Extended megadroughts in the southwestern United States during Pleistocene interglacials.

The potential for increased drought frequency and severity linked to anthropogenic climate change in the semi-arid regions of the southwestern United States (US) is a serious concern. Multi-year droughts during the instrumental period and decadal-length droughts of the past two millennia were shorter and climatically different from the future permanent, 'dust-bowl-like' megadrought conditions, lasting decades to a century, that are predicted as a consequence of warming. So far, it has been unclear whether or not such megadroughts occurred in the southwestern US, and, if so, with what regularity and intensity. Here we show that periods of aridity lasting centuries to millennia occurred in the southwestern US during mid-Pleistocene interglacials. Using molecular palaeotemperature proxies to reconstruct the mean annual temperature (MAT) in mid-Pleistocene lacustrine sediment from the Valles Caldera, New Mexico, we found that the driest conditions occurred during the warmest phases of interglacials, when the MAT was comparable to or higher than the modern MAT. A collapse of drought-tolerant C(4) plant communities during these warm, dry intervals indicates a significant reduction in summer precipitation, possibly in response to a poleward migration of the subtropical dry zone. Three MAT cycles ∼2 °C in amplitude occurred within Marine Isotope Stage (MIS) 11 and seem to correspond to the muted precessional cycles within this interglacial. In comparison with MIS 11, MIS 13 experienced higher precessional-cycle amplitudes, larger variations in MAT (4-6 °C) and a longer period of extended warmth, suggesting that local insolation variations were important to interglacial climatic variability in the southwestern US. Comparison of the early MIS 11 climate record with the Holocene record shows many similarities and implies that, in the absence of anthropogenic forcing, the region should be entering a cooler and wetter phase.

Temporal variation in delta(13)C of ecosystem respiration in the Pacific Northwest: links to moisture stress.

We measured seasonal and interannual variations in delta(13)C values within the carbon reservoirs (leaves and soil) and CO(2) fluxes (soil and ecosystem respired CO(2)) of an old growth coniferous forest in the Pacific Northwest USA with relation to local meteorological conditions. There were significant intra-annual and interannual differences in the carbon isotope ratios of CO(2) respired at both the ecosystem (delta(13)C(R)) and the soil levels (delta(13)C(R-soil)), but only limited variations in the carbon isotope ratios of carbon stocks. The delta(13)C(R) values varied by as much as 4.4 per thousand over a growing season, while delta(13)C(R-soil )values changed as much as 6.2 per thousand. The delta(13)C of soil organic carbon (delta(13)C(SOC)) and needle organic carbon (delta(13)C(P)) exhibited little or no significant changes over the course of this study. Carbon isotope discrimination within leaves (Delta(p)) showed systematic decreases with increased canopy height, but remained fairly constant throughout the year (Delta(p)=17.9 per thousand -19.2 per thousand at the top of the canopy, Delta(p)=19.6 per thousand -20.9 per thousand at mid-canopy, Delta(p)=23.3 per thousand -25.1 per thousand at the canopy base). The temporal variation in the delta(13)C of soil and ecosystem respired CO(2) was correlated ( r=0.93, P<0.001) with soil moisture levels, with dry summer months having the most (13)C-enriched values. The dynamic seasonal changes in delta(13)C of respired CO(2) are hypothesized to be the result of fast cycling of recently fixed carbon back to the atmosphere. One scaling consequence of the seasonal and interannual variations in delta(13)C(R) is that inversion-based carbon-cycle models dependent on observed atmospheric CO(2) concentration and isotope values may be improved by incorporating dynamic delta(13)C(R) values to interpret regional carbon sink strength.

Rapid 18O analysis of small water and CO2 samples using a continuous-flow isotope ratio mass spectrometer.

High-frequency throughput is often needed in isotopic studies in biological and medical fields. Here we report that high-precision oxygen isotope ratio measurements of water (+/-0.13 per thousand) were rapidly and routinely made on small samples (40-100 microL) using an isotope ratio mass spectrometer operated in continuous-flow mode. Simple modifications to existing instrumentation allow for rapid manual analyses of dilute CO2 (10% CO2/90% N2), including the addition of a septum port and water trap prior to the gas chromatography (GC) column (elemental analyzer column in this study) and the extension of fused-silica capillary tubing between the mass spectrometer source and the effluent tubing from the GC column (located within the CONFLO unit on Finnigan mass spectrometers). We routinely analyzed 20 small-volume samples per hour using this technique, without sacrificing precision of the oxygen isotope ratio measurement.

Age-related variations in delta(13)C of ecosystem respiration across a coniferous forest chronosequence in the Pacific Northwest.

We tested the hypothesis that forest age influences the carbon isotope ratio (delta13C) of carbon reservoirs and CO2 at local and regional levels. Carbon isotope ratios of ecosystem respiration (delta13C(R)), soil respiration (delta13C(R-soil)), bulk needle tissue (delta13C(P)) and soil organic carbon (delta(13)C(SOC)) were measured in > 450-, 40- and 20-year-old temperate, mixed coniferous forests in southern Washington, USA. Values of delta13C(R), delta13C(R-soil), delta13C(P) and delta13C(SOC) showed consistent enrichment with increasing stand age. Between the youngest and oldest forests there was an approximately 1 per thousand enrichment in delta13C(P) (at similar canopy levels), delta13C(SOC) (throughout the soil column), delta13C(R-soil) (during the wet season) and delta13C(R) (during the dry season). Mean values of delta13C(R) were -25.9, -26.5 and -27.0 per thousand for the 450-, 40- and 20-year-old forests, respectively. Both delta13C(R-soil) and the difference between delta13C(R) and delta13C(R-soil) were more 13C enriched in older forests than in young forest: delta13C(R) - delta13C(R-soil) = 2.3, 1.1 and 0.5 per thousand for the 450-, 40- and 20-year-old forests, respectively. Values of delta(13)C(P) were proportionally more depleted relative to delta13C(R): delta13C(R) - delta13C(P) = 0.5, 2.2 and 2.5 per thousand for the 450-, 40- and 20-year-old forests, respectively. Values of delta13C(P) were most 13C-enriched at the top of the canopy and in the oldest forest regardless of season (overall values were -26.9, -28.7 and -29.4 per thousand for the 450-, 40- and 20-year-old forests, respectively). Values of delta13C(SOC) from shallow soil depths were similar to delta13C(P) values of upper- and mid-canopy needles. All delta13C data are consistent with the hypothesis that a decrease in stomatal conductance associated with decreased hydraulic conductance leads to increased CO2 diffusional limitations in older coniferous trees. The strong associations between delta13C(P) in needles with delta13C(R) and delta13C(R-soil) at the forest level suggest that 13C observations scale between leaf and ecosystem levels.