N. A. Streck, I. Lago, F. B. Oliveira, A. B. Heldwein, L .
MODELING THE DEVELOPMENT OF CULTIVATEDRICE AND WEEDY RED RICEN. A. Streck, I. Lago, F. B. Oliveira, A. B. Heldwein, L. A. de Avila, L. C. BoscoABSTRACT. Development models are an important part of crop simulation models and are useful tools for field operations.The objective of this study was to simulate the development of cultivated rice and weedy red rice with a linear model (thermaltime model) and with a non‐linear model (WE model). Data from a four‐year experiment conducted during the 2003‐2004,2004‐2005, 2005‐2006, and 2006‐2007 growing seasons in Santa Maria, Rio Grande do Sul State, Brazil, with nine cultivatedrice genotypes and two weedy red rice biotypes were used. Plants were grown in 12 L pots during the four years and in a paddyrice field during the 2006‐2007 growing season. Developmental data available and used in this study were dates of emergence(EM), panicle differentiation (R1), anthesis (R4), and all grains with brown hulls (R9) recorded in five plants per replicationin each sowing date. Data collected in the 2004‐2005 and 2005‐2006 growing seasons were used to estimate the coefficientsof the two models, and data collected in the 2003‐2004 and 2006‐2007 growing seasons were used as an independent dataset for models evaluation. The root mean square error (RMSE) for all developmental stages varied from 4.9 to 10.5 days withthe thermal time model and from 4.3 to 10.9 days with the WE model. The WE model gave better predictions in six out of elevengenotypes, with better predictions for early (R1) than for later (R4 and R9) developmental stages. The WE model describedproperly the development of rice plants grown in pots and in a paddy rice field and can be recommended as an alternativemodel to the thermal time model for rice in situations when the air temperature falls into the non‐linear range of the responseof development to temperature.Keywords. Modeling, Oryza sativa L., Phenology, Temperature.Plant growth and development are independent pro‐cesses that may occur simultaneously or not. Whilegrowth involves irreversible increase in physical di‐mensions of organs, such as area, length, width,height, volume, density, and diameter, development refers toontogenetic processes at different levels of organization,such as cell differentiation, organ initiation (organogenesis),and appearance (morphogenesis), and extends to crop senes‐cence (Hodges, 1991; Wilhelm and McMaster, 1995). Thesimulation of crop development is an important part of cropsimulation models because partitioning of photoassimilatesto different plant organs is dependent on the developmentalstage, and because many physiological processes areswitched on or off and speed up or down depending on the de‐Submitted for review in February 2009 as manuscript number BE 7921;approved for publication by the Biological Engineering Division ofASABE in December 2010.The authors are Nereu Augusto Streck, Professor, Departamento deFitotecnia, Isabel Lago, Research Technician, Programa de Pós‐Graduação em Engenharia Agrícola, Felipe Brendler Oliveira, ResearchTechnician, Curso de Graduação em Agronomia, and Arno BernardoHeldwein, Professor, Departamento de Fitotecnia, Universidade Federalde Santa Maria, Santa Maria, Brazil; Luis Antonio de Avila, Professor,Departamento de Fitossanidade, Faculdade de Agronomia Eliseu Maciel,Universidade Federal de Pelotas, Pelotas, Brazil; and Leosane CristinaBosco, Research Technician, Programa de Pós‐Graduação em Agronomia,Centro de Ciências Rurais, Universidade Federal de Santa Maria, SantaMaria, Brazil. Corresponding author: Nereu Augusto Streck,Departamento de Fitotecnia, Centro de Ciências Rurais, UniversidadeFederal de Santa Maria, 97105‐900, Santa Maria, RS, Brazil; phone: 55‐55‐3220‐8179; fax: 55‐55‐3220‐8899; e‐mail: nstreck2@yahoo.com.br.velopmental stage (Penning de Vries et al., 1989; Goudrianand Van Lar, 1994). Development simulation models mayalso help in cultivar selection and in planning field operationssuch as fertilization, irrigation, pest control, and crop harvest(Johnson et al., 1986, 1987; Connel et al., 1999). Regardingweeds, development models may help to understand and pre‐dict the complex interactions of weed‐crop competition andthus are important tools for weed management. Developmentmodels are also important for assessing the response of cropsunder climate change scenarios (Weiss et al., 2003; Streck etal., 2006a).Temperature is a major environmental factor that drivesrice (Oryza sativa L.) development (Hodges, 1991; Gao et al.,1992; Infeld et al., 1998). The thermal time model, with unitsof C day, is often used to describe the effect of temperatureon crop development (Russele et al., 1984). The thermal timemodel as a measure of biological time has been applied toplants and insects for a long time (Arnold, 1960; Wang, 1960;Pruess, 1983). Advantages of the thermal time model includethe simplicity in the calculation of degree‐days (usually dailymean air temperature is subtracted from a lower threshold orcritical base temperature) and the fact that it is a better timedescriptor than calendar days to tell time in plants (Gilmoreand Rogers, 1958; Russele et al., 1984; McMaster and Smika,1988).One criticism of the thermal time model is that this modelusually performs well in many “normal” situations, but itmay fail in some situations when temperatures fall outsidethe linear range of the response of plant development to tem‐perature (Xue et al., 2004; Streck et al., 2007), which oftenoccur during heat waves and under climate change scenarios.The assumption in the thermal time model is a linear responseTransactions of the ASABEVol. 54(1): 371-384E 2011 American Society of Agricultural and Biological Engineers ISSN 2151-0032371
of development to temperature. The response of biologicalprocesses to temperature is better described with a non‐linearmodel considering the three cardinal temperatures (mini‐mum, optimum, and maximum) for development. The re‐sponse is linear only in a small range of the response betweenthe minimum and the optimum temperature (Shaykewich,1995; Yin et al., 1995). Another criticism of the thermal timemodel is the lack of consistency among reports regarding theway the calculation of degree‐days is implemented (McMas‐ter and Wilhelm, 1997): in some reports, the temperature issubtracted from the mean daily temperature, in some the basetemperature is subtracted from the minimum and maximumair temperatures and then averaged, and sometimes there isno indication which of the above methods is used in the cal‐culation. An alternative approach to the thermal time modelis a multiplicative model with non‐linear temperature re‐sponse functions (Yin et al., 1995; Streck et al., 2003, 2007;Setiyono et al., 2007).In general, rice development simulation models can be di‐vided into two groups: linear models, which use the thermaltime model (Alocilja and Ritchie, 1991; Infeld et al., 1998;Steinmetz et al., 2004), and non‐linear models (Gao et al.,1992; Yin et al., 1995, 1997). Yin et al. (1995) found that thenon‐linear model was superior to the linear model (thermaltime model) in predicting flowering time in rice. A non‐lineartemperature response function widely used for describing de‐velopmental response in several crops, including rice, is thebeta function (Yin et al., 1995, 1997; Yan and Hunt, 1999).There are several versions of the beta function, which havedifferent numbers of parameters, and the version of the betafunction used by Wang and Engel (1998) is particularly inter‐esting because of the simplicity of its parameters. The Wangand Engel (WE) model was first developed for simulatingwheat development (Wang and Engel, 1998; Streck et al.,2003) and was further extended to simulate potato develop‐ment (Streck et al., 2007). The WE model has not been evalu‐ated for simulating developmental stages in rice, whichconstituted a rationale for this effort. Furthermore, to the bestof our knowledge, rice development models have been usedto simulate the development of cultivated rice (CR), notweedy red rice (RR) (Oryza sativa L.), a major weed in paddyrice fields worldwide (Noldin et al., 1999; Gealy et al., 2003;Ziska and McClung, 2008), which constituted another ratio‐nale for this effort.The objective of this study was to simulate the develop‐ment of CR and RR with linear and non‐linear models. Thehypothesis was that linear and non‐linear models will resultin different predictions of the timing of rice developmentalstages.MATERIALS AND METHODSEXPERIMENTAL DATAData used in this study are from four‐year experimentsconducted in the field research area, Plant Science Depart‐ment, Federal University of Santa Maria, Santa Maria, RioGrande do Sul State, Brazil (29 43′ S, 53 43′ W, 95 m alti‐tude) during the 2003‐2004, 2004‐2005, 2005‐2006, and2006‐2007 growing seasons. This site has a subtropical hu‐mid climate (Cfa, according to Köppen's system) and is rep‐resentative of the rice growing area in southern Brazil. Soiltype at the experimental site was a Rhodic Paleudalf (USDA372Table 1. Growing seasons and sowing dates of cultivated riceand weedy red rice used in the study. Santa Maria,Rio Grande do Sul State, Brazil.SeasonSowing Dates 006‐200701/09/03, 20/10/03, 21/11/03, 05/01/04, 29/01/0402/09/04, 07/10/04, 04/11/04, 03/12/04, 02/03/0526/09/05, 25/11/05, 02/02/0608/11/06, 13/12/06, 16/01/07taxonomy). There were five sowing dates in the 2003‐2004and 2004‐2005 seasons and three sowing dates in 2005‐2006and 2006‐2007 (table 1). The wide range of sowing dateseach year was selected to have plants growing and develop‐ing under different temperatures, which is important for mod‐el parameterization and testing, and they correspond tosowing dates before, during, and after the recommended sow‐ing time for this location, which is from 1 October to 10 De‐cember.Cultivated rice genotypes and RR biotypes used in thisstudy are listed in table 2. The Hybrid, EEA 406, and two RRbiotypes were not used in the 2003‐2004 growing season. TheCR genotypes are of subspecies indica, semi‐dwarf, exceptgenotype EEA 406, which is of subspecies japonica, a tall,broad‐leaf genotype used in Brazil in the 1960s and oftenused as a red rice plant in crop‐weed competition trials(Agostinetto et al., 2004). These genotypes were selected be‐cause they are widely grown in southern Brazil, they have abroad range of rate of development, varying from very early(95 days) (IRGA 421) to late (142 days) (EPAGRI 109), andthey represent both subspecies of Oryza sativa. The two RRbiotypes (blackhull and strawhull) are usually found in ricefields in this location and they have the morphological char‐acteristics that usually differentiate RR from modern CR,i.e., tall plants, light‐green leaves, long and more slender pan‐icles, and heavy shattering of grains (Diarra et al., 1985; Nol‐din et al., 1999). The seeds of the RR biotypes came from asingle plant of natural occurrence that grew in a paddy ricefield near the experimental site during the 2003‐2004 grow‐ing season. The seeds used in the experiments were collectedeach year.In the four years, CR and RR were grown in 12 L (0.30 mdiameter and 0.26 m height) pots filled with local soil, whichhad 2.2% organic matter, 3.2 mg L‐1 P, and 46.0 mg L‐1 K. Soiltype at the experimental site was a Rhodic Paleudalf (USDAtaxonomy). The pots were buried in the soil to maintain potTable 2. Cultivated rice genotypes and weedyred rice biotypes used in the study.Days forMaturationGroupGenotypeSubspecies MaturationIRGA 421IRGA 416IRGA 417IRGA 420BR‐IRGA 409BRS 7 TAIMEPAGRI 109HybridEEA 9115120126130142‐‐140‐‐‐‐Very id‐lateUnknownUnknownYear ‐‐‐‐Not released yet.ABHRR awned blackhull red rice.AYHRR awned yellowhull red rice.TRANSACTIONS OF THE ASABE
Figure 1. Schematic representation of the cultivated rice and weedy red rice developmental cycle and the Wang and Engel (WE) model used in this study:EM emergence, R1 panicle differentiation, R4 anthesis, R9 all grains with brown hulls, r daily developmental rate, rmax,v daily maximumdevelopmental rate during vegetative phase, rmax,r daily maximum developmental rate during reproductive phase, rmax,gf daily maximum develop‐mental rate during the grain filling phase, and f(T) temperature response function.soil temperature similar to the soil temperature of the sur‐rounding area. Pots spacing was 1.5 m 0.8 m, so that shad‐ing was not a major factor affecting plants growth (Petersonet al., 1984). The experiment was conducted in a completelyrandomized design with 11 genotypes and four replications(replication pot), or a total of 44 pots per sowing date.Thirty seeds were sown per pot. At V3 stage of the Counceet al. (2000) scale, the plants were thinned to 15 plants perpot, which corresponded to a plant density of about 200 plantsm‐2 in the pots. This is a plant density commonly used in com‐mercial rice fields in southern Brazil.We also used data from another experiment conductedduring the 2006‐2007 growing season using two cultivars(IRGA 421 and EPAGRI 109) grown in a 10 ha paddy ricefield located about 500 m from the pots on two sowing dates(13/12/06 and 16/01/07). In this paddy rice field, there werefour plots with four rows of 1 m length with 0.17 m spacingamong rows, and the plant density was 200 plants m‐2. Hav‐ing rice plants of these two cultivars growing in pots and ina paddy field simultaneously allowed an assessment ofwhether potted plants have their development affected by re‐stricting rooting volume compared to field‐grown rice plants.Agronomic practices followed local recommendations forrice. In the pots, 20 g per pot of a 7‐11‐9 N‐P‐K fertilizer wasused at sowing, and additional nitrogen was added as a side‐dress application at beginning of tillering (V4 stage of theCounce et al., 2000 scale) and at R1 growth stage of the Counceet al. (2000) scale with urea at a rate of 8.5 g per pot. The samefertilizer rates and top‐dressings were used in the field plots, cor‐responding to 300 kg ha‐1 of 7‐11‐9 N‐P‐K fertilizer and 222 kgha‐1 of urea. Irrigation in both pots and field plots was performedstarting at V3, and a continuous 5 to 7 cm water layer above thesoil surface (flooded soil) was kept until R9.Emergence day was assumed when 50% of the plants werevisible at the soil surface. Five plants in the center part of thepot and five plants in the central rows of the field plots weretagged with colored wires one week after emergence. In se‐lecting plants from the central part of the pot and the rows,we tried to achieve a solar radiation balance of red and far‐redsimilar to the balance encountered in the canopy of a ricefield, which is known to affect development in small grain ce‐reals (Wilhelm and McMaster, 1995). On the main stem oftagged plants, the date of R4 (anthesis, one or more floretsopen) and R9 (all grains with brown hulls) developmentalstages of the Counce scale (Counce et al., 2000) were re‐corded. The R1 (panicle differentiation) growth stage was de‐termined by daily sampling four plants (one plant perreplication) daily from the outside plants in the pots and fromthe border rows in the field plots. The day when 50% of thesampled plants were at R1 (Counce et al., 2000) was consid‐ered the date of R1. Daily minimum (TN) and maximumVol. 54(1): 371-384(TX) air temperatures were measured with a standard meteo‐rological station (Brazilian National Weather Service) lo‐cated about 100 m from the pots and about 500 m from thepaddy rice field. Soil temperatures at 5 cm depth were mea‐sured in the center of one pot and in the center of one plot inthe paddy field with mercury‐in‐glass thermometers (accura‐cy 0.1 C). Soil temperature readings were taken during day‐time (from 6:00 a.m. to 8:00 p.m.) every 2 h except from 1:00p.m. to 4:00 p.m. when readings were taken every hour, dur‐ing four clear‐sky sunny days: 30 January 2007 (one day be‐fore flooding), 1 February 2007 (one day after flooding),12 February 2007 (twelve days after flooding), and 12 March2007 (72 days after flooding in the pots and 66 days afterflooding in the paddy field). The soil temperatures on the firstthree dates were recorded in cultivar IRGA 421 sown on16/01/07, and soil temperatures on the fourth date were re‐corded in cultivar EPAGRI 109 sown on 13/12/06. Thesedates covered a wide range of soil moisture, canopy cover,and crop ontogeny.RICE DEVELOPMENTAL CYCLE AND CARDINALTEMPERATURESThe developmental cycle of CR and RR was divided intothree phases (Gao et al., 1992; Infeld et al., 1998); the vegeta‐tive phase from emergence (EM) to panicle differentiation(R1), the reproductive phase from R1 to anthesis (R4), andthe grain filling phase from R4 to all grains with brown hulls(R9) (fig. 1).Table 3. Cardinal temperatures (Tmin minimum temperature,Topt optimum temperature, and Tmax maximum temperature)for developmental phases of cultivated rice genotypesand weedy red rice biotypes used in the study.Cardinal ypesIRGA 421, IRGA 416, IRGA 417, IRGA 420,BR‐IRGA 409, BRS 7 TAIM, EPAGRI 109,and HybridVegetative (EM‐R1)11[b]30[c]40[c]Reproductive (R1‐R4)15[d],[e]25[f]35[f],[g]Grain filling (R4‐R9)15[[d],[e]23[g]35[f],[g]EEA 406, ABHRR, and AYHRRVegetative (EM‐R1)Reproductive (R1‐R4)Grain filling (R4‐R9)[a][b][c][d][e][f][g]61310302523353030EM emergence, R1 panicle differentiation, R4 anthesis,and R9 all grains with brown hulls.Infeld et al. (1998).Gao et al. (1992).Buriol et al. (1991).Steinmetz (2004).Pedro et al. (1995).Venkataraman et al. (2007).373
Cardinal temperatures for the developmental phases ofCR genotypes and RR biotypes are listed in table 3. The setsof cardinal temperatures were determined by assuming thatminimum and maximum cardinal temperatures for EEA 406were lower than for the indica cultivars because japonica cul‐tivars can usually tolerate lower temperatures (Zeng et al.,2005; Venkataraman et al., 2007). We also assumed the sameset of cardinal temperatures for EEA 406 and the two RR bio‐types because of their similarities in morphology and growthhabit (Streck et al., 2008), because CR and RR biotypes havethe same genome (Zisca and McClung, 2008), and becauseno reports on difference in cardinal temperatures between RRand CR were found.RICE DEVELOPMENT SIMULATION MODELSLinear ModelThe linear model used to predict CR and RR developmen‐tal stages was the thermal time (TT) model. Daily values ofTT, with units of C day, were calculated using the methodof degree‐days presented by Matthews and Hunt (1994) andby Streck et al. (2007):T Tb 0, Tb T Topt (T Tb ) 1 day,TT (Tmax T ) (Topt Tb ), Topt T Tmax (Tmax Topt ) 1 day 0,T Tmax(1)where Tb , Topt , and Tmax are cardinal temperatures (minimum,optimum, and maximum) for each developmental phases,and T is the air temperature.The graphical representation of the TT calculation for dif‐ferent genotypes and developmental phases is shown in fig‐ure 2. The TT was calculated using daily TN and TX and thenaveraged. The accumulated thermal time (ATT, C day) foreach developmental phase (EM to R1, R1 to R4, and R4 toR9) was calculated by accumulating TT, i.e., ATT STT.Figure 2. Thermal time calculation and temperature response function, f(T), of the WE model for developmental phases of cultivated rice and weedyred rice. Cardinal temperatures of the developmental phases are listed in table 3. Graphs a, b, and c are for the indica rice genotypes (IRGA 421, IRGA416, IRGA 417, IRGA 420, BRS 7 TAIM, BR‐IRGA 409, EPAGRI 109 and Hybrid), and graphs d, e, and f are for the japonica genotype (EEA 406)and the two weedy red rice biotypes.374TRANSACTIONS OF THE ASABE
Non‐Linear ModelThe non‐linear development si
Transactions of the ASABE Vol. 54(1): 371-384 2011 American Society of Agricultural and Biological Engineers ISSN 2151-0032 371 MODELING THE DEVELOPMENT OF CULTIVATED RICE AND WEEDY RED RICE N. A .
Texts of Wow Rosh Hashana II 5780 - Congregation Shearith Israel, Atlanta Georgia Wow ׳ג ׳א:׳א תישארב (א) ׃ץרֶָֽאָּהָּ תאֵֵ֥וְּ םִימִַׁ֖שַָּה תאֵֵ֥ םיקִִ֑לֹאֱ ארָָּ֣ Îָּ תישִִׁ֖ארֵ Îְּ(ב) חַורְָּ֣ו ם
conservación en el lago Caballo Cocha (2002-2008) Cuadro No. 3: Calendario de actividades del pescador artesanal de Caballo Cocha por especies de mayor consumo Cuadro No. 4: Desembarque Histórico de la pesca del lago Caballo Cocha (2000-2008) Cuadro No. 5: Total de desembarque del Lago Caballo Cocha periodo 2000- 2009
A1 A9 A4 A8 A4 A26 Lago di Lugano Lago Maggiore Lago di Como TORINO GENOVA VOLTRI GENOVA OGNA VENEZIA MILANO AOSTA IVREA CASELLE ALAGNA ALESSANDRIA VARALLO GRIGNASCO BIELLA PRATO S. . AD-Merkblatt HP0 AD 2000A4 APPROVED All data, subject to technical changes, are only for information and not for official use. 19 4 3 1 13 14 18 5 6 2 20 8 15 7 .
Scheda di geografia – Il fiume e il lago 1 IL FIUME E IL LAGO 1. Osserva il disegno. Nella figura c’è un percorso che parte dalla montagna ma non è una strada! Descrivi cosa vedi e dove finisce il percorso secondo te.
La Donna del Lago In Focus Premiere: Teatro San Carlo, Naples, 1819 One of Rossini’s most beautiful and expressive scores, La Donna del Lago was a success in its premiere season and quickly travelled the world’s operatic stages. The extraordinary vocal writing—a rare combination of lyricism and virtuosity
Lago di Lecco e Como 21 26 Lago di Annecy 30 26-02 Ischia 76 26-02 Sardegna Baia Chia 103 AGOSTO 01-08 Puglia Villaggio le Dune 77 02-08 La Magia dei Fiordi 140 02-09 Costa Brava soggiorno mare 118 03-09 Trentino e Dolomiti 42 03-10 Reno e Mosella 132 03-10 Austria 138 07-16 Sicilia ferragosto 89 08-15 Polonia 144
director de El Eco dEl RoEdoR Tea STilTon aventurera y decidida, enviada eSPecial de El Eco dEl RoEdoR 032-LA ISLA DEL TESORO 001-006.indd 4 16/11/10 13:28:35 EL SECRETO DEL LAGO DESAPARECIDO T1_10139520 El secreto lago desaparecido.indd 5 30/03/16 17:24
Agile describes a set of guiding principles that uses iterative approach for software development Agile is a practice that helps continuous iteration of development and testing in the software development process. In this model, development and testing activities are concurrent, unlike the Waterfall model. This mindset allows more communication .
