NSCP Research - Project F
Open Category REP under the NSCP Research Program
The research program under NSCP is inviting unique proposals for research on soil quality issues which have not been identified in the foregoing five. They should be confined to the areas of work set out on page 1 of this guideline and subject to the preferences identified there. It is fully expected that any submission will be a serious contribution to our understanding of the conservation of soil quality, its measurement or extension to the agricultural community.
Although the expected scope of the work is undefined here, the scope of the work in the proposal should be clearly stated. The proposal should fall within the guidelines given above, observing the two year time frame, and the suggested level of effort. On farm demonstrations are discouraged in favour of more technically sophisticated studies which relate to our understanding of sustainable and/or restorable soil quality.
The Effect of Soil Quality on Field Scale Runoff under Conventional and Conservation Tillage Systems
Beak Consultants Limited
The National Soil Conservation Program (NSCP) is a research program managed and funded by Agriculture Canada. The program focuses on the improvement of agricultural soil and water quality through education and implementation of farm conservation practices. The program is multi-faceted in design and approach and includes several topics of study primarily in the soil quality and soil conservation areas.
As part of the NSCP, Beak Consultants Limited (BEAK) conducted a study examining soil quality, agricultural runoff quality and groundwater quality, under conventional and conservation tillage systems. This research study focused upon field scale processes and encompasses two growing seasons. The study's primary objective was to examine the transport and fate of nutrients (primarily nitrate-N and phosphorus) and metolachlor, a commonly used pesticide in Southern Ontario, under two differing tillage management systems.
This final report includes sections on literature review, study methods, study sites description and design, meteorology, soil quality assessment, surface water and groundwater monitoring and the effects of conservation and conventional tillage on soil and water quality within the study sites.
1.1 Background and Literature Review
Soil and Water Quality
Research over the past forty years in North America has produced conflicting evidence with respect to the effect of no-till farm production upon soil quality, surface runoff quality and groundwater quality. It is well recognized that certain agricultural chemical inputs, primarily nitrogen, (Keeney, 1986), phosphorus (Wall et al., 1982; Logan, 1982; Miller et al., 1978) and pesticides (Agriculture Canada, 1991) are found outside their target soil and plant zones - and some, such as nitrate nitrogen contamination, has been called ubiquitous (Gillham, 1991). Researchers have suggested that the means to controlling the reduction of nitrate transport in surface water, tile flow and groundwater is to examine carbon-nitrogen cycling and increase carbon contents in soils to promote better soil quality and associated reduced nitrate transport (Shirmohammadi, et al.; 1991, Gillham, 1988). Current research conducted by the Waterloo Centre for Groundwater Research has found that nitrate contamination of the shallow aquifers is widespread and that most deep aquifers generally have low nitrate concentrations (Russell et al., 1992).
Other studies have examined means to quantify pesticide leaching through modelling techniques (Carsel and Jones, 1990; Sjoerd and Zee, 1990). These modelling techniques take into account adsorption/desorption processes and leaching potential and provide a framework for evaluation of agricultural management systems such as no-till. However, the data used in these studies are representative of United States climatic conditions and therefore the modelling evaluations are not directly applicable to Southern Ontario. Sjoerd and Zee (1990) identified a stochastic technique for evaluating pesticide transport which accounted for spatial variability with relatively little data. The model accounts for spatial variability through the use of an approximation of apparent residence time variance.
Researchers in the United States have examined surface runoff volumes between conventional systems (usually mouldboard plow followed by discing) and conservation systems which included no-till and minimal tillage systems. During an eight year study, Laflen et al. (1990) found that less than half as much runoff was generated from a ridge till fields on steep loess slopes compared to a conventionally tilled fields (30-60 hectares) . Similarly, Laflen et al. (1990) reported a 90% reduction (compared to conventional tillage) in surface runoff in a maize-cowpea rotation with no till on slopes ranging from 1 to 15%. In these studies, soil loss was drastically reduced with no-till.
Long-term research in Coshocton USDA Hydrologic Station in Ohio found that no-till fields, on steep silt loam slopes (10-15%), generated less surface runoff during the critical summer thunderstorm period than did plowed fields (Edwards, 1991). Annual erosion was also much less. These results were consistent between well drained fields and fields with poorly drained subsurface soils.
Other U.S. experiences confirm these results (Berg et al. 1988; Blevens et al. 1983). Explanations for reduced surface runoff in conservation systems include reduced surface flow velocities due to mulch (Edwards, 1991) and higher macroporosity and infiltration rates due to earthworm tunnels (Hendrix et al, 1988).
In Ontario, similar studies conducted by Vyn et al. (1979) over a six year period (1971-76) were in general agreement with the U.S. results. No-till in Guelph loam had 35% less runoff than conventionally tilled fields. In these studies, benefits were credited to reduced soil sealing and crusting and impeded runoff which provide for more infiltration. However, recent Tillage 2000 studies and TED research has contradicted these observations (O'Neill et al. 1990). This study found that plots with a nine year history of no-till management had higher bulk density, lower total porosity, lower micro-porosity, lower saturated hydraulic conductivity, lower matric flux potential and lower infiltration rates than similar plots in which the moldboard plow was used. Study site soils ranged from fine sandy loam to clay loam. Sites with minimum till were not significantly different than the mouldboard sites. These studies indicated that no till systems may result in higher surface runoff due to reduced soil infiltration and hydraulic conductivity as a result of altered soil structure. It was felt, however, that soil loss and adsorbed chemical loss (i.e. phosphorus and pesticides) were reduced in no till due to the higher surface protection and reduced particle detachment inherent to no till. Increased groundwater contamination in no till was felt to be unlikely since water movement through soils may be reduced in no till. It was also suggested that surface runoff of soluble chemicals may be increased in a no till system.
It is well accepted that conservation tillage reduces soil loss and losses of soil adsorbed chemicals from tilled fields as well as resulting in soil quality improvements. The reasons for the soil quality improvements and reduced soil loss under these soil conservation systems is generally credited to the effect of plant residue sheltering soil particles against raindrop impact and impeding surface runoff velocity (Agriculture Canada, 1983).
Immunoassays for Pesticides Analysis
Immunoassays are analytical procedures based upon the specific binding of animal-derived antibodies to a target molecule. When coupled with a visualization method, immunoassays provide simple, specific, sensitive and rapid detection at low cost. The most common use of immunoassays is in clinical laboratories to identify drugs, hormones, viruses and bacteria. Within the past 15 years there has been a heightened interest in generating antibodies for residue analysis of pesticides. Recent work was reviewed by Hammock et al. (1987), Lankow et al. (1987), and Vanderlaan et al. (1988). Immunoassays for triazine and phenoxy herbicide detection and measurement in the environment have only recently been established. Bushway et al. 1988; Fleeker, 1987; Hall et al., 1987; Schlaeppi et al., 1989; Vanderlaan et al. 1988; Hall et al., 1989). Laboratory and initial field data generally show excellent correlation with conventional analytical tests. For example, Schlaeppi et al. (1989) could detect atrazine in water samples as low as 0.05 µg/L. For 25 samples containing atrazine (>0.05 µg/L), the correlation between immunoassay and conventional chemistry measurements, calculated by regression analysis, was excellent (r=0.91, p <0.0005).
More recently, Thurman et al. (1990) compared ELISA immunoassay to GC procedure for the analysis of triazine herbicides in surface water and groundwater in Kansas, U.S.A. Apparent recoveries from natural water and spiked water by both methods were comparable at 0.2 - 2 µg/L. The authors of this study concluded that the combination of screening analysis by ELISA, which requires no sample preparation, and confirmation by GC was designed for rapid, inexpensive analysis of triazine herbicides in water.
Clearly, immunoassays may provide for the analysis of a large number of samples without extraction (hence, low cost), rapidly and accurately. Among other potential benefits, immunoassays may provide an opportunity to greatly increase the scope of water monitoring programs, or to increase the frequency of tests, without added costs.
Pesticide Monitoring and Interpretation
Pesticide monitoring at the mouth of major watersheds in Ontario has been ongoing since the mid 1970's (Frank and Logan, 1988) and continues today as part of the enhanced effort (60 samples/year) at the mouth of the Grand, Thames and Saugeen Rivers and at 85 other sites in Ontario on a less frequent basis (pers. comm. L. Logan). Pesticide monitoring has also been conducted at the mouths of smaller agricultural watersheds as part of PLUARG (PLUARG, 1978a, Frank et al., 1982) and more recent specialized studies in the Nissouri Watershed (BEAK, 1989) and the Kintore Watershed presently monitored by Environment Canada, Inland Waters (pers. comm. D. Draper). The PLUARG study was conducted on 11 selected small agricultural watersheds in Southern Ontario. On average, atrazine was consistently found in runoff samples with an average load of 0.2 g/ha/yr. The Nissouri Creek study reported atrazine concentrations as high as 350 µg/L in surface waters and estimated that 0.84 to 3.3% of the atrazine which was applied in 1985-1986 was detected in runoff. This corresponded to a loading rate of 0.06 to 0.31 ha/yr. A crude extrapolation of these results to all of Southern Ontario indicates that loadings of atrazine to Lake Erie, the primary basin for Southern Ontario class 1 agricultural land, could approach 1 tonne/year.
There are no studies which have been specifically aimed at identifying empirical relationships between pesticide runoff and spatial factors such as watershed size, treated areas and other physical factors in Ontario or elsewhere. The frequency of sampling is usually insufficient to provide reliable estimates, better than an order of magnitude for watershed loadings. The Nissouri Creek study of atrazine runoff (BEAK, 1989) did provide a high level of atrazine runoff loading information which in turn was useful in examining event based, seasonal and annual runoff loadings. Baker (1987) monitored atrazine and nine other pesticides in three Ohio watersheds ranging in size from 386 km2 to 16,359 km2. While this study did not focus upon watershed scale, results did indicate that the time weighted mean concentration of atrazine (April to August, 1983-85) was inversely proportional to watershed size. All watersheds had similar agricultural use.
Numerous studies have reported the observed relationship between runoff hydrology and conventional water quality parameters such as suspended solids and phosphorus (Bodo and Unny, 1983; PLUARG, 1978b). These relationships are relatively well understood in the context of small headwater areas as well as at the mouth of large watersheds where flow and water quality attenuation effects are significant. No studies have attempted to relate pesticide runoff timing and impact (concentrations and durations) to the hydrologic response of various catchment sizes on a seasonal basis, recognizing that the runoff behaviour of these materials differs from that of conventional parameters.
In the past, these types of investigations have been prohibitively expensive due to analytical costs. Yet detailed information of this nature is required in order to design herbicide monitoring programs, identify stream courses at risk in terms of herbicide contamination, and evaluate the effects and benefits of alternative land and pesticide management alternatives.
Hydrologic flow data from streams, rivers and channels is very rarely normally distributed over time. In most single flow event cases, the rising limb of a flow hydrograph is more abrupt and rises faster than does the falling or recessional limb of the hydrograph. This type of data is said to be skewed, (i.e., not normally distributed) (Snedecor and Cochran, 1982). Special care must be taken when analyzing flow quantity and water quality data because most standard statistical methods require data to be distributed normally - particularly when determining yearly water quality mass loadings.
Often water quality parameters such as phosphorus, suspended solids and several organic compounds are highly skewed with respect to time and flow. Furthermore, generalities regarding the shape of the runoff curves for various pollutants cannot be made. Even persistent pesticides have seasonal peak periods corresponding to periods of application. Therefore, the whole year's runoff data cannot be lumped to estimate relationships between flow rate and concentration.
Past studies on pesticide residues in surface water generally centred around a single representative flow event and did not take into consideration the analysis of several flow events spaced over a season or year for the purpose of estimating seasonal or annual loads or parameter mass totals. To estimate yearly mass loadings from data that has been collected from selected rainfall-flow events requires special analysis of the data in order to accurately estimate water quality mass loadings for other non-sampled flow events. Traditionally, mass loads have been calculated using the sum of the product of flow and concentration. However, this can lead to biased estimates of mean loads, particularly if the flows and parameter concentrations vary by orders of magnitude. Bodo and Unny (1983), developed a method in which the data is broken into homogeneous, near normally distributed strata. In this method data segments which are similar in character are analyzed using standard techniques developed during PLUARG (IJC, 1977 and Tin, 1977). The result is a better estimate of means of water quality parameters for the determination of water quality mass loads.
SUMMARY AND CONCLUSION
A summary of the main results of monitoring and analyses conducted for the NSCP study are presented below. Conclusions are based on interpretation of significant trends and differences in the aforementioned data sets.
No significant differences were noted for crop type or yield when comparing test microbasins to control. There were also no significant differences observed with respect to applications of fertilizers and pesticides, including metolachlor. The only difference between test and control microbasins involved surface crop residue. As expected, preplant and post-harvest residue counts in the test (no-till) microbasins were significantly higher than in the control (conventional tillage). The higher surface residues are incorporated into the soil surface resulting in an overall improvement of soil quality (higher organic matter content, moisture retention capacity, etc.). In theory, this affects soil microbial populations and subsequently affects microbial degradation of agrochemicals such as metolachlor.
The physical and chemical properties of the soils of the test and control microbasins differed in a number of respects. The soils of the test microbasins (KTB1 and KTB2) 19 exhibited the expected characteristics of soil under conservation management (no-till) when examined in comparison to the soils from the control microbasins (KCB1 and KCB2). These characteristics include:
There were further differences evident in the chemistries of test and control soils, including lower levels of calcium in the test soils, which is also expected for no-till soils.
These general soil conditions for the test areas, especially elevated organic matter, provide an environment with a higher capacity for supporting soil micro-organisms than do the soil conditions found in the control areas. The test soils are therefore likely to have higher populations and activity levels of microbes. These micro-organisms, acting as decomposers, can in turn enhance the rate of degradation of agrochemicals such as metolachlor. Analyses of the soils from the Kettle Creek study area support the contention that metolachlor is degraded more readily in the test soils. Levels of metolachlor in soil samples from the test microbasins were much lower than those for control soil samples for almost every period of sampling. The difference in soil metolachlor concentration is most prominent during the periods closely following the application of the herbicide.
The main results from groundwater monitoring are as follows:
Surface water flow in the four study microbasins was quite variable over the NSCP study period. General findings concerning flow include the following:
Overall, further examination of flow data during periods of adequate precipitation is required to accurately assess any differences between test and control microbasins with respect to flow. Flow during the NSCP study was too infrequent for a thorough assessment.
Analyses of surface runoff samples collected from the four Kettle Creek microbasins during the study period reveal several differences in water quality between the test and control microbasins. The two main conclusions with respect to concentrations of water quality parameters are:
Microbasin water quality loads were also computed for six primary water quality parameters. The main findings are as follows:
Overall, both soil and water quality analysis appears to indicate that the conservation tillage implemented in the test microbasins is having a positive effect on environmental quality with respect to residues of the herbicide metolachlor. The main reason for the improvement is theorized to be the enrichment of soil organic matter and subsequent increase in populations of microorganisms (relying on organic matter to thrive) that act to degrade metolachlor and other compounds in the soil. Lower levels of soil metolachlor translate to lower water borne metolachlor delivered by runoff.
Recommendations for further study include:
Literature Review Pertaining to Buffer Strips
Helen Lammers-Helps and Douglas M. Robinson
Buffer strips historically have been used for the improvement of surface water runoff from logging and surface mine operations. Most recently they have been promoted in the U.S. and now Ontario for feedlot and cropland runoff.
Buffer stirps are bands of planted or indigenous vegetation situated downslope from cropland or animal production facilities to provide localized erosion protection and filter nutrients, sediment and other pollutants from agricultural runoff before they reach receiving waters. Buffer strips are also known as vegetative filter strips, grass filters, grass strips, riparian plantings and combinations thereof.
The two major removal mechanisms at work in vegetative filter strips are deposition and infiltration. As runoff enters the filter strip, its flow is retarded by the increased surface roughness and resistance of the vegetation. The decrease in velocity results in a decrease in the sediment transport capacity of the flow. If the resultant transport capacity is less than the inflow sediment load, sediment is deposited at the interface between the filter and the upslope area. The deposition wedge is typically 30-50 cm wide and occurs immediately upslope of the filter. Once this deposition zone fills up, the deposition front moves downslope in 50 cm intervals until the buffer strip is completely full. Sediment-bound pollutants are also deposited.
Soluble nutrients and some fine particles enter the soil profile with runoff infiltrating into the buffer strip. After entering the soil profile they can be removed by a combination of chemical, physical and biological processes. Mobile water soluble nutrients such as nitrate may leach through the soil profile.
Other mechanisms presumed to be at work are filtration of suspended solids, adsorption to plant and soil surfaces and absorption of soluble pollutants by plants. However, these mechanisms are not well understood at this time.
Research in the U.S. has shown that both tree and grass buffer strips can effectively remove coarse sediments if the runoff flow is shallow and uniform. Buffer strips are less efficient removal of the small particle sizes. Buffer strips do remove sediment-bound nutrients but with a slightly less efficiency than sediment.
Removal of soluble nutrients by buffer strips is highly variable. The concentration can actually increase due to the re-release of previously trapped nutrients as flow passes through the buffer strip.
There is very little information currently available with respect to the abilities of buffer strips to remove pesticides or pathogens from runoff water. Sediment-bound pesticides and pathogens are likely deposited to some extent but could be re-released at a later time. Some pathogens and pesticide would be removed with infiltrating water. More research is needed in this area.
Buffer strips have been credited with stabilizing streambanks and reducing in-channel erosion. Tillage implements are kept away from the watercourse edge, heavy equipment off the banks and vegetation roots stabilize the soil.
The buffer strip width required depends on many site, vegetation and climatic factors. In general terms, the width required increases as:
There are currently no simple design models available. According to James Krider, the National Environmental Engineer with the United States Department of Agriculture in Washington, D.C. a national handbook with design recommendations for site-specific conditions is in draft form and should be available later this year. The design criteria only consider sediment and surface flow. This may offer some guidance to extension personnel in Ontario.
There has not been any research to date on recommended species for buffer strips for Ontario conditions. Species recommended for grassed waterways could serve as a guide for grass buffer strips. There has not been any research yet on the most appropriate species selection for riparian tree plantings.
Since runoff must cross the buffer strip as sheet flow in order to be most effective, infield buffer strips or grasses waterways may be more appropriate in hilly areas where water tends to concentrate in natural drainageways prior to crossing the buffer strip.
No research to date has examined the effectiveness of buffer strips during the winter and early spring when vegetation is dormant. Runoff from snowmelt and winter/spring rains is very significant in Ontario.
Maintenance of a dense vegetation is essential to the long-term performance of the buffer strip. Mowing, fertilization and possibly herbicide application are necessary. Using the buffer strips as turn lanes or traffic lanes or for grazing livestock destroys the vegetation. Leaving a plough furrow parallel to the edge of the buffer strip results in water concentrating and flowing along the buffer strip edge and then crossing as concentrated flow at a low point. Furrows can be removed following ploughing with a light dishing. There is also a tendency for the strips to get narrower each year due to ploughing of the edge of the strip. This should be avoided.
More research is needed in several key areas in order to utilize buffer strips effectively in Ontario.