HESSHydrology and Earth System SciencesHESSHydrol. Earth Syst. Sci.1607-7938Copernicus PublicationsGöttingen, Germany10.5194/hess-26-5917-2022Simulating the hydrological impacts of land use conversion from annual crop
to perennial forage in the Canadian Prairies using the Cold Regions
Hydrological Modelling platformSimulating the hydrological impacts of land use conversion using CRHMCordeiroMarcos R. C.LiangKangkliang6@umd.eduhttps://orcid.org/0000-0002-8979-813XWilsonHenry F.VanrobaeysJasonLobbDavid A.FangXinghttps://orcid.org/0000-0002-4333-4815PomeroyJohn W.https://orcid.org/0000-0002-4782-7457Department of Animal Science, University of Manitoba, 12 Dafoe Road,
Winnipeg, MB R3T 2N2, CanadaScience and Technology Branch, Agriculture and Agri-Food Canada, 303 Main Street, Winnipeg, MB R3C 3G7, CanadaEarth System Science Interdisciplinary Center, University of Maryland,
5825 University Research Court, College Park, MD 20740, USAScience and Technology Branch, Agriculture and Agri-Food Canada, 2701
Grand Valley Road, Brandon, MB R7A 5Y3, CanadaScience and Technology Branch, Agriculture and Agri-Food Canada, 101
Route 100, Morden, MB R6M 1Y5, CanadaDepartment of Soil Science, University of Manitoba, 13 Freedman
Crescent, Winnipeg, MB R3T 2N2, CanadaCentre for Hydrology, University of Saskatchewan, 116a - 1151 Sidney
Street, Canmore, AB T1W 3G1, Canada
The Red River is one of the largest contributing sources
of discharge and nutrients to the world's 10th largest freshwater lake,
Lake Winnipeg. Conversion of large areas of annual cropland to
perennial forage has been proposed as a strategy to reduce both flooding
and nutrient export to Lake Winnipeg. Such reductions could occur either via a reduction in the concentration of nutrients in runoff or through changes in
the basin-scale hydrology, resulting in a lower water yield and the concomitant
export of nutrients. This study assessed the latter mechanism by using the
physically based Cold Regions Hydrological Modelling platform to examine the
hydrological impacts of land use conversion from annual crops to perennial
forage in a subbasin of the La Salle River basin in Canada. This basin is a typical
agricultural subbasin in the Red River Valley, characterised by flat
topography, clay soils, and a cold subhumid, continental climate. Long-term
simulations (1992–2013) of the major components of water balance were
compared between canola and smooth bromegrass, representing a conversion
from annual cropping systems to perennial forage. An uncertainty framework
was used to represent a range of fall soil saturation status (0 % to 70 %),
which governs the infiltration to frozen soil in the subsequent spring. The model
simulations indicated that, on average, there was a 36.5 ± 6.6 %
(36.5 ± 7.2 mm) reduction in annual cumulative discharge and a
29.9 ± 16.3 % (2.6 ± 1.6 m3 s-1) reduction in annual
peak discharge due to forage conversion over the assessed period. These
reductions were driven by reduced overland flow 52.9 ± 12.8 %
(28.8 ± 10.1 mm), increased peak snowpack (8.1 ± 1.5 %,
7.8 ± 1.6 mm), and enhanced infiltration to frozen soils (66.7 ± 7.7 %, 141.5 ± 15.2 mm). Higher cumulative evapotranspiration (ET)
from perennial forage (34.5 ± 0.9 %, 94.1 ± 2.5 mm) was also
predicted by the simulations. Overall, daily soil moisture under perennial
forage was 18.0 % (57.2 ± 1.2 mm) higher than that of crop simulation,
likely due to the higher snow water equivalent (SWE) and enhanced
infiltration. However, the impact of forage conversion on daily soil
moisture varied interannually. Soil moisture under perennial forage stands
could be either higher or lower than that of annual crops, depending on
antecedent spring snowmelt infiltration volumes.
Introduction
The Red River Valley in Manitoba, Canada, is prone to large overland flooding events
and is one of the largest sources of water and nutrients to Lake Winnipeg.
In recent decades, the frequency of flooding, the intensification of
agricultural activities in the basin, and environmental implications on
associated water courses have come into increased focus (Benoy et al.,
2016; Mccullough et al., 2012; Rattan et al., 2017; Painter et al., 2021;
Cordeiro et al., 2017). Since the mid-1990s, an increase in runoff during
the spring snowmelt season and an increase in the frequency of spring flooding has been
observed in the Red River Valley (Ehsanzadeh et al., 2012;
Schindler et al., 2012). This, combined with the amplified nutrient
availability as a result of the intensification of agricultural production
in the region, is considered to be the major driver of the eutrophication of
Lake Winnipeg (Mccullough et al., 2012; Schindler et al., 2012; Yates et
al., 2012). Conversion of some portions of land from annual cropping systems
to perennial forage in intensive agricultural basins has been proposed as
a means to increase agricultural system resilience in frequently flooded
locations, increase carbon sequestration, increase infiltration, and promote water
retention (Kharel et al., 2016; Hutchinson et al., 2007). However, the
hydrologic changes associated with the broad-scale conversion of large portions
of the Red River Valley to perennial forage remain understudied.
From a hydrological perspective, previous studies carried out in cold
regions suggest that nutrient export from cropland is mainly driven by
snowmelt runoff (Corriveau et al., 2013; Uusi-Kamppa et al., 2012;
Cade-Menun et al., 2013). Therefore, a reduction in nutrient loads could be
achieved by reducing agricultural runoff (Li et al., 2011; Liu et al.,
2014). Hydrological alterations that reduce runoff volume could also help to
address downstream flooding problems, which are also a significant challenge
associated with the flat topography of the Canadian Prairies under intensive
agriculture (Bower, 2007). Several major floods have occurred in recent years in the Canadian
Prairies, generating concern over causal factors ranging from climate change to
agricultural management practices (Buttle et al., 2016).
Conversion from cropland to perennial forage has been observed to cause
fundamental changes in the hydrology of small Canadian Prairie drainage
basins, such as increases in snow trapping, snowmelt infiltration to frozen
soils, and annual evapotranspiration as well as decreased soil moisture;
together, these changes have been attributed to causing reduced runoff and
declining wetland storage (van der Kamp et al., 2003). However, changes in
hydrology have been mainly described as a result of field-scale observations
in Saskatchewan and were made outside the higher rainfall and warmer climate
of the Red River Valley of Manitoba, which also has a high incidence of clay
soils. These differences make it difficult to extrapolate the impact of
forage conversion to broader scales due to the role of landscape
physiography (e.g., soils texture, topography) and climate on hydrology (van
der Kamp et al., 2003).
However, from a nutrient export perspective, research also suggests that
conversion from cropland to perennial forage could result in increased
nutrient losses in the years directly following conversion. For example, a
field experiment carried out by Liu et al. (2014) observed increased phosphorus and ammonia losses from perennial forage planted on former cropland and
attributed this pattern to increased concentrations following nutrient
release from forage residue due to freezing. Likewise, Cade-Menun et
al. (2013) found significantly more nitrogen in pasture runoff than cropland,
despite no significant difference in total phosphorus loss in runoff between
those land covers.
These contrasting perspectives suggest that comprehensive studies
integrating long-term land use (e.g., land cover and land management),
climate, and physiography (e.g., soil properties, topography, and drainage
conditions) are still required to understand the impacts of land conversion
on water quality in the Lake Winnipeg basin. Full investigation of nutrient
export is complex at large spatial scales, requiring available data on
nutrient management practices adopted at the field scale (e.g., fertiliser
application rates, times, and source; Mikkelsen, 2011). Research to more fully
define the factors controlling nutrient dynamics in the region is ongoing
(e.g., Liu et al., 2019), and continued research is
required before the influence of forage conversion on nutrient source can be
accurately represented in a modelling framework. Particularly, the relative
importance of the freeze–thaw release of nutrients from frozen vegetation,
stratification of nutrients near the soil surface, and legacy of past
nutrient inputs cannot be differentiated in those observational studies
cited above.
On the other hand, assessing hydrological dynamics at large scales is more
feasible due to the availability of ancillary data (e.g., soils databases
and weather records; Cordeiro et al., 2018, 2019),
hydrometric observations (ECCC, 2018), and modelling tools (Beven, 2011).
The Cold Regions Hydrological Modelling (CRHM) platform was specifically
developed to address the challenges of modelling cold-region hydrology in
the context of the Canadian Prairies' physiography (Pomeroy et al., 2007, 2022). CRHM adopts a physically based representation of key
hydrological processes in the Canadian Prairies such as blowing snow
transport, redistribution and sublimation of snow, infiltration to frozen
soils, energy balance snowmelt, snowmelt runoff, the combination of aerodynamic
and energy balance evapotranspiration, soil moisture redistribution, runoff,
and streamflow routing (Fang et al., 2010; Pomeroy et al., 2007). The
platform is also robust for scenario assessment of land use and climate
change (Fang and Pomeroy, 2020; He et al., 2021; Pomeroy and Krogh, 2019),
and it is under constant development to incorporate recent advances in
physically based descriptions of hydrological processes (e.g., Fang et al.,
2013; Harder and Pomeroy, 2014).
Location and land cover of the La Salle River basin
(LS-05OG008) used for model simulations (AAFC, 2013).
The objective of this research was to evaluate the basin-scale hydrological
impacts of land use conversion from annual crop to perennial forage in the
Canadian Prairies using the CRHM platform framework. A custom model was
developed using CRHM to represent the typical perennial forage and cropping
conditions in the Red River Valley. The hydrological impacts were assessed
by comparing simulations between annual crop and perennial forage models.
The analysis focused on changes in annual discharge volumes and peak
discharge rates but also considered other water balance components such as
surface runoff, snow water equivalent (SWE) accumulation, infiltration, soil
moisture, and seasonal evapotranspiration (ET) volumes.
Material and methodsStudy area
CRHM simulations were conducted in a 169 km2 subbasin of the La Salle
River basin (LS-05OG008), namely the La Salle River near Elie (05OG008;
Fig. 1) defined by Environment and Climate Change Canada's Water Survey
Canada (WSC). The La Salle River is a tributary to the larger Red River,
which drains northward to Lake Winnipeg. The basin is located near the
eastern edge of the Canadian Prairies in the central plains of Manitoba,
Canada (Graveline and Larter, 2006). The surficial geology of the area
consists of lacustrine clay deposits in the former bed of glacial Lake
Agassiz, which is a lower, dark-grey clay and a thinner upper unit of
lighter coloured, calcareous silty clay, with the surface texture being
predominantly clayey (La Salle Redboine Conservation District, 2007). Higher-order taxa in the Canadian System of Soil Classification (i.e., Vertisols)
correspond to Boroll soils in the US soils taxonomy (Agriculture and
Agri-Food Canada, 1998). The mean annual temperature in the study area is
2.5 ∘C, with a mean summer temperature of 16 ∘C and a mean
winter temperature of -13∘C, which are typical of the Canadian
Prairies; located in the eastern portion of this ecozone, precipitation
amounts are higher than those further west, with a mean annual
precipitation of around 560 mm (out of which approximately 25 % occurs as
snowfall) and an average mean annual potential evapotranspiration of about 834 mm
(La Salle Redboine Conservation District, 2007).
Annual crop condition simulations
A detailed description of the hydrological model used for annual crop
simulations, including input datasets, basin delineation, hydrological
response unit (HRU) definition, and model parameterisation, was given by
Cordeiro et al. (2017). Briefly, a 90 m digital elevation model (DEM)
derived from the NASA Shuttle Radar Topography Mission (SRTM) data and soil
datasets with scales ranging from 1:20 000 to 1:126 720 from the Manitoba
Land Initiative (MLI) database were used to delineate the modelled basin,
which consisted of four subbasins (Fig. 1). Cropland comprises 87 % of
the land use in the modelled basin (AAFC, 2013). Six annual crops (i.e.,
potato, carrot, soybean, spring wheat, winter wheat, and canola) and
alfalfa, which are usually grown in this area, were combined into five
different cropping systems (i.e., irrigated vegetables, pulse non-row,
oilseed–spring cereal, fall cereal, and perennial forage) to create 17
different crop HRUs using the land use split method (La Salle Redboine
Conservation District, 2007).
This method allows the representation of crop rotations in the model in a
static fashion by distributing the different crops within a cropping system
throughout the acreage of the cropping system in a single year (Cordeiro et
al., 2017). Canola and wheat were the major crops in these cropping systems.
Combined, these crops occupied a land basis ranging from 81 % to 95 % in
each subbasin.
List of hydrological response units (HRUs) in the La Salle
River basin used in annual crop and perennial forage simulations. The same
HRUs were present in each subbasin.
∗ The first two letters indicate the cropping system/land use, the third and
fourth letters indicate the crop, and letter(s) after the slash indicate the soil
texture. The abbreviations used for land use are as follows: CA – canola, AL – alfalfa, CR – carrot, FC – fall cereal, FY – feedlot, IV – irrigated vegetable, OS – oilseed, PF – perennial forage, PO – potato, PR – pulse non-row, RC – river channel, SB – smooth brome, SW – spring wheat, SY – soybean, URLD – urban (low density), URMD – urban (medium density), WETL/WA – wetland/water, and WW – winter wheat. The abbreviations used for soil texture are as follows: C – clay, SICL – silty clay loam, SIC – silty clay, and SIL – silty loam.
CRHM was used to develop a custom hydrological model for the LS-05OG008
(Fig. 2). A detailed description of the modules selected, their function,
the sequence in which they were entered into the customised model, and how
they were parameterised, is presented by Cordeiro et al. (2017). Similar
model structures have been used successfully to simulate hydrological
processes in other areas of the Canadian Prairies, including Smith Creek
basin in eastern Saskatchewan (Fang and Pomeroy, 2008; Fang et al., 2010)
and the South Tobacco Creek basin of southern Manitoba (Mahmood et al.,
2017; Van Hoy et al., 2020). The same model structure was applied in the
four subbasins of LS-05OG008. As the land use split approach was used,
the HRU distribution was held constant over the simulation period, which
allowed for a single set of parameters to be used in the model for each HRU.
Flowchart of the model structure of CRHM used in this
study (adapted from Cordeiro et al., 2017). PBSM represents the Prairie Blowing Snow
Module, and EBSM represents the Energy-Budget Snowmelt Module.
The streamflow for the LS-05OG008 simulated by CRHM using this annual crop
simulation was assessed by Cordeiro et al. (2017) using the Nash–Sutcliffe
efficiency (NSE) index (Nash and Sutcliffe, 1970) and was deemed good, with
an average NSE value of 0.76 in years when peak daily discharge and annual
discharge volumes were equal to or above the medians of 6.7
and 1.25×107 m3 s-1, respectively. The simulated
streamflow in below-average years was generally poor (NSE < 0). This
caveat was taken into consideration when comparing the simulations between
the annual crop and perennial forage simulations for dry years. However,
comparisons for years with larger-than-average discharges are of most
interest to the present study, as these years govern discharge volumes to
Lake Winnipeg.
Perennial forage simulation
A key premise of the changes for the forage simulation was that perennial
forage promotes enhanced infiltration to soils when compared with annual crops
because of drier soil conditions, deeper rooting and greater macropore
development (van der Kamp et al., 2003), and greater moisture detention due
to random soil surface roughness and greater surface vegetation cover. For
the two scenarios, the model structures were kept the same while certain
parameters were modified to differentiate the forage and crop simulations. A
key change in the forage model was the inclusion of a new parameter
“fallstat_correction” in the CRHM. This new parameter adjusts
the value of the “fallstat” parameter after it was set by the Volumetric Soil Moisture module (Fig. 2). The fallstat parameter defines the degree of soil
saturation in the fall and influences frozen soil infiltration in the
subsequent spring. Therefore, the fallstat_correction was
implemented in the model to modify the fallstat parameter in order to
simulate the influence of soil macropore development on soil saturation.
This influence is expected to be more prominent under forage than under annual
cultivated crops, as the conversion of cultivated crops to grassland has
shown increased infiltration in frozen soils in the Canadian Prairies due to
well-developed macropore networks (van der Kamp et al., 2003).
An ensemble of forage scenarios was implemented in CRHM by setting the
fallstat_correction parameter between 0 % and 70 % (in
10 % increments) on Julian date 305 (1 November in a non-leap year) to
represent different limited soil infiltration conditions. This range was
defined to capture the uncertainty in the hydrological response to the
macropore formation under forage. Under limited conditions, soil
infiltrability is governed primarily by the soil moisture content (water and ice) and soil temperature at the start of snow ablation and the infiltration
opportunity time (Gray et al., 2001). However, Gray et al. (2001) noted that
cracks and macropores can also result in infiltrability rates larger than those calculated by porous media flow models such as the algorithm
used in CRHM (Zhao and Gray, 1999).
The cracks and macropores that form with zero tillage can also increase
infiltration rates into frozen soils (Mohammed et al., 2019). As a result,
the “groundcover” parameter was changed from “row crop and small grains”
in the annual crop model to “good pasture” (Ayers, 1959) in the forage
simulation in order to enhance the soil infiltrability for rainfall.
Other changes in the simulations pertained to land use, in which all annual
crop HRUs were converted to perennial forage (Table 1) but the HRU areas did
not change. The forage simulation assumed smooth bromegrass (Bromus inermis) as the single
forage used, which is a commonly cultivated grass in Manitoba (Looman, 1983;
Satchithanantham et al., 2017). The forage cover was assumed to be already
established; thus, the initial crop height was set to 0.4 m to mimic the
lodging of the stand in the previous fall. A maximum plant height of 1.1 m
was also used in the simulation, which is similar to the leafy stem length
of smooth brome reported in the literature (Looman, 1983). A growth rate of
1.4×10-3 m d-1 between Julian date 129 (9 May in a
non-leap year; crop start parameter) and 249 (6 September in a non-leap year;
crop mature parameter) was defined for the vegetation height to go from the
initial to the maximum vegetation height. Although there was no harvesting
simulated in the forage model, the harvest date parameter was set to Julian
date 288 (15 October in a non-leap year) to represent the lodging of the
stand and the reduction in vegetation height from 1.1 to 0.4 m. The start and
end of the growing season were set to Julian date 129 and 249, respectively,
to capture the continuous forage ET, while the maximum and minimum values of the
leaf area index (LAI) were set to 7 and 4 to represent the growing season
mature LAI and winter season minimal LAI for bromegrass, respectively. The
“cov_type” parameter, used to set the rooting depth for soil
moisture withdrawal by ET, was changed from the upper recharge layer for a shallow
crop rooting depth in the annual crop model to the entire soil layer for a
deeper bromegrass rooting depth in the forage simulation. Finally, the
vegetation density number was set to 41 m-2 (Grilz, 1995), and the
Manning roughness coefficient was set to 0.06 (Cordeiro et al., 2017).
These values affect blowing snow transport and runoff velocities in CRHM.
Hydrological and meteorological observations
Both the annual crop and perennial forage models were forced by hourly
weather data during the 1990–2013 period; however, the first 2 years of data
were used for model spin-up and were not included in the model assessment. Data
were obtained from Environment and Climate Change Canada weather stations
located at Portage Southport Airport (station ID: 3518), Winnipeg
International Airport (station ID: 51097), and Marquette (station ID: 3619).
These stations are 26.6, 47.9, and 9.9 km from the geometric centre of the
study area, respectively (Cordeiro et al., 2017). Temperature, wind speed,
and relative humidity were obtained from the Portage Southport Airport,
solar radiation was acquired from the station located at the Winnipeg
International Airport, and precipitation was acquired from the weather
station in Marquette. Precipitation was only available at a daily time step
and was disaggregated to an hourly time step using the HyetosMinute R package (Kossieris et al., 2013; Koutsoyiannis and Onof, 2001). More
details about the meteorological data are provided by Cordeiro et al. (2019).
Daily streamflow observations between 1992 and 2013 were obtained from the
hydrometric data (HYDAT) database (Environment and Climate Change Canada, 2013) for the Water
Survey of Canada gauging station 05OG008 (La Salle River near Elie; Fig. 1) located at the outlet of the study basin. Data collection at this station
was seasonal from 1992 to 1996 and has been continuous from 2002 to present.
A gap in available flow data exists between 1997 and 2001 (Cordeiro et al.,
2017). Remarks in the HYDAT metadata also indicated equipment malfunctions
resulting in loss of data in 2004 and 2008. For this reason, 15 years of data
between 1992 and 2013 excluding 1997–2001, 2004, and 2008 were used for
model assessment.
Comparison of the annual discharge volume between
annual crop and forage simulations for the La Salle subbasin (LS-05OG008).
Error bars indicate the 95 % confidence interval of the forage simulation
ensemble. The years 1997–2001, 2004, and 2008 were not used for model
assessment due to missing data or equipment malfunctions.
Simulation comparison
Hourly output data from both annual crop and perennial forage simulations
were averaged or summed to daily values for simulation comparisons. The
outputs of the forage simulations with varying soil saturation status (i.e.,
the fallstat_correction changed from 0 % to 70 % in 10 %
intervals) were summarised as the average of the eight simulations, and the
95 % confidence interval of the mean was used to represent the uncertainty
arising from the range of possible soil moisture status under limited soil
infiltration conditions. Annual discharge volume and peak daily discharge
rate were compared between the annual crop and forage ensemble simulations
to investigate the effect of changes in land use on the hydrology of the study
basin. To contextualise the differences between simulations and to gain
insight into the impact of land use conversion on key components of the
water balance, annual overland flow, peak SWE, annual infiltration, daily
soil moisture status, and annual ET were also compared. The comparison of
these water balance components was made for the crop (HRU) with the largest
area in the annual crop simulation (i.e., canola; HRU 16 in subbasin 1;
Table 1). Canola is also a provincially representative crop, being the
insured crop with the largest acreage in Manitoba (35.4 % of the insured
crop acreage), followed by soybean (24.7 % of the insured crop acreage)
and spring wheat (23.0 % of the insured crop acreage) (Dawson, 2018).
Comparison of the simulations was conducted for 1992–2013, excluding years
with missing observed streamflow or equipment malfunctions (1997–2001, 2004,
and 2008). Model performance was assessed according to discharge volumes and
rates (i.e., above or below average) as described in Cordeiro et al. (2017).
Sensitivity analysis
A sensitivity analysis was performed to examine how major hydrological
processes respond to changes in the stomatal resistance parameter, which
is used by the Penman–Monteith (PM) method (Monteith, 1965) in CRHM. For both the
annual crop and perennial forage models, the initial value of stomatal
resistance in the PM method was adjusted to 50 s m-1 (Beven, 2011),
which is within the range of 25 to 100 s m-1 reported for crops and
grasses (Allen et al., 1998; Beven, 2011; Verseghy et al., 1993). We
examined the sensitivity by adjusting the value of stomatal resistance to
25, 75, and 100 s m-1 for each of the forage scenarios and then
compared the simulated cumulative discharge, peak discharge, runoff,
infiltration, SWE, and ET with those under 50 s m-1.
Results
The overall annual discharge volume decreased, on average, by 36.5 ± 6.6 % (36.5 ± 7.2 mm) for the 15 years studied due to the conversion
from annual crops to perennial forage (Fig. 3). The annual discharge
volume from the annual crop simulation was higher than the upper confidence
interval of the forage simulation ensemble in all 15 years (Fig. 3),
indicating the unequivocal effect of perennial forage conversion on
reducing discharge volumes. Reductions in annual discharge ranged from
16.4 ± 6.2 % (28.7 ± 10.8 mm) in 2005 to 52.0 ± 7.7 %
(49.4 ± 7.3 mm) in 2007.
Similar to annual discharge, the peak daily discharge also decreased
consistently (i.e., 14 out of 15 years with conversion to forage; 93 % of
the time; Fig. 4). Reductions in peak daily discharge ranged from
4.0 ± 20.9 % (0.3 ± 1.3 m3 s-1) in 1993 to 59.3 ± 12.3 % (5.5 ± 1.1 m3 s-1) in 2007. The only year that peak
discharge increased with land use conversion was 1996, in which this
variable increased by 1.4 ± 26.3 % (0.2 ± 3.8 m3 s-1).
The uncertainty in peak discharge due to forage conversion was larger than
that for annual discharge volumes, as the peak discharge of the annual crop
model was within the 95 % confidence interval of the forage model ensemble
in 3 out of 15 years (20 % of the time; Fig. 4). Nonetheless, on
average, there was a 29.9 ± 16.3 % (2.6 ± 1.6 m3 s-1)
reduction in the peak daily discharge rate in the 15 years due to the forage
conversion.
Comparison of peak daily discharge between annual crop and
forage simulations for the La Salle River subbasin (LS-05OG008). Error bars
indicate the 95 % confidence interval of the forage simulation ensemble.
The years 1997–2001, 2004, and 2008 were not used for model assessment due
to missing data or equipment malfunctions.
Similar to reductions in the annual discharge volumes and peak discharge
rates, annual overland flow declined when the land use was converted from
canola to smooth bromegrass (Fig. 5). Annual overland flow from the annual
crop simulation was consistently higher than the upper 95 % confidence
interval for those from the forage model ensemble, indicating the
unequivocal effect of the forage conversion on decreasing overland flow. On
average, overland flow was reduced by 52.9 ± 12.8 % (28.8 ± 10.1 mm) in the forage simulation ensemble compared with the annual crop
simulation.
Comparison of annual overland flow between annual crop and
forage simulations for the La Salle River subbasin (LS-05OG008). Error bars
indicate the 95 % confidence interval of the forage simulation ensemble.
The years 1997–2001, 2004, and 2008 were not used for model assessment due
to missing data or equipment malfunctions.
In contrast to the variables presented above, snow accumulation increased
when converting the land use from canola to smooth bromegrass, with
an 8.1 ± 1.5 % (7.8 ± 1.6 mm) average increase in peak SWE (Fig. 6). This was due to the effect of tall standing grass in trapping snow and
preventing its wind erosion, transport, and sublimation during blowing snow
(Pomeroy and Gray, 1995). For dry years with peak daily discharge rates ≤2.7 m3 s-1, there were very minor reductions in peak SWE depth,
ranging from 0.1 % in 2012 to 0.5 % in 1994, as a result of conversion
from canola to forage, due to the role of exposed grass in increasing
turbulent transfer to snow and its overwinter sublimation in very dry years.
However, this effect was very small. It is worth noting that there is no
variation in peak SWE depth for the eight forage scenarios, indicating that,
as expected, snow accumulation is insensitive to the infiltration status of
soil.
Comparison of peak snow water equivalent (SWE) between
annual crop and forage simulations for the La Salle River subbasin
(LS-05OG008). Error bars indicate the 95 % confidence interval of the
forage simulation ensemble. The years 1997–2001, 2004, and 2008 were not
used for model assessment due to missing data or equipment malfunctions.
Infiltration depths increased substantially when converting canola to smooth
bromegrass in the forage model; on average, the annual infiltration depth
increased by 66.7 ± 7.7 % (141.5 ± 15.2 mm) due to the forage
conversion (Fig. 7). The enhanced infiltration in the forage simulation is
the combination of increased SWE and higher soil infiltrability for snowmelt
and rainfall under this land use. The annual infiltration depth in the annual
crop simulation was below the lower 95 % confidence interval of that
variable in the forage model ensemble in all years, indicating the
unmistakable effect of forage conversion on promoting infiltration.
Enhanced infiltration in the forage simulation led to similar or higher
spring soil moisture conditions when compared with the annual crop model
(Fig. 8; larger individual panels available in the Supplement). On average, soil moisture under forage was 18.0 ± 0.0 %
(57.2 ± 1.2 mm) higher than that of the annual crop simulation. This is
likely caused by the combined effect of higher SWE and enhanced infiltration
under forage. Figure 8 also displays consistent seasonal variation in soil
moisture. In late winter and early spring, soil moisture is constant due to
the frozen soil status during this period. As soil starts to thaw and snow
begins to melt due to higher temperature and increasing solar radiation in
late spring, soil moisture starts to rise due to increased infiltration.
With the increase in evapotranspiration in the summer because of higher
temperature and higher plant growth rates, soil moisture drops sharply under
both land use scenarios in all years except 1993. This could be explained by
the extremely high precipitation during the growing season (May–October) in
1993. From 1992 to 2013, about 70 % of annual precipitation occurred
during the growing season, whereas 85.1 % of annual precipitation
occurred during this period in 1993. This, combined with the cooler summer, led to
the lower ET and higher soil moisture availability in the summer of 1993. It
is also interesting to note that soil moisture under forage tended to
deplete faster than that under cropland during the summer. This result suggests higher
productivity of forage driven by higher evapotranspiration and consequent
faster depletion in soil moisture. Moreover, the antecedent soil moisture
conditions seemed to have played a critical role in the soil moisture profile in
the subsequent year. For example, higher soil moisture in the fall of 2005
led to high soil moisture during the spring and summer of 2006, whereas low
soil moisture in the fall of 2006 led to low soil moisture during the spring
and summer of 2007 (Fig. 8). This pattern was consistent over other
simulation periods (e.g., 1992–1996 and 2009–2013).
Comparison of annual infiltration depth between annual
crop and forage simulations for the La Salle River subbasin (LS-05OG008).
Error bars indicate the 95 % confidence interval of the forage simulation
ensemble. The years 1997–2001, 2004, and 2008 were not used for model
assessment due to missing data or equipment malfunctions.
The higher soil moisture depth resulted in increased annual actual ET depths
in the forage model when compared with the annual crop model across all years
(Fig. 9). Actual ET depths from the annual crop simulation were lower than
forage simulations in all years, indicating the sustained increase in water
demand of the forage simulation across variable weather conditions and the
longer growth and photosynthesis period for forage compared with the
annual crop. On average, ET increased by 34.5 ± 0.9 % (94.1 ± 2.5 mm) over the assessed period due to the conversion from annual crops to
perennial forage.
The sensitivity analysis indicated that cumulative discharge, peak
discharge, runoff, infiltration, SWE, and ET respond differently to the
change in stomatal resistance between 25 and 100 s m-1 (Fig. 10).
Among the key hydrological processes, cumulative discharge and cumulative ET
were most sensitive to the changes in stomatal resistance, whereas SWE was
insensitive to changes in this parameter. Annual cumulative discharge, peak
discharge, and cumulative runoff decreased by 26.9 %, 3.0 %, and 0.5 %, respectively, as the stomatal resistance value decreased from 50 to 25 s m-1. Cumulative
infiltration and cumulative ET increased by 4.0 % and 17.5 %, respectively, under the
same change in stomatal resistance. In comparison, an increase in cumulative
discharge, peak discharge, and cumulative runoff as well as a decrease in
cumulative infiltration and cumulative ET were observed when increasing
stomatal resistance from 50 to 75 and 100 s m-1. When increasing
stomatal resistance to 75 and 100 s m-1, cumulative discharge increased
by 38.0 % and 59.1 %, peak discharge increased by 7.1 % and 10.4 %,
cumulative runoff increased by 4.6 % and 9.9 %, cumulative infiltration
decreased by 8.1 % and 16.9 %, and cumulative ET decreased by 19.2 %
and 32.9 %, respectively.
Comparison of soil moisture storage between annual crop
model and forage simulations for the La Salle River subbasin (LS-05OG008).
The shaded area indicates the 95 % confidence interval of the forage model
ensemble. The years 1997–2001, 2004, and 2008 were not used for model
assessment due to missing data or equipment malfunctions.
Discussion
During the study period, surface runoff under annual cropland contributed
72.2 % of the stream discharge, which is consistent with previous studies
performed in this region (Dibike et al., 2012; Tiessen et al., 2010).
Under the perennial forage scenario, this contribution was decreased to
54.4 %. This reduction in surface runoff, combined with an increase in
evapotranspiration, resulted in reduced annual discharge from perennial
forage being simulated by CRHM at the basin scale, which agrees with hydrological
observations at the field scale in the Canadian Prairies (van der Kamp et al.,
2003). Reduced overland flow in perennial forage is primarily caused by
enhanced infiltration (Rachman et al., 2004; Self-Davis et al., 2003;
Tricker, 1981). By measuring infiltration to fine loamy soils during
snowmelt in Saskatchewan using single-ring infiltrometers, van der Kamp et al. (2003) found that the infiltrability of frozen soil was much higher
in grasslands than in cultivated fields. Their results at most of the
infiltration test locations showed that the frozen soil in the grassed areas
had an infiltration rate in excess of the typical snowmelt rates (i.e., ≤10 mm h-1), whereas all of the infiltration tests on frozen soil in
cultivated fields indicated an infiltrability that was considerably less than the
typical snowmelt rate. Enhanced infiltrability in perennial forage was
attributed to the development of macropores, such as root holes, desiccation
cracks, and animal burrows (van der Kamp et al., 2003). The results
demonstrated that the model simulations presented here were able to capture
the increased infiltration in frozen soils due to macropore formation under
forage.
The higher soil moisture content for perennial forage in some years (i.e.,
1994–1996, 2002–2006, and 2011) is contrary to the trends reported by field
investigations in the Canadian Prairies (Christie et al., 1985; van der Kamp
et al., 2003), where grasses had lower soil moisture than cultivated fields.
Such contrasts could be due to the more western and drier locations as well as the
short period of field investigations (1990 and 2000 for van der Kamp et al., 2003, and seemingly 1975 and 1981 for Christie et al., 1985), which may not
cover the full range of climate conditions, including very dry and wet years
experience in Manitoba. Thus, the impact of perennial forage on soil
moisture may not be unequivocal as suggested by previous short-term field
research, and this land cover may show variation between periods of low and
high soil moisture dictated by antecedent conditions. These differences in
soil moisture may also be a result of differences in ET calculation,
although the mean annual precipitation in the present study (560 mm) is
larger than those reported by Christie et al. (1985) for Lethbridge in Alberta
(350–400 mm) and by van der Kamp et al. (2003) for the St. Denis National
Wildlife Area in Saskatchewan (358 mm).
Recent field studies in the western Canadian Prairies have indicated that
differences in annual ET values between cropland and bromegrass grassland can be
attributed to differences in their phenological response to precipitation
and air temperature (Morgan et al., 2019). In the present study, differences
in ET between annual crop and perennial forage were mainly caused by
differences in the length of the growing season, plant height, and growth
rates in the CRHM models, which were parameterised using the PM method
(Monteith, 1965), with a Jarvis-style resistance formulation (Verseghy et
al., 1993). The PM method estimated stomatal and aerodynamic resistances
that represent the diffusion path lengths through the vegetation and boundary
layer, respectively, and both resistances controlled the water vapour
transfer to the atmosphere. It is noteworthy that the fixed value of stomatal
resistance does not account for seasonal variations in the biophysical
properties of vegetation (i.e., leaf area index and plant height) nor for the effects of
environmental stress factors (i.e., light limitation, vapour pressure
deficit, soil moisture tension or air entry pressure, and air temperature),
which leads to uncertainties in the PM method for this study. The initial
stomatal resistance value represents the minimum unstressed vegetation
resistance and is difficult to measure. Moreover, there is no consensus on an
accepted approach to estimate four environmental stress factors, and they
are determined from correlation and regression analysis (Verseghy et al., 1993).
Thus, these uncertainties in the PM method could affect the ET flux
estimations and should be considered when interpreting the results. These
uncertainties were evidenced through the sensitivity analysis carried out in
the present study. Further investigations on canopy resistance formulation
and field campaigns to measure canopy resistance are needed to improve the ET
estimations for a number of vegetation types in the Canadian Prairies.
Comparison of annual cumulative evapotranspiration (ET)
between annual crop and forage simulations for the La Salle River subbasin
(LS-05OG008). The years 1997–2001, 2004, and 2008 were not used for model
assessment due to missing data or equipment malfunctions.
The changes in the water balance described in this study are conducive to
reductions in nutrient export from agricultural lands. Previous studies
have indicated that reductions in sediment and nutrient transport are closely
associated with the reduction in surface runoff (Aksoy and Kavvas, 2005;
Chen et al., 2016; Corriveau et al., 2013; Liang et al., 2020; Sharpley and
Williams, 1990). Previous modelling exercises in the region also
corroborate this conclusion. For example, simulations of land use
conversion from annual crop to perennial forage using the Soil and Water Assessment
Tool (SWAT) model
conducted in the entire La Salle River subbasin (where the study area in the
present study is located) reported reductions of 37 %, 58 %, and 72 %
in sediment, total nitrogen (TN), and total phosphorus (TP) loads,
respectively (Yang et al., 2014). The lower reduction in sediment compared
with TN and TP was due to the majority of cropland being located on very flat terrain
with clay soils, making soil erosion and sediment transport processes less
significant in that basin (Yang et al., 2014). However, parameterisation in
the nutrient dynamics module of SWAT, not discussed in detail in the study,
could also have influenced these results. A stepwise calibration of stream
discharge and sediment, followed by calibration of TN and TP, was achieved
using the sequential uncertainty fitting (SUFI-2) calibration algorithm in
SWAT-CUP software. This calibration procedure implies a dependency of TN and
TP on sediment transport, which is not usually the case in the Canadian
Prairies, where most of nutrient transport from basins occurs in dissolved
form (Cade-Menun et al., 2013; Liu et al., 2013; Tiessen et al., 2011).
Another potential concern in that study was the SWAT version used in the
simulations (i.e., SWAT 2012), which does not include modules for simulating
nutrient release from vegetation. As discussed above, not accounting for the
contribution of perennial forage to the runoff nutrient concentration could
underestimate the nutrient export from these landscapes. In fact, nutrient
leaching from plant residues has not been represented in water quality
models, which has led to the development of process-based algorithms in the
Canadian Prairies to address this gap (Costa et al., 2019).
Sensitivity of major hydrological processes to changes in
stomatal resistance (s m-1).
Despite the hypothetical positive water quality impacts due to land use
conversion from annual crops to perennial forage, this conversion is
challenged by current trends in agricultural lands. According to the 2021
Plowprint Report, over 1×106 ha of grasslands were converted between 2018
and 2019 alone, mostly to crop agriculture (World Wildlife, 2021).
Conversion to cropland is mostly driven by recent increases in grain prices
due to increased demands created by the rapid economic development in Asian
countries (Montossi et al., 2020). Grassland conversion in the US Upper
Midwest in the past decade has resulted in a substantial degradation of soil
quality, with implications for air and water quality (Zhang et al.,
2021). Such environmental impacts are likely related to hydrological
alterations, as indicated by the analysis presented in this study. However,
these hypotheses should be validated though field and modelling research
efforts in the future. With respect to the former, field monitoring
investigating the interplay between hydrology and nutrient release is
required, as stated previously. Regarding the latter, future model
development to better represent the hydrological behaviour of perennial
forage is needed. The methodology adopted in the present study (i.e.,
falsification of the fallstat parameter) was meant as a “proof-of-concept”
approach, but a more rigorous model development based on field research is
warranted.
Conclusions
Hydrologic changes due to land use conversion in the Canadian Prairies were
assessed at the basin scale within a modelling framework using the Cold
Regions Hydrological Modelling platform (CRHM), which has physically based
modules specifically developed to simulate cold-region hydrological
processes. An annual crop model and a perennial forage model were set up in
CRHM to simulate current agricultural conditions in a subcatchment of the
La Salle River basin, which is a subbasin of the Red River Valley. The model
simulations indicated that many of the hydrological changes from land use
conversion observed at the field scale would also take place at larger scales.
On average, there was a 36.5 ± 6.6 % (36.5 ± 7.2 mm) reduction
in the annual discharge volume and a 29.9 ± 16.3 % (2.6 ± 1.6 m3 s-1) reduction in the peak discharge rate due to conversion to forage over the period assessed. Reductions in the cumulative and peak discharge
under forage were driven by reduced overland flow (52.9 ± 12.8 %)
(28.8 ± 10.1 mm), increased infiltration to both frozen and unfrozen
soils (66.7 ± 7.7 %) (141.5 ± 15.2 mm), and higher cumulative ET
(34.5 ± 0.9 %) (94.1 ± 2.5 mm), despite increased peak SWE
(8.1 %) (7.8 mm). The impact of higher rates of snowmelt infiltration
more than compensated for higher SWE and resulted in reduced overland flow,
which mostly occurs during the spring snowmelt season in this basin. The
higher SWE due to the suppression of blowing snow erosion under the taller
bromegrass and enhanced infiltration led to the higher soil moisture due to
greater macropore formation under untilled bromegrass for the majority of the
simulation period. The average daily soil moisture under forage was 18.0 %
(57.2 ± 1.2 mm) higher than that under annual cropland. While the simulations
reported in this study do agree with results from field studies, they also
warrant further evaluation at the field scale to reconcile the contrasting
aspects of the water balance that persist among some field studies and model
simulations. Long-term monitoring of macropore network development (e.g.,
via infiltration measurement), spring infiltration, soil moisture
dynamics, evapotranspiration, and runoff volume at the edge of field since
forage establishment would cast some light on the temporal effect of
perennial forage on these variables. Moreover, parallel monitoring of
nutrient concentrations and weather patterns would also help devise the
release of nutrients from forage due to the breakdown of plant material, which
(combined with runoff volumes) determines the exported loads from perennial
forage. This monitoring would aid not only an assessment of the temporal
consistency of the forage impact on the water balance variables but also on
nutrient export.
Code and data availability
CRHM codes are available at https://research-groups.usask.ca/hydrology/modelling/crhm.php#TechnicalDetails, (last access: 15 November 2022).
The supplement related to this article is available online at: https://doi.org/10.5194/hess-26-5917-2022-supplement.
Author contributions
MRCC and KL performed the model simulations and data analysis, created the figures, and
prepared the original manuscript. MRCC, KL, HFW, and JWP conceived the
modelling objectives, scope, and strategy. MRCC, KL, JWP, and XF developed
the custom model for analysis. JV and DAL contributed to the methodology, review,
and editing.
Competing interests
The contact author has declared that none of the authors has any competing interests.
Disclaimer
Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Acknowledgements
Collaboration in the preparation of model input data with Zhiqiang Yu and our discussions about characteristics of the basin are
greatly appreciated.
Financial support
This research was supported by funding within the framework of the Agriculture and Agri-Food Canada’s Growing Forward 2 and Living Laboratories programs and by the Beef Cattle Research Council (grant no. ENV.07.19).
Review statement
This paper was edited by Christian Stamm and reviewed by two anonymous referees.
ReferencesAgriculture and Agri-Food Canada (AAFC): The Canadian System of Soil
Classification, NRC Research Press, Ottawa, 188 pp., 10.1139/9780660174044,
1998.Agriculture and Agri-Food Canada (AAFC): Annual Crop Inventory 2013, https://open.canada.ca/data/en/dataset/ba2645d5-4458-414d-b196-6303ac06c1c9/resource/7971855d-c2e0-422b-8da5-5bb87978522f (last access: 9 February 2021), 2013.Aksoy, H. and Kavvas, M. L.: A review of hillslope and watershed scale erosion
and sediment transport models, Catena, 64, 247–271,
10.1016/j.catena.2005.08.008, 2005.
Allen, R. G., Pereira, L. S., Raes, D., and Smith, M.: Crop
evapotranspiration-Guidelines for computing crop water requirements-FAO
Irrigation and drainage paper 56. Fao, Rome, 300, D05109, 1998.
Ayers, H. D.: Influence of soil profile and vegetation characteristic on
netrainfall supply to runoff, in: Spillway Design Floods: Proceeding of
HydrologySymposium No. 1, National Research Council of Canada, 198–205,
1959.Benoy, G. A., Jenkinson, R. W., Robertson, D. M., and Saad, D. A.: Nutrient
delivery to Lake Winnipeg from the Red – Assiniboine River Basin – A
binational application of the SPARROW model. Can. Water Resour. J., 41,
429–447, 10.1080/07011784.2016.1178601, 2016.
Beven, K. J.: Rainfall-Runoff Modelling: The Primer, Wiley, West Sussex, UK, ISBN 978-0-470-71459-1,
2011.Bower, S. S.: Watersheds: Conceptualizing Manitoba's Drained Landscape,
1895–1950, Environ. Hist., 12, 796–819, 10.1093/envhis/12.4.796, 2007.Buttle, J. M., Allen, D. M., Caissie, D., Davison, B., Hayashi, M., Peters,
D. L., Pomeroy, J. W., Simonovic, S., St-Hilaire, A., and Whitfield, P. H.:
Flood processes in Canada: Regional and special aspects, Can. Water Resour.
J., 41, 7–30, 10.1080/07011784.2015.1131629, 2016.Cade-Menun, B. J., Bell, G., Baker-Ismail, S., Fouli, Y., Hodder, K.,
McMartin, D. W., Perez-Valdivia, C., and Wu, K.: Nutrient loss from
Saskatchewan cropland and pasture in spring snowmelt runoff, Can. J. Soil
Sci., 93, 445–458, 10.4141/cjss2012-042, 2013.Chen, L., Wang, G., Zhong, Y., and Shen, Z.: Evaluating the impacts of soil data
on hydrological and nonpoint source pollution prediction, Sci. Total
Environ., 563, 19–28,
10.1016/j.scitotenv.2016.04.107, 2016.Christie, H. W., Graveland, D. N., and Palmer, C. J.: Soil and subsoil moisture
accumulation due to dryland agriculture in southern Alberta, Can. J. Soil
Sci., 65, 805–810, 10.4141/cjss85-084, 1985.Cordeiro, M. R. C., Lelyk, G., Kröbel, R., Legesse, G., Faramarzi, M., Masud, M. B., and McAllister, T.: Deriving a dataset for agriculturally relevant soils from the Soil Landscapes of Canada (SLC) database for use in Soil and Water Assessment Tool (SWAT) simulations, Earth Syst. Sci. Data, 10, 1673–1686, 10.5194/essd-10-1673-2018, 2018.Cordeiro, M. R. C., Wilson, H. F., Vanrobaeys, J., Pomeroy, J. W., Fang, X., and The Red-Assiniboine Project Biophysical Modelling Team: Simulating cold-region hydrology in an intensively drained agricultural watershed in Manitoba, Canada, using the Cold Regions Hydrological Model, Hydrol. Earth Syst. Sci., 21, 3483–3506, 10.5194/hess-21-3483-2017, 2017.
Cordeiro, M. R. C., Vanrobaeys, J. A., and Wilson, H. F.: Long-term weather,
streamflow, and water chemistry datasets for hydrological modelling
applications at the upper La Salle River watershed in Manitoba, Canada,
Geosci. Data J., 6, 41–57, 2019.
Corriveau, J., Chambers, P. A., and Culp, J. M.: Seasonal variation in
nutrient export along streams in the northern Great Plains, Water, Air, Soil
Pollut., 224, 1–16, 2013.
Costa, D., Liu, J., Roste, J., and Elliott, J.: Temporal dynamics of
snowmelt nutrient release from snow–plant residue mixtures: An experimental
analysis and mathematical model development, J. Environ. Qual., 48, 869–879,
2019.Dawson, A.: Despite dry weather Manitoba's 2017 harvest set many new yield
records, Yield Manitoba. Manitoba Agricultural Services Corporation,
Winnipeg, MB, 12–16, https://www.masc.mb.ca/masc.nsf/mmpp_index, (last access: 26 March 2020), 2018.Dibike, Y., Prowse, T., Shrestha, R., and Ahmed, R.: Observed trends and
future projections of precipitation and air temperature in the Lake Winnipeg
watershed, J. Great Lakes Res., 38, 72–82, 10.1016/j.jglr.2011.04.005, 2012.
Ehsanzadeh, E., van der Kamp, G., and Spence, C.: The impact of climatic
variability and change in the hydroclimatology of Lake Winnipeg watershed,
Hydrol. Proc., 26, 2802–2813, 2012.Environment and Climate Change Canada (ECCC) 2018: National Water Data
Archive: HYDAT,
https://collaboration.cmc.ec.gc.ca/cmc/hydrometrics/www/ (last access: 17 August 2021), 2018.Environment and Climate Change Canada (ECCC): HYDAT Database – Water –
Environment and Climate Change Canada, https://www.canada.ca/en/environment-climate-change/services/water-overview/quantity/monitoring/survey/data-products-services/national-archive-hydat.html, (last access: 26 March 2020), 2013.Fang, X., Pomeroy, J. W., Ellis, C. R., MacDonald, M. K., DeBeer, C. M., and Brown, T.: Multi-variable evaluation of hydrological model predictions for a headwater basin in the Canadian Rocky Mountains, Hydrol. Earth Syst. Sci., 17, 1635–1659, 10.5194/hess-17-1635-2013, 2013.Fang, X., Pomeroy, J. W., Westbrook, C. J., Guo, X., Minke, A. G., and Brown, T.: Prediction of snowmelt derived streamflow in a wetland dominated prairie basin, Hydrol. Earth Syst. Sci., 14, 991–1006, 10.5194/hess-14-991-2010, 2010.Fang, X. and Pomeroy, J. W.: Diagnosis of future changes in hydrology for a Canadian Rockies headwater basin, Hydrol. Earth Syst. Sci., 24, 2731–2754, 10.5194/hess-24-2731-2020, 2020.Fang, X. and Pomeroy, J. W.: Drought impacts on Canadian prairie wetland snow
hydrology, Hydrol. Proc., 22, 2858–2873, 10.1002/hyp.7074, 2008.Graveline, P. G. and Larter, J.: La Salle Redboine Conservation District: La
Salle River watershed assessmiment survey – with emphasis on La Salle
River, Elm River, Elm Creek channel, and The King drain – 2005, Winnipeg,
MB, https://redboine.com/wp-content/uploads/2020/07/lasalleriverwatershedassessmentsurvey2006.pdf (last access: 26 March 2020), 2006.Gray, D. M., Toth, B., Zhao, L., Pomeroy, J. W., and Granger, R. J.: Estimating
areal snowmelt infiltration into frozen soils. Hydrol. Proc., 15,
3095–3111, 10.1002/hyp.320, 2001.
Grilz, P. L.: Management considerations for controlling smooth brome in
fescue prairie, Nat. Areas J., 15, 148–156, 1995.Harder, P. and Pomeroy, J. W.: Hydrological model uncertainty due to
precipitation-phase partitioning methods. Hydrol. Proc., 28,
4311–4327, 10.1002/hyp.10214, 2014.He, Z., Pomeroy, J. W., Fang, X., and Peterson, A.: Sensitivity analysis of
hydrological processes to perturbed climate in a southern boreal forest
basin, J. Hydrol., 601, 126706, 10.1016/j.jhydrol.2021.126706, 2021.
Hutchinson, J. J., Campbell, C. A., and Desjardins, R. L. Some perspectives
on carbon sequestration in agriculture, Agr. For. Meteorol., 142,
288–302, 2007.
Kharel, G., Zheng, H., and Kirilenko, A.: Can land-use change mitigate
long-term flood risks in the Prairie Pothole Region?, The case of Devils
Lake, North Dakota, USA. Reg. Environ. Change, 16, 2443–2456, 2016.Kossieris, P., Tyralis, H., Koutsoyiannis, D., and Efstratiadis, A.: HyetosR: A
package for temporal stochastic simulation of rainfall at fine time scales,
R package version 0.0-2, http://www.itia.ntua.gr/ (last access: 17 August 2021), 2013.Koutsoyiannis, D. and Onof, C.: Rainfall disaggregation using adjusting
procedures on a Poisson cluster model, J. Hydrol., 246, 109–122,
10.1016/S0022-1694(01)00363-8, 2001.Krogh, S. A. and Pomeroy, J. W.: Impact of Future Climate and Vegetation on the
Hydrology of an Arctic Headwater Basin at the Tundra–Taiga Transition, J.
Hydrometeorol., 20, 197–215, 10.1175/jhm-d-18-0187.1, 2019.La Salle Redboine Conservation District: La Salle River Watershed – State of
the Watershed Report, Holland, Manitoba, https://www.gov.mb.ca/sd/water/watershed/iwmp/la_salle_river/documentation/summary_lasalle.pdf (last access: 26 March 2020), 2007.Liang, K., Jiang, Y., Qi, J., Fuller, K., Nyiraneza, J., and Meng, F. R.:
Characterizing the impacts of land use on nitrate load and water yield in an
agricultural watershed in Atlantic Canada, Sci. Total Environ., 729,
p. 138793, 10.1016/j.scitotenv.2020.138793, 2020.Li, S., Elliott, J. A., Tiessen, K. H., Yarotski, J., Lobb, D. A., and Flaten,
D. N.: The effects of multiple beneficial management practices on hydrology
and nutrient losses in a small watershed in the Canadian Prairies, J.
Environ. Qual., 40, 1627–1642, 10.2134/jeq2011.0054, 2011.Liu, J., Macrae, M. L., Elliott, J. A., Baulch, H. M., Wilson, H. F., and
Kleinman, P. J. A.: Impacts of Cover Crops and Crop Residues on Phosphorus
Losses in Cold Climates: A Review, J. Environ. Qual., 48, 850–868,
10.2134/jeq2019.03.0119, 2019.
Liu, K., Elliott, J. A., Lobb, D. A., Flaten, D. N., and Yarotski, J.: Critical
factors affecting field-scale losses of nitrogen and phosphorus in spring
snowmelt runoff in the Canadian Prairies, J. Environ. Qual., 42, 484–496,
2013.Liu, K., Elliott, J. A., Lobb, D. A., Flaten, D. N., and Yarotski, J.: Nutrient and
Sediment Losses in Snowmelt Runoff from Perennial Forage and Annual Cropland
in the Canadian Prairies, J. Environ. Qual., 43, 1644–1655,
10.2134/jeq2014.01.0040, 2014.
Looman, J.: 111 range and forage plants of the Canadian Prairies,
A53-1751/1983E-PDF, Canadian Government Publishing Centre, Ottawa, 255 pp., 1983.Mahmood, T. H., Pomeroy, J. W., Wheater, H. S., and Baulch, H. M.: Hydrological
responses to climatic variability in a cold agricultural region, Hydrol.
Proc., 31, 854–870, 10.1002/hyp.11064, 2017.McCullough, G. K., Page, S. J., Hesslein, R. H., Stainton, M. P., Kling, H.
J., Salki, A. G., and Barber, D. G.: Hydrological forcing of a recent
trophic surge in Lake Winnipeg, J. Great Lakes Res., 38, 95–105,
10.1016/j.jglr.2011.12.012, 2012.
Mikkelsen, R. L.: The “4R” nutrient stewardship framework for
horticulture, HortTechnology, 21, 658–662, 2011.Mohammed, A. A., Pavlovskii, I., Cey, E. E., and Hayashi, M.: Effects of preferential flow on snowmelt partitioning and groundwater recharge in frozen soils, Hydrol. Earth Syst. Sci., 23, 5017–5031, 10.5194/hess-23-5017-2019, 2019.
Monteith, J. L.: Evaporation and environment, Symposia of the society for
experimental biology, Cambridge University Press (CUP) Cambridge,
205–234, 1965.Montossi, F., Ayala, W., and Díaz, R.: The challenges of cropping and
forestry intensification on grasslands livestock production systems: the
Uruguayan case, https://uknowledge.uky.edu/igc/21/9-1/2/ (last access: 2 July 2022), 2020.Morgan, L. R., Hayashi, M., and Cey, E. E.: Land-use comparison of
depression-focussed groundwater recharge in the Canadian
prairies, Hydrol. Proc., 35, e14379, 10.1002/hyp.14379, 2021.Nash, J. E. and Sutcliffe, J. V.: River flow forecasting through conceptual
models part I – A discussion of principles, J. Hydrol., 10, 282–290,
10.1016/0022-1694(70)90255-6, 1970.Painter, K. J., Brua, R. B., Chambers, P. A., Culp, J. M., Chesworth, C. T.,
Cormier, S. N., Tyrrell, C. D., and Yates, A. G.: An ecological causal
assessment of tributaries draining the Red River Valley, Manitoba, J. Great
Lakes Res., 47, 773–787, 10.1016/j.jglr.2020.05.004, 2021.
Pomeroy, J. W., Brown, T., Fang, X., Shook, K. R., Pradhananga, D., Armstrong, R., Harder, P., Marsh, C., Costa, D., Krogh, S. A., Aubry-Wake, C., Annand, H., Lawford, P., He, Z., Kompanizare, M., and Lopez Moreno, J. I.: The Cold Regions Hydrological Modelling Platform for hydrological diagnosis and prediction based on process understanding, J. Hydrol., 128711, https://doi.org/10.1016/j.jhydrol.2022.128711, 2022.
Pomeroy, J. W. and Gray, D. M.: The cold regions hydrological model: a platform for basing process representation and model structure on physical evidence, 21, 2650–2667, https://doi.org/10.1002/hyp.6787, 2007.Pomeroy, J. W., Gray, D. M., Brown, T., Hedstrom, N. R., Quinton, W. L.,
Granger, R. J., and Carey, S. K.: The cold regions hydrological model: a
platform for basing process representation and model structure on physical
evidence, Hydrol. Proc., 21, 2650–2667, 10.1002/hyp.6787, 2007.Rachman, A., Anderson, S. H., Gantzer, C. J., and Thompson, A. L.: Influence of
Stiff-Stemmed Grass Hedge Systems on Infiltration, Soil Sci. Soc. Am. J.,
68, 2000–2006, 10.2136/sssaj2004.2000, 2004.Rattan, K. J., Corriveau, J. C., Brua, R. B., Culp, J. M., Yates, A. G., and
Chambers, P. A.: Quantifying seasonal variation in total phosphorus and
nitrogen from prairie streams in the Red River Basin, Manitoba Canada, Sci.
Total Environ., 575, 649–659, 10.1016/j.scitotenv.2016.09.073, 2017.Satchithanantham, S., Wilson, H. F., and Glenn, A. J.: Contrasting patterns of
groundwater evapotranspiration in grass and tree dominated riparian zones of
a temperate agricultural catchment, J. Hydrol., 549, 654–666,
10.1016/j.jhydrol.2017.04.016, 2017.Schindler, D. W., Hecky, R. E., and McCullough, G. K.: The rapid
eutrophication of Lake Winnipeg: Greening under global change, J. Great
Lakes Res., 38, 6–13, 10.1016/j.jglr.2012.04.003, 2012.
Self-Davis, M. L., Moore, P. A., Daniel, T. C., Nichols, D. J., Sauer, T. J.,
West, C. P., Aiken, G. E., and Edwards, D. R.: Forage species and canopy cover
effects on runoff from small plots, J. Soill Water Conserv., 58, 349–359,
2003.
Sharpley, A. N. and Williams, J. R.: EPIC-erosion/productivity impact calculator:
1. Model documentation., Technical Bulletin-United States Department of
Agriculture (1768 Pt. 1), 1990.
Tiessen, K., Elliott, J. A., Yarotski, J., Lobb, D., Flaten, D., and Glozier, N.: Conventional and conservation tillage: Influence on seasonal runoff, sediment, and nutrient losses in the Canadian prairies, J. Environ. Qual., 39, 964–980, https://doi.org/10.2134/jeq2009.0219, 2010.Tiessen, K. H. D., Elliott, J. A., Stainton, M., Yarotski, J., Flaten, D. N., and
Lobb, D. A.: The effectiveness of small-scale headwater storage dams and
reservoirs on stream water quality and quantity in the Canadian Prairies, J.
Soill Water Conserv., 66, 158–171, 10.2489/jswc.66.3.158, 2011.Tricker, A. S.: Spatial and temporal patterns of infiltration, J. Hydrol.,
49, 261–277, 10.1016/0022-1694(81)90217-1, 1981.Uusi-Kamppa, J., Turtola, E., Narvanen, A., Jauhiainen, L., and Uusitalo,
R.: Phosphorus mitigation during springtime runoff by amendments applied to
grassed soil, J. Environ. Qual., 41, 420–426, 10.2134/jeq2010.0441, 2012.van der Kamp, G., Hayashi, M., and Gallén, D.: Comparing the hydrology of
grassed and cultivated catchments in the semi-arid Canadian prairies,
Hydrol. Proc., 17, 559–575, 10.1002/hyp.1157, 2003.Van Hoy, D. F., Mahmood, T. H., Todhunter, P. E., and Jeannotte, T. L.: Mechanisms
of Cold Region Hydrologic Change to Recent Wetting in a Northern Glaciated
Landscape, Water Resour. Res., 56, e2019WR026932,
10.1029/2019WR026932, 2020.Verseghy, D., McFarlane, N., and Lazare, M.: CLASS – A Canadian land surface
scheme for GCMs, II. Vegetation model and coupled runs, IJCli, 13,
347–370, 1993.
World Wild Life (WWF): The Plowprint Report,
https://www.worldwildlife.org/publications/2021-plowprint-report (last access: 30 August 2022), 2021.Yang, Q., Leon, L. F., Booty, W. G., Wong, I. W., McCrimmon, C., Fong, P.,
Michiels, P., Vanrobaeys, J., and Benoy, G.: Land use change impacts on water
quality in three Lake Winnipeg watersheds, J. Environ. Qual., 43,
1690–1701, 10.2134/jeq2013.06.0234, 2014.Yates, A. G., Culp, J. M., and Chambers, P. A.: Estimating nutrient
production from human activities in subcatchments of the Red River,
Manitoba, J. Great Lakes Res., 38, 106–114, 10.1016/j.jglr.2011.04.009,
2012.Zhang, X., Lark, T. J., Clark, C. M., Yuan, Y., and LeDuc, S. D.:
Grassland-to-cropland conversion increased soil, nutrient, and carbon losses
in the US Midwest between 2008 and 2016, Environ. Res. Lett., 16, 054018,
10.1088/1748-9326/abecbe, 2021.
Zhao, L. T. and Gray, D. W.: Estimating snowmelt infiltration into frozen soils,
Hydrol. Proc., 13, 1827–1842, 1999.