Submitted to: Meeting Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 10/15/2005
Publication Date: 10/15/2005
Citation: Mulla, D.J., Birr, A.S., Kitchen, N.R., David, M.B. 2005. Evaluating the effectiveness of agricultural management practices at reducing nutrient losses to surface waters. Meeting Proceedings. Interpretive Summary: Nonpoint source pollution is responsible for a large proportion of the impairment in surface water bodies throughout the Midwestern region of the U.S. Small amounts of sediment, phosphorus or nitrogen lost from a large number of agricultural fields can collect in lakes, rivers, or at the mouths of large watersheds to produce significant water quality impairment. As an example, one impairment given notable attention in recent years is a “hypoxia” (low oxygen) in the Gulf of Mexico. This condition is created when nutrients from the Mississippi River stimulate abnormally high algae growth in the Gulf. As the algae decays, it consumes oxygen in the water to very low levels and most marine wildlife can not survive. A “dead zone” in the late spring and early summer can occupy thousands of square miles. Agriculture has been targeted as a main contributor to this problem, and has been charged with making changes to help rectify it. A variety of best management practices (BMPs) have been developed to reduce the losses of sediment, phosphorus and nitrogen from agricultural fields. A BMP can be defined as a practice or combination of practices that is the most effective, technologically, and economically feasible means of preventing or reducing the pollutant generated by nonpoint sources to a level that meets water quality goals. Typically a BMP reduces the pollutant while maintaining similar crop or animal productivity as before the BMP was implemented. A number of different assessment studies have been tried in the upper Midwestern region to evaluate the impact of BMPs on water quality. The goal of this investigation was to review the success of those different BMP assessment studies, and to suggest how assessment of BMPs could be improved. We found that whether or not a BMP was deemed effective varied by geographic location, climate conditions, time of year the assessment was being conducted, size of the assessment area, other non-controlled management practices, and what assessment method was being used. Results from some of these studies indicated water quality was improved, however, a number of studies showed no changes (or a worsening) in water quality. The studies that failed to show improvements in water quality often attributed the failure to an insufficient long-term water quality monitoring record, the failure to implement BMPs that correct the most important sources of pollution, or the failure to implement BMPs in the most critical areas of the watershed. We conclude that more emphasis is needed on long-term watershed scale projects to evaluate impacts of BMPs on water quality, especially projects that involve paired watersheds (one watershed improved with BMPs, the other watershed left as a control). Also, more focus is needed to evaluate the effectiveness of BMPs targeted to portions of the landscape that contribute most to water quality degradation. The phrase “precision conservation” has recently been coined to refer to this targeted approach. As BMPs for high-priority water quality problems become more focused (i.e., they address the related soil and water processes, in the most vulnerable locations, and at the most vulnerable times of the year), water quality will improve. All U.S. citizens and wildlife in terrestrial and aquatic ecosystems will benefit as effective BMPs are employed to protect water resources.
Technical Abstract: Water quality impairments are widespread throughout the upper Mississippi River basin due in large part to agricultural production practices. Many agencies and organizations have worked with landowners to implement various agricultural management practices to reduce nutrient and sediment losses to streams and rivers. However, it has been difficult to document the effectiveness of these practices at the field, and more specifically, the watershed scales. This is due to five factors we’ve identified and discuss in this paper. One, climate is highly variable, which has a dominant effect on the transport of nutrients and sediment. Change in precipitation patterns lead to highly variable nutrient and sediment exports from one month to another, and one year to another. Therefore without long-term data, it is difficult to know when a change in nutrient and sediment export has occurred. Long-term data sets of sufficient monitoring intensity are generally not available, and short-term (1 to 5 year) data sets can give false impressions of the response. A commitment to long-term, baseline, monitoring can help differentiate between climate effects and documented management changes. Two, our inability to measure reduction of nutrient and sediment losses in conjunction with specific management practices once smaller watersheds are aggregated into larger watersheds. Nutrient and sediment concentrations can be highly variable in surface runoff and tile drainage from one watershed to the next, and intensive measurements are needed to obtain accurate loss rates. Most monitoring studies do not have sufficient intensity for long enough time periods and in enough locations to allow statistically significant differences to be determined. Three, long lag times are common in response to changes in management. Because of large and dynamic pools of soil nitrogen and phosphorus, agricultural landscapes are buffered and therefore response to implementation of altered management practices can take many years to elicit a change in water quality. In addition, stream or river response may be obscured by previous accumulation and transport of in-stream sediments and nutrients, that mask reduced export from fields. Four, implementation of many improved management practices at watershed scales has been sparse. Most management programs at the watershed only involve some of the fields and often do not target the most critical areas. Given the previous points, this can greatly reduce our ability to document change. And five, modeling limitations make projections uncertain. Limitations include uncertainty in many parameters (e.g., soil hydraulic properties, denitrification, mineralization rates, biological N fixation), incomplete representations of field and watershed processes, and limited data for validation. Future assessment of agricultural watersheds will need to address these five issues in order to understand the potential benefit current and future management practices will have on water quality.