Environmental variables determining the occurrence of the red-listed Carbonicola anthracophila and C. myrmecina in boreal forests

Abstract: The global biodiversity loss is mainly due to human activities such as an intensification of forestry. Boreal forest ecosystems in Fennoscandia are characterized by disturbances such as forest fires, storms and floods. However, industrial forestry practices suppress forest fires and change the landscape, leading to a loss of habitats and associated species. Particularly lichen species with slow adaptation abilities and a strong substrate specificity face an extinction debt in boreal forests since their substrates are long-lived but no longer created. In this study the species-substrate relationship of two red-listed forest-fire dependent lichen species Carbon-icola anthracophila and C. myrmecina is investigated as well as the environmental variables determining their occurrence. Three dead wood types (snags, stumps, logs) of Picea abies as well as “ordinary” and “resin-rich” Pinus sylvestris were sampled. 24 forest stands of four different forest types (8-19 years old, 20-64 years old, >65 years old, nature reserves and vol-untary set-asides) were inventoried in boreal forests in Hälsingland, Sweden. To record the frequency of potential Carbonicola harbouring substrates a dead wood inventory was con-ducted using a 200 x 10 m sample plot in which all wooden objects were measured. Lichen occurrence probability was modelled based on the explanatory variables using a generalized linear mixed model with a logit link function. The number of suitable dead wood objects in a stand was modelled based on stand categories using a zero-inflated Poisson regression. Sub-strate characteristics were the most important variables in species occurrence and therefore confirm the strong substrate specificity of Carbonicola anthracophila and C. myrmecina to-wards resin rich, hard, charred, dead pine wood without bark. Both lichen species were found within all stand categories, which demonstrates that stand level characteristics such as forest age or management type are not per se important for lichen occurrence. However, under the current management regime, new dead wood of the quality demanded by the species will not be formed and the current occurrence is due to historical factors. Resin rich pine dead wood was found more often in nature reserves and voluntary set asides than in middle-aged stands. Due to the high substrate specificity of Carbonicola effective conservation measures should aim at (i) the preservation of suitable substrates currently existing in managed forests and (ii) the creation of suitable Carbonicola harbouring substrates in the future with a focus on pro-tected areas. This could be achieved through an incorporation of a dead wood inventory in the national forest inventory in order to identify resin rich, fire-scarred pine dead wood, as well as harvesting and planting techniques which take old, fire-scarred pine dead wood into consider-ation and prescribed burning of protected forest stands with a fire frequency similar to those in undisturbed boreal forest ecosystems.