| Summary: | According to theory, the functional traits of species dictate how environmental selection affects them, and also the functioning of ecosystems that those species form. However, we lack a general understanding about how exactly environmental selection affects the trait composition of communities, and consequently, ecosystem functions. In this thesis, I study how the effects of environmental selection manifest in the functional composition of field-layer plant communities in the tundra and in boreal forests. My aims are 1) to sharpen our understanding about the effects of trait-based selection on plant communities by accounting for the microenvironment in models of trait composition, 2) to elucidate the effects of that selection on tundra carbon cycling, and 3) to reveal how forestry and reindeer husbandry, two forms of human land use, modulate long-term vegetation changes by favouring certain trait combinations over others. The study areas span four tundra landscapes in Finnish Lapland, Greenland, Svalbard, and the southern Indian Ocean, and hundreds of herb-rich boreal forest patches in Northern Finland. I use linear modelling to study how the results of vegetation surveys, visual, sensor-based and laboratory measurements of traits and the environment, and carbon flux chamber measurements relate to each other. My results suggest the following. 1) The environment strongly determines the functional composition of plant communities when accounting for microenvironmental conditions. Warm, ungrazed and unshaded conditions favor larger plants. Leaf traits that confer fast returns on invested resources are favoured in conditions of high soil resource availability, in ungrazed areas, and in the shade. 2) In the tundra, communities consisting of larger plants cycle carbon more rapidly and have larger above-ground carbon stocks. Communities with “fast” leaf traits also cycle carbon with higher intensity, but they have lower above-ground carbon stocks than communities with “slow” leaf traits. 3) In boreal forests, forestry ...
|
|---|