Aboveground autotrophic respiration in a Spanish black pine forest: Comparison of scaling methods to improve component partitioning.

Abstract

Total wood CO2 efflux (Rw) varies vertically within individual trees, and leaves experience large variations in foliar respiration (Rf) rates over their life spans and during daily periods. Therefore, accurate sampling approaches are required to improve aboveground autotrophic respiration (RAa) estimations in stand-scale carbon cycling studies. We scaled-up Rw (comprising stem and branch CO2 efflux; ES and EB, respectively) and Rf from biometric and flux-chamber measurements taken between 2011 and 2013 in a Spanish black pine (Pinus nigra Arn. ssp. salzmannii) forest at an unburnt (UB) site and a low burn-severity (LS) site. We measured seasonal ES at breast height (1.30m) on 9 trees at each site, which was also vertically examined on 5 of those trees. We also measured seasonal Rf in current- and previous-year needles on 3 trees at each site, and quantified Rf variations in darkness and light. Finally, we compared complex and simple scale-up methods which did or did not account for the vertical variation in Rw and the effects of leaf ageing and light inhibition on Rf, respectively. The simple methods underestimated the annual stand-level stem, branch, and total wood respiration ≈35%, 55%, and 41%, respectively, and overestimated annual stand-level whole-canopy foliage respiration ≈43% at both sites. Both methods provided similar annual stand-level RAa estimates, although the complex methods improved estimations of the relative contribution of RAa components. Thus, based on the complex methods the mean annual RAa at the stand-level was 4.53±0.25 and 4.45±0.12MgCha-1year-1 at the UB and LS sites, respectively. Our data also confirmed that the low-severity fire did not alter the RAa rates. Collectively, this study reveals that complex approaches, applicable in other forest ecosystems, enhance the accuracy of partitioning RAa sources by reducing the error in scaling-up in chamber-based measurements.