The objectives of the study components
documented in this report were as follows:
- select three matched pairs of
agricultural watersheds in the Lake Erie drainage basin suitable for
conversion to conservation farming systems;
- compile physical and social
information on these watersheds to allow for agronomic and
environmental analysis; and
- monitor meteorology as a primary
causative factor in crop performance and erosion - transport
The three pairs of sub-watersheds were
monitored in a similar manner. Each pair of sub-watersheds had one main
weather station located in one of the sub-watersheds. The main station
included automatic instrumentation for monitoring the following
parameters at indicated frequencies:
- rate of rain or snow (15 minutes);
- wind speed (5 minutes);
- air temperature (15 minutes);
- solar radiation (15 minutes);
- potential evaporation (daily); and
- volume of rain (daily).
In addition to the main stations, each
test sub-watershed outlet had secondary meteorological stations where
the following measurements and measurement frequencies occurred:
- rate of rain (hourly);
- air temperature (hourly);
- water temperature (hourly);
- soil temperature at 3 depths (hourly);
- relative humidity (hourly) at Kettle
In addition to these two stations, each
pair of sub-watersheds had a set of three or four standard rain gauges
spatially distributed to address the heterogeneity of precipitation.
All instruments functioned, with minor
upsets, over the period January 1989 to June 1992.
In general, all study areas experienced a
cooler than normal 1989 and warmer than normal temperatures in 1990 and
1991. Most winter periods were warmer than normal resulting in frequent
winter rains and minimum snowpack. Growing season air temperatures did
not differ from the normal values significantly. In all, air
temperatures were within 1 degree C of normals in all areas on an annual
Temperature data were used to calculate
Corn Heat Units (CHU) for all sub-watersheds. CHU were found to be below
average in all study areas in 1989 and 1990 while 1991 values marginally
exceeded the normals. Essex displayed the highest corn heat units,
followed by Kettle Creek and then Pittock in all years. This was also
the order of recorded total solar radiation, discussed below.
Soil temperatures at depths of 5, 50 and
150 cm were reported for each sub-watershed and compared to air
temperature. The shallow 5 cm temperature probe was shown to clearly
relate to air temperature as expected with a typical annual range
encompassing 26 to 27° C. The deepest probe was below the frost line and
displayed a typical annual range covering approximately 10° C.
Similar total precipitation trends were
evident in each of the study areas. All areas had much lower than normal
annual precipitation in 1989, 1991 and the first six months of 1992. All
study areas had higher than normal precipitation in 1990.
In general, winter precipitation was
lower than normal while growing season totals were near normal. In
conjunction with warmer than normal temperatures, winter periods were
generally characterized by minimal snow cover and frequent mid-winter
thaws. This fact has implications in terms of winter erosion potential
and spring runoff volumes discussed in (PWS) Reports #5 and #6 (SWEEP
Report #73 and SWEEP Report #74).
Significant differences were noted on a
storm-by-storm basis and in the long term between study areas and
between single event rain volume measured at different locations within
the same study area.
Extreme hourly and daily precipitation is
reported for the study period. Maximum hourly and daily precipitation
were 66.5 mm/hr and 95.6 mm/day, both in Kettle Creek.
The total number of significant
precipitation events, exceeding 60 mm/day was one in each study area.
The number of precipitation events exceeding 40 mm/day was 12 in total.
The frequency of extreme high precipitation events and associated
erosion-runoff events was low during the course of this study.
Wind speeds recorded during the study were
slightly below normal in all areas. Wind speed affects the rate of
potential evaporation and snowmelt. The lower than normal wind speeds
have not likely been a significant factor in the study results.
Solar radiation has direct affects in terms
of crop productivity, potential evaporation and rate of snowmelt.
Essex received the most solar radiation
through the study period, Kettle Creek received the second most and
Pittock receiving the least. This trend corresponds with differences in
latitude and is expected. Total solar radiation received was near normal
in all areas.
The Pilot Watershed
Study was designed to be conducted in the Lake Erie watershed within
small sub-watersheds. The sub-watersheds were to be approximately 400 ha
in size and representative of broader physiographic units. Agriculture
within these sub-watersheds was to be carried out on a commercial scale
and typify the predominant crops and livestock systems found in
southwestern Ontario. It was particularly important that the selected
sub-watersheds be representative of the landscapes that were
contributing phosphorus to Lake Erie.
The five phases of the basin selection
screening exercise are presented in Figure 2.1. Figure 2.1 illustrates
the progression from a general to a more detailed analysis as the focus
moved from large candidate watershed areas to specific sub-watersheds.
Along the way, selection criteria and exclusionary factors were applied.
Initially these were based primarily on physical attributes that were
readily evaluated with published information applied at a smaller scale.
Each subsequent phase introduced a
greater level of detail and expanded the range of criteria from the
strictly physical watershed criteria to encompass socio-economic factors
that had a bearing on the final selection. The mapping analysis was
completed at a larger scale and the data inputs were drawn increasingly
from primary data collection efforts. The final phase focused entirely
on farm operator characteristics and relied, for these, on inputs from
the local farm communities.
By the end of April, 1988 the watershed
and the sub-watershed designations had been finalized. The final
||5th Concession drain
||2nd Concession drain
The virtue of the basin selection approach
was that it eliminated most of the unsuitable watershed and
sub-watershed areas at an early stage before detailed and costly data
collection efforts were initiated. Sub-watersheds that were carried into
the final screening phases generally satisfied the physical criteria and
further distinctions among them could be made based on farm operator
It was not enough to assure physical
homogeneity across the control and test study areas as would be required
under a typical experimental situation. The human dimension was of
paramount importance, for without the cooperation of farm operators in
each study the PWS would not be feasible.
The screening process had to locate
paired sub-watersheds that exhibited physical similarity, that were
representative of major agricultural regions in southwestern Ontario and
that were farmed by individuals who, for the most part, were likely to
cooperate with the study team. These demanding requirements necessitated
an innovative methodology that enabled the screening of a very large
number of candidate areas while still allowing for a degree of site
specific investigation that would enable an assessment of local farm
communities. The phased approach to screening fulfilled these
requirements while making best use of available study resources.
The three selected Pilot Watershed Study
areas represented typical farming communities in southwestern Ontario.
Cash cropping and livestock operations predominated. Interest in the
Study was high in all areas, with only two out of 78 farmers declining
to participate in any way. Similar numbers of farmers operated in each
area and the majority of farms were larger commercial operations.
Cultural practices were conventional, though experimentation with
conservation practices had occurred in all the sub-watersheds.
From a socio-economic perspective, the
three pairs of sub-watersheds fulfilled project requirements in that
they contained commercial farm operators who were familiar with
conservation issues, but were generally committed to conventional
farming methods at the time of the selection. This enabled a more
critical examination of a pro-active approach to encourage the use of
soil conservation practices.
From a physical standpoint, the
sub-watersheds are considered to be as closely paired as could be
expected given the various constraints and objectives that were in
effect during the selection process. Differences among sub-watershed
parts in shape, size and soils caused somewhat similar hydrologic
responses. These differences necessitated a more critical approach in
the analysis of monitoring data and encouraged greater reliance on field
scale investigations and modelling analysis to help interpret monitoring
results from the watershed outlets. Results
In-stream total suspended solids (TSS) and
total phosphorus (TP) concentrations at the six watershed outlets are
quite variable both temporally and with respect to location (i.e. test
or control sub-watershed). TP concentration at the watershed outlets
follows similar patterns to TSS, since a large proportion of the total
phosphorous content is sediment bound. In the Essex watershed, the test
sub-watershed exhibits higher TSS concentrations than the control
throughout the study period except for the first six months of 1992. In
general however, TSS concentrations consistently decreased in the test
sub-watershed over the time conservation practices had been phased in.
The control sub-watershed did not exhibit this trend. In the Kettle
Creek watershed, TSS concentrations were again higher in the test than
in the control except in 1991 when average annual TSS concentration was
higher in the control. Neither the test nor the control sub-watershed
exhibited a consistent increase or decrease in TSS over the study
period. TSS concentrations at the Pittock watershed were lower at the
test outlet than at the control except for during the first six months
of 1992, when both the test and control TSS concentrations were very
low. Overall, the lower Pittock test TSS concentrations are positive
evidence of the potential water quality enhancement of conservation
As with concentrations of water quality
parameter concentrations, loads are quite variable in time, especially
in the fall season, and with respect to location. In the Essex
watershed, annual unit area TSS loads where similar for the test and
control sub-watersheds in every year of the study. In Kettle Creek, the
test sub-watershed outlet had higher unit area TSS loads than the
control except during 1991 - a very dry year. A partial explanation for
low Kettle Creek control TSS loads is the sediment trap in the low
lying, wetland area situated in the control sub-watershed. In the
Pittock watershed, TSS loads were not consistently higher in either of
the sub-watersheds and therefore inconclusive.
Loadings of TP were similar in pattern to
TSS loads, since a large portion of the TP is delivered in a sediment
General trends between water quality and
conservation farm benefits were more apparent at the microbasin scale as
confounding spatial factors tended to decrease with a corresponding
decrease in size or scale of monitoring.
In Essex, TSS concentrations consistently
decreased in the test microbasins while showing no trends in the control
microbasins. The same decreasing trend was observed for one of the test
microbasins in Kettle Creek while the other microbasins (test and
control) was inconsistent. This decrease in TSS in test microbasins is
evidence of potential positive effects of land conservation practices on
water quality. The lack of flow events in the Pittock control
microbasins makes comparisons less conclusive. Microbasin TP
concentrations follow a very similar pattern to TSS concentrations.
Overall, unit area microbasin loads show no
distinct trends with respect to time or location (test or control).
Again, the rarity of significant microbasin flow events makes it
difficult to make firm conclusions and underscores the need for further
Correlation analysis of lumped microbasin
loading data did however, provide for a comparison of test and control
loads for both Essex and Kettle Creek watersheds. In both cases, TSS and
TP loads were higher in the test at low flows but were higher in the
control at moderate to high flows. This indicates a possible advantage
to land conservation practices in the enhancement of water quality
during critical large rainfall/flow events when a very large proportion
of the total annual soil loss usually occurs.
The rainfall simulation component of the PWS
water quality evaluation produced the most obvious and predictable
results. This is due to the controlled nature of these tests in terms of
area, land management, rainfall intensity and time, and precise
measurement of runoff, soil moisture, slope and residue.
Test plots had significantly lower loads
of TSS and TP, often due to lower TSS and TP concentrations in test plot
runoff particulary in Essex. This was particularly true for the
Pre-Plant period which generated the highest unit area loads by far and
for the Post-Tillage period which generally produced the next highest
loads of the three time periods examined.
The Post-Tillage and Pre-Plant periods
bracket a significant period of the year usually from early fall to
early summer and represent an erosion prone period with minimal live
plant cover. Results show that during this period conservation tillage
reduces TSS and TP loadings.
- The watershed scale proved to be the
most complex in terms of inherent variability, of climatic, soil,
runoff and erosion process factors. However, even with the inherent
variability at this monitoring scale, at four of the six subwatersheds
there was an indication that an increase in percent cover decreased
water quality loads.
- The positive effects of conservation
practices in the test watersheds were most evident during the November
through April period when the percent cover was much higher than the
- In Essex test sub-watershed, average
and median yearly TSS concentration decreased consistently since the
onset of the project. In contrast, Essex control sub-watershed average
and median concentrations were variable. Other watersheds did not show
this trend but may with time.
- Dry weather of the study period has
not provided for a comparison of test versus control during years with
higher than normal precipitation.
- During high flows (in the November to
April periods) the high percent cover in the test microbasins of Essex
and Kettle Creek proved to be effective in reducing TSS loads when
compared to control microbasins (low percent cover).
- Pittock microbasin loads were very low
in both test and control. Due to the lack of moderate to high flow
events, conclusions could not be drawn from the Pittock microbasins.
- Conservation tillage resulted in
significantly reduced loadings of TSS and TP in general with some
exceptions. These exceptions include Post-Tillage periods wherein some
cases, test plots produced significantly more runoff due to soil
surface differences between test and control plots.
- Pre-Plant periods produced the highest
loadings in all areas due to much higher TSS and TP concentrations and
generally higher runoff volumes. Post-Harvest measurements are of
little value since in conventional systems the condition is short
lived and crop residue levels are similar in test and control plots.
- Crop live and dead cover is an
effective level of protection against interrill (sheet) erosion as
determined in intense simulated rainfall measurements.
- Rainfall simulation measurements are
an effective means of evaluating conservation farming systems and
identifying the effect of various independent factors (i.e. cover,
soil moisture, tillage practices).