Carpathian forests represent unique and well-preserved ecosystems in relatively intensively managed forests of Europe. Habitat use, foraging assemblages and activity patterns of a bat community were investigated in semi-natural beech-oak forest by monitoring echolocation calls and mist-netting at three localities during the summers of 2003 and 2004. Six different forest habitat types were studied: oak forest, beech forest, stream, road, forest edge and open area within the forest. Bats were detected in all habitats. Sixteen species were found. Habitats were used differently by the individual species. The highest species diversity was observed in the forest interior. The first peak of flight activity was after sunset which then declined and was relatively even through the night until the second peak before sunrise, which was recorded in the forest interior, open area and on the road. The highest flight activity was recorded at the forest edge, forest stream and in open area. Recorded activity was 3× lower in the oak forest interior compared to the forest edge, but if the extent of the forests is considered, forest interior is the most important foraging habitat. Consequently future forest management should consider the needs of this endangered group of animals.
The species composition and relative density of bats were compared in forests of various sizes occurring as “islands” in the agricultural landscape of central Poland. The following island categories were distinguished: very small (0.3–0.7 km2), small (1.0–1.5 km2), medium (2.0–3.5km2) and large (approx. 18 km2). Bats flying over lanes were caught at 34 mist net stations in 13 islands at the end of June and beginning of July (period I) and again at the end of July and beginning of August (period II). Twelve species of bats were recorded (Plecotus auritus, Eptesicus serotinus and Barbastella barbastellus were the dominant species), and the number of species in specific categories of islands ranged from 8–9, except in the very small islands, where only 4 species were confirmed. Species diversity rose with the size of the islands. Nyctalus leisleri and Myotis mystacinus were caught only in the large island. The frequency of B. barbastellus and Nyctalus noctula clearly increased with island size as opposed to E. serotinus and P. auritus. The relative density (mean numbers of individuals caught at one location on one night) during period I increased with island size from very small (1.8) to large (8.1), while during period II, the highest values were achieved in the medium-sized islands (13.4). The mean number of species for one location and night rose in a similar manner. Forest fragmentation to very small units of less than 1 km2 in size negatively influences bat ensembles.
Probabilistic mixtures provide flexible "universal'' approximation of probability density functions. Their wide use is enabled by the availability of a range of efficient estimation algorithms. Among them, quasi-Bayesian estimation plays a prominent role as it runs "naturally'' in one-pass mode. This is important in on-line applications and/or extensive databases. It even copes with dynamic nature of components forming the mixture. However, the quasi-Bayesian estimation relies on mixing via constant component weights. Thus, mixtures with dynamic components and dynamic transitions between them are not supported. The present paper fills this gap. For the sake of simplicity and to give a better insight into the task, the paper considers mixtures with known components. A general case with unknown components will be presented soon.
This text describes a method of estimating the hazard rate of survival data following monotone Aalen regression model. The proposed approach is based on techniques which were introduced by Arjas and Gasbarra \cite{gasbarra}. The unknown functional parameters are assumed to be a priori piecewise constant on intervals of varying count and size. The estimates are obtained with the aid of the Gibbs sampler and its variants. The performance of the method is explored by simulations. The results indicate that the method is applicable on small sample size datasets.
This article deals with the objective Bayesian analysis of random censorship model with informative censoring using Weibull distribution. The objective Bayesian analysis has a long history from Bayes and Laplace through Jeffreys and is reaching the level of sophistication gradually. The reference prior method of Bernardo is a nice attempt in this direction. The reference prior method is based on the Kullback-Leibler divergence between the prior and the corresponding posterior distribution and easy to implement when the information matrix exists in closed-form. We apply this method to Weibull random censorship model and compare it with Jeffreys and maximum likelihood methods. It is observed that the closed-form expressions for the Bayes estimators are not possible; we use importance sampling technique to obtain the approximate Bayes estimates. The behaviour of maximum likelihood and Bayes estimators is observed via extensive numerical simulation. The proposed methodology is used for the analysis of a real-life data for illustration and appropriateness of the model is tested by Henze goodness-of-fit test.
The paper presents the stopping rule for random search for Bayesian model-structure estimation by maximising the likelihood function. The inspected maximisation uses random restarts to cope with local maxima in discrete space. The stopping rule, suitable for any maximisation of this type, exploits the probability of finding global maximum implied by the number of local maxima already found. It stops the search when this probability crosses a given threshold. The inspected case represents an important example of the search in a huge space of hypotheses so common in artificial intelligence, machine learning and computer science.