Genetic distances of inbred lines and correlation with single cross performance in maize, using AFLP and SSR molecular markers

Dois grupos de linhagens de milho tropical, um derivado da população BR105 e outro da população BR106, foram avaliados com os marcadores AFLP, para estimar as distâncias genéticas entre as linhagens, visando a alocação destas a grupos heteróticos e as relações entre as distâncias genéticas e as performances de híbridos simples para produção de grãos. As distâncias genéticas (DG) entre todos os pares de linhagens foram determinadas utilizando o coeficiente de Jaccard (dados de AFLP). As análises de agrupamento foram feitas utilizando o método UPGMA (unweighted pair-group method with arithmetical averages). As correlações entre as distâncias genéticas e as performances dos híbridos F1 e heteroses para os 148 híbridos simples foram estimadas através do coeficiente de correlação de Pearson. Cinco cruzamentos foram excluídos dessa análise: quatro correspondentes a parcelas perdidas e um correspondente a linhagens irmãs (todos da BR106). Os híbridos simples intra e interpopulacionais foram obtidos segundo um esquema dialélico e avaliados experimentalmente em delineamentos em látice em três ambientes. As distâncias genéticas foram em média maiores para os cruzamentos interpopulacionais. As análises de agrupamento com AFLP concordaram em grande parte com as informações das genealogias das linhagens. A linhagem 16, oriunda da população BR106, foi alocada ao grupo da BR105, concordando com os dados de produção de grãos dos híbridos intra e interpopulacionais. Os coeficientes de correlação entre as distâncias genéticas e as performances dos híbridos F1 e heteroses foram altos para os cruzamentos intrapopulacionais da BR106 (0,91), médios para os cruzamentos intrapopulacionais da BR105 (0,52) e baixos para os cruzamentos interpopulacionais (0,29). A menor correlação observada nos híbridos interpopulacionais foi devida à menor variação para DG. Os resultados mostraram que o marcador AFLP é eficiente para alocar linhagens de milho a grupos heteróticos e que as distâncias genéticas obtidas com AFLP podem ser úteis na predição das performances de híbridos simples, reduzindo a quantidade de trabalho e custos relacionadas com a avaliação experimental de híbridos.Two set of tropical inbred lines, one derived from BR105 population and other derived from BR106 population, were assayed for AFLP and for microsatellites (SSR), in order to investigate genetic distances among lines and their relationship to heterotic group allocation and single cross yield performance. Genetic distances for all possible pairs of inbreds were determined using Jaccard's similarity coefficient (AFLP data) and modified Roger's genetic distance (SSR data). Cluster analyses were performed using UPGMA (unweighted pair-group method with arithmetical averages). Pearson correlation coefficients were estimated for genetic distances (GD) with F1 grain yield and heterosis for the 148 single crosses derived from intra- and interpopulation crosses. Five crosses were excluded from this analysis: four corresponded to missing plots and one corresponded to sister lines (from BR106). The single crosses and lines were crossed following a diallel scheme at both intra- and interpopulation levels and evaluated in lattice designs across three environments. Genetic distances (GD) was on the average greater for inter population than intrapopulation crosses for both, AFLP and SSR. Cluster analysis from AFLP was almost in accordance with pedigree information. The inbred line 16, derived from BR106, was assigned to BR105 set. The assignment of inbred 16 to BR105 set is in accordance with single cross yield performance and heterosis from intra- and interpopulation crosses. However, the same pattern was not observed for SSR, where 3 lines from BR106 (9, 10 and 16) were assigned to BR105 set. One possible reason for this result is the low number of primers assayed (68). Correlation coefficients of genetic distances with F1 grain yield and heterosis were high for BR106 x BR106 crosses (0.91 and 0.82 for AFLP and SSR, respectively), medium for BR 105 x BR 105 crosses (0.52 and 0.52 for AFLP and SSR, respectively) and low for BR 105 x BR 106 crosses (0.29 and 0.16 for AFLP and SSR, respectively). The reason for the lower correlation at interpopulation level was the smaller variation for GD. General results showed that AFLP molecular marker is efficient in allocating maize lines to heterotic groups and that AFLP-based GD are suitable for predicting the performance of maize single crosses in order to reduce the costs related with experimental evaluation of plots. The same general pattern was observed for SSR molecular marker. However, more information is needed on SSR molecular marker, by increasing the number of primers