Implicit-factorization preconditioning and iterative solvers for regularized saddle-point systems

We consider conjugate-gradient like methods for solving block symmetric indefinite linear systems that arise from saddle-point problems or, in particular, regularizations thereof. Such methods require preconditioners that preserve certain sub-blocks from the original systems but allow considerable flexibility for the remaining blocks. We construct a number of families of implicit factorizations that are capable of reproducing the required sub-blocks and (some) of the remainder. These generalize known implicit factorizations for the unregularized case. Improved eigenvalue clustering is possible if additionally some of the noncrucial blocks are reproduced. Numerical experiments confirm that these implicit-factorization preconditioners can be very effective in practice.

Geographical distribution of the endemic Sardinian brook salamander, Euproctus platycephalus, and implications for its conservation

The Sardinian brook salamander, Euproctus platycephalus, is a cryptically coloured urodele found in streams, springs and pools in the main mountain systems of Sardinia, and is classified as critically endangered by IUCN. General reviews of the mountainous range where salamanders occur are numerous, but very few field-based distribution studies exist on this endemic species. Through a field and questionnaire survey, conducted between 1999 and 2001, we report a first attempt to increase data on the present distribution of E. platycephalus. A total of 14 localities where Sardinian salamanders are represented by apparently stable and in some cases abundant populations have been identified, as well as 30 sites where species presence has been recorded after 1991. Some 11 historical sites were identified which are no longer inhabited by the species. The implications of this distributional study for the conservation of the species and for the realization of an updated atlas are discussed.

Effects of Crossing Saddle Coil Conductors: Electric Length X Mutual Inductance

In magnetic resonance imaging (MRI), either on human or animal studies, the main requirements for radiofrequency (RF) coils are to produce a homogeneous RF field while used as a transmitter coil and to have the best signal-to-noise ratio (SNR) while used as a receiver. Besides, they need to be easily frequency adjustable and have input impedance matching 50 Omega to several different load conditions. New theoretical and practical concepts are presented here for considerable enhancing of RF coil homogeneity for MRI experiments on small animals. To optimize field homogeneity, we have performed simulations using Blot and Savart law varying the coil`s window angle, achieving the optimum one. However, when the coil`s dimensions are the same order of the wave length and according to transmission line theory, differences in electrical length and effects of mutual inductances between adjacent strip conductors decrease both field homogeneity and SNR. The problematic interactions between strip conductors by means of mutual inductance were eliminated by inserting crossings at half electrical length, avoiding distortion on current density, thus eliminating sources of field inhomogeneity. Experimental results show that measured field maps and simulations are in good agreement. The new coil design, dubbed double-crossed saddle described here have field homogeneity and SNR superior than the linearly driven 8-rung birdcage coil. One of our major findings was that the effects of mutual inductance are more significant than differences in electrical length for this frequency and coil dimensions. In vitro images of a primate Cebus paela brain were acquired, confirming double-crossed saddle superiority. (C) 2010 Wiley Periodicals, Inc. Concepts Magn Reson Part B (Magn Reson Engineering) 37B: 193-201, 2010

Repeatability of foraging tactics in young brook trout, Salvelinus fontinalis

In this study, we repeatedly observed individually marked Brook Trout in a field setting to determine if the bimodal variation in foraging tactics previously observed in young salmonid populations (without distinct morphological differences) is generated by individuals specializing at different foraging tactics. We found significant but low repeatability in several foraging movement parameters. This indicated that while individuals did have tendencies to be sedentary versus highly active, there was considerable variation in foraging activity within individuals. These results suggest that relatively consistent differences among individuals may facilitate selection for specialized morphology and that there may be a heritable component to activity.

Does the rate of foraging attempts predict ingestion rate for young-of-the-year brook trout (Salvelinus fontinalis) in the field?

To assess the costs and benefits of young fish adopting different behavioural tactics, field studies of juvenile salmonines have assumed that (but did not test whether) the rate of foraging attempts predicts ingestion rate. We tested this assumption by quantifying capture, ingestion, and rejection rates of potential prey items for individual young-of-the-year brook trout (Salvelinus fontinalis) in a lake. Overall, capture rate (a conservative estimate of the rate of foraging attempts) was only a fair predictor of overall ingestion rate (Kendall's 1 = 0.54) and only 46% of captured items (number/minute) were ingested. Surface capture rate was a poor predictor of surface ingestion rate (T = 0.27) and only 1% of captured items were ingested. In contrast, subsurface capture rate was an excellent predictor of subsurface ingestion rate (T = 0.75) and 93% of captured items were ingested. No benthic prey captures were observed. Fish that ingested a low proportion of captured items spent a greater proportion of time moving, moved faster, and pursued prey further than fish that ingested a higher proportion of captured items. Rejection of captured items can represent a significant and little appreciated component of the foraging cycle for young salmonid fishes.

The central-place territorial model does not apply to space-use by juvenile brook charr Salvelinus fontinalis in lakes

1. Territoriality is widely accepted as the mechanism responsible for density-dependent mortality, emigration, and 'self-thinning' of populations of juvenile salmonine fishes in streams. Numerous studies have focused on territoriality exclusively in stream (lotic) environments and thus have fostered a stereotyped view of juvenile salmonines as sedentary and territorial. We term this behavioural paradigm the central-place territorial model (CPTM).

2. We tested predictions characterizing the CPTM for young-of-the-year (YOY) brook charr (Salvelinus fontinalis) in two Canadian lakes to determine if territoriality may also potentially limit space and population size of brook charr in lakes.

3. Our findings were not consistent with the CPTM. Fish in both lakes were not central-place forages. Maximum displacement distance did not increase with body length as predicted by the general salmonine model of Grant and Kramer (1990). Net displacement distanced increased with the proportion of time spent moving. Aggressive frequency was greatest for fish which spent large proportions of time moving and did not defend from a central-place.

4. Fish in both lakes were rarely aggressive, highly active, and often moved back over the same areas. However, lake fish which migrated to a tributary stream had no net displacement (central-place foraging) illustrating the immediate effects of current on foraging tactics and space-use.

5. The effect of hydrodynamic environment (flowing vs. still water) on fish behaviour needs to be explicitly considered in future models of salmonine behaviour.

Early microhabitat use by age 0 year brook charr Salvelinus fontinalis in lakes

The early habitat use of age 0 year brook charr Salvelinus fontinalis in three north temperate lakes which differ in terms of shoreline physical habitat is described. In the two lakes, which contained abundant shoreline woody debris and inundated vegetation, brook charr were observed in extremely close proximity with these habitat features, near shore and near the surface. Fish were absent from open areas away from shore unless in close proximity with fallen floating logs near the surface, extending offshore. In a third lake that had no woody debris or inundated shoreline vegetation, brook charr were observed exclusively in close proximity with the shoreline itself, and near the surface. In all three lakes, fish were most closely associated with the shoreline and with woody debris and inundated vegetation (when present) shortly after emergence, and significantly farther from shore and deeper in the water column thereafter.

Individual variation in foraging movements in a lake population of young-of-the-year brook charr (Salvelinus fontinalis)

We examined the variation in foraging movements in a population of young-of-the-year (YOY) brook charr living in the near-shore littoral zone of a lake. By repeating the methodology of an earlier stream study, we made direct comparisons between data from lake and stream populations. In general, the pattern of variation in foraging movements was similar between the two sites with greater variability and activity observed in the lake population. The dichotomous nature of the proportion of time spent moving in the stream was also observed in the lake population but in a reversal of the stream pattern. Charr that moved constantly while foraging represented the largest movement category in the lake. In general, variation in foraging movements were more strongly related to the rate of prey rejection, whereas environmental factors, such as distance from shore, submerged objects, and the amount of overhead riparian cover, were more strongly related to prey ingestion. This last finding directly contrasts with that found in the stream literature for YOY charr in still water where ingestion rate, as estimated using feeding attempt rate, increases with the mobility of YOY charr.

ON THREE-PARAMETER FAMILIES of FILIPPOV SYSTEMS - THE FOLD-SADDLE SINGULARITY

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)