In recent years considerable interest has been shown in the diversity of ammonia-oxidizing bacteria in soil communities. The majority of the research has been carried out in Northern Europe where soils have received high atmospheric inputs of nitrogen over the past two centuries. In contrast, although much work has been conducted on nitrogen cycling processes in nitrogen limited forest ecosystems in western North America, no studies have examined the characteristics of ammonia-oxidizing communities in those environments. I was interested in measuring nitrification potential along a high-elevation temperate meadow-to-forest gradient, and characterizing the ammonia-oxidizing communities along that gradient using both molecular and culturing methods. Two experimental sites (Lookout and Carpenter) were chosen in the H.J. Andrews Experimental Forest, located in the western Cascade Range of Oregon, at elevations of approximately 1500 meters. Although nitrification potential rates (NPRs) between sites were not significantly different (P=0.544), variation was observed both within and between sites for specific vegetation types. NPRs were significantly lower in forest (F) soil samples than in meadow (M) soil samples, averaging 5 and 2% of meadow NPRs at Lookout and Carpenter, respectively. In meadow soil samples, most probable number (MPN) population densities of ammonia-oxidizers ranged from 0.6 to 2.6 x 10⁴ cells gram⁻¹ of oven dry soil and 0.9 x 10³ to 1.1 x 10⁵ cells g⁻¹ OD soil at Lookout and Carpenter, respectively. In forest soil samples, population densities ranged from undetectable to 1.1 x 10⁴ cells g⁻¹ OD soil, and 0.9 x 10² to 2.3 x 10³ cells g⁻¹ OD soil at Lookout and Carpenter, respectively. Microbial community DNA was amplified using primers to the ammonia monooxygenase subunit A. Terminal restriction fragments polymorphism analysis with three different restriction enzymes (CfoI, TaqI, and AluI) revealed community profiles dominated by Nitrosospira species. One fragment from CfoI (66 bp) and one fragment from AluI (392-bp) were prominent in 47 soil samples from both sites, and represented between 32 to 100% of the Genescan fragment analyses of PCR products. A full length fragment from AluI digests (491-bp), and three fragments from CfoI (68, 100, and 135- bp) were found sporadically in fewer soil sample T-RFLPs, and within those samples represented smaller percentages of total peak areas. The CfoI 135-bp fragment length was associated primarily with M and meadow/forest (M/F) soils where it was observed in approximately 58 and 100% of the respective transect locations. Eight isolates recovered from soil samples were analyzed using the same molecular methods as the field samples. The T-RFLP patterns of the isolates corresponded with many of those found in the community fingerprints. Four unique amoA sequences were identified among these isolates, including one that possessed the dominant T-RFLP amoA fingerprint in soil samples. This sequence shared 99.8% similarity with Nitrosospira sp. Ka4, a cluster 4 ammonia oxidizer isolated in Norway. Sequence analysis phylogenetically associated the other three isolates (with unique amoA sequences) near Nitrosospira sp. Nsp 1 and Nitrosospira briensis, both cluster 3 ammonia oxidizers. Cloning and sequencing of soil DNA confirmed that ammonia oxidizers with these amoA sequences were present in the soil samples. Two additional amoA sequences were identified in clones that were 95% similar and paraphylogenetically positioned between representatives of clusters 3 and 4. So far, these sequences have not been found in any of the isolates analyzed. / Graduation date: 2003

A biological inventory of the butterflies of the H. J. Andrews Experimental Forest [HJA] in Linn/Lane County, Oregon was conducted during 1994 and 1995. It was the first comprehensive survey of HJA butterflies for the site and serves as a baseline for future butterfly research. A detailed ecological account is provided for each species documented during this and previous studies from the HJA. Patterns of butterfly richness and abundance are addressed both temporally and spatially. Within-year and between year differences in butterfly richness and abundance are explained. Butterfly richness and abundance were compared between forest, clear-cut, and meadow habitats, as well as along the roads within these habitats. Butterfly richness and abundance comparisons were also made between local butterfly hotspots and immediately adjacent areas. Lastly, the HJA butterfly fauna was compared to those of five other Oregon sites to put it into a regional perspective. Seventy-two species were recorded during this two year period and increased the total documented butterfly fauna of the HJA to seventy-nine species. Butterfly species richness was high from June through early August. Butterfly abundance increased gradually over the season and peaked in early August. Each butterfly species displayed one of four patterns of combined relative abundance and distribution: common and widespread, rare and local, common only at low elevations or common only at high elevations. The results of standardized butterfly counts suggested that subalpine meadows were much higher than clear-cuts or forests in butterfly richness and abundance, and that roads served to increase butterfly richness and abundance on a local scale in most cases. Butterfly hotspots on the HJA appear as relatively small areas of high butterfly richness and abundance and have a correspondingly high number of plant species when compared to adjacent areas. With virtually one-half of all butterfly species known for the state of Oregon, the HJA ranks among the most species-rich locations for its size within the state. This diversity originates from several biogeographical regions of origin, as defined within this study. A total of 31 HJA species have a generalized Western North American distribution, but several other biogeographical regions are also well represented. Some butterfly species appear to be at or near their geographical limits on the HJA. The assemblage of HJA butterfly species is virtually inclusive of those from Mary's Peak and McDonald Forest in northwestern Oregon, whereas it differs by 30% or more from the more biogeographically distinct faunas of Crater Lake National Park, Mount Ashland and Steens Mountain. Future butterfly work on the HJA is recommended. Oregon butterfly distribution maps suggest that several additional butterfly species should be found there. More biogeographical analyses combined with long term monitoring of HJA butterflies could help to both predict and document changes in the Pacific Northwest butterfly fauna due to human disturbance and global climate change. / Graduation date: 2004

Thesis (M.S.)--Oregon State University, 2002. / Typescript (photocopy). Includes bibliographical references (leaves 131-140). Also available on the World Wide Web.

The link between aboveground net primary productivity (ANPP) and resource gradients generated by complex terrain (solar radiation, nutrients, and moisture) has been established in the literature. Belowground ecosystem stocks and functions, such as soil organic carbon (SOC), dissolved organic carbon (DOC), and belowground productivity have also been related to the same topography and resource distributions, and therefore it is expected that they share spatial and temporal patterns with ANPP. However, stand structure on complex terrain is a function of multiple trajectories of forest development that interact with existing resource gradients, creating feedbacks that complicate the relationships between resource availability and ANPP. On a 96 ha forested watershed in the H.J. Andrews Experimental Forest in the Western Cascades range of Oregon, spatiotemporal heterogeneity in the secondary succession of a replanted Pseudotsuga menziesii stand following harvest results from the interaction of stand composition and abiotic drivers and may create unique "hot spots" and "hot moments" that complicate gradient relationships. In this dissertation, I tested the hypotheses that (chapter 3) multiple successional trajectories exist and can be predicted from a general linear model using specific topographic, historical, and biological parameters and that an estimated "maximum ANPP" may better represent stand characteristics than ANPP measured at a particular moment in time. I also test that (chapter 4) the distribution of light fraction carbon (LFC; C with a density of less than 1.85 g/cm��) is spatially variable, elevated on hardwood-initiated sites (hardwood biomass > 50% of biomass), and positively correlated with litter fall and ANPP. Chapter 4 also tests that heavy fraction carbon (HFC; C with a density of greater than 1.85 g/cm��) is a function of both soil mineralogy, stand composition, and ANPP, such that edges observed spatially in site mineralogy (changes in soil type) are reflected in sharp changes in the composition of the forest community and the magnitude of HFC stores. Finally, I hypothesized (chapter 5) that in complex terrain, dissolved organic carbon (DOC) export can be predicted from landform characteristics, relates to ANPP, and may be measured by several methods which are well-correlated with one another. In chapter 6, I discuss how litter fall measurements can be extrapolated to a watershed extent, and use litter fall as an example of the error that can occur in scaling up measurements taken at a small scale, within a heterogeneous stand on complex terrain, to a landscape scale extent. / Graduation date: 2013

Thesis (Ph. D.)--Oregon State University, 2001. / Typescript (photocopy). Includes bibliographical references (leaves 99-110). Also available on the World Wide Web.

Thesis (M.S.)--Oregon State College, 1954. / Typescript. Includes bibliographical references (leaves 50-51). Also available on the World Wide Web.

Spring plant phenology is often used as an indicator of a community response to climate change. Remote data and low-resolution climate models are typically used to predict phenology across a landscape; however, this tends to miss the nuances of microclimate, especially in a mountainous area with heterogeneous topography. I investigated how inter-annual variability in regional climate affects the distribution of microclimates (i.e., areas <100m2) and spring plant phenology across a 6400-hectare watershed within the Western Cascades in Oregon. Additionally, I created species-specific models of bud break at the microclimate scale, that could then be applied across a wider landscape. I found that years with warm winters, few storms and low snowpack have a homogenizing effect on microclimate and spring phenology events, and that bud break models developed at a local scale can be effectively applied across a broader landscape. This thesis includes previously unpublished coauthored material.

ix, 94 p. : ill. A print copy of this thesis is available through the UO Libraries. Search the library catalog for the location and call number. / This thesis examines the Long-Term Ecological Reflections at H.J. Andrews Experimental Forest, Oregon, a project that promotes creative inquiry at an ecological research station. I analyze how reflective writing contributes to a deeper understanding of ecosystem processes and fosters a closer connection to nature, referred to as "ecological awareness." I argue that the greatest ecological awareness comes from a combination of scientific and artistic ways of knowing as well as physical immersion in nature. I critically analyze the work of four writers: Jane Coffey, Vicki Graham, Freeman House, and Robin Kimmerer. These works create a strong connection to place for the reader and facilitate an understanding of the forest ecosystem through a discussion of scientific facts. They also create a sense of place through vivid and metaphorical language. These reflections will become an increasingly valuable mode of inquiry as the Andrews Forest develops a stronger writer-scientists interaction. / Committee in Charge: Dr. Louise V. Westling, Chair; Dr. Brendan Bohannan; Dr. Scott D. Bridgham

Mat-forming ectomycorrhizal (EcM) fungi represent a prevalent constituent of many temperate forest ecosystems and create dramatic changes in soil structure and chemistry. EcM mat soil have been shown to have increased microbial respiration rates and have been hypothesized to harbor unique assemblages of fungi and bacteria. The objectives of this dissertation were to characterize and examine the fungal and bacterial communities associated with EcM mats in old-growth forests of the H.J. Andrews Experimental Forest located in the Oregon Cascades. Additionally, this work assessed the application of traditional, emerging, and novel molecular sampling techniques for determining microbial communities of environmental samples. This research investigated the microbial communities associated with two common EcM mat genera found in old-growth Douglas fir stands in the Pacific Northwest; Piloderma (Atheliales, Basidiomycota) and Ramaria (Gomphales, Basidiomycota). Soil samples were collected from Piloderma and Ramaria mats and surrounding non- mat soil for molecular analysis of nucleic acids. First, a comparative study was conducted to determine the most appropriate rDNA molecular sampling technique for microbial community characterization. Two next-generation sequencing methods, Roche 454 pyrosequencing and Illumina-based environmental sequencing, the latter developed by the author, were compared to a more traditional sequencing approach, i.e., Sanger sequencing of clone libraries. These findings informed the subsequent sampling of the fungal ITS and bacterial 16S rDNA fragment with 454 pyrosequencing to determine the microbial communities within mat and non-mat soils. Second, this work utilized a pyrosequencing approach to explore fungal community structure in EcM mat and non-mat soils. This work concluded that differences in microbial communities do exist between Piloderma mat, Ramaria mat, and non-mat soils, but the differences are largely quantitative with relatively few distinct taxonomic shifts in microbial constituents. Piloderma, Ramaria and Russula, in addition to being the dominant taxa found on mycorrhizal root tips, were found to be the most abundant taxa in bulk soils within their respective mat types or non-mat sample. The background fungal communities within the EcM mats in this study exhibited considerable taxonomic overlap with the exception of Piloderma vs. non-mat comparisons; Russula species dominated nonmat soils but tended to be excluded or significantly underrepresented in Piloderma mats. Lastly, this study explored the bacterial communities associated with Piloderma and Ramaria mats using lower- coverage 454-Jr pyrosequencing. Bacterial communities exhibited significant structure as a function of mat-type, soil horizon and pH, but this finding should be interpreted with respect to the nonrandom distribution of Piloderma-mats in the O- horizon and the Ramaria-mats in the A-horizon, and the tendancy for EcM mats to be more acidic than surrounding soils. Nonetheless, the total microbial (bacterial and fungal) community was typically dominated by the mat-forming taxa, or Russula, in the case of non-mat soils. While the presence of Piloderma mats did enrich or restrict some bacterial groups, soil pH was also found to be a significant driver of bacterial richness and taxonomic diversity. Fungal and bacterial richness were also found to be positively related to one another, regardless of soil horizon or EcM mat type. This work, taken together, contributes to the understanding of hyperdiversity and heterogeneity of microbial communities of temperate forest soils and highlights the potential for fungal and bacterial communities to be influenced by the presence of EcM mats. / Graduation date: 2012

Soil organic matter (SOM) is the terrestrial biosphere's largest pool of organic carbon (C) and is an integral part of C cycling globally. Soil organic matter composition typically can be traced directly back to the type of detrital inputs; however, the stabilization of SOM results as a combination of chemical recalcitrance, protection from microbial decomposition within soil structure, and organo-mineral interactions. A long-term manipulative field experiment, the Detrital Input and Removal Treatment (DIRT) Project, was established to examine effects of altering detrital inputs (above- vs. below-ground source, C and nitrogen (N) quantity, and chemical quality) on the stabilization and retention of SOM. Surface mineral soil was collected from two DIRT sites, Bousson (a deciduous site in western Pennsylvania) and H.J. Andrews (a coniferous site in the Oregon Cascade Mountains), to examine the influence of altering detrital inputs on decomposability and mean residence time of soil organic matter and different organic matter fractions. Soil organic matter was physically separated into light fraction (LF) and heavy fraction (HF) organic matter, by density fractionation in 1.6 g mL⁻¹ sodium polytungstate (SPT). Density fractionation in SPT resulted in the mobilization and loss of ~25% of total soil organic C and N during the physical separation and rinsing of fractions during recovery, which was also the most easily decomposed organic matter present in the bulk soil. At H.J. Andrews, this mobilized organic matter had a short mean residence time (MRT), indicating that it originated from fresh detrital inputs. In contrast, at Bousson, the organic matter mobilized had a long MRT, indicating that it originated from organic matter that had already been stabilized in the soil. Mean residence times of LF from Bousson varied widely, ~3 y from doubled litter and control plots and 78-185 y for litter removal plots, while MRT of HF was ~250 y and has not yet been affected by litter manipulations. Results from long term incubation of LF and HF material supported these estimates; respiration was greatest from LF of doubled litter and control plots and least from HF of litter removal plots. In contrast, MRT estimated for LF and HF organic matter from H.J. Andrews were similar to each other (~100 y) and were not affected by litter manipulation. These estimates were also supported by the incubation results; there was not a difference in cumulative respiration between detrital treatments or density fractions. The results from the coniferous site may be due to a legacy of historically large inputs of coarse woody debris on the LF and it may be decades before the signal of detrital manipulations can be measured. Alternatively, these highly andic soils may be accumulating C rapidly, yielding young HF ages and C that does not differ substantially in lability from coniferous litter-derived LF. The DIRT Project was intended to follow changes in soil organic matter over decades to centuries. As expected, manipulation of detrital inputs has influenced the lability and mean residence time of the light fraction before the heavy fraction organic matter; however, it will be on much more lengthy time scales that clear differences in organic matter stabilization in response to the alteration of detrital inputs will emerge. Soil CO₂ efflux is a compilation of CO₂ from many sources, including root respiration and the decomposition of different organic matter fractions, roots, and exudates. If the sources of CO₂ have different isotopic signatures, the isotope analysis of CO₂ efflux may reveal the dominant sources within the soil profile. In a short incubation experiment of density fractions from both sites, respired CO₂ reflected the isotopic signature of the organic matter fraction after 30 days, but was more enriched in ¹³C. Initially CO₂ was isotopically depleted in ¹³C relative to the organic matter fraction and the period of depletion related to the amount of easily degraded organic matter present at H.J. Andrews only. / Graduation date: 2006