Esta Dissertação de Mestrado propõe dois novos algoritmos para tornar mais eficiente o escalonamento on-line de tarefas com dependências estritas em agregados de computadores que usam como middleware para troca de mensagens alguma implementação da MPI (até a versão 2.1). Esses algoritmos foram projetados tendo-se em vista programas construídos no modelo de programação fork/join, onde a operação de fork é usada sobre uma chamada recursiva da função. São eles: 1. O algoritmo RatMD, implementado através de uma biblioteca de primitivas do tipo map-reduce, que funciona para qualquer implementação MPI, com qualquer versão da norma. Utilizado para minimizar o tempo de execução de uma computação paralela; e 2. O algoritmo RtMPD, implementado através de um sistema distribuído sobre daemons gerenciadores de processos criados dinamicamente com a implementação MPICH2 (que implementa a MPI-2). Utilizado para permitir execuções de instâncias maiores de programas paralelos dinâmicos. Ambos se baseiam em roubo de tarefas, que é a estratégia de balanceamento de carga mais difundida na literatura. Para ambos os algoritmos apresenta-se modelagem téorica de custos. Resultados experimentais obtidos ficam dentro dos limites teóricos calculados. RatMD provê uma redução no tempo de execução de até 80% em relação ao algoritmo usual (baseado em round-robin), com manutenção do speedup próximo ao linear e complexidade espacial idêntica à popular implementação com round-robin. RtMPD mantém, no mínimo, o mesmo desempenho que a implementação canônica do escalonamento em MPICH2, dobrando-se o limite físico de processos executados simultaneamente por cada nó. / This Master’s Dissertation proposes two new algorithms for improvement on on-line scheduling of dynamic-created tasks with strict dependencies on clusters of computers using MPI (up to version 2.1) as its middleware for message-passing communication. These algorithms were built targeting programs written on the fork-join model, where the fork operation is always called over an recursive function call. They are: 1. RatMD, implemented as a map-reduce library working for any MPI implementation, on whatever norm’s version. Used for performance gain; and 2. RtMPD, implemented as a distributed system over dynamic-generated processes manager daemons with MPICH2 implentation of MPI. Used for executing larger instances of dynamic parallel programs. Both algorithms are based on the (literature consolidated) work stealing technique and have formal guarantees on its execution time and load balancing. Experimental results are within theoretical bounds. RatMD shows an improvement on the performance up to 80% when paired with more usual algorithms (based on round-robin strategy). It also provides near-linear speedup and just about the same space-complexity on similar implementations. RtMPD keeps, at minimum, the very same performance of the canonical MPICH2 implementation, near doubling the physical limit of simultaneous program execution per cluster node.

A trinucleotide repeat (TNR) is a 3 base pair (bp) DNA sequence tandemly repeated in an array. In humans, TNR sequences have been found to be associated with at least 14 severe neurological diseases including Huntington disease, myotonic dystrophy and several of the spinocerebellar ataxias. Such diseases are caused by an expansion of the repeat sequence beyond a threshold length and are characterized by non-Mendelian patterns of inheritance which lead to genetic anticipation. Although the mechanism of the genetic instability in these arrays is not yet fully understood, various models have been suggested based on the in vitro observation that TNR sequences can form secondary structures such as pseudo-hairpins. In order to investigate the mechanisms responsible for instability of TNR sequences, a study was carried out on Escherichia coli cells in which TNR arrays had been integrated into the chromosomal lacZ gene. This genetic assay was used to identify proteins and pathways involved in deletion and/or expansion instability. Deletion instability was clearly dependent on orientation of the TNR sequence relative to the origin of replication. Interestingly, it was found that expansion instability is not dependent on the orientation of the repeat array relative to the origin of replication. The replication fork reversal pathway and the RecFOR mediated gap repair pathway were found to have no statistically significant influence on the instability of TNR arrays. However, the protein UvrD was found to affect the deletion instability of TNR sequences. The roles of key helicase genes were investigated for their effects on instability of chromosomal CTG•CAG repeats. Mutation of the rep gene increased deletion in the CTG leading-strand orientation of the repeat array, and expansion in both orientations - destabilizing the TNR array. RecQ helicase was found to have a significant effect on TNR instability in the orientation in which CAG repeats were present on the leading-strand relative to the origin of replication. Mutation of the recQ gene severely limited the number of expansion events in this orientation, whilst having no effect on deletions. This dependence of expansions on RecQ was lost in a rep mutant strain. In a rep mutant expansions were shown to be partially dependent on the DinG helicase. All together, these results suggest a model of TNR instability in which expansions are due to events occurring at either the leading or lagging strand of an arrested replication fork, facilitated by helicase action. The identity of the helicase implicated is determined by the nature of the arrest.

Shifting perspectives on restoration and management of public lands in the inland West have resulted in an increased need for maps of potential natural vegetation which cover large areas at sufficient scale to delineate individual stands . In this study, classification tree modeling was used to predictively model and map the plant association types of a relatively undisturbed wilderness area in the Blue Mountains of northeastern Oregon. Models were developed using field data and data derived from a geographic information system database. Elevation, slope, aspect, annual precipitation, solar radiation, soil type, and topographic position were important predictor variables. The model predicted plant association types with a relatively high degree of accuracy for most plant association types, with the lowest accuracy for the types within the grand fir series. Fuzzy confusion analysis was used to analyze model performance, and indicated the overall model accuracy was 72%.

The majority of interior cutthroat trout (Oncorhynchus clarki) subspecies have been extirpated from large rivers by anthropogenic activities that have fragmented habitats and introduced non-native competitors. Selective pressures against migratory behaviors and mainstream river occupation and conservation schemes that isolate genetically pure populations above barriers have restricted gene flow and prevented the expression of fluvial life history traits in many populations. Existing knowledge about the movements and home range requirements of fluvial cutthroat trout is therefore limited. We implanted a total of 55 Bonneville cutthroat trout (BCT) in the Thomas Fork River, Idaho, with radio transmitters and located them weekly or bimonthly from October to April of both 1999/2000 and 2000/2001. Half of these fish were located above a seasonal diversion barrier and half were located below. We found fish to be more mobile than previously reported. Individuals located above the diversion barrier in 2000/2001 occupied significantly larger home ranges (median 3,675 m, range 2,500-8,900 m) and moved more frequently (mean 0.89 movements/contact, range 0.57-1.0) than other fish. Fish occupied habitats in the lower Thomas Fork and Bear River during the winter that were marginal or uninhabitable during other seasons. During the spring of both years we located fish in both upstream and neighboring tributaries up to 84 km away from our study site. Our results document the existence of a fluvial component of BCT in the Bear River and its tributaries and suggest that successful efforts at conservation of these fish will focus on mainstream habitats and the maintenance of seasonal migration corridors.

The purposes of this paper are threefold: 1) To provide a justification for the increased use of native plants in mountain land development, not only in northeastern Utah, but in the entire western United States. 2) To investigate and delineate what the physical needs of a plant materials list are. 3) To provide a nearly comprehensive list of basically unused plants native to northeastern Utah and to highlight their landscaping potentials based on the earlier investigation. Also touched upon, to help landscape architectural students and professionals alike, are two cursory case studies and a chapter which deals briefly with the problems of nursery supply of native plants and the potential and future research.

Movement patterns and habitat use of Bonneville cutthroat trout (Oncorhynchus clarki Utah) in tributaries of the Logan River watershed are greatly affected by habitat alterations created by North American Beaver (Castor canadensis). Evaluation of cutthroat trout habitat use in these watersheds is also complicated by biotic interactions with invasive brown trout (Salmo trutta) and brook trout (Salvelinus fontinalis). My objectives in this thesis were to 1.) Evaluate the passage of beaver dams by each trout species in the Temple Fork watershed and 2.) Evaluate the habitat use of cutthroat trout in the presence of brown trout and brook trout over a range of spatial scales. To address these objectives, 1381 trout were fitted with passive integrated transponder tags. Their locations were recorded using a combination of annual capture/recapture surveys, stationary in-stream antennas, and monthly continuous mobile antenna surveys. To address objective 1, fish were located above and below 22 beaver dams to establish whether fish passed dams and to identify downstream and upstream passage; 187 individual trout were observed making 481 passes of all 22 beaver dams. Native Bonneville cutthroat trout passed dams more frequently than both non-native brown trout and brook trout. It was determined that spawning timing affected seasonal changes in dam passage for each species. Physical characteristics of dams such as height and upstream location affected the passage of each species. Movement behaviors of each trout species were also evaluated to help ex- plain dam passage. These data suggest beaver dams are not acting as barriers to movement for cutthroat and brook trout but may be impeding the movements of invasive brown trout. To address objective 2, a hierarchical classification of stream habitat was created using the River Styles framework. The River Styles framework not only establishes a relationship between habitats at different scales, but also attempts to understand the processes that create and maintain those habitats. The location of each fish observation was associated with habitats at the stream, landscape unit, River Style, and geomorphic unit levels. Habitat use of each species of fish was evaluated at each spatial scale using all fish observations over the four-year study period. Hotspot locations, or locations used consistently by a species of fish consistently through time, were also evaluated across the entire study period and over each year. It was found that brook trout almost exclusively use the beaver ponds and beaver-altered habitats in Spawn Creek. Brown trout were not found in the upper- most parts of both Spawn Creek and Temple Fork. They were also found more than the other species in pools created by bedrock or man-made control features, suggesting that they select highly stable habitats. Cutthroat trout were found more than brown trout in beaver-altered habitats and lateral scour pools, suggesting that they select more dynamic, naturally occurring habitats.

The landscape evolution of the northeastern Uinta Mountains and the manner in which climatic and tectonic forcing have influenced it are not well constrained. Surficial deposits covering ~325 km2 below the glacial termini in the Henrys Fork and ~50 km2 along the Green River in western Browns Park were mapped at 1:24,000 scale to develop a Quaternary stratigraphic framework for the northeastern Uinta Mountains. The Henrys Fork mapping area spans from late Wisconsinan moraines to Flaming Gorge Reservoir. The Henrys Fork stratigraphy contains 10 mainstem gravels, six piedmont gravels, and landslide deposits. Terraces preserved along the Henrys Fork converge downstream and are strath terraces underlain by clast-supported, cobble gravel derived from the Uinta Mountain Group and Paleozoic limestone units. The Henrys Fork terrace stratigraphy was correlated to the Wind River terrace stratigraphy for age control, and incision rates were estimated at 80-110 m/m.y. The Browns Park mapping area includes Little Hole and continues through lower Red Canyon into westernmost Browns Park, ending at the Warren Draw-Swallow Canyon quadrangle boundary. The Browns Park stratigraphy includes eight mainstem gravels, five piedmont gravels, and various landslide, colluvial, and eolian deposits. A tuffaceous bed with Lava Creek Bash (640 ka) was identified near the top of a deposit at Little Hole that was previously mapped as Miocene basin fill. Minimum Green River incision rates were estimated between 90 and 115 m/m.y. using the Lava Creek Bash for age control. These rates are comparable to estimates for the Henrys Fork, but are about half of the rates reported for the south flank of the Uintas and other central Rocky Mountain ranges. A series of three distinct deposits in western Browns Park are interpreted as evidence for the landslide impoundment and subsequent outburst flooding of the Green River. These include slackwater deposits at Little Hole, an outburst flood deposit in western Browns Park, and a large paleolandslide deposit that lies between them. Estimates of sediment accumulation rates behind the paleolandslide dam suggest it was stable for ~605 years. Peak discharge estimates from impounded water volume estimates and paleoflow competence indicators suggest that the resulting outburst flood was ~22,000 m3/s.

<p><p><p>This thesis aims toward developing strategies in the area of structural optimization and to implement these strategies in design processes. At</p><p> </p><em>GM Powertrain Sweden </em>where powertrains are designed and developed, two designs of a differential housing have been chosen for this thesis. The main tasks have been to perform a topology optimization of a model early in a design process, and a shape optimization on a model late in a design process. In addition the shape optimization strategies have also been applied on a fork shifter. This thesis covers the theory of different optimization strategies in general. The optimization processes are explained in detail and the results from the structural optimization of the differential housings as well as the fork shifter are shown and evaluated. The evaluation of the thesis provides enough arguments to suggest an implementation of the optimization strategies in design processes at <em>GM Powertrain</em><p>. A Structural Optimization group has great potential of closing the gap between structural designers and structural analysis engineers which in long terms mean that better structures can be developed in less time. To be competitive in the automotive industry these are two of the most important factors for being successful.</p></p></p>

The aim of Part One was to discover and investigate the physical and chemical properties of coal minesoils occurring within the Big South Fork National River and Recreation Area (BSF) in order to better understand the nature of these anthropogenic soils. In particular, this information was to identify which, if any, un-reclaimed or recently reclaimed minesoils were actively producing acid mine drainage (AMD) and what properties were inhibiting revegetation. Historical knowledge and maps were used to locate more than 30 un-reclaimed and reclaimed minesoil sites, which were mapped with GPS. Soil profiles were exposed on 18 sites and grab samples taken on another 12. The morphological properties of each full profile were described according to the National Soil Survey Handbook and samples were taken from each horizon. Chemical properties analyzed for include: particle size, acid-base account, pH, exchangeable aluminum, manganese oxides, soil organic carbon, cation-exchange capacity, exchangeable bases, Mehlich I-extractable elements, and total elemental concentrations. Significant differences in the following properties (averaged) were discovered between the un-reclaimed and reclaimed minesoils: slope, percent rock fragments, dominant lithology, net neutralization potential, pH, extractable aluminum, base saturation, several Mehlich I-extractable nutrients and total elemental concentrations. Hierarchical clustering analysis revealed similar findings and also highlighted instances where reclaimed minesoils were statistically more similar to un-reclaimed minesoils than to other reclaimed minesoils. This indicated that reclamation efforts may not have been completely successful on these sites. In Part Two, minesoil profiles were classified according to Soil Taxonomy and according to proposed amendments by the International Committee for Anthropogenic Soils (ICOMANTH). The ICOMANTH amendments provided more informative classifications for coal minesoils in the BSF however, shortcomings were noted. Additional recommendations were made and the minesoils were again classified according to these recommendations. Compared to both the Soil Taxonomy and the ICOMANTH classifications, those according to the proposed additional amendments revealed more of the unique properties of the minesoils studied in this project. The results of this study can aid the National Park Service with future land management of the minesoils located within the BSF boundaries and other users of drastically disturbed minesoils.

<p>Rivers interact with their valleys from headwaters to mouth, but nowhere as dynamically as in their floodplains. Rivers deliver water, sediments, and solutes onto the floodplain land surface, and the land in turn supplies solutes, leaves, and woody debris to the channel. These reciprocal exchanges maintain both aquatic and terrestrial biodiversity and productivity. In this dissertation I examine river-floodplain exchanges on the well-studied Nyack Floodplain, a dynamic, gravel-bedded floodplain along the Middle Fork Flathead River in the mountains of northwest Montana. I quantify exchanges at multiple timescales, from moments to centuries, to better understand how connectivity between aquatic and terrestrial habitats shapes their ecology.</p><p>I first address connectivity in the context of a long-standing question in ecosystem ecology: What determines the rate of ecosystem development during primary succession? Rivers have an immediate effect on floodplains when scouring floods remove vegetation and nutrients such as nitrogen (N) and leave only barren soils, but they might also affect the ensuing primary succession through the gradual delivery of N and other materials to floodplain soils. I quantify N inputs to successional floodplain forest soils of the Nyack Floodplain and find that sediment deposition by river flood water is the dominant source of N to soils, with lesser contributions from dissolved N in the river, biological N fixation, and atmospheric deposition. I also synthesize published rates of soil N accumulation in floodplain and non-floodplain primary-successional systems around the world, and I find that western floodplains often accumulate soil N faster than non-floodplain primary successional systems. My results collectively point to the importance of riverine N inputs in accelerating ecosystem development during floodplain primary succession.</p><p>I next investigate the role of river-floodplain exchanges in shaping the spatial distribution of a suite of soil properties. Even after flood waters have receded, dissolved N, carbon (C), and moisture could be delivered from the river to floodplain soils via belowground water flow. Alternatively, C inputs and N withdrawals by floodplain vegetation might be a dominant influence on soil properties. To test these hypotheses, I excavated and sampled soil pits from the soil surface to the water table (50-270 cm) under forests, meadows, and gravel bars of the Nyack Floodplain. Near-surface soils had C and N pools and N flux rates that varied predictably with vegetation cover, but soil properties below ~50 cm reflected influence by neither vegetation cover nor aquifer delivery. Instead, soil properties at these depths appear to relate to soil texture, which in turn is structured by the river's erosional and depositional activities. This finding suggests the revised hypothesis that soil properties in gravel-bedded alluvial floodplains may depend more on the decadal-scale geomorphic influences of floods than on short-term vertical interactions with floodplain vegetation or aquifer water. </p><p>Lastly, I explore the potential sources of organic C to the diverse and active community of aquatic organisms in the floodplain aquifer, where the lack of light prohibits in-situ organic C production by photosynthesis. I quantify floodplain carbon pools and the fluxes of organic carbon connecting the aquifer, river, and overlying forest. Spring flood waters infiltrating the soil are responsible for the largest dissolved carbon flux into the aquifer, while very large floods are essential for the other major C input, the burial of woody carbon in the aquifer. These findings emphasize the importance of a dynamic river hydrograph - in particular, annual floods and extreme annual floods - in delivering organic C to the aquifer community. </p><p>Overall, this dissertation draws our attention not just to the current exchanges of C, N, water, and sediment but to the episodic nature of those exchanges. To fully understand floodplain ecosystems, we have to consider not just present-day interactions but also the legacies of past floods and their roles in delivering solutes, eroding forests, depositing sediments, and physically shaping the floodplain environment.</p> / Dissertation