Effects of Nitrogen Rate and Harvest Moisture Content on Physicochemical Properties and Milling Yields of Rice

Transcription

1 Effects of Nitrogen Rate and Harvest Moisture Content on Physicochemical Properties and Milling Yields of Rice Brandon C. Grigg, 1 Terry J. Siebenmorgen, 1, and Richard J. Norman 2 ABSTRACT Cereal Chem. 93(2): Field studies were conducted from 2011 to 2013 near Stuttgart, Arkansas. The impacts of nitrogen rate (0, 45, 90, 135, and 180 kg of N/ha) and harvest moisture content (HMC) (23, 19, and 15%, wet basis) on physicochemical properties and milling yields were determined. Trends were similar for the cultivars evaluated: Cheniere, CL XL745, and Wells. Milled rice yields were only minimally impacted by either N rate or HMC level. Increasing N rate reduced kernel length and thickness of brown rice, chalkiness of brown rice and head rice, and viscosities of head rice flour, and it increased brown rice and head rice crude protein content and head rice yield (HRY). In terms of milling yields and head rice functionality, these data suggest that N rates as low as 90 kg of N/ha could be utilized, should production recommendations be changed. Significant interactions between N rate and HMC level were infrequent and were associated with the 0 kg of N/ha rate, unrealistic for rice production. Decreasing HMC from 23 to 19% reduced kernel length and thickness and increased crude protein content and chalkiness; further decreasing HMC to 15% also increased kernel fissuring and decreased HRY. Rice (Oryza sativa L.) is vital to food security for much of the world. The United States accounts for a relatively small portion of world production but ranked fifth worldwide in rice exported in 2014 (USDA-ERS 2014). Recently, U.S.-grown long-grain rice has faced increased marketing challenges in relation to quality factors such as chalkiness (McClung 2013). With respect to appearance, kernel dimensions and chalkiness impact rice quality (Lisle et al. 2000; Bergman et al. 2004; USDA- FGIS 2009), and increased chalkiness may affect sensory and cooking properties and end-use functionality of rice (Kim et al. 2000; Lisle et al. 2000; Ashida et al. 2009; Chun et al. 2009). However, the primary indices used to assess rice quality are from a milling perspective, comprising milled rice yield (MRY) and head rice yield (HRY). MRY represents the mass fraction of unprocessed, rough rice that remains as milled rice, which includes both head rice and broken kernels. HRY represents the mass fraction of rough rice that remains as head rice, synonymous with whole kernels (USDA- FGIS 2009) and defined as the well-milled rice kernels three-fourths or more of the original kernel length, with well-milled referring to the degree of milling and defined as the extent of bran removal from brown rice during the milling operation. Immature, fissured, and chalky kernels may reduce milling yields. Harvesting prematurely at elevated harvest moisture contents (HMCs) can lead to increased immature kernels (Kocher et al. 1990), which can break during milling. Delaying harvest, with consequently decreasing HMC, has also been shown to reduce HRY (Kester et al. 1963; Calderwood et al. 1980; Geng et al. 1984; Dilday 1989). However, such harvest delays are prevalent, most often a result of logistical issues or of extended in-field drying to reduce postharvest drying costs. Siebenmorgen et al. (2007) indicated a HMC range of 19 22% (wet basis) as optimal for maximizing HRY of long-grain rice grown in the mid-south United States. When harvesting rice below optimal HMCs, the potential for kernel breakage during milling increases with exposure to rapid moisture adsorption events and subsequent kernel fissuring (Siebenmorgen et al. 1998, 2007). In addition to the increased potential for fissuring, harvesting rice below 15% HMC has been Corresponding author. Phone: tsiebenm@uark.edu 1 Department of Food Science, University of Arkansas, 2650 N. Young Ave., Fayetteville, AR 72704, U.S.A. 2 Department of Crop, Soil and Environmental Sciences, University of Arkansas, 115 Plant Science Bldg., Fayetteville, AR 72701, U.S.A AACC International, Inc. shown to reduce agronomic yield as a result of shattering and lodging (Lu et al. 1995). Chalkiness is linked to the process of starch accumulation in the rice endosperm (Kim et al. 2000; Lisle et al. 2000; Patindol and Wang 2003; Ashida et al. 2009; Patindol et al. 2014). In addition to the previously mentioned marketing issues, chalkiness leads to increased breakage during milling, thus reducing HRY (Counce et al. 2005; Lanning et al. 2011). Whereas chalkiness of rice is affected by climate during reproductive growth (Counce et al. 2005; Ambardekar et al. 2011; Lanning et al. 2011), chalkiness may also be affected by production management. In terms of nitrogen management, Ahmad et al. (2009) showed that increasing the N rate resulted in decreased opaque (chalky) kernels. Nangju and De Datta (1970) reported increased milling quality of chalky rice cultivars in response to increased N fertilizer rates, but they did not directly report chalkiness levels in the remaining head rice. Reports indicate a positive relationship between increased milling yields and increased protein content resulting from increased N application (Nangju and De Datta 1970; Seetanun and De Datta 1973; Perez et al. 1996; Ghosh et al. 2004; Ahmad et al. 2009). Dilday (1988) surmised that increased HRY associated with increasing N application could result from increased protein content and improved structural integrity of the rice kernel. In terms of harvest management, Kester et al. (1963) reported fewer chalky kernels in early harvested rice compared with late-harvested rice. Cooking quality and end-use functionality are impacted by the highly variable nature of rice starch (Bergman et al. 2004). In addition, paste viscosities (cooking quality) of rice have been shown to decrease with increased protein content (Martin and Fitzgerald 2002) and increased chalkiness (Ashida et al. 2009). When compared with chalky kernels, Singh et al. (2003) attributed the firmer texture of cooked, translucent rice kernels to greater amylose content. Thus, N rate and HMC may impact not only milling yields but also end-use functionality of rice by altering the protein content and chalkiness. Although the impacts of N rate and HMC on milling yields have been separately reported, little research has been reported detailing the interrelated manner in which these management factors affect kernel physicochemical properties and end-use functionality of rice. Thus, a multiyear field study was conducted to determine the impacts of HMC and N fertilizer management, and their potential interaction, on the physicochemical properties and milling yields of rice. MATERIALS AND METHODS Cultural Management. The field study was conducted in 2011, 2012, and 2013 near Stuttgart, Arkansas, at the University of 172 CEREAL CHEMISTRY

2 Arkansas Rice Research and Extension Center. Individual plot area was 8.4 m 2, 1.8 m wide and 4.9 m long. Three long-grain cultivars (Cheniere, Wells, and CL XL745) were used each year of the study. In May of each year, plots were planted with a nine-row drill with a 0.2 m row spacing at recommended seeding rates for each cultivar (Hardke 2013). In mid-june of each year, N rates of 0, 45, 90, 135, and 180 kg of N/ha were applied to a dry soil as a single application of urea one day prior to flood application. A single, well-timed application of N fertilizer (an optimum preflood N rate of approximately 145 kg of N/ha, depending on cultivar) is currently an option for Arkansas rice producers, in comparison with the longstanding recommendation for split application (approximately 170 kg of N/ha, depending on cultivar) (Roberts and Hardke 2015). The study was organized as a complete factorial design (3 cultivars 5N rates 4 replications), with a total of 60 plots. Harvesting and Conditioning of Rough Rice. The target HMCs (regardless of property measured, all mass percentages are reported on a wet basis [as is]) were consistent for all cultivars, at 23, 19, and 15%, all ± 1 percentage point (pp) and corresponding to early, optimal, and late harvest, respectively. For each harvest year, a maximum of 180 rough rice samples (3 HMCs 60 plots) were collected over an approximately 45 day period. Immediately prior to harvest, HMC was determined for each plot by first hand-stripping 200 kernels of rough rice from six randomly selected panicles, and then promptly after collection measuring the moisture content (MC) of the kernels with a single-kernel moisture meter (CTR 800E, Shizuoka Seiki, Shizuoka, Japan). Within a cultivar, mean and standard deviation of single-kernel MCs varied little with respect to year or N rate (data not shown). Thus, to illustrate the effects of HMC level, summary statistics for the 180 kg of N/ha rate are shown for all cultivars harvested in the 2011 harvest (Table I). Upon verification of appropriate HMC, sufficient panicles were handharvested from each plot and threshed on location by using a portable thresher (SBT, Almaco, Nevada, IA, U.S.A.), yielding approximately 1.5 kg of rough rice. The average MC of the 200 individual kernels was used as the 1.5 kg sample HMC. Threshed samples of rough rice were maintained at ambient temperature in cloth bags for transport to the laboratory. Rough rice samples were cleaned with a dockage tester (XT4, Carter-Day, Minneapolis, MN, U.S.A.). Cleaned rough rice was conditioned to 12.5 ± 0.5% MC in a climate-controlled chamber (26 C and 56% relative humidity) regulated by a stand-alone conditioner (5580A, Parameter Generation and Control, Black Mountain, NC, U.S.A.). Conditioning of rough rice required from one to three days, depending on the HMC level. Rough rice MCs were measured by drying duplicate 10 g subsamples at 130 C for 24 h in a convection oven (1370FM, Sheldon Manufacturing, Cornelius, OR, U.S.A.) (Jindal and Siebenmorgen 1987). Conditioned rough rice samples were stored in air-tight, zippered plastic bags at 4 ± 1 C and were equilibrated to room temperature (22 ± 1 C) for at least 24 h prior to processing. Brown Rice Properties. Brown rice properties were determined by first dehulling a 100 g subsample of rough rice with a laboratory sheller (THU 35B, Satake, Hiroshima, Japan) with a clearance of cm between the rollers. Any remaining rough rice or broken kernels were manually removed prior to subsequent analyses. Dimensions, comprising length, width, and thickness of 400 brown rice kernels, were quantified with a rice image analyzer (1A, Satake). For total lipid content, 20 g of each brown rice subsample was ground in a cyclone mill ( , UDY, Fort Collins, CO, U.S.A.) fitted with a 100 mesh (0.5 mm) sieve to produce brown rice flour. Total lipid content was determined for a 3 g subsample of brown rice flour with a lipid extraction system (Avanti 2055, Foss North America, Eden Prairie, MN, U.S.A.) according to AACC International Approved Method , with modifications to the petroleum ether washing duration, as described by Matsler and Siebenmorgen (2005). Then, total lipid content was expressed as the mass percentage of extracted lipid relative to the mass of brown rice. Crude protein content of brown rice was determined as the average of two scans of a 50 g subsample of intact kernels with nearinfrared reflectance (NIR, DA7200, Perten Instruments, Hägersten, Sweden). This methodology was conducted with intact brown rice kernels, in a similar manner to AACCI Approved Method for whole-grain wheat. Prior to NIR detection of crude protein content for this study, NIR results from a sample set comprising brown rice and milled rice samples of various cultivars and milling durations (n = 40) were calibrated with corresponding determination of N following AACCI Approved Method and subsequently converted to crude protein content (N 5.95). The resulting formula, CP = 0:747 CP NIR + 1:893 ðr 2 (1) = 0:954; P < 0:0001Þ represents the relationship between the approved and NIR methodologies, where CP denotes the crude protein content as determined by the approved method and CP NIR denotes that determined with the NIR method. Crude protein was reported as a mass percentage of protein relative to the mass of brown rice. Defects of brown rice kernels (fissures and chalkiness) were quantified. Fissured kernels were visually determined for brown rice kernels by using a grain scope (TX-200, Kett Electric Laboratory, Tokyo, Japan) designed specifically for observing the exterior quality of rice kernels. On evaluation of 100 intact kernels, fissuredkernel percentage was reported as a number percentage of brown rice kernels with one or more fissures. A digital scanning system (WinSeedle Pro 2005a, Regent Instruments, Sainte-Foy, Quebec, Canada) was used to determine chalkiness of brown rice. Chalkiness was reported as a percentage of total scanned area of 200 brown rice kernels in the manner of Ambardekar et al. (2011). Milling Yield and Head Rice Properties. Milling properties were determined for each sample. One milling subsample (150 g) of rough rice was taken from each sample and dehulled as described. The resultant brown rice was milled (McGill number 2, RAPSCO, Brookshire, TX, U.S.A.; equipped with a 1.5 kg weight on the lever arm, situated 15 cm from the milling chamber centerline) for a 30 s duration. The resulting milled rice mass was quantified, after which head rice was separated from brokens with a sizing device (61, Grain Machinery Manufacturing Corp., Miami, FL, U.S.A.), and the resulting mass of head rice was quantified. After the surface lipid content of head rice was estimated (see equation 2), MRY and HRY were then adjusted to a consistent degree of milling (0.4% surface lipid content) in the manner of Pereira et al. (2008). Head rice surface lipid content was determined as an indicator of degree of milling by using the previously described NIR system as TABLE I Summary Statistics for Single-Kernel Moisture Content (SKMC) a SKMC b Cultivar HMC Mean SD Min. Max. Cheniere CL XL Wells a Data include target harvest moisture content (HMC), mean SKMC (Mean), standard deviation of the mean (SD), minimum recorded SKMC (Min.), and maximum recorded SKMC (Max.). All moisture content data are reported on a wet basis. b Data for SKMC represents only the 180 kg of N/ha rate during the 2011 harvest year. Vol. 93, No. 2,

3 described by Saleh et al. (2008), with the sampling procedure described herein for brown rice crude protein content. Prior to NIR detection of surface lipid content for this study, NIR results from a sample set comprising various cultivars and milling durations (n = 40) were calibrated to surface lipid content results determined following AACCI Approved Method as modified by Matsler and Siebenmorgen (2005), for which lipid from a 5 g subsample of intact head rice kernels was extracted. The resulting formula, SLC = 0:871 SLC NIR _ 0:092 ðr 2 (2) = 0:976; P < 0:0001Þ represents the relationship between the approved and NIR methodologies, where SLC denotes the surface lipid content as determined by the approved method and SLC NIR denotes that determined with the NIR method. Surface lipid content was reported as the mass percentage of extracted lipid relative to the mass of head rice. Head rice crude protein content was determined simultaneously with surface lipid content in the manner previously described for brown rice. Head rice crude protein content was reported as a mass percentage of crude protein relative to the mass of head rice. Head rice chalkiness was determined with the digital scanning system as described for brown rice and was reported as a percentage of total scanned area of 200 head rice kernels. Paste viscosities were determined according to AACCI Approved Method , for which 15 g of each head rice subsample was ground in the previously described cyclone mill to produce flour. Moisture content of each flour sample was determined by drying duplicate 2 g subsamples at 130 C for 2 h in the previously mentioned convection oven. Adjusted for MC, viscosities were determined on a paste of one 3 g subsample of rice flour in 25 ml of distilled water with a viscometer (RVA Super 4, Newport Scientific, Warriewood, NSW, Australia). The flour paste was held at 50 C for 1.0 min, heated to 95 Cat12 C/min, held at 95 Cfor2.5min,and cooled to 50 C at 12 C/min. Peak viscosity and final viscosity are reported as dynamic viscosity (cp). Statistics. Statistical software (JMP Pro version 11, SAS Institute, Cary, NC, U.S.A.) was used for analysis of data. Coefficients of determination (R 2 ) for regression analyses and associated statistical significance (P values) were determined. For comparison of data, analysis of variance was conducted, and means were compared with the Tukey Kramer honestly significant difference procedure (JMP Pro version 11). Statistical significance (P < 0.05) was determined, and significant differences were indicated with separateletter reporting. Although there were year interactions with both HMC and N rate, these were the result of year-to-year differences in magnitude of measured properties rather than differences in response trends. As a result, data were averaged across the 2011, 2012, and 2013 production years. Statistical significance of N rate, HMC, and N rate HMC effects on brown rice properties are detailed in Table II, and milling yields and head rice properties are detailed in Table III. Statistical significance of the N rate HMC interaction was infrequent and inconsistent across cultivars; moreover, such interactions resulted from N rates of 0 and 45 kg of N/ha, highly unlikely in U.S. production of rice. Thus, discussion will focus on main effects of N rate and HMC. RESULTS AND DISCUSSION Brown Rice Properties. Increasing N rates resulted in decreased kernel length for all cultivars and in decreased kernel thickness for Cheniere and Wells (Fig. 1A and E). Kernel width was not affected by N rate, regardless of cultivar (Fig. 1C). Limited reports of the impacts of N rate on rice kernel length were available, with focus on long-grain aromatic rice cultivars (Khalid and Chaudhry 1999; Mahajan et al. 2011). Mahajan et al. (2011) reported static to slightly increasing kernel length in response to TABLE II Statistical Significance of Model Terms for Brown Rice Properties a Brown Rice Cultivar Effect Length (mm) Width (mm) Thickness (mm) TLC (% mass) CP (% mass) FKP (% number) Chalkiness (% area) Cheniere N rate *** ns *** ns *** ns ns HMC ** ns ** ns *** *** *** N rate HMC ns ns ns ns ns ns * CL XL745 N rate *** ns ns ** *** ns *** HMC *** ns ** * * *** *** N rate HMC ns ns ns ns ns * ns Wells N rate ** ns *** ns *** *** *** HMC *** ns *** ** ** *** *** N rate HMC ns ns ns ns ns *** ns a Statistical significance indicated as * (P < 0.05), ** (P < 0.01), *** (P < ), or ns (not significant). Brown rice properties included kernel length, width, and thickness, total lipid content (TLC), crude protein (CP), fissured-kernel percentage (FKP), and chalkiness. TABLE III Statistical Significance of Model Terms for Milling Yields and Head Rice Properties a Yield b Head Rice Cultivar Effect MRY (% mass) HRY (% mass) SLC (% mass) CP (% mass) Chalkiness (% area) PV (cp) FV (cp) Cheniere N rate ** ** *** *** ns *** *** HMC ** *** *** *** ** ** * N rate HMC ** ns ns ns ns * * CL XL745 N rate *** ** *** *** ** *** ns HMC *** *** * ** ns * ** N rate HMC ns ns ns ns ns ns ns Wells N rate *** *** *** *** *** *** *** HMC *** *** ** ** ** ** *** N rate HMC * *** ns ns ns * ns a Statistical significance indicated as * (P < 0.05), ** (P < 0.01), *** (P < ), or ns (not significant). Properties evaluated include milled rice yield (MRY), head rice yield (HRY), surface lipid content (SLC), crude protein (CP), chalkiness, peak viscosity (PV), and final viscosity (FV). b Yield data were adjusted to a degree of milling of 0.4% SLC according to the method of Pereira et al. (2008). 174 CEREAL CHEMISTRY

4 increasing N rates from 0 to 60 kg of N/ha, whereas Khalid and Chaudhry (1999) reported peak kernel length at 80 kg of N/ha followed by decreased length at 120 kg of N/ha. Decreasing HMCs from 23 to 15% levels also reduced kernel length for all cultivars, and for the CL XL745 and Wells cultivars, the 15% HMC level resulted in significantly reduced kernel length compared with either the 19 or 23% HMC levels (Fig. 1B). Exhibiting a similar trend, kernel thickness was generally the greatest at the 23% HMC level, significantly declining at the 19% HMC level for the Cheniere and CL XL745 cultivars and at the 15% level for the Wells cultivar (Fig. 1F). As observed with N rate, HMC level had no significant impact on kernel width (Fig. 1D). Bautista et al. (2007) measured brown rice kernel dimensions at the actual HMC level, attributing a similar trend for decreasing kernel dimensions to increased kernel shrinkage associated with decreasing HMC. However, for this current report, all samples were dried to 12.5 ± 0.5% MC prior to measurement of kernel dimensions; thus, these observed differences were not attributable to kernel shrinkage. It is surmised that changing kernel dimensions are a response to a shift in harvested kernel population, discussed in greater detail in conjunction with results of brown rice chalkiness. With respect to N rate, total lipid content was statistically different only at the lowest level (0 kg of N/ha) for a single cultivar, CL XL745 (Fig. 2A). Both the 0 and 45 kg of N/ha rates are unlikely for production of rice in the mid-south United States; thus, total lipid content was considered to be unaffected by N rate. Decreased total lipid content of brown rice would be expected in more mature, completely filled kernels, because the lipid-rich bran layer forms early in kernel development, in advance of endosperm development (Lu and Luh 1991). In response to reduced HMC levels, limited differences were observed only for the CL XL745 and Wells cultivars (Fig. 2B), thus suggesting minimal increases in overall maturity of the harvested kernels in response to decreasing HMC levels. Brown rice crude protein content increased consistently with increasing N rates for all cultivars (Fig. 2C), agreeing with previous reports (Nangju and De Datta 1970; Seetanun and De Datta 1973; Perez et al. 1996; Ghosh et al. 2004; Ahmad et al. 2009). When increasing the N rate from 0 to 180 kg of N/ha, differentials of 1.2, 0.9, and 0.6 pp crude protein content were observed for the Cheniere, CL XL745, and Wells cultivars, respectively. Brown rice crude protein content also increased with decreasing HMC level Fig. 1. Kernel length (A and B), width (C and D), and thickness (E and F), measured at 12.5 ± 0.5% moisture content, of the indicated long-grain rice cultivars as affected by nitrogen fertilizer rate (N rate) or harvest moisture content (HMC). Data are averaged across 2011, 2012, and 2013 production years. Standard error of the mean is denoted by a capped bar at the top of each column. Comparisons are valid within a cultivar, and columns with different letters are significantly different (P < 0.05). Vol. 93, No. 2,

5 (Fig. 2D); however, the 0.3 pp differences between the 15 and 23% HMC levels were unlikely to be of practical significance, as previously suggested by Champagne et al. (2005). Brown rice fissured-kernel percentages generally declined in response to increasing N rates (Fig. 3A), although significant differences were observed only with the Wells cultivar. The fissuredkernel percentage data in response to increasing N rates suggest improved structural integrity of rice kernels, warranting further investigation. In contrast, the impact of HMC level on brown rice fissured-kernel percentages was significant for all cultivars (Fig. 3B), with the greatest fissured-kernel percentages observed at the HMC level of 15%. Differences in fissured-kernel percentages were minimal between the 19 and 23% HMC levels, with statistical difference observed only for the Cheniere cultivar. Increased fissured-kernel percentages in response to HMCs below optimal levels agree with the findings of Siebenmorgen et al. (1998, 2007). Increasing N rates resulted in significantly decreased chalkiness of brown rice for the cultivars CL XL745 and Wells (Fig. 3C), agreeing with the results of Ahmad et al. (2009). The impact of increasing N rates was not significant for Cheniere, a typically low-chalk cultivar. However, brown rice chalkiness of all cultivars increased as HMC levels decreased from 23 to 15% (Fig. 3D). Increased chalkiness is associated with loosely packed starch (Ashida et al. 2009). Thus, brown rice chalkiness likely decreased structural integrity of the rice kernel, exacerbating the impacts of fissured-kernel percentages at the 15% HMC level (Fig. 3B). Moreover, increased brown rice chalkiness in response to decreasing HMC level suggests that the associated harvest delays resulted in a population shift to more late-maturing kernels of secondary quality. Espinosa-Mendoza et al. (2012), Mohapatra and Sahu (1991), and Mohapatra et al. (1993) indicate that early maturing kernels in superior positions on the rice panicle result in higher-quality kernels when compared with kernels in other, inferior panicle positions that mature later. Bautista and Siebenmorgen (2005) reported a bimodal distribution of individual kernel MCs at the equivalent of the 23% HMC level, suggesting that individual kernels associated with the lower-mc mode were kernels of advanced maturity and dryness. Illustrated with single-kernel MC data from the 2011 harvest (CL XL745 at the 180 kg of N/ha rate), the bimodal distribution is evident for the 23 and 19% HMC levels, with a remnant of the higher-hmc mode at the 15% HMC level (Fig. 4). Mohapatra et al. (1993) also found that the earliest-developing kernels exhibited greater dry matter content, starch accumulation, and density than later-developing kernels. In relation to greater starch accumulation reported by Mohapatra et al. (1993), Chen et al. (2008) found that kernels in superior panicle positions had greater HRY and reduced chalkiness than those in inferior positions. Lu et al. (1995) also reported increased shattering and predation of early maturing, high-quality kernels in association with harvest delays. The loss of early developing, high-quality kernels to shattering would exacerbate the shift to harvest of a greater proportion of lesser-quality kernels in response to reduced HMCs. The data presented here, including both decreasing kernel length and thickness (Fig. 1B and F) and increasing brown rice chalkiness in response to declining HMCs (Fig. 3D), support such a shift to decreased overall kernel quality. Milling Yields and Head Rice Properties. Increasing the N rate from 0 to 90 kg of N/ha resulted in consistently increased MRYs for the CL XL745 cultivar, with MRYof Cheniere and Wells only increasing when the N rate increased from 0 to 45 kg of N/ha (Fig. 5A). Otherwise, when considering the kg of N/ha rates that may be utilized in the mid-south United States, MRYs were stable. Depending on cultivar, MRYs increased from 0.8 to 2.3 pp in response to increasing N rates from 0 to 180 kg of N/ha, agreeing with trends reported by Nangju and De Datta (1970). Regardless of cultivar, the 23% HMC resulted in reduced MRYs, ranging from Fig. 2. Total lipid content (TLC) (A and B) and crude protein (CP) (C and D) of brown rice, measured at 12.5 ± 0.5% moisture content, of the indicated long-grain rice cultivars as affected by nitrogen fertilizer rate (N rate) or harvest moisture content (HMC). Data are averaged across 2011, 2012, and 2013 production years. Standard error of the mean is denoted by a capped bar at the top of each column. Comparisons are valid within a cultivar, and columns with different letters are significantly different (P < 0.05). 176 CEREAL CHEMISTRY

6 0.3 to 1.1 pp less than MRYs at either the 19 or 15% HMC levels (Fig. 5B). The reduced MRYs at the 23% HMC level were likely a reflection of a larger proportion of immature kernels, as has been previously reported by Kocher et al. (1990) and Huck et al. (2012). There were no statistical differences between MRYs of rice harvested at the 19 and 15% HMC levels (Fig. 5B), likely reflecting only a limited increase in overall kernel maturity, as indicated by consequent declines in total lipid contents (Fig. 2B). The impacts of both N rate and HMC on HRYs (Fig. 5C and D) were generally of greater magnitude than observed with MRYs (Fig. 5A and B). In response to increasing N rate, HRYs (Fig. 5C) followed a similar trend to that of MRYs (Fig. 5A). Depending on cultivar, HRY increases ranged from 4.6 to 11.4 pp as N rate increased, likely reflecting improved kernel integrity in response to increased crude protein content, as well as decreased fissured-kernel percentage and chalkiness of the brown rice kernels (Figs. 2C, 3A, and 3C). With respect to HMC, HRYs of rice harvested at the optimal HMC of 19% were numerically greatest but were not significantly different from rice harvested at the 23% HMC level (Fig. 5D), with the slight differences likely the result of a greater proportion of immature kernels lost during dehulling and milling, as previously discussed. However, with further harvest delays to reach the HMC of 15%, HRYs declined significantly from maximum levels by 4.1, 11.1, and 13.2 pp for the cultivars Cheniere, CL XL745, and Wells, respectively (Fig. 5D). Decreased HRYs at the 15% HMC level were attributable to both increased brown rice fissured-kernel percentage and chalkiness (Fig. 3B and D). Both N rate and HMC level impacted degree of milling, as indicated by surface lipid content (Fig. 5E and F). Regardless of cultivar, surface lipid contents increased by approximately 0.1 pp in response to increasing the N rate from 0 to 180 kg of N/ha. With respect to harvest management, decreasing the HMC level from 23 to 15% increased surface lipid contents by approximately 0.05 pp regardless of cultivar; however, surface lipid contents of the 19 and 15% HMC levels were not significantly different. All samples were milled for the same duration; thus, surface lipid contents increased in response to decreasing kernel dimensions (Fig. 1), agreeing with the results of Chen et al. (1998), who reported reduced degree of milling (i.e., greater surface lipid content) of thinner kernels fractioned after being milled in bulk. Thus, both increasing N rates and decreasing HMC levels would lead to extended milling durations to achieve a desired degree of milling. Trends for crude protein content of head rice were similar to those of brown rice and increased with increasing N rate (Fig. 6A). A Fig. 4. Distributions (percentages of kernels) of single-kernel moisture content as affected by the target harvest moisture content (HMC). Singlekernel moisture content data are for the CL XL745 cultivar at the 180 kg of N/ha rate. Fig. 3. Fissured-kernel percentage (FKP) (A and B) and chalkiness (C and D) of brown rice, measured at 12.5 ± 0.5% moisture content, of the indicated long-grain rice cultivars as affected by nitrogen fertilizer rate (N rate) or harvest moisture content (HMC). Data are averaged across 2011, 2012, and 2013 production years. Standard error of the mean is denoted by a capped bar at the top of each column. Comparisons are valid within a cultivar, and columns with different letters are significantly different (P < 0.05). Vol. 93, No. 2,

7 small proportion of the increased crude protein content in response to increasing N rate was attributable to the consequently greater surface lipid content. Greater surface lipid contents indicate greater amounts of bran, and consequently greater crude protein content, remaining on milled kernels, because the crude protein content of the bran layer is greater in relation to the endosperm (Juliano 1993). Head rice crude protein contents were not significantly affected by decreasing HMCs from 23 to 19% for any cultivar (Fig. 6B). However, decreasing HMC to the 15% level resulted in significantly greater crude protein contents; this increase was likely in response to the previously discussed shift to harvest of late-developing kernels. Chalkiness of head rice also followed similar trends to those of brown rice. Excepting the Cheniere cultivar, head rice chalkiness decreased in response to increasing N rates (Fig. 6C), agreeing with the results of Perez et al. (1996). Excepting the CL XL745 cultivar, head rice chalkiness increased with decreasing HMC levels (Fig. 6D), agreeing with the results of Mohapatra et al. (1993). Thus, increasing either N rate or HMC levels would decrease head rice chalkiness, potentially improving both the visual quality and enduse functionality. Peak viscosities decreased by approximately 320, 400, and 270 cp for the Cheniere, CL XL745, and Wells cultivars, respectively, in response to increasing the N rate from 0 to 180 kg of N/ha (Fig. 7A). These decreases in peak viscosity with increasing N rate were likely owing to a combination of increasing crude protein content (Juliano et al. 1964; Martin and Fitzgerald 2002) and increasing surface lipid content (Perdon et al. 2001). However, when considering only the potentially applicable N rates of 90, 135, and 180 kg of N/ha, the differentials in peak viscosity were minimal, in the range of cp. In response to decreasing HMC, significant increases in peak viscosity were observed between the HMCs of 23 and 15% (Fig. 7B); however, the overall increase in peak viscosities in response to decreasing HMC levels was minimal, approximately 100 cp, regardless of cultivar. In the same manner as peak viscosities, final viscosities also decreased as N rate increased (Fig. 7C). In response to increasing N rate, decreases were of lesser magnitude than those of peak Fig. 5. Milled rice yield (MRY) (A and B), head rice yield (HRY) (C and D), and surface lipid content of head rice (SLC) (E and F) of the indicated long-grain rice cultivars as affected by nitrogen fertilizer rate (N rate) or harvest moisture content (HMC). Data are averaged across 2011, 2012, and 2013 production years. Samples were milled for 30 s, and milling yields were adjusted to a degree of milling of 0.4% SLC according to the method of Pereira et al. (2008). Standard error of the mean is denoted by a capped bar at the top of each column. Comparisons are valid within a cultivar, and columns with different letters are significantly different (P < 0.05). 178 CEREAL CHEMISTRY

8 Fig. 6. Crude protein content (CP) (A and B) and chalkiness (C and D) of head rice, measured at 12.5 ± 0.5% moisture content, of the indicated longgrain rice cultivars as affected by nitrogen fertilizer rate (N rate) or harvest moisture content (HMC). Data are averaged across 2011, 2012, and 2013 production years. Standard error of the mean is denoted by a capped bar at the top of each column. Comparisons are valid within a cultivar, and columns with different letters are significantly different (P < 0.05). Fig. 7. Peak viscosity (PV) (A and B) and final viscosity (FV) (C and D) of rice flour of the indicated long-grain rice cultivars as affected by nitrogen fertilizer rate (N rate) or harvest moisture content (HMC). Data are averaged across 2011, 2012, and 2013 production years. Standard error of the mean is denoted by a capped bar at the top of each column. Comparisons are valid within a cultivar, and columns with different letters are significantly different (P < 0.05). Vol. 93, No. 2,

9 viscosities, with final viscosities decreasing approximately 270, 120, and 200 cp for the Cheniere, CL XL745, and Wells cultivars, respectively. Again, when considering only the 90, 135, and 180 kg of N/ha rates, the differentials in final viscosity were further minimized, in the range of cp. The impacts of decreasing HMC (i.e., delaying harvest) on final viscosities were similar to those on peak viscosities, with final viscosities increasing approximately 90, 70, and 120 cp for the Cheniere, CL XL745, and Wells cultivars, respectively (Fig. 7D). Decreased peak and final viscosities were attributed to increased crude protein content resulting from either increasing N rate or decreasing HMC level, as previously reported by Juliano et al. (1964) and Martin and Fitzgerald (2002). Decreased chalkiness was also associated with increasing N rate (Fig. 6C), which would have contributed to increased viscosities, as previously reported by Lisle et al. (2000), Ashida et al. (2009), and Chun et al. (2009). Thus, decreased chalkiness likely muted the effects of increased head rice crude protein content (Fig. 6A). Peak and final viscosities increased minimally in response to decreasing HMCs (Fig. 7B and D). Kester et al. (1963) attributed significantly increased viscosities at reduced HMCs to advanced kernel maturity and associated decreased amylase activity. However, the results reported herein indicate lesser increases in viscosities than would be expected for a kernel population of advanced maturity. Siebenmorgen et al. (2006) also reported only minor increases in peak viscosities of three long-grain rice cultivars in response to declining HMC levels, as well as no differences in a-amylase activities or amylose contents. Thus, the trends reported by Siebenmorgen et al. (2006) are inconsistent with a static kernel population of increased maturity; rather, they agree with kernel dimensions, chalkiness, crude protein content, and viscosity data presented here, supporting a shift to a population containing a larger proportion of later-developing kernels. Less-mature, latedeveloping kernels of increased a-amylase activity would likely contribute to reduced viscosities when admixed with kernels of advanced maturity, also supported by a-amylase activities and amylose contents reported by Siebenmorgen et al. (2006). Moreover, increased crude protein content (Juliano et al. 1964; Martin and Fitzgerald 2002) and chalkiness levels (Lisle et al. 2000; Singh et al. 2003; Ashida et al. 2009; Chun et al. 2009) would have further suppressed viscosities of low-hmc rice samples. CONCLUSIONS With the exception of decreasing kernel length and thickness, increasing the N rate from 0 to 180 kg of N/ha positively impacted crude protein content, chalkiness, and milling yields of all three cultivars. Currently, in Arkansas, the approximate recommended N rate is 145 kg of N/ha for a single, well-timed application (optimum preflood N rate) (Roberts and Hardke 2015). However, at the 135 and 180 kg of N/ha rates bracketing the recommended rate, kernel length and thickness were stable. Increasing N rate also resulted in increased surface lipid contents for all cultivars, indicating that longer milling durations would be required for rice receiving greater N rates. For all cultivars, increasing N rate produced trends of decreasing peak viscosity and final viscosity, reflecting increased crude protein content and decreased chalkiness. Trends in observed milling yields and physicochemical properties were consistent across the full 0 to 180 kg of N/ha range of N rates; however, when considering only the 90 to 180 kg of N/ha rates, significant differences were infrequent and inconsistent with respect to cultivar. As HMCs decreased, kernel dimensions decreased and crude protein content and chalkiness increased. Reducing HMC levels from 23 to 19% resulted in slightly increased MRYs and stable HRYs, with expected gains in HRY resulting from increased kernel maturity being offset by increased chalkiness. Further delaying harvest to decrease the HMC to 15% did not impact MRYs but significantly reduced HRYs, a direct response to increased fissured-kernel percentage and chalkiness. In addition to the obvious negative impact to HRY, harvesting at the 15% HMC level would also decrease kernel size and increase chalkiness of head rice. In contrast to the decreased viscosities associated with increasing the N rate, decreasing HMC levels had little impact on paste viscosities. Significant interactions between N rate and HMC level were infrequent and inconsistent with respect to the property measured and cultivar. Moreover, such N rate HMC interactions were the result of the 0 and 45 kg of N/ha rates, not utilized in U.S. rice production. Thus, at rates potentially used in U.S. production of rice (90 kg of N/ha or greater), changing N rate did not exacerbate or minimize the effects of either the 23% (i.e., early harvest) or the 15% HMC (i.e., delayed harvest) levels. The response of measured indices to the 15% HMC level (i.e., delayed harvest) suggests a shift in kernel population, inclusive of a greater proportion of late-maturing, chalky kernels; thus, harvest delays to achieve HMCs less than the 19% HMC level likely sacrifice rice quality in the process. Moreover, in-field drying to the 15% HMC level greatly increased the potential for kernel fissuring in response to rapid moisture adsorption events. Thus, economic gains realized from extensive in-field drying to HMCs below the optimal 19% level (Siebenmorgen et al. 2008) would likely be countered by reduced HRYs and diminished quality of the harvested rice. ACKNOWLEDGMENTS The authors express their thanks to the staff of the University of Arkansas Rice Processing Program for sample collection, processing, and analysis. The authors acknowledge the Arkansas Rice Research and Promotion Board and the corporate sponsors of the University of Arkansas Rice Processing Program for financial support of this project. LITERATURE CITED AACC International. Approved Methods of Analysis, 11th Ed. Method Crude fat in wheat, corn, and soy flour, feeds, and mixed feeds. Approved April 13, Method Near-infrared reflectance method for protein content in whole-grain wheat. Approved November 8, Method Crude protein Improved Kjeldahl method, copper-titanium dioxide catalyst modification. Approved October 12, Method Determination of the pasting properties of rice with the Rapid Visco Analyser. Approved October 15, Available online only. AACCI: Saint Paul, MN. Ahmad, S., Zia-Ul-Haq, M., Ali, H., Ahmad, A., Khan, M. A., Khaliq, T., Husnain, Z., Hussain, A., and Hoogenboom, G Morphological and quality parameters of Oryza sativa L. as affected by population dynamics, nitrogen fertilization and irrigation regimes. Pak. J. Bot. 41: Ambardekar, A. A., Siebenmorgen, T. J., Counce, P. A., Lanning, S. B., and Mauromoustakos, A Impact of field-scale nighttime air temperatures during kernel development on rice milling quality. Field Crops Res. 122: Ashida, K., Iida, S., and Yasui, T Morphological, physical, and chemical properties of grain and flour from chalky rice mutants. Cereal Chem. 86: Bautista, R. C., and Siebenmorgen, T. J Individual rice kernel moisture content variability trends. Appl. Eng. Agric. 21: Bautista, R. C., Siebenmorgen, T. J., and Counce, P. A Rice kernel dimensional variability trends. Appl. Eng. Agric. 23: Bergman, C. J., Battacharya, K. R., and Ohtsubo, K Rice end-use quality analysis. Pages in: Rice: Chemistry and Technology, 3rd Ed. E. T. Champagne, ed. American Association of Cereal Chemists: Saint Paul, MN. Calderwood, D. L., Bollich, C. N., and Scott, J. E Field drying of rough rice: Effect on grain yield, milling quality, and energy saved. Agron. J. 72: Champagne, E. T., Bett-Garber, K. L., Thompson, J., Mutters, R., Grimm, C. C., and McClung, A. M Effects of drain and harvest dates on rice sensory and physicochemical properties. Cereal Chem. 82: CEREAL CHEMISTRY

10 Chen, H., Siebenmorgen, T. J., and Griffin, K Quality characteristics of long-grain rice milled in two commercial systems. Cereal Chem. 75: Chen, S.-Q., Jin, F., Wang, J.-Y., Liu, B.-L., Dong, D., Xue, J.-F., Zhang, W.-Z., Xu, Z.-J., and Chen, W.-F Comparison of grain chalky characters in different positions between two type panicles of japonica rice. Acta Agric. Boreali-Sinica. 22:1-8. Chun, A., Song, J., Kim, K.-J., and Lee, H.-J Quality of head and chalky rice and deterioration of eating quality by chalky rice. J. Crop Sci. Biotechnol. 12: Counce, P. A., Bryant, R. J., Bergman, C. J., Bautista, R. C., Wang, Y.-J., Siebenmorgen, T. J., Moldenhauer, K. A. K., and Meullenet, J.-F. C Rice milling quality, grain dimensions, and starch branching as affected by high night temperatures. Cereal Chem. 82: Dilday, R. H Effect of nitrogen fertilizer on milling quality of rice (Oryza sativa). Proc. Arkansas Acad. Sci. 42: Dilday, R. H Milling quality of rice: Cylinder speed vs. grainmoisture content at harvest. Crop Sci. 29: Espinosa-Mendoza, R. E., Solorza-Feria, J., Arenas-Ocampo, M. L., Camacho-Díaz, B. H., Del Villar-Martinez, A. A., Vanegas-Expinoza, P. E., and Jiménez-Aparicio, A. R Morphostructural characterization of rice grain (Oryza sativa L.) variety Morelos A-98 during filling stages. Sci. World J. 2012: Geng, S., Williams, J. F., and Hill, J. E Harvest moisture effects on rice milling quality. Calif. Agric. 38: Ghosh, M., Mandal, B. K., Mandal, B. B., Lodh, S. B., and Dash, A. K The effect of planting date and nitrogen management on yield and quality of aromatic rice (Oryza sativa). J. Agric. Sci. 142: Hardke, J. T., ed Arkansas Rice Production Handbook. Misc. Publ University of Arkansas Cooperative Extension Service: Little Rock, AR. Huck, A. C., Lanning, S. B., and Siebenmorgen, T. J Milling characteristics of high and low quality rice. Discovery. 13: Jindal, V. K., and Siebenmorgen, T. J Effects of oven drying temperature and drying time on rough rice moisture content determination. Trans. ASAE 30: Juliano, B. O Rice in Human Nutrition. Food and Agriculture Organization of the United Nations: Rome, Italy. Juliano, B. O., Bautista, G. M., Lugay, J. C., and Reyes, A. C Rice quality, studies on physicochemical properties of rice. J. Agric. Food Chem. 12: Kester, E. B., Lukens, H. C., Ferrel, R. E., Mohammad, A., and Finrock, D. C Influence of maturity on properties of western rices. Cereal Chem. 40: Khalid, M., and Chaudhry, A. U Response of kernel dimensions of fine rice to different levels and methods of nitrogen application. Pak. J. Agric. Sci. 360: Kim, S. S., Lee, S. E., Kim, O. W., and Kim, D. C Physicochemical characteristics of chalky kernels and their effects on sensory quality of cooked rice. Cereal Chem. 77: Kocher, M. F., Siebenmorgen, T. J., Norman, R. J., and Wells, B. R Rice kernel moisture content variation at harvest. Trans. ASAE 33: Lanning, S. B., Siebenmorgen, T. J., Counce, P. A., Ambardekar, A. A., and Mauromoustakos, A Extreme nighttime air temperatures in 2010 impact rice chalkiness and milling quality. Field Crops Res. 124: Lisle, A. J., Martin, M., and Fitzgerald, M. A Chalky and translucent rice grains differ in starch composition and structure and cooking properties. Cereal Chem. 77: Lu, R., Siebenmorgen, T. J., Costello, T. A., and Fryar, E. O., Jr Effect of rice moisture content at harvest on economic return. Appl. Eng. Agric. 11: Lu, S., and Luh, B. S Properties of the rice caryopsis. Pages in: Rice: Production and Utilization. Vol. 2. B. S. Luh, ed. Van Nostrand Reinhold: New York, NY. Mahajan, G., Gill, M. S., Sharma, N., and Bharaj, T. S Effect of nitrogen fertilization on yield and quality of aromatic rice in northwestern Indo-Gangetic Plains. J. Agric. Sci. Technol. B. 1: Martin, M., and Fitzgerald, M. A Proteins in rice grains influence cooking properties! J. Cereal Sci. 36: Matsler, A. L., and Siebenmorgen, T. J Evaluation of operating conditions for surface lipid extraction from rice using a Soxtec system. Cereal Chem. 82: McClung, A Report to the USA rice federation rice marketability and competitiveness task force: Quality assessment of rice samples produced in USA Rice Federation: Arlington, VA. Mohapatra, P. K., Patel, R., and Sahu, S. K Time of flowering affects grain quality and spikelet partitioning within the rice panicle. Aust. J. Plant Physiol. 20: Mohapatra, P. K., and Sahu, S. K Heterogeneity of primary branch development and spikelet survival in rice panicle in relation to assimilates of primary branches. J. Exp. Bot. 42: Nangju, D., and De Datta, S. K Effect of time of harvest and nitrogen level on yield and grain breakage in transplanted rice. Agron. J. 62: Patindol, J., and Wang, Y.-J Fine structures and physicochemical properties of starches from chalky and translucent rice kernels. J. Agric. Food Chem. 51: Patindol, J. A., Siebenmorgen, T. J., Wang, Y.-J., Lanning, S. B., and Counce, P. A Impact of elevated nighttime air temperatures during kernel development on starch properties of field-grown rice. Cereal Chem. 91: Perdon, A. A., Siebenmorgen, T. J., Mauromoustakos, A., Griffin, V. K., and Johnson, E. R Degree of milling effects on rice pasting properties. Cereal Chem. 78: Pereira, T., Cooper, N., and Siebenmorgen, T. J Effects of storage temperature and duration on the milling properties of rice. Discovery. 9: Perez, C. M., Juliano, B. O., Liboon, S. P., Alcantara, J. M., and Cassman, K. G Effects of late nitrogen fertilizer application on head rice yield, protein content, and grain quality of rice. Cereal Chem. 73: Roberts, T. L., and Hardke, J. T recommended nitrogen rates and distribution for rice cultivars in Arkansas. crops-commercial-horticulture/rice/. University of Arkansas: Fayetteville, AR. Saleh, M. I., Meullenet, J. F., and Siebenmorgen, T. J Development and validation of prediction models for rice surface lipid content and color parameters using near-infrared spectroscopy: A basis for predicting rice degree of milling. Cereal Chem. 85: Seetanun, W., and De Datta, S. K Grain yield, milling quality, and seed viability of rice as influenced by time of nitrogen application and time of harvest. Agron. J. 65: Siebenmorgen, T. J., Bautista, R. C., and Counce, P. A Optimal harvest moisture contents for maximizing milling quality of long- and medium-grain rice cultivars. Appl. Eng. Agric. 23: Siebenmorgen, T. J., Bautista, R. C., and Meullenet, J.-F Predicting rice physicochemical properties using thickness fraction properties. Cereal Chem. 83: Siebenmorgen, T. J., Cooper, N. T. W., Bautista, R. C., Counce, P. A., Wailes, E., and Watkins, K. B Estimating the economic value of rice (Oryza sativa L.) as a function of harvest moisture content. Appl. Eng. Agric. 24: Siebenmorgen, T. J., Perdon, A. A., Chen, X., and Mauromoustakos, A Relating rice milling quality changes during adsorption to individual kernel moisture content distribution. Cereal Chem. 75: Singh, N., Sodhi, N. S., Kaur, M., and Saxena, S. K Physicochemical, morphological, thermal, cooking and textural properties of chalky and translucent rice kernels. Food Chem. 82: USDA-ERS Rice Yearbook USDA, Economic Research Service: Washington, DC. USDA-FGIS United States standards for rice. USDA, Federal Grain Inspection Service: Washington, DC. [Received June 26, Accepted October 9, 2015.] Vol. 93, No. 2,