| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a The Asian Vegetable Res. & Dev. Ctr., P.O. Box 42, Shanhua Tainan, Taiwan People's Republic of China
b IRRI, P.O. Box 933, Manila 1099, Philippines
c Dep. of Plant Nutrition, Technische Universität München, Freising-Weihenstephan, D-85350 Germany
schmidhalter{at}weihenstephan.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Abbreviations: AVRDC, Asian Vegetable Research and Development Center • BRCI, Bukidnon Resources Corporation, Inc. • DS, dry season • IRRI, International Rice Research Institute • GM, [legume] green manure • MMSU, Mariano Marcos State University • SOM, soil organic matter • WS, wet season
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
The objective of this study was to monitor legume GM decomposition and determine the timing and quantity of GM N release in fields grown to tomato crops (Thönnissen Michel, 1996) at three locations and two seasons (wet season, WS; dry season, DS) in Taiwan and the Philippines. In the tropical WS in Taiwan, nitrate leaching losses were estimated in tomato plots amended with GM and N fertilizer. To trace the fate of GM N at one of the three locations, 15N-labeled GM was traced in soil and labile fractions of soil organic matter.
|
Materials and methods |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Environmental Monitoring
Soil moisture was monitored with tensiometers placed in GM and control treatments, at the 15-, 30- and 45-cm depths following tomato transplanting at AVRDC and MMSU.
Decomposition Study
Nylon bags (mesh size 1 mm) containing 15 g fresh plant material (4.7–5.5 g dry wt.) were used to determine biomass breakdown of incorporated or mulched soybean and indigofera GM at AVRDC in both the WS and the DS and at MMSU in the DS only. Bags were filled with root and shoot material on the same day as GM application. Mulch treatments contained shoot material only. At the time of GM application, all bags were either buried 10 cm deep for incorporation treatments or left on the surface for the mulch treatments. Litter bags were sampled at the same dates as soil sampling for inorganic N: at AVRDC WS 0, 2, 5, 8, 14, 29, 42, 62 and 75 d after GM application, at AVRDC DS 0, 7, 21, 35, 56, 98 d after GM application, and at MMSU 0, 5, 21, 36, 58, 77, 113 d after GM application. On each date two randomly chosen bags per treatment were retrieved, oven-dried at 60°C for 48 h, and weighed. Samples were ashed by dry combustion in a muffle furnace (500°C) for 8 h to determine original ash-free dry weight remaining (Aber et al., 1990). Biomass loss data for soybean and indigofera were fitted into the first-order single exponential model Mt = M0e-kt described for litter decomposition by Wieder and Lang (1982). The higher the k-value (decomposition rate), the faster the decomposition of the organic matter. Decomposition rates were calculated for a period of 77 d in the WS and 94 d in the DS at AVRDC and 113 d at MMSU.
Inorganic Nitrogen
The effects of legume species and GM placement treatments on the quantity and the timing of N release to soil were evaluated in all six field experiments. Inorganic N in the soil was monitored in plots planted to tomato in five treatments; control (Ck0), soybean incorporation (Si), soybean mulch (Sm), and either indigofera incorporation (Ii) and indigofera mulch (Im) (at AVRDC and MMSU) or mungbean incorporation (Mi) and mungbean mulch (Mm) (at BRCI). These plots were sampled on the dates listed above for litterbag sampling and also at 0, 14, 28, 42, 56, 70, 84, 96, 110 d after GM application at BRCI. Soil samples were collected with a 5-cm-diameter auger at the 0- to 30-cm depth from the five treatments in all four blocks. Each sample was a mixed composite collected from four locations in each plot. Soil samples were passed through a 10-mm sieve and extracted with 1 M KCl (1:1.5 soil/water); inorganic N (NH4–N and NO3–N) was determined with an ammonia gas sensing electrode (Siegel, 1980). At MMSU, additional soil samples from the 30- to 60-cm soil depth were taken at -74 (legume seeding), 1, and 113 d after GM application.
To study the effect of living plants on N mineralization, the five treatment plots in Blocks I, II, and III were split into three subplot treatments after GM application: (i) unplanted, (ii) planted with tomato, and (iii) planted with cabbage (Brassica oleracea var. capitata L.) in the DS at AVRDC; and (i) unplanted, (ii) planted with tomato 1 d after GM application, and (iii) planted with tomato 2 wk after GM application at MMSU. Nitrogen mineralization was monitored in all three subplot treatments.
Estimation of Potential Nitrate Leaching
Nitrate leaching in the WS at AVRDC was estimated by the NaCl method (Cameron and Wild, 1982). Fifty grams of NaCl was broadcast on 1 m2 in the tomato plots in the treatments Si, Sm, Ck0, and 120 kg N ha-1 (Ck120) in four replications in two bed systems, low or raised beds (Thönnissen Michel, 1996). Sodium chloride was applied on the respective plots after soybean incorporation and mulch on 23 June 1993. Soil samples were taken on 21 June, 23 July, and 30 August from the 0- to 50-cm layer in the raised beds and from the 0- to 30-cm layer in the low beds. In each, the soil core was separated into 10-cm sublayers. Soil samples were air-dried and extracted (1:2 soil/water). Chloride in the water extracts was determined with a chloride analyzer (Chloride Analyzer 926, Coramed AG, Dietlikon, Switzerland).
Nitrogen-15 Experiment
Tomato N response to GM was low in the DS at AVRDC (Thönnissen Michel, 1996). To understand the fate of GM N, soybean GM was labeled with 15N (Thönnissen Michel, 1996) for a 15N microplot experiment at MMSU. Microplots (metal frames 0.8 by 0.8 by 0.3 m, length by width by height, pushed into the soil to a depth of 25 cm) were amended with 15N-labeled soybean GM. The GM was incorporated manually down to the 10- to 15-cm soil depth. Two tomato seedlings were transplanted into each microplot. Green manure 15N recovery in tomato was determined (Thönnissen Michel, 1996). Soil was sampled for organic matter extraction and soil 15N determination in control and soybean incorporation treatment plots at 1 and 113 d after GM application. Nitrogen-15 determination was conducted on mobile humic acids (MHA) and calcium humates (CaHA), which were considered as C pools representing early and later stages of the humification process (Olk et al., 1995).
Statistical Analysis
Data were analyzed by ANOVA procedure using JMP Version 2 (SAS Inst., 1989) and SAS version 6.03 (SAS Inst., 1991).
|
Results |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
|
|
Nitrate was the dominant form of inorganic N in the soil soon after legume application at all three locations. Soil NH4–N contents remained low (±5 kg NH4–N ha-1 at AVRDC and MMSU; ±20 kg NH4–N ha-1 at BRCI) and were comparable to those of the control (data not shown). Green manure application increased soil NH4–N contents significantly by 10 to 15 kg NH4–N ha-1 at AVRDC, 5 kg NH4–N ha-1 at MMSU, and 30 kg NH4–N ha-1 at BRCI in the first week after GM application, but NH4–N declined rapidly within 3 wk. With the exception of an increase in soil NH4–N by 1 to 8 kg NH4–N ha-1 in the low beds in the DS at AVRDC, NH4–N contents did not differ between planted and unplanted plots in the raised beds at AVRDC and at MMSU.
At all three locations, N release in soil peaked at 80 to 120 kg NO3–N ha-1 with soybean GM (Fig. 2) . This peak N release occurred 2 to 6 wk after GM application in both seasons at AVRDC and at BRCI. At MMSU, GM N release peaks were delayed relative to the two other locations, occurring after 5 to 8 wk. Nitrate contents declined after 5 to 8 wk at all locations. More NO3–N was released with incorporated GM than mulched GM at AVRDC and MMSU. Far more N was released with soybean than indigofera in the WS at AVRDC; in the DS, however, differences in N release between legume species were small. Nitrate released with soybean GM was comparable to that released with mungbean GM at BRCI. Basal N mineralization (NO3–N) in control treatment plots was low in the WS at AVRDC and at MMSU, but high in the DS at AVRDC and at BRCI.
|
Nitrate Leaching
The potential for nitrate leaching estimated from the movement of chloride followed similar patterns in control, 120 kg N ha-1, soybean mulch and incorporation treatments. Therefore, chloride loss (%) data of these four treatments were averaged for each sampling date and bed system (Table 2)
. The background Cl-concentration (21 June) in the 10- to 50-cm soil depth was rather low. Of the applied chloride, 42 and 50%, had been lost by 23 July 1993 from a soil depth of 30 and 50 cm, respectively, in the raised bed only 1 mo after application. The greatest net loss occurred at the 0- to 10-cm soil depth, whereas chloride accumulation occurred at soil depths of 10 to 20 cm and 20 to 30 cm. Chloride did not accumulate at the 30- to 50-cm depth in the raised beds.
|
|
At tomato harvest, estimations of N losses were greater calculated with 15N than with total N (Table 4) , due to lower N recoveries of 15N in both tomato and soil. Nitrogen-15 values for whole soils, MHA, and CaHA for all treatments except soybean at tomato harvest were too low to allow accurate measurement.
|
|
Discussion |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
In most comparisons, plant chemical composition appeared to affect the decomposition rate of GM. Faster decomposition of indigofera at AVRDC was probably caused by its smaller and more tender leaves and less lignified stems relative to those of soybean. The slower decomposition of incorporated soybean compared with indigofera at MMSU occurred despite similar plant chemical compositions (data not shown). Sixty-day-old soybean (maturity scale R5 to R6; Fehr et al., 1971) in both seasons at AVRDC decomposed at rates similar to those of incorporated indigofera at MMSU. The physical nature of older soybean plant material (R6 to R7) used at MMSU, with hardy stems and pods containing full size yellow beans, may have been one of the main reasons for the large differences in decomposition rates between soybean decomposition at AVRDC and indigofera at MMSU.
Many investigators have observed that organic residues decompose more slowly in soils with higher clay contents, especially clays having higher exchange capacities (Lynch and Cotnoir, 1956; Sorensen, 1975). Microbial activity is controlled by soil physical conditions such as compaction, temperature and oxygen; by chemical conditions such as substrate availability; and by biological conditions such as predatory or antagonistic organisms (Grant et al., 1993). Reduced soil aeration or oxygen in the clayey soil at MMSU compared with the loamy soil at AVRDC may have further contributed to a slower legume residue decomposition rate at MMSU.
The exponential weight loss pattern agrees with previous assumptions that residues contain labile and recalcitrant fractions having different degrees of resistance to microbial degradation. Reinertsen et al. (1984) associated the more rapid decay immediately after the burial of the residue with the decomposition of water-soluble organic constituents. Hunt (1977) described differences in decomposition patterns and rates among substrates as a function of the amount of the labile or rapidly decomposing fractions (sugars, starches, proteins) and the recalcitrant or slowly decomposing fraction (cellulose, lignin, fats, tannins, waxes). Decomposition processes can be predicted from initial litter chemistry (Aber et al., 1990; Neely et al., 1991). Seasonal effects on chemical composition of legumes (i.e., C/N, initial N, lignin, polyphenol, and tannin contents) have been shown within the same location (Thönnissen Michel, 1996). The statistical significance of each chemical component to the rate of GM degradation varied widely between seasons and locations (Thönnissen Michel, 1996). The relatively high polyphenol (3.7%) and tannin (1.6%) content of indigofera may have retarded decomposition compared with soybean (polyphenol 1.7%, tannin 0.2%) in the WS, whereas in the DS the lower C/N-ratio (10.6) and higher initial N content (4.2%) of indigofera may have determined its faster decomposition compared with soybean (C/N 12.2; N 3.9%). Results of this study confirm the complexity of decomposition processes where the interaction of both resource quality and microclimate influence the conditions and activity of decomposer communities and those in turn mediate processes of decomposition and nutrient release (Neely et al., 1991; Hunt, 1977).
High soil temperatures (20–30°C) and moisture conditions near optimum (-0.01 to -0.05 MPa; Cassman and Munns, 1980) were mainly responsible for the fast release of NO3 following GM application in all locations and seasons. Nitrate-N release at AVRDC mirrored the initial exponential loss of biomass, evidence for the causal linkage between these two processes. Higher decomposition rates of indigofera in the DS led to N release in soil comparable to that of soybean, although far less N (33 vs. 127 kg N ha-1, Thönnissen Michel, 1996) was incorporated with indigofera GM. Reduced mineralization rates of surface applied residues (mulch) can be attributed to poor soil–residue contact and drastic temperature and moisture fluctuations at the soil surface (McCalla and Duley, 1943). Numerous authors (e.g., Janzen and McGinn, 1991) have stressed the importance of volatilization losses when GM is applied as surface mulch, since drying and decomposing conditions enhance volatilization. The volatile loss of labile N from decomposing GM mulch may appreciably diminish its fertility benefit, whereas NH3 losses from incorporated GM have been reported to be negligible (Janzen and McGinn, 1991). If, however, lower mineralization rates are the cause of reduced inorganic N accumulation under no-tillage, then such a system could better conserve organic N in the long term (Sarrantonio and Scott, 1988). Slight increases in NO3 contents in the soil 10 wk after GM application in the WS at AVRDC and at BRCI (Fig. 2) may indicate remineralization of N that had been immobilized earlier, even though the process is considered to be relatively slow in temperate soils (Mary and Recous, 1994). Lowest NO3–N contents in the soil with indigofera mulch were likely due to the NO3–N uptake of the indigofera (Thönnissen Michel, 1996).
Although N release dynamics may have been driven by a combination of location and/or season-specific factors, N release patterns across locations and seasons are similar. High leaching and denitrification losses in the WS at AVRDC may have reduced the amount of GM N available to tomato plants, although temperature and soil moisture were more favorable for N mineralization than in the DS. Results of an incubation study comparing N release after addition of dry organic residues to these three soils (Thönnissen Michel, 1996) suggested that certain soil chemical and physical properties retarded N release in MMSU soil, relative to BRCI and AVRDC soil. Soil basal N mineralization was higher in the AVRDC and BRCI soils than in the MMSU soil (Thönnissen Michel, 1996). Significant amounts of inorganic N were detected in fallow plots lacking GM addition prior to vegetable crops in the DS at AVRDC and at BRCI, while leaching losses may have prevented nitrate accumulation in the WS at AVRDC. The higher the soil N supply, the more legumes derive N from soil rather than from biological N2 fixation. Low NO3 contents in legume plots can be explained by the effectiveness of legumes to assimilate NO3 derived from soil N mineralization (George et al., 1994; Ladha et al., 1996).
At all locations and seasons, the decline of soil inorganic N at 6 to 8 wk after GM application may result from a combination of the period of greatest N uptake by the tomato plants (Thönnissen Michel, 1996), lower rates of N mineralization (Griffiths et al., 1994) and biological N immobilization (Mary and Recous, 1994). The faster decline of soil nitrate in planted compared with unplanted plots suggests vegetable N uptake at AVRDC and at MMSU. Using 15N-labeled residues in the absence of growing plants, Chotte et al. (1990) found net immobilization in the organic residues, but net mineralization occurred when plants were grown in these soils. Root exudates of vegetable crops may have been an insignificant energy source for soil microbial growth (Martens, 1990) in our experiments because of the high degradability of our soybean and indigofera GM, and the favorable soil temperature and moisture conditions. Reduction of microbial activity and microbial N immobilization after consumption of the labile fractions of the residue in early decomposition stages may have occurred due to the recalcitrance of the remaining crop residue. It is possible that these recalcitrant organic fractions lead to the formation of soil humus (Wilson and Hargrove, 1986). If microbial N needs were large, available soil inorganic N would be rapidly depleted and the decomposition rate of organic compounds would decline (Mary and Recous, 1994), leading to N immobilization (6–8 wk) and delayed N remineralization. In experiments by Broadbent and Tyler (1962), NO3 was immobilized to a considerable extent when it was the only N form available to soil microorganisms. Mary and Recous (1994) described N immobilization–remineralization following organic residue incorporation as a function of the amount and nature of the residues and soil mineral N, whereas basal mineralization was explained as a function of soil texture and long-term C and N inputs.
It is probable that liming of the soil and the addition of poultry manure (Gallus sp.) led to a strong soil N mineralization in BRCI soil. Decay of plant residues and SOM are accelerated by liming of acid soils (Alexander, 1977).
The great loss of Cl and ostensibly NO3 within the first month of GM and N fertilizer application in the WS at AVRDC, probably resulted from two rainfall events within that period. The soil at >30 cm depth in the raised bed system was permanently submerged due to the standing water in the rice beds (Thönnissen Michel, 1996), so that Cl may have been leached with rice bed irrigation. Improved infiltration rate through soil in the raised beds may have also increased leaching losses (Shennan, 1992). Our results confirm those of Stute and Posner (1995), that potentially leachable soil NO3–N differed little following GM or N fertilizer.
Total Nitrogen and Nitrogen-15 Balance
The similarities of the ratios of total 15N to total N for MHA and CaHA fractions compared with humin suggest that the two fractions were no more labile than the rest of the SOM in this soil. Our results are comparable to those of He et al. (1988), in that a significant proportion of recently added 15N in the soil was not extractable (humin). Humin can be very young and much of it is composed of alkyl compounds and carbohydrates as microbial byproducts (He et al., 1988). Domination of soil C and N by humin may be especially pronounced in a soil where conditions are favorable for degradation. The rapid decomposition of the soybean residues and the small quantities of MHA and CaHA extracted from the MMSU soil in relation to other rice soils of the Philippines (Olk et al., 1995) demonstrate the favorable conditions for degradation in this soil. Organic molecules resulting from microbial degradation, such as microbial tissues, will be preserved in the soil only if they are stabilized and thereby are protected from further degradation. One such form of protection is chemical binding to the mineral surface of such strength that the organic material is not extractable and hence considered as humin. The extremely high Ca levels in MMSU soil may also contribute to the humin constituting a high proportion of total SOM.
Lower rates of 15N recovery could be due to mineralization–immobilization turnover. The 15N released from the legume residue into the soil inorganic pool could be exchanged for 14N in microbial biomass, which could lead to lower 15N recovery. Alternatively, lower rates of 15N recovery than total N may result partly from an overestimation of apparent total N recovery and partly from the importance of soil conditions during the rapid degradation of 15N-labeled material. Higher mineralization rates of labeled than of unlabeled organic materials may have contributed to lower rates of N recovery in 15N compared with total N balances, in agreement with other authors (Amato and Ladd, 1980; Chichester et al., 1975). Greater loss of 15N than total N (57 vs. 30%) may reflect volatilization and denitrification at the beginning of the crop cycle, as well as the low plant N uptake at that time. Total N loss would be lower on a percent basis during this time because of the low basal rate of SOM-N mineralization. Because labeled soybean was quickly decomposed, the fate of GM 15N would be disproportionately determined by soil conditions early in the crop cycle. Given the relatively wet yet aerated conditions in the MMSU soil, these large amounts of 15NH4 mineralized quickly from plant residues or young SOM fractions would be prone to losses via volatilization or nitrification–denitrification. This scenario is supported by the low total soil C and N levels, small quantities of extracted MHA and CaHA, high losses of 15N from the system, and the greater relative loss of 15N than total N.
Given its low total C and N contents, the MMSU soil may not have a large capacity to store added N, whether the N is added in organic or inorganic forms. We conclude that the lack of synchronization between N supply and demand caused by a single application of GM shortly before vegetable transplanting makes this treatment less successful than split applications of inorganic N (Thönnissen Michel, 1996).SAS Institute 1989; SAS Institute 1991
Received for publication August 12, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. T. MacKown, J. J. Heitholt, and S. C. Rao Agronomic Feasibility of a Continuous Double Crop of Winter Wheat and Soybean Forage in the Southern Great Plains Crop Sci., July 30, 2007; 47(4): 1652 - 1660. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Cherr, J. M. S. Scholberg, and R. McSorley Green Manure as Nitrogen Source for Sweet Corn in a Warm-Temperate Environment Agron. J., August 3, 2006; 98(5): 1173 - 1180. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. C. Olk A Chemical Fractionation for Structure-Function Relations of Soil Organic Matter in Nutrient Cycling Soil Sci. Soc. Am. J., April 19, 2006; 70(3): 1013 - 1022. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Cherr, J. M. S. Scholberg, and R. McSorley Green Manure Approaches to Crop Production: A Synthesis Agron. J., February 7, 2006; 98(2): 302 - 319. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Crop Science | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
