Geomorphology of Cement Creek and its Relation to Ferricrete Deposits

Latest Quaternary landforms, composed of gravel and peat, in the lower 11 kilometers of Cement Creek were mapped and dated in order to understand the geomorphic history of the area and thus provide a geomorphological context for understanding the formation of ferricrete, as part of the Animas River watershed study. The San Juan Mountains were nearly covered by alpine glaciers during the latest Pleistocene and presumably numerous times before. During the most recent episode, generally referred to as the Pinedale glaciation, ice was about 520 meters thick at the present site of Silverton and 430 meters thick at the present site of Gladstone. That glaciation ended about 12,000 14 C years ago, and except for till, the sediments along Cement Creek were likely deposited after that. Other deposits apparently formed during deglaciation, including a canyon-filling stream-gravel deposit, a localized stream terrace about 20 meters higher than the valley floor, and a large landslide. Prominent alluvial fans, located at the mouths of most tributaries, also probably began to accumulate as soon as ice disappeared from the floor of the Cement Creek valley. The fans probably accumulated rapidly until the surrounding hillslopes became stabilized by vegetation, which may have taken many centuries. The largest fans, or pairs of fans, aggraded the valley bottom perhaps more than 20 meters and caused Cement Creek to aggrade upstream of them. This resulted in a segmented longitudinal profile of the Cement Creek valley that is still present, because the bulk of the fan deposits have not been removed by Cement Creek. More than 6,000 years ago, however, Cement Creek partially breached the toes of the fans, incising through as much as 5 meters of fan sediment. This, in turn, caused the fans to become incised by their tributary streams. Between 3700 B.C. and A.D. 400 (5,650 to 1,550 cal. yr B.P.) Cement Creek migrated laterally, but its bed remained at the level of the present streambed. Starting about A.D. 400, which is slightly before the Medieval Warm Period, Cement Creek began to aggrade and rose as much as 3 meters above its previous level. Cement Creek incised back to its previous level, leaving behind a prominent stream terrace, between A.D. 1330 and A.D. 1700. This incision may have occurred before A.D. 1500 and perhaps about A.D. 1440. Incision of the Cement Creek terrace is roughly synchronous with the beginning of the Little Ice Age. Aggradation and incision of the Cement Creek terrace were not caused by local base level processes, and may have been the result of climate change, although the precise mechanisms involved are not known. Recent human activities, in contrast, have had little influence on the shape and physical processes of Cement Creek. The influence of human activities on geochemical process is beyond the scope of this study, with the exception of ferricrete formation. Iron compounds locally cement clastic sediment of all origins, creating conglomerate-type ferricrete. Most ferricrete exposures are dry and presumably are inactive, and most are prehistorical in age. Defining the timing of conglomerate-type ferricrete cementation is problematic, even where the depositional age of the clastic sediment is precisely known, because cementation could have occurred at any time after clastic deposition. Cementation of distal fan sediment must have occurred more than 6,000 years ago, because that is when the ground water in the gravel was permanently drained by stream incision. A spring-deposited ferricrete encased charcoal that is 4,500 cal. years old, so presumably conglomerate ferricrete also formed during the middle Holocene. Stream gravels deposited in the past 500 years are also cemented by iron compounds. We conclude that ferricrete formed in Cement Creek clastic sediment at various times and locations throughout the Holocene, and thus most ferricrete is unrelated to mining. Exposures of ferricrete are spatially discontinuous and do not correspond to the rate of emergence of ground water during the low-flow season. This suggests that the geochemical conditions necessary for ferricrete formation are not uniform in the watershed. Exposures of wet, possibly active ferricrete are almost invariably located where there are sedge wetlands (underlain by peat) on an adjacent terrace. We conclude that both wetlands and wet ferricrete result from the perennial emergence of ground water that originated in tributary subbasins, rather than from Cement Creek itself. One must take care when using peat to date stream incision because peat can continue to accrete in wetlands after stream incision. This continuing accretion occurs because the water supply to wetlands is from emerging ground water that originated in tributary subbasins, not from the main stream. Thus, the presence of sedge peat in subalpine settings like Cement Creek is a paleo-environmental indicator of the emergence of ground water from the valley sides.