Root morphology and belowground carbon storage in tanoak (Notholithocarpus densiflorus)

Sampling challenges have restricted research on biomass and carbon (C) in root systems of trees due to roots' variable distribution in the soil and underestimation of biomass in non-destructive sampling. Tanoak (Notholithocarpus densiflorus), a tree species threatened by an exotic disease called sudden oak death, sprouts from belowground burls after disturbances and stores an unknown amount of C belowground. We sought (i) to identify models describing root morphology in tanoak, (ii) to derive allometric models of belowground biomass and C in tanoak and, (iii) to test for relationships between root-to-shoot biomass ratio and variables such as stand density, species composition, and crown position. This project was conducted at the L.W. Schatz Demonstration Tree Farm in Humboldt County, CA. Twelve tanoak root systems were excavated using destructive sampling, and individual models of root morphology were developed. We found that root taper rate was best modeled using the root's circumference and path length from the root burl. Larger roots closer to the burl tapered more rapidly per unit of length. Tanoak roots forked frequently, with forking events most often separated by short distances. Therefore, data for the length of individual root sections between forks were skewed, and were modeled based on path distance from the root burl and the cumulative frequency of the observed section lengths. The sum of the small-end cross-sectional areas of forked root sections was positively correlated with the root section's large-end cross-sectional area. The mass of a forked root section increased when the fork was longer and had a greater large-end cross-sectional area. Occurrence and quantity of branches (small roots branching laterally from larger roots) was dependent upon the length of the parent root section. Models of tanoak root morphology were organized together to estimate belowground biomass lost during the excavation. Applying the morphological models to obtain estimates of roots lost during excavation provided better estimates than simply regressing a lateral roots' large-end cross-sectional area against biomass of the entire root including all forks and branches. Destructive sampling involved severing coarse lateral roots by digging a trench around each tanoak and extracting the root burl and associated roots, and unearthing a subsample of the severed lateral roots that extended outside the trench. The sum of the excavated biomass, roots sampled beyond the excavated area, and predicted biomass of roots not sampled beyond the excavated area gave total belowground biomass for the 12 sample trees. Carbon was predicted from biomass estimates. Tree height, crown volume, basal area (BA), stand density index (SDI), and dummy variables for crown position and multiple stemmed tanoak were tested as predictors of belowground biomass and C. A model using basal area, SDI, and crown position explained the most variation in belowground biomass and C (R2 = 0.96). The best models in terms of AICc used only BA to predict belowground biomass and C. These models predicted that a tanoak with BA of 0.05 m2 (25 cm dbh) for example, stored 70 kg of biomass and 34 kg of C in the root system. Biomass root-to-shoot ratio varied between 0.11 and 0.65 with a mean of 0.35. Root-to-shoot ratio decreased with tree size and stand density, indicating that although large tanoaks in open stands stored the most biomass and C belowground, small open-grown tanoaks allocated the greatest proportion of their biomass belowground. Models of belowground C were applied to 0.04 ha plots previously measured at the excavation site. Inventory plots where tanoak represented one to 19% of the plots' total BA had an average of 4.7 tons of belowground tanoak C ha-1 while plots with tanoak representing 33 to 100% of the total plot BA had an average of 58.7 tons of belowground tanoak C ha-1.