotting control seed volumes (0.5, 1.0, 1.5 mL) versus the change in control DO (initial - final DO) at each seed volume. A regression line was plotted and the change in DO at 1.0 mg/L seed was recorded. BOD was then calculated using the following equation: 37  BODM IL)- '- 'l(D  D  (3)  (B)f  S  Where: Di = initial DO of sample; D = final DO of sample; 2  P = decimal volumetric fraction of sample used; B - change in DO concentration at 1.0 mg/L seed obtained from regression line; and f = ratio of seed in sample to seed in dilution control water (0.5). Dilutions that result in a residual DO of at least 1 mg/L and a DO uptake of at least 2 mg/L after 5 day incubation produce the most reliable results. If more than one sample dilution met these criteria, and there was no evidence of toxicity at higher sample concentrations, results were averaged for the acceptable dilutions (APHA 1998). If higher sample concentrations indicated the presence of sample toxicity (i.e., lower BOD at higher sample concentrations), the lowest sample concentration was used to calculate BOD.  2.3.1.4 Total Suspended Solids (TSS)  Total suspended solids were analyzed within 7 days of sample collection according to standard methods (APHA 1998). Three glass fiber filters were placed in numbered aluminum planchets, dried for approximately 1 hour in a 100°C drying oven and placed in a dessicator for approximately 30 minutes prior to taking initial weight measurements. Filters were placed in suction flasks and three 100 mL volumes of well shaken sample were filtered. After filtration was complete, filters were carefully  38  removed and placed in the appropriate aluminum planchet and left to dry overnight at approximately 100 °C in a drying oven prior to taking final weight measurements. TSS was calculated as follows:  f  TSS  {mglL)  =  A -B  100  mL  ^  x 1000  (4)  Where: A = weight offilter+ dried residue (mg); and B = weight offilter(mg).  2.3.1.5 Dehydroabietic A cid  DHA analysis was conducted by extraction with organic solvent and Gas Chromatograph (GC) analysis within 7 days of sample collection. All glassware used in the analysis was pre-cleaned and fired overnight in a muffle furnace. A 50 mL volume of homogenized undiluted sample was adjusted to pH 9, poured into a separatory funnel and 0.5 mL of recovery standard (9.8 mg/L o-methyl podocarpic acid [OMPCA]) was added to the sample. The sample was extracted twice with 50 mL methyl tertiarybutyl ether (MTBE) and the organic phase decanted into a round-bottom flask. The organic phase was evaporated to approximately 0.5 mL with a Biichi Rotavapour <R>. The remaining sample was transferred to a GC vial with a pipette (including a rinse of the round-bottom flask with MTBE) and dried to a film with nitrogen gas. A volume of 0.5 mL of bis(trimethylsilyl)trifluoroacetamide (BSTFA) was added to the GC vial and the vial was heated gently for 30 minutes to ensure volatilization of resins. In addition to the sample, a GC standard was simultaneously prepared in a vial by drying down a 1:1 mixture of recovery standard and internal  39  standard (10.5 mg/L henecosanoic acid) with nitrogen gas, adding BSTFA and heating for 30 minutes. The sample and standard were then analyzed on a GC (HP 5990). The column was a HP5 fused silica capillary column (30 m x 0.32 mm x 0.525 um). The method involved heating at 150°C for 1 minute, ramping heat at 4°C per minute at 265°C, maintaining heat at 265°C for 3 minutes and then ramping heat at 12.5°C per minute at 290°. The carrier gas was helium, gas flow rate was 15 PSI and the split ratio was 50:1. Total DHA was calculated according to the following equation: DP DHA D  H  A  (  m  g  /  L  )  -  x P « peakarea JQ  TC concentration  eakare  ^  a  s  w  x  s  o  (  5  )  20 Where: RS = GC Standard Recovery Standard (O-MPCA); IS= Internal Standard (henecosanoic acid); RF = Recovery Factor (0.95); and RSS =Sample Recovery Standard (O-MPCA). 2.3.1.6 Alkalinity  The alkalinity of water is its ability to neutralize acid inputs and is the sum of all titratable bases (APHA 1998). Analyses (potentiometric titration of low alkalinity) were conducted according to standard methods described in APHA (1998). An aliquot (100 mL) of undiluted sample in a 150 mL glass beaker was titrated with 0.02 N sulphuric acid. The sample was gently stirred during 40  the titration using a stir bar and pH measurements were recorded with an Orion pH meter (Models 266 or 230A, Orion Research, Beverly, MA, USA). The titration was stopped when the pH reached 4.5 and the exact volume of acid used was recorded. Additional acid was carefully added until pH dropped to 4.2 (a drop 0.30 pH units) and the exact volume of acid was recorded. Total alkalinity was calculated as follows:  Alkalinity (mg/L  CaC0 )  {2B - C) x N x 5 0 0 0 0  3  mL sample  (6)  Where: B = total mL titrant to first recorded pH; C = total mL titrant to second recorded pH; and N = normality of acid.  2.3.1.7 Hardness  Total hardness is defined as the ability of a water sample to precipitate soap. It is mainly attributable to the sum of calcium and magnesium concentrations, both expressed as calcium carbonate (CaC03) in mg/L. However, other ions such as iron can also contribute to hardness. Hardness was calculated according the following equation (APHA 1998): Hardness (mg/L CaCO,)=2A91  (Ca,mg/l)+4.118(Mg,mg/L)  (7)  2.3.1.8 Total Metals  Total metals were analyzed by ALS Laboratories within 6 months of sample collection (Table 2.2). Analysis was carried out using procedures adapted from APHA (1998) and from "Test Methods for Evaluating Solid Waste" SW-846 published by the USEPA. Samples were acid digested prior to instrumental analysis with inductively coupled plasma-optical emission 41  spectrophotometry (ICP-OES) (USEPA Method 601 OB). Table 2.2: List of metals analyzed and associated ICP-OES  2.3.2 2.3.2.1  detection limits.  Metals  DL  Metals  DL  A l u m i n u m (Al)  0.2  Manganese (Mn)  0.005  Antimony (Sb)  0.2  Molybdenum (Mo)  0.03  Arsenic (As)  0.2  Nickel (Ni)  0.05  Barium (Ba)  0.01  P h o s p h o r u s (P)  0.3  Beryllium ( B e )  0.005  Potassium (K)  2  B i s m u t h (Bi)  0.1  Selenium (Se)  0.2  Boron (B)  0.1  Silicon (Si)  0.05  Cadmium (Cd)  0.01  Silver ( A g )  0.01  Calcium (Ca)  0.05  Sodium (Na)  2  C h r o m i u m (Cr)  0.01  Strontium (Sr)  0.005  Cobalt (Co)  0.01  T h a l l i u m (TI)  0.2  Copper (Cu)  0.01  Tin ( S n )  0.03  Iron ( F e )  0.03  T i t a n i u m (Ti)  0.01  Lead(Pb)  0.05  V a n a d i u m (V)  0.03  Lithium (Li)  0.01  Zinc (Zn)  0.005  Magnesium (Mg)  0.1  Toxicity Testing Ceriodaphnia  dubia  Toxicity tests were conducted at EVS in a controlled environment room maintained at 25°C. Testing was conducted within 5 days of sample collection. Culture Water Culture water was prepared by EVS staff according to in-house protocol and USEPA (1993) guidelines. Culture water was prepared using 20% Perrier (4-L), 80% de-ionized water (16-L), 2 ug/L selenium and 10 ug/L vitamin B12. Water quality measurements were recorded after preparation to ensure that culture water met the following requirements: •  DO: 90-100%  42  •  pH: 7.9-8.3  •  Temperature: 25 ±1 °C  •  Hardness: 80 to 100 mg/L CaC0  •  Alkalinity: 60 -70 mg/L CaC0  3  3  Brood Cultures EVS maintained individual  C. dubia  cultures (broodboards) according to USEPA (1993) and in-  house protocols. Daphnids were cultured in the controlled environment room (25°C) in culture water augmented with food under a light regime of 16:8 h light/dark photoperiod. Culture water was replaced daily and health of the daphnids monitored. Daphnids were fed a diet of dYCT (yeast, cereal flakes, and trout food) and green algae (Selenastrum  capricornutum)  combined in  1:1 ratio. Toxicity tests with a reference toxicant (zinc sulphate) were conducted on a monthly basis by EVS staff to monitor culture health. Toxicity Test Procedures Acute 48-hr C.  dubia  tests were conducted according USEPA (1993) protocols. Undiluted  sample was warmed to 25°C and a geometric series of five exposure concentrations (6.25, 12.5, 25, 50, 100% v/v) were prepared using culture water. A laboratory control prepared using culture water was used in each test. Water quality parameters for exposure concentrations and controls were measured and recorded. Each exposure concentration and control water was poured into four test containers (replicates), consisting of 30 mL plastic cups containing 15 mL solution. Neonate daphnids (< 24 hours old) obtained from brooding adults from individual cultures were carefully placed in a glass beaker filled with dilution water with a glass pipette. From the glass beaker, five daphnids were gently 43  transferred to each test container with a pipette and gently released under water to minimize handling damage. Care was taken to ensure that exactly five were introduced. Test containers were randomized and maintained at 25°C with a 16:8 h light/dark photoperiod without aeration for a 48 hour period. Daphnids were not fed during the test. After the 48-hr period, mortalities in each test container were recorded and water quality parameters were measured. Great care was taken to locate all five seeded daphnids, dead or alive. Since samples were quite turbid, contents of the test container were poured into a shallow dish to help in locating the daphnids. If all five daphnids could not be located, the test solution was carefully pipetted out and the contents of the pipette examined for daphnids. In some cases, all five daphnids could not be found. In these cases, daphnids were presumed to be dead (as opposed to not introduced, or alive, but not found) for the following reasons: (1) care was taken at test initiation to release exactly five daphnids into the solution; (2) dead daphnids usually fall to the bottom of the test container, often becoming smothered in sample solids; and, (3) each sample was examined closely to ensure that no live daphnids were uncounted. Test data were considered acceptable if: (1) minimum control survival for test acceptability was 90% and (2) water quality parameters were within the specified range. The LC50s and confidence intervals were estimated with the trimmed Spearman-Karber method using EFFL LC50 version 1.2 (Environment Canada, North Vancouver, BC, Canada).  Sample Water Quality Parameters DO, conductivity, temperature and pH measurements were recorded for test solutions at test initiation and termination to ensure compliance with test protocols. Water quality parameters were measured for all exposure concentrations and test controls at test initiation to ensure samples were properly diluted. The same parameters were measured at test termination for 100,  44  50 and 6.25% sample concentration and test controls to ensure test compliance with specified water quality parameters. DO and conductivity were measured with YSI instruments (Models 57 and 135A, respectively). pH was measured with an Orion pH meter (Model 266) and temperature was measured with a glass mercury thermometer. Test requirements stipulate that water quality parameters for sample DO, temperature and pH remain within a certain range. DO must be greater than 4 mg/L (but not supersaturated) for the duration of the test, temperature must remain with 25 ±1 °C for the duration of the test and sample pH must between 6.0 - 9.0 at the initiation of the test (USEPA 1993). If DO in a exposure concentration was < 4 mg/L at test initiation, the flask was gently swirled to increase the oxygen content. When test vessels were supersaturated with oxygen (DO readings greater than 100%), DO was reduced with aeration tubes placed at the bottom of the beakers to "bubble out" excess DO. This was done for the minimal time required to reduce the oxygen content of the sample. Since altering sample pH has the potential to alter sample chemistry, sample pH was not altered unless the pH values fell below 6.0. When necessary, pH was adjusted with 0.01 M sodium hydroxide (NaOH).  Sample Treatments All toxicity tests on treated samples were conducted with C. dubia at the EVS Laboratory in a controlled environment room maintained at 25°C. Samples were considered usable for treatment purposes as long as they remained toxic (i.e., baseline sample used in treatments was toxic to C. dubia).  45  Suspended Solids In order to determine whether suspended solids were a contributing factor to sample toxicity, the following sample treatments were conducted and samples subsequently re-tested for toxicity: •  Removal of suspended solids by centrifugation; and  •  Spiking of sample suspended solids into culture water.  Suspended solids from four of seven coastal runoff samples and two of two interior samples that exhibited toxicity to  C. dubia  were removed. Undiluted samples were centrifuged in clean 50 mL  capped centrifuge tubes at approximately 3000 rpm for 15 minutes to remove particulates. The resulting supernatant was tested according to the standard toxicity test procedures described above along with standard controls. An untreated sample was tested concurrently to ensure that sample toxicity had not dissipated subsequent to the initial acute toxicity test. Spiking of culture water with sample centrifuged solids was conducted twice to determine whether toxicity could be introduced to culture water by adding solids from the sample. For sample N-07/19, undiluted sample was centrifuged in 50 mL centrifuge tubes for 15 minutes. The centrate was decanted into a clean beaker, and the solids re-suspended with dilution control water directly in the plastic centrifuge tube. The overlying centrate, the solution of re-suspended solids, an untreated sample and test controls were tested according to standard protocols, but with two replicates. To ensure that centrifugation and subsequent re-suspension of particulates did not affect sample toxicity, a centrifuge control test was conducted. Undiluted sample in a 50 mL centrifuge tube was centrifuged for 15 minutes and then immediately shaken to re-suspend the settled solids. Two replicates of the centrifuge and standard control were tested according to standard protocols. 46  Soluble COD analyses were conducted on five samples to determine the effects of TSS removal. Soluble tannins and lignin analyses were conducted for samples N-07/19 and S-10/24 , soluble DHA analysis were conducted for sample N-07/19.  Metals In order to determine whether metals were responsible for sample toxicity, undiluted samples N 07/19 and S-10/24 were treated with three ethylenediaminetetraacetic acid (EDTA, a metal chelating agent) concentrations. The range of EDTA concentrations added was determined for each sample based on its hardness (Table 2.3). An untreated sample was tested concurrently to ensure that sample toxicity had not dissipated subsequent to the initial acute toxicity test. In addition to standard dilution water controls, controls were spiked with the two higher concentrations of EDTA. A predetermined volume of 0.1 M EDTA was added with a pipette to three 50 mL volumes of undiluted sample according to the hardness of the sample. After addition of the EDTA, the samples were gently swirled and left to equilibrate for one hour prior to testing. After allowing the samples to equilibrate, the pH of each solution was adjusted back to the pH of the untreated sample using 0.0IM NaOH. From this point, tests proceeded as described above. Table 2.3: Volume of EDTA added for C. dubia sample treatments. o  | P  N-07/19  S-10/24  Sample Hardness (mg/L)  Sample Volume (mL)  V o l u m e EDTA (mL)  294  50  0.5  300  50  0.3  200  50  0.16  100  50  0.113  75  50  0.075  50  50  0.03  20  40  47  EDTA  Concentration (mg/L)  2.3.2.2 Microtox®  Acute 5 and 15 min Microtox® toxicity tests were conducted at the Pulp and Paper Laboratory within 24 hours of sample collection for coastal samples and upon receipt of interior samples (within 48 hours of collection). Basic Microtox® testing was conducted according to guidelines described in Microbics Corporation Microtox® Manual, Volume 3 (1992). The Microtox® system is used to find the concentration of a sample at which the light output of the reagent (freeze-dried culture of specially developed strain of the marine bacterium, Vibrio fischeri) is reduced by a specific percentage. Tests were conducted using the Microtox® Model 500 Analyzer. Basic test protocols, with four 1:2 serial dilutions (45.0, 22.5, 11.25, and 5.626 %) and three controls, were followed. Undiluted samples were centrifuged at 10,000 rpm for 10 minutes to remove suspended solids and pH adjusted (if necessary) to a value between 6.5 to 7.5. using 0.01 M NaOH. Microtox® cuvettes were placed in the incubator and in the reagent well. Frozen reagent was mixed with 1000 uL Microbics reconstitution solution (specially prepared distilled water) and placed in the reagent well. Reagent was thoroughly mixed with a pipette and transferred to experimental cuvettes containing 500 U.L Microbics diluent (2% sodium chloride [NaCl] solution), mixed and allowed to sit for 15 minutes. In the meantime, controls and sample dilutions were prepared in sample preparation cuvettes containing 1000 uL Microbics diluent and Microbics osmotic adjustment solution (22% solution of NaCl). Five minutes were allowed for samples to equilibrate to the appropriate temperature. After the reagent was allowed to sit for 15 minutes, 500 uL of sample dilutions were transferred to the experimental cuvettes and mixed with a pipette. Immediately after sample transfer, the incubator was calibrated with a control experimental cuvette and control and sample light levels recorded at time 0. Control and sample light levels were then recorded at 5 and 15 minutes. EC50 concentrations (5 and 15 minute) were 48  calculated using the Microtox® software. Test data were considered acceptable if: (1) control light levels were within 20% and (2) the 95% confidence range did not exceed 30% of the EC50.  2.4  Data  Analyses  Censored data (data below analytical detection limits) were replaced with half the detection limit for summary calculations and statistical analyses. Statistical analyses were conducted with SPLUS 2000 (MathSoft, Inc, Seattle, Washington). There were not enough samples collected from the interior site to conduct correlation and regression analyses. Therefore, in order to identify potential differences and similarities that exist between the sites, analyses were conducted on two sets of data: (1) the interior and coastal data combined and (2) the subset of coastal data. Regression analyses were not conducted with Microtox® toxicological data as samples were centrifuged prior to testing. Centrifuging samples potentially removes contaminants from solution, and thus toxicity results do not necessarily correspond to water quality parameters.  2.4.1  Evaluation of Log Yard Stormwater Runoff  Water quality results were summarized and compared to water quality criteria (where applicable) for each site. Additionally, comparisons between Microtox® and C. dubia tests were conducted qualitatively and by calculating Spearman's rank correlation coefficients between Microtox® EC50s (5 and 15 minute) and C. dubia LC50s. Where test endpoints were greater than 100% sample, 100% was substituted as the endpoint.  2.4.2  Statistical Comparisons Between Sites  Water quality data were tested for normality of distribution using the Kolmogorov-Smirnov  49  Goodness of Fit Test. Since not all the data met assumptions of normality, and due to the small number of samples collected at the interior site, comparisons of contaminant concentrations and toxicity between the coastal and interior sites were conducted using the Mann-Whitney test, a non-parametric t-test that uses ranked data instead of measurements. The Mann-Whitney is one of the more powerful non-parametric tests and is about 95% as powerful as the parametric t-test when either of the tests is applicable (Zar 1996). Laboratory measured pH and conductivity values measured at the time of C. dubia toxicity test initiation were used for tests instead of field measured pH and conductivity values. The metals chromium, cobalt, lithium, nickel and vanadium were not tested for differences as they were below detection limits in coastal runoff. Boron was not tested for differences as it was below the detection limit in interior runoff.  2.4.3  Seasonal Trends  For the coastal site, seasonal patterns in parameter concentrations and toxicity were evaluated graphically (qualitatively) and statistically (using the Mann-Whitney test). In order to statistically evaluate seasonal patterns for the coastal site, samples were categorized into wet (> 5 mm precipitation in preceding 3 days) and dry weather events and tested for differences using the Mann-Whitney test. In this way, only two "seasons" were evaluated - wet and dry. It was assumed that if runoff concentrations and toxicity were different for wet weather and dry weather events, then differences in runoff concentration and toxicity would exist between seasons with high or low rainfall amounts. Seasonal patterns could not be evaluated at the interior site due to the lack of samples. However, qualitative comparisons were conducted between the snowmelt and rainfall runoff sample. To examine the effect of flow (i.e., discharge volume over the sampling event) on toxicity, the log of the total flow (m ) for each sampling event was regressed against C. dubia survival in 3  50  100% sample for the combined and coastal data. Flow was used as an estimate of dilution and substitute parameter for rainfall/seasons due to lack of detailed rainfall data available for the sites. Since samples collected from the coastal site during operational (8) and non-operational (2) days differed visually and in their toxicity to C. dubia, total flow for only operational days was also regressed against C. dubia survival to investigate the presence of confounding factors such as the level of activity on site. To examine the effect of flow on water quality parameters, total flow (logged) for the combined and coastal data, as well as the subset of coastal operational days, was regressed (simple linear regression) against water quality parameters found to correlate well (r > 0.80) with C. dubia 2  toxicity. These included manganese for the combined data and tannins and lignins, manganese, barium and phosphorus for the coastal data.  2.4.4  Calculation of Loadings for Coastal and Interior Sites  In order to evaluate the potential for cumulative effects from log yard runoff stormwater at each site, contaminant loadings were calculated for each site. Loadings were calculated for stormwater parameters present at both sites and for those that have the potential to accumulate in the environment over time. Export coefficients (loadings standardized by drainage area) were calculated to facilitate comparisons between the two sites.  2.4.4.1 Data Evaluation  Ideally, water quality samples should be collected during a storm which: (1) has a least 2.54 mm of rainfall; (2) occurs at least 72 hours following the last measurable rainfall; (3) has a rainfall between 50%> and 150% of the long-term mean for the site and (4) has a duration between 50%> and 150% of the long term mean for the site (USEPA 1992). Additionally, storm water samples are often evaluated based on storm hydrographs, which chart the hourly precipitation of the  51  storm, and indicate the time of sample collection on the chart. Unfortunately, detailed site-specific storm event precipitation information was not available for the exact site. Therefore, only the first and second criteria were used for data evaluation. The first criterion was evaluated based on rainfall gauge data collected during the sampling event. The second criterion was evaluated based on Environment Canada Atmospheric Services climatological station data.  2.4.4.2 Loadings  Stormwater monitoring programs are never able to sample every storm event. Typically, when calculating loadings, it is required to estimate the annual load from flow records available for the whole period and constituent concentrations available from a limited number of events (Marsalek 1991). Two loading calculation methods are summarized in El-Shaarawi et al. (1986) and are referred to as the direct average method and the modified direct average method in Marsalek (1991). El-Shaarawi et al. (1986) demonstrate that both methods are likely to give very similar results in the absence of correlation between event flow volumes and EMCs. If correlation exists, then the product may need to be corrected for potential bias by employing flow weighted EMCs. However, Athayde et al. (1983) and Harremoes (1988) report the independence of runoff event volumes and mean concentrations. In the direct average method, the annual load equals the number of runoff events times the mean event load. The mean event load is obtained by multiplying the mean event volume by the mean event EMC. This calculation requires detailed knowledge of the frequency distribution of storms and the measurement of water volume in each storm (Macdonald et al. 1997). The Modified Direct Average Method was employed in this study. The annual runoff load is estimated as: 52  (8)  L=RC  Where: L = annual load of a particular constituent; C = event mean concentration (EMC); and R = total runoff. Confidence intervals (C.I. 95%) are estimated as: C.I.= L±(C XR)  (9)  CJ  This method was advantageous over the direct average method for this study for two reasons: (1) the data collected represent partial storm events (SMCs) rather than EMCs and as such are not applicable on a per storm basis; and, (2) the method partially uncouples flow and concentrations by using precipitation data to calculate stormwater runoff volumes (Macdonald et al. 1997).  Calculation of Sampled Mean Concentrations (SMCs) Spearman's rank correlations between flow and water quality parameters were examined to determine whether SMCs needed to be corrected for flow (see Section 2.4.51 for calculation). Correlations greater than 0.80 were considered high, but correlations were also examined for those in the 0.50-0.80 range. SMCs were calculated from log domain statistics. Since many stormwater parameters have parent distributions that are log-normally distributed (Macdonald et al. 1997) and the use of the log-normal distribution has been recommended (Harremoes 1988, Athayde et al. 1983, Marsalek 1991), water quality data were log-transformed in order to achieve the least unbiased estimates of summary data (mean, standard deviations and confidence 53  intervals). The log-normal distribution occurs when the logarithms of the parameters are normally distributed. The log-normal mean (X ) for each parameter was calculated as follows (El-Shaarawi 1989): v, =lnx .  (10)  (  X = exp Y 1 r  v  s  2  ++  V  /2  /  A  (11)  J  Where: 7 = average of the variable yj (log transformed data); and S = sample variance of yj. 2  The 95% confidence intervals (C.I) were derived according to El-Shaarawi (1989):  CJ.,_  a05  =Zexp  st  + 7  y  V  i  n  2s:  (12)  y  n-\  Where: Z = standard normal variable; and S - sample standard deviation of yj.  Seasonal SMC concentrations were assumed to be the same as annual SMCs for the following reasons: (1) samples were not collected over the entire year; (2) a limited number of samples were collected from the interior site and seasonal patterns could not be assessed; (3) no  54  differences were found for wet and dry weather samples at the coastal site.  Total Runoff Volume Total annual and seasonal runoff volumes (m ) were calculated according to the following 3  equation:  R = PY C A J  j  (13)  j  Where: A = contributing area in m ; 2  j = land use type (impervious, pervious); P = annual (or seasonal) precipitation for 2001 in m; and C = volumetric runoff coefficient (Marsalek and Schroeter 1988). Since the coastal log yard area was small, the contributing area of the coastal site was determined by assessing drainage patterns on the site. For the interior site, it was assumed that all of the log yard areas eventually drained to the river. Only the log yards that were active in 2002 were included in the contributing area. The contributing area at both sites was measured by photogrammetrically transferring areas from 1:15,000 airphotos to 1:20,000 topographic maps. However, since the drainage area at the coastal site was small, it was also measured by walking the site due to error in calculating small areas using airphotos and maps. Results from walking the site were used in subsequent calculations rather than airphoto measurements due to greater precision. Permeable and impervious areas were estimated from these measurements. Precipitation data were obtained from nearest Environment Canada Atmospheric Environment  55  Service climatological measurement stations. Seasonal precipitation was assigned according to the calendar seasons, with the exception of snowfall. Since snowmelt does not necessarily occur in the same season as it falls, the timing and volume of snowmelt runoff was estimated by examining snow cover at the end of each season. This snow cover was then added to the next season's runoff. This was done only for the interior site, as any snowcover at the coastal site was melted at the end of each season. Runoff coefficients estimate the fraction of precipitation that flows away as stormwater, and are often described as a function of land use. The runoff coefficient is assumed to be a linear function of percent imperviousness (I), and can be calculated according to the following equation (Schueler 1987): C = 0.05+0.9(7)  (14)  This equation was derived from data collected during the Nationwide Urban Runoff Program (NURP) in the United States. Baseline runoff coefficients were established at both sites. Coefficients for pervious areas were then modified for rainfall and topography according to Stanley Associates Engineering Ltd. (1992). In areas with large annual rainfall, soil will remain moist and less runoff will infiltrate the soil. In dry areas, more runoff will be absorbed by the soil. Topography also has an influence on runoff volume. In very flat areas, overland runoff progresses slowly and has a longer period of time to infiltrate. For the coastal site, baseline runoff coefficients were estimated to be 0.95 for the impervious paved surfaces and 0.35 for the permeable areas (same as that used for railyards, American Society of Civil Engineers (ASCE) 1992). Ten percent of the runoff coefficient was added to the  56  permeable area coefficient (0.35) for high annual rainfall and 10% was added for the steep topography for a final coefficient of 0.42. As the entire contributing area of the interior site was permeable, one baseline runoff coefficient of 0.35 was assigned to the interior site. Five percent of the runoff coefficient was added for moderate annual rainfall, and 10% subsequently subtracted for the flat topography for a final coefficient of 0.33.  2.4.5  Relationships between Toxicological and Chemical Variables  Statistical analyses were conducted to explore the relationships between toxicological (C. dubia toxicity) and chemical variables and identify potential toxic constituents in log yard runoff. Statistical analyses were not conducted with Microtox® toxicological data as samples were centrifuged prior to testing. Centrifuging samples potentially removes contaminants from solution, and thus toxicity results do not necessarily correspond to water quality parameters. Analyses were conducted on two sets of data: (1) the interior and coastal data combined and (2) the subset of coastal data. Analyses were not conducted on the subset of interior data as there were not enough data points. Parameters that were consistently below detection limit in both coastal and interior samples (Sb, As, Be, Bi, Cd, Pb, Mo, Se) or only present in interior samples (Cr, Co, Li, Ni and V) were not included in analyses. Three types of statistical analyses were conducted: (1) Spearman's rank correlation analyses; (2) linear regression analyses and (3) principal components analyses (PCA) with regression. Correlation analyses were performed to determine the level of correlation among water quality parameters. Regressions were conducted to identify potential sample toxicants. PCA with regression was used as an aid in removing autocorrelation between variables prior to conducting regression analyses with the resulting principle components and sample toxicity. 57  For parameters which were measured both in the field and laboratory (pH and conductivity), laboratory parameters were used in statistical analyses as they were measured directly before toxicity testing.  2.4.5.1  Spearman's Rank Correlation  Analyses  As not all data conformed to a normal distribution, and since sample sizes were small, Spearman's rank correlation coefficients were calculated for a matrix of 24 water quality chemical parameters to determine the association between the independent variables used in the regression analyses. Spearman's rank correlation coefficients are calculated by ranking the data and applying the following equation (Zar 1996): n i=\  (15)  Where: dj = difference between the ranks of the two variables. Correlations above 0.80 were considered high.  2.4.5.2  Linear  Regression  Simple least-squares regression of twenty-three water quality parameters vs. C. dubia sample toxicity were conducted to identify potential sample toxicants. Since pH is known to affect the toxicity of numerous water quality parameters, and was found to be significantly different (p < 0.05) between the two sites, laboratory measured pH values (the average values of initial and final pH measurements) were added to the regression models as a second variable for tannins and lignins, DHA and metals (COD, BOD, conductivity, hardness, alkalinity and general ions were excluded from this analysis). Since hardness is known to affect metal toxicity, and was generally 58  higher at the interior site, it was also included as a second variable in metal analyses although it was not significantly different between the two sites. General ions such as calcium, magnesium, potassium, phosphorus and sodium were not included in this analysis. Unfortunately, detailed information on the volume of wood on site at the time of sampling was not available. Monthly volumes of logs on site were recorded by site operations, but this information was not specific enough to examine relationships between volume of wood on site and toxicity. The underlying assumptions of simple linear regression analysis include: (1) a linear model exists for the data; (2) The y values for any given x have a normal distribution; (3) The spread of these normal distributions are homogeneous; and, (4) a given y is independent from all other y values (Zar 1996). In order to obtain a linear model for the data, water quality parameters and total sample flow (independent (x) variables) were log transformed to obtain a linear relationship to sample toxicity (dependent (y) variable). C. dubia toxicity was expressed as percent survival in undiluted sample. Percent survival in undiluted sample was used instead of LC50 values as they more accurately correspond with the chemical analysis data. The linear coefficient of determination (r ) and standard error of the estimate (SEE) were calculated for each relationship between the x and y variables. Scatter-plots of the log transformed water quality parameters and the sample toxicity were plotted to identify potential outliers in the data. Sample residuals generated during the regression were plotted to examine normality of distribution. Since r values tend to increase when the slope of the regression line is 2  steep, or when there are wide gaps between data, the SEE was also used to compare among the regressions. Relationships with r values greater than 0.80 and comparatively low SEEs were 2  59  considered good.  2.4.5.3  Principle  Components  and Regression  Analyses  Highly correlated variables can cause logical and statistical problems in multivariate statistical analyses, creating unstable results (Tabachnick and Fidell 2001). Principal components analysis (PCA) is often used in conjunction with regression when significant multicollinearity exists in a data set. PCA is a statistical technique that calculates a few orthogonal (i.e., statistically unrelated) principal components (PCs) from the original data. The PCs describe the variation in the data. Variables that are correlated with one another but largely independent of other subsets of variables are combined into these PCs. Each of these variables has a loading factor for each PC that represents the correlation between that variable and the PC. Additionally, each sample receives a score for each PC based on its chemical composition. Various orthogonal rotations of the principal components can be conducted to simplify the PC data structure (Tabachnick and Fidell 2001). PCA with logged transformed variables and covariance matrices was conducted on 19 independent water quality variables (BOD, DO, pH, conductivity, alkalinity and hardness were excluded. Metals which were consistently below the analytical detection limit were also not included). Varimax rotation (an orthogonal rotation) of the PCs was conducted for the first few PCs that cumulatively explained greater than 95% of the variation in the data. Parameter loadings were used to interpret the composition of each PC. Parameters were considered correlated with the PC at loadings greater than or equal to 0.32 (Tabachnick and Fidell 2001). Sample scores were subsequently used in simple linear regression analysis with sample toxicity to examine relationships between the PCs and sample toxicity.  60  3  Results  Ten and three composite samples were collected from the coastal and interior sawmill sites, respectively, instead of the planned twelve samples per site. The coastal sawmill was permanently shut down in 2002. Only three samples were collected at the interior sawmill due to the lack of runoff events (Table 3.1). Complete data tables are provided in Appendix A.  3.1 Evaluation of Log Yard Stormwater Runoff Visual observations during sample collection and/or during laboratory analysis indicated some differences between coastal and interior runoff. As the coastal sawmill log yard area was largely paved, TSS from the coastal site runoff consisted primarily of wood and bark particles, as well as sawdust. Due to the unpaved nature of the interior log yard area, TSS consisted of soil as well as wood and bark particles. Coastal runoff ranged from a light amber to a dark tea in colour, while the interior runoff was a pale yellow. When samples were centrifuged to remove sample solids, the centrate of the coastal samples was usually (but not always) lighter in colour than the whole sample (light amber), while the centrate of the interior samples remained pale yellow.  3.1.1  Runoff Chemistry  Water quality parameters were compared to BC Aquatic Life Criteria and Guidelines (BC Ministry of Environment Lands and Parks (BCMELP) 1999a, 1999b) where possible. These criteria are ambient criteria (i.e., no dilution is expected) and are relevant to sample toxicity.  3.1.1.1 General Runoff Characteristics  COD in coastal runoff ranged from 121 - 1231 mg/L with a mean of 541 mg/L (Table 3.2). In interior runoff, COD ranged from 218 to 750 mg/L with a mean of 557 mg/L (Table 3.2, Figure 3.1). Only one of the interior samples (August 23, N-08/23) had a low COD (218 mg/L), the COD concentrations of the other two interior samples were above 700 mg/L.  61  0 C L |  E  co CO cu  >> 0 0  co  Q  CD  4>  5 5  f  f  0 0  D is cu  1  o cz  CO  05  Q  CO N CN C CN O  CO CO  Zi  CN o d  ^—'  T—  0  01  00  d  00  csi  s?  ZJ  o CL rz & o  1  CO DC  CD O rr' CD T - CN  +  g > cu  g 'co  o co d  Q  q d  CN CN  I'  ai  co  CN  O 4  O od  CO CO  O-l  CD ^ ui  2  00  cb  CD  > CO 0  00  CO  0 > £  E  CO  CO  c?  o  1  co T3 T3 cz C CZ  1_ ZJ  LO  ^  CN ©  Q  o  o  N  (D  °. o N  o  o  ^  C\i  00 °  *  aS rz >co £  o  CN  CO  so  g  JZ  CO X 3 — 0 CO o ZJ 0  cu E & o  0  JO  rz  E o  cz  "c  I  cz c CO CO  cz c c c cz cz cz cz cz CZ CO CO CO CO co CO CO CO CO CO  0 .£  CO  co  o  ^ CO ^= co £ CO  T3 o cr I  CO  CD  05  Oi  Oi  Oi  Oi  Oi  Oi  Oi  CD  Oi  Oi  Oi  o  O  co E T0 3 LAlEO i± 0 co  0 != 0 cz O CO ca O 0 5 c  o o  o 66 L O CO o LO o C N CSj CO LO csi d  0  —  0  "co Q  0 Q-l E co CO  00  oi  c  CO  o  LO  ^~  T—  T—  CO  0  O  T—  co  o o CN co LO ib LO o o LO o csi cb  LO  ZJ - 3  CT5  O  T—  co CN O)  T—  t—  x—  Oi  LO  CN  CN i— CO  ZJ  o CL  CO CN 45 00 <0 O CO o O  Oi CN CO  q  LO  CO ib in LO CN  LO CO  T—  T—  •  LO LO  o  CO  00  T—  CL  CO  CO  ~CN CN  o  O  CO CO W  >%  o  LO  o  LO LO  o  CO  o  csi o o LO o  cb oi  oi  —>  "o O  co o  5  °co > ~ ^.9^-g_ ~  CO  o  CO E CO _ O O CO CL /ri CO 0 ctiS a> CL ca co Q. o c "° E cz o o co o m  CN  <  o 0 cz 0 E 2 5 o, o «o !c5 o c>  CO (O  q  CSI CSI c z r z r z CZ JO ZJ ZJ ZJ  CO i t  ~  ni* O  T—  i  Z Z (/} z  q LO  LO CO  CN  CSI  q LO  o  LO LO  o  o  LO CO  d  T—  CSI  < <  <  LO  CO  q  CO CO CO co  CN  CN  CN  CO  00  O t—  o  O  CO CO CO  CO T3 ~ ~  X  co  o m i—  co  CM c co  0 0 "c  Q  CO  CD co in  -3-  CO GO O 1CM  r-  ^ ^  CN  P  o o  1^ co  CD CD  ^  CD  O  °  T  £  o  5  CM CD ^  O  O) CO  o  o  o  CM  |S  T3  "  £  "to  O  CD cn CM ^ CM  cs CM  00 CD  •3-  CN  o  o  in  1^ d co  CM O  00  a>  •3-  o  m  CN in  cn  CO  2  CJ CO CD CD  in o  CM  T-  cn  £  «  CN  cn CD  o  o c to co  co  -3-  V  CO  CD  l_  co  CD  "co  CD Q. E  in m  CD to  CD  _g CO CD in in  CD CM CO  co  2  o  r>-  ?  CD-  in  o  CO  in  0)  CO CN CN  n CD T-  CD  co  co  T-  co >  CN CD  "O CD i_  ZD to  CO  CO  CO  CO CO  CO CM  CO  CO CD  co  2  CD  CO 00 X CO  D co M C cn C CO 1^ CN C M CO CM CD CD o  <£  °  CN CN  o in in  co cb  CN  co in  CM co  c  CO CD CN  to  CD  o c oo  CM  LD CD  T3  in  CD  V-  ^  co  LD  ^ t CD cn  io  ^  !"  o  CD CD  CD  O  O)  CO  CN  w  d  d  tCN  CT>  in  CN  CD CO in  to  CD O Q  C/3  a  CO CD  0 3 CO CO  CM CO  i : £ CD S £ 5?  m  -  CO cb  Ln °> d g  3  cn  O  O  co O  co O  cn  a>  CD  cn o  O co  CD CO  m 0 3  o v  c  CD cn •Q to •4= in co o o -0  CO  VI  X  Q.  o  CD CD  cn  ^  o  T-  O CM  00  "co L_  E  co  o o cn to c u CD "3 CD C. T_ to to CD X CD  CO  E  _o  „—  C SD 3C ^ r-O C C CD o ^ <u .2 ^ CO 3 i l l •> ^ CD COI I S Q u o >, to c cT CD  —•  °  .o CO  O  CO E  CO E  cn E  "co  CO *  CD  0) 0) E  to to  CO  i ro Q-  E Q O O  Q O  CO  co  CD c  XI CO X X  X  O Q  oo O O l^ 13 3 O T3 O T3 C D C 0 C Q. O Q. O CO O  E to o co <u  V  >  CD .O * I— _C0  5  CD  o  _ o CO  jfi  d  C X  S  8.8  co  -r-  CD  cn co  .-~ -o to „  in  o m  o  ^ cj)  o  ab  O  o in  CD  CD E  CO O  C i_D  •o c  _ 2 "co  CO  i_  CD E  0  CO CO  < x Q  i t co .c?  o  o  £  = O " iS to CO X CO S CD ® -§ £ S _l "D  co^  0  •4—'  o  800  600 H  Q O  o 400 200 Coast.  Int.  Coast.  Int.  Coast.  Int.  Coast.  Int.  7.5 X Q.  7.0 H  6.5  Coast.  Int.  Coast.  Int.  Coast.  Int.  1500  > 1000  500  Figure 3.1: Comparison of mean general water quality characteristics from coastal and interior log yard stormwater runoff (error bars are standard deviations).  64  BOD in coastal and interior runoff ranged from 5 to 290 mg/L and 174 to 586 mg/L, respectively (Table 3.2, Figure 3.1). Note that for three coastal samples, BOD results did not have a DO uptake of at least 2 mg/L. BOD values were still calculated, but were considered estimates (Table A.3). These tests were not repeated as samples were at least five days old - past the recommended analysis date. Alkalinity of interior runoff was much greater than the coastal runoff, with interior runoff ranging from 63 to 157 mg/L CaC03 (mean of 108 mg/L CaC03) and coastal runoff ranging from 6 to 32 mg/L CaC03 (mean of 16 mg/L CaC03 ). The mean hardness concentration of interior runoff (180 mg/L CaC0 ) was also greater than in coastal runoff (85 mg/L C a C 0 3 ) , but 3  the range of concentrations varied widely, from 70-294 and 31-236 mg/L C a C 0 3 for interior and coastal runoff, respectively (Table 3.2, Figure 3.1). The field measured pH of coastal runoff was slightly acidic, ranging from 5.7 to 6.7, with a mean pH of 6.1. It is difficult to compare the field measured pH of the interior runoff due to the use of a less sensitive method, but generally pH was higher, around pH 7. Laboratory pH measurements taken within 5 days of sample collection ranged from 6.1 to 6.8 and 7.4 to 7.7 for coastal and interior runoff, respectively (Table 3.2, Figure 3.1). DO saturation levels were high (above 98% saturation) for all coastal samples. Only two field measurements were collected for the three interior samples due to a malfunctioning meter at the time of sampling. Sample N-03/13 was taken during a snowmelt event and had a lower DO (66%>), than the other sample (Table 3.2). Specific conductivity (25°C) field measurements for the coastal runoff were high, ranging from 259 to 2121 uS/cm, with a mean of 739 uS/cm. Only two field measurements were collected for the three interior samples due to a malfunctioning meter at the time of sampling. The two 65  measurements of specific conductivity taken for interior samples were lower, 100 and 290 uS/cm (Table 3.2, Figure 3.1). Laboratory conductivity measurements taken within 5 days of sample collection were slightly higher than the field measurements and ranged from 520 to 2550 and from 154 to 309 u.S/cm for coastal and interior runoff, respectively. It is unclear why the increase in conductivity occurred, as both laboratory and field instruments were calibrated directly prior to use. Changes in the sample matrix may have occurred during storage prior to toxicity testing. Field temperature measurements varied seasonally and ranged from 4.0 to 16.8°C in coastal runoff and approximately 2.0 to 19.0°C in interior runoff (Table 3.2). The mean TSS was higher at the interior site than at the coastal site, concentrations ranged from 123 to 1407 mg/L, with a mean of 699 mg/L. TSS at the coastal site ranged from 36 mg/L to 578 mg/L, with a mean of 240 mg/L (Table 3.2, Figure 3.1).  3.1.1.2 Tannins and Lignins  Tannin and lignin concentrations were higher at the coastal site than at the interior site, ranging from 45 to 263 mg/L, with a mean concentration of 136 mg/L. Concentrations at the interior site were lowest for Sample N-08/23 and ranged from 43 to 75 mg/L, with a mean concentration of 62 mg/L (Table 3.2). Results of the manganese and ferric chloride addition tests (see Section 2.3.1.1) indicate that when ferric and manganese chloride was added to oxygen saturated distilled water and tested for tannin and lignin concentrations, there was minimal interference by iron and manganese concentrations (overestimation of tannin and lignin concentrations) (Table 3.3). Therefore, tannin and lignin concentrations were not adjusted for iron and manganese interference.  66  Table  3.3: Iron  and manganese  interference  in tannin  and lignin  analytical  method.  Interior Parameter  Coastal  Min  Max  Mean  Min  Max  Mean  7.47  69.60  32.82  0.64  8.79  3.71  Interference for 5 m g / L F e C I ( m g / L )  0.31  0.31  0.31  0.31  0.31  0.31  T o t a l i n t e r f e r e n c e for s a m p l e ( m g / L )  0.47  4.34  2.05  0.04  0.55  0.23  0.530  2.190  1.600  0.239  1.450  0.764  0.10  0.10  0.10  0.10  0.10  0.10  0.01  0.05  0.03  0.00  0.03  0.02  Iron ( m g / L ) 2  Manganese (mg/L) Interference for 5 m g / L M n C I  2  (mg/L)  T o t a l i n t e r f e r e n c e for s a m p l e ( m g / L ) Note:  T o t a l i n t e r f e r e n c e ( m g / L ) is the c o n c e n t r a t i o n in t a n n i n a n d lignin e q u i v a l e n t s that iron o r m a n g a n e s e p r o d u c e s in t h e tannin a n d lignin c o l o u r i m e t r i c m e t h o d . S i n c e interference for both iron a n d m a n g a n e s e w a s m i n i m a l , tannin a n d lignin c o n c e n t r a t i o n s w e r e not a d j u s t e d for interference.  3.1.1.3 Dehydroabietic acid (DHA)  BC Aquatic Life Criteria for DHA in freshwater are pH dependant and range from 0.001 mg/L at pH 5.0 to 0.014 mg/L at pH 9.0. In the pH range found for coastal and interior runoff (6.0-8.0), criteria range from 0.002 to 0.013 mg/L. There is no criterion for marine aquatic life. The majority of samples exceeded 0.013 mg/L with the exception of one coastal and one interior sample that were below the detection limit (note that the detection limit is above the lowest criterion of 0.002 mg/L). Concentrations at the coastal site ranged from undetected (O.01 mg/L) to 1.32 mg/L, with a mean of 0.48 mg/L. Concentrations at the northern site ranged from undetected (<0.01 mg/L) to 2.89 mg/L, with a mean of 1.11 mg/L (Table 3.1). The highest DHA concentration (2.89 mg/L) was found in the snowmelt sample taken from the interior site  3.1.1.4 Metals  Metals present above analytical detection limits in coastal runoff include aluminum, barium, boron, copper, iron, lithium (one sample), manganese, silicon, strontium, titanium and zinc. The same metals were detected in interior runoff with the exception of boron and the addition of chromium, cobalt, nickel and vanadium. Most average metal concentrations were higher at the  67  interior site than at the coastal site (Table 3.4, Figure 3.2). A subset of metals for which BC Aquatic Life Criteria exist and detection limits were low enough to facilitate comparisons was compared to criteria (Table 3.5, Figure 3.3). Coastal metal concentrations were compared to both freshwater and marine criteria (due to marine receiving environment), while interior concentrations were compared only to freshwater criteria. Criteria used were maximum criteria (designed to protect aquatic life from short-term lethal effects), with the exception of beryllium, where only a chronic criterion (designed to protect aquatic life from long-term sublethal effects) exists. At the coastal site, aluminum, copper, iron, titanium and zinc met or exceeded existing freshwater criteria in some or all of the runoff samples (Table 3.5). Barium, copper, iron, manganese and zinc exceeded existing marine criteria. At the interior site, aluminum, chromium, cobalt, copper, iron, titanium and zinc concentrations met or exceeded criteria. In general, metals that exceeded freshwater criteria in both coastal and interior runoff were aluminum, copper, iron, titanium and zinc. Aluminum, iron and zinc consistently exceeded freshwater criteria in all coastal and interior samples, with the exception of zinc in coastal runoff sample S-06/02. Titanium and copper exceeded the freshwater criteria in all three interior runoff samples. Titanium exceeded the freshwater criterion in four coastal runoff samples (S-04/05, S05/04, S-06/11 and S-06/27). Copper also exceeded the freshwater criteria in four ofthe coastal samples (S-04/05, S-05/04, S-05/14 and S-06/27).  3.1.1.5 General Ions - Calcium, Magnesium, Potassium, Phosphorus and Sodium  The main differences between the coastal and interior sites were for sodium and calcium (Table 3.4, Figure 3.4). Calcium concentrations were higher at the interior site, ranging from 24.50 to 61.20 mg/L, with a mean of 46.43 mg/L. The mean calcium concentration in coastal runoff was 4 68  co  _i Q V  _l Q V  _J CM Q d V  5  Q  Q  Q  c to  Q Q V V _ i  ^—  Q V  o  o  XJ  0  Q  CO  00  co  00  CD  Q V  o  K  3  o  CO 1 _ l CD _ i 00 Q Q CO Q CO V V d V  d  _ i  Q V o  d  _ i  _ i  Q  Q  V  V  CO CO  o  Q v  _ j  v  _ i  d  Q Q V V  in i—  o  CM  d  d  CM _i CM Q d V  o  d  v  _ l  Q V o  d  co  v  _ l  Q V o  d  _ i  _ i  _ i  O  Q  Q  V  V  V  o in CM  o  d  CD o CO d in  CM  d  d in o  V  d  o  o  d  00  o  co oo CM o o cb d d o  Oi  CD  h -  Q V _ l  r- V o  XT  d  CM co oo o O o oi d d d  CO CO CO CO  o  CM  o  DL  X  _l Q V  o  CD O CM O CO  M oo C CO o CM Q d CO V  o  CO  2 o  o  d  cb  CO in  Q ^ _ l  Q V  o  .  CM CO O d °- 1=1 CM V d 0 0  CM O CM  00  LO T-  CM Oi  O O  CD  o  _ i  Q V  d  o  m o  Q V  CO  _ l  Q V  C O O ^  v  d  t  T -  d  d  in o  Oi  d  d  CM  O  r-  "  a _ i  Q V  _J Q -  v o  d  Q v  O  CO  q o o o o q o o o o o o o o q; q; q> q O q o cb cb cb cd cb cb cb cb cb cb cb cb cb cb CO cb cd cb cb cd cb cb o  k. CD C3  s X CO  <3 =\  o  in  .o  d  cu  XJ CD  C3 CJ  _l Q V  _ l  Q V  K  •2  C3  _ i  _l Q Q V V _ i  CM  Q  Q  o  V V  CM  CO  co QCO o  o  o  o  _ i  Q V  _ i  Q V _ l  Q V  2 9 CM _Ql d V  _l Q Q CM Q V V d V o  cz  CD CM Q Q V V  CO CD  co o  o o  _ l  _ i  Q  Q  cb V CM _ l  4  CM CO  Q V  V  _ i  Q V  _ J  _ l  Q  Q  V  V  CM  o  cb Q  Q  V V  S v  C O o d  _l V  T—  _J CD Q d V  _i Q V q  d  LO _ i CO Q cb V  o  CO CO  o o  CM  O  _ l  Q V o  d  in  o  O  »i  CD O Q d V  CN  s  C3 O  _ l  cb Q V  co Q d v  Q  o  Oi  Q 00  1^  Q  v  d  cb v  o  CD O in  d  d  o  d  Oi  Q V  Oi  T—  CD  ^1-  Q V  ^Q 1 d V  O  CO d  00  T -  -  d  _ J  _ l  D  D  v  v  Q  v  co Q v  v  CM  N  CD CM  v  Q ^o  O  CO  m  CO CM Q v  in ,_'  Q  Q  v  CD  O  d  "  Q  K  CO CJ  K O CJ  C3  Q V  o  o  o ^—  CO  o  o  o  o  o  o  o  IO  o  o  Oi  o  o  o  o  o  o  O  o  o  q  o  o  o  q  d  d  d  d  d  d  d  q  o  o  q  o  o  o  q  q  o  o  o  q  q  O  T—  IO  o  o  o  CM CM CM o d d d V V V  0>  C3  0 0  E  co CO  Q_  LO  o  Q d V  d V  d V  d o  d V  d o  d  d d V V  d  d V  d  d V  d  d  CO oi n o d d V  d o  d d V  JO  ro "53 EZJ ro  o h-  % <  E  E  cz  ZJ •—  cz  <  ^  CO  CQ  5= ZJ  E  o5 « CQ CQ  ZJ  ZJ CZ  o  O  CQ  E  "D CO  O  d  ZJ CO  O  I JZ  O  1 o O  E  CD  CX CX o  o  XJ  ro CD  ZJ JZ  ZJ CO  0  d in o  o d  0  CO  0  d  d  CO CO  ro  d  d  CO LO CM CO o o d d d CM d V V V CO ZJ  ZJ  cz  cz 0  cz cz CO  ro  d  XJ JZ!  o  2 0  o  CDCO O JZ  0_  CO CO  ro o  0.  E ZJ  'cz _0 0  CO  X  co  o  CO ai co o co  in  Q v  co  a  CO  CD o  c co  TJ CD  3  Q CO  co  CD  o o  CM CNJ CM  00 CO CM  in o  a v  CO  CM  d  aV  _l Q  _j Q  O CM  _i Q V  _l Q V  _l Q V  CO  in  CO  d  d  o  V O  CO  CD O  V  V  CO  aV  V  cp  d  •»—  o  CO CM  CO  CM  CO  d  0 o  d  d CD  a  m  .  °  Q  d  V  v  Q V  a> o  .O CO CD o  CO o  co  CO in  CM c o  O 0 0 TJ 0 >  Q V  o  CO  o  CO  p  CO  o  o  p  o  CO CO  CO  o  CO  o  CO  o  CO  CO  0 co  0  O 0 CO  O  d oo  CM  in  a V  Q  CO CO  CO T3 0  CO  o o o  a  CM  00 CM CM  V  c:  o  CM  CM  1^ O  V  _l Q V  _i  aV  _i Q V  _i Q V  _i  Q V _i Q V  in  _j  d  V  CM  CM O  d o  d  Q  _i Q V  _i Q V  CO 00  1  0  CO g  s >^ c ~'  CO CM  0 CO 0 co <: 0 0 o in 0  o  a>  o  it  11  O TJ  o  VI  -£  Q CO  co -4—'  CO CO  a> Q CO  CO  CD  a>  CD o  o  o  o O  co  CD  in  CM  a  V  in  oo oo o  _l  Q V  _l  Q V  i—  o  d O)  q d  in o  m  o  £  _1  Q V  TJ  co oo o  CO CO O 0  *±z ~~ c  Q V  o  ^  lo  o  o  o  o  o  O  o  d  d  d  d  .g>  o  o c  "5 CD £ 0  l | o  d  o  d  o  d  o  d  o  C TJ  d  ° s co  in o  d  o v  0 0  ico co  CL  c o o  CO  CM  d  E CD > CO  in o o  TJ o  CO  CM  CO  V  v  d  o  d  o  d  co o  in o  V  o  d  P  CO  .£ -I Q  o  co  O TJ C O" 0 C 0 JS 0  E c o i  .1 2 • 0  Q t o l m "  co  TJ CO c  CO  >  o c  N  0 —i a  a co  o  o  Coast.  Int.  Coast.  Int.  Coast.  Int.  Coast.  Int.  Coast.  Int.  Coast.  Int.  2.5 2.0 E  1-5 1.0 0.5  Coast.  Int.  0.4  0.3 E  0.2  0.1  Coast.  Figure  3.2:  stormwater  Int.  Coast.  Comparison  of mean  runoff  bars  (error  metal  concentrations  are standard  Int.  from  deviations).  71  coastal  and interior  log  yard  Table 3.5: Summary of metal criteria (or guideline) exceedances in coastal and interior log yard stormwater runoff No. of Exceedances BC Criteria - maximum (mg/L) Parameter (mg/L)  DL  DHA  Interior  Coastal  freshwater  marine  freshwater  marine  freshwater  0.002 at pH 6.0 0.013 at pH 8.0  nc  9  -  2  Total Metals Aluminum' '  <0.2  0.005-1  nc  10  -  3  Barium  <0.01  5  0.02  0  7  0  <0.005  0.0053 (chronic)  0.1  0  0  0  Boron  <0.1  nc  5  -  0  -  Chromium  <0.01  0.001  na  na  3  Cobalt  <0.01  0.0009  na  na  Copper  <0.01  Hardness dependant (see Fig 3-3)  0.003  Iron  <0.03  0.3  0.05  Lithium  <0.01  5  Manganese  <0.005  Molybdenum  1  Beryllium  (3)  2  (2)  P)  3  10  10  3  nc  0  0  0  Hardness dependant (see Fig 3-3)  0.1  0  10  0  <0.03  2  nc  0  -  0  Nickel  <0.05  Hardness dependant (see Fig 3-3)  0.075  <4>  0  0  Titanium  <0.01  0.1  nc  4  -  3  Vanadium  <0.03  nc  10  -  0  -  Zinc  <0.005  Hardness dependant (see Fig 3-3)  0.01  9  10  3  nc na DL 1  2  3  4  4  0  5  No criterion Applicable criteria were below the detection limit No comparison; no criterion available Detection Limit Canadian Council of Ministers of the Environment (CCME) guidelines The applicable criterion was below the detection limit for one sample Applicable criteria were below the detection limit for 5 samples. Of the remaining 5 samples which exceeded the detection limit; 5 exceeded the marine criterion and 4 exceed the freshwater criterion All concentrations were below the detection limit. In five samples, the applicable criterion was below the detection limit. The remaining 5 samples can be considered below the criterion.  72  0.14 0.12  1  °-  1  & 0.08 2 0.06 o 0.04 o 0.02  o  0 0.00  200.00  400.00  600.00  0.00  100.00  Hardness (mg/L) ©  o  I  —  J2 |  0.4 -1 O) 0.3 E. u 0.2 c N 0.1 -  0.00  100.00  200.00  300.00  0.00  400.00  100.00  •  Nickel Concentrations Criteria (Hardness A d j u s t e d ) Detection Limit  3.3: Freshwater log yard  criteria stormwater  200.00  300.00  400.00  Hardness (mg/L)  Hardness (mg/L)  and interior  Manganese Concentrations  0.5 -,  0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0  Figure  400.00  Criteria (Hardness A d j u s t e d )  Detection Limit  1  300.00  Copper Concentrations Criteria (Hardness A d j u s t e d  S  200.00  Hardness (mg/L)  comparisons runoff  Zinc c o n c e n t r a t i o n s  o ---  for  (circles  copper,  Criteria (Hardness A d j u s t e d )  manganese,  are coastal  samples)  73  samples,  nickel triangles  and zinc for are  interior  coastal  Coast.  Int.  Coast.  Int.  Coast.  Int.  Coast.  Int.  cn E  Coast.  Int.  Figure 3.4: Comparison of mean general ion concentrations from coastal and interior log yard stormwater runoff (error bars are standard deviations)  74  times lower (10.31 mg/L) and ranged from 4.44 to 26.00 mg/L. In contrast, sodium was higher in coastal runoff and ranged from 72 to 385 mg/L, with a mean of 152 mg/L. The mean sodium concentration in interior runoff was 30 times lower (5 mg/L), with a range of 3 to 7 mg/L.  3.1.2  Runoff Toxicity  3.1.2.1 Acute 48-hr Ceriodaphnia dubia Toxicity  Coastal runoff samples varied widely in toxicity with LC50s ranging from 32.95 to > 100% (v/v). Of the ten samples collected, half were non-toxic to C. dubia (LC50 > 100%). For interior runoff, toxicity was similar for two out of three samples (LC50s of 60.54 and 58.70). Sample N 08/23 was non-toxic (LC50 > 100%) (Table 3.6). In general, runoff was not highly acutely toxic to C. dubia. With the exception of sample S-05/04, all runoff samples had LC50s > 50% (v/v) log yard runoff. All samples, with the exception of S-03/29 and S-04/05 met test requirements for control mortality, DO, pH and temperature measurements. Sample S-03/29 testing was originally initiated April 1, 2001 but was re-tested on April 4, 2000 due to unusually high mortality in the 6.25%o exposure concentration. Sample S-04/05 was re-tested on April 10, 2001 after a control failure during the initial test conducted April 8, 2001. Sample S-08/21 and N-08/23 had to be "bubbled out" to reduce DO to below 100%). Results from reference toxicant tests (zinc sulphate) conducted by EVS indicated that brood cultures used for C. dubia tests were healthy (i.e., LC50s fell into the appropriate range).  3.1.2.2 Runoff Treatments with Ceriodaphnia dubia  Suspended Solids For coastal runoff, centrifugation reduced toxicity in three of four samples, toxicity of sample S06/11 was not improved by centrifugation (Table 3-7). Toxicity was reduced completely in 75  CO CN  co o  oo 0  1^ o  CO  o  o o  o o  t  A  O  co LO  LO  d CD  o o  CO  CN  CD o E  A  T—  CD 00  cn LO  CD  oo  CN  t-~oo co  d  cn  CN CD d 1^-  CN CN CN CN  LO  oo  1^  o  d  co  CO CO  LO LO  CN CN  CO CO CD CN  LO  oo  CD  CN  LO 1^ CN  CM 00 O  •<-  LO  d  T3  CD LO  X  CD CO  c o  c CD o c o o  K  s  "3CN  cn  CO  00  -•^  o  CN  CO  CO  CO 1^ CN CO o co  —>  •2  co CO  <0  CN g  0  CD .O  ro "53  CO  CO o o  T3  o  T—  d  d  c\i s-  LO  1  co CN CN  o CO LO  g g CO  cn CN CO  g CO  CN  CN CN  ^ co  CO LO  LO O)  cn CO  CD  c\i  CO CN  LO  "3"  o o  00 CO LO  /\  co  cn LO  CN 00  00 CO  CN  od  •sf  c  UJ  o _J  OH  1^ CN  X— LO  o  CO  CO  d  cn cn  •a  T3  "CS  CN CD  cn  1  o  CD  CO  cn 00 CD  O 3  CN d CN  1^  CN LO CO  LO  cn  CO CO CO  1  cn  CN  O O  1  '"  -  1  "5 >  CN LO  0  o 0  o  CO CD CO CO  T3 0 O  CO  o LO O LU  o o  T— A  CO T  T  CO  o  T—  XT  LO  cn  CO CO  CN LO  CO  cn  CO CN  cn  CO  o  1^  00 LO  CN  CN -3"  d  CN LO  LO  LO  LO  oq  CO  CD 00  CN  CN CN •tf  CN  00  CO CN CO  CO LO  CD CO CO CD CN LO  o  cn LO  csi  LO  T—  CO  CO  d  LO CN  o  LO LO  co  £  o  CN CN  co  T3  CO CD CD "CS i_D CD C ro co CO CO  0 0  CO to to o O Q. o CO  CO  'c ro ro CO  o  m  °  o  O 3  E  X  o o o  is  T3  LO  O  LO LO  T3 0  5 ro  LO  d  CO  0  1^  CO  co  0  •o O 0  E  ca g  1 §  0  E to o g 8 2 o co  ."5 ^ o  o 3  cz o CO to o to 0 o Q_ ° ° E O- o 3 2 o 2  to  co o £ 0  o  0  vfc LU  o  —J  LO LO _ J  O O o O JLU  J=  E  0  IO  o LU —  X  1  o  ro  T3  co  o  LO  o o  VO  T3 C  1  T—  I co  CO  o  LO  LO  o  CN CN CO  •<fr  o  CU  00  CO  A  A  LO  CO CO  T—  T—  o o d  LO  £ T3  LO CO  ~P O O  1  O  CO  •a c  CO  — f  O  o  O  LO  CN  IO  CN CO  *<—  CO  o o d  Z3  CD CO  A  LO  vo  LO  co  o  cn io  d  o o  CD .  LO  CN  o o  O o  o  •S  CD  d  CD  o CO  e  cu  cp  1  CO  o d  J  D  T3 C  < r-  Q LU _J  Q LU  ~cb 0  CL  cz q  ro co  x:  O < I— Q LU  o o  E ro c/)  LO  XJ  c  ro >  o"  CO CD  C CD  o o  o  5  O CM  <D CZ  o"  "0 CO  O  LO XJ  c  XJ  cz  XJ  oo ro  i_ £  co  o"  cz £S ro > '£  JZ)  ZJ  co  w  CD >  o  £  CM  CO  CO  zz  o O LO  CD >  £  X  ZJ CO  CL  C3) 1^  CO  00  CM  1^  CN  o  o> 0  CD I— 0 >  00  CD XJ 0 CO ZJ  X  U3  CL  LO  o  co CT) cb ib  CN  cz  0  CO  M—  cb  0  ro >  0  O  '£  0  ZJ CO  *E ro ro  > co  LO  o  CT)  iz ^  XJ 0  •c o  -—~  ro  Q.  >  0  rvi  C0§  CT) '—* CD  ZJ CO  CO 0  ro o  >  £  X  CL  LO  1^  oo XJ  CN  T-  o  ^  5? o  O 00  Q.  v  0  5^  oAo '5 x '—' >> o  X CL  ro 0 L—  c  ZJ CO  LO  T—•  T—  co  cb cb  CO  cb  XJ  0  0  >  o o o E X  XJ 0  o ZJ XJ  cz o o  0  >  0  o o cz E  * i ro ro  LO  LO  co  > CO  o  0  0  XJ XJ  c z cz o o  CO c  ro ro  CO o  0  co - K ro "cz "cz o _ O o z < CO ZJ  CO ZJ  RO  CO  co 0 c  XJ  co  i  XJ  ro X 0  E ro  co  CO  rio  CL  0  CO  o  o CM  CT)  q  O  Z  z  CO CD O  O  o CN  ^  Lo q  CO o  CM CO o  CN O  co  CO  CO  CO  q  T—  0  0  0 .>=  r--  sample N-07/19 and partially removed in sample N-03/13. In general, some reduction of toxicity was observed when suspended solids were removed. Attempts to create toxicity by introducing sample solids (from sample N-07/19) to culture water showed variable results, with toxicity occurring in one of two attempts at spiking culture water. In the first test, survival in the spiked dilution water was 0%, as in the untreated sample. Unfortunately, this result could not be repeated in a subsequent test, where solids spiked dilution water exhibited high survival rate (60%) as compared to the untreated sample (0% survival). Centrifuge control results did not indicate modification of toxicity through centrifugation and resuspension of sample solids. Survival in the centrifuge control was 20% and similar to survival in the untreated sample 10% (Table 3.8).  Table  3.8: Results  water  and centrifuge  of TSS additions control  results  from for  log yard C. dubia  stormwater toxicity  runoff  (Sample  N-07/19)  to  control  tests.  % M e a n Survival of C. dubia at 1 0 0 % (two replicates) Date Tested  Untreated Sample  Dilution W a t e r with S a m p l e T S S  Centrifuge Control  29-Jul-01  0  0  nd  05-Aug-01  0  60%  nd  14-Aug-01  10  nd  20%  nd  No data  For the limited COD soluble analyses conducted, removal of TSS removed approximately 62 to 81%o of COD for three coastal samples and 17% and 71% of COD in two interior samples. (Table 3.9).  In general, removal of sample solids improved  C. dubia  survival in some samples but not in  others, indicating that TSS itself is likely not the sample toxicant, but may be associated with the toxicant.  78  Table  3.9:  Summary  stormwater  of total  to soluble  COD,  tannins  and lignins  and DHA  in log  yard  runoff Samples  Parameter (mg/L)  S-06/27  S-08/21  S-10/24  N-07/19  N-08/23  Total C O D  629  12,1  349  703  218  Soluble C O D  239  22  129  205  180  T o t a l T a n n i n s a n d Lignins  113  43  Soluble Tannins and Lignins  87  18  Total D H A  0.42  Soluble D H A  0.34  Metals The addition of EDTA to undiluted sample S-10/24 caused an increase in toxicity from 60% survival in the untreated sample to 50, 30 and 20% survival at 20, 50 and 75 mg/L EDTA, respectively (Table 3.7). It is not clear what caused this effect. EDTA added to the control at 50 and 75 mg/L showed no toxicity to  C. dubia.  pH, temperature and DO saturation were well  within the parameters of the test. Conductivity between the baseline (605 to 748 pS/cm) and EDTA addition samples was similar (607 to 745 pS/cm). Unfortunately, the test could not be repeated due to a dissipation in sample toxicity. Sample observations noted that the sample was turbid (solids were suspended) prior to losing toxicity. After toxicity dissipated, solids appeared to be coagulated on the bottom of the sampling container. Note that the toxicity of the S-10/24 EDTA chelation test was slightly lower (60% survival) than in the original baseline toxicity test (40% survival), possibly due to the initiation of the EDTA chelation test eight days after runoff sampling. For sample N-07/19, the addition of EDTA to undiluted sample did not reduce toxicity at any of the three EDTA concentrations added (100, 200 and 300 mg/L) (Table 3.7). However, results from the 300 mg/L EDTA addition cannot be used due to the failure of the 300 mg/L EDTA control (20% survival). pH, DO and temperature were well within test parameters. Conductivity  79  between the baseline (234-295 uS/cm) and EDTA addition samples (262-354 uS/cm) was similar. Although test results were not completely conclusive due to potential confounding factors (increased toxicity with EDTA addition, failure of the 300 mg/L EDTA control), EDTA did not reduce toxicity at any concentration in either sample, indicating that metals are likely not the source of toxicity.  3.1.2.3 Microtox® Toxicity  Coastal log yard runoff samples 5 minute EC50s ranged from 27.13 to > 100% (v/v). Only one sample (S-06/02) was completely non-toxic to Microtox® (EC50 > 100%>). Fifteen minute EC50s were similar to 5 min EC50s, ranging from 24.98 to > 100% (Table 3.6). One sample S08/21 was re-tested as an EC50 could not be calculated due to invalid data points - the response in lower concentrations was higher than in the control. However, the re-test was also invalid for the same reason. Therefore, no Microtox® test results are available for this sample. Interior runoff samples 5 and 15 minute EC50s were similar with 5 minute EC50s ranging from 22.22 to > 100%. Control replicates were within 20%, suggesting acceptable pipetting and mixing techniques, however, confidence intervals often exceeded 30% of the EC50 in each case. Wide confidence intervals often suggest pipetting inconsistencies, however, they may also occur when the EC50 is not bracketed by two dilutions or when the dose-response relationship is not linear. With runoff samples, the EC50 was often high, and outside of concentrations tested by the Microtox® Basic Protocols (highest concentration of 45%).  3.1.2.4 Relationships between Microtox® and Ceriodaphnia dubia toxicity  Coastal log yard stormwater runoff was generally more toxic to Microtox® (both 5 and 15 80  minute tests) than to C. dubia with the exception of two samples (S-05/04, S-06/02). The relationship between the two toxicity tests for interior runoff samples was not as well defined, with Microtox® toxicity (both 5 and 15 minute tests) equal to C. dubia toxicity in one sample, greater than in one sample and less than in another (Table 3.6). This is reflected in Spearman's rank correlation coefficients, which indicate that Microtox® (5 minute test) and C. dubia toxicity were well correlated (0.80) for the coastal data, but less so for the combined data (0.56). Correlations between the Microtox® 15 minute test and C. dubia toxicity were similar to those for the 5 minute Microtox® test.  3.2 Statistical Comparisons Between Sites BOD, alkalinity, laboratory measured pH and metals (Al, Ba, Ca, Cu, Fe, Si, Ti, and Zn) were significantly higher (p < 0.05) at the interior site. Conductivity (laboratory measured) and sodium were significantly higher (p < 0.05) at the coastal site (Tables 3.2 and 3.4).  3.3 Seasonal Trends It was difficult to evaluate seasonal trends for runoff samples as the majority of coastal samples were collected in spring 2001 and due to the small number of samples collected from the interior site. In general, there did not appear to be any seasonal trends for coastal samples (Appendix B, Figure B.l). Alkalinity appeared to increase slightly towards the end of the year. Field measured pH values also seemed to increase slightly toward the end of the year in coastal samples, however, laboratory measured pH values did not follow this trend. The increase in field measured pH values may be due to a malfunctioning pH sensor. No significant differences (p < 0.05) were found for contaminant concentrations or toxicity in wet and dry weather samples collected at the coastal site indicating that concentrations are not linked to seasonal weather 81  patterns (Table 3.10). However, this may be an artifact of sampling due to the difficulty in obtaining samples during the first flush of a storm event and the lack of samples in the fall season. Seasonal trends were not assessed in the interior samples due to the lack of samples. However, as sample N-03/13 was the only sample collected during snowmelt rather than rainfall, qualitative comparisons were made with the other two interior log yard runoff samples. No specific trends were observed, however, DHA (2.89 mg/L) and BOD (586 mg/L) were much higher and DO was lower (66%) in the snowmelt sample (Appendix B, Figure B.l). The interior sample N 08/23 was higher in pH (laboratory measurement) but lower in most conventional parameters including TSS, COD, hardness, alkalinity and specific conductivity (laboratory measurement). Sample N-08/23 was also lower in tannins and lignins (43 mg/L), DHA (< 0.01 mg/L) and most metals (Appendix A, Table A.l). There were not enough samples to statistically assess differences between wet and dry weather events. Simple linear correlation coefficients (r ) and standard of error estimates (SEE) were calculated for: (1) total sample flow (m ) and 3  C. dubia  survival in 100% whole effluent; and (2) flow and  select water quality parameters. Note that a negative correlation with C.  dubia  survival indicates  a positive correlation with toxicity and data are discussed in this context. No significant correlations (p < 0.05) were found for either combined or coastal data, suggesting that flow was not directly related to toxicity or contaminant concentrations. However, when sampling events where the mill was not operating were removed from analysis for the coastal subset data (two events), the correlation between sample flow and toxicity was negative (i.e., high flow and low toxicity) and significant (p < 0.05) although the r was not particularly high (0.50) (Table 3.11). 2  Additionally, the correlation between sample flow and tannins and lignins, manganese and phosphorus was significant (p < 0.05) and negative (Table 3.12). 82  X  ro  03 ro  O  cn  •tf  LO  co  co cn OJ  CN  1^  CN CN  co  •<-co  LO  •<-  CN  CD  LO  T-  LO  CO  CD  o CN LO  o LO  CD d  LO  T—  ^ tM  CN CO  T-  co o CO  cn  CD  CD CD  LO o d  co d  o  ^  ro  o  LO d  o d  CN d  tf-  o d  CN d  CD  o d  o CN  LO  d  LO  CN d  CO -  CD  ^  IO LO  ™. '~  h  •tf  cn LO  CN O d  o d  o d  CO •tf  o  cn  CN  •tf  •tf  CD d  OO  cn •tf  CD  od  00 d  CD •<-  •tf  co  o LO  ,CN  cn co CN  CD O LO  d  o d  oo CN O  LO d  00 CN  CN d  oo  CD d  o o co  oo  cn  CN LO O  Q Q CO  c:  co CD  X  co  T-  LO  ^  0 0  CN CD  CD d  O CD CO  CN CO  CO CO CN  00 d  o LO LO CN  1-  N-tf  CD  o 00  •tf d  o  CO d  1^ cn  CD  LO CO  CD  ^  CO  °  CN  tf"  2  cn •tf  CN  LO  CO  co  S ^  CN  CO LO  O O CD  tf•tf  CN  cn  co co CM  O  CO  d  LO tf" d  •<- oo  CD O -r^  CO CD CN  2  LO  O  d  -*  -  *T d  CD •>-  cn  g d  r  ^  q d  g d  •tf  CN LO °  CD o d  IO d  o d  CN d  CO o  CN d  o q d CN  O d  o d  CO o d  tf- -CN d  o CD  LO  LO CN CN  LO d  LO  T  CO d  cn  CD  o LO •tf  oo  •tf  ^  LO  O d  tfd  LO  ,J  CD  CD CN  CO CD  o CN  °  LO  o  CO  CO  cn CN  •tf  •tf  d  co CO o  o CD O  o CN  CO tfo d  CN  00  1^  CO  o  -r. CO O  CM CO o  CO  •<-  CD  "co co 0 0  •<00 LO  Q CO  CM •tf CO  CO  CD CD CD  0  00 CN  T-  tf-  CD  CO  °o g ^  ^-  1^  o CN  d  <* 12  co  92  ••-  CD  tfd  CN  CO d  LO  CN O  «  CJO  •tf  1^  LO  ^  CD -  co o  «  cq d  q d  00 CO  oo  CN CO  o d  CO CO CM  LO c\i  oo CD  co  co c\i  o d  I—  00 d  o LO  •tf d  •tf  cn cn d  •tf  CM  o CO CM  LO  o 00 d  tf  1  ' ~ r  LO LO  J3 CO  -i—i  co  O  'c cn  T3 C o  "O  |>  0 0  13  CO  co CO  0_  Q Q O O O CD  CO  0 c: T3 iC_O I  O .Q ro I CL  o CO < 0 CO- CO x CO Q  cz  c CO co  Z3  £ c CO  E  E J3 <  c o i o  E  0 o.  o CL ro O ca ca O o  co  co  0  c cn ro  0 co 0  c CO cn  co Zi CL CO  O  Zi CO CO  E O  ZJ  E "c o  ro o T3 O 1_ •4—' O CO CO CO CL  X  CM  CSI  CO  CSI  CSI  CO CO  CD O  CM  o o  o o o  z co c CO CD  1_  T3 CD  LO  o o  1  Q Q CO  c z co CD  LO O  CSI CO  o d  o o d  CD O  CN  o LO o d d  CO -p co E -D 0 £  O  x CO  LO CS|  d  CO  CM  o  CO  CD 00  3  > > CO  CN  "o o  Z! O CD CD  c z o o 1c cz CO c o co CD **=  CO CM  o o o  co CD  rz  Q CO  c z co CD  5 .£ ro 2  CN  o  CO T3 CD  JZ  _ o B E I E  o o o CD o d  CL LO  1^  C)  I  AI  a3  CL LO  CM O  £  v ro  TT-  CD  >  CD  "S3  JZ  CO  0 _j .2  C^ D  .t; ro 5  | ^  IP" "P  "CD CO  co \  CL  CD CO  1co i3coE— E T3 c z CrOz o co cz co >  Sera  COD CO  coc  JS  "3-  oo  Table 3.11: Simple linear coefficients of determination for total sample flow vs. C. dubia toxicity from combined and coastal log yard stormwater runoff Source  DF  SS  MS  F  1  16  16  0.01  11  22,707  2,064  Regression  1  2,176  2,176  Residual  8  13,837  1,730  Combined  SEE  r  2  ( 1 )  Regression Residual  0.001 45.43  Coastal 1.26  0.136 41.59  C o a s t a l (operational d a y s ) Regression  1  4,918  4,918  Residual  6  4,832  805  6.11  0.504* 28.38  C o m b i n e d d a t a includes d a t a f r o m both t h e c o a s t a l a n d interior log yard sites Significant correlation (p < 0.05) Degrees Freedom S u m of Squares M e a n S q u a r e Error S t a n d a r d Error o f E s t i m a t e  ( 1 )  * DF SS MS SEE  Table 3.12: Significant simple linear coefficients of determination (p <0.05) between total flow and select water quality parameters from combined and coastal log yard stormwater runoff Coastal (Operational Days) Parameters T a n n i n s a n d Lignins  Phosphorus  Manganese  Note:  3.4 3.4.1  Source  DF  SS  MS  F  Regression  1  31449.26  31449.3  21.10  Residual  6  8,943  1490.5  Regression  1  0.476  0.476  Residual  6  0.284  0.047  Regression  1  0.826  0.826  Residual  6  0.729  0.122  SEE  r  2  0.78 38.61  10.06  0.63 0.22  6.80  0.53 0.35  S e l e c t w a t e r quality p a r a m e t e r s that c o r r e l a t e d well ( r > 0 . 8 0 ) with C. dubia toxicity w e r e r e g r e s s e d a g a i n s t s a m p l e flow for c o m b i n e d a n d c o a s t a l d a t a . T h e s e i n c l u d e d m a n g a n e s e for t h e c o m b i n e d d a t a a n d t a n n i n s a n d lignins, m a n g a n e s e , b a r i u m a n d p h o s p h o r u s f o r t h e c o a s t a l d a t a . T h e r e w e r e no significant correlations f o r t h e c o m b i n e d a n d c o a s t a l s t o r m w a t e r d a t a . 2  Calculation of Loadings for Coastal and Interior Sites Site Precipitation  Total precipitation at the coastal and interior sites for the year 2001 was 2435 and 1235 mm, respectively. Snow comprised approximately 13% of total precipitation for the coastal site and 60% for the interior site. As expected, precipitation was greatest in the fall in the coastal region 85  due to heavy rainfall. Precipitation was high in fall and winter months at the interior site due to snowfall and was also fairly high in the summer due to thunderstorms in the region (Table 3.13). Table 3.13: Summary of2001 precipitation for the coastal and interior log yard sites (data from Environment Canada Atmospheric Environment Service) Coastal Season  Rain  Snow  Spring  420.7  0  Summer  343.2  Fall Winter Total  3.4.2  Interior  %  Total  Rain  Snow  % (rain:snow)  Total  100:0  420.7  120.2  15  89:11  135.2  0  100:1  343.2  247.7  0  100:0  247.7  972.2  302  76:24  1274.2  86.4  392  18:82  478.4  377.3  20  95:5  397.3  36.4  337  10:90  373.4  2113.4  322  87:13  2435.4  490.7  744  40:60  1234.7  (rain:snow)  Data Evaluation  Due to the lack of site-specific precipitation data, data evaluation was difficult to conduct. For the coastal site, eight of ten sampling events had greater than or equal to 2.5 mm of rain fall on site during sampling. Two events S-06/02 and S-06/11 did not meet this criteria (only 2 mm rain fell on site during sampling). Only four events (S-04/05, S-04/22, S-05/14 and S-08/21) met the criteria of the 72 hour antecedent dry period, however, two of these events, S-05/14 and S-08/21, had small amounts of rainfall (< 2.0 mm) fall the day prior to sampling (Table 3.1). Of the three samples collected from the interior site, one was collected during snowmelt and two were collected after precipitation stopped but the runoff had just started flowing. At the interior site, low precipitation amounts and rainfall infiltration into the soil through unpaved surfaces, contributed to low runoff volumes. Often, runoff occurred after a few days of previous light rainfall or very heavy downpour. Therefore, rainfall gauge data for these sampling events was not collected. Additionally, neither of the samples met the criteria of a 72 hour antecedent dry period (Table 3.1). Due to the small number of samples collected at each site and the difficulty conducting sample evaluation, all the sample data were used when calculating loadings. 86  3.4.3 Loadings 3.4.3.1 Sampled Mean Concentrations  Stormwater parameters and flow were not highly correlated (Spearman's rank correlation coefficients < 0.80) for both combined and coastal datasets (Appendix C, Tables C.l and C.2) and thus SMC values did not need to be corrected for flow. SMCs were calculated for only those stormwater parameters that have the potential to accumulate in the environment over time (Table 3.14). Table  3.14:  interior  Summary  log yard  of Sample  stormwater  Mean  Concentrations  (SMC) for  contaminants  Coastal Parameter (mg/L) TSS  in coastal  and  runoff Interior  SMC  SD  LCL  UCL  SMC  SD  LCL  UCL  267  324  99  718  985  1853  20  48936  *  *  *  DHA  0.89  2.817  0.03  31.34  37.76  T&L  140.6  105.2  85.6  230.9  63.4  19.1  45.3  88.8  2.8  2.4  1.5  5.0  22.6  33.1  1.5  330.8  Barium  0.03  0.04  0.01  0.08  0.26  0.27  0.06  1.24  Copper  0.01  0.01  0.01  0.02  0.09  0.04  0.05  0.15  Iron  3.99  3.90  1.95  8.17  41.64  65.55  2.06  841.82  Manganese  0.785  0.557  0.494  1.248  1.854  1.770  0.472  7.284  Nickel  <0.05  0.00  0.03  0.03  0.06  0.06  0.01  0.28  Silicon  4.65  3.18  2.98  7.24  25.56  17.61  10.93  59.80  Strontium  0.087  0.059  0.057  0.135  0.110  0.051  0.065  0.186  Titanium  0.10  0.09  0.05  0.19  1.42  3.27  0.01  236.81  0.090  0.071  0.053  0.153  0.273  0.171  0.128  0.580  Aluminum  Zinc * SD LCL UCL  S D a n d C o n f i d e n c e Intervals a r e unreliable d u e t o l a r g e variation in s a m p l e c o n c e n t r a t i o n s . Standard Deviation L o w e r C o n f i d e n c e Limit ( 9 5 % ) U p p e r C o n f i d e n c e Limit ( 9 5 % )  3.4.3.2 Runoff Volumes  Annual runoff volumes at the coastal and interior sites were approximated to be 17,236 and 73,137 m , respectively. The runoff volume approximated for the interior site was only four 3  times greater than that approximated for the coastal site, despite the drainage area being approximately 20 fold greater at the interior site. This is partially due to the lower precipitation at 87  the interior site, but also due to the flat topography and permeable nature of the interior log yard. As expected, the largest runoff at the coastal site is expected to occur in the fall months. For the interior site, the largest runoff was calculated to occur by the end of the winter season (March) due to winter and fall snowfall and subsequent melt (Table 3.15). Table 3.15: Summary of total and seasonal runoff volumes of log yard stormwater from the coastal and interior sites. Impervious Area Seasons Coastal  Pervious Area  Drainage Area ( m )  Precipitation (m)  Area (m )  Runoff coefficient  Area (m )  Runoff coefficient  8922  2.4354  6283  0.95  2639  0.42  2  2  2  Total % Runoff ( m ) Runoff 3  17,236  Spring  0.4207  2,977  17  Summer  0.3432  2,429  14  Fall  1.2742  9,018  52  Winter  0.3973  2,812  16  Interior  170775  1.2347  na  na  1  7  f °  0.33  73,137  Spring  0.1352  8,009  11  Summer  0.2477  14,673  20  Fall  0.3284  ( 1 )  19,453  27  Winter  0.5234  ( 1 )  31,004  42  na ( 1 )  N o i m p e r v i o u s a r e a s in c o n t r i b u t i n g d r a i n a g e a r e a . A d j u s t e d for r e m a i n i n g s n o w c o v e r at t h e e n d of t h e s e a s o n  3.4.3.3 Loadings (Export Coefficients)  Annual Export Coefficients (ECs, loadings standardized by size, kg/year/ha) rather than loadings were compared between sites as the interior site was much larger than the coastal site (Table 3.16). The TSS EC was slightly higher at the coastal site, the tannin and lignin EC was approximated to be 11 fold higher at the coastal site and DHA was approximately 9 fold greater at the interior site. Aluminum, barium, copper, iron, silicon and titanium concentrations ranged from 1.4 to 3.0 fold greater at the interior site than at the coastal site. Manganese (2 fold), 88  strontium (3.8 fold) and zinc (1.6 fold) export coefficients were greater at the coastal site. Seasonally, 42% of these loadings will occur by the end of the winter at the interior site and 52% will occur in the fall at the coastal site (Table 3.15). Table 3.16: Summary of select stormwater runoff export coefficients for contaminants from the coastal and interior log yard sites Coastal Parameter  Interior  Export Coefficients (kg/year/ha)  LCL  UCL  Export C o e f f i c i e n t s (kg/year/ha)  LCL  UCL  TSS  5149  1911  13877  4011  81  199360  DHA  17.22  0.49  605.42  153.83  *  *  T&L  2716  1654  4460  258.3  184  362  Aluminum  53.2  29.2  96.9  92.2  6.3  1347.6  Barium  0.55  0.20  1.53  1.07  0.23  5.04  Copper  0.25  0.14  0.45  0.35  0.21  0.59  Iron  77.07  37.64  157.83  169.63  8.39  3429.48  Manganese  15.170  9.545  24.109  7.554  1.923  29.676  Silicon  89.79  57.66  139.83  104.13  44.51  243.62  Strontium  1.688  1.093  2.608  0.448  0.266  0.757  Titanium  1.93  0.99  3.75  5.81  0.03  964.74  1.735  1.017  2.960  1.111  0.523  2.362  Zinc LCL UCL *  3.5 3.5.1  L o w e r C o n f i d e n c e Limit U p p e r C o n f i d e n c e Limit C o n f i d e n c e limits are e x t r e m e l y large d u e to large variation in s a m p l e c o n c e n t r a t i o n s  Relationships between Toxicological and Chemical Variables Spearman's Rank Correlation Analyses  In general, water quality data (both sites combined and coastal subset) were highly correlated, with numerous correlations greater than 0.80 (Appendix C, Tables C l and C.2). For the combined data, hardness was associated with COD, calcium and magnesium, as well as other metals, particularly strontium. TSS was correlated with metals (particularly Al, Fe, Mn, Si, Ti and Zn) and phosphorus. BOD was correlated with hardness and metals (Al, Ba, Ca, Fe, Sr  89  and Ti). COD was correlated with DHA , metals (K and Sr) and phosphorus. DHA was well correlated with COD, phosphorus and potassium. Tannins and lignins were not particularly well correlated with other contaminants, with the exception of potassium (0.88). pH was not highly correlated with most contaminants, the strongest correlations were negative correlations with laboratory conductivity (-0.68), tannins and lignins (-0.74), boron (-0.71) and sodium (-0.68). Metals showed high correlations amongst each other, with exception of sodium and boron. Sodium and boron were correlated together, but not well correlated (or were negatively correlated) with other metals. Both were highly correlated with laboratory conductivity. Both sodium and boron were largely associated with coastal log yard runoff. For the coastal data, hardness was associated with calcium, magnesium and strontium, but also with COD, conductivity, sodium, phosphorus and potassium (ions present in seawater). TSS was still associated with metals but also tannins and lignins. BOD was correlated with tannins and lignins, phosphorus and metals (Al, Ba and Ca). COD was correlated with numerous contaminants including hardness, DHA, tannins and lignins, conductivity and most metals. DHA was correlated with tannins and lignins, phosphorus and metals (Al, Ba, Ca, Mn and K). Tannins and lignins were highly correlated with most metals (> 0.80) and phosphorus, but less so with boron, copper, magnesium and sodium. Laboratory pH was negatively associated with all other contaminants, although not strongly. Metals were highly correlated amongst each other. In general, the main difference between the combined and coastal analysis was the number of strong correlations between contaminants and tannins and lignins for the coastal analysis including correlations with TSS, BOD, COD, DHA, phosphorus and metals.  3.5.2 Linear Regression Simple linear correlation coefficients (r ) and standard of error estimates (SEE) were calculated 2  90  for a subset of water quality parameters and C. dubia survival in 100% whole effluent. Analyses were conducted for both the combined data and the subset of coastal data. Note that the negative correlation of a contaminant with C. dubia survival indicates a positive correlation with toxicity (i.e., high contaminant concentration and high toxicity) and data are discussed in this context. Results from the combined data set indicate that a number of parameters were positively correlated with toxicity (Table 3.17). Manganese had the highest correlation with C. dubia toxicity (0.81) and lowest SEE (19.7). Results from the coastal data set are different, potentially indicating the role of different toxicants. Tannins and lignins, barium and manganese were positively correlated with toxicity, had the highest r values and the lowest SEE (Table 3.17). Of 2  these variables, tannins and lignins had the highest r (0.92) and the lowest SEE. 2  The addition of pH to the combined data regression models was significant (p < 0.05) for a number of contaminants including tannins and lignins, aluminum, barium, copper, iron, manganese, silicon, titanium and zinc. However, regression coefficients only increased to above 0.80 for barium (0.86) and manganese (0.91), of which manganese already correlated well with 2  2  toxicity without the addition of pH (r =0.88). The addition of pH increased the r for barium from 0.50 to 0.86. The addition of pH did not significantly (p < 0.05) improve regression models for any of the parameters for the subset of coastal data, likely due to minimal pH variability within the coastal data (Table 3.18). The addition of hardness was significant (p < 0.05) for boron (r = 0.61) and copper (r =0.58) for 2  2  the combined data and for strontium for the coastal subset (r =0.72). However, regression 2  coefficients were below 0.80 for all (Table 3.18).  91  tf-  CD CD  1^  LO  LU LU CO  CD  CO  csi  oo tf-  oo  CO  tf  1  CSI CN  oo  o  o  CO  tf•tf LO  o  CO CO  •tf tfLO  o"  i -  00  tf00 CO  OO CD  tfio"  csi o"  00 CSI  o  co  00 CD 00  CO  S  LO CO  LO  o  oo  1^- o  CM 00  ib  cb  , -  OO tf"  CSI  tf"  CM  CSI  0 0  CD CO  tfLO  tf-  oo LO 00  tf-  o  LO  tf-  CO  00  T-  00  LO CSI  OO  q  CO  °.  tf"  •tf  CM  00  LO  oo  OO CO 00 CO  CM"  LO  -  tf-  ^  » g £ !  C  O  T-  •tf  o  C30  o  CD O  CD  o oo  0O CO CD  CO  o  CO-  tf-  oo d  CD CM CM CD  CO CO  00  00  d  CO CSI  oo •tf  cb  CO  LO CD  CD  CO  o 0O LO CO  LO"  i -  d  CD LO  CO 00  oo  CM  tf- ;±  00  oo  d  CSI  LO CSI  CSI  00  ib  LO  LO •>-..  <° oo_  00 LO  ib  coo  CD  CSI  O CM  tf-  •tf  I -  LO"  00  CSI  LO  L_  CSJ  oo d  CD CD  CM  CO  CO CO  00 00  CD  CD  oo  co  ib  CD  co  CSJ  CSI  tf  •tf  00  cb  CD CSJ  CSI  tf-  coo  CD  00  t-  00  CD CD  d  <N  os  LU LU CO  o  o tfCM  CD  g  CD LO  o cb  CM  cb CM  CM  CSJ  o d  00 CM  CD  C7)  CSI  CD LO  CO  •tf  CO  CM CM  !Q  E o  o  co  CO CO CO  co  co  LO  tf-  I -  •tf oo"  oo  LO CD CD  00 LO  •tf LO  CO  CO CO  co CO ~  tf  LO  oo  CO CD  tf-"  LO  S  tfoo"  o  CO 00 00 CO  1^  •tf CM CO IO  CO CO ^  oo  o  CD CM  LO 00  •tf co  tf-  00  ^  CT>  o 1  o co  0 "53 co  CL  cz g CO CO CD  l_  CO CD  DC  CO  g CO CD  DC  cz g CO CO CD  l_  CO  DC CD  CZ CO ZJ  g  CO CD  DC  q  CO CO  CO CO  g  0 0 1 DC cn 0 CO  DC  Q O O  CO ZJ  08  h-  cz g co  CO  Zi  cz q co  CO ZJ  g co  cz g CO CO  co ZJ  g  cz g CO CO  "co ZJ TO-  cz o CO CO  "co ZJ TJ  o  s COO LO  tf-"  T— T—  T—  CD  LO  cz o CO CO  "co ZJ TO  cz o co CO  "co Z5 TO  c o  CO CO  CO ZJ  g 0 0 0 0 0 0 0 0 co 0 Icn 0 0 0 C 0 l_ 0 CO DC 0 cn DC 0 C 1—O DC DC co DC 0 DC CO DC 0O DC 0 0 0 DC DC DC DC CO 0 0 DC DU 0 DC CO  -g  CO  CO  CO  L_  CO  CO  L_  CO  k_  CO  L  DC  co CQ  co O  CO  CO CO  CO  co  X !  CZ  co LU LU CO  CO  o  co  oci  O  O  CN  1^  CN O  co CO o  o  1^  CSI  co  CO  cz CO  CO CO  o  T-  o  LO  o  CO  o  "o  oo co  0 0  X !  g  CO CO  CO CO  o  CO CN CO •tf"  o 0 o  T  o  CO  o  0  LO"  I—  CD g 0 00  T-  00  CO CO  CO  CO  05  0  CL  &| ro Lo  I  I  I  co £ co c— "0 £ "a r •t; co CD-o.b  I  X I  LU LU  3  CO  -  o g -2 c: 0 o 0) J 0)  CO  C  [-  0  ^ co ^  IN >  T3 CD  c o  O  ro  5o  ro CO  I  I  I  I  I  1  o O) c o < o 0  co co  _Q  ro - iz 0  — Oco — '0 CL co . ro  14 ^  I  X I  O  CO  0  ro S c0o  J3 X I  CO  0  CD O  c o CO  ZJ  0  i_  o co  CO I— CO  0  CC  CO ZJ "O CO  0  CC  o CO CO  0 L_ CD  0  Cd  £  "CD  E co  X I CO ZJ  u  CO  0  or  O ^ CO  0  p  cz  N  LU  2 'cz .£  (D  CO — £ 2 o ro LUro ^ _co zz k_ Z J crxi ro !^ LL CT to lo "Q> o co W c x> 0 ^ $ o ro £ obE (D CO CO o P i 0 Z J 2 CO 0 0  ?a  I!  Q CO  CO  CL  E O  cz  X3 CO 3 CZ O  I- O  LU I LL CO CO LU oS Q CO ^ CO I—;  CO  0  Table 3.18: Summary of significant (p <0.05) pH and hardness additions to coastal and combined data pH Additions Combined Parameter  r  2  Hardness Additions Combined  ( 1 )  SEE  T&L  0.73  24.58  Al  0.77  23.03  Ba  0.86  18.01  r  B Cu  0.49  33.88  Fe  0.72  25.02  Mn  0.91  14.63  Si  0.74  24.35  2  Coastal  < 1 )  SEE  0.61  29.2  0.58  31.04  Sr  r  2  0.72  Ti  0.68  26.92  Zn  0.67  27.52  SEE  25.43  C o m b i n e d d a t a includes d a t a f r o m both t h e c o a s t a l a n d interior log yard sites  3.5.3 3.5.3.1  Principle Components and Regression Analyses Combined  Data  Principal Component 1 (Comp. 1) was associated (loadings > 0.320) with metals, particularly aluminum, barium, iron and titanium. Comp. 2 was almost entirely negatively associated with DHA (-0.938). Comp. 3 was largely associated with sodium (0.675), but also tannins and lignins and boron (Table 3.19). Comp. 2 was the only component significantly (p < 0.05) correlated with toxicity (Table 3.20) in 2  2  linear regression analysis, but the r was low (r =0.44). However, Figure 3.5 shows that the correlation between toxicity and Comp. 2 is strongly influenced by two outliers, samples N 08/23 and S-08/21. Both of these samples had below detection limit DHA concentrations. When these two points were removed from analysis, correlation was not significant (r =0.19, Table 3.20). 94  Table 3.19: Rotated principal component (PC) loadings between PCs and C. dubia sample toxicity for combined and coastal log yard stormwater runoff Combined Parameter  Comp.1  COD TSS  0.269  DHA  -0.124  Coastal  ( 1 )  Comp.2  Comp.3  -0.174  0.146 0.255  Comp.1  Comp.3  -0.217  0.133  0.457  -0.938  T&L  Comp.2  -0.926 0.32  0.185  0.124  Al  0.321  Ba  0.351  B  -0.126  Cu  0.317  Ca  0.224  Fe  0.363  Mn  0.113  Mg  0.119  0.309  0.479  Na  -0.28  0.675  0.435  0.162  0.189  0.15  0.266  P  0.323 -0.137  0.267 0.252 0.329 0.374  -0.179  -0.2  Si  0.276  Sr  0.127  Ti  0.36  0.377  Zn  0.233  0.304  T&L ( 1 )  Comp Note:  -0.15  0.336  -0.156  K  0.341  0.186  0.272 0.246  0.448  T a n n i n s a n d Lignins C o m b i n e d d a t a includes d a t a f r o m both t h e coastal a n d interior log y a r d sites R o t a t e d Principal C o m p o n e n t L o a d i n g s g r e a t e r t h a n 0.32 a r e c o n s i d e r e d c o r r e l a t e d with t h e c o m p o n e n t  95  -0.108  in  CM tf-  co  CO CM  Q  "ro H—'  co ro o  m CO  tf; CM  r-  CD CD  O  CO CO  in CD  T—  O IO  CO CO  m m o q d  •tf  CM 1^ CD CD  CM CD d  CM CO  •tf  CO CT>  in  oo CO  ro  •tf  co  ro o o  oo io  CM CD Cvi  uri  JS oo  d  CO CO CO CO  LU LU  co o  CO  CM O  d *  CO CD  1^  o m Oi •tf  CD O CM  CO  co  co •tf  d  oo  CD  CM  CO 00  co  CT>  oo oo oo  tf-  CO CM  CM 00  o  q  ch XJ  1^  CO  cq  oo  oo  oo oo  d  c ro  tfo  co  -o 0 c  LO OO  CM  CO  •<-  O O  CO  CO CM CO O  o  •tf •tf  CM  d  LU LU CO  LL.  ro ro Q  LO CM d  oo co CD LO  m  CM  CD  CM  o  co CO  CM CO d  oo  •tf  CO CO  LO io  O0 0O LO tf" O0  CO  oo in co CM  o in •tf  ro o  °s  I0 t  CO V- tn  C\i  oo OO CO oo CM in co T— in LO cq CD co CO d in g CD •tf in co o LO !Q LO LO C30 T— LO CD C30 E tfm o O CD  E oo  CO  co co q d co CO oo 00  CM CO d  -o  " c\i  o tf;  o o tf-  o  •tf d CO  T3  0 CO  TZJ " °  ro E to £  CM  oo  00  •tf d  00 00 CM  CM CO  o  ro  8 0) ° 0 0 C5O  LO  in oo in  —  tri  CM  .Q  —  in  o  ro -*  0 ro roVI "D £ Q. CO "co 0 w  c  cz  o o ro o 0 o CO ro CO 1_  ZJ  o  CO  0  Zi  •g co  CO  -g  co co  CC  cr T3  E o O  E o O  CO  d  E o o  — i i ro U_ CO  CM  iO C  ol  E o O O  O ,_ LU  CT-D co- CO  co  cbE  0  m  ZJ  8o  CO  "ZJ  o  t 0 CO LU 00  co  Rei  pon d.  CM  E o  «O  E  • ^  co  ~u CO  0 0 0 cr coo  0  0  E o O  CO CO  •g  CO 0 0 0 l_ CC sCD 0 0 CO l_ CC cr 0 0 CO cr CC 0 CO  cz o  CO 13  AO  tn "cz  CO  Zi  CZ X i O ZJ  £ 2  ro Q ro -cz  TZ>  ^ TT3j S  CO  0  ro  CO ?) <D CO  2 CO  . _  CO O  co  o  Q .  E  ro  CO  ZJ  LU Q CO LL. CO CO LU  „  _  Combined Data  Outliers Removed  R2 = 0.44  R2 = 0.19  o o  o o  o  o  00  CD >  00  CO  o  >  CO  c/) CD  Id  o CO  ZJ  CO  CO  o -tf  b  o  o -tf  o CM  CN  o o o  O  -2 Comp.2  -1  o 0  Comp.2  Figure 3.5: Presence of outliers in the correlation between toxicity and DHA dominated Principal Component 2 (Comp. 2) of the combined PCA. Toxicity and DHA concentrations decrease along the y and x axes  3.5.3.2 Coastal Data  Comp.l was largely associated with TSS and metals, particularly aluminum, barium, iron and titanium. Comp. 2 was almost entirely negatively associated with DHA (-0.926) while Comp. 3 was associated with general ions calcium, magnesium and sodium as well as strontium (Table 3.19). Comp. 1 was the only component significantly correlated (p < 0.05) with toxicity (Table 3.20) in linear regression analysis, however, the r was still quite low (r =0.42). Since the correlation 2  2  between Comp. 1 and toxicity was positive, then toxicity increased with increasing concentrations of TSS, aluminum, barium, iron and titanium. As for the combined data, the correlation between Comp. 2 and toxicity (although not significant) in linear regression analysis was strongly influenced by one point, sample S-08/21 (Figure 3.6), which had a below detection 97  limit DHA concentration. When this point was removed, correlation between Comp. 2 and toxicity decreased from r =0.25 to 0.08 (Table 3.20). 2  Overall, for both the combined data and the subset of coastal data, metals such as aluminum, barium, iron and titanium were correlated and varied together. DHA, for both the combined and coastal data, was not associated with other variables. General ions, such as sodium, magnesium and calcium varied together for the coastal data. PCA did not identify any strong contenders for the toxicant to C. dubia as none of the principal components correlated strongly with toxicity. High concentrations of TSS, aluminum, barium, iron and titanium were weakly implicated in toxicity at the coastal site, but not with the combined data.  Coastal Data  Outliers R e m o v e d  R2 = 0.25  R2 = 0.08  O O  o  oo  CD >  CO  >  o  CO ZJ  ZJ  CD  CO CD  CO  !o •o 6  JU ZJ  6  o tf  ZJ  o  CN  Figure 3.6: Presence of outliers in the correlation between toxicity and DHA dominated (negative relationship) Principal Component 2 (Comp. 2) of the coastal PCA. Toxicity and DHA concentrations decrease along the y and x axes  98  4  Discussion  4.1 Comparison of Water Quality between Sites It is difficult to determine differences in water quality between the coastal and interior sites due to the small number of samples collected from the interior site. In general, the mean of most water quality parameters was higher at the interior site with the exception of tannins and lignins, conductivity, and sodium, which were present in higher concentrations at the coastal site. Mean magnesium, phosphorus and potassium concentrations were similar between sites. Mean BOD was significantly higher (p < 0.05) in interior runoff (326 mg/L) than in coastal runoff (134 mg/L). The BOD for one sample in particular (snowmelt BOD sample N-03/13) was high (586 mg/L) compared to the other coastal and interior samples. The snowmelt sample also had an elevated DHA (2.89 mg/L) and lower DO content (66%) compared to all other rainfall samples, coastal and interior. Mean alkalinity was significantly higher (p < 0.05) in interior runoff (108 mg/L) than in coastal runoff (16 mg/L) likely due to calcium carbonates commonly present in soils (Manahan 2000), particularly interior soils, which have a higher potential to reduce acidity than coastal soils (Hall et al. 1991). Subsequently, the significantly (p < 0.05) higher pH (7.4 to 7.7, laboratory measurements) in the interior samples may be due the greater capacity of interior samples to buffer rainwater and neutralize organic acids present in solution. Unpolluted rainfall usually has a slightly acidic pH of 5.5-6.0 due to the solution of carbon dioxide (Hall et al. 1991). The pH values for coastal log yard runoff samples (ranging from 6.1-6.8, laboratory measurements) were lower than for interior runoff, but also higher than pH values for unpolluted rainfall. Mean conductivity was significantly higher at the coastal site (1049uS/cm, laboratory measurement) due to the proximity of the site to a saltwater bay and the storage of logs in the  99  bay. Mean conductivity at the interior site was 223 uS/cm (laboratory measurement). Mean sodium was also significantly (p < 0.05) higher at the coastal site (152 mg/L) than at the interior site (5 mg/L). Sodium is a major contributing factor to conductivity, as well as its complimentary anion (chloride). Since the most common minerals in soil are composed of oxygen, silicon, aluminum, iron, calcium, sodium, potassium and magnesium, as well as manganese and titanium oxides (Manahan 2000), it is not surprising that stormwater runoff from the interior site, which has unpaved log yards, had significantly higher (p < 0.05) concentrations of aluminum, calcium, iron, silicon, and titanium. Average magnesium and potassium concentrations (common ions in both soil and seawater) were similar between sites, likely from seawater inputs at the coastal site and soil sources at the interior site. Average phosphorus concentrations were also similar between sites. Phosphorus is present as a macronutrient in soil and is also present in small amounts in seawater, as well as in plants (in this case, logs stored on site). Mean copper, barium and zinc concentrations were significantly higher (p < 0.05) at the interior site (0.08, 0.22 and 0.258 mg/L, respectively) than at the coastal site (0.01, 0.03 and 0.086 mg/L, respectively). These are naturally found in soils, but other sources can include metal alloys, rubber, electrical equipment and galvanized roofing. Some metals were found at concentrations above the detection limit at one site, but were below analytical detection limits at the other. Boron was found to be present above the detection limit in coastal samples but not interior samples. Since boron is found in soil, this suggests that interior concentrations should be higher. However, other sources of boron include glass, fiberglass, washing products, alloys and metals, wood treatment chemicals, insecticides and microbiocides 100  (Woods 1994). Chromium, cobalt, lithium (with the exception of one measurement at the coastal site), nickel and vanadium were only detected at the interior site. These metals are naturally found in soils, but also in vehicle sources, metal alloys and electroplating, batteries, oil and grease contaminants and high temperature lubricants. Despite these differences in water quality, there were no statistically significant differences in toxicity (either Microtox® EC50 or C.  dubia  LC50 endpoints) between sites.  4.2 Relationships between Microtox® and Ceriodaphnia dubia toxicity Toussaint et al. (1995) and Choi and Meier (2001), considered the Microtox® assay to have an acceptable overall sensitivity when compared to the standard acute tests. In this study, Microtox® was found to be more sensitive to log yard runoff than  C. dubia.  However, the  sensitivity of a test organism is largely driven by the type of contaminant. Choi and Meier (2001) found that for metal plating wastewater, the Microtox® assay was not as sensitive as the acute and chronic cladoceran tests (Daphnia  magna  and C.  dubia).  Data summarized in Toussaint  et al.  (1995) and Nelson and Roline (1998) also indicate the 5 minute Microtox® assay is less sensitive than the C. In this study, C.  dubia  dubia  48-hr test to zinc, cadmium and copper, phenol, malathion and 2,4,-D.  were tested with the whole sample, while the Microtox® bacteria were  tested with centrifuged sample (particulates were removed) according to basic protocols. This means the Microtox EC50s are reflective of the soluble portion of the runoff. Despite this, C. dubia  and Microtox® toxicity was correlated for coastal runoff data (0.80), but less so for the  combined data (0.56). Choi and Meier (2001) found that the 7-day chronic C.  dubia  correlated with the 30 minute Microtox® test (0.74) for metal plating wastewater.  101  test was well  4.3 Comparisons to Other Log Yard Runoff Studies Results from this study were compared to several other studies found in the literature (Table 4.1). BOD, COD and TSS results from this study were in the low range of values found for other sites. Conductivity at the coastal site was much higher than that found by Taylor (1994) due to the proximity to an ocean bay (the study conducted by Taylor was in the BC interior). DHA concentrations were lower than results reported by Zenaitis and Duff (2002). Tannin and lignin concentrations were in the middle range of values found for other studies. The toxicity to C. dubia  from log yard stormwater in this study was in the range of toxicity reported by Taylor  (1994) for another cladoceran, Daphnia  magna.  Microtox® toxicity was lower than toxicity  reported by Zenaitis and Duff (2002), but in the range reported by AFPA (1999).  4.4 Relationships between Toxicological and Chemical Variables A weight of evidence approach, examining results from treatments, regression analyses and criteria comparisons, was adopted to explore the relationships between chemical variables and the toxic response of C.  4.4.1  dubia  to log yard runoff.  Sample Treatments  Treatment results were inconclusive, and did not identify a specific contaminant as the sole cause of log yard runoff toxicity. Results from TSS removal and EDTA chelation were similar for both the coastal and interior sites.  102  3 to S<  -s:  o  ro -a co cz co  OO  o  to  _l  _  CSI CO  o  O  co  o  CD  CO  ro ^ "a  oo  O  LO  c  0  CO  < O  E CN LO CO  CD  cz  0  R 3  c o ^_ > cz LU  o CO LO  O  to  <3 ii)  i»  < co  C3  Q_ OO LL CO  bo •5  LO CN  CO  1^-  CM V  CO  3? ^3 S  o LO LO CM I  CN CO I  o  CO  CO  CO  o  o o  cn o  00  CC <  CO  •tf o' v  tf-  CL  O o  -3  J?  1^-  co tf-  -tf  O  >,co oo ro  I  CO CO  o CO LO CO I CO  co tf-  o o  •tf  I  A  CO  tf-  JZ o 1 03  0 0  CO  LO  O  cc c  0  £  CO CO  ro 92  cy  s a  CO  LO  o  <  CO  T3 CZ  CN  ro o  £8  'ro 3= c ZJ 0 Q N  s s v.  o  •tf CN  o CO  CO  CO CN 00  OO 00  1  £  o  LO  0  1 CO •tf  c  S  1  cz o  o  CM 00 ^  cd  o  o  I  o  CO  I  o  CM 1 Q  tf-  00  d  LO 1  T—  CN  0 >  CN  CO"  o  OO CN  O o  _ CO  o tf-  co  • CO CN  O O CN CN CN CN  a. o o  o o  00 LO  CO  co CO  •2 c»  CO  ro o  O  co CN I LO tf"  o CO CN  CO CN  CM  O V  co  CN CN i CO LO CN  CO  LO CO CO  O O i CO  o cz o  CO  0  CJ  cz 'c  cz co < CL LU  0  CO  OJ  0  c\i CO  CO  CO  0  o a> T—  LO 00 CM  0  0  5 o o  A  LO  cz  O "O  o  o  cz o  0  V  CM CO  CO  LO  0 0 CO CN  LO P0 LO  o CL  LO  oo  -*^»  CO  ro  I  CO  0  _ZJ  ro >  z ^ -a 0  -o  0  sz = ro  ,_ > 0 J3> TJ 0 CL 0 ro CL 0  ZJ CO  co  0  i  O CZ ZJ  to  O  CO  E, o  e o  .to  I a  0 0  E ro  ro D_  E  < I Q  c  -g o <  oo  o  cz ro CO  c 'cz cz ro H  CO  o  cz  l_  -o >^ .c  0  D  E E  co  E  Q  03  © X  o ZJ  -a cz o  O O O  I  Q.  CO  co  o o o  oo tfco c cn co  .co  5  -c  Q  —  0  CO _ZJ  S roro >  0  LO  co  £  ro >-  "3  0  !o  CL  C  UJ  co  = E S  O o LO  CO  T3  LO  r-  (0  s  "§ ro  Zi  ro » CO X I ro ro ro S co t 3 ro ro 0 ro J3 "O o ro " D CZ - r r ro Z Q CZ Z J .2 O ro co L L £  E  Removal of TSS from runoff samples reduced or removed toxicity in all but one coastal sample. Of the five remaining samples treated, the removal of TSS completely removed toxicity in one interior sample, almost completely in two coastal samples (survival > 90%), and partially removed toxicity in one interior and one coastal sample. Although EDTA addition results were confounded for one sample by an increase (rather than a reduction) in toxicity, results do not indicate that a divalent cation(s) is the toxic substance. It is possible that TSS can cause toxicity through physical means (i.e., suspended solids are themselves physically harmful to C. dubia), particularly in a test situation with small test containers (30 mL). However, Weltens et al. (2000) established that test concentrations of 250 to 500 mg/L uncontaminated particles did not cause Daphnia magna mortality within 48 hours. However, concentrations greater than 500 mg/L were not tested. In this study, two coastal log yard runoff samples exceeded 500 mg/L of suspended solids, as did two interior runoff samples. Additionally, as C. dubia neonates are smaller than Daphnia magna neonates used by Weltens et al. (2000), they may be more physically affected by particulates. However, since removal of TSS did not completely remove toxicity in all treated samples, TSS does not appear to be the sole cause of toxicity. This seems to indicate that physical interference by TSS to C. dubia is minimal. In general, it is assumed that the dissolved fraction of a toxic substance is responsible for toxicity to aquatic organisms and toxic substances bound to TSS are not considered biologically available to organisms. However, toxic substances bound to TSS can be available upon ingestion (USEPA 1991). A study by Weltens et al. (2000) investigating the potential adverse effects to filter feeding organisms (Daphnia magna) of exposure to contaminated particles indicated that contaminated particles have toxic potency by acting as a source of dissolved xenobiotics, but also by becoming available within the body of particle feeding organisms. 104  The toxicant (or toxicants) present in log yard runoff may exist both in solution or in particulate form with  C. dubia  exposed to the toxicant through both the particulate and soluble phases. TSS  may also affect the rate of toxicant transfer if the toxicant is adsorbed onto the suspended solids. Removal of TSS from the sample may remove a large sink of the particulate associated toxicant(s), while leaving some of the toxic substance in solution. For example, centrifugation of log yard stormwater runoff reduced COD by 17% to 81% and tannin and lignin concentrations by 20 to 70%. Total metal concentrations in several runoff samples were above maximum freshwater criteria for several metals (Table 3.5), however, the addition of EDTA did not remove toxicity. EDTA is a strong chelating agent and its addition to water solutions produces relatively non-toxic complexes with metals. The metals typically complexed by EDTA are aluminum, barium, cadmium, cobalt, copper, iron, lead, manganese, nickel, strontium and lead. EDTA has been shown to reduce the toxicity to  C. dubia  due to copper, cadmium, zinc, manganese, lead and  nickel in both dilution water and effluents (USEPA 1991). Note that copper (ranging from < 0.01 to 0.12 mg/L in coastal and interior samples) and zinc (ranging from 0.023 to 0.423 mg/L) were found above freshwater criteria levels in log yard stormwater runoff. Since EDTA did not reduce toxicity, metals able to complex with EDTA may already have been complexed (and thus exhibit reduced toxicity) by metal - organic complexes. Humic substances, which are thought to be derived from lignins, are known to be good complexing agents (Schevchenko and Bailey 1996, Pandey et al. 2000), however, EDTA is likely a stronger complexing agent. Knezovich et al. (1981) found that although both EDTA and humic matter reduced the toxicity of copper (40 pg/L to embryos of the Pacific Oyster (Crassostrea  gigas),  EDTA reduced toxicity at a higher copper concentration (100 pg/L). However, metal binding properties of humic substances and EDTA are likely dependent on the specific metal ion and 105  characteristics of the specific substances involved (Pandey et al. 2000, USEPA 1991). Toxicity may have been caused by a metal for which EDTA does not reduce toxicity (i.e., a metal that is not chelated by EDTA). For example, EDTA does not seem to reduce the toxicity of silver, aluminum or chromium when tested using moderately hard water (80-100 mg/L CaCOs) and C. dubia (USEPA 1991). In this study, aluminum concentrations were above the freshwater criterion for all coastal and interior samples. Chromium exceeded the criterion only in the interior samples (note that the chromium detection limit was too high to compare the criterion for coastal samples). Detection limits for silver were too high for criterion comparisons. It is unclear why addition of EDTA appeared to increase toxicity in sample S-10/24. Excess EDTA in solution becomes toxic when present above a certain concentration. At a hardness of 40-48 mg/L CaC0 , the LC50 of EDTA to C. dubia is 110 mg/L (USEPA 1991). The EDTA 3  concentration range added to sample S-10/24 (hardness of 40 mg/L CaCOs) was well below this, at 25 to 75 mg/L. Additionally, dilution water controls had 100% survival at all three concentrations of EDTA.  4.4.2  Regression Analyses  Strong correlations between chemical variables and toxicological response can indicate potential relationships between variables, but cannot always be interpreted as causal (Zar 1996). Also, it is difficult to interpret the results of the simple linear regression analyses and PCA with regression due to: (1) numerous correlations and interactions between the independent water quality variables; and (2) the large number of water quality variables that exhibited a high degree of correlation with toxic response of C. dubia. However, the analyses of the combined data set (coastal and interior) versus the coastal data subset indicate the role of different toxicants at the two sites (Table 4.2).  106  Table 4.2: Summary table of significant linear coefficients of determination (p <0.05 and r > 0.80) for contaminants vs. C. dubia toxicity in log yard stormwater runoff. 2  Combined Parameter  Coastal  ( 1 )  SEE  r  SEE  r  T&L  -  -  12.78  0.92  Ba  -  -  14.86  0.89  Mn  19.70  0.81  15.58  0.88  -  -  19.56  0.81  2  2  Linear Regressions  P SEE ( 1 )  S t a n d a r d Error o f E s t i m a t e C o m b i n e d d a t a includes d a t a f r o m both t h e c o a s t a l a n d interior log y a r d sites C o r r e l a t i o n coefficients b e l o w 0.80  For combined data, simple linear regression analysis identified manganese as a potential toxicant (r > than 0.80). PCA on combined data with subsequent regression analysis did not indicate strong relationships between any of the principal components and toxicity. For coastal data, simple linear regression analysis indicated that manganese was still highly correlated with toxicity, but so were tannins and lignins, barium, and phosphorus (r > than 0.80). 2  All of these variables were positively correlated with each other (Table 4.3), although TSS was less correlated with manganese and phosphorus. PCA with regression analysis did not identify strong contenders for potential toxicants. A weak relationship (r = 0.42) was identified with 2  toxicity and TSS, aluminum, barium, iron and titanium. Of these contaminants, barium was the only contaminant that strongly correlated (r > 0.80) with toxicity in simple linear regression 2  analyses. In general, the differences in simple linear regression results for combined and coastal data indicate the potential role of different toxicants at each site. Manganese was identified as a potential toxicant at both sites, however, other contaminants (particularly tannins and lignins, barium and phosphorus) also exhibited strong correlations with toxicity when just the coastal data was examined. 107  Table  4.3: Spearman's  contaminants  highly  rank  correlation  correlated  (r  2  TSS  4.4.3  coefficients  > than  between  0.80) with  C. dubia  TL  Ba  coastal  log yard  stormwater  runoff  toxicity.  Mn  TSS  1.0  TL  0.85  1.0  Ba  0.81  0.91  1.0  Mn  0.66  0.88  0.86  1.0  P  0.73  0.94  0.89  0.94  P  1.0  Criteria  DHA was above criteria in all samples where it was detected. Metals that exceeded freshwater criteria in both coastal and interior samples included aluminum, copper, iron, titanium and zinc, none of which correlated well with toxicity. Manganese, which did correlate well with toxicity, did not exceed freshwater criteria in either coastal or interior samples. Barium, which correlated with toxicity in coastal runoff, also did not exceed freshwater criteria (Table 3.5). Other metals may be responsible for toxicity, however, high detection limits and the limited number of samples from interior site make this impossible to assess. There are no acute criteria for tannins and lignins or phosphorus, however, it is unlikely that phosphorus (a nutrient) is the toxic agent.  4.4.4  Weight of Evidence  Regression analyses indicate that manganese was correlated with toxicity for both combined (r = 0.81) and coastal data (r = 0.88). However, manganese was below the freshwater criteria in both 2  coastal and interior stormwater runoff (Table 3.5). Additionally, chelation with EDTA, which typically complexes with manganese (USEPA 1991), did not reduce toxicity. Tannins and lignins (r = 0.92), as well as barium (r = 0.89) and P (r = 0.81) were also 2  2  2  correlated with toxicity for the coastal data. Barium, like manganese, was below the freshwater criterion in both coastal and interior stormwater runoff (Table 3.5). Additionally, treatment with  108  EDTA, which is also thought to complex with barium (USEPA 1991), did not reduce toxicity. Assuming that phosphorus (a nutrient) is not the toxic agent, tannins and lignins may be implicated in toxicity for coastal log yard stormwater runoff. TSS was also implicated in toxicity in treatment results, possibly as a source of dissolved xenobiotics or through particulate ingestion. Since: (1) TSS was correlated with tannins and lignins; and (2) results from the limited solubility analyses indicate that tannins and lignins were only partially removed with solids removal, it is likely that tannins and lignins are present both in particulate and soluble form. This corresponds with the only partial removal of toxicity achieved by removing solids in some of the treated samples. In addition, metals may also be bound with TSS (indicated by high correlations with TSS), however, no dissolved metals analyses were conducted. There is not enough evidence to implicate one particular substance as the toxic agent. Tannins and lignins have been implicated as potential toxic substances from coastal log yard areas (Bailey et al. 1999). They may play a role in log yard runoff toxicity, particularly where western hemlock, a coastal species with a high tannin content is processed (Samis et al. 1999). However, other potentially toxic contaminants that were not measured (other resin acids, tropolones) may be responsible.  4.5 Factors affecting Log Yard Stormwater Toxicity The pH of water is known to modify the toxicity of numerous substances, including metals, tannins and lignins and DHA. Since the pH of log yard runoff was significantly higher at the interior site, it may have modified sample toxicity. Hardness is also a potential modifier of toxicity, particularly for metals. Changes in pH affect the proportion of metals existing as free aquo ions (conventionally 109  regarded as the most bioavailable fraction), thus increasing toxicity. In certain sawmill runoff samples, higher pH (pH 8.5) ameliorated the toxicity to rainbow trout (suspected to be caused by tannins and lignins) until tannin and lignin concentrations exceeded approximately 50 mg/L (Bailey et al. 1999). This is consistent with the effects of pH on weak acids such as tannins and lignins (Bailey et al. 1999). DHA toxicity is also potentially modified by pH. For example, BC criteria are based on pH, with higher DHA criteria set for higher pH. Zanella (1983) found that increasing pH decreased the toxicity of DHA (i.e. increased the LC50) to Daphnia  magna,  fathead minnow and bluegill sunfish by decreasing the concentration of free acid. However, Borga et  al.  (1996) observed that the relative toxicity of dehydroabietic acid to Daphnia  magna  was six times higher at pH 9 than at pH 7, and 15 times higher than at pH 6.5. Hardness (a measure of mainly calcium and magnesium ions, but also iron, manganese and strontium) can also affect toxicity of certain substance by: (1) competing with toxic metals and other substances for binding sites on biological receptor sites, and (2) competing for binding sites on organic ligands, thus potentially increasing the free concentration (bioavailability) of other substances (Manahan 2000). Many metal criteria are based on hardness concentrations (BCMELP 1999a). The addition of pH to the combined data significantly (p < 0.05) improved the regression for tannins and lignins and numerous metals. However, improvements in regression were greatest for barium, where the r increased from 0.50 to 0.86. The improvement of regression fit by the 2  addition of hardness to the metals data was significant for boron and copper in the combined analysis and for strontium in the coastal analysis, but r values were below 0.80. 2  Humic substances present in solution can also modify stormwater toxicity. Since humic substances are known to interact with metals, and since lignins are thought to be a precursor to  110  humus (Manahan 2000, Schevchenko and Bailey 1996), lignins may also bind metals and potentially moderate toxicity. Oikari and Kukkonen (1990) found that dissolved organic carbon can moderate the effects of DHA on Daphnia  magna,  actually causing a decrease in  bioaccumulation of DHA. There are also indications that the presence of metal ions may reduce the toxic effects of wood extractives. Peters et al. (1976) observed that heartwood tropolones from western redcedar were rendered acutely non-toxic to juvenile coho salmon by iron concentrations of > 0.43 mg/L. High rainfall volumes can dilute runoff contaminant concentrations thus reducing contaminant concentrations and toxicity. While for the combined and coastal data, flow did not correlate well with toxicity or contaminant concentrations, for the subset of coastal data, when the two nonoperational days were removed from the regression analysis, flow and toxicity were significantly negatively correlated (p < 0.05, Table 3-11). Additionally, the relationships between flow and tannins and lignins, manganese and phosphorus were significant and negative (p < 0.05, Table 312). This suggests that high flow does reduce toxicity and contaminant concentrations, but also suggests that other factors are important. For example, the amount of vehicle traffic and wood processing occurring prior to and on site during the storm event. Observationally, it was noticed that during the two coastal events where the sawmill was not operating, runoff was low in TSS and a clear light tea colour. Unfortunately, relationships between volume of wood on site and toxicity could not be examined due to a lack of detailed information on log volume during sampling events. However, clearly, wood volume stored on site will affect the concentrations of wood extractives present in the runoff. Ill  The tree species stored on site may also affect toxicity. Western hemlock and Douglas-fir were predominantly processed at the coastal site, while white spruce, subalpine fir and lodgepole pine were processed at the interior site. Western hemlock has a high tannin content (Samis et al. 1999), perhaps accounting for the relatively high tannin and lignin concentrations (45-263 mg/L) found in the coastal log yard runoff. Resin acids constitute as much as 48% of Douglas-fir resins and only between 30 and 40% of pine and spruce resins (Samis et al. 1999). DHA was not different between sites, however, other resin acids were not measured. Lignin concentrations in the wood species processed at both sites are similar, approximately 27.2, 28.8 and 29.4%> for lodgepole pine, subalpine fir and white spruce and approximately 27.7 and 29.4% in Douglas-fir and western hemlock (Isenberg 1980).  4.6 Effects from Log Yard Stormwater 4.6.1  Acute Effects  Stormwater runoff from either site was not particularly toxic. Half of the coastal samples had no acute toxicity to C. dubia and only one sample had an LC50 below 40%. Microtox® 5 min EC50's were not particularly toxic at either coastal or interior site, with EC50s above 20%> for all samples. Since tannin and lignin concentrations were relatively high (ranging from 45 to 263 mg/L) at the coastal site, and metals and DHA concentrations exceeded criteria at both sites (Table 3.5), there is potential for acutely toxic effects in the receiving environment. Metal concentrations were generally higher at the interior site than at the coastal site, increasing the potential for toxic effects (Figure 3.4). Zinc - which has been linked to sawmill runoff toxicity by Bailey et al. (1999) - exceeded the criteria in runoff from both sites but was three fold higher in interior runoff. DHA has been linked to toxicity in several studies (Leach and Thakore 1976, Borga et al. 1996) as have tannins and lignins (Bailey et al. 1999, Field et al. 1988). However, since the 112  runoff was not highly toxic, high dilution ratios from river or ocean water would reduce the size of the potential effects zone. Assuming that no seasonal differences in average contaminant concentrations exist (see Section 3.3) acute effects would be most likely to occur in the fall at the coastal site and in late winter at the interior site (periods of high runoff volume and snowmelt).  4.6.2  Contaminant Loadings  Contaminants such as TSS, organic compounds (DHA, tannins and lignins) and metals can cause long-term cumulative effects if they accumulate in sediments in the aquatic environment. Although the interior site was approximately 20 fold larger than the coastal site, the unpaved log yard area and smaller precipitation volumes led to a relatively modest (4-fold) increase in estimated runoff and subsequently loadings as compared to the coastal site (i.e., more runoff was generated per square meter at the coastal site). When loadings were standardized per area (export coefficients), export coefficients were similar between sites with the exception of tannins and lignins (11-fold higher at the coastal site). This suggests that the continual introduction of small runoff volumes with elevated contaminant concentrations may result in high export coefficients to the receiving environment potentially leading to cumulative effects. Export coefficients from both sites were compared to coefficients obtained from urban runoff studies for available parameters - TSS, copper, manganese and zinc (Table 4.4). Relatively large export coefficients for TSS at both sites suggest that log yards contribute to problems with debris entering aquatic habitats. Compared to urban stormwater runoff export coefficients of TSS, loadings from log yard areas were approximately 4 to 8 fold higher. Metal loadings at both sites suggest that stormwater runoff from log yards contribute to aquatic metal loads. Total copper and zinc inputs from the log yards were similar to urban export coefficients, but total manganese  113  from coastal sites and interior sites were approximately 15 and 7 kg/year/ha, much higher than the average street runoff export coefficients of 0.911 kg/year/ha calculated for the Brunette Watershed area (Macdonald et al. 1997). Table 4.4-.Comparison of export coefficients from urban runoff studies to log yard runoff export coefficients  (kg/year/ha) Urban Runoff  Parameter  Stanley 1992  ( 1 )  NURP  ( 1 )  Log Yard Runoff  Macdonald et al. 1997  ( 2 )  Coastal  Interior  TSS  952  640  995  5149  4010  Total Copper  0.27  0.15  0.44  0.25  0.35  0.91  15.17  7.55  1.93  1.74  1.11  Total Manganese Total Zinc ( 1 )  <2)  1.14  0.72  Taken from Macdonald ef al. (1997) Average of four street runoff stations  Potential effects from debris accumulation include habitat alteration, clogging of spawning areas in streams, fungal and bacterial growth, hydrogen sulphide production, negative effects on suspension feeding organisms and reduction of benthic communities. Tannins and lignins have been implicated in toxicity, although the persistence of these compounds and their long term effects in the environment are unclear. Lignin is thought to be precursor to humus and is the most abundant naturally occurring source of aromatic compounds and second only to cellulose as the most abundant organic carbon source in the biosphere. In general, lignin is fairly resistant to microbial degradation, with degradation occurring faster in aerobic environments (via soil fungi or lignocellulolytic bacteria in aquatic environments) than in anaerobic environments (Benner et al. 1986). It appears that some consortia of bacteria can decompose lignin compounds in anaerobic environments (such as marine sediments), however, these structures are generally stable under anaerobic conditions (Young and Frazer 1987, Schevchenko and Bailey 1996). However, Hedges et al. (1997) suggest that both dissolved and particulate organic matter of terrestrial origin are rapidly oxidized following discharge into the 114  ocean, as most organic matter in the marine environment appears to be of marine origin, despite the large globalriverineinputs of terrestrially derived organic carbon from continents to the oceans. Additionally, Sedell and Duvall (1985) suggest that the toxicity of leachates in seawater is thought to be negligible due to precipitation in combination with chloride ions. For log yards discharging directly to aquatic receiving environments, the rate of degradation of lignins contained in stormwater runoff (either in soluble or as yet unbroken down wood particles) will likely be largely dependent on the oxygen and salinity content of the aquatic system and of the bottom sediments. Resin acids can be found in waters and aquatic sediments receiving pulp and paper wastewater discharges (Leppanen et al. 2000) and in sediments adjacent log handling areas (Tian et al. 1998, Healy et al. 1997). Despite observed biodegradability in aerated secondary treatment, resin acids have been shown to persist in the environment (Tavendale et al. 1997). Leppanen et al. (2000) found that the resin acid composition of sedimenting particles was similar to the resin acid composition of sediments at the same area, suggesting that most of the resin acids in the water column were re-suspended material or that the biotransformation of resin acids in surface sediment is negligible.  4.7  Conclusions  and  Recommendations  In general, the acute toxicity of log yard stormwater runoff was similar between sites and not particularly toxic. Acute effects are likely to be substantially reduced upon runoff dilution in the receiving environment. Maximum runoff volumes were estimated to occur during the fall at the coastal site and in late winter at the interior site (rainfall and snowmelt), potentially increasing chances for acute toxic effects during those times. There is not enough evidence to implicate one particular substance as the toxic agent, however, runoff toxicity from both sites seems to be at least partially associated with TSS. 115  Due to drier weather conditions and the more permeable (unpaved) nature of the interior log yard, proportionally less runoff per square meter was generated at the interior site. Runoff volumes generated from the interior site were estimated to be only 4-fold higher than coastal runoff volumes, despite the 20-fold difference in size between the two sites. This suggests that log handling areas that receive low precipitation and are unpaved, are less likely to discharge high volumes of runoff directly to the aquatic environment (unless sprinkling systems are used on a continual basis) potentially reducing acute effects. Therefore, where groundwater sources are not at risk, infiltration of runoff into the soil may reduce the chances of high runoff volumes and acute environmental effects to nearby aquatic systems. This is contrary to the WDOE BMP guidelines (WDOE 1995), which recommend site paving as a source control BMP for log yards (Section 1.2.4). When loading estimates were standardized for log yard size, export coefficients were generally similar between sites, with the exception of tannins and lignins and DHA. This suggests that although proportionally less runoff per square meter was generated at the interior site, most contaminant concentrations were higher at the interior site with the exception of tannins and lignins. Therefore, the continual introduction of small runoff volumes with elevated contaminant concentrations may result in high export coefficients to the receiving environment potentially leading to cumulative effects in the receiving environment. The complex contaminant matrix of log yard stormwater runoff makes it difficult to manage on a province-wide basis. Similar to urban runoff, it contains a variety of contaminants with the addition of numerous wood extractives, which can vary depending on the type and volume of wood processed on site. Additionally, although unpaved sites seem to generate less runoff due to infiltration thus reducing the potential for acute effects, small runoff volumes with high contaminant concentrations may still result in high export coefficients of contaminants over time. 116  Therefore, since results from sample treatments suggest that suspended solids are at least partially responsible for toxicity and since other substances, such as metals, are often bound to suspended solids, simple source control and filtration/infiltration treatment options (Section 1.2.5) should be implemented on a site-specific basis. Simple treatment measures may reduce some of the contaminant load, thus decreasing the potential for acute effects and the cumulative build-up of debris and other contaminants in the receiving environment. Source control, such as continuous cleaning of debris from log yards may also help reduce the volume of debris in stormwater runoff. In general, there appear to be management tradeoffs between paved and unpaved log yards. Paved log yards generate more runoff, thus requiring greater treatment capacity. However, debris may be easier to clean thus reducing contaminants released from crushed wood debris. Unpaved log yards generate less runoff, requiring less treatment capacity. However, debris may be more difficult to clean, thus increasing the potential for release of wood leachates from crushed wood debris. Recommendations for further study include toxicity identification evaluations (TIEs) on runoff from log yard areas from a number of coastal and interior log yard sites. TIEs would help identify whether: (1) wood leachate contaminants such as tannins and lignins and resin and fatty acids are causing toxicity; and (2) whether the same contaminants are causing toxicity in different regions of the province where different tree species are stored on site. Results of such investigations would help identify the most effective source control and treatment BMPs.  117  5  References  Alberta Forest Products Association (AFPA). 1999a. Characterization of Surface Water Run-Off from Log Yard Sites in Alberta., Edmonton, Alberta. Alberta Forest Products Association (AFPA). 1999b. Evaluation and Control of Environmental Effects from Log Yard Sites in Alberta., Edmonton, Alberta. American Public Health Association (APHA). 1998. Standard Methods for the Examination of Water and Wastewater. Fifteenth Edition. Washington, DC. American Society of Civil Engineers (ASCE). 1992. Design and Construction of Urban Stormwater Management Systems. Practice No. 77, American Society of Civil Engineers, New York, NY. Athayde, D. N., P. E. Shelley, E. D. Driscoll, D. Gaboury, and G. Boyd. 1983. Results of the Nationwide Urban Runoff Program, Volume 1: Final Report. PB84-185552, Water Planning Division, US Environmental Protection Agency, Springfield, VA. Bailey, H. C , J. R. Elphick, A. Potter, E. Chao, D. Konasewich, and J. B. Zak. 1999. Causes of toxicity in stormwater runoff from sawmills. Environmental Toxicology and Chemistry 18:1485-1491. BC Ministry of Environment Lands and Parks (BCMELP). 1999a. British Columbia Approved Water Quality Guidelines (Criteria). Water Management Branch, Environment and Resource Management Department, Ministry of Environment, Lands and Parks. BC Ministry of Environment Lands and Parks (BCMELP). 1999b. A Compendium of Working Water Quality Guidelines for British Columbia. Water Management Branch, Environment and Resource Management Department, Ministry of Environment, Lands and Parks. BC Ministry of Forests. 2001. 2000/2001 Annual Performance Report., Victoria, BC.  118  BC Research. 1975. Identification and Treatment of Toxic Materials in Mechanical Pulping Effluents. Co-operative Pollution Abatement Research (CPAR) 149-3,, Vancouver, BC. BC Research. 1977. Identification and Treatment of Toxic Materials in Pulp and Paper Woodroom Effluents. Cooperative Pollution Abatement Research (CPAR) 148-3,, Vancouver, BC. Bellevue Utilities. 1995. Characterization and Source Control of Urban Stormwater Quality: Volume 1 - Technical Report. City of Bellevue Utilities Department, Bellevue, WA. Benner, R., M . A. Moran, and R. E. Hodson. 1986. Biogeochemical cycling of lignocellulosic carbon in marine and freshwater ecosystems: relative contributions of prokaryotes and eukaryotes. Limnology and Oceanography 31:89-100. Borga, P., T. Elowson, and K. Liukko. 1996. Environmental loads from water-sprinkled softwood timber: 2. Influence of tree species and water characteristics on wastewater discharges. Environmental Toxicology and Chemistry 15:1445-1454. Buchanan, D. V., P. S. Tate, and J. R. Moring. 1976. Acute toxicities of spruce and hemlock bark extract to some estuarine organisms in Southeastern Alaska. Journal of Fisheries Research Board of Canada 33:1188-1193. Choi, K., and P. G. Meier. 2001. Toxicity evaluation of metal plating wastewater employing the Microtox® assay: a comparison with cladocerans and fish. Environmental Toxicology 16:136-141. Conlan, K. E., and D. V. Ellis. 1979. Effects of woodwaste on sand-bed benthos. Marine Pollution Bulletin 10:262-267. Council of Forest Industries (COFI). 2001. 2000 Fact Book., Vancouver, BC. El-Shaarawi, A. H. 1989. Inferences about the mean from censored water quality data. Water Resources Research 25:685-690.  119  El-Shaarawi, A. H., K. W. Kuntz, and A. Sylvestre. 1986. Estimation of loading by numerical integration, in A. H. El-Shaarawi and R. E. Kwiatkowski, editors. Statistical Aspects of Water Quality Monitoring. Elsevier, Amsterdam. Field, J. A., M . J. H. Leyendeckers, R. Sierra-Alvarez, G. Letting, and L. H. A. Habets. 1988. The methanogenic toxicity of bark tannins and the anaerobic biodegradability of water soluble bark matter. Water Science and Technology 20:219-240. Gibb, A., B. Bennet, and A. Birkbeck. 1991. Urban Runoff Quality and Treatment: A Comprehensive Review. British Columbia Research Corporation, Vancouver, BC. Gibb, A., T. R. Schueler, R. Horner, J. Simmler, and J. Knutson. 1999. Best Management Practices Guide for Stormwater. Greater Vancouver Sewerage and Drainage District, Burnaby, BC. Good, J. C. 1993. Roof runoff as a diffuse source of metals and aquatic toxicity in storm water. Water Science and Technology 28:317-321. Hall, K. J., D. A. Levy, A. H. J. Dorcey, I. Yesaki, E. A. Power, P. L. Wentzell, B. Christensen, D. Runesson, and H. D. Klassen. 1988. Log handling in the Aquatic Environment: Impact Assessment and Research on Babine Lake. Technical Report Westwater Research Center, University of British Columbia, Vancouver, BC. Hall, K. J., H. Schreier, and S. J. Brown. 1991. Water quality in the Fraser River Basin, in A. H. J. Dorcey and J. R. Griggs, editors. Water in Sustainable Development: Exploring our Common Future in the Fraser River Basin. Westwater Research Centre, University of British Columbia, Vancouver, BC. Harremoes, P. 1988. Stochastic models for estimation of extreme pollution from urban runoff. Water Research 22:1017-1026. Hatfield Consultants Ltd. 1996. Strait of Georgia Fisheries Sustainability Review. Hatfield Consultants Ltd., West Vancouver, BC. 120  Healy, T. R., A. L. Wilkins, and T. Leipe. 1997. Extractives from a coniferous bark dump in coastal estuarine sediments. Journal of Coastal Research 13:293-296. Hedges, J. I., R. G. Keil, and R. Benner. 1997. What happens to terrestrial organic matter in the ocean. Organic Geochemistry 27:195-212. Horner, R. R. 1988. Biofdtration Systems for Storm Water Runoff Water Quality Control. Washington State Department of Ecology, Municipality of Metropolitan Seattle, King County, City of Bellevue, City of Mountlake Terrace, City of Redmond, Washington. Inamori, Y., K. Nishiguchi, N. Matsuo, H. Tsujibo, K. Baba, and N. Ishida. 1991. Phytogrowthinhibitory activities of tropolone and hinokitiol. Chemical Pharmaceutical Bulletin 39:2378-2381. Isenberg, I. H. 1980. Pulpwoods of the United States and Canada: Volume 1 - Conifers, Third edition. Institute of Paper Chemistry, Appleton, Wisconsin. Jackson, R. G. 1986. Effects of bark accumulation on benthic infauna at a log transfer facility in southeast Alaska. Marine Pollution Bulletin 17:258-262. Knezovich, J. P., F. L. Harrison, and J. S. Tucker. 1981. The influence of organic chelators on the toxicity of copper to embryos of the Pacific Oyster Crassostrea  gigas.  Archives of  Environmental Contamination and Toxicology 10:241-249. Leach, J. M., and A. N. Thakore. 1976. Toxic constituents in mechanical pulping effluents. Tappi 59:129-132. Leppanen, H., J. Kukkonen, and A. Oikari. 2000. Concentration of retene and resin acids in sedimenting particles collected from a bleached kraft mill effluent receiving lake. Water Research 34:1604-1610. Liu, S. D., M . D. Nassichuk, and S. C. Samis. 1996. Guidelines on Storage, Use and Disposal of Wood Residue for the Protection of Fish and Fish Habitat in British Columbia. 95-18, Fraser River Action Plan (FRAP), Vancouver, BC.  Macdonald, R., K. J. Hall, and H. Schreier. 1997. Water quality and Stormwater Contaminants in the Brunette River Watershed, British Columbia, 1994/95. Westwater Research Unit, Institute for Resources and Environment, University of British Columbia. Manahan, S. E. 2000. Environmental Chemistry, Seventh edition. CRC Press LLC, Boca Raton, Florida. Marsalek, J. 1991. Pollutant loads in urban stormwater: review of methods for planning-level estimates. Water Resources Bulletin 27:283-291. Marsalek, J., and H. Schroeter. 1988. Annual loadings of toxic contaminants in urban runoff from the Canadian Great Lakes Basin. Water Pollution Research 23:360-376. McDougall, S. 1996. Assessment of Log Yard Runoff in Alberta: Preliminary Evaluation. Alberta Environmental Protection. Meidinger, D., and J. Pojar. 1991. Ecosystems of British Columbia. Special Report Series 6, British Columbia Ministry of Forests, Victoria, BC. Moore, K. 1992. Wood Waste Leachate Characterization Study. BC Environment, Lower Mainland Region, Industrial Section, Surrey, BC. National Council of the Paper Industry for Air and Stream Improvement Inc. (NCASI). 1992. Storm Water from Log Storage Sites: A Literature Review and Case Study. Technical Bulletin 637,, New York, NY. Nelson, S. M., and R. A. Roline. 1998. Evaluation of the sensitivity of rapid toxicity tests relative to daphnid acute lethality tests. Bulletin of Environmental Contamination and Toxicology 60:292-299. O'Connor, B. I., T. G. Kovacs, and R. H. Voss. 1992. The effect of wood species composition on the toxicity of simulated mechanical pulping effluents. Environmental Toxicology and Chemistry 11.  122  Oikari, A., and J. Kukkonen. 1990. Bioavailability of benzo(a)pyrene and dehydroabietic acid from a few lake waters containing varying dissolved organic carbon concentrations to Daphnia magna. Bulletin of Environmental Contamination and Toxicology 45:54-61.  Pandey, A. K., S. Pandey, and V. Misra. 2000. Stability constants of metal-humic acid complexes and its role in environmental detoxification. Ecotoxicology and Environmental Safety 47:195-200. Pease, B. C. 1974. Effects of log dumping and rafting on the marine environment of southeast Alaska. General Technical Report PNW-22, US Department of Agriculture Forest Service. Peng, G., and J. C. Roberts. 2000. Solubility and toxicity of resin acids. Water Research 34:2779-2785. Peters, G. B., H. J. Dawson, B. F. Hrutfiord, and R. R. Whitney. 1976. Aqueous leachate from western red cedar: effects on some aquatic organisms. Journal of Fisheries Research Board of Canada 33:2703-2709. Samis, S. C , S. D. Liu, B. G. Wernick, and M . D. Nassichuk. 1999. Mitigation of Fisheries Impacts from the Use and Disposal of Wood Residue in British Columbia and the Yukon. Habitat Enhancement Branch, Fisheries and Oceans Canada and Environmental Protection Branch, Environment Canada, Vancouver BC. Schermer, E. D., and J. B. Phipps. 1976. A Study of Woodwaste Leachate. Grays Harbour College, Aberdeen, Washington. Schevchenko, S. M., and G. W. Bailey. 1996. Life after death: lignin-humic relationships reexamined. Critical Reviews in Environmental Science and Technology 26:95-153. Schueler, T. R. 1987. Controlling Urban Runoff: A Practical Manual for Planning and Designing Urban BMPs. Department of Environmental Programs, Metropolitan Washington Council of Governments, Washington, DC. 123  Schueler, T. R. 1997. Comparative pollutant removal capability of urban BMPs: a reanalysis. Watershed Protection Techniques 2. Schueler, T. R., and M. Helfrich. 1989. Design of extended detention wet pond systems. Pages 180-202 in Design of Urban Runoff Quality Control: Proceedings of an Engineering Foundation Conference on Current Practice and Design Criteria for Urban Quality Control. Trout Lodge, Potosi, Missouri, July 10-15. American Society of Civil Engineers, New York, NY. Sedell, J. R., and W. S. Duvall. 1985. Influence of Forest and Rangeland Management on Anadromous Fish Habitat in Western North America. 5. Water Transportation and Storage of Logs. General Technical Report PNW-186, US Department of Agriculture, Forest Service. Sj5strom, E. 1993. Wood Chemistry: Fundamentals and Applications, second edition. Academic Press, Inc., Toronto, Ontario. Stahre, P., and B. Urbonas. 1990. Stormwater Detention for Drainage, Water Quality and CSO Management. Prentice-Hall, Inc., Englewood Cliffs, New Jersey. Stanhope, M . J., and C. D. Levings. 1985. Growth and production of Eogammarus  confevicolus  (Amphipoda: Anisogammaridae) at a log storage site in areas of undisturbed habitat within the Squamish Estuary. Canadian Journal of Fisheries and Aquatic Science 42:1733-1740. Stanley Associates Engineering Ltd. 1992. Urban Runoff Quantification and Contaminants Loading in the Fraser Basin and Burrard Inlet. DOE FRAP 1993-19,, Surrey, BC. Sweet, H. R., and R. H. Fetrow. 1975. Ground-water pollution by wood waste disposal. Ground Water 13:227-231. Tabachnick, B. G., and L. S. Fidell. 2001. Using Multivariate Statistics, fourth edition. Allyn & Bacon, Needham Heights, MA. 124  Tavendale, M . H., P. N. McFarlane, and K. L. Mackie. 1997. The fate of resin acids. 1. The biotransformation and degradation of deuterium labelled dehydroabietic acid in anaerobic sediments. Chemosphere 35:2137-2151. Taylor, B. R. 1994. Toxicity of Aspen Wood Leachate to Aquatic Life. Part II: Field Study. Environmental Protection Branch, Northern Interior Region, BC Ministry of Environment, Calgary, Alberta. Taylor, B. R., J. S. Goudey, and N. B. Carmichael. 1996. Toxicity of aspen wood leachate to aquatic life: laboratory studies. Environmental Toxicology and Chemistry 15:150-159. Thomas, P. R. 1977. Consequences of Leaching from Pulp and Paper Mill Landfdl Operations. Co-operative Pollution Abatement Research (CPAR) 363, Co-operative Pollution Abatement Research Program of the Canadian Forestry Service, Environment Canada. Tian, F., A. L. Wilkins, and T. R. Healy. 1998. Accumulation of resin acids in sediments adjacent to a log handling area, Tauranga Harbour, New Zealand. Bulletin of Environmental Contamination and Toxicology 60:441-447. Toussaint, M. W., R. R. Shedd, and W. H. Van der Schalie. 1995. A comparison of standard acute toxicity tests with rapid-screening toxicity tests. Environmental Toxicology and Chemistry 14:907-915. United States Environmental Protection Agency (USEPA). 1991. Methods for Aquatic Toxicity Identification Evaluations (Second Edition). EPA/600/6-91/003, Environmental Research Laboratory, Office of Research and Development, US Environmental Protection Agency, Duluth, MN. United States Environmental Protection Agency (USEPA). 1992. NPDES Storm Water Sampling Guidance Document. EPA 833-B-92-001, Office of Water, US Environmental Protection Agency.  125  United States Environmental Protection Agency (USEPA). 1993. Methods for Measuring the Acute Toxicity of Effluents and Receiving Waters to Freshwater and Marine Organisms: Fourth Edition. EPA/600/4-90/027F, Office of Research and Development, Washington, DC. Washington State Department of Ecology (WDOE). 1995. Best Management Practices to Prevent Stormwater Pollution at Log Yards. 95-53, Stormwater Unit, Olympia, WA. Weltens, R., R. Goossens, and S. V. Puymbroeck. 2000. Ecotoxicity of contaminated suspended solids for filter feeders (Daphnia  magna).  Archives of Environmental Contamination and  Toxicology 39:315-323. Woods, W.G. 1994. An introduction to boron: history, sources, uses and chemistry. Environmental Health Perspectives 102 (Supplement 7):5-11. Young, L. Y., and A. C. Frazer. 1987. The fate of lignin and lignin-derived compounds in anaerobic environments. Geomicrobiology 5:261-293. Yousef, Y. A., M . P. Wanielista, and H. H. Harper. 1986. Design and effectiveness of urban retention basins. Pages 338-350 in Urban Runoff Quality: Proceedings of an Engineering Foundation Conference, New England College, Henniker, New Hampshire, June 23-27'. American Society of Civil Engineers, New York, NY. Zanella, E. 1983. Effect of pH on acute toxicity of dehydroabietic acid and chlorinated dehydroabietic acid on fish and Daphnia. Bulletin of Environmental Contamination and Toxicity 30: 133-140. Zar, J. H. 1996. Biostatistical Analysis, third edition. Prentice-Hall, Inc., Upper Saddle River, NJ. Zenaitis, M.G., and J.B. Duff. 2002. Ozone for removal of acute toxicity from logyard run-off Ozone Science and Engineering 24: 83-90.  126  Appendix A: Data Tables Table A.l: General water quality characteristics in interior log yard stormwater runoff Samples Parameter  Units  N-03/13  N-07/19  N-08/23  COD  mg/L  750  703  218  BOD  mg/L  586  218  174  Alkalinity  mg/L ( C a C 0 3 )  104  157  63  Hardness  mg/L ( C a C 0 3 )  176  294  70  p H (field)  pH  6.0  7.0  7.0  pH (lab)  PH  7.36  7.38  7.72  % saturation  66.0  nd  94.0  uS/cm  100  290  nd  uS/cm  309.00  205.00  154.00  C  2.0  19.0  16.0  TSS  mg/L  568  1407  123  DHA  mg/L  2.89  0.43  <0.01  T a n n i n s a n d Lignins  mg/L  75  68  43  DO Specific C o n d u c t i v i t y (field) Specific C o n d u c t i v i t y (lab) Temperature  nd meters not working Hardness calculated based on C a  2 +  and M g  2 +  concentrations  127  Table A.2: Metals concentrations in interior log yard stormwater runoff Samples Parameter (mg/L)  N-03/13  N-07/19  N-08/23  Aluminum  11.7  38.8  4.6  Antimony  <0.2  <0.2  <0.2  Arsenic  <0.2  <0.2  <0.2  Barium  0.24  0.36  0.07  <0.005  <0.005  <0.005  Bismuth  <0.1  <0.1  <0.2  Boron  <0.1  <0.1  <0.1  <0.01  <0.01  <0.01  Calcium  53.6  61.2  24.5  Chromium  0.04  0.17  0.02  Cobalt  0.01  0.04  <0.01  Copper  0.05  0.12  0.08  Iron  21.4  69.6  7.47  Lead  <0.05  <0.05  <0.05  Lithium  0.02  0.04  0.01  Magnesium  10.3  34.4  6.5  Manganese  2.08  2.19  0.53  Molybdenum  <0.03  <0.03  <0.03  Nickel  <0.05  0.11  <0.05  Phosphorus  1.1  1.5  <0.3  Potassium  14  11  7  <0.2  <0.2  <0.2  Silicon  21  39.3  11.3  Silver  <0.01  <0.01  <0.01  7  5  3  0.105  0.151  0.063  <0.2  <0.2  <0.2  <0.03  <0.03  <0.03  Titanium  0.35  2.63  0.2  Vanadium  0.03  0.13  <0.03  0.217  0.423  0.134  Beryllium  Cadmium  Selenium  Sodium Strontium Thallium Tin  Zinc  128  tfCM  CO  00  CD  tfco  LO CD  CM  LO  o CO CM CO  o CO  coo  T-  CM CD  O •<-  CO  LO  tf-  cb  CM  CM CM  LO LO  co cb  co cb  II  CD  ^  co cb  CO  CO  tf" CD  co CD  O  oo  o o  LO OO  CO LO  tf  r^-  00  o tfo  LO  o cb  CO  CM CO  tf"  CM  -tf  tfoo  oo  tf" °°  CO  O V  CD  LO  co  T -  CM  o to CD  CD  o CO  CO CM  CM  CM CO  q  CD CO CM  o  T  -  CM CM  o  LO LO CM  •tf  LO  CD T -  LO  tf"  CD  o  CO CO CM  CO  0 TJ 0  i_ 0  CM  q CD CO  0 CL  q CO  E ca CO  O CD CO  CD LO  COO  cb  cb  s - »  CM  cb  § S £  LO  •tf tf"  tf  g CO cz o o ZJ -Q TO*  LO  o CO  00  LO  CO CO  CD •tf  oq LO  CM CD  00  00 LO  oo  LO CO  ib  CO  CM CM  tf o CO  tfcb  o o  CD LO CM  co  CD  d  CD LO  co LO  0  CM  d  co CD  oo tfcb  0  oo CD  00 CD CO  CO 1-  00  ^  oo  co  T -  O  I LO  ^ CZ»  tf" CM  T—  S  O CM CM  •tf LO  T-  CO  cb  CO  cb  CD CD  CD tf" tf"  O CM LO  CD CO  r8 ^  a o O LO  O COO CM  CM  «  oo ib  CO CD  o d o  *  CM  CO  LO  00  T -  o •tf  •rT -  CM CO  CM  CD CD  ^  CQ  — co  ro c as aj CO  ~  >4-  CO CO LO  LO LO  T-  CM  CM  o  co ib  1^ •tf  o o  LO  CD  CJ  -- -tf O  CD  ib  CM  o 00  3  CO co =d O =d O C0Q TOO E E 8 O  CO  E  X CL  X  tf"  io T-  O  CM  If  CL  JJ JJ  CD  CO CO  CO  E  ZJ  0  E  O  ZJ  CO  CO  CO  E  E  E  CO  c  O Zi TO CZ  £ 0  E co co  CO CO CD  Q.  Q O CQ  ca  c  "D  co  X  o O  TJ  "CD  CJ L^  o  -S  ^ ^ 9 CL  CL  Q  CD  Oft ro co  cz o "co i_ ZJ -t—•  CD  O "  O  Q O O  L .  ro 5  00 o CO  0  _ g  cn co  LO  CO  OS CN  CO  ro >  q tf= oo CO q CO  CM  ZJ  o CD CJ  tf-  q q  LO  'cz  o  CO  ZJ TJ CZ  o OC J  L^  CD (ZL  'CJ CD CL  CO  to  S TJ  ^  JZ) CO  O  CD CL  E CD  H  CO CO  < X Q  g  'cz  TJ  CO  CD  CO  g ro o  z! *  0 TJ CZ CO CO  <">  co o "co O CCO D ro gcz j 3  o a  ZJ  co co 0  cz T J  s  co  X  Table A.4: Metals concentrations in coastal log yard stormwater  runoff  Samples Parameter (mg/L)  S-03/29 S-04/05 S-04/22 S-05/04 S-05/14 S-06/02 S-06/11 S-06/27 S-08/21 S-10/24  Aluminum  2.6  3.5  0.5  6  2  1  4.2  2.8  1.1  2.2  Antimony  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  Arsenic  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  Barium  0.03  0.05  <0.01  0.06  0.02  <0.01  0.04  0.02  <0.01  0.02  <0.005  <0.005  <0.005  <0.005  <0.005  <0.005  <0.005  <0.005  <0.005  <0.005  <0.1  <0.1  <0.1  <0.1  <0.1  <0.1  <0.1  <0.1  <0.2  <0.2  0.3  0.3  0.2  0.2  0.2  0.2  0.5  0.2  0.2  0.2  O.01  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  13.6  13.7  4.44  12.6  6.4  6.24  26  8.17  5.52  6.46  Chromium  O.01  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  Cobalt  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  Copper  <0.01  0.02  <0.01  0.03  0.01  <0.01  0.02  0.02  <0.01  <0.01  Iron  3.01  4.43  0.64  8.79  2.26  1.11  7.25  4.27  1.67  3.69  Lead  <0.05  <0.05  <0.05  <0.05  <0.05  <0.05  <0.05  <0.05  <0.05  O.05  Lith ium  <0.01  <0.01  <0.01  <0.01  <0.01  <0.01  0.01  <0.01  <0.01  <0.01  Magnesium  16.5  19.7  4.9  12.9  7.2  9.7  41.6  13.4  10  7.5  Manganese  1.45  1.21  0.48  0.945  0.531  0.296  1.45  0.499  0.239  0.536  Molybdenum  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  Nickel  <0.05  <0.05  <0.05  <0.05  <0.05  <0.05  <0.05  <0.05  <0.05  <0.05  Phosphorus  0.9  0.9  0.4  0.8  0.6  0.4  1.5  0.7  0.4  0.6  Potassium  17  16  8  12  9  8  26  12  8  9  Selenium  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  Silicon  4.22  5.1  1.28  10.1  3.54  2.16  7.27  5.31  2.08  4.03  Silver  <0.01  <0.01  <0.01  <0.01  O.01  <0.01  <0.01  <0.01  <0.01  <0.01  184  211  72  120  86  119  385  146  109  84  0.085  0.11  0.028  0.089  0.044  0.06  0.27  0.083  0.059  0.049  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.2  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  0.08  0.12  0.02  0.25  0.08  0.03  0.13  0.11  0.03  0.09  Vanadium  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  <0.03  O.03  <0.03  Zinc  0.064  0.103  0.033  0.186  0.053  0.023  0.162  0.108  0.037  0.094  Beryllium Bismuth Boron Cadmium Calcium  Sodium Strontium Thallium Tin Titanium  130  Appendix B: Water Quality Trend Figures  03/13  05/04  06/27  08/21  10/24  03/13  05/04  Date  06/27 Date  08/21  10/24  03/13  05/04  06/27  08/21  Date  Figure B.l: Trends for general water quality characteristics in interior and coastal log yard stormwater runoff (circles are interior runoff samples, squares are coastal runoff samples)  131  10/24  03/13  05/04  06/27  08/21  10/24  03/13  05/04  Date  06/27 Date  08/21  10/24  03/13  05/04  06/27  08/21  10/24  Date  Figure B . l (continued): Trends for general water quality characteristics in interior and coastal log yard stormwater runoff (circles are interior runoff samples, squares are coastal runoff samples)  132  o  CQ  A/V/s,. -n—i—n—n—i  03/13  05/04  06/27  08/21  03/13  05/04  Date  -n—ill  03/13  05/04  05/04  06/27  Date  i  06/27  06/27  05/04  06/27  05/04  06/27  08/21  10/24  03/13  05/04  Date  08/21  10/24  03/13  Date  03/13  n—n—i—i—i—n—i—  03/13  Date  Date  03/13  r  06/27  05/04  06/27  06/27  08/21  Date  08/21  10/24  03/13  Date  05/04  06/27  08/21  Date  08/21  Date  Figure (circles  B.2: Trends for metal concentrations are interior  runoff samples,  in interior  and coastal log yard stormwater  squares are coastal runoff  133  samples)  runoff  03/13  05/04  06/27  08/21  03/13  08/21  03/13  05/04  03/13  05/04  Date  III 03/13  05/04  ll  n — n — i  06/27  i i  05/04  Date  Figure  B.3: Trends  runoff  (circles  for  06/27  08/21  Date  general  are interior  runoff  ion concentrations samples,  in interior  squares  134  are coastal  and coastal runoff  log yard samples)  stormwater  Figure BA: Trends for C. dubia and Microtox® toxicity in interior and coastal log yard stormwater runoff (circles are interior runoff samples, squares are coastal runoff samples)  135  Appendix C: Spearman's Rank Correlations Table C.l: Spearman's rank correlations for the combined log yard stormwater runoff contaminants and flow data TSS  DHA  TL  labpH  0.46  0.09  -0.28  -0.54  0.64  0.84  0.58  0.63  0.65  0.31  0.28  1.00  0.70  0.20  0.78  0.89  0.72  -0.27  0.84  0.70  1.00  0.61  0.66  0.64  0.34  0.22  0.46  0.58  0.20  0.61  1.00  0.39  0.03  -0.24  0.58  TSS  0.09  0.63  0.78  0.66  0.39  1.00  0.67  0.58  -0.14  DHA  -0.28  0.65  0.89  0.64  0.03  0.67  1.00  0.69  -0.21  TL  -0.54  0.31  0.72  0.34  -0.24  0.58  0.69  1.00  -0.74  labpH  0.64  0.28  -0.27  0.22  0.58  -0.14  -0.21  -0.74  1.00  labcond  -0.74  -0.02  0.48  0.15  -0.41  0.08  0.41  0.76  -0.68  Al  0.32  0.84  0.65  0.83  0.61  0.84  0.57  0.30  0.25  Ba  0.37  0.88  0.57  0.81  0.60  0.75  0.55  0.22  0.35  B  -0.75  -0.12  0.27  -0.02  -0.51  -0.13  0.29  0.68  -0.71  Ca  0.23  0.87  0.64  0.93  0.62  0.74  0.61  0.30  0.22  Cu  0.48  0.76  0.47  0.69  0.62  0.65  0.37  0.05  0.38  Fe  0.35  0.82  0.64  0.80  0.61  0.86  0.55  0.29  0.24  Mg  -0.21  0.57  0.77  0.80  0.23  0.65  0.68  0.68  -0.24  Mn  0.04  0.78  0.72  0.80  0.39  0.80  0.78  0.52  -0.01  Na  -0.73  -0.01  0.47  0.15  -0.40  0.05  0.40  0.76  -0.68  P  -0.14  0.68  0.85  0.78  0.27  0.81  0.86  0.68  -0.23  K  -0.44  0.55  0.83  0.63  -0.04  0.62  0.88  0.88  -0.49  Si  0.31  0.78  0.61  0.82  0.62  0.80  0.50  0.23  0.26  Sr  -0.06  0.80  0.84  0.96  0.49  0.73  0.74  0.52  0.05  Ti  0.34  0.81  0.64  0.78  0.62  0.85  0.54  0.27  0.24  Zn  0.29  0.79  0.67  0.77  0.56  0.89  0.59  0.35  0.13  H a r d n e s s Alkalinity  Flow  BOD  COD  Flow  1.00  0.35  -0.37  0.08  BOD  0.35  1.00  0.57  COD  -0.37  0.57  Hardness  0.08  Alkalinity  136  labcond  Al  Ba  B  Ca  Cu  Fe  Mg  Flow  -0.74  0.32  0.37  -0.75  0.23  0.48  0.35  -0.21  BOD  -0.02  0.84  0.88  -0.12  0.87  0.76  0.82  0.57  COD  0.48  0.65  0.57  0.27  0.64  0.47  0.64  0.77  Hardness  0.15  0.83  0.81  -0.02  0.93  0.69  0.80  0.80  Alkalinity  -0.41  0.61  0.60  -0.51  0.62  0.62  0.61  0.23  TSS  0.08  0.84  0.75  -0.13  0.74  0.65  0.86  0.65  DHA  0.41  0.57  0.55  0.29  0.61  0.37  0.55  0.68  TL  0.76  0.30  0.22  0.68  0.30  0.05  0.29  0.68  labpH  -0.68  0.25  0.35  -0.71  0.22  0.38  0.24  -0.24  1.00  -0.14  -0.23  0.91  -0.06  -0.33  -0.17  0.59  Al  -0.14  1.00  0.97  -0.35  0.92  0.92  0.99  0.55  Ba  -0.23  0.97  1.00  -0.37  0.93  0.90  0.96  0.46  B  0.91  -0.35  -0.37  1.00  -0.20  -0.51  -0.38  0.41  Ca  -0.06  0.92  0.93  -0.20  1.00  0.81  0.91  0.63  Cu  -0.33  0.92  0.90  -0.51  0.81  1.00  0.92  0.34  Fe  -0.17  0.99  0.96  -0.38  0.91  0.92  1.00  0.52  Mg  0.59  0.55  0.46  0.41  0.63  0.34  0.52  1.00  Mn  0.05  0.77  0.82  0.00  0.86  0.56  0.75  0.65  Na  0.99  -0.15  -0.23  0.91  -0.07  -0.31  -0.18  0.59  P  0.35  0.65  0.61  0.23  0.73  0.43  0.63  0.86  K  0.66  0.45  0.41  0.56  0.55  0.19  0.42  0.82  Si  -0.19  0.98  0.95  -0.41  0.93  0.93  0.98  0.50  Sr  0.35  0.80  0.74  0.14  0.86  0.65  0.77  0.89  Ti  -0.18  0.99  0.95  -0.39  0.90  0.93  1.00  0.50  Zn  -0.14  0.98  0.92  -0.34  0.88  0.89  0.98  0.54  labcond  137  Mn  Na  P  K  Si  Sr  Ti  Zn  Flow  0.04  -0.73  -0.14  -0.44  0.31  -0.06  0.34  0.29  BOD  0.78  -0.01  0.68  0.55  0.78  0.80  0.81  0.79  COD  0.72  0.47  0.85  0.83  0.61  0.84  0.64  0.67  Hardness  0.80  0.15  0.78  0.63  0.82  0.96  0.78  0.77  Alkalinity  0.39  -0.40  0.27  -0.04  0.62  0.49  0.62  0.56  TSS  0.80  0.05  0.81  0.62  0.80  0.73  . 0.85  0.89  DHA  0.78  0.40  0.86  0.88  0.50  0.74  0.54  0.59  TL  0.52  0.76  0.68  0.88  0.23  0.52  0.27  0.35  labpH  -0.01  -0.68  -0.23  -0.49  0.26  0.05  0.24  0.13  labcond  0.05  0.99  0.35  0.66  -0.19  0.35  -0.18  -0.14  Al  0.77  -0.15  0.65  0.45  0.98  0.80  0.99  0.98  Ba  0.82  -0.23  0.61  0.41  0.95  0.74  0.95  0.92  B  0.00  0.91  0.23  0.56  -0.41  0.14  -0.39  -0.34  Ca  0.86  -0.07  0.73  0.55  0.93  0.86  0.90  0.88  Cu  0.56  -0.31  0.43  0.19  0.93  0.65  0.93  0.89  Fe  0.75  -0.18  0.63  0.42  0.98  0.77  1.00  0.98  Mg  0.65  0.59  0.86  0.82  0.50  0.89  0.50  0.54  Mn  1.00  0.04  0.89  0.73  0.73  0.76  0.75  0.76  Na  0.04  1.00  0.35  0.66  -0.20  0.34  -0.19  -0.15  P  0.89  0.35  1.00  0.88  0.60  0.83  0.62  0.68  K  0.73  0.66  0.88  1.00  0.40  0.73  0.40  0.47  Si  0.73  -0.20  0.60  0.40  1.00  0.77  0.98  0.96  Sr  0.76  0.34  0.83  0.73  0.77  1.00  0.76  0.77  Ti  0.75  -0.19  0.62  0.40  0.98  0.76  1.00  0.98  Zn  0.76  -0.15  0.68  0.47  0.96  0.77  0.98  1.00  138  Table C.2: Spearman's rank correlations for the coastal log yard stormwater runoff contaminants and flow data Flow  BOD  COD  Hardness  Alkalinity  TSS  DHA  TL  labpH  Flow  1.00  -0.06  -0.59  -0.50  -0.14  -0.25  -0.43  -0.42  0.23  BOD  -0.06  1.00  0.64  0.76  0.30  0.58  0.75  0.83  -0.27  COD  -0.59  0.64  1.00  0.87  0.19  0.77  0.84  0.83  -0.44  Hardness  -0.50  0.76  0.87  1.00  0.34  0.59  0.76  0.79  -0.25  Alkalinity  -0.14  0.30  0.19  0.34  1.00  0.19  -0.15  0.17  0.10  TSS  -0.25  0.58  0.77  0.59  0.19  1.00  0.62  0.85  -0.67  DHA  -0.43  0.75  0.84  0.76  -0.15  0.62  1.00  0.83  -0.40  TL  -0.42  0.83  0.83  0.79  0.17  0.85  0.83  1.00  -0.62  labpH  0.23  -0.27  -0.44  -0.25  0.10  -0.67  -0.40  -0.62  1.00  labcond  -0.48  0.71  0.84  0.99  0.28  0.55  0.73  0.76  -0.30  Al  -0.28  0.80  0.89  0.79  0.19  0.92  0.81  0.94  -0.53  Ba  -0.28  0.85  0.84  0.74  0.09  0.81  0.86  0.91  -0.42  B  -0.47  0.72  0.61  0.81  0.20  0.31  0.72  0.67  -0.30  Ca  -0.43  0.81  0.89  0.90  0.17  0.76  0.85  0.92  -0.54  Cu  -0.10  0.66  0.78  0.56  0.13  0.71  0.69  0.69  -0.41  Fe  -0.22  0.75  0.87  0.73  0.19  0.95  0.77  0.90  -0.56  Mg  -0.39  0.77  0.79  0.98  0.29  0.56  0.72  0.78  -0.31  Mn  -0.49  0.74  0.74  0.71  0.02  0.66  0.83  0.88  -0.56  Na  -0.45  0.74  0.83  0.98  0.30  0.50  0.72  0.75  -0.29  P  -0.44  0.83  0.85  0.86  0.09  0.73  0.88  0.94  -0.59  K  -0.46  0.79  0.83  0.85  0.06  0.70  0.85  0.93  -0.63  Si  -0.39  0.67  0.93  0.77  0.15  0.92  0.77  0.90  -0.59  Sr  -0.48  0.77  0.90  0.99  0.37  0.65  0.77  0.81  -0.24  Ti  -0.23  0.72  0.88  0.71  0.20  0.95  0.76  0.87  -0.56  Zn  -0.22  0.69  0.82  0.65  0.08  0.95  0.76  0.90  -0.66  139  labcond  Al  Ba  B  Ca  Cu  Fe  Mg  Flow  -0.48  -0.28  -0.28  -0.47  -0.43  -0.10  -0.22  -0.39  BOD  0.71  0.80  0.85  0.72  0.81  0.66  0.75  0.77  COD  0.84  0.89  0.84  0.61  0.89  0.78  0.87  0.79  Hardness  0.99  0.79  0.74  0.81  0.90  0.56  0.73  0.98  Alkalinity  0.28  0.19  0.09  0.20  0.17  0.13  0.19  0.29  TSS  0.55  0.92  0.81  0.31  0.76  0.71  0.95  0.56  DHA  0.73  0.81  0.86  0.72  0.85  0.69  0.77  0.72  TL  0.76  0.94  0.91  0.67  0.92  0.69  0.90  0.78  labpH  -0.30  -0.53  -0.42  -0.30  -0.54  -0.41  -0.56  -0.31  labcond  1.00  0.76  0.68  0.81  0.89  0.53  0.70  0.99  Al  0.76  1.00  0.94  0.52  0.89  0.85  0.99  0.77  Ba  0.68  0.94  1.00  0.58  0.89  0.77  0.91  0.68  B  0.81  0.52  0.58  1.00  0.81  0.24  0.45  0.81  Ca  0.89  0.89  0.89  0.81  1.00  0.64  0.85  0.88  Cu  0.53  0.85  0.77  0.24  0.64  1.00  0.85  0.53  Fe  0.70  0.99  0.91  0.45  0.85  0.85  1.00  0.71  Mg  0.99  0.77  0.68  0.81  0.88  0.53  0.71  1.00  Mn  0.67  0.75  0.86  0.80  0.91  0.45  0.71  0.66  Na  0.99  0.75  0.68  0.81  0.88  0.57  0.67  0.98  P  0.85  0.88  0.89  0.82  0.98  0.63  0.83  0.85  K  0.85  0.85  0.85  0.82  0.97  0.59  0.79  0.85  Si  0.75  0.98  0.89  0.45  0.87  0.85  0.96  0.73  Sr  0.96  0.84  0.78  0.74  0.89  0.65  0.79  0.95  Ti  0.67  0.98  0.91  0.42  0.84  0.87  0.99  0.67  Zn  0.62  0.96  0.86  0.38  0.79  0.85  0.98  0.65  140  Mn  Na  P  K  Si  Sr  Ti  Zn  Flow BOD  -0.49 0.74  -0.45 0.74  -0.44  -0.46  -0.39  -0.48  -0.23  -0.22  0.83  0.79  0.67  0.77  0.72  0.69  COD  0.74  0.83  0.85  0.83  0.93  0.90  0.88  0.82.  Hardness  0.71  0.98  0.86  0.85  0.77  0.99  0.71  0.65  Alkalinity  0.02  0.30  0.09  0.06  0.15  0.37  0.20  0.08  TSS  0.66  0.50  0.73  0.70  0.92  0.65  0.95  0.95  DHA  0.83  0.72  0.88  0.85  0.77  0.77  0.76  0.76  TL  0.88  0.75  0.94  0.93  0.90  0.81  0.87  0.90  labpH  -0.56  -0.29  -0.59  -0.63  -0.59  -0.24  -0.56  -0.66  labcond  0.67  0.99  0.85  0.85  0.75  0.96  0.67  0^62  Al  0.75  0.75  0.88  0.85  0.98  0.84  0.98  0.96  Ba  0.86  0.68  0.89  0.85  0.89  0.78  0.91  0.86  B  0.80  0.81  0.82  0.82  0.45  0.74  0.42  0.38  Ca  0.91  0.88  0.98  0.97  0.87  0.89  0.84  0.79  Cu  0.45  0.57  0.63  0.59  0.85  0.65  0.87  0.85  Fe  0.71  0.67  0.83  0.79  0.96  0.79  0.99  0.98  Mg  0.66  0.98  0.85  0.85  0.73  0.95  0.67  0.65  Mn  1.00  0.66  0.94  0.93  0.72  0.68  0.69  0.67  Na  0.66  1.00  0.85  0.85  0.73  0.95  0.65  0.60  P  0.94  0.85  1.00  0.99  0.85  0.84  0.81  0.80  K  0.93  0.85  0.99  1.00  0.83  0.81  0.77  0.77  Si  0.72  0.73  0.85  0.83  1.00  0.82  0.96  0.95  Sr  0.68  0.95  0.84  0.81  0.82  1.00  0.77  0.71  Ti  0.69  0.65  0.81  0.77  0.96  0.77  1.00  0.96  Zn  0.67  0.60  0.80  0.77  0.95  0.71  0.96  1.00  141  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0090410/manifest

Comment

Related Items