Thesis (Ph. D. - Geosciences)--University of Arizona, 1985. / Includes bibliographical references (leaves 187-204).

The late Holocene flood history and associated channel changes were reconstructed for the Escalante River in south-central Utah. Analyses of flood deposits at 8 sites in the bedrock canyon indicate that the frequency of large floods was at a maximum 1100 to 900 yrs BP and in historic times in a 2000 year record. The largest flood occurred approximately 900 to 1000 yrs BP and was 7 times the largest flood recorded at a gaging station. The paleoflood discharges were close to the "maximum expected flood" derived from a regional flood envelope curve, and the 100-yr flood was increased 220% to 800 cubic meters per second (cms) with the addition of four historic flood discharges. Possible nonstationarity in the distribution due to channel changes and climatic shifts reduced the reliability of statistical flood-frequency analyses. The additional parameters of the "largest recorded flood" in 2000 years of paleoflood record -- 720 ems -- and the "maximum expected flood" -- 1180 cms -- were added to the flood-frequency summary. Channel changes in the upstream alluvial channel were related to flood-frequency changes. Valley-margin stratigraphy representing 1600 years of deposition indicated that after 1100 yrs BP, a time of increased frequency of large floods, a marshy floodplain was converted to a dry, fire-swept meadow and an arroyo 24-m wide and 2.5-m deep formed. This arroyo quickly filled with sediments between 500 and 400 yrs BP and a smaller channel then formed and persisted until settlement of the basin. Floods between 1909 and 1940 transformed the small channel into an arroyo up to 100-m wide and 17-m deep. The cause for flood-frequency and consequent channel changes on the Escalante River is complicated. Land-use practices caused pronounced changes in watershed and floodplain conditions. A subtle shift in climate increased the amount of summer precipitation and intensity of storms. The inability to test either the land-use practices or climatic shift hypotheses independently precludes the determination of a regional cause for arroyos.

Arroyos are steep-walled, entrenched, typically ephemeral streams commonly found in dryland river systems that form when streams incise into previously deposited alluvial fill. Arroyos in the southwestern U.S. have been studied extensively following the historic period of arroyo cutting in the late 1800s and early 1900s A.D. The upper Escalante River in south-central Utah similarly began incising in 1909, and records evidence for past cut and fill cycles in well-exposed walls along the now continuous arroyo. Establishing robust geochronologies of past arroyo cycles in these fluvial settings has been difficult. Recent improvements in accelerator mass spectrometry (AMS) radiocarbon (14C) and optically stimulated luminescence (OSL) dating provide an opportunity to link more highly resolved fluvial records to existing paleoclimate records. This allows hypotheses regarding the causes of arroyo cycles to be tested, and for the role of climate versus intrabasinal characteristics to be examined. One major objective of this research was to examine the applicability of OSL and AMS 14C dating in the upper Escalante, as both methods have proved problematic in similar settings. In total, 37 ages were obtained, 21 OSL and 16 14C ages. The Holocene fluvial history of the upper Escalante River was reconstructed using these age results and stratigraphic relationships. The chronostratigraphic record developed in this study suggests that at least six arroyo cycles have occurred in the upper Escalante since the middle Holocene, with incision occurring ~4.4 – 4.2 ka, ~2.6 – 2.4 ka, ~1.8 – 1.5 ka, ~1.0 – 0.9 ka, ~0.5 - 0.4 ka, and during the historic period of arroyo entrenchment. While semi-synchronous arroyo cutting (indicative of a climate signal) appears to have occurred in the Paria and Escalante drainages over the last 1 ka, correlations between proximal drainages are less clear prior to 1 ka, although this may be due in part to preservation effects. Overall, linkages to specific climate regimes and correlations to regional drainages are difficult to identify, suggesting that internal geomorphic thresholds are important in determining when arroyo entrenchment occurs in individual catchments.

Studies of environmental and riparian vegetation interrelationships along the lower Escalante River were conducted during 1974 and 1975. The study area included the Escalante River flood terrace between Harris Wash and Coyote Gulch. Methods were developed compatible for wilderness use . Sixty nine 1 x 10 meter macro-plots were taken at 23 locations. Measurements were taken of major environmental parameters thought to influence riparian vegetation : stratigraphy, soil texture and moisture, river bank angle and aspect, and height and width of flood terraces. Canyon profile was found to be strongly influenced by geologic strata. Resistance to erosion by these strata determines canyon fluctuations and flash floods further modify the river bank profile by creating three distinct flood terraces. Vegetation distribution was found to be generally correlated with stratigraphy and flood terraces. The most dominant semi-aquatic species, Scirpus americanus and Eguisetum laevigatum were found on soils saturated to supersaturated with water on the low flood terraces. Baccharis emoryi was found in the medium and high flood terraces where the Chinle Formation was exposed. Different time periods between river inundations and flash floods created significantly different population age structures for the three major tree species (Salix exigua, Populus fremontii, and Tamarix pentandra). Tree longevity increased from low to high flood terraces. Population age structure differences were attributed to flooding which dynamically maintains each cohort. Regeneration by root suckers for Salix and Populus have survival rates greater than Tamarix whose seedling source is washed away by periodic fluctuations in river level. Implications of this research are that reduction in river flow or regulation of floods would remove population equilibrium controls. Tamarix, without the effec t of its seed source being washed away , would have a successional advantage over the other two native tree species whose densities in the young age classes are lower than that of Tamarix. Furthermore, tree populations would shift to an older age structure with greater density. Deleterious effects would be increased evapotranspiration and inaccessability to river recreation. The effects of such vegetation changes on wildlife are unknown.

A Fremont site distribution model for the Grand Staircase-Escalante National Monument during the period A.D. 500—1050/1100 posits that the Fremont subsistence strategy (seasonal mobility with dependence on both agriculture and hunting/foraging) is reflected by a site pattern of low-investment, seasonal or short-term habitation sites and isolated storage facilities at "lowland" elevations, and high-investment, long-term residence sites at "upland" elevations (McFadden 1998, 2000). This research assesses the model to evaluate its general precision, looking particularly at its success in modeling site locations for long-term residential versus seasonal/short-term habitation sites. A database including more than 400 Fremont sites was created to evaluate the model. Data variables examined in this thesis included elevation, distance-to-water, and primary landform. Analysis of the elevation data demonstrates that the McFadden model does not fit the actual distribution of Fremont sites identified from survey. Further analysis also established that distance-to-water is not an effective variable in accurately modeling Fremont site patterning over this region. The association between functional site types and primary landforms, however, does appear to more accurately reflect site distribution as observed on the ground. Based on these results, a new model for Fremont site distribution in the upper Escalante River drainage is proposed.