| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Alterra, Dep. of Water and Environment, P.O. Box 47, 6700 AA Wageningen, the Netherlands
b Dep. of Agronomy and Range Science, University of California-Davis, 1 Shield Avenue, Davis, CA 95616
* Corresponding author (J.W.vanGroenigen{at}Alterra.wag-ur.nl)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
13C and 15N) on a Lethent Clay Loam (fine, smectitic, thermic typic natrargid) in the San Joaquin Valley (California) for a nonhalophyte, C3 plant and soil organic matter (SOM) fractions. A total of 101 plant (Littleseed Canarygrass, [Phalaris minor Retz.]) and soil samples were collected from a 10-ha area. Electrical conductivity in a 1:5 soil/water paste (EC1:5) ranged from 2.7 to 8.9 dS m-1. The 13Cplant values varied from -29.8 to -24.0
, and 15N from 2.2 to 19.1. Average values for 13C increased from -26.9 in the plant, to -25.3 in the light fraction (LF) and -24.1 in the SOM. Salinity explained 57% of variance in 13Cplant, 16% of 13CLF and 6% of 13CSOM. For 15N, these numbers were 41, 56, and 0%, respectively. There was a clear spatial pattern match between salinity, 13Cplant, 15Nplant, and 15NLF. The lack of any salinity-induced signature in total SOM probably indicates that the salinity was of recent origin. The high positive correlation between salinity and 15N in crop and LF might be because of higher NH3 volatilization caused by high pH, combined with a relative increase of NH+4-uptake by the plant under saline conditions. Under certain conditions, 13C and 15N signatures of recalcitrant SOM fractions may be used to reconstruct historic salinity patterns.
Abbreviations: CLF, C from light fraction • Cplant, C from plants • CSOM, C from soil organic matter • EC, electrical conductivity • EC1:5, electrical conductivity of a 1:5 soil/water paste • LF, light fraction • NLF, N from LF • Nplant, N from plants • NSOM, N from soil organic matter • SOM, soil organic matter • 13C, natural abundance of C • 15N, natural abundance of N
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
50% of all irrigated land experience some degree of salinity stress. Close to 10 million ha has to be taken out of production yearly because of salinity-related problems (Rhoades and Loveday, 1990). Although 15% of all agricultural land is irrigated, this land contributes to 40% of food production. Relatively little is known about the effects of salinity stress on C and (especially) N cycling in agroecosystems. Such knowledge is crucial in devising remedial action, either by improving crop management or by developing new cultivars. Since many biochemical and physiochemical reactions have the potential for some sort of isotopic discrimination, natural variations in stable isotopes of C and N could elucidate many nutrient cycling processes in agroecosystems. However, because isotopic concentrations are almost always the result of a combination of processes, it is often difficult to interpret results. The 13C concentrations in SOM, for example, can be the result of photosynthetic activity in past vegetation (C3 vs. C4), age of SOM, depth in the profile, and the amount of fossil fuel derived CO2 in the atmosphere (Shearer and Kohl, 1986). However, since plants derive all their C from the air, modeling 13C concentrations in the plant is relatively easy. Nitrogen, which can be taken up from 4 different sources (NH+4, NO-3, organic N, and atmospheric N2) via the roots and foliarly is much more difficult to model (Marriott et al., 1997; Farquhar et al., 1980).
A relationship between 13C of crop tissue and salinity (measured as NaCl concentration in nutrient solution) was first reported by Guy et al. (1980) for the C3 halophytes Salicornia europaea L. ssp. rubra (Nels.) Breitung and Puccinellia nuttalliana (Schultes) Hitchc. In a controlled experiment with different concentrations of NaCl, they found positive correlation coefficients (r2) of 0.88 and 0.96, respectively. These findings were later confirmed under field conditions for the same species (Guy et al., 1986a and 1986b). Along a soil water potential gradient ranging from -1.3 to -6.0 MPa, 13C values ranging from -27 to -21 were found for Puccinellia nuttalliana. The corresponding r2 value was 0.90. Flanagan and Jefferies (1989) reported a significant shift from -32.4 to -28.3 because of salinity in the C3 halophyte Plantago maritima L. under controlled conditions. Chmura et al. (1987) found a range of 13C values in sediment from -27.9 in fresh water marshes to -16.2 in salt marshes over a variety of plant species. However, these differences were probably primarily because of a corresponding shift from C3 plants to C4 plants. Walker and Sinclair (1992) reported r2 values around 0.66 between electromagnetic measurements and 13C values of leaves of the C4 halophytes Atriplex vesicaria and A. stipitata, although they did not detect an effect in whole plant 13C.
Initially, the observed relation was attributed to a shift in the use of the primary enzyme for CO2 fixation from ribulose 1,5 bisphosphate (RuBP) carboxylase towards phosphoenolpyruvate (PEP) carboxylase because of high levels of salinity (Guy et al., 1980). However, Farquhar et al. (1982) and Guy and Reid (1986) showed that the positive correlation was more likely an indirect consequence of reduced stomatal conductance under stress conditions. This results in a lower intercellular CO2 pressure. Since 12C is preferentially assimilated because of discrimination during diffusion and carboxylation, the reduced CO2 pressure will lead to an enrichment of the assimilated material with 13C. Farquhar et al. (1982) derived the following formula for 13C in a C3 plant:
 | [1] |
Much less progress has been made on the relationship between stress and 15N in the plant, mostly because the observed variation was often attributed to a shift in source 15N, rather than discrimination in the plant (Nadelhoffer and Fry, 1994; Högberg, 1997; Gebauer et al., 1994; Gebauer and Schulze, 1991). Foliar 15N discrimination has been observed in plants under cold stress (Kohls et al., 1994) and in slow growing plants (Domenach et al., 1989; Högberg, 1997). Handley et al. (1997) reported a decrease in foliar 15N because of moderate salt stress in a controlled experiment with barley (Hordeum spp.). They reported that this effect would be consistent with a decrease in nitrate reductase activity.
Limited work has been done on analyzing spatial variability of 15N and 13C in agricultural fields. Androsoff et al. (1995) and Stevenson et al. (1995) measured N2 fixation by pea (Pisum sativum L.) across several landscape units using 15N. They concluded that, although topography partially controls N2 fixation, most of the variability was at the microscale. Sutherland et al. (1993) and van Kessel et al. (1994) showed that topography had a significant influence on 15N in both plant and soil. The 15N patterns could be linked to denitrification. Velthof et al. (2000) confirmed these findings using geostatistical analyses. Karamanos et al. (1981) reported an increase in 15N values in saline seep soils of 5.2, as compared with similar, nonsaline soils. Using geostatistical analysis, Marriott et al. (1997) found that spatial correlation of 15N and 13C in an upland grassland soil was confined to 7 and 12 m, respectively.
The main objective of this study was to determine whether the relationship between salinity stress and 13C values in plant tissue that was observed earlier in natural systems for C4 plants and halophytes is valid in cultivated agroecosystems. In addition, we tested the negative relationship reported by Handley et al. (1997) between salinity and 15N in plant tissue in the field. The salinity signature was followed from the crop into the LF, and finally to the SOM. A California field with high variability in salinity was selected, and plant, LF, and SOM samples were analyzed for 13C and 15N. Salinity data and isotopic signatures were correlated using statistical and geostatistical techniques.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
10 ha in size. The field was cropped with Cotton (Gossypium hirsutum L.) for many years, and no record of any C4 crop was found. An experiment on the reuse of saline drainage water for irrigation of salt tolerant crops was initiated at this location. Littleseed Canarygrass, a C3 plant which was well established throughout the field, was used for plant analysis.
Soil and Crop Sampling
One hundred one soil and crop samples were collected on 3 and 4 Jan. 2000 using the sampling scheme shown in Fig. 1a
. The sampling scheme was part of a larger sampling effort, including surrounding fields. It was established as a combination of an optimized sampling scheme for spatial interpolation as described by van Groenigen et al. (1999), with an equal number of random observations for short-term variability assessment. All sampling sites were located using a Trimble Pathfinder Pro XRS differential Global Position System (GPS) (Trimble Navigation Ltd., Sunnyvale, CA) with real-time differential correction, resulting in an accuracy of 0.3 m. At all 101 locations, three soil cores were sampled from a 0- to 10-cm depth. Soil samples were dried for 48 h at 60°C. Three whole aboveground plants were collected at all 101 locations, dried for 48 h at 60°C and ball-milled.
|
Total C, total N, 13C, and 15N were determined at the UC Davis Stable Isotope Facility using a continuous flow, isotope ratio mass spectrometer (CF-IRMS, Europa Scientific, Crewe, UK) in the dual-isotope mode, interfaced with a CN sample converter. Concentrations of 13C and 15N are reported using differential notation, showing differences between the observed concentration and that of a common standard in (Handley and Scrimgeour, 1997). The standard for 13C was Pee Dee Belemnite 
, and for 15N, atmospheric N2. Purified (NH4)2SO4 with a 15N value of 1.33, calibrated against IAEA N1 and IAEA N2 standards served as the working standard for N. Beet (Beta vulgaris L.) sucrose with a 13C value of -23.83, calibrated against NIST SRM 8539 and NIST SRM 8542 standards, was the working standard for 13C.
For the LF of the SOM, a subsample of soil was milled in a Wiley mill, and 10 g of soil was weighed out. The sample was put into a 50-ml centrifuge bottle, with 25 ml of a 1.75 Mg m-3 NaI solution. The bottle was hand shaken to wet soil, shaken for 5 min. on a mechanical shaker and centrifuged for 20 min. at 2820 x g. The supernatant was decanted into a test tube, and 25 ml of the NaI solution was added for a new cycling of shaking, centrifuging, and decanting. This cycle was repeated twice, yielding a total of 75 ml of supernatant.
Subsequently, the solution was decanted under a vacuum over an ashed GF/A filter. The filter was thoroughly rinsed with demineralized water, and dried at 105°C for 36 h. The part of the filter containing the LF was removed and split in half, yielding two pieces of filter of 4 by 6 mm each. Both halves were rolled up, placed into a tin capsule and analyzed for 13C and 15N.
Salinity measurements of a 1:5 soil/water paste were conducted using the method outlined in Rhoades (1996). A soil subsample of 5 g was measured out in a flask with 25 ml of distilled water. One drop of 0.1% (NaPO3)6 was added to prevent precipitation of CaCO3, and the flask was shaken for 1 h in a mechanical shaker. The soil was allowed to settle overnight and electrical conductivity (EC) was measured in the clear supernatant using a standard EC meter. The EC meter was calibrated using a KCl standard series.
Geostatistical Analysis
Data were checked for normal or lognormal distribution. Descriptive statistics were collected and correlation coefficients were calculated between all measured properties. Lognormal data was transformed for geostatistical analysis. The data were checked for linear trends, and geostatistical analysis was performed on residuals. The experimental variograms were calculated up to a range of 160 m (half the width of the field), and modeled using Variowin software (Pannatier, 1996). Subsequently, interpolated maps were created by ordinary kriging using GSLIB (Deutsch and Journel, 1998). If necessary, the data were back transformed and the trend was added, yielding the final interpolated map. To better facilitate comparisons between the different properties and to enhance contrast, most maps were presented using contour maps of the quartiles of the interpolated values.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
13Cplant values ranged from -29.8 to -24.0, with an average of -26.9. The 13CLF values were more enriched, ranging from -26.9 to -19.1, and an average of -25.3. The whole SOM was even more enriched, with corresponding values of -25.3, -21.9, and -24.1, respectively. There was a notable decline in variation in 13CSOM, with the standard deviation only half of that for 13Ccrop and 13CLF.
|
15N, there was no such clear enrichment during decomposition, with average values of 7.8, 5.8, and 7.6 for plant, LF, and SOM, respectively. The maximum values for 15Nplant and 15Nsoil that were not rejected as outliers (19.1 and 18.5) are high, but not without precedent (Sutherland et al., 1993). It is noteworthy that the location with the highest EC1:5 value (8.93 dS m-1) also has the highest 15Nplant (19.1), the lowest total C and N in the soil (2.6 and 0.4 mg g-1, respectively) and the lowest C/N ratio (6.3, data not shown).
Apart from a well-known relation (total C and N in the soil), the three highest correlation coefficients are EC1:5 with 13Cplant (r = 0.752), 15Nplant (r = 0.637), and 15NLF (r = 0.748) (Table 2) highly significant correlations include a negative one between EC1:5 and total C (r = -0.572), and negative ones between total C and both 13Cplant (r = -0.504) and 15Nplant (r = -0.515). Finally, 15NLF is positively correlated with both 13Cplant and 13CLF.
|
13C and 15N in crop and LF positively correlated with EC1:5, and negatively correlated with total soil C and N. The 13CSOM and 15NSOM were only weakly or not correlated with values in crop and LF and with total C and N or EC1:5.
Variography
Table 3 shows the parameters of the modeled variograms for all properties of interest. Only total soil C and N showed a trend, which was subtracted for variography and interpolation. Three properties (13CSOM, 15NLF, and the residuals for total soil C) were approximately lognormally distributed, and were therefore logtransformed.
|
13CSOM, 15NSOM, Total C, and Total N) except EC1:5 have a relatively short range (i.e., scale of spatial correlation), ranging between 29 and 36 m. In contrast, all isotope data from crop and LF, as well as EC1:5, have a range of 90 m or larger, with EC1:5 and 15Nplant having no range within 160 m.
There is a reduction in scale of spatial correlation corresponding with increased decomposition for both 13C and 15N. For 13C the range of the modeled variogram decreases from 101 m in the crop, through 90 m in the LF, to 36 m in the soil. For 15N, these numbers are >160, 136, and 29 m, respectively.
Mapping
Figure 1b shows the interpolated map for EC1:5 by ordinary kriging. The interpolated values, ranging from 2.7 to 8.9 dS m-1, show a clear hotspot in the lower left corner of the field. Notable lowspots are in the lower right corner, the upper half of the right boundary and the middle of the left boundary. Overall, it is a smooth pattern, indicative of the strong spatial correlation of EC1:5 (Table 3).
Figure 2a
shows the same EC1:5 map, with only the quartiles of the interpolated values as contours to better accommodate comparisons with other properties. Figures 2b through 2d show interpolated values for 13C in plant, LF, and SOM, respectively. The maps for EC1:5 and 13Cplant show a strong positive correlation, with the upper quartiles extending almost exactly over the same area (Fig. 2a,b). Most of the lower quartiles of the maps also correspond. This is consistent with the strong positive correlation between EC1:5 and 13CPlant that was recorded in Table 2.
|
13CLF and 13CSOM that were reported in Table 2 are reflected in Fig. 2a,c, and d. The correspondence between the quartiles of EC1:5 and 13CLF is weak. Although parts of the lower and upper quartiles correspond, the overall patterns do not match. There is no visual correspondence between EC1:5 and 13CSOM. In addition, the short range of the variogram for 13CSOM (36.4 m; Table 3) results in a patchy interpolated surface. The other three properties, which all have ranges larger than 90 m, show a similar, much smoother variability.
Figure 3
shows the quartile map for EC1:5, together with 15N values for plant, LF, and SOM. As with 13C, there is a clear positive relationship between EC1:5 and 15Nplant (Fig. 3a and 3b, respectively), with almost identical upper quartiles and similar lower quartiles. In contrast to 13C, the EC1:5 pattern is also reflected in 15NLF quartile map (Fig. 3c). This is in line with the high positive correlation coefficients reported in Table 2.
|
13C data, there is no visual correspondence between EC1:5 and 15NSOM. The much shorter range of the variogram for 15NSOM (30 m) results in a more patchy interpolated surface, as compared with the other three maps with associated variogram ranges of 136 m and higher.
The quartile maps for total soil C and N (Fig. 4a and 4b
, respectively) look similar, as can be expected and is reflected by the high reported correlation coefficient (Table 2). The trend that has been detected increases from the right to the left side of the field. The negative correlation between total soil C and N on the one hand, and EC1:5, 13Ccrop, 15Ncrop, and 15NLF on the other hand can be seen most clearly by comparing high quartiles in Fig. 4a (total soil C) with low quartiles in Fig. 3c (15NLF), and vice versa.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
13Cplant and 15Nplant, and the fact that this relation is absent in the soil. We believe that these results can be explained by the spatio-temporal variability of fractionation processes induced directly and indirectly by salinity stress.
Salinity and Natural Abundance of Plant Carbon-13
To our knowledge, this is the first field study that reports a clear positive relation between salinity and 13C values of a C3, nonhalophyte plant. The high correlation coefficient between these two properties (Table 2), combined with the interpolated quartile maps of Fig. 2a and 2b, clearly shows such a relationship. We believe that this can be explained by the fractionation process outlined by Farquhar et al. (1982) and Guy and Reid (1986). Because of salt stress-related closing of the stomata, the partial 12CO2 pressure in the plant decreases, forcing it to assimilate more 13CO2, thereby making 13C signatures of the newly formed plant tissue less negative. This process, which has before been observed under laboratory conditions (e.g., Flanagan and Jefferies, 1989) and in natural ecosystems (e.g., Guy et al., 1980; Walker and Sinclair, 1992), manifested itself in managed fields.
It is noteworthy that the positive relation between salinity and both isotopes in the plant tissue is clearest for the higher salinity values. Although the lower quartiles of 13Cplant (Fig. 2b) and 15Nplant (Fig. 3b) roughly coincide with areas of lower salinity (Fig. 2a and 3a, respectively), it is the higher quartiles that show the best resemblance. This can be explained by the presence of a threshold salinity level (Rhoades and Loveday, 1990). Below this level, salinity stress will be mostly absent, and 13Cplant values will reflect other fractionation processes. At higher EC1:5 values, however, salinity stress becomes the strongest environmental factor and 13Cplant values will reflect salinity patterns almost exclusively. Although no salinity threshold for Canarygrass is available from the literature, it will certainly be within the range of observed salinity levels in our study.
Since 13Cplant values reflect salinity stress in the crop rather than salinity in the soil, it might be useful for agricultural purposes. Soil salinity on its own has its limits for explaining yield variation, as different crops and different cultivars have different levels of salt tolerance (e.g. Handley et al., 1997). Therefore, 13Cplant values, which reflect the actual salinity stress that the plant has experienced over the season, have the potential for becoming an important diagnostic tool for salinity stress.
Salinity and Natural Abundance of Plant Nitrogen-15
The positive relation between salinity and 15Nplant is puzzling, since no clear pathway for such an effect is known. Our results seem to contradict Handley et al. (1997), who reported a salinity-induced depletion of foliar 15N in barley under controlled conditions. However, the effect reported by Handley et al. (1997), and later by Robinson et al. (2000), was observed over a variety of genotypes. It might be possible that, within one genotype, this effect would not occur, or would even be more in line with our results.
Our observed effect has to be caused either by a shift in the 15N supply (external effect) or by a fractionation in the plant because of either 14N losses, increased 15N uptake, or reallocation within the plant (internal effects).
An external effect could be because of high pH in the more saline areas of the field. Loss of NH3 through volatilization from aqueous NH+4 in the soil increases with higher pH. Both the equilibrium isotope effect for NH3 + H+ = NH+4 and the vaporization of NH3 can contribute to fractionation (Shearer and Kohl, 1986). Mariotti (1982), as cited in Shearer and Kohl (1986), found N isotopic fractionation factors (ß) for the combined effects ranging from 1.02676 to 1.02453. The 15N enrichment because of gaseous losses of NH3 can occur from decomposing plant material (Turner et al., 1983), after fertilizer application (Medina et al. [1982], as quoted in Handley and Scrimgeour, 1997), as well as after N mineralization (Mizutani et al., 1991). These effects are potentially strong enough to explain the observed positive correlation with salinity. The presence of a buffered, neutral pH of the hydroponics solution in the experiment reported by Handley et al. (1997) would prohibit the detection of such an effect. However, pH-H2O values from a 1:5 soil/water in a subset of 40 of our samples varied only from 7.6 to 8.4, and showed a very weak positive correlation with EC1:5. Although the pH values are high enough to facilitate substantial NH3 losses, the absence of a significant correlation with salinity makes this pathway less likely.
Increased denitrification in the more saline areas, possibly related to drainage (e.g., Sutherland et al., 1993), would result in an enrichment with 15N in the soil. In our case, there is no relationship between salinity and 15NSOM (Table 3). However, a recent change in the salinity pattern might account for this absence.
Several pathways for internal discrimination of 15N are described in the literature. Evans et al. (1996) reported an enrichment of 5.8 in leaves as compared with roots for tomato (Lycopersicon esculentum Mill. cv. T-5) grown on NO-3. This enrichment was attributed to fractionation of the mineral N during partial assimilation in the roots. The enriched residual mineral NO-3 will be assimilated in the leaves, leading to higher 15N values for leaves. Since we only sampled aboveground parts of the plant this effect might have been measured. The NH+4, which was all assimilated in the roots, did not show such an effect in tomato. However, this would only explain our results if the mechanism of 15N discrimination is either controlled or enhanced by salinity. Such a mechanism remains to be found.
Handley et al. (1997) observed a salinity-induced depletion of 1.5 for foliar 15N in barley under controlled conditions. They suggested, based upon work by Mariotti et al. (1982) and Yoneyama (1995), a pathway based upon stress-related down-regulation of nitrate reductase. However, since we report an enrichment of up to 15 to 17, rather than a moderate depletion, this pathway would not explain our results.
Aslam et al. (1984) reported a decrease of up to 83% in NO-3 assimilation in barley because of salinity under controlled conditions. They concluded that the decrease was almost entirely because of a decrease in NO-3 uptake by the roots, and that NO-3 and NO-2 reduction by nitrate reductase was largely unaffected. In our case, reduced uptake of NO-3 by the plant would lead to an increase in the fraction of NH+4-N in the plant. Since NH+4 is enriched because of volatilization from the soil, this would lead to higher 15Ncrop values. This effect would not have been detected by Handley et al. (1997), since they used only NO-3 as N source. Although it is not yet possible to reliably measure 15N signatures of NH+4 and NO-3 in soil, it has been suggested they can differ significantly (Handley and Scrimgeour, 1997, and references therein; Feigin et al., 1974).
Other explanations could be based upon loss of 14N from roots (efflux) or leaves (plant volatilization). Mariotti et al. (1980) showed a slight enrichment in 15N for shoots as compared with roots in several species. This was attributed to plant volatilization of N. Since volatilization of N would occur more with open stomata, this enrichment would be strongest in plants not affected by salinity stress (although% foliar N would also play a role). Therefore, N volatilization losses from leaves are at odds with our results.
Evans et al. (1996) list a number of studies where significant efflux of NO-3 from the roots was reported, leaving slightly 15N-enriched plant tissue. However, they concluded that efflux is normally minimal. Yamashita and Matsumoto (1996) reported an increase of anion efflux from roots under salinity. However, if 14N leakage from the leaves or roots because of salinity stress were an explanation for the results, the plants should show a marked decrease in the percentage of N because of salinity. However, no relation between salinity and percentage of N in the plant was found. Since the percentage of N in the plant was not significantly related to the other measured variables, these data are not shown.
In summary, the most likely explanation for the observed effect is enrichment of soil mineral N because of preferential volatilization of 15N depleted NH3 throughout the field, because of a high pH, combined with a relative increase of NH+4 uptake because of decreased NO-3 uptake by the roots in areas with higher salinity. This mechanism seems to be the only one with the potential to cause the large range of 15N values in both plant and LF.
Salinity and Isotopic Signatures in the Soil
Salinity showed an increasing trend from east to west. This is probably related to irrigation practices in the past. Our studies showed that the salinity pattern was strongly present in the 15N of the LF, which is normally associated with the more labile pool of SOM (Table 2; Fig. 3a,c). The salinity signature was mostly absent from the 13CLF (Fig. 2c) and totally absent from the 13CSOM and 15NSOM (Fig. 2d and 3d). This pattern is also clear in the variogram data, which shows decreasing ranges for modeled variograms for both 13C and 15N. This phenomenon may reflect a recent change in the salinity pattern in the field. If the salinity pattern is of recent origin, it will be expressed in the plant and (partly) in the LF. However, changing the overall isotopic signature of the SOM requires a long time (Stevenson, 1994), and therefore a salinity pattern of recent origin would not reflect itself in the SOM yet. The relatively weak negative correlation between salinity and total soil C and N (Fig. 3a, 4a, and 4b) is an additional indication that the present salinity pattern might be relatively recent. We believe that this is the most likely explanation for the absence of a salinity signature in the total SOM.
The 15N and 13C signatures in the LF of the soil are unrelated to the signatures in the whole SOM. This is clear both from the correlation coefficients (r), which are lower than 0.12 (Table 2) and from the interpolated quartile maps (Fig. 2c vs. 2d, and 3c vs. 3d, respectively). Since the LF of the soil is normally associated with the most labile, mostly recently deposited material (e.g., Stevenson, 1994), this can be explained by the recent change in salinity hypothesized above. The strong relation between 15Nplant and 15NLF (Table 2: Fig. 3b,c) contributes to this argument. However, the relation between 13Cplant and 13CLF is weaker, with a correlation coefficient of 0.39 and largely different quartile maps (Fig 2b,c). The reason for this weaker correlation remains unclear.
The added value of a spatial analysis is well illustrated by the 13C values of plant, LF, and SOM. The average values of these C pools (-26.9, -25.3, and -24.1) show an enrichment that has often been reported (Ehleringer et al., 2000 and references therein). This could lead one to conclude that the relation between 13C in these properties is straightforward, reflecting only a slight enrichment because of decomposition. However, interpolated maps of these properties (Fig. 2b–d) reveal that there is virtually no relationship at all between 13Cplant and 13CSOM, and therefore suggested other, salinity-based explanations for the observed variability. With a conventionally designed experiment, such more complex relationships could easily have been missed.
The distinctive patterns of 15N and (especially) 13C in pools of increasing age (plant < LF < SOM), offers an intriguing possibility for reconstructing the salinity history of a field. In theory, the most recalcitrant fractions of SOM should reflect 13C and 15N patterns of salinity stress in the past. Using physical fractionations, such a history could be constructed using mineral-associated organic matter in aggregates, with smaller aggregates associated with older organic matter (e.g., Tisdall and Oades, 1982). Chemical fractionations could include fulvic and humic acids (Nissenbaum and Schallinger, 1974). However, such a method would be subjected to several constraints. It would be most applicable in a field that has been continuously cropped with either a C3 or C4 crop for a long time, preferably hundreds of years. In addition, the past history of salinity has to be well documented to validate the results. In our view, an appropriate agroecosystem to test out such an hypothesis would be a long-term rice cropping system in Europe, located in a marine estuary. Such cropping systems have been established in Spain since the 8th century, were cropped exclusively with rice, and experienced different phases of salt water intrusion from the Mediterranean (Fasola and Ruiz, 1996).
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
13C and 15N values of Littleseed Canarygrass. The relation with 13C is because of partial closing of stomata during salt stress, resulting in a lower partial 12C pressure in the crop and subsequent enrichment of assimilated compounds. We hypothesize that the relation with 15N might be because of the high pH, which increases discriminating volatilization of NH3. Because of salinity, the relative NO-3 uptake might decrease, resulting in a relative increase of enriched NH+4 assimilation. However, this hypothesis has to be tested under controlled conditions. The salinity signatures could be partially followed into the LF of the SOM, but became absent in the total SOM. This led us to conclude that the salinity pattern is probably of recent origin. Using chemical or physical fractionation of SOM, it should in theory be possible to reconstruct historic salinity patterns from the more resistant fractions. |
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
13C value of halophytes—Evidence for diffusional isotope fractionation determined by the ratio of intercellular/atmospheric partial pressure of CO2 under different environmental conditions. Oecologia 52:121–124.[ISI]13C values of leaves of Plantago maritima L. developed at low and high NaCl levels. Planta 178:377–384.13C values in a C3 halophyte. Plant Cell Environ. 9:65–72.15N correlates with genotype and salt stress in barley. Planta 201:100–102.15N and 13C) to integrate the stress responses of wild barley (Hordeum spontaneum C. Koch) genotypes. J. Exp. Bot. 51:41–50.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
