During the post-harvesting process, coffee berries are dried and separated into green commercial beans and husks. The dynamics of dry matter (DM) accumulation in the berry components along the maturation process are important for the definition of the most adequate moment for the harvest, which is genotype-dependent. The DM accumulation dynamics in the berries, beans, and husks in six Coffea canephora genotypes were studied during the fruit maturation process, with the aim of identifying the fruit harvesting stage when the highest bean yield can be obtained. Berry samples were collected every two weeks at nine maturation stages starting from 33 weeks after flowering (green berry stage). Second-order polynomial regressions were used to analyse berry and bean DM accumulation over time, while temporal husk DM accumulation was compared using ANOVA and the Tukey test. DM accumulation was the highest in the berries and beans following the initial sampling, while the highest husk DM accumulation occurred at the final stages of maturation. In general, DM accumulation of all components increased as fruit maturation progressed, attaining the highest DM values in the final stages of red berries, but occurred earlier for early/medium and medium maturation cycle genotypes. The Beira Rio 8 genotype showed the highest DM accumulation in all components. Bamburral and P1 genotypes showed the lowest berry fresh mass (FM) to bean DM ratios. The A1 genotype showed the greatest berry FM to bean DM ratio, being a genotype with the lowest DM and bean mass performances and bean yield. Our data revealed that not only should the absolute berry and bean yield be considered for highly productive genotype selection but also the bean DM dynamics in the characterization of commercial coffee yield.

Nonlinear common factor models with polynomial regression functions, including interaction terms, are fitted by simultaneously estimating the factor loadings and common factor scores, using maximum-likelihood-ratio and ordinary-least-squares methods. A Monte Carlo study gives support to a conjecture about the form of the distribution of the likelihood-ratio criterion.