JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Nord, E. A.
Right arrow Articles by Lanyon, L. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Nord, E. A.
Right arrow Articles by Lanyon, L. E.
Agricola
Right arrow Articles by Nord, E. A.
Right arrow Articles by Lanyon, L. E.
Related Collections
Right arrow Watershed and Landscape Processes
Right arrow Ecosystem Management
Right arrow Nutrient Cycling
Right arrow Nutrient Management
Right arrow Animal Waste
Journal of Environmental Quality 32:562-570 (2003)
© 2003 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORTS
Landscape and Watershed Processes

Managing Material Transfer and Nutrient Flow in an Agricultural Watershed

E. A. Norda and L. E. Lanyon*,b

a Intercollege Program in Ecology and Dep. of Crop and Soil Sciences, The Pennsylvania State Univ., 116 ASI Building, University Park, PA 16802
b Dep. of Crop and Soil Sciences, The Pennsylvania State Univ., 116 ASI Building, University Park, PA 16802

* Corresponding author (lel{at}psu.edu)

Received for publication February 20, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 STUDY SITE
 METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Place-based resource management, such as watershed or ecosystem management, is being promoted to replace the media-focused approach for achieving water quality protection. We monitored the agricultural area of a 740-ha watershed to determine the nature and scale of farm material transfers, N and P balances, and farmer decisions that influenced them. Using field data and farmer interviews we found that 3 of 15 farms, emphasizing hog, dairy, or cash crops with poultry production, accounted for more than 80% of the inputs and outputs of N and P for the 362-ha agricultural area (332 ha of managed cropland and animal facilities). Feed for hogs (38% each of total N and P) and manure applied to fields as part of the cash crop and poultry operation (28 and 38% of total N and P, respectively) were the dominant inputs. No crops grown in the watershed were fed to animals in the watershed and more manure nutrients were applied from animals outside than from those in the watershed. A strategic decision by the hog farmer to begin marketing finished hogs changed the material transfers and nutrient balances more than tactical decisions by other farmers in allocating manure to cropland. Since the components of agricultural production were not all interconnected, the fundamental assumption of place-based management programs is not well-suited to this situation. Alternative approaches to managing the effect of agriculture on water quality should consider the organization of agricultural production and the role of strategic decisions in controlling farm nutrient balances.

Abbreviations: DP, dissolved orthophosphate phosphorus • MCRSW, Mahantango Creek Research Subwatershed • TN, total nitrogen • TP, total phosphorus


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
STUDY SITE
 METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
AGRICULTURE HAS BEEN identified as a major source of nutrients that impair water resources (USEPA, 1997). Eutrophication of lakes and the freshwater portions of estuaries and bays due to P runoff can lead to the loss of valuable fisheries and degradation of wildlife habitat (Carpenter et al., 1998). Pollution of ground water by NO3 leaching from cropland can lead to the loss of safe drinking water, since high concentrations of NO3 in drinking water may cause methemoglobinemia (also known as blue baby syndrome) (Robillard et al., 1991).

Because productive agriculture is fundamental to sustaining our society, reducing the effect of agriculture on the environment while meeting the array of needs and expectations of society is a challenge. The goal of nutrient management as a water quality protection activity is to ensure that the nutrient requirements of crops are met by balancing the nutrients that are already in the soil with those from manure, biosolids, and commercial fertilizers (USDA and USEPA, 1999). Voluntary landowner participation in developing and implementing nutrient management plans is anticipated as part of locally led conservation efforts (USDA and USEPA, 1999).

The assumptions that form the basis for this expected voluntary management response are sometimes not clear, resulting in lack of participation that can limit program effectiveness (van Es, 1983). Perceptions that anything other than crop requirement–based material applications are due to farmer mismanagement are often at the core of agency programs (USDA and USEPA, 1999; USEPA, 2001). However, nutrient loading to farms specializing in animal production may not be related to the needs of the crops at all (Bacon et al., 1990) and constraining nutrient applications to crop requirements can actually limit the economic success of animal-based farms (Westphal et al., 1989).

Outcomes of the hierarchy of farmer daily operational, annual tactical, and multiple-year strategic management decisions influence material transfers and nutrient flow with implications for potential farm effects on water quality (Lanyon and Beegle, 1993). In contrast to the full range of management actions, traditional best management practices focus primarily on modifying farm operations, such as manure storage and land application, or influencing tactical management, such as the development of nutrient management plans, with cost-sharing to offset implementation costs (USEPA, 2001; United States General Accounting Office, 1995). These operational and tactical activities use or allocate available farm resources. In contrast, the strategic or structural dimension of farms that deals with the acquisition of external resources for crop and animal production is rarely addressed by environmental protection programs. Yet, it may be farmer commitment to a particular production strategy, such as relying heavily on purchased or contracted feed for animals, that sets the constraints on within-farm management options to limit nutrient applications to crop requirements. This strategic commitment develops based on a broad range of external technological, social, and financial factors, not just within-farm considerations.

Government agencies, scientific groups, and other professionals with water quality protection interests and responsibilities are shifting their focus from media-based protection programs to advocacy of place- or landscape-based approaches, such as watershed management or ecosystem management. These latter approaches generally emphasize the role of local nutrient cycling among interconnected components of agricultural production systems as the means to achieve program goals (Committee on Watershed Management, 1999; USEPA, 1996; Christensen et al., 1996; Gonzalez, 1996). However, nutrient loading of landscape areas, such as watersheds or ecosystems, may not be related to the potential crop nutrient requirements in those areas anymore than it is for individual farms. Animal production on farms or in regions that is supported by imported feeds can result in manure production in excess of the local crop nutrient requirement (Lanyon, 1995; Lanyon and Thompson, 1996).

Understanding the interactions among agricultural production system components, including soils, plants, animals, and humans, has been described as essential for effective environmental management of landscape areas, whether watersheds (Reimhold, 1998) or ecosystems (Christensen et al., 1996). As residents, farmers, unlike nonresident government agency, scientific group, or other resource professionals, make many of the final land use decisions that determine the interactions among components of these areas. Therefore, in order for nonresident professionals to create effective place-based programs, they should understand the agricultural production systems and the array of factors influencing the hierarchy of decisions residents make that influence material transfers and the associated nutrient flows.

Tracking typical managed farm material transfers in watershed areas may provide some insight into the organization or structure of farms, the management dynamics that influence nutrient flows, and the scope of the agricultural production systems within which those decisions are made. Key steps in the monitoring process are determining appropriate reference boundaries for material transfers, measuring or estimating the inputs and outputs of nutrients, and determining the nutrient balances, or difference between inputs and outputs. Nutrient balance can be an indicator of potential nutrient losses in unmanaged flows from agriculture to surrounding water resources (Goss and Goorahoo, 1995; Frink, 1969; Lanyon and Beegle, 1989; Schröder et al., 1996).

This study examined the farm material transfers and estimated the associated flow of N and P to, from, and within in a small central Pennsylvania watershed with multiple farms. The objective was to monitor actual farmer-managed material transfers and to achieve some insight into the nature and scale of farmer decisions that influenced the transfers, nutrient balances, and potential effects on water quality.


   STUDY SITE
TOP
 ABSTRACT
 INTRODUCTION
 STUDY SITE
METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
This study focused on a 362-ha agricultural area with 332 ha of managed cropland near Leck Kill, PA (about 60 km north of Harrisburg, PA), located in a 740-ha subwatershed of the Mahantango Creek and the Susquehanna River basins (Fig. 1) . The Mahantango Creek Research Subwatershed (MCRSW, also referred to elsewhere as MCR [Pionke et al., 1996] or WE-38 [Pionke et al., 1999]) is on a south-facing slope within the Ridge and Valley Physiographic Province (Sevon, 1995). Elevations range from about 460 to 240 m above mean sea level. The climate is humid and temperate with precipitation averaging 1128 mm yr-1. Additional land uses within the MCRSW were approximately 37% forest, 9% permanent pasture, and 5% buildings and roads. All ground water recharge emerges as surface flow within the watershed (Pionke and Urban, 1985).



View larger version (43K):
[in this window]
[in a new window]
  		Fig. 1. Locations of the Mahantango Creek watershed within the Chesapeake Bay and the Susquehanna River watersheds (a), the Mahantango Creek Research Subwatershed (MCRSW) within the Mahantango Creek watershed (b), and the distribution of land uses in MCRSW (c).

 
The soil series in the MCRSW include: Hazleton (loamy-skeletal, siliceous, active, mesic Typic Dystrudept) and Clymer (fine-loamy, mixed, active, mesic Typic Hapludult) derived from a residual sandstone parent material along the northern watershed; Laidig (fine-loamy, siliceous, active, mesic Typic Fragiudult) and Meckesville (fine-loamy, mixed, active, mesic Typic Fragiudult) developed in a band of colluvial deposits along the footslope of the watershed ridge; and Berks and Calvin (both loamy-skeletal, mixed, active, mesic Typic Dystrudepts), Klinesville and Weikert (both loamy-skeletal, mixed, active, mesic Lithic Dystrudepts), and Leck Kill (fine-loamy, mixed, semiactive, mesic Typic Hapludult) derived from shale residuum in the valley where most of the agricultural area is located (Eckenrode, 1985).


   METHODS
TOP
 ABSTRACT
 INTRODUCTION
 STUDY SITE
 METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Detailed farm management data were collected during 1994, 1995, and 1996 by staff from the Pasture Systems and Watershed Management Research Laboratory (PSWMRL) of the USDA Agricultural Research Service (ARS) during annual structured personal interviews with each of the farmers in the area. They recorded specific management practices, including fertilizer and manure applications, and crop yields, for each of the fields. Corn (Zea mays L.); small grains including wheat (Triticum aestivum L.), barley (Hordeum vulgare L.), and oat (Avena sativa L.); soybean [Glycine max (L.) Merr.]; and alfalfa (Medicago sativa L.) were grown on more than 80% of the cropland. Fields or parts of fields from 15 farms were in the watershed. Total field area in the watershed per farm ranged from 2 to 80 ha. Additional farmer interviews were conducted as part of this project to collect animal production information. Agricultural commodity suppliers and buyers were interviewed about the materials they delivered to or purchased from the farmers to determine the sources or destinations of material inputs and outputs. The watershed surroundings were classified as either local (<=10 km) or distant (>10 km). Only those material transfers to, from, or within the MCRSW were quantified, although additional transfers were reported by farmers during the interviews.

Nutrient flows were estimated by the methods of Meisinger and Randall (1991)(with modifications from Lanyon and Beegle [1989] and Bacon et al. [1990]) using standard nutrient concentration values for the farm materials (Lanyon and Schlauder 1988; Anonymous, 1998). Annual nutrient flow to and from each group of similarly treated fields was calculated using the fields planted to each crop, crop yield, and fertilizer and manure application rates from the farmer interviews. Area of each of the 339 fields was calculated using Arc View GIS Version 3.0 (Environmental Systems Research Institute, 1997) after digitizing from an aerial photograph and the final total areas crosschecked with the farmer records. For fields that crossed the watershed boundary, only the portion within the watershed was included. Net nutrient flow from soil to crop was assumed to equal the nutrients removed in the harvested portion of the crop. Biological N fixation by legume crops, including alfalfa (10 ha yr-1) and soybean (65 ha yr-1), was estimated as 60% of the total crop N content (Peoples et al., 1995). No manure was applied directly to the legume crops. Soybean generally received fertilizer at 15 kg N ha-1 yr-1. Nutrients in animals sold and feed purchased were also estimated using standard values as described in Lanyon and Schlauder (1988). Nutrient throughflow was calculated as the nutrient outputs from a farm (that could include both crop fields and animal production units) divided by the nutrient inputs for that farm. Stream flow was monitored at a permanent weir by the PSWMRL staff with a continuous recorder and expressed as mm per unit of land area. They also collected water samples for NO3–N (Wendt, 1995) and dissolved orthophosphate phosphorus (DP) (Pritzlaff, 1996) analyses three times weekly. Additional stream water samples were collected at the outlets of 14 subwatersheds within the MCRSW in the spring of 2001. The subwatersheds varied in land use ranging from dominantly forested (87% land area) to dominantly cropped (77% land area). The concentration of total N and P was determined on unfiltered stream samples following digestion with a semimicro Kjeldahl procedure (Bremner and Mulvaney, 1982; Taylor, 2000). Nutrient export in stream discharge was estimated from flow volume and nutrient concentration as follows: T = {sum}(i = 1 to n)[F(i) x C(i) x P], where T is the total annual nutrient transport, n is the total number of samples per year, F(i) is the stream flow volume when the sample is collected, C(i) is nutrient concentration when sample i is collected, and P is the period of time represented by the sample equal to 365 divided by n. (Since the n samples were nearly evenly spaced, each sample was assumed to be representative of an equal period of time.) The relationship between total nitrogen (TN; total Kjeldahl N plus NO3–N) and NO3–N used in mass balance accounting for N in the watershed stream discharge was derived from measured concentrations for the MCRSW spring 2001 sampling [TN = (1.03 ± 0.02) x NO3–N]. Similarly, the relationship between total phosphorus (TP) and DP concentrations for the MCRSW spring 2001 sampling [TP = 3.55 ± 0.41) x DP] was used in mass balance accounting for P. These concentration relationships are consistent with reports for a small watershed in the lower Chesapeake Bay drainage basin (Koerkle et al., 1997) for TN to NO3–N ratio and for the Mahantango Creek (McDowell et al., 2001) for TP to DP ratio. Nitrate N and DP contributions from forests were estimated using concentrations measured at the outlet of forested subwatersheds during the MCRSW spring 2001 sampling (0.90 ± 0.08 mg L-1 NO3–N and 0.008 ± 0.0007 mg L-1 DP). These concentrations are consistent with other reports for the Mahantango Creek (Pionke and Urban, 1985). All calculations were performed using standard spreadsheet software.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 STUDY SITE
 METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Material Transfers and Nutrient Flows at the Watershed Level
The mean annual inputs of nutrients to the watershed in the managed material transfers exceeded the corresponding outputs (Fig. 2) . The majority of the managed nutrient inputs (66% of N and 78% of P) was associated with the transfers of animal feed and manure while the managed nutrient outputs were primarily in harvested crops (78% of N and 72% of P). The differences between the managed nutrient inputs and outputs, or the balances, averaged 35.2 Mg N yr-1 and 13.6 Mg P yr-1, resulting in an apparent net storage in the agricultural area (fields) of 106 kg N ha-1 yr-1 and 41 kg P ha-1 yr-1. These net nutrient inputs were offset by stream nutrient exports of 32.3 Mg N yr-1 and 0.7 Mg P yr-1. No MCRSW crops were fed to animals within the watershed because all crops either were sold or used outside the watershed. Of the 15 farms with fields in the MCRSW, three farms accounted for 60% of the cropland for the years 1994 through 1996, and 90 and 93% of the input N and P, respectively, and 80% of the output flows for both N and P. A detailed description of these farms illustrates the diversity of the agricultural production systems and the farmer decisions influencing material transfers and nutrient flow in this watershed during the period of the investigation.



View larger version (29K):
[in this window]
[in a new window]
  		Fig. 2. The annual mean flow of N and P in the material transfers within the Mahantango Creek Research Subwatershed (MCRSW) in central Pennsylvania and to and from the local (<=10 km) and distant (>10 km) surroundings for 1994 through 1996.

 
Farm H emphasized hog production with imported feed accounting for 93 and 99% of the N and P input, respectively, to the Farm H component of the MCRSW (Fig. 3a) . Hog feed for approximately 250 sows and market hogs was purchased from a feed mill in Camp Hill, PA (70 km). Industry information indicated that approximately 50% of the hog feed ingredients would typically be purchased in Pennsylvania with the remainder from Ohio, Indiana, and Illinois. Crops from Farm H were sold to a feed mill–processor in Lebanon, PA (50 km). These crops, accounting for 36 and 28% of the Farm H nitrogen and phosphorus output, respectively, were sold as feed for other animals in southeastern Pennsylvania, New York, and Vermont. In 1994 and 1995, 2000 feeder pigs yr-1 were sold (at 22 kg each) to finishing operations in southeastern Pennsylvania. Beginning in 1996 with the construction of a new feeding facility, all farrowed pigs were retained within the farm until reaching market weight (114 kg). This change in farm operation dramatically influenced the amount of imported feed and the associated nutrient flows (Table 1). Approximately 3300 market hogs yr-1 for the three-year period and cull sows were sold to a pork processor in Hatfield, PA (100 km). Outputs from the MCRSW in the animals marketed from Farm H were 51 and 46% of the total N and P farm exports, respectively. Manure from the hogs on Farm H was applied to crops on the 43 ha of Farm H cropland within the MCRSW in all three years (Fig. 1c) and on rented land outside the MCRSW in both 1995 and 1996. Manure application to rented land in the local surroundings represented 13 and 25% of the Farm H nitrogen and phosphorus outputs, respectively. Supplementary fertilizer, 7 and 1% of the total N and P input, respectively, was purchased for crop production from a local farm supply business in Pitman, PA (10 km). This business purchased fertilizer from distributors in Lancaster, PA (80 km) and Baltimore, MD (160 km). Nitrogen inputs from estimated biological N fixation were less than 5% of the total N inputs. Significant quantities of N and P were unaccounted for in the managed material transfers of this farm as reflected in the 47 and 35% throughflow for N and P, respectively, and the apparent change in nutrient stock of 17.2 and 5.1 Mg yr-1 for N and P, respectively. Several unexplored losses in the livestock facilities and during the manure applications, including gaseous N losses, were aggregated in these amounts. Manure N applied to fields can accumulate in relatively stable organic fractions or leach from the crop root zone when applied in excess of crop utilization potential. Nitrogen balance in crop fields can be a significant factor influencing the potential N loss from farms (Frink, 1969; Saporito and Lanyon, 1998). The apparent difference in P stock also reflects potential losses in the manure handling process as well as accumulations in soil reserves. Repeated manure applications in excess of crop utilization (for many agronomic crops approximately 20 to 30 kg P ha-1) often result in the buildup of soil P reserves (Mallarino et al., 1991; McCollum, 1991).



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 3. The annual mean flow of N and P in the material transfers for the individual dominant farms and the remaining group of other farms within the Mahantango Creek Research Subwatershed (MCRSW) in central Pennsylvania and to and from the local (<=10 km) and distant (>10 km) surroundings for 1994 through 1996. (a) Farm H; (b) Farm D; (c) Farm CP; and (d) other farms.

 

View this table:
[in this window]
[in a new window]
  		Table 1. Nutrient flows (TN, total nitrogen; TP, total phosphorus) and balances (managed inputs - managed outputs) for the farms, the total agricultural area, and the watershed stream discharge estimated for the agricultural area of the Mahantango Creek Research Subwatershed (MCRSW) in central Pennsylvania for 1994 to 1996.

 
Farm D was a dairy farm with approximately 40 ha of cropland in the MCRSW (Fig. 1c) and dairy facilities plus an additional 69 ha of cropland that supported a herd of 75 lactating cows and 30 dry cows plus heifers located approximately 0.25 km southeast of the MCRSW. The crops produced in the MCRSW provided about 35% of the feed for the dairy animals with the remainder of the locally produced feed from the area outside the MCRSW. Approximately 10% of the dairy feed was purchased, mostly as soybean oil meal, from a supplier in Mount Joy, PA (70 km). Milk was sold to a processor in Reading, PA (80 km), and cull cows to a meat packing operation in Souderton, PA (130 km). Manure from the dairy herd was spread on cropland with approximately 20% of the annual manure production spread within the MCRSW. Supplementary fertilizer, 27 and 50% of the total N and P input, respectively (Fig. 3b), was purchased from the same local farm supply business in Pitman, PA as Farm H. All Farm D MCRSW inputs were for the production of crops and 67 and 50% of the N and P, respectively, were dairy manure. Only about 6% of the N input was estimated to be from biological N fixation by legume crops. All Farm D managed MCRSW outputs were in harvested crops and delivered locally to the dairy facilities. Because of the conservative manure applications and the management of the agricultural land in the MCRSW as a crop operation, the change in nutrient stocks was an order of magnitude less for Farm D compared with the other dominant farms in the agricultural area. The throughflow was 83 and 50% for N and P, respectively. The difference in Farm D year-to-year crop allocation to corn production in the MCRSW and the dairy manure applications for corn explained the nutrient flows for the three years of the study (Table 1).

Farm CP was a cash crop farm with a poultry layer enterprise. The poultry house was located approximately 0.25 km west of the MCRSW with the 109 000 layers under contract to a company that provided the hens and the feed, and marketed the eggs. The contracting feed mill in Spring Glen, PA (15 km) purchased roughly 80% of the corn for feed in Pennsylvania and the remaining 20% from New York, and all soybean oil meal from Ohio, Indiana, and Illinois. Of the 425 ha of cultivated cropland on Farm CP, 115 ha were within the MCRSW (Fig. 1c). Poultry manure was spread on an average of 27 ha yr -1 in the MCRSW at a rate of 22.7 Mg ha-1, primarily to "build up" the fields as described by the farmer after a period of cash-crop production relying on inorganic fertilizers. This transfer of manure from the local surroundings accounted for 56 and 80% of the N and P input flows, respectively (Fig. 3c). About 12% of the N input was estimated to be from biological N fixation by legume crops. Supplementary fertilizer, 32 and 20% of the total Farm CP nitrogen and phosphorus input, respectively, was purchased for crop production from the same local farm supply business in Pitman, PA as Farm H. Crops were sold for animal feed elsewhere including central and southeastern Pennsylvania, New York, and Vermont, after processing by feed mills in Spring Glen or Lebanon, PA. Because of the relatively extensive CP crop area in the MCRSW, the apparent changes in nutrient stock for this farm were the largest, approximately 60% of the total MCRSW change, even though the unit area loadings of 176 and 65 kg ha-1 for N and P, respectively, were less than those estimated for Farm H. The throughflow was only 40 and 22% for N and P, respectively. Whereas the unaccounted for nutrients for Farm H could include those associated with the animal operation, the processes for the Farm CP component of the MCRSW could only be associated with the fate of nutrients following field applications. The area of corn in the watershed receiving poultry manure increased from 22.2 ha in 1994 and 1995 to 36.4 ha in 1996, even as the area of corn grown as part of Farm CP in the MCRSW decreased from 66 to 37 ha. This change in manure allocation between the areas of CP within and outside the MCRSW accounted for the increased watershed N and P inputs (Table 1). Although the Farm CP nitrogen and phosphorus surpluses in the MCRSW were greater in 1996 than in 1994, the difference was not as great as for Farm H.

The other farms in the watershed using 134 ha, or 40% of the agricultural land (Fig. 1c), were cash crop operations, purchasing fertilizer and selling crops (Fig. 3d). The sources of these fertilizers and destinations of these crops were not investigated in detail, but are not likely to be very different from the dominant farms of the watershed. No manure use was reported for these operations. They represent small fractions of the total nutrient inputs or outputs in the MCRSW and are not connected to the local surroundings of the watershed. The throughflow was approximately 105% for N and 81% for P. The apparent change in nutrient stock for these other farms that relied on purchased fertilizer for crop needs was less than for the other three farms.

Stream Discharge of Nutrients
The average nutrient concentrations in stream water were 5.4 (standard deviation = 1.7) mg L-1 NO3–N and 0.02 (standard deviation = 0.02) mg L-1 DP. Stream base flow was 55 L s-1 in the period 1984 through 1996 (Pionke et al., 1999). The average annual nutrient exports in the stream flow attributed to agriculture were equivalent to 97 kg ha-1 yr-1 TN and 2 kg ha-1 yr-1 TP. These nutrient exports (yields) were substantially greater than the yields for the Susquehanna River in normal flow years (3.1 kg ha-1 yr-1 TN and 0.1 kg ha-1 yr-1 TP; Langland et al., 1995). The TN yields also were greater than the yields for agricultural watersheds (>50%) with similar bedrock in the Chesapeake Bay drainage basin (20.4 kg ha-1 yr-1 TN; Langland et al., 1995), but the TP yields were about the same (2.2 kg ha-1 yr-1 TP; Langland et al., 1995).

Nitrogen in stream flow from agriculture was approximately 41% for NO3–N and 43% for TN of the managed inputs and greater than 90% of the difference between the managed inputs and outputs for both NO3–N and TN. Phosphorus in stream flow was a much smaller proportion of the managed inputs (<1% for DP and >3% for TP) than for N and less than 10% of the P surplus in the managed transfers as either DP or TP. Phosphorus is probably accumulating in the agricultural soils since Mehlich-3 P ranges from 80 to greater than 800 kg ha-1 in the MCRSW (Sharpley et al., 1999), while 67 kg ha-1 is the threshold for optimum crop production (Anonymous, 1998). Year-to-year differences in nutrient export in the stream flow were proportionately smaller for TN (±65%) than for TP (±200%) (Table 1). Extraordinarily high precipitation and stream flow of 1261 mm (about twice the mean flow for the period; Pionke et al., 1996) in 1996 was associated with the greatest nutrient export, especially P. Actually, an individual 50-yr return period storm in January 1996, with 698 mm of flow, resulted in nearly 65% of the TP export for the entire three years of this study, but only 16% of the TN. Farmer decisions on material flows may have relatively rapid, but short-duration impact on N losses (N balance/N output in streamflow = 1.1) because the shallow ground water system in the watershed is the dominant path of nitrate transport to the streams that responds to changes in land use intensity within months rather than years (Gburek and Folmar, 1999). The impacts of the decisions are likely to be different for P losses in surface flow due to the differential susceptibility of landscape areas to rainstorms of variable frequency and intensity (Sharpley et al., 1999). The consequences for P losses may persist longer into the future (P balance/P output in streamflow = 19.4) and be of greater magnitude and variability as the retained P continues to interact with the runoff event–driven surface flows.

Managing Material Transfers and Nutrient Flow
Decisions made by different farmers at different levels of the management decision hierarchy (Lanyon and Beegle, 1993) influenced material transfers and nutrient flow to, from, and within the MCRSW from 1994 through 1996. Tactical decisions were made by farmers from outside the MCRSW about allocating MCRSW cropland between corn and other crops and using animal manure or fertilizer for corn production. For most of these farmers, the crops grown and the nutrient inputs to the MCRSW were based on their typical management protocols for the crops produced and selections from their suite of usual inputs. Most of these farmers also managed additional cropland nearby, so goals to better balance nutrient applications with crop utilization in the MCRSW could readily be incorporated into farm operations. For instance, manure could be applied at lower rates to a greater area for corn production, such as for Farm CP, to reduce the soil P loads in the MCRSW. Manure storage to facilitate more timely application might be an option for Farm D to improve the efficiency of nutrient utilization. These practices to limit nutrient inputs to the MCRSW would not influence the fundamental organization of these farms. As a result, nutrient management recommendations could be compatible with voluntary implementation for enhanced water quality protection as suggested by proposed programs (USDA and USEPA, 1999). Nutrient management programming focused on perceived risks associated with new practices or reducing the costs of relevant management information, such as soil and manure testing or specialized consulting, might reduce barriers to operational and tactical practice adoption (VanDyke et al., 1999).

In contrast to the other managers of cropland areas in the MCRSW, the hog farmer (Farm H) made a strategic decision when he decided to change his operation from farrowing to farrow-to-finish that resulted in more intensive land use for animal production, but not crop production. The intensity of agricultural land use, including cattle and livestock density, can be a significant factor influencing water quality. The decision by the farmer to increase the livestock density in his operation was based on his intent to increase the value added to hog farrowing by finishing the feeder pigs. This was a response to the dynamic and increasingly competitive hog market for independent hog producers. It had little to do with the nutrient management tactics of the existing farm operation based on local soil and crop management, but more to do with global conditions of hog production. The ability of this farmer to respond successfully to recommendations for controls to protect water quality based on MCRSW conditions, soil nutrient stocks, or crop nutrient requirements may be much more limited than the other farmers operating in the MCRSW. Voluntary development and implementation of a nutrient management plan to accomplish water quality protection for the MCRSW is less likely to be as readily incorporated into a farm business plan with strategically defined structural barriers such as a commitment to heavy reliance on purchased or contracted feed for animals (Napier et al., 2000) than for the other farms. These barriers that deal with the organization of agricultural production systems in response to external factors are difficult to overcome with improved or different tactical and/or operational decisions and actions.

Whereas place-based resource management programs developed by nonresident professional managers focus on voluntary commitments from the resident agricultural managers, most farms, as determined for the MCRSW, are also nodes on commercial networks. It is through these non-nested, spatially distributed, place-independent networks spanning large distances (>100 km) that feed and fertilizers are delivered to locations like the MCRSW and that the crops and animal products are distributed. The commercial productive function and farm structure are not tied to the characteristics of the place as might be assumed by the place-based management approaches (Office of Technology Assessment, 1995). Therefore, the ability of some resident managers to comply with the functional expectations of natural resource management programs can conflict with the demands of the commercial network. Tellarini and Caporali (2000) observed that information from the economic market discouraged nutrient recycling and rewarded a farm that depended heavily on off-farm inputs. The strategic environment of the commercial network may also make it difficult for professional resource managers to achieve their place-based program goals. For instance, the Dutch determined that nutrient accounting by itself was not an effective tool for inducing the necessary reduction in manure surplus when application rates would decrease as a result and contribute to greater manure surpluses (Breembroek et al., 1996).


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 STUDY SITE
 METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
A small number of farms (20%) dominated the material transfers and accounted for more than 80% of the N and P associated nutrient flows in the agricultural area of the watershed. Feed for hogs and poultry manure applied to cropland were the dominant watershed inputs, but the outputs were largely in crops. Most of the crops were transported >10 km from the watershed. No crops produced in the watershed were fed to the animals in the watershed and less than 50% of the manure nutrients applied in the watershed were produced by animals there. The components of agricultural production—soil, crops, and animals—were not all interconnected in this watershed. Consequently, the fundamental assumption of place-based management programs to protect water quality that these connections exist and can be the basis of management recommendations does not correspond to the organization and function of these farms. The balance of N in the managed material transfers to and from the watershed was comparable with the N yield in the stream discharge while the P balance was more than 19 times greater than the corresponding P yield. The balances of N and P were determined by farmer decisions spanning the operational, tactical, and strategic management decision hierarchy, but a strategic animal production decision had the greatest impact. Understanding farms as nodes in spatially distributed commercial networks in addition to their readily observable place-based characteristics could improve the water quality management programs developed by professional resource managers.


   ACKNOWLEDGMENTS
 
This study would not have been possible without the field data collected by the staff and scientists of the USDA-ARS Pasture Systems and Watershed Management Regional Laboratory at University Park, PA and Klingerstown, PA. Assistance with the location figure was provided by the Land Analysis Laboratory, Department of Crop and Soil Sciences, The Pennsylvania State University. The farmers in the study area courteously answered many questions about their operations, and the agribusinesses shared significant information with us.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 STUDY SITE
 METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2003 32: 377-382. [Full Text]  



This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
N. A. Slaton, K. R. Brye, M. B. Daniels, T. C. Daniel, R. J. Norman, and D. M. Miller
Nutrient Input and Removal Trends for Agricultural Soils in Nine Geographic Regions in Arkansas
J. Environ. Qual., September 1, 2004; 33(5): 1606 - 1615.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Nord, E. A.
Right arrow Articles by Lanyon, L. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Nord, E. A.
Right arrow Articles by Lanyon, L. E.
Agricola
Right arrow Articles by Nord, E. A.
Right arrow Articles by Lanyon, L. E.
Related Collections
Right arrow Watershed and Landscape Processes
Right arrow Ecosystem Management
Right arrow Nutrient Cycling
Right arrow Nutrient Management
Right arrow Animal Waste

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
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome