No abstract

Understanding physiological and behavioral mechanisms underlying the diversity of observed life-history strategies is challenging because of difficulties in obtaining long-term measures of fitness and in relating fitness to these mechanisms. We evaluated effects of experimentally elevated testosterone on male fitness in a population of dark-eyed juncos studied over nine breeding seasons using a demographic modeling approach. Elevated levels of testosterone decreased survival rates but increased success of producing extra-pair offspring. Higher overall fitness for testosterone-treated males was unexpected and led us to consider indirect effects of testosterone on offspring and females. Nest success was similar for testosterone-treated and control males, but testosterone-treated males produced smaller offspring, and smaller offspring had lower postfledging survival. Older, more experienced females preferred to mate with older males and realized higher reproductive success when they did so. Treatment of young males increased their ability to attract older females yet resulted in poor reproductive performance. The higher fitness of testosterone-treated males in the absence of a comparable natural phenotype suggests that the natural phenotype may be constrained. If this phenotype were to arise, the negative social effects on offspring and mates suggest that these effects might prevent high-testosterone phenotypes from spreading in the population.

No abstract

No abstract

Recognizing human activities from common color image sequences faces many challenges, such as complex backgrounds, camera motion, and illumination changes. In this paper, we propose a new 4-dimensional (4D) local spatio-temporal feature that combines both intensity and depth information. The feature detector applies separate filters along the 3D spatial dimensions and the 1D temporal dimension to detect a feature point. The feature descriptor then computes and concatenates the intensity and depth gradients within a 4D hyper cuboid, which is centered at the detected feature point, as a feature. For recognizing human activities, Latent Dirichlet Allocation with Gibbs sampling is used as the classifier. Experiments are performed on a newly created database that contains six human activities, each with 33 samples with complex variations. Experimental results demonstrate the promising performance of the proposed features for the task of human activity recognition.