Root productivity of deciduous and evergreen species identified using a molecular approach

The linkage between leaf traits and root structure may explain how plants integrate above and belowground traits into whole plant adaptations to environmental stresses. In dry seasonal forests, the lack of dry season precipitation dries out the relatively nutrient-rich shallow soil, leaving shallow soil water and nutrients inaccessible to uptake until the wet season. In tropical or subtropical seasonal dry forests, deciduousness may allow for the survival of shallow fine roots during the dry season. Losing leaves during the dry season reduces aboveground plant water demand, and a greater proportion of water extracted from deep soil can be used to maintain shallow roots until the wet season. Higher shallow root survival through the dry season than evergreen species means that deciduous species can take advantage of the nutrient pulse associated with the onset of the wet season. To test the above hypothesis, fine roots were collected from soil cores in a seasonally dry forest during the dry season, onset of the wet season, and the wet season and were identified to selected evergreen and deciduous study species. The fine roots of two of the selected species (Lyonia ferruginea and Carya floridana) could be identified from visual characteristics. The other three study species, which were all from the genus Quercus (Q. geminata, Q. myrtifolia, and Q. laevis), were impossible to separate visually. We developed a PCR-based restriction fragment length polymorphism (PCR-RFLP) technique, which provided a quick, simple, low-cost way to identify the species of all fine roots of our study species. We extracted DNA from all roots that were not visually identified, amplified the internal transcribed spacer region (ITS), digested the ITS region with the restriction enzyme TaqαI, and used gel electrophoresis to separate DNA fragments. Using a PCR-RFLP based root identification key that we developed for the species at Archbold Biological Station, all species that could not be identified visually were separated based on each species ' unique banding pattern of restriction fragments. Approximately 2,500 roots were identified using PCR-RFLP and approximately 1,500 more roots were identified visually. Identifying fine roots to species allows for species-level analysis of root productivity in this in situ study.