Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada

Journal of Great Lakes Research xxx (xxxx) xxx

Contents lists available at ScienceDirect

Journal of Great Lakes Research journal homepage: www.elsevier.com/locate/ijglr

Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada Cameron Irvine, Merrin Macrae ⇑, Matthew Morison, Richard Petrone University of Waterloo, Department of Geography and Environmental Management, 200 University Avenue W, Waterloo, Ontario N2L 3G1, Canada

a r t i c l e

i n f o

Article history: Received 23 January 2019 Accepted 3 October 2019 Available online xxxx Communicate by Joseph Makarewicz

Keywords: Phosphorus Nitrate Tile drainage Hot spots Hot moments Agriculture

a b s t r a c t Algal blooms in the Great Lakes are a concern due to excess nutrient loading from non-point sources; however, there is uncertainty over the relative contributions of various non-point sources under different types of land use in rural watersheds, particularly over annual time scales. Four nested subwatersheds in Southern Ontario, Canada (one natural woodlot, two agricultural and one mixed agricultural and urban) were monitored over one year to identify peak periods (‘hot moments’) and areas (‘hot spots’) of nutrient (dissolved reactive phosphorus, DRP; total phosphorus, TP; and nitrate, NO3 ) export and discharge. Annual nutrient export was small at the natural site (0.001 kg DRP ha 1; 0.004 kg TP ha 1; 1 ) compared to the agricultural and mixed-use sites (0.10–0.15 kg DRP ha 1; 0.70– 0.04 kg NO— 3 N ha 1 ). Temporal patterns in P concentrations were similar through0.94 kg TP ha 1; 9.15–11.55 kg NO— 3 N ha out the sites, where spring was the dominant season for P export, irrespective of land use. Within the Hopewell Creek watershed, P and N hot spots existed that were consistently hot spots across all events with the location of these hot spots driven by local land use patterns, where there was elevated P export from a dairy-dominated sub-watershed and elevated N export from both of the two agricultural subwatersheds. These estimates of seasonal- and event-based nutrient loads and discharge across nested sub-watersheds contribute to the growing body of evidence demonstrating the importance of identifying critical areas and periods in which to emphasize management efforts. Ó 2019 International Association for Great Lakes Research. Published by Elsevier B.V. All rights reserved.

Introduction The eutrophication of surface water bodies worldwide has detrimental impacts on aquatic ecosystems (Banner et al., 2009; Carpenter et al., 1998). Enhanced growth of algae and cyanobacteria, caused by elevated nutrient loads from watersheds, results in lake-bottom hypoxia (Scavia et al., 2014) and can result in fish mortality (Sharpley et al., 2001) and can also be associated with health risks to humans and animals. In the Great Lakes region of North America, Lake Erie is particularly vulnerable to these processes and experiences frequent algal blooms (IJC, 2012, 2014; Michalak et al., 2013) due to the increased tributary soluble reactive phosphorus loads from agriculturally-dominated tributaries (Kane et al., 2014). Management efforts resulting in targeted point-source load reductions initiated in 1972 resulted in lower phosphorus concentrations, leading to reduced phytoplankton biomass and associated bottom-water hypoxia (Makarewicz and Bertram, 1991). However, there has been a recent re-

⇑ Corresponding author.

eutrophication of Lake Erie beginning in the mid-1990s and continuing until present day (Watson et al., 2016) attributed to increases in both total and dissolved reactive phosphorus loads from nonpoint sources (Jarvie et al., 2017; Scavia et al., 2014; Kane et al., 2014), although the role of nitrogen dynamics may play a key role as well (Steffen et al., 2014). There is significant pressure to reduce the occurrences of harmful and nuisance algal blooms; however, the drivers of these blooms in large systems such as Lake Erie are complex and may vary in space and time. Non-point sources of pollution, particularly agricultural inputs of nutrients, have been identified as one of the largest contributors of nutrients downstream, exporting both nitrogen (N) and phosphorus (P) (IJC, 2014; Arbuckle and Downing, 2001). An improved understanding of spatial and temporal variability in nutrient loads, chemical speciation, and the relative importance of drivers such as climate, land use, and management practices in controlling these loads is required to aid both managers and modellers with identifying critical sources and periods for nutrient loading to the lake. This information will also provide insight into the potential to achieve the objectives of the Nutrients Annex (Annex 4) of the 2012 Great Lakes Water

E-mail address: [email protected] (M. Macrae). https://doi.org/10.1016/j.jglr.2019.10.005 0380-1330/Ó 2019 International Association for Great Lakes Research. Published by Elsevier B.V. All rights reserved.

Please cite this article as: C. Irvine, M. Macrae, M. Morison et al., Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada, Journal of Great Lakes Research, https://doi.org/10.1016/j.jglr.2019.10.005

2

C. Irvine et al. / Journal of Great Lakes Research xxx (xxxx) xxx

Quality Agreement and the targets set in 2016 under this agreement. Watershed-scale characteristics, such as land use, slope, soil texture and tile drain density, all impact nutrient export loads; but it is not clear if such drivers consistently lead to elevated loads or if this is restricted to specific events or periods. Nor is it clear if peak periods differ for different land uses. Many studies in agricultural systems have examined nutrient export at the field or plot scale (Lam et al., 2016a; Macrae et al., 2011; Rozemeijer et al., 2010; Stenberg et al., 2012; Van Esbroeck et al., 2016, 2017; Williams et al., 2016), which provides insight into the field-scale mechanisms driving nutrient fluxes. However, observations at small scales (i.e., plots, fields or sub-watersheds) are not always consistent with patterns observed at watershed outlets. More recently, there has been a need to determine how smaller-scale mechanisms scale to watershed outlets (Baker et al., 2017; Beckert et al., 2011; Evans et al., 2014; Jarvie et al., 2017; Mineau et al., 2015). Studies that have investigated relationships between land use and nutrient concentrations or export in watersheds that have multiple land uses have often employed a geographic information systems (GIS) approach (Agnew et al., 2006). For example, Evans et al. (2014) found a strong correlation between the amount of agricultural land use and the concentration of dissolved nutrients in Oregon, USA. A similar relationship was reported by Beckert et al. (2011) in Maryland, particularly in areas with high density of animal feeding operations. The watersheds that had the highest proportion of row crop agriculture had a strong correlation with mean baseflow total nitrogen (TN) concentrations (Beckert et al., 2011). Studies that investigate relationships between land use type and nutrient export require many watersheds within the same physiographic area and often necessitate pre-existing datasets that may be temporally coarse (Mehaffey et al., 2005). The influence of land use on nutrient export can be scale-dependent where large, higher-order streams can be impacted by land use far upstream in the headwater reaches, while smaller first- and second-order streams are strongly impacted by land use that is directly adjacent (Buck et al., 2004). Horizontal subsurface export of dissolved reactive P (DRP) via tile drainage is of particular concern as this bioavailable form of P is rapidly transported from tile drains into tributaries and subsequently into lakes (King et al., 2015; Vidon and Cuadra, 2011). Tile drain density in a watershed may also contribute to N loads due to the solubility and mobility of NO3 through vertical leaching, entering the tile drains through either matrix (Li et al., 2010) or macropore flow (Perks et al., 2015). At smaller scales, paired watershed studies that directly compare subcatchments with similar physiographic characteristics can provide insight into local controls on water quality and nutrient export. For example, Coulter et al. (2004) found that a Kentucky watershed with primarily agricultural land use exported significantly higher NO3 and DRP concentrations compared to mixed and urban watersheds, which had higher temperatures and turbidity. Pieterse et al. (2003) reported similar results in The Netherlands and Belgium, where many agricultural tributaries exceeded water quality standards for both TN and TP. Although paired watershed studies such as these are rare, they not only provide important insight into how subtle differences in land use may drive elevated nutrient loads, but also how nutrient loads may evolve throughout a watershed. The magnitude and speciation of nutrient loads also vary temporally, particularly in cold regions with nival (snowmeltdominated) flow regimes. P export in agricultural catchments can be highly episodic, and loss is largely event-based (Chen et al., 2015; Macrae et al., 2007a). For example, in a Kansas watershed, Banner et al. (2009) reported that 88% of TP loss occurred during high discharge events covering only 10% of the study time. In an Ontario study, Macrae et al. (2007a) reported that 70–80% of

annual TP loss and 82–83% of annual DRP loss were associated with storm events, and 38% of DRP loss and 32% of TP loss over a 2-year period occurred via three peak flow events. NO3 losses are less episodic than P, although they tend to increase under higher discharge events, particularly following fertilizer application (Macrae et al., 2007a), and generally have a strong relationship with discharge (Liu et al., 2014). These findings point to the need to use intensive, storm-based sampling procedures as opposed to regular-interval sampling methods when estimating nutrient losses and evaluating the impacts of land use on nutrient export (Grant et al., 1996; Williams et al., 2015). Sampling programs should also include the non-growing season, particularly the winter snowmelt period, as large exports of nutrients tend to occur during snowmelt events (Algoazany et al., 2007; Ford et al., 2018; Macrae et al., 2007a; Van Esbroeck et al., 2017), in the spring months due to large rain events (Vidon and Cuadra, 2011) or when rainfall occurs on soils with high antecedent moisture contents (Macrae et al., 2010). The role of winter weather has been identified as playing an important role in controlling riverine TP and DRP export to Lake Erie (Baker and Richards, 2002). Work from Ohio in the Maumee River basin using long-term data from the National Center for Water Quality Research has shown that, in particular, late winter/early spring loads have been increasing from 1990 to 2014 (Stow et al., 2015). At present, there is a paucity of field data collected during the winter period to compare with existing data on spring and summer loads. Although the episodic nature of nutrient export has been shown previously, an evaluation of whether or not such patterns are consistent across land uses and throughout nested watersheds has not been done. This research design uses an intensive, event-based runoff sampling approach to investigate spatiotemporal variability in agricultural nutrient export in a mixed land use watershed in Southern Ontario, Canada, within the Grand River watershed that discharges into Lake Erie. We addressed two key research questions: (1) Are observed temporal patterns consistent in space throughout the watershed? (2) Are spatial patterns consistent in time at both event-based and seasonal scales? More specifically, following the nomenclature and approach of McClain et al. (2003), the research objectives were to determine critical times (‘hot moments’) and critical locations (‘hot spots’) of DRP, TP and NO3 export within the mixed land use Hopewell Creek watershed and to infer possible causes for water chemistry observations using land use and physiographic sub-watershed GIS data. Methods Research design Four monitoring sites with slightly different land uses, nested throughout a watershed, were selected to monitor streamflow and hydrometric variables as well as water quality over a 1-year study period (November 2014–October 2015) to address the research objectives. Samples were collected year-round, both regularly during baseflow and intensively during events (rain storms and thaws) to determine nutrient concentrations, loads and speciation and to relate differences to hydroclimatic drivers and/or land use patterns. Study site The study was conducted in the Hopewell Creek watershed, a mixed land use watershed located 15 km east of KitchenerWaterloo, Ontario, Canada. Hopewell Creek is a third-order stream that drains into the Grand River, that subsequently drains into Lake Erie. The Hopewell Creek watershed is 72 km2, containing gray

Please cite this article as: C. Irvine, M. Macrae, M. Morison et al., Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada, Journal of Great Lakes Research, https://doi.org/10.1016/j.jglr.2019.10.005

C. Irvine et al. / Journal of Great Lakes Research xxx (xxxx) xxx

brown luvisols, melanic brunisols and humic gleysols, and is dominantly sandy-loam textured (Presant and Wicklund, 1971; Fig. 1). The catchment receives substantial groundwater inputs throughout the year (Jyrkama and Sykes, 2007), but also receives overland flow contributions to the streams during high flow events from the ponding of water at the surface in microtopographic lows (Macrae et al., 2007a), as well as inputs from tile drains, which are common throughout the watershed (Table 1). The Hopewell Creek watershed experiences a cool, temperate climate, with 916.5-mm precipitation (17% as snowfall) annually (Environment Canada, 2017). The 30-year mean air temperatures are 20.0 °C in July and 6.5 °C in January, with air temperatures typically at or below freezing between December and March. The primary land use in the catchment is 46% agriculture, of which 24% is tile-drained (Table 1; see TE watershed). Other land

3

uses in the watershed include 41% natural areas, including forested areas, hedgerows and riparian areas, and 9% residential lands. Row crops are the most common agricultural practice in the watershed and consist of corn-soybean-cereal rotations. Other agricultural practices include pasture land and livestock (dairy and poultry). Four monitoring sites were established within the Hopewell Creek watershed (Fig. 1). Headwater (HW) is the furthest upstream site located at the headwaters of the watershed with its catchment area being predominantly forested and is the only subcatchment in the study with no tile drainage (Fig. 1). The stream is an ephemeral first-order stream that flows primarily during the spring freshet and during heavy rain events, usually in the spring and fall seasons. HW was chosen to serve as a relatively ‘‘undisturbed” reference site to act as a natural analog to give an idea of pre-development conditions before artificial subsurface drainage and agriculture

Fig. 1. Map of the study area and monitoring sites for discharge and chemistry, with a view of the (A) Hopewell Creek watershed, (B) the Grand River watershed with the Hopewell Creek watershed outlined in black, and (C) the Great Lakes region with the Grand River watershed outlined in beige.

Please cite this article as: C. Irvine, M. Macrae, M. Morison et al., Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada, Journal of Great Lakes Research, https://doi.org/10.1016/j.jglr.2019.10.005

4

C. Irvine et al. / Journal of Great Lakes Research xxx (xxxx) xxx

Table 1 Land use characteristics of contributing areas of the four monitoring sites within the Hopewell Creek Watershed (see Fig. 1). Site Headwater (HW) Strawberry (ST) Maryhill (MH) Terminus (TE)

Drainage Area (km2)

Tiled farmland (%)

Natural farmland (%)

Total farmland (%)

Dominant soil type

Stream Order

Average slope (%)

Drainage density (m ha

1.08 2.61 14.77 72.20

0 65 41 24

37 25 23 22

37 90 63 46

Sandy Sandy Sandy Sandy

1 1 2 3

3.09 1.43 1.68 2.21

14.46 10.42 12.88 13.08

were introduced throughout the landscape. Strawberry Creek (ST) and Maryhill (MH) are two streams adjacent to agricultural, tiledrained fields that drain primarily agricultural subcatchments. ST is a first-order stream with a contributing area of 3 km2, while MH is a second-order stream with a contributing area of 15 km2. These two monitoring sites both have predominantly agricultural land uses, although the land use varies. ST has primarily cash crops including soybean (Glycine max), corn (Zea mays), winter wheat (Triticum aestivum) and strawberries (Fragaria  ananassa) and has the highest proportion of tile-drained fields, while the MH site also contains cash crops (e.g., corn, soybean), but also contains dairy cattle and grazing pastures in its subcatchment and has a lower proportion of tile drains relative to ST. Terminus (TE) is located at the outlet of the watershed and drains the entire 72 km2 Hopewell Creek catchment. Three of the four subcatchments are dominated by agricultural land use with some residential and natural lands, although the percentage of total cropland differs across sites (Table 1). One of the subcatchments Strawberry Creek (ST) has been the site of previous agricultural nutrient export studies (e.g., Harris, 1999; Macrae et al., 2007a,b, 2010; Mengis et al., 1999) and provides a history of reference data in which to contextualize our results. Field methods Hydrometric variables were measured continuously (30-min intervals) at all four sites. Streamflow was recorded using Doppler Ultrasonic flow sensors (Starflow Model 6526, Unidata Ltd.) at the MH and TE sites and using pressure transducers (HOBO U20, Onset Ltd.) at the HW and ST sites. Rating curves were developed for sites with pressure transducers, and flow rates estimated by the ultrasonic sensors were validated using manual gauging measurements (Swoffer Model 3000 Current Velocity Meter) under a wide range of flow conditions over the 1-year study period. Any gaps in data (all short in duration) were filled using linear interpolation of established relationships between tributary streams and the basin outlet station (TE). Micrometeorological variables were recorded at 30-min intervals (Sutron XLite 9210B data logger) at each site and included sensors for air temperature (Vaisala HMP155A), soil temperature (LiCor LI-7900–180) and soil moisture (LiCor LI-7900-175) at depths of 5, 10, and 15 cm. Rainfall was recorded with a tipping bucket rain gauge at MH as well as from the Environment Canada (EC) monitoring station at the Region of Waterloo International Airport. The EC monitoring station is 3 km south of the TE monitoring site. Snowfall was monitored at ST, MH and TE with a Belfort All Environment Universal Precipitation Gauge. Prior to spring snowmelt, snow surveys were conducted at all four sites in February 2015. The winter of 2014–2015 experienced no winter thaws, suggesting that the snow survey data was an accurate representation of a seasonal total snow water equivalent. During storm or thaw events, water samples were collected at all four sites between November 2014 and October 2015 using portable automated water samplers (Teledyne ISCO 6712) and acidwashed (10% H2SO4 acid), triple-rinsed, polyethylene sample bottles. A total of 16 runoff generating events were captured during

Loam Loam Loam Loam

1

)

the study period, along with periodic baseflow sample collections on an approximately monthly basis (after Johnes (2007) and Williams et al. (2015)). Macrae et al. (2007a) also sampled baseflow at the ST watershed outlet at approximately 2–4 week intervals and did not find it to be substantially variable, particularly in comparison to event-based nutrient concentrations which dominated annual nutrient loads. Additional baseflow samples were captured both prior to and following the onset of storm conditions. Water sample collections spanned the rising and falling limbs of each event hydrograph, and event-based sampling intervals ranged from 2 to 6 h depending on storm characteristics as well as expected duration and response. Over the course of the study period, 136 samples were collected from TE, 160 from MH, 110 from ST and 96 from HW. Sample processing and laboratory analysis Water samples were packed on ice in coolers and transported to the Biogeochemistry Lab at the University of Waterloo and processed immediately. Subsamples were filtered through 0.45-mm cellulose acetate filters (Flipmate, Delta Scientific) and stored in the dark at 4 °C until the subsequent determination of dissolved nutrient species. An unfiltered subsample was preserved with acid (0.2% H2SO4 final concentration) and subsequently digested using acid (Kjeldahl) digestion (Seal Analytical Hot Block Digestion System BD50) for the determination of TP. DRP and TP samples were analyzed using standard colorimetric methods in the Biogeochemistry Lab at the University of Waterloo using a Bran Luebbe AA3, detection limit 0.001 mg P L 1 (Seal Analytical). NO3 was analyzed using ion chromatography (DIONEX ICS 3000 with Ion Pac AS18 analytical column, detection limit 0.12 mg N L 1). Approximately 5% of all samples were analyzed in replicate and found to have analytical precision within 5% of reported values. Data and statistical analysis At each site, streamflow was separated into baseflow and event flow components using an automated separation procedure (‘EcohydRology’ package within R) to permit the delineation of events (Electronic Supplementary Material (ESM) Fig. S1). Runoff events were identified from hydrograph analysis and deemed to have commenced when a measurable increase (within the detection magnitude of the deployed sensor) in stream flow was observed and deemed to have ended upon a return to seasonal baseflow conditions. Stream responses with multiple peaks were treated as separate events if the falling limb of the hydrograph was closer to baseflow conditions than to the peak flow of the event. In such cases, events were delineated using a synthetic recession curve. Events were sampled when autosamplers were triggered manually using the ‘‘delayed start” setting on the ISCO autosamplers. Consequently, 12 of the 16 events were captured by our autosamplers, whereas four were missed when autosamplers failed to trigger or events were shorter than our programmed sampling interval. These missed events were the four smallest events in terms of total discharge, and load-discharge relationships from the remaining data were used to estimate loads for the four events with missing

Please cite this article as: C. Irvine, M. Macrae, M. Morison et al., Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada, Journal of Great Lakes Research, https://doi.org/10.1016/j.jglr.2019.10.005

C. Irvine et al. / Journal of Great Lakes Research xxx (xxxx) xxx

hydrochemical data (ESM Fig. S2). Further, several storms occurred during which the samples which were scheduled to be obtained for the analysis of NO3 were not completed (see results section), during which loads were estimated as above. Event-specific flow-weighted mean concentrations (FWMC) of TP, DRP and NO3 were calculated using the continuous streamflow data (both event flow and baseflow combined) and the 6–24 samples collected throughout the event (method described by Williams et al., 2015). For each event, a nutrient load (kg ha 1, hereafter referred to as ‘loads’) was calculated by multiplying the FWMC by the total flow for that event (following Williams et al. (2015)). Events missed by our autosamplers were interpolated using flow-load regressions (ESM Fig. S2). Nutrient loads in this paper are not split into baseflow and event flow because it is not possible to chemically separate baseflow loads from our data set. We have instead divided the flow and load data into ‘‘event” and ‘‘between-event” conditions (after Macrae et al., 2007a). A seasonal between-event load was also determined for each of the four seasons during which samples were collected. Between-event loads were determined by multiplying the streamflow between events by a seasonal mean concentration of grab samples collected approximately monthly between events (i.e. during baseflow conditions) as well as those captured prior to event onset or following events when conditions returned to baseflow. Total seasonal loads were then calculated as the sum of between-event and event loads occurring in June–August (summer), March–May (spring), September–November (fall), or December–February (winter). Seasons were defined using the MAM, JJA, SON, DJF convention due to the March climatic conditions, where little precipitation was observed and a 2-week long spring freshet driven by radiation melt was observed. Results Hydroclimatic patterns and drivers of nutrient export Annual precipitation during the study period was 694.6 mm (Fig. 2), 24% below the 30-year average for the region. Air temperatures were typical of long-term averages throughout most of the year, with the exception of the winter period, which averaged 8.6 °C between December and February, 3.1 °C cooler than the

5

30-year average. Streamflow was highly variable both spatially and temporally throughout the year in response to precipitation and thaw-melt events (Fig. 2), which is common for the region (Macrae et al., 2007a). The highest flows occurred from early March to late April when a large snowmelt event of 45 mm snow water equivalent was followed by three spring rain events (6.6– 16 mm in magnitude) within 7 days on saturated soils. The dominance of the snowmelt and spring period for the region has been shown previously (e.g., Macrae et al., 2010). June 2015 was considerably wetter than average with 124 mm of rainfall during the month (historical average of 82.4 mm for June), while the other growing season months from May–October (excluding June) were drier than average (a total of 298.5 mm, compared to the historical average of 420.0 mm). Overall, autumn (September, October and November) was drier than average for Southern Ontario (200.6 mm of rainfall, most of which fell as one major storm in November 2014, with low-flow conditions in October 2014 and September 2015), causing low flow conditions as opposed to a typical fall (Fig. 2). Spatial variability in stream flow responses Three of the four sites (ST, MH and TE) had persistent baseflow conditions (mean values of 0.3, 0.5, and 0.4 mm d-1, respectively), while HW (forested reference site) was ephemeral and exhibited no flow conditions during most of the growing season (May–October) although hydrograph responses were observed during the spring (March and April) and following large rain events in June (Fig. 2 and ESM S1). The four sites responded similarly to rain and melt events throughout the year (Fig. 2), although ST (the first-order stream) was flashier than MH or TE. At the MH and TE sites, bank-full flow was frequently observed following large events, whereas streamflow rarely exceeded bank-full flow at ST. Spatially, the MH site had the highest runoff ratios throughout the study period, with an annual runoff ratio of 0.78 (event + baseflow, 0.52 for events only), while annual runoff ratios for the other sites were 0.01 at HW, 0.60 (event + baseflow) and 0.46 (events only) at ST and 0.68 (event + baseflow, 0.44 for events only) at TE. These differences in runoff between the four sites were consistent across the study year. The durations of event responses were similar throughout the watershed between three of the four sites

Fig. 2. Total discharge at each subwatershed, with numbered runoff events which were sampled across the watershed. Precipitation is also shown.

Please cite this article as: C. Irvine, M. Macrae, M. Morison et al., Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada, Journal of Great Lakes Research, https://doi.org/10.1016/j.jglr.2019.10.005

6

C. Irvine et al. / Journal of Great Lakes Research xxx (xxxx) xxx

(MH, ST and TE), with the reference forested site (HW) showing much more rapid runoff responses which were less sustained than the other sites. Temporal variability in stream flow responses Over the study period, several runoff events comprised a disproportionate contribution of annual flow (Figs. 2–5). Snowmelt had the largest total discharge at all sites (spanning 1.5–69 mm between the four sites, comprising between 11 and 16% of the total annual event flow at the ST, MH and TE sites, and 18% of annual flow at the HW site) over a 15-day period from March 11 to March 25. Although snowmelt represented the largest magnitude event in terms of flow, peak flow rates during the snowmelt event were less than peak flow rates observed during some fall or spring rainfall events (Fig. 2). The highest annual peak flow events were simultaneous at three of the sites (ST, MH and TE), with the exception of HW, which had peak flows during spring rainfall but was ephemeral throughout most of the growing season. Strong seasonal patterns were observed with regards to streamflow. The spring season (March–May) had the greatest discharge, which was consistent at all four sub-watersheds. In contrast, flow was much lower throughout the summer season (June–August), despite the fact this season had the greatest rainfall (Figs. 3–6). Summer storms were characterized by high-intensity precipitation with rapid discharge responses. However, summer flows were lower than those occurring in spring or autumn due to the drier antecedent conditions.

Fig. 4. Total event discharge (solid grey bars) and FWMC (DRP; hollow circles) at each monitoring site for all of the sampled events (see Fig. 2) during the study period.

Nutrient concentrations and loads: spatial variation

Fig. 3. Total event discharge (solid grey bars) and FWMC (TP; hollow circles) at each monitoring site for all of the sampled events (see Fig. 2) during the study period.

Across the nested subcatchments, spatial hot spots were identified for nutrient export, in which loads remained disproportionately high compared to the rest of the catchment, but the locations of these hot spots were different between N and P. During events, MH frequently had the greatest DRP and TP concentrations across all seasons and most events, with the exception of snowmelt (event 5) where TP FWMC was highest at TE and a single summer event where TP and DRP FWMC were highest at ST (Figs. 3 and 4). This suggests that MH is a spatial ‘hot spot’ for both P species. ST frequently had the greatest NO3 concentrations compared to the other sites, although MH had greater NO3 concentrations during a few events (Fig. 5). Consistently elevated NO3 loads from ST suggest that ST is a ‘hot spot’ for NO3 across temporal scales, with the highest instream NO3 concentrations of all sites at least once during each season (autumn, winter, spring and summer). The NO3 loads from ST were comparable in magnitude to those from MH, despite the fact that MH supplied more discharge. Unfortunately, NO3 concentrations were determined on a fewer number of storms at the MH site than at the ST site, and the differences between the two subwatersheds should therefore be treated with some caution. Indeed, elevated NO3 loads were observed in both of the agricultural subwatersheds relative to the other sites. Discharge at HW was negligible at the yearly time scale (8.4 mm) when compared to the other three sites (417–543 mm total flow and 304–368 mm event flow). This, coupled with the

Please cite this article as: C. Irvine, M. Macrae, M. Morison et al., Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada, Journal of Great Lakes Research, https://doi.org/10.1016/j.jglr.2019.10.005

C. Irvine et al. / Journal of Great Lakes Research xxx (xxxx) xxx

7

in agricultural systems in cold-regions. Outside of snowmelt, DRP concentrations were larger during storm events compared to baseflow conditions, during which concentrations were near the detection limit across all seasons and at all sites. Flow-weighted mean concentrations (FWMC) of DRP exhibited little seasonal variation (Fig. 4), although the snowmelt event had the largest DRP FWMC of all events at three sites (ST, MH and TE). Loads at these three sites (ST, MH and TE) were strongly influenced by discharge (Fig. 6 and ESM Table S2). The winter period had lowest P loads due to the fact that most flow occurred as non-event related flow, and DRP and TP concentrations were particularly low during baseflow conditions. The relative contribution of the spring period (dominated by snowmelt) to annual loads was greater for DRP than it was for TP. The two agricultural sites exported greater NO3 loads per area than the basin average (TE) both annually and during each of the four seasons. Differences in NO3 loads were less seasonally variable than P loads and closely resembled temporal patterns in discharge. While TP concentrations were low throughout the summer, DRP was relatively lower as a fraction of TP (low DRP:TP ratio) at all sites, suggesting high PP export during high-intensity precipitation events on dry soils during the summer months, while three rain events in the fall and early winter (November and December 2014) fell on wet, unfrozen soils and also resulted in low DRP:TP ratios throughout the four sites.

Discussion

Fig. 5. Total event discharge (solid grey bars) and FWMC (NO3 ; hollow circles) at each monitoring site for all of the sampled events (see Fig. 2) during the study period.

relatively low FWMC of DRP, TP and NO3 at this site (ESM Table S1), demonstrates that there is no significant nutrient export originating from the forested headwaters of the Hopewell Creek watershed (0.001 kg DRP ha 1; 0.004 kg TP ha 1; 0.04 kg NO3 N ha 1). In contrast, nutrient loads at the agricultural and mixed land use sites were elevated (0.10–0.15 kg DRP ha 1; 0.70– 0.94 kg TP ha 1; 9.15–11.55 kg NO3 -N ha 1) relative to the natural (HW) site (Fig. 6). The ST, MH and TE sites had similar DRP concentrations during low flow periods. During event-related flow, ST showed the least variability in DRP of all of the sites, whereas MH and TE had greater DRP concentrations coinciding with large discharge events (Fig. 4). TE had TP concentrations that were generally intermediate of MH and ST, a pattern observed during all four seasons, further suggesting that the MH subcatchment is the major contributor of TP in the watershed as the MH basin is fully nested within the TE basin. Spatial patterns of TP loading were consistent with those of DRP, where MH contributed greater TP loads than both ST and TE during most seasons. Moreover, TP concentrations at TE had the same temporal patterns as both ST and MH on both short (during events) and longer (seasonal) time scales (Fig. 3).

Nutrient concentrations and loads: temporal variation The greatest concentrations were observed during the March snowmelt at all sites for all nutrient species, indicating that snowmelt can act as a ‘hot moment’ and is a key driver of N and P export

This research has shown that agricultural land practices elevate nutrient concentrations in streamflow (e.g., SW, HM and TE sites) above natural concentrations for this region (HW site). The pristine headwater subcatchment (HW) was selected as a monitoring site to serve as an analog of background/baseline conditions to get a sense of pre-settlement nutrient dynamics. It is unclear whether the ephemeral nature of flow at HW is site-specific or indicative of wider pre-development runoff conditions. The nutrient loads from HW were negligible as a result of both the stream drying up during baseflow conditions and remaining dry during most of the summer months and the consistently low nutrient concentrations. The forested area in the headwater catchment is likely responsible for such minimal event runoff, in which a greater proportion of precipitation is directed towards infiltration, storage, canopy interception and evapotranspiration. Differences in runoff regimes between MH and SW are less pronounced and likely owing to differences in hydrogeologic conditions, where SW and much of the eastern lobe of the Hopewell Creek watershed are underlain by stone-poor, carbonate-derived silty to sandy till, and, in contrast, the surficial geology of the western lobe of the sub watershed, where MH is located, has a greater area of glaciofluvial deposits of sandy/gravelly deposit (Karrow, 1987). These data demonstrate that land use, specifically agriculture, has an impact on N and P export. This has been shown in studies in other regions including the Midwest United States (Arbuckle and Downing, 2001; Coulter et al. 2004; Southeastern United States (Beckert et al., 2011; Brion et al., 2011; Niño de Guzmán et al., 2012), prairie landscapes (Dodds and Oakes, 2006), Europe (Peršic´ et al., 2013), tropical regions (Castillo, 2010) and New Zealand (Abell et al., 2011), as well as in other watersheds in Southern Ontario (Sliva and Williams, 2001). This work shows that spatially distributed, event-based sampling regimes are important for load estimation in agricultural and mixed land use watersheds due to heterogeneity in subwatershed export dynamics. Indeed, the loads observed at the watershed outlet may not be representative of the overall watershed and often originate from critical source areas (hot spots) (Pionke et al., 2000), and annual loads may be largely driven by

Please cite this article as: C. Irvine, M. Macrae, M. Morison et al., Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada, Journal of Great Lakes Research, https://doi.org/10.1016/j.jglr.2019.10.005

8

C. Irvine et al. / Journal of Great Lakes Research xxx (xxxx) xxx

Fig. 6. Seasonal and annual loads of DRP, TP, and NO3 at each monitoring site, split by season.

critical events (hot moments) (Macrae et al., 2007a). This was particularly true for P in the current study, given contributions of the two agricultural sub-watersheds, particularly the MH site (Figs. 3 and 4). Although this study only spans one study year, leaving limited sampling opportunities to capture seasonal variability among storm events, the patterns shown here, particularly the dominance of the spring snowmelt period, are similar to what has been shown in the ST watershed in other years (e.g. Macrae et al., 2007a) and in edge-of-field studies across Ontario (e.g. Van Esbroeck et al., 2017). In addition, the speciation of P differed both spatially and temporally (Figs. 3 and 4), demonstrating the importance of highfrequency field monitoring programs. The identification of critical source areas for specific nutrients and critical time periods (hot moments) for nutrient loss can help land managers optimize BMPs throughout the watershed.

Within the agricultural areas of the mixed land use watershed, this study has identified both ‘hot spots’ and ‘hot moments’ for nutrient loading. The MH site represents a ‘hot spot’ in the watershed, with almost consistently greater P concentrations and greater rates of export than the other sites. This spatial pattern was observed during all seasons and suggests that localized land use at the field scale is important. The two agricultural subcatchments (ST and MH) have similar dominant land use (agriculture), soil type and topography (Table 1). However, the MH monitoring site is located immediately downstream of a dairy farm, a factor linked to high P concentrations in other studies (Niño de Guzmán et al., 2012) and is adjacent to pasture land, although cattle do not access the stream which is fenced off. The floodplain of the MH stream is flat and receives runoff from adjacent sloped fields. The floodplain areas are more prone to frequent surface

Please cite this article as: C. Irvine, M. Macrae, M. Morison et al., Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada, Journal of Great Lakes Research, https://doi.org/10.1016/j.jglr.2019.10.005

C. Irvine et al. / Journal of Great Lakes Research xxx (xxxx) xxx

inundation in comparison to the other sites. It is likely that much of the P losses at this site are generated by surface runoff, as the high instantaneous P concentrations and loads were observed during peak flow events during which surface overland flow was observed. The significance of overland flow in event-related P loss has been observed by others in the same region (e.g., Macrae et al., 2007a,2010). This site appears to be a critical source area for P within the overall watershed where the large P source can become connected to the stream, leading to significant P loads. In contrast, the ST site also has a stream but experiences less overland flow and does not have a significant P source (e.g., livestock) adjacent to the stream. This emphasizes the fact that not all agricultural fields (even within small watersheds) behave the same way. Indeed, two agricultural sub-watersheds located in close proximity to one another exhibited different nutrient export patterns throughout this study, which were likely a function of both supply and transport mechanisms from the farms. On an annual basis, P loads at the TE site were intermediate between the other subwatersheds. This was also true during most events, with the exception of Event 5 (the major snowmelt event), when P loads were very high at the TE site, likely due to the cumulative effects of the combined upstream sources, but possibly also due to an additional P source downstream of the other two sub-watersheds. Scale is an important consideration when discussing nutrient ‘hot spots’. This study has identified the MH sub-watershed as a ‘hot spot’ in the Hopewell Creek watershed, where elevated TP concentrations and loads were observed relative to the other monitored sub-watersheds. It is unclear if these elevated loads originate from numerous low-intensity ‘hot spots’ throughout the watershed (i.e., cumulative effects of upstream sources) or, from a small number of key hot spots within the watershed (for example, a livestock farm located within the watershed, near the stream). Ongoing work is exploring this. Previous work suggests that first and second order streams are highly influenced by land use immediately adjacent to the stream (Buck et al., 2004), which supports the idea that livestock operations may be an important P source at MH. In the future, additional field-scale studies within sub-watersheds with high nutrient export may be able to provide more information on land management specific hot spots and identify targeted BMPs to more efficiently limit nutrient export. This study has also identified the occurrence of ‘hot moments’ within a mixed land use watershed. Discharge was a strong control on nutrient loads during both short temporal scales (events) and long temporal scales (seasons). However, this was more evident during events and seasons that had consistently higher soil moisture (i.e., excluding summer months). Seasonally, the dominant pattern was that spring had greater nutrient loads than any other season, attributed to the large spring snowmelt period of March 2015. There were no mid-winter thaws during the winter of the 2014/2015 study period, even though these are typical in Southern Ontario (Environment Canada, 2017). This resulted in high flows in March during a 2-week period of radiation-melt with little rainfall. The dominance of the snowmelt period has been noted in previous studies in agricultural landscapes at both watersheds (e.g., Macrae et al., 2007a, 2010) and field scales (Lam et al., 2016b; Van Esbroeck et al., 2017), in addition to the importance of inter-year total precipitation variability controlling export loads (Richards et al., 2010). This study has shown that the dominance of this snowmelt period in annual hydrochemical export persists across multiple sub-watersheds in this northern region of the Lake Erie watershed, irrespective of land use. As such, management efforts across a range of land uses should target the non-growing season, particularly the spring snowmelt event. The large nutrient loads that were observed following snowmelt illustrate the importance of year-round, event-based stream sampling designs to capture these ‘hot moments’ and to determine more precise estimates of

9

annual loads, particularly in snowmelt-dominated regions. Such information is crucial to the successful evaluation of progress towards nutrient reduction targets. The export of P and export of N did not share the same hot spots, suggesting particular areas of the landscape contributing disproportionate P loads may not necessarily be problematic for N export, and vice versa. This is important for management strategies that may target different nutrient types and suggests that management efforts within a larger watershed may need to be optimized for different sub-watersheds. The MH sub-catchment was the major contributor of DRP and TP to the watershed outlet, while both agricultural sites MH and ST contributed elevated NO3 loads. Although some highly elevated NO3 concentrations were observed at ST, these were during small events not captured at the MH site (Fig. 5). MH has greater area of pasture land and more livestock, which has been shown to have lower N:P ratios compared to agricultural areas that are predominantly row crops (Arbuckle and Downing, 2001; Niño de Guzmán et al., 2012). The similar NO3 concentrations between the ST and MH site, and the elevated P concentrations at the MH site support this. The ST site has a higher proportion of tile-drained agricultural fields, 65% compared to 47% at MH, and a greater proportion of farmland which is being used to grow cash crops. Both of these factors have been shown to be a significant transport mechanism for NO3 (Macrae et al., 2007b; Stenberg et al., 2012). As well, evidence from Wood Country, Ohio, within the Western Lake Erie Basin, demonstrates the potential longevity of legacy P export maintained throughout several consecutive storm events, contrasting with the depletion effect in NO3 concentrations (King et al., 2017). Thus, future land management and identification of key nutrient export contributing areas (hot spots) and periods (hot moments) should consider the potential for differences between N and P export regimes and develop strategies for achieving target loads for either nutrient individually. In contrast to the framework of classifying particular seasons or events as ‘hot moments’ and particular sub-watesheds as ‘hot spots’ offered by McClain et al. (2003), Bernhardt et al. (2017) propose an alternative concept of ‘control points’ within hydrobiogeochemical systems. These ‘points’ may be regarded as both temporal and spatial while also not restricted to a binary state of either ‘hot’ or ‘not’. For instance, in this study the role of the spring snowmelt period as a major nutrient loading event across all sites, particularly from MH and ST, likely represents an ‘export control point’, as sufficient nutrients are available for export at that time while remaining transport-limited until sufficient snowmelt allows (Bernhardt et al., 2017). While acknowledging that a reclassification of findings for nutrient export from agricultural subwatersheds is possible under this new conceptual framework, further within-field and within-plot scales studies are required to properly elucidate a mechanistic understanding of the suite of complex processes responsible for nutrient export at the larger sub-watershed scale. Conclusion The results of this study show that multi-scale nested subwatershed studies can identify local ‘hot spots’ for P export in mixed land use watersheds. Temporally, a considerable proportion of TP and NO3 was exported during the winter months, not including the major role of the large melt event in March. Year-round monitoring of nutrient export is required in order to better understand annual nutrient export dynamics as related to hydrometeorological conditions and to more accurately inform watershed-scale nutrient models. Moreover, this is of particular importance in agricultural landscapes where management activities include fertilizer application in the fall, which can lead to large N and P exports during

Please cite this article as: C. Irvine, M. Macrae, M. Morison et al., Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada, Journal of Great Lakes Research, https://doi.org/10.1016/j.jglr.2019.10.005

10

C. Irvine et al. / Journal of Great Lakes Research xxx (xxxx) xxx

winter and spring thaw events. Further, catchment hydrological properties controlling the peak and timing of discharge and land management practices (e.g., livestock and timing of fertilizer application) are stronger controls of P export than the presence and ubiquity of tile drainage at these sites. The local hydrologic regime should be considered when land managers and farmers are determining site-specific best management practices. This nested sub-watershed study sheds light on the importance of subcatchment variation in nutrient loads for future design of watershed-scale targets. If data were only taken at the watershed outflow, the measured P loads would be used to set targets for P and N throughout the watershed. Unfortunately, there are occasionally trade-offs between the best management practices of TP, DRP and N (Her et al., 2016; Jarvie et al., 2017; Kokulan et al., 2019). Knowledge of spatial patterns throughout the watershed allows us to see where the most majority of the P and N areis originating and consequently allows us to optimize specific land management strategies to smaller regions within the broader watershed. This will presumably lead to improved nutrient reduction rather than applying broad practices across the entire watershed. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This research was funded by the Ontario Ministry of Food, Agriculture and Rural Affairs (Canada-Ontario Agreement for Great Lakes Program) and Natural Sciences and Engineering Research Council (Macrae DG). The Southern Ontario Water Consortium is acknowledged for equipment and logistical support. W. Strenzke and J. Nederend are thanked for logistical support. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.jglr.2019.10.005. References Abell, J.M., Ozkundakci, D., Hamilton, D.P., Miller, S.D., 2011. Relationships between land use and nitrogen and phosphorus in New Zealand lakes. Mar. Freshw. Res. 62, 162–175. Agnew, L.J., Lyon, S., Gérard-Marchant, P., Collins, V.B., Lembo, A.J., Steenhuis, T.S., Walter, M.T., 2006. Identifying hydrologically sensitive areas: bridging the gap between science and application. J. Environ. Manage.. 78, 63–76. Algoazany, A.S., Kalita, P.K., Czapar, G.F., Mitchell, J.K., 2007. Phosphorus transport through subsurface drainage and surface runoff from a flat watershed in east central Illinois, USA. J. Environ. Qual. 36 (3), 681–693. Arbuckle, K.E., Downing, J.A., 2001. The influence of watershed land use on lake N:P in a predominantly of watershed agricultural landscape. Limnol. Oceanogr. 46, 970–975. Baker, D.B., Johnson, L.T., Confesor, R.B., Crumrine, J.P., 2017. Vertical stratification of soil phosphorus as a concern for dissolved phosphorus runoff in the Lake Erie basin. J. Environ. Qual. 46 (6), 1287–1295. Baker, D.B., Richards, R.P., 2002. Phosphorus budgets and riverine phosphorus export in northwestern Ohio watersheds. J. Environ. Qual. 31 (1), 96–108. Banner, E.B.K., Stahl, A.J., Dodds, W.K., 2009. Stream discharge and riparian land use influence in-stream concentrations and loads of phosphorus from central plains watersheds. Environ. Manage. 44, 552–565. Beckert, K.A., Fisher, T.R., O’Neil, J.M., Jesien, R.V., 2011. Characterization and comparison of stream nutrients, land use, and loading patterns in Maryland Coastal Bay Watersheds. Water, Air, Soil Pollut. 221, 255–273. Bernhardt, E.S., Blaszczak, J.R., Ficken, C.D., Fork, M.L., Kaiser, K.E., Seybold, E.C., 2017. Control points in ecosystems: moving beyond the hot spot hot moment concept. Ecosystems 20 (4), 665–682.

Brion, G., Brye, K.R., Haggard, B.E., West, C., Brahana, J.V., 2011. Land-use effects on water quality of a first-order stream in the Ozark Highlands, Mid-Southern United States. River Res. Appl. 27, 772–790. Buck, O., Niyogi, D.K., Townsend, C.R., 2004. Scale-dependence of land use effects on water quality of streams in agricultural catchments. Environ. Pollut. 130, 287– 299. Carpenter, S.R., Caraco, N.F., Correll, D.L., Howarth, R.W., Sharpley, A.N., Smith, V.H., 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecol. Appl. 8, 559–568. Castillo, M.M., 2010. Land use and topography as predictors of nutrient levels in a tropical catchment. Limnol. Ecol. Manag. Inl. Waters 40, 322–329. Chen, N., Wu, Y., Chen, Z., Hong, H., 2015. Phosphorus export during storm events from a human perturbed watershed, southeast China: implications for coastal ecology. Estuar. Coast. Shelf Sci. 166, 1–11. Coulter, C.B., Kolka, R.K., Thompson, J.A., 2004. Water quality in agricultural, urban, and mixed land use watersheds. J. Am. Water Resour. Assoc. 40, 1593–1601. Dodds, W.K., Oakes, R.M., 2006. Controls on nutrients across a prairie stream watershed: land use and riparian cover effects. Environ. Manage. 37, 634–646. Environment Canada, 2017. Historical Data Record Accessed: http://climate. weather.gc.ca/historical_data/search_historic_data_e.html. Evans, D.M., Schoenholtz, S.H., Wigington, P.J., Griffith, S.M., Floyd, W.C., 2014. Spatial and temporal patterns of dissolved nitrogen and phosphorus in surface waters of a multi-land use basin. Environ. Monit. Assess. 186, 873–887. Ford, W., Williams, M.R., Young, M.B., King, K.W., Fischer, E., 2018. Assessing intraevent phosphorus dynamics in drainage water using phosphate stable oxygen isotopes. T. ASABE 61 (4), 1379–1392. Grant, R., Laubel, A., Kronvang, B., Anderson, H.E., Svendsen, L.M., Fuglsang, A., 1996. Loss of dissolved and particulate phosphorus from arable catchments by subsurface drainage. Water Resour. 30, 2633–2642. Harris, M., 1999. Nitrate Attenuation in a Narrow Non-forested Riparian Buffer Zone in an Agricultural Watershed in Southern Ontario (Thesis). Wilfrid Laurier University, p. 318. Her, Y., Chaubey, I., Frankenberger, J., Smith, D.R., 2016. Effect of conservation practices implemented by USDA programs at field and watershed scales. J. Soil Water Conserv. 71 (3), 249–266. International Joint Comission (IJC), 2012. Great Lakes Water Quality Agreement. Available online: https://www.ijc.org/en/who/mission/glwqa. (last accessed August 1, 2019). International Joint Comission (IJC), 2014. A Balanced Diet for Lake Erie: Reducing Phosphorus Loadings and Harmful Algal Blooms. Report of the Lake Erie Ecosystem Priority. Available online: https://legacyfiles.ijc.org/publications/ 2014%20IJC%20LEEP%20REPORT.pdf (last accessed August 1, 2019). Jarvie, H.P., Johnson, L.T., Sharpley, A.N., Smith, D.R., Baker, D.B., Bruulsema, T.W., Confesor, R., 2017. Increased soluble phosphorus loads to Lake Erie: unintended consequences of conservation practices? J. Environ. Qual. 46, 123–132. Johnes, P.J., 2007. Uncertainties in annual riverine phosphorus load estimation: impact of load estimation methodology, sampling frequency, baseflow index and catchment population density. J. Hydrol. 332 (1–2), 241–258. Jyrkama, M.I., Sykes, J.F., 2007. The impact of climate change on spatially varying groundwater recharge in the Grand River watershed (Ontario). J. Hydrol. 338 (3–4), 237–250. Kane, D.D., Conroy, J.D., Richards, R.P., Baker, D.B., Culver, D.A., 2014. Reeutrophication of Lake Erie: correlations between tributary nutrient loads and phytoplankton biomass. J. Great Lakes Res. 40 (3), 496–501. Karrow, P.F., 1987. In: Quaternary Geology of the Hamilton-Cambridge Area. Ontario Geologic Survey Open File Report 5429, pp. 1–147. King, K.W., Williams, M.R., Macrae, M.L., Fausey, N.R., Frankenberger, J., Smith, D.R., Kleinman, P.J.A., Brown, L.C., 2015. Phosphorus transport in agricultural subsurface drainage: a review. J. Environ. Qual. 44 (2), 467–485. King, K.W., Williams, M.R., Johnson, L.T., Smith, D.R., LaBarge, G.A., Fausey, N.R., 2017. Phosphorus availability in Western Lake Erie Basin drainage waters: legacy evidence across spatial scales. J. Environ. Qual. 46 (2), 466–469. Kokulan, V., Macrae, M.L., Lobb, D.A., Ali, G.A., 2019. Contribution of different flowpaths to hydrochemical export in the Northern Great Plains. J. Environ. Qual. 48 (4), 959–965. Lam, W.V., Macrae, M.L., English, M.C., O’Halloran, I.P., Plach, J.M., Wang, Y., 2016a. Seasonal and event-based drivers of runoff and phosphorus export through agricultural tile drains under sandy loam soil in a cool temperate region. Hydrol. Process. 2656, 2644–2656. Lam, W.V., Macrae, M.L., English, M.C., O’Halloran, I.P., Wang, Y.T., 2016b. Effects of tillage practices on phosphorus transport in tile drain effluent under sandy loam agricultural soils in Ontario, Canada. J. Great Lakes Res. 42, 1260–1270. Li, H., Sivapalan, M., Tian, F., Liu, D., 2010. Water and nutrient balances in a large tile-drained agricultural catchment: a distributed modeling study. Hydrol. Earth Syst. Sci. 14, 2259–2275. Liu, R., Wang, J., Shi, J., Chen, Y., Sun, C., Zhang, P., Shen, Z., 2014. Runoff characteristics and nutrient loss mechanism from plain farmland under simulated rainfall conditions. Sci. Total Environ. 468–469, 1069–1077. Macrae, M.L., English, M.C., Schiff, S.L., Stone, M., 2007a. Capturing temporal variability for estimates of annual hydrochemical export from a first-order agricultural catchment in southern Ontario, Canada. Hydrol. Process. 21, 1651– 1663. Macrae, M.L., English, M.C., Schiff, S.L., Stone, M., 2007b. Intra-annual variability in the contribution of tile drains to basin discharge and phosphorus export in a first-order agricultural catchment. Agric. Water Manage. 92, 171–182.

Please cite this article as: C. Irvine, M. Macrae, M. Morison et al., Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada, Journal of Great Lakes Research, https://doi.org/10.1016/j.jglr.2019.10.005

C. Irvine et al. / Journal of Great Lakes Research xxx (xxxx) xxx Macrae, M.L., English, M.C., Schiff, S.L., Stone, M., 2010. Influence of antecedent hydrologic conditions on patterns of hydrochemical export from a first-order agricultural watershed in Southern Ontario, Canada. J. Hydrol. 389, 101–110. Macrae, M.L., Zhang, Z., Stone, M., Price, J.S., Bourbonniere, R.A., Leach, M., 2011. Subsurface mobilization of phosphorus in an agricultural riparian zone in response to flooding from an upstream reservoir. Can. Water Resour. J. 36, 293– 311. Makarewicz, J.C., Bertram, P., 1991. Evidence for the restoration of the Lake Erie ecosystem. Bioscience 41 (4), 216–223. McClain, M.E., Boyer, E.W., Dent, C.L., Gergel, S.E., Grimm, N.B., Groffman, P.M., Hart, S.C., Harvey, J.W., Johnston, C.A., Mayorga, E., McDowell, W.H., 2003. Biogeochemical hot spots and hot moments at the interface of terrestrial and aquatic ecosystems. Ecosystems 6 (4), 301–312. Mehaffey, M.H., Nash, M.S., Wade, T.G., Ebert, D.W., Jones, K.B., Rager, A., 2005. Linking land cover and water quality in New York City’s water supply watersheds. Environ. Monit. Assess. 107, 29–44. Mengis, M., Schif, S.L., Harris, M., English, M.C., Aravena, R., Elgood, R.J., MacLean, A., 1999. Multiple geochemical and isotopic approaches for assessing ground water NO3 elimination in a riparian zone. Groundwater 37 (3), 448–457. Michalak, A.M., Anderson, E.J., Beletsky, D., Boland, S., Bosch, N.S., Bridgeman, T.B., Chaffin, J.D., Cho, K., Confesor, R., Daloglu, I., Depinto, J.V., Evans, M.A., Fahnenstiel, G.L., He, L., Ho, J.C., Jenkins, L., Johengen, T.H., Kuo, K.C., Laporte, E., Liu, X., McWilliams, M.R., Moore, M.R., Posselt, D.J., Richards, R.P., Scavia, D., Steiner, A.L., Verhamme, E., Wright, D.M., Zagorski, M.A., 2013. Record-setting algal bloom in Lake Erie caused by agricultural and meterological trends consistent with expected future conditions. Proc. Natl. Acad. Sci. U.S.A. 110, 6448–6452. Mineau, M.M., Wollheim, W.M., Stewart, R.J., 2015. An index to characterize the spatial distribution of land use within watersheds and implications for river network nutrient removal and export. Geophysical Res. Lett. 42, 6688–6695. Niño de Guzmán, G.T., Hapeman, C.J., Prabhakara, K., Codling, E.E., Shelton, D.R., Rice, C.P., Hively, W.D., McCarty, G.W., Lang, M.W., Torrents, A., 2012. Potential pollutant sources in a Choptank River (USA) subwatershed and the influence of land use and watershed characteristics. Sci. Total Environ. 430, 270–279. Perks, M.T., Owen, G.J., Benskin, C.M.H., Jonczyk, J., Deasy, C., Burke, S., Reaney, S.M., Haygarth, P.M., 2015. Dominant mechanisms for the delivery of fine sediment and phosphorus to fluvial networks draining grassland dominated headwater catchments. Sci. Total Environ. 523, 178–190. Peršic´, V., Kocˇic´, A., Horvatic´, J., 2013. Water quality and algal growth potential of watercourses draining agricultural and forested catchments in eastern Croatia (Middle Danube Basin). Fundam. Appl. Limnol. 182, 31–46. Pieterse, N.M., Bleuten, W., Jorgensen, S.E., 2003. Contribution of point sources and diffuse sources to nitrogen and phosphorus loads in lowland river tributaries. J. Hydrol. 271, 213–225. Pionke, H.B., Gburek, W.J., Sharpley, A.N., 2000. Critical source area controls on water quality in an agricultural watershed located in the Chesapeake Basin. Ecol. Eng. 14 (4), 325–335. Presant, E.W., Wicklund, R.E., 1971. The soils of Waterloo County. Report No. 44 of the Ontario Soil Survey. Research Branch, Canada Department of Agriculture.

11

Richards, R.P., Baker, D.B., Crumrine, J.P., Stearns, A.M., 2010. Unusually large loads in 2007 from the Maumee and Sandusky Rivers, tributaries to Lake Erie. J. Soil Water Conserv. 65 (6), 450–462. Rozemeijer, J.C., van der Velde, Y., van Geer, F.C., Bierkens, M.F.P., Broers, H.P., 2010. Direct measurements of the tile drain and groundwater flow route contributions to surface water contamination: from field-scale concentration patterns in groundwater to catchmentscale surface water quality. Environ. Pollut. 158, 3571–3579. Scavia, D., Allan, J.D., Arend, K.K., Bartell, S., Beletsky, D., Bosch, N.S., Brandt, S.B., Briland, R.D., Dalog˘lu, I., DePinto, J.V., Dolan, D.M., 2014. Assessing and addressing the re-eutrophication of Lake Erie: central basin hypoxia. J. Great Lakes Res. 40 (2), 226–246. Sharpley, A.N., Mcdowell, R.W., Kleinman, P.J.A., 2001. Phosphorus loss from land to water: integrating agricultural and environmental management. Plant Soil 237, 287–307. Sliva, L., Williams, D.D., 2001. Buffer zone versus whole catchment approaches to studying land use impact on river water quality. Water Res. 35, 3462–3472. Steffen, M.M., Belisle, B.S., Watson, S.B., Boyer, G.L., Wilhelm, S.W., 2014. Status, causes and controls of cyanobacterial blooms in Lake Erie. J. Great Lakes Res. 40 (2), 215–225. Stenberg, M., Ulén, B., Söderström, M., Roland, B., Delin, K., Helander, C.A., 2012. Tile drain losses of nitrogen and phosphorus from fields under integrated and organic crop rotations. A four-year study on a clay soil in southwest Sweden. Sci. Total Environ. 434, 79–89. Stow, C.A., Cha, Y., Johnson, L.T., Confesor, R., Richards, R.P., 2015. Long-term and seasonal trend decomposition of Maumee River nutrient inputs to western Lake Erie. Environ. Sci. Technol. 49 (6), 3392–3400. Van Esbroeck, C.J., Macrae, M.L., Brunke, R.R., McKague, K., 2016. Annual and seasonal phosphorus export in surface runoff and tile drainage from agricultural fields with cold temperate climates. J. Great Lakes Res. 42 (6), 1271–1280. Van Esbroeck, C.J., Macrae, M.L., Brunke, R.R., McKague, K., 2017. Surface and subsurface phosphorus export from agricultural fields during peak flow events over the nongrowing season in regions with cool, temperate climates. J. Soil Water Conserv. 72, 65–76. Vidon, P., Cuadra, P.E., 2011. Phosphorus dynamics in tile-drain flow during storms in the US Midwest. Agric. Water Manage. 98, 532–540. Watson, S.B., Miller, C., Arhonditsis, G., Boyer, G.L., Carmichael, W., Charlton, M.N., Confesor, R., Depew, D.C., Höök, T.O., Ludsin, S.A., Matisoff, G., 2016. The reeutrophication of Lake Erie: harmful algal blooms and hypoxia. Harmful Algae 56, 44–66. Williams, M.R., King, K.W., Ford, W., Buda, A.R., Kennedy, C.D., 2016. Effect of tillage on macropore flow and phosphorus transport to tile drains. Water Resour. Res. 52, 2868–2882. Williams, M.R., King, K.W., Macrae, M.L., Ford, W., Van Esbroeck, C., Brunke, R.I., English, M.C., Schiff, S.L., 2015. Uncertainty in nutrient loads from tile-drained landscapes: effect of sampling frequency, calculation algorithm, and compositing strategy. J. Hydrol. 530, 306–316.

Please cite this article as: C. Irvine, M. Macrae, M. Morison et al., Seasonal nutrient export dynamics in a mixed land use subwatershed of the Grand River, Ontario, Canada, Journal of Great Lakes Research, https://doi.org/10.1016/j.jglr.2019.10.005