Introduction

Europe’s forests have a long history of human intervention (Rackham 1990). The elimination of predators, clearance of land for agriculture, introduction of domestic grazing stock, utilization of forests for wood products and the introduction of non-native species, which sometimes proved invasive, have all disturbed natural cycles of woodland regeneration (Huffel 1926). As a result, in many places, regeneration of forests is now less likely to succeed without some form of human intervention. One of the key problems facing young regenerating tree seedlings is competition from other vegetation for the acquisition of major resources like light, water and mineral nutrients. In the worst cases, young tree seedlings do not establish or do not survive this competition and thus stand regeneration is compromised; however, even when survival is not affected, undesirable plant competition can seriously affect early tree growth (Balandier et al. 2006) as well as have long-term impacts on stand production for up to 50 years or more (Wagner et al. 2006).

Since the late 1940s, in many countries of the world, competing vegetation has been controlled either by mechanical treatments or increasingly by the use of synthetic herbicides (Fortier et al. 2005; McCarthy and McCarthy 2005; Schütz 2004; Willoughby 1999; Willoughby and McDonald 1999). However, environmental concerns have led to public pressure to reduce the use of herbicides generally (Wagner et al. 1998), regardless of the actual risks to the environment of using specific artificial chemicals (Frochot et al. 2003). In fact, in many developed countries, there is a trend to consider any interventions in a forest, such as clearcuts, the use of herbicides and insecticides, or controlled burning, as undesirable (Newton and Balandier 2004). There is therefore a public demand for the adoption of novel approaches to vegetation management and particularly alternatives to the use of synthetic herbicides. Among the potential alternatives, most commonly advocated are biological control, canopy manipulation to regulate shade in the forest understorey, cover plants, cover crops or nurse plants and mechanical control. All are based on the principle of mimicking natural processes to some extent, which may be more acceptable to public opinion than the use of synthetic pesticides.

Biological control, or biocontrol, being the use of invertebrate or pathogenic natural organisms to limit the growth of competing vegetation to an acceptable level, has received increasing attention from researchers in recent years. For example, the fungal pathogen Chondrostereum purpureum is evaluated as a means of controlling sprouting of birch (Betual sp.) (Vartiamäki et al. 2008) and, as reported in this Special Issue, poplar (Populus sp.) (Hamberg et al. 2010). In the United Kingdom, Japanese knotweed (Fallopia japonica) and Rhododendron (Rhododenron ponticum) are identified as potential targets for biocontrol (Green 2003), and research is underway searching appropriate control agents in the United Kingdom and Ireland (Seier 2009). Examples of successful implementation of programmes of biological control can be found in Canada, Australia and South Africa. However, the need for high specificity between the pathogen and the target plant, and strict regulations on the development and use of biological control agents, have so far limited the use of the technique to the most frequent and often invasive plants (Seier 2009).

Partial cutting, shelterwood and group selection systems are all silvicultural approaches to stand management that aim to regulate shade in the forest understorey to limit the colonization of competing vegetation following harvesting operations. As reported in this Special Issue, the aim of stand manipulation is to change the environmental conditions at the forest floor level by manipulation of the tree canopy level, shifting the competitive strength in favour of crop trees over competing vegetation. Canopy effects are complex interplays between light, water and nutrient availability (Wagner et al. 2010). The goal is not to completely eliminate unwanted plants, but rather to limit their development such that they do not prevent the natural or artificial regeneration of tree seedlings (Gaudio et al. 2010). Although an interesting concept attempts to control understorey vegetation by canopy manipulation are often unsuccessful (Thompson and Pitt 2003), in many cases due to a lack of understanding of the ecological requirements and characteristics of particular forest plants (Gaudio et al. 2008), or because some canopy species such as pines are unable to provide sufficient shade in late stand phases to control certain competing understorey species (Wagner et al. 2010).

The techniques of using cover plants, cover crops or nurse plants, are all based on the aim of establishing plants that may themselves only be weak competitors with crop trees, but that could potentially limit the development of more competitive weed vegetation and that might also help to create an improved microclimate for the growth of tree seedlings and saplings (Balandier et al. 2009a). The use of naturally occurring or cultivated herbaceous species has sometimes led to good results but also to a long list of failures to control competing vegetation or to establish tree seedlings (Willoughby 1999; Balandier et al. 2009b). Woody plants, shrubs or trees often lead to better results (Prévosto and Balandier 2007).

Mechanical techniques such as cutting, screefing, mounding, trenching, subsoiling, clearing are still widely used in many countries (Thompson and Pitt 2003). However, mechanical treatments alone do not always succeed in controlling competing vegetation (Thiffault and Roy 2010).

Despite adverse public opinion, and the opposition of many non-governmental organisations, the use of herbicides still remains the primary technique for controlling unwanted weed vegetation in forests in many countries in the world, with glyphosate being the most commonly used compound (Thompson and Pitt 2003).

Just as the underlying climate, forest type and social and economic framework varies across the world, the techniques adopted to control weeds may also by necessity differ considerably between countries. However, to date there has been little attempt to document the similarities or differences in techniques and solutions that may have been developed in different countries across Europe. This paper therefore reports on a review of current forest vegetation management (FVM) practices in Europe, which aimed to address this information gap and hence facilitate the sharing of best practice, particularly on potential non-chemical alternatives. It is based on additional analysis of the review data presented in a book issued by COST Action E47 (Willoughby et al. 2009) and also refers to some of the research papers presented at the final scientific conference of COST Action E47, which themselves form this Special Issue of the European Journal of Forest Research.

Methodology and data

Through COST Action E47, a review was carried out on forest vegetation management practices in 18 European countries (Willoughby et al. 2009). This revealed major differences between countries in forest vegetation management methods applied, in the use of herbicides in forestry and in the uptake of different certification schemes. In this paper, we summarize the results from the review and carry out additional analysis on the underlying data. The hypothesis behind this additional analysis was that European countries are likely to apply different forest vegetation management principles and methods and that patterns in applied methods might be explained by factors such as the underlying climatic conditions, geographical location, economic development and political tradition. For the purposes of this paper, forest vegetation management methods are defined as actions taken and policies adopted on a local or national level, encompassing the vegetation management methods applied, the use of herbicides in forest management and the application of forest certification schemes. Quantitative data available from Willoughby et al. (2009) report on herbicide use in forestry and the relative uptake of certification schemes. As there are many different certification schemes in Europe, some strictly national and others international, the scope is narrowed to cover only two of the major international schemes, that of the Forest Stewardship Council (FSC 2009) and the Programme for the Endorsement of Forest Certification Schemes (PEFC).

Following FAO nomenclature (FAO 2009), the countries participating in the review can be grouped into four geographical regions. These are eastern Europe: Bulgaria, Czech Republic, Romania, Slovakia; northern Europe: Denmark, Finland, Iceland, Ireland, Lithuania, Norway, Sweden, United Kingdom; southern Europe: Greece, Italy, Serbia, Spain; and western Europe: France, Germany. The geographical regions largely but not completely cover the political regions of Europe. Namely, eastern Europe mainly covers the former Warsaw pact, northern Europe consists mainly of the Scandinavian region and the western European countries being the co-founders of the predecessor to the European Union.

Europe also is covered by several climatic zones. In this paper, countries are grouped according to the Köppen–Geiger Climate Classification (Peel et al. 2007). As climate zones do not follow national borders, we allocated countries to the climate classification group that covered the largest part of the country.

The influence of climatic conditions and geographical region on the application of different forest vegetation management methods was analysed using generalized linear models. Differences between observations were tested using Tukey’s HSD test or Student’s T-test with significance level α = 0.05. Data for herbicide use in forestry exhibited a non-normal distribution and were log transformed before analysis.

Results

Summary statistics

There is a huge variation across European countries in forest area and the proportion of the country covered by forest (Fig. 1). National forest areas range from 149,000 ha in Iceland to 27 million ha in Sweden, with a European average of 8.4 Million ha. Forest cover measured as the proportion of land covered by forest ranges from 1% in Iceland to 86% in Finland, with a European average of 33%.

Fig. 1
figure 1

Forest area and proportionate forest cover of the 18 European countries in the survey. The horizontal line indicates the average relative forest cover of the 18 countries. The vertical line indicates the average national forest area of the 18 countries

Full size image

The forests of Europe are made up of natural, semi-natural and planted woodland of a variety of species (Fig. 2). A breakdown into the relative cover of conifers, broadleaves and mixed forests shows conifer coverage ranging from more than 80% of the total forest cover in Sweden to less than 10% of the total forest cover in Serbia.

Fig. 2
figure 2

Composition of species groups in the 18 European countries in the survey

Full size image

A public/private ownership breakdown also reveals considerable variation across Europe, ranging from Norway with 12% public ownership to Bulgaria with 88% public ownership (data not shown).

The impact of competing vegetation

Across all countries, perennial grass and herbaceous species (e.g., Calamagrostis sp., Deschampsia sp., Molinia caerulea, Epilobium sp.) are identified as the main competitors on regeneration sites. Some shrubs such as Rubus fruticosus, Ulex europaeus, Calluna vulgaris can also be significant competitors on some sites. Betula sp., Populus tremula and a range of other broadleaved trees are also identified as problematic in later stages of the regeneration cycle, particularly in conifer forests. Pteridium aquilinum is also often cited as preventing tree regeneration. In drier regions of southern Europe where competition from grass and herbaceous species may be less of an issue, woody vegetation also contributes to increased fuel loads and raises the risk of fire.

Another reported trend from the country-by-country review is that the variety and competitiveness of weeds increases with soil fertility, and this is a particular challenge for afforestation of ex-agricultural sites. Particular alien invasive weeds are also identified as common problems in several countries; examples include Phytolacca sp., Rhododendron ponticum, Impatiens glandulifera, Prunus serotina and Robinia pseudoacacia.

Chemical vegetation management

The availability of detailed statistics on forest pesticide use is limited; 13 out of 18 European countries surveyed were able to quantify pesticide and/or herbicide use in their forests (Fig. 3). Herbicide use is identified as the main method for vegetation control in France, Ireland, Serbia and the United Kingdom (Table 1).

Fig. 3
figure 3

Forestry pesticide and herbicide use in 2009 in 13 out of 18 European countries surveyed. Data on total pesticide use in forestry is not available for Iceland and Ireland. Data on herbicide use in forestry is not available for Greece. Data on pesticide and herbicide use in forestry is not available for Denmark, France, Germany, Italy and Spain

Full size image
Table 1 Summary of vegetation control methods used in the 18 European countries surveyed in 2009
Full size table

Non-chemical vegetation management

In addition to herbicides, a range of non-chemical methods are used in all countries surveyed for controlling unwanted weed vegetation in forests. The main non-chemical method used is the mechanical control of unwanted plants (i.e. cutting weeds by machine or hand tools), practised in almost all European countries. This is followed by silvicultural approaches (i.e. group selection, partial cutting, shelterwood) aiming at controlling light transmitted by the overstorey to restrict weed development in the understorey, and then cultivation. Mulches are only occasionally used in some countries. Aside from grazing, biocontrol techniques are not widely adopted at an operational level (Table 1).

The impact of climatic conditions

We found a significant relation (P < 0.012) between underlying climatic conditions and the use of herbicides in forestry (Fig. 4). The region with cold overall climate and cold summers (Finland, Iceland, Norway and Sweden) appeared to have lower annual per hectare use of herbicides than other regions.

Fig. 4
figure 4

Average annual per hectare use of herbicides in forestry in four of five climate regions covering the surveyed countries. No data are available for climate class temperate/dry season/hot summer. Data were log transformed prior to analysis

Full size image

Regional trends

There are regional patterns in the uptake of the two certification schemes. FSC is to a greater or lesser extent adopted in 17 out of 18 countries surveyed. PEFC is present in 10 of the surveyed countries. PEFC, however, covers nearly four times as much forest area as FSC (55 Million ha vs. 15 Million ha). Within the western region, the average national forest area covered by PEFC is significantly (P < 0.045) higher than that covered by FSC (Fig. 5). Apparent differences between PEFC and FSC observed in other regions are not statistically significant.

Fig. 5
figure 5

Average national forest area covered by FSC or PEFC certification schemes in 2009 in four geographical regions in Europe

Full size image

Discussion

The impacts of competing vegetation

In the majority of forests across Europe, as in other regions across the world, controlling competing vegetation to favour tree seedlings is viewed as a critical silvicultural operation. Tree regeneration may be impossible or at the very least severely delayed without some form of intervention, be it through management of an existing overstorey of trees to favour seedlings or to create conditions where weeds compete less vigorously, manipulation of browsing pressure, biocontrol, or the direct physical control of vegetation.

Throughout the European countries reviewed, the impact of weed competition on the growth and survival of tree seedlings is an important consideration; however, to date the precise nature of competitive effects has only been studied in detail in relatively few countries. In this Special Issue, Gaudio et al. (2010) report new work in this area involving three understorey plant species common in central France—Calluna vulgaris, Pteridium aquilinum and Molinia caerulea—which were found not to interfere with the survival of pine (Pinus sylvestris) seedlings, but to significantly reduce diameter growth. Detrimental plant density level varied between the three understorey species, with 16 plants m−2 of Calluna and Molinia appearing to permit acceptable tree growth compared to a threshold of only 10 plants m−2 for Pteridium inducing already a significant loss (38%) in pine diameter (Gaudio et al. 2010). Also in this Special Issue, Prévosto et al. (2010) reports that in southern France grass cover is the main factor limiting seedling survival of seeded Holm oak (Quercus ilex) and downy oak (Q. pubescens). Again in this Special Issue, Dodet et al. (2010) report that on fertile sites, targeting herbaceous vegetation provides the biggest improvement in crop tree regeneration and diversity, but on less fertile sites only the most intensive vegetation control proved beneficial.

The interactions between tree seedlings and the surrounding vegetation often imply a complex balance of positive (facilitation) and negative (competition) effects (Callaway and Walker 1997), and depending on the strength of the two processes, the presence of the vegetation can facilitate seedling establishment and growth. The vegetation can act directly, for instance by providing a protective microclimate around the young trees (Balandier et al. 2009a) or indirectly when a third component modifies the interactions between tree seedlings and the surrounding vegetation. Such an effect of indirect facilitation can, for example, be found in the case of canopy manipulation, where the shade provided by the overstorey limits the development of the understorey vegetation, and thus indirectly reduces the competition between young trees and the surrounding vegetation (Pages and Michalet 2003). Because the magnitude and precise nature of competition and facilitation is so dependent on local site conditions, plant life stages and sizes (Callaway and Walker 1997), with our current level of knowledge, it is difficult to make broad generalizations that can be applied to a wide range of sites across Europe. However, the role of facilitation is often identified as crucial in southern Europe, where tree seedlings could be exposed to high temperatures and dry air, which can increase water stress. Numerous studies have thus highlighted the beneficial role of nurse shrubs in facilitating tree regeneration in the Mediterranean area (Gómez-Aparicio et al. 2005; Smit et al. 2008).

Pesticide use in European forestry

Herbicides offer one effective means of controlling vegetation identified as problematic in the review (Willoughby et al. 2004; Gama et al. 2006). However, there has been increasing pressure in recent years to reduce reliance on chemical use alone. Policy drivers for this pesticide reduction fall into three main categories: (1) National policies restricting pesticide use; (2) Independent forest certification schemes and (3) European Union policy on pesticide approvals, which has led to a review of the safety and efficacy of all registered pesticides resulting in the withdrawal of a large number of products (Willoughby et al. 2009). It is difficult to accurately quantify how effective these policies have been in restricting the use of pesticides in European forestry. The availability of detailed statistics on forest pesticide use is limited—only 13 out of 18 European countries surveyed were able to quantify pesticide use in forests. Although the estimates obtained are only indicative, do not give details of trends over time and are often based on expert judgement, they nevertheless provide a valuable comparison of indicative usage throughout Europe.

All countries surveyed reported that chemical agents were still used in their forestry management. However, for some countries, data were not sufficiently detailed to allow forestry use to be identified separately from that of other land use types. The use of pesticides in forestry reported ranges from 0.0002 to 0.69 kg a.i. ha−1year−1 Footnote 1 averaged across total forest area in the countries. By comparison, in the same countries, agriculture uses significantly higher quantities of pesticides, 0.24–1.84 kg a.i. ha−1 year−1 (Willoughby et al. 2009). However, a perennial crop forest is not in fact treated with pesticides every year, but only very limited data are available on pesticide treatment frequencies in European forestry. A review of herbicide use in forestry in the United States suggests an application rate of 0.6 kg a.i. ha−1 year−1 on area treated with herbicides in national forests (Shepard et al. 2004). In 2008, Canadian forestry applied 351,832 kg a.i. herbicides on 122,878 ha of forest corresponding to 2.9 kg a.i. ha−1 on forest area treated with herbicides (National Forestry Database 2010). Glyphosate made up 96% of the applied amount. Data from Willoughby et al. (2009) covering seven countries suggest herbicide application rates from 0.1 to 2.2 kg a.i. ha−1 year−1 on forest area treated with herbicides.

Pesticide use in forestry is not confined to vegetation control. In Finland and United Kingdom for example, the use of urea for stump treatment against Heterobasidion annosum is far higher than the use of herbicides. In the Czech Republic, 84% of the total pesticide use is as repellents, whereas herbicides account for only 7%.

Some countries in the survey, predominantly in Scandinavia and in Central Europe, have enacted voluntary restrictions on pesticide use in forests. These restrictions may encompass the entire forest area of a country or only parts of the forest area relating to ownership or protection status. In these countries, non-chemical vegetation management is usually the primary method of vegetation control adopted.

Forest certification and pesticide use

For the purposes of this paper, we have considered forest certification to be a forest vegetation management ‘method’, because forest owners and managers adopt certification schemes on a voluntary basis and in doing so restrict themselves in their potential future management actions. For example, certification by FSC or PEFC requires that pesticide use for vegetation management is reduced and eventually eliminated (Forest Stewardship Council 2007; Programme for the Endorsement of Forest Certification Schemes 2009). However, our review of the current status of vegetation management shows no evidence that pesticide use is actually lower in countries with a high proportion of certified forests. This may be for several reasons. Firstly, as already noted, our data are based in part on expert judgement and do not show trends over time. Neither FSC nor PEFC dictates an instant exclusion of pesticide use, but instead require that managers work towards eventual elimination of all usage. A considerable time lag can thus be expected between the adoption of a certification scheme and the point at which significant reductions in pesticide use are observed, and our data may not be suitable for identifying such trends over time in any case. Another factor may be that pesticide use may have been largely eliminated at a national level without reference to or long before certification was adopted, which is the case for example in Norway and Denmark. In Norway, the ‘Living Forest Project’ (2009) stating that herbicide use in forests should be avoided was established in 1998, before large parts of Norway’s forests were certified. State-owned forests in Denmark have since 1996 strived towards a reduction in pesticide use. By 2002, the pesticide use had dropped to 10% of that in 1995 (Kristoffersen et al. 2004), and by 2003, a voluntary agreement covering all publicly owned area came into force that banned the routine use of pesticides. FSC and PEFC certification of the Danish state forests was enacted 4 years after this national initiative, in 2007 (Danish Forest and Nature Agency 2009).

Non-chemical vegetation management

Results from the review suggest that contributing countries with broadly similar weed species, climate and economic conditions already appear to have adopted similar approaches to the non-chemical management of competing vegetation. Cultivation appears to be most effective and hence most widely practised on the driest and least fertile sites (Hytönen and Jylhä 2008). Eight countries spread across Europe report that mulches are used occasionally as a means of controlling vegetation. No country report that mulches are the main method or main alternative. In countries where higher value roadside or arboricultural plantings are common, the use of mulches tend to be more common (Van Lerberghe 2004). Biological control using invertebrate or pathogenic natural organisms is not widely used as a vegetation control method in European forestry; however, research takes place in several countries such as Finland, United Kingdom and Ireland. In Greece, biocontrol means grazing by cattle or sheep as a means of controlling competing vegetation (Papachristou and Platis 2009). Grazing is also seen as an important tool in other southern European countries (Harper et al. 1999). Stand manipulation is most widely practised in the countries with a high proportion of forest cover and a long tradition of silviculture supported by forest policy. The main barriers to the further adoption of proven methods of non-chemical weed management therefore seem to be excessive cost or perceived lack of efficacy in particular circumstances (Willoughby et al. 2004).

The impact of climate

The composition of competing vegetation in forestry is a function of climatic conditions (temperature and precipitation) and natural migration or artificial introduction of species. The competitive interactions between vegetation and trees also depend on tree species. Countries surveyed span five different climate classes according to the Kobben–Geiger Classification (Peel et al. 2007). Data on herbicide use were available for four climate classes and show that countries with cold summers (Finland, Iceland, Norway and Sweden) on average use less herbicide than countries with warm or hot summers, suggesting more competitive conditions in the latter case. However, the causal relation is not unambiguous as the majority of countries with colder summers have also enacted voluntary national restrictions on the use of pesticides in forestry (Willoughby et al. 2009).

Future European research collaboration

COST Action E47 successfully brought together practitioners and scientists from 18 European countries to share the best practice and the latest scientific advances in the field of forest vegetation management. The primary means for achieving this exchange of information was through participation in a series of field meetings and conferences that examined research and operational practice in a range of contrasting European conditions. The discussions and research that were fostered by those meetings are recorded by Willoughby et al. (2009), and scientific papers from the final conference of the COST Action (Bentsen 2009) are presented in this Special Issue of the European Journal of Forest Research. Summing up the work of the COST action, Willoughby et al. (2009) suggested that if a ban on the use of pesticides in forestry across Europe was to occur in the near future, this could have significant and unintended negative impacts on the health of many of Europe’s forests, unless considerable extra funding was made available to forest managers to adopt existing non-chemical weed control methods. The development of more cost-effective practical guidance for managers across Europe on non-chemical control methods can best be brought about by future collaborative research into more integrated methods of managing forest vegetation, through the identification of silvicultural approaches to reduce or eliminate pesticide use rather than viewing control methods isolated from forest management (see also Little et al. 2006). In addition, gaining a better understanding of the fundamental mechanisms and impacts of competition between species is of paramount importance in the development of efficient control methods for a more integrated approach to vegetation and forest management. Such a move towards integrated forest vegetation management would support important new European policy initiatives on the sustainable use of pesticides, and benefit Europe by ensuring its forests are being managed and regenerated in a more sustainable fashion.

Conclusions

This review of forest vegetation management practices has shown that many European countries face the same challenges regarding establishment and tending of young tree stands, but countries with different climates, forest types and economies by necessity address the problem by different means.

National policies targeted at reducing the dependence on pesticides, be they certification or other national initiatives, are enacted in most countries, but we found that the instruments chosen differ across regions and across countries. Herbicides are still utilized in all countries, but the extent to which they are used varies greatly.

Any reductions in herbicide use achieved do not seem to have been driven solely by participation in forest certification schemes. Other factors, such as national initiatives or the availability of additional resources to implement more expensive non-chemical approaches may be equally important.

Several common information gaps relating to forest vegetation management in Europe can be identified from this work. From these, four high priority areas for future European research collaboration have been proposed: alternative control methods; environmental impacts of non-chemical methods; ecosystem responses to specific problem species, with a special emphasis on invasive species, or functional groups of weeds; and social issues and public perceptions (see Willoughby et al. 2009 for further detail).