Journal Pre-proof
Severe western Canadian wildfire affects water quality even at large basin scales
Craig A. Emmerton, Colin A. Cooke, Sarah Hustindss, Uldis Silins, Monica B. Emelko,
Ted Lewis, Mary K. Kruk, Nadine Taube, Dongnan Zhu, Brian Jackson, Michael
Stone, Jason G. Kerr, John F. Orwin
PII:
S0043-1354(20)30608-4
DOI:
https://doi.org/10.1016/j.watres.2020.116071
Reference:
WR 116071
To appear in:
Water Research
Received Date: 23 January 2020
Revised Date:
4 June 2020
Accepted Date: 15 June 2020
Please cite this article as: Emmerton, C.A., Cooke, C.A., Hustindss, S., Silins, U., Emelko, M.B., Lewis,
T., Kruk, M.K., Taube, N., Zhu, D., Jackson, B., Stone, M., Kerr, J.G., Orwin, J.F., Severe western
Canadian wildfire affects water quality even at large basin scales, Water Research (2020), doi: https://
doi.org/10.1016/j.watres.2020.116071.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier Ltd.
Following a wildfire, shortlived increases in chemicals
associated with ash occurred
in the waters of very large,
low-relief rivers
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Severe western Canadian wildfire affects water quality even at large basin
scales
Emmerton, Craig A.1,2*, Cooke, Colin A.1,3*, Hustins, Sarah1, Silins, Uldis4, Emelko, Monica B.5,
Lewis, Ted6, Kruk, Mary K.1, Taube, Nadine1, Zhu, Dongnan1, Jackson, Brian 1, Stone, Michael 7,
Kerr, Jason G.1, Orwin, John F.1
Affiliations:
1
Alberta Environment and Parks, Edmonton, Alberta, Canada
Dept. of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada
3
Dept. of Earth and Atmospheric Sciences, University of Alberta, Edmonton, Alberta, Canada
4
Dept. of Renewable Resources, University of Alberta, Edmonton, Alberta, Canada
5
Dept. of Civil and Environmental Engineering, University of Waterloo, Waterloo, Ontario, Canada
6
Hatfield Consultants, North Vancouver, British Columbia, Canada
7
Dept. of Geography and Environmental Management, University of Waterloo, Waterloo, Ontario, Canada
2
Corresponding Author (*) Emails:
Emmerton: craig.emmerton@gov.ab.ca
Cooke: colin.cooke@gov.ab.ca
20
21
Abstract
22
However, our understanding of these impacts is founded primarily from studies of small
23
watersheds with well-connected runoff regimes. Despite the predominance of large, low-relief
24
rivers across the fire-prone Boreal forest, it is unclear to what extent and duration wildfire-
25
related material (e.g., ash) can be observed within these systems that typically buffer upstream
26
disturbance signals. Following the devastating 2016 Fort McMurray wildfire in western Canada,
27
we initiated a multi-faceted water quality monitoring program that suggested brief (hours to
28
days) wildfire signatures could be detected in several large river systems, particularly following
29
rainfall events greater than 10 millimeters. Continuous monitoring of flow and water quality
30
showed distinct, precipitation-associated signatures of ash transport in rivers draining expansive
31
(800–100,000 km2) and partially-burned (<1–22 percent burned) watersheds, which were not
32
evident in nearby unburned regions. Yields of suspended sediment, nutrients (nitrogen,
33
phosphorus) and metals (lead, others) from impacted rivers were 1.2 to 10 times greater than
34
from those draining unburned regions. Post-fire suspended sediment concentrations in impacted
35
rivers were often larger than pre-fire 95% prediction intervals based on several years of water
36
sampling. These multiple lines of evidence indicate that low-relief landscapes can mobilize
37
wildfire-related material to rivers similarly, though less-intensively and over shorter durations,
38
than headwater regions. We propose that uneven mixing of heavily-impacted tributaries with
39
high-order rivers may partially explain detection of wildfire signals in these large systems that
40
may impact downstream water users.
41
Keywords: wildfire, river, water quality, suspended sediment, Boreal, continuous monitoring
Wildfires can have severe and lasting impacts on the water quality of aquatic ecosystems.
42
43
1. Introduction
Communities within the wildland-urban interface of forested regions are experiencing
44
more frequent and destructive wildfires (Hanes et al., 2018; Mell et al., 2010). This changing
45
wildfire regime is associated with a warming climate that is intensifying dangerous fire weather
46
and drought conditions (Diffenbaugh et al., 2015; Jolly et al., 2015), exposing these communities
47
to increased risk of catastrophic impacts (Radeloff et al., 2018). Besides the evident
48
consequences to lives, natural capital and infrastructure, wildfires often impact downstream
49
surface water resources resulting from changes in runoff and degraded, increasingly variable
50
water quality that can affect aquatic ecosystem functioning and drinking water treatability
51
(Bladon et al., 2014; Gresswell, 1999; Hohner et al., 2019; Smith et al., 2011; Writer et al.,
52
2014).
53
Wildfires can enhance meteoric runoff from watersheds through chemical soil sealing
54
and reduction in evapotranspiration due to vegetation removal (Larsen et al., 2009; Neary et al.,
55
2005). Depending on fire severity and the timing of storm events, delivery of ash and sediment
56
(Moody and Martin, 2001; Santín et al., 2015), organic matter and nutrients (Burd et al., 2018;
57
Earl and Blinn, 2003; Lane et al., 2008; Rhoades et al., 2011), major ions (Bayley et al., 1992;
58
Chanasyk et al., 2003; Mast et al., 2016), and trace contaminants (Kelly et al., 2006) into
59
receiving waters can occur and cause deteriorated water quality (Bladon et al., 2014; Smith et al.,
60
2011). Ensuing biogeochemical processing of this mobilized wildfire-related material can induce
61
secondary ecological impacts on aquatic ecosystems resulting from eutrophication and
62
sedimentation (Earl and Blinn, 2003; Gresswell, 1999; Silins et al., 2014). However, our
63
understanding of fire-related impacts on surface waters has been overwhelmingly established
64
from investigations conducted at relatively small catchment scales (e.g., <10s–100s km2) and in
65
catchments with responsive hydrologic connectivities between landscapes and receiving waters
66
(Belillas and Rodà, 1993; Earl and Blinn, 2003; Gerla and Galloway, 1998; Inbar et al., 1998;
67
Lane et al., 2008; Minshall et al., 2001; Neary et al., 2005; Prepas et al., 2003; Scott et al., 1998;
68
Smith et al., 2011).
69
Landscape disturbances and related impacts on downstream river conditions are difficult
70
to detect as watershed and river sizes increase (Blöchl et al., 2007; Elhadi et al., 1984). High-
71
order river systems aggregate water and material from broad drainage networks and dampen
72
biogeochemical signals from upstream landscapes through lateral mixing, dilution, and other in-
73
stream processes (Temnerud and Bishop, 2005). Consequently, assessments of wildfire-related
74
impacts on water quality in rivers draining diverse watershed sizes—especially very large ones—
75
are scant (Smith et al., 2011). Watershed hydrologic connectivity also influences river response
76
to landscape disturbance (Bracken and Croke, 2007). For instance, steeper topographic settings
77
and well-drained soils can promote rapid hydro-geochemical connections between disturbed
78
landscapes and rivers, and contribute to sustaining downstream and temporal wildfire impacts on
79
water quality (e.g.,(Rust et al., 2018). Though several studies have described wildfire impacts on
80
water quality in larger (Burke et al., 2005; Emelko et al., 2016; Rhoades et al., 2011) or low-
81
relief (Townsend and Douglas, 2000) watersheds, these investigations have been conducted in
82
landscapes that are hydrologically well-connected. Critically, wildfire impacts on water are not
83
well understood in large river systems with relatively low hydrologic connectivity, in which
84
contaminants may be more efficiently retained on land surfaces conferring greater apparent
85
storage or watershed buffering capacity (Ebel and Mirus, 2014; Prepas et al., 2003). These
86
attributes are characteristic of many areas across the circumpolar Boreal forest, where low
87
topographic relief, heterogeneous glacial deposits and extensive peatland landscapes result in
88
relatively low landscape-water coupling (Devito et al., 2017). While wildfires have been
89
investigated at plot- and subwatershed-scales in these Boreal landscapes (e.g.,(Depante et al.,
90
2018; Olefeldt et al., 2013; Prepas et al., 2003; Wilkinson et al., 2018), wildfire impacts on water
91
quality at large basin scales in these regions have not been previously reported. This a critical
92
knowledge gap considering the high future probability of more severe fires occurring across the
93
circumpolar Boreal forest.
94
The primary objective of this study is to assess the impacts of the May 2016 Fort
95
McMurray (Horse River) wildfire in northeastern Alberta, Canada on the water quality of large
96
(>1,000 km2 drainage areas) and low-relief (<5% drainage area slope) rivers in the region. Using
97
a multi-tiered monitoring approach to target signal detection in these large rivers, we first
98
compared suspended sediment, dissolved material (ion), nutrient, and metal concentrations and
99
yields in rivers draining burned areas with an unburned control station. Second, at each station,
100
we assessed if precipitation events and subsequent river flow mobilized more wildfire-related
101
material relative to non-precipitation periods. Finally, we evaluated suspended sediment,
102
dissolved material and flow relationships in rivers draining burned landscapes with pre-fire data
103
from the same station. We hypothesized that rivers draining burned watersheds would be
104
resistant to unbuffered overland flow and therefore show no detectable differences in water
105
quality relative to rivers draining unburned landscapes or extensive historical data.
106
107
108
2. Methods
2.1 Setting
Our study region is contained within the Boreal Plains ecozone, which is characterized by
109
flat and rolling landscapes with thick glacial till soils, and extensive coverage of wetlands and
110
coniferous forests (Chanasyk et al., 2003; Devito et al., 2017). Due primarily to the Rocky
111
Mountains barrier ~600 km to the west, much of this ecozone is relatively dry (419 mm annual
112
precipitation at Fort McMurray; 1981–2010) with a cold, continental climate (Environment and
113
Climate Change Canada, 2018). Human disturbances in the region are mostly from local
114
urbanization in Fort McMurray and oil sands strip-mining and exploration-related impacts to
115
landscapes (Figure 1).
116
The Athabasca River drains approximately 140,000 km2 of western Canada and occupies
117
23% of the province of Alberta (Figure 1). This Strahler-order nine river is large and voluminous
118
(~500 m wide; >500 m3 s-1 mean annual flow) in its lower reaches and exhibits a spring
119
snowmelt-freshet seasonality in flow (Figure 2). In this portion of the river, concentrations of
120
suspended sediment and associated water quality parameters (e.g., organic carbon, nutrients,
121
metals) derived from upstream tributaries and channel re-suspension can be high relative to
122
national water quality guidelines during much of the year (Glozier et al., 2018). Several high-
123
order river systems empty into the lower Athabasca River including the Clearwater and
124
Hangingstone rivers (Figures 1, 2).
125
The 2016 Fort McMurray wildfire (Figure 1; May 01/16–August 02/17) was the most
126
costly natural disaster in Canadian history ($8.9 billion in direct costs;(Alam et al., 2018). By
127
May 03/16, the wildfire advanced quickly from the southwest of Fort McMurray and triggered
128
the dramatic evacuation of ~90,000 people, resulting in a loss of 3,000 structures and eventual
129
consumption of ~600,000 hectares of forest and peatland. Within the final fire perimeter (nearly
130
established by May 17/16), 77% of the area was impacted by fire, half of which burned at a high
131
severity (see supplementary material for methodology).
132
2.2 Monitoring and historic data
133
Post-fire river water quality monitoring in this study focused on three rivers: the
134
Athabasca, Clearwater and Hangingstone. Three burned (impacted) stations were established: (1)
135
on the Athabasca River upstream of Fort McMurray (left bank); (2) the Clearwater River near
136
the mouth; and (3) the Hangingstone River at Fort McMurray (Figure 1). These locations drain a
137
gradient of watershed size (Hangingstone<Clearwater<Athabasca) and reflected a contrasting
138
gradient of relative burned area and area burned at a high severity in their watersheds
139
(Athabasca<Clearwater< Hangingstone; Table 1). An unburned (control) station was also
140
initiated on the Athabasca River at Grand Rapids, upstream of the wildfire perimeter. Each of
141
these four water quality stations incorporated multiple measurement approaches and, where
142
possible, were co-located near meteorological and flow stations to better understand hydrological
143
influences on post-fire water quality.
144
2.2.1 Water quality measurements
145
A multiple-method approach for sampling river chemistry of each of the four water
146
quality stations was implemented after the fire to assess differences between unburned and
147
burned locations (Figure 1, Table S1). Nearly two weeks after the fire began, we deployed multi-
148
probe data sondes (Hydrolab DS5X, OTT Hydromet, USA) just offshore at the two Athabasca
149
stations (burned and unburned), and near mid-channel at the smaller Clearwater and
150
Hangingstone stations. Two additional sondes were deployed downstream of Fort McMurray just
151
offshore of each bank of the Athabasca River to assess river mixing conditions (Figure 1). Each
152
calibrated sonde recorded general water quality including water temperature, pH, specific
153
conductivity, turbidity and dissolved oxygen at 15-minute intervals. Sondes were retrieved and
154
replaced with calibrated instruments every week to limit impacts from fouling and sensor drift.
155
All sonde monitoring concluded in mid-October 2016.
156
In addition to the sondes, we collected daily water quality samples at the four main
157
stations (Figure 1, Table S1). Surface water autosamplers (Teldyne ISCO, USA) were deployed
158
on the river bank at each station (burned: May–August, 2016; control: June–August, 2016) with
159
water intakes placed mid-column just offshore at the two Athabasca locations, and mid-channel
160
in the Clearwater and Hangingstone rivers. Each autosampler collected equal aliquots of river
161
water every three hours into 1 L ProPak® plastic bags in holders for a time-integrated sample
162
each day. Weekly grab samples of surface waters were manually collected at each station from
163
May to August 2016 to supplement automated sampling. After wading waist-deep at each
164
location, water samples were collected at 30-cm depth directly into pre-cleaned or new plastic or
165
glass bottles, depending on chemical methods. At the three burned stations (Athabasca-Fort
166
McMurray, Clearwater-mouth, Hangingstone-Fort McMurray), grab sampling continued
167
monthly through 2016, 2017 and 2018 via other programs.
168
All water samples were analyzed at accredited laboratories for suspended sediment, total
169
ion (Na, K, Mg, Ca, Cl), total nutrient (total organic carbon, total Kjeldahl nitrogen, total
170
phosphorus) and total metal (29-metal scan) concentrations. Only total measurements were
171
considered for autosampler water samples to limit sample chemical transformation concerns
172
during storage.
173
2.2.2 Precipitation and river flow measurements
174
175
To identify periods when wildfire-related material may be mobilized in burned
watersheds, we delineated notable local precipitation events and subsequent river flow response
176
periods at each water quality station. Total daily precipitation data from meteorological stations
177
nearest to sampling locations (Figure 1, Table S1) were used to identify noteworthy rainfalls
178
equalling or exceeding 10 mm (94th percentile of daily rain events at Fort McMurray A, 1999–
179
2015). Precipitation events at each station (see supplementary material) are defined in time here
180
as the day of the noteworthy rainfall plus five subsequent days to encompass flow and
181
biogeochemical responses of each river. Days not within a precipitation event period are defined
182
as non-event periods.
183
Hourly and daily flow data before and after the wildfire were retrieved from hydrometric
184
stations close to water quality locations, where possible (Figures 1, 2, Table S1). Discharge at the
185
Athabasca River upstream of Fort McMurray (left bank) station was calculated by difference in
186
measured daily flows between the nearby Athabasca River downstream of Fort McMurray and
187
Clearwater River at Draper stations. Flow at Athabasca River at Grand Rapids was calculated
188
using daily runoff yields from the upstream Athabasca River at Athabasca hydrometric station
189
scaled to the drainage area at the Grand Rapids location (Table 1, Figure 1).
190
2.2.3 Historic water quality data
191
Historic water quality data from long-term provincial and oil sands monitoring programs
192
(2001-2016) were used to quantify pre- and post-fire water quality conditions at the three
193
impacted stations (Table S1). Monthly surface water sampling has been ongoing in the
194
Athabasca River upstream of Fort McMurray for many decades, while the Clearwater and
195
Hangingstone rivers have been monitored for several years. Monthly grab sampling also
196
occurred in 2015 (pre-fire) and 2016 (post-fire) at burned (High Hills, Christina, Clearwater,
197
Hangingstone) and unburned (Calumet, Mackay, Ells, Firebag) river stations via regional
198
monitoring networks (Table S1, Figure S1a). Water collection and analytical methods from
199
historical programs were consistent with those used during the post-fire sampling program.
200
2.3 Quantitative analyses
201
2.3.1 Water quality assessment
202
River turbidity and conductivity are typically negatively correlated during high-water
203
events due to flow-related sediment concentration and ion dilution (Schlesinger, 2007).
204
However, after wildfires, ash-laden runoff pulses during storms can trigger concurrent increases
205
in turbidity and conductivity in receiving waters (Dahm et al., 2015; Earl and Blinn, 2003). To
206
identify existence of concentration-dilution and concentration-concentration periods in each
207
river, mean hourly sonde and flow data were included in an unconstrained principal components
208
analysis (PCA; Canoco v.5.03; Biometris, The Netherlands). Six-hour covariances (using a 60-
209
minute moving window) of turbidity and conductivity were quantified at each station (including
210
sonde-only locations) to directly compare the magnitude and frequency of suspected ash
211
mobilization periods (i.e., positive covariances) between stations. We also performed a
212
laboratory ash-mixing experiment to demonstrate how ash affected turbidity and conductivity
213
measurements in an aqueous solution. We collected loose grey and black ash from the soil
214
surface of burned forested plot (0.25 m2) in the Clearwater River watershed (N56.6849; W-
215
111.2560) on May 13/16, prior to post-fire rainfall events. Later in the laboratory, we
216
incrementally added ~500 mg of ash to 10 L of deionized water and continuously stirred the
217
solution while turbidity, conductivity and pH were measured using a calibrated sonde.
218
Descriptive statistics were calculated to quantify differences in yields of selected particle-
219
bound and dissolved chemicals between impacted and control river stations on the Athabasca
220
River. To link assessments of selected chemicals at each station to other measured parameters,
221
we correlated all parameter concentrations with concurrent measurements of flow, turbidity,
222
conductivity and pH to organize parameters as particle-bound type (positive correlation with
223
flow and turbidity) or dissolved type (positive correlation with conductivity and pH).
224
225
2.3.2 Precipitation and flow assessment
All water quality data (i.e., sonde, samples) at each station were binned into precipitation
226
event (≥10 mm) and non-event (<10 mm) periods for comparisons. This approach was used on
227
sonde turbidity-conductivity covariance data and concentrations and yields of sampled
228
chemicals. Assessment of hydrologic responses of impacted rivers before and after the fire is
229
described in the supplementary material.
230
2.3.3 Historic water quality assessment
231
Post-fire changes in particle-bound and dissolved river water chemistry, relative to
232
historic records, were evaluated in the three rivers draining burned watersheds (Table S1) using
233
concentration-discharge (C-Q) relationships. Monthly, pre-fire water quality data (suspended
234
sediments as a proxy of particle-bound chemicals, total calcium as a proxy for specific
235
conductivity/dissolved chemicals; see Table S2) from each river station were log-transformed
236
and linearly regressed with log-transformed, pre-fire daily discharge data. Subsequently, 95%
237
prediction intervals were calculated (Sigmaplot v13, Systat, USA) to serve as a boundary of what
238
C-Q relationships would be expected in the river based on historic conditions. When post-fire C-
239
Q relationships (2016–18) were compared with historic relationships, data points that were
240
outside of the bounds of the prediction intervals were considered noteworthy. Other monthly
241
grab sample and flow data from four unburned and four burned river stations were used in a
242
repeated measures ANOVA (SPSS v.25; IBM, USA) to assess lower-resolution pre-fire (2015)
243
and post-fire (2016) impacts on water quality.
244
3. Results
245
3.1 Water quality
246
3.1.1 Sonde water quality
247
High-frequency water quality monitoring by sondes at the four main stations typically
248
showed flow-concentration for turbidity and flow-dilution for conductivity (Figure 3) and pH, as
249
well as contrasting seasonal changes for water temperature and dissolved oxygen. These trends
250
were evident in PC1 and PC2 scores of the PCA (Table S3). However, sharp peaks in turbidity
251
and conductivity together occurred during short periods (hours to days) often along the rising
252
limb of the hydrograph at burned stations only. The third principal component at each burned
253
station reflected this covariation between turbidity and conductivity in the absence of association
254
with flow. At the Athabasca River control station, PC3 scores also showed periods of concurrent
255
increases in turbidity and conductivity, but were positively associated with flow, perhaps
256
suggesting an influence of saline groundwater (Gibson et al., 2016). Mean six-hour covariance
257
scores of turbidity and conductivity more distinctly showed differences between stations than did
258
the PCA. Covariances from burned stations were 3–25 times higher than the control station,
259
though means were influenced by the highest values (Figure 4). Highest covariances at impacted
260
stations typically occurred during precipitation events (mean: 14-113) and were 2 to 22 times
261
higher than non-event periods. These observations were consistent with results of the ash mixing
262
experiment that demonstrated concurrent increases in turbidity, specific conductivity and pH as
263
ash was added to deionized water (Figure S2). Further, though pre-fire grab sampling in
264
impacted rivers showed consistent negative associations between turbidity concentrations and
265
specific conductivity, 11-38% of post-fire, precipitation event turbidity and conductivity samples
266
were greater than the upper historic 95% prediction interval, suggesting a more positive
267
association between the parameters during those periods (Figure S3).
268
Within the burned reach of the Athabasca River, turbidity and conductivity covariances
269
on the left bank decreased from a mean of 7.7 (maximum: 1,071) upstream to a mean of 7.0
270
(maximum: 574) downstream. However, the downstream right bank station observed much
271
higher covariance values (mean: 17.2; maximum: 3,782) than either station on the left bank,
272
which together suggests an unmixed chemical profile longitudinally and laterally in the river
273
(Figure 3,4).
274
3.1.2 Sample water quality
275
Daily automated and grab sample water quality results indicated only slight differences
276
between burned and control stations when aggregated to the entire sampling season (Figure 5).
277
For example, seasonal mean daily sediment yields (June 22–August 30/16) from the directly
278
comparable stations of the Athabasca River at Grand Rapids (unburned; GR) and the Athabasca
279
River upstream of Fort McMurray (burned; FMM) were within one standard error of each other
280
(GR: 0.68±0.10 kg ha-1 d-1; FMM: 0.75±0.09 kg ha-1 d-1). Results of other sediment-associated
281
yield comparisons including total lead (GR: 8.1±1.0; FMM: 10.3±1.8 mg ha-1), total Kjeldahl
282
nitrogen (GR: 3.8±0.3; FMM: 3.3±0.2 g ha-1 d-1), and total phosphorus (GR: 0.54±0.07; FMM:
283
0.45±0.04 g ha-1 d-1) were consistent with the suspended sediment results post-wildfire and likely
284
other correlates (Table S2). Similarly, monthly suspended sediment yields across a subset of
285
several rivers draining burned and unburned watersheds in the region before and after the fire
286
indicated no fire effect (Figure S1b). Total calcium, used as a proxy measure for dissolved
287
material (Table S2), also showed similar yields between the impacted (0.23±0.01 kg ha-1 d-1) and
288
the control station (0.21±0.01 kg ha-1 d-1) on the Athabasca River (Figure 5). Differences in
289
monthly calcium yields between a subset of rivers draining burned and unburned watersheds
290
were also statistically-similar (Figure S1b). Alternatively, total organic carbon (TOC), most of
291
which is in the dissolved form in the Athabasca River (Glozier et al., 2018), had mean yields
292
greater at the control (91±5 g ha-1 d-1) relative to the impacted (67±3 g ha-1 d-1) station.
293
Substantial differences occurred within and between stations when classifying data by
294
precipitation event and non-event periods. At impacted stations only, peaks of concentrations
295
(Figure 6) and yields during precipitation events were observed relative to non-event periods for
296
suspended sediment (concentrations: 1.7–6.6 times higher; yields: 1.8–10.3x), total phosphorus
297
(1.9–2.6x; 2.0–5.3x), total Kjeldahl nitrogen (1.3–1.4x; 1.2–2.4x), and total lead (2.1–3.3x; 2.0–
298
6.3x). In contrast, such ratios at the control station were lower than impacted stations and were
299
often near unity (concentrations: SS=0.9, TP=1.3, TKN=0.9, Pb=1.1x; yields: SS=1.0, TP: 1.4;
300
TKN: 0.9; Pb: 1.2x). Total calcium and TOC yields were largely unchanged between
301
precipitation and non-event periods (0.8–1.1x) at all stations except for the burned Hangingstone
302
River (Ca: 1.7x; TOC: 1.5x).
303
Despite observing measureable impacts to water quality at burned stations, most
304
monitored parameters rarely exceeded national surface water quality guidelines for the protection
305
of aquatic life (Government of Canada, 2019), or the relative number of exceedances were
306
comparable to pre-fire conditions or those at control stations (Table S4).
307
308
3.2 Historic water quality
Elevated suspended sediment concentrations observed at burned stations did not result
309
from higher flows than sampled historically, nor wholesale changes in the hydrological
310
conditions of impacted river systems (see supplementary material; Figures S4-S6). Specifically,
311
the prediction intervals calculated using historic C-Q relationships for suspended sediment from
312
the burned Athabasca River upstream of Fort McMurray station were exceeded on four
313
occasions after the wildfire (Figure 7). Two of these exceedances occurred on June 09 and July
314
31/16, when the two largest recorded rainfalls in the Fort McMurray region occurred during that
315
summer. Sampling during subsequent years showed no similar notable concentrations. Five
316
exceedances of the upper historic prediction interval occurred in the Clearwater River during
317
precipitation events in 2016, but none during the following years. The Hangingstone River
318
experienced 17 exceedances relative to its historic concentration-discharge relationship.
319
However, these exceedances occurred during precipitation event and non-event periods in 2016,
320
as well as across multiple years after the fire, including during spring runoff in 2017. These C-Q
321
results were likely similar to other particle-bound measures (Table S2). Compared to suspended
322
sediment concentrations, only four exceedances of the upper historic prediction interval for total
323
calcium concentration were observed across all rivers, and likely similar for other dissolved-type
324
chemicals (Table S2). Collectively, suspended sediment concentration exceedances of historic
325
95% prediction intervals in impacted rivers became more frequent as watershed size decreased,
326
and the proportion of burned area and area burned at high severity across watersheds increased
327
(Table 1, Figure 7).
328
4. Discussion
329
330
4.1 Episodic detections of fire-related material in large river systems
Despite (a) naturally turbid and organic-rich water qualities prior to disturbance (Figures
331
3,5,6;(Glozier et al., 2018), (b) the potential for considerable dilution including from substantial
332
subsurface contributions to regional river flow (Gibson et al., 2016), and (c) poor hydrologic
333
connectivity between the landscapes and low-relief rivers (Chanasyk et al., 2003; Devito et al.,
334
2017; Gibson et al., 2016; Ireson et al., 2015), wildfire impacts on river water quality were
335
detectable at the very large basin scale following the 2016 Fort McMurray wildfire. Three
336
independent monitoring approaches reported brief, precipitation-related increases in wildfire-
337
related material (i.e., suspended and dissolved material, nutrients, metals;(Burton et al., 2016;
338
Earl and Blinn, 2003; Reale et al., 2015) in rivers draining burned watersheds relative to a
339
control station or historic, pre-fire data (Figures 3-4,6-7). These findings were also supported by
340
an ash mixing experiment (Figure S2) and other historic data. Together, this weight-of-evidence
341
suggests rainfall-induced surface erosion delivered ash and soil from local, burned hillslopes into
342
large rivers that briefly impacted water quality without shifting overall metrics of river water
343
quality relative to unburned regions (Figure 5). The role of precipitation events in mobilizing
344
particle-bound chemicals across burned landscapes is well-established in the literature
345
(e.g.,(Bladon et al., 2014; Burton et al., 2016; Langhans et al., 2016; Townsend and Douglas,
346
2000) and in more extensively-burned regions of the Boreal Plains (Burke et al., 2005).
347
However, detection of suspended chemicals in large, wetland-dominated, low-slope river
348
watersheds with low proportions of areas burned (Table 1) was a key outcome of this study.
349
Alternatively, total measurements mostly in dissolved form (e.g., total calcium, total organic
350
carbon; others; Table S2) showed inconsistent changes between impacted and control stations, as
351
well as between precipitation and non-event periods, perhaps reflecting uneven burning of peat at
352
depth where non-erosive flows dominate (Devito et al., 2012; Wilkinson et al., 2018).
353
354
4.2 Factors shaping wildfire signal detection in large river systems
Following wildfire, loss of ground cover, decreased surface roughness, and soil crusting
355
and sealing (Larsen et al., 2009) can impede infiltration and expose soils to splash and fluvial
356
erosion (Shakesby and Doerr, 2006). These processes often allow storms to mobilize loads of
357
sediment, nutrients and contaminants downstream (Bayley et al., 1992; Betts and Jones, 2009;
358
Moody and Martin, 2001; Neary et al., 2005). However, the magnitude and duration of these
359
water quality impacts are shaped by the intensity of precipitation events and the erosive potential
360
of impacted landscapes (Langhans et al., 2016). Headwater systems, with typically small,
361
extensively-burned, steep and flashy landscapes, often show substantial sediment yields from
362
burned watersheds relative to those unburned (typically >100x;(Smith et al., 2011) with water
363
quality impacts often detected for months to years (e.g.,(Inbar et al., 1998; Owens et al., 2013).
364
In this study of lower-relief and wetland-dominated landscapes, catchments are large, burn less-
365
extensively and runoff storage is prevalent (Table 1;(Devito et al., 2012). These conditions
366
resulted in suspended sediment yield ratios (impacted:control) ≤1.3x with detectable water
367
quality impacts at only hourly and daily scales (Figures 3-4,6). However, fire-related impacts on
368
water quality in this study (Table 1, Figure 6), as well as those in headwater regions (Rhoades et
369
al., 2011) demonstrate that differing basin sizes, proportions of area burned, and area burned at
370
high severity, can impact water quality of river systems differently. This suggests that
371
mobilization of fire-related materials downstream is likely driven by similar processes across
372
differing landscapes, but are shaped by catchment- and fire-related characteristics. However, we
373
postulate that incomplete hydro-chemical mixing may also play an important role in wildfire
374
signal detection in these large rivers expected to buffer landscape chemical signals (Blöchl et al.,
375
2007).
376
Large rivers can substantially modify upstream and locally-derived terrestrial material via
377
mixing, dilution and sedimentation processes (Temnerud and Bishop, 2005). However, we
378
suggest that the complexities of mixing river waters of different qualities and quantities may
379
have also been responsible for the detection and downstream persistence of wildfire chemical
380
signals observed in these large rivers, particularly in the Athabasca River. Other studies have
381
reported a persistence of plumes/mixing zones downstream of river confluences because of
382
differences in water densities, particle loads, or the chemical composition of sediments (Bouchez
383
et al., 2010; Herrero et al., 2018; Stone et al., 2011). Measurements assessing cross-sectional
384
changes in water quality of the Athabasca River suggest incomplete mixing occurs well
385
downstream of anthropogenic or tributary inputs (Glozier et al., 2018). Following the wildfire
386
and precipitation events, we observed that both low- and high-order tributaries draining severely
387
burned watersheds delivered plumes of ash to the Athabasca River that were visible for several
388
kilometers downstream of confluences (Figure 8). Empirical evidence from the three Athabasca
389
River sondes supported these visual observations. For example, we observed strong differences
390
in turbidity, conductivity, and turbidity-conductivity covariance values measured on the left and
391
right banks of the Athabasca River at the same longitudinal location (Figures 3,4). Higher
392
turbidity values on the right bank were coincident with turbidity and conductivity spikes in the
393
smaller but more extensively-burned Clearwater and Hangingstone rivers that empty upstream
394
from the right bank. The entry of these rivers established a distinct, fire-impacted plume on the
395
right bank of the Athabasca River through to where sonde measurements occurred. Though it
396
was unclear where the right bank plume mixed fully with the Athabasca River downstream, we
397
did observe likely mixing and dilution of fire-influenced plumes on the left bank of the
398
Athabasca. Near the upstream of Fort McMurray station, several severely-burned low-order
399
streams entered the Athabasca River on the left bank and passed though the sonde just
400
downstream, resulting in defined, precipitation-related turbidity and conductivity spikes up to
401
1000 NTU and 350 µS cm-1, respectively (Figure 3,8). Downstream of Fort McMurray on the
402
same bank, another sonde measured turbidity and conductivity spikes often 10-60% as large with
403
lower turbidity-conductivity covariances (Figure 4) compared to the upstream location, and
404
similar to the unimpacted Grand Rapids station. These results suggest efficient mixing of the
405
small, fire-impacted plumes on the left bank with larger volumes of unimpacted Athabasca River
406
water over just a few kilometers of this reach. Regardless, this uneven mixing of waters from
407
impacted tributaries with the large Athabasca River may simultaneously explain observed
408
downstream persistence of wildfire signals, but also overall signal attenuation due to dilution
409
(i.e., relative to headwater regions), both of which were defining observations of this study.
410
411
4.3 Human-related impacts of wildfires in large river systems
The occurrence of and damage from wildland-urban interface fires has increased over the
412
previous decade in western North America (Schoennagel et al., 2017; Westerling et al., 2006).
413
As such an event, the Fort McMurray wildfire burned regions both within and surrounding the
414
remote urban center of Fort McMurray, Alberta. However, this wildfire presented a unique
415
situation where very large, low-relief watersheds were expected to buffer wildfire impacts on
416
local surface water resources. Interestingly, though we observed relatively muted water quality
417
changes in these very large rivers compared to headwater regions, these arguably inconsequential
418
basin-scale effects still resulted in impacts to post-fire drinking water treatment costs borne by
419
the local municipality, similar to headwater regions (Bladon et al., 2014; Emelko et al., 2011;
420
Hohner et al., 2019; Thurton, 2017). Preliminary analyses (unpublished) suggest unmixed
421
plumes in the Athabasca River from heavily-impacted, low-order tributaries discharging
422
upstream of the drinking water treatment plant intake on the river may have been responsible.
423
Thus, exclusive focus on the magnitude and persistence of the impacts of wildfire on surface
424
water quality may fail to uncover its true stresses on surface water resources. This consideration
425
may be particularly applicable to future wildland-urban interface fires as export of complex
426
chemicals (e.g., plastics) from residential, commercial, and industrial settings may present
427
unique water quality conditions in receiving waterbodies of human importance (Oliver et al.,
428
2012).
429
430
5. Conclusions
This study presents a multi-tiered approach to monitoring the flow and water quality of
431
large rivers in a low-relief region affected by an expansive and severe wildfire. While broader
432
wildfire impacts to water quality of regional rivers were markedly lower than commonly
433
reported in smaller watersheds of greater relief, water quality impacts were nonetheless observed
434
across a range of very large river basin scales. Novel use of continuous flow and general
435
chemical monitoring was key to identifying impacted periods in the rivers studied herein, similar
436
to few other investigations (Cooke et al., 2016; Dahm et al., 2015; Mast et al., 2016). In
437
particular, our study emphasizes the importance of considering the spatial frame of reference
438
(entire basin vs. local) when evaluating wildfire impacts on water quality in large river basins.
439
The proximity and scale of this and other recent wildfires to major urban municipalities further
440
highlights the importance of shifting North American wildfire regimes as increasing threats to
441
community surface water resources.
442
443
444
445
446
447
448
449
450
451
452
6. Acknowledgements
Funding for this work was provided by Canada’s emergency disaster recovery fund and
the Alberta Government. We would like to thank the tireless work during the emergency by
Alberta Environment and Parks water quality technicians, particularly Chris Ware, Monica
Polutranko, Trina Ball, Shelley Manchur, Brittany Kereliuk, Meghan House, Sarah Lamb, Tye
Dubrule and Jessica Pope. We also thank government Data Management staff including Chris
Rickard, Doreen LeClair, Lisa Reinbolt and Jenny Pham. Burn severity mapping was graciously
provided by the Wildfire Operations sections of Alberta Agriculture and Forestry. We appreciate
input by Dr. Bill Donahue, Dr. Paul Drevnick, Dr. Jennifer Graydon and Dr. Fred Wrona during
its development. Martin Davies provided valuable interpretation of sonde water quality data.
Finally, we thank two anonymous reviewers for taking the time to improve this manuscript.
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
7. References
Alam, R., Islam, S., Mosely, E., Thomas, S., Dowdell, V., Doel, D., 2018. Rapid impact
assessment of Fort McMurray Wildfire. Edmonton.
Bayley, S.E., Schindler, D.W., Parker, B.R., Stainton, M.P., Beaty, K.G., 1992. Effects of forest
fire and drought on acidity of a base-poor boreal forest stream: similarities between climatic
warming and acidic precipitation. Biogeochemistry 17, 191–204.
https://doi.org/10.1007/BF00004041
Belillas, C.M., Rodà, F., 1993. The effects of fire on water quality, dissolved nutrient losses and
the export of particulate matter from dry heathland catchments. J. Hydrol. 150, 1–17.
https://doi.org/10.1016/0022-1694(93)90153-Z
Betts, E.F., Jones, J.B., 2009. Impact of Wildfire on Stream Nutrient Chemistry and Ecosystem
Metabolism in Boreal Forest Catchments of Interior Alaska. Arctic, Antarct. Alp. Res. 41,
407–417. https://doi.org/10.1657/1938-4246-41.4.407
Bladon, K.D., Emelko, M.B., Silins, U., Stone, M., 2014. Wildfire and the future of water
supply. Environ. Sci. Technol. 48, 8936–8943. https://doi.org/10.1021/es500130g
Blöchl, G., Ardoin-Bardin, S., Bonell, M., Dorninger, M., Goodrich, D., Gutknecht, D.,
Matamoros, D., Merz, B., Shand, P., Szolgay, J., 2007. At what scales do climate variability
and land cover change impact on flooding and low flows? Hydrol. Process. 21, 1241–1247.
https://doi.org/10.1002/hyp.6669
Bouchez, J., Lajeunesse, E., Gaillardet, J., France-Lanord, C., Dutra-Maia, P., Maurice, L., 2010.
Turbulent mixing in the Amazon River: The isotopic memory of confluences. Earth Planet.
Sci. Lett. 290, 37–43. https://doi.org/10.1016/j.epsl.2009.11.054
Bracken, L.J., Croke, J., 2007. The concept of hydrological connectivity and its contribution to
understanding runoff-dominated geomorphic systems. Hydrol. Process. 21, 1749–1763.
https://doi.org/10.1002/hyp.6313
Burd, K., Tank, S.E., Dion, N., Quinton, W.L., Spence, C., Tanentzap, A.J., Olefeldt, D., 2018.
Seasonal shifts in export of DOC and nutrients from burned and unburned peatland-rich
catchments, Northwest Territories, Canada. Hydrol. Earth Syst. Sci. 22, 4455–4472.
https://doi.org/10.5194/hess-22-4455-2018
Burke, J.M., Prepas, E.E., Pinder, S., 2005. Runoff and phosphorus export patterns in large
forested watersheds on the western Canadian Boreal Plain before and for 4 years after
wildfire. J. Environ. Eng. Sci. 4, 319–325. https://doi.org/10.1139/s04-072
Burton, C. a., Hoefen, T.M., Plumlee, G.S., Baumberger, K.L., Backlin, A.R., Gallegos, E.,
Fisher, R.N., 2016. Trace Elements in Stormflow, Ash, and Burned Soil following the 2009
Station Fire in Southern California. PLoS One 11, e0153372.
https://doi.org/10.1371/journal.pone.0153372
Chanasyk, D.S., Whitson, I.R., Mapfumo, E., Burke, J.M., Prepas, E.E., 2003. The impacts of
forest harvest and wildfire on soils and hydrology in temperate forests: A baseline to
develop hypotheses for the Boreal Plain. J. Environ. Eng. Sci. 2, S51–S62.
https://doi.org/10.1139/s03-034
Cooke, C.A., Schwindt, C., Davies, M., Donahue, W.F., Azim, E., 2016. Initial environmental
impacts of the Obed Mountain coal mine process water spill into the Athabasca River
(Alberta, Canada). Sci. Total Environ. 557–558, 502–509.
https://doi.org/10.1016/j.scitotenv.2016.03.058
Dahm, C.N., Candelaria-Ley, R.I., Reale, C.S., Reale, J.K., Van Horn, D.J., 2015. Extreme water
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
quality degradation following a catastrophic forest fire. Freshw. Biol.
https://doi.org/10.1111/fwb.12548
Depante, M., Petrone, R.M., Devito, K.J., Kettridge, N., Macrae, M.L., Mendoza, C.,
Waddington, J.M., 2018. Potential influence of nutrient availability along a hillslope:
Peatland gradient on aspen recovery following fire. Ecohydrology 11, 1–13.
https://doi.org/10.1002/eco.1955
Devito, K., Mendoza, C., Qualizza, C., 2012. Conceptualizing water movement in the Boreal
Plains. Implications for watershed reconstruction. Synthesis report preapred for the
Canadian Oil Sands Network for Research and Developement 164.
Devito, K.J., Hokanson, K.J., Moore, P.A., Kettridge, N., Anderson, A.E., Chasmer, L.,
Hopkinson, C., Lukenbach, M.C., Mendoza, C.A., Morissette, J., Peters, D.L., Petrone,
R.M., Silins, U., Smerdon, B., Waddington, J.M., 2017. Landscape controls on long-term
runoff in subhumid heterogeneous Boreal Plains catchments. Hydrol. Process. 31, 2737–
2751. https://doi.org/10.1002/hyp.11213
Diffenbaugh, N.S., Swain, D.L., Touma, D., 2015. Anthropogenic warming has increased
drought risk in California. Proc. Natl. Acad. Sci. 112, 3931–3936.
https://doi.org/10.1073/pnas.1422385112
Earl, S.R., Blinn, D.W., 2003. Effects of wildfire ash on water chemistry and biota. Freshw. Biol.
1015–1030.
Ebel, B.A., Mirus, B.B., 2014. Disturbance hydrology: challenges and opportunities. Hydrol.
Process. 28, 5140–5148. https://doi.org/10.1002/hyp.10256
Elhadi, N., Harrington, A., Hill, I., Lau, Y.L., Krishnappan, B.G., 1984. River mixing—A stateof-the-art report. Can. J. Civ. Eng. 11, 585–609. https://doi.org/10.1139/l84-076
Emelko, M.B., Silins, U., Bladon, K.D., Stone, M., 2011. Implications of land disturbance on
drinking water treatability in a changing climate: Demonstrating the need for “ source water
supply and protection” strategies. Water Res. 45, 461–472.
https://doi.org/10.1016/j.watres.2010.08.051
Emelko, M.B., Stone, M., Silins, U., Allin, D., Collins, A.L., Williams, C.H.S., Martens, A.M.,
Bladon, K.D., 2016. Sediment-phosphorus dynamics can shift aquatic ecology and cause
downstream legacy effects after wildfire in large river systems. Glob. Chang. Biol. 22,
1168–1184. https://doi.org/10.1111/gcb.13073
Environment and Climate Change Canada, 2018. Canadian Climate Normals 1981-2010 [WWW
Document]. URL http://climate.weather.gc.ca/climate_normals/ (accessed 2.14.18).
Gerla, P.J., Galloway, J.M., 1998. Water quality of two streams near Yellowstone Park,
Wyoming, following the 1988 Clover-Mist wildfire. Environ. Geol. 36, 127–136.
https://doi.org/10.1007/s002540050328
Gibson, J.J., Yi, Y., Birks, S.J., 2016. Isotope-based partitioning of streamflow in the oil sands
region, northern Alberta: Towards a monitoring strategy for assessing flow sources and
water quality controls. J. Hydrol. Reg. Stud. 5, 131–148.
https://doi.org/10.1016/j.ejrh.2015.12.062
Glozier, N.E., Pippy, K., Levesque, L., Ritcey, A., Armstrong, B., Tobin, O., Cooke, C.A.,
Conly, M., Dirk, L., Epp, C., Gue, A., Hazewinkel, R., Keet, E., Lindeman, D., Maines, J.,
Syrgiannis, J., Su, M., Tumber, V., 2018. Surface Water Quality of the Athabasca, Peace
and Slave Rivers and Riverine Waterbodies within the Peace-Athabasca Delta. Oil Sands
Monit. Progr. Tech. Rep. Ser. No. 1.4. 64pp.
Government of Canada, 2019. Canadian Environmental Quality Guidelines Summary Table
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
[WWW Document]. URL http://st-ts.ccme.ca/ (accessed 9.20.18).
Gresswell, R.E., 1999. Fire and Aquatic Ecosystems in Forested Biomes of North America.
Trans. Am. Fish. Soc. 128, 193–221. https://doi.org/10.1577/15488659(1999)128<0193:FAAEIF>2.0.CO;2
Hanes, C.C., Wang, X., Jain, P., Parisien, M.-A., Little, J.M., Flannigan, M.D., 2018. Fireregime changes in Canada over the last half century. Can. J. For. Res. 49, 256–269.
https://doi.org/10.1139/cjfr-2018-0293
Herrero, H.S., Díaz Lozada, J.M., García, C.M., Szupiany, R.N., Best, J., Pagot, M., 2018. The
influence of tributary flow density differences on the hydrodynamic behavior of a confluent
meander bend and implications for flow mixing. Geomorphology 304, 99–112.
https://doi.org/10.1016/j.geomorph.2017.12.025
Hohner, A.K., Rhoades, C.C., Wilkerson, P., Rosario-Ortiz, F.L., 2019. Wildfires alter forest
watersheds and threaten drinking water quality. Acc. Chem. Res. 52, 1234–1244.
https://doi.org/10.1021/acs.accounts.8b00670
Inbar, M., Tamir, M., Wittenberg, L., 1998. Runoff and erosion processes after a forest fire in
Mount Carmel, a Mediterranean area. Geomorphology 24, 17–33.
https://doi.org/10.1016/S0169-555X(97)00098-6
Ireson, A.M., Barr, A.G., Johnstone, J.F., Mamet, S.D., van der Kamp, G., Whitfield, C.J.,
Michel, N.L., North, R.L., Westbrook, C.J., DeBeer, C., Chun, K.P., Nazemi, A., Sagin, J.,
2015. The changing water cycle: the Boreal Plains ecozone of Western Canada. Wiley
Interdiscip. Rev. Water 2, 505–521. https://doi.org/10.1002/wat2.1098
Jolly, W.M., Cochrane, M. a., Freeborn, P.H., Holden, Z. a., Brown, T.J., Williamson, G.J.,
Bowman, D.M.J.S., 2015. Climate-induced variations in global wildfire danger from 1979
to 2013. Nat. Commun. 6, 1–11. https://doi.org/10.1038/ncomms8537
Kelly, E.N., Schindler, D.W., St. Louis, V.L., Donald, D.B., Vladicka, K.E., 2006. Forest fire
increases mercury accumulation by fishes via food web restructuring and increased mercury
inputs. Proc. Natl. Acad. Sci. 103, 19380–19385. https://doi.org/10.1073/pnas.0609798104
Lane, P.N.J., Sheridan, G.J., Noske, P.J., Sherwin, C.B., 2008. Phosphorus and nitrogen exports
from SE Australian forests following wildfire. J. Hydrol. 361, 186–198.
https://doi.org/10.1016/j.jhydrol.2008.07.041
Langhans, C., Smith, H.G., Chong, D.M.O., Nyman, P., Lane, P.N.J., Sheridan, G.J., 2016. A
model for assessing water quality risk in catchments prone to wildfire. J. Hydrol. 534, 407–
426. https://doi.org/10.1016/j.jhydrol.2015.12.048
Larsen, I.J., MacDonald, L.H., Brown, E., Rough, D., Welsh, M.J., Pietraszek, J.H., Libohova,
Z., de Dios Benavides-Solorio, J., Schaffrath, K., 2009. Causes of Post-Fire Runoff and
Erosion: Water Repellency, Cover, or Soil Sealing? Soil Sci. Soc. Am. J. 73, 1393.
https://doi.org/10.2136/sssaj2007.0432
Mast, M.A., Murphy, S.F., Clow, D.W., Penn, C.A., Sexstone, G.A., 2016. Water-quality
response to a high-elevation wildfire in the Colorado Front Range. Hydrol. Process. 30,
1811–1823. https://doi.org/10.1002/hyp.10755
Mell, W.E., Manzello, S.L., Maranghides, A., Butry, D., Rehm, R.G., 2010. The wildland urban interface fire problem - current approaches and research needs. Int. J. Wildl. Fire 19,
238. https://doi.org/10.1071/wf07131
Minshall, G.W., Brock, J.T., Andrews, D.A., Robinson, C.T., 2001. Water quality, substratum
and biotic responses of five central Idaho (USA) streams during the first year following the
Mortar Creek fire. Int. J. Wildl. Fire 10, 185–199. https://doi.org/10.1071/WF01017
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
Moody, J.A., Martin, D.A., 2001. Initial hydrologic and geomorphic response following a
wildfire in the Colorado front range. Earth Surf. Process. Landforms 26, 1049–1070.
https://doi.org/10.1002/esp.253
Neary, D.G., Ryan, K.C., Debano, L.F., 2005. Wildland Fire in Ecosystems Effects of Fire on
Soil and Water, United States Department of Agriculture Gen. Tech. Rep. Ogden, Utah.
https://doi.org/10.1021/ac102195x
Olefeldt, D., Turetsky, M.R., Blodau, C., 2013. Altered Composition and Microbial versus UVMediated Degradation of Dissolved Organic Matter in Boreal Soils Following Wildfire.
Ecosystems 16, 1396–1412. https://doi.org/10.1007/s10021-013-9691-y
Oliver, A.A., Reuter, J.E., Heyvaert, A.C., Dahlgren, R.A., 2012. Water quality response to the
Angora Fire, Lake Tahoe, California. Biogeochemistry. https://doi.org/10.1007/s10533-0119657-0
Owens, P.N., Giles, T.R., Petticrew, E.L., Leggat, M.S., Moore, R.D., Eaton, B.C., 2013. Muted
responses of streamflow and suspended sediment flux in a wildfire-affected watershed.
Geomorphology 202, 128–139. https://doi.org/10.1016/j.geomorph.2013.01.001
Prepas, E.E., Burke, J.M., Chanasyk, D.S., Smith, D.W., Putz, G., Gabos, S., Chen, W., Millions,
D., Serediak, M., 2003. Impact of wildfire on discharge and phosphorus export from the
Sakwatamau watershed in the Swan Hills, Alberta, during the first two years. J. Environ.
Eng. Sci. https://doi.org/10.1139/s03-036
Radeloff, V.C., Helmers, D.P., Kramer, H.A., Mockrin, M.H., Alexandre, P.M., Bar-Massada,
A., Butsic, V., Hawbaker, T.J., Martinuzzi, S., Syphard, A.D., Stewart, S.I., 2018. Rapid
growth of the US wildland-urban interface raises wildfire risk. Proc. Natl. Acad. Sci. 115,
3314–3319. https://doi.org/10.1073/pnas.1718850115
Reale, J.K., Van Horn, D.J., Condon, K.E., Dahm, C.N., 2015. The effects of catastrophic
wildfire on water quality along a river continuum. Freshw. Sci. 34, 1426–1442.
https://doi.org/10.1086/684001
Rhoades, C.C., Entwistle, D., Butler, D., 2011. The influence of wildfire extent and severity on
streamwater chemistry, sediment and temperature following the Hayman Fire, Colorado.
Int. J. Wildl. Fire 20, 430. https://doi.org/10.1071/wf09086
Rust, A.J., Hogue, T.S., Saxe, S., McCray, J., 2018. Post-fire water-quality response in the
western United States. Int. J. Wildl. Fire 27, 203–216. https://doi.org/10.1071/wf17115
Santín, C., Doerr, S.H., Otero, X.L., Chafer, C.J., 2015. Quantity, composition and water
contamination potential of ash produced under different wildfire severities. Environ. Res.
142, 297–308. https://doi.org/10.1016/j.envres.2015.06.041
Schlesinger, W., 2007. Biogeochemistry: an analysis of global change, 2nd Ed. ed. Academic
Press, London.
Schoennagel, T., Balch, J.K., Brenkert-Smith, H., Dennison, P.E., Harvey, B.J., Krawchuk, M.
a., Mietkiewicz, N., Morgan, P., Moritz, M. a., Rasker, R., Turner, M.G., Whitlock, C.,
2017. Adapt to more wildfire in western North American forests as climate changes. Proc.
Natl. Acad. Sci. 114, 4582–4590. https://doi.org/10.1073/pnas.1617464114
Scott, D., Versfeld, D.B., Lesch, W., 1998. Erosion and sediment yield in relation to
afforestation and fire in the mountains of the western cape province, south africa. South
African Geogr. J. 80, 52–59. https://doi.org/10.1080/03736245.1998.9713644
Shakesby, R.A., Doerr, S.H., 2006. Wildfire as a hydrological and geomorphological agent.
Earth-Science Rev. 74, 269–307. https://doi.org/10.1016/j.earscirev.2005.10.006
Silins, U., Bladon, K.D., Kelly, E.N., Esch, E., Spence, J.R., Stone, M., Emelko, M.B., Boon, S.,
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
Wagner, M.J., Williams, C.H.S., Tichkowsky, I., 2014. Five-year legacy of wildfire and
salvage logging impacts on nutrient runoff and aquatic plant, invertebrate, and fish
productivity. Ecohydrology 7, 1508–1523. https://doi.org/10.1002/eco.1474
Smith, H.G., Sheridan, G.J., Lane, P.N.J., Nyman, P., Haydon, S., 2011. Wildfire effects on
water quality in forest catchments: A review with implications for water supply. J. Hydrol.
396, 170–192. https://doi.org/10.1016/j.jhydrol.2010.10.043
Stone, M., Emelko, M.B., Droppo, I.G., Silins, U., 2011. Biostabilization and erodibility of
cohesive sediment deposits in wildfire-affected streams. Water Res. 45, 521–534.
https://doi.org/10.1016/j.watres.2010.09.016
Temnerud, J., Bishop, K., 2005. Spatial variation of streamwater chemistry in two Swedish
boreal catchments: Implications for environmental assessment. Environ. Sci. Technol. 39,
1463–1469. https://doi.org/10.1021/es040045q
Thurton, D., 2017. Fort McMurray seeing big spike in water treatment costs. Can. Broadcast.
Corp. https//www.cbc.ca/news/canada/edmonton/fort-mcmurray-wildfire-water-treatmentcosts-contaminants-1.3973249.
Townsend, S.A., Douglas, M.M., 2000. The effect of three fire regimes on stream water quality,
water yield and export coefficients in a tropical savanna (northern Australia). J. Hydrol.
https://doi.org/10.1016/S0022-1694(00)00165-7
Westerling, A.L., Hidalgo, H.G., Cayan, D.R., Swetnam, T.W., 2006. Warming and earlier
spring increase Western U.S. forest wildfire activity. Science (80-. ).
https://doi.org/10.1126/science.1128834
Wilkinson, S.L., Moore, P.A., Flannigan, M.D., Wotton, B.M., Waddington, J.M., 2018. Did
enhanced afforestation cause high severity peat burn in the Fort McMurray Horse River
wildfire? Environ. Res. Lett. 13, 014018.
Writer, J.H., Hohner, A., Oropeza, J., Schmidt, A., Cawley, K., Rosario-Ortiz, F.L., 2014. Water
treatment implications after the High Park Wildfire in Colorado. J. Am. Water Works
Assoc. 106, 85–86. https://doi.org/10.5942/jawwa.2014.106.0055.
663
664
665
8. Tables
Table 1 Drainage area and burn metrics at water quality stations operating on rivers draining
burned (B) and unburned (U) areas following the 2016 Fort McMurray wildfire.
666
9.
Figures
667
668
669
670
Figure 1 Map of the lower Athabasca River in northeastern Alberta showing the extent and
severity of the May 2016 Fort McMurray wildfire (see supplementary material for methods) and
water monitoring stations. Inset (a.) shows monitoring station clustering around Fort McMurray.
671
672
673
674
Figure 2 Log-scaled historic flow quantiles and 2016 mean daily hydrographs and hyetographs
of monitored hydrometric stations (Q) and their closest meteorological stations (P) of unburned
(top panel) and burned watersheds in the lower Athabasca River of northeastern Alberta. Daily
precipitation events ≥10 mm and five subsequent days are highlighted in yellow.
675
676
677
678
Figure 3 Hourly time series of sonde turbidity and specific conductivity, and discharge in
unburned (top panel) and burned river stations in the lower Athabasca River region following the
2016 Fort McMurray wildfire. Precipitation events (daily ≥10 mm + 5 days) are highlighted in
yellow.
679
680
681
682
683
Figure 4 Schematic of sonde turbidity and specific conductivity covariance, mean daily flow and
precipitation events (≥10 mm and five subsequent days) measured at an unburned control station
and five downstream stations within burned landscapes. All graphs share the same axis metrics
and time periods as the control station graph. LB and RB denote left bank and right bank of the
river. *denote four, 15-minute measurements beyond the scale on the covariance axes.
684
685
686
687
Figure 5 Box and whisker plots (median-mean [dashed]-25/75-10/90 percentiles-outliers) of
Athabasca River watershed yields of selected chemicals from an upstream unburned station and
a downstream burned station from daily sampling (May–August) following the 2016 Fort
McMurray wildfire.
688
689
690
691
Figure 6 Time series of mean daily concentrations of selected chemicals collected by automated
and grab samples for unburned (top panel) and burned river stations in the lower Athabasca
River region following the 2016 Fort McMurray wildfire. Precipitation events (daily ≥ 10 mm +
5 days) are highlighted in yellow.
692
693
694
695
Figure 7 Log-scaled concentration-discharge relationships of suspended sediment and total
calcium (Ca) of historic and post-fire data from three rivers draining burned watersheds in the
lower Athabasca River region following the 2016 Fort McMurray wildfire. 95% prediction
intervals and linear fit lines on historic data are also shown.
696
697
698
Figure 8 Photos of high-order (left panel; Horse River) and low-order (right panel; Little
Fisheries Creek) rivers draining recently burned watersheds and discharging into the Athabasca
River just upstream of Fort McMurray after a June 2016 precipitation event. (Credit: S. Hustins).
Station
Athabasca R. at Grand Rapids (U)
Athabasca R. u/s Ft. McMurray (B)
Clearwater R. near the mouth (B)
Hangingstone R. at Ft. McMurray (B)
Watershed
area
(km2)
Mean
watershed
slope
(%)
Burned area:
watershed area
(%)
High severity
burned area:
burned area
(%)
94,464
98,013
31,936
903
3.3
3.3
3.4
3.5
0
0.4
10.2
21.5
45.6
52.1
57.1
Highlights
•
•
•
•
Large rivers show wildfire water quality signatures after precipitation events
Continuous monitoring captures short-term wildfire impacts on river water quality
Suspended sediment concentrations after fire were higher than historic records
Incomplete mixing within large rivers impacts detection of altered water quality
Declaration of interests
☒ 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.
☐The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests: