Forest pest insects may cause large‐scale tree growth reductions and tree mortality during outbreaks. The large‐scale development of outbreaks has frequently been studied, while the colonization‐extinction dynamics during outbreaks is less known. We study the colonization‐extinction dynamics of a severe tree‐killing bark beetle,Ips typographus, during an outbreak across a 130 000 ha forest landscape in southern Sweden. We recorded annual colonization and extinction events in 1 ha pixels across the landscape by helicopter surveys during three consecutive years, and modeled colonization and extinction probabilities based on focal pixel quality, local population size, connectivity to surrounding beetle populations, and tree composition in the surrounding landscape. The local populations had a high turnover; 81–93% of all occupied pixels were the result of colonizations in the same year and the annual extinction rates were 84–90%. The colonization probability increased and the extinction probability decreased with increasing spruce (host‐tree) volume in focal pixels and the connectivity. The fitted spatial scaling parameter of the connectivity measure suggested that the colonization probability was mainly affected by the number of surrounding occupied pixels within a few hundred meters. The colonization probability also decreased with increasing volume of birch (a nonhost) in the focal pixel, while the extinction probability decreased with increasing local population size. In conclusion, the local population size and quality of the focal pixel explained more of the colonization and extinction probability than the connectivity and composition of surrounding forest. The distribution of tree‐killing bark beetles during outbreaks can be patchy and highly dynamic. Two reasons for this are the increased probability of successful attacks when exceeding a critical attack density and the fragmented distribution of large spruce volumes throughout the forest landscape.
The fungal pathogen Batrachochytrium dendrobatidis (Bd) poses a major threat to amphibian populations. To assist efforts to address such threats, we examined differences in Bd host infection prevalence among amphibian species and its relations to both local environmental factors in breeding habitats and landscape variables measured at three scales (500, 2000 and 5000 m radii) around breeding sites in southernmost Sweden. We sampled 947 anurans of six species in 31 ponds and assessed their infection status. We then examined correlations of infection prevalence with canopy cover, pond perimeter and pH (treated as local-scale pond characteristics), and the number of ponds, area of arable land, area of mature forest, number of resident people and presence of sea within the three radii (treated as landscape variables). The Bd infection prevalence was very low, 0.5–1.0%, in two of the six anuran species (Bufo bufo and Rana temporaria), and substantially higher (13–64%) in the other four (Bombina bombina, Bufotes variabilis, Epidalea calamita, Rana arvalis). In the latter four species Bd infection prevalence was positively associated with ponds’ pH (site range: 5.3–8.1), and negatively associated with areas of mature forest and/or wetlands in the surroundings. Our results show that the infection dynamics of Bd are complex and associated with host species, local pond characteristics and several landscape variables at larger spatial scales. Knowledge of environmental factors associated with Bd infections and differences in species’ susceptibility may help to counter further spread of the disease and guide conservation action plans, especially for the most threatened species.
Context: Biodiversity is highly affected by industrial forestry, which leads to the loss and fragmentation of natural habitats. To date, most conservation studies have evaluated associations among a single species group, forest type, or spatial scale.Objective: The objective was to evaluate the richness of multiple species groups across various forest types and characteristics at multiple scales.Methods: We used the occurrence data for 277 species of conservation interest from 455 stands of high conservation value, including four species groups and four forest types.Results: Local, landscape, and regional forest characteristics influenced biodiversity in a non-uniform pattern among species groups and forest types. For example, an increased local spruce basal area in spruce forests was associated with higher vascular plant and bryophyte richness values, whereas macrofungi and lichen richness were positively correlated with deadwood availability, but negatively correlated with the spruce volume in the landscape. Furthermore, landscapes with twice as much mature forest as the average, had more than 50% higher richness values for vascular plants, macrofungi, and lichens.Conclusion: Among sessile species groups in northern forests, a uniform conservation strategy across forest types and scales is suboptimal. A multi-faceted strategy that acknowledges differences among species groups and forest types with tailored measures to promote richness is likely to be more successful. Nevertheless, the single most common measure associated with high richness across the species groups and forest types was mature forest in the landscape, which suggests that increasing old forests in the landscape is a beneficial conservation strategy.