Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Assessment of cumulative effects of urban and agricultural land uses on indicators of water and stream… Wilson, Julie Elizabeth 2010

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

Item Metadata

Download

Media
24-ubc_2011_spring_wilson_julie.pdf [ 29.27MB ]
Metadata
JSON: 24-1.0071509.json
JSON-LD: 24-1.0071509-ld.json
RDF/XML (Pretty): 24-1.0071509-rdf.xml
RDF/JSON: 24-1.0071509-rdf.json
Turtle: 24-1.0071509-turtle.txt
N-Triples: 24-1.0071509-rdf-ntriples.txt
Original Record: 24-1.0071509-source.json
Full Text
24-1.0071509-fulltext.txt
Citation
24-1.0071509.ris

Full Text

ASSESSMENT OF CUMULATIVE EFFECTS OF URBAN AND AGRICULTURAL LAND USES ON INDICATORS OF WATER AND STREAM QUALITY by Julie Elizabeth Wilson B.Sc., The University of British Columbia, 2007  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)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December 2010  © Julie Elizabeth Wilson, 2010  Abstract This study assessed the cumulative effects of intensive urban and agricultural land use in Marshall Creek, B.C. The focus was to document how groundwater from the Abbotsford Aquifer and surface water from urban and agricultural tributaries contribute to cumulative stream quality downstream. Landscape stressors such as livestock numbers, Total Impervious Area (TIA) and human population numbers were related to the quality of water, bed sediment, suspended sediment and biofilms. Biofilms and suspended sediments were selected to provide an integrated indicator of potentially bioavailable contaminants. Nitrate-N concentrations in Marshall Creek are strongly influenced by groundwater discharges from the Abbotsford Aquifer, which has elevated nitrate-N concentrations due to non-point source agricultural activities. Nitrate decreased in the downstream due to significantly lower nitrate-N inputs from the tributaries (M-W U, p<0.015). In contrast, Soluble Reactive Phosphorus (SRP) and TOC increased downstream from agricultural runoff. Cu, Mn and Zn have increased in bed sediments in Marshall Creek since 1993. P concentrations have not changed markedly over time, but are significantly higher at the mouth than in the headwaters. SRP, NH4+-N, Cl-, turbidity and TOC were significantly higher in water in the agricultural tributary than the urban tributary (M-W U, p < α=0.0167). Differences in bed sediments, suspended sediments and biofilms showed that concentrations of P were significantly higher in the agricultural tributary than in the urban tributary (M-W U, p < 0.001), which is the result of nutrient and organic inputs from manure in the agricultural subwatershed. Biofilm chl-a and trace element concentrations were significantly different between sites and seasons. In the summer, chl-a and P were highest in the agricultural tributary, whereas in the winter chl-a was highest in the urban tributary due to warmer water temperatures in urban runoff.  ii  Trace element concentrations (e.g. Cu, Ni, P) were significantly higher in suspended sediments than in bed sediments or biofilms at all sites as a result of the higher surface area and greater organic matter associated with fine suspended sediments. Current agricultural activities are key contributors to the cumulative effects observed on Marshall Creek. Thus, improved livestock density and manure management practices are needed.  iii  Table of Contents Abstract .............................................................................................................................. ii Table of Contents ............................................................................................................. iv List of Tables .................................................................................................................... vi List of Figures ................................................................................................................. viii List of Abbreviations and Symbols ................................................................................. x Acknowledgements ......................................................................................................... xii 1  Introduction ................................................................................................................ 1 1.1 1.2 1.3 1.4 1.5  2 3  Objectives .................................................................................................................. 10 Background ............................................................................................................... 11 3.1 3.2  4  Cumulative Effects ...........................................................................................................1 Watershed CEA Framework ..........................................................................................3 Study Area ........................................................................................................................4 Land use and stream quality ...........................................................................................5 Methods of Assessment ....................................................................................................7  Study Site ........................................................................................................................11 Sampling Site Selection ..................................................................................................14  Methods ..................................................................................................................... 17 4.1 Land use indices and climate data ................................................................................17 4.1.1 Meteorological and hydrometric data .......................................................................17 4.1.2 Geographic Information Systems (GIS) data ............................................................17 4.1.3 Agricultural census data and Animal Unit Equivalents (AUEs) ...............................17 4.1.4 Population data ..........................................................................................................18 4.2 Field Sampling and Laboratory Analysis ....................................................................18 4.2.1 Water sampling and laboratory analysis ...................................................................18 4.2.2 Bed and suspended sediment sampling and laboratory analysis ...............................19 4.2.3 Biofilm collection and laboratory analysis................................................................20 4.3 Data analysis methods....................................................................................................22 4.3.1 Quality analysis and quality control ..........................................................................22 4.3.2 Statistical analyses ....................................................................................................22  5  Results and Discussion ............................................................................................. 25 5.1 Land Use Change – Landscape Stressors ....................................................................25 5.1.1 GIS and impervious area calculations .......................................................................25 5.1.2 Animal Numbers and AUEs......................................................................................25 5.1.3 Population Estimates .................................................................................................29 5.2 Stream Quality – Measurement Endpoints .................................................................31 5.2.1 Water Quality ............................................................................................................31 5.2.1.1 Overall .............................................................................................................................31 5.2.1.2 Upstream to Downstream Trends ....................................................................................32 5.2.1.3 Correlations between water quality parameters ..............................................................36 5.2.1.3.1 Wet Season Correlations .........................................................................................37 5.2.1.3.2 Dry Season Correlations ..........................................................................................38 5.2.1.4 Site comparison ...............................................................................................................38  iv  5.2.1.5 Historic Nitrate-N............................................................................................................41 5.2.1.5.1 Time-Series Analysis...............................................................................................42  5.2.2 5.2.3  Bed Sediments – Historic comparison ......................................................................44 Suspended Sediments ................................................................................................47  5.2.3.1 5.2.3.2  5.2.4  Overall .............................................................................................................................47 Site Comparison ..............................................................................................................48  Biofilms .....................................................................................................................50  5.2.4.1 Site and Seasonal Comparison ........................................................................................50 5.2.4.1.1 Chlorophyll a ...........................................................................................................50 5.2.4.1.2 Dry Mass .................................................................................................................55 5.2.4.1.3 Trace Elements ........................................................................................................59  5.2.5  Comparisons between in-stream indicators...............................................................60  5.2.5.1 5.2.5.2 5.2.5.3  6 7  Correlations .....................................................................................................................60 Trace elements in sediments and biofilms.......................................................................61 Biofilm : sediment ratios .................................................................................................63  Summary ................................................................................................................... 67 Conclusions................................................................................................................ 70 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8  Land use trends ..............................................................................................................70 Seasonal and spatial trends in water quality ...............................................................71 Trends in bed sediment quality .....................................................................................72 Interactions between water, sediments and biofilms ..................................................73 Seasonal effects ...............................................................................................................73 Suspended sediments and biofilms as indicators of cumulative effects .....................74 Cumulative effects assessment ......................................................................................74 Limitations and future research opportunities ...........................................................76  References ........................................................................................................................ 78 Appendices ....................................................................................................................... 89 Appendix A: Maps of the Marshall Creek watershed ..........................................................89 Appendix B: Water sampling and analysis results ...............................................................95 Appendix C: Sediment sampling and analysis results ........................................................103 Appendix D: Biofilm sampling and analysis results ...........................................................108 Appendix E: Spearman Rank Correlations .........................................................................113 Appendix F: Photographs of field sites ................................................................................120  v  List of Tables Table 1. List of water, sediment and biofilm sampling stations and date range they were sampled. ............................................................................................................ 15 Table 2. Wet and dry season classifications during study period. .................................... 23 Table 3. Summary of Total Impervious Area (TIA) data for the Marshall Creek watershed and subwatersheds of interest. .......................................................................... 25 Table 4. 1991-2006 animal numbers from Statistics Canada for the Abbotsford Aquifer and Sumas Prairie regions of the lower Fraser Valley. .................................... 26 Table 5. Estimated 2006 livestock numbers in the Marshall Creek Watershed. .............. 27 Table 6. Estimated 2006 livestock numbers in the agricultural subwatershed within the Marshall Creek watershed. ............................................................................... 27 Table 7. Average Animal Unit Equivalents (AUEs) calculated for the Abbotsford Aquifer and Sumas Prairie regions of the lower Fraser Valley, 1996-2006. ................. 28 Table 8. Average 2006 Animal Unit Equivalents in the Marshall Creek Watershed and the agricultural subwatershed. .......................................................................... 29 Table 9. Population estimate for 12 Census Tracts that overlap the Marshall Creek watershed boundary (StatsCan, 2010). The * indicates those CTs that fall entirely within the watershed boundary. .......................................................... 30 Table 10. Historic population estimates for 12 Census Tracts that overlap the Marshall Creek catchment area (StatsCan, 2001b; StatsCan, 1996b). ............................ 30 Table 11. Population estimates for the urban and agricultural subwatersheds of the Marshall Creek watershed. ............................................................................... 31 Table 12. Seasonal water quality averages across the Marshall Creek watershed. .......... 31 Table 13. Significant Spearman’s Rank correlations (|Spearman’s coefficient (R)| > 0.4, p<0.05) for water parameters (wet season). ..................................................... 37 Table 14. Significant Spearman’s Rank correlations (|Spearman’s coefficient (R)| > 0.4, p<0.00091) for water parameters (wet season). ............................................... 37 Table 15. Significant Spearman’s Rank correlations (|Spearman’s coefficient (R)| > 0.4, p<0.05) for water parameters (dry season). ...................................................... 38 Table 16. Significant Mann-Whitney U results (p<0.0167) for water quality parameters by site. .............................................................................................................. 39  vi  Table 17. Sediment quality averages in the Marshall Creek watershed and Interim Sediment Quality Guidelines (ISQG) for particular trace elements (Canadian Council of Ministers of the Environment, 1999). ............................................ 48 Table 18. Significant Mann-Whitney U results (p<0.0167) for parameters in suspended sediments by site. ............................................................................................. 49 Table 19. Significant Mann-Whitney U results (p<0.0167) in the wet season for parameters in suspended sediments by site. ..................................................... 49 Table 20. Mean chlorophyll a concentrations (mg/L ± 1 standard error) of biofilms collected at the end of each season in the Marshall Creek watershed. ............. 54 Table 21. Summary of trace elements measured in biofilms collected in the Marshall Creek watershed. (n=100) ................................................................................ 60 Table 22. Spearman’s rank correlations between average biofilm chl-a and water quality indicators. (R=correlation coefficient, p=probability, n=sample size)............. 60 Table 23. Average metal concentrations (mg/kg) ± 1 S.E. in suspended sediments and biofilms collected from three sites in the Marshall Creek watershed. ............. 61 Table 24. Trace element ratios for the pooled dataset of biofilms and sediments in the Marshall Creek watershed. ............................................................................... 63 Table 25. Trace element ratios in biofilms and sediments by site (Marshall mainstem (Mar.), urban tributary (Urb.) and agricultural tributary (Ag.)). ...................... 64 Table 26. Summary of land use indicators for the Marshall Creek watershed and subwatersheds of interest.................................................................................. 67 Table 27. Summary of in-stream parameters measured at three sites. (Mean ± 1. S.E.) .. 67  vii  List of Figures Figure 1. Map of the Marshall Creek watershed, located in the lower Fraser Valley of British Columbia. ............................................................................................. 12 Figure 2. Map of sampling sites in the Marshall Creek watershed. .................................. 15 Figure 3. Box-whisker plot showing upstream to downstream trend of nitrate-N in water collected in Marshall Creek August 2008 – May 2010. Error bars represent maximum and minimum values. (n=2-16) ....................................................... 32 Figure 4. Box-whisker plot showing upstream to downstream trend of SRP in water collected in Marshall Creek August 2008 – May 2010. Error bars represent maximum and minimum values (n=2-16). ....................................................... 34 Figure 5. Box-whisker plot showing upstream to downstream trend of TOC in water collected in Marshall Creek August 2008 – May 2010. Error bars represent maximum and minimum values (n=1-13). ....................................................... 35 Figure 6. Box-whisker plot showing upstream to downstream trend of TOC in water collected in Marshall Creek August 2008 – May 2010. Error bars represent maximum and minimum values. ...................................................................... 36 Figure 7. Average concentrations of 4 water quality parameters (± 1 S.E.) in the wet season and dry season collected at 4 sites in the Marshall Creek watershed (HW=Headwaters – Site 11). ........................................................................... 40 Figure 8. Nitrate-N data from water samples collected at 2 sites on Marshall Creek between 1994 and 2010. Note the breaks in the time series on the x-axis, indicated by the vertical lines. .......................................................................... 43 Figure 9. Trace element concentrations: a) Cu, b) Mn, c) Zn and d) P, measured in bed sediments collected in late summer in the Marshall Creek watershed. Each column represents one sample. No Mn data for 2008 at the agricultural tributary. (ISQG=Interim Sediment Quality Guideline) .................................. 46 Figure 10. Average chl-a concentrations (mg/L) in 3 sample sites over three seasonal sample collections. Error bars represent max/min values. .............................. 51 Figure 11. Average chlorophyll a concentrations of biofilms collected from urban, agricultural and Marshall mainstem stations in summer 2009 (a, n=2), winter 2009-10 (b, n=3) and spring 2010 (c, n=3) after x days of colonization. Mean ± 1 standard error. Note: Log scale in Figures a and b. ....................................... 53 Figure 12. Water temperatures at three sites during the winter biofilm collection period. .......................................................................................................................... 55  viii  Figure 13. Average biofilm dry mass (mg) in 3 sample sites over three seasonal sample collections. Error bars represent max/min values. ........................................... 56 Figure 14. Average dry mass (mg) of biofilms collected from urban, agricultural and Marshall mainstem stations in summer 2009 (a, n=2), winter 2009-10 (b, n=3) and spring 2010 (c, n=3) after x days of colonization. Mean ± 1 standard error. .......................................................................................................................... 58 Figure 15. Box-whisker plot showing P concentrations in bed (n=1) and suspended sediments (n=7) and biofilms (n=34) and SRP (n=17) in water samples collected from three sites in the Marshall Creek watershed. (Error bars represent min/max values). Note: Water SRP concentrations are on the righthand vertical axis. ............................................................................................. 62  ix  List of Abbreviations and Symbols Al ALR ANOVA AUE BC Ca Cd CEA Chl-a ClCr CT Cu CWN DA DO EIA EU Fe GIS GW ha HCl HNO3ICP-AES ISQG K Ln Log M-W U Mg mg/kg mg/L mm Mn N  Aluminum Agricultural Land Reserve Analysis of Variance Animal Unit Equivalent British Columbia Calcium Cadmium Cumulative Effects Assessment Chlorophyll a Chloride Chromium Census Tract Copper Canadian Water Network Dissemination Area Dissolved Oxygen Environmental Impact Assessment European Union Iron Geographic Information Systems Groundwater Hectare Hydrochloric acid Nitric acid Inductively Coupled Plasma - Atomic Emission Spectroscopy Interim Sediment Quality Guidelines Potassium Natural logarithm Logarithm Mann-Whitney U Test Magnesium Milligrams per kilogram Milligrams per liter Millimeter = 10-3 meters Manganese Nitrogen x  n NH4+-N Ni NO3--N NTU P p Pb Pct ppm R R2 SE Sp. Cond. SRP TIA TOC US EPA USDA VEC Z Zn α µg/L µm µS/cm ºC  Sample size Ammonia-Nitrogen Nickel Nitrate-Nitrogen Nephelometric Turbidity Units Phosphorus Probability level (statistics) Lead Percentile Parts per million Correlation coefficient (statistics) Coefficient of determination (statistics) Standard Error Specific Conductivity Soluble Reactive Phosphorus Total Impervious Area Total Organic Carbon United States Environmental Protection Agency United States Department of Agriculture Valued Ecosystem Component Standard score (statistics) Zinc Alpha (significance level) Micrograms per liter Micrometer = 10-6 meters MicroSiemens per centimeter - units of conductivity Degrees Celsius  xi  Acknowledgements I am grateful to the following agencies for providing research funding: the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Canadian Water Network (CWN) National Centre of Excellence. The following individuals assisted by providing information and expertise: Monique Dubé, Nicole Nadorozny, Gila Somers and Kelly Munkittrick, CWN; Frank Wright and Siamak Meghdadi, City of Abbotsford; and Steve Arnold and Charlotte Rosenau, Fraser Valley Trout Hatchery. My supervisor, Hans Schreier, has provided an immense amount of guidance and wisdom over the years and I have to credit him for being the inspiration for my entrance into the “water world”. I cannot thank him enough for his unfailing support and for being such a pleasure to work with. I would also like to thank my other committee member, John Richardson for guiding me into the new realm of biofilms and for making his lab available to me. I thank Les Lavkulich for all his knowledge and assistance, and for many a chat over coffee. A large number of people helped me out in the field, in the lab and around the office. They include: Sandra Brown, Emma Holmes, Trudy Naugler, Mathieu Beaulieu, Nathalie Maurer, Maria-Cecilia Roa, Maureen Soon, Santiago Larrañaga, Martin Hilmer, Paula Parkinson and Gina Bestbier. I am grateful to my Mom and Dad for providing me with support and encouragement in all my endeavours. Special thanks to Brock for his love and understanding and for helping to keep things in perspective.  xii  1  Introduction  Growth in human populations and material well-being is resulting in unprecedented and largely unknown impacts on the natural resource base. Two activities that are considered to have impact are intensification of agriculture and the increase in the area and density of urban development (Foley et al., 2005; Alig, Kline & Lichtenstein, 2004). To sustain this growth reality, the management of water resources, including water quality and security, are paramount. The environmental impacts of these human pressures have been studied extensively in the literature (Kuusisto-Hjort, 2009; Cianfrani, Hession & Rizzo, 2006; Wang & Kanehl, 2003; Schueler, 1994) , but are poorly understood at a watershed scale and emergent issues continue to surface. Although the categorization of water pollution by origin of point and non-point sources is helpful, they are not exclusive of each other, and it is generally agreed that the effects of point sources are more readily assessed and mitigated by design and by the application of appropriate technology. However, these management actions often transfer the problem from one resource sector to another. For example, the application of manure to croplands to increase productivity can result in surface water contamination if applied in excess or at the wrong time of year (BC Ministry of Environment, 1992).  1.1  Cumulative Effects  Non-point sources of pollution are generally considered to be an accumulation of smallscale point sources. Therefore, they might not have a measurable effect individually, but when considering the incremental, accumulating and interacting impacts of all point and non-point sources in an entire watershed, there is a measurable cumulative effect (Canadian Environmental Assessment Agency (CEAA), 1999). At a watershed scale, cumulative effects are difficult to relate back to a specific land use activity due to the extensive nature of non-point source pollution. Thus, preventing non-point sources of pollution and assessing cumulative effects are some of the most challenging problems facing watershed managers. Cumulative effects 1  assessment (CEA) is an important component of environmental impact assessment (EIA) required under the Canadian Environmental Assessment Act (Duinker & Greig, 2006). EIAs are performed to inform decision-makers about the likely environmental consequences of alternative development options (Duinker & Greig, 2006). In CEA, proponents of development are responsible for examining the cumulative effects associated with their own proposed development together with relevant past, present and potential future projects in the watershed. The prediction of cumulative effects is very complicated however, as these effects are often not simply additive but may be synergistic or antagonistic (Dubé, 2009). When performing EIAs and CEAs, valued ecosystem components (VEC) are identified to help provide focus. VECs are components or indicators of interest to the “community” that require protection from any potential effects of the proposed project activities (Dubé, 2009). Examples of VECs in aquatic ecosystems include cultural VECs such as aesthetic resources, biological VECs such as fish and fish habitat, physical VECs such as surface water quantity and quality. The latter are often related to established water quality guidelines for specific chemical or microbiological parameters (i.e. nitrate or fecal coliform). It is important to note that these VECs are interconnected in aquatic ecosystems. There are several criticisms of the current system of CEA in Canada (Duinker & Greig, 2006; Kennett, 2000). EIAs are currently required on a project-by-project basis; however, because of the large-scale, integrated nature of cumulative effects, it has been argued that CEA is not suitable for assessments of individual projects. It is important to investigate the full range of human-generated stresses on the VEC, including point and non-point sources of pollution. VEC conservation depends upon ensuring that the cumulative effects of all stresses are kept within tolerable and acceptable levels.  2  1.2  Watershed CEA Framework  Because of the shortcomings of the current CEA process in Canada, Dubé (2009) has proposed a new framework for watershed-scale CEA. The aim of this framework is to integrate information from existing watershed management programs, inform these sources when additional information is required and provide relational information to these sources for effects assessment and resource management. The watershed CEA framework development is a Canadian Water Network (CWN) project involving researchers and Canadian graduate students. Portions of this framework are implemented in 6 different watersheds across Canada. This study examines the cumulative effects in a small watershed in the lower Fraser Valley with mixed land uses, which is a particularly challenging setting for CEA. The first component of the watershed CEA framework is a “state of the watershed” assessment which involves comparison of the current state of a VEC to a previous state or condition (historical reference), an alternate state (spatial reference) or to a statement of acceptability (guideline or objective) (Dubé, 2009). The condition of VECs can be measured by monitoring indictors known as measurement endpoints. The connections between the measurement endpoint and the VEC have often been estimated through experimentation. For instance, high concentrations of phosphorus measured in stream bed sediments (measurement endpoint) do not necessarily have an impact on local fish abundance (VEC); however, under anoxic conditions the release of phosphorus to the water column has been shown to cause aquatic ecosystems to shift to a eutrophic state, which reduces dissolved oxygen levels and can directly affect the health of fish and also fish abundance (Schindler, 2006). The measurement endpoint value can be compared to a guideline or alternate state and used to calculate the magnitude of change in the watershed (Dubé, 2009). Indicators of land use activities or landscape stressors are also monitored to track changes in the assumed stressors. The state of the watershed assessment is an imperative first step, as the monitoring program produces data that is required for the second component of the CEA framework: relational analysis.  3  Relational analysis determines how or if a stressor (x) affects a response (y). In reality, it is not likely a single stressor affecting a single response, but most responses in stream systems are related to multiple stressors. The relative contributions of each stressor to the observed response are also largely unknown in many systems. An example of a crude stressor-response relationship that is accepted by watershed managers is the relationship between imperviousness and several indices of biotic integrity (Schueler, Fraley-McNeal & Cappiella, 2009; Cianfrani et al., 2006; Novotny et al., 2005). It is generally agreed that when impervious surface area in a watershed exceeds 8-12%, the aquatic biodiversity drops significantly (Wang & Kanehl, 2003; Wang, Lyons & Kanehl, 2001; Schueler, 1994). However, the input of contaminants into an urban stream is highly variable, and depends on not only imperviousness but also on land use activities, traffic density, local topography and climatic conditions.  1.3  Study Area  The lower Fraser Valley in British Columbia serves as an ideal study location. It is a region of rapid urban growth and agricultural intensification. The area’s major agricultural products are poultry, dairy, forage crops, turf and specialty food crops including berries and vegetables (Statistics Canada (StatsCan), 2006). It is an economically important agricultural region and commodity production is enhanced by the applications of chemical amendments and manure. Both of these land-use activities are primarily non-point sources of pollution and it is therefore difficult to identify “cause and effect” relationships between stressor and response variables in the system. The University of British Columbia has conducted land use-water quality studies for several decades in the region. Much of this work has been summarized by Lavkulich, Hall & Schreier (1999), Berka (1996) and Smith (2004). These studies provide an excellent base line against which continued research efforts may be compared. One major convergent conclusion of these studies is that cumulative effects, notably from intensive agriculture in the productive lowlands and rapid urbanization of the uplands, are a serious concern for the region (Schreier, 2009; Smith, 2004; Lavkulich et al., 1999; Berka, 1996). This review of the historical information provided guidelines for the identification of a 4  watershed, Marshall Creek, as an exemplary study site to focus research on cumulative effects of urban development and intensifying agricultural operations.  1.4  Land use and stream quality  Transportation, industrial production, housing, building construction, the discharge of treated wastewater and combined sewer overflows in urban areas are known to pollute freshwater systems through point and non-point sources. In addition, the covering of soils by impervious surfaces interferes with the functioning of soils and the natural hydrological cycle by reducing infiltration and filtering of rainwater (Kuusisto-Hjort, 2009). During dry periods, metals and organic materials such as oil and grease from different sources accumulate on impervious surfaces. When rainfall commences, these pollutants are effectively washed off the surfaces to the stormwater conveyance system and receiving water bodies. This is particularly true during storm events. Trace elements (i.e. Cu, Ni, P, Zn) often adsorb to fine sediments and can also form complexes with dissolved organic matter in runoff (Murakami et al., 2009). Within the stream system, these particles can settle out and enrich bed sediments with trace element concentrations, which can impact benthic communities (Courtney & Clements, 2002). Association with suspended sediments affects the bioavailability, transport and fate of trace elements (Bibby & Webster-Brown, 2006; Boenigk, Weidlroither & Pfandl, 2005) and nutrients in the stream system (Stutter, Langan & Demars, 2007). It is well documented that in most fluvial systems, suspended sediment is transported primarily as larger aggregates (or flocs) composed of a heterogeneous mix of particle sizes and also have organic particles associated with them, such as microorganisms and detritus (Bibby & Webster-Brown, 2006; Phillips, Russell & Walling, 2000; Droppo & Ongley, 1994). Suspended sediments can become trapped in biofilms and organisms that feed on these biofilms may ingest associated trace elements and lead to bioaccumulation (Boenigk et al., 2005; Courtney & Clements, 2002). Turf grass in residential, commercial and public landscapes is commonly fertilized with inorganic N and P fertilizers during production and after application. Richards et al. (2008) have shown that this can contribute 5  significant loadings of total P, nitrate/nitrite-N and organic N to urban streams, particularly due to increased surface runoff and sediment loading compared to undisturbed conditions. Intensive agriculture poses a high risk of soil and water contamination due to excess nutrient, organic and trace element inputs (Smith et al., 2007). Livestock and poultry manure is often enriched in nutrients (N and P), trace elements and antibiotics added to the feedstock to facilitate weight increase and disease prevention (Han et al., 2000). Livestock number, type and density in a watershed can be used to indicate pollution from manure leaching and runoff (United States Department of Agriculture (USDA), 2008). The Animal Unit Equivalent (AUE) is a means of expressing different types and classes of livestock in a common form, based on their live weight, the amount of manure they produce and the amount of nitrogen in manure (USDA, 2008; Werry, 2003). For example, one horse produces an equivalent amount of manure as 150 laying hens or 6 veal calves. Stockpiling of manure and poorly timed land application of manure can lead to the contamination of surface and ground waters via runoff and leaching of nutrients and trace elements (Steinmetz et al., 2009; 87 Vadas, Haggard & Gburek, 2005; Han et al., 2000). Over-application of inorganic fertilizers on agricultural land contributes to nutrient contamination of surface and ground waters. Sediments from agricultural areas are responsible for the supply of nutrients, pesticides and heavy metal contaminants to streams (Ward, 2008). Excess concentrations of nutrients, particularly P, in surface waters from agricultural land use have led to eutrophication in rivers and estuaries in many parts of the world (Maier et al., 2009; Smith, 2004). Nitrate-N is a negatively charged ion that readily leaches through soils into unconfined aquifers. Nitratecontaminated groundwater supplies are a problem in agricultural watersheds worldwide (McMahon et al., 2008; Almasri & Kaluarachchi, 2004; Wassenaar, 1995).  6  1.5  Methods of Assessment  Traditional approaches to assessing the effects of contaminants from different land use activities have focused on measuring them in the water column and sediments in aquatic systems. These are useful to make preliminary assessments of water contamination but are, at best, snapshots of events and of specific concern when researching streams and other similar water bodies. Historically aquatic organisms, including benthic macroinvertebrates and small fish, have been utilized to assess aquatic ecosystem health (Munkittrick & Dixon, 1989; Karr & Dudley, 1981). It is generally concluded that various species or taxa respond differently to metal contents in the water body (Courtney & Clements, 2002; Clements, 1994) and therefore the use of a single taxon may not represent the overall health of the system. Common indicators in aquatic systems have focused on those that might be traced back to the origin, e.g. N and P from agricultural activities, or metals such as Cr and Mn that are more commonly associated with urban sources (Kuusisto-Hjort, 2009). These are but indicators and not unique in their ability to separate alternate land use activities. Another reality is that most studies collect samples from the aquatic systems at fixed time intervals. These “grab samples” have limitations when one wishes to compare stream water quality indicators temporally and spatially. This is of particular concern when attempts are made to assess sediment quality. Of the parameters commonly used, many are very dynamic and change with stream conditions such as temperature or oxidationreduction status (e.g. nitrate-N). Other parameters, such as many of the metals, have complex reaction pathways in aquatic and terrestrial systems by being subject to partitioning between different components within the aqueous system (Ryder et al., 2007; Holding, Gill & Carter, 2003). For example some metals such as Cu, and Zn, will partition between the organic (including the living biota) and inorganic components of the system (Lofts & Tipping, 1999). Thus the identification and application of techniques that are both indicative of the status of the stream and provide some temporal integration are needed. Two emerging techniques are suspended sediment and biofilm collection.  7  Increasingly, researchers have suggested that what is needed is a biological measurement endpoint that integrates contaminant effects on a temporal scale, yet is not destroyed in the sampling process, and one that is easily implemented and replicated in the field (Burns & Ryder, 2001). One such technique has been the adoption of the use of biofilms as an indicator of aquatic ecosystem health. In this study, biofilms are a collective term referring to the organic coatings that consist of bacteria, periphyton and extracellular polymeric substances that are ubiquitous in aquatic systems (Holding et al., 2003; Schorer & Eisele, 1997). As Battin et al. (2003) state, “Biofilms bring hydrodynamic retention and biochemical processes into close spatial proximity and influence biogeochemical processes and patterns in streams”. Trace metal concentrations in sediments have been found to be positively correlated with concentrations in biofilms in streams, which would indicate metal exchange through sorption and desorption processes between the sediments and biofilms (Holding et al., 2003; Schorer & Eisele, 1997). The use of artificial materials of known composition and size as a medium for biofilm colonization has been shown to provide quality information on nutrients and metal pollutants in stream water (Kröpfl et al., 2006). Although several studies have investigated water and sediment quality in catchments with mixed land uses (Stutter et al., 2007; Coulter, Kolka & Thompson, 2004; Dougherty, Dymond & Zipper, 2004), very few studies have linked measurement endpoints to particular landscape stressors in separate tributaries, and also looked at their cumulative effects downstream. Also, few studies have looked at the relationship between trace elements in suspended sediments and biofilms (Hodgson, 1983). Some studies that have demonstrated these linkages involved both point and non-point sources of pollution, where effects from point sources are relatively easy to isolate and identify (Bowes et al., 2005). The Marshall Creek watershed provides a good model watershed in the lower Fraser Valley because it contains examples of both point and non-point sources of pollution and therefore one that lends itself to cumulative effects assessment. It also provides the unique opportunity to look at urban and agricultural land use activities in isolation and combined together in the same catchment as non-point sources of pollution.  8  In addition, historic water quality data aid in the determination of changes in the stream related to land use change over time.  9  2  Objectives  This study investigates the cumulative effects of both urban and agricultural activities through the integration of multiple response indicators and relating them to surrounding land use stressors within the Marshall Creek watershed within a seasonal and historical variability framework. The specific objectives are to: 1. Determine upstream to downstream trends and cumulative effects on selected indicators in the Marshall Creek main channel. 2. Estimate the contribution of nutrients in groundwater versus surface water from agricultural land use. 3. Document historical changes of nitrate-N over time at particular sample sites (19942010). 4. Determine quantities of nutrients in water and trace elements in bed and suspended sediments at sites influenced by urban, agricultural and mixed land uses. 5. Determine the growth, productivity and trace element accumulation in biofilms at sites influenced by urban, agricultural and mixed land uses and relate them to water and sediment quality indicators. 6. Determine seasonal differences in these indicators between agricultural, urban and mixed sites.  10  3  Background  3.1  Study Site  This study was conducted in the Marshall Creek watershed, which is in the southeastern portion of the City of Abbotsford in the lower Fraser Valley of British Columbia, Canada (49º01’57” N 122º15’23” W). (Marshall Creek is also known locally as Lonzo Creek). Marshall Creek is a third order stream that flows in a northeast direction into the Sumas River, and eventually joins the Fraser River (Figure 1). Marshall Creek is approximately 13 km long. The watershed is approximately 3,800 hectares in area and contains about 15 small tributary branches that feed into the mainstem of Marshall Creek. The majority of the mainstem is located in the lowland portion of the watershed and winds its way through agricultural land. The other smaller creek tributaries feed down from the steep, sloped upland areas, which are being urbanized at a rapid rate. The topography ranges from El. 460 m (Sumas Mountain) to El. 5 m (agricultural lowlands) above mean sea level.  11  Figure 1. Map of the Marshall Creek watershed, located in the lower Fraser Valley of British Columbia.  Baseflow in the Marshall Creek mainstem originates from the Abbotsford Aquifer that passes through the Fraser Valley Trout Hatchery. The hatchery discharges approximately 50 L/second to the creek year-round (Wright, personal communication, 2009). This significantly improves summer flows and upstream water temperature is also quite cool, as it originates from deep wells in the Abbotsford Aquifer and from direct groundwater discharge. Surface runoff also enters Marshall Creek, particularly during the winter when precipitation is high. The impacts of these two different water sources are of interest because seasonal shifts in water quality have potential to stress the aquatic biota. Environment Canada and Fisheries & Oceans Canada recognize Marshall Creek as a fishbearing stream (Integrated Resource Consultants, 1994). Marshall Creek surveys have 12  observed brook trout (Salvelinus fontinalis), rainbow trout (Oncorhynchus mykiss) and cutthroat trout (Oncorhynchus clarki) in the past (Bryan & Larkin, 1972), and more recent surveys by the BC Ministry of Environment have found sticklebacks (general), sculpins (general) and some salmon species including chum (Oncorhynchus keta) and coho (Oncorhynchus kisutch) (BC Ministry of Environment, 2007). The water level in Marshall Creek is regulated by the City of Abbotsford in the form of gravity drain floodgates at Barrowtown Pump Station, which is located approximately 9 km downstream of the confluence of Marshall Creek with the Sumas River. The floodgates are closed between May and September each year (exact dates vary) for the purpose of storing irrigation water in the dry season (Wright, personal communication, 2009). During this time the water at the mouth of Marshall Creek is quite stagnant and has been observed to flow in reverse at times (Wright, personal communication, 2009). The floodgates are always opened by September 15 each year to allow passage of migrating salmon into the Sumas River and its tributaries, including Marshall Creek (Wright, personal communication, 2009). The lower Fraser Valley has the highest livestock density in Canada and the City of Abbotsford has intensified many of its livestock operations, particularly poultry (Beaulieu, 2001). As a result, the Abbotsford area has a surplus of nitrogen and phosphorus from excess manure production (Schreier, Bestbier & Derksen, 2003). Kowalenko, Schmidt & Hughes-Games (2007) determined that 11-47% of agricultural soils in the Abbotsford area fall into the high or very high environmental class risk based on kg of NO3-N per hectare, and that 82-89% of soils fall into the high or very high class risk based on mg P per kg of soil. These excess nutrients have contributed to eutrophication in Marshall Creek, which is commonly observed in late summer (Smith, 2004). In addition, the City of Abbotsford’s population grew from 68,000 to 130,000 between 1986 and 2006 (StatsCan, 2010). This growth has been restricted to the upland areas of Sumas Mountain, since much of the lowland area is in the Agricultural Land Reserve (ALR) and is restricted to agricultural uses (Wright, personal communication, 2009). Marshall Creek receives runoff from this rapidly growing urban residential 13  development on Sumas Mountain. This watershed is influenced by multiple, diffuse sources of pollution and the linkages between different land uses and in-stream measurement endpoints have not been investigated. This study focused on three primary landscape stressors: 1) Intensive agricultural production over the Abbotsford Aquifer. Although much of this land base is largely outside of the watershed boundary, groundwater from the aquifer provides continuous baseflow to the Marshall Creek mainstem. 2) Intensive agricultural production in the Marshall Creek watershed produces contaminants that enter surface water via surface runoff, drainage tiles and/or subsurface discharge. 3) Urban residential development in the Marshall Creek watershed. Population growth on the uplands has resulted in forest clearing and significant increases of impervious surfaces, which alters the runoff regime and contributes contaminants to Marshall Creek via stormwater discharge.  3.2  Sampling Site Selection  Sampling stations were selected in the Marshall Creek watershed to represent the distinct influences of the groundwater discharge from the Abbotsford Aquifer and from urban and agricultural land uses (Figure 2). Two sites on the mainstem of Marshall Creek (11 and 2) were chosen based on locations sampled in the past in order to make historical comparisons of the results. A total of 18 stations were sampled for water quality and 14 for bed sediments from summer 2008 to spring 2010. Seven of these sites are located on the mainstem of Marshall Creek, 2 are tributaries influenced by groundwater springs, 5 are in urban tributaries (& ponds) and 2 are in agricultural tributaries (Table 1). Biofilms and suspended sediments were sampled at site #2 (Marshall Creek directly upstream from Sumas River confluence), site #8 (urban tributary in DeLair Park) and site #22 (agricultural tributary on Sumas Prairie). These three sites are the focus of the majority of data analyses. 14  Figure 2. Map of sampling sites in the Marshall Creek watershed.  Table 1. List of water, sediment and biofilm sampling stations and date range they were sampled.  Station  Site ID  Date range sampled  Groundwater (Hatchery well) Tributary (GW influence) Tributary (GW influence) Marshall Creek (Hatchery discharge) Marshall Creek Marshall Creek  12  Aug 08 – Jun 09  13  Aug 08 – Dec 08  10  Aug 08 – Sep 09  14  Aug 08 – May 10  15 11  Aug 08 – May 10 Aug 08 – May 10  Marshall Creek  17  Aug 08 – May 10  # grab samples (w - water, s - bed sediments) w–2  Biofilms/Suspended Sediments  w–2 s–1 w–3 s–2 w–8 s–2  No  w–5 w – 12 s–3 w–7 s–3  No No  No  No No  No  15  Station  Site ID  Date range sampled  Marshall Creek  18  Aug 08 – May 10  Marshall Creek  7  Aug 08 – May 10  Marshall Creek (Downstream) Urban tributary  2  Aug 08 – May 10  3  Aug 08 – Sep 09  Urban tributary  4  Aug 08 – Sep 09  Urban Detention pond Urban Tributary (Junction Creek) Urban tributary  5  Aug 08 – Dec 08  23  Jun 09 – Sep 09  8  Aug 08 – May 10  Agricultural tributary Agricultural tributary Total  16  Aug 08 – Sep 09  22  Jun 09 – May 10  17 sites  # grab samples (w - water, s - bed sediments) w – 11 s–3 w–7 s–3 w – 17 s–3 w–4 s–2 w–4 s–1 w–2 s–1 w–2 s–1 w – 18 s–3 w–4 s–3 w – 16 s–2 w – 124 samples (not including historic) s – 33 samples  Biofilms/Suspended Sediments No No Yes No No No No Yes No Yes  16  4  Methods  4.1 4.1.1  Land use indices and climate data Meteorological and hydrometric data  Precipitation data were obtained from the Abbotsford Airport climate station about 7 km west of the Marshall Creek watershed (Environment Canada, 2010). There are no public hydrometric stations in the Marshall Creek watershed. Odyssey water depth loggers were installed in the stream channel in the Marshall mainstem (site 2), in the urban tributary (site 8) and in the agricultural tributary (site 22) (Dataflow Systems Ltd., Christchurch, NZ). 4.1.2  Geographic Information Systems (GIS) data  All GIS data were analyzed using ArcGIS (v.9). A 2008 digital orthophoto and spatial data were provided by the City of Abbotsford for land use assessment. The Marshall Creek watershed boundary was delineated based upon the City of Abbotsford’s Integrated Stormwater Management Plan (Wright, personal communication, 2009). Subwatershed boundaries for the urban and agricultural tributaries were delineated with the assistance of the City of Abbotsford (Wright, personal communication, 2009). Impervious surfaces were delineated and dwellings/farm operations were counted for the urban and agricultural subwatersheds using ArcMap v. 9.3. 4.1.3  Agricultural census data and Animal Unit Equivalents (AUEs)  All census data were obtained from Statistics Canada’s Agricultural Censuses (StatsCan, 2006; StatsCan, 2001a; StatsCan, 1996a; StatsCan, 1991). Statistics Canada data for the Abbotsford Aquifer (Matsqui South) and the Sumas Prairie (Abbotsford) sub-regions of the Fraser Valley Census Division were used to explore livestock densities. A map of the sub-region boundaries for the 2006 census is presented in Appendix A. For 2006, census data from dissemination area (DA) 59090721 (Appendix A) were used to estimate the number of animals in the Sumas Prairie portion of the Marshall Creek 17  watershed, due to its greater overlap with the watershed boundary. The data from this DA were used to estimate the livestock numbers within the Marshall Creek watershed. The South Matsqui sub-region was used to estimate animal numbers in the Abbotsford Aquifer portion of the watershed. An average number of each type of livestock per farm was calculated for each area. Then, using GIS, the number of farms in the watershed was counted. With this known number of farms (assuming minimal change between 2006 and 2008), the 2006 livestock numbers in the Marshall Creek catchment were estimated. Estimations were made for 2001 and 1996 as well based on the number of animals and farm dwellings in those years. In order to compare the livestock intensity between different areas, each type of livestock was converted to an Animal Unit Equivalent (AUE). The number of AUEs per hectare of farmland for each area was calculated to help determine livestock intensity. The calculated AUEs/ha for each area were then compared to AUE guidelines set by agricultural regulators (USDA, 2008; European Commission 2002). 4.1.4  Population data  Human population data were estimated for 1996, 2001 and 2006 using 11 Statistics Canada Census Tracts (CTs) that overlap with the Marshall Creek watershed boundary (StatsCan, 2010; StatsCan, 2001b; StatsCan, 1996b). (See Appendix A for a map of the CT boundaries). The 2006 population within the watershed was estimated by counting the number of dwellings in each CT that lie inside the watershed and multiplying it by an estimate of the number of people per dwelling. 4.2 4.2.1  Field Sampling and Laboratory Analysis Water sampling and laboratory analysis  Water samples were collected from various sites throughout the Marshall Creek watershed on 18 occasions (summer & winter 2008, summer 2009 – spring 2010). Grab samples were collected in clean acid-washed 250 mL Nalgene bottles and were stored on ice in a cooler during transportation to the laboratory. All samples were collected after rinsing the bottle thoroughly with stream water. Parameters that were measured in situ include specific conductivity and temperature using a YSI 30 conductivity meter and 18  dissolved oxygen using a YSI 95 dissolved oxygen meter (YSI Inc., Yellow Springs, OH). Samples were refrigerated prior to analysis. In the laboratory, pH and turbidity were measured within 12 hours of sample collection using the Orion model 420A pH meter (Thermo Scientific, Waltham, MA) and the Hach 2100P turbidimeter (Hach Co., Düsseldorf, Germany), respectively. All samples were measured 3-5 times and averaged for a representative reading. The water samples were divided into two, where one half was immediately frozen and later analyzed for total and organic carbon (TOC). Samples were thawed and filtered using Whatman #41 filter paper and analyzed using a Shimadzu (TOC-500) Total Organic Carbon Analyzer. The second half of the samples were filtered using Whatman #42 filter paper within 12 hours of sample collection. These filtered samples were then frozen for up to 2 months before analysis. Chloride (Cl-), nitrate/nitrite-nitrogen (NO3--N), total ammonia as ammonium (NH4+-N) and SRP were analyzed on a LaChat XYZ QuickChemAE autoanalyzer. NO3--N was analyzed using method #12-107-04-1-B, method #10-107-062-A for NH4+-N, method #10-115-01-1-A for SRP and method #10-117-07-1-A for Cl-. 4.2.2  Bed and suspended sediment sampling and laboratory analysis  Grab samples of bed sediment were collected on August 18, 2008 (n=13), June 18, 2009 (n=10) and September 17, 2009 (n=13) during low flow conditions. These sites were a subset of the water sampling sites. The samples were selected from the surface sediment layer of the streambed in the centre of the waterway. The samples were placed into labeled plastic bags and transported to the laboratory on ice in a cooler. Suspended sediments were collected at three sites in the watershed using a timeintegrated suspended sediment sampler (Phillips et al., 2000). The samplers were installed in the stream channel and were sampled 9 times between July 2009 and May 2010. The suspended sediments were emptied from the samplers after 27 – 75 days of  19  collection. The contents of the samplers were emptied into 4 L acid-washed Nalgene bottles and transported to the laboratory on ice in a cooler. In the laboratory, the sediment samples were wet-sieved through stainless steel sieves in order to obtain the <63 µm fraction, which includes both silt and clay particles. This fraction of sediment usually contains the greatest concentrations of trace elements and represents the fraction of sediment that is transported in suspension (Walling, 1989; Horowitz & Elrick, 1987). The sieved sediment was placed into acid-washed beakers, which were oven dried at approximately 70°C overnight. Once fully dried, the beakers were allowed to cool for 30 minutes. A mortar and pestle was used to break apart large aggregates of the dried sediment and the samples were stored in sealed labeled plastic containers until analysis. The bed and suspended sediment samples were prepared for elemental analysis using the US EPA method #200.2 for sediments (United States Environmental Protection Agency (US EPA), 1994). This aqua-regia digestion method provides data in the form of total recoverable elements. Appendix C contains a summary of this method. The digested samples were analyzed using Inductively Coupled Plasma – Atomic Emission Spectroscopy (ICP–AES) (Varian 725 – ES). 4.2.3  Biofilm collection and laboratory analysis  Unglazed, ceramic (terracotta) tiles (100 cm2) were used as artificial substrates for biofilm colonization. At each sampling station, wooden trays housing 4 sets of tiles were immersed in the water column. A photograph of the tile apparatus is presented in Appendix D. Sets of 4 or 6 tiles were collected after four successive time durations to examine colonization and growth over time. One set was collected after 11, 19, 27 and 34 days of colonization (Aug – Sep ‘09), 10, 18, 27 and 38 days (Dec ‘09 – Jan ‘10), and after 11, 19, 27 and 35 days (Apr – May ‘10). During the first set run in August 2009, only 4 were used per set. Half of the tiles in each set were analyzed for chlorophyll a (2 or 3 replicates), while the other half were analyzed for total metal concentrations and dry mass (2 or 3 replicates).  20  On each collection day, one set of tiles was randomly removed from the wooden trays at each sampling station. The biofilm material was scraped off of each tile with an acidwashed, medium bristle toothbrush. The material was rinsed with deionized water into a dark, 60 mL acid-washed Nalgene bottle. The samples were stored in a cooler during transportation to the laboratory. Water samples were collected at each site when the biofilms were collected. The terracotta tiles were acid-bathed in 5% nitric acid between each seasonal immersion in the field to remove any traces of biofilm matter that could contaminate future samples. For chlorophyll a (chl-a) analysis: Within 12 hours of sample collection, the biofilm samples were vacuum-filtered through 0.4 µm glass fiber filters (Millipore, Billerica, MA) in the dark. Biofilms and filters were stored in Petri dishes wrapped in aluminum foil and frozen in the dark for up to 2 months. Biofilms were analyzed for chl-a using a TD-700 fluorometer (Turner Designs, Sunnydale, CA). Samples were thawed for approximately two hours then prepared for analysis in the dark. To extract chl-a for analysis the filter was added to 10 mL of 90% acetone in a plastic test tube. Parafilm was wrapped around the lid of the test tube and the sample was agitated for about 15 seconds. Each sample was wrapped in foil to block the light then left to steep in the refrigerator overnight. The samples were agitated again after the steeping period and warmed to room temperature (approximately 30 minutes). Approximately 1 mL of the supernatant was diluted with 90% acetone to fall within the instrument’s range. The solution was poured into a sample cuvette (>75% full) and was analyzed with the fluorometer. For dry mass and metal analysis: Glass fiber filters were pre-weighed and the biofilms were vacuum-filtered in the dark. Biofilms and filters were stored in Petri dishes and dried in a drying oven at 45°C overnight. Dried samples were weighed and filter weights were subtracted to obtain dry mass. 21  The biofilm samples were prepared for elemental analysis using a variation on the US EPA method #200.2 for sediments. In step 1, 0.5 g of dried biofilm plus filter was weighed out. Triplicate blank filters were also digested and analyzed for total recoverable elements and then values were subtracted from sample results. The acid-digested biofilm samples were analyzed using ICP–AES (Varian 725 – ES).  4.3 4.3.1  Data analysis methods Quality analysis and quality control  In the field, duplicate water samples were collected at a random site during each collection to account for within site variability. Duplicate or triplicate tiles were collected from each site during biofilm sample collections. In the laboratory, several duplicate analyses of water and sediment samples were used to determine the variability caused by instrument analysis. Accuracy of water sample analysis was determined by measuring standard samples with known concentrations on the Lachat autoanalyzer during each sampling run. Six standards of known trace element concentrations were prepared in the same matrix as the sediment and biofilm samples (5% HNO3-) and used to calibrate the ICP-AES. Blanks were included in the water and sediment analyses to ensure that solutions were not contaminated. The TD-700 fluorometer was calibrated with a known concentration of chlorophyll a and random samples were run in triplicate during each sampling run. 4.3.2  Statistical analyses  All statistical analyses, including descriptive statistics, normality tests, time-series analyses, correlations and comparisons of means were done using SPSS software v. 16.0. All statistical analyses used an α-value of 0.05. In order to facilitate data analysis, different sites were grouped together. Upstream to downstream analysis looked only at sites 14, 15, 11, 17, 18, 7 and 2, in addition to site 12, which is groundwater. Historic trend analysis was done only on data from sites 11 and 2, for which historic data exists. The remainder of the analysis focused on one urban 22  tributary (site 8), one agricultural tributary (site 22) and on the mouth of Marshall Creek (site 2). The water and sediment results were separated into wet and dry seasons based on several characteristics (changes in precipitation and water depth). The wet and dry seasons are summarized in Table 2. Table 2. Wet and dry season classifications during study period.  Dry Season  Wet Season  August 18, 2008; June 18 – October 14, 2009; June 13 – July 26, 2010  December 15, 2008; October 15, 2009 – June 12, 2010  Biofilms were sampled over three distinct time periods (see section 4.2.3). The results remained categorized in this fashion for seasonal comparisons. Graphs were created in order to visualize the trends in water quality parameters in the Marshall Creek mainstem from the upstream to the downstream direction. Bed sediment data was also analyzed visually for historic trends due to small sample sizes. The data for most variables followed non-normal distributions and therefore data transformation was attempted. Both the log and natural log (ln) transformations failed to produce normal distributions in the data. As a result, all further analyses were performed on non-transformed data using non-parametric statistical testing. Time series analyses were run on dissolved N data collected between 1994 and 2010 at sites 11 and 2. These analyses included tests for randomness and auto-correlation. Non-parametric Spearman-Rank correlations were run between water quality, sediment and biofilm parameters on a pooled set of data for the entire watershed, as well as on data collected from specific sites. The Bonferroni adjustment was used to determine the significance level for these correlations in order to reduce the Type I error that may occur when testing several pairs of samples. This method divides the initial α level (0.05) by the possible number (n) of pair-wise comparisons for a resulting significance level of α/n. The non-parametric Kruskal-Wallis test, which can be used as a substitute for the  23  parametric ANOVA test, was performed on the water, suspended sediment and biofilm data (pooled and seasonal datasets) to determine whether any differences existed between the means of the parameters by site. The three sites that were the focus of these comparisons were the urban (site 8) and agricultural (site 22) tributaries and the site at the mouth of Marshall Creek (site 2). If the results from the Kruskal-Wallis test indicated significant differences between sites, a Mann-Whitney U (M-W U) test was performed to determine which pairs of sites specifically differed from one another. The M-W U test is the non-parametric equivalent to the z and t tests for independent samples. The pair-wise M-W U significance level was determined by using the Bonferroni adjustment for a resulting significance level of α/3 = 0.0167. Seasonal differences in parameters were tested within each site using the M-W U test at a significance level of α=0.05, since there is only one pair to compare (wet versus dry). The assumption of independent samples in the M-W U test is not entirely met when comparing water and suspended sediment parameters in site 2 to upstream tributaries because of possible autocorrelation due to upstream to downstream effects. However, tests for autocorrelation were run on those parameters that produced significant differences between Marshall mainstem and at least one of the tributaries (Spearman Rank). Where no autocorrelations were detected, it is accepted that the assumption of independence is met and that the results of the M-W U tests hold true. Biofilms in each site are assumed to be independent of each other, since they are sessile and therefore do not experience upstream to downstream effects.  24  5  Results and Discussion  5.1 5.1.1  Land Use Change – Landscape Stressors GIS and impervious area calculations  As of 2006, the Total Impervious Area (TIA) in the Marshall Creek watershed was 874 ha, or 23% (Wright, personal communication, 2009). Based on known relationships between TIA and indices of aquatic health, this would classify the watershed as impacted (Schueler et al., 2009). Impervious areas were manually delineated for both the urban and agricultural subwatersheds in ArcGIS and the results are summarized in Table 3. The urban tributary would be considered severely impacted, considering the amount and quality of runoff from an urban residential area with 35% TIA (Schueler et al., 2009). Approximately one third of the TIA in the urban subwatershed consists of paved roads. GIS maps showing TIA delineation in both of the subwatersheds can be found in Appendix A. Table 3. Summary of Total Impervious Area (TIA) data for the Marshall Creek watershed and subwatersheds of interest.  Watershed Urban Agricultural Marshall Creek  5.1.2  Area (ha) 165 73 3800  TIA (ha) 57.5 5.1 874  % TIA 35 7.1 23  Animal Numbers and AUEs  The Marshall Creek Watershed overlaps with the South Matsqui (Abbotsford Aquifer region) and Sumas Prairie sub-regions in the Agricultural Census of Canada (StatsCan, 2006) (Appendix A). Livestock numbers for these sub-regions are summarized in Table 4.  25  Table 4. 1991-2006 animal numbers from Statistics Canada for the Abbotsford Aquifer and Sumas Prairie regions of the lower Fraser Valley.  Abbotsford Aquifer Region Poultry Poultry Farms Average # Poultry per Farm  1991 2,724,660 129 21,121  1996 2,709,513 139 19,493  2001 3,176,719 116 27,386  2006 3,116,341 90 34,626  Beef and Dairy Cattle Cattle Farms Average # Cattle per Farm  2,196 102 22  1,957 94 21  1,334 46 29  1,371 41 33  Pigs Pig Farms Average # Pigs per Farm  6,015 11 547  5,807 12 484  6,134 7 876  832 9 92  Sumas Prairie Region Poultry Poultry Farms Average # Poultry per Farm  1991 488,976 52 9,403  1996 872,075 71 12,283  2001 1,816,021 86 21,117  2006 1,761,956 63 27,968  Beef and Dairy Cattle Cattle Farms Average # Cattle per Farm  18,535 180 103  18,293 161 114  19,778 142 139  20,100 118 170  Pigs Pig Farms Average # Pigs per Farm  38,862 36 1,080  41,429 28 1,480  60,086 26 2,311  50,105 18 2,784  Table 4 shows that the Abbotsford Aquifer region is dominated by poultry production, with a major increase in poultry numbers between 1996 and 2001. A decline in poultry numbers can be observed between 2001 and 2006, possibly related to an outbreak of avian flu in the region in the year 2003, which may have resulted in a number of operations shutting down temporarily. It is also noted that over the aquifer the intensity of poultry and cattle has increased (based on the number of animals per farm) but there has also been a sharp decrease of pigs and pig farms recently. The Sumas Prairie region contains considerably lower poultry numbers, but has much larger populations of cattle and pigs. Again, there is a decline in poultry numbers between 2001 and 2006, which is likely due to avian flu.  26  Animal numbers within the Marshall Creek watershed were estimated for 2006 using data from census DA 59090721, the South Matsqui census EA and from GIS. The data is summarized in Table 5.  Table 5. Estimated 2006 livestock numbers in the Marshall Creek Watershed.  Sumas Prairie portion 56,408 2 28,204 8.7%  Abbotsford Aquifer portion 621,440 18 34,626 26%  Total  Beef and Dairy Cattle Cattle Farms Average # Cattle per Farm Cattle farms as % of total farms  3,050 12 254 52%  273 8 33 12%  3,323 20  Pigs Pig Farms Average # Pigs per Farm Pig farms as % of total farms  1,334 2 889 6.5%  166 2 92 2.6%  1,500 4  Poultry Poultry Farms Average # Poultry per Farm Poultry farms as % of total farms  677,848 20  A similar calculation was done in the agricultural subwatershed. Based on the orthophoto and ground-truthing, there were four farms inside the sub-catchment area: three dairy and one chicken operation. 2006 livestock numbers were calculated for the agricultural subwatershed using this information and data from DA 59090721 (see above) (Table 6).  Table 6. Estimated 2006 livestock numbers in the agricultural subwatershed within the Marshall Creek watershed.  Agricultural Subwatershed Poultry Poultry Farms Average # Poultry per Farm  2006 28,204 1 28,204  Dairy Cattle Dairy Farms Average # Cattle per Farm  762 3 254  27  AUEs were calculated for the 2006 agricultural census year and compared to previous census years. Table 7 shows this historic comparison for the Abbotsford Aquifer and Sumas Prairie regions of the lower Fraser Valley. Table 7. Average Animal Unit Equivalents (AUEs) calculated for the Abbotsford Aquifer and Sumas Prairie regions of the lower Fraser Valley, 1996-2006.  Abbotsford Aquifer Region 1996 2001 2006  AUEs 20,279 21,634 20,783  Area (ha) 3,875 5,216 6,088  Sumas Prairie Region 1996 2001 2006  30,504 7,596 40,834 8,124 39,163 8,048  AUEs/ha 5.23 4.15 3.41  4.02 5.03 4.87  The area under agriculture in Table 7 changes between census years, particularly in the Abbotsford Aquifer region. This is due to the changing boundaries of the Agricultural Census, and unfortunately, makes it difficult to compare data across years. Because of the increasing land area in the Abbotsford Aquifer region, the AUEs/ha appears to be decreasing over time. The total animal numbers data (Table 4) provides a better picture of change in the Abbotsford Aquifer region, with an increase in poultry numbers and decrease in pig numbers. This shift between livestock types could also contribute to the discrepancies in AUEs/ha across census years. The Sumas Prairie census area has experienced minor changes compared to the Abbotsford Aquifer area, so the AUEs/ha are more comparable across years. There was a major increase in AUEs between 1996 and 2001, followed by a slight decline in 2006. A similar calculation was done for the Marshall Creek watershed and for the agricultural subwatershed. Based on the estimated 2006 livestock numbers in the watershed in Tables 5 and 6, AUEs were calculated within the watershed areas. Table 8 shows the AUEs calculated for the Marshall Creek watershed and for the agricultural subwatershed.  28  Table 8. Average 2006 Animal Unit Equivalents in the Marshall Creek Watershed and the agricultural subwatershed.  AUEs Abbotsford Aquifer Region 4,170 Sumas Prairie Region 3,727 Marshall Creek Watershed 7,897  Area (ha) 1,214 840 2,054  AUEs/ha 3.44 4.44 3.85  Agricultural Subwatershed  72.66  15.5  1,129  According to the USDA (2008), the general rule of thumb is that it takes 1-2 acres of land to support one AUE. This equates to a recommended range of 1.2-2.5 AUE/ha. The EU Council Directive concerning the protection of waters against pollution caused by nitrates from the agricultural sector set a guideline of 1.7 AUE/ha (European Commission, 2002). Based on these classification systems the Abbotsford Aquifer and Sumas Prairie regions are characterized as having high livestock densities for 1996, 2001, and 2006 census years. Assessment at a subwatershed scale reveals that the AUE density is extremely high at 15.5 units per hectare. This would indicate that this subwatershed area is a hot spot for excess manure production in the area.  5.1.3  Population Estimates  The Marshall Creek watershed area falls within 12 Statistics Canada Census Tracts (CTs) (StatsCan, 2010). A map of these CTs is presented in Appendix A. Three of these tracts fall completely inside the watershed boundary, while the remaining 9 overlap partially. Using the 2006 census data, an average number of people per private dwelling was calculated for each CT. Then, with a 2008 orthophoto of the Marshall Creek watershed area, the number of dwellings in the watershed was counted for each CT, except for the three that fall entirely within the watershed. With this known number of dwellings, the 2006 Marshall Creek catchment population was estimated to be approximately 25,181 people (Table 9).  29  Table 9. Population estimate for 12 Census Tracts that overlap the Marshall Creek watershed boundary (StatsCan, 2010). The * indicates those CTs that fall entirely within the watershed boundary.  Census Tract # 9320001.00 9320005.01 9320011.00 9320012.00 9320100.00 9320101.00* 9320102.00 9320103.00 9320104.00 9320105.00* 9320106.02 9320106.03*  Average pop. per dwelling (from census) 4.09 2.87 2.91 3.18 3.22 2.16 2.07 1.71 2.95 2.84 2.71 2.94 Total  # of dwellings within catchment (from GIS) 40 720 41 426 253 1014* 2028 17 1267 1422* 2320 62* 9610  Estimated pop. within catchment 164 2068 119 1354 814 2187 4193 29 3740 4042 6289 182 25,181  To obtain a historic estimate of the Marshall Creek watershed population, the population and number of dwellings from the 2001 and 1996 censuses were compared to the 2006 estimate. The change in population and dwellings was determined between the census years in all 12 CTs and the change within the Marshall Creek watershed was estimated as a % of the total change. The historic population estimates are shown in Table 10. Table 10. Historic population estimates for 12 Census Tracts that overlap the Marshall Creek catchment area (StatsCan, 2001b; StatsCan, 1996b).  Census Tract # 2001 estimated pop. within catchment 9320001.00 157 9320005.01 1955 9320011.00 118 9320012.00 1066 9320100.00 799 9320101.00* 1890 9320102.00 4303 9320103.00 31 9320104.00 3667 9320105.00* 3815 9320106.02 5060 9320106.03* 175 Total 23,037  1996 estimated pop. within catchment 153 2045 114 926 726 1749 4156 30 3467 3549 3746 136 20,797  30  Similar calculations were done for both the urban and agricultural subwatersheds of interest. The population estimates for these two subwatersheds are shown in Table 11. The population in the urban subwatershed increased by approximately 40% between 1996 and 2006, but the agricultural subwatershed population remained constant. Table 11. Population estimates for the urban and agricultural subwatersheds of the Marshall Creek watershed.  Subwatershed 2006 2001 1996 Urban 1716 1471 1205 Agricultural 29 28 26  5.2 5.2.1  Stream Quality – Measurement Endpoints Water Quality 5.2.1.1 Overall  Table 12 summarizes the water quality data that was collected at 17 sites throughout the Marshall Creek watershed (refer to Appendix A for site locations). Table 12. Seasonal water quality averages across the Marshall Creek watershed.  Parameter  Wet Season Range Mean NO3--N (mg/L) 0.088 – 19.5 4.40 NH4+-N (mg/L) 0.051 – 7.74 0.790 SRP (mg/L) <0.001 – 0.323 0.0280 Cl- (mg/L) 6.59 – 171 38.0 Turbidity (NTU) 0.96 – 125 21.8 TOC (mg/L) 0.96 – 19.4 4.37 DO (mg/L) 3.4 – 11.5 8.05 pH 6.16 – 7.79 7.03 Sp. Cond (µS/cm) 12.8 – 660 257 Temperature (ºC) 0.10 – 17.0 7.79  N 83 83 83 83 67 72 67 83 83 79  Dry Season Range Mean <0.05 – 11.7 3.78 <0.05 – 1.94 0.285 <0.001 – 0.0545 0.0136 9.43 – 137 23.0 N/A N/A 1.27 – 22.3 3.98 1.7 – 11.1 8.50 6.34 – 7.83 7.13 119 – 568 257 12.4 – 18.0 14.7  N 43 30 42 30 0 30 27 43 43 27  In all of the indicators except for pH, there is order of magnitude variability. Because of this variability, subsets of sites were selected for upstream to downstream analysis and site comparison.  31  5.2.1.2 Upstream to Downstream Trends The concentration of nitrate-N gradually decreases from the headwaters of Marshall Creek to the confluence with the Sumas River (Figure 3). See Figure 2 for a map of the sample sites. This trend is due to the high concentrations of nitrate-N in the groundwater of the Abbotsford Aquifer, which is the primary source of baseflow in Marshall Creek. The decrease in nitrate-N downstream can be attributed to dilution from relatively low N inputs from tributaries that are not influenced by the Abbotsford Aquifer groundwater (Smith, 2004). Nitrate-N was significantly higher in the dry season along Marshall Creek (M-W U, p=0.012) than in the agricultural tributary (site 22). This indicates the influence of groundwater on the Marshall Creek mainstem, whereas surface runoff processes are the major contributors of nutrients from the land into a typical agricultural stream (Stutter et al., 2007). Mean  Drinking Water Guideline  Freshwater Aquatic Life Guideline  12  Nitrate-N (ppm)  10 8 6 4 2 0 12  14  15  11  17  18  7  2  Upstream to Downstream Sites Figure 3. Box-whisker plot showing upstream to downstream trend of nitrate-N in water collected in Marshall Creek August 2008 – May 2010. Error bars represent maximum and minimum values. (n=2-16)  32  Site 12 is a groundwater well that the Fraser Valley Trout Hatchery draws from for its operations and site 14 is located just below the discharge from the hatchery. These two sites approached the Canadian drinking water guideline of 10 ppm nitrate-N but did not exceed the guideline during the sample period (Health Canada, 2008). However, this guideline has been exceeded at site 14 in the past (Khan et al., 2009). All sites contained nitrate-N concentrations that consistently exceeded the Freshwater Aquatic Life Guideline of 2.9 ppm nitrate-N (Canadian Council of Ministers of the Environment, 2007), except for two samples that dropped below the guideline at sites 17 and 18. This is interesting because several species of fish are known to be present in Marshall Creek including young of the year fish, which are more sensitive to poor water quality; therefore, the fish that are present do not appear to be suffering from direct toxic effects suggested by the nitrate-N guideline (Wilson, unpublished data, 2009). This observation may be the result of the limited time that the fish are exposed to the higher N levels or that the fish selectively avoid the areas of highest N concentrations (Larrick et al., 1978). Soluble reactive phosphorus (SRP) did not show a marked upstream to downstream trend (Figure 4). There was a slight increase in concentrations towards the downstream sites, with the highest concentrations found at site 2. This peak in concentration just exceeded the trigger concentration of 0.075 ppm total phosphorus for eutrophic conditions in rivers and streams (Canadian Council of Ministers of the Environment, 2005). However, the greatest variability in SRP concentrations was also found at site 2. No Canadian aquatic quality guidelines currently exist for SRP, but a guideline of 0.02 ppm was proposed as a recommended ecosystem target for upland streams in the UK (Neal et al., 2003).  33  Mean  Eutrophic Trigger Range (TP)  0.08 0.07  SRP (ppm)  0.06 0.05 0.04 0.03 0.02 0.01 0 12  14  15  11  17  18  7  2  Upstream to Downstream Sites Figure 4. Box-whisker plot showing upstream to downstream trend of SRP in water collected in Marshall Creek August 2008 – May 2010. Error bars represent maximum and minimum values (n=2-16).  This difference between the SRP trend and the nitrate-N trend can be attributed to the different chemistries of the two compounds. Phosphorus is generally not labile in the soil environment and it ends up adsorbing to clay particle surfaces and is held in the soil. It is not until soil particles are eroded and enter the stream environment through surface runoff that most P becomes available to aquatic biota as SRP. As indicated at site 12 (groundwater well), there is very little SRP in the Abbotsford Aquifer groundwater, and the amount that enters Marshall Creek does not change substantially downstream. Total organic carbon (TOC) increased steadily from headwaters to mouth of Marshall Creek (Figure 5). One groundwater sample (Site 12) showed a lower concentration than any of the stream samples, indicating anthropogenic or natural organic compounds do not likely contaminate it. The recommended BC water quality guideline for TOC is the 30day median ± 20% of the median background concentration (BC Ministry of Environment, 2001). Because samples were collected only a few times a year for some sites, this cannot be shown in Figure 4. 34  Mean  10 9 8 TOC (ppm)  7 6 5 4 3 2 1 0 12  14  15  11  17  18  7  2  Upstream to Downstream Sites  Figure 5. Box-whisker plot showing upstream to downstream trend of TOC in water collected in Marshall Creek August 2008 – May 2010. Error bars represent maximum and minimum values (n=1-13).  The increase in TOC in a downstream direction can be attributed to organic inputs via erosion and runoff and decomposition of in-stream biota. Agricultural activities, particularly excess manure production and application, are one major component that could be attributed to this increase in TOC towards the mouth of Marshall Creek. Over the Abbotsford Aquifer, TOC is likely degraded and consumed by soil communities, resulting in groundwater that is relatively devoid of organic C (Wilson et al., 1983). Dissolved oxygen (DO) concentrations remain generally constant, except for a slight decrease after site 17 (Figure 6). DO is temperature-dependent and no observed upstream to downstream trend was found in temperature in Marshall Creek. The four upstream sites have an average DO that is close to the fresh water aquatic guideline of 9.5 ppm DO as a minimum requirement for early life-stages of cold-water biota (Canadian Council of Ministers of the Environment, 2007), whereas the three downstream sites fall consistently below this guideline. All sites generally fall above the guideline of 5.5 ppm DO as a  35  minimum requirement for ‘other’ life stages of warm-water biota (Canadian Council of Ministers of the Environment, 2007). Mean  Cold-water biota minimum requirement  Warm-water biota minimum requirement  12 10  DO (ppm)  8 6 4 2 0 14  15  11  17  18  7  2  Upstream to Downstream Sites Figure 6. Box-whisker plot showing upstream to downstream trend of TOC in water collected in Marshall Creek August 2008 – May 2010. Error bars represent maximum and minimum values.  Site 18 is just downstream of the confluence of the agricultural tributary (site 22) with Marshall Creek. The agricultural tributary has an average DO concentration of 5.9 ppm, which could contribute to the slight downward shift in average DO in the lower reaches of Marshall mainstem. There was no observed upstream to downstream trend in specific conductivity.  5.2.1.3 Correlations between water quality parameters Spearman’s rank correlations were run on all the water quality parameters collected at each of the three sites in the Marshall Creek watershed (α=0.05). The water quality data was split into wet season (n=40) and dry season (n=8) data. The dry season sample was comparatively small and so these correlations have greater error associated with them.  36  5.2.1.3.1 Wet Season Correlations Several significant correlations (α=0.05) emerge between water quality parameters during the wet season. The results are summarized in Table 13. Table 13. Significant Spearman’s Rank correlations (|Spearman’s coefficient (R)| > 0.4, p<0.05) for water parameters (wet season).  Parameter(s) Dissolved N (NO3-/NO2-N) Soluble Reactive Phosphorus (SRP) ClNH4+-N pH  Correlation + or +  Dissolved Oxygen (DO)  -  Turbidity  +  + + -  With Parameters(s) Temperature Dissolved N (NO3-/NO2-N & NH4+-N), Cl-, TOC NH4+-N, Sp. Cond, TOC Sp. Cond, TOC NO3-/NO2-N, SRP, Cl-, Turbidity, TOC SRP, Cl-, NH4+-N, Sp. Cond, Turbidity, TOC SRP, Cl-, NH4+-N, Sp. Cond, TOC  These correlations were particularly apparent in the agricultural (site 22) and urban (site 8) tributaries, where there is no groundwater influence from the Abbotsford Aquifer, and surface runoff is the dominant process affecting these parameters. After the Bonferroni adjustment was applied to the wet season correlations (55 pairs), the α value became 0.05/55 = 0.00091. This reduced the number of significant correlations, but the ones that remained significant are displayed in Table 14. Table 14. Significant Spearman’s Rank correlations (|Spearman’s coefficient (R)| > 0.4, p<0.00091) for water parameters (wet season).  Parameter(s) Soluble Reactive Phosphorus (SRP) Cl-  Correlation + or +  With Parameters(s) NH4+-N, TOC  +  NH4+-N pH Dissolved Oxygen (DO)  + -  Turbidity  +  NH4+-N, Sp. Cond, Turbidity TOC TOC, Turbidity SRP, TOC SRP, Cl-, NH4+-N, Sp. Cond, Turbidity, TOC Sp. Cond, TOC  37  5.2.1.3.2 Dry Season Correlations During the dry season, the correlations are different. These changes are likely due to a combination of seasonal effects as well as the lower sample size for the dry season. The dry season results are summarized in Table 15. Table 15. Significant Spearman’s Rank correlations (|Spearman’s coefficient (R)| > 0.4, p<0.05) for water parameters (dry season).  Parameter(s) Dissolved N (NO3-/NO2-N) SRP Cl-  Correlation + or + +  With Parameters(s) Cl-, Sp. Cond NH4+-N, Sp. Cond, TOC  After the Bonferroni adjustment was applied to the dry season correlations (45 pairs), the α value became 0.05/45 = 0.0011. The only correlation that remained after this adjustment was between nitrate-N and specific conductivity.  5.2.1.4 Site comparison Non-parametric Mann-Whitney U tests were run to compare average values of water quality parameters from grab water samples (α/3=0.0167). An overall comparison of means is summarized in Table 16. There were no significant differences in average water depth or temperature among the three sites. Water quality parameters were tested for autocorrelation between sites using Spearman’s rank correlations. Significant positive correlations were found in nitrate-N (R=0.847, p<0.001) and SRP (R=0.649, p=0.007) between the Marshall mainstem site and the urban tributary. Therefore, these sites are not independent of one another and so one must consider the following when interpreting the results of the significance tests presented in Table 16. The urban tributary is obviously contributing to Marshall Creek because the sites are autocorrelated, but when the mean nitrate-N concentrations are compared (1.82 mg/L versus 4.15 mg/L, respectively), the greater contribution is likely from the groundwater, as indicated by the upstream to downstream trend (Figure 3). As for SRP, again, the urban tributary is contributing to Marshall Creek as indicated by the  38  autocorrelation, but because the concentration is significantly lower than in the agricultural tributary, the SRP loadings from this site are likely low. There was also a significant autocorrelation in turbidity between the Marshall mainstem and the agricultural tributary (R=0.582, p=0.037). These sites are therefore not independent and the assumption of the Mann-Whitney U test is violated. In this case, the agricultural tributary is significantly contributing to the turbidity in the Marshall Creek mainstem, which is diluted from other sources. Table 16. Significant Mann-Whitney U results (p<0.0167) for water quality parameters by site.  Parameter  Site  NO3--N NH4+-N, SRP  Marshall Mainstem Urban trib.  Value < or > > <  Cl-, Turbidity, TOC pH, DO DO Sp. Cond  Agricultural trib. Marshall mainstem Urban trib. Marshall mainstem Agricultural trib.  > > > > >  In comparison to Urban trib. and agricultural trib. Agricultural trib. and Marshall mainstem Marshall mainstem Urban trib. Marshall mainstem Agricultural trib. Marshall mainstem and urban trib.  Wet season comparisons showed the same significant relationships except for one, which was that pH was not significantly different between the urban tributary and Marshall mainstem. Sample sizes were not large enough in the dry season to perform between-site statistical comparisons. Wet season and dry season graphical comparisons for key water quality parameters are summarized in Figure 7.  39  0.05 0.04 SRP (mg/L)  Nitrate-N (mg/L)  8 7 6 5 4 3 2 1 0  0.03 0.02 0.01 0  Wet Dry Wet Dry Wet Dry Wet Dry Mouth  Ag. Trib.  Urban Trib.  HW  12  12  10  10  8  8  TOC (mg/L)  DO (mg/L)  HW  Wet Dry Wet Dry Wet Dry Wet Dry  6 4 2  Mouth  Ag. Trib.  Urban Trib.  6 4 2  0  0 Wet Dry Wet Dry Wet Dry Wet Dry HW  Mouth  Ag. Trib.  Urban Trib.  Wet Dry Wet Dry Wet Dry Wet Dry HW  Mouth  Ag. Trib.  Urban Trib.  Figure 7. Average concentrations of 4 water quality parameters (± 1 S.E.) in the wet season and dry season collected at 4 sites in the Marshall Creek watershed (HW=Headwaters – Site 11).  Nitrate-N and DO have similar relationships in the wet season and the dry season at all four sites. This is because nitrate-N, as an oxygen-containing molecule, is dependent on the concentration of dissolved oxygen in the water. In this system, nitrate-N is a good indicator of groundwater contribution due to the high concentrations in the Abbotsford Aquifer. At the Marshall Creek headwater site (site 11), the nitrate-N concentrations in the dry season exceed those in the wet season, because groundwater makes up a greater proportion of total discharge at this time. Once rain falls in the wet season, the nitrate-N concentrations at this site are diluted. The agricultural tributary (site 22) is not affected by groundwater from the Abbotsford Aquifer and it shows the opposite dynamic in nitrate-N concentrations. In the dry season, organisms take up nearly all of the nitrate-N, leaving 40  very little in the water column. This supply is replenished in the winter when precipitation causes runoff and leaching of nitrate-N from the soil and subsurface into the stream channel (Coulter et al., 2004). SRP enters the stream dominantly by surface runoff processes when it is adsorbed to soil particles as P, and organisms take it up during the productive dry season (Stutter et al., 2007). This is corroborated in Figure 7, where the SRP concentrations are higher in the wet season than in the dry season at all four sites, regardless of groundwater influence. TOC follows a similar dynamic to SRP at the mouth of Marshall Creek (site 2), but otherwise is more variable in its wet season-dry season dynamic between sites. The complete water quality dataset is presented in Appendix B. 5.2.1.5 Historic Nitrate-N A historic comparison was made with water quality data collected from previous studies on Marshall Creek (Lyautey et al., In Press; Khan et al., 2009; Smith, 2004; Berka, 1996). Two sites on the Marshall mainstem have historic data: site 11 near the headwaters and site 2 at the mouth. Figure 8 shows nitrate-N collected from grab water samples in 199495, 2002-04, 2006-07 and the current study period. Both time series show variability. The greatest range of nitrate-N concentrations were found at site 11. In the year 2007, samples were collected more frequently at site 11 (biweekly) and this greater sample frequency captured much greater variability in the concentrations of nitrate-N. There were also particularly high concentrations of nitrate-N measured at this site in 2007 (Figure 8), which could be indicative of a “hot moment” where an event occurred that released additional nitrate-N into Marshall Creek near the headwaters. The sampling and analytical methodologies between the 1990s and the present day have remained consistent, so the associated error is minimal. There were significant differences in the mean nitrate-N concentrations between the two sites in the wet season (M-W U, p=0.02) and in the dry season (M-W U, p<0.001). (There were no significant autocorrelations in the nitrate-N data between sites 11 and 2). The more upstream site (11) had consistently higher average nitrate-N concentrations than the 41  downstream site in both the wet and dry seasons, but the difference was more marked in the dry season. 5.2.1.5.1 Time-Series Analysis A time-series analysis was run on the historic water quality data (averaged monthly) collected at site 11 to determine if there has been any significant change over the last 15 years. Based on the input dataset, SPSS selected the Simple Seasonal model as the model that best described the nitrate-N time-series (Manly, 2001). This model is appropriate for a series with no long-term trend and a seasonal effect that is constant over time. This model produced an R2 = 0.624 and p <0.001. This would mean that there is no long-term trend in the nitrate-N data, only seasonal variability (Figure 8). An autocorrelation analysis on the time-series showed that there is positive autocorrelation at approximately 12-13 month time lags and negative autocorrelation at 6-7 month time lags, which supports that there is seasonal variability. The Simple Seasonal model fit and the autocorrelation test outputs are presented in Appendix B. This dataset was also tested for randomness. A random time series is one that consists of independent values from the same distribution (Manly, 2001). There is no serial correlation and this is the simplest type of data that can occur. The non-parametric ‘runs above and below the median’ test produced a Z-value of -0.691; the probability of a value this far from zero is 0.489, therefore we cannot reject hypothesis that data is random.  42  05/31/10  05/04/10  04/15/10  01/18/10  12/29/09  12/11/09  10/28/09  06/18/09  08/18/08  12/03/07  11/05/07  10/09/07  09/10/07  07/23/07  06/25/07  05/28/07  04/30/07  04/02/07  11/06/06  10/10/06  09/11/06  08/15/06  07/17/06  06/19/06  05/23/06  04/24/06  06/07/04  04/19/04  03/01/04  01/12/04  11/24/03  10/14/03  07/07/03  08/22/02  02/01/95  10/03/94  07/26/94  05/09/94  02/20/94  Nitrate-N (mg/L)  Site 11 Site 2  25  20  15  10  5  0  Figure 8. Nitrate-N data from water samples collected at 2 sites on Marshall Creek between 1994 and 2010. Note the breaks in the time series on the xaxis, indicated by the vertical lines.  43  It is not unexpected that the time-series analysis shows no long-term change in nitrate-N concentrations at site 11 in Marshall Creek. The main source of nitrate-N is groundwater discharge from the Abbotsford Aquifer into Marshall Creek as baseflow from the Fraser Valley Trout Hatchery. The hatchery draws groundwater from wells between 45-65 meters deep, which is drawing up water at least 10 years old (Wassenaar, Hendry & Harrington, 2006). Wassenaar et al. (2006) showed that for groundwaters >10 years old in the Abbotsford Aquifer (determined through 3H/3He dating) that there were no significant changes in nitrate-N concentrations from 1992-2004. This relatively unchanged average nitrate-N concentration is also reflected at site 11. That being said, the seasonal variability of nitrate-N concentrations in Marshall Creek is apparent, with higher average concentrations of nitrate-N in the summer dry season, when baseflow makes up a larger proportion of total stream flow. In the wet winter season, there is usually a release of nitrate-N from surface runoff in agricultural areas. However, there is a greater proportion of rainfall and surface runoff with low nitrate-N values from nonagricultural areas and this results in a dilution, since the proportional contribution from groundwater declines. Historic nitrate-N data for site 2 at the mouth of Marshall mainstem did not produce anything of significance in the time series analysis. This could be due to the lower sample size and greater variability in concentrations found at this site (see boxplot in Figure 3).  5.2.2  Bed Sediments – Historic comparison  Bed sediments were collected at sites throughout the Marshall Creek watershed during the 2008-09 sampling period and were analyzed for trace elements. Historic bed sediments were collected at sites 11 and 2 in 1993-94 and 2003-04 (Smith 2004; Berka, 1996). Historic bed sediment data was compared to current data at these sites and also at the agricultural tributary (site 22) and the urban tributary (site 8). Because single samples were collected in the late summer of each sample year, statistical analyses were not done, but graphs were created to visualize trends and differences between sites (Figure 9).  a)  ISQG 70 60  Cu (mg/kg)  50 40 30 20 10 0 1993 1994 2003 2004 2008 2009 1993 1994 2003 2004 2008 2009 2008 2009 2008 2009 Headwaters (11)  b)  Mouth (2)  Ag. Trib. Urb. Trib. (22) (8)  1400 1200  Mn (mg/kg)  1000 800 600 400 200 0 1993 1994 2003 2004 2008 2009 1993 1994 2003 2004 2008 2009 2008 2009 2008 2009 Headwaters (11)  Mouth (2)  Ag. Trib. Urb. Trib. (22) (8)  45  c)  ISQG 250  Zn (mg/kg)  200 150 100 50 0 1993 1994 2003 2004 2008 2009 1993 1994 2003 2004 2008 2009 2008 2009 2008 2009 Headwaters (11)  d)  Mouth (2)  Ag. Trib. Urb. Trib. (22) (8)  3000 2500  P (mg/kg)  2000 1500 1000 500 0 1993 1994 2003 2004 2008 2009 1993 1994 2003 2004 2008 2009 2008 2009 2008 2009 Headwaters (11)  Mouth (2)  Ag. Trib. Urb. Trib. (22) (8)  Figure 9. Trace element concentrations: a) Cu, b) Mn, c) Zn and d) P, measured in bed sediments collected in late summer in the Marshall Creek watershed. Each column represents one sample. No Mn data for 2008 at the agricultural tributary. (ISQG=Interim Sediment Quality Guideline)  46  In Cu, Mn, and Zn there are not notable differences in trace element concentrations between the headwaters and mouth of Marshall Creek. However, at both of these sites, these elements have increased in concentration over time. The Interim Sediment Quality Guidelines (ISQG) for Cu (Fig. 9a) and Zn (Fig. 9c) were both exceeded in Marshall mainstem since 2003 (Canadian Council of Ministers of the Environment, 1999). In addition, the Marshall Creek mainstem has higher concentrations of these elements in bed sediments in 2008-09 than do the urban or agricultural tributaries. The Cu concentration in bed sediments from the urban tributary is slightly higher than from the agricultural tributary (Fig. 9a). In contrast, the Zn concentrations are similar but higher in the agricultural tributary than the urban tributary (Fig. 9c). This can be attributed to the fact that Zn and Cu originate from vehicle wear in the urban environment (Cu from brakes and Zn from tires) (Kuusisto-Hjort, 2009) and both trace metals are used as growth supplements in animal feed for pigs and chickens in the agricultural environment (Han et al., 2000). Bed sediment P concentrations are consistently higher at the mouth of Marshall Creek than at the headwater site, except for in one sample collected in 1994 (Fig. 9d). The concentrations of P haven’t changed markedly since the 1990s in the headwaters of Marshall Creek, but have increased over time at the mouth. The agricultural tributary P concentrations are similar to the mouth of Marshall Creek, but there is variability between the 2008 and 2009 sample years. The urban tributary overall has the lowest P concentrations in bed sediments. It appears that the elevated P in the agricultural tributary is contributing to the P concentrations in the mouth of Marshall Creek. This observation is further investigated in suspended sediment and biofilm results. The complete bed sediment results are presented in Appendix C.  5.2.3  Suspended Sediments 5.2.3.1 Overall  Sediment samplers were installed at 3 sites in the Marshall Creek watershed: the agricultural tributary (site 22, n=8) the urban tributary (site 8, n=9) and at the mouth of the Marshall Creek mainstem (site 2, n=5). Trace element and dry mass data for all fine 47  (<63 µm) suspended sediments are summarized and presented with Canadian Interim Sediment Quality Guidelines (where they exist) in Table 17 (Canadian Council of Ministers of the Environment, 1999). There was no significant difference in any of the parameters between the wet season and dry season. Table 17. Sediment quality averages in the Marshall Creek watershed and Interim Sediment Quality Guidelines (ISQG) for particular trace elements (Canadian Council of Ministers of the Environment, 1999).  Parameter Dry mass (mg) Al (mg/kg) Ca (mg/kg) Cd (mg/kg) Cr (mg/kg) Cu (mg/kg) Fe (mg/kg) K (mg/kg) Mg (mg/kg) Mn (mg/kg) Ni (mg/kg) P (mg/kg) Pb (mg/kg) Zn (mg/kg)  Range 400 – 51050 7892 – 92600 3714 – 42150 0 – 11.14 15.55 – 384.9 45.54 – 508.1 17970 – 195000 605.4 – 8514 4791 – 135100 312.6 – 11430 21.16 – 1305 507.2 – 5534 5.893 – 364.3 69.80 – 1074  Mean 9450 22600 15280 1.731 81.42 111.7 49880 1893 17220 1731 134.1 2032 67.50 228.5  ISQG N/A N/A N/A 0.6 37.3 35.7 N/A N/A N/A N/A N/A N/A 35 123  For the trace metals where ISQGs exist, all average concentrations in suspended sediments in the Marshall Creek watershed exceed the guidelines. These exceedances agree with the impervious surface data in section 5.1.1, which predicts that the aquatic health in Marshall Creek is negatively impacted. 5.2.3.2 Site Comparison Non-parametric Mann-Whitney U tests were run to compare average dry mass and metal concentrations of fine suspended sediments collected in the traps (α/3=0.0167). An overall comparison of means is summarized in Table 18. No significant autocorrelations were found in the suspended sediment parameters between the Marshall mainstem and either of the tributaries (p>0.0167).  48  Table 18. Significant Mann-Whitney U results (p<0.0167) for parameters in suspended sediments by site.  Parameter Dry mass (g) Cu (mg/kg) Mg (mg/kg)  Site Agricultural trib. Agricultural trib. Agricultural trib.  Value < or > > < >  Mn (mg/kg)  Agricultural trib.  <  Ni (mg/kg) P (mg/kg)  Urban trib. Urban trib.  < <  Zn (mg/kg)  Agricultural trib.  <  In comparison to Urban trib. Urban trib. Urban trib. and Marshall mainstem Urban trib. and Marshall mainstem Agricultural trib. Agricultural trib. and Marshall mainstem Marshall mainstem  The urban tributary has the highest concentration of Cu in the suspended sediments. Roads make up approximately one third of the TIA in the urban subwatershed and brake wear from vehicles are a potentially significant source of Cu contamination via stormwater runoff (Kuusisto-Hjort, 2009). Ni concentrations in suspended sediments were highest in the agricultural tributary. Soil amendments and phosphate fertilizers contain traces of nickel, as it is a nutritionally essential metal for some plants (Curtis & Smith, 2002). Runoff from agricultural land containing these types of amendments is a potential source of Ni to the agricultural tributary. The next step was to see if these between-site relationships varied with season. In the wet season, several significant relationships changed, as summarized in Table 19. There were no significant differences between sites in the dry season. This is likely due to the smaller sample sizes in the dry season (n=2 or 3). Table 19. Significant Mann-Whitney U results (p<0.0167) in the wet season for parameters in suspended sediments by site.  Parameter Mn (mg/kg)  Site Agricultural trib. Urban trib.  Value < or > < <  In comparison to Urban trib. Marshall Mainstem  Ni (mg/kg)  Urban trib.  <  Zn (mg/kg)  Agricultural trib.  <  Agricultural trib. and Marshall mainstem Urban trib. and Marshall mainstem  49  There is no difference in dry mass between the urban and agricultural tributaries in the wet season, although overall the agricultural tributary had greater amounts of suspended sediments in the trap. This result suggests that runoff processes become more dominant in the urban tributary during the wet season, contributing greater amounts of suspended sediment during storm events when there is rapid runoff from the steeply sloping urban area. Overall there is more sediment entering the agricultural tributary than the urban tributary, which could be attributed to greater soil and stream bank erosion via disturbance from agricultural practices, as well as organic matter inputs through manure application and storage (Ward, 2008; Smith et al., 2007). The complete suspended sediment dataset is presented in Appendix C.  5.2.4  Biofilms 5.2.4.1 Site and Seasonal Comparison 5.2.4.1.1 Chlorophyll a  Average chl-a concentrations were compared between each of the three time periods within each collection site (n=4 within each site, per season) (Figure 10). Chl-a concentrations were significantly different between at least two sample times in all of the sites (Kruskal Wallis, p<0.001). In the Marshall mainstem site, chl-a was significantly lower in the winter than in either the spring (M-W U, p<0.001) or summer samples (M-W U, p<0.001). In the urban tributary, the average chl-a concentration was significantly higher in the spring than in either the winter (M-W U, p=0.007) or summer (M-W U, p=0.011). In the agricultural tributary, the chl-a concentrations were significantly different between all three sample times (M-W U, p<0.003); the highest concentrations occurred in the summer, followed by the spring and finally the winter (Figure 10). Two parameters, in addition to light, have a major impact on stream biofilm biomass production and chlorophyll a: temperature and nutrients (Hill, Mulholland & Marzolf, 2001; Kiffney & Bull, 2000). If one assumes that light availability and quality are the same among the three sites at each season, the data suggests that nutrients are the 50  dominant agent controlling the production of chl-a in the sites affected by agriculture, and most notably in the summer. This same relationship may be seen for the Marshall Creek mainstem site. Chl-a concentrations at the urban site are less distinct among the seasons, suggesting that nutrient supply is relatively constant throughout the season and that water temperature is also less variable through the three seasons. The data variability was greatest in summer and least in winter, again suggesting a temperature response.  Average chl-a concentration (mg/L)  Mean 65 60 55 50 45 40 35 30 25 20 15 10 5 0 Urban  Ag.  Marshall Urban  Summer '09  Ag.  Marshall Urban  Winter '09-'10  Ag.  Marshall  Spring '10  Figure 10. Average chl-a concentrations (mg/L) in 3 sample sites over three seasonal sample collections. Error bars represent max/min values.  Within each season, biofilms were collected at four different intervals, on average about 9 days each of colonization duration. The changes in chl-a concentrations during the colonization periods for the three stations in late August-September 2009, December 2009-January 2010 and April-May 2010 are shown in Figure 11.  51  a)  Summer 2009 Urban  Agriculture  Marshall  Mean Chl-a (mg/L)  100  10  1  0.1 11  19  27  34  Days of colonization  b)  Winter 2009-10 Urban  Agriculture  Marshall  Mean Chl-a (mg/L)  10  1  0.1  0.01 10  18  27  38  Days of Growth  52  c)  Spring 2010 Urban  Agriculture  Marshall  14  Mean Chl-a (mg/L)  12 10 8 6 4 2 0 11  19  27  35  Days of colonization  Figure 11. Average chlorophyll a concentrations of biofilms collected from urban, agricultural and Marshall mainstem stations in summer 2009 (a, n=2), winter 2009-10 (b, n=3) and spring 2010 (c, n=3) after x days of colonization. Mean ± 1 standard error. Note: Log scale in Figures a and b.  Statistical tests did not show any significant differences between sites in the summer 2009 dataset (Figure 11a) and this is likely due to low sample sizes (n=2 for each time period). However, there are observable differences in the average chl-a concentrations between sites that are indicated by the logarithmic scale in Figure 11a. In addition, it can be observed that the trends of chl-a concentrations through time are very similar, although their magnitudes are different. There is a sharp increase in chl-a production between 11 and 19 days of growth, then either a steady state (Marshall Creek) or a modest and gradual decrease between 19 and 34 days of growth. This decreasing trend could be due to a saturation point being reached in biofilm. Bothwell (1989) found that biomass growth rates in periphytic diatom communities reached saturation when exposed to phosphate concentrations ranging from 0.1 to 1.0 µg/L, which are several orders of magnitude lower than what was measured in the Marshall Creek watershed (SRP ranges from <0.001-0.323 mg/L). In the winter 2009 dataset (Figure 11b), the chl-a from biofilms collected at the urban site is significantly higher than the chl-a from the mainstem Marshall Creek site and from the 53  agricultural tributary site across all days of growth (M-W U, p<0.015). Marshall mainstem is significantly lower in chl-a than the agricultural site only after the first 10 days of growth (M-W U, p<0.015), and then there are no significant differences in chl-a concentrations from 18 to 38 days of growth. The chl-a trend through time appears to increase in a nearly exponential fashion in the urban tributary (masked by log scale in Figure 11b). The chl-a trends at each site in the spring of 2010 are more varied than in the other two datasets. After 11 days of growth, there are no significant differences in chl-a between any of the sites (Figure 11c). After 19 and 27 days of growth, the agricultural site has significantly lower chl-a than Marshall mainstem and the urban site (M-W U, p<0.015). By 35 days of growth, there are no significant differences in chl-a between the sites. As indicated by the different scales in Figures 11a, b and c, there were significant differences in chl-a production between seasons (M-W U, p<0.05). Each site had a different chl-a dynamic, which is likely controlled by different drivers at each site. Table 20 summarizes the mean chl-a concentrations at the end of each season for each site. After approximately 5 weeks of growth, the chl-a at the Marshall mainstem site appears to follow a similar dynamic to the agricultural site, although at a lower magnitude. The urban tributary however, follows a different seasonal dynamic, with greater chl-a in the winter than in the late summer. This is likely due to the warmer winter water temperatures in the urban tributary (Figure 12), where stormwater discharges tend to have higher temperatures relating to flow over relatively warm asphalt surfaces (Herb et al., 2008). Temperature-chlorophyll a relationships are further investigated in section 5.2.5.1. Table 20. Mean chlorophyll a concentrations (mg/L ± 1 standard error) of biofilms collected at the end of each season in the Marshall Creek watershed.  Site Urban Trib. Agricultural Trib. Marshall Mainstem  Sep. 10/09 (34 days of colonization, n=2) 1.405 ± 0.3857 18.55 ± 2.342 10.55 ± 8.992  Jan. 18/10 (38 days of colonization, n=3) 7.643 ± 1.179 0.4435 ± 0.03617 0.3130 ± 0.07708  May 31/10 (35 days of colonization, n=3) 10.58 ± 3.171 8.320 ± 1.442 7.557 ± 0.1927  54  Urban Tributary  Agricultural Tributary  Marshall Mainstem  10 Water Temperature (ºC)  9 8 7 6 5 4 3 2 1 01/18/10  01/16/10  01/14/10  01/12/10  01/10/10  01/08/10  01/06/10  01/04/10  01/02/10  12/31/09  12/29/09  12/27/09  12/25/09  12/23/09  12/21/09  12/19/09  12/17/09  12/15/09  12/13/09  12/11/09  0  Figure 12. Water temperatures at three sites during the winter biofilm collection period.  5.2.4.1.2 Dry Mass Average biofilm dry mass was compared between each of the three time periods within each collection site (n=4 within each site, per season) (Figure 13). Dry mass was significantly different between at least two sample times in all of the sites (Kruskal Wallis, p<0.005). In the Marshall mainstem site, dry mass was significantly higher in the spring than in the summer (M-W U, p=0.0003), but there were no significant differences between the winter season (M-W U, p>0.0167). In the urban tributary, the average dry mass was significantly lower in the summer than in the winter (M-W U, p=0.0002) or spring (M-W U, p=0.004). In the agricultural tributary, the biofilm dry mass was significantly different between all three sample times (M-W U, p<0.001); the highest dry mass occurred in the winter, followed by the spring and finally the summer (Figure 13).  55  Mean Average biofilm dry mass (mg)  12000 10000 8000 6000 4000 2000  Summer '09  Winter '09-'10  Marshall  Ag.  Urban  Marshall  Ag.  Urban  Marshall  Ag.  Urban  0  Spring '10  Figure 13. Average biofilm dry mass (mg) in 3 sample sites over three seasonal sample collections. Error bars represent max/min values.  Across all seasons, biofilm dry mass is significantly higher in the urban site than in the Marshall mainstem (M-W U, p=0.008). From field observations, the biofilms in the urban tributary appeared to trap more fine sediments on the tile medium than in the other sites. Withers & Jarvie (2008) noted that in urbanized watersheds with a flashier hydrologic regime and higher suspended sediment concentrations, that in-stream macrophytes and their associated biofilms help to reduce flow and trap incoming sediments that enter the stream via stormwater. This trapping phenomenon may have contributed to the higher dry mass in the biofilms at the urban site. The changes in biofilm dry mass during the colonization periods for the three stations in late August-September 2009, December 2009-January 2010 and April-May 2010 are shown in Figure 14.  56  a)  Summer 2009 Urban  Agriculture  Marshall  700 Mean dry mass (mg)  600 500 400 300 200 100 0 11  19  27  34  Days of colonization  b)  Winter 2009-10 Urban  Agriculture  Marshall  Mean dry mass (mg)  8000 7000 6000 5000 4000 3000 2000 1000 0 10  18  27  38  Days of colonization  57  c)  Spring 2010 Urban  Agriculture  Marshall  Mean dry mass (mg)  3500 3000 2500 2000 1500 1000 500 0 10  18  27  38  Days of colonization  Figure 14. Average dry mass (mg) of biofilms collected from urban, agricultural and Marshall mainstem stations in summer 2009 (a, n=2), winter 2009-10 (b, n=3) and spring 2010 (c, n=3) after x days of colonization. Mean ± 1 standard error.  Statistical tests did not show any significant differences between sites in the summer 2009 dataset (Figure 14a) and, similarly with chl-a, this is likely due to low sample sizes (n=2 for each time period). However, the biofilm dry mass is observably greater in the urban site than in the other two sites. It appears that the biofilm dry mass increases in a linear fashion between 11 and 19 days of colonization, and then dry mass plateaus or decreases slightly between 27 and 34 days. The dry mass plateaued, but the decrease in chl-a (Figure 14a) indicates that there might be a shift away from photosynthetic biomass to non-photosynthetic. This shift could be due to a number of micro-ecological relationships, such as competition between algae and bacterial colonies, the growth rate reaching a saturation point, preferential grazing or decomposition (Battin et al., 2007; Hill et al., 2001; Bothwell, 1989). In the winter 2009 dataset (Figure 14b), the dry mass decreases sharply between 19 and 27 days of colonization at the urban site, but the agricultural tributary and Marshall Creek biofilm dry masses increase in a linear fashion over the entire colonization period. This drop and recovery of dry mass in the urban tributary could be due to the more flashy 58  response to precipitation, which is expected in urbanized streams. Between 19 and 27 days of colonization, 85 mm of precipitation fell in Abbotsford with the majority of the precipitation falling in two storm events (Environment Canada, 2010). This major influx of water over a short period of time may have eroded/scoured the biofilms from the tiles, and then followed by a period where biofilm growth recovered (Battin et al., 2003). The other two sites drain lower % IA in their catchments and so are less flashy and would not experience runoff as intense as in the urban tributary. Overall, in the winter season, biofilm dry mass was significantly higher in the urban site than in the Marshall mainstem site after 10 days of colonization, as well as the agricultural tributary after 18 days of colonization (M-W U, p<0.015). The agricultural tributary also had significantly higher dry mass than the Marshall mainstem site at all times except after 18 days (M-W U, p<0.015). In the spring 2010 dataset, overall biofilm dry mass in the urban tributary is significantly higher than in the Marshall Creek mainstem (M-W U, p<0.015) and in the agricultural tributary (except for after 27 days of colonization) (Figure 14c). There is no significant difference in biofilm dry mass between the agricultural tributary and Marshall Mainstem site after 11 and 18 days of colonization, but after 27 days, Marshall mainstem dry mass drops significantly below the agricultural tributary, and then after 35 days, Marshall mainstem dry mass is significantly higher than the agricultural tributary (M-W U, p<0.015). The urban tributary and Marshall mainstem appear to follow similar growth trends, with the mainstem site at a lower magnitude. In the spring season, the drop in dry mass between 18 and 27 days of growth cannot be attributed to any detectable precipitation event, like in the winter season. 5.2.4.1.3 Trace Elements Table 21 summarizes the trace element data measured from biofilms that were collected at 3 sites in the Marshall Creek watershed between July 2009 and June 2010.  59  Table 21. Summary of trace elements measured in biofilms collected in the Marshall Creek watershed. (n=100)  Element Cr (mg/kg) Cu (mg/kg) Fe (mg/kg) Mg (mg/kg) Mn (mg/kg) Ni (mg/kg) P (mg/kg)  Range 0 – 398 0 – 98.2 2880 – 88800 81.4 – 26500 86.9 – 2020 0 – 260 118 – 3970  Mean 30.4 25.0 22100 4200 593 45.6 1320  Guidelines do not currently exist for trace element concentrations in biofilms (Canadian Council of Ministers of the Environment, 1999). These trace element concentrations are discussed in detail in section 5.2.5.2. The complete biofilm dataset is presented in Appendix D.  5.2.5  Comparisons between in-stream indicators 5.2.5.1 Correlations  The biofilm data collected at the three sample sites were pooled together and Spearman’s Rank Correlations were run against other indicators measured at the sites (α/12 pairs = 0.00417). Significant positive correlations were found between average biofilm chlorophyll a and temperature and pH (Table 22, in bold). Table 22. Spearman’s rank correlations between average biofilm chl-a and water quality indicators. (R=correlation coefficient, p=probability, n=sample size).  Average biofilm dry mass (mg) Average R -0.374 chl-a p 0.012 (mg/L) n 44  NO3--N (mg/L)  SRP (mg/L)  ClNH4+-N (mg/L) (mg/L)  pH  -0.455 0.022 25  -0.422 0.035 25  -0.479 0.015 25  -0.537 0.734 0.807 <0.0001 <0.0001 0.007 24 25 25  -0.545 0.005 25  Temp. (°C)  Turbidity (NTU)  Correlations were then determined between biofilm and water quality data from each site individually (not pictured) and the significant positive correlation that was also  60  consistently found within each individual site was between chl-a and water temperature (Spearman’s rank, p < 0.00417). 5.2.5.2  Trace elements in sediments and biofilms  Comparisons were made between metal concentrations (Cr, Cu, Mn and Ni) in suspended sediments and biofilms collected at the same sites using the Mann-Whitney U test (α=0.05). In each of the sites, all the metal concentrations were significantly higher in the suspended sediments than in the biofilms (M-W U, p < 0.05). The metal concentrations are summarized in Table 23. Table 23. Average metal concentrations (mg/kg) ± 1 S.E. in suspended sediments and biofilms collected from three sites in the Marshall Creek watershed.  Metal Cr Cu Mn Ni P  Marshall Mainstem Sediment Biofilm 36.9 ± 3.92 28.0 ± 1.88 110 ± 11.7 30.4 ± 0.694 2450 ± 435 870 ± 16.0 92.8 ± 4.71 55.9 ± 1.51 2290 ± 130 1320 ± 21.2  Urban Tributary Sediment Biofilm 117 ± 13.4 28.7 ± 1.39 145 ± 15.5 24.3 ± 0.252 2240 ± 385 445 ± 8.93 61.8 ± 6.50 26.7 ± 0.735 1400 ± 176 665 ± 13.6  Agricultural Tributary Sediment Biofilm 69.6 ± 10.3 34.3 ± 2.19 75.2 ± 2.08 20.7 ± 0.336 715 ± 66.5 481 ± 10.7 241 ± 53.8 54.8 ± 1.41 2580 ± 138 1980 ± 22.5  Trace element concentrations were not measured in the stream water, but correlations were run between total phosphorus in biofilms and SRP in the water collected from the same site at the same time. No significant correlations were found between SRP in water and P in biofilms in any of the sites (Spearman’s Rank, p>0.05). Figure 15 shows the average overall concentrations of P in sediments, biofilms and SRP in water collected at the three different sites of interest.  61  6000  0.14  5000  0.12 0.1  4000  SRP in water (mg/L)  P in sediments and biofilms (mg/kg)  Mean  0.08 3000 0.06 2000  0.04  1000  0.02 0  Urban Trib.  Agricultural Trib.  Water  Biofilm  Sus. Sediment  Bed Sediment  Water  Biofilm  Sus. Sediment  Bed Sediment  Water  Biofilm  Sus. Sediment  Bed Sediment  0  Marshall Mainstem  Figure 15. Box-whisker plot showing P concentrations in bed (n=1) and suspended sediments (n=7) and biofilms (n=34) and SRP (n=17) in water samples collected from three sites in the Marshall Creek watershed. (Error bars represent min/max values). Note: Water SRP concentrations are on the right-hand vertical axis.  The concentrations of P in the suspended sediments were significantly higher than in the biofilms in the urban tributary (M-W U, p=0.005) and the Marshall mainstem site (p=0.008), but there was no significant difference between sediments and biofilms in the agricultural tributary (Figure 15, Table 23). It appears that there was greater P uptake/sorption in the biofilms at the agricultural tributary than in the other two sites. The SRP concentrations were lowest in the urban tributary and higher in agricultural tributary and Marshall Creek mainstem, where the influence of agricultural runoff is present. The late summer bed sediment P concentrations were generally more similar to the biofilm concentrations than the suspended sediments, except for in the Marshall mainstem site. The next section investigates these dynamics further.  62  5.2.5.3 Biofilm : sediment ratios The ratios of average trace element concentrations (Cu, Mn, P, Ni and Cr) measured in bed sediments, suspended sediments and biofilms are presented in Table 24. These ratios provide an indication of the exchange of these elements between the different forms in the aquatic system. Table 25 summarizes these ratios within each of the three sites: the urban tributary, agricultural tributary and Marshall Creek mainstem. Table 24. Trace element ratios for the pooled dataset of biofilms and sediments in the Marshall Creek watershed.  Cu Biofilm : Sus. Sediment 0.224 Biofilm : Bed Sediment 0.508 Sus. Sediment : Bed Sediment 2.27  Mn 0.343 0.801 2.34  P 0.650 0.948 1.46  Ni 0.340 0.219 0.645  Cr 0.373 0.339 0.910  63  Table 25. Trace element ratios in biofilms and sediments by site (Marshall mainstem (Mar.), urban tributary (Urb.) and agricultural tributary (Ag.)). Cu Biofilm : Sus. Sediment Biofilm : Bed Sediment Sus. Sediment : Bed Sediment  Mn  P  Ni  Cr  Mar.  Urb.  Ag.  Mar.  Urb.  Ag.  Mar.  Urb.  Ag.  Mar.  Urb.  Ag.  Mar.  Urb.  Ag.  0.276  0.167  0.275  0.355  0.199  0.673  0.576  0.475  0.767  0.602  0.432  0.227  0.758  0.246  0.492  0.475  0.549  0.492  0.870  0.998  0.521  0.584  0.840  1.38  0.615  0.810  0.410  0.448  0.423  0.230  1.72  3.28  1.79  2.45  5.01  0.774  1.01  1.77  1.80  1.02  1.87  0.0932  0.591  1.72  0.466  The specific elements appear to be behaving differently in the aquatic system, as indicated by the different ratios in Table 24. This is due to the variable stability of elements when binding on charged functional groups in the biofilm matrix (Schorer & Eisele, 1997). Cu and Ni are expected to more readily react with the organic coatings in the biofilms, whereas Mn behaves like Fe and readily forms inorganic precipitates (Lion et al., 1988). However, the ratios shown here do not support the theory. Cu appears to have a greater affinity for the suspended sediment fraction because the suspended sediment is not entirely inorganic, but contains organic matter with high surface area. The significant differences in turbidity and TOC between the three sites likely play a role in the exchange of trace elements between the sediments and biofilms. Higher TOC could contribute to a greater affinity for trace elements to remain in the suspended fraction via organic complexing (particularly for Cu, less so for P) (Schorer & Eisele, 1997). The high variations show that sorption is not only dependent on the amount of trace elements available in the system, but also on the conditions of the sediments and biofilms. P is a conservative element/nutrient, whose fractionation among the three compartments is controlled dominantly by its inorganic chemistry. There is almost a 1:1 relationship (or equilibrium) between the concentration in biofilms and bed sediments (Table 24). The higher P concentration in the suspended sediments than the biofilms suggests that P is acting both as a nutrient and as an inorganic precipitate. Al or Fe could dominate the inorganic precipitate form at a pH of 7 (average pH in the watershed = 6.96) (Brady & Weil, 1996). Except for P in the agricultural tributary there are higher concentrations of trace elements in the bed and suspended sediments than in the biofilms (Table 25). There is a narrowing between the amount of trace elements in suspended and bed sediments (suspended : bed is closer to 1) from that found in the biofilms and suspended sediment (biofilm : suspended further from 1), suggesting that equilibrium is slowly becoming achieved between the sediment compartments. The discrepancy between biofilm and suspended sediment concentrations reveal that biofilms integrate the trace elements over a longer time period and are a slower sink in the stream environment than suspended sediments. They have a saturation point and can switch from a sink to a source for these elements  depending on the surrounding chemical environment, flow velocity and mixing (Ryder et al., 2007). Bed sediments also act as a sink for many trace elements in the stream environment, but can rapidly become a source when re-suspended into the water column. There is considerable potential to move (and remove from the watershed) trace elements and excess nutrients via suspended sediments, because they are mobile and they contain the highest concentration of trace elements of the three compartments measured in this system.  66  6  Summary  Table 26 summarizes the land use indicator data collected in the Marshall Creek watershed and the urban and agricultural subwatersheds. The in-stream results collected from each of the three sampling sites draining each watershed are presented in Table 27. Table 26. Summary of land use indicators for the Marshall Creek watershed and subwatersheds of interest.  Watershed  Area (ha)  % TIA  Average # Poultry per farm  Average # Cattle per farm  AUE/ha 2006 Population (est.)  Urban Agricultural Marshall Creek  165 73 3800  35 7 23  0 28,204 33,892  0 254 166  0 15.5 3.85  2006 Population Density (est.) (#/ha) 10.4 0.397 6.63  1716 29 25,181  Table 27. Summary of in-stream parameters measured at three sites. (Mean ± 1. S.E.)  Sample  Parameter  Water (n=17)  Bed Sediments (n=3)  Suspended Sediments (n=7)  NO3--N (mg/L)  Urban Tributary (Site 8) 1.82 ± 0.0306  Agricultural Tributary (Site 22) 1.97 ± 0.183  Marshall Creek Mainstem (Site 2) 4.15 ± 0.0428  SRP (mg/L)  0.00771 ± 0.0004  0.0435 ± 0.00215  0.0307 ± 0.00129  DO (mg/L)  9.43 ± 0.176  5.39 ± 0.110  7.24 ± 0.0929  pH  7.19 ± 0.00856  6.77 ± 0.0182  7.03 ± 0.0145  TOC (mg/L)  2.18 ± 0.0641  10.3 ± 0.369  4.41 ± 0.132  Turbidity (NTU)  3.09 ± 0.310  51.3 ± 1.15  13.0 ± 0.379  Sp. Cond. (µS/cm)  253 ± 3.12  443 ± 7.60  254 ± 1.90  Temperature (ºC)  9.88 ± 0.176  9.65 ± 0.343  8.99 ± 0.324  Cu (mg/kg)  44.2 ± 3.28  42.1 ± 7.86  64.0 ± 0.714  Ni (mg/kg)  33.0 ± 1.59  588 ± 345  90.9 ± 0.467  Mn (mg/kg)  446 ± 43.2  924 ± 336  1000 ± 195  P (mg/kg)  792 ± 25.9  1430 ± 390  2260 ± 206  Zn (mg/kg)  93.5 ± 7.44  116 ± 37.1  210 ± 3.97  Cu (mg/kg)  145 ± 15.5  75.2 ± 2.08  110 ± 11.7  Ni (mg/kg)  61.8 ± 6.50  241 ± 53.8  92.8 ± 4.71  Mn (mg/kg)  2240 ± 385  715 ± 66.5  2450 ± 435  P (mg/kg)  1400 ± 176  2580 ± 138  2290 ± 130  Zn (mg/kg)  271 ± 33.9  169 ± 6.93  249 ± 13.3  67  Sample  Parameter Chl-a (µg/L)  Urban Tributary (Site 8) 4500 ± 126  Agricultural Tributary (Site 22) 9260 ± 436  Marshall Creek Mainstem (Site 2) 5240 ± 203  Biofilms (n=34)  Dry mass (mg)  2000 ± 76.4  987 ± 47.7  444 ± 16.9  Cu (mg/kg)  24.3 ± 0.251  20.7 ± 0.336  30.4 ± 0.694  Ni (mg/kg)  26.7 ± 0.735  54.8 ± 1.41  55.9 ± 1.51  Mn (mg/kg)  445 ± 8.93  481 ± 10.7  870 ± 16.0  P (mg/kg)  665 ± 13.6  1980 ± 22.5  1320 ± 21.2  The urban subwatershed is the most densely populated and has the highest % total impervious area. The receiving tributary has a typically flashy hydrological regime, with large influxes of stormwater runoff containing fine sediments that quickly settle to the bed of the channel when flows have subsided. In periods between storm events, the water in the urban tributary has very low turbidity and TOC, and is relatively low in specific conductivity. The chlorophyll a in biofilms collected from the urban tributary is the highest in the winter, which is due to the higher water temperatures entering this tributary from pavement-warmed stormwater runoff. There are also elevated metals in biofilms in the urban tributary, which is likely due to the biofilms trapping incoming fine suspended sediments in the biofilm matrix. The agricultural subwatershed has the lowest population density and % TIA. However, it contains the highest AUE density at 15.5, which is considered extremely intense (USDA, 2008). There are three dairy cattle operations and one poultry operation in the agricultural subwatershed, which are producing a surplus of manure on a relatively small land base. This has led to high TOC, specific conductivity and turbidity in the agricultural tributary. There are also elevated concentrations of P in the suspended and bed sediments as well as biofilms. Biofilm chl-a is the highest in this tributary in the summer, which is related to warm temperatures and eutrophic conditions from high nutrients in the water. The Marshall Creek watershed’s overall population density falls between that of the urban and agricultural tributaries, which was expected. There are greater numbers of poultry per farm in the Marshall Creek watershed than in the agricultural watershed,  68  which is due to the higher poultry density over the Abbotsford Aquifer than in the Sumas Prairie, where the agricultural watershed is located. However, there are fewer AUEs/ha in the Marshall Creek watershed (3.85 versus 15.5), which is due to the larger area of land to assimilate manure from the additional animals. Still, 3.85 AUEs/ha exceeds the recommended level of 1.2-2.5 (USDA, 2008). The mouth of Marshall Creek is elevated in nitrate-N concentrations, due to the groundwater influence from the Abbotsford Aquifer on the main channels flows. The dynamics of the in-stream parameters measured at this site are more complicated than in the two tributaries not only because of the groundwater influence, but also because of the flow regulation downstream from the Barrowtown Pump Station. Because the floodgates are closed between May and September, the water level in Marshall Creek is artificially high and flows are minimal in the summer, which affects the temperature and dissolved oxygen. These two parameters are the drivers of many chemical reactions in the stream and therefore the nutrient and trace element data measured at this site are considered atypical. That being said, many of the parameter ranges measured on Marshall Creek (site 2) fall between the ranges measured in the tributaries during this study period, showing that the tributaries are still contributing significantly to the downstream site.  69  7  Conclusions  Examining the cumulative effects of urban and agricultural land uses in the Marshall Creek watershed provided considerable information from water, bed sediments, suspended sediments and biofilms. Great variability was found in the levels of indicators from the headwaters to the mouth of Marshall Creek and between tributaries. This examination was done to increase the understanding of the influence of three primary landscape stressors identified in this watershed: 1) Effects of agriculture outside of the watershed boundary through groundwater contamination and discharge; 2) Effects of runoff within the watershed from agricultural activities; and from 3) upland urban development. In addition, compartmentalization of trace elements between different instream indicators was examined to describe the complexity of pathways of contamination. Based on the findings, the following main conclusions can be drawn: 7.1  Land use trends  The Marshall Creek watershed is a representative example of rapid development in lower Fraser Valley of BC, particularly in urban residential growth and agricultural intensification. Although the area under agriculture has remained constant for decades, the number of livestock is growing. The general trend is that the number of farms is shrinking but the total number of animals is growing, resulting in larger more intensive livestock operations per hectare. This is particularly notable in poultry production. Based on an animal unit equivalent calculation of 3.85 for the watershed, these high densities of livestock are producing more waste than the land can assimilate. In the agricultural subwatershed, the AUEs were calculated to be 15.5, which is unsustainable. This example indicates that there are hot spots of potential contamination throughout the watershed that need to be investigated and managed accordingly. Some measures have been taken to export animal manure out of the watershed area and use it to fertilize crops elsewhere, but a capacity has been reached and there is an excess manure issue. The groundwater and surface water resources of the area are being affected by the nutrients that are leaching and running off the excess manure generated in the watershed. Although a large proportion of the Abbotsford Aquifer falls outside of the Marshall Creek watershed boundary, the aquifer does contribute to the quality and amount of water 70  within the boundaries of the study area. This example demonstrates that it is important to consider land use outside the traditional watershed when groundwater plays a role in stream discharge. In terms of urban development, there is also limited land available for growth in the watershed. The BC government has zoned most of the arable lands in the lower Fraser Valley within the ALR, thereby preventing urban development in the lowlands. This has left the upland areas in the watershed as the only available land for urban residential growth for the City of Abbotsford. Sumas Mountain is rapidly being deforested and developed with dominantly single and multiple-family residences. The population has increased from approximately 20,800 to 25,180 residents in the watershed from 19962006. The high impervious surface area and flashy response of the urban tributary during the wet season demonstrates the effect of this growth. 7.2  Seasonal and spatial trends in water quality  There were several trends in upstream to downstream water quality in Marshall Creek. Most notably was the decreasing trend in nitrate-N from the headwaters to the mouth, which is indicative of the decreasing proportion of groundwater to the total discharge of the creek downstream. The source of baseflow to Marshall Creek is the unconfined Abbotsford Aquifer, which is contaminated with nitrate-N via leaching from intensive agricultural activities over recharge zones. Elevated concentrations of nitrate-N exist near the headwaters of Marshall Creek. There are no other major groundwater resources entering the stream. Therefore, any major changes in water quality in Marshall Creek are related to surface runoff mechanisms from elsewhere in the watershed. The tributaries are diluting the nitrate-N concentrations downstream in the mainstem. On the other hand, TOC concentrations increase in Marshall Creek in the downstream direction. This suggests that the mechanism of TOC entering Marshall Creek is via surface runoff, particularly from agricultural tributary inputs and less from urban inputs. A potential source of TOC might be manure that is applied on croplands in the watershed. Over the Abbotsford Aquifer, the potential sources would be generally the same, but TOC does not appear to be entering Marshall Creek via groundwater discharge as it is assimilated into the soil. 71  There is not a clear upstream to downstream trend of SRP, although there is greater variability in SRP at the mouth of Marshall Creek than in the headwaters. Similarly to TOC, phosphorus generally enters surface waters via surface runoff since P has a high affinity to adsorb to soil and sediment particles. The potential sources of P to Marshall Creek are likely organic P in manure and inorganic P in fertilizers from agricultural and urban activities. Phosphorus is not a labile nutrient like nitrate-N and so it is not found in high concentrations as SRP in groundwater in the Abbotsford Aquifer. Although there have been considerable changes in land use in the Marshall Creek watershed and over the Abbotsford Aquifer since the early 1990s, this has not been reflected in nitrate-N concentrations in Marshall Creek. A historic time-series of nitrate-N data collected at site 11 showed no long-term trend but showed seasonal variability. The timescales over which groundwater chemistry changes are considerably longer than for surface water and so baseflow concentrations have a relatively constant nitrate-N average through time. That being said, the seasonal variability of nitrate-N concentrations in Marshall Creek is apparent, with higher average concentrations of nitrate-N in the summer dry season, when the majority of the baseflow is from groundwater. In the wet winter season, there is a greater proportion of surface runoff, which generally contains lower concentrations of nitrate-N than the groundwater in this system. 7.3  Trends in bed sediment quality  There is clear evidence that Cu, Mn, Zn and P concentrations in bed sediments have increased since the early 1990s. Both urban and agricultural land uses are contributing to these increases. Cu and Mn concentrations have increased at about the same rate in the headwaters and mouth stations and the contribution of Cu is similar from the urban and agricultural tributaries. In contrast, the Zn and P concentrations represent a cumulative effect at the mouth of Marshall Creek and agriculture appears to be the dominant stressor that is contributing to these increases over time. The use of Zn in poultry feed and the increasing amounts of manure generated from livestock production are likely the main reasons for this increase.  72  7.4  Interactions between water, sediments and biofilms  There were significant differences in many measurement endpoints (in-stream indicators) among sampling sites in the Marshall Creek watershed, particularly TOC, P, DO, Cu, Ni and biofilm chl-a. Phosphorus concentrations in suspended sediments at the mouth of Marshall Creek were not significantly different from those collected in the agricultural tributary but were significantly higher than in the urban tributary. This would indicate that runoff from agriculture is affecting the downstream sediments and potentially making more P available to biota in the form of SRP. The lower P in suspended sediments at the urban tributary indicates that potential sources of P, such as inorganic fertilizers, are not reaching the stream via stormwater runoff. This is reflected in low SRP concentrations in the water as well. P concentrations in biofilms further support this observation: P is highest in the biofilms at the agricultural tributary, followed by the Marshall mainstem and lowest at the urban tributary. This would indicate that there is greatest uptake of P by biofilms in the agricultural tributary. In the agricultural tributary, there is no significant difference between P in suspended sediments and in biofilms, whereas in the other two sites, there is significantly higher P in the suspended sediments. The elevated TOC in the agricultural tributary likely had an influence on the P dynamics between the sediments and biofilms, with more organic P existing in both compartments. Chl-a in biofilms provided an indicator of primary productivity and varied between site and season. Chl-a production was positively correlated with temperature, which is clearly seen in the winter season when warmer water is entering the urban tributary compared to the others and this results in higher chl-a. Elevated chl-a concentrations were found in the agricultural tributary in the summer due to overall warmer temperatures driving productivity. 7.5  Seasonal effects  There is seasonal variability in many of the in-stream indicators measured in Marshall Creek. Nitrate-N showed seasonal variability in the short term (2008-2010) and the long73  term (1994-2010). Nitrogen enters Marshall Creek through leaching into groundwater followed by discharge, and also by surface and subsurface runoff. Nitrogen is generally labile in the NO3- form, and yet is held by the soil and sediment in the NH4+ form. Both of these forms are influenced by the chemical state of the surrounding stream environment (e.g. DO, temperature, pH). The variability in the forms that nitrogen takes (NO3-, NH4+) in the stream system does not make it a reliable indicator of cumulative effects. Phosphorus is a more conservative indicator in that it is less influenced by DO, etc.; therefore, its reaction rates, in relation to environmental fluctuations, are slower and are better indicators of effects from stressors. P is associated primarily with surface runoff and can be predicted more reliably on a seasonal basis. 7.6  Suspended sediments and biofilms as indicators of cumulative effects  This study employed relatively new indicators, suspended sediments and biofilms, in addition to traditional water quality and bed sediment indicators to assess the state of the Marshall Creek watershed. If the biological condition of a stream ecosystem is of concern (VEC), then the closer the indicator to biological processes, the more likely the indicator will be successful in providing the information needed to meet the objectives. Thus, biofilms are better indicators of stream health than snap shots of water quality. For example, the aquatic health guideline for nitrate-N was exceeded throughout Marshall Creek, but fish are still present, thereby showing that nitrate-N is not the best indicator to use for aquatic health. Therefore, the isolated guideline is not applicable here. Suspended sediments and biofilms provide an inexpensive and integrated indicator of the more bioavailable contaminants in a stream system over time. Ongoing monitoring of stream quality using integrated indicators, such as biofilms, is recommended to better grasp long term and seasonal changes in the bioavailable fraction of contaminants.  7.7  Cumulative effects assessment  This study is a precursor to Watershed Cumulative Effects Assessment, as proposed by the new CWN framework. This framework has been shown to be a useful approach 74  (conceptual model), as it aids in the identification of causal factors. The fact that the responses of parameters cannot be unequivocally assigned to a specific land use as a stressor is not unexpected. However, the incorporation of measurements associated with suspended sediments and biofilms with the more traditional analyses of the water column and bed sediment, provides a more integrated assessment over time. In the Marshall Creek watershed, stressor-response relationships are not developed well enough to predict the effects of a new development in the watershed. The extent of non-point sources of pollution within the watershed makes it very difficult to link effects to a particular stressor on the land. The framework and the results of this study have aided in the understanding of the different factors influencing stream dynamics, such as the temperature moderation of urban runoff on stream productivity during the winter months and the confirmation that P is a reliable marker for aqueous system dynamics in that there is a tendency to establish equilibrium conditions between suspended sediments, biofilms and bed sediments. Water and sediment quality results from the urban and agricultural tributaries were generally as expected based on the current literature, but the influence of agriculturally impacted groundwater at the headwaters of Marshall Creek make this analysis more complex and partially masks the cumulative effects downstream (site 2). However, in many instances, the data showed the Marshall mainstem to fall in between the ranges of the two tributaries for particular indicators (e.g. Chl-a in biofilms, Cu in suspended sediments, etc.) In these cases, the response was similar to what was expected after the groundwater influence is considered. In contrast, Zn and P indicators show a response that is dominated by agricultural inputs. Based on the overall results, it appears that the conditions at the mouth of Marshall Creek are more similar to those at the agricultural tributary than those at the urban tributary. If this is the case, this suggests that agricultural activity is controlling the health of the Marshall Creek watershed and it should be the target of better management practices to reduce the impacts on the aquatic health of the system. Focusing watershed management efforts on agricultural activity over the Abbotsford Aquifer and in the Sumas Prairie will work towards protecting the VECs in the Marshall 75  Creek watershed. However, just because the urban tributary appears to be having less of an effect on Marshall Creek at present does not mean that preventative measures should not be taken to reduce future impact. The rate of urban growth on Sumas Mountain is rapid and development is relatively new and ongoing. It is possible therefore, that more observable effects might be delayed. Better management practices for these uplands as well as for the agricultural lowlands are recommended.  7.8  Limitations and future research opportunities  In retrospect, there were some challenges in assessing cumulative effects in the Marshall Creek watershed. These challenges provide opportunities for future research. 1. Stream discharge measurements and loading calculations The flow at the mouth of Marshall Creek is controlled by the floodgates at the Barrowtown Pump Station, which results in a stagnant flow regime at certain times of the year and occasional reversed flows. The water level changes that were monitored were insufficient to arrive at a reliable stage-discharge rating curve. This, together with the groundwater contribution at the headwaters of Marshall Creek made it impossible to determine contaminant loadings. When investigating contaminant movement from landscapes into the aquatic environment, it is important to quantify flow patterns in a stream so that loadings can be estimated. This data is particularly important for watershed managers, who can set recommendations and guidelines for loadings of a contaminant from a specific land use. Loadings measure the amount of a contaminant entering the system and thus provide a more refined indication of the relative contributions of urban, agricultural and groundwater discharges to stream health. 2. Organic matter content The amount of organic matter in a stream system is highly variable and has significant impacts on trace element analyses. This is particularly important when investigating trace element exchanges between different compartments in a 76  stream, such as biofilms, suspended sediments, bed sediments and the water column. In this study, the major differences in trace element concentrations are mostly the result of organic matter differences. Thus, it is important to quantify these finer scale dynamics in future research to help elucidate what form of P, for example, that is dominant in the stream at different times of year. 3. Biofilm experiments There is evidence that light, water temperature, turbidity, nutrients and flow variations have a significant influence on biofilm growth and productivity. Much of the research involving stream biofilms is done in small headwater streams where nutrients are a limiting factor. Further research involving biofilms as a cumulative effects indicator needs to be conducted in watersheds that are affected by different types and intensities of development. This is particularly important when landscape stressors exhibit seasonal variation (e.g. spring-time manure application), which alters the expected biofilm growth dynamic.  77  References  Alig, R. J., Kline, J. D. & Lichtenstein, M. (2004). Urbanization on the US landscape: Looking ahead in the 21st century. Landscape and Urban Planning, 69, 219-234. Almasri, M. N. & Kaluarachchi, J. J. (2004). Assessment and management of long-term nitrate pollution of ground water in agriculture-dominated watersheds. Journal of Hydrology, 295, 225-245. Battin, T. J., Kaplan, L. A., Newbold, J. D. & Hansen, C. M. E. (2003). Contributions of microbial biofilms to ecosystem processes in stream mesocosms. Nature, 426, 439442. Battin, T. J., Sloan, W. T., Kjelleberg, S., Daims, H., Head, I. M., Curtis, T. P. & Eberl, L. (2007). Microbial landscapes: New paths to biofilm research. Nature, 5, 76-81. Environmental Management Act, S.B.C. Agricultural Waste Control Regulation. B.C. Reg. 131/92U.S.C. Part 4 (1992). Retrieved from http://www.bclaws.ca/EPLibraries/bclaws_new/document/ID/freeside/10_131_92 BC Ministry of Environment. (2001). Ambient water quality guidelines for organic carbon. Retrieved 11/22, 2010, from http://www.env.gov.bc.ca/wat/wq/BCguidelines/orgcarbon/ocarbon_over.html BC Ministry of Environment. (2007). Habitat wizard. Retrieved 06/17, 2009, from http://www.env.gov.bc.ca/habwiz/ Beaulieu, M. S. (2001). Intensive livestock farming: Does farm size matter? Agriculture and Rural Working Paper series, Working Paper No. 48). Ottawa, ON: Statistics Canada.  78  Berka, C. S. (1996). Relationships between agricultural land use and surface water quality using a GIS: Sumas river watershed, Abbotsford, B.C. M.Sc. Thesis, The University of British Columbia, Vancouver, B.C. Bibby, R. L. & Webster-Brown, J. G. (2006). Trace metal adsorption onto urban stream suspended particulate matter (Auckland region, New Zealand). Applied Geochemistry, 21(7), 1135-1151. Boenigk, J., Wiedlroither, A. & Pfandl, K. (2005). Heavy metal toxicity and bioavailability of dissolved nutrients to a bacterivorous flagellate are linked to suspended particle physical properties. Aquatic Toxicology, 71(3), 249-259. Bothwell, M. L. (1989). Phosphorus-limited growth dynamics of lotic periphytic diatom communities: Areal biomass and cellular growth rate responses. Canadian Journal of Fisheries and Aquatic Sciences, 46, 1293-1301. Bowes, M. J., Hilton, J., Irons, G. P. & Hornby, D. D. (2005). The relative contribution of sewage and diffuse phosphorus sources in the River Avon catchment, southern England: Implications for nutrient management. Science of the Total Environment, 344(1-3), 67-81. Brady, N. C. & Weil, R. R. (1996). Soil phosphorus and potassium. The nature and properties of soils (11th ed., pp. 445-473). Upper Saddle River, NJ: Prentice-Hall, Inc. Bryan, J. E. & Larkin, P. A. (1972). Food specialization by individual trout. Journal of the Fisheries Research Board of Canada, 29(11), 1615-1624. Burns, A. & Ryder, D. S. (2001). Potential for biofilms as biological indicators in Australian riverine systems. Ecological Management & Restoration, 2(1), 53-63. Canadian Council of Ministers of the Environment. (1999). Protocol for the derivation of Canadian sediment quality guidelines for the protection of aquatic life. (No. CCME  79  EPC-98E.). Ottawa: Prepared by Environment Canada, Guidelines Division, Technical Secretariat of the CCME Task Group on Water Quality Guidelines. Canadian Council of Ministers of the Environment. (2005). Canadian water quality guidelines – Phosphorus. Environment Canada. Canadian Council of Ministers of the Environment. (2007). Canadian water quality guidelines for the protection of aquatic life. Environment Canada. Canadian Environmental Assessment Agency. (1999). Operational policy statement: Addressing cumulative environmental effects under the Canadian Environmental Assessment Act. No. Cat No. OPS_EOP/3-1999). Hull, QC. Cianfrani, C. M., Hession, W. C. & Rizzo, D. M. (2006). Watershed imperviousness impacts on stream channel condition in southeastern Pennsylvania. Journal of the American Water Resources Association, 42(4), 941-956. Clements, W. H. (1994). Benthic invertebrate community responses to heavy metals in the upper Arkansas River Basin, Colorado. Journal of the North American Benthological Society, 13, 30-44. Coulter, C. B., Kolka, R. K. & Thompson, J. A. (2004). Water quality in agricultural, urban, and mixed land use watershed. Journal of the American Water Resources Association, 40(6), 1593-1601. Courtney, L. A. & Clements, W. H. (2002). Assessing the influence of water and substratum quality on benthic macroinvertebrate communities in a metal-polluted stream: An experimental approach. Freshwater Biology, 47, 1766-1778. Curtis, L. R. & Smith, B. W. (2002). Heavy metal in fertilizers: Considerations for setting regulations in Oregon. Salem, OR: Oregon Department of Agriculture.  80  Dougherty, M., Dymond, R. L. & Zipper, C. E. (2004). Quantifying NPS pollutant delivery in an urbanizing headwater basin. Water Resource Management for the Commonwealth, 12. Droppo, I. G. & Ongley, E. D. (1994). Flocculation of suspended sediment in rivers of southeastern Canada. Water Research, 28(8), 1799-1809. Dubé, M. G. (2009). Toward a framework for watershed-based cumulative effects assessment. Personal communication with author. Duinker, P. N. & Greig, L. A. (2006). The impotence of cumulative effects assessment in Canada: Ailments and ideas for redeployment. Environmental Management, 37(2), 153-161. Environment Canada. (2010). National climate data and information archive. Retrieved 07/14, 2010, from http://www.climate.weatheroffice.gc.ca/climateData/canada_e.html European Commission. (2002). Implementation of council directive 91/676/EEC concerning the protection of waters against pollution caused by nitrates from agricultural sources. Italy: European Communities. Retrieved from http://ec.europa.eu/environment/water/water-nitrates/pdf/91_676_eec_en.pdf Foley, J. A., DeFries, R., Asner, G. P., Barford, C., Bonan, G., Carpenter, S. R., Chapin, F. S., Coe, M. T., Daily, G. C., Gibbs, H. K., Helkowski, J. H., Holloway, T., Howard, E. A., Kucharik, C. J., Monfreda, C., Patz, J. A., Prentice, I. C., Ramankutty, N. & Snyder, P. K. (2005). Global consequences of land use. Science, 309, 570-574. Han, F. X., Kingery, W. L., Selim, H. M. & Gerard, P. D. (2000). Accumulation of heavy metals in a long-term poultry waste-amended soil. Soil Science, 165(3), 260-268. Health Canada. (2008). Guidelines for Canadian drinking water quality - summary table. Health Canada. 81  Herb, W. R., Janke, B., Mohseni, O. & Stefan, H. G. (2008). Thermal pollution of streams by runoff from paved surfaces. Hydrological Processes, 22, 987-999. Hill, W. R., Mulholland, P. J. & Marzolf, E. R. (2001). Stream ecosystem responses to forest leaf emergence in spring. Ecology, 82(8), 2306-2319. Hodgson, G. W. (1983). Biogeochemistry of stream waters: A comparison of the effects of suspended sediments and river-bed biofilms. In R. Hallberg (Ed.), Environmental Biogeochemistry Ecological Bulletin (Volume 35 ed., pp. 85-97) Holding, K. L., Gill, R. A. & Carter, J. (2003). The relationship between epilithic periphyton (biofilm) bound metals and metals bound to sediments in freshwater systems. Environmental Geochemistry and Health, 25, 87-93. Horowitz, A. J. & Elrick, K. A. (1987). The relation of stream sediment surface area, grain size and composition to trace element chemistry. Applied Geochemistry, 2(4), 437-451. Integrated Resource Consultants Inc. (1994). Agricultural land use survey in the Sumas River watershed summary report. (No. DOE FRAP 1994-21). Vancouver, BC: Environment Canada. Karr, J. R. & Dudley, D. R. (1981). Ecological perspective on water quality goals. Environmental Management, 5(1), 55-68. Kennett, S. A. (2000). The future for cumulative effects management: Beyond the environmental assessment paradigm. (No. 69). Calgary, AB: Resources: The Newsletter of the Canadian Institute of Resources Law. Khan, I. U. H., Gannon, V., Loughborough, A., Jokinen, C., Kent, R., Koning, W., Lapen, D. R., Medeiros, D., Miler, J., Neumann, N., Phillips, R., Robertson, W., Schreier, H., Topp, E., van Bochove, E. & Edge, T. A. (2009). A methods comparison for the isolation and detection of thermophilic Campylobacter in agricultural watersheds. Journal of Microbiological Methods, 79(3), 307-313. 82  Kiffney, P. M. & Bull, J. P. (2000). Factors controlling periphyton accrual during summer in headwater streams of southwestern British Columbia, Canada. Journal of Freshwater Ecology, 15(3), 339-351. Kowalenko, C. G., Schmidt, O. & Hughes-Games, G. (2007). Fraser Valley soil nutrient study, 2005. Abbotsford, BC: British Columbia Agriculture Council. Kröpfl, K., Vladár, P., Szabó, K., Ács, É., Borsodi, A. K., Szikora, S., Caroli, S. & Záray, G. (2006). Chemical and biological characterization of biofilms formed on different substrata in Tisza River (Hungary). Environmental Pollution, 144, 626-631. Kuusisto-Hjort, P. (2009). Controls on trace metals in urban stream sediments implications for pollution monitoring using sediment chemistry data. Ph.D. Dissertation, University of Helsinki – Helsinki, Finland. Larrick, S. R., Dickson, K. L., Cherry, D. S. & Cairns, Jr., J. (1978). Determining fish avoidance of polluted water. Hydrobiologia, 61(3), 257-265. Lavkulich, L. M., Hall, K. & Schreier, H. (1999). Land and water interactions: Present and future. In M. C. Healey (Ed.), Seeking sustainability in the lower Fraser Basin: Issues and choices. Vancouver, BC: Institute for Resources and Environment, The University of British Columbia. Lion, L. W., Shuler, M. L., Hsieh, K. M., Ghiorse, W. C. & Corpe, W. A. (1988). Trace metal interactions with microbial biofilms in natural and engineered systems. Critical Reviews in Environmental Science and Technology, 17(4), 273-306. Lofts, S. & Tipping, E. (1999). Modelling the solid-solution partitioning of metals in environmental systems. Environmental Geochemistry and Health, 21, 299-304. Lyautey, E., Wilkes, G., Miller, J., van Bochove, E., Schreier, H., Koning, W., Edge, T., Lapen, D. R. & Topp, E. (In Press). Variation of an indicator of Escherichia coli persistence from surface waters of mixed-use watersheds, and relationship with population density and other factors. International Journal of Limnology. 83  Maier, G., Nimmo-Smith, R., Glegg, G., Tappin, A. & Worsfold, P. (2009). Estuarine eutrophication in the UK: Current incidence and future trends. Aquatic Conservation: Marine and Freshwater Ecosystems, 19(1), 43-56. Manly, B. F. J. (2001). Statistics for environmental science and management. New York: Chapman & Hall/CRC. McMahon, P. B., Böhlke, J. K., Kauffman, L. J., Kipp, K. L., Landon, M. K., Crandall, C. A., Burow, K. R. & Brown, C.J. (2008). Source and transport controls on the movement of nitrate to public supply wells in selected principal aquifers of the United States. Water Resources Research, 44(W04401), 1-17. Munkittrick, K. R. & Dixon, D. G. (1989). A holistic approach to ecosystem health assessment using fish population characteristics. Hydrobiologia, 188/189, 123-135. Murakami, M., Fujita, M., Furumai, H., Kasuga, I. & Kurisu, F. (2009). Sorption behavior of heavy metal species by soakaway sediment receiving urban road runoff from residential and heavily trafficked areas. Journal of Hazardous Materials, 164(2-3), 707-712. Neal, C., Reynolds, B., Neal, M., Hughes, S., Wickham, H., Hill, L., Rowland, P. & Pugh, B. (2003). Soluble reactive phosphorus levels in rainfall, cloud water, throughfall, stemflow, soil waters, stream waters and groundwaters for the upper River Severn area, plynlimon, mid Wales. The Science of the Total Environment, 314-316, 99-120. Novotny, V., Bartosova, A., O'Reilly, N. & Ehlinger, T. (2005). Unlocking the relationship of biotic integrity of impaired waters to anthropogenic stresses. Water Research, 39, 184-198. Phillips, J. M., Russell, M. A. & Walling, D. E. (2000). Time-integrated sampling of fluvial suspended sediment: A simple methodology for small catchments. Hydrological Processes, 14, 2589-2602.  84  Richards, C. E., Munster, C. L., Vietor, D. M., Arnold, J. G. & White, R. (2008). Assessment of a turfgrass sod best management practice on water quality in a suburban watershed. Journal of Environmental Management, 86(1), 229-245. Ryder, D., Vink, S., Bleakley, N. & Burns, A. (2007). Managing sources, sinks and transport of natural contaminants in regulated rivers: A case study in the Murrumbidgee River catchment, NSW. Proceedings of the 5th Australian Stream Management Conference. Australian Rivers: Making a Difference. Charles Sturt University, Thurgoona, NSW, 5 354-359. Schindler, D. W. (2006). Recent advances in the understanding and management of eutrophication. Limnology and Oceanography, 51(1), 356-363. Schorer, M. & Eisele, M. (1997). Accumulation of inorganic and organic pollutants by biofilms in the aquatic environment. Water, Air and Soil Pollution, 99, 651-659. Schreier, H. (2009). Agricultural water policy challenges in British Columbia. Policy Options, 30(7), 56-60. Retrieved from http://www.irpp.org/po/archive/jul09/schreier.pdf Schreier, H., Bestbier, R. & Derksen, G. (2003). Agricultural nutrient management trends in the lower Fraser Valley, B.C. Vancouver, B.C.: Institute for Resources, Environment and Sustainability, The University of British Columbia. Schueler, T. (1994). The importance of imperviousness. Watershed Protection Techniques, 1(3), 100-111. Schueler, T. R., Fraley-McNeal, L. & Cappiella, K. (2009). Is impervious cover still important? Review of recent research. Journal of Hydrologic Engineering, 14(4), 309-315. Smith, I. M. (2004). Cumulative effects of agricultural intensification on nutrient and trace metal pollution in the Sumas river watershed, Abbotsford, B.C. M.Sc. Thesis, The University of British Columbia, Vancouver, B.C. 85  Smith, I. M., Hall, K. J., Lavkulich, L. M. & Schreier, H. (2007). Trace metal concentrations in an intensive agricultural watershed in British Columbia, Canada. Journal of the American Water Resources Association, 43(6), 1455-1467. SPSS. (2010). Statistical model output, 11/25/2010. StatsCan. (1991). Agricultural census data for 1991. Ottawa, ON: Statistics Canada. StatsCan. (1996a). Agricultural census data for 1996. Ottawa, ON: Statistics Canada. StatsCan. (1996b). Population and dwelling counts, 1996 census. Ottawa, ON: Statistics Canada. StatsCan. (2001a). Agricultural census data for 2001. Ottawa, ON: Statistics Canada. StatsCan. (2001b). Population and dwelling counts, 2001 census. Ottawa, ON: Statistics Canada. StatsCan. (2006). Agricultural census data for 2006. Ottawa, ON: Statistics Canada. StatsCan. (2010). Census tract (CT) profiles, 2006 census of population. Retrieved 09/01, 2010, from http://www12.statcan.gc.ca/census-recensement/2006/dp-pd/prof/92597/index.cfm?Lang=E Steinmetz, R., Kunz, A., Dressler, V., Flores, E. & Martins, A. (2009). Study of metal distribution in raw and screened swine manure. CLEAN - Soil, Air, Water, 37(3), 239-244. Stutter, M. I., Langan, S. J. & Demars, B. O. L. (2007). River sediments provide a link between catchment pressures and ecological status in a mixed land use Scottish river system. Water Research, 41, 2803-2815. United States Department of Agriculture. (2008). Natural resources conservation service - Agricultural waste management field handbook. (No. H_210_NEH_651), USDA. Retrieved from http://policy.nrcs.usda.gov/viewerFS.aspx?hid=21430 86  United States Environmental Protection Agency. (1994). Sample preparation procedure for spectrochemical determination of total recoverable elements. Method 200.2, Revision 2.8. Cincinnati, OH: United States Environmental Protection Agency. Retrieved from http://water.epa.gov/scitech/swguidance/methods/bioindicators/upload/2007_07_10_ methods_method_200_2.pdf Vadas, P. A., Haggard, B. E. & Gburek, W. J. (2005). Predicting dissolved phosphorus in runoff from manured field plots. Journal of Environmental Quality, 34, 1347-1353. Walling, D. E. (1989). Physical and chemical properties of sediment: The quality dimension. International Journal of Sediment Research, 4(1), 27-39. Wang, L. & Kanehl, P. (2003). Influences of watershed urbanization and instream habitat on macroinvertebrates in cold water streams. Journal of the American Water Resources Association, 39(5), 1181-1196. Wang, L., Lyons, J. & Kanehl, P. (2001). Impacts of urbanization on stream habitat and fish across multiple spatial scales. Environmental Management, 28(2), 255-266. Ward, P. J. (2008). River Meuse suspended sediment yield: A new estimate and past estimates revisited. Netherlands Journal of Geosciences, 87(2), 189-193. Wassenaar, L. I. (1995). Evaluation of the origin and fate of nitrate in the Abbotsford aquifer using the isotopes of 15N and 18O in NO3-. Applied Geochemistry, 10, 391405. Wassenaar, L. I., Hendry, M. J. & Harrington, N. (2006). Decadal geochemical and isotopic trends for nitrate in a transboundary aquifer and implications for agricultural beneficial management practices. Environmental Science & Technology, 40, 46264632.  87  Werry, M. (2003). Match stocker numbers with resources & management style. Retrieved 11/17, 2010, from http://www.omafra.gov.on.ca/english/livestock/beef/facts/info_match.htm Wilson, J. E. (2009). Fish assemblage data collected by backpack electrofishing in Marshall Creek. Unpublished data. Wilson, J. T., McNabb, J. F., Balkwill, D. L. & Ghiorse, W. C. (1983). Enumeration and characterization of bacteria indigenous to a shallow water-table aquifer. Ground Water, 21(2), 134-142. Withers, P. J. A. & Jarvie, H. P. (2008). Delivery and cycling of phosphorus in rivers: A review. Science of the Total Environment, 400(1-3), 379-395. Wright, F. (2009). City of Abbotsford - Stormwater and agricultural drainage in the Marshall Creek watershed. Personal communication with author.  88  Appendices Appendix A: Maps of the Marshall Creek watershed Figure A1. Map of 2006 Census Enumeration Area boundaries (South Matsqui & Sumas Prairie) and Dissemination Area 59090721. Figure A2. Map of Census Tract boundaries for Population data. Figure A3. Map of Marshall Creek watershed with delineated urban and agricultural subwatersheds. Figure A4. Map of the Total Impervious Area (TIA) delineation for the urban subwatershed. Figure A5. Map of the Total Impervious Area (TIA) delineation for the agricultural subwatershed.  89  Figure A1. Map of 2006 Census Enumeration Area boundaries (South Matsqui & Sumas Prairie) and Dissemination Area 59090721.  90  Figure A2. Map of Census Tract boundaries for Population data.  91  Figure A3. Map of Marshall Creek watershed with delineated urban and agricultural subwatersheds.  92  Figure A4. Map of the Total Impervious Area (TIA) delineation for the urban subwatershed.  93  Figure A5. Map of the Total Impervious Area (TIA) delineation for the agricultural subwatershed.  94  Appendix B: Water sampling and analysis results Table B1. Summary of water quality results. Figure B1. Simple Seasonal Time-Series model output on historic nitrate-N data collected at site 11. Figure B2. Autocorrelation function graph of historic nitrate-N data collected at site 11.  95  Table B1. Summary of water quality results.  Site # 2  3  4  5  Site Description Marshall Creek mainstem mouth  Small urban tributary Small urban tributary  Urban detention pond  Date 8/18/08 12/15/08 06/18/09 09/17/09 10/28/09 11/25/09 12/11/09 12/21/09 12/29/09 01/07/10 01/18/10 02/10/10 04/15/10 04/26/10 05/04/10 05/12/10 05/31/10 8/18/08 12/15/08 06/18/09 8/18/08 12/15/08 06/18/09 09/17/09 8/18/08 12/15/08  NO3--N (mg/L) 4.54 4.79 4.20 3.42 19.2 5.60 4.75 4.19 4.50 4.85 3.81 4.43 3.59 3.53 3.11 3.63 3.38 1.01 1.39 0.608 0.398 0.852 0.572 0.366 0.000 0.714  SRP (mg/L) 0.00951 0.00418 0.0101 0.0194 0.323 0.0498 0.0139 0.0494 0.0362 0.0535 0.0495 0.0199 0.0759 0.0158 0.0405 0.0141 0.0299 0.00608 0.0158 0.000 0.00342 0.00350 0.00350 0.0266 0.00898 0.00377  Cl(mg/L) 20.7 29.3  NH4+-N (mg/L) 0.296 0.455  10.6 25.7 29.6 33.3 34.4 37.8 36.8 32.1 35.0 39.5 30.6 24.9 27.6 28.9 13.3 6.59  0.138 0.150 0.230 0.310 2.56 7.74 5.58 3.93 0.110 7.69 0.284 0.831 0.190 0.451 0.00 0.0829  30.3 17.1  0.110 0.121  39.0 18.8 6.75  0.0569 0.00620 0.0531  pH 7.10 6.96 6.99 7.14 6.16 6.58 6.85 6.97 6.93 6.87 6.87 6.87 7.18 7.17 7.05 7.64 7.24 7.34 7.39 6.99 7.52 7.44 7.54 7.45 6.91 7.28  Sp. Cond. (µS/cm) 233 205 245 269 257 253 231 221 288 258 259 249 329 258 217 271 278 194 249 170 228 135 240 283 119 98.0  Temp. (ºC) 0.200 15.9 18.0  DO (mg/L)  9.20 4.20 7.00 4.90 3.20 8.00 6.90 11.0 11.0 9.20 12.6 13.5  6.59 9.10 9.79 6.25 10.2 7.48 7.30 8.20 5.90 7.80 6.40 7.42 6.40 7.00 5.30  5.50 14.7  7.72  0.100 15.8 15.7 6.40  Turbidity TOC (NTU) (mg/L) 2.80 4.23  14.1 8.69 10.3 11.4 20.8 11.7 15.1 15.9 15.9 9.68 20.5 9.13 7.23  2.59 10.0 6.96 2.68 6.25 5.51 6.18 3.64 5.29  2.77 4.02 1.59 1.20 3.44 2.12  9.71 10.1  3.39 3.83 1.64  96  Table B1. Continued  Site Site Description # 7 Marshall Creek mainstem  8  Urban tributary DeLair Park  Date 8/18/08 12/15/08 06/18/09 09/17/09 12/11/09 01/18/10 05/31/10 8/18/08 12/15/08 06/18/09 09/17/09 10/28/09 11/25/09 12/11/09 12/21/09 12/29/09 01/07/10 01/18/10 02/10/10 04/15/10 04/26/10 05/04/10 05/12/10 05/20/10 05/31/10  NO3--N (mg/L) 5.60 5.27 4.26 5.66 5.44 4.61 3.00 1.97 2.34 2.00 1.66 8.67 2.55 2.54 0.929 2.03 2.32 2.30 1.96 1.59 1.58 1.36 1.60 1.51 0.760  SRP (mg/L) 0.0127 0.00406 0.00881 0.00330 0.0137 0.0273 0.0312 0.00622 0.00363 0.000168 0.00600 0.0167 0.0117 0.00650 0.00140 0.00760 0.00990 0.0154 0.00433 0.00877 0.00765 0.00649 0.00571 0.0296 0.00940  Cl(mg/L) 23.7 29.2 16.8 34.2 35.5 14.1 26.6 18.2 26.6 22.5 22.1 34.7 21.0 28.3 28.8 22.8 28.3 20.9 20.6 16.0 21.5 18.8 7.97  NH4+-N pH (mg/L) 0.00 7.20 0.596 6.95 7.03 0.0340 7.13 0.420 6.83 0.540 6.76 0.335 7.08 0.108 7.20 0.528 7.23 6.99 0.0860 7.37 0.0800 6.50 0.0600 7.07 0.0500 7.11 0.0800 7.06 0.0700 7.25 0.0800 7.09 0.0500 7.19 0.138 6.92 0.314 7.27 0.0757 7.31 0.0796 7.37 0.0665 7.47 0.234 7.11 0.199 7.14  Sp. Cond. (µS/cm) 235 196 245 252 226 236 208 309 227 315 343 319 268 279 139 263 240 229 262 257 262 242 272 260 136  Temp. DO Turbidity TOC (ºC) (mg/L) (NTU) (mg/L) 2.42 1.30 4.20 15.2 9.02 14.0 9.10 2.42 5.00 9.03 65.8 2.73 8.20 6.50 17.9 5.78 12.3 6.80 19.6 3.56 4.31 4.70 1.72 15.2 8.92 15.3 8.30 2.88 11.5 1.00 2.55 10.6 11.0 3.77 2.12 7.90 11.3 1.26 1.57 7.90 11.0 16.1 1.80 7.70 10.5 0.96 1.32 7.00 9.10 1.38 9.20 10.1 1.08 1.71 8.20 9.30 2.23 1.71 10.2 8.70 1.71 2.00 10.1 9.37 1.55 9.50 9.10 1.51 10.8 8.60 1.58 1.63 11.4 7.80 2.96 3.17 12.4 8.20 4.14 2.44  97  Table B1. Continued  Site 10  11  12 13 14  Site Description Tributary with groundwater influence Marshall Creek mainstem historic site  Groundwater Well (Hatchery) Tributary with groundwater influence Marshall Creek mainstem Hatchery discharge  8/18/08 12/15/08 09/17/09 8/18/08 12/15/08 06/18/09 09/17/09 10/28/09 11/25/09 12/11/09 01/07/10 01/18/10 02/10/10 04/15/10 05/31/10 8/18/08 06/18/09 8/18/08 12/15/08  NO3--N (mg/L) 1.31 1.24 1.30 6.78 6.93 5.40 7.26 10.0 6.16 7.35 6.31 5.25 5.24 5.81 3.67 9.97 7.77 11.7 7.90  SRP (mg/L) 0.00731 0.0105 0.0545 0.0111 0.0189 0.00522 0.0237 0.0194 0.0150 0.0139 0.0141 0.0168 0.0124 0.0170 0.0222 0.00716 0.0137 0.0108 0.0153  Cl(mg/L) 24.2 38.9 22.6 15.5 15.8  9.42 11.9  0.0297 0.397  8/18/08 12/15/08 06/18/09 09/17/09 12/11/09 01/18/10 04/15/10 05/31/10  9.90 9.75 8.04 9.14 9.54 7.78 7.84 6.16  0.0134 0.0141 0.00595 0.0265 0.0107 0.0007 0.0190 0.0266  13.9 14.4  0.00 0.369  Date  14.4 17.4 17.6 19.7 19.9 19.7 17.2 18.1 10.67 13.3  13.3 14.5 13.1 12.8 17.6  NH4+-N (mg/L) 0.191 0.637 0.247 0.00 0.362 0.210 0.140 0.100 0.250 0.120 0.100 0.131 0.360 0.463 0.0314  0.158 0.250 0.120 0.300 0.295  pH 7.25 6.80 6.87 7.29 7.19 7.19 7.01 6.87 6.92 7.70 7.13 7.13 7.10 7.37 7.25 6.97 7.24 7.34 6.98  Sp. Cond. (µS/cm) 196 176 220 208 182 230 243 202 206 210 194 205 213 228 170 230 270 293 192  7.14 7.02 7.26 7.10 7.63 7.11 7.29 7.13  248 212 260 262 227 214 252 230  Temp. DO (ºC) (mg/L) 1.40 16.1 4.30 12.8 12.4 9.50 7.00 6.90 9.60 8.40 11.0 12.2  8.47  10.6 10.4 10.8 10.0 10.9 10.6 9.30 9.20 8.90 8.10  Turbidity TOC (NTU) (mg/L) 6.45 3.92 3.94 1.62 1.67  3.58 8.20 12.0 8.76 5.13 5.21 3.22 8.87  3.28 2.69 2.48 1.23 1.75 1.43 1.80 3.33 0.500  9.25 1.42 1.65  4.80 5.00 13.6 12.5 7.80 9.50 12.0 11.5  2.74 1.38 11.1 9.70 10.1 8.54 9.30 7.50  5.12 2.89 1.69 6.82  1.45 0.96 1.35 1.24 2.29  98  Table B1. Continued  Site 15  16  17  18  Site Description Marshall Creek mainstem  Small agricultural tributary Marshall Creek mainstem  Marshall Creek mainstem  Date 8/18/08 12/15/08 06/18/09 09/17/09 05/31/10 8/18/08 12/15/08 06/18/09 09/17/09 8/18/08 12/15/08 06/18/09 09/17/09 12/11/09 01/18/10 05/31/10 8/18/08 12/15/08 06/18/09 09/17/09 11/25/09 12/11/09 01/07/10 01/18/10 02/10/10 04/15/10 05/31/10  NO3--N (mg/L) 8.78 5.58 7.85 9.00 6.33 0.0240 4.11 0.0320 0 6.14 6.51 4.88 6.55 6.63 4.84 2.71 6.18 5.73 4.77 5.88 7.34 5.83 6.29 5.18 4.72 4.52 2.34  SRP (mg/L) 0.0167 0.0211 0.00505 0.0243 0.0270 0.0140 0.0233 0.0173 0.00370 0.00996 0.0125 0.00703 0.00 0.0135 0.0179 0.0370 0.0144 0.0231 0.00956 0.00390 0.0461 0.0164 0.0255 0.0350 0.0471 0.0271 0.0250  Cl(mg/L) 14.8 14.5  NH4+-N (mg/L) 0.109 0.424  13.4 25.5 20.4 18.0  0.138 0.231 1.70 0.595  16.8 16.1 17.7  1.94 0.00 0.690  15.3 21.3 22.5 9.78 19.2 27.7  0.209 0.160 0.330 0.579 0.00 1.36  17.0 36.3 35.1 39.6 36.9 38.5 39.9 9.47  0.133 0.560 0.530 0.460 0.510 0.144 0.436 0.488  pH 7.35 7.35 7.26 7.11 7.14 6.70 6.65 6.52 6.34 7.26 7.07 7.04 7.02 7.79 6.94 7.05 7.33 6.98 7.03 6.89 6.54 7.65 6.85 6.81 6.94 7.24 7.08  Sp. Cond. (µS/cm) 229 209 250 256 241 242 221 210 297 221 174 250 243 209 207 155 227 197 270 250 316 231 233 244 249 277 147  Temp. DO Turbidity (ºC) (mg/L) (NTU) 5.00 13.5 12.4 11.2  11.0 10.4 7.80  2.40 13.8 13.5  3.73 3.40  3.10 13.4 13.1 6.40 9.10 12.5  9.80 10.4 11.0 8.70 7.40  15.9 8.34 13.2  1.80 14.0 13.8 9.80 5.60 4.70 8.70 7.30 13.0 12.4  8.73 7.50 5.10 9.80 9.30 7.00 7.90 8.30 7.20  13.5 63.5 22.6 17.7 122 14.7 19.3  9.18  TOC (mg/L) 1.65 1.56 1.27 2.20 9.33 9.30 6.12 2.75 3.04 2.22 1.94 3.65 6.32 2.27 4.56 2.44 7.83 2.80 5.76 3.57 2.96 4.74  99  Table B1. Continued  Site  Site Description  Date  NO3--N (mg/L)  SRP (mg/L)  22  Agricultural tributary - Kenny Rd.  06/18/09 09/17/09 10/28/09 11/25/09 12/11/09 12/21/09 12/29/09 01/07/10 01/18/10 02/10/10 04/15/10 04/26/10 05/04/10 05/12/10 05/20/10 05/31/10 06/18/09 09/17/09  0.0470 0.0690 19.5 7.71 0.852 7.88 1.14 4.78 3.66 0.447 0.130 0.282 0.895 0.0890 0.0880 1.45 1.72 1.11  0.0345 0.0466 0.0743 0.0793 0.0222 0.127 0.0178 0.0305 0.0643 0.0262 0.0104 0.0292 0.0684 0.00990 0.0195 0.0675 0.00316 0.0182  23  Small urban tributary  Cl(mg/L)  NH4+-N (mg/L)  137 51.9 78.9 98.9 66.5 127 120 127 143 167 113 86.7 128 171 52.5  1.48 0.740 1.19 1.63 1.13 1.47 0.960 1.60 2.15 1.34 1.14 1.11 1.04 1.70 1.25  38.0  0.0310  pH  Sp. Cond. (µS/cm)  Temp. (ºC)  6.91 6.54 6.34 6.31 7.27 6.32 6.63 6.89 6.64 6.60 7.00 6.99 6.71 6.98 7.04 6.77 7.83 7.76  365 568 320 369 466 279 325 357 332 457 660 580 352 542 541 450 400 390  17.3 16.4 9.20 2.60 5.90 2.30 2.90 7.70 5.80 17.0 10.2 9.30 12.4 12.4 13.4 15.2 14.9  DO (mg/L ) 3.75 1.70 7.42 3.40 6.80 4.80 6.25 7.40 4.60 5.20 6.10 7.41 7.50 5.70 5.70 4.50 10.1 10.9  Turbidity TOC (NTU) (mg/L)  15.4 26.5 54.2 37.2 56.1 69.6 48.0 76.9 52.6 42.2 60.4 50.2 125 42.0  10.4 15.0 12.9 6.37 15.2 6.82 10.1 7.69 6.84  8.23 9.58 19.4 2.68  100  Figure B1. Simple Seasonal Time-Series Series model output on historic nitrate nitrate-N data collected at site 11.  101  Figure B2. Autocorrelation function graph of historic nitrate-N N data collected at site 11.  102  Appendix C: Sediment sampling and analysis results Table C1. Summary of USEPA Method 200.2 – Total recoverable metals from sediments (USEPA, 1994). Table C2. Summary of bed sediment results. Table C3. Summary of suspended sediment results.  103  Table C1. Summary of USEPA Method 200.2 – Total recoverable metals from sediments (USEPA, 1994).  Step  Instruction  1  Approximately 0.5 g of the dried sample is added to a 250 mL Erlenmeyer flask.  2  10 mL of 1:4 HCl and 4 mL of 1:1 HNO3- (aqua regia) is added to the flask and covered in a watch glass.  3  In a fume hood, flasks are placed on a hot plate at 85°C and refluxed for approximately 30 minutes.  4  The hot plate is turned off and flasks are allowed to cool for 30 minutes.  5  The solution is filtered through a Whatman filter #42 into a 100 mL volumetric flask.  6  Solution is brought up to 100 mL using 5% HNO3-, covered with Parafilm and shaken 6 times.  7  Sample is transferred into a 60 mL Teflon sample container for analysis.  104  Table C2. Summary of bed sediment results. Site # 2 3 4 5  7  8 10  11 13 14  16  17  18 22 23  Date 08/18/08 09/17/09 06/18/09 09/17/09 08/18/08 08/18/08 06/18/09 09/17/09 08/18/08 06/18/09 09/17/09 08/18/08 09/17/09 08/18/08 06/18/09 09/17/09 08/18/08 08/18/08 09/17/09 08/18/08 06/18/09 09/17/09 08/18/08 06/18/09 09/17/09 08/18/08 06/18/09 09/17/09 06/18/09 09/17/09 09/17/09  Sample ID 102 302 203 304 105 107 207 307 108 208 308 110 310 111 211 311 113 114 314 116 216 316 117 217 317 118 218 318 222 322 323  Al (mg/kg) 20973 27806 15161 25310 24089 19562 21921 24480 12710 13749 20326 13134 17404 14939 43900 16389 18834 14999 20382 20244 14734 16555 15606 13820 19752 21469 15908 29007 7004 27442 15300  Ca (mg/kg) 6900 8122 6506 12012 4125 6241 6634 9268 5063 5225 9401 4801 8326 6424 10585 9277 5565 8725 13350 5873 6286 5006 6385 6065 9070 8113 6101 11843 2841 6343 7863  Cd (mg/kg) 8.4 1.46 0.89 1.07 4.3 8.6 1.98 1.60 3.9 1.20 1.344 7.6 1.58 3.6 1.45 1.38 5.8 8.7 1.62 12.2 2.29 2.51 6.7 1.50 1.91 7.8 1.79 1.78 0.80 1.64 1.24  Cr (mg/kg) 63.0 61.8 42.1 88.5 68.1 64.7 64.2 63.4 63.7 56.5 82.9 41.4 57.6 51.1 81.1 50.0 54.3 49.9 68.3 52.9 41.8 47.9 71.1 44.0 58.3 57.9 50.5 71.3 233.5 65.2 55.5  Cu (mg/kg) 63.0 65.0 34.4 38.4 42.1 98.4 78.3 80.9 42.4 35.5 54.9 78.0 64.7 66.3 59.7 58.8 325 81.9 84.5 60.5 74.9 73.8 58.7 49.4 59.3 68.9 51.5 82.4 30.9 53.2 38.0  Fe (mg/kg) 58669 41524 20620 27849 29234 61964 56083 50320 26283 25070 31670 52723 42913 38429 40347 34369 38217 59226 41660 84990 103579 104624 45809 43244 43200 53656 63609 49682 47665 38913 23107  K (mg/kg) 1183 2588 744 2390 1049 990 1061 1654 746 680 1305 682 876 759 3986 715 1080 938 1479 1170 798 632 805 541 853 1234 792 2303 307 1249 740  Mg (mg/kg) 8276 8313 4848 7661 6011 9129 9141 9748 5137 4446 6298 4865 7110 8953 15523 10712 7750 7416 9082 9599 7288 8605 8377 7146 9825 9716 7902 11069 118822 10583 5495  Mn (mg/kg) 724 1277 272 1156 328 560 605 623 382 361 595 1327 2026 964 511 1296 462 1685 1168 464 424 429 753 639 1396 716 455 867 1398 449 601  Ni (mg/kg) 91.5 90.2 38.3 54.7 55.1 91.3 89.3 85.2 36.3 27.5 35.0 35.2 56.3 91.7 98.0 82.3 47.4 50.8 52.8 115 82.8 99.5 85.3 62.1 89.0 102 75.0 100 1076 99.5 35.4  P (mg/kg) 2548 1965 847 553 876 2890 2899 2754 738 756 881 1131 1149 1267 752 1400 1338 2507 2743 3844 4708 4913 1625 1697 2032 2636 3305 3078 881 1984 633  Pb (mg/kg) 115.1 21.6 10.4 17.4 55.8 121.0 29.8 37.2 50.9 11.4 15.0 111.4 37.6 70.5 10.8 29.4 98.6 126.5 43.1 154.3 22.1 24.3 99.2 24.6 43.4 110.0 25.8 28.7 14.6 13.1 12.8  Zn (mg/kg) 215.4 204.2 71.4 113.9 95.7 370.7 232.7 271.1 82.0 79.2 119.2 189.2 326.2 198.5 118.3 191.5 244.8 270.4 282.5 121.7 112.2 121.8 181.0 144.4 290.5 210.5 156.1 234.4 63.7 168.7 87.8  105  Table C3. Summary of suspended sediment results.  Site #  Site Description  Date  2  Marshall Creek mainstem mouth  8  Urban tributary DeLair Park  22  Agricultural tributary Kenny Rd.  07/31/09 09/10/09 01/07/10 04/15/10 05/12/10 07/13/09 09/10/09 10/28/09 12/11/09 01/07/10 02/10/10 04/15/10 05/12/10 07/26/10 07/16/09 09/10/09 10/28/09 01/07/10 02/10/10 04/15/10 05/12/10 07/26/10  Days of Collection 57 41 27 64 27 39 59 48 44 27 34 64 27 75 42 56 48 27 34 64 27 75  Dry mass (mg) 400 660 1190 20830 1770 6540 1660 2180 1750 9810 4420 5380 1210 4710 15310 4380 1120 13930 11300 9400 51050 38880  Al (mg/kg) 19000 14111 15585 22000 17398 18201 20680 23115 11500 12788 36290 16945 14672 92600 19446 21119 22939 16174 25327 26235 23167 7892  Ca (mg/kg) 15008 16234 7265 15120 12554 13441 32420 17307 3714 7896 21834 8458 9406 42150 11718 26483 29866 5560 11416 16662 7027 4669  Cd (mg/kg) 2.76 1.54 1.79 0.00 0.00 1.20 2.80 4.20 0.00 1.77 0.00 0.00 0.00 11.1 1.43 1.68 1.43 1.32 0.00 0.00 0.00 5.03  Cr (mg/kg) 59.6 42.9 49.2 16.0 17.0 65.9 240 116 46.9 87.4 53.3 23.0 32.0 385 57.5 52.2 55.8 57.5 15.6 16.7 33.0 269  Cu (mg/kg) 203 116 99.8 45.5 85.8 74.3 142 140 84.5 90.4 93.8 54.9 118 508 64.4 80.0 108 67.8 51.8 67.7 79.1 82.9  Fe (mg/kg) 65779 32075 48717 67360 51574 25299 34480 44652 17973 23069 70226 36882 29745 195030 41420 45140 23459 25519 52327 54725 58594 53210  106  Table C3. Continued  Site #  Site Description  Date  2  Marshall Creek mainstem mouth  8  Urban tributary DeLair Park  22  Agricultural tributary Kenny Rd.  07/31/09 09/10/09 01/07/10 04/15/10 05/12/10 07/13/09 09/10/09 10/28/09 12/11/09 01/07/10 02/10/10 04/15/10 05/12/10 07/26/10 07/16/09 09/10/09 10/28/09 01/07/10 02/10/10 04/15/10 05/12/10 07/26/10  Days of K Mg Mn Ni P Pb Zn Collection (mg/kg) (mg/kg) (mg/kg) (mg/kg) (mg/kg) (mg/kg) (mg/kg) 57 1204 9975 6271 122.9 2314.1 45.23 363.1 41 9617 893 60.9 1350.6 18.52 202.6 1467 27 822 7872 1640 92.2 2308.5 31.32 214.8 64 9412 2001 106.1 3184.9 142.53 250.6 955 27 2300 10683 1447 82.1 2283.9 117.12 212.9 39 6839 782 35.3 556.8 16.33 161.3 1460 59 1824 12916 997 102.3 1005.3 19.40 208.3 48 9092 1541 47.1 1394.8 22.75 213.8 1772 44 688 4791 710 21.2 507.2 5.89 79.1 27 691 5390 711 35.5 698.4 18.24 124.9 34 1686 15216 1434 58.3 1411.1 139.93 234.1 64 730 6920 1293 26.7 651.6 71.42 165.6 27 841 6921 1215 26.5 833.6 60.66 173.5 75 4768 33700 11431 203.0 5534.0 364.32 1073.6 42 1501 11416 537 83.4 2123.2 9.64 175.1 56 4890 19328 662 81.3 2812.5 11.19 174.6 48 8514 19229 501 73.6 2034.1 12.19 169.9 27 800 7196 313 74.5 1601.8 16.98 108.9 34 1490 12684 533 101.1 3461.1 84.15 192.5 64 1613 12775 750 112.5 3649.7 98.54 222.6 27 1021 11690 438 97.6 4084.2 109.89 236.5 75 605 135120 1990 1305.3 893.6 68.83 69.8  107  Appendix D: Biofilm sampling and analysis results Figure D1. Photograph of tile apparatus used for biofilm colonization. Table D1. Summary of biofilm results.  108  Figure D1. Photograph of tile apparatus used for biofilm colonization.  109  Table D1. Summary of biofilm results. Site Site Sample Number Description Date ID 08/18/09 102 08/18/09 102 08/26/09 202 08/26/09 202 09/03/09 302 09/03/09 302 09/10/09 402 09/10/09 402 12/21/09 502 12/21/09 502 12/21/09 502 12/29/09 602 12/29/09 602 12/29/09 602 01/07/10 702 01/07/10 702 01/07/10 702 01/18/10 802 01/18/10 802 01/18/10 802 04/26/10 902 04/26/10 902 04/26/10 902 05/04/10 1002 05/04/10 1002 05/04/10 1002 05/20/10 1102 05/20/10 1102 05/20/10 1102 Marshall 05/31/10 1202 Creek mainstem 05/31/10 1202 2 mouth 05/31/10 1202  Tile Replicate 1 2 1 2 1 2 1 2 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3  Chl-a (mg/L) 4.81 3.00 19.86 14.92 11.80 14.97 2.07 27.50 0.02 0.08 0.02 0.33 0.08 0.24 0.24 0.18 0.05 0.18 0.58 0.18 2.38 7.79 6.62 2.12 6.81 4.58 4.40 4.29 4.72 7.39 8.20 7.08  Dry mass (mg) 44 16 101 86 51 112 143 61 78 29 16 108 628 901 1542 996 177 2075 1936 957 196 297 422 183 373 552 204 135 171 490 609 521  Cr (mg/kg) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 66.9 0.0 22.9 17.2 26.1 13.7 3.3 0.0 11.6 14.0 0.0 0.0 0.0 0.0 0.0 0.0 11.0 240.0 235.2 72.6 105.6 23.7 31.6  Cu (mg/kg) 0.0 1.5 34.9 65.3 16.9 98.2 16.4 11.4 17.4 50.0 42.7 37.6 43.6 88.7 24.6 7.8 17.8 18.1 11.9 1.3 22.5 31.0 31.6 14.4 28.2 28.4 33.3 39.4 28.0 38.1 35.2 37.9  Fe (mg/kg) 9241 3060 16404 17752 7492 10896 12902 8330 19838 7750 13230 17489 27149 88847 18049 7698 14912 21042 10328 2882 33969 37293 37935 31217 31200 32955 40485 30969 32778 40475 40938 40119  Mg (mg/kg) 368 81 2073 2119 898 1752 2018 976 2713 1344 2844 3533 5716 26459 5138 4185 3016 5119 4738 448 4563 5622 5876 4287 5304 5630 4818 4692 4420 6218 6637 6290  Mn (mg/kg) 846 448 1467 1340 916 899 1000 929 773 441 526 719 590 280 302 134 275 364 267 87 1103 668 890 870 1182 993 1354 1202 1159 2023 1852 1943  Ni (mg/kg) 0.0 5.3 46.2 55.6 25.1 16.2 33.0 28.2 28.3 63.4 26.7 43.3 77.5 167.7 43.3 20.9 11.0 38.3 28.0 7.1 40.2 51.2 52.1 37.0 49.3 52.6 183.7 185.7 92.9 125.4 73.4 79.9  P (mg/kg) 1013 531 2269 2382 1274 1061 1710 1251 784 1799 1456 1190 1148 3971 716 670 1021 968 857 118 1093 1204 1385 948 1391 1389 1383 832 1137 1778 1743 1667  110  Site Number  8  Site Description  Urban tributary DeLair Park  Date 07/13/09 07/13/09 08/18/09 08/18/09 08/26/09 08/26/09 09/03/09 09/03/09 09/10/09 09/10/09 12/21/09 12/21/09 12/21/09 12/29/09 12/29/09 12/29/09 01/07/10 01/07/10 01/07/10 01/18/10 01/18/10 01/18/10 04/26/10 04/26/10 04/26/10 05/04/10 05/04/10 05/04/10 05/12/10 05/12/10 05/12/10 05/20/10 05/20/10 05/20/10  Sample ID 8 8 108 108 208 208 308 308 408 408 508 508 508 608 608 608 708 708 708 808 808 808 908 908 908 1008 1008 1008 1108 1108 1108 1208 1208 1208  Tile Replicate 1 2 1 2 1 2 1 2 1 2 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3  Chl-a (mg/L) 2.48 2.21 0.43 0.56 11.11 0.70 5.07 0.18 1.04 2.13 0.23 0.19 0.24 1.18 2.25 1.84 4.14 3.86 4.02 6.76 11.54 4.63 4.06 3.72 7.57 6.94 7.29 6.42 4.63 7.34 6.46 21.54 4.51 5.68  Dry mass (mg) 415 685 56 30 430 194 438 639 630 452 1165 1085 970 6476 8970 2895 3273 1607 1378 11791 5362 1559 423 961 1003 1425 2304 3909 591 1615 305 1346 2314 1153  Cr (mg/kg) 55.9 39.6 0.0 0.0 0.0 0.0 0.0 0.0 5.0 0.0 34.3 41.5 25.8 9.9 14.3 13.1 5.1 19.8 19.3 1.6 13.3 25.9 0.0 15.9 13.6 56.4 29.1 31.0 264.8 103.7 19.0 35.7 34.9 45.9  Cu (mg/kg) 20.1 28.6 35.2 37.4 24.9 34.7 18.3 13.1 26.8 27.4 41.7 33.9 40.3 17.9 11.5 17.1 7.5 13.4 19.9 9.3 14.3 23.4 23.3 29.1 25.9 25.3 30.5 24.8 25.7 32.1 21.6 23.2 24.1 22.9  Fe (mg/kg) 22980 17715 12561 11073 13480 14172 14421 10440 17104 17784 21845 22597 21349 12778 10273 14522 7143 10508 9734 6824 15107 20240 18119 17877 17689 17020 19657 21459 22570 24691 19513 26852 25566 25216  Mg (mg/kg) 4484 4331 1791 1071 3096 2781 3291 3578 3330 3143 4229 4513 4514 3732 3456 3618 2962 3278 3876 3541 3661 3841 3035 3884 3713 3713 3959 3400 3280 4032 2912 3336 3483 3323  Mn (mg/kg) 334 331 1313 1604 651 949 472 399 585 632 386 378 394 183 216 184 136 245 270 152 254 329 346 355 332 362 492 427 418 501 384 411 363 343  Ni (mg/kg) 27.1 25.1 38.3 63.6 23.2 37.9 12.1 13.6 16.8 20.3 29.1 41.2 26.5 12.9 11.8 13.6 4.5 12.7 28.0 5.3 10.2 19.1 13.4 17.5 18.0 39.6 20.6 17.1 146.5 58.0 20.8 16.3 19.7 26.4  P (mg/kg) 487 435 2089 2478 990 1484 624 569 618 649 710 648 733 387 471 465 488 561 601 484 528 583 269 470 460 449 455 439 519 520 379 509 432 625  111  Site Number  Site Description  22  Agricultural tributary Kenny Rd.  Date 07/13/09 07/13/09 08/18/09 08/18/09 08/26/09 08/26/09 09/03/09 09/03/09 09/10/09 09/10/09 12/21/09 12/21/09 12/21/09 12/29/09 12/29/09 12/29/09 01/07/10 01/07/10 01/07/10 01/18/10 01/18/10 01/18/10 04/26/10 04/26/10 04/26/10 05/04/10 05/04/10 05/04/10 05/12/10 05/12/10 05/12/10 05/20/10 05/20/10 05/20/10  Sample ID 22 22 122 122 222 222 322 322 422 422 522 522 522 622 622 622 722 722 722 822 822 822 922 922 922 1022 1022 1022 1122 1122 1122 1222 1222 1222  Tile Replicate 1 2 1 2 1 2 1 2 1 2 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3  Chl-a (mg/L) 28.71 24.74 14.79 10.34 59.80 45.04 41.46 1.18 21.86 15.23 0.15 0.09 0.14 0.08 0.29 0.09 0.15 0.20 0.17 0.56 0.44 0.34 0.62 9.07 2.38 1.25 0.59 1.13 3.22 3.24 2.42 13.30 5.51 6.14  Dry mass (mg) 380 220 56 58 93 95 110 115 126 91 512 1088 737 1398 1367 703 2355 3337 3399 3299 7794 3556 100 214 114 177 263 128 266 348 243 230 225 360  Cr (mg/kg) 64.9 66.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 14.9 16.8 31.2 23.9 12.1 6.3 3.2 19.5 7.0 14.9 8.3 0.0 17.2 0.0 0.0 0.0 13.0 73.6 0.0 13.2 98.2 179.5 13.4 398.3 69.9  Cu (mg/kg) 26.8 25.5 28.3 30.4 4.7 50.3 6.9 21.5 5.1 3.7 31.3 36.7 32.9 17.1 9.9 13.4 24.0 12.0 21.3 16.4 22.6 17.4 7.4 15.3 8.3 18.2 28.8 16.8 0.5 29.7 37.1 25.7 24.3 33.4  Fe (mg/kg) 41684 38132 13239 14231 23002 13248 17642 20975 18304 13176 27625 29373 29741 18801 9093 13844 23419 10380 17074 13069 29224 11298 25596 32424 28698 32908 41566 28910 8686 44554 42778 44514 41943 43333  Mg (mg/kg) 11029 10509 1299 1450 927 189 1205 1261 1424 1427 6677 7952 7542 6007 5296 4591 7475 6604 6791 6359 8102 6575 3377 4448 3307 4010 5289 3551 743 5843 4897 4680 4766 6169  Mn (mg/kg) 408 401 1217 1255 489 415 421 654 370 416 300 283 286 168 127 106 208 137 155 131 298 163 338 303 419 1038 959 862 164 1524 904 657 385 395  Ni (mg/kg) 95.5 92.5 41.2 40.6 1.7 9.3 14.9 23.3 19.3 43.6 61.2 74.5 65.8 40.4 21.8 24.8 56.1 38.5 44.5 35.2 64.2 45.0 16.3 31.4 13.5 50.0 89.4 35.9 24.6 105.7 143.4 47.6 260.3 90.7  P (mg/kg) 2698 2097 3266 3189 2857 2296 2575 3642 3147 2000 2421 2088 2141 1027 1370 820 1691 1634 1443 1870 1434 1813 793 1436 1052 1421 2166 1322 310 2257 2010 2400 2186 2433  112  Appendix E: Spearman Rank Correlations Table E1. Spearman rank correlation coefficients for all water quality data in the Marshall Creek watershed. Table E2. Spearman rank correlation coefficients for water quality data collected on the mainstem sites of Marshall Creek. Table E3. Spearman rank correlation coefficients for biofilm and water quality data collected at three sites in the Marshall Creek watershed. Table E4. Spearman rank correlation coefficients for suspended sediment data collected at three sites in the Marshall Creek watershed.  113  Table E1. Spearman rank correlation coefficients for all water quality data in the Marshall Creek watershed. NO3--N  SRP  Cl-  NH4+-N  pH  Sp. Cond  Temp.  DO  Turb.  TOC  Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N  NO3--N 1 . 124 0.134 0.137 124 -0.297 0.002 110 -0.132 0.168 110 -0.016 0.859 124 -0.178 0.047 124 -0.266 0.006 104 0.366 0.000 95 -0.134 0.281 67 -0.306 0.002 99  SRP 0.134 0.137 124 1 . 124 0.365 0.000 110 0.524 0.000 110 -0.402 0.000 124 0.261 0.003 124 -0.066 0.506 104 -0.494 0.000 95 0.502 0.000 67 0.484 0.000 99  Cl-0.297 0.002 110 0.365 0.000 110 1 . 110 0.493 0.000 110 -0.410 0.000 110 0.582 0.000 110 -0.146 0.166 91 -0.468 0.000 81 0.683 0.000 67 0.644 0.000 99  NH4+-N -0.132 0.168 110 0.524 0.000 110 0.493 0.000 110 1 . 110 -0.506 0.000 110 0.244 0.010 110 -0.229 0.029 91 -0.753 0.000 81 0.721 0.000 67 0.641 0.000 99  pH -0.016 0.859 124 -0.402 0.000 124 -0.410 0.000 110 -0.506 0.000 110 1 . 124 -0.143 0.113 124 0.174 0.078 104 0.482 0.000 95 -0.431 0.000 67 -0.607 0.000 99  Sp. Cond -0.178 0.047 124 0.261 0.003 124 0.582 0.000 110 0.244 0.010 110 -0.143 0.113 124 1 . 124 0.355 0.000 104 -0.345 0.001 95 0.339 0.005 67 0.337 0.001 99  Temp. -0.266 0.006 104 -0.066 0.506 104 -0.146 0.166 91 -0.229 0.029 91 0.174 0.078 104 0.355 0.000 104 1 . 104 -0.030 0.780 90 -0.236 0.062 63 0.064 0.571 80  DO 0.366 0.000 95 -0.494 0.000 95 -0.468 0.000 81 -0.753 0.000 81 0.482 0.000 95 -0.345 0.001 95 -0.030 0.780 90 1 . 95 -0.581 0.000 67 -0.775 0.000 70  Turb. -0.134 0.281 67 0.502 0.000 67 0.683 0.000 67 0.721 0.000 67 -0.431 0.000 67 0.339 0.005 67 -0.236 0.062 63 -0.581 0.000 67 1 . 67 0.704 0.000 56  TOC -0.306 0.002 99 0.484 0.000 99 0.644 0.000 99 0.641 0.000 99 -0.607 0.000 99 0.337 0.001 99 0.064 0.571 80 -0.775 0.000 70 0.704 0.000 56 1 . 99  114  Table E2. Spearman rank correlation coefficients for water quality data collected on the mainstem sites of Marshall Creek. NO3--N  SRP  Cl-  NH4+-N  pH  Sp. Cond  Temp.  DO  Turb.  TOC  Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N  NO3--N 1 . 69 -0.282 0.019 69 -0.337 0.008 61 -0.460 0.000 61 0.151 0.216 69 -0.057 0.641 69 -0.146 0.273 58 0.678 0.000 54 -0.372 0.020 39 -0.574 0.000 56  SRP -0.282 0.019 69 1 . 69 0.360 0.004 61 0.513 0.000 61 -0.286 0.017 69 0.184 0.131 69 -0.167 0.210 58 -0.564 0.000 54 0.368 0.021 39 0.567 0.000 56  Cl-0.337 0.008 61 0.360 0.004 61 1 . 61 0.429 0.001 61 -0.373 0.003 61 0.370 0.003 61 -0.432 0.002 51 -0.322 0.029 46 0.575 0.000 39 0.543 0.000 56  NH4+-N -0.460 0.000 61 0.513 0.000 61 0.429 0.001 61 1 . 61 -0.268 0.037 61 0.006 0.966 61 -0.382 0.006 51 -0.549 0.000 46 0.466 0.003 39 0.566 0.000 56  pH 0.151 0.216 69 -0.286 0.017 69 -0.373 0.003 61 -0.268 0.037 61 1 . 69 -0.066 0.589 69 0.277 0.035 58 0.312 0.022 54 -0.300 0.063 39 -0.561 0.000 56  Sp. Cond -0.057 0.641 69 0.184 0.131 69 0.370 0.003 61 0.006 0.966 61 -0.066 0.589 69 1 . 69 0.450 0.000 58 -0.250 0.069 54 0.197 0.230 39 0.147 0.279 56  Temp. -0.146 0.273 58 -0.167 0.210 58 -0.432 0.002 51 -0.382 0.006 51 0.277 0.035 58 0.450 0.000 58 1 . 58 0.001 0.997 51 -0.361 0.028 37 -0.087 0.564 46  DO 0.678 0.000 54 -0.564 0.000 54 -0.322 0.029 46 -0.549 0.000 46 0.312 0.022 54 -0.250 0.069 54 0.001 0.997 51 1 . 54 -0.265 0.103 39 -0.668 0.000 41  Turb. -0.372 0.020 39 0.368 0.021 39 0.575 0.000 39 0.466 0.003 39 -0.300 0.063 39 0.197 0.230 39 -0.361 0.028 37 -0.265 0.103 39 1 . 39 0.508 0.002 34  TOC -0.574 0.000 56 0.567 0.000 56 0.543 0.000 56 0.566 0.000 56 -0.561 0.000 56 0.147 0.279 56 -0.087 0.564 46 -0.668 0.000 41 0.508 0.002 34 1 . 56  115  Table E3. Spearman rank correlation coefficients for biofilm and water quality data collected at three sites in the Marshall Creek watershed. Biofilms  Chl-a  Dry mass Cr  Cu  Fe  Mg  Mn  Ni  P  NO3--N  Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N  Water  Chla  Dry mass  Cr  Cu  Fe  Mg  Mn  Ni  P  NO3-N  SRP  Cl-  1  -0.356  -0.163  0.040  0.136  -0.421  0.373  -0.080  0.264  -0.485  -0.449  -0.436  . 38 -0.356  0.028 38 1  0.328 38 0.408  0.810 38 -0.246  0.416 38 0.022  0.008 38 0.518  0.021 38 -0.743  0.631 38 -0.163  0.110 38 -0.573  0.019 23 0.221  0.032 23 -0.238  0.028 38 -0.163  . 38 0.408  0.011 38 1  0.137 38 0.171  0.895 38 0.553  0.001 38 0.574  0.000 38 -0.201  0.328 38 0.507  0.000 38 -0.178  0.310 23 -0.319  0.328 38 0.040  0.011 38 -0.246  . 38 0.171  0.305 38 1  0.000 38 0.371  0.000 38 0.137  0.225 38 0.512  0.001 38 0.513  0.286 38 0.258  0.810 38 0.136  0.137 38 0.022  0.305 38 0.553  . 38 0.371  0.022 38 1  0.413 38 0.605  0.001 38 0.212  0.001 38 0.616  0.416 38 -0.421  0.895 38 0.518  0.000 38 0.574  0.022 38 0.137  . 38 0.605  0.000 38 1  0.201 38 -0.297  0.008 38 0.373  0.001 38 -0.743  0.000 38 -0.201  0.413 38 0.512  0.000 38 0.212  . 38 -0.297  0.021 38 -0.080  0.000 38 -0.163  0.225 38 0.507  0.001 38 0.513  0.201 38 0.616  0.631 38 0.264  0.328 38 -0.573  0.001 38 -0.178  0.001 38 0.258  0.110 38 -0.485  0.000 38 0.221  0.286 38 -0.319  0.019 23  0.310 23  0.138 23  NH4+N  pH  Sp. Cond  Temp.  DO  Turb.  TOC  Depth  -0.554  0.741  -0.056  0.824  0.187  -0.571  -0.379  .449*  0.038 23 -0.339  0.006 23 -0.393  0.000 23 0.153  0.799 23 -0.246  0.000 23 -0.372  0.393 23 0.328  0.006 22 -0.359  0.164 15 -0.352  0.017 28 -0.221  0.274 23 -0.230  0.114 23 -0.059  0.063 23 -0.175  0.485 23 0.227  0.258 23 0.164  0.080 23 0.399  0.126 23 -0.053  0.101 22 -0.095  0.198 15 -0.043  0.258 28 -0.062  0.138 23 0.136  0.292 23 -0.051  0.790 23 -0.208  0.423 23 -0.025  0.297 23 0.209  0.454 23 -0.199  0.059 23 0.219  0.809 23 -0.076  0.673 22 -0.065  0.879 15 0.023  0.753 28 0.108  0.118 38 0.214  0.535 23 -0.265  0.816 23 0.039  0.342 23 0.152  0.911 23 0.075  0.338 23 0.092  0.364 23 0.335  0.316 23 0.537  0.730 23 -0.322  0.774 22 0.269  0.934 15 0.102  0.584 28 0.047  0.000 38 0.552  0.196 38 -0.012  0.222 23 0.260  0.861 23 0.371  0.488 23 0.511  0.733 23 0.362  0.675 23 -0.457  0.118 23 0.413  0.008 23 -0.207  0.134 23 -0.583  0.225 22 0.612  0.718 15 0.502  0.812 28 -0.196  0.070 38 1  0.000 38 0.390  0.942 38 0.409  0.231 23 -0.211  0.082 23 0.007  0.013 23 -0.126  0.090 23 -0.015  0.028 23 0.278  0.050 23 -0.087  0.344 23 0.624  0.003 23 -0.130  0.002 22 -0.011  0.056 15 -0.034  0.318 28 0.412  0.070 38 0.552  . 38 0.390  0.015 38 1  0.011 38 0.495  0.333 23 0.045  0.975 23 0.252  0.565 23 0.460  0.946 23 0.369  0.199 23 -0.237  0.693 23 0.426  0.001 23 0.223  0.556 23 -0.625  0.962 22 0.522  0.904 15 0.377  0.029 28 0.094  0.000 38 0.214  0.000 38 -0.012  0.015 38 0.409  . 38 0.495  0.002 38 1  0.837 23 0.215  0.246 23 0.640  0.027 23 0.752  0.084 23 0.658  0.277 23 -0.638  0.043 23 0.493  0.307 23 -0.073  0.001 23 -0.790  0.013 22 0.730  0.166 15 0.817  0.636 28 0.226  0.118 38 0.136  0.196 38 -0.265  0.942 38 0.260  0.011 38 -0.211  0.002 38 0.045  . 38 0.215  0.324 23 1  0.001 23 0.551  0.000 23 0.002  0.001 23 0.311  0.001 23 -0.302  0.017 23 -0.260  0.742 23 -0.568  0.000 23 -0.207  0.000 22 0.003  0.000 15 0.147  0.247 28 -0.051  0.535 23  0.222 23  0.231 23  0.333 23  0.837 23  0.324 23  . 23  0.006 23  0.993 23  0.148 23  0.161 23  0.231 23  0.005 23  0.343 23  0.990 22  0.615 14  0.844 17  116  Biofilms  SRP  Cl-  NH4+N pH  Sp. Cond Temp.  DO  Turb.  TOC  Water depth  Water  Chla  Dry mass  Cr  Cu  Fe  Mg  Mn  Ni  P  NO3-N  SRP  NH4+N  pH  Sp. Cond  Corr. Coef. Sig. N  -0.449  -0.238  -0.230  -0.051  0.039  0.371  0.007  0.252  0.640  0.551  1  0.481  Temp.  DO  Turb.  TOC  Depth  0.734  -0.767  0.214  -0.331  -0.610  0.529  0.662  -0.414  0.032 23  0.274 23  0.292 23  0.816 23  0.861 23  0.082 23  0.975 23  0.246 23  0.001 23  0.006 23  . 23  0.020 23  0.000 23  0.000 23  0.326 23  0.122 23  0.002 23  0.011 22  0.010 14  0.098 17  Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N Corr. Coef. Sig. N  -0.436  -0.339  -0.059  -0.208  0.152  0.511  -0.126  0.460  0.752  0.002  0.481  1  0.711  -0.751  0.743  -0.210  -0.704  0.810  0.833  -0.559  0.038 23 -0.554  0.114 23 -0.393  0.790 23 -0.175  0.342 23 -0.025  0.488 23 0.075  0.013 23 0.362  0.565 23 -0.015  0.027 23 0.369  0.000 23 0.658  0.993 23 0.311  0.020 23 0.734  . 23 0.711  0.000 23 1  0.000 23 -0.796  0.000 23 0.365  0.337 23 -0.342  0.000 23 -0.654  0.000 22 0.656  0.000 14 0.662  0.020 17 -0.473  0.006 23 0.741  0.063 23 0.153  0.423 23 0.227  0.911 23 0.209  0.733 23 0.092  0.090 23 -0.457  0.946 23 0.278  0.084 23 -0.237  0.001 23 -0.638  0.148 23 -0.302  0.000 23 -0.767  0.000 23 -0.751  . 23 -0.796  0.000 23 1  0.087 23 -0.381  0.110 23 0.567  0.001 23 0.597  0.001 22 -0.799  0.010 14 -0.829  0.055 17 0.456  0.000 23 -0.056  0.485 23 -0.246  0.297 23 0.164  0.338 23 -0.199  0.675 23 0.335  0.028 23 0.413  0.199 23 -0.087  0.277 23 0.426  0.001 23 0.493  0.161 23 -0.260  0.000 23 0.214  0.000 23 0.743  0.000 23 0.365  . 23 -0.381  0.073 23 1  0.005 23 0.134  0.003 23 -0.554  0.000 22 0.622  0.000 14 0.618  0.066 17 -0.591  0.799 23 0.824  0.258 23 -0.372  0.454 23 0.399  0.364 23 0.219  0.118 23 0.537  0.050 23 -0.207  0.693 23 0.624  0.043 23 0.223  0.017 23 -0.073  0.231 23 -0.568  0.326 23 -0.331  0.000 23 -0.210  0.087 23 -0.342  0.073 23 0.567  . 23 0.134  0.542 23 1  0.006 23 -0.055  0.002 22 -0.280  0.019 14 -0.189  0.013 17 0.227  0.000 23 0.187  0.080 23 0.328  0.059 23 -0.053  0.316 23 -0.076  0.008 23 -0.322  0.344 23 -0.583  0.001 23 -0.130  0.307 23 -0.625  0.742 23 -0.790  0.005 23 -0.207  0.122 23 -0.610  0.337 23 -0.704  0.110 23 -0.654  0.005 23 0.597  0.542 23 -0.554  . 23 -0.055  0.804 23 1  0.207 22 -0.653  0.517 14 -0.849  0.381 17 0.270  0.393 23 -0.571  0.126 23 -0.359  0.809 23 -0.095  0.730 23 -0.065  0.134 23 0.269  0.003 23 0.612  0.556 23 -0.011  0.001 23 0.522  0.000 23 0.730  0.343 23 0.003  0.002 23 0.529  0.000 23 0.810  0.001 23 0.656  0.003 23 -0.799  0.006 23 0.622  0.804 23 -0.280  . 23 -0.653  0.001 22 1  0.000 14 0.852  0.295 17 -0.449  0.006 22 -0.379  0.101 22 -0.352  0.673 22 -0.043  0.774 22 0.023  0.225 22 0.102  0.002 22 0.502  0.962 22 -0.034  0.013 22 0.377  0.000 22 0.817  0.990 22 0.147  0.011 22 0.662  0.000 22 0.833  0.001 22 0.662  0.000 22 -0.829  0.002 22 0.618  0.207 22 -0.189  0.001 22 -0.849  . 22 0.852  0.000 13 1  0.071 17 -0.342  0.164 15 0.449  0.198 15 -0.221  0.879 15 -0.062  0.934 15 0.108  0.718 15 0.047  0.056 15 -0.196  0.904 15 0.412  0.166 15 0.094  0.000 15 0.226  0.615 14 -0.051  0.010 14 -0.414  0.000 14 -0.559  0.010 14 -0.473  0.000 14 0.456  0.019 14 -0.591  0.517 14 0.227  0.000 14 0.270  0.000 13 -0.449  . 15 -0.342  0.303 11 1  0.017 28  0.258 28  0.753 28  0.584 28  0.812 28  0.318 28  0.029 28  0.636 28  0.247 28  0.844 17  0.098 17  0.020 17  0.055 17  0.066 17  0.013 17  0.381 17  0.295 17  0.071 17  0.303 11  . 28  Cl-  117  Table E4. Spearman rank correlation coefficients for suspended sediment data collected at three sites in the Marshall Creek watershed. Dry weight Al  Ca  Cd  Co  Cr  Cu  Fe  K  Mg  Mn  Na  Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N  Dry weight 1 . 22 0.163 0.468 22 -0.360 0.100 22 -0.229 0.306 22 -0.235 0.293 22 -0.058 0.797 22 -0.723 0.000 22 0.251 0.259 22 -0.296 0.180 22 0.095 0.673 22 -0.249 0.264 22 -0.418 0.229 10  Al 0.163 0.468 22 1 . 22 0.660 0.001 22 -0.105 0.643 22 -0.185 0.411 22 -0.089 0.695 22 -0.031 0.891 22 0.602 0.003 22 0.704 0.000 22 0.553 0.008 22 -0.015 0.946 22 0.818 0.004 10  Ca -0.360 0.100 22 0.660 0.001 22 1 . 22 0.279 0.209 22 0.180 0.423 22 0.198 0.377 22 0.401 0.064 22 0.276 0.214 22 0.871 0.000 22 0.532 0.011 22 0.225 0.313 22 0.794 0.006 10  Cd -0.229 0.306 22 -0.105 0.643 22 0.279 0.209 22 1 . 22 0.922 0.000 22 0.838 0.000 22 0.609 0.003 22 0.017 0.940 22 0.153 0.498 22 0.299 0.177 22 0.360 0.100 22 -0.112 0.757 10  Co -0.235 0.293 22 -0.185 0.411 22 0.180 0.423 22 0.922 0.000 22 1 . 22 0.765 0.000 22 0.506 0.016 22 0.047 0.835 22 0.102 0.653 22 0.263 0.237 22 0.375 0.085 22 -0.407 0.243 10  Cr -0.058 0.797 22 -0.089 0.695 22 0.198 0.377 22 0.838 0.000 22 0.765 0.000 22 1 . 22 0.547 0.008 22 -0.119 0.597 22 0.083 0.713 22 0.189 0.399 22 0.213 0.342 22 -0.079 0.829 10  Cu -0.723 0.000 22 -0.031 0.891 22 0.401 0.064 22 0.609 0.003 22 0.506 0.016 22 0.547 0.008 22 1 . 22 -0.019 0.934 22 0.298 0.179 22 0.146 0.516 22 0.420 0.052 22 0.309 0.385 10  Fe 0.251 0.259 22 0.602 0.003 22 0.276 0.214 22 0.017 0.940 22 0.047 0.835 22 -0.119 0.597 22 -0.019 0.934 22 1 . 22 0.244 0.273 22 0.580 0.005 22 0.522 0.013 22 0.370 0.293 10  K -0.296 0.180 22 0.704 0.000 22 0.871 0.000 22 0.153 0.498 22 0.102 0.653 22 0.083 0.713 22 0.298 0.179 22 0.244 0.273 22 1 . 22 0.607 0.003 22 -0.028 0.903 22 0.818 0.004 10  Mg 0.095 0.673 22 0.553 0.008 22 0.532 0.011 22 0.299 0.177 22 0.263 0.237 22 0.189 0.399 22 0.146 0.516 22 0.580 0.005 22 0.607 0.003 22 1 . 22 0.083 0.713 22 0.345 0.328 10  Mn -0.249 0.264 22 -0.015 0.946 22 0.225 0.313 22 0.360 0.100 22 0.375 0.085 22 0.213 0.342 22 0.420 0.052 22 0.522 0.013 22 -0.028 0.903 22 0.083 0.713 22 1 . 22 -0.164 0.651 10  Na -0.418 0.229 10 0.818 0.004 10 0.794 0.006 10 -0.112 0.757 10 -0.407 0.243 10 -0.079 0.829 10 0.309 0.385 10 0.370 0.293 10 0.818 0.004 10 0.345 0.328 10 -0.164 0.651 10 1 . 10  Ni 0.258 0.246 22 0.370 0.090 22 0.213 0.342 22 0.351 0.109 22 0.408 0.059 22 0.119 0.597 22 0.004 0.986 22 0.720 0.000 22 0.204 0.363 22 0.679 0.001 22 0.318 0.149 22 0.030 0.934 10  P 0.199 0.374 22 0.693 0.000 22 0.324 0.142 22 -0.005 0.981 22 0.023 0.918 22 -0.258 0.246 22 -0.098 0.665 22 0.732 0.000 22 0.461 0.031 22 0.565 0.006 22 0.036 0.875 22 0.503 0.138 10  Pb 0.244 0.273 22 0.462 0.030 22 0.147 0.513 22 -0.230 0.304 22 -0.305 0.168 22 -0.308 0.164 22 -0.024 0.915 22 0.805 0.000 22 0.080 0.725 22 0.312 0.157 22 0.532 0.011 22 0.406 0.244 10  Zn -0.126 0.577 22 0.680 0.000 22 0.513 0.015 22 0.033 0.883 22 0.046 0.839 22 -0.142 0.529 22 0.266 0.232 22 0.787 0.000 22 0.455 0.034 22 0.343 0.118 22 0.459 0.032 22 0.455 0.187 10  Cum. Precip. 0.144 0.524 22 0.167 0.457 22 0.002 0.994 22 -0.132 0.558 22 -0.234 0.295 22 -0.170 0.451 22 -0.183 0.414 22 0.085 0.708 22 -0.132 0.558 22 0.043 0.849 22 0.130 0.563 22 -0.088 0.810 10  118  Ni  P  Pb  Zn  Cum. Precip.  Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N Corr. Coef. Sig. (2-tailed) N  Dry weight 0.258 0.246 22 0.199 0.374 22 0.244 0.273 22 -0.126 0.577 22 0.144 0.524 22  Al 0.370 0.090 22 0.693 0.000 22 0.462 0.030 22 0.680 0.000 22 0.167 0.457 22  Ca 0.213 0.342 22 0.324 0.142 22 0.147 0.513 22 0.513 0.015 22 0.002 0.994 22  Cd 0.351 0.109 22 -0.005 0.981 22 -0.230 0.304 22 0.033 0.883 22 -0.132 0.558 22  Co 0.408 0.059 22 0.023 0.918 22 -0.305 0.168 22 0.046 0.839 22 -0.234 0.295 22  Cr 0.119 0.597 22 -0.258 0.246 22 -0.308 0.164 22 -0.142 0.529 22 -0.170 0.451 22  Cu 0.004 0.986 22 -0.098 0.665 22 -0.024 0.915 22 0.266 0.232 22 -0.183 0.414 22  Fe 0.720 0.000 22 0.732 0.000 22 0.805 0.000 22 0.787 0.000 22 0.085 0.708 22  K 0.204 0.363 22 0.461 0.031 22 0.080 0.725 22 0.455 0.034 22 -0.132 0.558 22  Mg 0.679 0.001 22 0.565 0.006 22 0.312 0.157 22 0.343 0.118 22 0.043 0.849 22  Mn 0.318 0.149 22 0.036 0.875 22 0.532 0.011 22 0.459 0.032 22 0.130 0.563 22  Na 0.030 0.934 10 0.503 0.138 10 0.406 0.244 10 0.455 0.187 10 -0.088 0.810 10  Ni 1 . 22 0.717 0.000 22 0.461 0.031 22 0.521 0.013 22 0.004 0.986 22  P 0.717 0.000 22 1 . 22 0.495 0.019 22 0.732 0.000 22 -0.003 0.988 22  Pb 0.461 0.031 22 0.495 0.019 22 1 . 22 0.649 0.001 22 0.237 0.288 22  Zn 0.521 0.013 22 0.732 0.000 22 0.649 0.001 22 1 . 22 -0.058 0.798 22  Cum. Precip. 0.004 0.986 22 -0.003 0.988 22 0.237 0.288 22 -0.058 0.798 22 1 . 22  119  Appendix F: Photographs of field sites Figure F1. Discharge near Fraser Valley Trout Hatchery into Marshall Creek (site 14). Figure F2. Site 11 where water and bed sediments were sampled between 1993-2010. Figure F3. Suspended sediment sampler, water level probe and tiles for biofilms at the agricultural tributary (site 22). Figure F4. Suspended sediment sampler, water level probe and tiles for biofilms at the urban tributary (site 8). Figure F5. Suspended sediment sampler, water level probe and tiles for biofilms at the mouth of Marshall mainstem (site 2). Figure F6. Agricultural tributary (site 22) showing signs of eutrophication during the summer of 2009. Figure F7. Mouth of Marshall Creek (site 2) showing signs of eutrophication in the summer of 2009. Figure F8. The mouth of Marshall Creek mainstem (site 2) looking downstream towards the confluence of the Sumas River in the winter of 2008. Figure F9. Large chicken operation near the agricultural tributary (site 22) in the fall of 2009. Figure F10. Liquid manure application in the Marshall Creek watershed in the spring of 2009. Figure F11. Dairy manure storage facility in the agricultural subwatershed. Figure F12. Forest clearing and urban residential development on Sumas Mountain in the Marshall Creek watershed.  120  Figure F1. Discharge near Fraser Valley Trout Hatchery into Marshall Creek (site 14).  121  Figure F2. Site 11 where water and bed sediments were sampled between 1993-2010.  122  Figure F3. Suspended sediment sampler, water level probe and tiles for biofilms at the agricultural tributary (site 22).  123  Figure F4. Suspended sediment sampler, water level probe and tiles for biofilms at the urban tributary (site 8).  Figure F5. Suspended sediment sampler, water level probe and tiles for biofilms at the mouth of Marshall mainstem (site 2).  124  Figure F6. Agricultural tributary (site 22) showing signs of eutrophication during the summer of 2009.  Figure F7. Mouth of Marshall Creek (site 2) showing signs of eutrophication in the summer of 2009.  125  Figure F8. The mouth of Marshall Creek mainstem (site 2) looking downstream towards the confluence of the Sumas River in the winter of 2008.  Figure F9. Large chicken operation near the agricultural tributary (site 22) in the fall of 2009.  126  Figure F10. Liquid manure application in the Marshall Creek watershed in the spring of 2009.  Figure F11. Dairy manure storage facility in the agricultural subwatershed.  127  Figure F12. Forest clearing and urban residential development on Sumas Mountain in the Marshall Creek watershed.  128  

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.24.1-0071509/manifest

Comment

Related Items