Trail analysis under multicollinearity of Nopalea cochenillifera in annual and biennial harvest

Forage cactus is widely cultivated in environments with low rainfall, due to its adaptability to the climate and richness in water and carbohydrates for ruminants. Knowing the morphology of these plants and their responses to management practices through biometric data can be used to estimate their quality and productivity. Thus, the study aimed to evaluate, through trail analysis under multicollinearity, the direct and indirect effects of two basic variables of productive yield between the annual and biennial Palmepa-PB01 forage cactus variety variables. The experiment was carried out in a randomized block design, with two treatments: annual and biennial harvest frequency; and three replications each one. The plants were spaced 1.0 m x 0.50 m for cultivation, to achieve plant density of 20,000 plants ha -1 . For the establishment of the cactus, a cladode per pit were buried 2/3 of the seedling in the soil, in a vertical position during the final third of the dry period. In 4 plants selected in the plot, their height (PH), plant width (PW) and cladode area index (CAI) were measured. The analysis of the trail under multicollinearity was carried out and corrected with the value of K to 0.1002 between the variables response of green matter (GMY) and dry matter yield (DMY) and the explanatory variables: PH, PW and CAI. The coefficient of determination of the trail analysis model (R 2 ) explains for the annual and biennial harvest 96.40 and 92.25% of variation in basic variable GMY, 96.44 and 92.25% for SMP. The characters PH, PW of the plant and CAI alone do not explain the variations of GMY and DMY through the multicollinearity trail analysis test, and their relationships vary according to the annual and biennial harvest of the forage cactus.