Retention Tree Groups Research Articles

Canopy gaps are less susceptible to disturbance-related and invasive herbs than clear-cuts: Temporal changes in the understorey after experimental silvicultural treatments

Forest management has a major impact on the understorey vegetation, with the intensity and type of the applied silvicultural treatments driving variable vegetation responses. We compared understorey variables across different experimental silvicultural treatments in a temperate oak-hornbeam forest in Central Hungary. Five treatment types were used in six replicates representing rotation and selection silvicultural systems: control (C), clear-cutting (CC), gap-cutting (G), preparation cutting (P), and retention tree group (R). The response of several understorey variables was investigated to the treatments in the first six years after their implementation in 2014.We assessed how understorey variables change in response to different forestry treatments, how these responses vary with time, and how game exclusion affects them. We then evaluated how well the treatments can preserve the forest character of the vegetation.We found a large temporal variability in the understorey variables over the study period. In all cases, the interventions led to an initial increase in species richness, followed by a decline later, where the regeneration layer started to close. The regeneration layer grew most intensively in G and CC. At the end of the study, R had the highest average species number, comprising a heterogeneous group of perennial forb species. The interventions all resulted in a rapid increase in total herb layer cover, mainly in favour of graminoid and perennial species. The extent of cover increase depended primarily on the amount of additional light received (CC > G > P > R > C). Turnover and beta diversity values also decreased in a similar order. The effect of game exclusion was especially pronounced in the case of the CC and G, where game browsing significantly slowed the regeneration outside the fences. The most significant changes in almost all variables were in the CC. It had the highest number of indicator species, many of them annual, disturbance-related, and invasive. G preserved the forest character of the vegetation better and proved to be less susceptible to the mass appearance of disturbance-related and invasive herbaceous species.Increasing the share of continuous cover forestry methods is crucial to preserve the forest herb layer. Rotation forestry with large cutting areas is not recommended or should be kept at low landscape rates, as these areas are highly exposed to disturbance-related and invasive species. Leaving retention tree groups can be key to the survival of numerous forest plant species.

Functional plasticity of carabids can presume better the changes in community composition than taxon-based descriptors.

Although the functional trait approach can facilitate the understanding of mechanisms that underline community responses to habitat alteration, only a few studies used this way on exploring the structure of insect assemblages compared to taxon-based analyses. We compared the descriptive power of medium-term effects (2014-2018) of forestry treatments in a temperate managed oak-dominated forest on taxon- vs. trait-based descriptors of ground beetle assemblages. The treatments included rotation forestry (partial preparation cutting, clear-cutting, retention-tree group, and mature closed forest as control) and continuous cover forestry (gap cutting) operations. The species composition was only slightly influenced by the treatments; on the ordination biplot, the control, retention tree group, and clear-cutting treatments formed relatively homogeneous groups, well separated from each other, while the others were scattered randomly in the ordination space. Over time, the species richness decreased in all treatments, but it was higher in the retention tree group treatment than in others in 2016 and 2017. The activity density also declined between years, but an immediate mass effect was revealed after the implementation of treatment types especially in the control, gap, and preparation cuts. We found that assemblages in the clear-cutting and retention-tree group had similar characteristics: high functional diversity; more open-habitat, generalist, and omnivore species and fewer carnivore species; while those in the control, gap, and preparation cutting groups had the opposite: lower functional diversity, more forest species, and more carnivorous species. Our findings will demonstrate that the simultaneous use of the two approaches will allow the most articulate understanding of the status of ground beetles assemblages in managed forests.

Unfolding the effects of different forestry treatments on microclimate in oak forests: results of a 4-yr experiment.

A stable below‐canopy microclimate of forests is essential for their biodiversity and ecosystem functionality. Forest management necessarily modifies the buffering capacity of woodlands. However, the specific effects of different forestry treatments on site conditions, the temporal recovery after the harvests, and the reason for the contrasts between treatments are still poorly understood. The effects of four different forestry treatments (clear‐cutting, retention tree group, preparation cutting, and gap‐cutting) on microclimatic variables were studied within a field experiment in a managed oak‐dominated stand in Hungary, before (2014) and after (2015–2017) the interventions by complete block design with six replicates. From the first post‐treatment year, clear‐cuts differed the most from the uncut control due to the increased irradiance and heat load. Means and variability of air and soil temperature increased, air became dryer along with higher soil moisture levels. Retention tree groups could effectively ameliorate the extreme temperatures but not the mean values. Preparation cutting induced slight changes from the original buffered and humid forest microclimate. Despite the substantially more incoming light, gap‐cutting could retain the cool and humid air conditions and showed the highest increase in soil moisture after the interventions. For most microclimate variables, we could not observe any obvious trend within 3 yr. However, soil temperature variability decreased with time in clear‐cuts, while soil moisture difference continuously increased in gap‐ and clear‐cuts. Based on multivariate analyses, the treatments separated significantly based mainly on the temperature maxima and variability. We found that (1) the effect sizes among treatment levels were consistent throughout the years, (2) the climatic recovery time for variables appears to be far more than 3 yr, and (3) the applied silvicultural methods diverged mainly among the temperature maxima. Based on our study, the spatially heterogeneous and fine‐scaled treatments of continuous cover forestry (gap‐cutting, selection systems) are recommended. By applying these practices, the essential structural elements creating buffered microclimate could be more successfully maintained. Thus, forestry interventions could induce less pronounced alterations in environmental conditions for forest‐dwelling organism groups.

Initial regeneration success of tree species after different forestry treatments in a sessile oak-hornbeam forest

Ecological, economic, and social demands triggered a shift in the management of temperate deciduous forests from rotation forestry system towards more nature-based forest management techniques such as continuous cover forestry. However, there is insufficient knowledge on the regeneration success of different tree species—especially oaks—within this management system. Through a systematic experiment, we compared the regeneration processes of a sessile oak-hornbeam forest after gap-cutting (as an element of continuous cover forestry system) to regeneration after clear-cutting, preparation cutting, and in retention tree groups (treatments of rotation forestry system). A managed, closed, mature forest was used as control. Several different aspects of the regeneration were studied: (1) seed supply of sessile oak—Quercus petraea (Matt.) Liebl., (2) species number and abundance of the natural regeneration, (3) survival and growth of individual saplings of five tree species (sessile and Turkey oak—Quercus cerris L., hornbeam—Carpinus betulus L., beech—Fagus sylvatica L., and common ash—Fraxinus excelsior L.). The number of acorns was high in closed forest, intermediate in preparation cutting and retention tree group, low in gaps, and zero in clear-cutting. Four years after the interventions, there was no detectable treatment effect on the species number of regeneration. Survival increased in every treatment compared to control, but there was no significant difference in this measure between the differently treated sites. Height growth was highest in the gaps and clear-cuts, intermediate in preparation cuts, and lowest in retention tree groups and controls. Species with different seed dispersal mechanisms responded differently to treatments: oaks were dispersal-limited in the gaps and clear-cuts, while anemochorous species (e.g., hornbeam and manna ash) were present in every treatment. The survival and growth pattern of the particular species proved to be similar, but the intensity of the response differed: shade-tolerants (hornbeam, beech, and ash) showed better survival than oaks in most treatments, and their height growth was larger. According to our results, oak regeneration establishes successfully in oak-hornbeam forests not only in the case of rotation forestry, but also during continuous cover forestry (gap-cutting). The survival and growth of the saplings are similar in cutting areas and gaps, but keeping in mind other considerations (such as preserving forest continuity, balanced site conditions, and forest biodiversity), continuous cover forestry should be preferred.

Green tree retention enhances negative short-term effects of clear-cutting on enchytraeid assemblages in a temperate forest

We studied the response of enchytraeid (Annelida: Oligochaeta) community structure to different forest management treatments via an open-field forestry experiment in a managed sessile oak-hornbeam forest in Hungary. The applied forestry treatments were 1) clear-cutting 2) clear cutting with a small patch of retention tree group, 3) gap-cutting, 4) preparation cutting and 5) control (closed mature stand).Our questions were: how did the selected forestry treatments influence the abundance and species richness of enchytraeids two years after the intervention, and how could the worms compensate the altered environmental conditions by vertical movements in topsoil layers (0–12 cm). The effects of the treatments on the species composition of the assemblages and species-specific responses were also analysed.Two years after the treatments, a strong response of enchytraeids was observed in clear-cutting and even more in retention tree group: in both treatments the abundance and the species richness of enchytraeids were reduced. Species did not show vertical movements into deeper layers and were not able to tolerate the altered soil conditions caused by changed microclimate. Gap-cutting and preparation cutting did not differ significantly from control plots. Our study showed that right after the interventions tree retention at the size of one tree height in diameter had no sheltering effect on this important soil decomposer animal community. These results are in contrast with earlier findings in boreal zones, where soil organic layer is considered a well buffered habitat against environmental changes. Oppositely, enchytraeid assemblages in a temperate deciduous forest are more diverse but seem more vulnerable to management-related alterations in soil conditions (soil temperate increment, reduced soil moisture).

Initial understory response to experimental silvicultural treatments in a temperate oak-dominated forest

In recent decades, alternative management techniques integrating conservation concerns into industrial forestry have become increasingly widespread. In order to compare the effects of various management methods on forest site and biodiversity, a systematic forestry experiment was conducted in a managed, mature oak–hornbeam forest. The present work introduces the 2-year responses of environmental variables and understory vegetation to different silvicultural treatments. These belong either to clear-cutting (clear-cutting, retention tree group), to shelterwood (preparation cutting), or to continuous cover forestry systems (gap-cutting). The experiment follows a complete block design with four replicates. Light availability was significantly higher in all the treatments than in the uncut control, with highest values in the clear-cuts. Soil moisture was highest in the gap-cuts and clear-cuts, while in the retention tree group, it did not differ from the control. Species richness, cover, and height of the understory considerably differed from the control in the clear-cuts and gap-cuts, while in the retention tree group, only species richness was higher. The establishment of ruderal, non-forest species altered the species composition in the clear-cuts. Based on these short-term responses, we conclude that as a result of the extreme environmental changes, clear-cutting in oak–hornbeam forests changes the understory vegetation considerably. Retention tree groups can maintain legacies of the understory composition for some years. Despite the experienced high vegetation cover, gap-cutting preserves the forest characteristics of the understory better than clear-cutting. These results confirm that in oak–hornbeam forests, continuous cover forestry may be more sustainable from conservational aspect than clear-cutting system.

Taxon-specific responses to different forestry treatments in a temperate forest

There are only few studies that explore the ecological consequences of forest management on several organism groups. We studied the short-term effects of four forestry treatments including preparation cutting, clear-cutting, retention tree group and gap-cutting in a temperate managed forest on the assemblage structure of understory plants, enchytraeid worms, spiders and ground beetles. Here we show, that the effect of treatments on the different facets of assemblage structure was taxon-specific. Clear-cutting and retention tree group strongly impoverished enchytraeids assemblages. Even if the species richness and cover of plants increased in clear-cutting and gap-cutting, their species composition moderately changed after treatments. For spiders only their species composition was influenced by the treatments, while the response of ground beetles was slightly affected. Short-term effect of forest management interventions on biodiversity might be compensated by the dispersal (spiders, ground beetles) and resilience (plants) of organism groups, however sedentary soil organism showed high sensitivity.

The Short-Term Effects of Experimental Forestry Treatments on Site Conditions in an Oak–Hornbeam Forest

Forest management alters forest site; however, information is still limited about how different silvicultural treatments modify abiotic conditions. We compared the effects of four treatments from three different forestry systems on forest microclimate, litter, and soil conditions. The clear-cutting, retention tree group, preparation cutting, and gap-cutting treatments were experimentally established in a European oak-dominated forest, following a complete block design with six replicates. In this study, we show the results of the quantitative analyses of 21 variables, one year after the interventions. Strong treatment effects were observed for the microclimate and litter variables, whereas the soil characteristics remained similar. The increase in light was the highest in the clear-cuts with intermediate effects in the gap-cuts. The means and variances of the air and soil temperature as well as the vapor pressure deficit were the highest for the clear-cutting treatment. An increase in soil moisture, litter pH, and litter moisture was significant in the gap-cuts and, to a smaller extent, in the clear-cuts. The soil pH increased in the retention tree groups. Microclimatic differences between the treatments were the largest during the summer, which demonstrates the buffering effect of the canopy. Our study confirms that less intensive and more spatially heterogeneous silvicultural treatments (e.g., gap-cutting) preserve a stable below-canopy microclimate more effectively. These results can support and might be useful for both forest management and conservation planning.

Factors driving tree mortality in retained forest fragments

Green tree retention is often applied in forests otherwise harvested by clearcutting. Its aim is to support biodiversity by contributing to a continuity of living trees and large-diameter dead wood in the new generation of a forest. However, high levels of mortality can undermine the aims of tree retention and pose a concern for forest managers. Therefore, knowledge about the mortality of retention trees over the long term is needed. We calculated cumulative tree mortality for the seven most common tree species up to 20years after clearcutting based on a chronosequence of dead and living trees in 583 tree groups on 206 clearcuts distributed over a wide area in boreal central Sweden. For all tree taxa combined and for the three most common tree taxa (Scots pine, Norway spruce and two birch species) we modeled tree mortality based on structural and environmental variables measured for each tree group and characteristics of the clearcut using generalized linear mixed effects models. The cumulative mortality averaged over all clearcut ages was 12% for Scots pine, 25% for Norway spruce and 16% for birch. Only 10% of all retention tree groups had a mortality >50%. Key factors reducing mortality for all tree taxa combined were tree density, tree volume of a retention group and the position at a former stand edge abutting open habitat, while a high wind exposure increased tree mortality. For the three most common species (pine, spruce, birch), the same factors as for all taxa combined were of importance, except tree density that resulted in species specific responses. For pine and spruce, the presence of seed trees on the clearcut reduced mortality. Increased mortality was additionally observed on wet soils. Mortality increased with tree height in birch and with diameter and slenderness ratio in spruce. Generally, a focus on retaining trees in groups with large tree volumes and tree density in less wind exposed positions and on forest edges will decrease the mortality after clearcutting and thus also reduce dead wood input.

The effect of stand-level habitat characteristics on breeding bird assemblages in Hungarian temperate mixed forests

The effects of stand structure, tree species composition, proportion of habitat types and land use history on breeding bird assemblages in temperate mixed forests in Western Hungary were studied. The species richness, the abundance and the composition of the whole breeding bird assemblage and of some groups formed on the basis of nesting site and rarity were examined. Stand structural variables had the highest impact on the breeding bird assemblage, while tree species composition, the varying proportion of vegetation types and land use history had no significant effect. In the case of the species richness, the abundance and the composition of the whole assemblage, the most important variables were the mean diameter of trees, the vegetation cover of the forest floor and the dead wood volume. The explained variance in the linear models of different groups varied between 20% and 60%, and the relative importance of these three variables also differed considerably. These results indicate that forest management may considerably influence the diversity and the composition of birds, as all the structural elements affecting birds deeply depend on it. Within the shelterwood management system, the elongation of the rotation and regeneration periods, and the relatively high proportion of retention tree groups after harvest could contribute to the conservation of forest birds. Our results also showed that for the forest bird communities, both the prevalence of big trees and the presence of a dense understory layer are important. Management regimes which apply continuous forest cover might be more appropriate for providing these structural elements simultaneously on small spatial scales, and for the maintenance of a more diverse bird community, thus healthier forest ecosystems.

Effects of forest heterogeneity on the efficiency of caterpillar control service provided by birds in temperate oak forests

Controlling herbivore insects by insectivorous birds is a major ecosystem service, nevertheless little is known about how local habitat features and forest management influence the efficiency of this service and about how the pest control service birds provide can be maintained and improved. We conducted an experiment in temperate oak forests in the Mátra Mountains, northern Hungary to measure bird predation rate of artificial caterpillars resembling winter moth (Operophtera brumata L.) larvae, to evaluate the relationships among insectivorous bird communities, caterpillar populations and leaf damage caused by caterpillars and to assess the effect of forest heterogeneity on these processes. We found, that structurally heterogeneous forests maintained a significantly higher abundance of insectivorous birds. Especially the tree size heterogeneity increased bird abundance. The rate of bird predation was positively related to the abundance of insectivorous birds as well as to caterpillar abundance, which indicates that birds were able to respond to caterpillar density. We were not able to demonstrate a direct negative effect of bird predation on caterpillar abundance and a positive effect of caterpillar abundance on leaf damage. Structurally heterogeneous forests, however, suffered from less leaf damage than did homogeneous forests, which result may indicates that the higher activity of insectivorous birds in heterogeneous stands resulted in lower activity of insect herbivores. Thus, we concluded that forest management can contribute to the mitigation of insect damages by maintaining the suitability of forest stands to the insectivorous bird communities through the maintenance of high stand heterogeneity and the presence of some key elements (e.g. retention tree groups, tree diversity, shrub layer).

Effects of Retention Felling on Epixylic Species in Boreal Spruce Forests in Southern Finland

We studied the effects of patch retention felling and soil scarification by harrowing on the coverage and species richness of epixylic species in boreal Norway spruce (Picea abies) forests in Southern Finland. The epixylics were investigated from both the retention tree groups (RTGs) and the surrounding felling areas before and after fellings and after scarification on consecutive years. The cover percentage of all included species groups was shown to decrease after the felling, especially in the felling areas (vascular plants, −0.4%; mosses, −27.8%; liverworts, −4.0%; and lichens,−2.1%). The decrease was considerable also in the RTGs. The amount of dead moss increased in both the areas indicating microclimatic changes. Species richness also declined rapidly after the first year, especially in the felling areas (vascular plants, −2.2%; mosses, −27.3%; liverworts, −30.3%; and lichens, −22.9%). Scarification also decreases covers and species richness of bryophytes. After the second year, the covers of the species groups generally started to regain, especially in the untreated RTGs. The size of RTG was in positive correlation with the total species number. Another main result indicates that it is possible to maintain much higher initial vegetation abundance and diversity in the RTGs than in the felling areas. Coarse woody debris formed by the frequent tree uprootings may also enhance the long-term survival of epixylics over forest regeneration period. RTGs should be at least 10 times larger than the size used in current Finnish forestry, so that they could function as species refugia.

Changes in soil fauna 10 years after forest harvestings: Comparison between clear felling and green-tree retention methods

We studied the responses of soil decomposer animals to clear felling and alternative, green-tree retention harvesting methods (GRT) in Norway spruce forests in Finland. The study plots which were sampled for immediate treatment effects (up to three years) were resampled after 10 years. We hypothesized that responses of decomposers still depends on the level of GTR. The treatments, in addition to untreated controls (100% retained), were: (1) selection felling (70% dispersed tree retention), (2) and (3) gap felling with and without site preparation, respectively (three small gaps were felled in a 1-ha area and 50% of the stand volume was retained), (4) retention felling (10% of the stand volume was retained in three small tree groups) and (5) traditional clear felling (0% retention). Ten years after the treatments the strong increase of enchytraeids immediately after the harvestings was leveled out and the numbers on all treated plots were at the control level. Total numbers or community structure of soil dwelling collembolans were not affected either. The abundance of macroarthropods on the harvested plots, except selection felling, was ca. 50% lower compared to the control forest. Their community structure was also changed due to the harvestings. The most sensitive animal groups were herbivores, microbivores and some detritivores. After 10 years, predators (mainly spiders and certain coleopterans) were decreased only under retention tree groups, while microbidetritivores (dipteran and coleopteran larvae) were unaffected. Our study showed that dispersed tree retention, at least when only 30% of the stand volume was harvested, induced no significant changes in soil decomposer animal community in boreal coniferous forest. On the other hand, decomposer community in clear-felled areas (either large or small) and in small aggregated retention tree groups is still different from that in unfelled forest 10 years after the harvestings. Mesofauna living in the soil organic layer (humus) seem to be better buffered against environmental changes induced by the harvesting compared to animals dwelling close to or on the soil surface. These animals are affected long after harvestings.

The effects of green tree retention and subsequent prescribed burning on ground beetles (Coleoptera: Carabidae) in boreal pine‐dominated forests

We studied how two methods to promote biodiversity in managed forests, i.e. green tree retention and prescribed fire, affect the assemblages of carabid beetles. Our experiment consisted of 24 study sites, each 3–5 ha in size, which had been prepared according to factorial design. Each of the eight treatment combinations determined by the two factors explored – tree retention level (0, 10, 50 m3/ha−1 and uncut controls) and prescribed use of fire (yes/no) – was replicated three times. We sampled carabids using pitfall traps one year after the treatments. Significantly more individuals were caught in most of the burned sites, but this difference was partially reflective of the trap‐catches of Pterostichus adstrictus. The fire did not increase no. of P. adstrictus in the uncut sites as much as in the other sites. Species richness was significantly affected by both factors, being higher in the burned than in the unburned sites and in the harvested than in the unharvested sites. Many species were concentrated in the groups of retention trees in the burned sites, but only a few were in the unburned sites. The species turnover was greater in the burned than in the unburned sites, as indicated by the NMDS ordinations. Greater numbers of smaller sized species and proportion of brachypterous species were present in the burned sites. Fire‐favored species, and also the majority of other species that prefer open habitats were more abundantly caught in the burned sites than in the unburned sites. Dead wood or logging waste around the traps did not correlate with the occurrence of species. We conclude that carabids are well adapted to disturbances, and that frequent use of prescribed fire is essential for the maintenance of natural assemblages of carabid beetles in the boreal forest. Small retention tree groups can not maintain assemblages of uncut forest, but they can be important by providing food, shelter and breeding sites for many species, particularly in the burned sites.

Retention-tree groups in clear-cuts: Do they constitute ‘life-boats’ for spiders and carabids?

Green-tree retention aims at maintaining biodiversity while harvesting timber, but knowledge of its efficiency is still limited. Green-tree retention may function as a ‘life-boat’ for organisms, maintain structural features, or create stepping-stones in the harvested landscape. The aim of this study is to determine whether larger retention-tree groups established in clear-cuts maintain forest species better than smaller ones. We posed hypotheses for different habitat affinity groups of spiders and carabids, and tested them in Eastern Finland with 11 retention-tree groups, 0.09–0.55 ha in size (47–385 trees per tree group). Although some of the analysed species responded as expected, we did not find unambiguous support for the hypotheses that (1) species requiring forest and/or (medium-) moist habitat should be more abundant in larger retention-tree groups, (2) species requiring (semi-) open and/or (medium-) dry habitat should be more common in smaller retention-tree groups, and (3) species indifferent with respect to light intensity and moisture should be caught in equal numbers in all retention-tree groups irrespective of their size. However, the mean numbers of individuals of many forest species and/or species requiring (medium-) moist habitat decreased over the 3 years after logging in all retention-tree groups, whereas the mean numbers of many species of (semi-) open and/or (medium-) dry habitat increased over the 3 years after logging. Furthermore, species composition changed from the pre-treatment year for spiders. We conclude that green-tree retention is justified because (1) retention-tree groups tend to maintain some forest species, thereby mitigating the effects of clear-cutting, at least in the short term, (2) they provide dead wood, large old trees, and other structural features, and (3) there are gaps in our ecological knowledge that warrant the adoption of the precautionary principle. However, as there are economic costs associated with the establishment of retention-tree groups, their ecological benefits as compared to other approaches aimed at maintaining biodiversity should be assessed carefully.

Impacts of retention felling on coarse woody debris (CWD) in mature boreal spruce forests in Finland

We studied the immediate changes in pre-treatment coarse woody debris (CWD) after retention felling and mechanical site preparation (scarification) in mature Picea abies-dominated boreal forest. Retention felling and scarification were hypothetised to affect the amount of CWD. The disturbance caused to CWD was assumed to depend on species and decay class. Logs were inventoried before fellings, after fellings, and after scarification, estimating the damage percentage for each log. After felling, 7.8% of the total pre-treatment volume of CWD was destroyed in the felled area. After scarification, the decline from pre-treatment volume was 67.6% in the felled area. The amount of CWD decreased also inside the retention tree groups; in the 1st post-treatment season, 4.6% was destroyed of the pre-treatment volume and 20% in the 2nd season. Of the retained trees, 40% were uprooted by the end of the 2nd season. If the majority of the initial downed CWD is destroyed by scarification, as our results show, we can estimate that since scarification became a widely used regeneration method in the 1960s, at least from 4 to 6 million m3 of CWD has already been destroyed in Finland. The role of CWD as a key element for biodiversity in boreal forest is generally accepted, which has led to retention of trees in fellings instead of clear felling. We suggest that at least as important as leaving trees in order to maintain continuum in CWD and species diversity is to preserve existing CWD in fellings over the regeneration phase. This can be done using less destructive harvesting methods, reduced use of scarification and placing retention tree groups in patches with high amounts of CWD.

Green Tree Retention in Fennoscandian Forestry

Boreal forest management has changed rapidly during recent years, and new forestry practices, such as green tree retention (GTR), which aim at ecological sustainability have been developed. With our present knowledge, we cannot fully assess the ecological benefits of prevailing Fennoscandian GTR levels (5-10 trees/ha). Small retention tree groups cannot be expected to provide habitat for interior forest species because of edge effects. However, even individual trees will increase the amount of snags and logs, which are important habitats for many species. According to our own results, the response of understory vegetation depends clearly on retention level. With 7% retention of stand volume, or about 50 trees per ha, we found no remarkable differences in vegetation response as compared with clear-felling. This indicates that a higher level of retention is needed to maintain late successional plant species on the site. Our data also show that there is much within-stand type variation, which correlates with the amount of CWD and stand characteristics. By locating larger retention tree groups in species-rich areas, and avoiding disturbance to CWD, a higher proportion of the resident species diversity may be ''lifeboated'' over the regeneration phase on the stand level. In this paper we review existing literature on green tree retention, with special emphasis on Fennoscandian forest management, biodiversity and productivity issues. We present results of two ongoing Finnish experimental studies that focus on the effects of various forestry practices on understory vegetation including epixylic taxa.