| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. de Ciencias y Recursos Agrícolas y Forestales, Univ. of Córdoba, P.O. Box 3048, Córdoba, Spain
b Dep. de Ciencias y Recursos Agrícolas y Forestales, Univ. of Córdoba, P.O. Box 3048, Córdoba, Spain
c Dep. de Producción Vegetal, Univ. of Extremadura, Badajoz, Spain
d F.J. 
opez-Bellido, Dep. de Producción Vegetal y Tecnología Agraria, Univ. of Castilla-La Mancha, 13071 Ciudad Real, Spain
cr1lobel{at}uco.es
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMaterials and methodsResults and discussionGrain Yield ComponentsTotal Dry Matter and...SummaryREFERENCES |
|---|
Abbreviations: ANOVA, analysis of variance • CP, chickpea • CT, conventional tillage • CW, continuous wheat • F, fallow • FB, faba bean • LSD, least significant difference • NT, no tillage • ROT, rotation • S, sunflower • TILL, tillage • W, wheat • WCP, wheat–chickpea • WF, wheat–fallow • WFB, wheat–faba bean • WS, wheat–sunflower
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionGrain Yield ComponentsTotal Dry Matter and...SummaryREFERENCES |
|---|
Mediterranean climate is characterized by a high seasonal variability of rainfall, with 85% of annual rainfall ocurring during the months of October to April. Mediterranean soils contain little organic mater (9–12 g kg-1) and are highly prone to wind and water erosion due to the presence of irregular steep slopes. The factors most strongly influencing crop yields—particularly grain yields—in the region are soil moisture and N, the former of which depends on rainfall and its distribution during the growing season (Cooper et al., 1987).
One of the more important rainfed Mediterranean regions, the so-called "Campiña", is located in Andalusia (southern Spain). This region, characterized by Vertisols, produces mostly wheat at yield levels greater than the Mediterranean area average (2–2.5 Mg ha-1). Farmers in the region use high levels of inputs, particularly N fertilizer. Typically, wheat is rotated with sunflower, and to a lesser extent faba bean and chickpea. Bare fallow is seldom used on these soils, which are fertile and possess soil texture favorable for water retention. Fallow is only encountered as a result of set-aside measures of the European's Community Agricultural Policy Reform. Conventional tillage with moldboard plowing is commonly used but conservation tillage practices have lately been introduced.
The long-term effects of conservation tillage and no tillage, under rainfed Mediterranean conditions, have scarcely been studied. By contrast, no tillage practices in other areas such as USA, Australia, and Canada are well-documented (Baker et al., 1996). The effects of tillage on wheat yield under rainfed conditions have been thoroughly studied in the U.S. Great Plains (Halvorson and Reule, 1994; Norwood, 1994; Unger, 1994; Wiese et al., 1994). No tillage has been found to increase soil moisture and diminish soil erosion. This, however, has not always been found to be beneficial for grain yield. Rao and Dao (1996) found the no tillage system to occasionally diminish yield through decreased N availability. Potential causes of this yield depression included slow mineralization, increased N immobilization, denitrification, leaching, volatilization, and surface run-off losses (Terman, 1979; Phillips et al., 1980; Rice and Smith, 1984). According to Rasmussen et al. (1997), high levels of cereal residues on the soil surface can reduce wheat yield, with the reduction variously attributed to disease, weed competition, or decreased light intensity. Hay et al. (1978) claimed that surface residues lower soil temperature by 2 to 6°C during early spring.
Under a dry climate, Vertisols pose unique soil moisture problems as impacted by tillage (Probert et al., 1987). Intensive conventional tillage is known to degrade soil structure; however, the sustained use of no tillage with fine-textured soils can result in adverse physical conditions (e.g., soil compaction) and poor drainage (Probert et al., 1987; Blevins and Frye, 1993).
The positive effects of crop rotations over cereal monocultures are also well-documented (Pierce and Rice, 1988; Mcewen et al., 1989; Christen et al., 1992). In dry areas, the grain–fallow system uses water and N less efficiently than when wheat is grown in rotations with other plants (Halvorson and Reule, 1994; Unger, 1994). Loomis and Connor (1992) identified a Mediterranean climate in southern Australia as an example of an area ideal for incorporating legumes in wheat rotations, and they claimed that the efficiency of fallow depends on the particular soil type, rainfall and its distribution, and evaporation. Yields are higher when cereals follow a legume as this saves N and breaks the disease cycle of grains. The effect of the preceding legume on grain yield is frequently quantified in terms of the amount of N required to obtain an equivalent yield from a grain monoculture (Herridge, 1982; Peterson and Varvel, 1989).
Papastylianou (1993) and López-Bellido et al. (1996) found legumes to increase grain yield under rainfed Mediterranean conditions. Sustainable agriculture has revitalized the interest in crop rotations and their effect on N use efficiency to promoting profitable, and efficient, agriculture. The amount of N fertilizer required under semiarid climates is largely dictated by the seasonal rainfall (Myers, 1984). Long-term experiments have revealed a wide variability in the response of wheat to N fertilizer; however, very few of these experiments have included evaluation for more than 3 to 5 yr that also accounted for acumulation of and within the soil profile (Westerman et al., 1994). Under semiarid conditions, periods of intensive accumulation of mineral N can be followed by N depletion, depending on rainfall and crop growth. According to Corbeels et al. (1998), the carryover effect of N fertilizer from a growing season to the next in soils under rained Mediterranean conditions can be substantial. Losses of residual N fertilizer between cropping and the following growing season are usually small. Some studies have revealed changes in the efficiency of available N under reduced tillage and no tillage conditions (Peoples and Herridge, 1990; Rao and Dao, 1992; Campbell et al., 1993).
According to Cassman et al. (1995) and Steiner (1995), the results of long-term experiments are reliable indicators of sustainability. In fact, such experiments are amenable to data extension and to pattern recognition in crop yields and productivity indices for agricultural systems. They also allow determination of causal relations before they are perceived by farmers.
Previous work (López-Bellido et al., 1996) had documented the results of tillage, rotation, and N rate on wheat yield for average and drier than average years (1987–1994). Our interest was to assess these management practices during a series of wet years (1995–1998) as well as throughout the long-term experiment (1986–1998) in a rainfed Mediterranean agricultural system.
|
Materials and methods |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionGrain Yield ComponentsTotal Dry Matter and...SummaryREFERENCES |
|---|
No tillage plots were seeded with a no-till drill. Weeds were controlled with glyphosate [N-(phosphomethyl)glycine] + MCPA [(4-chloro-2-methylphenoxy)acetic acid] at a rate of 0.5 + 0.5 L a.i. ha-1 before planting. The CT treatment included moldboard plowing and disk harrowing and/or vibrating tine cultivator several times to prepare a proper seedbed.
Hard red spring wheat (cv. Cajeme) was planted in 18 cm wide rows in December at a seeding rate of 150 kg ha-1. Sunflower (various hybrid cultivars) was planted in 50 cm wide rows in February at a seeding rate of 5 kg ha-1. Winter chickpea (CV FLIP-84/1SC) was planted in 35 cm wide rows in December at a seeding rate of 80 kg ha-1. Faba bean (cv. Alameda) was planted in 50 cm wide rows in November at a seeding rate of 170 kg ha-1. Fallow weeds were controlled with glyphosate + MCPA at the above-stated rates two or three times throughout the year in NT and by spring-tooth harrowing in CT.
Nitrogen fertilizer was applied to wheat plots as ammonium nitrate. At all application rates, half was applied before sowing (incorporated by disk harrowing in CT plots and surface-broadcast in NT plots). The remaining N was applied as top dressing at the beginning of wheat tillering corresponding to stage 21 of Zadoks scale (Zadoks et al., 1974).
Weeds within the growing season were controlled by means of specific herbicides, namely: diclofop methyl [2-(4-(2,4-dichlorophenoxy)phenoxy)propanoic methyl] + tribenuron [methyl-2-(((N-4-methoxy-6-methyl-1,3,5-triazin-2-yl)-N-methylamino)carbonyl)amino)sulphonyl)benzoate] at 0.9 L a.i. ha-1 and 15 g a.i. ha-1, respectively, for wheat; and cyanazine [2-(4-chloro-6-ethylamino-1,3,5-triazin-2-yl-amino)-2-methyl propionitrile] at 2 kg a.i. ha-1 for chickpea and faba bean. Glyphosate was applied at rate of 0.065 L a.i. ha-1 as a postemergence spray on faba bean plot when broomrapes (Orobanche crenata Forsk) were about 0.5 to 1 cm high (García-Torres et al., 1987). Every year wheat plots were also supplied with P fertilizer at a rate of 65 kg P2O5 ha-1, the fertilizer was incorporated in CT soil and banded with drilling in NT plots. Soil-available K was adequate. At harvest a 0.5 m2 portion at the center of each wheat sub-subplot was sampled (harvest years 1996–1998 only). From this sample total dry matter and grain weight were determined by drying the sampled plants at 80°C to constant weight. The harvest index (HI, the ratio at harvest of grain dry wt. to total aboveground dry wt.) and yield components (heads m-2, seeds m-2, and mean seed wt.) were measured. Wheat grain was harvested in early June using a 1.5 m wide Nurserymaster Elite Plot Combine (30 m2 per sub-subplot).
In the last 3 yr of the study, soil samples were taken on all plots in autumn, before wheat-sowing, to a depth of 90 cm. Soils were analyzed for nitrate content using the Griess-Illosvay colorimetric method as modified by Bremner (1965).
Annual wheat yields (1995–1998), total dry matter, HI, and yield components (1996–1998) were subjected to analysis of variance (ANOVA) using a randomized complete block design combined over years, according to Mcintosh (1983). Treatment means were compared using Fischer's protected least significant difference (LSD) test at P 
0.05. Least significant differences for different main effect and interaction comparisions were calculated using the appropriate standard error term according to Gómez and Gómez (1984). The results of ANOVA of wheat grain yield for the period from 1987–1988 to 1993–1994, and from 1987–1988 to 1997–1998 (the latter of which included the 4 yr of this study) were compared. The Statgraphics Plus v. 7.0 software suite (Manugistics, 1993) was used for this purpose.
|
Results and discussion |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionGrain Yield ComponentsTotal Dry Matter and...SummaryREFERENCES |
|---|
|
|
Grain Yield
Response of wheat yield to tillage methods, crop rotation, and N fertilizer varied with prevailing weather conditions in the particular growing season (Tables 2 and 3)
. Overall, only the 1995–1996 season had high yield (4.08 Mg ha-1). Yields for the 1997–1998 season was less (2.47 Mg ha-1) and considerably less for the 1994–1995 and 1996–1997 seasons (1.09 and 1.58 Mg ha-1, respectively).
|
|
|
Wheat yield varied significantly with crop rotation, both for the individual years and for the 4-yr as a whole (Tables 3 and 4). The WFB rotation resulted in the highest grain yields during this research. The other rotations (WS, WCP, and WF) produced similar grain yields, and continuous wheat (CW) resulted in the lowest yields. This trend was also observed in the results for the individual years, albeit with some differences resulting from the effects of rainfall during growing season (Table 3). During predominantly wet years (3 of the 4 study years), differences among rotations were minimal since water was not limiting and differed from results of López-Bellido et al. (1996) for the dry years of this long-term study. Even under wet conditions, however, faba bean was found to exert a positive residual effect on grain yield. By contrast, chickpea had little positive effect on yield from the beginning of the experiment (López-Bellido et al., 1996). Differences among legume species in their ability to fix atmospheric N2 and hence the amount of residual N that is supplied to soil have been reported to influence cereal yield (Herridge, 1982). During the 4 yr of this study wheat monoculture was significantly lower than those systems in rotation with other crops. In contradiction with the result of Mcewen et al. (1989), Power (1990), and Christen et al. (1992) under different growing conditions, we observed no visible signs of an increased disease incidence in continuous wheat.
The N fertilizer rate exerted a significant influence on wheat yield when averaged across years as well as in three out of the four individual years of this study (Tables 2, 3, and 4). Differences in grain yield among N fertilizer rates were significant except between the 100 and 150 kg ha-1 treatments (Tables 2 and 3). These results are consistent with those obtained by López-Bellido et al. (1996) in the previous years of this long-term study that showed that wheat failed to respond to N fertilizer in dry years but responded in wet years, even when there was a carryover effect of fertilizer N from one growing season to another. Work by Corbeels et al. (1998) found N carryover quite substantial under semiarid Mediterranean conditions. Wheat yield response was more pronounced by the introduction of the zero N fertilizer rate from 1994–1995. Differences between 100 and 150 kg ha-1 were not significant as a result of the gradual increase in the amount of available N in the system as a whole (i.e., the combined effects of fertilization, rotation, and tillage), all in agreement with the results of Campbell et al. (1993).
Analyses for soil nitrate carried out in the autumn of 1995, 1996, and 1997 at depths between 0 and 90 cm under the different treatments (only means presented here) provided average contents of 50, 54, and 92 kg N ha-1, respectively. The WS rotation gave the lowest contents, with 41, 43, and 72 kg N ha-1, and CW (to which N fertilizer was applied every year) the highest, with 57, 61, and 112 kg N ha-1 in 1995, 1996, and 1997, respectively. The marked increase in soil content during 1997 can be ascribed to intensive mineralization of organic matter in 1996 and 1997 due to favorable soil moisture during fall months. We presume the droughty conditions of the previous years (between 1992–1993 and 1994–1995) hindered mineralization. Slight differences in soil nitrate among N fertilizer rates (data not shown) indicate potential N loss (e.g., volatilization, leaching, and runoff through Vertisol cracks). However, grain N removal as well as immobilization in the soil was presumed greater with more intensively fertilized plots (MacMahon and Thomas, 1976; Terman, 1979; Corbeels et al., 1998).
Compared with NT, grain yield under CT was higher at all N rate applied to wheat. In constrast to CT, yield for the NT increased significantly when the N rate increased from 100 to 150 kg N ha-1 (Tables 2 and 4). From these same plots, López-Bellido (1998) found significantly higher contents in the 0 to 90 cm deep soil profile under CT than under NT, with the difference increasing with time. The decreased availability of mineral N in the NT system can be ascribed to a number of documented processes including slower mineralization, increased mobility of N, denitrification, leaching, volatilization and runoff (both across the surface and through Vertisol cracks) (Terman, 1979; Phillips et al., 1980; Rice and Smith, 1984).
With CW and WS, the effect of N fertilizer on wheat yield was more marked since there was no residual N contributed by a legume crop. Wheat yield for WFB at zero N was significantly higher than yield for CW at any N rate. While generally greater, WFB was not significantly different from yield for the WS, WCP, and WF rotations at any N rate (Tables 3 and 4). In the wet years, differences in wheat yield between CT and NT was greater at lower N rates (Tables 2 and 4). This is consistent with the increased N mineral contents found under CT relative to NT.
Finally, the significant YR x ROT x N rate interaction (Tables 3 and 4) indicates that in wet years (1995–1996 and 1997–1998), which correspond to the higher yields, N application resulted in yield differences between rotations. These differences were not observed in the years where yield was lower, either due to drought (1994–1995) or to excesive rainfall (1996–1997) (Table 3).
|
Grain Yield Components |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionGrain Yield ComponentsTotal Dry Matter and...SummaryREFERENCES |
|---|
|
The number of heads m-2 increased with increasing N rate for the studied period as a whole, with no difference between the rates 100 and 150 kg N ha-1. The number of seeds head-1 was also greater at the higher N rates. On the other hand, seed weight decreased slightly with increasing N fertilizer rate, even though differences between rates were not always significant (Tables 4 and 5). Rasmussen et al. (1997) also found increasing N fertilizer rate increased the number of heads, seeds head-1, and decreased seed weight.
The interaction effects for grain yield components are shown in Fig. 2 and 3 . The number of heads m-2 increased only with increased N rate in NT (Fig. 3c), but not all N rates displayed significant differences between the two tillage systems. The effect of rotations on number of heads m-2 also differed depending on N rate (Fig. 3e); increases in the N rate were not always associated with a greater number of heads m-2. The WF rotation yielded the clearest response to higer doses of N. Seeds head-1 was the only component where year displayed a significant interaction with all treatments (Table 4, Fig. 2c, d, and e). In the NT system, rotations including legumes and fallow yielded a higher value for seeds head-1 than other rotations (Fig. 3d), whereas the highest value using CT was found for wheat monoculture, although differences were not significant. The rotation x N rate interaction showed that number of seeds head-1 differed only in wheat monoculture (Fig. 3f), with higher values recorded at doses of 100 and 150 kg N ha-1 than at 0 and 50 kg N ha-1. Finally, seed weight varied between extreme N rates only in 1997 (Fig. 2f).
|
|
|
Total Dry Matter and Harvest Index |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionGrain Yield ComponentsTotal Dry Matter and...SummaryREFERENCES |
|---|
|
Total dry matter behaved similary to grain yield; it peaked at the N fertilizer rates 100 and 150 kg N ha-1 and varied little among rates (Table 6), although years also exerted a marked influence (Fig. 2b).
Effects of the 1994–1995 to 1997–1998 Period on Long-Term Results
Assessing consistency in the results of a long-term experiment entails gradually incorporating new data and determining whether their inclusion alters results through changes in the cropping conditions over time. In this study, the new data allowed for an examination of the effect of alternate wet and dry periods under semiarid Mediterranean conditions. Compared with the initial 7-yr study (1987–1988 to 1993–1994) of this long-term experiment (López-Bellido et al., 1996), which was predominantly dry, 3 of the 4 subsequent years (1995–1996 to 1997–1998) had above average precipitation (Table 1) and provided an opportunity to examine now inclusion of these wet years affected the overall long-term results on grain yield.
Table 7 gives the results of the ANOVA for the two periods studied. Main effect of rotation and N fertilizer rate were significant in a similar way for both two periods. Tillage was not significant during the former period (1987–1988 to 1993–1994), but was significant with the wet years included because grain yield under CT was greater than NT during these years (Tables 7 and 8) . The CW and WS rotations responded similarly in both periods, with significantly lower grain yields in the former rotation (Table 8). However, inclusion of the wetter years resulted in relative greater decrease grain yield with WCP in the latter period relative to the former (Table 8). On the other hand, the wet years had a favorable effect on WFB, which exhibited significantly high grain yields relative to WF in the latter period (Table 8).
|
|
The traditional use of high N fertilizer rates by farmers in the region therefore appears to be wasteful; while part of the N applied may be retained by the soil profile that is explored by roots, N also can easily be lost through denitrification or through runoff and leaching in the rainy years typical of the Mediterranean climate.
As can be seen from Table 7, interactions among treatments (tillage, crop rotation, and N rate) responded similarly. Only the YR x TILL x N interaction, insignificant during the former period, was significant in the latter. The presence of wetter years, which promoted a favorable response of wheat to N fertilizer, offset the difference in number between dry and wet years in the studied period (11 yr) and established interactions between them and both N fertilizer and tillage system.
|
Summary |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionGrain Yield ComponentsTotal Dry Matter and...SummaryREFERENCES |
|---|
The 2-yr WFB rotation was the most productive rotation. Including a legume itself did not improve production (e.g., WCP). The wheat monoculture exhibited the lowest yield, but this was not a factor of pests and disease incidence.
The response of wheat to the N fertilizer was absent in the dry years but significant up to 100 kg N ha-1 in the wet ones. The carryover effect of N fertilizer under the rainfed conditions of Mediterranean Vertisols can be substantial. The traditional use of high rates of N fertilizer by farmers in the region to meet crop requirements in wet years may be excessive with greater N loss to denitrification, runoff, and leaching during the periods of heavy winter rain. Timing of N applications is probably more critical for these years.
|
ACKNOWLEDGMENTS |
|---|
Received for publication March 26, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionGrain Yield ComponentsTotal Dry Matter and...SummaryREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. J. Lopez-Bellido, L. Lopez-Bellido, J. Benitez-Vega, and F. J. Lopez-Bellido Tillage System, Preceding Crop, and Nitrogen Fertilizer in Wheat Crop: I. Soil Water Content Agron. J., January 1, 2007; 99(1): 59 - 65. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Lopez-Bellido, L. Lopez-Bellido, J. Benitez-Vega, and F. J. Lopez-Bellido Tillage System, Preceding Crop, and Nitrogen Fertilizer in Wheat Crop: II. Water Utilization Agron. J., January 1, 2007; 99(1): 66 - 72. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. J. Wiatrak, D. L. Wright, and J. J. Marois Tillage and Residual Nitrogen Impact on Wheat Forage Agron. J., November 1, 2004; 96(6): 1761 - 1764. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. J. Wiatrak, D. L. Wright, and J. J. Marois Influence of Residual Nitrogen and Tillage on White Lupin Agron. J., November 1, 2004; 96(6): 1765 - 1770. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Lopez-Bellido, L. Lopez-Bellido, F. J. Lopez-Bellido, and J. E. Castillo Faba Bean (Vicia faba L.) Response to Tillage and Soil Residual Nitrogen in a Continuous Rotation with Wheat (Triticum aestivum L.) under Rainfed Mediterranean Conditions Agron. J., September 1, 2003; 95(5): 1253 - 1261. [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 | |||
