Skip booklet index and go to page content

Biodiversity in Canadian Lakes and Rivers

Trends in Hydrological Regimes

Return to Table of Contents

Page Contents

The Water Survey of Canada coordinates the national database, HYDAT (Environment Canada, 2006b), which, as of 2006, contained national hydrometric information for >2500 active water level and discharge monitoring stations and an additional >5500 discontinued stations located on lakes, rivers, and streams across Canada. The majority of stations are located in the southern half of the country near centres of population and development, with fewer stations with fewer consecutive years of data in the north (Figure 8).

Figure 8. Distribution of existing and historically gauged water monitoring stations across Canada: (a) natural lakes; (b) regulated lakes; (c) natural rivers; and (d) regulated rivers.
Figure 8 image is comprised of four maps of Canada showing the distribution of existing and historically gauged water monitoring stations across Canada
Source: data from Environment Canada (2006b)
Long Description for Figure 8.

This figure is comprised of four maps of Canada with dots representing the location of existing and historically gauged water monitoring stations located on lakes, rivers, and streams across Canada. There is a separate map for natural lakes, regulated lakes, natural rivers, and regulated rivers. The majority of stations are located in the southern half of the country near centers of population and development, with fewer stations located in the north.

Top of Page

Hydrometric stations are designated as either “natural” or “regulated” by the Water Survey of Canada. Natural sites reflect gauging stations with minimal regulation or impact upstream of the site (Environment Canada, 2010b). Regulated sites vary regionally in their hydrological alteration in terms of water abstraction, impoundment, or diversion of flow. While daily observations have been recorded for some sites since the early 1800s, the majority of data in HYDAT were collected from the 1970s to the early 1990s (Figure 9). Network rationalisation during the 1990s resulted in significant station loss, reducing the spatial coverage of the network. Early monitoring appears to have focused on regulated river and lake systems. Although all site types show a recent increase in monitoring, the strongest increase appears to be for natural river systems. The majority of sites within the HYDAT database have less than 18 consecutive years of data (Figure 10). With the need for long-term continuous records for monitoring current trends and projecting future trends in hydrological regimes, it is imperative that sites with longer records are maintained and that additional sites are brought online where trends monitoring is required, for example for the ongoing assessment of climate change.

Figure 9. Number of sites with hydrological records for regulated and natural lakes and rivers, 1800–2006.
Figure 9 is a line graph showing the number of sites with hydrological records for regulated and natural lakes and rivers for the years 1800 to 2006
Source: data from Environment Canada (2006b)
Long Description for Figure 9.

This line graph shows the number of sites with hydrological records for regulated and natural lakes and rivers for the years 1800 to 2006. While observations have been recorded for some sites since the early 1800s, the majority of data were collected from the 1970s to the early 1990s. During the 1990s there was significant station loss, reducing the spatial coverage of the network. Early monitoring appears to have focused on regulated river and lake systems. Although all site types show a recent increase in monitoring, the strongest increase appears to be for natural river systems.

Top of Page

Figure 10. Frequency histogram for the total number of years of data at each hydrometric site.
Figure 10 is a bar graph showing the number of years of hydrological record
Source: data from Environment Canada (2006b)
Long Description for Figure 10.

This bar graph shows the following information:

Number of years of hydrological recordFrequency
31222
6942
9671
12500
15384
18384
21349
24282
27298
30257
33242
36248
39287
42236
45279
48168
51197
54144
57116
6082
6351
6674
6962
7248
7535
7834
8162
8446
8751
9039
9345
96110
9985
10228
10514
1087
1112
1141
1171
1201
1230
1260
1293
1320
1351
1380
1410
1440
1470
1502
1531
1561
1590
1620
1650
1680
1710
1740
1770
1800
1830
1860
1890
1920
1951
1982

Top of Page

Exploring hydro-ecological trends

Aquatic communities are generally adapted to natural inter-annual variability in the flow regime and the magnitude, timing, and predictability of high and low flow periods (Jowett and Duncan, 1990). Previous studies have quantitatively explored the ecological significance of components of the flow regime for hydro-ecological research and for environmental flow recommendations and related methods development (for example, Monk et al., 2008). Additionally, research has explored trends in hydrological regimes across North America (see examples in Table 2).  Schindler and Donahue (2006) reported, for example, that mean summer (May to August) flows of the Athabasca River decreased by 20% since 1958. In addition, changes in the drivers of the hydrological regime have been reported--for example, the total precipitation falling as snow declined in western Canada and the Prairies from 1900 to 2003 as a result of climate warming trends (Vincent and Mekis, 2006).

Table 2. Summary of published scientific papers exploring statistical trends in streamflow and runoff in Canadian rivers.
Study catchment(s)Area of studyEcozone+Number of sites
Analysis method
Analysis period
with # sites for period)
Results / direction of trendsReference
Liard, Peace, AthabascaYK, AB, BC, SK, NTBoreal Cordillera,
Taiga Plains,
Boreal Plains,
Montane Cordillera
26 hydrometric
Mann-Kendall
1970–2005 (26)
1965–2005 (21)
1960–2005 (18)
  • Spring freshet occurring earlier with the timing shift stronger in headwater catchments
  • Part attributed to effects by Pacific Decadal Oscillation, North Atlantic Oscillation, North Pacific Index and Atlantic Multi-decadal Oscillation
Burn (2008)
LiardYT, NT, BC and ABTaiga Plains,
Taiga Cordillera, Boreal Plains,
Boreal Cordillera
12 hydrometric
Mann-Kendall
1975–1999 (13)
1970–1999 (12)
1965–1999 (8)
1960–1999 (7)
  • Increasing winter flow and increasing minimum flow
  • Slight decreasing trend in the summer months' flow
  • Spring freshet occurring earlier and earlier spring peak flow related to increases in air temperatures
  • Increased winter flows related to Pacific Decadal Oscillation
  • Clear spatial differences along catchment, e.g., upper catchment experiencing decreasing trends overall but lower catchment presents increasing trends
Burn (2004)
Athabasca, Peace, Great Slave, Liard, Mackenzie, Northern Peel, CoppermineAB, SK, BC, NT, YTTaiga Shield,
Boreal Shield,
Boreal Plains, Montane Cordillera, Taiga Plains,
Taiga Cordillera, Boreal Cordillera
54 hydrometric
10 meteorological
Mann-Kendall
1975–2000 (54)
1970–2000 (46)
1965–2000 (34)
1960–2000 (21)
  • Strong increasing trends over the winter months (December to April)
  • Increasing annual minimum flow
  • Weak decreasing trends in annual mean flow and the early summer and late fall flows
  • Earlier onset of spring freshet
Aziz and Burn (2006)
248 sites across CanadaAll areasAll ecozones+ except Arctic Cordillera portion of Arctic Ecozone+248 RHBN hydrometric
1940–1997
1950–1997
1960–1997
1970–1997
All available records at stations
  • Spatial patterns in significant trends suggest impacts are not spatially uniform
  • Decreasing trend in annual maximum flow in the south and an increasing trend in the north
  • Date of ice breakup is earlier, probably as a result of earlier onset of spring melt conditions
  • Strong increasing trend in March and April flow indicative of an earlier onset of spring snowmelt
  • June flow displays a strong decreasing trend
  • October flows display increasing trends in the east and north and decreasing trends in the west
Burn and Hag Elnur (2002)
25 streams across the PrairiesAB, SK, MBPrairies,
Boreal Plains,
Boreal Shield,
Taiga Shield,
Taiga Plains
25 hydrometric
16 meteorological
1976–2005 (26)
1971–2005 (24)
1966–2005 (17)
  • Decreasing trends in the spring snowmelt runoff volume and peak flow
  • Earlier spring snowmelt peak
  • Decreasing trends in seasonal (March–October) runoff volume
  • Attributed to reductions in snowfall and increases in winter air temperatures
Burn et al. (2008)
26 subcatchments within the Churchill–NelsonBCMontane Cordillera, Pacific Maritime, Western Interior Basin26 hydrometric
19 air temperature
18 precipitation
Mann-Kendall
1960–1999
  • Decrease in magnitude of hydrological events
  • Earlier snowmelt runoff events
  • Spring mean monthly flow increased due to greater snow melt potential
  • Timing of hydrological event strongly influenced by changes in air temperature
  • Decreasing trends in the southern region, while increasing trends in the northern regions
Cunderlik and Burn (2004)
156 hydrometric stations across CanadaAll areasAll ecozones+ except Arctic Cordillera portion of Arctic Ecozone+156 RHBN hydrometric
1974–2003 (156)1964–2003 (102) 1954–2003 (49)
  • Fluctuating trends in minimum flows (1-, 7-, 15-, and 30-day annual and seasonal low flow regimes)
  • Sensitivity of results to analysis time frame
Khaliq et al. (2008)
Winnipeg RiverON, MBBoreal Shield9 hydrometric
Adjusted Historical Canadian Climate Data and gridded climate dataset
Mann-Kendall
1924–2003
  • Increasing trend in streamflow with winter streamflow increasing by 60–110% over the entire basin
  • Changes related to climate as accounted for regulated vs. natural gauges
  • Basin hydrology has amplified coincident but smaller increases in summer / autumn precipitation
St. George (2007)
Miramichi RiverNBAtlantic Maritime2 hydrometric
3 meteorological
Linear regression
1970–1999
  • Spring and summer air temperature increases
  • Discharge relatively unchanged in winter / autumn
  • Increasing duration of low flow conditions reflecting greater evaporation rates from increased air temperatures
  • Earlier spring peak flows
Swansburg et al. (2004)
151 hydrometric stations across CanadaAll areasAll ecozones+ except Arctic Cordillera portion of Arctic Ecozone+151 RHBN hydrometric
Mann-Kendall
1967–1996 (151)
1957–1996 (71)
1947–1996 (47)
  • General decrease in mean annual streamflow with significant decreases in the southern part of Canada
  • Large decreases in August and September flow
  • Increase in spring (March / April) flow
  • Significantly earlier spring freshet in British Columbia.
  • Earlier freeze up, particularly in Eastern Canada
Zhang et al. (2001)
248 hydrometric stationsAll areasAll ecozones+ except Arctic Cordillera portion of Arctic Ecozone+248 RHBN hydrometric
Statistical trend estimate. Divided into ten geospatial regions. Each region with differing data length.
Max. time series = 1908–1997 (Mountain region)
Min. time series = 1961–1997 (NWT/NU)
  • Significant increase in July (Prairie and Pacific region) and December (NWT/NU) flow
  • Decreasing trend for annual mean flow (Central, Mountain-North and Pacific region)
  • Significant decrease in annual maximum daily flow (Central/East, Mountain-North, Pacific, NWT/NU)
  • Significant increasing trend in annual maximum daily flow (Central and Prairie)
  • Significant increasing annual minimum flow (Western Quebec/Southern Ontario, Mountain-North and Pacific)
  • Decreasing trend in annual minimum flow (Central/East region)
Adamowski and Bocci (2001)
Mackenzie basinAB, SK, BC, NT, YTTaiga Shield,
Boreal Shield,
Boreal Plains, Montane Cordillera, Taiga Plains,
Taiga Cordillera, Boreal Cordillera
16 hydrometric
9 meteorological
Spearman's rank correlation
1972–1999
  • At the scale of the Mackenzie basin, no obvious streamflow trends were present at either monthly or annual time scale
  • At the scale of individual rivers, evidence of earlier ice break up possibly linked to increasing air temperatures for the snowmelt months (April to June)
  • Overall, date and magnitude of peak flow show no trend but greater variability for lower Mackenzie and Peace rivers
  • Tendency of increasing streamflow variability
Woo and Thorne (2003)
64 hydrometric sites draining to high-latitude oceansYT, NT, NU, QC, ON, MB, NLNorthern Arctic, Southern Arctic,
Arctic Cordillera Taiga Cordillera, Taiga Plains,
Hudson Plains,
Taiga Shield
64 hydrometric
Mann-Kendall
1964–2003
  • Significant decreased trend in total annual freshwater discharge leading to a 10% decrease in the total annual discharge to the Arctic and North Atlantic Oceans
  • Attributed to decreased trend in precipitation over same period and suggests that changes in river discharge over northern Canada are driven primarily by precipitation rather than evapotranspiration
Déry and Wood (2005)
42 rivers draining into Hudson, James and Ungava baysNU, ON, SK, AB, MB, QCNorthern Arctic, Southern Arctic, Taiga Shield,
Taiga Plains,
Hudson Plain,
Boreal Shield,
Boreal Plains
42 hydrometric
Mann-Kendall
1964–2000
  • Decreasing trends in discharge for 36 out of 42 rivers
  • Total annual freshwater discharge in 2000 into Hudson, James and Ungava bays decreased by 13% from its value in 1964
  • Peak discharge rate associated with snowmelt has advanced by eight days and diminished in intensity
  • Spring freshet varies by five days for each degree of latitude
Déry et al. (2005)
56 rivers across North America (14 flow into Arctic Ocean and 42 flow into Hudson, Ungava and James bays)YT, NT, NU, QC, ON, MB, NLTaiga Cordillera, Taiga Plains,
Southern Arctic, Northern Arctic,
Arctic Cordillera
Taiga Shield,
Hudson Plains,
56 hydrometric
Mann-Kendall
1964–2000
  • Discharge to Arctic Ocean decreased from sites in North America
  • Discharge from sites draining Hudson, Ungava and James bays decreased by about 2.5 km3/y/y during 1964–2000
  • Reconstructed discharge from the Yukon River showed no significant change in discharge
  • Suggest concomitant decreases in precipitation and river discharge
McClelland et al. (2006)

Top of Page

A subset of hydrometric gauging stations form the Reference Hydrological Basin Network (RHBN) (Environment Canada, 2010b). This hydrological network of unimpacted catchments across Canada provides the national contribution to the World Meteorological Organization Monitoring Program for Climate Change. Brimley et al. (1999) and Harvey et al. (1999) defined the six original RHBN station selection criteria: i) stations should be minimally impacted with less than 10% modification from natural conditions; ii) absence of significant regulations or diversions upstream of the gauging station; iii) a minimum of 20 years of hydrological data; iv) future longevity of the station in its current pristine or stable state; v) accuracy of data records assessed by local experts for both open-water and ice-cover conditions; and vi) breadth of coverage of the different types of hydrometric stations (seasonal, continuous, streamflow, and lake level). From the original list of 255 RHBN gauging stations, 7 are lake level stations,   37 are seasonal stations, and 211 are continuous discharge monitoring stations (Environment Canada, 2006b).

Trends in lake levels were not considered in the analysis for this report because of the limited number within the RHBN database precludes any meaningful derivation of national trends at this time. There are, however, notable regional analyses on trends in lake levels. Van der Kamp et al.(2008), for example, explored patterns in long-term water level changes in 16 closed-basin lakes in the semi-arid prairie region of Canada. Their results indicated an overall declining trend in lake levels of 4 to 10 metres from circa 1920 to 2005. However, some east-central lakes demonstrated rising water levels from the 1960s onwards, which was linked to either higher precipitation or lower evaporation, in addition to sensitivity to changing land use relating to their low-lying relief (Van der Kamp et al., 2008). This further underscores the influence of regional climate variation in masking habitat change.

Daily mean flow data from 1969 to 2005 (which corresponds to the 1970 to 2005 hydrological years) were extracted from the Water Survey of Canada HYDAT database (Environment Canada, 2006b) for the 211 RHBN continuous flow river monitoring stations across Canada. Data were assessed for missing data and data quality. With catchment areas ranging from 3.9 to 145,000 km2, the majority of stations had more than 30 years of data over the 1970 to 2005 time period, with the remaining stations having more than 20 years of data. Following the approach of Burn and Hag Elnur (2002), no more than five years of data during the common period (1970 to 2005) could be missing for a station to be included in the analysis. A minimum of 31 years of data, with less than five years of missing records, were selected for this analysis as this record length ensured maximum time series length (Figure 11). Selecting a common period of record for the analysis allows investigation of variable climatic conditions.

The spatial coverage of hydrological gauging stations encompassed all ecozones+ (Figure 11); however, the distribution of stations across ecozones+ was non-uniform with the highest density by area of suitable gauging stations in the Western Interior Basin and the Atlantic Maritime ecozones+. More northerly ecozones+ presented lower densities. In addition, the spatial bias reflected a lack of stations with suitable data in northeastern Canada and the Prairies, where seasonal operation of gauging stations was common. Due to the low number of stations in some regions, caution should be observed when interpreting the results of the trend analyses.

Figure 11. Map of stations with suitable hydrological data used in trend analyses and table summarizing the number of suitable stations by ecozone+.
Graphic
Long Description for Figure 11.

This map and accompanying table (below) show the location of stations with suitable hydrological data to use in trend analysis. There are 172 stations with suitable data across the country. The map shows the spread of stations provides a spatial coverage of hydrological gauging stations encompassing all ecozones+. In general, coverage is better in the southern portion of the country and near population centres.

Ecozone+No. of
stations issued
Arctic Northern2
Arctic Southern3
Taiga Cordillera1
Taiga Plains11
Boreal Cordillera9
Taiga Shield6
Pacific Maritime11
Montane Cordillera27
Boreal Plains6
Boreal Shield31
Newfoundland Boreal12
Hudson Plains2
Prairies4
Western Interior Basin8
Atlantic Maritime34
Mixedwood
Plains
5

Top of Page

Richter et al. (1996) identified 32 annual hydro-ecological variables, known as the Indicators of Hydrologic Alteration (IHA), which represent ecologically important flow regime components (Table 3 and Table 4). These 32 IHA variables were calculated for hydrometric stations for each hydrological year from 1970 to 2005 (defined as October 1st to September 30th). The IHA variables quantify the magnitude (size), frequency, timing, duration, and flashiness (rate of change) of the flow regime. The IHA variables were calculated using the Nature Conservancy's IHA software (The Nature Conservancy, 2007). In the absence of long‑term ecological data, these IHA variables describing the hydrological regime offer a surrogate assessment of the habitat suitability for aquatic communities.

Table 3. Description of flow regime components, their instream ecological impacts, and exemplar variables.
Flow regime componentDescriptionEcological impactsExemplar variables
MagnitudeA measure of the amount of water passing a fixed point per unit time. Flow can vary with climatic conditions and catchment size both within and among river systems. Magnitude can be used as an indicator of the suitability of a habitat.
  • Habitat availability for species
  • Soil moisture availability for plants
  • Influences water temperature, oxygen levels, and photosynthesis in the water column
Mean monthly flow
Maximum or minimum flow
FrequencyA measure of the rate of recurrence of hydrological events above a given magnitude over a specified time interval.
  • Availability of floodplain habitats
  • Nutrient and organic matter exchanges between river and floodplain
  • Soil mineral availability
  • Influences bedload transport, channel sediment textures
Number of low or high flow events per year above a certain magnitude
DurationA measure of the period of time over which a hydrological condition, such as an extreme event or a normal condition, persists. Duration can be associated with a particular flow event or defined as a composite expressed over a given time period.
  • Duration of stressful conditions, for example, low oxygen and concentrated chemicals in aquatic environments
  • Distribution of plant communities in lakes, ponds, and floodplains
  • Duration of high flows for aeration of spawning beds in channel sediments
Number of days per year of a specified flow magnitude
TimingA measure of the regularity of hydrological conditions of a defined magnitude.
  • Predictability/avoidability of stress for organisms
  • Spawning cues for migratory fish
  • Evolution of life history strategies and behavioural mechanisms
The Julian date of the annual 1 day maximum flow
Rate of change (flow variability)Refers to the speed at which conditions change from one magnitude to another, for example, ‘stable’ streams have slow rates of change compared to ‘flashy’ systems that have rapid rates of change in flow conditions.
  • Stranding of species in isolated habitat patches (falling levels)
  • Entrapment on islands and floodplains (rising levels)
Number of daily positive/negative changes in flow
Source: Richter et al. (1997) and Poff et al. (1996)

Top of Page

Table 4. Ecologically relevant hydrological parameters used in the Indicators of Hydrologic Alteration (IHA) and their characteristics.
IHA groupHydrological regime componentHydrological Parameters
Group 1
Magnitude of monthly runoff
  • Magnitude
  • Timing
  • Median value for each calendar month (October–September)
Group 2
Minimum and maximum of annual runoff
  • Magnitude
  • Duration
  • Annual mean 1-day, 3-day, 7-day- 30-day and 90-day minimum
  • Annual mean 1-day, 3-day, 7-day- 30-day and 90-day maximum
  • Baseflow (7-day minimum / mean annual flow)
Group 3
Timing of annual one day minimum and one day maximum runoff
  • Timing
  • Julian date of each annual 1-day minimum
  • Julian date of each annual 1-day maximum
Group 4
Frequency and duration of high and low runoff
  • Magnitude
  • Frequency
  • Duration
  • No. of low pulses each year (where a pulse threshold is the median -25%)
  • No. of high pulses each year (where a pulse threshold is the median +25%)
  • Median duration of low pulses within each year
  • Median duration of high pulses within each year
Group 5
Variability in runoff
  • Frequency
  • Rate of change
  • Median of all negative difference between consecutive daily means
  • Median of all positive difference between consecutive daily means
  • No. of hydrograph rises (daily positive changes in flow)
  • No. of hydrograph falls (daily negative changes in flow)
  • No. of reversals (number of switches between positive and negative flow)
Source: adapted from Richter et al. (1996)

Top of Page

River flow data (m3/s) were converted to runoff (mm/day) to standardize the effects of differing drainage areas. Most variables were calculated using non-parametric (percentile) statistics because of the naturally skewed nature of many hydrological data records. The exception was the moving average variables (1-day to 90-day minimums and maximums) as these are always calculated as means. The variables were calculated for individual hydrological years and were used in the trend analysis. The presence of trends at each station for each variable was analysed using the Mann-Kendall methods to determine the significance of the detected trends using a permutation procedure (see Box). The analysis was completed using the MAKESENS program, a Microsoft Excel worksheet-based application (Salmi et al., 2002). Results were deemed statistically significant at the 10% level where p<0.1, consistent with other studies of this type (for example, Aziz and Burn, 2006).

Box: Details of Mann-Kendall trends analysis

A non-parametric statistic method was used to detect trends in the IHA variables because of the naturally skewed nature of many hydrological data records. The Mann Kendall trend analysis was applied to minimize the problems associated with differentiating between natural variability and data trends (Burn and Hag Elnur, 2002; Burn and Cunderlik, 2004; Kundzewicz and Robson, 2004). Originally derived by Mann (1945) and subsequently developed by Kendall (1975), the Mann-Kendall non parametric test for trend analysis has been applied by other researchers in similar studies (e.g., Burn and Hag Elnur, 2002; Chu et al., 2003; Burn and Cunderlik, 2004; Bonsal et al., 2006). The Mann-Kendall test is non-parametric and compares the relative magnitudes of input data as opposed to actual data values (Mann, 1945; Khaliq et al., 2008). The test statistic can be given by:

Non-parametric statistic test for trend analysis (formula)

Graphic

Long description for Box: Details of Mann-Kendall trend analysis (above)

The data are listed in the order in which they were collected over time (i.e. x1, x2, x3,… xn obtained at times 1, 2, 3,…n). The sign of all n(n-1)/2 possible differences xj – xk where j > k (i.e. x2 – x1, x 3 – x1, … xn – xn-2, xn – xn-1). The sign of all possible differences is calculated (i.e. sign(xj – xk) = 1 if xj – xk >0, sign(xj – xk) = 0 if xj – xk = 0, sign(xj – xk) = -1 if xj – xk <0). The test statistic, S, is equal to the total number of positive differences minus the number of negative differences for each time step.

Top of Page


Results: national summary and spatial distribution of hydrological trends, 1970 to 2005

Published research has explored the correlations in hydrological trends with changes in the climate (see examples in Table 2). However, as the results have been variable, it is important to interpret longer-term trends in streamflow with inter-decadal shifts in climate in addition to accounting for basin characteristics (Woo and Thorne, 2008). A national summary of the results of the application of the non-parametric Mann-Kendall trend analysis is presented in Table 5 and on Figure 12, and results are summarized by ecozone+ in Figure 13 and Figure 14. The results are then examined at the national and ecozone+ level in three sections: magnitude of runoff; timing frequency and duration; and flashiness. It is difficult to draw conclusions regarding trends for several ecozones+ because of the limited number of suitable hydrometric RHBN stations within them (that is, they have less than 10 stations). These ecozones+, Mixedwood Plains, Hudson Plains, Taiga Shield, Boreal Plains, Prairies, Taiga Cordillera, Arctic, and Boreal Cordillera, are not considered further in this section. In order to restrict potential bias and reduce error in conclusions the data for all ecozones+ are provided in Appendix 1.

Table 5a. Trend results for the Indicators of Hydrologic Alteration (IHA) variables for 172 RHBN stations used in this analysis, using data for hydrological years 1970–2005.
Group 1
- Magnitude of monthly runoff
IHA variables% of stations with significant increasing trend (p<0.1)% of stations with significant decreasing trend (p<0.1)
October4.78.1
November8.73.5
December16.97.6
January18.68.1
February14.08.7
March12.22.9
April29.13.5
May2.322.1
June5.819.8
July6.413.4
August4.728.5
September5.818.0
Table 5b. Trend results for the Indicators of Hydrologic Alteration (IHA) variables for 172 RHBN stations used in this analysis, using data for hydrological years 1970–2005.
Group 2
- Minimum and maximum of annual runoff
IHA variables% of stations with significant increasing trend (p<0.1)% of stations with significant decreasing trend (p<0.1)
1-day minimum12.826.2
3-day minimum13.425.6
7-day minimum14.025.0
30-day minimum15.723.3
90-day minimum16.321.5
Baseflow12.215.7
1-day maximum6.418.0
3-day maximum5.818.0
7-day maximum6.418.0
30-day maximum5.216.9
90-day maximum6.414.5
Table 5c. Trend results for the Indicators of Hydrologic Alteration (IHA) variables for 172 RHBN stations used in this analysis, using data for hydrological years 1970–2005.
Group 3
- Timing of annual one day minimum and one day maximum runoff
IHA variables% of stations with significant increasing trend (p<0.1)% of stations with significant decreasing trend (p<0.1)
Date of 1-day minimum16.38.1
Date of 1-day maximum6.410.5
Table 5d. Trend results for the Indicators of Hydrologic Alteration (IHA) variables for 172 RHBN stations used in this analysis, using data for hydrological years 1970–2005.
Group 4
- Frequency and duration of high and low runoff
IHA variables% of stations with significant increasing trend (p<0.1)% of stations with significant decreasing trend (p<0.1)
Low pulse number11.62.3
Low pulse duration7.014.5
High pulse number4.17.0
High pulse duration5.26.4
Table 5e. Trend results for the Indicators of Hydrologic Alteration (IHA) variables for 172 RHBN stations used in this analysis, using data for hydrological years 1970–2005.
Group 5
- Variability in runoff
IHA variables% of stations with significant increasing trend (p<0.1)% of stations with significant decreasing trend (p<0.1)
Rise rate8.120.9
Fall rate15.15.8
Hydrograph reversal30.210.5

Top of Page

Figure 12. Summary of the total number of stations displaying significant (p<0.1) increasing and decreasing trends for each IHA variable, using data for hydrological years 1970–2005.
Figure 12 is a graph summarizing the total number of stations displaying significant increasing and decreasing trends
Long Description for Figure 12.

This bar chart shows following information

Count
Variabledecreasingincreasing
October-148
November-615
December-1329
January-1432
February-1524
March-521
April-650
May-384
June-3410
July-2311
August-498
September-3110
1-day min-4522
3-day min-4423
7-day min-4324
30-day min-4027
90-day min-3728
Base flow-2721
1-day max-3111
3-day max-3110
7-day max-3111
30-day max-299
90-day max-2511
Date min-1428
Date max-1811
Low pulse #-420
Low pulse dur-2512
High pulse #-127
High pulse dur-119
Fall rate-1026
Rise rate-3614
Reversals-1852

Top of Page

Figure 13. Summary of the total number of stations displaying significant (p<0.1) increasing and decreasing trends for each IHA variable for the Atlantic Maritime, Taiga Plains, Boreal Shield, and Pacific Maritime ecozones+, using data for hydrological years 1970–2005.
Figure 13 is a graph summarizing the total number of stations displaying significant increasing and decreasing trends
Note the different x-axis scales. Only stations presenting significant trends are included.
Long Description for Figure 13.

This figure is composed of four bar charts displaying the total number of stations displaying significant increasing or decreasing trends for each indicator of hydrological alteration variable for the years 1970 to 2005 for the Atlantic Maritime, Taiga Plains, Boreal Shield and Pacific Maritime ecozones+. The results are a wide range of increasing and decreasing trends by variable. The Atlantic Maritime Ecozone+ had more decreasing trends than increasing trends across the indicators while in the Taiga Plains had more increasing trends than decreasing trends. The Boreal Shield and Pacific Maritime ecozones+ both had more decreasing trends than increasing trends across the indicators.

Top of Page

Figure 14. Summary of the total number of stations displaying significant (p<0.1) increasing and decreasing trends for each IHA variable for the Montane Cordillera and Newfoundland Boreal ecozones+, using data for hydrological years 1970–2005.
Figure 14 is a graph summarizing the total number of stations displaying significant increasing and decreasing trends
Only stations presenting significant trends are included.
Long Description for Figure 14.

This figure is composed of two bar charts displaying the total number of stations displaying significant increasing or decreasing trends for each indicator of hydrological alteration variable for the years 1970 to 2005 for the Montane Cordillera and Newfoundland Boreal ecozones+. For the Montane Cordillera Ecozone+, the number of increasing and decreasing trends varies across the indicators. The Newfoundland Boreal Ecozone+ had more decreasing trends than increasing trends across the indicators.

We focus primarily on the statistically significant (p<0.1) directional trends, although we have also reported non-significant (p>0.1) directional tendencies. Although we acknowledge that these are much more likely to be as a result of chance because of the inherently noise associated with hydrological data, they do provide a method for visualising regional consistency where a majority of stations have demonstrated a directional pattern. In addition, these non-significant results do not mean that trends are absent, but rather indicate that significant trends cannot be detected with the short data series available. Thus, these are used in the context for broad spatial characterisation within each ecozone+ rather than to draw specific conclusions about individual station response. This approach has been used by other researchers, for example Hannaford and Marsh (2006) who explored regional hydrological trends within the United Kingdom.

Top of Page

Trends in the magnitude of runoff

The magnitude, or size, of runoff can reflect differences in the availability of suitable habitat for aquatic communities in addition to influencing water temperature and oxygen levels, especially in areas with seasonal ice cover. Due to the large variability in driving factors (such as climatic and natural watershed differences), there is a lack of a consistent pattern across the country causing differences in the monthly timing of hydrological events, such as initiation of the spring freshet. There was large variability in the percentage of trends in runoff among different months (Table 5 and Figure 15). For example, a greater proportion of stations demonstrated statistically significant increasing trends (p<0.1) for the winter months (December to February) than demonstrated a decrease for the same months. The pattern for spring and summer runoff was mixed, with significant increases for median April runoff while the majority of trends showed significant decreases for runoff between May and September. In particular, August runoff significantly decreased for over 28% of the stations. There were a few spatial patterns for monthly runoff trends, for example the majority of stations in western Canada demonstrated significantly increasing trends in April runoff with a cluster of stations showing decreasing trends in the Great Lakes region.

Figure 15. Trends in long-term monthly runoff for RHBN stations using data for hydrological years 1970–2005.
Figure 15 is a graph displaying trends in long-term monthly runoff of RHBN stations
Long Description for Figure 15.

This bar graph shows the following information:

Monthly runoff% of sites with significant increasing trend (p<0.1)% of sites with significant decreasing trend (p<0.1)
Oct4.658.14
Nov8.723.49
Dec16.867.56
Jan18.608.14
Feb13.958.72
Mar12.212.91
Apr29.073.49
May2.3322.09
Jun5.8119.77
Jul6.4013.37
Aug4.6528.49
Sep5.8118.02

Minimum flows are ecologically important as they can limit the availability of specific aquatic habitats and also influence water temperatures and dissolved oxygen levels. Within the dataset, a greater proportion of stations demonstrated significant (p<0.1) decreasing trends than significant increasing trends in minimum runoff for all variables analyzed (Table 5 and Figure 16). These significant trends were more prominent, however, for variables describing minimum flows over shorter durations. For example, 26.2% of stations demonstrated significantly decreasing 1-day minimum flow compared with 21.5% of stations for minimum flow over 90 days. The tendency towards shorter duration trends could reflect the larger-scale overriding influences of seasonality over the longer term. Geographically, stations with significant trends towards reductions in 1-day minimum runoff were concentrated in southeastern and Atlantic Canada, and in southwestern Canada (Figure 17). Stations which demonstrated significant increases in 1-day minimum runoff were predominantly located in northwestern Canada and the few stations across northern Canada.

Top of Page

Figure 16. Trends in the magnitude of the 1-day, 3-day, 7-day, 30-day, and 90 day minimum runoff and in baseflow for RHBN stations using data for hydrological years 1970–2005.
Figure 16 is a graph displaying trends in the magnitude of the 1-day, 3-day, 7-day, 30-day, and 90 day minimum runoff
Long Description for Figure 16.

This bar graph shows the following information:

Minimum day runoff% of sites with significant decreasing trend (p<0.1)% of sites with non-significant decreasing tendency% of sites with significant increasing trend (p<0.1)% of sites with non-significant increasing tendency
1 day min26.1633.7212.7923.84
3 day min25.5836.0513.3723.26
7 day min25.0032.5613.9526.74
30 day min23.2633.1415.7026.16
90 day min21.5131.9816.2829.07
baseflow15.7038.3712.2132.56

Top of Page

Figure 17. Map showing trends in 1- day minimum river flow in natural rivers across Canada, using data for hydrological years 1970–2005.
Figure 17 is a map showiing the trends in the one-day minimum river flow in natural rivers across Canada
+ sign significant increasing trend (p<0.1), + sign increasing tendency (p>0.1), Yellow square form significant decreasing trend (p<0.1), white square form decreasing tendency (p>0.1), white circle form no trend
Long Description for Figure 17.

This map shows the trends in the one-day minimum river flow in natural rivers across Canada for the years 1970 to 2005. The map is marked with icons that indicate whether a monitoring station has observed a significant increase trend, an increasing tendency, a significant decreasing trend, a decreasing tendency, or no trend. Geographically, stations with significant decreasing trends in 1-day minimum runoff were concentrated in southeastern and Atlantic Canada, and in southwestern Canada. Stations which demonstrated significant increases in 1-day minimum runoff were predominantly located in northwestern Canada and the few stations across northern Canada.

Top of Page

For all of these variables, observed trends could reflect real change in aquatic processes, for example there could be a decreased amount of nutrient exchange between rivers and floodplains, in addition to more protracted and stressful low flow periods. Overall fewer stations showed statistically significant trends in maximum runoff, regardless of duration (Figure 12 and Figure 18). However, there appeared to be a tendency towards lower maximum runoff between 1970 and 2005 (Figure 18 and Figure 19). Spatially, there appeared to be significant decreasing trends in 1-day maximum runoff around the Great Lakes and St. Lawrence area but the remaining trends did not show a clear spatial pattern (Figure 19). Shorter and mid-duration (3-day to 30-day) maximum runoff variables demonstrated decreasing trends for groups of stations in eastern and lower western Canada. Spatial patterns in the 90-day maximum runoff reflect decreasing trends in lower latitude stations but greater number of significant increasing trends in western coastal areas and upper latitude stations. Strong spatial patterns in decreasing trends were noted in eastern and lower western Canada while increasing trends were noted in northwestern parts of Canada.

Figure 18. Trends in the magnitude of the 1-day, 3-day, 7-day, 30-day, and 90 day maximum runoff for RHBN stations using data for hydrological years 1970–2005.
Figure 18 is a graph showing trends in the magnitude of maximum runoff for RHBN stations
Long Description for Figure 18.

This bar graph shows the following information:

Variable% of sites with significant decreasing trend (p<0.1)% of sites with non-significant decreasing tendency% of sites with significant increasing trend (p<0.1)% of sites with non-significant increasing tendency
1 day max18.0250.006.4025.00
3 day max18.0249.425.8126.74
7 day max18.0248.266.4027.33
30 day max16.8654.655.2323.26
90 day max14.5353.496.4025.58

Top of Page

Figure 19. Map showing trends in the 1-day maximum river flow in natural rivers across Canada, using data for hydrological years 1970–2005.
Figure 19 is a map showing trends in the 1-day maximum river flow in natural rivers across Canada
+ sign significant increasing trend (p<0.1), + sign increasing tendency (p>0.1), Yellow square form significant decreasing trend (p<0.1), white square form decreasing tendency (p>0.1), white circle form no trend
Long Description for Figure 19.

This map shows the trends in the one-day maximum river flow in natural rivers across Canada for the years 1970 to 2005. The map is marked with icons that indicate whether a monitoring station has observed a significant increase trend, an increasing tendency, a significant decreasing trend, a decreasing tendency, or no trend. Overall, fewer stations showed statistically significant trends in maximum runoff than minimum runoff.

Top of Page

Summary of trends by ecozone+

Atlantic Maritime Ecozone+ (n = 34) (Figure 17, Figure 19, Figure 13)
Few stations presented significant (p<0.1) trends in median monthly runoff with the exception of May and June, with 19 and 17 of 34 stations, respectively, significantly decreasing, and August, where 19 of 34 stations significantly decreased. All stations with a statistically significant change (an average of 20 of 34 stations) demonstrated a decrease in minimum flow, regardless of duration. In addition, the majority of the remaining non-significant stations (an average of 13 of 14) had a tendency towards a reduction in the minimum flow. Although the variable describing baseflow conditions did not present strong significant trends (10 of 34), most stations (23 of 34) showed a tendency towards a reduction in value. The majority of stations also demonstrated reductions in maximum flow, regardless of duration with an average of five stations presenting significant decreasing trends with no stations showing a significant increasing trend. An additional average of 19 stations demonstrated a non-significant decreasing tendency.
Newfoundland Boreal Ecozone+ (n = 12) (Figure 17, Figure 19, Figure 14)
Few stations demonstrate significant trends in monthly runoff with the exception of August runoff where half of the stations showed a statistically significant decrease. The majority of stations showed a decrease in minimum runoff, regardless of duration. There were no clear patterns in maximum runoff where stations were split between tendencies towards increasing or decreasing directions.
Boreal Shield Ecozone+ (n = 31) (Figure 17, Figure 19, Figure 13)
Few stations showed a statistically significant trend in monthly runoff at the 10% level, with the exception of late summer runoff when 10 and 9 of 31 stations demonstrating decreasing trends for August and September runoff, respectively. The lack of clear directional trends could reflect the large longitudinal gradient in the spatial extent of this ecozone+. A greater proportion of stations presented significantly decreasing trends in both minimum and maximum runoff variables with the majority of the remaining stations showing a tendency towards a reduction.
Taiga Plains Ecozone+ (n = 11) (Figure 17, Figure 19, Figure 13)
7 out of 11 stations demonstrated a statistically significant increase in winter and early spring runoff (January to March). Strong significant increasing trends in minimum runoff were observed for an average of six stations within this ecozone+ with the majority of the remaining stations also showing a tendency towards an increase. In addition, five stations demonstrated a significant increase in baseflow. There were few significant trends in maximum runoff.
Montane Cordillera Ecozone+ (n = 27) (Figure 17 , Figure 19, Figure 14)
The majority of stations did not demonstrate significant trends in monthly median runoff with the exception of April which showed a strong increasing trend. Although not statistically significant, stations trended towards an increase in minimum runoff, particularly with longer duration variables. The majority of stations demonstrated decreasing maximum runoff but these were often non-significant.
Pacific Maritime Ecozone+ (n = 11) (Figure 17, Figure 19, Figure 13)
Few trends in monthly median runoff variables were apparent although there did appear to be an overall decreasing trend in late summer runoff (July, August, and September). There was a clear decreasing trend in minimum runoff, regardless of duration, while the majority of stations showed a significant decrease in baseflow. Conversely, the majority of stations showed a tendency towards increased maximum runoff but these were generally not significant.

Top of Page

Trends in the timing, frequency, and duration of extreme hydrological events

Many species respond to the timing of low and high flows to determine life cycle processes, for example as spawning cues for migratory fish or to provide access to marginal habitats during reproduction. Few stations demonstrated significant trends in the timing of the 1-day minimum and 1-day maximum runoff (Table 5 and Figure 20). However, the majority of stations (78 of 172) showed a tendency towards a later date in the annual minimum flow, while 85 out of 172 stations demonstrated a tendency towards an earlier date in the annual maximum flow. The latter is particularly important as it relates to the annual spring freshet for most river systems. The suggestion of trends for an earlier annual peak flow reflects previously reported earlier ice breakup trends (see Trends in river and lake ice break-up/freeze-up). Geographically, stations presenting a later 1-day minimum runoff were predominantly located in eastern Canada and the Great Lakes region while stations with earlier 1-day minimum were located in lower and coastal western Canada in addition to the Great Lakes and northeastern Canada. Later trends are noted in stations in the northern latitudes and lower eastern Canada.

Figure 20. Trends in date of annual 1-day minimum and 1-day maximum runoff for RHBN stations using data for hydrological years 1970–2005.
Figure 20 is a graph displaying how exposure to extreme high and low flow periods can cause stress to aquatic communities
Long Description for Figure 20.

This bar graph shows the following information:

Proportion of sites
Variableearlier trend (p<0.1)non-significant earlier tendency (p>0.1)later trend (p<0.1)non-significant later tendency (p>0.1)
Date of annual 1 day minimum8.1427.3316.2845.35
Date of annual 1 day maximum10.4749.426.4025.00

Top of Page

Exposure to extreme high and low flow periods can cause stress to aquatic communities in addition to affecting the abiotic environment. For example, these extreme flows can influence bedload transport of aquatic sediment, thereby affecting sediment composition and disturbance. Longer low flow periods may affect access for water birds for feeding, resting, and reproduction. Few stations demonstrated statistically significant increasing or decreasing trends with a split between trend direction in variables quantifying these extreme conditions (Figure 21). Despite a low number of significant trends in the number of low pulses, stations in eastern Canada demonstrated an increasing trend (both significant and non-significant) in the duration of low pulses (Figure 22). Stations in western and northwestern Canada presented shorter duration (both significant and non-significant) in the low flow periods with the exception of the southern portion of the Montane Cordillera which presented trends in longer durations of low flow periods.

Figure 21. Trends in the frequency and duration of low and high pulses for RHBN stations using data for hydrological years 1970–2005.
Figure 21 is a phraph showing trends in the frequency and pulses for RHBN stations
Long Description for Figure 21.

This bar graph shows the following information:

Proportion of sites
Variablesignificant decreasing trend (p<0.1)non-significant decreasing tendencysignificant increasing trend (p<0.1)non-significant increasing tendency
Number of low pulses2.334.6511.638.72
Duration of low pulses14.5332.566.9834.88
Number of high pulses6.9810.474.076.40
Duration of high pulses6.4027.915.2333.72

Top of Page

Figure 22. Map showing trends in the duration of low pulses for natural rivers across Canada, using data for hydrological years 1970–2005.
Figure 22 is a map showing the trends in the duration of low and high pulses recorded at Reference Hydrological Basin Network monitoring stations for the years 1970 to 2005
+ sign significant increasing trend (p<0.1), + sign increasing tendency (p>0.1), Yellow square form significant decreasing trend (p<0.1), white square form decreasing tendency (p>0.1), white circle form no trend
Long Description for Figure 22.

This map shows the trends in the duration of low and high pulses recorded at Reference Hydrological Basin Network monitoring stations for the years 1970 to 2005. The map is marked with icons that indicate whether a monitoring station has observed a significant increase trend, an increasing tendency, a significant decreasing trend, a decreasing tendency, or no trend. In general, stations in eastern Canada demonstrated increases or tendencies toward increases (non-significant) in the duration of low pulse events. Stations in western and northwestern Canada demonstrated decreases or tendencies toward decreases (non-significant) in the duration of low pulse events with the exception of the southern portion of the Montane Cordillera, which presented tendencies toward increases in the duration of low flow events.

Top of Page

Summary of trends by ecozone+

Atlantic Maritime Ecozone+(n = 34) (Figure 13, Figure 22)
Stations showed a tendency towards a reduction in the number of high pulses (20 of 34 stations). In addition, the majority of stations demonstrated a tendency towards an increase in the duration of low pulse events but the pattern for the number of low pulse events was not clear.
Newfoundland Boreal Ecozone+ (n = 12) (Figure 14, Figure 22)
The majority of stations showed a tendency towards a later date of both the annual maximum and the annual minimum. The tendency towards a later annual maximum is not reflected across the country where the majority of stations are demonstrating an earlier spring freshet.
Boreal Shield Ecozone+ (n = 31) (Figure 13, Figure 22)
Over half of the stations had a tendency towards an earlier date of annual maximum runoff (18 of 31 stations) suggesting an earlier spring freshet. A large proportion of stations (19 of 31) showed a tendency towards increased duration of high pulse events.
Taiga Plains Ecozone+ (n = 11) (Figure 13, Figure 22)
Stations within the region did not demonstrate any clear patterns in the number of low and high pulses nor in the duration of high and low flow pulse events. However, 7 out of 11 stations showed a statistically significant (p<0.1) decreasing trend in the duration of low pulse events.
Montane Cordillera Ecozone+ (n = 27) (Figure 14, Figure 22)
Although not statistically significant, the majority of stations demonstrated a tendency towards an earlier date of maximum runoff. There did not appear to be any clear trends in high and low pulse events with the exception of a tendency towards a decreased duration of high pulse events.
Pacific Maritime Ecozone+ (n = 11) (Figure 13, Figure 22)
Stations demonstrated a tendency towards an earlier date of maximum runoff, potentially reflecting an earlier spring freshet. Stations demonstrated a tendency towards a decreased duration of low pulse events.

Top of Page

Trends in flashiness

Variability in flows can alter the availability of habitat and nutrients. Although naturally flashy systems are often considered ecologically disturbed or 'harsh' systems, decreases in flashiness can still stress aquatic communities, as organisms attempt to adapt to new conditions. Several stations demonstrated significant trends (both increases and decreases) in the variables quantifying the variability of the annual flow regime (50 for rise rate, 36 for fall rate, and 70 for number of hydrograph reversalsFootnote8, out of 172 stations) (Table 5 and Figure 23). Interestingly, a greater number of stations presented statistically significant increases in the variability of annual runoff, as quantified by the number of reversals (52 of 172 stations with significant trends and an additional 43 stations demonstrated a tendency towards an increase) (Figure 23). Spatially, these stations were largely found in western and northwestern Canada, in addition to some stations in southeastern Canada (Figure 24).

Top of Page

Figure 23. Trends in the variability of runoff for RHBN stations using data for hydrological years, 1970–2005.
Figure 23 is a graph showing trends in the variability of runoff for RHBN stations
Long Description for Figure 23.

This bar graph shows the following information:

Proportion of sites
Variablesignificant decreasing trend (p<0.1)non-significant decreasing tendencysignificant increasing trend (p<0.1)non-significant increasing tendency
Rise rate20.9334.888.1425.00
Fall rate5.8115.1215.1231.40
Reversals10.4724.4230.2325.00

Top of Page

Figure 24. Map showing trends in the number of hydrograph reversals in natural rivers across Canada, using data for hydrological years 1970–2005.
Figure 24 is a map showing trends in the hydrograph reversals in natural rivers recorded at Reference Hydrological Basin Network monitoring stations for the years 1970 to 2005
+ sign significant increasing trend (p<0.1), + sign increasing tendency (p>0.1), Yellow square form significant decreasing trend (p<0.1), white square form decreasing tendency (p>0.1), white circle form no trend
Long Description for Figure 24.

This map shows the trends in the hydrograph reversals in natural rivers recorded at Reference Hydrological Basin Network monitoring stations for the years 1970 to 2005. The map is marked with icons that indicate whether a monitoring station has observed a significant increase trend, an increasing tendency, a significant decreasing trend, a decreasing tendency, or no trend. A large number of stations presented statistically significant increases in the variability of annual runoff, as quantified by the number of reversals. The map shows that the stations showing reversals were largely found in western and northwestern Canada, in addition to some stations in southeastern Canada.

Top of Page

Summary of trends by ecozone+

Atlantic Maritime Ecozone+ (n = 34 stations) (Figure 13, Figure 24)
Decreasing trends in rise rates were demonstrated for 11 stations and an additional 16 stations showed a decreasing tendency. Conversely, 10 stations showed an increasing trend in fall rates with an additional   20 stations showing a tendency towards an increase. The number of reversals did not show a directional trend.
Newfoundland Boreal Ecozone+ (n = 12) (Figure 14, Figure 24)
There were few significant trends in variability (pulse rates, rise and fall rates, or number of reversals). However, the majority of stations demonstrated a tendency towards an increase in the fall rate.
Boreal Shield Ecozone+ (n = 31) (Figure 13, Figure 24)
A quarter of stations demonstrated a significant increase in the number of flow reversals reflected in an increase in the fall rate and a decrease in the rise rate. These results suggest increased variability in flows, and therefore on habitat, in rivers within this ecozone+.
Taiga Plains Ecozone+ (n = 11) (Figure 13, Figure 24)
The majority of stations (9) presented a significant increasing trend in the number of reversals with the remaining two stations showing a tendency towards an increase. This demonstrates variability in runoff has increased and suggests increased levels of potential hydrological stress within the system.
Montane Cordillera Ecozone+ (n = 27) (Figure 14, Figure 24)
The majority of stations showed a tendency towards a decrease in rise rates with five being significant (p<0.1). In addition, there was a tendency towards an increase in fall rate. This is reflected in the majority of stations demonstrating a tendency towards an increase in flow variability as quantified by the number of flow reversals, with 13 being statistically significant (p<0.1).
Pacific Maritime Ecozone+ (n = 11) (Figure 13, Figure 24)
Stations demonstrated a tendency towards an increase in the rise and fall rates in addition to the number of reversals. Reflecting the other variables, this suggests an increased variability within these rivers.

Top of Page

Summary of hydroecological trends

The application of the non-parametric Mann-Kendall analysis to identify trends in hydro-ecological variables from the Canadian RHBN network has resulted in the identification of spatial and temporal patterns for the period 1970 to 2005. Using IHA variables (Richter et al., 1996), we have attempted to quantify trends in ecologically important hydrological habitat.

Few monitored rivers demonstrated significant trends in monthly median runoff with the exception of April and August. Examination of trends in maximum and minimum flows revealed limited numbers of statistically significant trends at the national scale. However, for both minimum and maximum runoff, the majority of significant trends were decreasing. In addition to the significant stations, a third of stations demonstrated a tendency towards decreased minimum flows while just over half of the stations showed a tendency towards decreased maximum flows. The implications for decreased magnitude in runoff are significant, for example effects on habitat availability for aquatic communities. However, there is some spatial variability in these minimum and maximum runoff trends, for example, stations within the Taiga Plains and Pacific Maritime ecozones+ presented an increase in these variables while the remaining studied ecozones+ demonstrated an overall decrease.

A greater proportion of significant stations showed a significant trend towards earlier date of the annual 1-day maximum with an additional 56% of stations demonstrating a tendency towards this. The annual 1-day maximum runoff often occurs during the spring freshet thus this suggests that ice break-up is occurring earlier. Combined with a tendency towards later ice freeze-up (see Trends in river and lake ice break-up/freeze-up section below), this suggests an extension to the open water season which will affect the aquatic ecosystem. The majority of stations do not show significant trends in the frequency and duration of extreme high and low flow events. However, the majority of stations showed a tendency towards a decrease in the rise rate, an increase in the fall rate, and an increase in the number of reversals, suggesting a trend towards an increase in flow variability.

Top of Page

Footnotes

Footnote 8

The number of hydrograph reversals quantifies the variability of the flow regime by calculating the number of times that flow switches either from a rising to a falling condition or from a falling to a rising condition. This reversal may directly affect aquatic communities, for example macroinvertebrates often lack the mobility to respond to rapidly changing conditions (The Nature Conservancy, 2007)

Return to reference 8 referrer

Top of Page

Return to Table of Contents