UVM Theses and Dissertations
Format:
Print
Author:
Sosa-Gonzalez, Veronica
Dept./Program:
Natural Resources
Year:
2012
Degree:
MS
Abstract:
Knowledge of background erosion rates and the controls on them is important for evaluating land management practices in Panama, where the proper function of the Panama Canal is linked to water-storage reselVoir sedimentation and thus the rate at which sediment leaves the adjacent landscape. This research determined long-term erosion rates in selected watersheds (n=40) in Panama using cosmogenic nuclide analysis in river sediments. Sampled watersheds stretch from east to west across the country, and most include at least some of the mountainous spine of the isthmus. This research is the first attempt to quantify erosion rates in Panama at a broad scale.
A total of 44 variables (physiographic, climatic, seismic, geology and a land use proxy) were quantified and their relationship to erosion rates was assessed using linear regressions and ANDVA. I also assessed the grain size effect on ¹⁰Be concentrations in stream and landslide samples. Three landslide-related samples were collected and split into seven size fractions to quantify the effect of such rapid sediment inputs on the measured ¹⁰Be concentration.
Cosmogenic-inferred erosion rates ranged from 26.1 m/Myr ± 0.6 to 597 ± 62m/Myr. The strongest and most significant relationship in the dataset was found between erosion rate and silicate weathering rate, the mass of material leaving the basin in solution (R²= 0.726, p= 0.004). None of the physiographic variables showed a significant relationship to erosion (at the 95% significance level). The number of seismic events in a 10-km buffer is weakly and positively related to erosion rates (R²= 0.338, p= 0.033); the average magnitude of seismic events in a 100-km buffer (R² = 0.316, p= 0.047) is weakly and negatively related to erosion rates. This may indicate that the energy released during seismic events weakens rocks and increases the rate of erosion up to a certain threshold distance. Several bioclimatic variables showed weak but positive relationships with erosion rates including temperature seasonality (R²= 0.445, p= 0.004) and precigitation during both the driest month (R²= 0.319, p= 0.045) and the driest quarter (R² = 0.376, p= 0.017).
Two bioclimatic variables showed weak negative relations to erosion rates: isothermality (R²= 0.381, p= 0.015) and precipitation seasonality (R²= 0.394, p= 0.012). Watersheds were clustered into regions according to their location. Spatial analysis at the regional scale strengthened the relationships between variables and erosion rates, but decreased the statistical significance of those relationships. An inverse relationship was found between ¹⁰Be concentration and grain size in a set of landsliderelated samples. These samples demonstrate both that deep-seated material, which enters streams when such slides happen, carries less ¹⁰Be than surface material, and that finegrained material is preferentially sourced from near the land surface. Due to the small number of grain-size specific samples, no statistical testing can be done to prove differences in isotopic concentration according to their source. Erosion rates in Panama are higher than all other published cosmogenic-derived erosion rates in tropical climates. The lack of relationship between erosion and physiography observed in Panama is uncommon.
A total of 44 variables (physiographic, climatic, seismic, geology and a land use proxy) were quantified and their relationship to erosion rates was assessed using linear regressions and ANDVA. I also assessed the grain size effect on ¹⁰Be concentrations in stream and landslide samples. Three landslide-related samples were collected and split into seven size fractions to quantify the effect of such rapid sediment inputs on the measured ¹⁰Be concentration.
Cosmogenic-inferred erosion rates ranged from 26.1 m/Myr ± 0.6 to 597 ± 62m/Myr. The strongest and most significant relationship in the dataset was found between erosion rate and silicate weathering rate, the mass of material leaving the basin in solution (R²= 0.726, p= 0.004). None of the physiographic variables showed a significant relationship to erosion (at the 95% significance level). The number of seismic events in a 10-km buffer is weakly and positively related to erosion rates (R²= 0.338, p= 0.033); the average magnitude of seismic events in a 100-km buffer (R² = 0.316, p= 0.047) is weakly and negatively related to erosion rates. This may indicate that the energy released during seismic events weakens rocks and increases the rate of erosion up to a certain threshold distance. Several bioclimatic variables showed weak but positive relationships with erosion rates including temperature seasonality (R²= 0.445, p= 0.004) and precigitation during both the driest month (R²= 0.319, p= 0.045) and the driest quarter (R² = 0.376, p= 0.017).
Two bioclimatic variables showed weak negative relations to erosion rates: isothermality (R²= 0.381, p= 0.015) and precipitation seasonality (R²= 0.394, p= 0.012). Watersheds were clustered into regions according to their location. Spatial analysis at the regional scale strengthened the relationships between variables and erosion rates, but decreased the statistical significance of those relationships. An inverse relationship was found between ¹⁰Be concentration and grain size in a set of landsliderelated samples. These samples demonstrate both that deep-seated material, which enters streams when such slides happen, carries less ¹⁰Be than surface material, and that finegrained material is preferentially sourced from near the land surface. Due to the small number of grain-size specific samples, no statistical testing can be done to prove differences in isotopic concentration according to their source. Erosion rates in Panama are higher than all other published cosmogenic-derived erosion rates in tropical climates. The lack of relationship between erosion and physiography observed in Panama is uncommon.