Search

The increasing collaboration between physicists and biologists in recent years has led to a series of breakthroughs enabled, in part, by the use of lasers in biological experiments. Once such recent development is the biological laser where a living cell containing a fluorescent protein or dye acts as a laser gain medium. This thesis presents work designed to develop the idea of the living laser leading to their implementation as a research tool. This work has consisted of two main areas of research; microfluidics and biological lasers. The use of microfluidics enables the miniaturisation of many existing types of biological diagnostics. In this thesis devises are demonstrated for use in temperature sensing and flow cytometry. These were fabricated through the use of Ultrafast Laser Inscription (ULI) and selective chemical etching. As part of this work we have also investigated the integration of silver nanoparticles into microfluidic devices, using photo-reduction, for the enhancement of Raman sensing. Several types of living laser have been fabricated containing different fluorescent dyes and Enhanced Green Fluorescent Protein (eGFP). Lifetime extension has been achieved using vitamin C and work has been conducted towards demonstrating calcium sensing inside cells. As an alternative to the dyes used in these experiments we have also investigated the use of upconverting nanoparticles.

This research investigates the vegetation of Great Smoky Mountains National Park (GRSM) using three different techniques: 1) analysis of vegetation data collected circa 1930, 2) remote sensing of current (1984) vegetation using Landsat Thematic Mapper (TM) satellite imagery, and 3) the integration of gradient analysis with forest succession modeling. Analysis of the 1930's data revealed that Castanea dentata was the dominant species in the GRSM at that time, even with the introduction of the chestnut blight to the park in 1926. Classification of the 1930's plot data identified 16 unique vegetation types ranging from the low elevation, xeric Pinus rigida type to the high elevation, mesic Abies fraseri type. Ordination of the data indicated that species are responding to a complex moisture - elevation gradient. Unsupervised classification of the satellite data was able to identify 14 vegetation types. The classification process used all seven TM bands and had an overall classification accuracy of 83%. The types identified were Spruce - Fir, Northern Hardwood, Cove Hardwood, Mesic Oak, Mixed Mesic Hardwood, Tulip Poplar, Xeric Oak, Pine - Oak, Heath Bald, Grassy Bald, Grape Thicket, Treeless, and Water. The Cove Hardwood type was the most prominent type and occupied 33% of the park's area. The integration of gradient analysis with a forest succession model yielded a model that was capable of replicating vegetation types throughout the entire GRSM landscape. The model was used to provide insights into the dynamics of Castanea dentata prior to the chestnut blight and patterns of replacement of Castanea dentata, and its effect on biomass after the blight, took place via multiple pathways depending on site conditions. This research emphasizes the vegetation of GRSM as a whole rather than as a study of specific types or locations. It provides insights into vegetation dynamics in the context of both space and time.