Studies of animal behavior are crucial to understanding animal-ecosystem interactions, but require substantial efforts in visual observation or sensor measurement. We investigated how classifying behavioral states of grazing livestock using global positioning data alone depends on the classification approach, the preselection of training data, and the number and type of movement metrics. Positions of grazing cows were collected at intervals of 20 seconds in six upland areas in Switzerland along with visual observations of animal behavior for comparison. A total of 87 linear and cumulative distance metrics and 15 turning angle metrics across multiple time steps were used to classify position data into the behavioral states of walking, grazing, and resting. Five random forest classification models, a linear discriminant analysis, a support vector machine, and a state-space model were evaluated. The most accurate classification of the observed behavioral states in an independent validation dataset was 83%, obtained using random forest with all available movement metrics. However, the state-specific accuracy was highly unequal (walking: 36%, grazing: 95%, resting: 58%). Random undersampling led to a prediction accuracy of 77%, with more balanced state-specific accuracies (walking: 68%, grazing: 82%, resting: 68%). The other evaluated machine-learning approaches had lower classification accuracies. The state-space model, based on distance to the preceding position and turning angle, produced a relatively low accuracy of 64%, slightly lower than a random forest model with the same predictor variables. Given the successful classification of behavioral states, our study promotes the more frequent use of global positioning data alone for animal behavior studies under the condition that data is collected at high frequency and complemented by context-specific behavioral observations. Machine-learning algorithms, notably random forest, were found very useful for classification and easy to implement. Moreover, the use of measures across multiple time steps is clearly necessary for a satisfactory classification.

Long-term research on storm areas demonstrates the potential and the limits of natural regeneration After windthrow, questions arise about the appropriate silvicultural management. Answers can be derived from long-term studies on 19 storm-damaged areas caused by Vivian (1990) and Lothar (1999), which encompass cleared, cleared and planted as well as uncleared subareas. Forest succession on these areas was studied using repeated regeneration inventories. Ten resp. 20 years after the storms, the resulting young forests were 3–12 m tall and had a stem density of 500 to 31,400 per ha. Many tree species grow in the storm areas, with climax species like European beech (in the lowlands) and Norway spruce (in high-altitude forests) being most frequent. Advance regeneration has only a small share of the young stands, since seedlings were scarce in the pre-storm stands. Regeneration is slightly more dense on cleared than on uncleared storm areas. The yearly increase in seedling density ranged from 25 to 4,000 trees per ha, with low values occurring mainly if dense vegetation of tall forbs, bramble or bracken covered the ground. The increase in density has fallen since the storms, and in thickets with high stem numbers, the regeneration density has even started to decrease. Pionieer trees as well as sycamore maple and ash grow fastest, and climax species like Norway spruce and silver fir slowest. For spruce, planting results in an advance of 1.0 to 2.4 m after 20 years in high montane storm areas; moreover, gaps, which are widespread in storm areas even 10 or 20 years after the storm event, can be avoided. On areas with total damage, cluster planting should be considered, in particular in protection forests and in cases with scarce advance regeneration, missing seed trees and dense ground vegetation.