The European spruce bark beetle Ips typographus is the most important insect pest in Central European forests. Under climate change, its phenology is presumed to be changing and mass infestations becoming more likely. While several studies have investigated climate effects across a latitudinal gradient, it remains an open question how phenology will change depending on elevation and topology. Knowing how an altered climate is likely to affect bark beetle populations, particularly across diverse topographies and elevations, is essential for adaptive management. We developed a time-varying distributed delay model to predict the phenology of I. typographus. This approach has the particular advantage of capturing the variability within populations and thus representing its stage structure at any time. The model is applied for three regional climate change scenarios, A1B, A2 and RCP3PD, to the diverse topography of Switzerland, covering a large range of elevations, aspects and slopes. We found a strong negative relationship between voltinism and elevation. Under climate change, the model predicts an increasing number of generations over the whole elevational gradient, which will be more pronounced at low elevations. In contrast, the pre-shift in spring swarming is expected to be greater at higher elevations. In comparison, the general trend of faster beetle development on steep southern slopes is only of minor importance. Overall, the maximum elevation allowing a complete yearly generation will move upwards. Generally, the predicted increase in number of generations, earlier spring swarming, more aggregated swarming, together with a projected increase in drought and storm events, will result in a higher risk of mass infestations. This will increase the pressure on spruce stands particularly in the lowlands and require intensified management efforts. It calls for adapted long-term silvicultural strategies to mitigate the loss of ecosystem services such as timber production protection against rockfall and avalanches and carbon storage. K E Y W O R D Sforest pest, infestation risk, Norway spruce (Picea abies), phenology, temperature, timevarying distributed delay model, voltinism

In South Africa, a plan was launched to manage separate sub-populations of endangered African wild dogs (Lycaon pictus) in several small, geographically isolated conservation areas as a single meta-population. This intensive management approach involves the re-introduction of wild dogs into suitable conservation areas and periodic translocations among them. Despite the initial failures and high costs associated with wild dog re-introductions and translocations, there is no predictive framework available to quantify which management protocol is the most efficient. We therefore developed an individualbased model of wild dog population and pack dynamics, which accounts for the wild dogs' social complexity. The model appeared to capture the essential characteristics of a real wild dog population from Hluhluwe-iMfolozi Park, South Africa and to be relatively robust to parameter uncertainty, suggesting that the model is valid enough for addressing management problems. The model enabled us to quantify a critical initial number of packs (two) and individuals per pack (six) necessary for a re-introduced wild dog population to establish itself in the release area. We also found a practically feasible intervention regime at which a re-introduced wild dog population had the best chance of persistence: intermittently adding packs (at least every 6 years) and harvesting disperser groups (as often as every 4 years) for translocation to other release sites, without threatening the small source population. This study demonstrates that individual-based models can be a powerful decision-support tool in re-introduction planning and provides insight into how populations made up of social groups have dynamics, and ultimately persistence, determined by individual behaviour.

No abstract

No abstract