Barley breeding program in Brazil has focused on characteristics associated with malting for beer purposes as the main economic application for this crop. The breeding process focused on selection for grain yield, disease resistance and malting quality. The objective of this work was to quantify the genetic gain in barley grain yield from 1968 and 2008 in Brazil and to identify the physiological characteristics associated with the increase of grain yield. Field experiments with five 2-row barley cultivars were tested from 2011 to 2013 in the absence of biotic and abiotic stresses and with mechanical restriction to lodging. The ANOVA showed no genetic gain until 1980 with average grain yield of 4.632 kg/ha. After 1980, there was a productivity increase of 59.9 kg/ha/year. No correlation was observed between total maturity biomass and the year of release of the cultivars, while harvest index and plant height, were significantly improved. The main component associated with grain yield was the number of grains/m 2 , due to the higher number of spikes/m 2 associated to a greater contribution of the tillers in the modern cultivars.

A longer stem elongation phase (from the terminal spikelet-TS to the anthesis-ANT phases) increases grain production due to an increase in spikelet fertility. However, the mechanism behind the greater number of fertile flowers that occur when the duration of stem elongation is modified by photoperiod and vernalization is not fully understood. The goal of this study was to investigate the effect of combinations of photoperiod and vernalization on the duration of stem elongation and spikelet fertility in wheat. Thus, a greenhouse experiment was performed by subjecting 13 wheat genotypes to two vernalization regimes (V0-non-vernalized plants and V40-plants with 40 vernalization days) and to two photoperiod regimes (NP-natural and NP + 6-photoperiod extended by six hours), during the stem elongation phase. The natural photoperiod increased the spikelet fertility of eight cultivars without a corresponding increased duration of the TS-ANT phase, suggesting the existence of a direct effect of photoperiod on increased spikelet fertility. Vernalization increased the duration of the TS-ANT phase, without influencing spikelet fertility. There was genetic variability in the responses to photoperiod and vernalization.

Despite recognizing the importance of genetic improvement in the production of barley grains, little information is available on the contribution of phenological development to the genetic improvement of Brazilian barley. Field experiments were carried out between 2011 to 2013, in the absence of biotic and abiotic stresses and with preventive lodging control. Five two-rowed spring barley cultivars, released between 1968 and 2008, were evaluated. Although there was no significant association in the cycle length (Emergence - Anthesis) of the cultivars with the year of release, the genetic improvement increased the proportion of the Doble ridge - Maximum number of spikelet primordia/Maximum number of spikelet primordia - anthesis period to the total time to anthesis. The period between DR-MNP was increased in modern cultivars, to the detriment of the Doble ridge - Maximum number of spikelet primordia period. However, the duration of the period between emergences to the double ridge (vegetative phase) was not altered in the analyzed period. Barley breeding in Brazil did not change the total number of leaves on the main stem but caused an increase in the number of leaves earlier in the development, favoring the high level of tillering. The leaf architecture of modern barley was altered towards a more vertical inclination (erectophilic canopy), allowing the penetration of photosynthetically active radiation into the crop canopy.

Uma obra acabada é o resultado de um esforço coletivo. Esta tese não é exceção. Várias pessoas contribuíram de uma forma ou de outra para a realização deste trabalho, seja prestando apoio técnico, material, estrutural ou mesmo moral e aos quais eu deixo meu agradecimento: Ao Instituto de Química da UNICAMP, pela excelente infra-estrutura, aos seus funcionários e professores.Ao meu orientador, prof. Carlos Roque Duarte Correia, pelo exemplo de dedicação à pesquisa, pelo incentivo e pelo apoio intelectual. Aos professores LuizCarlos Dias, Anita Marsaioli e Lúcia Baptistella pelas valiosas sugestões apresentadas durante o exame de qualificação. E, aos professores Ronaldo Pilli e Fernando Coelho sempre prestativos, colaborando com sugestões, empréstimos de reagentes e equipamentos. Aos professores Mara Braibante e Vítor Ferreira, que prontamente aceitaram o convite para compor a banca examinadora da defesa de tese. Aos órgãos financiadores: CAPES, pela bolsa concedida e ao CNPq e FAPESP, pelo financiamento do projeto.A EMBRAPA, em especial ao CNPT (Centro Nacional de Pesquisas do Trigo), que foi sensível ao meu pedido de licença para a conclusão deste trabalho. Ao Ângelo, por ter "segurado as pontas" durante minha licença para a conclusão deste trabalho. E, em extensão, aos colegas do setor de fisiologia vegetal do CNPT: Erivelton, Mauro, Osmar, Décio, Jorge e batatinha, pelo apoio e confiança recebidos.