Accurate knowledge of stellar parameters such as mass, radius, effective temperature, and composition inform our understanding of stellar evolution and constrain theoretical models. Binaries and, in particular, eclipsing binaries make it possible to measure directly these parameters without reliance on models or scaling relations. In this dissertation we derive fundamental parameters of stars in close binary systems with and without (detected) tertiary companions to test and inform theories of stellar and binary evolution. A subsample of 41 detached and semi-detached short-period eclipsing binaries observed by NASA’s Kepler mission and analyzed for eclipse timing variations form the basis of our sample. Radial velocities and spectroscopic orbits for these systems are derived from moderate resolution optical spectra and used to determine individual masses for 34 double-lined spectroscopic binaries, five of which have detected tertiaries. The resulting mass ratio M2/M1 distribution is bimodal, dominated by binaries with like-mass pairs and semi-detached classical Algol systems that have undergone mass transfer. A more detailed analysis of KIC 5738698, a detached binary consisting of two F-type main sequence stars with an orbital period of 4.8 days, uses the derived radial velocities to reconstruct the primary and secondary component spectra via Doppler tomography and derive atmospheric parameters for both stars. These parameters are then combined with Kepler photometry to obtain accurate masses and radii through light curve and radial velocity fitting with the binary modeling software ELC. A similar analysis is performed for KOI-81, a rapidly-rotating B-type star orbited by a low-mass white dwarf, using UV spectroscopy to identify the hot companion and determine masses and temperatures of both components. Well defined stellar parameters for KOI-81 and the other close binary systems examined in this dissertation enable detailed analyses of the physical attributes of systems in different evolutionary stages, providing important constraints for the formation and evolution of close binary systems.

Hobby activities can be viewed through the lens of informal, free-choice learning. A wide range of hobbies combine fun and learning-intensive practices, and can contribute to scientific literacy. Hobby learning involves clear goal orientation, persistence and effort, and often results in more richly and strongly connected knowledge; traits highly valued in both in and out-of-school science learning. In this study, I used koi hobbyists as subjects to discover and explore hobbyists' information-seeking strategies under different learning scenarios. I approached koi hobbyists' learning about koi and their koi hobby in both quantitative and qualitative ways. I designed a Stage of Engagement Model to illustrate koi hobbyists' engagement with their hobby, and adapted Falk and Dierking's Contextual Model of Learning to explain how personal, socio-cultural and physical contextual factors affect koi hobbyists' learning. An instrument was developed to assess koi hobbyists' experience with keeping koi, knowledge about the hobby, motivation/goals, interaction with other hobbyists, and the information-seeking strategies they used under different learning scenarios. I administered this questionnaire to koi hobbyist communities in the U.S. Pacific Northwest and online. Based on the quantitative analysis, the results supported my hypotheses that koi hobbyists chose different information-seeking strategies based on personal contextual factors such as previous experience, motivation and learning goals; socio-cultural contextual factors such as interactions with other koi hobbyists; and physical contextual factors such as the nature of the problems they encounter. Koi hobbyists also chose different information-seeking strategies based upon their stage of engagement with their hobby. The long-term potential of this study is to offer insights into how learners construct their knowledge by applying different learning strategies under different personal, socio-cultural and physical circumstances, and to provide a framework for the future study of other kinds of hobbies and hobbyists that will help to promote public scientific literacy. / Graduation date: 2012

The expanding aquaculture industry continues to encounter major challenges in the form of highly contagious aquatic viruses. Control and eradication measures targeting the most lethal and economically damaging virus-induced diseases, some of which are notifiable, currently involve ‘stamping out’ policies and surveillance strategies. These approaches to disease control are performed through mass-culling followed by restriction in the movement of fish and fish products, resulting in considerable impacts on trade. Although effective, these expensive, ethically complex measures threaten the sustainability and reputation of the aquatic food sector, and could possibly be reduced by emulating innovative vaccination strategies that have proved pivotal in maintaining the success of the terrestrial livestock industry. DIVA ‘differentiating infected from vaccinated animal’ strategies provide a basis to vaccinate and contain disease outbreaks without compromising ‘disease-free’ status, as antibodies induced specifically to infection can be distinguished from those induced in vaccinated animals. Various approaches were carried out in this study to assess the feasibility of marker/DIVA vaccination for two of the most important disease threats to the global Atlantic salmon and common carp/koi industries, i.e. infectious salmon anaemia (ISA) and koi herpesvirus disease (KHVD), respectively. Antibody responses of Atlantic salmon (Salmo salar L.), following immunisation with an ISA vaccine, administered with foreign immunogenic marker antigens (tetanus toxoid (TT), fluorescein isothiocyanate (FITC) and keyhole limpet hemocyanin (KLH)) were assessed by antigen-specific enzyme linked immunosorbent assay (ELISA). Although antibodies were induced to some markers, these were unreliable and may have been affected by temperature and smoltification. Detectable antibodies to ISAV antigen were also largely inconsistent despite low serum dilutions of 1/20 being employed for serological analysis. The poor antibody responses of salmon to the inactivated ISA vaccine suggested that DIVA vaccination is not feasible for ISA. A similar approach for KHV, utilising green fluorescent protein (GFP) as the marker, similarly failed to induce sufficiently detectable antibody responses in vaccinated carp (Cyprinus carpio L.). However, as high anti-KHV antibody titres were obtained with an inactivated KHV vaccine (≥1/3200), alternative approaches were carried out to assess the feasibility of DIVA vaccination for carp. Investigations of early KHV pathogenesis in vivo and antigen expression kinetics in vitro (0-10 days post infection (dpi)) provided valuable data for the diagnostics necessary for DIVA surveillance strategies. Following viral infection, molecular methods were shown to be the most effective approach for early detection of KHV infected fish prior to sero-conversion, during which time antibodies are not detectable. An experimental immersion challenge with KHV, however, revealed complications in molecular detection during early infection. The KHV DNA was detected in external biopsies of skin and gills, but also internally in gut and peripheral blood leukocytes ≤ 6 hours post infection (hpi), suggesting rapid virus uptake by the host. The gills and gut appeared to be possible portals of entry, supported by detection of DNA in cells by in situ hybridisation (ISH). However, many false negative results using organ biopsies occurred during the first 4 dpi. The gills were the most reliable lethal biopsy for KHV detection by various polymerase chain reaction (PCR) assays, with a PCR targeting a glycoprotein-gene (ORF56) and a real-time PCR assay being the most sensitive of the 7 methods investigated. Importantly, non-lethal mucus samples reduced the number of false negative results obtained by all KHV PCR assays during the earliest infection stages with large levels of viral DNA being detected in mucus (up to 80,000 KHV DNA genomic equivalents 200 μL-1). KHV DNA was consistently detected in the mucus as a consequence of virus being shed from the skin. Determining the expression kinetics of different viral structural proteins can be useful for DIVA serological tests. Analysis of KHV antigen expression in tissues by immunohistochemistry and indirect fluorescent antibody test was inconclusive, therefore 2 novel semi-quantitative immunofluorescence techniques were developed for determining KHV antigen expression kinetics in susceptible cell lines. During the course of KHV infection in vitro, a greater abundance of capsid antigen was produced in infected cells compared to a glycoprotein antigen (ORF56), as determined by detection with antigen-specific monoclonal antibodies (MAbs). The capsid antigen was characterised as a ~100 kDa protein by SDS-PAGE and identified as a product of KHV ORF84 by matrix-assisted laser desorption ionisation time-of-flight mass spectrometry (MALDI-TOF/TOF MS). This antigen was subsequently detected in the serum of >25% of KHV infected/exposed carp (6/17), as well as in carp vaccinated with a live attenuated vaccine (3/4), but not with an inactivated vaccine (0/7), by Western blot making it a potential DIVA target for an inactivated vaccine. Attempts were made to improve the sensitivity of KHV serological testing by taking advantage of recombinant proteins specific for KHV (CyHV-3), rORF62 and rORF68 and eliminating any interference by cross-reacting antibodies to carp pox (CyHV-1). These proteins successfully reacted with anti-KHV antibodies. The feasibility of DIVA strategies for KHVD was determined using these recombinant antigens to coat ELISA plates. Differential antibody responses were detected from carp sera to an internal virus tegument protein (rORF62) and external region of a transmembrane protein (rORF68). Fish vaccinated with an inactivated vaccine produced significantly lower antibody responses to rORF62 than to rORF68, whereas infected, exposed and live attenuated vaccinated fish recognised both proteins allowing differentiation between vaccinated and infected carp. However, the sensitivity of the assay was limited, possibly by high levels of natural antibodies detected at the relatively low serum dilutions (1/200) used. As the capsid antigen (ORF84) and tegument protein (ORF62) are derived from internal KHV structural proteins, they induce non-neutralising antibodies, which may be useful for DIVA strategies. Such antibodies are longer lasting than neutralising antibodies and often comprise the majority of fish anti-viral antibodies. This was noted in a fish surviving experimental challenge, which had an antibody titre of 1/10,000, but neutralising titre of 1/45. Such antigens may therefore hold potential for developing effective serological diagnostic tests for KHV and provide the potential for DIVA strategies against KHVD. Natural antibodies will, however, continue to present a challenge to the development of sensitive and reliable KHV serological tests, and hence the application of DIVA strategies.