Aims: To quantify the magnitude of the genotype x harvest cycle interaction (GxC) of sugarcane during three harvest cycles and to select superior clones for cultivation on the Coast of the Southern Forest of Pernambuco. Study Design: The experiment was conducted in a randomized complete block design. Place and Duration of Study: Evaluated during the 2011/2012, 2012/2013 and 2013/2014 harvest years in the agricultural area of the Cucaú Plant, located in the Municipality of Rio Formoso (8°39' 49" S and 35º09'31" W, altitude of 5 m), Microregion of the Southern Forest of Pernambuco. Methodology: 11 genotypes Republic of Brazil of the RB 2004 series and three RB cultivars were evaluated. Each plot was represented by five grooves of 8.0 m in length, spaced in 1.0 m, totaling 40 m². The crops were harvested 15 months after planting (MAP) for the first crop cycle and 12 MAP during the two subsequent cycles were evaluated tons of sugarcane per hectare (TCH), tons of pol per hectare (TPH) and total recoverable sugar (ATR). The variance analyses, the Scott and Knott clustering test, the estimative of the simple and complex parts of the G x C interaction and the Pearson correlation coefficient were processed in the Genes program. Results: The genotypes showed a significant reduction of TCH from the first to the second cycle and that only the genotype UFRPE11 showed a significant decrease for the third. The genotypes UFRPE10, UFRPE6, UFRPE11, UFRPE7, UFRPE2, UFRPE9 and UFRPE1 exceeded all commercial varieties in TPH. It was observed for the variable ATR that there were no significant differences between the genotypes in the third cycle. The simple fraction of the interaction G x C were predominant between cycles C1 and C2 for TCH (67.91%) and TPH (69.35%), while for ATR (56.42%) the complex fraction was predominant. For the pair C2 x C3, the simple fraction of the interaction G x C predominated only in the TCH (62.85%) and TPH (62.41%) variables, but was not significant for the variable ATR. It is worth mentioning that the C1 x C3 cycle pair presented predominantly complex type interactions for all variables TCH (50.42%), TPH (52.20%) and ATR (59.66%). Conclusion: The simple fraction of the genotype x harvest cycles (G x C) interaction provides genetic gain for yield of sugarcane and sugar in selection in subsequent pairs of harvest cycles, year by year. The complex fraction of G x C interaction reduces the predictability of genetic gain, making it difficult to select new cultivars. Local selection favors expressive genetic gain in a few selection cycles. However, it does not favor the selection of genotypes with high adaptability and phenotypic stability, requiring tests in several environments. The UFRPE06 and UFRPE10 clones can be selected to continue the selection cycles for the southern coastal conditions of the Mata de Pernambuco.

The objective of this research was to evaluate the correlations between variables related to the biological nitrogen fixation (BNF) in segregating generations of cowpea and to unfold these correlations in direct and indirect effects, through path analysis. An outdoor bench experiment was conducted at Carpina Experimental Sugarcane Station of, located at the Zona da Mata region of the State of the Pernambuco, Northeast of Brazil (Federal Rural University of Pernambuco), between March and April 2016. The seeds were planted in 20 cm × 30 cm polyethylene bags, using a substrate composed of a mixture of vermiculite and sand washed in a ratio of 1:1. Bradyrhizobium references, recommended for culture, were used as a mixture of two strains. Parental and F2, F3 and F4 generations were evaluated in a randomized block design with four replicates. Data collection was performed 45 days after the emergency (DAE). Phenotypic correlations and path analysis of the number of nodules per plant (NN), nodules dry mass (NDM), dry roots mass (DRM), dry mass of aerial part (DMAP), nodulation efficiency (NODE) and nitrogen accumulated in the aerial part (NAAP). The phenotypic correlations between the variables related to the BNF showed high magnitudes, demonstrating that there is a great influence of each of the variables on the others, furthermore the path analysis of the coefficients indicated that all the primary components (NN, NDM, DRM, DMAP and NODE) must be considered when it is desired to increase the NAAP in segregating generations of cowpea.