The objective of this work was to build weighted correlation networks, in order to discover correlation structures and link patterns among 28 morphoagronomic traits of chili pepper related to seedling, plant, inflorescence, and fruit. Phenotypic and genotypic information of 16 Capsicum genotypes were analyzed. Correlation structures and link patterns can be easily identified in the matrices using the Fruchterman-Reingold algorithm with correlation network information. Both types of correlations showed the same general link pattern among fruit traits, with high broad-sense heritability values and high aptitude of the genotypes for agronomic and ornamental breeding. Leaf dimensions are correlated with a cluster of fruit traits. Correlation networks of chili pepper traits may increase the effectiveness of genotype selection, since both correlated traits and groups can be identified.

Umbu fruits from four municipalities of Paraíba were morphologically characterized in order to identify promising genotypes for fresh consumption and agro-industrial use. We used 400 fruits from parent plants located in São Vicente do Seridó, Queimadas, Boqueirão, and Cabaceiras and 40 commercial fruits from two open fairs in Campina Grande, PB. The experimental unit consisted of ten fruits. The fruits were analyzed for total mass (MTF), length (LC), diameter (DF), fresh matter (MMF), yield (RF), and presence of hairs. In the endocarps, mass (ME), length (CE), diameter (DE), and width (LE) were the evaluated parameters. A completely randomized design was adopted, the data were subjected to variance analysis, and the means were compared by the Tukey test and principal component analysis. The fruits from Queimadas and São Vicente do Seridó showed the best morphological characters with regard to mass, length, diameter, fresh matter, and fruit yield, and were the most promising for agro-industrial commercialization. The fruits from Boqueirão showed the highest endocarp mass. Most fruits showed hairs on the peel, except the commercial ones. Four groups of municipalities were formed, with the fruits from Queimadas showing the largest size, while the fruits from São Vicente de Seridó and Soledade showed the highest yield.

The leaf area is a parameter of fundamental importance in studies on plant growth and physiology. The objective of this study was to build allometric equations for the accurate and fast estimation of sapodilla leaf areas. In total, 250 leaves of different shapes and sizes were collected from sapodilla matrices trees growing at the Universidade Federal Rural do Semi-Árido, Mossoró-RN, Brazil. For each leaf, the length, width, product of length and width (LW), product of length and length, product of width and width, and leaf area were measured. Linear and nonlinear models were used to construct the allometric equations. The best equations were chosen on the basis of the following criteria: the highest coefficient of determination, Pearson’s linear correlation coefficient, and Willmott’s index of agreement; and the lowest Akaike information criterion and root mean square error. It was verified that the models that used the LW value presented the best criteria for estimating the leaf area. Specifically, the equations ŷ = 0.664 × LW1.018 and ŷ = 0.713 × LW, which use LW values, are the most suitable for estimating the leaf area of sapodilla quickly and accurately.

Peppers belong to the family Solanaceae and can be used for several different purposes. The use of morphological descriptors associated with multivariate techniques allows identifying pepper progenies that can be recommended for ornamental purposes. This study aimed to select segregating Capsicum annum progenies based on multivariate analyses. The study was conducted in a plant nursery at the Biotechnology and Plant Breeding Sector of the Center of Agricultural Sciences of the Federal University of Paraíba (CCA/UFPB). Twenty F3 progenies of ornamental peppers were used, namely: UFPB 5, 8, 10, 15, 18, 22, 25, 32, 33, 36, 40, 42, 49, 56, 60, 62, 64, 69, 75, and 77, plus four additional controls consisting of the parents UFPB 77.3 and UFPB137 and two commercial cultivars, Etna and Pirâmide. Seventeen plant and fruit traits were evaluated. The experimental design was completely randomized. The data were subjected to a multivariate analysis of variance. All analyses were performed with the software R, version 3.0.3. The results of the multivariate analysis of variance highlighted significant differences between the populations of ornamental peppers. Singh’s (1981) method made it possible to determine that eleven of the seventeen traits studied contributed with 73.7% of the genetic divergence, whereas six traits contributed with only 21.2%. There is genetic diversity among the evaluated progenies. Progenies 5, 10, 22, 33, 49, and 69 are recommended for selection for having ornamental traits for in-pot cultivation. Selection within these progenies is recommended to continue the breeding program of ornamental peppers for in-pot cultivation.