Shifts in plant respiration and carbon use efficiency at a large-scale drought experiment in the eastern Amazon.

*The effects of drought on the Amazon rainforest are potentially large but remain poorly understood. Here, carbon (C) cycling after 5 yr of a large-scale through-fall exclusion (TFE) experiment excluding about 50% of incident rainfall from an eastern Amazon rainforest was compared with a nearby control plot. *Principal C stocks and fluxes were intensively measured in 2005. Additional minor components were either quantified in later site measurements or derived from the available literature. *Total ecosystem respiration (R(eco)) and total plant C expenditure (PCE, the sum of net primary productivity (NPP) and autotrophic respiration (R(auto))), were elevated on the TFE plot relative to the control. The increase in PCE and R(eco) was mainly caused by a rise in R(auto) from foliage and roots. Heterotrophic respiration did not differ substantially between plots. NPP was 2.4 +/- 1.4 t C ha(-1) yr(-1) lower on the TFE than the control. Ecosystem carbon use efficiency, the proportion of PCE invested in NPP, was lower in the TFE plot (0.24 +/- 0.04) than in the control (0.32 +/- 0.04). *Drought caused by the TFE treatment appeared to drive fundamental shifts in ecosystem C cycling with potentially important consequences for long-term forest C storage.

Intraoperative radiotherapy in the conventional linear accelerator room for early breast cancer treatment: an alternative choice in developing countries.

Intraoperative radiotherapy with a single dose of electrons (ELIOT) in the conservative treatment of breast cancer is a possibility under evaluation in clinical trials. The costs of the mobile linear accelerator with a robotic arm, used in intraoperative radiotherapy, are prohibitive for poor countries. The aim of this study was to evaluate the feasibility of ELIOT in the accelerator room of the Radiotherapy Service for early breast cancer treatment. We analyzed 40 patients submitted to breast conservative surgery and ELIOT, in the accelerator room of the Radiotherapy Service at the Hospital of Pontificia Universidade Católica do Rio Grande do Sul in Brazil from January 2004 to July 2005. Patients with unifocal breast carcinoma smaller than 25 mm, aged over 45 years, who were candidates for conservative surgery were selected and a total dose of 21Gy was delivered, without further radiotherapy. In the short-term follow-up (median 18 months), six patients (15%) presented with some grade of fibrosis under the scar. One case (2.5%) of local recurrence was reported. There are no cases of contralateral carcinoma or distant metastases so far. Our data show that intraoperative radiotherapy with electrons can be safely performed in an accelerator room with a conventional machine.

Degradabilidade ruminal in situ de vagens de faveira (Parkia platycephala Benth.) em diferentes tamanhos de partículas / In situ ruminal degradability of faveira (Parkia platycephala Benth.) pods in different particle sizes

Estimaram-se os parâmetros de degradação ruminal da matéria seca (MS) e da proteína bruta (PB) de vagens de faveira, trituradas em partículas de 2 e 5mm, pelo método do saco de náilon in situ em ovinos, nos tempos de incubação 3, 6, 12, 24, 48, 72 e 96 horas, e determinou-se a degradabilidade efetiva, considerando-se taxas de passagem 2, 5 e 8 por cento/h. A fração a da MS e da PB foram 69,6 e 49,9 por cento, respectivamente, revelando elevada solubilidade da MS; a fração b para MS e PB foi 24,7 e 43,9 por cento, indicando baixa degradação da MS in situ, com estabilização da degradação da MS às 72h e da PB às 48h de incubação. O tanino de vagens de faveira não se mostrou depressor da degradabilidade in situ da PB

Soluble (a) and potentially degradable (b) fractions and degradation rate of b fraction (c) of dry matter (DM) and crude protein (CP) of Parkia platycephala pods in particle sizes 2 and 5mm were estimated by in situ nylon bag method in sheep. The times of incubation were 3, 6, 12, 24, 48, 72 and 96 hours, and the effective degradability (ED) was determined considering passage rates of 2, 5 and 8 percent/h. The a fractions for DM and CP was 69.6 and 49.9 percent, respectively, revealing elevated DM solubility: The b fractions for DM and CP was 24.7 and 43.9 percent, denoting reduced DM in situ degradation. The stabilization of the DM and CP degradation occurred at 72h and 48h after incubation, respectively. In situ degradability of constituents of P. platycephala pods, in special CP, were not depressed for their tannin contents

A method for extracting plant roots from soil which facilitates rapid sample processing without compromising measurement accuracy.

This study evaluates a novel method for extracting roots from soil samples and applies it to estimate standing crop root mass (+/- confidence intervals) in an eastern Amazon rainforest. Roots were manually extracted from soil cores over a period of 40 min, which was split into 10 min time intervals. The pattern of cumulative extraction over time was used to predict root extraction beyond 40 min. A maximum-likelihood approach was used to calculate confidence intervals. The temporal prediction method added 21-32% to initial estimates of standing crop root mass. According to predictions, complete manual root extraction from 18 samples would have taken c. 239 h, compared with 12 h using the prediction method. Uncertainties (percentage difference between mean, and 10th and 90th percentiles) introduced by the prediction method were small (12-15%), compared with uncertainties caused by spatial variation in root mass (72-191%, for nine samples per plot surveyed). This method provides a way of increasing the number of root samples processed per unit time, without compromising measurement accuracy.

Training SVMs with EDR algorithm.

The aim of this work is to present a new training algorithm for SVMs based on the pattern selection strategy called Error Dependent Repetition (EDR). With EDR, the presentation frequency of a pattern depends on its error: patterns with larger errors are selected more frequently and patterns with smaller error (or learned) are presented with minor frequency. Using a simple iterative process based on gradient ascent, SVM-EDR can solve the dual problem without any assumption about support vectors or the Karush-Kuhn-Tucker (KKT) conditions.

Recent advances in the MOBJ algorithm for training artificial neural networks.

This paper presents a new scheme for training MLPs which employs a relaxation method for multi-objective optimization. The algorithm works by obtaining a reduced set of solutions, from which the one with the best generalization is selected. This approach allows balancing between the training error and norm of network weight vectors, which are the two objective functions of the multi-objective optimization problem. The method is applied to classification and regression problems and compared with Weight Decay (WD), Support Vector Machines (SVMs) and standard Backpropagation (BP). It is shown that the systematic procedure for training proposed results on good generalization neural models, and outperforms traditional methods.

Inversion of simulated positron annihilation lifetime spectrum using a neural network.

Inversion of positron annihilation lifetime spectroscopy, based on a neural network Hopfield model, is presented in this paper. From a previous reported density function for lysozyme in water a simulated spectrum, without the superposition of statistical fluctuation and spectrometer resolution effects, was generated. These results were taken as the exact results from which the neural network was trained. The precision of the inverted density function was analyzed taking into account the number of neurons and the learning time of the neural network. A fair agreement was obtained when comparing the neural network results with the exact results. For example, the maximum of the density function, with a precision of 0.4% for the percentual relative error, was obtained for 64 neurons.

Neural networks learning with sliding mode control: the sliding mode backpropagation algorithm.

Based on the classical backpropagation weight update equations, sliding mode control theory is introduced as a technique to adapt weights of a multi-layer perceptron. As will be demonstrated, the introduction of sliding mode has resulted in a much faster version of the standard backpropagation. The results show also that the proposed algorithm presents some important features of sliding mode control, which are robustness and high speed of learning. In addition to that, this paper shows also how control theory can be applied to train neural networks.