| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Soil Science, University of Wisconsin-Madison, 1525 Observatory Dr., Madison, WI 53706
b USDA Dairy Forage Research Center, University of Wisconsin-Madison, 1925 Linden Drive West, Madison, WI 53706
* Corresponding author (kkelling{at}wisc.edu).
Received for publication January 26, 2003.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Abbreviations: Diff Meth, difference method • FE, fertilizer equivalence • GNU, grain nitrogen uptake • GY, grain yield • Rel Eff, relative effectiveness • TKN, total Kjeldahl nitrogen • WPNU, whole-plant nitrogen uptake • WPY, whole-plant yield
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Over the past 10 to 15 yr, there has been an increasing effort to improve the use of manure as a crop nutrient source for both environmental and economical reasons. Optimal manure N use that ensures adequate crop nutrition while avoiding pollution problems requires accurate and reliable estimates of manure N availability or recovery by the crop during the growing season. Several relatively simple chemical and biological indices have been proposed to predict N availability from native soil organic matter or organic amendments (Waring and Bremner, 1964; Stanford and Smith, 1972; Hong et al., 1990; Serna and Pomares, 1991; Paul and Beauchamp, 1993; Qafoku et al., 2001). However, it is widely acknowledged that these are more useful to compare relative availabilities than to provide absolute numbers (Keeney, 1982; Douglas and Magdoff, 1991). In addition, it is not likely that satisfactory fertilizer recommendations can be based on a single index (Bundy and Meisinger, 1994), and ultimately these predictions have to be corroborated by field experiments.
Two commonly used methods to determine manure N availability to crops are the fertilizer equivalence (FE) approach and measurement of apparent N recovery by the difference method (Diff Meth). The results from these indirect methods are often highly variable. For example, Motavalli et al. (1989), using the FE approach, measured 12 to 63% of dairy manure N as plant-available during the first season after application. Other estimates for dairy manure N availability have ranged from 10 to 57% (Castellanos and Pratt, 1981; Safley et al., 1986; Xie and MacKenzie, 1986; Jokela, 1992; Paul and Beauchamp, 1993).
Nitrogen recovery is the amount of applied N actually taken up by the plant and typically measured in aboveground tissue. Nitrogen availability is the amount of applied N that could be taken up by the plant in forms, concentrations, and locations that allow utilization by plants (Bundy and Meisinger, 1994), or compounds likely to convert to chemical forms accessible to plant roots during the growing season (Blackmer, 2000). The two indices of N availability used in this study compare manure N use by the crop to the use of fertilizer N, which is considered to be 100% plant-available. Hence, in this paper, available N can be thought of as N that behaves as fertilizer N, and therefore might be termed fertilizer N replacement value.
A direct assessment of manure N recovery can be attained by labeling manure with 15N and then measuring 15N in the crop (Kirchmann, 1990; Sørensen et al., 1994), but an estimate of availability also requires consideration of N recovery from 15N-enriched inorganic fertilizer. It is frequently assumed that the inorganic portion of manure N is as available as fertilizer N. However, Paul and Beauchamp (1995) found it to be about 59% as available as fertilizer N due to greater losses by NH3 volatilization, denitrification, and immobilization. We chose to refer N availability and recovery to total N in manure, to allow these losses to be reflected in the indices.
The objective of this research was to compare estimates of dairy manure N availability or recovery by corn using direct (15N labeled manure) and indirect (Diff Meth and FE) techniques in a field study.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Treatments were five levels of fertilizer N (45, 90, 135, 179, and 224 kg ha–1, applied as NH4NO3), two manure rates (estimated to provide approximately 90 and 180 kg available N ha–1 to corn the first year following application based on 40% availability), and a control receiving neither fertilizer N nor manure. There was a minimum of four replications of each treatment, arranged in four blocks to account for variations due to field topography. The plots were 10.6 by 6 m and separated by 1.5-m alleys, and contained eight corn rows, 0.75 m apart. For the 15N experiment, microplots of 1.5 by 2.3 m containing three corn rows were established within each of the low manure rate plots, following the design proposed by Jokela and Randall (1987).
Fertilizer and manure were applied about 5 d before planting. The field was disked twice (12–14 cm deep) within 3 to 20 h after application to incorporate the manure. All plots were planted to corn (cv. Lemke 6063) in each of the three study years. To ensure an adequate supply of P and K and optimize plant growth (Motavalli et al., 1993), all plots received starter fertilizer (band-applied, 5 cm to the side and 5 cm below the seed at planting, at 224 kg ha–1 of 9–23–30 in 1998 and 1999, and 168 kg ha–1 in 2000). About 40 d after planting, plants were thinned to a uniform population of 55000, 74000, and 60000 plants ha–1 in 1998, 1999, and 2000, respectively. The target population was 74000 plants ha–1, but poor seeding in 1998 and soil crusting in 2000 resulted in lower than optimal stands. The field received herbicides at (or shortly following) planting each year, and was cultivated at least once each season.
Corn aboveground tissue (henceforth referred to as "whole-plant") was harvested at approximately physiological maturity by cutting 10 adjacent plants 5 cm above the ground from one row in 1998 and 1999, and five from each of three rows (15 total) in 2000. The harvesting procedure was changed to obtain a more representative sample and reduce variability between replications of the same treatment. Grain was harvested from two entire rows (10.6 m) with a small plot combine. Three plants were cut from the middle row of each of the 15N microplots. Whole-plant and grain subsamples were oven-dried (55°C, 5–10 d) to determine tissue dry matter, ground in a stainless steel Wiley mill to pass a 2-mm screen, and stored in plastic bags until analyzed for total Kjeldahl nitrogen (TKN). Samples from the 15N microplots were reground in a Udy mill to pass a 1-mm screen, and analyzed for 15N and total N. Since samples from 15N microplots were very small, the more representative main plot data were used for whole-plant yield calculations in both main and microplots. Whole-plant yields are reported on a dry-matter basis, whereas grain yields are at 15.5% moisture. After sampling, the remaining plants were removed from the field. The site was chisel-plowed each fall.
Treatment Applications
Each rate of NH4NO3 was broadcast preplant to the same plots each year. Although some residual effect might be expected, this was not obvious in our experiment, as crop response slopes did not increase with time. Manure was applied to a new plot each year. The number of control plots diminished with time (e.g., plots receiving manure in 1999 for the first time were controls in 1998), but there was always at least one control per replication. The number of plots receiving manure was also different every year because this study used a subset of plots within a larger study. However, only plots receiving manure for the first time, with no other amendments applied since the onset of the experiment, were evaluated.
Fresh dairy manure (composite of feces, urine, and straw bedding) was collected from a stockpile where it had been stored for a few days. In 1998 and 1999, manure was applied with a small spreader. The amount applied was calculated by placing a small tarpaulin over the area where the 15N was to be applied. After spreading, the tarpaulins were removed and weighed, and a subsample of the manure was frozen until analyzed. To increase precision, in 2000, manure for each plot was individually weighed and uniformly forked on the plot. Manure analyses, as well as amounts applied each year, are presented in Table 1.
|
Chemical Analyses
Manure was analyzed following the procedures outlined by Combs et al. (2001). Total N in plant tissue was determined following a semimicro Kjeldahl digestion procedure adapted from Liegel et al. (1980). The digestions were performed on 250 mg of plant tissue in Pyrex Folin-Wu tubes. The digests were diluted, filtered, and analyzed for NH+4–N in an automated colorimeter (Lachat Instruments, Milwaukee, WI) using QuikChem Method 13-107-06-2-D (Lachat Instruments, 1992) with sodium phenate and 5.2% sodium hypochlorite. Total N and 15N concentrations in dairy feces, urine, and corn tissue samples from 15N microplots were determined using a Carlo Erba (Milan, Italy) elemental analyzer coupled with a mass spectrometer (Europa [Crewe, UK] 20/20 tracermass) on 5-mg samples.
Nitrogen Availability and Recovery Calculations
Corn grain nitrogen uptake (GNU) and whole-plant nitrogen uptake (WPNU) were determined by multiplying dry matter yields by respective tissue N concentrations. Since a small amount of N (15–20 kg ha–1) in starter fertilizer was consistently applied to all plots (including controls), this amount was not considered in the amounts of applied N. Manure N availability or recovery was estimated using three methods.
Fertilizer Equivalence
The FE method compares crop yield or N uptake in the manure treatments with those obtained from inorganic N fertilizer. Four crop parameters were used: whole-plant yield (WPY), WPNU, grain yield (GY), and GNU. Each year, each crop parameter was regressed against fertilizer N rate. These relationships were best described by linear functions in all cases, except for WPY and WPNU in 1999, where data were best-fit to an asymptotic response model adapted from Klausner and Guest (1981):
 | [1] |
To solve for FE for each crop parameter, data from manured plots were entered into the regression curves, and the fertilizer rate that would have produced the same yield or N uptake (the FE) was determined. This process is graphically illustrated in Fig. 1
, although FEs were calculated numerically. Fertilizer equivalents for replications of a given treatment were averaged. Percent nitrogen availability (NA) was calculated by dividing the FE by total applied manure N:
 | [2] |
|
 | [3] |
In the above equation, treatment N uptake and control N uptake were the N (kg ha–1) contained in the aboveground whole-plant for a given treatment and control plots, respectively. Applied N is the total rate of N applied in the treatment plot (kg ha–1). Calculations were made for each individual plot using the mean control N uptake for that year.
Apparent recovery of manure N can be compared with that of a fertilizer treatment providing an approximately similar amount of expected available N. An index of manure N availability termed "relative effectiveness" estimates the proportion of manure N that behaves as fertilizer N and therefore is more comparable with results from the FE method:
 | [4] |
The fertilizer treatments chosen were 90 kg N ha–1 rate for the low manure rate and 179 kg N ha–1 for the high manure rate, under the assumption that approximately 40% of newly applied manure N would be available during the first growing season. Although blocking did account for some variation, it actually increased variability of Rel Eff values, probably because there was only one plot per fertilizer rate per block. Mean apparent N recovery for any given treatment was the same with or without blocking; therefore, the mean N uptake for each fertilizer rate and controls was used.
Nitrogen-15 Recovery
Manure N recovery was estimated directly by measuring percentage 15N recovered in aboveground corn tissue at physiological maturity using the procedures outlined by Hauck and Bremner (1976):
 | [5] |
In this equation, P = total crop N uptake (yield data from main plot, N concentration from microplot), f = total manure N applied, a = atom % 15N in the manure applied, b = atom % 15N in the unlabeled manure (0.377 in 1998, and 0.366 in 1999 and 2000), c = atom % 15N in the treated crop, and d = atom % 15N in the control crop (0.366).
Statistical Analysis
Statistical analyses and regressions were performed using SAS (SAS Institute, 1990). Each crop parameter was analyzed as a randomized complete block design, with treatments and years as fixed effects. Blocks (replications) were treated as random effects. The same approach was used to analyze N availability and recovery estimates. For FE, the fixed effects considered were year, rate, and crop parameter; for the Diff Meth (both apparent N recovery and Rel Eff), year and rate; for 15N recovery, year; and for the comparisons among methods, year and method.
Type III F tests were used to assess the significance of fixed effects. Whenever interactions among fixed effects were not significant, they were eliminated from the model to gain degrees of freedom for the error term. When fixed effects were significant at 
= 0.10, selected orthogonal contrasts were performed to compare treatment means. The contrast labeled "manure vs. control" compared mean crop response in manured plots (both high and low rates) against control. The contrast labeled "fertilizer linear increase" assessed a linear increase in crop response to fertilizer N rates.
Regression analyses were performed by year for each crop parameter to determine crop response to fertilizer. Linear, quadratic plateau, and exponential models were used. For nonlinear regressions, the endpoint for successive iterations was determined using the Marquardt method (SAS Institute, 1990). The required initial values were estimated based on the graphs of crop response versus fertilizer rate. Single plot observations rather than averages by treatments were used to obtain the curves. The best model was chosen based on the highest R2, but whenever the inclusion of additional parameters did not result in a substantially better fit, the simpler model was used.
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
|
|
Over all three years of the study, there was a significant trend toward a linear increase in corn yield and N uptake with fertilizer rate (Table 4). Average crop responses were significantly higher in manure-amended than control plots (Table 4), and in general, crop responses to manure were about the same as those obtained for fertilizer between N application rates of 45 and 90 kg ha–1 (Tables 2 and 3). There was strong evidence that manure at the high rate increased the whole-plant and grain yield and N uptake compared with the control; however, this increase for the low manure rate was less distinct (Table 4). Crop differences between the high and low manure rates were only evident for the grain parameters (GY and GNU) (Table 4). Tables 2 and 3 show that whole-plant responses were similar for both manure rates, and sometimes slightly lower for the high manure rate. Motavalli et al. (1989) found similar or slightly lower corn silage yields for one of the six site-years at high dairy manure rates (about 150 Mg manure ha–1) compared with a moderate rate (90 Mg manure ha–1), although rates were higher than the ones used in the present experiment. Safley et al. (1986) and Vitosh et al. (1973) similarly reported little if any additional benefit from high rates of applied manure.
For this field, in these years, manure applied at the lower rate appeared to have provided sufficient N to the crop. The small increase in crop responses that resulted from doubling the manure rate generally did not justify the increased inputs.
Manure Nitrogen Availability and Recovery
Fertilizer Equivalence
Regressions for each crop parameter against fertilizer rate and the corresponding FE calculations were made by year because of the great across-year variation observed in crop responses. Large variability was also observed within a year (Fig. 1). This was probably due to the large field size (0.7 ha) with uneven slope and soil characteristics. We tried to reduce variability by blocking, but evaluating manure versus fertilizer responses within a block still showed considerable variation. In the end, we found it was more satisfactory to use one response curve for the whole field. The regression coefficients, R2, and p values are presented in Table 5. As an example of FE calculations, in 1998, the average WPNU was 275 kg ha–1 (Fig. 1) at the low manure rate (194 kg N ha–1). According to the regression line, a fertilizer rate of 60 kg ha–1 would have resulted in the same crop response; this is the FE for this manure rate. Using Eq. [2]: N availability % = (60 kg ha–1/194 kg ha–1) x 100 = 31%. This means that manure N (applied at 194 kg ha–1) had approximately 31% the effect of fertilizer N in increasing WPNU.
|
|
In all years, N availability estimates were lower for the high than the low manure rate, using WPY (p = 0.026) or WPNU (p = 0.008). In 2000, all estimates were lower for the high manure rate (p = 0.010). This was the consequence of small (if any) increases in crop responses after doubling the manure rate (Eq. [2]). Although not significant, it is noteworthy that N availability based on GY in 1998, and GY and GNU in 1999, was higher at the high than the low manure rate.
At the low manure rate, N availability estimates obtained in 2000 were higher than in 1998 (p = 0.045) and 1999 (p = 0.040). This was possibly due to a gradual lowering of the soil N supply and a reduced influence of the previous alfalfa crop. As a result, corn probably had to rely more heavily on the fertilizer or manure N inputs. Control plot yields and N uptake decreased significantly (statistics not shown) with time.
Evaluation of the various availability estimates using the different crop response parameters showed that care must be exercised when choosing a crop parameter to calculate these estimates. Grain yield and GNU sometimes appeared to be more sensitive indicators of crop response by giving more statistically significant treatment differences (Table 4), but their response curves (vs. fertilizer rate) resulted in less steep slopes compared with WPY or WPNU (Table 5). When the slope is shallow, a given change in the y axis (the crop response measured) translates into a much greater uncertainty in the x axis (the FE calculated). In general, this study showed that the steepest slopes and highest R2 and p values were found for WPNU. Other reasons to choose this parameter over the others as an indicator of manure N availability include (i) the limited increase in dry matter production as a result of additional N at high N inputs, whereas crop N uptake can still increase further (Klausner et al., 1994); (ii) the fact that N in the grain not only depends on the total amount of crop N uptake, but also on redistribution of N from vegetative tissue to grain, which might vary from one season to the next; and (iii) the fact that grain yield is highly affected by weather during anthesis, while N uptake is a cumulative parameter integrated over the entire cropping season (Meisinger, 1984). As was confirmed by this study, WPNU is the most reliable crop parameter to analyze crop response to N.
Difference Method
According to this method, the amount of N provided by manure or fertilizer was equaled to additional crop N uptake with respect to the control, and referenced to the total N applied (Eq. [3]). As discussed previously, an N availability index (Rel Eff) can be obtained by relating the apparent N recovery from the manure treatments to apparent recovery from an approximately similar fertilizer rate (Eq. [4]). Both indices for first-year manure and fertilizer treatments are presented in Table 7. Neither apparent N recovery nor Rel Eff of manure N varied significantly across years.
|
Manure N availability as estimated by Rel Eff at the low manure rate ranged from 24 to 61% with an average of 32%. This means that manure N was approximately 32% as effective as a similar rate of fertilizer N in increasing crop N uptake. Consistently lower estimates, although not significant (p = 0.14), were obtained at the high manure rate, which ranged from 4 to 27%, with a mean of 15%. Again, these results were a consequence of the small increase in crop N uptake after doubling the manure rate.
Nitrogen-15 Recovery
First-year recoveries of 15N in whole-plant (Eq. [5]) ranged from 10 to 22%, with an average across years of 14% (Table 8). These results are somewhat lower than those from other 15N experiments using sheep manure and barley as a crop, with 15N recoveries of 12 to 14% when feces only were applied (Sørensen et al., 1994), 22% (Jensen et al., 1999) (undersown with ryegrass), and 22% (Thomsen et al., 1997). Ryegrass amended with labeled fresh chicken manure recovered 26% (Kirchmann, 1990). These experiments were performed in pots or lysimeters where the manure was immediately covered with soil. Reduced N losses and possibly differences in type of manure and crop and soil conditions are likely causes of the higher 15N recoveries.
|
Comparison of Methods
Table 8 shows the N availability and recovery estimates according to the FE, difference, and 15N methods for each year. The ranges were usually quite large, especially for the FE and Diff Meth; however, to compare the indices, we averaged them. Only the low manure rate was considered, since this was the only level of application common to all three methods. Only results based on WPNU were taken into account for the FE method, since both the 15N and the Diff Meth were based on this crop parameter. Apparent N recovery by the Diff Meth and 15N recovery are estimates of the percent of manure N that is actually taken up by the corn. On the other hand, the FE and Rel Eff methods estimate available N from manure in comparison with fertilizer N use, or potentially utilizable by the crop. An estimate of manure N availability by the 15N method could have been obtained by including 15N-enriched fertilizer treatments and following a similar procedure to that of the FE method or computing a Rel Eff; however, those treatments were not included in the early years of this experiment.
Estimates of N recovery were not significantly affected by method or year. However, ranges for 15N recovery were somewhat narrower than for the Diff Meth, particularly in 1998. More importantly, several of the N recoveries as computed by the Diff Meth were negative (more commonly in 1998), meaning that crop N uptake in control plots exceeded those in manured plots. If during 1998 native N levels were high due to the previous alfalfa, then it is reasonable that no extra N was needed. However, some dairy farmers are faced with the need to dispose of manure wherever possible. Whereas the credit given to previous alfalfa is high the first year (110–215 kg ha–1), this is much smaller the second year (Voss and Shrader, 1984).
A comparison between Rel Eff and FE manure N availability estimates showed that the two methods did not statistically differ, and year also had no effect. Estimates of N availability by both methods were remarkably close at both manure rates (except in 1999; Tables 6 and 7) and averaged across rates: 19, 28, and 45% (FE, 1998–2000; Table 6) versus 14, 27, and 41% (Rel Eff, 1998–2000; Table 7). This was probably because both these methods estimate the proportion of manure N that behaves as fertilizer N, using slightly different approaches. The Rel Eff estimate of N availability uses actual measurements of N uptake in control, manure, and fertilizer (at a comparable rate) plots. Combining Eq. [3] and [4]:
 | [6] |
Nitrogen availability by the FE method can be expressed in a similar way:
 | [7] |
The Diff Meth only compares crop responses in manured or fertilized plots to controls. This approach has limited applicability in extreme situations where assumptions might not be met, such as when the soil is either high or severely deficient in available N. When using 15N-enriched manure, we positively know that excess 15N in the plant was provided entirely by applied manure. The 15N method does not require calibration curves and the control is the background level of 15N. Although this has to be measured, the value should be equal or very close to the natural abundance of this isotope, provided that no 15N-enriched material was applied. These measurements are very accurate and repeatable. The 15N method provides a more precise and direct estimate of manure N use by crops. However, this method does not allow for N availability estimates per se unless 15N-enriched fertilizer treatments are included.
In spite of the apparent lower accuracy of the Diff Meth, it provided virtually the same average estimate of manure N recovery estimates as the 15N method. This might suggest that, at least for our experimental conditions, the Diff Meth could be the most cost-effective approach for determining manure N recovery. However, considering the breadth of the N recovery ranges, sometimes going from negative to more than 100%, it is somewhat surprising that it has worked out so well. Although using 15N is costly and involves much more work, from experiment setup to sample analyses, manure N recovery measurements using this method are invariably more consistent and reliable.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Related articles in JEQ:
This article has been cited by other articles:
|
|
 |
|
  L. Bechini and P. Marino Short-Term Nitrogen Fertilizing Value of Liquid Dairy Manures is Mainly Due to Ammonium Soil Sci. Soc. Am. J., October 21, 2009; 73(6): 2159 - 2169. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Powell and J. H. Grabber Dietary Forage Impacts on Dairy Slurry Nitrogen Availability to Corn Agron. J., June 2, 2009; 101(4): 747 - 753. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. Ruiz Diaz and J. E. Sawyer Plant-Available Nitrogen from Poultry Manure as Affected by Time of Application Agron. J., August 11, 2008; 100(5): 1318 - 1326. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. R. Munoz, K. A. Kelling, K. E. Rylant, and J. Zhu Field Evaluation of Nitrogen Availability from Fresh and Composted Manure J. Environ. Qual., May 1, 2008; 37(3): 944 - 955. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. R. Cusick, K. A. Kelling, J. M. Powell, and G. R. Munoz Estimates of Residual Dairy Manure Nitrogen Availability Using Various Techniques J. Environ. Qual., October 27, 2006; 35(6): 2170 - 2177. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Sorensen and I. K. Thomsen Separation of Pig Slurry and Plant Utilization and Loss of Nitrogen-15-labeled Slurry Nitrogen Soil Sci. Soc. Am. J., August 25, 2005; 69(5): 1644 - 1651. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Powell, K. A. Kelling, G. R. Munoz, and P. R. Cusick EVALUATION OF DAIRY MANURE NITROGEN-15 ENRICHMENT METHODS ON SHORT-TERM CROP AND SOIL NITROGEN BUDGETS Agron. J., January 1, 2005; 97(1): 333 - 337. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Soil Science Society of America Journal | Journal of Plant Registrations | The Plant Genome | |||
