Explaining the Apparent Resiliency of Loblolly Pine Plantation to Organic Matter Removal.

Many long term forest productivity experiments have found that removal of OM (OM) typically does not lead to a decrease in productivity.  There have been few hypotheses posed regarding the apparent resiliency of forest productivity to these severe OM removal treatments.  We utilized 15 years of published and unpublished data and from an OM manipulation experiment in a loblolly pine plantation in the Upper Coastal Plain of Alabama.  Treatments were replicated 4 times and included complete removal of harvest residues and forest floor (REM), doubling of harvest residues and forest floor (ADD), and a standard harvest residue management (REF).  Bulk density and grab samples of O horizons and mineral soils (0-20) were collected from 5 areas per plot and composited during years 10 and 15. Litterfall was collected monthly at during year 10 and 15, and foliage was collected in year 10 and 14.  Density fractionation was conducted on soils collected in year 15.  All whole soils, fractions, litter, and foliage were analyzed for C and N using dry combustion.  The δ¹⁵N and CuO oxidation biomarker (e.g. lignin and cutin) composition of whole soils and fractions from year 15 were measured.  At year 10 and 15 there was no difference between REF and REM while the ADD exhibited higher volume as a result of nutrient additions through the addition of OM.  At year 15 the heavey fraction (i.e. recalcitrant) N on REM was lower than the other two treatments suggesting that this recalcitrant pool helped buffer the N loss.  The δ¹⁵N composition of the heavy fraction of mineral soil from the REM site was enriched supporting the hypothesis that the REM treatment has experienced higher rates of N mineralization.  Since this finding was only significant in the slower cycling heavy fraction it suggests that these higher rates of mineralization occurred at some point in the past, but may not presently be occurring.  Higher concentrations of biomarkers indicative of higher relative degradation of whole soils and density fractions support these assertions.  In both years that foliage and litterfall samples were collected the foliage samples had N concentrations that were not different between the treatments; however literfall N concentrations were significantly lower on the REM relative to the other treatments.  We hypothesize that the lack of difference between the REM and REF is a result of a higher rate of mineralization of labile N during the early part of stand development on the REM plots subsidizing the nutrient demand of the developing stand. Since that time the REM plots have compensated for low soil N supply by retranslocating a greater fraction of foliar N.  These results have implications for long term site productivity since the recovery of labile N after severe removal of OM may be slow and the productivity of subsequent rotations (with or without OM removal) may be negatively impacted.