Evaluating large numbers of genotypes and phenotypes in multi-environment trials is key to crop improvement for biomass performance in sorghum. In this dissertation, we developed an approach that integrates crop growth models with remote-sensing data and genetic information for modeling and predicting sorghum biomass yield. The goal of studies described in Chapter 2 was to parameterize the Agricultural Production Systems sIMulator (APSIM) crop growth models with remote-sensing and ground-reference data to predict variation in phenology and yield-related traits for 18 commercial grain and biomass sorghum hybrids. These studies showed that (i) biomass sorghum hybrids tended to have higher maximum plant height, final dry biomass and radiation use efficiency (RUE) than grain sorghum, (ii) photoperiod-sensitive sorghum hybrids exhibited greater biomass potential in longer growing environments and (iii) the parameterized APSIM models performed well in above-ground biomass simulations across years and locations. Crop growth models that integrate remote-sensing data offer an efficient approach to parameterize models for larger plant breeding populations. Understanding the genetic architecture of biomass productivity and bioenergy-related traits is another key aspect of bioenergy sorghum breeding programs. In Chapter 3, 619 sorghum genotypes from the sorghum diversity panel were individually crossed to ATx623 to create a half-sib population that was planted and evaluated in field trials in three consecutive years. Single-nucleotide polymorphisms (SNPs) were used in a genome-wide association study (GWAS) to identify genetic loci associated with variation in plant architecture and biomass productivity. A few SNPs associated with these traits were located in previously described genes including the sorghum dwarfing genes <i>Dw1</i> and <i>Dw3</i> and stay-green QTLs <i>Stg1</i> and <i>Stg4</i>. Of particular interest were seven genetic loci that were discovered for biomass yield. For three of these loci, the minor or uncommon allele exhibited a favorable effect on productivity suggesting opportunities to further improve the crop for biomass accumulation through plant breeding. Marker-assisted and genomic selection strategies may provide tools to introgress and exploit these genes for bioenergy sorghum development. Since parameterizing biophysical crop models requires extensive time and manual effort, a simple model was developed in Chapter 4 that used time-dependent measurements of RGB canopy cover and daily radiation coupled with end-of-season biomass for estimating seasonal radiation use efficiency (SRUE) in 619 sorghum hybrids. SRUE was shown to be a stable and heritable trait that has a positive relationship with aboveground dry biomass (ADB) over seasons. GWAS identified 11 SNPs associated with SRUE with the favorable effect represented by the minor allele for seven of these SNPs. Increasing the frequency of these favorable alleles may improve the breeding population. These results demonstrated that the simple model for calculating SRUE can be used in genetic studies and for parameterizing biophysical crop models. The studies integrating crop growth models with remote sensing technologies provide an opportunity to evaluate a large number of phenotypes for the target population to understand the underlying genetic variation of bioenergy sorghum.

Submitted by Roberta Lorenzon (roberta.lorenzon@fgv.br) on 2011-06-01T14:19:41Z No. of bitstreams: 2 65080100022.pdf: 2118563 bytes, checksum: e772406b416ab303d53d7648ac10d40b (MD5) 65080100022.pdf: 2118563 bytes, checksum: e772406b416ab303d53d7648ac10d40b (MD5) / Approved for entry into archive by Suzinei Teles Garcia Garcia(suzinei.garcia@fgv.br) on 2011-06-01T14:54:51Z (GMT) No. of bitstreams: 2 65080100022.pdf: 2118563 bytes, checksum: e772406b416ab303d53d7648ac10d40b (MD5) 65080100022.pdf: 2118563 bytes, checksum: e772406b416ab303d53d7648ac10d40b (MD5) / Approved for entry into archive by Suzinei Teles Garcia Garcia(suzinei.garcia@fgv.br) on 2011-06-01T14:55:48Z (GMT) No. of bitstreams: 2 65080100022.pdf: 2118563 bytes, checksum: e772406b416ab303d53d7648ac10d40b (MD5) 65080100022.pdf: 2118563 bytes, checksum: e772406b416ab303d53d7648ac10d40b (MD5) / Made available in DSpace on 2011-06-01T15:02:41Z (GMT). No. of bitstreams: 2 65080100022.pdf: 2118563 bytes, checksum: e772406b416ab303d53d7648ac10d40b (MD5) 65080100022.pdf: 2118563 bytes, checksum: e772406b416ab303d53d7648ac10d40b (MD5) Previous issue date: 2010-11-12 / Brazil is the second major soybean [Glycine max (L.) Merr.] producer and the seventh one on soybean oil. Brazilian production reached 61 million tons at 2008 and the forecast to 2020 is 105 million tons. Biodiesel consumption at 2008 was one million tons and the demand for this biofuel will reach 3,1 million tons at 2020. To amount this demand, the planting area on centerwest region of Brazil will increase, but also efforts on productivity must be required. Looking for a better knowledge on the climate variables temperature and global radiation over soybean development, yield and oil productivity was purposed a stochastic model with truncated normal distribution for maximum, minimum and average temperature data. Included in this model, a triangular asymmetric distribution to determine the probable oil productivity. Eight sowing dates were stipulated on Piracicaba/SP where the climate data was given from ESALQ/USP agrometeorologic station. The conclusions were: (i) there were decreases on soybean cycle duration with the average temperature increase; (ii) the soybean cycle decrease restricted soybean yield and oil productivity; (iii) the global radiation thirty days after antesis reflected on photo assimilates partition and soybean yield and oil productivity; (iv) stochastic models can be used for soybean yield and oil productivity forecast. / O Brasil é o segundo produtor mundial de soja [Glycine max (L.) Merr.] e o sétimo de óleo vegetal. A produção brasileira desta oleaginosa alcançou 61 milhões de toneladas na safra 2007/08 e projeta-se, para 2020, produção de 105 milhões de toneladas. O consumo de biodiesel em 2008 representou um milhão de toneladas e a demanda por este biocombustível deverá atingir 3,1 milhões de toneladas em 2020. Para atender esta demanda haverá ampliação da área plantada principalmente na região Centro-Oeste, mas também exigirá esforços no aumento de produtividade. Visando melhor conhecimento das inferências das variáveis climáticas temperatura e radiação global sobre o desenvolvimento da soja e sua produtividade de grãos e óleo, foi proposto um modelo estocástico com distribuição normal truncada para os dados de temperatura máxima, mínima e média. Também foi incluído neste modelo distribuição triangular assimétrica para determinação da produtividade de óleo mais provável. Foram estipuladas oito datas de semeadura para a localidade de Piracicaba/SP onde está localizada a estação meteorológica da ESALQ/USP, fornecedora dos dados climáticos utilizados neste estudo. Conclui-se que: (i) ao longo das datas de semeadura houve redução do ciclo com o aumento da temperatura média; (ii) a redução do ciclo da cultura de soja interferiu nas produtividades de grãos e de óleo; (iii) a radiação global média nos trinta dias após a antese refletiram-se na partição de fotoassimilados e na produtividade de grãos e óleo; (iv) os modelos estocásticos podem ser utilizados para a previsão das produtividades de soja e óleo.