| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a School of Natural Resources, The Ohio State University and the Ohio Agricultural Research and Development Center, Wooster, OH 44691
b Water Quality Laboratory, Heidelberg College, Tiffin, OH 44883
c School of Natural Resources, The Ohio State University, Columbus, OH 43210
* Corresponding author (calhoun.2{at}osu.edu)
Received for publication August 12, 2000.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONDATA SOURCES AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Abbreviations: SRP, soluble reactive phosphorus • WQL, Water Quality Laboratory
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONDATA SOURCES AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Water quality monitoring programs in the Lake Erie basin are currently the most detailed of their type in the United States (Richards et al., 2002). Interpretation of trends in these data during the last 25 yr has focused on agricultural management as a principal determining factor. Minimum consideration has been given to soil properties that can modify the effect of tillage and fertilizer management. Water quality data collected near the outlet of large watersheds reflect integrated pollutant export from each upstream watershed and loading to downstream receiving waters. The water quality data collected near the outlets of large rivers provide no clues as to the critical source areas for various pollutants within the upstream watershed.
The Water Quality Laboratory (WQL) at Heidelberg College initiated storm event sampling for total phosphorus in the Sandusky River basin in 1969 (Baker and Kramer, 1976). Continuous studies of nutrient and sediment export have been underway for the Sandusky River by the WQL since 1975 and for the Maumee River since 1976 (with a gap in the 1979, 1980, and 1981 water years). Two intermediate-size watersheds in the Sandusky River basin, Honey Creek and Rock Creek, have been continuously monitored since 1974 and 1983, respectively. Intermediate-size watershed monitoring (1000–50000 ha) seems to be predicated on the assumption that relatively steep slopes in northwestern Ohio are the major contributors of sediment and chemical pollutants to Lake Erie. This may not be a reasonable assumption once soil properties and principles of soil physics are considered.
Upstream monitoring of smaller watersheds in the Maumee basin reached a peak in the 1970s in response to excess phosphorus loading in Lake Erie (Logan, 1981; Jones et al., 1977). Jones et al. (1977) evaluated the chemistry of water and sediment leaving three nearly level watersheds (50–150 ha) with very poorly drained soils derived from the most extensive geologic deposits in the basin (lacustrine clays, glacial till, and beach sand). The monitoring period for this pioneer study by Jones et al. (1977) was 1970–1972. This is important because the study precedes the initiation of the WQL studies. During the period 1975–1980, the losses of nutrients and sediments from agricultural lands were monitored in the Maumee River basin using micro-watersheds and research plots selected to represent major soil series occurring within the basin (Logan, 1979). The project included a number of special studies that targeted soil loss, sediment export, and water quality. None of these smaller-watershed studies, with the exception of Lost Creek, continued into the 1980s and it was terminated in 1993, plagued by questionable data (Baker, 1994).
Myers et al. (2000) reported that between 1996 and 1998 unit area export of sediment (expressed as ton mi-2) for several tributaries of the Maumee River was highest in parts of the lake plain where soils were both poorly drained and high in clay. Soil loss estimates based on the Universal Soil Loss Equation (USLE) (Wischmeier and Smith, 1978) were much higher for the sloping moraines with better drained soils than they were for the more clayey lake plain soils. Implied is the need to carefully interpret USLE estimates compared with calculations of unit area export of suspended solids. Sediment-yield measurements and USLE estimates of soil loss appear to be contradictory even when tillage management is considered.
Short-term, small-watershed studies were completed (late 1970s) by the time continuous monitoring of the Maumee and Sandusky watersheds began. The data from the smaller watersheds were interpreted and published without the availability of the long-term water quality data for comparison. These studies have been largely ignored since their publication. Through reexamination of these studies we demonstrate the effect of intrinsic soil properties, including slope, surface soil texture, and internal drainage, on pollutant export. This paper emphasizes the importance of the soil series in containing information critical to water quality. The results of this review will bring into question the assumption that sloping glacial moraine areas are more important pollutant sources than are the flat lake plains.
|
DATA SOURCES AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONDATA SOURCES AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
This paper is a review of previous water quality studies in the Maumee and Sandusky River basins. The data are reinterpreted in light of information gained from WQL studies of the larger watersheds during the past 20 yr. We will examine water quality and soil interrelationships at four watershed scales: (i) large, (ii) intermediate, (iii) small, and (iv) microwatershed or plot. The approach that is used in this paper is to begin with data from the two largest watersheds (Maumee and Sandusky) and then to compare them on a unit area export basis with the data from the intermediate-size and smaller watersheds.
The watersheds that provided data in this report are shown in Fig. 1 and water quality data for the large and intermediate watersheds are summarized in Table 1. These include the WQL collection stations at Fremont, Ohio near the outlet of the Sandusky River and at Waterville, 30 miles inland from the outlet of the Maumee River into Lake Erie. The two intermediate-size watersheds, Rock and Honey Creek, are tributary to the Sandusky River. There has been no monitoring of intermediate-size watersheds in the Maumee River basin, although the program described by Myers et al. (2000) represents the initiation of monitoring at this scale. At all of the WQL stations, automatic samplers, located at United States Geological Survey (USGS) stream gauges, were used to collect three or four samples per day, with all samples analyzed during storm events and one sample per day analyzed under nonstorm conditions. Details concerning sample collection, laboratory analyses and statistical procedures are covered by Baker and Richards (2002). Unit area export of suspended solids and selected nutrients for the large and intermediate sized watersheds are shown in Table 2.
|
|
|
|
Between 1975 and 1980 water quality research was conducted on several field-size watersheds and research plots in the Maumee basin (Logan and Stiefel, 1979; Logan, 1981). The characteristics of these sites are shown in Table 4. Their approach was to measure pollutant export from several major soils of the Maumee basin under crop production practices. The field-size watersheds were equipped with rated flumes, water level recorders for measuring runoff, and Coshocton wheel samplers. Logan and Stiefel (1979) describe design and analysis details for the micro-watershed and plot-size studies.
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONDATA SOURCES AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
General relationships between pollutant levels at the river mouth and upstream soil conditions are vague and seemingly contradictory at the large-watershed scale.
Intermediate-Size Watersheds
Rock and Honey Creeks
Both Rock and Honey Creeks are tributary to the Sandusky River and flow through glacial till–derived soils that are medium-textured and range in drainage from well to very poorly. The headwaters of Honey Creek originate in Crawford County and pass through the Willard Marsh. Willard Marsh is dominated by organic soils. The contribution of nitrogen to drainage water is expected to be significant because of the intensive fertilization of vegetable production and the nature of organic soils. In addition to leaching and runoff of nitrogen fertilizer, there is transport of N as nitrate to Honey Creek due to the drainage and subsequent oxidation of the surrounding organic soils. Erosion rate is anticipated to be higher in the Rock Creek tributary in comparison with Honey Creek because of slopes associated with the Defiance End Moraine. Lower suspended solids export for Honey Creek is expected due to the dominance of gentler slopes associated with till plain and some lake plain topography within the watershed.
Unit area export, from 1976–1995, of all potential pollutants for Honey Creek is less (significantly less for suspended solids) than that for the Sandusky basin (Table 2). For Rock Creek the same comparisons are made for the period of 1984–1995 and the differences are not statistically significant for suspended solids and total P but they are significantly less for soluble reactive P and nitrate N. This suggests that other watersheds in the Sandusky basin are contributing greater amounts of suspended solids and other pollutants than these two tributaries.
Land use in both watersheds is nearly identical although in recent years land area set aside for the Conservation Reserve Program has been relatively greater in the Rock Creek watershed. Cropland accounts for 81 to 83%, pasture 1 to 2%, forest 10 to 12%, water–wetland 1%, and other uses 4 to 6% in both Honey Creek and Rock Creek watersheds. The Sandusky basin is covered by 80% cropland, 2% pasture, 9% forest, 2% wetlands, and 7% other uses. Assuming that fertilizer application rates have declined and implementation of reduced tillage has occurred at the same level in these intermediate watersheds as it has basinwide in the Sandusky, then the smaller pollutant export is due to intrinsic properties of the watersheds and not due to management.
If these two watersheds are representative of the glacial moraine areas characteristic of the headwaters of the Sandusky and Maumee Rivers, then it is not logical to assume that watersheds located in moraines are the major contributors to downstream sediment pollution. Even though surface runoff will be greater than more gently sloping downstream watersheds, it can be inferred that there is less clay available for dispersion and transport. A large percentage of the soils in both watersheds have silt loam and loam surface textures while clayey textures are more prevalent in the northern half of the Sandusky basin. It could be concluded that delivery of suspended solids from tributary watersheds is more a function of quantity and availability of dispersible colloids than is slope and land management. By examining smaller, more edaphologically uniform watersheds it is possible to test this hypothesis.
Lost Creek
This watershed in the Maumee basin is 1117 ha in size and is located in a sloping end moraine area composed of, primarily, Glynwood (fine, illitic, mesic Aquic Hapludalf) and Blount (fine, illitic, mesic Aeric Epiaqualf) soils (Baker, 1994). Pollutant export data seem to contradict the similarly composed Rock and Honey Creek watersheds. Unit area export for suspended solids and total P is twice that of the Maumee as a whole and of the Rock and Honey Creek watersheds, which are also located in moraines. Pollutants that are transported in solution (SRP and nitrate N) are not substantially higher in terms of unit area export (Table 2). Part of the apparent discrepancy may be explained by uncertainty in the discharge data for the watershed. The USGS rating curves for the flume changed markedly during the life of the study, which may have led to overestimates of discharge. Comparison of annual runoff with other northwestern Ohio watersheds showed that Lost Creek was higher by 50% or more. In the latest published discharge data for 1997, mean annual runoff for water years 1986–1997 is 35 cm, which now makes it around 20% higher than mean annual runoff compared with other Maumee basin watersheds (Shindel et al., 1998). This watershed is still monitored by USGS but it is now rated as poor in terms of records.
Small Watersheds
The only water quality study of small (50–150 ha) watersheds was that conducted by Jones et al. (1977). It is unique because it is confined to lake plain physiography and the very poor drainage class. The one variable is soil texture and the effect of this property on water quality is evaluated. The discussion that follows is restricted to the Jones et al. (1977) study.
Median soluble reactive P values for the Millgrove and Hoytville watersheds are lower than those at Waterville on the Maumee River (Fig. 2) . It was discovered at a later date that septic tank effluent was entering the Paulding drainage ditch via a subsurface field tile outlet. This explains the high soluble reactive P value for the Paulding site and also highlights a potential problem in selection of small watersheds for monitoring.
|
Paulding soils, due to their unfavorable physical properties, are not highly productive; consequently, they may not have been heavily fertilized during the agricultural history of the Maumee basin (Calhoun et al., 2002). Current plant-available P levels in northwestern Ohio are 76 and 223% higher, respectively, for the Hoytville and Millgrove soils compared with the Paulding soils (Calhoun et al., 2002).
Suspended solids concentration in the drainage water was directly related to amount of soil clay. The Paulding watershed yielded the highest concentration of suspended solids. This suggests that nearly level watersheds, especially fine-textured ones, may be important contributors of suspended solids entering the Maumee River. Additionally, it was discovered that clay from the Paulding watershed was more easily dispersed in distilled water than that from the Hoytville watershed. The reason for this is not readily explained. Paulding soils are lacustrine clays with very little sand while Hoytville soils have higher sand contents and have formed in glacial till. As flat as the land surface appears there is always some slope gradient on these fine-textured soils. Once saturated there is no infiltration and clay is easily dispersed with raindrop impact. There is usually no tile drainage with Paulding soils. Sheet erosion will carry large quantities of sediments overland and into drainage ditches during high-intensity, medium- and long-duration storms. As previously discussed, in the case of the Rock and Honey Creek watersheds, it appears that the moraine areas may not be the greatest contributors to estuary sediment and chemical pollution. The results from, especially, the Paulding watershed, indicate that low-gradient lake plain watersheds dominated by very poorly drained soils high in clay (>50%) are an important component in export of suspended solids.
It is apparent, at the small-watershed scale, that soil texture and internal drainage are the controlling factors for median pollutant concentrations. In this particular case soil texture is controlled by parent material and internal drainage can be controlled by man's intervention with tile. Natural internal drainage (very poorly) and slope (nearly level) are controlled by physiography (lake plain).
Micro-Watersheds and Plot Studies
Between 1975 and 1980 water quality research was conducted on several field-size watersheds and research plots in the Maumee basin (Logan and Stiefel, 1979; Logan, 1981). The characteristics of these sites are shown in Table 4. The results of this study provide complimentary information on the importance of soil properties on pollutant export. Table 5 summarizes mean annual unit area export values for suspended solids, total P, SRP, and nitrate N. In general, these data reinforce conclusions reached with the previously discussed study on agronomic watersheds (Jones et al., 1977). In this case, flow volume was determined, allowing calculation of unit area export. The data are calculated means for what was highly variable monthly and yearly export generated by rainfall events but tempered by soil properties.
|
Export of suspended solids from the outlet of small watersheds into the next level of tributary drainage is a function of the clay content of the soils that occupy the watershed. The Roselms watershed is a dissected part of the lake plain. Paulding soils occupy the low-gradient, undissected plain. With slopes ranging from 3 to 15%, suspended solids yields from the Roselms and that from the nearly flat Paulding were of the same magnitude. A large portion of the suspended solids exported from watersheds dominated by Paulding soils may never make it to the Maumee River. As flow declines following a runoff event, Ca + Mg concentration in ditch water increases, causing suspended clay to flocculate and settle to the bottom of the ditch. Flocculation apparently retards resuspension and transport during subsequent runoff events as drainage ditches in the lake plain region are frequently cleaned; consequently, a sizable but unknown amount of the sediment is piled along ditch banks and spread over field edges throughout the area. If there is a contribution of septic tank effluent or other waste water that is higher in Na, then some clay may stay in suspension much longer and, subsequently, make its way further downstream. Roselms landscapes result from eroding slopes retreating into the lake plain surface. Consequently, much of the eroded sediment is moving into natural outlets that are more directly connected to permanent streams than are most of the Paulding outlets into low-gradient surface drainage ditches. High flow rates and dilute Ca + Mg concentration in these streams associated with, especially spring season, storm events point to Roselms watersheds as important sources of suspended solids moving into the Maumee River.
Hoytville and Lenawee (fine, mixed, nonacid, mesic Mollic Epiaquept) soils are lower in clay and much more frequently tile-drained than are the Paulding soils. Unit area export of sediment for both soils is only about half of that for the Maumee watershed. Tile drainage of Hoytville and Lenawee soils dramatically reduces surface runoff. In contrast to the Paulding, which is much higher in clay content and lacks macropore continuity, Hoytville and Lenawee soils, once tile-drained, will rapidly transmit water. This results in higher unit area export of nitate N and soluble reactive P from the Hoytville plots. The magnitude of export is a function of N and P fertilization levels and rainfall amount and intensity but tempered by soil properties. The export of soluble reactive P from the Hoytville parallels that of nitate N for the same reason (Table 5). This phenomenon is less pronounced for the Lenawee, due probably to lower sand and higher silt content compared with the Hoytville.
The better-drained, sloping Blount soils have a sediment export rate three to four times greater than the Hoytville but only 40% of that of Paulding and Roselms. Export of suspended solids for the Blount watershed is 70% greater than the average unit area export for the Maumee basin between 1976 and 1978. Part of the additional sediment exported from the Blount probably contained silt and sand-sized particulates deposited within a short distance downstream from the watershed. Both the Roselms and Blount soils are somewhat poorly drained and have higher amounts of free iron oxides that fix phosphates. This may explain the lower export of soluble reactive P compared with the poorly drained soils that contain lower levels of free iron oxides. The relatively high mean unit area export of soluble reactive P and total P from the Paulding watershed was due to unusually high runoff in the spring of 1978, although rainfall amounts were not excessively high during the period. The original authors provided no additional explanation. Unit area export of total P is generally greater for each soil than it is for the Maumee basin, but does not appear to be related to export of sediment.
At the micro-watershed and research plot scale, the general conclusions are the same as those derived from the study of small watersheds by Jones et al. (1977). The additional variables introduced by Logan and Stiefel (1979) and Logan (1981) were slope and internal drainage. Although Paulding and Roselms soils occupy only 5% of the Maumee basin, they appear to generate 2.5 times as much sediment as the Blount, which occupies nearly 20% of the Maumee basin. Blount soils export nearly twice and Paulding–Roselms soils five times as much sediment per unit area as that basinwide. Tile drainage of very poorly drained soils that are formed from either glacial till or silty lake deposits reduces runoff and increases downward movement of soluble nutrients into the tile drains.
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONDATA SOURCES AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Careful examination of upstream water quality studies reveals the importance of the soil series in explanation of pollutant export from rural landscapes. Future collection, analysis, and interpretation of water quality data would benefit from a more thorough examination of soils in tributary watersheds in order to explain pollutant export differences between tributary watersheds.
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONDATA SOURCES AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONDATA SOURCES AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. J. A. Kleinman, A. N. Sharpley, T. L. Veith, R. O. Maguire, and P. A. Vadas Evaluation of Phosphorus Transport in Surface Runoff from Packed Soil Boxes J. Environ. Qual., July 1, 2004; 33(4): 1413 - 1423. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. P. Richards, F. G. Calhoun, and G. Matisoff The Lake Erie Agricultural Systems for Environmental Quality Project: An Introduction J. Environ. Qual., January 1, 2002; 31(1): 6 - 16. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. B. Moog and P. J. Whiting Climatic and Agricultural Factors in Nutrient Exports from Two Watersheds in Ohio J. Environ. Qual., January 1, 2002; 31(1): 72 - 83. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Vadose Zone Journal | Journal of Plant Registrations | ||||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal |
