Miha Adamič

The impact of high speed, high volume traffic axes on brown bears in Slovenia

Abstract The Ljubljana–Razdrto highway and the parallel Ljubljana–Trieste railway cut through critical brown bear ( Ursus arctos ) habitat in south-central Slovenia. These high speed, high volume traffic axes are located close to the main dispersal corridor for bears from the Dinaric Mountain range into the Alps. We analyzed radiotracking data of 15 individual bears that lived within 10 km of the highway, compared transportation related and overall known bear mortality, and analyzed the spatial distribution of bear–vehicle accidents. The highway posed a home range boundary to resident bears, but was not an absolute barrier. Transportation-related mortality was high in the vicinity of the highway and railway, and averaged 31% of the total known local mortality from 1992 to 1999. At present the detrimental impact of transportation routes on the bear population in Slovenia is modest due to the high density of bears and the low density of highways—but new highways are planned or already under construction. Managers have to be aware that, due to bears large home ranges and long dispersal distances, a single highway affects bears from a huge area—emphasizing the importance of international cooperation and a landscape approach in highway planning.

Habitat suitability modelling for red deer (Cervus elaphus L.) in South-central Slovenia with classification trees

We study and assess the potential habitats of a population of red deer in South-central Slovenia. Using existing data on the deer population spatial distribution and size, as well as data on the landscape and ecological properties (GIS) of the area inhabited by this population, we develop a habitat suitability model by automated data analysis using machine learning of classification trees. We assume that the recorded observations of deer approximate the actual spatial distribution of the deer population reasonably well. The habitat suitability models for individual animals have the form of classification trees. The induced trees are interpreted by domain experts and a generic model is proposed. The generic habitat suitability models can help determine potential unoccupied habitats for the red deer population and develop guidelines for managing the development of the red deer population and its influence on the environment.

Identifying brown bear habitat by a combined GIS and machine learning method

In this paper we attempt to identify brown bear (Ursus arctos) habitat in south-western part of Slovenia, a country lying on the north-western-most edge of the continuous Dinaric-Eastern Alps brown bear population. The knowledge base (in the form of a decision tree) for the expert system for identifying the suitable habitat, was induced by automated machine learning from recorded bear sightings, and then linked to the GIS thematic layers for subsequent habitat/non-habitat classification of the entire study area. The accuracy of the decision tree classifier was 87% (KHAT 73%). The decision tree mostly agreed with the existing domain knowledge. For the study area the main factors considered by the expert system to be important for brown bear habitat were the percentage of forest (positive), proximity to settlements (negative) and elevation above see (positive), however the decision tree did not account for habitat patch size. After filtering out habitat patches smaller than 5000 ha in GIS, the accuracy increased to 89% (KHAT 77%). Whereas 88% of the habitat was within forests, only 33% of all forests were considered suitable as habitat.

Modelling the population dynamics of red deer (Cervus elaphus L.) with regard to forest development

Abstract Recent advances in artificial intelligence in general, and in machine learning in particular, enable scientists to apply new machine learning technics to their specific areas. In our work we apply such a machine learning technique to the modelling of population dynamics of red deer for the 40 000 hectares co-natural manage forest area on high Karst of Notranjska in Slovenia. We used the RETIS program, a machine learning tool developed by A. Karalie at the Institute Jožef Stefan in Ljubljana. This program induces regression trees from data, and has already been applied to several ecological problems. RETIS was applied on data, collected in the period 1976–1994, which included several meteorological parameters, parameters about the state of the forest, and parameters about the population of the red deer. Given these data about the observed system, the system RETIS automatically induces a model which has the form of a regression tree. We evaluate our induced models qualitatively and quantitatively. For the qualitative evaluation, we present an expert interpretation of the models. We show that quantitatively, using the models (we use a relative prediction error) and given the meteorological parameters during winter and summer and an estimate of the number of red deer in the area, it is possible to predict the state of the forest in the near future. This is very important for maintaining the balance between red deer population and other parameters of the forest, which will allow sustainable development of the complex forest ecosystem.

Supplemental feeding with carrion is not reducing brown bear depredations on sheep in Slovenia

Abstract Supplemental feeding is often believed to be a successful tool for reducing human–bear (Ursus arctos) conflicts, especially in Europe. However, effectiveness of this measure is poorly understood and there is growing concern for potential negative side-effects. This is particularly true for supplemental feeding using livestock carrion. Carrion feeding is considered especially effective in reducing livestock depredations by diverting bears from pastures and meeting their protein needs. In Slovenia, year-round supplementary feeding of bears with livestock carrion and corn was intensive and in some areas practiced for over 100 years. However, in 2004 the use of livestock carrion was banned in accordance with European Union regulations. This provided an opportunity to study the effects of carrion feeding on livestock depredations by bears. We used sheep as they represented 97% of all depredation events by brown bears in Slovenia. We analyzed whether bears selectively used carrion feeding stations over corn feeding stations (i.e., indicating that carrion might be more effective in diverting bears from sheep pastures) during 1994–2011, and compared the annual frequency and seasonal distribution of sheep depredations 5 years before and after the ban on livestock carrion feeding during 1999–2009. We found no support that bears selected carrion feeding sites over feeding sites with corn. When controlled for changes in bear and sheep numbers, there was no indication that the ban on carrion feeding increased sheep depredations. Moreover, complementary data indicated that natural protein sources were considerably more important than livestock carrion and that use of carrion peaked in spring, when sheep are rarely outdoors and thus unavailable for depredation. Because of the observed lack of effectiveness, high costs, and potential negative side-effects, we discourage supplemental feeding with livestock carrion to reduce livestock depredations.

Fast food bears: brown bear diet in a human-dominated landscape with intensive supplemental feeding

Distribution, quantity and quality of food resources affect the diet and several other life-history traits of large mammals. Supplemental feeding of wildlife has high potential for influencing the behaviour and diet of opportunistic omnivores, such as bears. Supplemental feeding of brown bears Ursus arctos is a common practice in several European countries, but the effects of this controversial and expensive management measure on bear diet and behaviour are poorly understood. We analysed 714 brown bear scats collected throughout the year in three regions of Slovenia with different densities of supplemental feeding sites. Supplemental food was the most important food category in the bear diet and represented 34% of the annual estimated dietary energy content (maize: 22%, livestock carrion: 12%). The proportion of supplemental food in the diet varied with season and region, being highest in spring and in the region with the highest density of feeding sites. However, considerable seasonal changes in bear diet, despite year-round access to supplemental food, suggest that bears prefer high-energy natural food sources, particularly insects, fruits, and hard mast, when available. Despite high availability and use of supplemental food, human—bear conflicts are frequent in Slovenia. In addition, evidence from earlier studies suggests that changes in diet and foraging behaviour due to supplemental feeding may affect several aspects of bear biology and in some cases increase the probability of human—bear conflicts. Thus, we caution against promoting unconditional supplemental feeding as a measure to prevent or reduce human—bear conflicts.

Interactions among the red deer (Cervus elaphus, L.) population, meteorological parameters and new growth of the natural regenerated forest in Snežnik, Slovenia

Abstract Following a preliminary study (Stankovski et al., Ecol. Modelling, 108, 1998), we use machine learning techniques to conduct a more detailed analysis of the interactions among the red deer population, meteorological parameters and new forest growth. We use the machine learning program M5 (Quinlan, Proc. 10th Int. Conf. Machine Learning, Morgan Kaufmann, San Mateo CA, 1993) that learns regression trees to automate the modelling of dynamic interactions. An area of 40 000 hectares of naturally regenerated forest on the high Dinaric Karst of Notranjska, Slovenia, is studied. The analysis uses data collected during the period 1976–1993, which include several meteorological parameters, the degrees of browsing intensity of new growth of woody plants (beech and maple), and parameters about the population of red deer. Models of the degree of beech browsing and calf weight were studied earlier; here, we automatically induce models of the red deer population size, the degree of beech and maple browsing, calf weight for 1- and 2-year-olds, and hind weight. The induced models are evaluated in terms of predictive accuracy and interpreted for their explanatory power. The models show that the meteorological parameters, the parameters of the red deer population and the rates of the browsing intensity of the new growth form a complex system with closely related parameters. While these interactions can be mainly explained by our current knowledge, we still gain some new knowledge from the automatically induced models. The results emphasise the importance of a pluralistic approach and a holistic perception of the system formed by meteorological conditions, the red deer population and the new growth in a forest ecosystem.

Experience obtained from box trapping and handling wildcats in Slovenia

WildcatsFelis silvestris Schreber, 1775 were captured for radio collaring as a part of a study of their spatial distribution and social organisation in southern Slovenia between 1999 and 2001. Double-door box traps, with a roof that bears easily break (bear permeable traps), have been used to capture individuals. The distances between traps were between 550 to 2200 m. They were set out on logging roads and narrow trails in the forest. Nine wildcats, one lynx Lynx lynx (Linnaeus, 1758) and one feral cat were caught as target species and 19 badgersMeles meles and one bear cubUrsus arctos as non-targets. The catching success was 1 wildcat/58 trap-days and seems to be in correlation with the lunar cycle. Overall, 7 males and 2 female wildcats were captured which might indicate sex biased trapping selection. Methodological improvements shortened the time of handling procedures. Improved field protocols as well as restraining and immobilisation procedures increased reliability and safety of drug administrations, decreased potential chances for injuries and reduced overall stress of captured animals.