TY - THES T1 - Genotypic responses of upland rice to an altitudinal gradient A1 - Shrestha,Suchit Prasad Y1 - 2012/07/03 N2 - Adaptation strategies are required for crops to cope with changing climate. The impact of climate change on crop production is not straight forward to predict as extreme events comprise multiple combination of abiotic stresses and their impact differs in crop physiological growth stages. The mechanism on how new abiotic stress combinations translate into phenology and yield, and which cultivars are better adapted is yet unclear. Crop growth models are available that have been parameterized and validated for some aspects of possible climate change scenarios but in view of complex interactions crop responses to climate change are difficult to predict. On the other hand, prediction of the complex ideotype trait combinations may be interesting for breeders but physiological models are required that are well validated for the target environments. In upland rice grown under rainfed conditions without surface water accumulation methane emission is negligible and therefore greenhouse gas emission much lower compared to irrigated paddy rice systems. In addition, growing demand for rice and the increasing pressure on irrigated land leads to development of upland rice areas to supplement irrigated rice. Therefore, this study investigates genetically diverse upland rice genotypes from a wide range of origins across altitudinal gradient locations. The main objective of this study is to investigate genotypic responses of upland rice to different environments in order to calibrate crop growth models, which allow the evaluation of effects of climate change on upland rice systems. Multi-locational field (three locations: 1625, 965 and 25 m asl) trials comprising non-replicated phenological plots with five sowing dates (monthly staggered) in two consecutive years creating thirty different environments, and replicated physiological yield trials with two sowing dates (monthly staggered; early and late sowing) in two consecutive years creating twelve different environments were established in Madagascar. Ten contrasting upland rice genotypes were included in both field trials. Meteorological data were recorded on a daily basis during trial periods. Developmental stages were observed in the phenological plots; in the physiological plots yield and yield components were recorded. In addition, greenhouse trials were conducted with one upland rice genotype subjected to seven N-supply levels in a hydroponic system at the University of Hohenheim in order to understand the relationship between chlorophyll index, photochemical reflectance index and chlorophyll fluorescence parameters. Various statistical tools were applied to analyse field and greenhouse data sets. The phenological trial showed that duration to flowering was 117, 81 and 67 d in high (HA), mid (MA) and low (LA) altitudinal locations respectively. 90% of the total variance was explained by location when pooled over genotype, location, sowing dates and year. In HA, factors such as genotype, sowing date and year equally contributed to the observed variability whereas in MA year was the most determining factor and genotype had no significant contribution. Similarly, in LA sowing date was the main influencing factor and year had no significant effect. Aggregated data over locations, sowing dates and years indicated that each degree Celsius rise in mean air temperature decreased crop duration by 5 to 9 days depending upon genotype. Basic genotypic thermal constants Tbase ranged from 9.8 to 13.9 °C and Tsum from 816 to 1220 °C d within the selected genotypes. Cold tolerant genotypes were less affected by lower Tmin (14 °C) at booting to heading stage regarding spikelet sterility in HA, whereas others were highly affected at 15 °C (cold stress). Similarly, both cold sensitive and tolerant genotypes were affected by Tmax (above 30 °C) at flowering in MA and LA locations (heat stress). Grain yield and yield components were highly affected by location, year, sowing date, and genotypes and the interactions between these yield-determining factors were obvious. In HA, early sown cold tolerant genotypes had more than 5 t ha-1 grain yield and one month delay in sowing led to highly reduced yield whereas other genotypes had very poor yield on both sowing dates due to cold stress. In MA, yield difference between sowing date and genotypes was small (4.3 - 4.9 t ha-1). Grain yield in LA was vulnerable due to frequent tropical storms. Yield stability analysis showed that cold tolerant genotypes had above average stability. AMMI model for grain yield showed that environment and genotype by environment interactions were highly significant. Yield components determined during specific development stages of the genotype such as tillers per hill and percentage of filled spikelets were mainly influenced by environment, spikelets per panicle and thousand grain weight were influenced by genotype, and percentage of productive tillers was equally influenced by both genotype and environment. PCA biplots showed that all HA environments were equally influenced by all weather parameters with minimum air temperature having the strongest positive influence on genotypic performance. In all MA environments genotypic performance in all phenophases was strongly and positively influenced by rainfall, and strongly and negatively influenced by vapour pressure deficit, solar radiation and potential evapotranspiration. In the LA environments, main weather parameters influencing genotypic performance were maximum temperature and high rainfall accompanied by strong winds. The field measured SPAD values of the upper canopy leaves reflected the location specific N-remobilization and leaf senescence levels after flowering. Similarly, PRI values showed the abiotic stress responses among development stages and locations along the altitudinal gradient. These readings showed that genotypes were efficient in radiation use and N-remobilization after flowering in MA. The unsynchronized relationship between source (leaf) and sink (grain) explained the yield penalty. Emphasis on identification of morpho-physiological traits contributing to cold tolerance should be placed for further breeding. We conclude that genotypic responses of upland rice cultivars differed across altitudinal gradients. Genotypes that are well adapted in HA can easily be adapted in MA without yield decrease. But genotypes well adapted in MA may show a huge yield penalty in HA due to lower temperature during reproductive phase and consequently reduced sink formation. Frequent tropical storms and high temperature reduced yield potential in LA. Therefore, HA has a large potential for the future food security considering climate change scenarios. At present, MA is favorable for upland rice production systems, whereas LA is highly vulnerable and is expected to be even more vulnerable in future. Those results on genotype-specific responses to environmental conditions allow further improvement of crop models such as RIDEV and SAMARA (synthesis of SARRAH and EcoMeristem), which can be used to test a number of phenotypic traits x environments combinations to define ideotypes of upland rice varieties adapted to changing climate and cropping calendars. Genotypic responses of phyllochron, biomass production and crop growth rate, and radiation use efficiency across altitudinal gradients will be included to parameterize these models. In this regard, collaborations with AfricaRice, CIRAD and IRRI are ongoing. KW - Genotyp KW - Umwelt KW - Reis KW - Stabilität KW - Interaktion CY - Hohenheim PB - Kommunikations-, Informations- und Medienzentrum der Universität Hohenheim AD - Garbenstr. 15, 70593 Stuttgart UR - http://opus.uni-hohenheim.de/volltexte/2012/727 ER -