| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Social Sciences Division, International Rice Research Institute (IRRI), DAPO Box 7777, Metro Manila, Philippines
b CIMMYT, P.O. Box 5186, Kathmandu, Nepal
c Central Soil Science Division, NARC, Khumaltar, Nepal
* Corresponding author (J.K.LADHA{at}cgiar.org)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: DAS, days after sowing • DAT, days after transplanting • EC, electrical conductivity • FYM, farmyard manure • IRRI, International Rice Research Institute • LTE, long-term experiment • PMN, potentially mineralizable N • TC, total C • TK, total K • TN, total N • TP, total P • WSC, water-soluble C
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
75% of the country's total food demand (Nepal Agricultual Research Council [NARC], 1995). The Tarai belt is an extension of the Indo-Gangetic plains and has an altitude of 100 to 200 m above mean sea level. The productivity and profitability of the rice–wheat system in Nepal are quite low (NARC, 1995; Hobbs and Morris, 1996). Moreover, recent studies indicated declining trends of rice and wheat yield (Brar et al., 1998; Yadav et al., 1998, 2000; Duxbury et al., 2000). It is estimated that the population of Nepal will grow at 2.7% per annum (Food and Agricultural Organization of the United Nations [FAO], 1996), which would require a doubling of rice production in the next 25 yr. Since arable land is becoming limited, improving productivity and increasing cropping intensity by adopting the rice–rice–wheat system in potentially irrigated areas will be required to meet the food demand of the region's increasing population. At present, this system is practiced in Nepal only in the subtropical Tarai and lower foothill river basins. However, maintenance of soil fertility will be essential to improve and sustain yields from this intensive cropping system. One of the main constraints to rice and wheat production in South Asia is low soil organic matter content and low indigenous nutrient supply (Nambiar, 1994; Adhikari et al., 1999). Whether the soil resources of the intensive rice–rice–wheat system can support future increases in production is of concern. Long-term experiments (LTEs) are valuable for determining yield trends, estimating nutrient dynamics and balances, understanding changes in yield, predicting soil carrying capacity, and assessing system sustainability. Several LTEs began in the 1970s in double- and triple-cropped rice and rice–wheat systems in many tropical Asian countries to monitor yield trends and system sustainability. The data of many LTEs have been analyzed, but the analyses are mostly restricted to yield trends. The analyses of some of these experiments have shown declining rice and wheat yields (Cassman et al., 1995; Nambiar, 1994; Brar et al., 1998; Yadav et al., 1998, 2000; Duxbury et al., 2000), whereas in others, yields either increased or were maintained (Aggarwal et al., 2000a; Yadav et al., 2000; Dawe et al., 2000). Where a yield decline is reported, the major causes suggested are a gradual decline in the supply of soil nutrients because of inappropriate fertilizer applications, a decline in soil organic matter content, atmospheric pollution, pest and disease infestation, and negative changes in the biochemical and physical composition of soil organic matter (Nambiar, 1994; Yadav et al., 1998, 2000).
In the present study, we analyzed results from an LTE conducted over a 20-yr period at the Regional Agricultural Research Station, Bhairhawa, Nepal, with a continuous triple cereal crop system (rice–rice–wheat). This LTE provided an opportunity to monitor changes in soil nutrient status, because the soil archives were maintained. This study aimed to (i) examine the yield trends of rice and wheat under long-term organic and mineral fertilization, (ii) identify the causes of such yield trends, (iii) monitor changes in nutrient contents of soil under continuous cropping, and (iv) estimate apparent input-output balances of N, P, and K in a few selected treatments.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The experiment included three crops per year, first rice (April–July), second rice (July–November), and wheat (November–March), with nine treatments arranged in a randomized complete block design with three replications. Plots are 6 m long and 4 m wide. For the current study, the following six treatments were used: (i) Treatment 1, control (no added nutrients); (ii) Treatment 2, N only; (iii) Treatment 3, N and P (NP); (iv) Treatment 4, N and K (NK); (v) Treatment 5, N, P, and K (NPK); and (vi) Treatment 6, farmyard manure (FYM) only (Table 1). Fertilizers were applied as per the recommendations set by the Nepal Agricultural Council when the experiment began. All the P as (NH4)2HPO4 and K as KCl were applied in Treatments 2 to 5 as basal fertilizer on the day of planting. Nitrogen was applied in two splits, 50% at transplanting of rice and sowing of wheat as (NH4)2HPO4 plus urea and the remaining 50% topdressed as urea at 25 to 30 d after transplanting (DAT) of rice and at 21 to 25 d after sowing (DAS) of wheat. Farmyard manure containing 20.0 g kg-1 N, 4.5 g kg-1 P, and 10.0 g kg-1 K (dry-weight basis) was applied 4 to 7 d before transplanting or sowing. Two rice seedlings (20- to 30-d-old) were transplanted at 20 by 20-cm spacing. Wheat (120 kg seed ha-1) was sown in rows 25 cm apart. Cultivars of rice and wheat and dates of transplanting or sowing in different years are given in Table 2. Irrigation was given in rice to maintain a submerged condition (3- to 5-cm water layer). But, in the first rice because of high evapotranspiration resulting from high temperature, the soil moisture many times dropped below field capacity. This may have resulted in differences in oxidation-reduction status of soil in two rice crops. Plots were drained 15 d before the rice harvest. In wheat, three irrigations were given at crown root initiation (21 DAS), maximum tillering (55 DAS), and flowering (85 DAS). Hand weeding was done to manage the weeds and plant protection measures were applied when needed to control pests. Crops were harvested manually close to the ground using sickles and straw was removed from the field. The grains were separated from the straw using a plot thresher. Grain yield was measured from the whole plot in both crops at maturity and was adjusted to 140 and 120 g water kg-1 for rice and wheat, respectively.
|
|
Soil Sampling and Analysis
Soil samples of three treatments (control, NPK, and FYM) were collected in Year 10, 13, 14, 15, 16, 18, and 19 to study the effect of inorganic fertilizers and FYM on soil nutrient content. Soil samples from the 0- to 15-cm layer were collected from three sites in each plot with a bottomless 20 by 20 by 65 cm (width by length by depth) core sampler at 5 to 10 d after the harvest of second rice. The entire volume of soil was weighed and mixed thoroughly and a subsample was taken to determine dry weight. The fresh soil was air-dried for 7 d, sieved through a 2-mm screen, mixed, and stored in sealed plastic jars for analysis. Representative subsamples were drawn to determine total C and total N (Jimenez and Ladha, 1993), water-soluble C (Nelson and Sommers, 1982), KMnO4-oxidized C (Blair et al., 1995), potentially mineralizable N (Kempers, 1986), total P (Olsen and Sommers, 1982), Bray P (Bray et al., 1954), total K (Knudsen, 1982), and ammonium acetate-extractable K (Council on Soil Testing and Plant Analysis [CSTPA], 1974). Total C and N were analyzed by the dry-combustion method using the Perkin Elmer 2400 CHN analyzer (Perkin Elmer, Norwalk, CT). Analyses of archived soil samples were performed at the same time in the year 2000.
Nutrient Budgets
Annual apparent N, P, and K balances were estimated for the control, NPK, and FYM treatments using different inputs and outputs measured by us during the present experiment and by others in other studies. Average crop yields were considered for apparent balance estimations. Nitrogen, P, and K balances were calculated as:
Among inputs, N, P, and K contents in mineral fertilizer, organic manure and irrigation water were measured in the present study. The N, P, and K contributions of 3.4, 0.2, and 5 kg ha-1 yr-1 with rainfall were based on the data of Mishra (1980) and Brown et al. (1999). Nitrogen input from biological N2 fixation was considered at the rate of 10 kg ha-1 during rice and 5 kg ha-1 during wheat (Brown et al., 1999). In the treatment with FYM, biological-N2 fixation was considered to be higher as amendment of organics such as FYM is known to stimulate N2 fixation (Roper and Ladha, 1995). The quantities of N, P, and K added to the soil with rice seedlings (dry weight 75.6 kg ha-1) and wheat seed (120 kg ha-1) are obtained by considering N, P, and K contents as 41.0, 4.0, and 30.0 g kg-1 in seedlings (dry weight) and 20.0, 3.6, and 4.0 g kg-1 in seeds, respectively.
Plant uptake of N, P, and K was estimated using the QUEFTS (Quantitative Evaluation of Fertility of Tropical Soils) model calibrated for rice (Witt et al., 1999) and wheat (Pathak et al., 2002). Total loss (volatilization, denitrification, and leaching) of fertilizer N was taken to be 600 g kg-1 for rice (Brown et al., 1999; Smaling and Fresco, 1993) and 500 g kg-1 for wheat (Tandon, 1994). Losses of soil N and manure N were estimated based on data reported by Smaling and Fresco (1993), Dobermann and Cassman (1997), and Kundu and Ladha (1997). We assumed that there would be no loss of P through leaching or otherwise from the soil system. Leaching loss of K was taken to be 150 g kg-1 of K input (Smaling and Fresco, 1993).
Data Analyses
Linear regression analyses were done to determine trends (slopes) of grain yield and various soil parameters over the years using SAS systems (SAS Inst., 1995). The P values, t-statistics, and 95% confidence intervals on the slopes were used to test whether the observed changes were significantly different from 0. Analysis of variance for each year was done to determine the effects of treatment on yield and soil properties by PROC GLM (SAS Inst., 1995). Means for each year were compared using Duncan's multiple range test (DMRT) for yield. The slopes and y-intercepts were compared based on their 95% confidence intervals.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
5 Mg ha-1, indicating a similar yield potential of two seasons. The control yields were lowest in Year 1 and declined to zero with time (Fig. 1)
. Treatments 2 (N) and 4 (NK) maintained high yields similar to the NPK and FYM treatments only during the first few years then followed trends similar to the control treatment afterwards, decreasing to zero eventually (data not shown). From Year 5 on, grain yield of first rice, did not exceed 0.75 Mg ha-1 in treatments (control, only N and NK) where P was not applied, and from Year 11 on, no grain yield was produced. However, the second rice crop produced no grain only in Year 18 in the absence of P (data not shown). This indicated that the supply of P largely controlled yields of both rice crops but that the supply was more crucial for the first crop. The omission of K also reduced the yield of both rice crops but the yield loss was substantially less than without P. Compared with the NPK treatment, the maximum yield reduction because of K omission was 40 to 45% in both rice crops (data not shown).
|
|
Like rice yields, wheat yields were significantly influenced by treatments. Wheat responded to NPK application through either an inorganic or organic source. Yields were consistently higher in the NPK and FYM treatments than in treatments where one or more nutrients were lacking (Table 3). The highest wheat yield of 4.9 Mg ha-1 was obtained in Year 7 in the NPK treatment (Fig. 1). However, in other years, wheat yields were 32.8 to 86.3% lower than the highest yield obtained. The control and treatments without P (NK) and without P and K (N) had the lowest yields (Table 3). Wheat yields without P was 70% of NPK treatment in Year 1 and 10% in Year 13. However, unlike rice, (i) wheat yields did not reach zero in treatments without P and (ii) the yield response to K fertilizer was much stronger in wheat. Except in one year, the wheat yields were significantly lower (P < 0.01) in Treatment 3 without K than in the NPK treatment. These results suggest that soil supply of both P and K limited wheat yield.
Wheat yield declined linearly and the change was significant in all treatments (Fig. 1 and Table 3).
Rice Nutrient Concentration
Treatments significantly influenced all the nutrient concentrations except Mn (Table 4). Potassium was noticeably below the critical level of deficiency in all the treatments, with a very low concentration in the FYM treatment. Phosphorus was deficient in all except the NP and FYM treatments. In the NPK treatment also, the P concentration was lower than the critical level of deficiency. Calcium, Mn, and B were within the optimum range in all treatments. Magnesium was within the normal range in all treatments, with the highest concentration in the NP and FYM treatments, which were slightly above the critical toxic level.
|
Total Nitrogen and Potentially Mineralizable Nitrogen
Total soil N increased significantly over time, with the highest increase in the FYM treatment (64 mg kg-1 yr-1). The changes in N over time of the control and NPK treatments were similar (Table 5). In Year 19, total N in the NPK treatment was 57% of the total N in FYM treatment, and the control had only 43% of the FYM value (Fig. 2a)
. Although the N input through FYM was lower (240 kg ha-1 yr-1) than in the NPK treatment (300 kg ha-1 yr-1), the accumulation of N in soil was higher in the former than in the latter. This could be because of the slow release of N from FYM, resulting in lower losses of N (Bhandari et al., 1992; Yadav et al., 2000). Organic fertilizers such as FYM are also known to stimulate biological-N2 fixation in soil and may also be responsible for the increase in total soil N (Roper and Ladha, 1995). In addition, treatments receiving NPK and FYM produced higher crop biomass and thus extensive roots improved the organic matter and N content of the soil compared with those not receiving NPK or FYM. Like total soil N, PMN was highest in the treatment with FYM (Fig. 2b).
|
|
|
|
|
|
We analyzed the weather data from 1977 to 1998 to see whether there was any significant change in temperature, sunshine hours per day, and rainfall pattern, which can affect the potential yield of crops. Based on linear regression analysis, we found no appreciable change in weather variables during the study period. Pest and disease incidences were properly controlled in the experiment. Analysis of soil samples showed no measurable change in EC or pH after 20-yr of cropping. Therefore, effects of weather, pest infestation, and adverse soil conditions were ruled out as the causes of a yield decline.
Although initial data on total and available soil nutrient contents are lacking, the trends from Year 10 to 20 suggest that N, P, and C pools were either maintained or increased in the NPK and FYM treatments over time (Fig. 2, 3, and 5). In FYM, all the pools were maintained at a higher level than in the NPK treatment but still, the yield decline was not arrested. Therefore, the soil data seem to suggest that N, P, and C may not be the reasons for the yield declines of rice and wheat. However, declines in total and ammonium acetate-extractable soil K in both treatments do indicate K as a possible reason for a yield decline (Fig. 4). These conclusions based on soil data tend to be consistent with the annual N, P, and K balances (Table 6). On the other hand, the results of nutrient concentration in rice tissue suggest that, in addition to poor K supply in both the NPK and FYM treatments, the P level was also below the critical level of deficiency in the NPK treatment (Table 4). Furthermore, the crop yield data in treatments without P clearly suggested that soil P supply was a key constraint limiting yield. Yields of both rice crops dropped to zero when P was omitted but two rice crops behaved differently with P deficiency. The first rice crop produced negligible to no grain after Year 5 but the second rice crop continued to produce some grain throughout, except for a few years. The differential response to P by crops could be because of changes in the oxidation-reduction status of soil resulting from the continuous submergence in the second rice crop, intermittent wetting and drying in the first rice crop, and aerobic conditions during wheat crop. Reduced soil conditions result in increased P availability because of reduction of ferric iron phosphate compounds and increased solubility of Ca-P compounds because of a pH decrease in alkaline soils (Kirk et al., 1990; Sanyal and De Datta, 1991). Differences in P availability are also possible because of increased P sorption during the drying phase (Brandon and Mikkelsen, 1979; Sah and Mikkelsen, 1986) and differences in soil P diffusion in submerged and dry soils (Turner and Gilliam, 1976). Regmi (1994) observed increased P sorption because of drainage for the same soil. The above discussion suggests that available P measured in dry soil and at one stage of the cropping system may not be a good index of P availability (Dobermann et al., 1998).
It appears that insufficient application of K primarily limited yield in both NPK and FYM treatments. Crops fertilized with NPK averaged 47 kg ha-1 yr-1 more negative K balance than the FYM plots due almost entirely to lower K application. Regmi (1994), from a K adsorption isotherm study, showed that the soil from the FYM treatment had significantly higher K sorption capacity than the soil from the NPK treatment. In addition, plant growth may have been constrained by the excess of Mg and slight deficiency of Zn (Table 4).
In the current study, we observed a considerable fluctuation in dates of sowing of wheat but not so with rice (Table 2). Using linear regression analysis, we established the effect of a delay in planting on grain yield in the NPK treatment over the years taking the earliest date of sowing of wheat (18 November) as the base date (Fig. 6) . The analysis revealed a yield decline of 0.04 Mg ha-1 for each day delay in sowing of wheat. We adjusted the actual yields with the yield reduction caused by a delay in sowing and observed that no yield decline occurred in wheat in the NPK and FYM treatments (Fig. 7) . This shows that a delay in sowing of a crop such as wheat results in a yield loss because of a rise in temperature during grain filling. Hobbs et al. (1996) reported a yield decrease of 1% for every day delay in sowing of wheat beyond the optimum sowing day (15–20 November). We simulated the effect of the planting date on grain yield of wheat using the CERES model and observed that the sowing of wheat after the third week of November would lower the yield.
|
|
It is possible to bridge the yield gap and reverse the yield decline through appropriate nutrient management. A sustainable fertilizer management strategy must ensure high and stable overall productivity and sufficient nutrient supply for potential yield increases. The present recommendation of 100, 13, and 25 kg ha-1 N, P, and K for rice and 100, 17.5, and 25 kg ha-1 for wheat is inadequate for N and K but optimum for P. The studies conducted in several other locations in Nepal showed that higher yields can be obtained by applying higher rates of N, P, and K (R. Shreshtha, unpublished data, 2001). However, determination of optimum N application rate is critical because N-use efficiency varies with method and time of application. We therefore suggest applying N based on crop demands and using tools such as the chlorophyll (SPAD) meter or the leaf color chart (Balasubramanian and Morales, 1999). Application of K by farmers in the region is typically highly inadequate. But sometimes farmers do not get a response to K application and, as it involves more cost, they are reluctant to take the risk to apply K. The government should formulate policies of fertilizer pricing to encourage farmers to use more K fertilizer based on a balanced approach to ensure sustainable productivity as it is difficult to build up soil N, P, and K once they are depleted. Further studies, however, are required to validate the recommendations in farmers' fields and evaluate the economic feasibility of the fertilizer recommendations. Such a fertilizer management strategy must be revised after some years to account for the change in indigenous N, P, and K status that may have occurred because of a positive or negative N, P, and K input-output balance.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received for publication April 18, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Girma, S. L. Holtz, D. B. Arnall, B. S. Tubana, and W. R. Raun The Magruder Plots: Untangling the Puzzle: Previously Published in Agron. J. 99:1191-1198 (2007) Agron. J., May 7, 2008; 100(Supplement_3): S-11 - S-18. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Girma, S. L. Holtz, D. B. Arnall, B. S. Tubana, and W. R. Raun The Magruder Plots: Untangling the Puzzle Agron. J., August 10, 2007; 99(5): 1191 - 1198. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Pathak, C. Li, R. Wassmann, and J. K. Ladha Simulation of Nitrogen Balance in Rice-Wheat Systems of the Indo-Gangetic Plains Soil Sci. Soc. Am. J., August 3, 2006; 70(5): 1612 - 1622. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. C. Manna, A. Swarup, R. H. Wanjari, Y. V. Singh, P. K. Ghosh, K. N. Singh, A. K. Tripathi, and M. N. Saha Soil Organic Matter in a West Bengal Inceptisol after 30 Years of Multiple Cropping and Fertilization Soil Sci. Soc. Am. J., December 2, 2005; 70(1): 121 - 129. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Fan, B. A. Stewart, W. A. Payne, W. Yong, J. Luo, and Y. Gao Long-Term Fertilizer and Water Availability Effects on Cereal Yield and Soil Chemical Properties in Northwest China Soil Sci. Soc. Am. J., May 6, 2005; 69(3): 842 - 855. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Yadvinder-Singh, Bijay-Singh, J. K. Ladha, C. S. Khind, R. K. Gupta, O. P. Meelu, and E. Pasuquin Long-Term Effects of Organic Inputs on Yield and Soil Fertility in the Rice-Wheat Rotation Soil Sci. Soc. Am. J., May 1, 2004; 68(3): 845 - 853. [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 | |||
0.05.
