Bird predation and vegetation structure affecting spruce-living arthropods in a temperate forest

1. The impact of bird predation and vegetation structure, viz. needle density, on macro-arthropods living on spruce branches (Picea abies (L.) Karst.) was studied in a field experiment. The arthropod abundance and spider size distribution were recorded in autumn and spring in one pre-experimental (1988-89) and two experimental years (1989-91). 2. The experiment was conducted in SW Sweden. One control branch, subjected to bird predation and with unaltered needle density, and three experimentally manipulated categories were used: (i) needle-sparse branches, (ii) net-enclosed branches on which bird predation was prevented but arthropods were free to leave and enter, and (iii) branches that were needle-sparse and net-enclosed. 3. Spiders (Araneae) were the most common arthropod taxa on spruce branches making up between 36 . 4% and 96 . 6% of the arthropod median densities. 4. In 37 . 5% of the data sets there were positive correlations between insect and spider abundances. No negative correlations were found. 5. Bird predation reduced insect abundance in three out of four experimental samples, and spider abundance in four samples. Interaction between bird predation and needle density was found in the autumn samples for spiders, and in the spring samples for insects. However, bird predation effects were found on needle-sparse branches in some samples and on branches with unaltered needle density in others. 6. Spider median body lengths on bird predation-free branches were consistently larger than the median lengths on branches subjected to bird predation. Interaction between bird predation and needle density affected spider size but the direction of the impact varied with time. On bird predation-free branches, the median size was small on needle-sparse branches and larger on branches with unaltered needle density in the first sample. In the following three samples the median sizes in relation to needle density were reversed. 7. It is concluded that in the present system effects due to passerine birds predominate. Bird predation is strong and reduces arthropod density. Needle density has a less prominent impact in the system. Arthropod abundance, however, is often affected by interaction between bird predation and needle density but the direction of the effect is inconsistent. Spider size distribution is affected by needle density in the absence of bird predation, revealing that coarse vegetation structures, i.e. needle-sparse branches, favour large spiders.