| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Soil Science Department, North Carolina State University, Campus Box 7619, Raleigh, NC 27695-7619
b Crop and Soil Environmental Science Department, Virginia Polytechnic Institute and State University, 330 Smyth Hall, Blacksburg, VA 24061
* Corresponding author (amjohns2{at}unity.ncsu.edu)
Received for publication January 19, 2005.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: PLAT, Phosphorus Loss Assessment Tool • STP, soil test phosphorus
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
The revision of NRCS Standard 590 requires that sites receiving organic by-products or fertilized fields that are located in sensitive watersheds be assessed for P loss potential. Regulations for poultry operations utilizing dry litter are not tied to state standards. Based on the revised standard, all livestock operations in North Carolina that apply liquid animal waste, receive federal or state cost-share monies, or meet federal Confined Animal Feeding Operation standards (USEPA, 2003) will be required to use a P index known as the North Carolina Phosphorus Loss Assessment Tool (PLAT).
Since this P index will affect thousands of producers in North Carolina, a quantitative and easy-to-use tool was developed to provide a more comprehensive risk assessment of potential P loss than STP alone. Thus, the specific objective of this study was to estimate the percentage and types of farms that will be required to use P-based nutrient management as a result of new regulations requiring the use of PLAT. We have no actual field data quantitatively measuring P loss with which to compare to PLAT's predictions and, therefore, cannot perform a validation of the index. However, those sites with high STP levels or that occur in high animal density areas can be compared to those sites which PLAT identifies as being high risk to determine if PLAT is correctly identifying suspected problem areas throughout the state.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
A total of 1379 fields in the state were sampled. The maximum number of fields sampled from a given county was 36, the minimum number was 5, and the mean number sampled was 16. A list of farms in each county was obtained from county extension personnel or from the Farm Service Agency from which twice the number of farms needed to fulfill the sampling requirement in a given county were randomly selected. On a chosen farm, only one field was used unless the farm included both crop and pasture–hay fields; in the latter case, one field of each type was sampled. If the random selection of a farm produced a landowner who was not farming or who was unwilling to participate, the next farm on the list was used. Once a farm was chosen, the individual field closest to the geographic center of the county, according to digital orthoquads, was selected for sampling to ensure that soil samples were obtained without bias toward fields closest to the manure source.
Data collected from each field included: inputs used in the Revised Universal Soil Loss Equation (RUSLE) (vegetative cover type, cover condition, slope variables, and soil mapping unit); information on the presence of drainage tiles or ditches, and the depth of and distance between tile lines or ditches; water control structures; farm ponds; and buffers including buffer width. Length of depositional area (receiving slope), tillage, types and amounts of P application, and timing and method of application were also recorded. In addition, a composite soil sample was collected at the 0- to 20-cm depth and analyzed for Mehlich-3 P at the North Carolina Department of Agriculture and Consumer Services Soil Testing Lab. When analysis of a surface sample indicated the need for a subsurface sample (see description of PLAT below), a composite sample from the 70- to 80-cm soil depth was taken from the same site for measurement of Mehlich-3 P. Soil mapping units were used to obtain information on clay percentage, curve number, and transmissivity of the sampled soil and to determine classification into one of the four P threshold groups (organic, sand, loam, or clay). Data for each field were entered into a computerized version of PLAT and an estimation of potential P loss calculated.
Phosphorus Loss Assessment Tool
The PLAT index identifies four P loss pathways: (i) loss via erosion (sediment P); (ii) loss via surface runoff (dissolved P); (iii) loss via subsurface drainage (dissolved P); and (iv) loss via applied P source (particulate and dissolved P from applied sources). Source and transport factors are evaluated within each loss pathway in a multiplicative manner and results of the four parts are summed to estimate a relative amount of total P being delivered off-field. Although the tool estimates P loss in kg P ha–1 yr–1, it is unlikely to be an accurate approximation of actual P flux leaving a field and is only a relative indication of potential P loss. This allows the user to make comparative estimates of risk between different management practices and field conditions. Estimated values of P loss are converted to index values in PLAT by multiplying kg P ha–1 by 22.3 (25.0 when using lb P acre–1). This value was chosen to scale index ratings (when on lb P acre–1 basis), similar to the state soil test indexing system. Index values obtained from PLAT are then used to rate the potential for edge-of-field P loss by classifying them into one of four risk categories (Table 1).
|
The loss of sediment-bound P is a function of the amount of soil leaving the field through erosion, the total amount of P that is attached to those eroded soil particles, and any sediment redeposition that occurs within the field. As has been done in most states' P indices, PLAT determines erosion using RUSLE (Version 1.4) (Renard et al., 1991). The total P content of the eroded sediment is estimated by employing an exponential relationship between the clay content and STP of the soil (Cox, 1994). Multiplying the erosion rate by the total P factor yields an estimate of the particulate P that could potentially leave the field. Credits, or reductions in the amount of eroded material leaving the field, can be achieved through the presence of depositional areas, vegetative buffers, water control structures, sediment basins, or farm ponds.
Loss of P through surface runoff is predicted by considering the amount of STP that is likely to be dissolved in runoff water and the volume of runoff that occurs at a site. Determining runoff concentration is done by assigning a soil to one of four "P threshold groups" based on the assumption that soils have differing abilities to retain P. A "P threshold" is defined as the STP value at which 1 mg L–1 of dissolved P is expected to be lost to the soil solution (i.e., runoff). Four threshold soil groups are defined and, with their respective STP threshold value, are: clays, 500 mg kg–1; loams, 200 mg kg–1; sands, 100 mg kg–1; and organics, 50 mg kg–1. In general, assignment to one of the four threshold groups is based on soil texture, drainage, and depth to the Bt horizon, but for simplicity the groups are referred to by their generalized texture. Grouping all state soils into only four threshold groups is understood to be an oversimplification of the P retention characteristics of all soils, but it is thought to be a better approach than assuming all soils behave identically with regard to the relationship between STP and P solubility. A unique equation has been developed for each threshold group based on studies done on North Carolina soils and is used to estimate the amount of dissolved P for a given STP (Woodruff, 1963; Daughtrey, 1970; Novias, 1977; Cox, 1994; Cox and Hendricks, 2000; Kamprath, unpublished data).
Concentration of P in subsurface drainage is only considered in specific situations. Two conditions must be met in order for a subsoil sample to be required. First, STP of the surface (0–20 cm) soil must exceed the P threshold (50, 100, 200, or 500 mg kg–1) for that particular soil group (organic, sand, loam, or clay). If this does not occur, it is assumed that the surface soil is still able to retain a significant amount of P and very little dissolved P will reach the subsoil. Second, the soil must be designated as one prone to the loss of dissolved P through leaching. For certain types of soil with highly developed Bt horizons it is assumed that even if the surface is P saturated, a significant amount of dissolved P will not move downward due to the relatively high sorption capacity of the fine-textured Bt horizon. If, however, the soil surface is P-saturated, as indicated by the exceedance of the P threshold, and the soil type is not a well-drained soil with a stable Bt, than a subsoil sample is required and STP determined.
Estimates of surface runoff and subsurface drainage volumes are determined differently, depending on the drainage condition of a site. On well-drained soils that do not require improved drainage, surface runoff is determined using a modification of the SCS curve number method (USDA-SCS, 1985) and long-term rainfall data from each county. To estimate subsurface drainage volume on naturally drained soils a mass balance approach is used. In this case, average subsurface drainage is considered to be the volume of water after accounting for precipitation, runoff (as calculated for the surface runoff estimation), and evapotranspiration. Evapotranspiration is estimated for either pasture or row crop, using simulations with the GLEAMS model (Leonard et al., 1987) and long-term weather data. For poorly drained soils that require artificial drainage for crop production, a modification of the hydrologic model DRAINMOD, a water-balance model that partitions rainfall into the various hydrologic components of shallow water-table soils, is employed to predict the amount of surface runoff and subsurface drainage (Skaggs et al., 2004).
Predictions of P loss due to applied P source are made using the P content, application rate, method and timing of P application of the applied P source, and an estimation of rainfall that is moving over the soil surface as runoff, the "runoff fraction". Additionally, the total P of a given animal waste type is partitioned into liquid and solid fractions, each with its own attenuation factor according to waste type. Three categories of waste type exist: fresh manure/litter/scraped, sludge/slurry, and liquid. A P source such as lagoon liquid will have a relatively small amount of applied source P loss predicted by PLAT due to its high solubility and tendency to infiltrate the soil rapidly, whereas poultry litter, with its greater nonsoluble P content, will be more likely to remain on the surface and provide a longer-term source of P. Inorganic P fertilizer is assumed to be 100% soluble.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
Magnitude of PLAT Components
The mean amount of total P loss for our sample fields as estimated by PLAT was 0.72 kg P ha–1 which ranks as "low" when converted to an index value (Table 2). Predicted P losses for individual components of PLAT averaged 0.11 kg P ha–1 for erosion, 0.40 kg P ha–1 for surface runoff, 0.06 kg P ha–1 for subsurface drainage, and 0.18 kg P ha–1 for applied P source. It is important to be aware of the differences between mean and median values in Table 2. In the case of almost every parameter, the mean value is greater then the median indicating a nonsymmetrical distribution of sampled values. In general, a few sites with very high parameter values are skewing the means upward. However, both the mean and median PLAT rating for the 1379 samples are in the "low" risk category. Because the main goal of Tables 3–6 is to compare PLAT predictions on different soils, landscapes, and/or managements, only the means are reported in these tables.
|
|
|
|
|
Both the mean and median STP for the 1379 fields (114 mg kg–1) exceeded the 53 mg kg–1 considered as optimum for crop production by the state soil-testing lab. In the range of STP values for which additions of fertilizer or manure P applications would be recommended (0–53 mg kg–1), the majority of sites received a PLAT rating of "low", with less than 1% of the sites rating "high" or "very high" (Table 3). When STP increased to levels above which a crop response would not be expected, the percentage of fields that fell into risk categories requiring P-based management increased; 31% of the fields having STP levels of 200 mg kg–1 or greater exhibited a "high" or "very high" potential to contribute P to surface waters. When STP was 200 mg kg–1 or greater, 40% of sites were still predicted to have a "low" potential for P loss, indicating the importance of transport factors in PLAT. Despite a large contribution of P from either soil or manure, if the mechanisms to transport it as either dissolved or particulate P do not exist, it theoretically cannot present a threat to water quality.
The mean runoff P concentration as estimated by PLAT was approximately 0.5 mg L–1, a concentration at which eutrophication has been shown to occur (Sims et al., 1998). No best management practices (BMPs) exist for the control of surface runoff other than reducing STP or reducing the amount of runoff by improving soil cover conditions, thereby maximizing infiltration. Reducing STP by discontinuing P applications and optimizing crop removal of residual soil P has been shown to require many years (Kamprath, 1999). Kamprath found that, with no P fertilization, it took 26 yr for the STP of a sandy Coastal Plain soil with an initial Mehlich-3 STP value of 154 mg kg–1 to decrease to the critical agronomic STP of 53 mg kg–1.
Differences in P loss via surface runoff and subsurface drainage for different drainage situations are illustrated in Table 2. Greater volumes of both surface runoff and subsurface drainage are predicted under artificial drainage compared to natural drainage. This result is somewhat counterintuitive as surface runoff generally decreases and subsurface drainage increases when artificial drainage is implemented (Skaggs et al., 1980). The different methods used by PLAT to estimate volumes of water moving over and through the soil are apparently biasing prediction results. Whether the DRAINMOD method is overpredicting runoff volume on artificially drained sites or the Curve Number method is underpredicting runoff for naturally drained conditions is unknown at this time. It is interesting to note, however, that water quality studies done on the Albermarle–Pamlico Drainage Basin examining various watersheds grouped according to land use, drainage characteristics, and surficial geology have shown that the highest surface water P concentrations were observed in agricultural Coastal Plain basins with medium to poor drainage (Harned et al., 1995; McMahon and Harned, 1998). It is not known whether these increased concentrations of P in surface waters stem from P in drainage waters or surface runoff on these sites.
Artificially drained fields had a greater predicted P loss via surface runoff than naturally drained fields (0.59 vs. 0.35 kg P ha–1 for artificial and natural drainage, respectively), but a lower P loss due to subsurface drainage (0.02 vs. 0.07 kg P ha–1) despite the greater estimated volume of drainage. The reason for this finding is the inclusion of a transmissivity factor (T30/Tp) in PLAT's calculation of P loss through subsurface drainage. This factor acts to modify the amount of dissolved P moving through the soil profile. It was introduced because it is assumed that once dissolved P moves to a soil depth of 80 cm, it will be lost through subsurface drains rather than continue downward through the soil profile. No such factor exists for the computation of P loss via subsurface drainage on naturally drained sites. Thus, this factor acts to reduce the predicted P loss via subsurface drainage on artificially drained sites when compared to naturally drained sites, despite a greater drainage volume and equal drainage P concentrations. We recognize that the transmissivity factor, as well as the greater estimated volume of runoff from fields with artificial drainage, represent a potential problem and research attempting to validate the DRAINMOD model portion of PLAT is currently being performed.
Comparison by Phosphorus Source
When sample fields were segregated by applied P source, PLAT predicted that a substantially higher percentage of producers using animal waste would need to change to P-based management compared to fields receiving inorganic fertilizer. Between 20 and 25% of the fields receiving animal waste rated in the "high" or "very high" PLAT categories, whereas only 4% of fields receiving inorganic fertilizer as a P source resulted in either of these ratings (Table 3).
Mean total P loss predicted by PLAT was 1.58 kg P ha–1 for fields using animal waste versus 0.62 kg P ha–1 for fields on which inorganic fertilizer was applied (Table 4). Thus, the typical field on which inorganic fertilizer is applied represents a "low" potential for P loss while fields receiving animal waste were, on average, in the "medium" risk category. At this level, waste can be applied on an N basis, but the implementation of practices to reduce P loss should be considered.
Comparing P loss from only the applied source loss pathway shows that, among fields receiving swine waste, the predicted loss of P due to applied source was half that calculated from sites receiving other animal waste types (0.32 vs. 0.65 kg P ha–1 for dairy and 0.69 kg P ha–1 for poultry). Swine waste is typically applied as lagoon effluent, which has a greater proportion of dissolved P that can infiltrate the soil rapidly, leaving less P on the soil surface, available to runoff.
Differences in other loss pathways from fields receiving various animal waste types occurred, but these results are due to a variety of factors. Farms applying dairy waste had a relatively high loss of P through erosion, because these fields experienced a greater mean rate of soil loss, as predicted by the Revised Universal Soil Loss Equation (RUSLE) (7.4 Mg ha–1 vs. 2.3, 1.8, and 3.6 Mg ha–1 for poultry, swine, and inorganic fertilizer sites, respectively). All of the dairy sites occurred on erosion-prone upland soils. Presumably, a greater clay content is the reason why loss of P through surface runoff on dairy sites is relatively low, as PLAT assumes that soils with greater amounts of clay can sorb greater quantities of P before releasing dissolved P into the soil solution.
Most of the predicted P loss from sites receiving swine waste was through surface runoff and subsurface drainage pathways. Fields that had received swine waste had a greater subsurface STP, perhaps related to the proclivity of this type of waste to move downward through the soil profile. Additionally, 95% of sample fields receiving swine waste occurred in the Coastal Plain, a region which has a greater amount of artificially drained sites (49 vs. 3% in both the Mountains and Piedmont combined), from which PLAT tended to predict greater subsurface drainage volumes, although not necessarily greater P loss via drainage as described earlier. However, the combination of increased drainage volume and elevated subsurface STP levels caused a greater predicted P loss via subsurface drainage.
Fields receiving poultry waste tended to lose P predominantly through surface runoff and from the applied P source. Mean STP was highest on surface soils of these sites at 247 mg kg–1 as compared to 177 and 117 mg kg–1 for swine and dairy, respectively. Fields that did not receive animal waste had a mean STP of 94 mg kg–1.
Comparison by Physiographic Region
When samples were segregated by physiographic region, the proportion of sample fields in the "high" and "very high" risk categories decreased as elevation increased (Coastal Plain > Piedmont > Mountain). Approximately 11% of sites in the Coastal Plain would be required to switch to P-based management while 7% of sites in the Piedmont and only 1% of sites in the Mountains would be affected (Table 3). This same pattern of increasing P loss risk from west to east across the state occurred for predicted loss of P through surface runoff, subsurface drainage, and total P loss (Table 5). Mean soil test P also followed this trend: 134, 114, and 77 mg kg–1 for the Coastal Plain, Piedmont, and Mountain, respectively. The concentration of livestock in North Carolina is greatest in the Coastal Plain and Piedmont with the Mountain region having very little animal agriculture, with the exception of one county. Additionally, 99% of the clay sample sites occurred in the Mountains and Piedmont, while the majority of sites with sands were in the Coastal Plain. This is important because more P is required to raise STP by the same amount on a finer-textured soil versus a coarser-textured soil.
In general, the Coastal Plain region has more coarse-textured soils, higher water tables, and more poorly drained soils. These characteristics of the region, as well as being the site of the nation's second largest swine industry and its close proximity to sensitive estuaries, increases the threat to the environment. The PLAT index appeared to identify these potentially harmful conditions. As an example, sample fields in Duplin County, a Coastal Plain county which has the highest density of hogs in the state, had an average of 177 mg kg–1 STP and PLAT predicted over half of the fields to be in the "high" or "very high" risk category.
Fields located in the Piedmont region had, on average, greater estimated losses of P through erosion than sample fields in other regions. Fields in the Mountain region, which are expected to have relatively higher erosion rates, had low predicted loss via erosion and little P loss overall. This finding most likely results from the prevalent cropping system: 94% of fields sampled in the Mountain region were in pasture.
Comparison by Threshold Group
Because loss of P through surface runoff was the dominant loss pathway and is controlled partly by which threshold group a soil belongs to, we wanted to assess differences in PLAT results among threshold groups. The results show that the lower the P threshold of the soil on which a sample field occurs, the greater percentage of fields that fall in the "high" or "very high" risk category, with the exception of organic sites (Table 3). Fields with clay soil types, receiving both animal waste and inorganic fertilizer, and fields with loam soil types receiving inorganic fertilizer received an index rating placing them in the "low" category (Table 6). Fields with organic soils and receiving animal waste had a "very high" risk of P loss but this category included only one sample field so generalizations are not appropriate. All other soil type and P source type combinations fell in the "medium" risk category. Fields receiving animal waste and having loam or sandy soils were very close to index values which would place them into the "high" risk category at which P-based management is required. Where animal waste was applied, the index value was always greater than fields on which inorganic fertilizer was applied, regardless of soil type. This is due to the greater applied P source loss on these fields, but also to the buildup of soil P on sites receiving animal wastes as evidenced by the higher STP (Table 6).
Erosion P loss was greatest in clay soils where inorganic fertilizer was used. The reason for this is not clear as most clay sites were in the Mountain and Piedmont regions where pasture cropping systems dominated. Less erosion was predicted from clay and loam sites receiving animal waste versus those receiving inorganic fertilizer. This result may be due to the fact that more sites receiving animal waste in the Mountains and Piedmont were in pasture. Approximately 65% of loam fields receiving animal waste were under pasture while the same can be said of 93% of clay fields, versus 23 and 78% of sites receiving inorganic fertilizer for loams and clays, respectively. Loam and sand groups had a greater predicted risk of dissolved P loss through surface runoff when compared to soils in the clay threshold group. Clay soils are assigned a greater P threshold value than other soil groups in PLAT and are assumed to be capable of accumulating greater levels of STP before releasing dissolved P into solution.
Soil test P was greater in each soil group where animal waste was applied compared to inorganic fertilizer applications. With the exception of clays, all soil groups receiving animal waste exceeded the threshold P value for that soil group. The threshold value affects estimations of runoff P concentration in that even though clay soils receiving animal waste had similar STP values as other soil groups, they had relatively low surface runoff P loss. This is because the method used in PLAT for determining runoff concentration is based not only on a soil's STP but also to which threshold group it belongs to. For example, a clay site with a STP of 226 mg kg–1 was predicted to produce a much lower P concentration in surface runoff than a sandy field with a STP of 225 mg kg–1 (0.45 vs. 2.25 mg L–1). The organic soils had the greatest predicted loss of dissolved P through surface runoff due to the fact that these soils are treated very conservatively in PLAT, having a threshold P value of only 50 mg kg–1. Therefore, PLAT predicts high P concentrations in runoff from these soils at relatively low STP levels. This level of STP was set as a threshold in these unique soils because they have been shown to release large amounts of dissolved P (Daughtrey et al., 1972). More recent studies, however, suggest that shallow organic soils of the North Carolina Coastal Plain are able to retain more P than previously thought (Johnson, 2004). Currently, a number of studies are underway to help us better understand the dynamics of P movement in the state's organic soils.
Fields occurring on sands had a higher estimated subsurface drainage P loss component than did other soil types, which may have been confounded by the effect of location (Coastal Plain; Table 5) and animal waste type (swine lagoon effluent; Table 4).
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Overall, only a small percentage of producers in North Carolina can expect to be affected by new standards requiring the use of PLAT. Based on PLAT estimations, fields in the Coastal Plain represent a markedly higher risk of P loss from soil to water, and animal producers are more likely to be forced to adjust their management practices to comply with federal and state regulations. Therefore, more attention and resources should be focused on these sites as they are more likely to contribute to degraded water quality. Loss of P from fields receiving poultry waste was dominated by loss of P through surface runoff and loss of applied P source due to the high P content of the waste. Dairy waste had an equally high total P content and loss through applied source as well as a higher loss of P through erosion but a much smaller runoff loss. Sites receiving swine waste tended to lose P as dissolved P, to both surface runoff and subsurface drainage, because most of the P in this waste is already in a dissolved form. It is clear that where animal waste is being applied all soil types are experiencing a buildup of STP to levels which are considered excessive.
Loss of P through surface runoff, and the parameters that are used to estimate it, represents an area of PLAT that should be targeted for future research efforts, as this pathway accounted for a higher proportion of total P loss, especially from fields in the Coastal Plain. In soil types with lower P threshold values, losses of dissolved P through runoff were more significant, as it is assumed that these soils have lower P sorption capacities. Additionally, sandy fields were more susceptible to subsurface drainage P losses. Organic soils clearly represent an area in which our knowledge of P movement is lacking and which deserves further study. On average, PLAT appeared to identify problem sites as well as the vulnerable P loss pathways of an individual site.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Cahill, A. Johnson, D. Osmond, and D. Hardy Response of Corn and Cotton to Starter Phosphorus on Soils Testing Very High in Phosphorus Agron. J., May 7, 2008; 100(3): 537 - 542. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. R. Bond, R. O. Maguire, and J. L. Havlin Change in Soluble Phosphorus in Soils following Fertilization is Dependent on Initial Mehlich-3 Phosphorus J. Environ. Qual., August 9, 2006; 35(5): 1818 - 1824. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Vadose Zone Journal | Journal of Plant Registrations | ||||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal |
