Petrographic recognition of layer-bounding surfaces in stalagmites offers an important tool in constructing paleoclimate records. Previous petrographic efforts have examined thickness of layers (a possible proxy for annual rainfall) and alternation of layers in couplets (a possible indicator of seasonality). Layer-bounding surfaces, in contrast, delimit series of layers and represent periods of non-deposition, either because of exceptionally wet or exceptionally dry conditions.Two types of layer-bounding surfaces can be recognized according to explicitly defined petrographic criteria. Type E layer-bounding surfaces are surfaces at which layers have been truncated or eroded at the crest of a stalagmite. Keys to their recognition include irregular termination of layers otherwise present on the stalagmite’s flank, dissolutional cavities, and coatings of non-carbonate detrital materials. Type E surfaces are interpreted to represent wet periods during which drip water became so undersaturated as to dissolve pre-existing stalagmite layers, and thus they necessarily represent hiatuses in the stalagmite record. Type L layer-bounding surfaces are surfaces below which layers become thinner upward and/or layers have lesser lateral extent upward, so that the stalagmite’s layer-specific width decreases. They are thus surfaces of lessened deposition and are interpreted to represent drier conditions in which drip rate slowed so much that little deposition occurred. A Type L surface may, but does not necessarily, represent a hiatus in deposition. However, radiometric age data show that Type L surfaces commonly represent significant hiatuses.These surfaces are significant to paleoclimate research both for their implications regarding climate change (exceptionally wet or dry conditions) and in construction of chronologies in which other data, such as stable isotope ratios, are placed. With regard to climate change, recognition of these surfaces provides paleoclimatological information that can complement or even substitute for geochemical proxies. With regard to chronologies, recognition of layer- bounding surfaces allows correct placement of hiatuses in chronologies and thus correct placement of geochemical data in time series. Attention to changing thickness of annual layers and thus to accumulation rate can also refine a chronology. A chronology constructed with attention to layer-bounding surfaces and to changing layer thickness is much more accurate than a chronology in which hiatuses are not recognized at such surfaces

No abstract