Importance of Snow in Alpine Plant Ecology : And Remote Sensing Applications in Species Modeling

Abstract: Global warming, with pronounced increases toward the poles and high latitudes, has impacts on alpine vegetation. If snow patterns change in response to global warming, snow parameters are predictors of interest in alpine plant species distribution and traits modeling because snow is known to strongly control the local growing conditions of alpine plants. Also, with technical advances some environmental variables can be produced using remote sensing instead of using field measurements. In this thesis, I investigate the role of snow persistence and wintertime snow depth in ecological modeling of alpine vegetation, derived from a mixture of satellite data and field data. The possibilities of using predictor data derived from remote sensing versus field measurements is also investigated in species modeling, with Landsat derived Land Surface Temperature and topography-based Wetness Index in comparison to punctual field measurements of temperature and soil moisture. Species presence, coverage and heights were collected in alpine Vindelfjällen from 90 plots of 1 m2, covering a range of 900-1450 m.a.s.l. and at varying distances to snow patches. A total of 186 models were generated for 10 species, with varying response variables and predictor combinations. The effect snow patches had was species specific. Snow parameters were statistically significant in half of the species distribution models (mostly snow persistence). Most species’ traits were best estimated not using snow parameters. Both temperature and soil moisture collected in the field can be replaced by corresponding predictors generated through remote sensing to create functioning models, which are also better at predicting species distribution and traits.