Quantcast

Relationship of Land-Use/Land-Cover Patterns and Surface-Water Quality in the Mullica River Basin1

June 21, 2007

By Zampella, Robert A Procopio, Nicholas A; Lathrop, Richard G; Dow, Charles L

ABSTRACT: We describe relationships between pH, specific conductance, calcium, magnesium, chloride, sulfate, nitrogen, and phosphorus and land-use patterns in the Mullica River basin, a major New Jersey Pinelands watershed, and determine the thresholds at which significant changes in water quality occur. Nonpoint sources are the main contributors of pollutants to surface waters in the basin. Using multiple regression and water-quality data for 25 stream sites, we determine the percentage of variation in the water- quality data explained by urban land and upland agriculture and evaluate whether the proximity of these land uses influences water- quality/land-use relationships. We use a second, independently collected water-quality dataset to validate the statistical models. The multiple-regression results indicate that water-quality degradation in the study area is associated with basin-wide upland land uses, which are generally good predictors of water-quality conditions, and that both urban land and upland agriculture must be included in models to more fully describe the relationship between watershed disturbance and water quality. Including the proximity of land uses did not improve the relationship between land use and water quality. Ten-percent altered-land cover in a basin represents the threshold at which a significant deviation from reference-site water-quality conditions occurs in the Mullica River basin.

(KEY TERMS: New Jersey Pinelands; nonpoint-source pollution; urban land; upland agriculture; watersheds.)

INTRODUCTION

Numerous studies have assessed the relationship between land-use patterns in a watershed and water quality (Basnyat et al., 1999; Gburek and Folmar, 1999; Herlihy et al., 1998; Johnson et al., 1997; Jordan et al., 1997a, b; Osborne and Wiley, 1988; Rhodes et al., 2001; Tufford et al., 2003; among others). These studies focused primarily on agricultural or urban land uses. In urban areas, waterquality degradation may be associated with the extent of impervious area (Brown, 1988; Driver and Troutman, 1989; Arnold and Gibbons, 1996). Spatial patterns can also affect land-use/water- quality relationships, with the land use nearest a monitoring site rather than the aggregate land use over an entire watershed governing the relationships (Robinson et al., 1996; Basnyat et al., 1999; King et al., 2005).

Land-use-related watershed disturbances have a substantial effect on the water chemistry of streams in the New Jersey Pinelands. Streams draining forested watersheds are usually acidic and nutrient poor, whereas streams draining upland agriculture and developed lands display elevated pH and dissolvedsolid concentrations (Morgan and Good, 1988; Watt and Johnson, 1992; Zampella, 1994; Johnson and Watt, 1996; Dow and Zampella, 2000; Hunchak-Kariouk and Nicholson, 2001). The ecological consequences of watershed disturbance are significant. Water-quality degradation is associated with a shift from biological communities characterized by nativePinelands plant and fish species to those supporting both native and non-native species (Morgan and Philipp, 1986; Zampella and Laidig, 1997; Zampella and Bunnell, 1998).

In this study, we describe the relationship between water quality and land-use patterns in the Mullica River basin, a major New Jersey Pinelands watershed, and determine the thresholds at which significant changes in water quality occur. We also use multiple regression to determine which land-use class, urban land or upland agriculture, explains a greater percentage of the variation in water- quality data and to evaluate whether the proximity of these land uses influences water-quality/land-use relationships. A second, independently collected water-quality dataset is used to validate the statistical models.

METHODS

Study Area

The 1,474-km^sup 2^Mullica River basin (Figure 1), a coastal plain watershed, was the focus of our study. This major Pinelands watershed, which comprises several tributary systems, displays a range of natural and human-dominated landscapes. The unconfined Kirkwood-Cohansey aquifer system underlies the entire basin (Rhodehamel, 1979b; Zapecza, 1989; Johnson and Watt, 1996). The aquifer system consists of the Kirkwood Formation, which is composed primarily of sand, silt, clay, and some gravel, and the overlying Cohansey Sand, which is primarily quartz sand (Rhodehamel, 1979a). Ground-water discharge accounts for 89% of stream discharge in Pinelands streams (Rhodehamel, 1979b). With only one major point- source sewage discharge in the basin, nonpoint sources are the main contributors of pollutants to surface waters.

Streams selected for study ranged from reference streams that are minimally impacted by land-use-related disturbances to streams in basins that have been heavily altered by urban and upland agricultural land uses (Figures 1 and 2). The percentage of urban land and upland agriculture in a drainage basin ranged from

Water-Quality Data

We used water-quality data collected under baseflow conditions between 1995 and 1998 at 25 U.S. Geological Survey (USGS) water- quality monitoring sites to describe general land-use/water-quality relationships and develop regression models (Figure 1; Reed et al, 1997, 1998; DeLuca et al, 1999; Zampella et al., 2001). We calculated median values for selected water-quality variables, including pH, specific conductance, calcium, magnesium, chloride, sulfate, nitrite plus nitrate as nitrogen (NO^sub x^-N), ammonia as nitrogen, and total phosphorus.

To validate the regression models, we used waterquality data that we collected independently at 18 other Mullica River basin sites between November 1992 and October 1994 (Figure 1). Site locations and field and laboratory methods for the validation dataset are described in Dow (1996). Laboratory methods and sampling frequency varied among laboratories. Specific conductance and pH were measured at all 18 sites. Calcium, magnesium, and chloride were sampled at 14 of the 18 sites and sulfate was measured at only four sites.

Land-Use/Land-Cover Data

For each drainage basin, we determined the extent and position of aerial-photo-derived urban land, agriculture (upland agriculture), and agricultural wetlands using digital data obtained from the New Jersey Department of Environmental Protection (NJDEP, 1995/97 Land Use/Land Cover Update 2001). The NJDEP coverage, which uses a modified Anderson et al. ( 1976) system, was derived through visual interpretation of color infrared digital orthophotography with 1-m spatial resolution (acquired in March 1995) using a 1-acre (0.4-ha) minimum-mapping unit.

We used ArcView and Arclnfo software (ESRI, Inc., Redlands, CA) and a USGS Digital Elevation Model (DEM) to determine surface-water flow-path directions. The DEM consisted of individual 1:24,000scale, 7.5-min quadrangles composed of 30 ? 30-m grid cells or pixels that contained a terrain-elevation value. We assigned a flow-path direction (N, NE, E, SE, S, SW, W, or NW) to each grid cell in the DEM. Using the water-quality monitoring sites as downstream endpoints, the flow-path-direction grid cells were used to produce drainage basins for each site. A drainage basin included all of the pixels that contributed flow to a particular monitoring site. For each USGS monitoring site, we calculated the flow-path distance between the site and each grid cell in the associated drainage basin. Next, we cross-tabulated the flow-path distance data with the land-use data to determine the amount of drainage-basin area within each land-use class to create a set of basin-wide landuse variables.

To create a set of distance-weighted land-use variables, we summed the inverse of the flow-path distances for all cells in the basin classified as urban and divided this value by the sum of inverse distances of all cells in the basin. The same step was completed for cells classified as upland agriculture and agricultural wetlands. This transformation gives greater weight to land uses closer to a monitoring site. Both the basinwide and distance-weighted urban-land and uplandagriculture variables are expressed as a percentage.

General Land-Use/Water-Quality Relationships

We graphically analyzed the relationship between basin-wide altered land, composed of urban land and upland agriculture, and water quality. We did not graph ammonia or total phosphorus because both parameters were below detection limit at most sites. Neither of these two variables was analyzed statistically. For each of the other variables, we plotted median values against the percentage of altered land in a drainage basin.

Land- Use/Water-Quality Thresholds

We performed a separate analysis of variance (ANOVA) on ranks (Helsel and Hirsch, 1993) for each water-quality variable to assess differences among four altered-land (urban land and upland agriculture) categories. Because both upland agriculture and urban land were generally found co-occurring across our study basins, we did not separately evaluate individual thresholds for each land use. The altered-land categories were 50% altered land in a basin. No basins fell within the 20-29% range. Due to the small sample size (n = 2), we graphed values for basins falling within the 30-39% category, but did not include them in the analysis of variance. Post hoc tests were completed using the Tukey honest significant difference test for unequal sample sizes. Multiple-Regression Model Development

For model development, we used forward stepwise regression and the USGS dataset to relate median pH, specific conductance, calcium, magnesium, sulfate, and chloride values to the urban-land and upland- agriculture variables. We initially included agricultural wetlands in the models. Because this independent variable, which never exceeded 5% of any basin area, was not a significant predictor variable in any of models, we excluded it from the final analysis. Separate models were developed for the basin-wide and distance- weighted land-use variables. Except for pH, we performed the analyses on logtransformed water-quality data.

Due to closure among land-use variables, collinearity may introduce a bias when relating the percentage of a particular land use to water-quality characteristics (Barringer et al., 1990; King et al., 2005). We assessed the presence of multicollinearity among the independent land-use variables using the variance inflation factor (VIF) calculated for each variable. A VIF of

Model Validation

We used the multiple-regression models and land-use data to predict water-quality conditions at the model-validation sites. For each water-quality variable, we used the Wilcoxon matched-pairs test to compare the absolute residuals (observed minus predicted values) derived from the regression models to determine if the basin-wide and distance-weighted models produced significantly different results. We also used the Wilcoxon matched-pairs test to compare the observed validation-site values for each waterquality variable to the values predicted by the basinwide and distance-weighted models.

Nitrite Plus Nitrate as Nitrogen

We did not develop regression models for nitrite plus nitrate as nitrogen because median concentrations were below the detection limit at several sites. We used Spearman rank correlation to relate NO^sub x^-N concentrations to the percentage of basin-wide and distance-weighted urban land, upland agriculture, and altered land in a drainage basin.

Statistical Significance

An alpha level of 0.05 was used to assess significance for all tests. Statistical significance was determined after adjusting p- values for each set of related regressions, ANOVAS, Wilcoxon matched- pairs tests, and rank correlations using the sequential Bonferroni method (Rice, 1989, 1990). All statistical analyses were completed using STATISTICA for Windows (Statsoft Inc., Tulsa, OK, 1994).

RESULTS

General Land-Use/Water-Quality Relationships

Graphical analysis of the relationship between land use and water quality demonstrated that increases in pH, specific conductance, and dissolved solids are associated with an increase in the extent of urban and upland-agricultural lands in a drainage basin (Figure 2). Median NO^sub x^-N concentrations were below the 0.05 mg/l detection limit at eight sites. Median ammonia values were below detection limit (0.02 or 0.03 mg/l) at all but three stream sites, where the median concentration was 0.02 mg/l. Median total phosphorus concentrations equaled or exceeded the 0.01 mg/l detection limit at only five of the 25 streams. The highest total phosphorus concentrations were found in more heavily impacted streams, but neither this nutrient nor ammonia appeared to vary in relation to land-use intensity.

Land-Use/Water-Quality Thresholds

The ANOVAS comparing the four altered-land categories revealed that median specific conductance, pH, and calcium, magnesium, and chloride concentrations differed significantly between streams in basins in the 50% categories (Figure 3). Of special notice was the significant difference in these same five variables between the least-altered streams (

Multiple-Regression Models

Upland agriculture was correlated with urban land. However, the VIFs between the independent variables never exceeded 1.5 in any model, indicating that multicollinearity did not adversely affect the results. The basin-wide urban-land values were correlated with the distance-weighted values (r = 0.96). The basin-wide and distance- weighted upland-agriculture values were also correlated (r = 0.92).

Except for sulfate, the variability in water-quality characteristics explained by both the basin-wide and distance- weighted models was high (Table 1). All relationships were statistically significant at p

The variation in water quality explained by the basin-wide land- use models was consistently higher than that of the distance- weighted land-use models (Table 1). One major difference was a switch in major predictor variable in the specific conductance models. Upland agriculture rather than urban land was the significant predictor variable in the distance-weighted specific- conductance model.

Model Validation

Although R^sup 2^ values varied between the basin-wide and distance-weighted models, comparison of the absolute residuals produced using the model-validation data indicated that the two different models did not produce significantly different results for any water-quality variable (Table 2). Based on the Wilcoxon matched- pairs tests, we also found no significant difference between predicted and measured specific conductance, calcium, magnesium, chloride, and sulfate values for the model-validation sites (Table 2). The initial p-values for the basin-wide (p = 0.004) and distance- weighted (p = 0.011) pH analysis were significant, with only the former remaining significant following the Bonferroni adjustment. Nearly all of the measured pH values were lower than the predicted values, which was not the case for the other water-quality variables. The median difference in pH between the measured and predicted values was 0.37 and 0.41 units for the basin-wide and distance-weighted models, respectively. In both cases, the difference represented

Nitrite Plus Nitrate as Nitrogen

We found a significant correlation between NO^sub x^-N and each of the three land-use indices (Table 3). The strongest relationship was between this nutrient and total altered land, compared with urban land and upland agriculture. The results obtained using basinwide and distance-weighted land-use data were similar.

DISCUSSION

Our study demonstrates that water-quality degradation in the Mullica River basin is associated with upland land uses. Because upland agriculture was the major predictor variable in the calcium, magnesium, and sulfate models and contributed significantly to the pH, specific conductance, and chloride models, urban land cannot be used as the sole estimator of water-quality conditions. The regression analyses highlight the importance of including both urban land and upland agriculture in water-quality models. Inclusion of distance-weighted information did not improve the explanatory power of the regression models over that provided by raw basin-wide land- use percentages. The reasons for the lack of a flow-path distance signal are unclear but may have to do with the overall pattern of land use in the study basins where altered land was generally associated with the upper reaches of each basin. Basin ranks based on both raw basin-wide and distance-weighted land-use values were similar. Furthermore, our analysis did not distinguish between flow- path distances to the stream channel (associated primarily with groundwater flow) vs. stream-channel distances (associated with surface-water discharge) to the downstream monitoring station. Such a distinction may be needed to more conclusively determine if a distance-decay effect of upland land uses on downstream water quality exists.

Although we addressed multicollinearity and the proximity of land use in our analysis, we did not account for the possible effects of spatial autocorrelation and hydrologie connectivity (King et al., 2005). Because the urban and upland-agricultural landscapes are concentrated in the western portion of the Mullica River basin, a degree of spatial autocorrelation is expected. Furthermore, land- use and waterquality characteristics of upstream and downstream monitoring sites are not completely independent. In a Pinelands- wide study of pH and specific conductance, Dow and Zampella (2000) addressed hydrologie connectivity by using upstream and downstream sites for model development and validation, respectively. In the present study, deletion of downstream sites to deal with hydrologie connectivity was weighed against the effect of losing valuable information. Although the lack of complete independence among monitoring stations may be important statistically, from a practical viewpoint, the results of the validation exercise show that the regression models we developed can provide a fairly accurate characterization of water-quality conditions in the Mullica River basin streams. We cannot conclusively demonstrate cause and effect between land use and water quality, but there is sufficient evidence to effectively explain variations in stream chemistry in relation to land use. The results of the ANOVAS comparing water quality in basins with varying land-use profiles suggest that 10% altered-land cover in a basin represents a threshold at which a significant deviation from reference-site water-quality conditions occurs. Water quality also differed significantly between streams in basins with 10-19% altered land compared with streams draining basins with 40% or more altered land. The lack of streams with 20-29% altered land and the small number of streams in the 30-39% category prevented a more detailed characterization of thresholds.

Because elevated phosphorus and ammonia levels in Pinelands streams are usually associated with direct sewage discharges (Zampella, 1994), the general absence of point discharges in the study basins partly explains the low levels of both nutrients throughout the basin. Morgan and Good (1988), who also reported that ammonia and phosphorus did not appear in the Mullica River basin streams they studied, suggested that phosphorus may precipitate out as iron phosphate. Throughout the United States, measured phosphorus concentrations in streams do not agree with anticipated levels based on land-use practices (Smith et al., 1987). In contrast, elevated nitrate concentrations were associated with urban land and upland agriculture. Nonpoint sources of nitrites and nitrates include fertilizers applied on agricultural lands. Stackelberg et al. (1997) found that median nitrate as nitrogen concentrations in water samples drawn from shallow Kirkwood-Cohansey monitoring wells in undeveloped areas was 0.07 mg/l compared with median concentrations of 2.6, 3.5, and 13 mg/l for new urban areas, old urban areas, and agricultural land, respectively. Watt and Johnson (1992) and Szabo et al. (1997) also reported elevated nitrate concentrations in Kirkwood-Cohansey wells associated with agricultural land. Septic systems are another source of nitrates (Robertson et al., 1991; Steffy and Kilham, 2004). Based on a sample of 11 residential septic systems in the Pinelands, Bunnell et al. (1999) reported a mean total nitrogen concentration of 34 mg/1 at a depth of 1.2 m below the disposal bed, with nitrite plus nitrate as nitrogen comprising 94% of the total. Most ammonia was rapidly nitrified when discharged to the disposal beds.

Unlike other regions where geology is a major determinant of calcium and magnesium levels in streams (Patrick, 1996; Rhodes et al., 2001) and may confound the relationship between land use and water chemistry (Rhodes et al., 2001), weathering of sediments is not a significant source of dissolved solids in Pinelands streams. Although calcium and magnesium may be exported from Pinelands stream systems, concentrations of both ions are very low in unaltered watersheds (Yuretich et al., 1981; Morgan and Good, 1988). Land-use related watershed disturbance is the most likely explanation for significant variations in the concentrations of these constituents (Morgan and Good, 1988; Zampella, 1994). In the Pinelands, where elevated calcium and magnesium levels have been found in the Kirkwood-Cohansey aquifer beneath agricultural lands, liming associated with agricultural activities is one potential source of both base cations (Watt and Johnson, 1992; Johnson and Watt, 1996).

Marine aerosols are a source of chloride in Pinelands streams (Yuretich et al., 1981; Morgan and Good, 1988), but the results of our study suggest that land use overshadows the effect of atmospheric deposition. Urban land was the major predictor variable in the chloride models. The use of road salts in developed areas is a potential source of chloride in the associated drainage basins and may explain the patterns that we observed (Hay and Campbell, 1990; Robinson et al, 1996; Rhodes et al., 2001; Kaushal et al., 2005). Septic systems are another source of chloride (Bunnell et al., 1999), although the contribution from this source may be small compared with that of road salts (Kaushal et al., 2005). Herlihy et al. (1998), who showed that chloride concentrations in mid-Atlantic region streams were strongly related to land cover, suggested that chloride represents a surrogate for watershed disturbances associated with human activities. Our results contrast with those of Morgan and Good (1988), who concluded that watershed disturbance had no effect on chloride levels in Pinelands streams. This discrepancy is likely due to the narrow range of watershed conditions studied by Morgan and Good (1988).

The relationship between calcium, magnesium, sulfate, and chloride and altered land is reflected by parallel variations in specific conductance. Like chloride, a major contributor to specific conductance, specific conductance is most strongly associated with urban land. An increase in specific conductance in relation to land- use-related watershed disturbance has been demonstrated in all the major Pinelands watersheds (Dow and Zampella, 2000).

Undisturbed Pinelands streams have probably always been acidic, but acidity is currently controlled primarily by sulfate resulting from acid deposition (Kaufmann et al, 1988; Herlihy et al., 1991; Morgan, 1991). The positive relationship between sulfate concentrations and pH in Mullica River basin streams suggests that other processes are influencing pH in the more heavily altered basins. An increase in pH in degraded Pinelands waters may be related to enhanced primary productivity and nitrate assimilation (Morgan, 1985) or increases in calcium and magnesium, a relationship reported from other regions (Herlihy et al., 1998). However, the relationship between pH and both calcium and magnesium is confounded by our observation that urban land is the major predictor variable for pH, whereas agriculture explains a greater percentage of the variability in the two base cations. Szabo et al. (1997) suggested that nitrification of ammonia-based fertilizers generates enough hydrogen ions to counteract the acidneutralizing effect of lime, which is counter to the positive association of pH and both cations.

CONCLUSION

Water-quality degradation in the Mullica River basin is associated with basin-wide upland land uses, which are generally good predictors of water-quality conditions. Our results indicate that both urban land and upland agriculture must be included in models describing the relationship between watershed disturbance and water quality. Including the proximity of land uses did not improve predictive ability, indicating that the simplest approach represents an effective predictive tool. Ten-percent altered-land cover in a basin represents the threshold at which a significant deviation from reference-site water-quality conditions occurs in the Mullica River basin.

ACKNOWLEDGMENTS

Funding for this study was provided by the U.S. Environmental Protection Agency, the National Park Service, and the Pinelands Commission. Parts of this paper are based on reports submitted to these institutions. Eighteen of the 25 monitoring sites were operated through Pinelands Commission – U.S. Geological Survey cooperative agreements with funding from the National Park Service and the Camden County Municipal Utilities Authority. We thank John Bognar, who assisted with the generation of some geographic information system data.

Zampella, Robert A., Nicholas A. Procopio, Richard G. Lathrop, and Charles L. Dow, 2007. Relationship of Land-Use/Land-Cover Patterns and Surface-Water Quality in the Mullica River Basin. Journal of the American Water Resources Association (JAWRA) 43(3):594-604. DOI: 10.1111/j.1752-1688.2007.00045.x

1 Paper No. J05110 of the Journal of the American Water Resources Association (JAWRA). Received July 29, 2005; accepted August 1, 2006. (c) 2007 American Water Resources Association.

LITERATURE CITED

Andersen, J.R., E.E. Hardy, J.T. Roach, and R.E. Witmer, 1976. A Land Use and Land Cover Classification System for Use With Remote Sensor Data. U.S. Geological Survey Professional Paper 964.

Arnold, C.L. and C.J. Gibbons, 1996. Impervious Surface Coverage: The Emergence of a Key Environmental Indicator. Journal of the American Planning Association 62:243-258.

Barringer, T., D. Dunn, W. Battaglin, and E. Vowinkel, 1990. Problems and Methods Involved in Relating Land Use to GroundWater Quality. Water Resources Bulletin 26:1-9.

Basnyat, P., L.D. Teeter, K.M. Flynn, and B.G. Lockaby, 1999. Relationships Between Landscape Characteristics and Nonpoint Source Pollution Inputs to Coastal Estuaries. Environmental Management 23:539-549.

Brown, R.G., 1988. Effects of Precipitation and Land Use on Storm Runoff. Water Resources Bulletin 24:421-426.

Bunnell, J.F., R.A. Zampella, M.D. Morgan, and D.M. Gray, 1999. A Comparison of Nitrogen Removal by Subsurface Pressure Dosing and Standard Septic Systems in Sandy Soils. Journal of Environmental Management 56:209-219.

DeLuca, M.J., J.H. Oden, K.M. Romanok, and M.L. Riskin, 1999. Water Resources Data, New Jersey, Water Year 1998, Volume 3. Water- Quality Data. U.S. Geological Survey Water-Data Report NJ-98-3.

Dow, C.L., 1996. A Summary of Pinelands Surface Water Quality Data for Atlantic County, Burlington County, and Cape May County. Pinelands Commission, New Lisbon, New Jersey. Dow, C.L. and R.A. Zampella, 2000. Specific Conductance and pH as Indicators of Watershed Disturbance in Streams of the New Jersey Pinelands, USA. Environmental Management 26: 437-445.

Driver, N.B. and B.M. Troutman, 1989. Regression Models for Estimating Urban Storm-Runoff Quality and Quantity in the United States. Journal of Hydrology 109:221-236.

Gburek, W.J. and G.J. Folmar, 1999. Flow and Chemical Contributions to Streamflow in an Upland Watershed: A Baseflow Survey. Journal of Hydrology 217:1-18.

Hay, L.E. and J.P. Campbell, 1990. Water Quality Trends in New Jersey Streams. U.S. Geological Survey Water-Resource Investigations Report 90-4046.

Helsel, D.R. and R.M. Hirsch, 1993. Statistical Methods in Water Resources. Studies in Environmental Science 49. Elsevier, Amsterdam, the Netherlands.

Herlihy, A.T., P.R. Kaufmann, and M.E. Mitch, 1991. Stream Chemistry in the Eastern United States 2. Current Sources of Acidity in Acidic and Low Acid-Neutralizing Capacity Streams. Water Resources Research 27:629-642.

Herlihy, A.T., J.L. Stoddard, and C.B. Johnson, 1998. The Relationship Between Stream Chemistry and Watershed Land Cover Data in the Mid-Atlantic Region. U.S. Water, Air, and Soil Pollution 105:377-386.

Hunchak-Kariouk, K. and R.S. Nicholson, 2001. Watershed Contributions of Nutrients and Other Nonpoint Source Contaminants to the Barnegat Bay-Little Egg Harbor Estuary. Journal of Coastal Resources 32:28-81.

Johnson, M.L. and M.K. Watt, 1996. Hydrology of the Unconfined Aquifer System, Mullica River Basin, New Jersey, 1991-1992. U.S. Geological Survey Water-Resources Investigations Report 94-4234.

Johnson, L.B., C. Richards, G.E. Host, and J.W. Arthur, 1997. Landscape Influences on Water Chemistry in Midwestern Stream Ecosystems. Freshwater Biology 37:193-208.

Jordan, T.E., D.L. Correll, and D.E. Weller, 1997a. Effects of Agriculture on Discharges of Nutrients From Coastal Plain Watershed of Chesapeake Bay. Journal of Environmental Quality 26:836-848.

Jordan, T.E., D.L. Correll, and D.E. Weller, 1997b. Relating Nutrient Discharges From Watersheds to Land Use and Streamflow Variability. Water Resources Research 33:2579-2590.

Kaufmann, P.R., A.T. Herlihy, J.W. Elwood, M.E. Mitch, W.S. Overton, M.J. Sale, J.J. Messer, K.A. Cougan, D.V. Peck, K.H. Reckhow, A.J. Kinney, S.J. Christie, D.D. Brown, C.A. Hagley, and H.I. Jager, 1988. Chemical Characteristics of Streams in the Mid- Atlantic and Southeastern United States. Volume I: Population Descriptions and Physico-Chemical Relationships. EPA/600/3-88/012a. U.S. Environmental Protection Agency, Washington, District of Columbia.

Kaushal, S.S., P.M. Groffman, G.E. Likens, K.T. Belt, W.P. Stack, V.R. Kelly, L.E. Band, and G.T. Fisher, 2005. Increased Salinization of Fresh Water in the Northeastern United States. Proceedings of the National Academy of Sciences 102:13517-13520.

King, R.S., M.E. Baker, D.F. Whigham, D.E. Weller, T.E. Jordon, P.F. Kazyak, and M.K. Kurd, 2005. Spatial Considerations for Linking Watershed Land Cover to Ecological Indicators in Streams. Ecological Applications 15:137-153.

Morgan, M.D., 1985. Photosynthetically Elevated pH in Acid Waters With High Nutrient Content and Its Significance for the Zooplankton Community. Hydrobiologia 128:239-247.

Morgan, M.D., 1991. Sources of Stream Acidity in the New Jersey Pinelands. Verhandlungen Internationale Vereinigung Limnologie 24:1707-1710.

Morgan, M.D. and R.E. Good, 1988. Stream Chemistry in the New Jersey Pinelands: The Influence of Precipitation and Watershed Disturbance. Water Resources Research 24:1091-1100.

Morgan, M.D. and K.R. Philipp, 1986. The Effect of Agricultural and Residential Development on Aquatic Macrophytes in the New Jersey Pine Barrens. Biological Conservation 35:143-158.

Myers, R.H., 1990. Classical and Modern Regression With Applications. PWS-Kent Publishing Co., Boston, Massachusetts.

Osborne, L.L. and M.J. Wiley, 1988. Empirical Relationships Between Land Use/Cover and Stream Water Quality in an Agricultural Watershed. Journal of Environmental Management 26:9-27.

Patrick, R., 1996. Rivers of the United States of America. Volume III: The Eastern and Southeastern States. John Wiley and Sons, Inc., New York, New York.

Reed, T.J., G.L. Centinaro, M.J. DeLuca, J.T. Hutchinson, and J. Scudder, 1997. Water Resources Data, New Jersey, Water Year 1996, Volume 1, Surface-Water Data. U.S. Geological Survey Water-Data Report NJ-96-1.

Reed, T.J., G.L. Centinaro, M.J. DeLuca, and J.H. Olden, 1998. Water Resources Data, New Jersey, Water Year 1997, Volume 1, Surface- Water Data. U.S. Geological Survey Water-Data Report NJ-97-1.

Rhodehamel, B.C., 1979a. Geology of the Pine Barrens of New Jersey. In: Pine Barrens: Ecosystem and Landscape, R.T.T. Forman (Editor). Academic Press, New York, New York, pp. 39-60.

Rhodehamel, E.G., 1979b. Hydrology of the New Jersey Pine Barrens. In: Pine Barrens: Ecosystem and Landscape, R.T.T. Forman (Editor). Academic Press, New York, New York, pp. 147-167.

Rhodes, A.L., R.M. Newton, and A. Pufall, 2001. Influences of Land Use on Water Quality of a Diverse New England Watershed. Environmental Science and Technology 35:3640-3645.

Rice, W.R., 1989. Analyzing Tables of Statistical Tests. Evolution 43:223-225.

Rice, W.R., 1990. A Consensus Combined P-Value Test and Family- Wide Significance of Component Tests. Biometrics 46:303-308.

Robertson, W.D., J.A. Cherry, and E.A. Sudicky, 1991. GroundWater Contamination From Two Small Septic Systems on Sand Aquifers. Ground Water 29:82-92.

Robinson, K.W., T.R. Lazaro, and C. Pak, 1996. Associations Between Water-Quality Trends in New Jersey Streams and Drainage- Basin Characteristics, 1975-86. U.S. Geological Survey Water- Resources Investigations Report 96-4119.

Smith, R.A., R.B. Alexander, and M.G. Wolman, 1987. Water- Quality Trends in the Nation’s Rivers. Science 235:1607-1615.

Stackelberg, P.E., J.A. Hopple, and L.J. KaufFman, 1997. Occurrence of Nitrate, Pesticides, and Volatile Organic Compounds in the Kirkwood-Cohansey Aquifer System, Southern New Jersey. U.S. Geological Survey Water-Resource Investigation Report 97-4241.

Steffy, L.Y. and S.S. Kilham, 2004. Elevated delta^sup 15^N in Stream Biota in Areas With Septic Tank Systems in an Urban Watershed. Ecological Applications 14:637-641.

Szabo, Z., D.E. Rice, C.L. MacLeod, and T.H. Barringer, 1997. Relation of Distribution of Radium, Nitrate, and Pesticides to Agricultural Land Use and Depth, Kirkwood-Cohansey Aquifer System, New Jersey Coastal Plain, 1990-1991. U.S. Geological Survey Water- Resources Investigations Report 96-4165A.

Tufford, D.L., C.L. Samarghitan, H.N. McKellar Jr., D.E. Porter, and J.R. Hussey, 2003. Impacts of Urbanization on Nutrient Concentrations in Small Southeastern Coastal Streams. Journal of the American Water Resources Association 39:301-312.

Watt, M.K. and M.L. Johnson, 1992. Water Resources of the Unconfined Aquifer System of the Great Egg Harbor River Basin, New Jersey, 1989-90. U.S. Geological Survey WaterResources Investigations Report 91-4126.

Yuretich, R.F., D.A. Crerar, D.J. Kinsman, J.L. Means, and M.P. Borcsik, 1981. Hydrogeochemistry of the New Jersey Coastal Plain. 1. Major-Element Cycles in Precipitation and River Water. Chemical Geology 33:1-21.

Zampella, R.A., 1994. Characterization of Surface Water Quality Along a Watershed Disturbance Gradient. Water Resources Bulletin 30:605-611.

Zampella, R.A. and J.F. Bunnell, 1998. Use of Reference-Site Fish Assemblages to Assess Aquatic Degradation in Pinelands Streams. Ecological Applications 8:645-658.

Zampella, R.A. and K.J. Laidig, 1997. Effect of Watershed Disturbance on Pinelands Stream Vegetation. Journal of the Torrey Botanical Society 124:52-66.

Zampella, R.A., J.F. Bunnell, K.J. Laidig, and C.L. Dow, 2001. The Mullica River Basin: A Report to the Pinelands Commission on the Status of the Landscape and Selected Aquatic and Wetland Resources. Pinelands Commission, New Lisbon, New Jersey.

Zapecza, O.S., 1989. Hydrogeologie Framework of the New Jersey Coastal Plain. U.S. Geological Survey Professional Paper 1404-B.

Robert A. Zampella, Nicholas A. Procopio, Richard G. Lathrop, and Charles L. Dow2

2 Respectively, Chief Scientist and Research Scientist, Pinelands Commission, PO Box 7, New Lisbon, New Jersey 08064; Director, Grant F. Wallon Center for Remote Sensing and Spatial Analysis, Rutgers, the State University of New Jersey, 14 College Farm Road, New Brunswick, New Jersey 08091; and Director of Information Services, Stroud Water Research Center, 970 Spencer Road, Avondale, Pennsylvania 19311 (E-Mail/Zampella: Robert.Zampella@njpines.state.nj.us).

Copyright American Water Resources Association Jun 2007

(c) 2007 Journal of the American Water Resources Association. Provided by ProQuest Information and Learning. All rights Reserved.




comments powered by Disqus