This thesis presents an investigation aimed at the understanding of the effect of interfacial energy anisotropy on growth morphologies in Al-Zn alloys. Most low alloyed metals form dendritic structures that grow along directions that correspond to low index crystal axes, e.g., 〈100〉 directions in fcc aluminum. However, recent findings [Gon06, Gon08] have shown that an increase in the zinc content of Al-Zn continuously changes the dendrite growth direction from 〈100〉 to 〈110〉 in the (100) plane. At intermediate compositions, 〈320〉 dendrites and textured seaweeds were even observed. The reason for this dendrite orientation transition (DOT) is that this system exhibits a large solubility of the element zinc, which forms hexagonal crystals, in the primary fcc aluminum phase, thus modifying its weak solid-liquid interfacial energy anisotropy γsl. Due to the complexity of the phenomenology, there is still no satisfactory theory that predicts all the observed microstructures. The goal of this work was to measure the anisotropy of γsl, to gain further insight in the dendrite growth morphologies and mechanisms in Al-Zn via X-ray tomography and investigate the relationship between dendrite growth directions and the anisotropy of γsl via phase-field modeling. The values of the anisotropy in Al-82 wt.% Zn, obtained by equilibrium shape measurement, suggests a DOT in a regime with very low anisotropy, favoring the formation of seaweed structures. The enhanced experimental procedures and observation methods allowed an unprecedented characterization of such complex morphologies. Observations showed that seaweeds are far from random structures. They grow in a (100) symmetry plane by an alternating growth direction mechanism and are therefore less hierarchically ordered than common dendrites. Phase-field simulations of equiaxed growth of a dilute alloy showed that the DOT is not a continuous transition of growth directions, but rather a competition between the 〈100〉 and the 〈110〉 character of dentrites. However, crystallographic orientation measurements showed that the macroscopic texture evolves as evidenced by Gonzales and Rappaz. This showed that the preferred growth direction is not the same as the final texture of a sample, which is uncommon. Simulations of directional growth showed that seaweeds, which do not appear in equiaxed growth, are in fact produced by the interplay of low or degenerate crystalline anisotropy and the misorientation of the crystal with respect to the thermal gradient.