| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Dep. of Plant and Environmental Sciences, P.O. Box 5003, Norwegian Univ. of Life Sciences, N-1432 Aas, Norway
* Corresponding author (susanne.eich{at}umb.no)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: BS, base saturation • BSE, backscattered mode • CEC, cation exchange capacity • ICP–AES, inductively coupled plasma-atomic emission spectrophotometry • SEM, scanning electron microscope • XRD, x-ray diffraction
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Since roots represent the underground part of the plant, studying characteristics of the rhizosphere poses a methodological problem. Under controlled conditions, rhizosphere soil has been studied with different experimental set-ups restricting root growth by porous membranes either vertically (Helal and Sauerbeck, 1983) or horizontally (Kuchenbuch and Jungk, 1982; Gahoonia and Nielsen, 1991). Under natural conditions, however, accurate methods to evaluate rhizosphere conditions are scarce. In a lot of field studies, rhizosphere soil has been sampled by gently shaking roots with adhering soil and defining the soil that remains attached to the roots after this procedure as rhizosphere soil (e.g., Hendriks and Jungk, 1981; Schöttelndreier and Falkengren-Grerup, 1999; or related methods, e.g., Gobran and Clegg, 1996). While this is a simple and broadly applicable method, it may neglect parts of the rhizosphere that do not adhere tightly to the root. Also, it is not a method that is appropriate to use on massive or coarse-structured heavy soils.
As an alternative to the shaking method, the soil surrounding roots in a stable soil block may be removed with suitable tools. Since roots grow preferentially in soil that is less dense than the mean soil bulk density (Hinsinger et al., 1993) they do not exploit the full extent of the soil resources and are not spread evenly throughout the soil. Roots seem to grow preferentially in preexisting pores (Passioura, 1991; Puhe, 2003). Wang et al. (1986) showed that soybean roots follow biopores rather than creating new pores in the bulk soil. Other studies suggest that roots preferentially recolonize existing root channels created by other species (Rasse and Smucker, 1998). In structured soils, roots are often located between the aggregates. Furthermore, under certain conditions such as hardsetting clay soils, up to 80% of all roots are found within 1 to 2 mm from macropores (Stewart et al., 1999). However, this can be viewed equally as an affinity of roots for the preexisting pores or as an inhibition by the soil matrix (Hinsinger et al., 1993). It is possible to scrape soil from pore walls at millimeter scale, as has been done by Pierret et al. (1999). They sampled 1 to 3 mm soil from pore walls of hardsetting clay B horizons and termed this macropore sheath soil though it might also be termed rhizosphere soil when roots are present in the macropores. Continuous pores are required for roots to grow in them. It may be assumed that the distribution of roots in continuous pores is a totally random process. On the other hand, pore diameter, water availability, presence of nutrients or toxic compounds may determine whether or not the root enters the pore. Alternatively, the distribution may be determined by plant species and root morphology. To evaluate whether nutrient availability is a determining factor, chemical characteristics of the soil around the pores need to be studied. Pore wall material around pores containing roots may be considered to be a special kind of rhizosphere where root-soil contact is low because the root diameter is smaller than the pore diameter. Sampling soil at the millimeter scale for extraction and chemical analysis poses a methodological challenge because it requires relatively undisturbed soil samples, tools for sampling very small amounts of soil and suitable analytical methods. The purpose of this study was to evaluate two different techniques for examining the chemical properties of the rhizosphere in soils with a massive to coarse prismatic structure: 1. sampling of soil at the millimeter scale in combination with a sequential extraction method, and 2. studying resin-impregnated, undisturbed soil samples with the help of a SEM.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
The root distribution in the profile was visually assessed during profile description. The upper O and A horizons were highly (and randomly) interwoven with roots and contained the majority of the roots in the profile. Due to the relatively high bulk density of 1.7 g cm–3 in the Btg horizon, and the strong stratification, roots grew exclusively in preexisting cracks and pores in the lower part of the profile.
Soil Sampling
In the A horizon, the soil structure is not influenced by stratification and the bulk density is lower so that root growth is not restricted to preexisting pores. Since the goal of this study was to evaluate methods of investigating chemical characteristics of soil around pores with and without roots, soil samples were taken from the Btg horizon and downward where root growth occurred mainly in preexisting pores. Soil samples with an intact structure were collected using a cylindrical soil sampler, fitted with an inner PVC tube 7 cm in diameter and 15 cm in length. The cylinder was pressed into the soil, and care was taken to disturb the samples as little as possible during the sampling process. The PVC tubes allowed transport and storage of a relatively intact sample. A total of nine samples were taken in the soil profile, with three replicates from the Btg horizon (B, 20–35 cm), the Btg/Cg1 transition zone (BC, 37–52 cm), and the Cg1 horizon (C, 55–70 cm), respectively. Samples were stored at 4°C for up to 5 mo until analysis. Similarly, nine undisturbed samples for embedding were taken next to the soil cores, using 10 x 6 x 5 cm metal boxes.
Sequential Extraction
In the laboratory, the soil cores were removed from the PVC tubes and carefully divided into smaller units of 10 to 20 mm along the natural strata. Mainly vertical, continuous pores of 1 to 2 mm in diameter with and without roots were identified. Precision dental hand tools (LM 125–126 XSi GM Trimmer mesial 1.2, LM 631–641 XSi TA Excavator 1.5, LM 522–523 Si Excavating spoon 1.0, LM 5–9 Si Explorer; LM Instruments, Parainen, Finland), allowing work at a millimeter scale, were used to scrape off samples of soil material surrounding these pores at different distances from the pore. The distances used were 1, 2, 4, and 6 mm from the inner pore wall into the bulk soil, giving four subsamples per pore. On average, five to eight pores, sampled along a length of 10 to 20 mm each, were sufficient to supply approximately 5 g soil for chemical analysis and pH determination in H2O and CaCl2 (soil to solution ratio 1:2.5). The procedure resulted in three replicates for each pore type, horizon, and distance from pore wall, that is, altogether 72 samples.
A sequential extraction was performed to determine the availability of different elements in the soil. Ammonium nitrate was chosen as an extracting agent for exchangeable cations (Stuanes et al., 1984) because this method has the advantage of not requiring a buffer. For the extraction, 1 g soil was weighed into centrifuge tubes, 15 mL of H2O was added, and the samples were shaken for 2 h and left overnight before centrifugation and decanting. Thereafter, 15 mL of 0.01 M NH4NO3 was added to the same soil sample, shaken, left overnight, and treated as before. This step was repeated with 0.1 and 1 M NH4NO3, respectively. Between each extraction step, the samples were weighed to correct for the solution that could not be decanted.
All samples were adjusted from the respective extracting solutions to a common background concentration of 1 M NH4NO3, by adding required amounts of NH4NO3, to avoid problems for further analysis due to varying ionic strength in the background solutions. The samples were analyzed by inductively coupled plasma spectrophotometry (ICP–AES, Thermo Jarell Ash, Franklin, MA) for the following elements: Al, B, C, Ca, Co, Cu, Fe, K, Mg, Mn, Mo, Na, Ni, P, S, Si, Ti, and Zn.
Scanning Electron Microscope
For quantitative x-ray analysis in the SEM and imaging in the backscattered mode (BSE), resin impregnation and preparation of a polished surface is necessary (Adamo et al., 1998). The undisturbed soil samples were impregnated in epoxy-resin. The soil blocks were then sliced into units of approximately 5 mm thickness using a diamond saw. From each of the nine soil blocks that were impregnated, one subsample was chosen that contained one or more pores of a suitable dimension (between 0.5 and 2 mm). The respective subsamples were ground with progressively finer grain sizes (50–5 µm) of AlO3, and polished with SnO2 (Murphy, 1986) until their surface quality was sufficient for SEM observations. The final polished surfaces were thoroughly cleaned in an ultrasonic bath before mounting the samples on SEM stubs and coating with carbon.
The samples were analyzed with a JEOL JSM 840 SEM equipped with a Link ISIS 300 Microanalysis System (Oxford Instruments, Witney, UK) with the following operating conditions: working distance: 15 mm, accelerating voltage: 20 kV. Initially, BSE images of relevant pores and surrounding soil, as well as x-ray maps to show distribution of Mg, Al, Si, P, S, K, Ca, Ti, Mn, and Fe, were acquired from all the three samples in each of the horizons (B, BC, and C). Out of these samples, one pore with and one without roots was selected in every horizon for more detailed analysis. The criterion used for selecting pores was similarity of size (approximately 0.5 mm in diameter) and absence of visible artifacts. Arrays, each containing 30 points along a line of approximately 500 µm length, were placed at 0-, 100-, 200-, 300-, 400-, 600-, 900-, 1200-, and 1500-µm distance from the pore wall. Quantitative element analysis was performed at each of the 30 points per array at x50 magnification using ISIS' SEMQuant and Autobeam programs (Oxford Instruments, Witney, UK). This procedure was conducted in three different directions out from the pore to obtain three replicates at each pore.
Data Processing and Statistical Analysis
Since the elemental distribution maps in the SEM analysis reflected the mineralogy of the soil around pores, the elemental data from the x-ray microanalysis of the different arrays was used to attempt a rough classification of type and abundance of the minerals present. From the x-ray diffraction analyses we knew which minerals to expect. This information was combined with information on the element composition of these minerals (Ottonello, 1997) to set up a rough classification routine that could be used for classifying the results obtained from the element analysis of the SEM procedure (Table 3). Point observations showing extremely high Sn values were excluded as these were assumed to be remains of the SnO2 polish left after the ultrasonic cleaning. The classification procedure was then used to test whether or not the SEM observations matched the requirements of Group 1. If so, the observations were excluded from further classification, if not, they were subsequently tested for Group 2 and so on. Observations that did not fit any of the seven groups were classified as Group 8, unidentified minerals. Mineral Group 1 was assumed to consist mainly of quartz, Group 2 mainly of Fe oxides, Group 3 of biotite and similar minerals, Group 4 of pyroxenes, Group 5 of amphiboles, Group 6 of different types of feldspar, and Group 7 of CaCO3. Pyroxenes and amphiboles (Groups 4 and 5) were later grouped together as heavy minerals. In the following results and discussion section, the name of the main mineral constituent of each group, rather than the group number, is used. However, the reader is asked to bear in mind that this is a simplified mineralogical identification. All analyzed points, where identification according to the group characteristics was impossible, were listed as "unidentified."
|
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Summing the mineral contents of all four extractions independent of pore type (data not shown), showed that concentrations for Al, Fe, and Li decreased significantly with depth (p < 0.001) at all distances from the pore walls. Concentration decreases were also found for Ti from the B to the BC horizon at 2-mm distance and for C from the BC to the C horizon at 2- and 4-mm distance from the pore walls (p < 0.05). In the case of K, Ca, Si, and Ni, concentrations generally increased with depth at all distances from the pore walls (p < 0.05). At 2- and 4-mm distance from the pore walls, B, Mo, and Mn concentrations were found to be larger in the BC horizon than in both B and C horizon (p < 0.05). Statistical analysis of the concentrations along the gradient from the pore walls into the bulk soil, independent of pore type, revealed no significant differences in the B and BC horizon. In the C horizon, the concentrations of Ca and Mg were significantly greater in the first millimeter around the pores compared with the samples taken at 2-, 4-, and 6-mm distance (p < 0.05; data not shown).
There were few statistically significant differences between pores with and without roots in any of the four extraction steps, due to high variation between the three replicates per distance, pore type, and depth. Exceptions are Mg and S in the BC horizon of the H2O-extract (Table 4), where concentrations were greater (p < 0.05 or p < 0.1) around pores with roots compared with those without. Similarly in the 0.01 M NH4NO3–extract, both Mg and Ca in the BC horizon were found in significantly greater concentrations (p < 0.05) at 2-mm distance from the walls of pores with roots compared with those without roots (data not shown). In the 0.1 and 1 M NH4NO3–extracts, no examples of significantly greater concentrations of nutrients close to pores containing roots were found (data not shown).
The pH measurements revealed no trends in relation to either the gradient from the inner pore wall to the bulk soil or between pores with and without roots. In CaCl2, pH values varied between pH 6.0 and 6.4 in the B horizon, between pH 6.8 and 6.9 in the BC horizon, and between 6.9 and 7.0 in the C horizon. Values for pH in H2O were approximately 0.3 to 0.5 units above the pH values in CaCl2. The values generally corresponded with the bulk pH in these horizons as shown in Table 2. Only in the C horizon, the values measured around pores both with and without roots differed by 0.5 or more pH units from the bulk soil pH of 7.9.
Scanning Electron Microscope
The SEM analysis described here yields total concentrations of different elements, but, unlike the sequential extraction, no information on nutrient availability to roots. Elemental distribution maps of total concentrations revealed little information on differences between pores with and without roots or gradients from the pore wall into the bulk soil. At a scale where it was possible to view a part of a pore with a diameter of approximately 0.5 mm and at least 1 mm of the surrounding soil (maximum x50 magnification), only very strong concentration differences would be evident in the distribution maps. The elemental distribution maps do, however, reflect the mineralogy of the soil, and the BSE images can illustrate potential textural differences that may occur in connection with the pores. The BSE images of the soil surrounding the pore showed that there were no obvious textural differences between the pore surface and the bulk soil, suggesting that these pores did not have any coatings.
For the B horizon, our tentative classification of minerals surrounding the pores showed a significantly lower percentage of biotite and similar minerals close to the pore wall in the pore containing a root than in the one without, whereas quartz, Fe oxides, and heavy minerals were present in higher percentages around the pore with a root (0–200 µm; Table 5). The percentage of feldspars was similar for both pore types. Though CaCO3 was found in the C horizon, only very low amounts were detected by the SEM analysis. Unidentified minerals were observed at a higher percentage around the pore without roots at distances of more than 200 µm from the pore. In general, the group of unidentified minerals had very few counts in the first 200 µm around the pores compared with the other layers around the pores independent of root presence.
|
In the C horizon, no significant differences between the pore with and the one without roots were found. While most mineral groups showed a similar pattern to the B horizon, the opposite was true for quartz and Fe oxides with a trend toward slightly higher counts close to the pore without roots.
Generally, fewer counts were included in our results from the first 250 µm around pores due to a higher percentage of points being excluded from interpretation because of high amounts of Sn. This resulted from a stronger relief closer to the pore, which seemed to have prevented more accurate cleaning of the SnO2 used for polishing.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
In the sequential extraction, most elements were found in similar amounts around pores with roots compared with those without. Pore wall material around pores containing roots may be considered as a special kind of rhizosphere where root-soil contact is low because the pore diameter generally exceeds the root diameter. For comparison, most field studies analyzing rhizosphere soil use the soil that adheres to roots after gentle shaking, which by definition ensures a more intensive root-soil contact. In any case, the contact between root and solid phase occurs mainly through the soil solution (Tinker and Nye, 2001). Nevertheless, in our study we may assume that the indirect contact via the solution phase was very variable, depending on the amount of soil solution present at different times of the growing season. Since the roots examined were found in pores of mainly vertical direction, these pores also represented the pathway for downward movement of soil solution. In wet periods, the indirect contact between roots and soil via the soil solution may have been intensive due to a saturated pore system with little movement, whereas in dry periods little contact existed. For nutrients transported to roots mainly by mass flow, such as Ca and Mg, an accumulation in the rhizosphere may be due to uptake at a lower rate than water (Malzer and Barber, 1975; Tinker and Nye, 2001). The fact that differences in Ca and Mg concentration between the pore types were either small or nonexisting in the present study may have been due to less intensive contact between the root and the soil than in other rhizosphere investigations. However, not all parts of the root are active in nutrient uptake. The effect of Ca and Mg accumulation is only transient because it occurs along the active root parts. As the root grows through the soil, former active parts become less active and differences in nutrient concentration due to root uptake decrease. The data we collected in the sequential extraction represents the rhizosphere independent of root activity.
Nutrients, such as K and Fe, do not normally accumulate in the rhizosphere. Instead, depletion zones are often found, at least in the case of K (Jungk and Claassen, 1986; Tinker and Nye, 2001). Our study did not confirm this, which again may have been due to the type of root-soil-solution contact or the transient nature of the root effect discussed above. It is also possible that existing differences diminished during storage.
Pierret et al. (1999) used a similar scraping technique to sample soil around macropores containing roots. They found larger C, N, P, Fe, and Mn concentrations in soil from the macropore sheath than in the bulk soil whereas Ca and Mg concentrations were similar. Larger amounts of Fe and Mn in their experiment were suggested to be due to the presence of dark coatings. In our study, no dark coatings were observed. However, Fe hypo-coating, impregnating the matrix around the pores in different distances, was observed both in the field, when scraping the pores, and in the embedded soil samples. These hypo-coatings were interpreted as being connected to redox reactions. No statistically significant differences in Fe concentration were found for the sequential extraction. The SEM analysis, however, showed a significantly larger amount of points classified as Fe oxides around pores with roots. Under waterlogged conditions, redox processes are enhanced by the presence of organic material and biological activity (Marschner et al., 1987). This may explain the differences found between pores with roots and pores without roots in the SEM study.
An increase in Fe around pores containing roots compared with pores without roots may also indicate that weathering of minerals (e.g., chlorite and biotite) has taken place. Weathering of minerals is an important source for many nutrients in natural ecosystems. Since Fe is relatively mobile in the soil profile, Ti is often used as a more reliable indicator of weathering due to its low mobility in the soil (Milnes and Fitzpatrick, 1989). Titanium is less mobile and an accumulation therefore suggests that weathering has taken place. Titanium may originate from weathering of such minerals as biotite or hornblende. Our study only shows a weak tendency toward greater amounts of Ti around pores with roots than around those without; no results were statistically significant. Weathering could explain an accumulation of nutrients such as Mg, as found in the sequential extraction, or also Ca, K, and Fe, as well as other elements such as Si and Al, around pores with roots. However, weathering represents a slow process compared with ion transport in the soil solution from dissolution of minerals and ion exchange. The soil in our study is pedologically very young with a high pH and CaCO3 present at least in the lower horizon (C) suggesting that, in general, weathering will be a process of minor importance with respect to nutrient availability in this horizon.
For the upper horizons (B and BC) where pH is lower and no CaCO3 is present, the mineralogical data derived from the SEM analysis may suggest that some weathering has taken place. Still, we have no information on whether the differences found between the pore types are actually caused by the roots present or whether they existed before their presence. However, a lower number of easily weatherable minerals, such as biotite, but a higher percentage of more stable minerals, such as quartz and Fe oxides, were found adjacent to the pores containing roots in the B and BC horizons. This agrees with some of the findings by Courchesne and Gobran (1997), where the mineralogy of bulk soil was compared with that of rhizosphere soil. Rhizosphere soil in their experiment was defined as the soil adhering to the roots after shaking. The rhizosphere contained significantly lower amounts of amphiboles and expandable layer silicates relative to the bulk soil and this was paralleled by a systematic increase in oxalate-extractable Al and Fe at the root-soil interface. However, in our study, there were no significant effects on amphiboles and pyroxenes (i.e., heavy minerals) with respect to pore type. This may be due to the presence of more easily weatherable minerals, such as chlorite and biotite, which weather before these heavy minerals (Teveldal et al., 1990). The group of unidentified minerals may contain weathering products (including expandable layer silicates) from primary minerals. Higher counts in this group around pores with roots, as seen in the B and BC horizon, would therefore be in accordance with weathering being induced or accelerated by the presence of roots. However, some uncertainty is related to the statistical analysis of our study. Only one pore per pore type and horizon was used for element quantification in the SEM, and subsequent creation of mineral groups. The measurements are not truly independent since they were taken around the same pore. Still, they covered different areas around the pore in question. This weakness in replication poses a limitation to the significance of the results from this study, since they may be an artifact of the pore examined.
While the pH around pores in the B and BC horizon was below 7, it was higher in the C horizon with 7.4 around pore walls and even 7.9 in the bulk soil. Presence of CaCO3 as indicated by weak effervescence (see section on Site and Soil) and a pH higher than 7 suggest little weathering in the C horizon. This is in accordance with the lack of significant differences between pore types in the C horizon (Table 5). The fact that the mineral classification showed very low amounts of CaCO3 around pores in all horizons may have been due to the SEM investigations concentrating on the first 1.5 mm from the pore walls, and not investigating the bulk soil. It is possible that the presence of roots may have locally reduced the pH below 7 and thus dissolved CaCO3 even in the C horizon. This may be supported by the findings of the sequential extraction that significantly more Ca and Mg were present in the first millimeter around pores compared with, for example, 6-mm distance, when all extractions and both pore types were combined (see Results below). The Ca found in the extracts might have been bound to the exchange sites in the soil after dissolution of CaCO3. In general, any differences in nutrient availability between pores with and without roots that may be found with a sequential extraction method, may be due to interactions between the root present in the pore and the surrounding soil. However, they may also be the result of accumulated rhizosphere processes over many years in the case of recolonization of previous root channels. In the case of recolonization, one might expect increased amounts of organic matter around these pores. Nutrients connected to organic matter may be more easily accessible than nutrients connected to minerals (Pierret et al., 1999). However, the sequential extraction did not show any significant differences for C. Neither did we find any coating around these pores that could indicate organic matter. It is also possible that preexisting differences in texture and mineralogy between the two pore types would influence the results, though the pores investigated in the SEM analyses did not suggest any such differences. In our study, finding some evidence for weathering through the SEM analysis, however, supports the hypothesis that recolonization occurred since the influence of one root on weathering is likely to be too small to be detected by the SEM method.
The sequential extraction may reflect the ability of the present root in the pore to influence its surroundings. The method can successfully be used for showing differences between pore walls and bulk soil. That there were relatively few statistically significant differences in our study may have several reasons: (1.) The contact between the root and the solid phase (via the solution phase or direct) was too limited to cause significant influence on nutrient availability. (2.) A greater number of replicates may be necessary to overcome the high variability of the data. (3.) We do not know if roots had been present in pores in previous growing seasons that were in our study sampled as pores without roots. This would mean that the pore types were not actually different.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Resin impregnation of undisturbed soil blocks for further investigation in the SEM may be applicable to a wider range of soils. In our study, the elemental distribution maps generated by the SEM did not reveal enough detail to draw conclusions about possible differences in total elemental concentrations. In addition, a quantitative element analysis in the SEM of points at different distances from the pores was conducted and created a large data set on elemental composition of pore wall material that mainly reflected mineralogy. Scanning electron microscope data will be difficult to use for evaluation of chemical conditions around pores with and without roots unless the mineralogy of the soil is very homogeneous. Grouping elemental data into mineral groups according to the content of key elements allowed a comparison between the pore types with respect to differences in easily weatherable or stable minerals. The differences found suggested that weathering was enhanced around pores containing roots.
Both methods, though used for pores with and without roots in our study, are also applicable for rhizosphere and bulk soil investigations independent of pores, as long as the soil structure allows the respective sampling techniques.
|
ACKNOWLEDGMENTS |
|---|
Received for publication March 25, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Calvaruso, L. Mareschal, M.-P. Turpault, and E. Leclerc Rapid Clay Weathering in the Rhizosphere of Norway Spruce and Oak in an Acid Forest Ecosystem Soil Sci. Soc. Am. J., January 21, 2009; 73(1): 331 - 338. [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 | |||
