Epsilon Archive for Student Projects

Abstract

The longhorn beetles, LHBs, belonging to the Family Cerambycidae constitute an important part of the biodiversity spectrum in Swedish forests. The LHB community signifies high relevance of biodiversity and conservation status of several specific habitats as well as at the landscape level, owing to their ecological position in a variety of habitats across the landscape. The current study focused on LHBs, since there exists with well-organized and structured data sets on their distribution and abundance over the last two hundred years. The present study focused on discerning the influence of different land uses and land cover parameters on the distribution of certain long horn beetles over different landscapes in Sweden. The main objective was to understand the different aspects to be considered in suggesting inputs for biodiversity conservation in the Swedish forest landscape. The study is based on spatially explicit analysis methods to discern the distribution of selected species of LHBs both at individual and group levels by studying a grid network of 10x10km representing the different supporting habitat factors. The analysis was performed with the support of data already available in SMD - Swedish land cover data base, KNN forest data and Key biotopes data. Variables for prediction were selected based on the assumption that they should influence the distribution of LHBs. A combined LHB records over 65 years from Lindhe et al. 2010 and artportalen records were considered in this study as response variables. The geospatial analysis to quantify the predictors and response variables were accomplished with ArcMap 10.2.2. However, the influence of land use and land cover parameters on the distribution of LHBs were assessed with binary logistic regression models by considering predictor variables both individually and also in combined models. The results of the study indicate that among landscape variables considered in the study general structures such as arable land and pasture land exhibited consistently high statistical power in predicting the presence of most groups of LHB across the landscapes, whilst the area of nominal habitat was observed with very limited predictive power to explain the distribution of LHBs. However, the exception of general pattern is represented by LHBs breeding in oak habitat substrate, which had a relatively strong relation to deciduous forest cover when predictor variables were studied individually. Volumes of specific categories of trees as predictors were found to contribute significantly for the occurrence of particular beetles in respective forest habitats. Key biotope small scale owner area revealed positive significant relation with the occurrence of almost all species in both individual and combined models. In contrast key biotope large scale owner area showed negative relation with almost all of the species’ occurrence. Only one exception for conifer breeding beetles in individual model both key biotope area as predictors exhibited the opposite response pattern. No group of LHBs exhibited strong threshold effects but some individual cases, Trogosoma depsarium, Monochamus galloprovincialis, revealed threshold effects of nominal habitat area as well as volume and below that area or volume the existence of these particular beetles almost absence. Xylotrechus antelope showed to have strong threshold effect of coniferous forest area although it’s breeding substrate is oak habitat. It is concluded that among the land use and land cover parameters arable land, pasture land and clearcut area were represented to have higher relative contribution in regard to the distribution of most of the LHBs. Volume of respective nominal habitat as well as key biotopes also contributed to influence the distribution of LHBs. Therefore for biodiversity conservation we observe that indicators of general management and landscape factors appear to contribute equally much or more towards sustainable forest management than nominal habitat resources in the landscape.