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  ii 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.  iii  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.  iv 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 Cumulative Effects ...........................................................................................................1 1.2 Watershed CEA Framework ..........................................................................................3 1.3 Study Area ........................................................................................................................4 1.4 Land use and stream quality ...........................................................................................5 1.5 Methods of Assessment ....................................................................................................7 2 Objectives .................................................................................................................. 10 3 Background ............................................................................................................... 11 3.1 Study Site ........................................................................................................................11 3.2 Sampling Site Selection ..................................................................................................14 4 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  v 5.2.1.5 Historic Nitrate-N ............................................................................................................ 41 5.2.1.5.1 Time-Series Analysis............................................................................................... 42 5.2.2 Bed Sediments – Historic comparison ......................................................................44 5.2.3 Suspended Sediments ................................................................................................47 5.2.3.1 Overall ............................................................................................................................. 47 5.2.3.2 Site Comparison .............................................................................................................. 48 5.2.4 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 Correlations ..................................................................................................................... 60 5.2.5.2 Trace elements in sediments and biofilms ....................................................................... 61 5.2.5.3 Biofilm : sediment ratios ................................................................................................. 63 6 Summary ................................................................................................................... 67 7 Conclusions................................................................................................................ 70 7.1 Land use trends ..............................................................................................................70 7.2 Seasonal and spatial trends in water quality ...............................................................71 7.3 Trends in bed sediment quality .....................................................................................72 7.4 Interactions between water, sediments and biofilms ..................................................73 7.5 Seasonal effects ...............................................................................................................73 7.6 Suspended sediments and biofilms as indicators of cumulative effects .....................74 7.7 Cumulative effects assessment ......................................................................................74 7.8 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   vi 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  vii 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   viii 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  ix 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 right- hand vertical axis. ............................................................................................. 62   x List of Abbreviations and Symbols  Al Aluminum ALR Agricultural Land Reserve ANOVA Analysis of Variance AUE Animal Unit Equivalent BC British Columbia Ca Calcium Cd Cadmium CEA Cumulative Effects Assessment Chl-a Chlorophyll a Cl- Chloride Cr Chromium CT Census Tract Cu Copper CWN Canadian Water Network DA Dissemination Area DO Dissolved Oxygen EIA Environmental Impact Assessment EU European Union Fe Iron GIS Geographic Information Systems GW Groundwater ha Hectare HCl Hydrochloric acid HNO3- Nitric acid ICP-AES Inductively Coupled Plasma - Atomic Emission Spectroscopy ISQG Interim Sediment Quality Guidelines K Potassium Ln Natural logarithm Log Logarithm M-W U Mann-Whitney U Test Mg Magnesium mg/kg Milligrams per kilogram mg/L Milligrams per liter mm Millimeter = 10-3 meters Mn Manganese N Nitrogen  xi n Sample size NH4+-N Ammonia-Nitrogen Ni Nickel NO3--N Nitrate-Nitrogen NTU Nephelometric Turbidity Units P Phosphorus p Probability level (statistics) Pb Lead Pct Percentile ppm Parts per million R Correlation coefficient (statistics) R2 Coefficient of determination (statistics) SE Standard Error Sp. Cond. Specific Conductivity SRP Soluble Reactive Phosphorus TIA Total Impervious Area TOC Total Organic Carbon US EPA United States Environmental Protection Agency USDA United States Department of Agriculture VEC Valued Ecosystem Component Z Standard score (statistics) Zn Zinc α Alpha (significance level) µg/L Micrograms per liter µm Micrometer = 10-6 meters µS/cm MicroSiemens per centimeter - units of conductivity ºC Degrees Celsius   xii 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.   1 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 small- scale 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  2 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.     3 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.   4 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  5 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  6 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. Nitrate- contaminated groundwater supplies are a problem in agricultural watersheds worldwide (McMahon et al., 2008; Almasri & Kaluarachchi, 2004; Wassenaar, 1995).    7 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 oxidation- reduction 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.   8 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.  9 In addition, historic water quality data aid in the determination of changes in the stream related to land use change over time.   10 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 (1994- 2010). 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.   11 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.  12  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 fish- bearing stream (Integrated Resource Consultants, 1994). Marshall Creek surveys have  13 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  14 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.  15  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 # grab samples (w - water, s - bed sediments) Biofilms/Suspen- ded Sediments Groundwater (Hatchery well) 12 Aug 08 – Jun 09 w – 2  No Tributary (GW influence) 13 Aug 08 – Dec 08 w – 2 s – 1 No Tributary (GW influence) 10 Aug 08 – Sep 09 w – 3 s – 2 No Marshall Creek (Hatchery discharge) 14 Aug 08 – May 10 w – 8 s – 2 No Marshall Creek 15 Aug 08 – May 10 w – 5 No Marshall Creek 11 Aug 08 – May 10 w – 12 s – 3 No  Marshall Creek 17 Aug 08 – May 10 w – 7 s – 3 No   16 Station Site ID Date range sampled # grab samples (w - water, s - bed sediments) Biofilms/Suspen- ded Sediments Marshall Creek 18 Aug 08 – May 10 w – 11 s – 3 No  Marshall Creek 7 Aug 08 – May 10 w – 7 s – 3 No Marshall Creek (Downstream) 2 Aug 08 – May 10 w – 17 s – 3 Yes Urban tributary 3 Aug 08 – Sep 09 w – 4 s – 2 No  Urban tributary 4 Aug 08 – Sep 09 w – 4 s – 1 No  Urban Detention pond 5 Aug 08 – Dec 08 w – 2 s – 1 No Urban Tributary (Junction Creek) 23 Jun 09 – Sep 09 w – 2 s – 1 No Urban tributary 8 Aug 08 – May 10 w – 18 s – 3 Yes  Agricultural tributary 16 Aug 08 – Sep 09 w – 4 s – 3 No  Agricultural tributary 22 Jun 09 – May 10 w – 16 s – 2 Yes Total 17 sites  w – 124 samples (not including historic) s – 33 samples     17 4 Methods  4.1 Land use indices and climate data 4.1.1 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  18 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 Field Sampling and Laboratory Analysis 4.2.1 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  19 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 #10-107-06- 2-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 time- integrated 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  20 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).  21 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 acid- washed, 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.  22 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 Data analysis methods 4.3.1 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  23 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  24 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.   25 5 Results and Discussion 5.1 Land Use Change – Landscape Stressors 5.1.1 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 Area (ha) TIA (ha) % TIA Urban 165 57.5 35 Agricultural 73 5.1 7.1 Marshall Creek 3800 874 23  5.1.2 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.    26 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 1991 1996 2001 2006 Poultry 2,724,660 2,709,513 3,176,719 3,116,341 Poultry Farms 129 139 116 90 Average # Poultry per Farm 21,121 19,493 27,386 34,626   Beef and Dairy Cattle 2,196 1,957 1,334 1,371 Cattle Farms  102 94 46 41 Average # Cattle per Farm  22 21 29 33   Pigs 6,015 5,807 6,134 832 Pig Farms 11 12 7 9 Average # Pigs per Farm 547 484 876 92  Sumas Prairie Region 1991 1996 2001 2006 Poultry 488,976 872,075 1,816,021 1,761,956 Poultry Farms 52 71 86 63 Average # Poultry per Farm 9,403 12,283 21,117 27,968   Beef and Dairy Cattle 18,535 18,293 19,778 20,100 Cattle Farms  180 161 142 118 Average # Cattle per Farm  103 114 139 170   Pigs 38,862 41,429 60,086 50,105 Pig Farms 36 28 26 18 Average # Pigs per Farm 1,080 1,480 2,311 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.  27 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 Abbotsford Aquifer portion Total Poultry 56,408 621,440 677,848 Poultry Farms 2 18 20 Average # Poultry per Farm 28,204 34,626 Poultry farms as % of total farms 8.7% 26%  Beef and Dairy Cattle 3,050 273 3,323 Cattle Farms  12 8 20 Average # Cattle per Farm  254 33 Cattle farms as % of total farms 52% 12%  Pigs 1,334 166 1,500 Pig Farms 2 2 4 Average # Pigs per Farm 889 92 Pig farms as % of total farms 6.5% 2.6%  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 2006 Poultry 28,204 Poultry Farms 1 Average # Poultry per Farm 28,204   Dairy Cattle 762 Dairy Farms  3 Average # Cattle per Farm  254   28 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 AUEs Area (ha) AUEs/ha 1996 20,279 3,875 5.23 2001 21,634 5,216 4.15 2006 20,783 6,088 3.41   Sumas Prairie Region 1996 30,504 7,596 4.02 2001 40,834 8,124 5.03 2006 39,163 8,048 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.   29 Table 8. Average 2006 Animal Unit Equivalents in the Marshall Creek Watershed and the agricultural subwatershed.  AUEs Area (ha) AUEs/ha Abbotsford Aquifer Region 4,170 1,214 3.44 Sumas Prairie Region 3,727 840 4.44 Marshall Creek Watershed 7,897 2,054 3.85  Agricultural Subwatershed 1,129 72.66 15.5  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).  30 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 # Average pop. per dwelling (from census) # of dwellings within catchment (from GIS) Estimated pop. within catchment 9320001.00 4.09 40 164 9320005.01 2.87 720 2068 9320011.00 2.91 41 119 9320012.00 3.18 426 1354 9320100.00  3.22 253 814 9320101.00* 2.16 1014* 2187 9320102.00  2.07 2028 4193 9320103.00  1.71 17 29 9320104.00 2.95 1267 3740 9320105.00* 2.84 1422* 4042 9320106.02 2.71 2320 6289 9320106.03* 2.94 62* 182  Total 9610 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 1996 estimated pop. within catchment 9320001.00 157 153 9320005.01 1955 2045 9320011.00 118 114 9320012.00 1066 926 9320100.00  799 726 9320101.00* 1890 1749 9320102.00  4303 4156 9320103.00  31 30 9320104.00 3667 3467 9320105.00* 3815 3549 9320106.02 5060 3746 9320106.03* 175 136 Total 23,037 20,797   31 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 Stream Quality – Measurement Endpoints 5.2.1 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 Dry Season Range Mean N Range Mean N NO3--N (mg/L) 0.088 – 19.5 4.40 83 <0.05 – 11.7 3.78 43 NH4+-N (mg/L) 0.051 – 7.74 0.790 83 <0.05 – 1.94 0.285 30 SRP (mg/L) <0.001 – 0.323 0.0280 83 <0.001 – 0.0545 0.0136 42 Cl- (mg/L) 6.59 – 171 38.0 83 9.43 – 137 23.0 30 Turbidity (NTU) 0.96 – 125 21.8 67 N/A N/A 0 TOC (mg/L) 0.96 – 19.4 4.37 72 1.27 – 22.3 3.98 30 DO (mg/L) 3.4 – 11.5 8.05 67 1.7 – 11.1 8.50 27 pH 6.16 – 7.79 7.03 83 6.34 – 7.83 7.13 43 Sp. Cond (µS/cm) 12.8 – 660 257 83 119 – 568 257 43 Temperature (ºC) 0.10 – 17.0 7.79 79 12.4 – 18.0 14.7 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.   32 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).  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)  0 2 4 6 8 10 12 12 14 15 11 17 18 7 2 N itr a te - N  (p pm ) Upstream to Downstream Sites Mean Drinking Water Guideline Freshwater Aquatic Life Guideline  33 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).   34 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 30- day 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. 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 12 14 15 11 17 18 7 2 SR P (p pm ) Upstream to Downstream Sites Mean Eutrophic Trigger Range (TP)  35  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 0 1 2 3 4 5 6 7 8 9 10 12 14 15 11 17 18 7 2 TO C  (p pm ) Upstream to Downstream Sites Mean  36 minimum requirement for ‘other’ life stages of warm-water biota (Canadian Council of Ministers of the Environment, 2007).  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. 0 2 4 6 8 10 12 14 15 11 17 18 7 2 D O  (p pm ) Upstream to Downstream Sites Mean Cold-water biota minimum requirement Warm-water biota minimum requirement  37 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) Correlation + or - With Parameters(s) Dissolved N (NO3-/NO2-N) - Temperature Soluble Reactive Phosphorus (SRP) + Dissolved N (NO3-/NO2-N & NH4+-N), Cl-, TOC Cl- + NH4+-N, Sp. Cond, TOC NH4+-N + Sp. Cond, TOC pH - NO3-/NO2-N, SRP, Cl-, Turbidity, TOC Dissolved Oxygen (DO) - SRP, Cl-, NH4+-N, Sp. Cond, Turbidity, TOC Turbidity + 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) Correlation + or - With Parameters(s) Soluble Reactive Phosphorus (SRP) + NH4+-N, TOC Cl- + NH4+-N, Sp. Cond, Turbidity TOC NH4+-N + TOC, Turbidity pH - SRP, TOC Dissolved Oxygen (DO) - SRP, Cl-, NH4+-N, Sp. Cond, Turbidity, TOC Turbidity + Sp. Cond, TOC   38 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) Correlation + or - With Parameters(s) Dissolved N (NO3-/NO2-N) - Cl-, Sp. Cond SRP + NH4+-N, Cl- + 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  39 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 Value < or > In comparison to NO3--N Marshall Mainstem > Urban trib. and agricultural trib. NH4+-N, SRP Urban trib. < Agricultural trib. and Marshall mainstem Cl-, Turbidity, TOC Agricultural trib. > Marshall mainstem Marshall mainstem > Urban trib. pH, DO Urban trib. > Marshall mainstem DO Marshall mainstem > Agricultural trib. Sp. Cond 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.  40  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 0 1 2 3 4 5 6 7 8 Wet Dry Wet Dry Wet Dry Wet Dry HW Mouth Ag. Trib. Urban Trib. N itr a te - N  (m g/ L) 0 0.01 0.02 0.03 0.04 0.05 Wet Dry Wet Dry Wet Dry Wet Dry HW Mouth Ag. Trib. Urban Trib. SR P (m g/ L) 0 2 4 6 8 10 12 Wet Dry Wet Dry Wet Dry Wet Dry HW Mouth Ag. Trib. Urban Trib. D O  (m g/ L) 0 2 4 6 8 10 12 Wet Dry Wet Dry Wet Dry Wet Dry HW Mouth Ag. Trib. Urban Trib. TO C  (m g/ L)  41 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 1994- 95, 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  42 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.  43  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. 0 5 10 15 20 25 0 2 / 2 0 / 9 4 0 5 / 0 9 / 9 4 0 7 / 2 6 / 9 4 1 0 / 0 3 / 9 4 0 2 / 0 1 / 9 5 0 8 / 2 2 / 0 2 0 7 / 0 7 / 0 3 1 0 / 1 4 / 0 3 1 1 / 2 4 / 0 3 0 1 / 1 2 / 0 4 0 3 / 0 1 / 0 4 0 4 / 1 9 / 0 4 0 6 / 0 7 / 0 4 0 4 / 2 4 / 0 6 0 5 / 2 3 / 0 6 0 6 / 1 9 / 0 6 0 7 / 1 7 / 0 6 0 8 / 1 5 / 0 6 0 9 / 1 1 / 0 6 1 0 / 1 0 / 0 6 1 1 / 0 6 / 0 6 0 4 / 0 2 / 0 7 0 4 / 3 0 / 0 7 0 5 / 2 8 / 0 7 0 6 / 2 5 / 0 7 0 7 / 2 3 / 0 7 0 9 / 1 0 / 0 7 1 0 / 0 9 / 0 7 1 1 / 0 5 / 0 7 1 2 / 0 3 / 0 7 0 8 / 1 8 / 0 8 0 6 / 1 8 / 0 9 1 0 / 2 8 / 0 9 1 2 / 1 1 / 0 9 1 2 / 2 9 / 0 9 0 1 / 1 8 / 1 0 0 4 / 1 5 / 1 0 0 5 / 0 4 / 1 0 0 5 / 3 1 / 1 0 N i t r a t e - N  ( m g / L ) Site 11 Site 2  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 non- agricultural 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).   45    0 10 20 30 40 50 60 70 1993 1994 2003 2004 2008 2009 1993 1994 2003 2004 2008 2009 2008 2009 2008 2009 Headwaters (11) Mouth (2) Ag. Trib. (22) Urb. Trib. (8) C u  (m g/ kg ) ISQGa) 0 200 400 600 800 1000 1200 1400 1993 1994 2003 2004 2008 2009 1993 1994 2003 2004 2008 2009 2008 2009 2008 2009 Headwaters (11) Mouth (2) Ag. Trib. (22) Urb. Trib. (8) M n  (m g/ kg ) b)  46   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)  0 50 100 150 200 250 1993 1994 2003 2004 2008 2009 1993 1994 2003 2004 2008 2009 2008 2009 2008 2009 Headwaters (11) Mouth (2) Ag. Trib. (22) Urb. Trib. (8) Zn  (m g/ kg ) ISQGc) 0 500 1000 1500 2000 2500 3000 1993 1994 2003 2004 2008 2009 1993 1994 2003 2004 2008 2009 2008 2009 2008 2009 Headwaters (11) Mouth (2) Ag. Trib. (22) Urb. Trib. (8) P (m g/ kg ) d)  47 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  48 (<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 Range Mean ISQG Dry mass (mg) 400 – 51050 9450 N/A Al (mg/kg) 7892 – 92600 22600 N/A Ca (mg/kg) 3714 – 42150 15280 N/A Cd (mg/kg) 0 – 11.14 1.731 0.6 Cr (mg/kg) 15.55 – 384.9 81.42 37.3 Cu (mg/kg) 45.54 – 508.1 111.7 35.7 Fe (mg/kg) 17970 – 195000 49880 N/A K (mg/kg) 605.4 – 8514 1893 N/A Mg (mg/kg) 4791 – 135100 17220 N/A Mn (mg/kg) 312.6 – 11430 1731 N/A Ni (mg/kg) 21.16 – 1305 134.1 N/A P (mg/kg) 507.2 – 5534 2032 N/A Pb (mg/kg) 5.893 – 364.3 67.50 35 Zn (mg/kg) 69.80 – 1074 228.5 123  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).  49 Table 18. Significant Mann-Whitney U results (p<0.0167) for parameters in suspended sediments by site. Parameter Site Value < or > In comparison to Dry mass (g) Agricultural trib. > Urban trib. Cu (mg/kg) Agricultural trib. < Urban trib. Mg (mg/kg) Agricultural trib. > Urban trib. and Marshall mainstem Mn (mg/kg) Agricultural trib. < Urban trib. and Marshall mainstem Ni (mg/kg) Urban trib.  < Agricultural trib. P (mg/kg) Urban trib. < Agricultural trib. and Marshall mainstem Zn (mg/kg) Agricultural trib.  < 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 Site Value < or > In comparison to Mn (mg/kg) Agricultural trib. < Urban trib. Urban trib. < Marshall Mainstem Ni (mg/kg) Urban trib. < Agricultural trib. and Marshall mainstem Zn (mg/kg) Agricultural trib. < Urban trib. and Marshall mainstem  50 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  51 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.   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. 0 5 10 15 20 25 30 35 40 45 50 55 60 65 Urban Ag. Marshall Urban Ag. Marshall Urban Ag. Marshall Summer '09 Winter '09-'10 Spring '10 A v er a ge  ch l-a  c o n ce n tr a tio n  (m g/ L) Mean  52   0.1 1 10 100 11 19 27 34 M ea n  C hl - a  (m g/ L) Days of colonization Summer 2009 Urban Agriculture Marshall a) 0.01 0.1 1 10 10 18 27 38 M ea n  C hl - a  (m g/ L)  Days of Growth Winter 2009-10 Urban Agriculture Marshall b)  53  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 0 2 4 6 8 10 12 14 11 19 27 35 M ea n  C hl - a  (m g/ L) Days of colonization Spring 2010 Urban Agriculture Marshall c)  54 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 Sep. 10/09 (34 days of colonization, n=2) Jan. 18/10 (38 days of colonization, n=3) May 31/10 (35 days of colonization, n=3) Urban Trib. 1.405 ± 0.3857 7.643 ± 1.179 10.58 ± 3.171 Agricultural Trib. 18.55 ± 2.342 0.4435 ± 0.03617 8.320 ± 1.442 Marshall Mainstem 10.55 ± 8.992 0.3130 ± 0.07708 7.557 ± 0.1927   55  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). 0 1 2 3 4 5 6 7 8 9 10 12 /1 1/ 09 12 /1 3/ 09 12 /1 5/ 09 12 /1 7/ 09 12 /1 9/ 09 12 /2 1/ 09 12 /2 3/ 09 12 /2 5/ 09 12 /2 7/ 09 12 /2 9/ 09 12 /3 1/ 09 01 /0 2/ 10 01 /0 4/ 10 01 /0 6/ 10 01 /0 8/ 10 01 /1 0/ 10 01 /1 2/ 10 01 /1 4/ 10 01 /1 6/ 10 01 /1 8/ 10 W at er  Te m pe ra tu re  (ºC ) Urban Tributary Agricultural Tributary Marshall Mainstem  56  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. 0 2000 4000 6000 8000 10000 12000 U rb an A g. M ar sh al l U rb an A g. M ar sh al l U rb an A g. M ar sh al l Summer '09 Winter '09-'10 Spring '10 A v er a ge  bi o fil m  dr y m a ss  (m g) Mean  57   0 100 200 300 400 500 600 700 11 19 27 34 M ea n  dr y m a ss  (m g) Days of colonization Summer 2009 Urban Agriculture Marshall a) 0 1000 2000 3000 4000 5000 6000 7000 8000 10 18 27 38 M ea n  dr y m a ss  (m g) Days of colonization Winter 2009-10 Urban Agriculture Marshall b)  58  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 0 500 1000 1500 2000 2500 3000 3500 10 18 27 38 M ea n  dr y m a ss  (m g) Days of colonization Spring 2010 Urban Agriculture Marshall c)  59 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.  60 Table 21. Summary of trace elements measured in biofilms collected in the Marshall Creek watershed. (n=100) Element Range Mean Cr (mg/kg) 0 – 398 30.4 Cu (mg/kg) 0 – 98.2 25.0 Fe (mg/kg) 2880 – 88800 22100 Mg (mg/kg) 81.4 – 26500 4200 Mn (mg/kg) 86.9 – 2020 593 Ni (mg/kg) 0 – 260 45.6 P (mg/kg) 118 – 3970 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) NO3--N (mg/L) SRP (mg/L) Cl- (mg/L) NH4+-N (mg/L) pH Temp. (°C) Turbidity (NTU) Average chl-a (mg/L) R -0.374 -0.455 -0.422 -0.479 -0.545 0.734 0.807 -0.537 p 0.012 0.022 0.035 0.015 0.005 <0.0001 <0.0001 0.007 n 44 25 25 25 25 25 25 24  Correlations were then determined between biofilm and water quality data from each site individually (not pictured) and the significant positive correlation that was also  61 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.  Marshall Mainstem Urban Tributary Agricultural Tributary Metal Sediment Biofilm Sediment Biofilm Sediment Biofilm Cr 36.9 ± 3.92 28.0 ± 1.88 117 ± 13.4 28.7 ± 1.39 69.6 ± 10.3 34.3 ± 2.19 Cu 110 ± 11.7 30.4 ± 0.694 145 ± 15.5 24.3 ± 0.252 75.2 ± 2.08 20.7 ± 0.336 Mn 2450 ± 435 870 ± 16.0 2240 ± 385 445 ± 8.93 715 ± 66.5 481 ± 10.7 Ni 92.8 ± 4.71 55.9 ± 1.51 61.8 ± 6.50 26.7 ± 0.735 241 ± 53.8 54.8 ± 1.41 P 2290 ± 130 1320 ± 21.2 1400 ± 176 665 ± 13.6 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.   62  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. 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 1000 2000 3000 4000 5000 6000 B ed  Se di m en t Su s.  Se di m en t B io fil m W at er B ed  Se di m en t Su s.  Se di m en t B io fil m W at er B ed  Se di m en t Su s.  Se di m en t B io fil m W at er Urban Trib. Agricultural Trib. Marshall Mainstem SR P in  w a te r (m g/ L) P in  se di m en ts  a n d bi o fil m s (m g/ kg ) Mean  63 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 Mn P Ni Cr Biofilm : Sus. Sediment 0.224 0.343 0.650 0.340 0.373 Biofilm : Bed Sediment 0.508 0.801 0.948 0.219 0.339 Sus. Sediment : Bed Sediment 2.27 2.34 1.46 0.645 0.910  Table 25. Trace element ratios in biofilms and sediments by site (Marshall mainstem (Mar.), urban tributary (Urb.) and agricultural tributary (Ag.)).  Cu Mn P Ni Cr  Mar. Urb. Ag. Mar. Urb. Ag. Mar. Urb. Ag. Mar. Urb. Ag. Mar. Urb. Ag. Biofilm : Sus. Sediment 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 Biofilm : Bed Sediment 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 Sus. Sediment : Bed Sediment 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  66 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.  67 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.) 2006 Population Density (est.) (#/ha) Urban 165 35 0 0 0 1716 10.4 Agricultural 73 7 28,204 254 15.5 29 0.397 Marshall Creek 3800 23 33,892 166 3.85 25,181 6.63  Table 27. Summary of in-stream parameters measured at three sites. (Mean ± 1. S.E.) Sample Parameter Urban Tributary (Site 8) Agricultural Tributary (Site 22) Marshall Creek Mainstem (Site 2) Water (n=17) NO3--N (mg/L) 1.82 ± 0.0306 1.97 ± 0.183 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 Bed Sediments (n=3) 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 Suspended Sediments (n=7) 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  68 Sample Parameter Urban Tributary (Site 8) Agricultural Tributary (Site 22) Marshall Creek Mainstem (Site 2) Biofilms (n=34) Chl-a (µg/L) 4500 ± 126 9260 ± 436 5240 ± 203 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,  69 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.   70 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 in- stream 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  71 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 1996- 2006. 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.  72 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.   73 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 long-  74 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  75 (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  76 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  77 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.    78 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, 439- 442. 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.  79 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  80 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.  81 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.  82 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.  83 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.  84 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.  85 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.  86 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/92- 597/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  87 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, 391- 405. 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, 4626- 4632.  88 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.  89 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.  90  Figure A1. Map of 2006 Census Enumeration Area boundaries (South Matsqui & Sumas Prairie) and Dissemination Area 59090721.  91  Figure A2. Map of Census Tract boundaries for Population data.  92  Figure A3. Map of Marshall Creek watershed with delineated urban and agricultural subwatersheds.  93  Figure A4. Map of the Total Impervious Area (TIA) delineation for the urban subwatershed.  94  Figure A5. Map of the Total Impervious Area (TIA) delineation for the agricultural subwatershed.  95 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.  96 Table B1. Summary of water quality results. Site # Site Description Date NO3--N (mg/L) SRP (mg/L) Cl- (mg/L) NH4+-N (mg/L) pH Sp. Cond. (µS/cm) Temp. (ºC) DO (mg/L) Turbidity (NTU) TOC (mg/L) 2  Marshall Creek mainstem - mouth  8/18/08 4.54 0.00951 20.7 0.296 7.10 233    2.80 12/15/08 4.79 0.00418 29.3 0.455 6.96 205 0.200   4.23 06/18/09 4.20 0.0101   6.99 245 15.9 6.59 09/17/09 3.42 0.0194 10.6 0.138 7.14 269 18.0 9.10  2.59 10/28/09 19.2 0.323 25.7 0.150 6.16 257   9.79 14.1 10.0 11/25/09 5.60 0.0498 29.6 0.230 6.58 253 9.20 6.25 8.69 6.96 12/11/09 4.75 0.0139 33.3 0.310 6.85 231 4.20 10.2 10.3 2.68 12/21/09 4.19 0.0494 34.4 2.56 6.97 221 7.00 7.48 11.4 6.25 12/29/09 4.50 0.0362 37.8 7.74 6.93 288 4.90 7.30 20.8 5.51 01/07/10 4.85 0.0535 36.8 5.58 6.87 258 3.20 8.20 11.7 01/18/10 3.81 0.0495 32.1 3.93 6.87 259 8.00 5.90 15.1 6.18 02/10/10 4.43 0.0199 35.0 0.110 6.87 249 6.90 7.80 15.9 3.64 04/15/10 3.59 0.0759 39.5 7.69 7.18 329 11.0 6.40 15.9 5.29 04/26/10 3.53 0.0158 30.6 0.284 7.17 258 11.0 7.42 9.68 05/04/10 3.11 0.0405 24.9 0.831 7.05 217 9.20 6.40 20.5 05/12/10 3.63 0.0141 27.6 0.190 7.64 271 12.6 7.00 9.13 2.77 05/31/10 3.38 0.0299 28.9 0.451 7.24 278 13.5 5.30 7.23 4.02 3 Small urban tributary 8/18/08 1.01 0.00608 13.3 0.00 7.34 194    1.59 12/15/08 1.39 0.0158 6.59 0.0829 7.39 249 5.50   1.20 06/18/09 0.608 0.000   6.99 170 14.7 7.72 4 Small urban tributary 8/18/08 0.398 0.00342 30.3 0.110 7.52 228    3.44 12/15/08 0.852 0.00350 17.1 0.121 7.44 135 0.100   2.12 06/18/09 0.572 0.00350   7.54 240 15.8 9.71 09/17/09 0.366 0.0266 39.0 0.0569 7.45 283 15.7 10.1  3.39 5 Urban detention pond 8/18/08 0.000 0.00898 18.8 0.00620 6.91 119    3.83 12/15/08 0.714 0.00377 6.75 0.0531 7.28 98.0 6.40   1.64    97 Table B1. Continued Site # Site Description Date NO3--N (mg/L) SRP (mg/L) Cl- (mg/L) NH4+-N (mg/L) pH Sp. Cond. (µS/cm) Temp. (ºC) DO (mg/L) Turbidity (NTU) TOC (mg/L) 7 Marshall Creek mainstem 8/18/08 5.60 0.0127 23.7 0.00 7.20 235    2.42 12/15/08 5.27 0.00406 29.2 0.596 6.95 196 1.30   4.20 06/18/09 4.26 0.00881   7.03 245 15.2 9.02 09/17/09 5.66 0.00330 16.8 0.0340 7.13 252 14.0 9.10  2.42 12/11/09 5.44 0.0137 34.2 0.420 6.83 226 5.00 9.03 65.8 2.73 01/18/10 4.61 0.0273 35.5 0.540 6.76 236 8.20 6.50 17.9 5.78 05/31/10 3.00 0.0312 14.1 0.335 7.08 208 12.3 6.80 19.6 3.56 8 Urban tributary - DeLair Park 8/18/08 1.97 0.00622 26.6 0.108 7.20 309    4.31 12/15/08 2.34 0.00363 18.2 0.528 7.23 227 4.70   1.72 06/18/09 2.00 0.000168   6.99 315 15.2 8.92 09/17/09 1.66 0.00600 26.6 0.0860 7.37 343 15.3 8.30  2.88 10/28/09 8.67 0.0167 22.5 0.0800 6.50 319  11.5 1.00 2.55 11/25/09 2.55 0.0117 22.1 0.0600 7.07 268 10.6 11.0 3.77 2.12 12/11/09 2.54 0.00650 34.7 0.0500 7.11 279 7.90 11.3 1.26 1.57 12/21/09 0.929 0.00140 21.0 0.0800 7.06 139 7.90 11.0 16.1 1.80 12/29/09 2.03 0.00760 28.3 0.0700 7.25 263 7.70 10.5 0.96 1.32 01/07/10 2.32 0.00990 28.8 0.0800 7.09 240 7.00 9.10 1.38 01/18/10 2.30 0.0154 22.8 0.0500 7.19 229 9.20 10.1 1.08 1.71 02/10/10 1.96 0.00433 28.3 0.138 6.92 262 8.20 9.30 2.23 1.71 04/15/10 1.59 0.00877 20.9 0.314 7.27 257 10.2 8.70 1.71 2.00 04/26/10 1.58 0.00765 20.6 0.0757 7.31 262 10.1 9.37 1.55 05/04/10 1.36 0.00649 16.0 0.0796 7.37 242 9.50 9.10 1.51 05/12/10 1.60 0.00571 21.5 0.0665 7.47 272 10.8 8.60 1.58 1.63 05/20/10 1.51 0.0296 18.8 0.234 7.11 260 11.4 7.80 2.96 3.17 05/31/10 0.760 0.00940 7.97 0.199 7.14 136 12.4 8.20 4.14 2.44    98 Table B1. Continued Site  Site Description Date NO3--N (mg/L) SRP (mg/L) Cl- (mg/L) NH4+-N (mg/L) pH Sp. Cond. (µS/cm) Temp. (ºC) DO (mg/L) Turbidity (NTU) TOC (mg/L) 10 Tributary with groundwater influence 8/18/08 1.31 0.00731 24.2 0.191 7.25 196    6.45 12/15/08 1.24 0.0105 38.9 0.637 6.80 176 1.40   3.92 09/17/09 1.30 0.0545 22.6 0.247 6.87 220 16.1 8.47  3.94 11 Marshall Creek mainstem - historic site 8/18/08 6.78 0.0111 15.5 0.00 7.29 208    1.62 12/15/08 6.93 0.0189 15.8 0.362 7.19 182 4.30   1.67 06/18/09 5.40 0.00522   7.19 230 12.8 10.6 09/17/09 7.26 0.0237 14.4 0.210 7.01 243 12.4 10.4  3.28 10/28/09 10.0 0.0194 17.4 0.140 6.87 202  10.8 3.58 2.69 11/25/09 6.16 0.0150 17.6 0.100 6.92 206 9.50 10.0 8.20 2.48 12/11/09 7.35 0.0139 19.7 0.250 7.70 210 7.00 10.9 12.0 1.23 01/07/10 6.31 0.0141 19.9 0.120 7.13 194 6.90 10.6 8.76 01/18/10 5.25 0.0168 19.7 0.100 7.13 205 9.60 9.30 5.13 1.75 02/10/10 5.24 0.0124 17.2 0.131 7.10 213 8.40 9.20 5.21 1.43 04/15/10 5.81 0.0170 18.1 0.360 7.37 228 11.0 8.90 3.22 1.80 05/31/10 3.67 0.0222 10.67 0.463 7.25 170 12.2 8.10 8.87 3.33 12 Groundwater Well (Hatchery) 8/18/08 9.97 0.00716 13.3 0.0314 6.97 230    0.500 06/18/09 7.77 0.0137   7.24 270  9.25 13 Tributary with groundwater influence 8/18/08 11.7 0.0108 9.42 0.0297 7.34 293    1.42 12/15/08 7.90 0.0153 11.9 0.397 6.98 192 4.80   1.65 14 Marshall Creek mainstem - Hatchery discharge 8/18/08 9.90 0.0134 13.9 0.00 7.14 248    2.74 12/15/08 9.75 0.0141 14.4 0.369 7.02 212 5.00   1.38 06/18/09 8.04 0.00595   7.26 260 13.6 11.1 09/17/09 9.14 0.0265 13.3 0.158 7.10 262 12.5 9.70  1.45 12/11/09 9.54 0.0107 14.5 0.250 7.63 227 7.80 10.1 5.12 0.96 01/18/10 7.78 0.0007 13.1 0.120 7.11 214 9.50 8.54 2.89 1.35 04/15/10 7.84 0.0190 12.8 0.300 7.29 252 12.0 9.30 1.69 1.24 05/31/10 6.16 0.0266 17.6 0.295 7.13 230 11.5 7.50 6.82 2.29  99 Table B1. Continued Site  Site Description Date NO3--N (mg/L) SRP (mg/L) Cl- (mg/L) NH4+-N (mg/L) pH Sp. Cond. (µS/cm) Temp. (ºC) DO (mg/L) Turbidity (NTU) TOC (mg/L) 15 Marshall Creek mainstem 8/18/08 8.78 0.0167 14.8 0.109 7.35 229    1.65 12/15/08 5.58 0.0211 14.5 0.424 7.35 209 5.00   1.56 06/18/09 7.85 0.00505   7.26 250 13.5 11.0 09/17/09 9.00 0.0243 13.4 0.138 7.11 256 12.4 10.4  1.27 05/31/10 6.33 0.0270 25.5 0.231 7.14 241 11.2 7.80 9.18 2.20 16 Small agricultural tributary 8/18/08 0.0240 0.0140 20.4 1.70 6.70 242    9.33 12/15/08 4.11 0.0233 18.0 0.595 6.65 221 2.40   9.30 06/18/09 0.0320 0.0173   6.52 210 13.8 3.73 09/17/09 0 0.00370 16.8 1.94 6.34 297 13.5 3.40  6.12 17 Marshall Creek mainstem 8/18/08 6.14 0.00996 16.1 0.00 7.26 221    2.75 12/15/08 6.51 0.0125 17.7 0.690 7.07 174 3.10   3.04 06/18/09 4.88 0.00703   7.04 250 13.4 9.80 09/17/09 6.55 0.00 15.3 0.209 7.02 243 13.1 10.4  2.22 12/11/09 6.63 0.0135 21.3 0.160 7.79 209 6.40 11.0 15.9 1.94 01/18/10 4.84 0.0179 22.5 0.330 6.94 207 9.10 8.70 8.34 3.65 05/31/10 2.71 0.0370 9.78 0.579 7.05 155 12.5 7.40 13.2 6.32 18 Marshall Creek mainstem 8/18/08 6.18 0.0144 19.2 0.00 7.33 227    2.27 12/15/08 5.73 0.0231 27.7 1.36 6.98 197 1.80   4.56 06/18/09 4.77 0.00956   7.03 270 14.0 8.73 09/17/09 5.88 0.00390 17.0 0.133 6.89 250 13.8 7.50  2.44 11/25/09 7.34 0.0461 36.3 0.560 6.54 316 9.80 5.10 13.5 7.83 12/11/09 5.83 0.0164 35.1 0.530 7.65 231 5.60 9.80 63.5 2.80 01/07/10 6.29 0.0255 39.6 0.460 6.85 233 4.70 9.30 22.6 01/18/10 5.18 0.0350 36.9 0.510 6.81 244 8.70 7.00 17.7 5.76 02/10/10 4.72 0.0471 38.5 0.144 6.94 249 7.30 7.90 122 3.57 04/15/10 4.52 0.0271 39.9 0.436 7.24 277 13.0 8.30 14.7 2.96 05/31/10 2.34 0.0250 9.47 0.488 7.08 147 12.4 7.20 19.3 4.74   100 Table B1. Continued Site  Site Description Date NO3--N (mg/L) SRP (mg/L) Cl- (mg/L) NH4+-N (mg/L) pH Sp. Cond. (µS/cm) Temp. (ºC) DO (mg/L ) Turbidity (NTU) TOC (mg/L) 22  Agricultural tributary - Kenny Rd.  06/18/09 0.0470 0.0345   6.91 365 17.3 3.75 09/17/09 0.0690 0.0466 137 1.48 6.54 568 16.4 1.70  10.4 10/28/09 19.5 0.0743 51.9 0.740 6.34 320   7.42 15.4 15.0 11/25/09 7.71 0.0793 78.9 1.19 6.31 369 9.20 3.40 26.5 12.9 12/11/09 0.852 0.0222 98.9 1.63 7.27 466 2.60 6.80 54.2 6.37 12/21/09 7.88 0.127 66.5 1.13 6.32 279 5.90 4.80 37.2 15.2 12/29/09 1.14 0.0178 127 1.47 6.63 325 2.30 6.25 56.1 6.82 01/07/10 4.78 0.0305 120 0.960 6.89 357 2.90 7.40 69.6 01/18/10 3.66 0.0643 127 1.60 6.64 332 7.70 4.60 48.0 10.1 02/10/10 0.447 0.0262 143 2.15 6.60 457 5.80 5.20 76.9 7.69 04/15/10 0.130 0.0104 167 1.34 7.00 660 17.0 6.10 52.6 6.84 04/26/10 0.282 0.0292 113 1.14 6.99 580 10.2 7.41 42.2 05/04/10 0.895 0.0684 86.7 1.11 6.71 352 9.30 7.50 60.4 05/12/10 0.0890 0.00990 128 1.04 6.98 542 12.4 5.70 50.2 8.23 05/20/10 0.0880 0.0195 171 1.70 7.04 541 12.4 5.70 125 9.58 05/31/10 1.45 0.0675 52.5 1.25 6.77 450 13.4 4.50 42.0 19.4 23 Small urban tributary 06/18/09 1.72 0.00316   7.83 400 15.2 10.1 09/17/09 1.11 0.0182 38.0 0.0310 7.76 390 14.9 10.9  2.68   Figure B1. Simple Seasonal Time-Series model output on historic nitrate-N data collected at site 11. 101   Figure B2. Autocorrelation function graph of historic nitrate-N data collected at site 11. 102    103 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.  104 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.  105 Table C2. Summary of bed sediment results. Site # Date Sample ID 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) 2 08/18/08 102 20973 6900 8.4 63.0 63.0 58669 1183 8276 724 91.5 2548 115.1 215.4 09/17/09 302 27806 8122 1.46 61.8 65.0 41524 2588 8313 1277 90.2 1965 21.6 204.2 3 06/18/09 203 15161 6506 0.89 42.1 34.4 20620 744 4848 272 38.3 847 10.4 71.4 4 09/17/09 304 25310 12012 1.07 88.5 38.4 27849 2390 7661 1156 54.7 553 17.4 113.9 5 08/18/08 105 24089 4125 4.3 68.1 42.1 29234 1049 6011 328 55.1 876 55.8 95.7 7 08/18/08 107 19562 6241 8.6 64.7 98.4 61964 990 9129 560 91.3 2890 121.0 370.7 06/18/09 207 21921 6634 1.98 64.2 78.3 56083 1061 9141 605 89.3 2899 29.8 232.7 09/17/09 307 24480 9268 1.60 63.4 80.9 50320 1654 9748 623 85.2 2754 37.2 271.1 8 08/18/08 108 12710 5063 3.9 63.7 42.4 26283 746 5137 382 36.3 738 50.9 82.0 06/18/09 208 13749 5225 1.20 56.5 35.5 25070 680 4446 361 27.5 756 11.4 79.2 09/17/09 308 20326 9401 1.344 82.9 54.9 31670 1305 6298 595 35.0 881 15.0 119.2 10 08/18/08 110 13134 4801 7.6 41.4 78.0 52723 682 4865 1327 35.2 1131 111.4 189.2 09/17/09 310 17404 8326 1.58 57.6 64.7 42913 876 7110 2026 56.3 1149 37.6 326.2 11 08/18/08 111 14939 6424 3.6 51.1 66.3 38429 759 8953 964 91.7 1267 70.5 198.5 06/18/09 211 43900 10585 1.45 81.1 59.7 40347 3986 15523 511 98.0 752 10.8 118.3 09/17/09 311 16389 9277 1.38 50.0 58.8 34369 715 10712 1296 82.3 1400 29.4 191.5 13 08/18/08 113 18834 5565 5.8 54.3 325 38217 1080 7750 462 47.4 1338 98.6 244.8 14 08/18/08 114 14999 8725 8.7 49.9 81.9 59226 938 7416 1685 50.8 2507 126.5 270.4 09/17/09 314 20382 13350 1.62 68.3 84.5 41660 1479 9082 1168 52.8 2743 43.1 282.5 16 08/18/08 116 20244 5873 12.2 52.9 60.5 84990 1170 9599 464 115 3844 154.3 121.7 06/18/09 216 14734 6286 2.29 41.8 74.9 103579 798 7288 424 82.8 4708 22.1 112.2 09/17/09 316 16555 5006 2.51 47.9 73.8 104624 632 8605 429 99.5 4913 24.3 121.8 17 08/18/08 117 15606 6385 6.7 71.1 58.7 45809 805 8377 753 85.3 1625 99.2 181.0 06/18/09 217 13820 6065 1.50 44.0 49.4 43244 541 7146 639 62.1 1697 24.6 144.4 09/17/09 317 19752 9070 1.91 58.3 59.3 43200 853 9825 1396 89.0 2032 43.4 290.5 18 08/18/08 118 21469 8113 7.8 57.9 68.9 53656 1234 9716 716 102 2636 110.0 210.5 06/18/09 218 15908 6101 1.79 50.5 51.5 63609 792 7902 455 75.0 3305 25.8 156.1 09/17/09 318 29007 11843 1.78 71.3 82.4 49682 2303 11069 867 100 3078 28.7 234.4 22 06/18/09 222 7004 2841 0.80 233.5 30.9 47665 307 118822 1398 1076 881 14.6 63.7 09/17/09 322 27442 6343 1.64 65.2 53.2 38913 1249 10583 449 99.5 1984 13.1 168.7 23 09/17/09 323 15300 7863 1.24 55.5 38.0 23107 740 5495 601 35.4 633 12.8 87.8  106 Table C3. Summary of suspended sediment results. Site # Site Description Date Days of Collection Dry mass (mg) Al (mg/kg) Ca (mg/kg) Cd (mg/kg) Cr (mg/kg) Cu (mg/kg) Fe (mg/kg) 2 Marshall Creek mainstem - mouth 07/31/09 57 400 19000 15008 2.76 59.6 203 65779 09/10/09 41 660 14111 16234 1.54 42.9 116 32075 01/07/10 27 1190 15585 7265 1.79 49.2 99.8 48717 04/15/10 64 20830 22000 15120 0.00 16.0 45.5 67360 05/12/10 27 1770 17398 12554 0.00 17.0 85.8 51574 8  Urban tributary - DeLair Park  07/13/09 39 6540 18201 13441 1.20 65.9 74.3 25299 09/10/09 59 1660 20680 32420 2.80 240 142 34480 10/28/09 48 2180 23115 17307 4.20 116 140 44652 12/11/09 44 1750 11500 3714 0.00 46.9 84.5 17973 01/07/10 27 9810 12788 7896 1.77 87.4 90.4 23069 02/10/10 34 4420 36290 21834 0.00 53.3 93.8 70226 04/15/10 64 5380 16945 8458 0.00 23.0 54.9 36882 05/12/10 27 1210 14672 9406 0.00 32.0 118 29745 07/26/10 75 4710 92600 42150 11.1 385 508 195030 22 Agricultural tributary - Kenny Rd. 07/16/09 42 15310 19446 11718 1.43 57.5 64.4 41420 09/10/09 56 4380 21119 26483 1.68 52.2 80.0 45140 10/28/09 48 1120 22939 29866 1.43 55.8 108 23459 01/07/10 27 13930 16174 5560 1.32 57.5 67.8 25519 02/10/10 34 11300 25327 11416 0.00 15.6 51.8 52327 04/15/10 64 9400 26235 16662 0.00 16.7 67.7 54725 05/12/10 27 51050 23167 7027 0.00 33.0 79.1 58594 07/26/10 75 38880 7892 4669 5.03 269 82.9 53210    107 Table C3. Continued Site # Site Description Date Days of Collection K (mg/kg) Mg (mg/kg) Mn (mg/kg) Ni (mg/kg) P (mg/kg) Pb (mg/kg) Zn (mg/kg) 2 Marshall Creek mainstem - mouth 07/31/09 57 1204 9975 6271 122.9 2314.1 45.23 363.1 09/10/09 41 1467 9617 893 60.9 1350.6 18.52 202.6 01/07/10 27 822 7872 1640 92.2 2308.5 31.32 214.8 04/15/10 64 955 9412 2001 106.1 3184.9 142.53 250.6 05/12/10 27 2300 10683 1447 82.1 2283.9 117.12 212.9 8  Urban tributary - DeLair Park  07/13/09 39 1460 6839 782 35.3 556.8 16.33 161.3 09/10/09 59 1824 12916 997 102.3 1005.3 19.40 208.3 10/28/09 48 1772 9092 1541 47.1 1394.8 22.75 213.8 12/11/09 44 688 4791 710 21.2 507.2 5.89 79.1 01/07/10 27 691 5390 711 35.5 698.4 18.24 124.9 02/10/10 34 1686 15216 1434 58.3 1411.1 139.93 234.1 04/15/10 64 730 6920 1293 26.7 651.6 71.42 165.6 05/12/10 27 841 6921 1215 26.5 833.6 60.66 173.5 07/26/10 75 4768 33700 11431 203.0 5534.0 364.32 1073.6 22 Agricultural tributary - Kenny Rd. 07/16/09 42 1501 11416 537 83.4 2123.2 9.64 175.1 09/10/09 56 4890 19328 662 81.3 2812.5 11.19 174.6 10/28/09 48 8514 19229 501 73.6 2034.1 12.19 169.9 01/07/10 27 800 7196 313 74.5 1601.8 16.98 108.9 02/10/10 34 1490 12684 533 101.1 3461.1 84.15 192.5 04/15/10 64 1613 12775 750 112.5 3649.7 98.54 222.6 05/12/10 27 1021 11690 438 97.6 4084.2 109.89 236.5 07/26/10 75 605 135120 1990 1305.3 893.6 68.83 69.8  108 Appendix D: Biofilm sampling and analysis results  Figure D1. Photograph of tile apparatus used for biofilm colonization. Table D1. Summary of biofilm results.  109  Figure D1. Photograph of tile apparatus used for biofilm colonization.  110 Table D1. Summary of biofilm results. Site Number Site Description Date Sample ID Tile Replicate Chl-a (mg/L) Dry mass (mg) Cr (mg/kg) Cu (mg/kg) Fe (mg/kg) Mg (mg/kg) Mn (mg/kg) Ni (mg/kg) P (mg/kg) 2 Marshall Creek mainstem - mouth 08/18/09 102 1 4.81 44 0.0 0.0 9241 368 846 0.0 1013 08/18/09 102 2 3.00 16 0.0 1.5 3060 81 448 5.3 531 08/26/09 202 1 19.86 101 0.0 34.9 16404 2073 1467 46.2 2269 08/26/09 202 2 14.92 86 0.0 65.3 17752 2119 1340 55.6 2382 09/03/09 302 1 11.80 51 0.0 16.9 7492 898 916 25.1 1274 09/03/09 302 2 14.97 112 0.0 98.2 10896 1752 899 16.2 1061 09/10/09 402 1 2.07 143 0.0 16.4 12902 2018 1000 33.0 1710 09/10/09 402 2 27.50 61 0.0 11.4 8330 976 929 28.2 1251 12/21/09 502 1 0.02 78 0.0 17.4 19838 2713 773 28.3 784 12/21/09 502 2 0.08 29 66.9 50.0 7750 1344 441 63.4 1799 12/21/09 502 3 0.02 16 0.0 42.7 13230 2844 526 26.7 1456 12/29/09 602 1 0.33 108 22.9 37.6 17489 3533 719 43.3 1190 12/29/09 602 2 0.08 628 17.2 43.6 27149 5716 590 77.5 1148 12/29/09 602 3 0.24 901 26.1 88.7 88847 26459 280 167.7 3971 01/07/10 702 1 0.24 1542 13.7 24.6 18049 5138 302 43.3 716 01/07/10 702 2 0.18 996 3.3 7.8 7698 4185 134 20.9 670 01/07/10 702 3 0.05 177 0.0 17.8 14912 3016 275 11.0 1021 01/18/10 802 1 0.18 2075 11.6 18.1 21042 5119 364 38.3 968 01/18/10 802 2 0.58 1936 14.0 11.9 10328 4738 267 28.0 857 01/18/10 802 3 0.18 957 0.0 1.3 2882 448 87 7.1 118 04/26/10 902 1 2.38 196 0.0 22.5 33969 4563 1103 40.2 1093 04/26/10 902 2 7.79 297 0.0 31.0 37293 5622 668 51.2 1204 04/26/10 902 3 6.62 422 0.0 31.6 37935 5876 890 52.1 1385 05/04/10 1002 1 2.12 183 0.0 14.4 31217 4287 870 37.0 948 05/04/10 1002 2 6.81 373 0.0 28.2 31200 5304 1182 49.3 1391 05/04/10 1002 3 4.58 552 11.0 28.4 32955 5630 993 52.6 1389 05/20/10 1102 1 4.40 204 240.0 33.3 40485 4818 1354 183.7 1383 05/20/10 1102 2 4.29 135 235.2 39.4 30969 4692 1202 185.7 832 05/20/10 1102 3 4.72 171 72.6 28.0 32778 4420 1159 92.9 1137 05/31/10 1202 1 7.39 490 105.6 38.1 40475 6218 2023 125.4 1778 05/31/10 1202 2 8.20 609 23.7 35.2 40938 6637 1852 73.4 1743 05/31/10 1202 3 7.08 521 31.6 37.9 40119 6290 1943 79.9 1667    111  Site Number Site Description Date Sample ID Tile Replicate Chl-a (mg/L) Dry mass (mg) Cr (mg/kg) Cu (mg/kg) Fe (mg/kg) Mg (mg/kg) Mn (mg/kg) Ni (mg/kg) P (mg/kg) 8 Urban tributary - DeLair Park 07/13/09 8 1 2.48 415 55.9 20.1 22980 4484 334 27.1 487 07/13/09 8 2 2.21 685 39.6 28.6 17715 4331 331 25.1 435 08/18/09 108 1 0.43 56 0.0 35.2 12561 1791 1313 38.3 2089 08/18/09 108 2 0.56 30 0.0 37.4 11073 1071 1604 63.6 2478 08/26/09 208 1 11.11 430 0.0 24.9 13480 3096 651 23.2 990 08/26/09 208 2 0.70 194 0.0 34.7 14172 2781 949 37.9 1484 09/03/09 308 1 5.07 438 0.0 18.3 14421 3291 472 12.1 624 09/03/09 308 2 0.18 639 0.0 13.1 10440 3578 399 13.6 569 09/10/09 408 1 1.04 630 5.0 26.8 17104 3330 585 16.8 618 09/10/09 408 2 2.13 452 0.0 27.4 17784 3143 632 20.3 649 12/21/09 508 1 0.23 1165 34.3 41.7 21845 4229 386 29.1 710 12/21/09 508 2 0.19 1085 41.5 33.9 22597 4513 378 41.2 648 12/21/09 508 3 0.24 970 25.8 40.3 21349 4514 394 26.5 733 12/29/09 608 1 1.18 6476 9.9 17.9 12778 3732 183 12.9 387 12/29/09 608 2 2.25 8970 14.3 11.5 10273 3456 216 11.8 471 12/29/09 608 3 1.84 2895 13.1 17.1 14522 3618 184 13.6 465 01/07/10 708 1 4.14 3273 5.1 7.5 7143 2962 136 4.5 488 01/07/10 708 2 3.86 1607 19.8 13.4 10508 3278 245 12.7 561 01/07/10 708 3 4.02 1378 19.3 19.9 9734 3876 270 28.0 601 01/18/10 808 1 6.76 11791 1.6 9.3 6824 3541 152 5.3 484 01/18/10 808 2 11.54 5362 13.3 14.3 15107 3661 254 10.2 528 01/18/10 808 3 4.63 1559 25.9 23.4 20240 3841 329 19.1 583 04/26/10 908 1 4.06 423 0.0 23.3 18119 3035 346 13.4 269 04/26/10 908 2 3.72 961 15.9 29.1 17877 3884 355 17.5 470 04/26/10 908 3 7.57 1003 13.6 25.9 17689 3713 332 18.0 460 05/04/10 1008 1 6.94 1425 56.4 25.3 17020 3713 362 39.6 449 05/04/10 1008 2 7.29 2304 29.1 30.5 19657 3959 492 20.6 455 05/04/10 1008 3 6.42 3909 31.0 24.8 21459 3400 427 17.1 439 05/12/10 1108 1 4.63 591 264.8 25.7 22570 3280 418 146.5 519 05/12/10 1108 2 7.34 1615 103.7 32.1 24691 4032 501 58.0 520 05/12/10 1108 3 6.46 305 19.0 21.6 19513 2912 384 20.8 379 05/20/10 1208 1 21.54 1346 35.7 23.2 26852 3336 411 16.3 509 05/20/10 1208 2 4.51 2314 34.9 24.1 25566 3483 363 19.7 432 05/20/10 1208 3 5.68 1153 45.9 22.9 25216 3323 343 26.4 625  112 Site Number Site Description Date Sample ID Tile Replicate Chl-a (mg/L) Dry mass (mg) Cr (mg/kg) Cu (mg/kg) Fe (mg/kg) Mg (mg/kg) Mn (mg/kg) Ni (mg/kg) P (mg/kg) 22 Agricultural tributary - Kenny Rd. 07/13/09 22 1 28.71 380 64.9 26.8 41684 11029 408 95.5 2698 07/13/09 22 2 24.74 220 66.4 25.5 38132 10509 401 92.5 2097 08/18/09 122 1 14.79 56 0.0 28.3 13239 1299 1217 41.2 3266 08/18/09 122 2 10.34 58 0.0 30.4 14231 1450 1255 40.6 3189 08/26/09 222 1 59.80 93 0.0 4.7 23002 927 489 1.7 2857 08/26/09 222 2 45.04 95 0.0 50.3 13248 189 415 9.3 2296 09/03/09 322 1 41.46 110 0.0 6.9 17642 1205 421 14.9 2575 09/03/09 322 2 1.18 115 0.0 21.5 20975 1261 654 23.3 3642 09/10/09 422 1 21.86 126 0.0 5.1 18304 1424 370 19.3 3147 09/10/09 422 2 15.23 91 14.9 3.7 13176 1427 416 43.6 2000 12/21/09 522 1 0.15 512 16.8 31.3 27625 6677 300 61.2 2421 12/21/09 522 2 0.09 1088 31.2 36.7 29373 7952 283 74.5 2088 12/21/09 522 3 0.14 737 23.9 32.9 29741 7542 286 65.8 2141 12/29/09 622 1 0.08 1398 12.1 17.1 18801 6007 168 40.4 1027 12/29/09 622 2 0.29 1367 6.3 9.9 9093 5296 127 21.8 1370 12/29/09 622 3 0.09 703 3.2 13.4 13844 4591 106 24.8 820 01/07/10 722 1 0.15 2355 19.5 24.0 23419 7475 208 56.1 1691 01/07/10 722 2 0.20 3337 7.0 12.0 10380 6604 137 38.5 1634 01/07/10 722 3 0.17 3399 14.9 21.3 17074 6791 155 44.5 1443 01/18/10 822 1 0.56 3299 8.3 16.4 13069 6359 131 35.2 1870 01/18/10 822 2 0.44 7794 0.0 22.6 29224 8102 298 64.2 1434 01/18/10 822 3 0.34 3556 17.2 17.4 11298 6575 163 45.0 1813 04/26/10 922 1 0.62 100 0.0 7.4 25596 3377 338 16.3 793 04/26/10 922 2 9.07 214 0.0 15.3 32424 4448 303 31.4 1436 04/26/10 922 3 2.38 114 0.0 8.3 28698 3307 419 13.5 1052 05/04/10 1022 1 1.25 177 13.0 18.2 32908 4010 1038 50.0 1421 05/04/10 1022 2 0.59 263 73.6 28.8 41566 5289 959 89.4 2166 05/04/10 1022 3 1.13 128 0.0 16.8 28910 3551 862 35.9 1322 05/12/10 1122 1 3.22 266 13.2 0.5 8686 743 164 24.6 310 05/12/10 1122 2 3.24 348 98.2 29.7 44554 5843 1524 105.7 2257 05/12/10 1122 3 2.42 243 179.5 37.1 42778 4897 904 143.4 2010 05/20/10 1222 1 13.30 230 13.4 25.7 44514 4680 657 47.6 2400 05/20/10 1222 2 5.51 225 398.3 24.3 41943 4766 385 260.3 2186 05/20/10 1222 3 6.14 360 69.9 33.4 43333 6169 395 90.7 2433  113 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.  114 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 NO3--N Corr. Coef. 1 0.134 -0.297 -0.132 -0.016 -0.178 -0.266 0.366 -0.134 -0.306 Sig. (2-tailed) . 0.137 0.002 0.168 0.859 0.047 0.006 0.000 0.281 0.002 N 124 124 110 110 124 124 104 95 67 99 SRP Corr. Coef. 0.134 1 0.365 0.524 -0.402 0.261 -0.066 -0.494 0.502 0.484 Sig. (2-tailed) 0.137 . 0.000 0.000 0.000 0.003 0.506 0.000 0.000 0.000 N 124 124 110 110 124 124 104 95 67 99 Cl- Corr. Coef. -0.297 0.365 1 0.493 -0.410 0.582 -0.146 -0.468 0.683 0.644 Sig. (2-tailed) 0.002 0.000 . 0.000 0.000 0.000 0.166 0.000 0.000 0.000 N 110 110 110 110 110 110 91 81 67 99 NH4+-N Corr. Coef. -0.132 0.524 0.493 1 -0.506 0.244 -0.229 -0.753 0.721 0.641 Sig. (2-tailed) 0.168 0.000 0.000 . 0.000 0.010 0.029 0.000 0.000 0.000 N 110 110 110 110 110 110 91 81 67 99 pH Corr. Coef. -0.016 -0.402 -0.410 -0.506 1 -0.143 0.174 0.482 -0.431 -0.607 Sig. (2-tailed) 0.859 0.000 0.000 0.000 . 0.113 0.078 0.000 0.000 0.000 N 124 124 110 110 124 124 104 95 67 99 Sp. Cond Corr. Coef. -0.178 0.261 0.582 0.244 -0.143 1 0.355 -0.345 0.339 0.337 Sig. (2-tailed) 0.047 0.003 0.000 0.010 0.113 . 0.000 0.001 0.005 0.001 N 124 124 110 110 124 124 104 95 67 99 Temp. Corr. Coef. -0.266 -0.066 -0.146 -0.229 0.174 0.355 1 -0.030 -0.236 0.064 Sig. (2-tailed) 0.006 0.506 0.166 0.029 0.078 0.000 . 0.780 0.062 0.571 N 104 104 91 91 104 104 104 90 63 80 DO Corr. Coef. 0.366 -0.494 -0.468 -0.753 0.482 -0.345 -0.030 1 -0.581 -0.775 Sig. (2-tailed) 0.000 0.000 0.000 0.000 0.000 0.001 0.780 . 0.000 0.000 N 95 95 81 81 95 95 90 95 67 70 Turb. Corr. Coef. -0.134 0.502 0.683 0.721 -0.431 0.339 -0.236 -0.581 1 0.704 Sig. (2-tailed) 0.281 0.000 0.000 0.000 0.000 0.005 0.062 0.000 . 0.000 N 67 67 67 67 67 67 63 67 67 56 TOC Corr. Coef. -0.306 0.484 0.644 0.641 -0.607 0.337 0.064 -0.775 0.704 1 Sig. (2-tailed) 0.002 0.000 0.000 0.000 0.000 0.001 0.571 0.000 0.000 . N 99 99 99 99 99 99 80 70 56 99    115 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 NO3--N Corr. Coef. 1 -0.282 -0.337 -0.460 0.151 -0.057 -0.146 0.678 -0.372 -0.574 Sig. (2-tailed) . 0.019 0.008 0.000 0.216 0.641 0.273 0.000 0.020 0.000 N 69 69 61 61 69 69 58 54 39 56 SRP Corr. Coef. -0.282 1 0.360 0.513 -0.286 0.184 -0.167 -0.564 0.368 0.567 Sig. (2-tailed) 0.019 . 0.004 0.000 0.017 0.131 0.210 0.000 0.021 0.000 N 69 69 61 61 69 69 58 54 39 56 Cl- Corr. Coef. -0.337 0.360 1 0.429 -0.373 0.370 -0.432 -0.322 0.575 0.543 Sig. (2-tailed) 0.008 0.004 . 0.001 0.003 0.003 0.002 0.029 0.000 0.000 N 61 61 61 61 61 61 51 46 39 56 NH4+-N Corr. Coef. -0.460 0.513 0.429 1 -0.268 0.006 -0.382 -0.549 0.466 0.566 Sig. (2-tailed) 0.000 0.000 0.001 . 0.037 0.966 0.006 0.000 0.003 0.000 N 61 61 61 61 61 61 51 46 39 56 pH Corr. Coef. 0.151 -0.286 -0.373 -0.268 1 -0.066 0.277 0.312 -0.300 -0.561 Sig. (2-tailed) 0.216 0.017 0.003 0.037 . 0.589 0.035 0.022 0.063 0.000 N 69 69 61 61 69 69 58 54 39 56 Sp. Cond Corr. Coef. -0.057 0.184 0.370 0.006 -0.066 1 0.450 -0.250 0.197 0.147 Sig. (2-tailed) 0.641 0.131 0.003 0.966 0.589 . 0.000 0.069 0.230 0.279 N 69 69 61 61 69 69 58 54 39 56 Temp. Corr. Coef. -0.146 -0.167 -0.432 -0.382 0.277 0.450 1 0.001 -0.361 -0.087 Sig. (2-tailed) 0.273 0.210 0.002 0.006 0.035 0.000 . 0.997 0.028 0.564 N 58 58 51 51 58 58 58 51 37 46 DO Corr. Coef. 0.678 -0.564 -0.322 -0.549 0.312 -0.250 0.001 1 -0.265 -0.668 Sig. (2-tailed) 0.000 0.000 0.029 0.000 0.022 0.069 0.997 . 0.103 0.000 N 54 54 46 46 54 54 51 54 39 41 Turb. Corr. Coef. -0.372 0.368 0.575 0.466 -0.300 0.197 -0.361 -0.265 1 0.508 Sig. (2-tailed) 0.020 0.021 0.000 0.003 0.063 0.230 0.028 0.103 . 0.002 N 39 39 39 39 39 39 37 39 39 34 TOC Corr. Coef. -0.574 0.567 0.543 0.566 -0.561 0.147 -0.087 -0.668 0.508 1 Sig. (2-tailed) 0.000 0.000 0.000 0.000 0.000 0.279 0.564 0.000 0.002 . N 56 56 56 56 56 56 46 41 34 56    116 Table E3. Spearman rank correlation coefficients for biofilm and water quality data collected at three sites in the Marshall Creek watershed.     Biofilms Water Chl- a Dry mass Cr Cu Fe Mg Mn Ni P NO3- -N SRP Cl- NH4+- N pH Sp. Cond Temp. DO Turb. TOC Depth Chl-a  Corr. Coef. 1 -0.356 -0.163 0.040 0.136 -0.421 0.373 -0.080 0.264 -0.485 -0.449 -0.436 -0.554 0.741 -0.056 0.824 0.187 -0.571 -0.379 .449* Sig.  . 0.028 0.328 0.810 0.416 0.008 0.021 0.631 0.110 0.019 0.032 0.038 0.006 0.000 0.799 0.000 0.393 0.006 0.164 0.017 N 38 38 38 38 38 38 38 38 38 23 23 23 23 23 23 23 23 22 15 28 Dry mass Corr. Coef. -0.356 1 0.408 -0.246 0.022 0.518 -0.743 -0.163 -0.573 0.221 -0.238 -0.339 -0.393 0.153 -0.246 -0.372 0.328 -0.359 -0.352 -0.221 Sig.  0.028 . 0.011 0.137 0.895 0.001 0.000 0.328 0.000 0.310 0.274 0.114 0.063 0.485 0.258 0.080 0.126 0.101 0.198 0.258 N 38 38 38 38 38 38 38 38 38 23 23 23 23 23 23 23 23 22 15 28 Cr Corr. Coef. -0.163 0.408 1 0.171 0.553 0.574 -0.201 0.507 -0.178 -0.319 -0.230 -0.059 -0.175 0.227 0.164 0.399 -0.053 -0.095 -0.043 -0.062 Sig.  0.328 0.011 . 0.305 0.000 0.000 0.225 0.001 0.286 0.138 0.292 0.790 0.423 0.297 0.454 0.059 0.809 0.673 0.879 0.753 N 38 38 38 38 38 38 38 38 38 23 23 23 23 23 23 23 23 22 15 28 Cu Corr. Coef. 0.040 -0.246 0.171 1 0.371 0.137 0.512 0.513 0.258 0.136 -0.051 -0.208 -0.025 0.209 -0.199 0.219 -0.076 -0.065 0.023 0.108 Sig.  0.810 0.137 0.305 . 0.022 0.413 0.001 0.001 0.118 0.535 0.816 0.342 0.911 0.338 0.364 0.316 0.730 0.774 0.934 0.584 N 38 38 38 38 38 38 38 38 38 23 23 23 23 23 23 23 23 22 15 28 Fe Corr. Coef. 0.136 0.022 0.553 0.371 1 0.605 0.212 0.616 0.214 -0.265 0.039 0.152 0.075 0.092 0.335 0.537 -0.322 0.269 0.102 0.047 Sig.  0.416 0.895 0.000 0.022 . 0.000 0.201 0.000 0.196 0.222 0.861 0.488 0.733 0.675 0.118 0.008 0.134 0.225 0.718 0.812 N 38 38 38 38 38 38 38 38 38 23 23 23 23 23 23 23 23 22 15 28 Mg Corr. Coef. -0.421 0.518 0.574 0.137 0.605 1 -0.297 0.552 -0.012 0.260 0.371 0.511 0.362 -0.457 0.413 -0.207 -0.583 0.612 0.502 -0.196 Sig.  0.008 0.001 0.000 0.413 0.000 . 0.070 0.000 0.942 0.231 0.082 0.013 0.090 0.028 0.050 0.344 0.003 0.002 0.056 0.318 N 38 38 38 38 38 38 38 38 38 23 23 23 23 23 23 23 23 22 15 28 Mn Corr. Coef. 0.373 -0.743 -0.201 0.512 0.212 -0.297 1 0.390 0.409 -0.211 0.007 -0.126 -0.015 0.278 -0.087 0.624 -0.130 -0.011 -0.034 0.412 Sig.  0.021 0.000 0.225 0.001 0.201 0.070 . 0.015 0.011 0.333 0.975 0.565 0.946 0.199 0.693 0.001 0.556 0.962 0.904 0.029 N 38 38 38 38 38 38 38 38 38 23 23 23 23 23 23 23 23 22 15 28 Ni Corr. Coef. -0.080 -0.163 0.507 0.513 0.616 0.552 0.390 1 0.495 0.045 0.252 0.460 0.369 -0.237 0.426 0.223 -0.625 0.522 0.377 0.094 Sig.  0.631 0.328 0.001 0.001 0.000 0.000 0.015 . 0.002 0.837 0.246 0.027 0.084 0.277 0.043 0.307 0.001 0.013 0.166 0.636 N 38 38 38 38 38 38 38 38 38 23 23 23 23 23 23 23 23 22 15 28 P Corr. Coef. 0.264 -0.573 -0.178 0.258 0.214 -0.012 0.409 0.495 1 0.215 0.640 0.752 0.658 -0.638 0.493 -0.073 -0.790 0.730 0.817 0.226 Sig.  0.110 0.000 0.286 0.118 0.196 0.942 0.011 0.002 . 0.324 0.001 0.000 0.001 0.001 0.017 0.742 0.000 0.000 0.000 0.247 N 38 38 38 38 38 38 38 38 38 23 23 23 23 23 23 23 23 22 15 28 NO3--N Corr. Coef. -0.485 0.221 -0.319 0.136 -0.265 0.260 -0.211 0.045 0.215 1 0.551 0.002 0.311 -0.302 -0.260 -0.568 -0.207 0.003 0.147 -0.051 Sig.  0.019 0.310 0.138 0.535 0.222 0.231 0.333 0.837 0.324 . 0.006 0.993 0.148 0.161 0.231 0.005 0.343 0.990 0.615 0.844 N 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 22 14 17  117     Biofilms Water Chl- a Dry mass Cr Cu Fe Mg Mn Ni P NO3- -N SRP Cl- NH4+- N pH Sp. Cond Temp. DO Turb. TOC Depth SRP Corr. Coef. -0.449 -0.238 -0.230 -0.051 0.039 0.371 0.007 0.252 0.640 0.551 1 0.481 0.734 -0.767 0.214 -0.331 -0.610 0.529 0.662 -0.414 Sig.  0.032 0.274 0.292 0.816 0.861 0.082 0.975 0.246 0.001 0.006 . 0.020 0.000 0.000 0.326 0.122 0.002 0.011 0.010 0.098 N 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 22 14 17 Cl- Corr. Coef. -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 Sig.  0.038 0.114 0.790 0.342 0.488 0.013 0.565 0.027 0.000 0.993 0.020 . 0.000 0.000 0.000 0.337 0.000 0.000 0.000 0.020 N 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 22 14 17 NH4+- N Corr. Coef. -0.554 -0.393 -0.175 -0.025 0.075 0.362 -0.015 0.369 0.658 0.311 0.734 0.711 1 -0.796 0.365 -0.342 -0.654 0.656 0.662 -0.473 Sig.  0.006 0.063 0.423 0.911 0.733 0.090 0.946 0.084 0.001 0.148 0.000 0.000 . 0.000 0.087 0.110 0.001 0.001 0.010 0.055 N 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 22 14 17 pH Corr. Coef. 0.741 0.153 0.227 0.209 0.092 -0.457 0.278 -0.237 -0.638 -0.302 -0.767 -0.751 -0.796 1 -0.381 0.567 0.597 -0.799 -0.829 0.456 Sig.  0.000 0.485 0.297 0.338 0.675 0.028 0.199 0.277 0.001 0.161 0.000 0.000 0.000 . 0.073 0.005 0.003 0.000 0.000 0.066 N 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 22 14 17 Sp. Cond Corr. Coef. -0.056 -0.246 0.164 -0.199 0.335 0.413 -0.087 0.426 0.493 -0.260 0.214 0.743 0.365 -0.381 1 0.134 -0.554 0.622 0.618 -0.591 Sig.  0.799 0.258 0.454 0.364 0.118 0.050 0.693 0.043 0.017 0.231 0.326 0.000 0.087 0.073 . 0.542 0.006 0.002 0.019 0.013 N 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 22 14 17 Temp. Corr. Coef. 0.824 -0.372 0.399 0.219 0.537 -0.207 0.624 0.223 -0.073 -0.568 -0.331 -0.210 -0.342 0.567 0.134 1 -0.055 -0.280 -0.189 0.227 Sig.  0.000 0.080 0.059 0.316 0.008 0.344 0.001 0.307 0.742 0.005 0.122 0.337 0.110 0.005 0.542 . 0.804 0.207 0.517 0.381 N 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 22 14 17 DO Corr. Coef. 0.187 0.328 -0.053 -0.076 -0.322 -0.583 -0.130 -0.625 -0.790 -0.207 -0.610 -0.704 -0.654 0.597 -0.554 -0.055 1 -0.653 -0.849 0.270 Sig.  0.393 0.126 0.809 0.730 0.134 0.003 0.556 0.001 0.000 0.343 0.002 0.000 0.001 0.003 0.006 0.804 . 0.001 0.000 0.295 N 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 22 14 17 Turb. Corr. Coef. -0.571 -0.359 -0.095 -0.065 0.269 0.612 -0.011 0.522 0.730 0.003 0.529 0.810 0.656 -0.799 0.622 -0.280 -0.653 1 0.852 -0.449 Sig.  0.006 0.101 0.673 0.774 0.225 0.002 0.962 0.013 0.000 0.990 0.011 0.000 0.001 0.000 0.002 0.207 0.001 . 0.000 0.071 N 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22 13 17 TOC Corr. Coef. -0.379 -0.352 -0.043 0.023 0.102 0.502 -0.034 0.377 0.817 0.147 0.662 0.833 0.662 -0.829 0.618 -0.189 -0.849 0.852 1 -0.342 Sig.  0.164 0.198 0.879 0.934 0.718 0.056 0.904 0.166 0.000 0.615 0.010 0.000 0.010 0.000 0.019 0.517 0.000 0.000 . 0.303 N 15 15 15 15 15 15 15 15 15 14 14 14 14 14 14 14 14 13 15 11 Water depth Corr. Coef. 0.449 -0.221 -0.062 0.108 0.047 -0.196 0.412 0.094 0.226 -0.051 -0.414 -0.559 -0.473 0.456 -0.591 0.227 0.270 -0.449 -0.342 1 Sig.  0.017 0.258 0.753 0.584 0.812 0.318 0.029 0.636 0.247 0.844 0.098 0.020 0.055 0.066 0.013 0.381 0.295 0.071 0.303 . N 28 28 28 28 28 28 28 28 28 17 17 17 17 17 17 17 17 17 11 28   118 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 Ni P Pb Zn Cum. Precip. Dry weight Corr. Coef. 1 0.163 -0.360 -0.229 -0.235 -0.058 -0.723 0.251 -0.296 0.095 -0.249 -0.418 0.258 0.199 0.244 -0.126 0.144 Sig. (2-tailed) . 0.468 0.100 0.306 0.293 0.797 0.000 0.259 0.180 0.673 0.264 0.229 0.246 0.374 0.273 0.577 0.524 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Al Corr. Coef. 0.163 1 0.660 -0.105 -0.185 -0.089 -0.031 0.602 0.704 0.553 -0.015 0.818 0.370 0.693 0.462 0.680 0.167 Sig. (2-tailed) 0.468 . 0.001 0.643 0.411 0.695 0.891 0.003 0.000 0.008 0.946 0.004 0.090 0.000 0.030 0.000 0.457 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Ca Corr. Coef. -0.360 0.660 1 0.279 0.180 0.198 0.401 0.276 0.871 0.532 0.225 0.794 0.213 0.324 0.147 0.513 0.002 Sig. (2-tailed) 0.100 0.001 . 0.209 0.423 0.377 0.064 0.214 0.000 0.011 0.313 0.006 0.342 0.142 0.513 0.015 0.994 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Cd Corr. Coef. -0.229 -0.105 0.279 1 0.922 0.838 0.609 0.017 0.153 0.299 0.360 -0.112 0.351 -0.005 -0.230 0.033 -0.132 Sig. (2-tailed) 0.306 0.643 0.209 . 0.000 0.000 0.003 0.940 0.498 0.177 0.100 0.757 0.109 0.981 0.304 0.883 0.558 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Co Corr. Coef. -0.235 -0.185 0.180 0.922 1 0.765 0.506 0.047 0.102 0.263 0.375 -0.407 0.408 0.023 -0.305 0.046 -0.234 Sig. (2-tailed) 0.293 0.411 0.423 0.000 . 0.000 0.016 0.835 0.653 0.237 0.085 0.243 0.059 0.918 0.168 0.839 0.295 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Cr Corr. Coef. -0.058 -0.089 0.198 0.838 0.765 1 0.547 -0.119 0.083 0.189 0.213 -0.079 0.119 -0.258 -0.308 -0.142 -0.170 Sig. (2-tailed) 0.797 0.695 0.377 0.000 0.000 . 0.008 0.597 0.713 0.399 0.342 0.829 0.597 0.246 0.164 0.529 0.451 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Cu Corr. Coef. -0.723 -0.031 0.401 0.609 0.506 0.547 1 -0.019 0.298 0.146 0.420 0.309 0.004 -0.098 -0.024 0.266 -0.183 Sig. (2-tailed) 0.000 0.891 0.064 0.003 0.016 0.008 . 0.934 0.179 0.516 0.052 0.385 0.986 0.665 0.915 0.232 0.414 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Fe Corr. Coef. 0.251 0.602 0.276 0.017 0.047 -0.119 -0.019 1 0.244 0.580 0.522 0.370 0.720 0.732 0.805 0.787 0.085 Sig. (2-tailed) 0.259 0.003 0.214 0.940 0.835 0.597 0.934 . 0.273 0.005 0.013 0.293 0.000 0.000 0.000 0.000 0.708 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 K Corr. Coef. -0.296 0.704 0.871 0.153 0.102 0.083 0.298 0.244 1 0.607 -0.028 0.818 0.204 0.461 0.080 0.455 -0.132 Sig. (2-tailed) 0.180 0.000 0.000 0.498 0.653 0.713 0.179 0.273 . 0.003 0.903 0.004 0.363 0.031 0.725 0.034 0.558 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Mg Corr. Coef. 0.095 0.553 0.532 0.299 0.263 0.189 0.146 0.580 0.607 1 0.083 0.345 0.679 0.565 0.312 0.343 0.043 Sig. (2-tailed) 0.673 0.008 0.011 0.177 0.237 0.399 0.516 0.005 0.003 . 0.713 0.328 0.001 0.006 0.157 0.118 0.849 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Mn Corr. Coef. -0.249 -0.015 0.225 0.360 0.375 0.213 0.420 0.522 -0.028 0.083 1 -0.164 0.318 0.036 0.532 0.459 0.130 Sig. (2-tailed) 0.264 0.946 0.313 0.100 0.085 0.342 0.052 0.013 0.903 0.713 . 0.651 0.149 0.875 0.011 0.032 0.563 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Na Corr. Coef. -0.418 0.818 0.794 -0.112 -0.407 -0.079 0.309 0.370 0.818 0.345 -0.164 1 0.030 0.503 0.406 0.455 -0.088 Sig. (2-tailed) 0.229 0.004 0.006 0.757 0.243 0.829 0.385 0.293 0.004 0.328 0.651 . 0.934 0.138 0.244 0.187 0.810 N 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10  119  Dry weight  Al Ca Cd Co Cr Cu Fe K Mg Mn Na Ni P Pb Zn Cum. Precip. Ni Corr. Coef. 0.258 0.370 0.213 0.351 0.408 0.119 0.004 0.720 0.204 0.679 0.318 0.030 1 0.717 0.461 0.521 0.004 Sig. (2-tailed) 0.246 0.090 0.342 0.109 0.059 0.597 0.986 0.000 0.363 0.001 0.149 0.934 . 0.000 0.031 0.013 0.986 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 P Corr. Coef. 0.199 0.693 0.324 -0.005 0.023 -0.258 -0.098 0.732 0.461 0.565 0.036 0.503 0.717 1 0.495 0.732 -0.003 Sig. (2-tailed) 0.374 0.000 0.142 0.981 0.918 0.246 0.665 0.000 0.031 0.006 0.875 0.138 0.000 . 0.019 0.000 0.988 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Pb Corr. Coef. 0.244 0.462 0.147 -0.230 -0.305 -0.308 -0.024 0.805 0.080 0.312 0.532 0.406 0.461 0.495 1 0.649 0.237 Sig. (2-tailed) 0.273 0.030 0.513 0.304 0.168 0.164 0.915 0.000 0.725 0.157 0.011 0.244 0.031 0.019 . 0.001 0.288 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Zn Corr. Coef. -0.126 0.680 0.513 0.033 0.046 -0.142 0.266 0.787 0.455 0.343 0.459 0.455 0.521 0.732 0.649 1 -0.058 Sig. (2-tailed) 0.577 0.000 0.015 0.883 0.839 0.529 0.232 0.000 0.034 0.118 0.032 0.187 0.013 0.000 0.001 . 0.798 N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22 Cum. Precip. Corr. Coef. 0.144 0.167 0.002 -0.132 -0.234 -0.170 -0.183 0.085 -0.132 0.043 0.130 -0.088 0.004 -0.003 0.237 -0.058 1 Sig. (2-tailed) 0.524 0.457 0.994 0.558 0.295 0.451 0.414 0.708 0.558 0.849 0.563 0.810 0.986 0.988 0.288 0.798 . N 22 22 22 22 22 22 22 22 22 22 22 10 22 22 22 22 22   120 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.  121  Figure F1. Discharge near Fraser Valley Trout Hatchery into Marshall Creek (site 14).  122  Figure F2. Site 11 where water and bed sediments were sampled between 1993-2010.  123  Figure F3. Suspended sediment sampler, water level probe and tiles for biofilms at the agricultural tributary (site 22).  124  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).  125   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.  126  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.  127  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.    128  Figure F12. Forest clearing and urban residential development on Sumas Mountain in the Marshall Creek watershed.  

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0071509/manifest

Comment

Related Items