| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a PDK Projects Inc., 365 Wildwood Park, Winnipeg MB, Canada R3T 0E7
b Semiarid Prairie Agricultural Research Centre, Agriculture and Agri-Food Canada, Swift Current, SK, Canada S9H 3X2
c Western Land Resource Group, Semiarid Prairie Agricultural Research Centre, Agriculture and Agri-Food Canada, Ellis Building, Univ. of Manitoba, Winnipeg MB, Canada R3T 2N2
* Corresponding author (dmalley{at}pdkprojects.com)
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Abbreviations: NIR, near-infrared • NIRS, near-infrared spectroscopy • NSAS, Near-Infrared Spectral Analysis software • RER, the ratio of the range of the reference chemistry values for the prediction set to the SEP • RPD, the ratio of the SD of the reference chemistry values for the validation set to the SEP • SEP, standard error of validation • SRP, soluble reactive P • TDN, total dissolved N • TDP, total dissolved P
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Hog manure is gradually becoming a recognized fertilizer resource, largely because most of its N is in the form of NH4–N (North Carolina State University Extension, 1993), that binds to soil and is less prone to leach to ground water than is NO3–N. Nevertheless, management of manure from hog production is a particular challenge not only because of the associated odor, but also because of the large volume, high water content, and highly variable nutrient composition. Overfertilization and leaching problems have been documented to occur on some manured lands in Manitoba, Canada (Ewanek, 1995; Flynn and Cho, 1996).
Measuring nutrient loading to land reliably is more difficult for application of hog manure than for inorganic fertilizer. The manure varies in composition with type and age of animal, food, water content, storage and handling, climate, and amount of particulate material (Prairie Swine Centre Inc., 2002). Rapid settling of the particulates makes it difficult to uniformly mix the manure by agitation of a manure store. Determination of nutrient concentrations at numerous times, or continuously, during application is desirable to achieve the target loading.
Research is required to determine the best ways to combine hog manure, inorganic fertilizers, and the inherent nutrients in soil to meet but not exceed crop demands (Slevinski and Small, 1997). Technology has been developed to combine manure and inorganic fertilizer on-the-go (Lyseng, 1999). Long-term soil monitoring should be conducted to ensure that the manure application is appropriate to the soil conditions to reduce the risk of NO3–N leaching or P accumulation.
The optimal use of hog manure as a crop fertilizer thus requires methods of chemical analysis for two different matrices, i.e., manure and the soil. A method of chemical testing that is rapid, cost effective, and, ideally, field-portable and operative in-stream capable of analyzing N and P and salts in liquids and slurries would have wide use. The method should also be capable of providing low cost soil testing both before and after manure application.
Near-infrared spectroscopy is a rapid, nondestructive analytical technique used widely for the analysis of organic constituents and other properties in a wide range of commodities (Burns and Ciurczak, 1992; Williams and Norris, 2001). Near-infrared spectroscopy combines applied spectroscopy and statistics. Covalent chemical bonds between light atoms such C, N, O, and H, with primary absorbances in the infrared (IR) region, have strong vibrational overtones and combination bands that absorb light in the NIR region (780–2500 nm). The NIR region of the electromagnetic spectrum is mainly useful because a linear relationship between absorbance and concentration (i.e., the Beer-Lambert-Bouguer relationship) is exhibited in the majority of biological and agricultural applications, in contrast to the case in the IR region. Light in the NIR region reflected from samples is amplified, digitized, recorded as absorbance, and computed into composition data, using equations developed during the calibration procedure. Calibration equations, developed from spectral data and the results of conventional chemical analysis (termed reference data) on the same samples, are then used to predict concentrations of the constituents of interest in further unknown samples of a similar type. The technique analyzes intact samples that may be dried and ground, or analyzed "as is." The technique is thus rapid, nondestructive, and being introduced to the field with the development of field-portable or mobile NIR instrumentation (Case IH, 1999; Paul and Rode, undated). Near-infrared spectroscopy has the capability of measuring constituents in liquids (Gatin et al., 1996), slurries (Wust et al., 1996; Malley et al., 2000), and solids (Malley, 1998). A considerable amount of literature exists on the application of NIRS to the analysis of soil (Ludwig and Khanna, 2001; Malley et al. 1998). The technique has been used for the analysis of poultry (Reeves, 2001) and dairy manure (Asai et al., 1993; Nakatani, 1995; Kinoshita et al., 1997; Reeves and Van Kessel, 2000a, 2000b; Millmier et al., 2000). Hog manure has been analyzed by Malley et al. (1999) and Millmier et al. (2000).
The purpose of this study was to determine the feasibility of NIRS for the analysis of major nutrients, salts, and several other chemical parameters in hog manure and manure-amended soil.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Two studies are described. In the first study in September and October 1998, two soil types were sampled south-east of Winnipeg, Manitoba before and after the application of hog manure from two hog operations (Table 1). Samples were removed with a 5-cm diam. hand-held auger. At the clay site, soil was sampled from three locations within 10- to 20-m of each other varying slightly in topography. The three locations were marked and resampled 9 d after manure application. At the peaty site, four sampling locations within a 65-ha field were sampled on the same day. Two of these locations were sampled just before and just after manure was applied on that day. The other two locations had received manure 4 d previously and samples were post-application samples. Hog manure was sampled from the two earthen storage facilities being emptied at the time of this study and from five additional manure storage facilities representing several types of hog operations in south-eastern Manitoba. The manure (n = 64 samples) was grab sampled from the seven facilities and some samples were allowed to settle before subsampling to provide a wide range of particulate concentrations.
|
1 L in volume and were stored at 4°C until scanned or analyzed. All of the manure storage facilities were simple one or two-celled earthen structures. No treatment of the manure was carried out in the earthen storage facilities.
Chemical Analyses
Manure was analyzed for NO3–N, NH4–N, TDN, suspended N, SRP (essentially inorganic PO4–P), TDP (inorganic and organic dissolved P), suspended P, suspended C, Na, Mg, K, Ca, conductivity, and pH using methods modified from Stainton et al. (1977) and described by Malley et al. (2000). Except for moisture, pH and conductivity, samples were diluted 10000 to 100000 times for the analyses. Moisture was determined on 5 g of well-mixed samples as weight loss after drying for 4 h at 105 ± 2°C. pH was measured by pH electrode and conductivity by conductivity meter. Total dissolved N was determined after samples were decomposed with UV radiation in the presence of acid and adequate O2 and reduced to NH4–N. The NH4–N was measured by reaction with phenol and hypochlorite under alkaline conditions to form indophenol blue. The color intensity measured at 640 nm was proportional to NH4–N concentration. Nitrate-N was determined as the difference before and after the sample was reduced using a Cd-Cu couple. The couple reduced NO3–N to NO2–N that was determined using a Technicon Autoanalyzer (Bran + Luebbe, Hamburg, Germany; see http:www.branluebbe.com/company.html) by colorimeter measurement at 543 nm after reaction with aromatic amines. Suspended N and C were measured in the particulates collected on glass fiber filter paper combusted in a Control Equipment Corporation Model 240-XA Elemental Analyzer (Exeter Analytical, Inc., North Chelmsfor, MA). Soluble reactive P was determined with a Technicon Autoanalyzer after reaction with molybdate and reducing agents to form a blue compound absorbing at 885 nm. Total dissolved P was determined by the molybdate method (Strickland and Parsons, 1968) after the organic P compounds were photo-oxidized to orthophosphate with UV radiation. Suspended P was determined in the particulates collected on glass fiber filter paper ignited at low temperature to destroy organic matter. The P on the filter was extracted by acid and converted to orthophosphate and determined with the molybdate method. Sodium, Mg, K, and Ca were determined by atomic absorption and emission spectroscopy using a GBC 902 double beam atomic absorption spectrophotometer (GBC Scientific Equipment, Dandenong, Victoria, Australia).
Soil was analysed for moisture, organic matter, total N, available NO3–N, NH4–N, PO4–P, SO4–S, K, Na, Ca, and Mg by Norwest Labs, Winnipeg, MB. Moisture was determined by drying at 40°C and samples were ground for further analysis. Total N was determined by the Dumas combustion method using a Leco FP 428 N Analyzer (Leco Corp., St. Joseph, MI) and IR detection. Organic matter content was determined by the Walkley and Black (1934) method. Calcium chloride extraction followed by automated colorimetry using the Technicon Autoanalyzer was used for the determination of NH4–N and NO3–N. Ammonium acetate and acetic F extraction followed by automated molybdate colorimetry was used for the analysis of available PO4–P (Ashworth and Mrazek, 1995). Ammonium acetate and acetic F extraction followed by flame photometry was used for the analysis of K (Ashworth and Mrazek, 1995). Available Ca, Mg, and Na were determined by ammonium acetate extraction followed by analysis by inductively coupled plasma emission spectrometry. Except as noted, constituents are reported on a dry-weight basis.
For each of the sets of manure samples, the matrix of Pearson correlation coefficients, r, was calculated among concentrations of the constituents.
Near-Infrared Spectroscopy
Recording Spectra from Manure Samples
Absorbance of visible and NIR light was recorded using a Foss NIR Systems (Foss NIRSystems, Silver Spring, MD) Model 6500 visible/near-infrared spectrophotometer in the reflectance mode. Absorbance, as log 1/R where R is reflectance, was recorded at 2-nm intervals between 400 to 2500 nm for a total of 1050 absorbance values per sample. Software used to operate the instrument and record the spectra was Near-infrared Spectral Analysis Software (NSAS) (FossNIR Systems, Silver Spring, MD). A particular challenge in scanning slurries is that the particles can settle out during the scanning period of 50 s. Therefore, the instrument was placed on its back so that the sample cell was in a horizontal position and particles settled onto the transparent face of the cell and remained in the path of the light.
The manure samples were shaken well and 4-mL aliquots dispensed into a liquid sample cell of 2-mm path length for scanning. The cell was backed with an opaque ceramic. For samples with high suspended load, most of the light was reflected from the particles, whereas in clear samples, the light passed through the samples and was reflected by the ceramic. This mode is termed transflectance. It is a cross between transmittance (where the light passes through the sample to the detector) and reflectance (where the light is reflected from the sample to the detector). Between each sample scan, a reference ceramic was scanned and the reference spectrum subtracted from each subsequent sample scan. For each loading of the cell, triplicate scans were recorded, with the cell turned 120° between scans. Samples were loaded into the cell twice, resulting in six scans per sample.
Recording Spectra from Soil Samples
Soil samples were scanned by NIRS in both a field-moist ("as is") state and after drying and grinding. In both cases, soil was placed in a standard sample cell and scanned from 400 to 2500 nm, as for manure. Each soil sample was scanned three times between which the cell was turned 120°. Each field-moist soil sample was loaded into the cell three separate times, resulting in nine spectra per sample. This compensated for the variability in texture and color and microvariability in composition within the samples.
Dried and ground samples were more uniform than field moist. The cell was loaded once, and three scans were recorded.
Calibration Procedure by Multiple Linear Regression
The ability of NIRS to provide rapid analyses depends on the prior preparation of mathematical calibrations used to predict constituents, parameters, or functionality in unknown samples. A calibration is a statistical correlation model relating the spectral data for a set of samples to the constituent data determined by conventional methods.
Averaging the replicates for samples compared with using all of the spectra in calibration development gave slightly better results for "as is" dairy manure (Reeves and Van Kessel (2000b). Spectra averaged from triplicate spectra on 99 samples gave validation results of r2 and root mean squared deviation, respectively, of 0.967 and 0.132 g kg-1 for NH4–N, 0.945 and 10.2 g kg-1 for moisture, 0.950 and 4.00 g kg-1 for total C, and 0.956 and 0.30 g kg-1 for total N compared with respective results using all 297 spectra of 0.960 and 0.145 g kg-1, 0.930 and 11.6 g kg-1, 0.940 and 4.39 g kg-1, and 0.951 and 0.32 g kg-1. The root mean squared deviation is a statistic equivalent to SEP (standard error of prediction, described below) but not bias (y-intercept) corrected.
Replicate spectra for each hog manure or soil sample were averaged to give one spectrum per sample. The reference results for all the constituents for each sample were added to the NIR spectral file. Concentrations of constituents in the manure were on a wet weight basis. For field moist soil, concentrations were converted from dry weight to wet weight using the dry weight/wet weight ratio. For each constituent, the spectra were sorted from lowest to highest constituent value and divided equally into two sets. Every other sample was allocated to the calibration set, and the remaining samples to the validation set. Each set therefore represented the full range of constituent concentrations. Using the calibration set, up to 288 calibration equations were developed for the wavelength range 400 to 2498 nm using the stepwise multiple linear regression (MLR) option in the NSAS software. For example, separate calibration equations were computed using the raw optical data (log 1/R) smoothed over 4, 10, 20, or 40 wavelength points, termed "segments" (where wavelength points were 2 nm apart). The optical data were also transformed using first or second derivative and derivative ("gap") sizes of 4, 10, 20, or 40 wavelength points. As for the raw optical data, the derivatized data were smoothed using the above wavelengths segments. For each combination of segment and gap, equations for one to eight wavelengths were calculated.
Each of the calibration equations developed from the calibration set was used to predict the constituent values for the independent spectra in the validation set. The NIR-predicted values for the validation set were correlated to their measured reference values. The calibration process was completed when one equation was selected as giving the best results. This equation was slope and bias-corrected so that the slope was 1 and the y-intercept was 0.
The best calibration was the one with the highest r2 (coefficient of determination) between NIR-predicted values and measured values, and lowest SEP (i.e., the standard deviation of the residuals about the 1:1 line). Other statistics used to evaluate the calibration were the RPD, that is the ratio of the SD of the reference chemistry values for the validation set to the SEP, and the RER, the ratio of the range of the reference chemistry values for the validation set to the SEP. The procedure was repeated for each constituent. Only the results from the validation process are reported in the Results section below.
The usefulness of NIRS for the determination of these parameters in manure and soil was evaluated separately for each constituent of the manure and the soil. In the successful analysis of agricultural commodities, usually r2 is >0.95, RPD is >5 and RER is >20. Nevertheless, in more variable samples such as manure or soil, values of r2 > 0.9, RPD >3, and RER >10 are considered to indicate successful calibrations. Calibrations with lower statistical performance may still be useful depending on the accuracy required in the field situation and the availability of better alternative methods. They are useful for screening purposes, such as for distinguishing among low, medium, and high values, or for selecting samples for more costly conventional chemical analysis.
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
|
|
|
|
|
|
|
0.8 (Table 6) but somewhat less successful than for the set of 64 samples. The calibration for conductivity was marginal. The poor calibration for suspended N was inconsistent with good results for this constituent obtained in other manure data sets (the set of 64 samples and unpublished data). From the high correlations among C, N, and P in the 64 samples (Table 3), it is not clear which of the constituents were spectrally active. For reliable prediction, constituents must either be spectrally active or reliably correlated with a constituent that is spectrally active. There are numerous known NIR absorbers involving C and N (Williams and Norris, 2001) that could explain the predictability of these elements, but it was less clear whether fractions of P were spectrally active, or predictable because of correlation with C and N. The correlation matrix for the set of 74 samples (Table 4) provides evidence for an independent spectral basis for P fractions. The P fractions in this set were less highly correlated with the C and N fractions than they were for the 64 samples, yet were generally successfully predicted by NIRS (Table 6). Near-infrared prediction of NH4–N has been reported previously. Hall et al. (1996) successfully predicted NH4–N, glycerol, and biomass in an industrial E. coli fermentation medium. Nutrient levels in a fermentation system were predictable by NIRS by Brimmer and Hall (1993).
Millmier et al. (2000) predicted solids in hog manure successfully (r2 = 0.90), but total Kjeldahl N (TKN) (r2 = 0.69), NH4–N (r2 = 0.62), K (r2 = 0.71), and P (r2 = 0.61) less successful. For dairy manure, Reeves and Van Kessel (2000b) predicted moisture, C, total N, and NH4–N (r2 from 0.95–0.97) well but K, less successfully (r2 = 0.60) and P, poorly (r2 = 0.34). Results from Millmier et al. (2000) for beef feedlot manure were generally better than for hog manure (solids, r2 = 0.91; TKN, r2 = 0.67; NH4–N, r2 = 0.95; K, r2 = 0.82; and P, r2 = 0.58). For poultry manure, Reeves (2001) successfully analyzed NH4–N, organic N, total N, moisture and dry matter (r2 from 0.85–0.97), achieved useful results for Ca (r2 = 0.80), and poor results for P, K, Mg, S, Mn, Zn, and Cu (r2 = 0.40–0.62).
Prediction of Constituents in Field Moist and Dry Soil by Near-Infrared Spectroscopy
Means and ranges of constituents in the soil samples, expressed on a dry-weight basis, except for moisture, are given in Table 7. These samples from two sites and seven sampling locations southeast of Winnipeg, MB represent several soil types, sampling depths, and timing in relation to application of hog manure (Table 1). The soil samples met the requirement of the NIR technique for substantial ranges in composition.
|
|
|
|
The prediction of P fractions could not be accounted for simply by the fact that their concentrations were correlated with fractions of C or N. The success of predicting fractions of P suggests the presence of P bonds that are spectrally active. Phosphorus is not generally predicted successfully in poultry (Reeves, 2001), dairy (Millmier et al., 2000; Reeves and Van Kessel, 2000b), and hog manure in some studies (Millmier et al., 2000). More work is warranted on the role of food quality on animal manure quality and predictability of the manure P fractions. Phosphorus in the manure arises from two major sources. Phosphorus bound in plant material in phytates (inositol hexaphosphate) in the feed is generally unavailable to hogs since they lack the enzyme phytase. Therefore, to meet nutritional requirements, inorganic phosphate, such as dicalcium phosphate, is added to the feed. Undigested phytate and excess dietary phosphate end up in the manure. In the present study, both dissolved and suspended P were found to be predictable in hog manure. Hog diet is an active developmental area and numerous changes, e.g., replacing grain with low-phytate corn (Zea mays L.), the addition of phytase to the feed, or the use of transgenic hogs that digest phytate, could affect the predictability of the P fractions by NIRS.
There are no absorption bands for cations such as Na, K, Ca, and Mg (Burns and Ciurczak, 1992). Nevertheless, salts may be detectable in high-moisture samples because they change H bonding resulting band shifts (Burns and Ciurczak, 1992). Mineral constituents may also be predictable if they are bound to organics or correlated with organic components. In this way, it may be possible to predict K, Ca, and Mg. For example, in hay, good results were obtained for Ca, but the results were relatively poor for K, and Mg (Burns and Ciurczak, 1992). In the present study, Mg was highly correlated with suspended C and N and probably predictable because of this relationship. On the other hand, Na and Ca in manure were not correlated highly with any other constituents in this study, yet were successfully predicted. The prediction of Na and K in manure by NIRS probably related to shifts in the absorbance peak for water, nonetheless, was not found to be repeatable in two subsequent studies (unpublished data).
The spectral basis for predicting moisture in field moist soil as in the manure is expected to be the O-H in free water. In the dry soil, it is likely that clay or organic matter are responsible for predicting the water that would normally be held in the field-moist soil, that is, the spectra reflect water-holding capacity associated with organic matter and clays. The spectral basis of predicting organic matter and total N is expected to be the C-H, C-N, and C=O groups of organic matter as for manure. The NH4–N and NO3–N were not predictable in the field-moist soil, and NO3–N were not predictable in moist or dry soil. These latter results are consistent with most soil studies. The basis of predicting Ca and Mg is expected to be their association with carbonate, that is spectrally active.
|
CONCLUSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Successful calibrations were developed for field-moist soil for moisture, organic matter, total N, and Mg. Useful calibrations were obtained for SO4–S. These results are useful in indicating that NIR has potential to be used for rapid on-site analysis of "as is" soil. Calibrations developed for the same samples when they were dried and ground gave better calibrations for organic matter, NH4–N, K, SO4–S, and Ca. This technology has been found useful with varying degrees of success for the prediction of organic matter, NH4–N, NO3–N, organic C, N, P, Ca, Mg, Fe oxide, Al, Si, and other constituents in soil (Malley et al., 1999; Krischenko et al., 1992). Near-infrared instrumentation is still evolving with the recent development of precise field instruments that will permit mobile on-site measurement (Case IH, 1999; Rode, 2000; Williams, 2000).
A combination of on-site and in-laboratory analyses of soil by NIRS may be feasible to provide rapid low-cost accurate soil analysis to monitor nutrient status before and after manure application. Further studies with a wider range of manure samples and field demonstration projects are next steps in evaluating its use for rapid, on-site, manure analysis to guide applicators at the actual time of manure application concerning the amounts of nutrients going on to the land. The real-time, portable capability of NIRS makes it a potential partner with GPS/GIS technology for variable application of manure to match the fertility needs of the soil or the nutrients needs of the crop.
|
ACKNOWLEDGMENTS |
|---|
We gratefully acknowledge the sharing of hog manure samples by G. Racz, J. Hicks, and G. Plohman from their study, Agricultural Research Development Initiative (ARDI) Project 98-124 entitled "Long-term effects of hog manure on soil quality and productivity".
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Received for publication September 19, 2000.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. J. Chen, L. Xing, and L. J. Han Quantitative determination of nutrient content in poultry manure by near infrared spectroscopy based on artificial neural networks Poult. Sci., December 1, 2009; 88(12): 2496 - 2503. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Nduwamungu, N. Ziadi, G. F. Tremblay, and L.-E. Parent Near-Infrared Reflectance Spectroscopy Prediction of Soil Properties: Effects of Sample Cups and Preparation Soil Sci. Soc. Am. J., September 11, 2009; 73(6): 1896 - 1903. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Madejova, H. Palkova, M. Pentrak, and P. Komadel NEAR-INFRARED SPECTROSCOPIC ANALYSIS OF ACID-TREATED ORGANO-CLAYS Clays and Clay Minerals, June 1, 2009; 57(3): 392 - 403. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Xing, L. J. Chen, and L. J. Han Rapid Analysis of Layer Manure Using Near-Infrared Reflectance Spectroscopy Poult. Sci., July 1, 2008; 87(7): 1281 - 1286. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. K. Sorensen, P. Sorensen, and T. S. Birkmose Application of Reflectance Near Infrared Spectroscopy for Animal Slurry Analyses Soil Sci. Soc. Am. J., June 29, 2007; 71(4): 1398 - 1405. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Guerrero, J. Mataix-Solera, V. Arcenegui, J. Mataix-Beneyto, and I. Gomez Near-Infrared Spectroscopy to Estimate the Maximum Temperatures Reached on Burned Soils Soil Sci. Soc. Am. J., May 16, 2007; 71(3): 1029 - 1037. [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 | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
