Sven G. Nilsson

Biodiversity, disturbances, ecosystem function and management of European forests

Abstract We review the effects of human impact on biodiversity of European forests in the light of recent views on disturbances and succession in ecosystems, and discuss recent ideas on how biodiversity affects ecosystem functions such as productivity and ecosystem stability. With this as a background we discuss how to better manage European forests for both production and biodiversity. We argue that the next generation of forestry practices need to understand and mimic natural disturbance dynamics much better than the present ones. Of particular importance is the fact that most species in European forests have evolved in forests that were to a large extent influenced by large grazers, first by megaherbivores and later, in historic times, by domestic animals. We highlight several areas where new knowledge and management tools are urgently needed: (i) How do species survive and adapt to the natural disturbance regimes in different regions and forest types? (ii) How can new and imaginative forest management practices be devised that take natural disturbance regimes into account? (iii) How does forest biodiversity affect ecosystem function and stability in a changing world, in particular in the light of predicted climate changes? (iv) How are ecological processes at different levels and scales related to diversity, and how do different management practices affect biodiversity? (v) How can efficient agroforestry methods be developed to preserve biodiversity? (vi) What is the role of humans and human behaviour for sustainable management of ecosystems?

The evolution of nest-site selection among hole-nesting birds: the importance of nest predation and competition

Minimum nest entrance widths were correlated with the size of the species. Depths of nesting holes were generally similar for different species, but the Blue Tit occupied shallower holes than did the Great Tit. The Starling Sturnus vulgaris, the Nuthatch Sitta europaea, and the Blue Tit occupied holes higher up in the trees than did the Great Tit and the Pied Flycatcher, in accordance with results of other studies. Contrary to what has been found in these, Marsh Tits nested very low. Total rates of nestfailure and predation were greater in low nests than in higher ones for the Starling, the Blue Tit, and the Marsh Tit. Differences between species in nest predation rates in natural holes could be explained by the vulnerability of the nest sites to predators. For the Starling, the Nuthatch and the Blue Tit average nest heights were negatively correlated with their respective population densities. For the Marsh Tit there was no such correlation, but its nest heights correlated negatively with the density of superior nest competitors. Independent nest preference and utilization data suggest that all four species that vary their nest heights in relation to density prefer to nest high. This indicates that there is competition for safe nest sites. The preference for nest sites higher up in the trees is adaptive and I suggest that predation is the factor that has selected for this preference. The Starling reduced the breeding success of the Nuthatch by taking over holes occupied by the latter. The proportion of Nuthatch nesting attempts that failed because of such interference was higher in high nests and when the population density of the Nuthatch was high.

Seed Predation and Cross‐Pollination in Mast‐Seeding Beech (Fagus Sylvatica) Patches

In this study we examined cross-pollination and levels of insect and verte- brate predation on the seeds of beech as a function of stand size and tree size. Cross- pollination in this wind-pollinated species is positively correlated with stand size and flower density. In two mast years, 6-14% of seeds contained no kernels, and in a nonmast year the figure was 24%. Predispersal destruction of seeds by a moth (Cydia fagiglandana) was 3.1% in a mast year and 38% in a nonmast year. The variation in seed destruction by vertebrates was much less between mast (5.7%) and nonmast (1 2%) years. Destruction by the moth was negatively correlated with tree size, while that by vertebrates was positively correlated with tree size. Vertebrate seed predation was negatively correlated with stand size. In a large stand, < 10% of the seeds were consumed by vertebrates by April in two mast years. Both cross-pollination and seed predation may have been important factors in the evolution of mast seeding in beech.

Densities of large living and dead trees in old-growth temperate and boreal forests

We recorded and reviewed densities and basal areas of large living and dead trees in old-growth forest in Europe. Recorded densities were similar to those reported from old-growth forests in eastern North America, but lower than in northwestem North America. Based on our results we suggest that, 10-20 living trees per ha with dbh > 70 cm may have been typical values for many central European and south Scandinavian virgin forests. In boreal forests, it was probably common with at least 20 living trees per ha with dbh > 40 cm. Basal areas of living trees in mixed old-growth forests in central Europe and southern Sweden were 34-40 m 2 per ha on dry ground and about 60 m(2) per ha in wet alder-ash-spruce forests. Densities of large trees (dbh > 40 cm) were twice as high in the latter forest type than on dry ground in Bialowieza forest, Poland. Based on our results, we propose the following generalizations to be further tested in other old-growth temperate and boreal forests: 1. Among all standing trunks (including high stumps) about 10% are dead. but this proportion increases for the largest trees. The proportion of standing trees that are dead seem to be independent of total basal areas. Based on this, we suggest that the volume of dead wood is directly proportional to the productivity of old-growth forests. 2. Standing dead trees (snags) are on average larger than downed dead trees. Trees with dbh >40 cm often dominate the basal area and volume of standing dead trees and living trees. 3. About 30% (20-40%) of the basal area and volume of dead trees is standing in old-growth forests. This proportion seems to be independent of total volume of dead wood. Large disturbances by fire, strong winds and insects may temporarily change these proportions considerably in individual stands

Biodiversity and its assessment in boreal and nemoral forests

We review species richness in major organism groups, mainly using examples from northern Europe. A high proportion of these species is forest living, and large numbers are dependent on decaying wood. Biodiversity can be assessed at various scales using two different principles. One is to use features, such as ancient and dead trees, known to be important for a large number of species. The other method is to choose species or groups of species known to indicate high biodiversity or presence of many red-listed species. We argue that any serious biodiversity assessment method should include the most species rich organism groups, for example insects. In the present paper we point out the most important features for high biodiversity (old trees and large dead trees), and review the quantities of these features in near-virgin forests. The natural disturbance regime of a region should be the basis for defining a suitable scale and the appropriate features for biodiversity assessment. Possible indicator species for high biodiversity in northern Europe are suggested, based on previous investigations. Among epiphytic lichens and wood-living beetles there are many potentially useful species in addition to vascular plants in the nemoral forest. Among vertebrates, woodpeckers and grouses seem to be the most useful. Validation tests for indicator structures and species are largely lacking but urgently needed. The implications of possible delayed local extinctions are important to bear in mind when managing for sustainable forestry. The knowledge of forest history is useful when developing cost-efficient measures.

Forests in the Temperate–boreal Transition—Natural and Man-made Features

To fit nature into categories is always difficult, especially with complex vegetation. In the northern hemisphere forest vegetation in the temperate zone usually consists of deciduous forests and in the boreal zone often of coniferous forests. In the transition zone there are forests with deciduous trees, coniferous trees and also very often deciduous and coniferous trees mixed (later in this chapter referred to as mixed forests). In the Nordic countries this transition is termed the ‘hemiboreal’ zone (Ahti et al., 1968; see map in chapter 7). In Sweden this zone is about 600 km wide and in the western Soviet Union even wider. Thus, the hemiboreal zone covers large areas. In spite of this its main ecological features are less well understood than those of more southern and northern zones. This is due to intensive exploitation in historic times and the complex dynamics between deciduous and coniferous trees. The separation of natural and man–made factors affecting vegetation dynamics has been especially difficult. In this short presentation of the hemiboreal zone in Sweden the description of some important features and processes must be very brief due to space limitation. I present major landscape changes and their main causes over the last 1000 years. Keystone species, i.e. species whose presence affects the distribution and/or abundance of several other species or whose absence would cause cascading effects throughout a community, are sometimes treated in more detail. Examples of such species are major forest trees and shrubs, common vertebrate predators and herbivores, pollinators and seed dispensers.

HABITAT DIVERSITY OR AREA PER SE? SPECIES RICHNESS OF WOODY PLANTS, CARABID BEETLES AND LAND SNAILS ON ISLANDS

SUMMARY (1) We examine patterns of species richness of woody plants, carabid beetles and land snails, respectively, on seventeen undisturbed forested islands (area range: 0-6-75 ha) in Lake Malaren, Sweden, in relation to area, isolation, vegetation structure, habitat diversity, habitat heterogeneity, and other factors. (2) The slopes of the species-area relations (z-values in the power model) are 0 10 for woody plants, 0-16 for land snails, 0-36 for carabid beetles, and 0-62 for forest birds. The z-values differ significantly from zero and also differ between all organism groups except woody plants and land snails on one hand, and land snails and carabid beetles on the other. (3) Island area is the best single predictor of species richness in the organism groups examined. Since habitat variables and island area are uncorrelated on the studied islands, we can reject the habitat diversity hypothesis as an explanation for the species-area relations found. (4) Total densities of woody plants and land snails were not correlated with island area, while the total density of carabid beetles was positively correlated with island area. (5) Islands with a high proportion of wet forests have relatively higher species richness of carabid beetles, and the number of land snail species relates positively to the proportion of deciduous forest. The patterns found are discussed in relation to general species richness theories.

Insect flower visitation frequency and seed production in relation to patch size of Viscaria vulgaris (Caryophyllaceae)

We examined frequency of pollinator visits, seed set, and seed predation in relation to plant patch size of the perennial caryophyllaceous herb Viscaria vulgaris in two areas in southern Sweden. The Dalsland area had approximately ten times more Viscaria plants per unit area than the Sma˙land area. Long-tongued bumblebees and to a lesser extent Lepidoptera were the most important pollinators in both areas. Honeybees and bumblebees with short proboscides were mainly nectar robbers and/or pollen collectors. Insect visitation per Viscaria plant and patch size were either unrelated or negatively correlated. However, bumblebee visitation was twice as high in the Dalsland area compared to Sma˙land, while no such difference was found in Lepidoptera

Seasonal variation in home-range size, and habitat area requirement of the lesser spotted woodpecker (Dendrocopos minor) in southern Sweden

Seasonal variation in home-range size and habitat area requirement of lesser spotted woodpeckers (Dendrocopos minor) were studied by radio-tracking in southern Sweden for 6 years. Home-range size did not vary between age-groups or sexes, but varied with season and decreased successively from 742 ha in winter (n = 10), 355 ha in early spring (n = 15), 103 ha in late spring (n = 22) to 43 ha during nesting (n = 10). The home-range in late spring (i.e. the 3-5 weeks preceeding egg-laying) represents the defended breeding territory. This included on average 39 ha of forest utilised for foraging (range 31-46 ha, n = 15). Since food availability in late spring has a significant influence on reproductive success, and mortality is highest in this period, we regard this as an estimate of the habitat area requirement. This estimate is valid primarily for birds in southern Sweden, but circumstantial evidence indicate that the area requirement may not be grossly different in other areas with different forest types. For conservation of lesser spotted woodpeckers, management should focus on a minimum of 40 ha of forest dominated by deciduous trees, which may be fragmented over a maximum of 200 ha, (C) 2001 Elsevier Science Ltd. All rights reserved. (Less)

Is local distribution of the epiphytic lichen Lobaria pulmonaria limited by dispersal capacity or habitat quality

By surveying and re-surveying 12 forest sites in southern Sweden for the epiphytic lichen Lobaria pulmonaria with a 9-year interval, and measuring tree-related habitat quality variables, we have investigated whether the local distribution of the lichen is limited by poor dispersal capacity or by habitat quality. Dispersal distances were measured indirectly as the distances between colonised trees and the nearest trees occupied by L. pulmonaria in both 1992 and 2001. To compare habitat quality between trees occupied by L. pulmonaria and neighbouring control trees, we recorded tree species and measured age and growth rate of trees, light conditions, bark structure and bryophyte cover. The estimated mean dispersal distance was 35 m, with a recorded maximum of 75 m. Occupied trees were larger and had a larger cover of bryophytes than unoccupied trees of similar size. The results indicate that dispersal capacity probably is the most important factor in limiting the local distribution of L. pulmonaria, but habitat-quality factors may be important on a smaller spatial scale.