To inform an individual-based forest stand model emphasizing belowground competition, we explored the potential of using the relative abundances of ribosomal PCR products from pooled and milled roots, to allocate total root biomass to each of three coniferous species and to non-coniferous understory species. The study stand in the Cascade Range in Oregon, USA, is dominated by 400 year-old Douglas-fir (Pseudotsuga menziensii), but there are much larger numbers of smaller Pacific silver fir (Abies amabilis) and western hemlock (Tsuga heterophylla) that mainly arrived with fire-suppression in the late 1800s. Previously we estimated the distribution of fine roots by tree species using single-root PCR-RFLP analysis on subsamples from each size class (<1mm, 1-2mm and >2mm) obtained from soil cores, but this approach is highly labor intensive. Here we applied a T-RFLP approach to DNA extracts from subsamples of pooled/milled roots. This approach has been well established in numerous studies of microbial community structure. Ribosomal internal transcribed spacer sequences were amplified and restricted for use in PCR-RFLP and T-RFLP (NSIP forward and 58CPR reverse ITS1 primers, labeled for T-RFLP with HEX and 6-FAM to be detected by Applied Biosystems 3100). Fluorescence was used as a proxy for relative abundance.
Results/Conclusions
The PCR product lengths for the understory plants varied around 500 bp, while the PCR products for the coniferous species were around 1500 bp in length. This large difference allowed for easy discrimination between the two groups so that on-gel quantification by fluorescence provided for initial partitioning of the total root biomass between coniferous and non-coniferous roots. In this study the roots were primarily coniferous and molecular identification showed all coniferous roots were Douglas-fir, Pacific silver fir or western hemlock. DNA yield from milled roots was very low on the >2mm size fraction, but good quality DNA was obtained from the two smaller size classes. The approach was calibrated with milled roots of known species, mixed at various proportions, to determine root-dry-weight to PCR-relative-abundance conversion factors by size class of each species. Samples prepared using equal dry-weight proportions of the three coniferous species, produced PCR amplicon mixes in which the fluorescence for Pacific silver fir averaged 64% of the total fluorescence in the <1mm size class and with Douglas-fir at 21% and hemlock at 15%. This over-representation relative to biomass was less dramatic in the 1-2mm size class where Pacific silver fir averaged 47% of the total fluorescence.