Printer Friendly

Land Management at the Major Watershed--Agroecoregion Intersection.

ABSTRACT: The watershed natural resources management framework is prevalent today because land use in watersheds is presumed to he reflected in receiving stream water quality. However, landscape characteristics affecting soil erosion and water quality (e.g., precipitation, geomorphology, slope, soil internal drainage, cropping system) often vary significantly within a single large watershed ([greater than] 200,000 ha). A uniform watershed best management practice would not account for this variability and would not be satisfactory for soil conservation, water quality, or socioeconomic returns. It is highly unlikely that stream water quality monitoring will take place on enough small streams within a large watershed to capture the landscape variation. We have developed "agroecoregions" to quantify this variation, based on empirical data from the Minnesota River Basin (MRB). This approach is needed to help target cleanup efforts to the most sensitive soils and landscapes within the most critical watersheds. Our work shows that soil erodibility index variability and stream biotic habitat scores were better represented by agroecoregions than by watersheds. Stakeholder characterization and economic analysis reveal a large variance in attitudes and beliefs about pollution issues and mitigation costs in the MRB, due in part to problems of scale perception (e.g., entire basin, major watershed, county, city, farm). We suggest that watershed management in highly agricultural watersheds will be most effective when hydrologic watersheds are used as a framework that is complemented by agroecoregions to identify and target regions where specific combinations of best management practices for agricultural sediment and phosphorus abatement are most appropriate.

Keywords: Ecoregions, Minnesota River, nonpoint source pollution, water quality, watershed management

Large point sources of river pollution have become regulated in the past three decades, and state governments now are placing increased emphasis on reducing nonpoint source (NPS) pollution. This situation has arisen in part because attaining water quality goals through further reductions in point source pollution is often cost prohibitive. To deal effectively with the large spatial domain inherent in NPS pollution, state agencies have developed management strategies based on various units such as ecoregions or watersheds (Brown and Marshall 1996; NRC 1999). It has been argued, however, that using either of these frameworks can fall short of management goals if implemented improperly (Omernik and Bailey 1997). Omernik and Bailey asserted that ecoregions and watersheds can serve as complementary management and/or research units if used in the proper context. For example, one would expect differences in water quality to be less for two watersheds in the same ecoregion than between two watersheds in different eco regions.

Scientific justification for using ecoregions to structure investigative and/or managerial goals lies in the fact that broad spatial patterns can strongly influence ecological processes (Turner 1989). For example, it has been demonstrated that riparian forests have the potential to reduce nutrient (nitrogen and phosphorus) flux from cropland to streams (Schlosser and Karr 1981; Peterjohn and Correll 1984). Hence, the combination of riparian forest with agricultural systems can reduce NPS pollution.

Several ecologists have argued that there is no single spatial and/or temporal scale appropriate for managing the wide range of properties in an ecosystem (Christensen et al. 1996; Perry and Vanderklein 1996). Ecosystems and the scale of management goals are often incongruent. For example, in a large agricultural basin one might need to develop management strategies for soil erosion at the scale of single farms to limit the effects of intense rainstorms when crop cover is sparse. Consider a fisheries manager who only has jurisdictional authority in small catch-and-release stream sections within the agricultural basin. If the sediments/nutrients impacting these sections originate upstream of the jurisdictional area, how can the fisheries manager address this situation? We propose that by knowing the water quality impacts of different land organizational units (i.e., watersheds and ecoregions), managers in large, complicated, multiple land use basins may be able to utilize their resources in a more effective m anner (e.g., targeting resources for more effective cleanup).

Here we demonstrate ways in which agricultural ecoregions and major watersheds serve as complementary units for information gathering and management in large agricultural basins. First, we discuss the development of agricultural ecoregions; then we demonstrate how soil erosion and stream macroinvertebrate metrics vary within the context of major watersheds and agricultural ecoregions of the Minnesota River Basin (MRB). Finally, we describe social, cultural, and economic factors/issues that managers need to consider when integrating agricultural ecoregions into their watershed management framework.

Study Area

The Minnesota River, southern and western Minnesota, (Figure 1) has 92% of its 38,419 [km.sup.2] basin area under intensive agriculture, and contains high levels of pollution (MPCA 1994). Significant point source and NPS loads of sediment, nitrogen, phosphorus (P), pesticides, and pathogens are a concern. For example, from 1968-1994 turbidity at the mouth of the Minnesota River exceeded 25 NTU 33% of the time, and total P exceeded 250 [micro]g [L.sup.-1] during 73% of the May-September season (Mulla and Mallawatantri 1997). Runoff from row-crop agriculture (corn/soybean rotation) and livestock feedlots is thought by many to be the main source of pollution to the river, but sewage treatment plants, septic tanks/leaching fields, and streambank erosion also are significant contributors.

With headwaters in eastern South Dakota, the Minnesota River flows 539 km to join the Mississippi River in the Minneapolis-St. Paul metropolitan area. Total elevation change is 84 m, and the average slope is 1.5% (MPCA 1997). The Minnesota River increases flow in the Mississippi River by 50% at its confluence from 225 [m.sup.3] [s.sup.-1] to 450 [m.sup.3] [s.sup.-1] mean annual flow (USGS 1996). Precipitation varies significantly across the Minnesota River basin, from 560 mm [yr.sup.-1] in the west to 860 mm [yr.sup.-1] in the east, and this results in 100 mm of annual runoff entering rivers in the west and 200 mm in the east (MPCA 1997). Nearly 41% of the soils in the basin have poor soil internal drainage, and many areas have significant surface and subsurface drainage systems installed. MRB soils have high organic matter content, often exceeding 4% (MPCA 1997).

In 1989, the Minnesota River Assessment Program (MNRAP) was initiated by government agencies (federal, state, and local) and non governmental organizations (e.g., business and environmental groups) to evaluate water quality in the MRB, set water quality goals, and develop cleanup plans (MPCA 1994). Land use and water quality databases were assembled, and water quality studies were conducted to evaluate the river and its tributaries (e.g., Payne 1994). Significant public interest in cleaning up the Minnesota River resulted from the MNRAP study. An ongoing implementation phase (Minnesota River Improvement Program (MNRIP)) includes federal, state, and local agencies.

In response to these actions, the 37 counties in the MRB formed a Joint Powers Board to coordinate information gathering and cleanup efforts. This action was partially due to local mistrust of potential "top-down" directives from state agencies. The MRB Joint Powers Board created watershed action groups for the basin's 12 major watersheds to develop comprehensive watershed management and implementation plans that include best management practices (BMPs) for soil conservation, improvement of septic tank and sewage treatment facilities, more effective use of pesticides and fertilizers, and the restoration of riparian and wetland areas. It is recognized, however, that these plans are based on inadequate data and an incomplete research base.

An overall goal for cleanup efforts in the MRB is to improve water quality in the Minnesota River system, the Mississippi River into which the Minnesota River discharges, and the Gulf of Mexico. Routine water quality monitoring data (total suspended solids, total P, nitrate) for major tributaries and Minnesota River mainstem sites from 1968-1994 were gathered from the U.S. Geological Survey, the Twin Cities Metropolitan Council for Environmental Services, and the U.S. Environmental Protection Agency (US EPA). Discharge measurements from U.S. Geological Survey gaging stations were used in conjunction with the FLUX model (Walker 1987) to calculate sediment and nutrient loads for the 12 major watersheds in the MRB (Mulla and Mallawatantri 1997). These calculations led to the conclusion that most of the sediment and nutrient loading at the mouth of the River comes from only three major watersheds in the eastern section of the MRB (Table 1). The Blue Earth River watershed, the Le Sueur River watershed, and the Low er Minnesota River watershed together cover only 26% of the MRB but they contribute nearly two-thirds of the total suspended sediment, total phosphorus, and nitrate load. This information is useful for prioritizing the 12 watersheds in terms of their sediment and phosphorus loads. Below, we show that within these three watersheds there are specific soil, landscape, and climate factors which cause more NPS pollution from these watersheds than all the others.

Development of Agroecoregions

One of the goals of our research is to determine the most appropriate spatial organizational units for managing land use in highly agricultural basins. The MRB Joint Powers Board and several state agencies in Minnesota have endorsed the watershed concept as the appropriate unit for management; however, other state agencies have supported management based on more landscape oriented criteria, including ecoregions. Because land use in the MRB is predominantly agricultural, we chose to create an overall classification scheme that reflects both the factors influencing rowcrop productivity and the factors influencing NPS pollution potential (hence, "agroecoregions"). The delineation of agroecoregions was based on data from the State Soil Atlas (16 ha minimum cell size) relating to classes of precipitation, soil geomorphology, slope steepness, soil internal drainage, and crop productivity (Table 2).

Precipitation classes. The basin was divided into "wetter" and "dryer" components based on long term precipitation. The eastern portion of the basin, receiving greater than 680 mm [yr.sup.-1], was classified as wetter, and the western portion, receiving less than 680 mm [yr.sup.-1], was classified as dryer. This demarcation separates soils that have calcareous surface horizons in the west (due to lower precipitation and leaching) from soils that are non calcareous in the east (due to higher precipitation and leaching).

Soil geomorphic classes. The Minnesota River Basin has 27 distinct geomorphic groups. These groups are partly differentiated on the basis of five types of parent material: glacial moraines, glacial till, lacustrine sediments, glacial outwash, and alluvial deposits. Each of these parent materials is associated with distinct patterns of relief; the geomorphic classes are based on both parent material and relief.

Slope steepness classes. Landscapes were divided into seven slope steepness classes (Table 2). A greater potential for water erosion and runoff is generally associated with greater land steepness. Land that has high potential for erosion has the potential to degrade water quality from sediments and phosphorus.

Internal drainage classes. MRB soils are divided into three natural and artificial drainage classes: 1.) poorly drained soils, 2.) poorly drained soils in which both surface and subsurface drainage was improved by tiling, and 3.) well drained soils. The poorly drained soils have slow infiltration and a tendency for surface water ponding, which are impediments to farming. Approximately 41% of the MRB consists of poorly drained soils; most of these soils have been riled. Subsurface tile drainage promotes the transport of nitrate to streams, and surface tile inlets promote the transport of sediments, phosphorus, and human/animal wastes to streams.

Crop productivity classes. Low, medium, and high crop productivity classes were made using precipitation (range: 560-860 mm [yr.sup.-1]), growing degree days (range: less than 1500-2750), slope steepness, soil texture, soil drainage, and rooting depth (Rust et al. 1984). Crop productivity increases from the western to the eastern part of the MRB, following increasing gradients of precipitation and growing degree days.

We used a GIS to develop maps of these five classes, which were then overlain and grouped into 13 agroecoregions. Each of the 12 major watersheds includes at least two agroecoregions, some have as many as six agroecoregions (Figure 1). For example, the Chippewa River watershed contains six agroecoregions, whose properties range from the steep, well drained loamy soils of the Steep Dryer Moraine agroecoregion to the flat, poorly drained, silty-clay soils of the Dryer Clays and Silts agroecoregion (MPCA 1997). The Alluvium and Outwash, Steep Streambanks, Steep Valley Walls, and Steeper Till agroecoregions are also present. Each of the 12 major watersheds, therefore, contains significant physiographic variability that influences both the potential for NPS pollution and the potential for adoption of BMPs (due to the varying erosion control techniques needed).

Data Collection

The ability of agroecoregions to improve our understanding of biophysical and socioeconomic processes in the MRB was evaluated by collecting additional data. We examined these processes in the context of three land classifications: 1.) major watersheds, 2.) agroecoregions, and 3.) the major watershed--agroecoregion "intersection," which is defined as the area where an agroecoregion falls within a major watershed. Soil erosion potential across the MRB was calculated using the Universal Soil Loss Equation (USLE). We also collected stream macroinvertebrate samples and assessed stream habitat in the eastern half of the MRB to determine which of the three land classifications best described the variance of these variables. Finally, we assembled available data and collected new data for an economic assessment and sociocultural characterization of stakeholders in the MRB.

Soil erosion potential. The Erodibility Index (EI) for the MRB was calculated using the USLE for map units in the State Soil Geographic Database. This database contains soil associations derived from generalized soil survey data. Potential soil erosion depends on rainfall erosivity, soil erodibility, and slope steepness and length. Slope length (L) factors were calculated using average slope lengths for each region from NRCS NRI data. Land cover and land management were not considered in these erosion estimates and so they represent bare soil (i.e., a worst case scenario). This means that the cover/management (C) factor in the USLE and the conservation practices (P) factor were both set to unity. Mean EI values were area weighted to compensate for the different sizes of the watersheds and agroecoregions.

Stream habitat and macroinvertebrates. We collected data on stream benthic macroinvertebrates in three major watersheds, Blue Earth, Le Sueur, and Lower Minnesota, that span two agroecoregions (Rolling Moraine, Wetter Clays and Silts) to determine which spatial unit best predicted macroinvertebrate habitat and diversity. These two agroecoregions are the dominant units by area in agricultural portions of each watershed. The Steep Wetter Moraine was not studied because it is in an area of rapid urban growth. Hence we examined three major watersheds, two agroecoregions, and their six intersections. All second and third order streams were identified, and 68 sites were chosen in a stratified random design, with 12 sites in each of the six intersections (except for the Blue Earth -- Rolling Moraine intersection which only had eight sites).

Aquatic macroinvertebrate sampling took place during the summer of 1998 under low flow conditions. We sampled macroinvertebrates and habitat data in a discrete 100 m segment of each stream channel following the US EPA Low Gradient Stream Method (US EPA 1997). Three productive macroinvertebrate habitats were sampled: banks, woody snags, and submerged macrophytes. The sampling effort was distributed proportionately based on habitat occurrence. Macroinvertebrates were sampled with 20 "jabs" using a D-frame dip net: a jab consisted of an initial thrust of the net into the target habitat for a distance of approximately 1 m, followed by 2-3 sweeps of the same area to collect dislodged organisms. Net contents from each jab were composited; the sample was sieved, rinsed to remove fine sediment, and preserved in 90% ethanol. Organisms were identified to genus. The habitat assessment consisted of visual scoring of seven habitat quality parameters including general characteristics of the in-stream, stream bank, and rip arian zone areas.

Social and economic parameters. Major stakeholder groups in the MRB were characterized by telephone surveys and personal interviews of parties interested in the MRB during the summer of 1998. Five question segments were used for extensive telephone interviewing, including style of decision-making, driving values, focus of attention, management preferences, and scale of interest. These data were evaluated to categorize major stakeholder groups and to postulate their views on MRB issues. In addition, a sample of farmers and the staff of government agencies, environmental groups, and industrial organizations were sent a mail questionnaire to determine their ranking of conservation practices and associated costs.


Soil erosion potential across the Minnesota River Basin. Erodibility index (EI) values have a greater range for comparisons between the agroecoregions than for comparisons between the major watersheds. Watershed values ranged from 2.2-6.1 t [ac.sup.-1] [yr.sup.-1], while agroecoregion values ranged from 1.9-17.2 t [ac.sup.-1] [yr.sup.-1] (Table 3). Paired t-test comparisons of EI values across the 12 watersheds resulted in 23 (34%) pairs with significant differences (ANOVA 0.05 level) and 43 (66%) pairs with no significant differences. A similar paired t-test comparison for the 13 agroecoregions resulted in 57 (73%) pairs with significant differences and 21 (27%) pairs with no significant differences. Hence agroecoregions offer a much higher percent of unique EI values than do watersheds. Although the agroecoregions do not always differ significantly in EI, they are different from each other when considering such other characteristics as soil drainage and productivity.

Differences in EI values for major watershed--agroecoregion intersections indicate the ability of agroecoregions to capture the variability of erosion potential across the landscape (Table 4). For example, for the six agroecoregions in the Lower Minnesota River watershed, the 15 comparisons yielded 10 significant differences (ANOVA 0.05 level). In addition, the three largest area agroecoregions in the Lower Minnesota watershed (Rolling Moraine at 21%, Steep Wetter Moraine at 28%, and Wetter Clays and Silts at 32%) were all significantly different from one another with respect to EI values.

In the relatively dry western MRB, the Chippewa River watershed has six agroecoregions (Table 4). Eight of fifteen inter-agroecoregion EI comparisons were different (ANOVA 0.05 level). The largest Chippewa River watershed agroecoregion (Steep Dryer Moraine at 45%) has an EI value that is significantly different than the next three largest agroecoregions (Alluvium and Outwash at 18%, Dryer Clays and Silts at 20%, and Steeper Till at 15%). The large range in EI values for the Stream Banks agroecoregion is most likely the cause for the lack of statistically significant differences between that area and the other agroecoregions.

Stream macroinvertebrate variability. Stream habitat assessment scores varied considerably among the spatial units (Table 5). For example, major watershed scores varied from 80-88, agroecoregion scores varied from 79-89, and major watershed--agroecoregion intersection scores varied from 65-99. Standard deviations also varied according to spatial units. Coefficients of variation ranged from 31--40% for major watersheds (highest: Lower Minnesota), 31-39% for agroecoregions (highest: Wetter Clays and Silts), and 25-49% for their intersection (highest: Lower Minnesota - Wetter Clays and Silts, lowest: Lower Minnesota - Rolling Moraine).

A two-way ANOVA was used to assess the influence of major watersheds, agroecoregions, and the specific watershed -- agroecoregion interaction (i.e., intersection) on stream habitat assessment scores in the eastern MRB (Table 5). The ability of a given spatial unit classification to explain macroinvertebrare variability is indicated by the level of significance. Habitat scores differed more widely between agroecoregions (p [less than] 0.05) than between major watersheds (p [less than] 0.10). Habitat scores among major watershed agroecoregion intersections, however, were most distinct (p [less than] 0.01).

In contrast, significant correlations were not found between 1.) the biotic index or stream macroinvertebrate population variables, and 2.) any of the three land unit types (watershed, agroecoregion, their intersection). We did find a significant relationship between habitat assessment scores and the corresponding biotic integrity index (Figure 2). A lower biotic index score indicates a higher quality species assemblage (e.g., stoneflies, mayflies); a higher biotic index indicates a more pollution tolerant assemblage (e.g., chironomid larvae). A strong relationship therefore exists between the agroecoregion-major watershed intersection and stream habitat quality, and in turn stream habitat is strongly related to the stream macroinvertebrate index.

Stakeholder characterization and the economic considerations. We have demon-strated a connection between biophysical variables and land area designations. However, using these results to recommend agricultural best management practices must take into account the many social, cultural, and economic factors influencing agricultural pollution. Two important factors include the opinions and actions of the major stakeholder groups, and the public and private costs associated with reducing agricultural NPS pollution.

Analysis of the telephone and personal interviews conducted by the social scientists in our group began by grouping similar responses for a given question, regardless of the person's group affiliation. This approach allowed us to define three categories of stakeholder groups based on response patterns: non government utilitarian (e.g., corn, soybean, and beet growers; pork, poultry, and beef producers), government (e.g., US EPA, Minnesota Pollution Control Agency, Minnesota Department of Natural Resources, Minnesota Board of Water and Soil Resources, Minnesota River Joint Powers Board, Redwood-Cottonwood Rivers Control Area, Blue Earth River Basin Initiative), and non government ecological (e.g., Clean Up our River Environment, Concerned Citizens for the Minnesota River, Friends of the Minnesota River Valley, Friends of Big Stone Lake, Audubon Society).

Examples of the attributes and opinions that led us to identify these stakeholder groups are given in Table 6. For example, the scales of interest for non-government utilitarian organizations are local (e.g., farms) and in the present (e.g., this year's crop), being associated with individual farming operations. Government organizations are interested in joint areas (e.g., watersheds) and the near future (decades). Non government ecological organizations, however, are interested in all spatial scales and ecologically associated temporal scales (decades to centuries). It is apparent that Minnesota River water quality issues are culturally constructed in significantly different ways and at different scales according to different stakeholder groups. It is important to note that recent government is based on collaborative partnerships between agencies, whereas government previously had a generally bureaucratic method of organization (e.g., MPCA 1994).

The opinions and beliefs of members of these different groups have the potential to play a significant role in how these groups interact in regulatory and economic matters. For example, farmers were surveyed by mail regarding their production practices and their opinions about alternative policies for reducing phosphorus loading (McCann and Easter 1999a). The policies most acceptable to them were: 1.) a requirement for conservation tillage on highly erodible land, and 2.) educational programs on best management practices. The two least acceptable policies were taxes on phosphate fertilizers and manure. A survey of persons from government agencies, farm organizations, and environmental organizations indicated that expected farmer resistance was a more important determinant of agency preference for policies than either farmer costs or administrative costs. A requirement for conservation tillage on highly erodible land also was the preferred policy of government agency personnel in the survey, and a tax on manur e was the least preferred policy. A comparison of farmers from the eastern and western halves of the MRB resulted in a significant difference in acceptability of farming practices only for the category "conservation tillage on all land." In this case, farmers in the eastern MRB found conservation tillage on all land more acceptable, suggesting that for many of the approaches to reducing P loading, farmer locations in the MRB does not have a big effect on how options to reduce P pollution are ranked.

Interviews with agency staff were conducted to estimate transaction costs associated with four policies to reduce agricultural phosphorus pollution in the Minnesota River by 40% (McCann and Easter 1999b). The tax on phosphate fertilizers had the lowest transaction costs ($0.94 million), followed by educational programs on best management practices ($3.11 million), the requirement for conservation tillage on all cropped land ($7.85 million), and expansion of the permanent conservation easement program ($9.37 million). These results suggest that if either of the latter two strategies are used, they will need to be targeted at the most erosion prone regions to hold down their high costs.


Soil erodibility. Based on our soil erodibility estimates, the following major watershed-agroecoregion intersections have the greatest potential for soil erosion (Figure 1): Steep Wetter Moraine, Rolling Moraine, and Steep Valley Walls in the Lower Minnesota watershed; Rolling Moraine in the Le Sueur watershed; Steep Stream Banks and Rolling Moraine in the Blue Earth watershed; Steep Dryer Moraine in the Chippewa and Pomme de Terre watersheds; Coteau in the Lac Qui Parle, Hawk Creek - Yellow Medicine, Redwood, and Cottonwood watersheds; and Wetter Blue Earth Till, and Steep Valley Walls in the Middle Minnesota watershed. These areas generate most of the pollution. Managers wanting to control "hot spots" of soil erosion in the MRB would achieve better results by targeting BMPs to these major watershed--agroecoregion intersections. Mitigation of stream sediment transport (and associated phosphorus) transported from these regions may also benefit from such BMP targeting. Specific BMPs can be developed along with a complementary educational program to enhance implementation.

The group of intersection areas above pertain primarily to soil and phosphorus erosion potentials. Nitrogen pollution (in the form of nitrate losses in tile drains), however, can be significant in areas with low soil erodibility values and poor soil drainage characteristics. Although we have nor addressed nitrate transport in this study, BMPs specific to nitrate loss in tile drain systems also might be better targeted to major watershed -- agroecoregion intersections that have these soil conditions (e.g., Wetter Clays and Silts).

Stream invertebrates. Many investigators (Larson et al. 1986, 1988 [Ohio]; Hughes et al. 1987 [Oregon]; Rohm et al. 1987 [Arkansas]; Whittier et al. 1988 [Oregon]; Lyons 1989 [Wisconsin]; Matthews et al. 1992 [Arkansas]; Poff and Allan 1995 [Minnesota and Wisconsin]) have used watersheds or ecoregions to classify aquatic characteristics (e.g., water chemistry, fish assemblages, physical habitat). We focused our efforts at the unique intersection of these two landscape classes, which allowed us to better explain stream macroinvertebrate habitat variability (Table 5). A strong relationship between habitat and the stream macroinvertebrate community also was found (Figure 2), but the major watershed-agroecoregion intersection did not predict the variance in stream macroinvertebrate communities better than the individual metrics.

Perhaps two spatial scales need to be addressed when assessing factors that affect stream habitat. "Macro-habitat" variables, such as stream geomorphology, may be more effectively described at a coarser, intersection scale, while "micro-habitat" variables, such as woody debris and stream pebble characteristics, may be more effectively described at a finer, local scale. These two scales are not necessarily independent, but may be connected at different temporal scales. For example, woody debris may serve as a refuge for macroinvertebrates (local, short term processes), but woody debris may be determined in part by geomorphological channel shifts (intersection, long term processes).

Socioeconomic implications. Because our work indicated that sediment and phosphorus loads (Table 1) and soil erosion potential (Tables 3 and 4) in the MRB are not uniform across the landscape, application of a uniform policy for land management across the entire MRB would not be effective. It has been argued that greater environmental benefits may result from targeting BMPs where the greatest phosphorus loss and erosion potentials are located in the United States (Sharpley et al. 1994). However, given the large number of farming operations in highly agricultural basins, governmental management of BMPs on an individual farm basis would be prohibitive due to excessive administrative and personnel costs. Voluntary adoption of BMPs, therefore, is crucial if environmental improvements are to be significant. This could be achieved if agency grant assistance programs (e.g., USDA NRCS Environmental Quality Incentives Program, Conservation Reserve Program, etc.) could be targeted to the most sensitive agroecoregions w ithin the most polluted watersheds. In the MRB, this is already starting to happen: the Redwood - Cottonwood Rivers Control Area organization has divided the Cottonwood watershed into three priority areas which correspond to agroecoregion boundaries. Goals and implementation activities are different in each priority area.

One must also consider the impact of mitigation efforts on crop yield: although a certain BMP might result in significant environmental improvements, it is unlikely that the practice will be adopted if crop returns per acre are reduced significantly. Additional concern is that BMPs for phosphorus reduction by erosion control are not necessarily compatible with BMPs for nitrogen reductions through subsurface drainage management. These qualifications must be considered. when formulating government policy to control sediment, phosphorus, and other pollutants.


Agroecoregion delineations represent a step beyond traditional ecoregions because they focus on factors affecting both agricultural production and soil loss. We have demonstrated the utility of agroecoregions (and their intersection with watersheds) for describing soil erosion and stream habitat/biota. These results suggest that agroecoregion-based management practices may provide more specific recommendations for the control of erosion and the enhancement of stream habitat/biota.

However, like many other states, Minnesota already has adopted the major watershed as the level of management and organization. There is an established institutional framework for watershed management already in place for the MRB. It has also been argued that managing at the watershed level is most politically suitable because it is easily understood by the public and has a long history with regard to environmental policy (Ruhl 1999). Restructuring regulatory organizations to adopt an agroecoregion only management framework would not be cost-effective, and perhaps not biophysically] effective either. Still, it is possible to directly integrate agroecoregions into management plans for the purpose of targeting subregions of watersheds where BMPs for agricultural sediment and phosphorus can be implemented with the least resistance from farmers. By targeting BMPs to specific agroecoregions within major watersheds (i.e., their intersection), we anticipate that local managerial resources will be used more effective ly and costs of reducing sediment and P pollution will be minimized. We are trying to develop a framework for broad targeting of federal, state, and local nonpoint pollution control programs. This framework can be used to ensure that money flows to the most critical areas.

Additional criteria can be used to describe which farm or which minor watershed in the targeted intersection has highest cleanup priority. This targeted approach of BMP adoption and monitoring is consistent with the present government approach of agency partnerships and seems to have the greatest potential for public acceptance among available choices.

Lorin K. Hatch is with the Water Resources Center and the Department of Forest Resources at the University of Minnesota. Ananda Mallawatantri, Dan Wheeler, and David Mulla are in the Department of Soil, Water, and Climate at the University of Minnesota, Anne Gleason and James Perry are in the Department of Forest Resources at the University of Minnesota. K. William Easter is in the Department of Applied Economics at the University of Minnesota. Richard Smith and Luther Gerlach are in the Department of Anthropology at the University of Minnesota. Patrick Brezonik is in the Water Resources Center and the Department of Civil Engineering at the University of Minnesota.


This work was supported by the US EPA STAR Grant Water and Watersheds Program (R825290). Although the information in this document has been funded wholly or in pare by the Environmental Protection Agency, it may flat necessarily reflect the views of that agency and no official endorsement should be inferred. Development of the agroeoregion framework was supported by funding from the Minnesota Department of Agriculture, Sc. Paul, Minnesota. Financial assistance from the Minnesota Department of Agriculture (Sc. Paul, MN) was used to develop agroecoregions in the Minnesota River Basin.


(1.) See MPCA (1997) for a more detailed discussion.


Brown, R. and K. Marshall. 1996. Ecosystem management in state governments. Ecological Applications 6(3):721-723.

Christensen, N., A. Bartuska, J. Brown, S. Carpenter, C. D'Antonio, R. Francis, J. Franklin, J. MacMahon, R. Noss, D. Parsons, C. Peterson, M. Turner, and R. Woodmansee. 1996. The report of the Ecological Society of America Committee on the scientific basis for ecosystem management. Ecological Applications 6(3):665-691.

Hughes, R., E. Rexstad, and C. Bond. 1987. The relationship of aquatic ecoregions, river basins, and physiographic provincea to ichthyogeographic regions of Oregon. Copeia 161(2):423-432.

Larsen, D., D. Dudley, and R. Hughes. 1988. A regional approach for assessing attainable surface water quality: an Ohio case study. Journal of Soil and Water Conservation Society 43(2):171-176.

Larsen, D., J. Omernik, R. Hughes, C. Robin, T. Whittier, A. Kinney, A. Gallant, and D. Dudley. 1986. The correspondence between spatial patterns in fish assemblages in Ohio streams and aquatic ecoregions. Environmental Management 10:815-528.

Lyons, J. 1989. Correspondence between the distribution of fish assemblages in Wisconsin streams and Omernik's ecoregions. America Midland Naturalist 122 (1):163-182.

Matthews, W., D. Hough, and H. Robison. 1992. Similarities in fish distribution and water quality patterns in streams of Arkansas: congruence of multivariate analyses. Copeia 166(2):296-305.

McCann, L. and K. Easter. 1999a. Differences between farmer and agency attitudes regarding policies to reduce phosphorus pollution in the Minnesota River Basin. Review of Agricultural Economies 21 (1):189-207.

McCann, L., and K. Easter. 1999b. Transaction costs of policies to reduce agricultural phosphorus pollution in the Minnesota River. Land Economies 75 (3):402-414.

Minnesota Pollution Control Agency (MOCA) 1994. Minnesota River assessment project. Vols. I-III. St. Paul: Minnesota Pollution Control Agency.

Minnesota Pollution Control Agency (MPCA). 1997. Minnesota River Basin Information Document. Norman Senjem (ed). St. Paul: Minnesota Pollution Control Agency.

Mulla, D. and A. Mallawatantri. 1997. Minnesota River basin water quality overview. St. Paul: University of Minnesota. Minnesota Extension Service Bull. FO-7079-E.

National Research Council (NRC). 1999. New strategies for America's watersheds. Washington, D.C.: National Academy Press.

Omernik, J. and R. Bailey. 1997. Distinguishing between watersheds and ecoregions. Journal of American Water Resources Association 33 (5):935-949.

Payne, G. 1994. Sources and transport of sediment, nutrients, and oxygen-demanding substances in the Minnesota River basin, 1989-1992. Mounds View: U.S. Geological Survey Water Resources Investigations Report 93-4232.

Perry, J. and E. Vanderklein. 1996. Water quality: management of a natural resource. Cambridge: Blackwell Scientific.

Peterjohn, W. and D. Correll. 1984. Nutrient dynamics in an agricultural watershed: observations on the role of a riparian forest. Ecology 65:1466-1475.

Poff, N. and J. Allan. 1995. Functional organization of stream fish assemblages in relation to hydrologic variability. Ecology 76 (2):606-627.

Rohm, C., J. Geise, and C. Bennett. 1987. Evaluation of an aquatic ecoregion classification of streams in Arkansas. Journal of Freshwater Ecology 4:127-140.

Ruhl, J. 1999. The (political) science of watershed management in the ecosystem age. Journal of American Water Resources Association. 35 (3):519-526.

Rust, R, L. Hanson, and J. Anderson, 1984. Productivity factors and crop equivalent ratings for soils of Minnesota. St. Paul: University of Minnesota. Minnesota Extension Service Pub. No. AG-BU-2199.

Schlosser, I. And J. Karr. 1981. Riparian vegetation and channel morphology impact on spatial patterns of water quality in agricultural watersheds. Environmental Management 5:233-243.

Sharpley, A., S. Chapra, R Wedepohl, 3. Sims, T. Daniel, and K. Reddy. 1994. Managing agricultural phosphorus for protection of surface waters: issues and options. Journal of Environmental Quality 23:437-451.

Turner, M. 1989. Landscape ecology: the effect of pattern on process. Annual Review of Ecological Systematics 20:171-197.

United States Environmental Protection Agency (US EPA). 1997. Field and laboratory methods for macroinvertebrate and habitat assessment of low gradient, nontidal Streams. Wheeling: Mid-Atlantic Coastal Streams Workgroup. Environmental Services Division, Region 3.

United States Geological Survey (USGS). 1996. Contaminants in the Mississippi River, 1987-1992. Washington, D.C.: U.S. Geological Survey Circular 1133.

Walker, W. 1987. Empirical methods for predicting eutrophication in impoundments; report 4, phase III: applications manual. Vicksburg: U.S. Army Corps of Engineers. Technical Report No. E-81-9.

Whittier, T., B.. Hughes, and D. Larsen. 1988. The correspondence between ecoregions and spatial patterns in stream ecosystems in Oregon. Canadian Journal of Fisheries and Aquatic Science 45 (7):1264-1278.
                    Suspended sediment, phosphorus, and
                   nitrogen load contributions from the
              major watersheds of the Minnesota River Basin.
                                  Suspended      Total Nitrate +
                              MRB    Solids Phosphorus Nitrite-N
Major Watershed            % area         %          %         %
Upper Minnesota                12       2.1        4.8       0.4
Pomme de Terre                  5       0.7        1.3       0.1
Lac Qui Parle                   7       0.8        1.9       0.1
Hawk Creek-Yellow Medicine     12       4.3        4.5       2.5
Chippewa                       12       4.0        5.6       1.6
Redwood                         4       1.0        2.1       1.9
Middle Minnesota                8       8.6        5.2       9.9
Cottonwood                      8       8.1        5.2       9.4
Watonwan                        5       4.3        5.4      11.7
Blue Earth                      9      19.0       14.6      25.6
Le Sueur                        6      20.7       16.7      25.6
Lower Minnesota                11      26.4       32.5      12.0
(*.)Source: Mulla and Mallawatantri (1997).
Values are given as a percent of
the total load at the mouth of
the Minnesota River (values
rounded to nearest tenth percent).
         Range of areas for various land classification coverages
          across agroecoregions in the Minnesota River Basin, MN.
Slope Steepness
    %                        Internal Drainage           Crop Productivity
Class             Range               Class        Range       Class
  0-2           3.0-62.2 [*]           Poor    0.6-46.0%         Low
  2-6           3.4-81.9        Poor, Tiled    0.9-51.3%      Medium
 2-12           0.1-41.0       Well Drained    8.0-91.0%        High
 6-12             0-51.7              Water       0-2.6%
 6-45             0-80.2
12-45             1-16.6
Water              0-2.6
Slope Steepness
Class               Range
  0-2           0.1-34.5%
  2-6           3.2-97.4%
 2-12           0.2-92.9%
(*.)For example, with respect to the Slope Steepness
0-2% Class, coverages range from 3.0-62.2% of the land
within each of the 13 agroecoregions.
               Erodibility index according to watershed and
                agroecoregion in the Minnesota River Basin.
                    Erodibility Index
                      Weighted Mean
                      t [ac.sup.-1]
Watershed                             Agroecoregion
Upper Minnesota           3.33        Steep Dryer Moraine
Pomme De Terre            4.18        Steeper Till
Lac Qui Parle             4.12        Coteau
Hawk Cr./Yell. Med.       2.61        Dryer Blue Earth Till
Chippewa                  2.93        Dryer Clays and Silts
Redwood                   2.55        Steep Wetter Moraine
Middle Minnesota          6.14        Rolling Moraine
Cottonwood                3.04        Wetter Blue Earth Till
Watonwan                  2.59        Wetter Clays and Silts
Blue Earth                3.43        Alluvium and Outwash
Le Sueur                  2.24        Steep Stream Banks
Lower Minnesota           4.56        Steep Valley Walls
                                      Stream Banks
                    Erodibility Index
                      Weighted Mean
                      t [ac.sup.-1]
Upper Minnesota           5.30
Pomme De Terre            2.46
Lac Qui Parle             4.53
Hawk Cr./Yell. Med.       1.94
Chippewa                  2.79
Redwood                   7.78
Middle Minnesota          3.92
Cottonwood                3.30
Watonwan                  1.94
Blue Earth                3.77
Le Sueur                  5.98
Lower Minnesota          17.18
                       Erodibility index comparisons
                  between agroecoregions within the Lower
                     Minnesota River watershed and the
               Chippewa River watershed (ANOVA 0.05 level).
Lower Minnesota Watershed
                          Rolling Moraine Steeper Till Steep Valley Walls
Alluvium and Outwash            no             no             yes
Rolling Moraine                                no             yes
Steeper Till                                                  yes
Steep Valley Walls
Steep Wetter Moraine
Wetter Clays and Silts
Lower Minnesota Watershed
                          Steep Wetter Wetter Clays Area Within
                            Moraine     and Silts    Watershed
Alluvium and Outwash           no          yes           8%
Rolling Moraine               yes          yes          21%
Steeper Till                  yes           no           8%
Steep Valley Walls            yes          yes           3%
Steep Wetter Moraine                       yes          28%
Wetter Clays and Silts                                  32%
Chippewa Watershed
                      Dryer Clays and              Steep Dryer Steep Valley
                           Silts      Steeper Till   Moraine      Walls
Alluvium and Outwash        no             no          yes         yes
Dryer Clays and Silts                      no          yes         yes
Steeper Till                                           yes         yes
Steep Dryer Moraine                                                yes
Steep Valley Walls
Stream Banks
Chippewa Watershed
                                   Area Witin
                      Stream Banks Watershed
Alluvium and Outwash       no         18%
Dryer Clays and Silts      no         20%
Steeper Till               no         15%
Steep Dryer Moraine        no         45%
Steep Valley Walls        yes          1%
Stream Banks                           1%
                     Stream habitat assessment scores
                       according to major watershed,
                  agroecoregion, and the intersection of
                    major watershed and agroecoregion.
                            Stream Habitat
                           Assessment Score
Spatial Unit                     Mean       S.D. n
Lower Minnesota River (LM)        82         33  24 Major Watershed
Le Sueur River (LS)               88         31  24
Blue Earth River (BE)             80         25  20
Wetter Clays & Silts (WCS)        79         31  36 Agroecoregion
Rolling Moraine (RM)              89         28  32
LM-WCS                            65         32  12 Intersection
LM-RM                             99         25  12
LS-WCS                            83         33  12
LS-RM                             93         29  12
BE-WCS                            89         25  12
BE-RM                             68         21  8
                           Two-way ANOVA Level
Spatial Unit                 of Significance
Lower Minnesota River (LM)        0.10
Le Sueur River (LS)
Blue Earth River (BE)
Wetter Clays & Silts (WCS)        0.05
Rolling Moraine (RM)
LM-WCS                            0.01
                     Minnesota River Basin stakeholder
                   groups and selected relationship. [*]
                      Stakeholder Group
                       Non Government      Government
Topic                    Utilitarian       New Style
Decision-making          individual        committee
Driving Values            expansion        ecosystem
Focus of Attention       successful       partnership
Management Preference  local/political   eco-political
Scale of Intereset      local/present   joint areas/near
Targeting Mitigation?        yes            depends
                          Non Government
Topic                       Ecological
Decision-making          moral authority
Driving Values         cultural integration
                          the ecosystem
Focus of Attention        protection and
Management Preference       ecological
Scale of Intereset    all scales, ecological
Targeting Mitigation?           no
(*.)See text for examples of stakeholder groups.
COPYRIGHT 2001 Soil & Water Conservation Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2001 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Hatch, L.K.; Mallawatantri, A.; Wheeler, D.; Gleason, A.; Mulla, D.; Perry, J.; Easter, K.W.; Smith,
Publication:Journal of Soil and Water Conservation
Article Type:Statistical Data Included
Geographic Code:1U4MN
Date:Jan 1, 2001
Previous Article:Influence of Thermal Gradients on the Rates of Heating and Cooling of Streams.
Next Article:Comparison of Profit Maximization and Multiple Criteria Models for Selecting Farming Systems.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters