Parallel geostatistics for sparse and dense datasets

Very large spatially-referenced datasets, for example, those derived from satellite-based sensors which sample across the globe or large monitoring networks of individual sensors, are becoming increasingly common and more widely available for use in environmental decision making. In large or dense sensor networks, huge quantities of data can be collected over small time periods. In many applications the generation of maps, or predictions at specific locations, from the data in (near) real-time is crucial. Geostatistical operations such as interpolation are vital in this map-generation process and in emergency situations, the resulting predictions need to be available almost instantly, so that decision makers can make informed decisions and define risk and evacuation zones. It is also helpful when analysing data in less time critical applications, for example when interacting directly with the data for exploratory analysis, that the algorithms are responsive within a reasonable time frame. Performing geostatistical analysis on such large spatial datasets can present a number of problems, particularly in the case where maximum likelihood. Although the storage requirements only scale linearly with the number of observations in the dataset, the computational complexity in terms of memory and speed, scale quadratically and cubically respectively. Most modern commodity hardware has at least 2 processor cores if not more. Other mechanisms for allowing parallel computation such as Grid based systems are also becoming increasingly commonly available. However, currently there seems to be little interest in exploiting this extra processing power within the context of geostatistics. In this paper we review the existing parallel approaches for geostatistics. By recognising that diffeerent natural parallelisms exist and can be exploited depending on whether the dataset is sparsely or densely sampled with respect to the range of variation, we introduce two contrasting novel implementations of parallel algorithms based on approximating the data likelihood extending the methods of Vecchia [1988] and Tresp [2000]. Using parallel maximum likelihood variogram estimation and parallel prediction algorithms we show that computational time can be significantly reduced. We demonstrate this with both sparsely sampled data and densely sampled data on a variety of architectures ranging from the common dual core processor, found in many modern desktop computers, to large multi-node super computers. To highlight the strengths and weaknesses of the diffeerent methods we employ synthetic data sets and go on to show how the methods allow maximum likelihood based inference on the exhaustive Walker Lake data set.

Egg quality and fecundity in rainbow trout: the determining factors and mechanisms of control

This thesis considers the factors involved in the determination of egg quality and fecundity in farmed stocks of rainbow trout ( Salmo gairdneri R) • Measurements of egg quality, ie. percentage survivals of eggs and fry, from the production batches of eggs of seven fish farms, showed mean survivals of 70% to eying but levels of only 35% to 4.5g fry (approx. 130 days post-fertilisation). Under optimum conditions survivals may reach 85% suggesting that husbandry methods exert significant influences on egg quality. Chemical analyses of the protein, fat, vitellogenin, ash, amino acids, free fatty acid and mineral levels of eggs of varying quality and from parents of different strains showed compositional differences even between individuals of the same stock. However, none of these differences were correlated with egg quality. Egg size showed similar variations but, again under hatchery conditions there was no correlation with differences in egg quality. The only factor which has been shown to exert a significant influence on egg quality is the time of stripping after ovulation. At 1 0°C eggs should be removed from gravid females within ten days of ovulation to achieve optimum egg and fry survival. Studies of egg production from approximately 10,000 broodstock revealed that total fecundity and egg size increased and relative fecundity decreased with increasing fish size. In general, most fish appeared to produce a constant volume of eggs. This is consistent with a hypothesis that egg size can only be increased by parallel reductions in fecundity. Feeding broodstock at half-ration (0.35% body weight day- 1 ) did not affect egg quality but reduced total fecundity and egg size and increased relative fecundity when compared with eggs produced by fish on full-ration. Comparisons of regressions of total fecundity against fish weight for three strains using ANOCO revealed that one strain was significantly more fecund than two other strains considered. Trout of the same strain maintained on different farms behaved similarly suggesting there was some reproducibility of strain characteristics.

An investigation into the synaptic conditions and celluar mechanisms underlying cerebellar synaptic plasticity

The direction of synaptic plasticity at the connection between parallel fibres (PFs) and Purkinje cells can be modified by PF stimulation alone. Strong activation (Hartell, 1996) or high frequency stimulation (Schreurs and Alkon, 1993) of PFs induced a long-term depression (LTD) of PF-mediated excitatory postsynaptic currents. Brief raised frequency molecular layer stimulation produced a cAMP-dependent long-temi potentiation (LTP) of field potential (FP) responses (Salin et al., 1998). Thin slices of cerebellar vermis were prepared from 14-21 day old male Wistar rats decapitated under Halothane anaesthesia. FP's were recorded from the Purkinje cell layer in response to alternate 0.2Hz activation of stimulating electrodes placed in the molecular layer. In the presence of picrotoxin, FPs displayed two tetrodotoxin-sensitive, negative-going components termed N1 and N2. EPs were graded responses with paired pulse facilitation and were selectively blocked by 101AM 6-cyano-7-nitroquinoxaline-2,3-dicne (CNQX) an antagonist at iy,-amino-3-hydroxy-5-methyl-4-isoxazolepropionate-type ionotropic glutamate receptors (AMPAR) suggesting that they were primarily PE-mediated. The effects of raised stimulus intensity (RS) and/or increased frequency (IF) activation of the molecular layer on FP responses were examined. In sagittai and transverse slices combined RS and IF molecular layer activation induced a LTD of the N2 component of FP responses. RSIF stimulation produced fewer incidences of LTD in sagittal slices when an inhibitor of nitric oxide synthase (NOS), guanylate cyclase (GC), protein kinase G (PKG) or the GABAB receptor antagonist CGP62349 was included into the perfusion medium. Application of a nitric oxide (NO) donor, a cyclic guanosine monophosphate (cGMP) analogue or a phosphodiesterase (PDE) type V inhibitor to prevent cGMP breakdown paired with IF stimulation produced an acute depression, Raised frequency (RF) molecular layer stimulation produced a slowly emerging LTD of N2 in sagittal slices that was largely blocked in the presence of NOS, cGMP or PKG inhibitors. In transverse slices RE stimulation produced a LTP of the N2 component that was prevented by an inhibitor of protein kinase A or NOS. Inhibition of cGMP-signalling frequently revealed an underlying potentiation suggesting that cGMP activity might mask the effects of cAMP. In sagittal slices RE stimulation resulted in a potentiation of FPs when the cAMP-specific PDE type IV inhibitor rolipram was incorporated into the perfusion medium. In summary, raised levels of PE stimulation can alter the synaptic efficacy at PF-Purkinje cell synapses. The results provide support for a role of NO/cGMP/PKG signalling in the induction of LTD in the cerebellar cortex and suggest that activation of GABAa receptors might also be important. The level of cyclic nucleotide-specific PDE activities may be crucial in determining the level of cGMP and CAMP activity and hence the direction of synaptic plasticity.

On the mechanisms of cerebellar synaptic plasticity

Changes in the strength of signalling between neurones are thought to provide a cellular substrate for learning and memory. In the cerebellar cortex, raising the frequency and the strength of parallel fibre (PF) stimulation leads to a long-term depression (LTD) of the strength of signalling at the synapse between PFs and Purkinje cells (PCs), which spreads to distant synapses to the same cell via a nitric oxide (NO) dependent mechanism. At the same synapse, but under conditions of reduced post-synaptic calcium activity, raised frequency stimulation (RFS) of PFs triggers a long-term potentiation of synaptic transmission. The aims of the work described in this thesis were to investigate the conditions necessary for LTD and LTP at this synapse following RFS and to identify the origins and second messenger cascades involved in the induction and spread of LTP and LTD. In thin, parasagittal cerebellar slices whole cell patch clamp recordings were made from PCs and the effects of RFS of one of two, independent PF inputs to the same PC were examined under a range of experimental conditions. Under conditions designed to reduce post-synaptic calcium activity, RFS to a single PF input led to LTP and a decreases in paired pulse facilitation (PPF) in both pathways. This heterosynaptic potentiation was prevented by inhibition of protein kinase A (PKA) or by inhibition of NO synthase with either 7-nitroindazole (7-NI) or NG Nitro-L-argenine methyl ester. Inhibition of guanylate cyclase (GC) or protein kinase G (PKG) had no effect. A similar potentiation was observed upon application of the adenylyl cyclase (AC) activator forskolin or the NO donor spermine NONOate. Both of these treatments also resulted in an increase in the frequency of mEPSCs, which provides further evidence for a presynaptic origin of LTP. Forskolin induced potentiation and the increase in mEPSC frequency were blocked by 7-NI. The styryl dye FM1-43, a fluorescent reporter of endo- and exocytosis, was also used to further examine the possible pre-synaptic origins of LTP. RFS or forskolin application enhanced FM1-43 de-staining and NOS inhibitors blocked this effect. Application of NONOate also enhanced FM1-43 de-staining. When post-synaptic calcium activity was less strictly buffered, RFS to a single PF input led to a transient potentiation that was succeeded by LTD in both pathways. This LTD, which resembled previously described forms, was prevented by inhibition of the NO/cGMP/PKG cascade. Modification of the AC/cAMP/PKA cascade had no effect. In summary, the direction of synaptic plasticity at the PF-PC synapse in response to RFS depends largely on the level of post-synaptic calcium activity. LTP and LTD were non-input specific and both forms of plasticity were dependent on NOS activity. Induction of LTP was mediated by a presynaptic mechanism and depended on NO and cAMP production. LTD on the other hand was a post-synaptic process and required activity of the NO/cGMP/PKG signalling cascade.

Cortical mechanisms of attention in time:neural correlates of the Lag-1-sparing phenomenon

If humans monitor streams of rapidly presented (approximately 100-ms intervals) visual stimuli, which are typically specific single letters of the alphabet, for two targets (T1 and T2), they often miss T2 if it follows T1 within an interval of 200-500 ms. If T2 follows T1 directly (within 100 ms; described as occurring at 'Lag 1'), however, performance is often excellent: the so-called 'Lag-1 sparing' phenomenon. Lag-1 sparing might result from the integration of the two targets into the same 'event representation', which fits with the observation that sparing is often accompanied by a loss of T1-T2 order information. Alternatively, this might point to competition between the two targets (implying a trade-off between performance on T1 and T2) and Lag-1 sparing might solely emerge from conditional data analysis (i.e. T2 performance given T1 correct). We investigated the neural correlates of Lag-1 sparing by carrying out magnetoencephalography (MEG) recordings during an attentional blink (AB) task, by presenting two targets with a temporal lag of either 1 or 2 and, in the case of Lag 2, with a nontarget or a blank intervening between T1 and T2. In contrast to Lag 2, where two distinct neural responses were observed, at Lag 1 the two targets produced one common neural response in the left temporo-parieto-frontal (TPF) area but not in the right TPF or prefrontal areas. We discuss the implications of this result with respect to competition and integration hypotheses, and with respect to the different functional roles of the cortical areas considered. We suggest that more than one target can be identified in parallel in left TPF, at least in the absence of intervening nontarget information (i.e. masks), yet identified targets are processed and consolidated as two separate events by other cortical areas (right TPF and PFC, respectively).