| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Natural Resource Sci. and Landscape Architecture, Univ. of Maryland, 1112 H.J. Patterson Hall, College Park, MD 20742
b Pasture Systems and Watershed Management Research Unit, USDA-ARS, Curtin Rd., University Park, PA 16802
c Dep. of Crop and Soil Sci., The Pennsylvania State Univ., 116 A.S.I. Building, University Park, PA 16802
* Corresponding author (bneed{at}umd.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: VSA, variable source area
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Variable source area (VSA) hydrology has been proposed as an effective framework to describe the response of humid temperate watersheds to precipitation (Ward, 1984). Under VSA hydrology, a small percentage of the landscape is the source of stormflow; source areas contract and expand within and between events. Variable source areas are a function of topography, soils, geology, climate, and management. Within VSA hydrology, surface runoff generation is classified as either infiltration excess or saturation excess (Sklash, 1990). Infiltration excess, or Hortonian overland flow, occurs when rainfall intensities exceed the infiltration capacity of a soil. Variable infiltration capacities in the landscape cause partial areas of infiltration-excess surface runoff (Betson, 1964). Saturation excess occurs when the water table rises to saturate the soil profile, filling storage zones and minimizing infiltration capacity. Ideally, P management strategies should differ depending on the dominant surface runoff generation mechanism in a watershed. An infiltration-excess-based approach should target soils with a low infiltration capacity, while a saturation-excess-based strategy should target near-stream and other zones that are subject to surface saturation (Gburek et al., 1996) regardless of infiltration capacity at these sites.
The saturation-excess mechanism has been studied primarily under forest and grassland vegetation (Bonell, 1993). In most temperate forested landscapes, infiltration excess is a minor mechanism of runoff generation; rather, stormflow is dominated primarily by shallow subsurface flow, with contributions from saturation-excess surface runoff (Anderson and Burt, 1990). The lack of infiltration-excess surface runoff in forested landscapes is due to the high infiltration capacity of forest surface soil horizons, including O horizons. Few field studies have been conducted in agroecosystems to investigate the relative occurrence of infiltration-excess and saturation-excess runoff (Betson and Marius, 1969). Typically, agricultural soils, particularly seasonally bare soils, exhibit lower infiltration capacities and less macroporosity than do forest soils. These factors favor the infiltration-excess mechanism over the saturation-excess mechanism (Burch et al., 1987).
This field study was located within a subwatershed of the Mahantango Creek watershed. The Mahantango Creek watershed has been reported as a VSA watershed (Pionke et al., 1996). Variable source areas are thought to occur primarily in the near-stream zones in response to the close proximity of the water table to the land surface, which causes seep zones and high antecedent soil water content contents (Gburek and Sharpley, 1998). Stormflow from these VSAs is thought to be primarily saturation-excess surface runoff and rapid subsurface flow (Gburek et al., 1996). The VSA-based simulation modeling of a 26-ha mixed land use watershed located near the field site for the present study predicted that 98% of the runoff volume is produced from 14% of the area (Zollweg et al., 1995). These studies in the Mahantango Creek watershed have been conducted only under grassland vegetation. In cropped soils, cultivation, wheel traffic, and other management practices may decrease infiltration capacities so that infiltration-excess runoff may be of greater relative importance in runoff generation.
Subsurface hydrologic properties may be critical to the occurrence of surface saturation because of the central role of shallow water table dynamics. Subsurface properties that may influence the occurrence of surface saturation include the presence of a fragipan or other soil layers that may cause episaturation, the hydraulic conductivity of subsurface horizons, depth to bedrock, depth of topsoil, and the depth to a water table. The occurrence of the water table rise that causes surface saturation may be correlated to soil hydromorphology and natural drainage class (Engman and Rogowski, 1974). Runoff and drainage class have been indirectly linked through soil water content measurements in a study of similar soils as those in the present study (Henninger et al., 1976). A good correlation was observed between the extent of somewhat poorly drained soils and the maximum seasonal extent of the surface-saturated zone in a small catchment in northeastern Vermont (Dunne et al., 1975; Moore et al., 1976).
The objectives of this study were to assess the surface runoff response to rainfall at four colluvial and two residual soil sites and investigate the factors influencing runoff generation at these sites including rainfall event characteristics, surface saturation, management, and antecedent soil water content. For additional information on this study see Needelman (2002).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Instrumentation
Two hillslopes were instrumented and continuously monitored during nonwinter periods to compare the hydrologic response of colluvial and residual soils. The study could not be conducted during the winters because many of the sensors are susceptible to frost damage. Three sites were established on each hillslope at distances of about 15, 35, and 65 m from the stream channel (Fig. 2
and Table 1). Sites on the south side of the channel are labeled S1, S2, and S3, with S1 closest to the stream, and on the north side were labeled N1, N2, and N3, with N1 closest to the stream.
|
|
Sites were monitored during the periods of July through December 2000, April through December 2001, and April through December 2002. At each site, a surface H-flume was installed to monitor surface runoff (Fig. 3) . A berm was constructed to give the flume a runoff collection width of 10 m. This width was chosen to be sufficiently large to aggregate small-scale runoff processes such as rill flow, yet small enough to allow for intense data collection in the upslope source area.
|
To prevent intersite effects, sites were staggered along the hillslopes. This was not possible between Sites S1 and S2 because of local topography (Fig. 2). A network of roughly 7- x 7-cm ditches were excavated to redistribute runoff water passing through the flume at S2 into the area upslope of flume at S1.
Eight paired 45-cm wells and surface runoff indicators were installed upslope from each flume (Fig. 3). Wells were constructed and installed in the same manner as were the piezometers, except tube diameters were 3.75 cm and slotting extended from 5 to 60 cm (rather than 5 cm of slotting at 30 or 60 cm). The first two well runoff indicator pairs were installed 2 m on both sides and 1 m upslope of the piezometers. The other six pairs were installed on a 10- x 10-m grid upslope from the flume. Wells were monitored with subsurface saturation sensors following Srinivasan et al. (2000). These sensors indicate whether the water table is at or above 1, 5, 15, 30, or 45 cm from the soil surface. Surface runoff indicators (Srinivasan et al., 2000) are miniature V-notch weirs (40-cm width) with an electrical conductivity sensor indicating the presence or absence of surface water. All data were collected on a 5-min interval with a Campbell CR-10 datalogger to capture rapid water table changes during events (Calmon and Day, 1999).
Data Analysis
This study was observational rather than inferential. It was not feasible to monitor a sufficient number of sites with the intensity of data collection performed at each site to address objectives with statistical hypothesis testing. Data were analyzed on a site-by-site basis across time, comparing runoff response under varying precipitation and site conditions. Conclusions on the comparative runoff generation between sites were based on broad, clearly discernible differences.
A topographically based surface-runoff-contributing area was calculated for each flume (Table 1). All flumes except S1 were located in a hillslope position with a plan curvature near zero. Therefore, contributing areas were calculated by extending the berm width to a road, which forms a surficial topographic boundary on each hillslope. Site S1 is located within a convergent position. The S1-contributing area was delineated by the upslope road and a topographic bench in the landscape.
Note that the topographic contributing area conflicts with the concept of a VSA. Nonetheless, the topographic contributing area is necessary for the calculation of runoff yields, an important comparative measure between sites. Runoff yields were calculated as the volume of runoff divided by the volume of rainfall falling within the topographically defined contributing area.
Soils of Study Site
An intensive soil survey was conducted along the two hillslopes through a cooperative agreement with the Pennsylvania Natural Resources Conservation Service (NRCS) (Needelman, 2002). Survey work included pit and soil core descriptions and laboratory characterization. The soils are formed in shale, siltstone, and sandstone residuum and colluvium. Two contrasting soil groups are found along the hillslopes, differing primarily in subsurface hydrologic characteristics including the presence of a fragipan, the clay content of the argillic horizons, and natural drainage class (Table 2). Ap horizons are generally 25 to 30 cm thick and are underlain directly by Bt horizons. The upper portions of both hillslopes have well-drained residual soils classified as fine-loamy, mixed, semiactive, mesic Typic Hapludults (Leck Kill series). The bottom of the northern hillslope has moderately well-drained colluvial soils classified as fine-loamy, mixed, active, mesic Oxyaquic Fragiudalfs (Hustontown series). The southern hillslope has somewhat poorly drained colluvial soils classified as fine-loamy, mixed, semiactive, mesic Aquic Fragiudalfs (Albrights series). On the southern hillslope, the colluvial soils extend about 90 m from the stream channel. There is a 15-m transition zone where the fragipan grades out and redoximorphic features decrease. Sites S1, S2, and S3 are colluvial soil sites (Fig. 2). On the north side of the channel, the bottom 15 m of the hillslope are colluvial soils, followed by a 15-m transition zone; the remaining hillslope is residual soils. Site N1 is a colluvial soil site while N2 and N3 are residual soil sites (Fig. 2).
|
Detailed Site Characteristics
Cropping for the sites is presented in Table 1. Sites S2 and S3 were moldboard plowed and disc harrowed each spring before planting. The N1 site was chisel and disc tilled in the spring of 2000 and 2001 and was not tilled in 2002. The N2 site was disc harrowed in the fall of 1999. The N3 site was chisel and disc tilled in the spring of 2000 and the fall of 2001.
Site S1
Site S1 was converted to mowed grass in 1996 when a former runoff generation study was initiated (Fig. 2). In the former study, a network of saturation detectors were monitored throughout the grassed area. Results of the previous study remain unpublished. The conclusion of the previous study was that saturation-excess runoff occurred throughout the sampled area during major events, indicating that the extent of area monitored was not large enough to delineate the VSA. Fragipans at site S1 are typically shallow with minimum depths of 44 cm in some locations.
Site S2
Site S2 is in a toeslope position. Fragipan depth is generally deep in this area (80–150 cm), though depth is difficult to estimate because of broken argillic-fragipan boundaries and the presence of clay lenses within the fragipans in this area.
Site S3
Site S3 is located in the footslope position of a slope portion that grades from the shoulder with skeletal residual soils 40 m above the flume (upper extent of wells and runoff indicators) to lower rock fragment colluvial soils near the flume. While installing instrumentation, a French drain was discovered 2 m upslope from the flume running parallel to the berm. The bank of sensors located above the flume is just downslope of the French drain. The infrequent occurrence of a high water table during the study, despite the presence of redoximorphic features directly below the Ap horizon, may be because of this drain. This site therefore represents the artificial condition of a drained soil that is naturally somewhat poorly drained. Analysis of data from this site cannot be extrapolated to similar sites without French drains; nonetheless, this site proved worthwhile to elucidate the role of surface saturation in runoff generation. French drains are ubiquitous in this area and have not been mapped; it is possible that there are other drains in the study area. However, the sites with off-drained soils (S1, S2, and N1) had sustained high water tables during the study period, an indication that the effect of artificial drainage is minimal if present.
Site N1
Site N1 is located within a colluvial draw at the base of a primarily residual hillslope. Colluvium extends about 25 m from the stream at this location. A minor seep is located about 15 m directly upslope of the flume. Although this seep did not generate substantial overland flow during the study period, there is likely a channelized natural or artificial subsurface flow system that feeds the ground water in the site area.
Sites N2 and N3
Sites N2 and N3 have similar residual soils and are topographically similar: Both are located along an upland backslope that extends nearly to the watershed divide, though a road forms the upper boundary of the surface runoff contributing area. Site N2 was removed from the study following the 2001 monitoring period because of farmer participation and resource constraint limitations.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Surface Runoff
Aggregated runoff results at the six monitoring sites are given in Table 3 based on all 111 rainfall events, except that Site N2 was not in operation in 2002. Total runoff volumes were greatest at Sites S1 and S2 and minimal at the residual soil sites. Rainfall characteristics and runoff volumes for larger events (>2.0-cm depth) and selected smaller events (
2.0 cm) are given in Table 4. The colluvial soils, Sites S1, S2, S3, and N1, produced dramatically greater runoff volumes than did the residual soils, Sites N2 and N3, for all major runoff events during the study period. For larger rainfall events, runoff volumes were generally greatest at Sites S1 and S2. The two sites located on residual soils, N2 and N3, produced little or no runoff during any event. Three events in the larger event group (1 Sept. 2000, 1 June 2001, 25 Nov. 2001, and 15 Sept. 2002) produced little or no runoff at all sites. There were several smaller events that produced runoff volumes equal to or greater than some of the large events. For example, at Site S2 the 22 June 2001 (2) event generated the largest volume of any smaller event, yet only had a rainfall depth of 0.94 cm.
|
|
All sites except S1 had similar topographic contributing areas; the topographic contributing area for Site S1 is five to seven times larger than the area of the other sites. However, the actual contributing areas, as indicated by the runoff indicators, only included a small percentage of the full topographic contributing area. This follows the basic precepts of VSA hydrology, that contributing areas are dynamic during events and may not be directly related to the topographic contributing area.
Factors Influencing Surface Runoff Generation
Surface Saturation
Surface saturation was operationally defined as cases when at least two of the eight 45-cm wells recorded a water table within 5 cm of the soil surface. Sites S1, S2, and N1 were subject to surface saturation for many of the larger flow events during the study period (Table 4). At Site S1, there was a one-to-one correspondence between surface saturation and the occurrence of any surface runoff. Runoff was recorded at Site S2 under both surface-saturation and non-surface-saturation conditions; the largest runoff-producing events (>500 L) were all surface-saturating events at Site S2. There were four surface-saturating events that did not generate substantial runoff volumes (<300 L). The largest volumes at Site N1 were also restricted to surface-saturating events. Sites S3 and N2 each experienced surface saturation during only one event during the study period; Site N3 did not experience surface saturation at any point.
To quantify the effect of surface saturation, total runoff yields were calculated separately for surface-saturated and unsaturated events (Table 3). Sites S2 and N1 were the only sites with multiple surface-saturated and unsaturated runoff events. The ratio of total runoff yield generated under saturated conditions to total runoff yield generated under unsaturated conditions was 40 and 15 for Sites S2 and N1, respectively. These ratios are dramatic, yet they are somewhat biased because larger events tended to generate both surface saturation conditions and larger runoff yields. The effect of surface saturation can better be observed by comparing similarly sized events with differing saturation status. This analysis is presented in Fig. 4 , in which rainfall depths are plotted against peak rainfall intensity with symbol size proportional to runoff volume at Site S2. The effect of surface saturation can be assessed independent of rainfall depth by comparing the events with rainfall depths between 0.9 to 2.0 cm: the range of intermediate-sized events smaller than those that were exclusively saturation-inducing and larger than those that produced little runoff. There were 28 events in this size class, 10 of which caused surface saturation at Site S2. The non-surface-saturating events in this group had an average runoff yield of 0.1% with a maximum yield of 1.0%. In contrast, the surface saturation events in this size class had an average runoff yield of 8.5% with two events yielding over 20%.
|
|
The occurrence of surface saturation was strongly related to runoff generation at these sites. Surface saturation was observed almost exclusively at colluvial soil sites without subsurface drainage.
Management
Surface runoff from agricultural lands can be affected by surface cover, tillage, and other management practices. Management decisions were made by farmers in this on-farm study (Table 1). Therefore, site management was an uncontrolled variable in the experiment, which is a limitation of this study. Although the design of this study does not allow for a detailed analysis of management influences, broad trends were discernible.
There were three generalized cropping practices applied to the study sites: cultivated row crops, cultivated small grains and legumes, and managed grassland (Site S1). There was no discernible difference in runoff production between sites under cultivated row crops vs. those under cultivated small grains and legumes (Tables 1, 4). There were some minor management influences discernible within sites. For example, crop management may have caused the reduced runoff production at Site N3 during 2002 in comparison with 2000 and 2001. However, these differences were not substantial within the study (Table 4).
The greatest management variation was the grassland vegetation at Site S1. If this site had little runoff production, it would not have been clear whether the cause was management or other factors. However, this site was among the largest runoff-producing sites (along with S2). There was a one-to-one correspondence between surface runoff and the surface saturation at this site; the soil-dependent occurrence of surface saturation outweighed the runoff mitigating effect of the grassland vegetation.
The strongest runoff generation factor in this study, surface saturation, was independent of management practices and was related only to the presence of colluvial or residual soils and subsurface drainage at these sites (Table 4). For example, both Sites S2 and N3 were in small grains in 2002, yet three events with runoff yields >10% were recorded at Site S2 while there wasn't a single runoff-generating event observed at N3 (Table 4).
Antecedent Soil Water Content
Antecedent soil water content was associated with runoff production at these sites indirectly; the occurrence of surface saturation was dependent on antecedent soil water content for all except the largest events. The two large-volume, high-intensity events (>2.0 cm, >8.0 cm h–1 peak intensity) both caused surface saturation regardless of antecedent soil water content, though surface saturation was delayed. In all cases of surface saturation occurrence during smaller events (2.0 cm), there was a minimum antecedent surface soil water content of 0.35 m3 m–3 at the given site. The occurrence of a surface-saturated layer under these small rainfall depths was likely related to the presence of a capillary fringe before the event (Gillham, 1983; Novakowski and Gillham, 1988; Abdul and Gillham, 1989). The S1 and S2 sites exhibited the highest levels of surface soil water content during most of the study period. In general, surface soil water content levels are correlated to soil drainage class in this area (Henninger et al., 1976). These observations are consistent with positive correlations observed between soil water content levels and clay content, the presence of fragipan-perched water tables (Knuteson et al., 1989), and unsaturated flow of soil water to footslope and toeslope positions (Hewlett and Hibbert, 1963; Jackson, 1992).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We were not able to study winter conditions such as frozen soils or snowmelt events in this study because of equipment limitations. This is a major limitation of this study because hydrologic conditions during the winter (frozen soil, snowmelt events, high water table and soil water content) are favorable for large runoff events. For example, winter runoff from residual soils may have occurred during the study period. The results of this study suggest that nonwinter P management for the residual soils in this watershed should focus on rare, large events.
The methods common in current runoff indexing methods (such as curve number, soil hydrologic group, slope class) do not capture the important role of the subsurface soil properties and upslope soil characteristics that was observed in this study. For example, the method used in the Pennsylvania P index, a site assessment tool to rank agricultural fields by their potential for P export (Lemunyon and Gilbert, 1993), estimates that all of these sites are in the moderate class, except S1 which is low (Weld et al., 2002). This study was not designed to test the P index, and therefore we cannot make conclusions on the success of the P index at the monitoring locations. However, there are several implications of these results regarding indexing methods that warrant discussion. The flumes at Sites S1 and N1 are located within the same soil mapping unit and both frequently experienced surface saturation, but the S1 site generated substantially greater runoff yields than did the N1 site. This can be attributed to the contrasting soils upslope of these flumes (S1 primarily colluvial, N1 primarily residual). This comparison has important applied ramifications. For example, the Pennsylvania P index would estimate that Sites N1 and S1 have the same transport potential because they are the same distance to a stream and lie within the same soil mapping unit (therefore, identical Runoff Class and soil erodibility factor) (Gburek et al., 2000). However, the total runoff yield at Site S1 was nearly five times that of N1 (Table 3). For larger events, the contributing area for S1, composed of colluvial soils, was hydrologically active (as detected by distributed runoff indicators) while the N1 upslope residual soils were relatively inactive. To accurately estimate the runoff production of a site, upslope hydrologic characteristics should be integrated with other site-characteristic information.
While designing this study, we considered the rise of the fragipan-perched water table to the soil surface as the likely mechanism for surface saturation under the saturation-excess runoff generation mechanism. During the course of the study, we recognized and observed a second possibility, the perching of a water table within the argillic horizon. An argillic-perched transient water table may have developed as the advancing wetting front met sharply decreasing hydraulic conductivity rates at the Ap-Bt1 horizon boundary (p. 220–221 in Daniels and Hammer, 1992). A good example of this phenomenon occurred during the 12 June 2001 event at Site S1 (Table 4). The watershed was relatively dry before this event, stream baseflow was low, both piezometers were dry, and surface soil water content at the S1 site was 0.29 m3 m–3. Forty-eight hours after this event, the surface soil water content at this site was 0.43 m3 m–3. There is a monitored tile drain that outlets 5 m downslope of Flume 3. Before this event, this tile drain was not flowing (a rare occurrence during the study period). We deduced from this set of conditions that the subsurface soil horizons were not near field capacity before the event, and therefore a rapid rise of the water table caused by the wetting of a capillary fringe could not occur (Gillham, 1983). After only 2.0 cm of rainfall, surface saturation was recorded extending from 30-cm (piezometer response) to the soil surface (well data). Realistic estimates of porosity and antecedent profile water content are not consistent with the rise of a water table from the fragipan (60 cm) to the soil surface through a water-depleted profile with only 2.0 cm of rainfall. Rather, this free water must have been perched near the Ap-Bt1 boundary, which is normally 25 to 30 cm deep at these sites.
We may assume that there is a sharp hydraulic conductivity decrease at the Ap-Bt1 horizon boundary due to the strongly contrasting soil texture, bulk density, and structure at this boundary (Bonell, 1993). The degree of this drop is likely very strong in the colluvial soils and weaker in the residual soils because of the greater clay contents of the argillic horizons in the colluvial soils (37–44% in colluvial soils, 24% in residual soils). There is also likely a more general decrease in hydraulic conductivity with depth within the argillic horizon, because of both increasing clay content and increasing size of the subangular and angular blocky structure. A saturated layer similar to this argillic-perched water table was observed along the A-B horizon boundary in soils with thin A horizons in a field study on an agricultural watershed (Betson and Marius, 1969). It seems clear that surface saturation induced from the rise of a fragipan-perched water table is a case of saturation-excess runoff generation. However, the case of an argillic-perched water table inducing surface saturation has characteristics of both saturation-excess and infiltration-excess runoff generation.
All the colluvial soil sites were hydrologically active throughout the study period; the residual soil sites were hydrologically inactive. The lower portion of the colluvial footslope (Sites S1 and S2) produced the greatest runoff volumes and yields. The primary factor related to runoff production at the study sites was the occurrence of surface saturation, which was restricted to undrained colluvial soils. Antecedent soil water was also related to runoff production, both directly and because the occurrence of surface saturation was dependent on high antecedent soil water contents for smaller events. Agricultural management practices had minimal influence on runoff production at these sites in comparison with the strong influence of surface saturation. Nutrient management planning could be improved if surface runoff indexing methods, such as runoff class, were to better integrate information on the upslope and subsurface hydrologic properties (fragipans and high clay content argillic horizons) that differentiate the colluvial and residual soils in this study.
|
ACKNOWLEDGMENTS |
|---|
Received for publication March 3, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. E. Winzeler, P. R. Owens, B. C. Joern, J. J. Camberato, B. D. Lee, D. E. Anderson, and D. R. Smith Potassium Fertility and Terrain Attributes in a Fragiudalf Drainage Catena Soil Sci. Soc. Am. J., September 1, 2008; 72(5): 1311 - 1320. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. M. Easton, M. T. Walter, and T. S. Steenhuis Combined Monitoring and Modeling Indicate the Most Effective Agricultural Best Management Practices J. Environ. Qual., August 8, 2008; 37(5): 1798 - 1809. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. N. Sharpley, P. J. A. Kleinman, A. L. Heathwaite, W. J. Gburek, J. L. Weld, and G. J. Folmar Integrating Contributing Areas and Indexing Phosphorus Loss from Agricultural Watersheds J. Environ. Qual., June 23, 2008; 37(4): 1488 - 1496. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. N. Sharpley, P. J. A. Kleinman, A. L. Heathwaite, W. J. Gburek, G. J. Folmar, and J. P. Schmidt Phosphorus Loss from an Agricultural Watershed as a Function of Storm Size J. Environ. Qual., March 1, 2008; 37(2): 362 - 368. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. W. McDowell Phosphorus and sediment loss in a catchment with winter forage grazing of cropland by dairy cattle. J. Environ. Qual., March 1, 2006; 35(2): 575 - 583. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
