ALBERT

Notes: We studied net ecosystem CO2 exchange (NEE) dynamics in a high-elevation, subalpine forest in Colorado, USA, over a two-year period. Annual carbon sequestration for the forest was 6.71 mol C m−2 (80.5 g C m−2) for the year between November 1, 1998 and October 31, 1999, and 4.80 mol C m−2 (57.6 g C m−2) for the year between November 1, 1999 and October 31, 2000. Despite its evergreen nature, the forest did not exhibit net CO2 uptake during the winter, even during periods of favourable weather. The largest fraction of annual carbon sequestration occurred in the early growing-season; during the first 30 days of both years. Reductions in the rate of carbon sequestration after the first 30 days were due to higher ecosystem respiration rates when mid-summer moisture was adequate (as in the first year of the study) or lower mid-day photosynthesis rates when mid-summer moisture was not adequate (as in the second year of the study). The lower annual rate of carbon sequestration during the second year of the study was due to lower rates of CO2 uptake during both the first 30 days of the growing season and the mid-summer months. The reduction in CO2 uptake during the first 30 days of the second year was due to an earlier-than-normal spring warm-up, which caused snow melt during a period when air temperatures were lower and atmospheric vapour pressure deficits were higher, compared to the first 30 days of the first year. The reduction in CO2 uptake during the mid-summer of the second year was due to an extended drought, which was accompanied by reduced latent heat exchange and increased sensible heat exchange. Day-to-day variation in the daily integrated NEE during the summers of both years was high, and was correlated with frequent convective storm clouds and concomitant variation in the photosynthetic photon flux density (PPFD). Carbon sequestration rates were highest when some cloud cover was present, which tended to diffuse the photosynthetic photon flux, compared to periods with completely clear weather.The results of this study are in contrast to those of other studies that have reported increased annual NEE during years with earlier-than-normal spring warming. In the current study, the lower annual NEE during 2000, the year with the earlier spring warm-up, was due to (1) coupling of the highest seasonal rates of carbon sequestration to the spring climate, rather than the summer climate as in other forest ecosystems that have been studied, and (2) delivery of snow melt water to the soil when the spring climate was cooler and the atmosphere drier than in years with a later spring warm-up. Furthermore, the strong influence of mid-summer precipitation on CO2 uptake rates make it clear that water supplied by the spring snow melt is a seasonally limited resource, and summer rains are critical for sustaining high rates of annual carbon sequestration.

Notes: Unlithified and partly lithified carbonate sequences are ideally suited to the application of ground-penetrating radar (GPR), augmented by percussion augering and shallow seismic techniques, all tied to present-day topography using global positioning system (GPS) methods. This methodology provides the first clear information on the distribution and geometry of lithofacies within buried tufa complexes. The approach has been applied to a thick succession of Holocene tufas filling a gorge site along a 3·5-km length of the River Lathkill, north Derbyshire. Earlier studies have demonstrated the presence of up to 16 m of tufas and sapropels associated with two transverse tufa dams (barrages). These strata have been accumulating throughout the Holocene, although tufa developments at present are of minor extent. Internal tufa morphologies are recorded by GPR as ‘bright’, laterally continuous reflections for lithified, concretionary and lithoclast-rich horizons. The ‘brightest’ reflectors occur within well-cemented barrages and delineate core areas and prograding buttress zones. In contrast, unlithified lime muds and sapropels produce low-contrast reflections. Lithostratigraphic control and depth calibration of the GPR profiles was provided by percussion augering at selected sites. Six distinct lithofacies and four secondary barrages are identified in the study. Constructional and destructional events can be identified and correlated within the GPR profile network, and the internal growth morphologies of the barrages are apparent. GPR profiles also clearly define the evolution of the facies geometries. Three phases of tufa development can be recognized within the GPR data and greatly extend our understanding of Holocene tufa-forming processes in valley sites: (a) Early Holocene barrage build-ups but with limited paludal deposition; (b) Middle Holocene ponding and sapropel accumulation under ‘warm’ conditions; and (c) Late Holocene barrage termination and valley levelling, probably coincidental with anthropogenic activity. This type of multidisciplinary approach should be considered as an essential prerequisite to all biostratigraphic and geochemical studies of Holocene freshwater carbonate sites.