HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (8) Citing Articles via Google Scholar Google Scholar Articles by Moog, D. B. Articles by Whiting, P. J. Search for Related Content PubMed PubMed Citation Articles by Moog, D. B. Articles by Whiting, P. J. Agricola Articles by Moog, D. B. Articles by Whiting, P. J. Related Collections Soil Conservation Surface Water Quality Watershed and Landscape Processes Agroclimatology Water Pollution

SPECIAL SUBMISSIONS
Findings from the USDA-sponsored Lake Erie Agricultural Systems for Environmental Quality Project

Climatic and Agricultural Contributions to Changing Loads in Two Watersheds in Ohio

Department of Geological Sciences, Case Western Reserve University, 10900 Euclid Ave., Cleveland, OH 44106-7216

* Corresponding author (dbm3{at}po.cwru.edu)

Received for publication August 12, 2000.

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Trends in climatic variables, streamflow, agricultural practices, and loads of nutrients and suspended solids were estimated for 1976–1995 in the Maumee and Sandusky watersheds, two large agricultural basins draining to Lake Erie. To understand the contributions that various factors may have made to the trends in loads, earlier results of models linking loads to explanatory variables were combined with estimated trends in those variables. The study period was characterized by increases in temperature, wintertime precipitation and streamflow, conservation farming, and loads of nitrate and total suspended solids; decreases in snowfall and snow cover, fertilizer, manure from livestock, and loads of soluble reactive phosphorus; and relatively steady exports of total phosphorus. After removing the effects of trends in streamflow, nitrate loads increased much less while total suspended solids and total phosphorus loads declined. The analysis suggests that the nitrate increases were due largely to climatic factors, particularly increases in winter streamflow, decreases in snowfall and snow cover, and declining annual precipitation. Decreases in soluble reactive phosphorus were associated with changes in agricultural practices, particularly declines in fertilizer deliveries and head of livestock.

Abbreviations: CRP, Conservation Reserve Program • NO₂₊₃, nitrate plus nitrite • SRP, soluble reactive phosphorus • TP, total phosphorus • TSS, total suspended solids

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
THE MAUMEE RIVER and Sandusky River drain large, contiguous, agricultural watersheds in northwest Ohio. They supply large quantities of nitrate, phosphorus, and suspended solids to Lake Erie (Richards et al., 2002b). Efforts to reduce these species have been made over the last several decades (Myers et al., 2000a, b; Casey et al., 1997). The International Joint Commission called for an 11000 metric ton (Mg) annual limit on phosphorus discharges to Lake Erie; the Maumee watershed alone contributes about a quarter of this value (Myers et al., 2000b). Success in reducing point phosphorus sources has turned attention to agricultural nonpoint sources, which also dominate nitrate loads in the Maumee and Sandusky watersheds (Myers et al., 2000b). Fertilizer use appears to have declined (Richards and Baker, 2002). Reduction in sediment loss (and adsorbed nutrients) has been an objective of best management practices and conservation programs, chiefly conservation tillage and the Conservation Reserve Program (CRP), both of which have become widespread since the mid-1980s (Forster, 2002). Loads also may be influenced by factors largely beyond local human control, including trends in climate and streamflow. Simultaneous analysis of all these trends by a common analytical technique permits more reliable comparison and a broader perspective of factors affecting export from agricultural watersheds.

Richards and Baker (1993)(2002) examined trends in flow-weighted concentrations of nitrate plus nitrite (NO₂₊₃), soluble reactive phosphorus (SRP), total suspended solids (TSS), and total phosphorus (TP) in the Maumee River and Sandusky River since 1975. Using log-linear least-squares regression, they found highly significant decreases in SRP and TP concentrations. For the 1975–1990 period, TSS loads were fairly steady, and NO₂₊₃ load increased significantly. Declines since 1990 led to an overall steady NO₂₊₃ load and decreasing TSS load for the period 1975–1995. Noting concurrent trends in agricultural practices (Richards et al., 2002a), they attributed the declines in loads of SRP, TP, and TSS to increases in conservation tillage and decreases in fertilizer and manure application.

This study explicitly links trends in loads of NO₂₊₃, TSS, TP, and SRP to trends in explanatory variables by extending the statistical model of Moog and Whiting (2002). The set of explanatory variables is expanded to include climatic variables and streamflow, providing a test of the explanatory power of changes in agricultural practices. The analysis employs a method of trend detection and estimation different from that used by Richards and Baker (2002), and considers seasons separately, since they exhibit varying hydrologic characteristics.

The objectives of this study were twofold. The first objective was to detect and estimate trends and changes in loads, streamflow, climatic variables, and agricultural practices for the 1976–1995 study period. The second objective was to determine the changes in load implied by the changes in the other variables and compare them with the observed changes in load. These results suggest which factors may have contributed the most to changes in loads from 1976 to 1995.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Background
This study views the loads and the explanatory variables as random variables. Following standard statistical practice, the values of these random variables are regarded as samples drawn from a population of possible realizations—for example, a set of possible daily rainfalls, having a probability distribution that reflects the current climate. This distribution—and thus the population—may change with time.

The meanings of specific terms used in this paper are based on this paradigm. Trend refers to the slope of a linear variation with time of the mean of the population. Estimation refers to the process of fitting a line to data to derive an estimate of the trend (i.e., an "estimated trend"). Detection refers to the process of determining whether a trend is nonzero to a specified level of confidence. Change is the difference in a quantity due to an estimated trend over a specified time period.

Confusion may arise between estimation and detection. Lack of detection does not mean a trend does not exist (i.e., is zero-valued), and a trend that is not detected may still be estimated. All estimates involve a degree of uncertainty. Detection is a matter of determining whether this uncertainty includes the possibility that the trend is zero, but lack of detection does not render the estimated trend meaningless or less precise than estimates of other, detected trends. For example, one may precisely obtain an estimated trend close to zero.

Two primary techniques have been employed in detection and estimation of trends in climate and streamflow (Salas, 1993): linear regression on time, and the nonparametric Mann–Kendall test, described and extended to serially correlated data as the "seasonal Kendall test" by Hirsch et al. (1982). Richards and Baker (1993) applied both techniques in examining concentrations in the Maumee and Sandusky Rivers. Kunkel et al. (1994) used regression to analyze trends in Illinois. Gan (1998) applied the Mann–Kendall test to temperature, precipitation, and evapotranspiration in the Canadian Prairies. Associated with the seasonal Kendall test is its slope estimator (Hirsch et al., 1982), based on the median change between all pairs of months over the period.

The Mann–Kendall test and estimator are insensitive to short-term variations, while linear regression integrates their effects. Consider a steady 15-year rise followed by an equal fall occurring within 5 years. The Kendall estimator will mostly reflect the longer rise, whereas linear regression, influenced more by the drop, will convey a much steadier change.

Trend Detection and Estimation
While both the Mann–Kendall test and linear least-squares regression were employed in this study, detection and estimation of trends were based on linear regression of the variables on time. This technique was chosen to emphasize the changes that actually occurred, even if they were the "random" outcome of a steady climate. A change in load, even if it does not arise from a trend, still may be associated with changes in the explanatory variables. Thus, the detection of a trend is not a necessary condition for analysis of changes.

The explanatory variables for loads were streamflow, climatic variables (precipitation, temperature, snowfall, snow depth, and variables derived from them), and agricultural variables (fertilizer deliveries to dealers, manure from livestock, and acreage in conservation tillage and the Conservation Reserve Program). The time steps were annual for agricultural variables and monthly for all others. Streamflow, load, precipitation, and rainfall were log-transformed before regression. Data were averaged spatially over each watershed. The specific explanatory variables, stations, data sources, and averaging techniques are described by Moog and Whiting (2002).

In this study, detection employed a one-sided t test for significance of the regression slope. In cases involving all months of the year, the t statistic was corrected for autocorrelation of the residuals using the method described by Richards and Baker (1993). In other cases, only two or three consecutive months of each year were included; in these cases, the effect of autocorrelation on the t statistic was accounted for conservatively by counting one degree of freedom per year. Changes are reported as percentages over the 20-year period of study, water years 1976–1995, in the manner of Richards and Baker (2002), except for temperature changes, which are reported in degrees. For example, if the estimated trend in the natural log of load was m per month, then the percent change P over 20 years was:

[1]

Contributions of Changes in Explanatory Variables to Changes in Loads
To compare changes in loads with changes in the explanatory variables, we used results from a study reported in this issue (Moog and Whiting, 2002), in which we developed a statistical model expressing the loads as functions of streamflow and climatic and agricultural variables. Thus, a change in an explanatory variable may be transformed to a corresponding change in load using the model equations. While the results do not prove causation, they estimate the contributions to load changes produced by changes in explanatory variables under the hypothesis that the relationship is causal.

Time derivatives of the model equation from Moog and Whiting (2002) lead to an expression for estimated load trends:

[2]

where L is load, t is time, x are explanatory variables, N is the number of explanatory variables, b are fit parameters, and subscripts 1 and 2 refer to untransformed and log-transformed explanatory variables. For SRP, TP, and TSS Q' equals the streamflow rate Q, while for NO₂₊₃, they are related as:

[3]

Trends in "load–discharge residuals" may be estimated as:

[4]

where L_Q is the set of residuals of a fit by least-squares regression of loads to streamflow discharge, as described in this issue by Moog and Whiting (2002). The regression was log-quadratic for NO₂₊₃, and log-linear for the other species.

Calculation of load trends and changes were based on Eq. [2], which expresses the estimated trend as a linear combination of the load change due to discharge and each element of x₁ and x₂. The time derivatives were the slopes calculated using linear least-squares regression, retaining the individual values for each month in a season. The b coefficients were derived by Moog and Whiting (2002).

Trends and changes in load–discharge residuals, based on Eq. [4], may be interpreted as trends and changes in loads that have been adjusted to a constant streamflow discharge, referred to herein as "adjusted loads". Load–discharge residuals were employed because Moog and Whiting (2002) modeled the explanatory variables as predictors of the load–discharge residuals. The difference between a change in load and a change in the adjusted load indicates the magnitude of the contribution to changes in load from changes in discharge. Under the hypothesis that changes in the explanatory variables caused changes in loads, the contribution to changes in load–discharge residuals due to each variable may be taken as the value of the corresponding term on the right-hand side of Eq. [4].

Seasonal Disaggregation
Estimation of trends in annual average values may hide very different trends for individual seasons; for example, Karl et al. (1996) found seasonal differences in precipitation trends for the United States. Such differences are important, because the climate and hydrologic system in this region change extensively among the seasons.

To gain more insight into the dynamics of climate, agriculture, and loads, we disaggregated the year into five seasonal groups as in Moog and Whiting (2002): (1) January and February, (2) March and April, (3) May, June, and July, (4) August, September, and October, and (5) November and December. These seasons correspond to groups of months that were similar in load, streamflow, snow, and other climatic factors, as well as the relationships between climatic variables and loads.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Trends in Annual Averages
The estimated trends in agricultural variables for 1976–1995 are not presented in this paper, because they were similar to the estimated trends presented by Richards et al. (2002a) in this issue for virtually the same data. They showed that the primary nutrient sources—fertilizer use and manure from livestock—decreased, while conservation tillage grew from almost nothing to more than half the acreage. Most pronounced among nutrient sources was the drop in phosphorus fertilizer. The CRP enrollment showed a fairly steady climb beginning in 1988.

Mean annual flow rates were estimated to increase by 21% in the Maumee River and to decrease by 3% in the Sandusky River for the study period, 1976–1995. Neither change was a significant trend.

Estimated trends in actual loads and adjusted loads, calculated from the concentration and flow records, are shown in Table 1. Examination of actual loads shows that significant downward trends were detected in SRP load in both rivers. The NO₂₊₃ load trended upward in the Maumee River at a 90% confidence level (p < 0.1). Other trends were not detected, and estimates were less than 20% except for a sizeable increase in Maumee TSS load.

The sharper decline in SRP load as compared with TP load may indicate decreasing levels of soil phosphorus. Soluble reactive P appears in soils where phosphorus concentrations exceed the adsorption capacity of the soil (Stevenson, 1986). Therefore, it is a marginal indicator of soil phosphorus and is much more sensitive than TP to changes in soil phosphorus concentrations.

Changes in Climatic Variables and Load by Season
The largest changes in climatic variables and streamflow occurred in winter (November to February), which grew much warmer, particularly in January–February (Fig. 1) . Precipitation increased in November–December, while snowfall and snow cover decreased (Fig. 2) . It seems likely that the observed increases in wintertime runoff and decreases in springtime runoff are attributable to this winter warming and its associated decreasing snowpack storage of precipitation. Also evident are drier late summers in the Sandusky watershed.

View larger version (31K): [in this window] [in a new window]   Fig. 1. Temperature changes by season, 1976–1995. Numbers on bars indicate significance level of trend: 1, p < 0.05; 2, p < 0.01; 3, p < 0.001.

View larger version (27K): [in this window] [in a new window]   Fig. 2. Changes in selected climate variables, by season, 1976–1995.

View larger version (38K): [in this window] [in a new window]   Fig. 3. Changes in flow-adjusted loads, by season, 1976–1995.

View larger version (45K): [in this window] [in a new window]   Fig. 4. Changes in actual loads, by season, 1976–1995.

It is possible to estimate how much of the decline in adjusted TP loads was due to the soluble component. The decline was similar in both watersheds, averaging 28.5%. The decline in adjusted SRP load was about 80%. Since, on average, 15% of the total load was soluble, these SRP load declines would contribute to a decline of about 12% in TP load, or roughly 40% of the observed decrease.

Contributions to Changes in Loads
Solving for the terms in Eq. [4] does not directly reveal any causal relationships. We do hypothesize that changes in the explanatory variables are responsible for changes in the loads and load–discharge residuals, and in this sense refer to their "contributions to changes in loads."

Tables 2 and 3 list changes in the flow-adjusted loads over the 1976–1995 study period, accompanied by explanatory variables and their calculated contributions to load changes. Adjusted loads are listed instead of actual loads because the relatively large contributions from streamflow in all seasons would dominate the tables. The contributions of changes in streamflow variations are conveyed by a comparison of Fig. 3 and 4. The tables include only those seasons in which the adjusted load in question exhibited a trend that was significant at an 80% confidence level (lower than usual in hypothesis testing, but reasonable for separating those cases that probably do and do not exhibit changes large enough to lead to meaningful calculations of load contributions).

The NO₂₊₃ adjusted load trends are included in Table 2. The calculated adjusted load increases are generally greater than might be explained by the listed variables, but they agree in direction. The results suggest that the influence of decreasing snow cover on increasing January–February NO₂₊₃ exports is particularly strong. It appears to be greater than that of decreasing prior precipitation and streamflow, which makes more sense as an explanation of interannual variation. An exception to increasing NO₂₊₃ load is November–December in the Sandusky watershed. The fact that the load contributions are consistently lower than the total load change suggests that the model of Moog and Whiting (2002) is an incomplete description of NO₂₊₃ load variations. Owing to limitations on availability and accuracy of data, one or more important factors was not sufficiently represented to meet the criterion for significance as an explanatory variable.

The consistently highly significant downward trends in SRP loads (Table 3) lead to more numerous discernible changes and load contributions. The agreement in magnitude between the calculated contributions and adjusted load decreases is generally close. Decreases in phosphorus fertilizer deliveries and phosphorus from manure appear to explain the adjusted load decreases very well in 6 of 10 seasons in Table 3. Interestingly, these seasons are different in the two watersheds, so that each season is represented by fertilizer or manure reductions in at least one watershed. It is most likely that phosphorus source reductions were effective throughout the year in both watersheds. Conservation practices (tillage and reserve) appear in Table 2 under the Sandusky watershed. The results support the hypothesis that changes in agricultural practices may account for decreases in SRP load from the Maumee and Sandusky watersheds (Richards and Baker, 1993).

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
For 1976–1995, significant annual trends included a large decline in SRP load in both watersheds, and an increase in NO₂₊₃ load in the Maumee watershed. Adjustments for flow variation revealed significant decreases in TP load. Increasing flow in the Maumee River appears to have been responsible for amplifying the NO₂₊₃ load increase, negating what would have been a decrease in TP load, and turning what would have been a decrease in TSS load to an increase. It was estimated that roughly 40% of the flow-adjusted TP load decreases could be accounted for by decreases in the soluble component.

The most notable climatic changes occurred during the cold months: higher temperatures, increased precipitation and streamflow, and decreased snowfall, snow depth, and snow cover. These results imply increased wintertime rainfall and surface runoff, and decreased snowpack storage. March–April experienced decreased flows due to both decreased precipitation and decreased snowmelt.

Combining changes in climatic variables with changes in agricultural practices (Richards et al., 2002a) and relationships between loads and the explanatory variables (Moog and Whiting, 2002) suggested possible sources of changes in NO₂₊₃ and SRP loads over the 1976–1995 period. Increasing cold-season streamflows led to much greater increases in NO₂₊₃ and TSS exports during the winter, and were responsible for increasing TP in January–February. Nitrate export also may have been enhanced by decreasing snow cover and precipitation. Changes in agricultural practices were able to statistically explain decreases in soluble reactive phosphorus.

	ACKNOWLEDGMENTS

 
The support of the U.S. Department of Agriculture is gratefully acknowledged. Lynn Forster and Don Eckert are thanked for their reviews and insights. Agricultural data were compiled with the assistance of the School of Natural Resources at Ohio State University, in particular Phil Levison. Pete Richards of Heidelberg College helped to provide and process data. We would like to thank the reviewers for valuable contributions.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 

• Casey, G.D., D.N. Myers, D.P. Finnegan, and M.E. Wieczorek. 1997. National water-quality assessment of the Lake Erie–Lake St. Clair basin, Michigan, Indiana, Pennsylvania, and New York—Environmental and hydrologic setting. Water Resour. Investigations Rep. 97-4256. U.S. Geol. Survey Info. Serv., Denver, CO.
• Forster, D.L. 2002. Effects of conservation tillage on the performance of Lake Erie basin farms. J. Environ. Qual. 31:32–37 (this issue).[Abstract/Free Full Text]
• Gan, T.Y. 1998. Hydroclimatic trends and possible climatic warming in the Canadian Prairies. Water Resour. Res. 34:3009–3015.
• Hirsch, R.M., J.R. Slack, and R.A. Smith. 1982. Techniques of trend analysis for monthly water quality data. Water Resour. Res. 18:107–121.
• Karl, T.R., R.W. Knight, D.R. Easterling, and R.G. Quayle. 1996. Indices of climate change for the United States. Bull. Am. Meteor. Soc. 77:279–292.
• Kunkel, K.E., W.M. Wendland, and S.J. Vermette. 1994. Climate trends in Illinois. p. 7–89. In Illinois Dep. of Energy and Natural Resour. (ed.) The changing Illinois environment: Critical trends. Tech. Rep. Vol. 1: Air resources. ILENR/RE-EA-94/05(1). Illinois Dep. of Energy and Natural Resour., Springfield.
• Moog, D.B., and P.J. Whiting. 2002. Climatic and agricultural factors in nutrient exports from two watersheds in Ohio. J. Environ. Qual. 31:72–83 (this issue).[Abstract/Free Full Text]
• Myers, D.N., K.D. Metzker, and S. Davis. 2000a. Status and trends in suspended-sediment discharges, soil erosion, and conservation tillage in the Maumee River Basin—Ohio, Michigan, and Indiana. Water Resour. Investigations Rep. 00-4091. U.S. Geol. Survey Info. Serv., Denver, CO.
• Myers, D.N., M.A. Thomas, J.W. Frey, S.J. Rheaume, and D.T. Button. 2000b. Water quality in the Lake Erie–Lake St. Clair Drainages. Circ. 1203. U.S. Geol. Survey Info. Serv., Denver, CO.
• Richards, R.P., and D.B. Baker. 1993. Trends in nutrient and suspended sediment concentrations in Lake Erie tributaries, 1975–1990. J. Great Lakes Res. 19:200–211.
• Richards, R.P., and D.B. Baker. 2002. Trends in water quality in LEASEQ rivers and streams (northwestern Ohio), 1975–1995. J. Environ. Qual. 31:90–96 (this issue).[Abstract/Free Full Text]
• Richards, R.P., D.B. Baker, and D.J. Eckert. 2002a. Trends in agriculture in the LEASEQ watersheds, 1975–1995. J. Environ. Qual. 31:17–24 (this issue).[Abstract/Free Full Text]
• Richards, R.P., F.G. Calhoun, and G. Matisoff. 2002b. The Lake Erie Agricultural Systems for Environmental Quality project: An introduction. J. Environ. Qual. 31:6–16 (this issue).
• Salas, J.D. 1993. Analysis and modeling of hydrologic time series. p. 19.1–19.72. In D.R. Maidment (ed.) Handbook of hydrology. McGraw–Hill, New York.
• Stevenson, F.J. 1986. Cycles of soil: Carbon, nitrogen, phosphorus, sulfur, micronutrients. John Wiley & Sons, New York.

This article has been cited by other articles:

				D. Ginting and M. Mamo Measuring Runoff-Suspended Solids Using an Improved Turbidometer Method J. Environ. Qual., April 3, 2006; 35(3): 815 - 823. [Abstract] [Full Text] [PDF]

				P. L. Bishop, W. D. Hively, J. R. Stedinger, M. R. Rafferty, J. L. Lojpersberger, and J. A. Bloomfield Multivariate Analysis of Paired Watershed Data to Evaluate Agricultural Best Management Practice Effects on Stream Water Phosphorus J. Environ. Qual., May 11, 2005; 34(3): 1087 - 1101. [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]

				R. P. Richards, D. B. Baker, and D. J. Eckert Trends in Agriculture in the LEASEQ Watersheds, 1975-1995 J. Environ. Qual., January 1, 2002; 31(1): 17 - 24. [Abstract] [Full Text] [PDF]

				F. G. Calhoun, D. B. Baker, and B. K. Slater Soils, Water Quality, and Watershed Size: Interactions in the Maumee and Sandusky River Basins of Northwestern Ohio J. Environ. Qual., January 1, 2002; 31(1): 47 - 53. [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]

				R. P. Richards and D. B. Baker Trends in Water Quality in LEASEQ Rivers and Streams (Northwestern Ohio), 1975-1995 J. Environ. Qual., January 1, 2002; 31(1): 90 - 96. [Abstract] [Full Text] [PDF]

				D. B. Baker and R. P. Richards Phosphorus Budgets and Riverine Phosphorus Export in Northwestern Ohio Watersheds J. Environ. Qual., January 1, 2002; 31(1): 96 - 108. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (8) Citing Articles via Google Scholar Google Scholar Articles by Moog, D. B. Articles by Whiting, P. J. Search for Related Content PubMed PubMed Citation Articles by Moog, D. B. Articles by Whiting, P. J. Agricola Articles by Moog, D. B. Articles by Whiting, P. J. Related Collections Soil Conservation Surface Water Quality Watershed and Landscape Processes Agroclimatology Water Pollution