Recent work has shown that the relatively crude classification of “fine roots” using a 2mm diameter threshold is not appropriate when trying to understand root function, lifespan, and carbon and nutrient fluxes.  Branching order may be a more functionally relevant classification system for roots.  Guo et al. (2008) found that 1^st through 3^rd order roots had little secondary development and high rates of mycorrhizal colonization (indicating absorptive capacity), while 4^th and 5^th order roots lacked mycorrhizal colonization but had both secondary and xylem development (indicating  a shift in function from absorption to transport and possibly storage).

Due to the inherent difficulty of studying fine roots and their associated mycorrhizal fungi, ecosystem models of carbon and nitrogen often lack adequate incorporation of belowground processes.  Measurements of carbon and nitrogen stable isotopes (δ¹³C and δ¹⁵N) by root order may provide a better understanding of carbon and nitrogen sources and sinks across species, and the movement of C and N between above and belowground components.  We measured the δ¹³C and δ¹⁵N of roots by root order for both arbuscular and ecto-mycorrhizal trees at different soil depths during three times throughout the growing season.

Results/Conclusions

Clear patterns emerged in the δ¹⁵N signatures of roots across species, order, depth, and time.  Root δ¹⁵N increased with depth in both species, tracking the δ¹⁵N of soil with a 5‰ depletion relative to surrounding soil.  The decrease in δ¹⁵N with root order was three times greater for ecto-mycorrhizal Larix gmelinii, 0.6‰ per order, than in the arbuscular mycorrhizal species, Fraxinus mandschurica, 0.2‰ per order.  The pattern in δ¹⁵N appears to be driven by the relatively larger mass of ¹⁵N-enriched fungal material on lower order ecto-mycorrhizal roots.  Temporal changes in the slope of δ¹⁵N versus root in Larix suggests addition of ¹⁵N-depleted N (presumably resorbed foliar N) to 4^th and 5^th order roots beginning in September.We hypothesized that δ¹³C would be more variable in lower order roots due to their shorter lifespan and the temporally variable δ¹³C signature of photosynthate.  However, δ¹³C showed little variance across all depths, plots, and dates in both species (within measurement error).  The lack of systematic variance in δ¹³C implies that root carbon is derived from a large internal carbon pool which integrates the δ¹³C of photosynthate produced over multi-year time periods.