Speakers Club: Paul Alessio and Jack Mitchell

Event Date: 

Thursday, March 3, 2016 - 2:00pm

Paul Alessio and Jack Mitchell, both UCSB graduate students, will each give a talk at this week's Speakers Club.  Paul's talk is entitled Characterizing the Spatial Variability of Saturated Hydraulic Conductivity in Historic Landslides to Determine Thresholds for Instability at Sedgwick Ranch, Ca and Jack's talk is entitled Application of micro-CT to Quantitatively Assess Tooth Crown Outline in Notoungulates.  The event begins at 2:00 PM sharp on Thursday March 3 in Webb Hall 1100.

 
Paul's abstract:
 
Steepland areas commonly experience storms that can cause sudden shallow landslides, threatening communities. In general, shallow landslides are interrelated with the generation of positive pore-fluid pressures caused by heavy rainfall and hydraulic discontinuities. In order to determine the triggering mechanisms and thresholds for landsliding, it is necessary to understand what rainfalls lead to soil water pressures that drive failure. I visited three historic landslide sites at Sedgwick Ranch, Ca,sat). It was found that a low conductivity layer is correlative with the failure plane at the soil-bedrock interface and was the primary cause of all three debris flows. Infiltration data also indicate that Ksat has a strong dependence on soil horizon. Ksat values at failure planes ranged from 4 to 33 mm/hr, whereas overlying soils of the A and B/Bt horizons often had higher infiltration rates, ranging from 37 to 138 mm/hr and 24 to 127 mm/hr, respectively. It was hypothesized that the low-conductivity layer at the base of the soil column was formed by a reduction in grain size from illuviation and decreasing pore space due to compaction. However, measurements of grain size distributions and bulk densities do not provide conclusive evidence that this is the case. Based on observation, the low conductivity layer at the soil bedrock interface is most likely formed from the infilling of bedrock fractures with sediments and weathered material. Infiltration measurements, soil properties, and hill slope parameters that control slope stability provide substantial information to perform a stability analysis. In this analysis, infiltration in the vadose zone will be modeled to estimate the depth of the wetting front and the timing of the pore-fluid pressure response that caused failure. Based on this analysis, a rainfall-rate threshold for debris flow initiation may be estimated in order to forecast landslide-causing storms for this area.
 
Jack's abstract:
 
The acquisition of high-crowned cheek teeth (hypsodonty) among various Cenozoic herbivorous mammals is classically thought to be a response to the spread of open habitats, increasing the functional longevity of teeth in conditions that favor high tooth wear (e.g., phytoliths in grasses, grit on food items). Notoungulates, an extinct clade of South American mammals, are one of the earliest groups to exhibit hypsodonty in the fossil record. Given the vertical height of their crowns, hypsodont teeth often undergo considerable change in outline and form as they wear. Typical mammalian dental studies use the tooth crown morphology to make inferences about ecology and phylogeny; however, crown morphology is constantly changing through wear.  While crown morphology is labile, crown outline is genetically determined and is stable until that particular section of the tooth comes into occlusion.  The present study uses 3D micro-computed tomography (CT), a non-destructive technique, to digitally reconstruct fossil notoungulate teeth.  Once a dental arcade is reconstructed, digital cross-sections parallel to the occlusal surface of each tooth are created to form a wear series, attempting to mimic gross macrowear.  The progressive cross-sections of the teeth are analyzed using semi-landmarks to quantitatively assess the change in tooth outline.  This technique may be used to corroborate phylogenetic studies of hypsodont species that are limited in the number of specimens. 
Paul Alessio and Jack Mitchell