Banana is one of the most abundantly consumed fruits in the world for its contribution to human nutrition, including an important content of potassium.  Given the social contribution of this crop in the cropping areas of Colombia, it is necessary to develop agronomic management strategies that contribute to increasing productivity and offering alternatives that regulate the effects of climate seasonality on banana production in the Urabá region of Colombia.  For this purpose, different planting densities (2,000, 2,500, 3,000 and 3,500 plants ha-1) and planting arrangements (triangle, rectangle and double furrow), plus a relative control (1,700 plants ha-1 under triangle arrangement), were evaluated in the banana cv. Williams.  The experimental design corresponded to randomized complete blocks in a 4 x 3 + 1 (densities x arrangements + control treatment at 1,700 plants ha-1) factorial arrangement.  During harvest, the following parameters were monitored:  bunch weight, number of hands and fingers per bunch, finger length and diameter, bunch ratio and percentage of loss per bunch.  Two analyses of variance were carried out, a general one including the control treatment, and an additional one excluding it. The double furrow planting arrangement favored the productivity of the crop in terms of a greater assimilation and lesser loss records at the bunch level.  The strong correlation of the studied variables with the double furrow arrangement became evident.

Banana is after apple, the most consumed fruit in the world for its nutritional contribution. Given its importance, it is necessary to advance in the development of agronomic management strategies that contribute to increase productivity and offer alternatives to regulate the consequences derived from the seasonality of production in the Urabá area (Colombia). For this, the clone Williams (Musa AAA Simmonds), at different planting densities (2000, 2500, 3000 and 3500 plants ha-1), different population arrangements (triangle, rectangle and double furrow), compared with a relative control (1700 plants ha-1 in triangle) was evaluated. The design used was randomized complete blocks in factorial arrangement of 4 x 3 + 1 (densities x arrays + control). The variables height and perimeter of the pseudostem, number of leaves from planting to harvest, leaf area index (IAF), light extinction coefficient, number of suckers and developmental stages (days elapsed after flowering and harvest) were determined. An analysis of general variance was carried out and another one excluding the control. The Tukey test (P <0.05) was applied to the qualitative variables and contrasts of orthogonal polynomials were made to test the linear and quadratic effects of the population density factor. The results indicated that the high planting densities and the population arrangements had a negative effect on the perimeter of the pseudostem and the height of the main sucker, favored the LAI, influenced the extinction coefficient, but did not affect the development of the plants.