Contrasting responses to drought of forest floor CO2 efflux in a Loblolly pine plantation and a nearby Oak-Hickory forest

Forest floor CO2 efflux (F-ff) depends on vegetation type, climate, and soil physical properties. We assessed the effects of biological factors on F-ff by comparing a maturing pine plantation (PP) and a nearby mature Oak-Hickory-type hardwood forest (HW). F-ff was measured continuously with soil chambers connected to an IRGA during 2001-2002. At both sites, F-ff depended on soil temperature at 5 cm (T-5) when soil was moist (soil moisture, theta > 0.20 m(3) m(-3)), and on both T-5 and theta when soil was drier. A model (F-ff (T-5, theta)) explained >= 92% of the variation in the daily mean F-ff at both sites. Higher radiation reaching the ground during the leafless period, and a thinner litter layer because of faster decomposition, probably caused higher soil temperature at HW compared with PP. The annual F-ff was estimated at 1330 and 1464 g C m(-2) yr(-1) for a year with mild drought (2001) at PP and HW, respectively, and 1231 and 1557 g C m(-2) yr(-1) for a year with severe drought (2002). In the wetter year, higher soil temperature and moisture at HW compared with PP compensated for the negative effect on F-ff of the response to these variables resulting in similar annual F-ff at both stands. In the drier year, however, the response to soil temperature and moisture was more similar at the two stands causing the difference in the state variables to impel a higher F-ff at HW. A simple mass balance indicated that in the wetter year, C in the litter-soil system was at steady state at HW, and was accruing at PP. However, HW was probably losing C from the mineral soil during the severe drought year of 2002, while PP was accumulating C at a lower rate because of a loss of C from the litter layer. Such contrasting behavior of two forest types in close proximity might frustrate attempts to estimate regional carbon (C) fluxes and net C exchange.