The value of plantation forests for plant,invertebrate and bird diversity and the potential for cross-taxon surrogacy

As the area of plantation forest expands worldwide and natural, unmanaged forests decline there is much interest in the potential for planted forests to provide habitat for biodiversity. In regions where little semi-natural woodland remains, the biodiversity supported by forest plantations, typically non-native conifers, may be particularly important. Few studies provide detailed comparisons between the species diversity of native woodlands which are being depleted and non-native plantation forests, which are now expanding, based on data collected from multiple taxa in the same study sites. Here we compare the species diversity and community composition of plants, invertebrates and birds in Sitka spruce- (Picea sitchensis-) dominated and Norway spruce- (Picea abies-) dominated plantations, which have expanded significantly in recent decades in the study area in Ireland, with that of oak- and ash-dominated semi-natural woodlands in the same area. The results show that species richness in spruce plantations can be as high as semi-natural woodlands, but that the two forest types support different assemblages of species. In areas where non-native conifer plantations are the principle forest type, their role in the provision of habitat for biodiversity conservation should not be overlooked. Appropriate management should target the introduction of semi-natural woodland characteristics, and on the extension of existing semi-natural woodlands to maintain and enhance forest species diversity. Our data show that although some relatively easily surveyed groups, such as vascular plants and birds, were congruent with many of the other taxa when looking across all study sites, the similarities in response were not strong enough to warrant use of these taxa as surrogates of the others. In order to capture a wide range of biotic variation, assessments of forest biodiversity should either encompass several taxonomic groups, or rely on the use of indicators of diversity that are not species based.     

Plantations of exotic tree species in Britain: irrelevant for biodiversity or novel habitat for native species?

Novel or emergent ecosystems arising from human action present both threats and opportunities for biodiversity. It has been suggested that exotic species can “facilitate” or “inhibit” native biodiversity through habitat modification. In Britain, there is a discussion over the contribution to biodiversity of plantations of exotic conifer species as these are commonly thought to have little relevancy as a habitat for native biodiversity. To address this we compared the species richness of a range of different taxonomic groups (lichens, bryophytes, fungi, vascular plants, invertebrates and songbirds) in exotic and native forest stands of differing structural stages in northern and southern Britain. In terms of overall native species-richness there was no significant difference between the exotic and the native stands. In the north, six species groups showed higher values in the exotic Sitka spruce (Picea sitchensis) stands with the remaining six showing higher values in the native Scots pine (Pinus sylvestris) stands. Most notably, lichen species richness was much lower in the exotic stands compared to the native stands, whereas bryophyte and fungal species richness was proportionately higher in the exotic stands. In the south, five species groups (all invertebrate taxa) showed higher species richness in exotic Norway spruce (Picea abies) stands compared to native oak (Quercus robur) stands. Five species groups had higher species-richness in the oak stands, in particular lichens and fungi. It is concluded that emergent ecosystems of exotic conifer species are not irrelevant to biodiversity. Where already well-established they can provide habitat for native species particularly if native woodland is scarce and biodiversity restoration is an immediate priority.     

Are bird density,species richness and community structure similar between native woodlands and non-native plantations in an area with a generalist bird fauna?

This study compared the bird assemblages of native semi-natural woodlands and non-native Sitka spruce (Picea sitchensis) plantations in Ireland to identify what vegetation variables most influenced birds and to identify management targets in plantations to maximise future bird conservation. Point counts were conducted in 10 Oak (Quercus spp.) and 10 Ash (Fraxinus excelsior) native woodlands and in five Mid-rotation (20–30 years old) and five Mature (30–50 years old) Sitka spruce plantations. Ordination was used to characterise woodland types according to their constituent bird species. Total bird density (calculated using Distance software) and species richness were assessed for the different woodland types. Oak and Ash woodland bird assemblages were separated from Mid-rotation and Mature plantations by the ordination. There was no difference in total bird density between any of the woodland types. Oak woodlands had significantly higher species richness than either Mid-rotation or Mature Sitka spruce plantations. Ash had higher species richness than Mature Sitka spruce plantations. Understorey vegetation was negatively associated with total bird density, which also varied with survey year. Understorey vegetation was positively associated with species richness. Reasons for the relationships between vegetation and bird assemblages are discussed. Management should seek to increase shrub and understorey vegetation in the Mid-rotation phase to improve the contribution of plantations to bird conservation.     

Testing indicators of biodiversity for plantation forests

In many parts of the world, plantations make up a considerable proportion of the total forest area. In such regions, the identification of high biodiversity value stands and of management practices to enhance biodiversity is essential if the goals of Sustainable Forest Management are to be achieved. Since complete biodiversity assessments are rarely possible, efforts have been increasingly focussed on the use of indicators. Of particular interest are indicators applicable to individual stands that require no specialist taxonomic or technical knowledge to assess. Candidate biodiversity indicators had been identified in a previous study using data from Irish Sitka spruce (Picea sitchensis) and ash (Fraxinus excelsior) plantations but had yet to be tested on independent data. In the present study, the provisional indicators for vascular plant, bryophyte, spider and bird diversity were tested on data from Irish Scots pine (Pinus sylvestris), oak (Quercus petraea/Quercus robur), Sitka spruce and lodgepole pine (Pinus contorta) plantations. Conifer canopy cover was confirmed as an important biodiversity indicator, due to its influence on below-canopy microclimatic and structural conditions. Bryophyte species richness was higher in relatively high canopy cover plantations on poorly drained soils, while bird species richness was higher in more open plantations with high shrub cover. Coarse woody debris was an important substrate for forest-associated bryophytes, with higher species richness at higher volumes of deadwood. Both proximity to old woodland and stand age were confirmed as positive indicators for forest-associated vascular plants. This is related to dispersal limitation in these species, with nearby woodlands acting as important seed sources and colonisation increasing with time. Stand age was also confirmed as a positive indicator for forest-associated spiders and is related to the development of suitable habitat as the plantation matures. All of the confirmed indicators can be assessed without need for specialist knowledge, are ecologically meaningful and applicable to a range of forests managed under a clearfelling system. They can be used to assess the potential value of stands for the taxonomic groups to which they apply, as well as giving insights into management practices to enhance diversity in these groups.     

The role of planted forests in the provision of habitat: an Irish perspective

The continued decline of natural forests globally has increased interest in the potential of planted forests to support biodiversity. Here, we examine the potential conservation benefits of plantation forests from an Irish perspective, a country where remaining natural forests are fragmented and degraded, and the majority of the forest area is comprised of non-native Sitka spruce (Picea sitchensis (Bong.) Carr.) and Norway spruce (Picea abies (L.) Karst.) plantations. We examine the true value of Irish plantation forests to native biodiversity, relative to remaining natural forest fragments, and to prior and alternative land use to afforestation. We find that plantation forests provide a suitable surrogate habitat primarily for generalist species, as well as providing habitat for certain species of conservation concern. However, we find that plantation forests provide poor habitat for native forest specialists, and examine potential management strategies which may be employed to improve habitat provision services for this group.     

Investigating the abundance and flight period of bark beetles (Coleoptera: Curculionidae: Scolytinae) over elevational gradients in Sitka spruce forests

1. A warming climate, as predicted under current climate change projections, is likely to influence the population dynamics of many forest insect species. Numerous bark beetle species in both Europe and North America have already responded to a warming climate by significantly expanding their geographical ranges.
2. The aim of the current study was to investigate how populations of bark beetles within stands of Sitka spruce, a widely planted non-native commercial plantation tree species in the U.K., were likely to respond to a warming climate. Experimental plots were established in stands of Sitka spruce over elevational gradients in two commercial forest plantations, and the abundance and emergence times of key bark beetle species were assessed over a 3-year period using flight interception traps. The air temperature difference between the lowest and highest experimental plot in each forest was consistently >1°C throughout the 3-year period.
3. In general, the abundance of the most dominant bark beetle species (e.g. Trypodendron, Dryocoetes, Hylastes spp.) was higher, and emergence times tended to be earlier in the year at the lower elevation plots, where temperatures were higher, although not all bark beetle species responded in the same manner.
4. The results of the study indicated that, under the projected future climate warming scenarios, monoculture Sitka spruce stands at low elevations may potentially be more vulnerable to significant outbreak events from existing or invasive bark beetle species. Hence, consideration of establishing more resilient forests of Sitka spruce by diversifying the species composition and structure of Sitka spruce stands is discussed.

     

Plant community ecology of Pinus sylvestris, an extirpated species reintroduced to Ireland

Plantation forests can make a significant contribution to the conservation of native biodiversity, especially where native forest cover is low. Ireland is used as a case study to explore the contribution to biodiversity made by stands of Pinus sylvestris (Scots pine), a reintroduced species. Despite its disputed native status, P. sylvestris is being widely planted in semi-natural habitats in Ireland. The associated vegetation communities have not previously been described and their conservation value is unknown. Baseline information is needed to inform conservation and forest management strategies. Botanical surveys were carried out at 20 plots of P. sylvestris-dominated woodland and scrub throughout the Republic of Ireland. Vegetation, structural and environmental data were recorded. Data were analysed using non-parametric and multivariate statistical techniques and a synoptic table was prepared. P. sylvestris was found to be a non-specialist in terms of its environmental tolerances. β diversity among plots was high while α diversity within plots was low to moderate. The plots surveyed contained 14.2% of the Irish native flora. There was a low level of constancy of species. Four reasonably well defined vegetation communities were identified. Soil pH, altitude and slope had important roles in partitioning these vegetation types and soil pH was positively correlated with species richness. P. sylvestris is well established, well integrated and naturalising in Irish semi-natural habitats. Some of the associated vegetation communities corresponded to habitats of international conservation importance. This research demonstrates that stands of P. sylvestris represent an important resource for Ireland’s native botanical and habitat diversity.     

Non weaving spiders on native woodlands and <Emphasis Type="Italic">Eucalyptus</Emphasis> plantations in Western Mexico: diversity and distribution patterns

Eucalyptus spp. are commonly planted, forming non-native plantations, including the tropics and their wildlife conservation value is relatively unknown. Recent studies have concluded that secondary forests and tree plantations are less diverse than well-developed tropical rain forests. However, introduced Eucalyptus stands harbored similar species richness to surrounding native woodland in temperate woodlands in North America though the identity of the species present may differ. Species composition, as well as dominance curves and differences in community structure add additional insight to understanding faunistic responses to replacement of native woodland by Eucalyptus plantations. Here, we compared species richness, diversity patterns, and the distribution of non-weaving spiders between native woodlands and Eucalyptus plantations in a temperate region of Mexico. We found more Lycosidae species in all plantation stands. Other community attributes were not consistently different between plantations and native woodlands. This is explained by similarities between, and differences within, the understory of the two main vegetation types. Multivariate analyses identified three spider groups and five spider species could be identified as indicators of these groups. A comparison of the number of species of the wandering spiders between the two vegetation types suggests a compensation pattern that is reported here for the first time.     

Restoring planted ancient woodland sites — Assessment,silviculture and monitoring

Summary

This paper deals with three aspects of the process of restoring planted ancient woodland sites(PAWS) to semi-natural conditions. Firstly, we describe a baseline assessment of botanical interest within a PAWS. This survey has been undertaken to determine the impact of clearfelling, particularly on lower plants and the subsequent colonisation of ground vegetation into areas currently dominated by needle litter. Secondly, we discuss some of the main considerations when undertaking restoration through alternative silvicultural systems to clearfell. Finally, we describe the main requirement for successful site monitoring for management purposes.

The last decade has seen considerable attention given to the benefits of restoring plantations on ancient woodland sites (PAWS) to semi-natural conditions. The survival of species and communities associated with ancient woodland through the process of conversion will be a critical measure of success for restoration practice. It is generally assumed that a gradual approach will improve the chances of such success.

Glencripesdale National Nature Reserve (NNR) is a heterogeneous area of semi-natural open ground, ancient woodland and Sitka spruce plantation. The plantation blocks contain occasional discrete elements of semi-natural vegetation, including some veteran trees. Because of difficult access and poor stability, silvicultural options are limited and clear felling is the only practical option. We describe a simple method of monitoring changes to cryptogamic communities and ground layer vegetation prior to and following clearfelling of the plantation matrix. Baseline data are presented.

In more stable and accessible stands, there are a number of alternative silvicultural approaches to consider when gradually restoring a PAWS. This paper addresses the question of how to secure ancient woodland remnants. We outline some of the initial silvicultural considerations such as stability, thinning/felling pattern and light requirements of native tree species. An approach to site monitoring is presented to allow managers to assess whether conditions are improving or declining and whether they are delivering objectives.     

Macrofungal diversity and ecology in four Irish forest types

The macrofungal communities of Irish native tree species (ash and oak) and exotic tree species (Scots pine and Sitka spruce) forests were examined through the collection of sporocarps over 3 yr. Sampling of 27 plots revealed 186 species of macrofungi, including 10 species new to Ireland. The species richness of non-native Sitka spruce and Scots pine forests was similar to that of native oak forests. However, specific communities of macrofungi existed in each of the forest types as confirmed by non-metric multidimensional scaling and multi-response permutation procedure. Indicator species analysis was used to identify macrofungi which are indicative of the four forest types. The oak community lacked certain species/genera known to be distinctive of oak woods in Britain, possibly due to low inoculum availability as a result of historic removal of Ireland’s oak forests. Our results indicate that, while being similar to native forests in species richness, non-native forests of Sitka spruce and Scots pine in Ireland harbour many fungal species which are not typical of native forests, particularly members of the genus Cortinarius.     

Successful reproduction and pheromone production by the spruce bark beetle in evolutionary naïve spruce hosts with familiar terpenoid defences

相似文献   

Identifying practical indicators of biodiversity for stand-level management of plantation forests

Identification of valid indicators of biodiversity is a critical need for sustainable forest management. We developed compositional, structural and functional indicators of biodiversity for five taxonomic groups—bryophytes, vascular plants, spiders, hoverflies and birds—using data from 44 Sitka spruce (Picea sitchensis) and ash (Fraxinus excelsior) plantation forests in Ireland. The best structural biodiversity indicator was stand stage, defined using a multivariate classification of forest structure variables. However, biodiversity trends over the forest cycle and between tree species differ among the taxonomic groups studied. Canopy cover was the main structural indicator and affected other structural variables such as cover of lower vegetation layers. Other structural indicators included deadwood and distances to forest edge and to broadleaved woodland. Functional indicators included stand age, site environmental characteristics and management practices. Compositional indicators were limited to more easily identifiable plant and bird species. Our results suggest that the biodiversity of any one of the species groups we surveyed cannot act as a surrogate for all of the other species groups. However, certain subgroups, such as forest bryophytes and saproxylic hoverflies, may be able to act as surrogates for each other. The indicators we have identified should be used together to identify stands of potentially high biodiversity or to evaluate the biodiversity effects of silvicultural management practices. They are readily assessed by non-specialists, ecologically meaningful and applicable over a broad area with similar climate conditions and silvicultural systems. The approach we have used to develop biodiversity indicators, including stand structural types, is widely relevant and can enhance sustainable forest management of plantations.     

Large‐scale assessment of regeneration and diversity in Mediterranean planted pine forests along ecological gradients

Aim There is increasing concern regarding sustainable management and restoration of planted forests, particularly in the Mediterranean Basin where pine species have been widely used. The aim of this study was to analyse the environmental and structural characteristics of Mediterranean planted pine forests in relation to natural pine forests. Specifically, we assessed recruitment and woody species richness along climatic, structural and perturbation gradients to aid in developing restoration guidelines. Location Continental Spain. Methods We conducted a multivariate comparison of ecological characteristics in planted and natural stands of main Iberian native pine species (Pinus halepensis, Pinus pinea, Pinus pinaster, Pinus nigra and Pinus sylvestris). We fitted species‐specific statistical models of recruitment and woody species richness and analysed the response of natural and planted stands along ecological gradients. Results Planted pine forests occurred on average on poorer soils and experienced higher anthropic disturbance rates (fire frequency and anthropic mortality) than natural pine forests. Planted pine forests had lower regeneration and diversity levels than natural pine forests, and these differences were more pronounced in mountain pine stands. The largest differences in recruitment – chiefly oak seedling abundance – and species richness between planted and natural stands occurred at low‐medium values of annual precipitation, stand tree density, distance to Quercus forests and fire frequency, whereas differences usually disappeared in the upper part of the gradients. Main conclusions Structural characteristics and patterns of recruitment and species richness differ in pine planted forests compared to natural pine ecosystems in the Mediterranean, especially for mountain pines. However, management options exist that would reduce differences between these forest types, where restoration towards more natural conditions is feasible. To increase recruitment and diversity, vertical and horizontal heterogeneity could be promoted by thinning in high‐density and homogeneous stands, while enrichment planting would be desirable in mesic and medium‐density planted forests.     

Do vascular plants and bryophytes respond differently to coniferous invasion of coastal heathlands?

Invasion by non-native conifers may pose a threat to local biodiversity, but knowledge about introduced conifer effects on Northern Hemisphere ecosystems is scarce. The coastal heathlands of north-west Europe are threatened by invasion of native and introduced tree species. We assess how spread of the introduced conifer Sitka spruce (Picea sitchensis (Bong.) Carr.) into European coastal heathlands affect two major functional groups; vascular plants and bryophytes, and how these effects relate to the environmental changes imposed by the developing tree canopies. We compared the impact of introduced Sitka spruce and native Scots pine (Pinus sylvestris L.) by analysing effects on species richness and turnover of vascular plants and bryophytes along fine-scale transects from individual tree stems into open heathland vegetation. Environmental impacts were assessed by measured environmental variables, and the responses of the two species groups were assessed by calculating changes in their respective mean Ellenberg indicator values. Species richness decreased beneath both conifers, related to decreased light and increased nitrogen and pH. Whereas vascular plants responded negatively to poor light conditions beneath dense and low Sitka spruce canopies, bryophytes were more negatively affected by the warmer and drier microclimates beneath Scots pine. Introduced Sitka spruce impacts the sub-canopy environment differently from the native Scots pine, and the two functional plant groups responded differently to these impacts. This suggests that future forests are likely to differ in species richness and composition, depending on whether succession is based on native or introduced coniferous trees.     

Native tree regeneration in native tree plantations: understanding the contribution of Araucaria angustifolia to biodiversity conservation in the threatened Atlantic Forest in Argentina

Deforestation is a global process that has strongly affected the Atlantic Forest in South America, which has been recognised as a threatened biodiversity hotspot. An important proportion of deforested areas were converted to forest plantations. Araucaria angustifolia is a native tree to the Atlantic Forest, which has been largely exploited for wood production and is currently cultivated in commercial plantations. An important question is to what extent such native tree plantations can be managed to reduce biodiversity loss in a highly diverse and vulnerable forest region . We evaluated the effect of stand age, stand basal area, as a measure of stand density, and time since last logging on the density and richness of native tree regeneration in planted araucaria stands that were successively logged over 60 years, as well as the differences between successional groups in the response of plant density to stand variables. We also compared native tree species richness in planted araucaria stands to neighbouring native forest. Species richness was 71 in the planted stands (27 ha sampled) and 82 in native forest (18 ha sampled) which approximate the range of variation in species richness found in the native forests of the study area. The total abundance and species richness of native trees increased with stand age and time since last logging, but ecological groups differed in their response to such variables. Early secondary trees increased in abundance with stand age 3–8 times faster than climax or late secondary trees. Thus, the change in species composition is expected to continue for a long term. The difference in species richness between native forest and planted stands might be mainly explained by the difference in plant density. Therefore, species richness in plantations can contribute to local native tree diversity if practices that increase native tree density are implemented.     

Community homogenization and the invasiveness of commensal species in Mediterranean afforested landscapes

The ecological consequences of homogenization remain relatively unexplored. One example of landscape-homogenizing is the establishment of plantations. We studied the effect of human-made forests by contrasting plant and small-mammal community composition between planted tree stands and adjacent natural habitat in two different Mediterranean habitats in Israel: (1) inland habitat where we focused on pine (Pinus halepensis) and carob (Ceratonia siliqua) stands, and (2) coastal sand dune habitat where we focused on planted acacia (Acacia saligna) stands. We first wanted to verify whether planted trees modify plant species composition, and second, if and how the small-mammal community is affected by the different habitat conditions created in plantations with different canopy cover. We were especially interested in the abundance of the commensal house mouse (Mus musculus). All tree stands underwent biotic homogenization indicated by abundance of house mice coupled with lower diversity of indigenous vegetation and small-mammal abundances and diversities. Habitat structural diversity was positively related with small-mammals diversity and was lower in artificial tree stands in both habitats. Our results suggest that using the abundance of commensal generalist species such as the house mouse relative to other more specialist small-mammals is a good approach to determine ecosystem integrity. Pre-commercial thinning treatment is a potential management tool to maintain a proportion of native tree species within the canopy of planted tree stands. However, until sufficient data is available for making generalizations, the exact level of thinning necessary to reverse the homogenization processes in man-made plantations and keeping indigenous small-mammal communities diverse and less prone to invasion must be determined empirically.     

Diversity of understory birds in old stands of native and Eucalyptus plantations

We compared the vegetation structure between old (>70 year) stands of planted diversified native forests and stands of Eucalyptus tereticornis embedded in a mosaic of Eucalyptus stands. We then tested for differences in the abundance, species richness, species composition, and ecological traits (forest dependence, sensitivity to forest fragmentation, and diet) of the understory bird assemblages inhabiting both kinds of stands. We expected differences in the structure of the bird assemblages because of the different origins and management strategies (contrary to native stands, Eucalyptus stands were selectively logged in the past). Three stands of each habitat (native and Eucalyptus) were sampled with mist nets during 11 months. Eucalyptus stands had a denser understory, whereas native plantations had a more developed vertical structure and a greater density of native trees. The abundance distribution of bird species was more homogeneous in Eucalyptus than in native stands. Eucalyptus had slightly higher species richness (36 species) than native stands (32 species). The composition of species and the occurrence of the diet, forest dependence, and sensitivity to forest fragmentation categories were similar between habitats. Some bird species (e.g. Turdus leucomelas), however, were more abundant in one habitat over the other. Old stands of Eucalyptus and planted native forest can harbor a diverse bird community similar in structure but not exactly equivalent for individual bird species. Planting native diversified forests and keeping set‐aside stands of the exotic tree should be viewed as complementary rather than alternative strategies for maintaining bird diversity within plantations.     

Protura in native and exotic forests in the North Island of New Zealand

Abstract

This study provides data on population and community ecology of Protura in native forests and Pinus radiata plantations in New Zealand. Abundance, age structure, sex ratios, biodiversity, and relationship with soil chemistry are discussed. Protura were significantly more abundant in pine plantations in comparison to native forests. Among native forests, Protura were most abundant under Southern beech (Nothofagus solandri). The abundance patterns may reflect the association between Protura and fungal communities in the soil. No correlation was found between soil organic matter, nitrogen, phosphorous, pH, cation exchange capacity, and the density of Protura. Protura assemblage composition was significantly related to forest type. New Zealand endemic species were associated with native forests; species with distribution outside New Zealand dominated in pine plantations. The distribution records within New Zealand were expanded for five species. The Protura fauna of New Zealand was increased to 18 species. Berberentulus capensis, Eosentomon australicum and Australentulus tillyardi are new records for New Zealand fauna.     

Woody Species Diversity in Forest Plantations in a Mountainous Region of Beijing,China: Effects of Sampling Scale and Species Selection

The role of forest plantations in biodiversity conservation has gained more attention in recent years. However, most work on evaluating the diversity of forest plantations focuses only on one spatial scale; thus, we examined the effects of sampling scale on diversity in forest plantations. We designed a hierarchical sampling strategy to collect data on woody species diversity in planted pine (Pinus tabuliformis Carr.), planted larch (Larix principis-rupprechtii Mayr.), and natural secondary deciduous broadleaf forests in a mountainous region of Beijing, China. Additive diversity partition analysis showed that, compared to natural forests, the planted pine forests had a different woody species diversity partitioning pattern at multi-scales (except the Simpson diversity in the regeneration layer), while the larch plantations did not show multi-scale diversity partitioning patterns that were obviously different from those in the natural secondary broadleaf forest. Compare to the natural secondary broadleaf forests, the effects of planted pine forests on woody species diversity are dependent on the sampling scale and layers selected for analysis. Diversity in the planted larch forest, however, was not significantly different from that in the natural forest for all diversity components at all sampling levels. Our work demonstrated that the species selected for afforestation and the sampling scales selected for data analysis alter the conclusions on the levels of diversity supported by plantations. We suggest that a wide range of scales should be considered in the evaluation of the role of forest plantations on biodiversity conservation.     

Importance of semi-natural habitats for the conservation of butterfly communities in landscapes dominated by pine plantations

While the area of plantation forests continues to increase worldwide, their contribution to the conservation of biodiversity is still controversial. There is a particular concern on the central role played by natural habitat remnants embedded within the plantation matrix in conserving species-rich insect communities. We surveyed butterflies in maritime pine plantation landscapes in south-western France in 83 plots belonging to seven habitat types (five successional stages of pine stands, native deciduous woodlands and herbaceous firebreaks). The effect of plot, habitat and landscape attributes on butterfly species richness, community composition and individual species were analysed with a General Linear Model (GLM), partial Canonical Correspondence Analysis (CCA) and the IndVal method. The most important factors determining butterfly diversity and community composition were the presence of semi-natural habitats (deciduous woodlands and firebreaks) at the landscape scale and the composition of understorey vegetation at the plot scale. Pure effects of plot variables explained the largest part of community variation (12.8%), but landscape factors explained an additional, independent part (6.7%). Firebreaks were characterized by a higher species richness and both firebreaks and deciduous woodlands harboured species not or rarely found in pine stands. Despite the forest-dominated landscape, typical forest butterflies were rare and mainly found in the deciduous woodlands. Threatened species, such as Coenonympha oedippus and Euphydryas aurinia, were found in pine stands and in firebreaks, but were more abundant in the latter. In the studied plantation forest, the conservation of butterflies depends mainly on the preservation of semi-natural habitats, an adequate understorey management and the maintenance of soil moisture levels.