Forest history and the development of old‐growth characteristics in fragmented boreal forests

Questions: Can small and isolated high‐conservation value forests (e.g. designated woodland key habitats) maintain old‐growth forest characteristics and functionality in fragmented landscapes? To what extent have past disturbances (natural and anthropogenic) influenced the development of old‐growth characteristics of these forests? How long does it take for selectively cut stands to attain conditions resembling old‐growth forests? Location: Southern boreal zone of central Sweden. Methods: We linked multiple lines of evidence from historical records, biological archives, and analyses of current forest structure to reconstruct the forest history of a boreal landscape, with special emphasis on six remaining core localities of high‐conservation value forest stands. Results: Our reconstructions revealed that several of these stands experienced wildfires up to the 1890s; all had been selectively harvested in the late 1800s; and all underwent substantial structural and compositional reorganization over the following 100‐150 years. This time interval was sufficient to recover considerable amounts of standing and downed dead wood (mean 60.3 m³ ha^?1), a range of tree ages and sizes (mean basal area 32.6 m² ha^?1), and dominance of shade‐tolerant spruce. It was insufficient to obtain clearly uneven tree age structures and large (>45 cm diameter) living and dead trees. Thus, these forests contain some, but not all, important compositional and structural attributes of old‐growth forests, their abundance being dependent on the timing and magnitude of past natural and anthropogenic disturbances. Our landscape‐level analysis showed marked compositional and structural differences between the historical forest landscape and the present landscape, with the latter having a greater proportion of young forests, introduction of non‐native species, and lack of large trees and dead wood. Conclusions: The remnant high‐conservation value stands were not true representatives of the pre‐industrial forests, but represent the last vestige of forests that have regenerated naturally and maintained a continuous tree cover. These traits, coupled with their capacity for old‐growth recovery, make them valuable focal areas for conservation.     

Disentangling patch and landscape constraints of nested assemblages

The role of the matrix for the functionality of residual patches and therefore for species persistence has been recently recognised. However its influence on assemblage structure and composition is still unclear. We studied the effect of both patch and matrix attributes on community assemblage structure and composition in a fragmented landscape of central Italy. We used bird assemblages as a study model. Birds were sampled in 24 wood patches and distribution patterns were analysed considering both assemblage and guild-level responses. Patch composition and structure together with landscape patterns quantified with GIS were used as explanatory variables. The assemblages appeared to be strongly influenced by the landscape context, both in terms of composition and configuration. The functional connectivity of the patches, in terms of amount of neighbouring borders between woods and hedgerows, the shape of the woods in the matrix, together with the age of the patches, expressed as tree diameter, acted as the essential determinants, and even if the assemblages presented a significant nested structure, seven out of 40 species showed an idiosyncratic distribution. The study of the association between functional guilds and environmental parameters highlighted the influence of ecological constraints on the occurrence of unexpected presences and absences, suggesting the ability of some species to perceive and exploit matrix habitats. The matrix holds the capability to at least mitigate the effects of isolation and habitat loss, and the semi-natural elements of the mosaic may represent a biodiversity reservoir in an otherwise deprived landscape.     

Understorey foraging and habitat selection by sheep in mixed Atlantic woodland

Abstract. The main objectives of this study were to investigate sheep foraging behaviour in mixed Atlantic woodland and to assess its impact on the forest understorey. We established 89 plots along four forest types: Fagus woodland, Quercus woodland, riparian gallery forest and conifer plantations. The presence of plant species in the forest understorey and their foraging damage was surveyed bimonthly from July 1996 to June 1997. In addition, we estimated the selection of woodland types by sheep through the pellet‐group count technique. The intensity of foraging by sheep was negligible for most of the plant species, however several species showed substantial damage in some woodland types. Among the species that were abundant and widespread in the entire study area, Rubus ulmifolius, graminoids and Ilex aquifolium were consumed most. Sheep selected only larch plantations, where grasses and Rubus were very abundant. This grazing behaviour reduced browsing damage of the understorey of woodland stands with higher conservation value, such as Quercus and Fagus woodlands.     

Pollen-derived history of timber exploitation from the Roman period onwards in the Romanche valley, central French Alps

The history of forestry in the Romanche river valley, south-east of Grenoble, France, is reconstructed for the past ca. 3000 years on the basis of detailed pollen analysis and AMS¹⁴C dating. Three deforestation phases are recorded during the last two millennia, each phase showing different features and also contrasting woodland succession in the post-clearance period. The first major deforestation is recorded at the Roman time whenAbies alba (fir) was selectively exploited, presumably for use by peoples living downstream of the site. Apart from the deforestation, there appears to have been little human activity in the vicinity of the site at this time. After the clearance fir gradually, and more or less fully, recovered. The second deforestation phase occurred in ca. the 5^th and 6^th century A.D. when there is also substantial evidence for local farming. At this time, both fir and beech (Fagus sylvatica) were non-selectively exploited and probably used locally. Beach subsequently recovers but there is no further regeneration of fir. The third deforestation phase in ca. the 12^th century A.D. is similar to the preceding phase but this time beech does not recover. With the decline in human activity, secondary forest that included spruce (Picea) and pine (Pinus), developed. Forest dynamics were controlled by local human activity and also the economic relationships between the local area and the wider region and especially the region downstream from the site.     

Effects of coppicing on demographic structure, fruit and seed set in Orchis mascula

Opening of the forest canopy may result in higher fitness of understory plants due to increased light intensity on the forest soil and increased nutrient availability. Especially, tuberous orchids may profit from increased light as photosynthetic assimilation products and nutrients are accumulated in their belowground storage organs. We investigated the effects of coppicing on demographic structure and fruit and seed set in 15 populations of the tuberous, perennial orchid Orchis mascula, seven of which were located in undisturbed, shaded woodland and eight in recently coppiced woodland. Coppicing resulted in a massive increase in flowering and in increased fruit set. On average, 42.9% of all individuals consisted of flowering plants in coppiced woodland, whereas in undisturbed woodland only 20.8% of all individuals flowered. The percentage fruit set varied between 20% and 55% in coppiced woodland and was strongly related to population size (measured as the number of flowering individuals). Mean fruit weight and fruit size were also significantly higher in populations located in coppiced woodland than in undisturbed woodland, whereas the proportion of viable seeds was not related to woodland type. Given that O. mascula is dependent on seed set to fill space with new individuals, these results suggest that the long-term survival of this species strongly depends on frequent opening of the canopy.     

Local extinctions and recolonisations of passerine bird populations in small woods

This paper considers, for eight species of woodland bird, the factors that influenced both local extinctions and recolonisations in 145 woods over 3 years. In all species, probability of local extinction was inversely related to population size; most local extinctions occurred in woods containing one to three breeding pairs. However, considerable variation in extinction probabilities occurred between species and between years. In addition, the suitability of habitat within a wood (more extinctions in less suitable woods) was important for wren Troglodytes troglodytes, song thrush Turdus philomelos and blue tit Parus caeruleus; also, the structure of the surrounding landscape was important for blue tit, great tit Parus major, and chaffinch Fringilla coelebs (more extinctions in localities with less woodland). In only two species was the probability of recolonisation related to any of the measured variables. Wrens were more likely to recolonise larger woods, whereas song thrushes were more likely to recolonise woods with a high habitat suitability rating and those which are more isolated from other woodland     

Identification of site-types important for rare ferns in an area of deciduous woodland in northwest Spain

The Eume woods in northwest Spain are an important refuge for a number of endangered ferns, including Culcita macrocarpa, Cystopteris diaphana, Dryopteris aemula, Dryopteris guanchica, Hymenophyllum tunbrigense and Woodwardia radicans. Because some of these species are inconspicuous or difficult to identify in the field, we attempted to develop simple floristic criteria to identify sites important or potentially important for these species (high-value sites). Having drawn up field-layer vascular plant inventories for 44 plots in the study area, we considered two major strategies. Under strategy 1, we looked for those conspicuous species which best discriminated between plots containing the endangered species (current-high-value plots) and all other plots. Under strategy 2, we first carried out a global plots-by-species classification using indicator species analysis (Hill et al. 1975) and then looked for those conspicuous species which best discriminated between groups containing current-high-value plots and all other groups. For our data set, the second strategy proved more effective. The best criterion identified was If two or more of Luzula sylvatica, Viola riviniana and Rubus spp. are present, then that plot is high-value. We include a listing of a BASIC program for the discriminant analysis algorithm, which we consider to have a number of potential applications in vegetation science and biogeography.     

Plant species as indicators of ancient woodland in northwestern Germany

Abstract. 315 isolated semi-natural and natural stands in NW Germany were investigated floristically, 285 stands of ancient woodlands and 30 recent ones. In the study area nearly all semi-natural and natural ancient woodlands are found on mesotrophic or eutrophic sites and can be assigned to the Fagetalia. 54 selected herbaceous and five woody Fagetalia-species have been tested in their association to ancient woodlands. 21 of the 59 selected plant species show a highly significant association to ancient woodlands and seven species show a weak significance. 31 plant species are not significantly correlated with ancient woodland sites, but 25 of them have a low frequency. The result shows that investigations of the historical ecology can help to understand floristic composition of present-day woodlands. Restriction of many woodland species to ancient woodlands, especially rare species, emphasizes the importance of woodlands with a long continuous history for the preservation of endangered species. Irrespective of ecological conditions, the restriction of plant species to ancient woodlands seems to be mainly caused by their low ability to colonize recent woodlands, especially isolated stands. Differences in the association to ancient woodlands between European countries are assumed to be a function of both time and degree of isolation of woodlands, of the dispersal mode of the plant species and of the availability of suitable habitats.     

Vegetation differentiation and secondary succession on a limestone hill in southern Poland

Abstract. The vegetation of the SkoHzanka Reserve, situated on a limestone hill near Krakow, was examined over a 30 year period, using the Braun-Blanquet method. With the help of a numerical classification eight types of grassland communities were distinguished, belonging to the classes Sedo-Scleranthetea, Nardo-Callunetea and Festuco-Brometea, and 11 types of woodland communities, both remnants of ancient woodlands from the classes Querco-Fagetea and Vaccinio-Piceetea and more recent woods. In the case of grasslands the results of detrended correspondence analysis (DCA) indicated that the largest variation was associated with the soil type. Woods on sandy soils are very different from those on calcareous soils, and the younger recent woods are very different from the ancient woodlands. An analysis of historical cartographic material and aerial photographs shows that in the last few decades the area of recent woods has increased considerably at the expense of grasslands. Changes in species diversity in the woods during secondary succession and their successional convergence were examined and discussed. It was found that young and mature woods on sandy soils are more similar to each other than young and mature woods on calcareous soils. These differences are mainly related to the fact that many woodland herb species growing in rich, broadleaved ancient woods are unable to colonize recent woods since they have a limited colonization capacity.