Forests and forest soils are some of the largest biologically active carbon reservoirs in the world. Therefore, understanding how disturbances, such as forest harvest, influence biogeochemical cycling is particularly important for managing forests sustainably. Timber harvest can have large impacts on forest soils, which may affect the long-term productivity and function of forest ecosystems. Soil microbes are crucial regulators of biogeochemical cycling, and they can be impacted by disturbance of soil organic matter and compaction caused by logging. Despite their vital roles in ecosystem function, few studies have examined both microbial activity and community composition in response to forest harvest, and the link between microbial function and composition remains unclear. The goal of this thesis was to examine the impacts of different harvest intensities on microbial biogeochemical activity and community structure of prokaryotic and fungal communities in a Douglas-fir forest in Oregon. Samples were collected 3 months and 3 years post-harvest in order to determine the short-term impacts of forest harvest on the soil microbial community and to help elucidate the links between microbial function and structure in forest soils.
The first objective was to compare microbially mediated nutrient cycling among harvest treatments with different levels of organic matter removal and compaction through examination of extracellular enzyme profiles and carbon and nitrogen mineralization. Enzyme activities generally differed little among harvest treatments but varied more through time. The only significant differences in enzyme activities among treatments occurred 3 years post-harvest: the activities of β-glucosidase, cellobiohydrolase, and peroxidase were elevated in harvest treatments compared to the unharvested reference. Carbon and nitrogen mineralization were determined using soil microcosm incubations. Respiration (carbon mineralization) did not differ among treatments 3 months post-harvest and only small differences were observed among treatments 3 years post-harvest. By contrast, harvest treatments typically had greater nitrate and total mineralized N than the reference both 3 months and 3 years post-harvest. As a whole, changes in activity took longer than 3 months to manifest with some differentiation in activity between harvest treatments visible after 3 years. Furthermore, forest harvest appeared to be the main impact on activity rather than organic matter or soil compaction manipulations.
The second objective was to examine soil microbial communities and identify any alterations in diversity and structure of these communities as a result of different harvest intensities. Communities were examined by extracting DNA from soils and performing amplicon sequencing using Illumina Miseq. Timber harvesting with different levels of organic matter removal and soil compaction led to distinct differences in prokaryotic and fungal communities. Microbial composition and structure varied more between years and compared to the unharvested reference than among harvest treatments. Relative abundance of many bacterial phyla, including the dominant Proteobacteria, Acidobacteria, and Verrucomicrobia, was significantly altered among harvest treatments 3 years post-harvest but not immediately after harvest. Less abundant phyla such as Nitrospirae (bacterial) and Thaumarchaeota (archaeal) were also impacted by harvest. In the fungal communities of harvest treatments, Basidiomycota abundance decreased whereas Ascomycota abundance increased. The unharvested reference had significantly greater ectomycorrhizal fungi, and harvest treatments had an enrichment of saprotrophs.
Finally, alterations in microbial activity were compared with changes in community structure to help identify impacts of harvest on soil microbes that may affect nutrient cycling and long-term productivity. Enzyme activity was generally correlated with bacterial communities more than fungal. Increases in soil nitrate and total mineralized nitrogen were attributed to increased abundance of nitrogen cycling autotrophs, including Nitrospirae and Thaumarchaeota. Neither bacterial nor fungal communities were well correlated with alterations in carbon cycling parameters such as respiration and fast and slow-cycling carbon pool parameters. Overall, concurrent changes in activity and community structure imply that alteration in community structure did impact community function. Although clear links between activity and community composition were difficult to discern, our results suggest that some bacterial and fungal groups have the potential to be used as indicators of disturbance and ecosystem status, at least in the short-term.