Nitrogen is the most important nutrient in rice systems,

Transcription

1 RICE Nitrogen Dynamics and Fertilizer Use Efficiency in Rice following Straw Incorporation and Winter Flooding Alison J. Eagle, Jeffrey A. Bird, James E. Hill, William R. Horwath, and Chris van Kessel* practice is problematic because of its contributions to air pollution (Bossio and Scow, 1995; Ponnamperuma, 1984). California legislation now restricts burning to 25% of the rice acreage, and allowable burning will likely decrease to zero. Alternatives to burning include soil incorporation or baling of the straw. Shallow flood- ing of fallow rice fields is commonly used in California because of its potential to enhance straw decomposition (Hill et al., 1999) and provide winter wetland habitat for migrating waterfowl (Elphick and Oring, 1998). Addition of organic matter to the soil in the form of straw can increase soil organic matter content in rice systems (Verma and Bhagat, 1992). In the long-term, straw incorporation has resulted in increased N mineral- ization potential in non-rice and rice systems (Bacon, 1990). Sustained increase in microbial biomass has been observed following many seasons of straw incorporation compared with burning (Bird et al., 2001; Powlson et al., 1987). Incorporation of straw with a high C/N ratio may initially immobilize inorganic N because of the nutrients required to sustain microbial growth (Verma and Bhagat, 1992). After an initial equilibration period that may last up to 3 yr following annual rice straw incorporation (Verma and Bhagat, 1992), plant available N supply in the soil tends to increase (Bacon, 1990). Studies on the impact of winter flooding on N dynam- ics in temperate rice systems have been lacking. The main effects of winter flooding, however, may be due to increased straw decomposition resulting from waterfowl activity (Bird et al., 2000) and to the effects of increased anaerobic conditions. The extended anaerobic time pe- riod during winter flooding has increased extractable inorganic N (Bird et al., 2001) after 4 yr compared with a non-winter-flooded fallow system. The increase in in- organic N is possibly due to the lower N immobilization after incorporating straw in anaerobic systems compared with aerobic systems (Acharya, 1935). It is also possible that the extended anaerobic time period during winter flooding may increase N availability to plants. Although other studies have found that incorporation of straw can increase soil N supply to rice, the impact in temperate rice systems is not as well known. Addi- tionally, the effect of these changes in soil N supply on N use efficiency and seasonal uptake has not been examined. The main objective of this study was to determine the impact of winter flooding and straw incorpora- ABSTRACT Incorporation of rice (Oryza sativa L.) straw, when compared with burning, affects soil N supply by increasing N and C inputs. This study determined the effects of long-term alternative rice straw management and winter flooding on seasonal crop uptake of N, fertilizer N use efficiency (FNUE), and crop N uptake from N-labeled straw. Microplots were established on two sites in California, Maxwell and Biggs, by applying N-labeled fertilizer during Year 4 of a long-term rice straw management study. At the end of the year, N-labeled straw was applied to assess crop uptake of straw N in the following season. Fertilizer use efficiency by N dilution (FUE N) over the growing season at Maxwell and at final harvest at Biggs was signifi- cantly higher when straw was burned rather than incorporated. Fertilizer use efficiency by N difference was significantly greater than FUE N, suggesting an apparent added N interaction (ANI). Straw management did not significantly affect the uptake of residual fertilizer N or of straw N in the subsequent year. Winter flooding had no significant effect on measured parameters. Data indicate a decrease in FNUE with a concomitant and more significant increase in the plant available soil N following the change in straw management from burning to incorporation. Belowground N pools appear to be a larger source of N for the following crop than aboveground residue. Nitrogen is the most important nutrient in rice systems, accounting for 67% of total fertilizer applica- tions to rice worldwide (Vlek and Byrnes, 1986). Nitrogen uptake patterns in rice over the growing season depend on the availability of soil N and fertilizer N (Bufogle et al., 1997; Norman et al., 1992). When fertil- izer N is applied preplant, fertilizer N uptake tends to be concentrated toward the beginning of the season, with soil N being the dominant N pool after the fertilizer N supply is depleted or immobilized (Bufogle et al., 1997). The relationship between fertilizer N uptake and total N uptake over the growing season depends on timing of the fertilizer N application (Guindo et al., 1994a) and the amount of fertilizer N available (Bufogle et al., 1997). The introduction of rice straw management, such as straw incorporation, often confounds these es- tablished relationships and requires additional research to elucidate the factors controlling N availability in rice. While burning is the traditional disposal method of straw and stubble in temperate and tropical rice-growing areas (Becker et al., 1994; Williams et al., 1972), this Abbreviations: ANI, added nitrogen interaction; FNUE, fertilizer ni- trogen use efficiency; FUE N, fertilizer use efficiency by nitrogen- dilution; FUE-ND, fertilizer use efficiency by nitrogen difference. A.J. Eagle, J.A. Bird, and W.R. Horwath, Dep. of Land, Air, and Water Resources, and J.E. Hill and Chris van Kessel, Dep. of Agron. and Range Sci., Univ. of California, Davis, CA *Corresponding author (cvankessel@ucdavis.edu). Published in Agron. J. 93: (2001). 1346

2 EAGLE ET AL.: NITROGEN DYNAMICS AND FERTILIZER USE EFFICIENCY IN RICE 1347 tion or burning on total N accumulation and seasonal Table 1. Characteristics of N-labeled rice straw applied to two N uptake and fertilizer N use efficiency (FNUE; N flooding treatments in the fall of 1997 at Maxwell, CA. and N difference methods). The fate of N-labeled fertilizer Winter flooded Not winter flooded in the second year and the use efficiency of N- Total N, kg ha 1 50 (4.0) 53 (2.0) labeled straw by the crop were also examined. Finally, Total C, Mg ha (0.21) 3.8 (0.21) Atom% excess N (0.032) (0.013) the contribution of N from below- and aboveground C/N ratio 69 (2.7) 73 (2.1) N sources to rice in the second year following the N,gkg (0.2) 5.3 (0.1) application of N-fertilizer was assessed. Straw was removed from the straw burn treatment microplots and placed onto new microplots in the incorporated treatment. Numbers in parentheses are standard error of the mean of four replicates. MATERIALS AND METHODS Field Sites labeled straw from the burned treatment was replaced with surrounding unlabeled straw, which was subsequently burned. Straw and winter flooding treatments were established at From the new N-residue microplots established in the fall of two sites in the Sacramento Valley of northern California. A 1997, the contribution of N-labeled surface straw, designated 28-ha study site near Maxwell, Colusa County, was established aboveground, to the subsequent crop could be measured comin fall A 10-ha site at the California Rice Research pared with the soil plus root N and unburned stubble N, Station near Biggs, Butte County, was established in fall designated belowground. The N-residue microplot in the in- Main-plot sizes were 42 by 180 m at Maxwell and by 142 m corporated treatment was used to determine the entire contriat Biggs. Important differences in soil characteristics between bution of below- plus aboveground residue pools to the subse- the two sites are the lower clay content, ph, and exchangeable quent rice crop. None of the straw treatments received N K at Biggs. Additional information on soil characteristics is fertilizer in Characteristics of the N-labeled straw apreported elsewhere (Eagle et al., 2000). plied to the new N microplots are summarized in Table 1. Treatments were arranged in a split-plot design, with winter The rate of N straw applied was equal to the amount of straw flooding and no winter flooding as main-plot treatments and produced under straw-incorporated conditions. In the fifth straw management practices as subplot treatments. Treatlished in the spring of the fourth year represented the uptake year of the experiment, N uptake in the N microplots estab- ments were replicated four times. The straw management treatments were baling and removal, rolling, incorporation, of the previous year s fertilizer N application from both soil or burning. Two of the straw management practices, burning N pools and above- and belowground residue. Fertilizer N and soil incorporation of straw, were addressed in this portion uptake in the fifth year through belowground N sources was of the study. Rice variety M202 was aerially seeded each spring calculated from the difference between N uptake in the fertil- onto fields that were flooded approximately 1 d before seedstraw was applied. izer N microplots and N uptake where only N-labeled ing. The fields remained flooded throughout the growing season until crop maturity and then were drained to allow drying for harvest. Straw treatments were applied in the fall, and the Plant Samples water level in winter-flooded plots was maintained at 5- to Plant samples were collected throughout the growing sea- -cm depth from November until drainage in March. son at Maxwell, including at plant maturity and final harvest, In the spring of the fourth year, 1997 at Maxwell and 1998 with sample dates concentrated toward the beginning of the at Biggs, N-fertilizer microplots were established at each site. season during rapid growth. Samples were collected at plant The N microplots were rectangular, 3 by 4 m, at Maxwell and maturity and final harvest at Biggs. Five or more N-labeled circular, 2-m diam., at Biggs. The N-labeled urea [(NH 2 ) 2 CO] plants were selected at random and harvested, by carefully was applied at a rate of 20 kg N ha 1 at 10 atom% N just pulling out the plant with roots, at each sampling time. A before preplant flooding. Additional unlabeled fertilizer N quadrant in the surrounding area was used for yield determinawas applied as aqueous NH 3 and monoammonium phosphate tion. The plants were washed to remove soil from roots; sepa- (NH 4 H 2 PO 4 ) to obtain a total rate of 188 kg N ha 1 with rated into roots, shoots, and panicles (when present); dried an N content of 1.07 atom% N at Maxwell. To reduce at 60 C; weighed; and ground to a powder in a ball mill. The denitrification, the nitrification inhibitor nitrapyrin {N-serve N-labeled plants were then analyzed for the concentration 24E [2-chloro-6-(tricholoromethyl) pyridine], Dow Elanco, of N and atom percent of N using a combustion continuous- Indianapolis, IN) was applied at a rate of 0.4 L ha 1. At Biggs, flow isotope ratio mass spectrometer (PDZ Europa, Crewe, ammonium sulfate [(NH 4 ) 2 SO 4 ] was applied to result in a total England). of 161 kg N ha 1, with an enrichment of 1.24 atom% N. To At final harvest, shoots were collected by cutting plants just prevent lateral movement of the labeled N, metal barriers above ground level. Total biomass and grain yield (1 m 2 ) were surrounding the microplots were inserted into the soil to a determined from both inside and outside of the N microplot. depth of approximately cm. During each year of the study, There were no significant differences between yield estimates additional microplots within each main plot received no fertil- from the N microplots and the main plot at Maxwell. Howizer N. Phosphorus was applied as triple superphosphate at ever, yield was significantly affected by microplots and deemed the same rate as in the N fertilized plots, 32 and 25 kg P ha 1 unreliable at Biggs, likely due to the small microplot size. at Maxwell (1997) and Biggs (1998), respectively. These plots Yields from the N microplot were used for all N-microplotwith zero N fertilizer were placed in a different location within related calculations at Maxwell in In 1998, the mainthe larger plot each year. plot yields were used because no N yield samples were taken The significance of below- vs. aboveground residue as a due to the lack of significant difference between main plot source of N for the following crop was determined in the and N yield in At Biggs, yields from the main plot second year at Maxwell. Following harvest of the N-labeled were used in calculations of total N uptake and fertilizer use crop at Maxwell in 1997, the straw from the N-microplots in efficiency by the N dilution method (FUE N). Yields from the burned treatments was transferred onto new N-residue the main plot and zero-n-fertilizer microplot were used for microplots in the straw-incorporated treatments. The N- calculating fertilizer use efficiency by N difference (FUE-ND).

3 1348 AGRONOMY JOURNAL, VOL. 93, NOVEMBER DECEMBER 2001 Plants from the final harvest were separated into grain and straw components, dried at 60 C, and weighed. The N-labeled samples were first ground in a Wiley mill, ball-milled for analysis, and analyzed for N and atom percent of N as described above. Fertilizer Nitrogen Use Efficiency Fertilizer use efficiency by the N dilution method (%) was calculated as follows: FUE N (atom% N excess plant )/ (atom% N excess fertilizer ) (N plant /N fertilizer ) 100 [1] where atom% N excess plant atom% N excess (over background levels) in the plant, atom% N excess fertilizer atom% N excess in the labeled fertilizer N, N plant total plant N (kg ha 1 ), and N fertilizer fertilizer N applied (kg ha 1 ). Straw N use efficiency (%), the proportion of straw N that ended up in the crop the following year, was calculated as follows: straw N use efficiency (atom% N excess plant )/ (atom% N excess straw ) (N plant /N straw ) 100 [2] where atom% N excess straw atom% N excess in the labeled straw and N straw total straw N applied (kg ha 1 ). Fertilizer N use efficiency by the N difference method (%) was calculated as follows: FUE-ND (NPlant fert NPlant zeron )/(N fertilizer ) 100 where NPlant fert total plant N uptake (kg ha 1 ) in N-fertil- ized plot and NPlant zeron total plant N uptake (kg ha 1 )in zero-n plots. Interactions between added fertilizer N and native soil N that change the N content in a given pool are called added N interactions (ANI) (Jenkinson et al., 1985). These interactions may result in different estimates for FUE N and FUE-ND. Statistical Analysis Data were analyzed using the PROC GLM procedure of SAS (SAS Inst., 1989). Analysis of variance (ANOVA) was used to determine treatment effects, and the flood block mean squared error was used as the error term for winter flooding. Repeated-measures ANOVA was used to determine treatment effects over time during the growing season. RESULTS Plant N during the growing season at Maxwell reached a maximum between 60 and 80 d after seeding in both 1997 and 1998 (Fig. 1). In 1997, there were no straw management effects on total N accumulation during the growing season, but in 1998, the incorpora- tion of straw significantly increased plant N when analyzed over the entire growing season (P 0.01). By final harvest, straw incorporation significantly increased total plant N in 1998 (Table 2). As with total N, fertilizer N uptake at Maxwell peaked at approximately 60 to 80 d after seeding, reaching a maximum FUE N during the season of 37 and 32% when straw was burned or [3] Fig. 1. Total above- and belowground N in rice during the 1997 and 1998 growing seasons at Maxwell, CA, as affected by straw management and winter flooding. Error bars are standard error of four replicates. incorporated, respectively (Fig. 2). Fertilizer use effi- ciency by the N dilution method over the season in 1997 was significantly greater when straw was burned than when it was incorporated (P 0.05). This trend was noted only in the grain at final harvest (Table 2), and neither straw management nor winter flooding had a significant effect on total plant FUE N at Maxwell at the final harvest (Table 3). While winter flooding again had no effect, straw management significantly af- fected fertilizer N uptake at Biggs (Table 4). The FUE N at Biggs averaged 31% when straw was burned and 24% when it was incorporated (P 0.01) (Table 3). At Maxwell in 1997, FUE-ND was greater when straw was burned rather than incorporated (Table 3). This indirect method of measuring FNUE was on average 1.7 and 1.5 times greater than FUE N at Maxwell and Biggs, respectively, indicating the presence of an ANI (Table 3). The difference in fertilizer N recovery be- tween FUE-ND and FUE N at Maxwell in 1997 corresponded to 49 kg N ha 1. In the fifth year of the straw management experiment, the year following the application of the N fertilizer at Maxwell, an average of 2.3% of the N fertilizer applied the previous year was accumulated by rice at plant maturity (Fig. 3). One month later, at final harvest, the average FUE N of the previous year s applied

4 EAGLE ET AL.: NITROGEN DYNAMICS AND FERTILIZER USE EFFICIENCY IN RICE 1349 Table 2. Soil and fertilizer N recovery in rice plants at final harvest as affected by rice straw management and winter flooding at Maxwell, CA. Grain Straw Total plant Treatment Total N Soil N Fert. N Total N Soil N Fert. N Total N Soil N Fert. N kg ha Burn and winter flood Burn and no winter flood Incorporate and winter flood Incorporate and no winter flood Analysis of variance Straw NS NS * * NS NS NS NS Winter flood NS NS NS NS NS NS NS NS NS Straw winter flood NS NS NS NS NS NS NS NS NS kg ha Burn and winter flood Burn and no winter flood Incorporate and winter flood Incorporate and no winter flood Analysis of variance Straw * * * Winter flood NS NS NS Straw winter flood NS NS NS * Significant at the 0.05 level. Significant at the 0.1 level. Nitrogen--labeled fertilizer was only applied in 1997, so total plant N in 1998 was not divided into soil and fertilizer N components. fertilizer was 3.0% (Table 5). The majority of the labeled ha 1, resulting in straw N use efficiency of 3.5% in the N uptake was from belowground sources rather than aboveground portion of the plants. Straw N use effiaboveground straw (Fig. 3A vs. Fig. 3B). Slightly greater ciency (roots and shoots) increased over the growing uptake of N fertilizer in grain was seen in incorporated season along with N uptake, and the majority of the compared with burned plots (P 0.078) in the sec- straw N was accumulated by 60 to 80 d after seeding (Fig. ond year. 4). There was no significant effect of winter flooding on Although plant N was greater in incorporated vs. straw N use efficiency. burned treatments by 13 and 23 kg N ha 1 in 1997 and Total N in plant pools was calculated at final harvest 1998, respectively (Table 2), only 1.8 kg N ha 1 came in 1997 and 1998 for Maxwell (Table 5). Following harvest directly from straw. Straw N inputs averaged 51 kg N in 1997, up to 39% of the fertilizer N applied in the Fig. 2. Fertilizer N use efficiency during the 1997 growing season at Maxwell, CA, as affected by straw management. Error bars are standard error of four replicates.

5 1350 AGRONOMY JOURNAL, VOL. 93, NOVEMBER DECEMBER 2001 Table 3. Fertilizer N use efficiency (FNUE) determined by N dilution (FUE N) and N difference method (FUE-ND) for the final harvest of rice as affected by alternative straw management and winter flooding at Maxwell, CA in 1997 and at Biggs, CA in Maxwell, 1997 Biggs, 1998 Treatment FUE N FUE-ND FUE N FUE-ND % Burn and winter flood Burn and no winter flood Incorporate and winter flood Incorporate and no winter flood Analysis of Variance Straw NS ** NS Winter flood NS NS NS NS Straw winter flood NS NS NS NS ** Significant at the 0.01 level. Significant at the 0.1 level. organic matter (Bird et al., 2001; Bossio and Scow, 1995). Total plant N and fertilizer N uptake reached a maximum at approximately 60 to 80 d after seeding at Max- well in both years. This is the time of maximum tillering and panicle initiation. Patrick and Reddy (1976) also found a large portion of fertilizer N uptake occurred early in the season. Other studies found N uptake to continue until much later in the growing season (Bufogle et al., 1997; Guindo et al., 1994a). This discrepancy in timing of maximum N uptake may be due to differences in soil N availability over the growing season, use of different rice varieties, climatic differences, or length of growing season. In the current study, both total N and fertilizer N recovery dropped slightly toward the end of the season. Guindo et al. (1994a) also found a similar drop in total fertilizer N recovery and total plant N content using a preflood fertilizer N application. Split- ting fertilizer N application has resulted in increasing total N and static fertilizer N content (Bufogle et al., 1997) or total plant N content (Guindo et al., 1994b) toward the end of the season. Incorporation of straw increased total plant N in N fertilized treatments by 13 and 23 kg N ha 1 at Maxwell in the fourth and fifth years of the study (Table 2). Although this was not significant in 1997, repeated measures ANOVA revealed a significant difference over five years (Eagle et al., 2000). The increase in N uptake due to straw incorporation in unfertilized microplots burned treatments was removed either through grain (28%) or by burning the straw (11%). When the straw was incorporated, only 24% of the fertilizer N applied was removed in the grain and none in the straw. At the Biggs site, there was a stronger effect of straw management on fertilizer N uptake in both the grain and the straw (Table 5). Due to low N label in the straw and the small size of the microplots, the residual N uptake in the second year after fertilizer N application was not examined at Biggs. DISCUSSION Nitrogen Accumulation In this study, the increased plant N following straw incorporation indicates an increase in plant available soil N (Table 2). In the first 2 yr of the experiment, a reduction in available N was measured when straw was incorporated rather than baled or burned (Hill et al., 1999). Immobilization of available N had taken place when straw was incorporated. However, in the subse- quent years, an increase in available N took place when straw was incorporated, which manifested itself clearly in the zero-n-fertilizer treatments (Eagle et al., 2000). No significant increases in total soil N were observed between straw treatments, so the increase in plant avail- able N could be due to the greater amount of N in the labile pools of soil organic matter (Bird et al., 2002) or to promotion of microbial activity following addition of Table 4. Soil and fertilizer N recovery in rice plants at final harvest as affected by rice straw management and winter flooding at Biggs, CA in Grain Straw Total plant Treatment Total N Soil N Fert. N Total N Soil N Fert. N Total N Soil N Fert. N kg ha 1 Burn and winter flood Burn and no winter flood Incorporate and winter flood Incorporate and no winter flood Analysis of Variance Straw NS NS ** NS NS NS ** Winter flood NS NS NS NS NS NS NS NS NS Straw winter flood NS NS NS NS NS NS NS NS NS ** Significant at the 0.01 level. Significant at the 0.1 level.

6 EAGLE ET AL.: NITROGEN DYNAMICS AND FERTILIZER USE EFFICIENCY IN RICE 1351 Fig. 3. Recovery of N added to rice in 1997 by rice in the next growing season (1998), as affected by alternative straw-manage- ment practices and winter flooding at Maxwell, CA. (A) Includes all above- and belowground sources of N from previous year s fertilizer. (B) Includes only the N through belowground pools (N through residue excluded). Error bars are standard error of four replicates. Nitrogen- Fertilizer Use Efficiency The FUE N values measured in this study are comparable to the 30 to 50% values reported in research on tropical lowland rice (Bronson et al., 2000; De Datta et al., 1968). These values tend to be lower than those reported for upland crops, which are also dependent on crop and soil types, production methods, and timing of fertilizer application (Macdonald et al., 1997). Other studies have also reported that application of fertilizer N later in the growing season increases FUE N (Patrick and Reddy, 1976). Bronson et al. (2000), however, did not notice any difference in FUE N between split fertilizer applications with different times of application. The FUE N values have been noted in the range of 72 to 79% when N fertilizer was applied 27 and/or 55 d after emergence (Norman et al., 1992). The FUE N was greater when straw was burned rather than incorporated over the growing season at Maxwell (Fig. 2) and at final harvest at Biggs (Table 3). Incorporating straw compared with burning it increased the soil N availability through an increase in net N mineralization and corresponding dilution of fertilizer N. At final harvest, the increase in fertilizer N uptake in the total plant at Biggs and in the grain at Maxwell when straw was burned was associated with an increase in soil N uptake in the straw at both locations when straw was incorporated (Tables 2 and 4). Therefore, the rate of fertilizer N application may be reduced when straw is incorporated. After four and five seasons of straw incorporation in situ, greater N immobilization shortly after residue incorporation and greater N mineralization during the growing season was observed compared with where straw was burned (Bird et al., 2001). Further, four sea- sons of residue incorporation increased the C and N contents of the active light and mobile humic fractions (Bird et al., 2002). Clearly, the incorporation of straw for a prolonged period of time changed the overall N dynamics and cycling in the soil and caused a net in- crease in the N supply power of the soil, reflected in higher yields and total N uptake of the unfertilized rice crop. Adjustment of fertilizer N application to better reflect soil N supply should be considered to increase was even more substantial (Eagle et al., 2000). However, on average only 3.5% of the straw N directly entered the following year s crop (Fig. 4). Therefore, the impact of straw incorporation on N availability is much larger than would be suggested from the recovery of one year s worth of straw-n in the subsequent crop. Additional benefits following the incorporation of organic material, such as mineralization of other nutrients and improved soil quality, may lead to an increase in total N accumulation in the crop by supplying other limiting nutrients and increasing microbial activity. FNUE in rice systems (Cassman et al., 1993). Table 5. Recovery of N added in 1997 at Maxwell, CA by rice at final harvest in 1997 and 1998, and recovery of N added in 1998 at Biggs, CA by rice at final harvest in 1998 as affected by straw management and winter flooding. Maxwell, 1997 Maxwell, 1998 Biggs, 1998 Treatment Grain Straw Grain Straw Grain Straw % of total fertilizer N applied Burn and winter flood Burn and no winter flood Incorporate and winter flood Incorporate and no winter flood Analysis of Variance Straw NS NS ** Winter flood NS NS NS NS NS NS Straw winter flood NS NS NS NS NS NS ** Significant at the 0.01 level. Significant at the 0.1 level.

7 1352 AGRONOMY JOURNAL, VOL. 93, NOVEMBER DECEMBER 2001 Fig. 4. Recovery of incorporated N-labeled rice straw (residue N use efficiency) by rice in the next growing season, as affected by winter flooding at Maxwell, CA. Error bars are standard error of four replicates. Added Nitrogen Interactions The recovery of fertilizer N varied widely whether based on the N isotope dilution or the N difference method (Table 3). Such a large difference in the recovery of fertilizer N indicates the strong presence of an ANI. The ANI observed at Maxwell and Biggs could be apparent or real. Apparent ANI is caused by mineral- ization immobilization turnover, in which newly miner- alized unlabeled N replaces fertilizer N ions in solution (Jenkinson et al., 1985). This process is microbially driven, with concomitant N immobilization of added fertilizer N and mineralization of native soil N. At the Maxwell site, a sustained, greater microbial biomass C and N pool was observed after Year 4 along with greater N fertilizer recovered as labile humic N (mobile humic acid) in incorporated plots compared with burned plots (Bird et al., 2001; Bird et al., 2002). These results suggest that the microbial stabilization of fertilizer N leads to the enhanced apparent ANI with incorporating rather than burning. Apparent ANI is also the most likely contributor when the ANI increases, with a longer con- tact period between fertilizer N and the soil N pools (Schnier, 1994). Because the uptake curves for fertilizer N and total N were similar in shape (Fig. 1 and 2), fertilizer N and soil inorganic N are likely in the same or similar pools, making pool substitution of labeled and unlabeled N probable (Hart et al., 1986). It has been suggested that apparent ANI likely constitutes the majority of observed ANI (Jenkinson et al., 1985). Studies using N-labeled fertilizer in rice systems have often found positive ANIs (Cassman et al., 1993). Added N interaction in tropical rice production systems ranged from 7.0 to 22.6 kg N ha 1 at various times throughout the growing season (Schnier, 1994). The ANI increased where fertilizer N was in contact with the soil for a greater period of time. The degree of pool substitution, and thus the nature of the observed ANI, depends on method of fertilizer application (Schnier, 1994). At Biggs, ANI was estimated at 14 kg N ha 1 while at Maxwell, values were much higher, i.e., 51 and 47 kg N ha 1 when straw was incorporated and burned, respectively. At Maxwell, the higher ANI, if it was associated with higher N turnover rates, might have been caused by higher organic matter content. Also, the greater yield response to fertilizer N at Maxwell could have resulted in a higher real ANI due to more root penetration or root exudates and turnover. The FUE N at Biggs was lower than at Maxwell and can be partially explained by the poor growing season in 1998 (El Nino) compared with the better grow- ing season in 1997, which had a warmer and dryer spring. However, differences in soil characteristics and manage- ment practices play a large role because comparisons of FUE-ND from 1995 through 1998 are consistently lower at Biggs than at Maxwell (average of 43% at Biggs vs. 66% at Maxwell). The low soil extractable K at Biggs (Hill et al., 1999) may also contribute to lower N use efficiencies. Residual Nitrogen- Fertilizer Tracing the fate of the N fertilizer through the second growing season indicated that belowground pools (root and microbial derived) are more significant sources of plant available N than incorporated straw N. In the spring before the second year (May 1998), 21% of the original N-labeled fertilizer was measured in the top cm of the soil profile (Bird et al., 2001). At final harvest of the second year, the crop had accumulated

8 EAGLE ET AL.: NITROGEN DYNAMICS AND FERTILIZER USE EFFICIENCY IN RICE 1353 % of that N, or approximately 3% of total amount Therefore, in our study, fertilizer N rates can be reduced of N fertilizer applied in the previous year (Table 5), by at least 25 kg N ha 1 when straw is incorporated compared with 39 and 36% of the fertilizer N taken up (Eagle et al., 2000). A reduction in N fertilizer input the year before in burned and incorporated treatments, without a reduction in yield would improve the FNUE. respectively. Therefore, as expected, by the second year, Therefore, fertilizer N input should be adjusted constantly, the N added as fertilizer was in less-available forms. depending on the increase in the supply of N In contrast, the incorporated straw alone contributed from residues. 3.5% of its total N, and 3.5% of the fertilizer N within A large difference in the FUE N and the FUE-ND the residue, to the subsequent crop. Only 13% of the was observed. The difference was likely caused by a N-labeled fertilizer in the second-year crop came from strong ANI, whereby unlabeled N in the microbial biothe rice residue. Therefore, the availability of the N mass was substituted for N-labeled fertilizer. Hence, residue and N soil pools appeared to be different, and the recovery of fertilizer by N isotope underestimates the belowground N pools were more important N sources the role of fertilizer N as a source of N for the rice crop. to the crop. Unfortunately, most other field studies only The belowground pools of residual N after application followed the fate of N-labeled straw (Becker et al., of N-labeled fertilizer were more important than incorporated 1994) or combined roots and shoots (Norman et al., straw in availability of N to the next rice crop, 1990). A separation of above- vs. belowground contribu- and the small uptake of straw N in the first year compared tions as a source of N for the subsequent crop is seldom with a much higher effect of straw N in later years made. In addition, belowground sources of N include indicates significant N cycling between various soil N root, crown, and microbially immobilized fertilizer N, pools in subsequent years. making it difficult to assess the importance of these pools in the years following fertilizer addition. ACKNOWLEDGMENTS From our study, it appears that in rice cropping sys- We thank the California Energy Commission, the California tems, the aboveground contribution may not be as im- Rice Research Board, and Ducks Unlimited for their generous portant a source of N as belowground sources of N such financial support. We also thank the California Rice Research as remaining fertilizer N, belowground N-labeled res- Station and Canal Farms for their collaboration and assistance idues (roots or exudates), or N immobilized by micro- in the field operations. The technical assistance of S. Scardaci, bial biomass during the year of fertilizer application. Dr. M. Hair, M. Llagas, P. Flaugher, P. Kong, S. Lo, and T. Nitrogen fractions, including mobile humic substances, Kraus; the statistical assistance of Dr. F.C. Stevenson; and the light fraction, and microbial biomass, are the most active valuable comments made by Dr. C.A. Beyrouty on an earlier soil N pools and likely contribute the majority of the version of the manuscript are highly appreciated. N label to the second-year crop (Bird et al., 2001; Bird et al., 2002). REFERENCES While the straw N use efficiency was low, the cumula- Acharya, C.N Studies on the anaerobic decomposition of plant tive effect of 4 and 5 yr of straw incorporation on N materials: I. The anaerobic decomposition of rice straw (Oryza nutrition was greater than direct N flow from straw to sativa L.). Biochem. J. 29: Bacon, P.E Effects of stubble and N fertilization management crop. Although only 1.8 kg N ha 1 straw N was directly on N availability and uptake under successive rice (Oryza sativa available to the crop in the year following incorporation, L.) crops. Plant Soil 121: total N uptake increased by 23 kg N ha 1 in Therelowland Becker, M., J.K. Ladha, and J.C.G. Ottow Nitrogen losses and fore, these results, combined with the evidence that yield rice yield as affected by residue nitrogen release. Soil Sci. and N uptake only began to be affected by straw man- Soc. Am. J. 58: Bird, J.A., W.R. Horwath, A.J. Eagle, and C. van Kessel Immoagement by the third year of the long-term study (Eagle bilization of fertilizer nitrogen in rice: Effects of straw management et al., 2000), suggest that active N pools in incorporated practices. Soil Sci. Soc. Am. J. 65: plots were increasing over time, thereby enhancing availwaterfowl Bird, J.A., G.S. Pettygrove, and J.M. Eadie The impact of foraging on the decomposition of rice straw: Mutual able N more as incorporation practices continued. benefits for rice growers and waterfowl. J. Appl. Ecol. 37: Neither straw management nor winter flooding af- Bird, J.A., C. van Kessel, and W.R. Horwath Nitrogen dynamics fected the total amount of N fertilizer remaining in the in humic fractions under alternative straw management in temperate rice. Soil Sci. Soc. Am. J. 66:(in press). system at Maxwell 2 yr after application (Bird et al., 2001) even though more fertilizer N was removed from Bossio, D.A., and K.M. Scow Impact of carbon and flooding the system where straw was burned, both in the grain on the metabolic diversity of microbial communities in soils. Appl. Environ. Microbiol. 61: and the burned straw. The low straw use efficiency may Bronson, K.F., F. Hussain, E. Pasuquin, and J.K. Ladha Use have been due to losses of the residue N following spring of -N-labeled soil in measuring nitrogen fertilizer recovery efficiency tillage. This could contribute to the lack of difference in transplanted rice. Soil Sci. Soc. Am. J. 64: between treatments, both in N plant uptake and total Bufogle, A., P.K. Bollich, R.J. Norman, J.L. Kovar, C.W. Lindau, and R.E. Macchiavelli Rice plant growth and nitrogen accumula- N recovery. tion in drill-seeded and water-seeded culture. Soil Sci. Soc. Am. J. 61: Cassman, K.G., M.J. Kropff, J. Gaunt, and S. Peng Nitrogen CONCLUSIONS use efficiency of rice reconsidered: What are the key constraints? Plant Soil 6: Due to increased soil N availability as a result of straw De Datta, S.K., C.P. Magnaye, and J.C. Moomaw Efficiency incorporation for 5 yr, FNUE declined when straw was of fertilizer nitrogen ( N-labelled) for flooded rice. p In incorporated compared with when straw was burned. Trans. Int. Soil Sci. Congr., 9th, Adelaide, Australia. Vol. 4.

9 1354 AGRONOMY JOURNAL, VOL. 93, NOVEMBER DECEMBER 2001 Eagle, A.J., J.A. Bird, W.R. Horwath, B.A. Linquist, S.M. Brouder, Norman, R.J., J.T. Gilmour, and B.R. Wells Mineralization of J.E. Hill, and C. van Kessel Rice yield and nitrogen utilization nitrogen from nitrogen- labeled crop residues and utilization by efficiency under alternative straw management practices. Agron. rice. Soil Sci. Soc. Am. J. 54: J. 92: Norman, R.J., D. Guindo, B.R. Wells, and C.E. Wilson, Jr Elphick, C.S., and L.W. Oring Winter management of Califor- Seasonal accumulation and partitioning of nitrogen- in rice. Soil nian rice fields for waterbirds. J. Appl. Ecol. 35: Sci. Soc. Am. J. 56: Guindo, D., R.J. Norman, and B.R. Wells. 1994a. Accumulation of Patrick, W.H., and K.R. Reddy Fate of fertilizer nitrogen in a fertilizer nitrogen- by rice at different stages of development. flooded rice soil. Soil Sci. Soc. Am. J. 40: Soil Sci. Soc. Am. J. 58: Ponnamperuma, F.N Straw as a source of nutrients for wetland Guindo, D., B.R. Wells, and R.J. Norman. 1994b. Cultivar and nitro- rice. p In Organic matter and rice. IRRI, Los Banos, Philgen rate influence on nitrogen uptake and partitioning in rice. Soil ippines. Sci. Soc. Am. J. 58: Powlson, D.S., P.C. Brookes, and B.T. Christensen Measure- Hart, P.B.S., J.H. Rayner, and D.S. Jenkinson Influence of pool ment of soil microbial biomass provides an early indication of substitution on the interpretation of fertilizer experiments with N. changes in total soil organic matter due to straw incorporation. J. Soil Sci. 37: Soil Biol. Biochem. 19: Hill, J.E., D.M. Brandon, S.M. Brouder, A.U. Eke, T.E.C. Kraus, SAS Inst SAS/STAT user s guide. Version 6. 4th ed. Vol. 2. M.A. Llagas, B.A. Linquist, and S.C. Scardaci Agronomic implications of alternative straw management practices. In Winter SAS Inst., Cary, NC. flooding and straw management: Implications for rice production Schnier, H.F Nitrogen- recovery fraction in flooded tropical Agron. Prog. Rep. Agric. Exp. Stn. and Coop. Ext., rice as affected by added nitrogen interaction. Eur. J. Agron. Univ. of California, Davis. 3: Jenkinson, D.S., R.H. Fox, and J.H. Rayner Interactions bepractices on yield, nitrogen uptake and soil properties in a wheat Verma, T.S., and R.M. Bhagat Impact of rice straw management tween fertilizer nitrogen and soil nitrogen the so-called priming effect. J. Soil Sci. 36: rice rotation in northern India. Fert. Res. 33: Macdonald, A.J., P.R. Poulton, D.S. Powlson, and D.S. Jenkinson. Vlek, P.L., and B.H. Byrnes The efficacy and loss of fertilizer Effects of season, soil type and cropping on recoveries, resi- N in lowland rice. Fert. Res. 9: dues and losses of N-labelled fertilizer applied to arable crops in Williams, W.A., M.D. Morse, and J.E. Ruckman Burning vs. spring. J. Agric. Sci. (Cambridge) 129: incorporation of rice crop residues. Agron. J. 64: Rice Parameters Describing Crop Performance of Four U.S. Cultivars James R. Kiniry,* Garry McCauley, Yun Xie, and Jeffrey G. Arnold Realistic description of key processes in crops provides a means of quantifying how cultivars differ J.R. Kiniry and J.G. Arnold, USDA-ARS, 808 E. Blackland Rd., Temple, TX 76502; G. McCauley, Texas A&M Agric. Exp. Stn., Box 717, Eagle Lake, TX 77434; and Y. Xie, Beijing Normal Univ., Beijing, China. *Corresponding author (kiniry@brc.tamus.edu). Published in Agron. J. 93: (2001). ABSTRACT Parameters describing processes of crop growth and yield produc- tion provide modelers with the means to simulate crops and provide breeders with a system of comparing cultivars. Such values for rice (Oryza sativa L.) are especially important for some regions in the southern USA. Accordingly, the objective of this study was to quantify key biomass and yield production processes of four rice cultivars common in this region. We measured the leaf area index (LAI), the light extinction coefficient (k) for Beer s law, N concentrations, and the harvest index (HI) for the main and ratoon crops in 1999 and 2000 at Eagle Lake, TX. Dry matter was linearly related to intercepted photosynthetically active radiation (IPAR) for all of the data sets. The mean radiation use efficiency (RUE) was 2.39 g aboveground biomass MJ 1 IPAR. Maximum LAI values ranged from 9.8 to 12.7, and the mean k value for the main crop was The highest main crop yields were 7.04 Mg ha 1 for Cocodrie in 1999 and 7.22 Mg ha 1 for Jefferson in Yield differences among cultivars were due to HI differences and were not related to RUE values. The mean HI was 0.32 for all four cultivars over the two harvests in each of the 2 yr. Consistency in values of RUE, k, N concentrations, and HI among the cultivars in this study and between this study and values reported in the literature will aid modelers simulating rice development and yield and aid breeders in identifying key traits critical to rice grain yield improvement. and helps provide a system of simulating grain yield production using crop models. Crops grow leaf area [as described with the leaf area index (LAI)], intercept light [as described with the light extinction coefficient (k) in Beer s law; Monsi and Saeki, 1953], and produce bio- mass as a function of intercepted light [radiation use efficiency (RUE)]. Biomass is partitioned into grain, as reflected by the harvest index (HI), and biomass production can be reduced by N deficiency, as described by plant concentrations less than values determined with adequate available N. Quantification of these processes for common cultivars of a crop is valuable for identifying key traits for selecting higher yielding cultivars in the future. Such is also important for simulating common cultivars under variable climatic conditions. As discussed by Amthor and Loomis (1996), mechanistic models simulating cropping systems at one level are best described by processes at a lower level. Likewise, Sinclair and Seligman (2000) discussed how crop-level simulation models should simulate processes at the whole-plant level and whole-plant simulation should be simulated at the organ level. Description of these processes in rice is needed for simulating how changes in production acreage in the USA influence regional environmental quality. Reductions in areas used for rice production can change water- shed hydrology, water quality downstream, and plant cover throughout the year. Accurate simulation of rice Abbreviations: AB, aboveground biomass; HI, harvest index; k, light extinction coefficient for Beer s law; IPAR, photosynthetically active radiation intercepted by plants; LAI, leaf area index; PAR, photosynthetically active radiation; RUE, radiation use efficiency.