EMBOLISM INCREASE AND ANATOMICAL MODIFICATIONS CAUSED BY A PARASITIC PLANT: PHORADENDRON CRASSIFOLIUM (SANTALACEAE) ON TAPIRIRA GUIANENSIS (ANACARDIACEAE)

Abstract

Parasitic plants are capable of causing a variety of effects to their hosts, including alterations in the process of wood formation. However, the majority of studies dealing with parasitic plant anatomy have focused on the host–parasite interface and the direct action of the haustorium, which is the organ responsible for attaching the parasite to the host. Considering this gap, we studied the anatomical and functional effects caused by a mistletoe species, Phoradendron crassifolium (Santalaceae), on the wood anatomy of the host tree Tapirira guianensis (Anacardiaceae). Both parasitized and non-parasitized branches were collected from host trees. Traditional wood anatomy procedures were employed, along with functionality experiments using the ascent of safranin solution through the xylem. Prior to the analysis, all sampled branches were divided in “upstream” and “downstream” portions, considering the direction of xylem sap flow inside the plant body. This design was chosen in order to avoid biased results derived from normal ontogeny-related wood anatomical and functional changes. Our results showed that infested wood expressed a higher density of embolized vessels, narrower vessel lumen diameter, higher vessel density, taller and wider rays, and fibers with thinner cell walls. All these responses were most conspicuous in the downstream sections of the parasitized branches. We propose that the wood anatomical and functional alterations were induced by the combination of water stress caused by water use by the parasite and consequent low turgor in differentiating cambial derivates; by unbalanced auxin/cytokinin concentrations originating at the infestation region due to phloem disruptions caused by the parasite’s penetration and action; and by higher than usual ethylene levels. Further analysis of hydraulic conductivity and hormonal changes in host branches are necessary to test this hypothesis.