Browsing by Author "Romme, William, advisor"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Open Access Initial and future stand development following mountain pine beetle in harvested and uncut lodgepole pine forests(Colorado State University. Libraries, 2010) Collins, Byron Joshua, author; Romme, William, advisor; Rhoades, Charles, advisor; Hubbard, Robert, committee member; Martin, Patrick, committee memberThe extent and severity of over tory lodgepole pine (Pinus contorta Dougl.) mortality from mountain pine beetle (Dendroctonus ponderosae Hopkins) has created management concerns associated with forest regeneration, wildfire risk, human safety, and scenic, wildlife and watershed resources in western North America. In northern Colorado and southern Wyoming the long-term ecological and socioeconomic consequences of the outbreak hinge upon the response of tree regeneration in both harvested and untreated forests. To characterize initial and future forest development following mountain pine beetle mortality I conducted two studies. First, I used historic U.S. Forest Service stand and seedling survey records to compare the density and species composition of advance regeneration in uncut stands and post-harvest recruits in clearcut harvest units during pre-outbreak (1980-1996) and outbreak (2002-2007) period. Second, I compared the effects of various intensities of forest management on site conditions, seedling establishment and growth of advance regeneration to uncut areas in beetle-infested lodgepole pine stands. Advance regeneration averaged 3,953 stems ha-1 and was at least as high in beetle-infested stands compared to the pre-outbreak period. Lodgepole pine advance regeneration showed increased leader growth from 2008 to 2009 in harvested and untreated stands in response to canopy removal and decreased canopy foliage following overstory mortality. The density of seedling recruitment was three times higher in harvested than untreated stands (6,487 versus 2,021 seedlings ha-1), and did not differ between outbreak and pre-outbreak stands. Growth simulations showed uncut and partial cut stands will be dominated by subalpine fir (Abies lasiocarpa), while clearcut stand will be dominated by lodgepole pine and have attributes similar to pre-outbreak stands within a century.Item Open Access Mapping Tamarix: new techniques for field measurements, spatial modeling and remote sensing(Colorado State University. Libraries, 2009) Evangelista, Paul H., author; Romme, William, advisor; Stohlgren, Thomas, advisorNative riparian ecosystems throughout the southwestern United States are being altered by the rapid invasion of Tamarix species, commonly known as tamarisk. The effects that tamarisk has on ecosystem processes have been poorly quantified largely due to inadequate survey methods. I tested new approaches for field measurements, spatial models and remote sensing to improve our ability measure and to map tamarisk occurrence, and provide new methods that will assist in management and control efforts. Examining allometric relationships between basal cover and height measurements collected in the field, I was able to produce several models to accurately estimate aboveground biomass. The best two models were explained 97% of the variance (R 2 = 0.97). Next, I tested five commonly used predictive spatial models to identify which methods performed best for tamarisk using different types of data collected in the field. Most spatial models performed well for tamarisk, with logistic regression performing best with an Area Under the receiver-operating characteristic Curve (AUC) of 0.89 and overall accuracy of 85%. The results of this study also suggested that models may not perform equally with different invasive species, and that results may be influenced by species traits and their interaction with environmental factors. Lastly, I tested several approaches to improve the ability to remotely sense tamarisk occurrence. Using Landsat7 ETM+ satellite scenes and derived vegetation indices for six different months of the growing season, I examined their ability to detect tamarisk individually (single-scene analyses) and collectively (time-series). My results showed that time-series analyses were best suited to distinguish tamarisk from other vegetation and landscape features (AUC = 0.96, overall accuracy = 90%). June, August and September were the best months to detect unique phenological attributes that are likely related to the species' extended growing season and green-up during peak growing months. These studies demonstrate that new techniques can further our understanding of tamarisk's impacts on ecosystem processes, predict potential distribution and new invasions, and improve our ability to detect occurrence using remote sensing techniques. Collectively, the results of my studies may increase our ability to map tamarisk distributions and better quantify its impacts over multiple spatial and temporal scales.