Primeval Forests, Natural Forests and Managed Forests in the Context of the Biodiversity Debate and Climate Protection

1.    Introduction

In recent years, numerous papers have been published that compare the ecosystem services of managed forests with those of primeval forests or former managed forests that have been left unmanaged for many years and so come close to their natural state (hereafter referred to as natural forests) (see also Box 1: What are primeval forests, what are natural forests?). While studies in tropical and boreal forests with increasing utilization have mostly found significant declines in forest-typical biodiversity (e.g. Alroya 2017, Giam 2017, Pyles et al. 2018, FAO & UNEP 2020), studies in temperate forests presented below have somewhat different findings (e.g. Schulze 2018, Dieler et al. 2017, Schall et al. 2020). These can essentially be summarized in the following points:
(1) primeval forests and natural forests show lower biodiversity than managed forests in terms of the respective taxa studied and would therefore not be of particular significance for nature conservation, and
(2) non-utilization of primeval forests and natural forests would be detrimental to climate protection because only utilization can contribute decisively to the reduction of greenhouse gas emissions.

Particularly in the context of climate policy recommendations, this topic is highly relevant and requires comprehensive scientific consideration.

---

Box 1: What are primeval forests, what are natural forests?

Definitions from Buchwald 2005, Biris & Veen (2005), Fanta (2005), Wirth et al. (2009), Commarmot et al. (2013), Barredo et al. (2021, Sabatini et al. 2021).

In the English-language scientific literature in particular, the German term Urwald – or synonymously Primärwald – is paraphrased with a large number of terms that have the same meaning (intact, mature, natural, primary, primeval, undisturbed, untouched, virgin). There one finds agreement on the defining descriptions: a primeval forest is a large-scale forest ecosystem for which no direct human interventions are known and the composition of the natural biotic communities and the forest-typical processes have never been significantly altered. A primeval forest consists of tree and shrub species which, in their various life cycle stages, are typical of the location and biogeography. Characteristic are usually large stocks of deadwood in different qualities (standing, lying, young, old) and a mostly complex vertical and horizontal forest structure resulting from undisturbed natural dynamics. 

Since primeval forests have not existed in Germany and Central Europe for a long time, near-natural forests that have remained unmanaged for a longer or shorter time (or show only minor traces of utilization) often serve as a reference in comparative studies of forest ecosystems: according to Vandekerkhove et al. (2007) and Commarmot et al. (2013), these can be referred to as “natural forests”. Typical descriptive terms for such woodland are “ancient”, “near-virgin”, “old-growth”, “quasi-virgin”, or “widely undisturbed”. Natural forests have emerged from natural regeneration and have developed freely over long periods of time without human intervention (Fig.  1). Unlike primeval forests, they display influences of utilization that can be documented to this day. Such forests contain those (tree) species that would also occur in the natural plant community at that particular location. Natural forests display the natural development cycle up to the decay phase, but, depending on the length of time without utilization, they still have an incomplete inventory of structures and stages of development typical of primeval forests. Natural forests can become very similar to primeval forests over long periods of time, and in landscapes with high levels of disturbance (e.g., floodplains), this process can also be relatively rapid. The conceptual basis for the idea of natural forests, i.e. forests taken out of utilization and to which dedicated research activities are assigned, was already established in the 1930s by Hesmer (1934) and Hueck (1937) under the term “natural forest reserves”.

In the political context, however, even the definition of the typological terms “primeval forest” and “natural forest”, which are actually sufficiently well delineated in the scientific literature, is the subject of intense and controversial debates. In a recent study by the European Forest Institute (EFI), it is stated that the protection of the remaining European primeval forests and natural forests is difficult because there are no clear and unambiguous terminological definitions and boundaries. The terminological discrepancies are in part due to the incomplete and inconsistent data and cartographic material on these categories. Thus, they are also difficult to apply in policy and in practical guidance (O'Brien et al. 2021). The EU’s draft post-2020 forest strategy also mentions that there is still no definitive agreement on the content of the terms primeval forest and natural forest (EU 2021). While this terminological debate rages on, the remaining primeval and natural forests are steadily being logged.

---

We see our essay as a contribution to currently topical political debates and to course-setting in nature conservation and climate policy: (1) at the EU level, intensive and controversial discussions are ongoing as to the instruments to be used in new EU biodiversity strategy; (2) the new EU climate protection targets to achieve the goals of the Paris Climate Agreement require adjustments to their instruments; (3) these EU climate protection targets along with the recent decision of the German Federal Constitutional Court, which declared key parts of the German Climate Protection Act of 2019 unconstitutional and emphasized the precautionary duty to future generations, require changes to legal norms and readjustments of political instruments.

2.    Background – Where are there still primeval forests in Europe?

According to an assessment by Forest Europe (2020), Europe, including the Eastern European countries and the European part of Russia, has approximately 227 million hectares of forests; this represents 33% of the land area. Only about 4.6 million hectares (2.2%) of European forests are still characterized as primeval forests or natural forests; of these, about 3.6 million hectares are in the EU (2.4%). Five years earlier, the Forest Europe study (2015) still classified 7.3 million hectares (3.3%) of European forests as primeval forests or natural forests; this would represent a decrease of 2.7 million hectares (about 40%) between the two reporting periods. We suspect that this massive decline is not only explained by actual new land utilization and thus losses, but also includes statistical effects. In the Romanian Carpathians, for example, large areas of primeval and natural forests are being downgraded by definition, because politicians want the land to be opened up for utilization (Luick 2021, Luick et al. 2021).

We know little about the exact distribution and condition of Europe's remaining primeval and natural forests. This is also clear from the study by Sabatini et al. (2018) according to whom there are an estimated 1.4 million hectares of primeval forests remaining in Europe, excluding Russia; of these, about 1.1 million hectares are boreal forests, about 0.2 million hectares are montane beech and beech-fir forests, and about 0.07 million hectares are subalpine temperate conifer forests. This is about 0.7% of the total forest area. Germany has not had a primeval forest for a long time. A study by the Joint Research Centre (JRC) of the EU on the existence of primeval forests and natural forests in EU countries determined that together they account for less than 3% of the total forest area and are also predominantly small-scale and fragmented (Barredo et al. 2021). In another analysis, Sabatini et al. (2020) report that six of the 54 differentiated European forest types no longer have any primeval forest reference areas and that 70% of the forest types have less than 1% of primeval forest reference areas.

In relation to this very small proportion, and without taking into account the boreal primeval forests in the northern regions of Scandinavia and European Russia, about 80% of the temperate primeval forests of Europe are located in the Carpathian arc in Ukraine, Romania and Slovakia (Fig. 2). In the EU, no member state has as many temperate, hardwood primeval forests as Romania. According to current estimates, these represent about two-thirds of the remaining temperate primeval forests, but even in Romania they represent only 0.5 to a maximum of 1% of the total forest area, and they continue to shrink (Biris 2017, Luick 2021, Luick et al. 2021). In the period 2001 to 2019 alone, according to a study by Global Forest Watch (2020), Romania lost about 350,000 hectares of primeval and natural forests through illegal (and also legal) logging. Large-scale clear-cutting takes place even in protected areas such as national parks, UNESCO World Heritage and Natura 2000 sites (Luick et al. 2021; see also Box 2 and Figs. 3, 4a to c). The importance of the last primeval forests and natural forests in the Carpathians).

---

Box 2: The importance of the last primeval forests and natural forests in the Carpathians

It may well be that cheap shelves in a German furniture store or planks and boards in any DIY store come from Romanian, Slovakian or Ukrainian primeval mountain spruce forests (Fig.  5). Even logs, wood pellets or wood briquettes, which are offered in sacks, containers and on pallets in many DIY stores or on the internet, often point via their labels to origins in Eastern Europe. Not infrequently, these products are made from centuries-old giant trees. The enormous global demand for packaging materials, due to the rapidly growing internet trade, so-called “fresh & tasty packaging” and seemingly “sustainable”, wood-based substitutes for packaging and transport materials previously made of plastic, are further drivers and increase the pressure to develop cheap wood resources. The formula for success is: cheap resources due to cheap labor and low concession levies and rents with high volumes of wood per unit area, minimal requirements for occupational safety and corruptible social structures. This guarantees high profits on our markets and explains why in recent years numerous international corporations with large factories that have enormous logging and processing capacities have settled around the Carpathian Arc. All of them have one thing in common: a high and increasing demand for wood as raw material (see Luick et al. 2021). 

In all the countries of the Carpathian region (Poland, Romania, Slovakia and Ukraine), this explains the massive pressure to exploit the last primeval forests and natural forests, up to and including large-scale deforestation. This is even affecting important protected area categories such as UNESCO World Heritage Sites, national parks and Natura 2000 sites. Germany, as an importer, processor and then exporter of such wood and wood-based products, is also partly responsible for the pressure to exploit these forests. This means that we are also involved as consumers. Yet it is equally true that Germany could provide support to permanently preserve the region’s natural heritage due to its huge significance for biodiversity. 

It is certainly possible to justify an ethical, scientific, and even self-interest-based commitment to protect the last remnants of large-scale (European) primeval forests. Important reasons for doing so are:

(1) Primeval forests are places (ecosystems) not directly influenced by civilization, which preserve genetic reserves vital for the adaptability of forests. Here, the intraspecific variability, as it has differentiated out over very long periods of time, has remained untouched by utilization-oriented selection. This also applies to species-specific adaptation to abiotic and biotic environmental factors unaffected by anthropogenic selection. The existence of genetically diverse populations has exceptional importance in the face of climate change and the search for climate-adaptive tree species and provenances. Primeval beech and fir forests exist in the Carpathians over a wide geographical and climatic gradient; some associated species are classified as tertiary relicts. The genetic diversity and thus the potential for “climate-adaptive” evolutionary development of these species is significantly lower in regions to which they migrated only in post-glacial times and over long distances (such as the present areas of distribution in Germany) than in their periglacial refugial areas.

(2) In their temporal and spatial dynamics, primeval forests are refugial- and source biotopes for highly specialized species that depend on the long-term stable habitat requisites and environmental conditions that exist in this quality only in primeval forests (e.g. fastidious xylobionts among the fungi, lichens, beetles, hymenopterans and dipterans). Only in large-scale primeval forests is there the necessary long-term continuity of habitat and thus the corresponding structures and processes that correlate with the maturity and age phase of the trees. 

(3) Primeval forests are rare learning sites for principles and knowledge that also have great practical and thus economic relevance for managed forest ecosystems. From primeval forest research come practice-oriented findings on threshold values for a minimum provision of deadwood, habitat trees, primeval trees, disturbance areas and microstructures to achieve forest-specific biodiversity in managed forests. They are indispensable objects of reference for developing concepts of sustainable forest management.

(4) Primeval forests are reference systems and important research laboratories in which long-term trends in environmental change can be documented and analyzed without being overshadowed by management cycles. They are also reference systems for comparing natural forest development with managed forests and thus serve to guide adaptive forest management through the iterative development of climate change adaptation and mitigation strategies.

---

3.    The notion that “primeval and natural forests do not make an important contribution to biodiversity conservation due to their low biodiversity”

To compare the biodiversity of managed and unmanaged (primeval and natural forests) European forests, reference is often made to the meta-study by Paillet et al. (2010). This summarizes, among other things, comparative studies in Central European deciduous forests that found higher vascular plant diversity in managed forests than in unmanaged forests. In a sectoral interpretation, this has established the narrative that primeval and natural forests would generally have lower a-biodiversity than managed forests. (see also Supplement 3: The biotic richness and condition of Central European forests). It is certainly interesting and useful to compare differently utilized forests with regard to their ecosystem services, which can also include biodiversity. However, it is problematic to deduce from this (1) that the high biodiversity desired from a nature conservation point of view is inherently aided by forest management, or even that forest management is a prerequisite for such biodiversity; (2) that managed forests would per se have an equally high or even higher nature conservation value than primeval or natural forests; and (3) that the designation of new protected areas (including wilderness and areas allowing natural processes) is unnecessary and should be rejected (e.g. Walentowski et al. 2013, Schulze et al. 2014, Schulze 2018, Dieter et al. 2020, Dieter 2021).

Comparative studies on the biodiversity of Central European forest ecosystems are confronted with the fact that almost every forest has a centuries-old history of utilization. Thus, potential reference systems are not primeval forests but at best natural forests. In this context, the forest area of the Hainich in Thuringia, Germany and especially the Hainich National Park play a central role (Fig. 6). The Hainich is a real laboratory for numerous studies; including, among others, the project "Biodiversity Exploratories in Germany" (see also Section 4 and Box 3). With an area of 130 km², the Hainich is the largest contiguous deciduous forest area in Germany. Its southern part with 75 km² was designated a national park in 1997. In 2011, 1,573 hectares in the central parts of the Hainich containing especially semi-natural old beech forests were recognized and designated part of the UNESCO World Heritage site "Ancient and Primeval Beech Forests and Primeval Beech Forests of the Carpathians and Other Regions of Europe". Regardless of site-specific variations, the forests in the Hainich are characterized by centuries of diverse forest utilization, display various developments, and are still clearly differentiated to this day. These features include, for example, afforestation, forests that emerged from coppice and coppice with standards, pasture forests, and more recent successions of open land with different starting points (NPH 2012, see also Figs. 7a and b). Forestry utilization in the deciduous forest of the Hainich National Park has been discontinued over a wide area for only about 20 years – specifically since 1998 – and only in very small subareas has no utilization taken place since the 1960s.

Thus, the Hainich is neither primeval forest nor large-scale natural forest since it lacks the influence of prolonged natural disturbances and (aging) processes. This fact is not always correctly taken into account in the interpretation of data collected from the Hainich National Park, or is ignored altogether. In the study by Schall et al. (2021), the heterogeneity of managed beech forests is seen as a key factor for biodiversity. Their assessment is based on a plot comparison of the occurrence and distribution of deadwood, microhabitats, vascular plants, and herbivorous and carnivorous arthropods in two differently managed beech forests in comparison with unmanaged beech forests. Data sets from the Hainich National Park serve as a reference for these natural beech forests, but they are not presented in a nuanced manner against the background of the history of the forest’s utilization and its heterogeneity. Also problematic is the fact that the individual plot data from the Hainich are summarily placed in an interpretative context with data from the federal forest inventory and generalizations are derived on this basis.

---

Box 3: The biotic richness and condition of Central European forests

The forests of Central Europe are relatively poor in tree species, but they possess a highly diverse fauna, fungal flora and flora of herbaceous and lower plants. However, this high diversity of species is only found in near-natural forests and historical-traditional forest forms such as pasture forests or oak coppice-with-standards forests. If only beech forests are considered, more than 11,000 previously known eukaryotic organism species (excluding bacteria) have been counted as indigenous forest species, including more than 6,800 animal species, 3,345 fungal species, 280 lichen species, 190 moss species, and 215 herbaceous layer plants, along with several hundred facultative vascular plant species that also occur outside the forest (Scherzinger 1996, Assmann et al. 2007, Dorow et al. 2007). In one Hessian beech natural forest alone, more than 6,000 animal species were recorded on 60 hectares (Dorow et al. 2007). In many cases, highly specialized species show higher densities in natural forests than in managed forests, and may even be completely absent in the latter (Friedel et al. 2006, Dörfelt 2007, Müller & Bütler 2010, Hauk et al. 2013, Dvorak et al. 2017, Kaufmann et al. 2017, Gerlach et al. 2019, Jacobsen et al. 2020). In the case of insects and fungi in particular, there are species which depend on forest wilderness that are generally extremely weak in dispersal and thus rare and highly endangered. Many deadwood dwellers require significantly higher supplies of deadwood than are commonly found in managed forests and, given a sufficiently large habitat, benefit significantly from increases in deadwood quantity and quality (Flade & Winter 2021, Rosenthal et al. 2021).

The forestry practices that have long dominated Central Europe (including even-aged forests with non-native spruce or pine) have led to losses of forest-typical species and habitats. Increasingly long Red Lists and Appendices of species protected by law (e.g. the EU-FFH-Directive and the various regional nature conservation and forest laws) indicate that xylobiont beetles, birds, fungi, lichens and mosses, in particular, which are dependent on deadwood and old trees, have been pushed back by forest management to a few small and mostly isolated refugia. Also highly rare and endangered are heliophilous biota that are sensitive to specific disturbances connected to natural dynamics (Fartmann et al. 2021, Rosenthal et al. 2021). 

Only in forest ecosystems rich in disturbances, such as in natural or near-natural mountain spruce forests (due to bark beetles and windthrow) or in floodplains (due to flooding and bedload), can the amount and heterogeneity of deadwood increase rapidly with sufficient likelihood. The significant and rapid increases in species numbers and in the abundance of disturbance-affine species illustrate this correlation (Rosenthal et al. 2015). This highlights the central role of natural forests of sufficient size that have remained unutilized for long periods of time in ensuring the protection of forest-typical flora and fauna. Conversely, forests that rely on specific traditional forms of forest utilization can be exceptionally rich in endangered species (e.g., wooded pastures and oak coppice-with-standards forests; see Schoof et al. 2018, Fartmann et al. 2021). In the context of biodiversity conservation, managed and protected forests are complementary concepts, not opposites. 

However, systematic monitoring of species diversity in the various forest communities and their alterations has only recently begun to be carried out – a problem that is not specific to Germany. Only for avifauna do we have longer term data series. The bird indicator for forests produced by the German Federal Agency for Nature Conservation shows a slight positive population trend over the last 30 years. However, only the populations of 11 more or less forest-bound species were considered (Kamp et al. 2021). Detailed figures are available for the forests of England since 1970: here, considerably more bird species were studied and distinguished according to those with broad (generalists) and those with narrow habitat requirements (specialists). In English forests, overall woodland bird populations have declined by 28% since 1970, those of specialists by as much as 41%, while generalist populations have increased by 7% (DEFRA-UK 2020). Similar developments can be expected in German forests, as initial evaluations of long-term population trends of all forest birds show (Gerlach et al. 2019). According to these, species tied to natural forests with low disturbance intensities, such as the white-backed woodpecker, gray woodpecker, collared flycatcher, and black woodpecker, have shown declines in recent decades; these species may benefit from more protection of natural processes. In saying this, we do not question that near-natural managed forests are also important for biodiversity conservation.

The "insect die-off” has apparently also affected Central European forests. As part of the Biodiversity Exploratories project (https://www.biodiversity-exploratories.de/de/), the insect fauna at 150 sites in the three areas of Hainich, Schorfheide-Chorin and the Swabian Alb were also studied annually between 2007 and 2017 (Seibold et al. 2019). The results were alarming: regardless of the type of forest management and also including forests taken out of use, a 41% decline in insect biomass and 36% decline in absolute species number was found for this 10-year interval, while there were no correlations with abundance. At the insect taxa level, heliophilous species and those dependent on tree senescence were most affected. The researchers cannot yet come up with clear explanatory causalities, but trends suggest: (1) that the environmental influence of intensive anthropogenic practices also affects forest areas, and (2) that the forest areas taken out of use are far too small in number and area and too isolated to compensate for negative effects. 

Data from the other two major forest biomes – tropical and boreal forests – show, through direct comparisons between primary forest and managed forest ecosystems, that the diversity of closely forest-bound taxa of mammals, birds, insects, and other animal groups declines even at low levels of forest utilization, i.e., characteristic species are lost (e.g., Burivalova et al. 2014, Franca et al. 2017, Niemalä 1997). At low intensities of utilization, the decrease in diversity is in some cases numerically compensated by immigration of non-forest species benefiting from disturbances, though this alters the forest-typical species composition (see, e.g., Schmidt 2005). There is no reason to assume that temperate primeval forests and natural forests do not respond to timber harvesting and associated disturbances with comparable loss of forest-typical biodiversity.

However, “species counting” neglects the fact that the original benchmarks of many forest ecosystems are simply unknown and that various (biotic and abiotic) ecosystem-shaping factors (including storms, fire, floods, calamities, megaherbivores, predators) have already been deliberately eliminated by humans a long time ago (e.g. Vera 2000). Thus neither tree species diversity nor herbaceous plant diversity are, by themselves, useful indicators for characterizing the species diversity of a Central European forest. Focusing on a specific a-diversity is also problematic because it is not a suitable assessment variable, especially for nature conservation objectives at higher spatial scales. Natural peatland forests or even boreal coniferous forests are undoubtedly poor in species, but the associated biocenoses are composed of many highly endangered species. Despite this obvious connection, comparative studies in the context of forestry versus process conservation lack this holistic perspective.

In a recent assessment of forest ecosystem services in the EU-28 countries, serious negative developments were identified and this despite the fact that since 1990 the forest floor area has increased by 13 million hectares through succession and afforestation (Maes et al. 2020). For example, according to Hansen et al. (2013), the indicator “tree cover loss”, which was 27% in the period 2000-2012, has increased to 74% in the period 2009 - 2018. Causes are considered to be the increased logging and the more frequent consequences of calamities, general pressures (mainly nitrogen emissions), diseases, extreme weather events and forest fires. Maes et al. (2020) assess the overall ecological status of forests as a concern: out of 81 forest habitat types, only 14% are in a favorable condition; 53% are in an unfavorable-insufficient condition and 31% are in an unfavorable-poor condition; for about 2%, the condition is unknown. Nitrogen inputs also pose special challenges for the protection of natural processes. In some regions, emissions (mainly from large livestock farms) are so high that renouncing any forest management (= no nitrogen removal) can also lead to eutrophication on sensitive sites (see Bobbink et al. 1998, Hermann et al. 2020).

---

Due to their rarity, there are only a few meaningful studies comparing the biodiversity of temperate primeval forests with that of managed forests under analogous geographical conditions. For this reason, the comparative study of three primeval beech forests in the Slovakian Western Carpathians, which are part of the UNESCO World Heritage Site "Ancient and Primeval Beech Forests and Primeval Beech Forests of the Carpathians and Other Regions of Europe", with three neighboring geographically-similar beech managed forests aged between 80 and 100 years (e.g. Kaufmann et al. 2017, Kaufmann et al. 2018, Leuschner et al. 2021) is particularly informative. The data are based on an analysis of a total of 150 recording plots, 30 of which were in managed forests and 10 each in the predominant phases of primeval forests (with development-, optimal- and decay phases). The important findings are:

• The diversity of lichens is twice as high in primeval forests as in managed forests.
• The diversity of mosses is 50% greater in primeval forests than in managed forests.
• For vascular plants, managed forests have higher species numbers at the plot level (a-diversity); at the landscape level (?-diversity) species richness is similar in primeval and managed forests

The comparative study demonstrates the great importance of the spatial heterogeneity of primeval forests through the occurrence of all tree age classes and forest development phases. Forest management can favor the immigration of non-forest plant species, but usually leads to a homogenization of the tree stand and the loss of the senescence phase, i.e. the aging and decay phases, of forest development with their associated habitat structures (ibid.).

How do these seemingly contradictory results on the biodiversity of forest ecosystems and the influence of forest management arise? In comparative biodiversity studies, it is often the case that only vascular plants are taken into account. From the consideration of this single taxon, values for biodiversity conservation are then inferred. This is demonstrated by the meta-study by Bernes et al. (2015) of comparisons of managed and natural forests in boreal and temperate regions in terms of their importance for nature conservation: in around 17,000 studies, it was almost exclusively forest structures and vascular plants that were considered as comparative parameters. 

High species numbers of a single selective taxon in relation to a certain area unit, scientifically expressed as a-diversity, are not a sufficient conservation criterion to assess an ecosystem and it is generally unsuitable to derive conservation recommendations on this basis alone. For a holistic assessment of habitats, further conservation criteria are necessary in addition to species numbers and abundance values; these include: (1) naturalness, i.e., the correspondence of the biocoenoses found with a natural or at least near-natural reference system; (2) the rarity and endangerment of the species; (3) the representativeness and probability of occurrence of natural disturbances and processes; (4) the resilience (stability) and restorability (elasticity) of the biocoenoses in the event of natural disturbances; (5) habitat connectivity, i.e., interconnectedness with sufficiently similar habitats in the surrounding landscape matrix; (6) representativeness of the ecosystem at a higher spatial scale; and (7) undisturbedness, which refers to the lowest possible indirect anthropogenic influences (e.g. by nitrogen emissions) (see also Schmidt et al. 2011, 2014, Schoof 2013, Opitz et al. 2015, Rosenthal et al. 2015, Brackhane et al. 2021).

However, this does not address the relevance and weighting of these criteria in relation to each other. The cardinal problem concerning the evaluative and decision-making basis for normative nature conservation lies in the agglomeration of evaluative criteria and a presumptive inclusion of non-comparable criteria. Even the choice of a single, specific criterion can lead to a completely contrary evaluation, as illustrated by the following example: A managed beech forest with diverse structures, such as clearings created by felling, fringes along forest roads and logging trails, ruderal areas due to deposits or compaction from vehicles, usually displays a relatively high species diversity for vascular plants, while it is poor in terms of the “dark optimal phases” of a primeval beech forest. Although the individual observation of such conditions is correct, the generalization and especially generalizing evaluation derived from it are invalid. As mentioned above, a balanced spectrum of different parameters is always needed for the evaluation of habitats. In the ecological assessment of forests, sensitive species groups such as old-growth and deadwood specialists must also be taken into account. 

---

Box 4: Legal obligations and goals for the protection of primeval and natural forests

In 2007, the German government adopted the National Strategy on Biological Diversity (NBS); this was 14 years after the ratification (1993) of the Convention on Biological Diversity (CBD), which had been signed at the UN World Summit in Rio in 1992 (BMUB 2007). For forests, one of the goals formulated in the NBS was to permanently renounce forestry utilization on 5% of the forest area by 2020 and to allow process-oriented natural development (and also to secure this by law in the long term). In short, the goal is officially “natural forest development” on 5% of the total forest area.

A further goal was that nature should be able to develop dynamically and undisturbed in large areas covering at least 2% of the country’s terrestrial surface. In addition to forest ecosystems, this also included peatlands, former mining landscapes and high mountains (BMUB 2007). The goals that should have been achieved by 2020 are modest: according to official sources on the achievement of the NBS targets, small-scale set-asides in accordance with the 5% target have so far been implemented on only 2% of forest areas, and the 2% wilderness target has been implemented on only about 0.8% of total terrestrial areas, with national parks included here on a blanket basis, even if in some cases they still practice forest conversion or hunting (BMU 2018, Höltermann et al. 2020).

In 2011, in response to the 1993 CBD commitments, the EU presented the EU Biodiversity Strategy 2020 with few concrete targets and these were largely non-binding (see EU 2011). However, on issue of forests, it formulated emphatically: “The development of forests in Europe is a cause for concern. Most managed forests are still operated as commercial plantation forests and are of limited value for biodiversity. Of the forest habitats and forest-dwelling species protected under the EU Habitats Directive, only 21% of habitats and 15% of species have favorable conservation status. Only 1-3% of forests in Europe are still in a natural condition”. Objective 3, increasing the contribution of agriculture and forestry to the conservation and enhancement of biodiversity, states: “By 2020, introduce forest management plans or equivalent instruments that are consistent with sustainable forest management. This applies to all state-owned forests and to forest holdings that exceed a certain size; the details of this are to be defined by the member states or regions. Measured against the 2010 EU reference scenario, the aim is to bring about a measurable improvement in the conservation status of species and habitats that depend on or are influenced by forestry.” To date, however, there has been no evaluation by the EU or its bodies such as the European Environment Agency of the progress or success of its own biodiversity strategy.

The new Biodiversity Strategy 2030, which is one of the central elements of the EU’s Green Deal project, formulates the following strategic objectives on the issue of forests (EU 2020a, b): (1) Establish a coherent network of well-managed protected areas on at least 30% of the land area; (2) Protect and restore forests in the EU; (3) Strictly protect all remaining primeval forests and old-growth forests, (4) Call for the development of an EU forest strategy; (5) Expand forest areas in the EU and plant at least 3 billion trees, while fully respecting ecological principles; and (6) Implement measures to improve the resilience of forests and their role in combating biodiversity loss and mitigating climate change. The Forest Strategy 2050 recently presented by the German government also contains a clear commitment to implement the forest-related goals of the EU 2030 Biodiversity Strategy. It explicitly emphasizes: “placing more land area under nature conservation and also a significant part of it under strict protection. Old-growth forests in particular should be strictly protected” (BMEL 2021).

In a study by the Thünen Institute for International Forestry and Forest Economics, by contrast, fears are expressed that the implementation of the EU biodiversity strategy would be associated with considerable restrictions on forest use and would have extremely negative effects on the forestry and timber industries in Germany as well as in the EU as a whole (Dieter et al. 2020, Dieter 2021). 

In general, the relevance of the EU nature conservation goals for forests (protection and setting-aside of primeval and natural forests and increased designation of protected areas) is questioned and changes in strategy, away from the demand for protection and towards utilization, are called for. A threatening scenario is presented that if European forests were not to be utilized there could be fatal ecological consequences: (1) that meeting projected future demand for raw timber would lead to an outsourcing to third countries with less sustainable wood production and overall lower political regulation; (2) that serious negative effects on forest biodiversity in these countries could thus be expected; and (3) that this would lead to an overall increase in deforestation pressure. These projections are made on the basis of the following modeled assumptions, and assuming that EU nature conservation targets were implemented normatively and instrumentally: 

(1) For EU countries, a 42% reduction in logging by 2050 is assumed compared to a baseline scenario (see also Section 5.4). 

(2) For Germany, it is assumed that the annual volume of raw timber would then fall by an average of approx. 24 million m³ for the period under consideration from 2018 to 2050 (-31 % compared to the baseline scenario). 

(3) For Germany, it is assumed that 10% of total forest area will be set-aside in future, and that this setting-aside would be distributed representatively, i.e. would also include highly productive sites.

On this "factual basis", a broad alliance of the forestry and timber industry argues that active climate protection requires abandoning further restrictions on forest use and instead increasing CO₂ sequestration through wood use (Verbändeposition der Forst- und Holzwirtschaft 2021). In our opinion, the shifts postulated by the Thünen Institute are not particularly realistic given the few remaining and still shrinking primeval forest areas in Europe and against the backdrop of the 5% target of "natural forest development” set by the National Biodiversity Strategy, which, though far from being achieved, is nevertheless legally codified, and which, it is to be hoped, will be adhered to at the political level.

---

4.    The discourse on primeval forests, natural forests and managed forests and their contribution to climate protection

The importance of forests for climate protection lies in their function as carbon reservoirs and sinks. During the growth phase, forests remove large amounts of CO₂ from the atmosphere and store it in the biomass (wood) and soil over the long term (e.g. Luyssaert et al. 2008, Gleixner et al., 2009, Nord-Larsen et al. 2019, Meyer et al. 2021). The global C storage of forest ecosystems comprises 300 Gt C in mineral soil organic carbon, 295 Gt C in living biomass, and 68 Gt C in dead wood and litter layer (Fig. 8a). Globally, forests are the largest terrestrial sink for CO₂, absorbing about 2 Gt CO₂ annually; equivalent to 0.55 Gt C (UN 2021) (see Box 5 and Carbon Inventory 2017, Riedel et al. 2019). However, this storage capacity is decreasing due to the destruction and overexploitation of forests. The main driver is the increasing utilization of wood for thermal energy (firewood, wood chips) (e.g. UNEP 2020, UN 2021). Increasingly, climate effects such as forest fires, calamities and drought stress, which reduce stands and productivity, are also contributing to this negative development.

---

Box 5: Carbon reservoirs and CO₂ equivalents

An average of 0.5 t of carbon is stored in 1 t of air-dry wood. According to the molar mass ratio of CO₂ to C (44/12 = 3.67), this amount of carbon corresponds to 1.83 t CO₂. The volume-related measure of solid cubic meters (fm = m³), which is commonly used in forest inventories, is simplified to an average wood weight of 0.7 t; this value is plausible for the beech-dominated forests of the Hainich, for example. This means that an additional growth of 1 m³ of wood corresponds to a CO₂ sequestration of 1.28 t CO₂. During harvesting and subsequent oxidation of carbon by burning, this amount of CO₂ is released to the atmosphere again. The same amount is also released when the wood decays (mineralizes and oxidizes) in the forest. However, this does not happen abruptly, but over a period of decades, depending on the tree species, the climate and the dimensions of the dead wood. Use of the wood in long-life products also delays the release of CO₂, for example in timber construction – on average 35 years (UBA 2020a, Figure 9). These figures refer exclusively to the aboveground wood-based biomass.

---

For the 11.4 million hectares of forest in Germany, the magnitude of C storage in biomass, deadwood and soil is calculated to be 2.6 billion t for the baseline year 2017 (which corresponds to 9.5 billion t CO₂). Of this, 1.23 billion C is stored in above-ground biomass, 0.034 billion t C in dead wood and the humus layer, and 1.335 billion t C in mineral soil and below-ground biomass (Fig.  8b).  Baseline values are a determined average (wood) stand of 358 m³ with a growth of 10.9 m³ per hectare per year (Riedel et al. 2019, BMEL 2021). Due to the extreme conditions between 2018 to 2020, it can be expected that the average annual growth in Germany was significantly lower in this period; this could be interpreted as an indication of declines in production or even a decrease in the stand in our forests that can be expected in the near future.

The accounting of carbon storage and thus also the CO₂ sink performance of forests (or trees) is complex in its details and strongly dependent on the forest type, development history and management. During tree development, the C-sink performance gradually decreases after reaching the culmination point of growth. Under natural conditions, if there are no significant disturbances, forest development continues steadily and then gradually enters the decay phase where release of CO₂ predominates. This is not an abrupt process but can takes decades or even centuries, and can vary greatly in duration depending on the tree species. C storage is  is therefore significantly higher in virgin forests at the landscape level than in commercial forests, while the increment in intensively managed (thinned) commercial forests can be higher if thinning strengthens the increment of productive dominant trees (target trees) at the expense of weaker-growing or silviculturally undesirable individuals of inferior quality. In the case of a high intensity of timber extraction, the C-sink performance of a forest can be very low at the landscape level without taking into account possible C-sinks in material products, since an almost comparably large amount of timber is removed as grows back. However, in a direct comparison of beech management forests and virgin beech forests in the Slovak Carpathians, an equally high productivity is also reported for both forest types (Glatthorn et al. 2017).

On productive sites, the stocks of wood can reach high amounts when the forest is set aside; for example, in the beech-dominated primeval forests of the Ukrainian and Romanian Carpathians, over 600 m³ per ha. In addition, there are high levels of deadwood, which can be around 200 m³ per ha, ten times the average value of German managed forests (Commarmot et al. 2013, Glatthorn et al. 2017, Kun et al. 2020). In total, an ecosystem C pool of 272 t carbon per hectare was measured in the Slovakian Carpathians for managed forests compared to 347 t in primeval forests (+ 27%) (Leuschner et al. 2021). The C sink was significantly larger than in the managed forest, especially in deadwood (+ 310%), but also in wood biomass (+ 20%) and soil (+ 17%) (Fig.  10). It should be taken into account that for the primeval forests the prevailing development phases were mapped, whereas for the managed forests it was the mature phase shortly before harvesting. Thus, if all phases of the beech managed forests are considered in one production cycle and compared with the C-storage of primeval forests, the difference in C storage would likely increase by more than 75 t C per hectare for the primeval forests in comparison to the managed forests. 

From a comparison of managed forests with natural forests that have been set aside for several decades, Schulze et al. (2020a, 2021) conclude that primeval forests and natural forests are disadvantageous for climate protection, because managed forests have a 10-fold higher climate protection effect than natural forests due to higher tree growth and a corresponding long-term wood product storage. These calculations and the evaluations derived from them are now often cited as grounds for discrediting attempts to protect primeval forests and to cease natural forest use. For example, forestry experts and politicians in Romania explicitly refer to the study by Schulze et al. (2020a) and recommend to the Romanian government that further protection of primeval forests and natural forests is inappropriate for ecological and climate policy reasons (UTB 2020a, b).

The empirical basis of the study by Schulze et al. (2020a) is a comparison of data sets from the German federal forest inventories 2002 (BWI 2) and 2012 (BWI 3) for managed forests (comprising about 60,000 permanent inventory points across Germany) with inventory data from the Hainich National Park for the years 2000 and 2010 (comprising 1,200 and 1,421 inventory points, respectively). From the comparison of the two federal forest inventories for German managed forests, a CO₂ mitigation effect of 3.2 to 3.5 t CO₂ equivalents per hectare and year is calculated. This calculation takes into account losses during harvesting and wood processing as well as substitution effects determined from life cycle assessment studies (see also part 2, section 2 of this paper). For the Hainich National Park, a CO₂ sink of only 0.37 t CO₂ equivalents per hectare and year was calculated by comparing the inventory data (and excluding utilization of wood).

The Hainich National Park Authority comments on this in a statement (NPH 2020, Welle et al. 2020), saying that Schulze et al.’s (2020a) data for the Hainich was given an incorrect analysis, evaluation and interpretive context, as the two time periods have different and non-comparable forest reference areas. To determine periodic growth, only the respective identical forest areas can be compared in a time series. In their comparative calculations, Schulze et al. mistakenly included about 220 samples from the Hainich National Park that were not yet forest areas during the first inventory and are also not listed in the data records of the initial inventory. These are several hundred hectares of clear-cuts from previous utilization and scrub areas of former open lands such as shooting ranges which were within the coverage threshold for forest in the initial inventory (Fig. 11). These areas, even after 10 years of development between the 2000 initial inventory and the 2010 replicate, still have very low timber inventories and are in no way comparable to natural forest. When adding up the identified stocks and averaging over all the samples, the impression was thus created that the forests in the Hainich, with only 0.37 m³ per hectare and year, would show practically no increase in stocks over a 10-year period. This is in clear contradiction to the growth in stocks of about 7.9 m3 per hectare per year determined for Slovakian primeval beech forests, a figure which was exactly as high in site-homologous managed forests.

The second mistake of the study is to use the forests of the Hainich as the only reference and benchmark for Central European natural forests. The correct evaluation of the inventory areas in the Hainich results, in a decade comparison, in a an average growth in stock of 8.6 m³ per hectare and year (NPH 2020); this value is of a similar order of magnitude as the average increment in German (managed) forests (Third Federal Forest Inventory: 10.3 m3 per hectare and year for beech forests); the figure can be reconciled with the prevalent highly productive forest development phases once it is noted that the Hainich forests are by no means yet in their optimal or terminal phase.

Transferring the calculation method used by Schulze et al. (2020a) to the unused forest areas in the Hainich, the growth in stock of 8.6 m³ per hectare and year results in a CO₂ sink of 8.0 t CO₂ equivalents per hectare and year. Since no utilization takes place, this results in a medium-term climate protection performance of the forests in the Hainich, which is even higher by a factor of 2.5 than the CO₂ mitigation effect of managed forests of 3.2 to 3.5 t CO₂ equivalents per hectare and year as determined by Schulze et al. (2020a). It becomes clear that the discussion about the climate protection performance of forests rests in part on incorrect data. 

5.    Outlook

The primeval forests in Europe - to be distinguished from today's unused natural forests, which display more or less pronounced characteristics and consequences of past use - have shrunk to tiny remnants of less than 1%, and even these tiny remnants are highly endangered. Four-fifths of temperate primeval forest areas are located in the Carpathian arc, and political commitments justifiably require their protection. At the same time, Germany's National Strategy on Biological Diversity sets clear targets of 5% of forest area for natural forest development (future natural forests) and 2% of the federal territory for wilderness development - targets that should have been achieved by 2020 but have been missed by a long way. In order to objectify the public debate, this article has analyzed the importance of primeval forests for the conservation of biodiversity and in their functions as carbon stores and sinks and the effects of abandoning utilization in forests.

As the WBGU’s main report (2020) on “Land Transition in the Anthropocene” has shown, natural process conservation in the forest is integral to any exemplary multiple-benefit strategy: it contributes significantly both to the conservation and promotion of biodiversity, in particular the specific species and biocenoses of old forest development stages, and to carbon storage and sequestration, i.e. to climate protection. Abandoning the utilization of representatively selected forest areas is thus best practice for sustainable combined nature- and land-based climate protection and also does not rule out the material utilization of wood when this has objectively positive effects on the climate.

In the second part of the article (see next issue), we address the narrative of the climate neutrality of wood as a resource. Establishing the CO₂ sink performance of wood in the context of logging and the substitution of wood products by other materials requires nuanced discussion and our article undertakes just that. The assumptions made concerning the sustainability of alternative scenarios for forest treatment and wood use made in the WEHAM study (WaldEntwicklungs- und HolzAufkommensModellierung – Forest Development and Wood Supply Modeling) are criticized. The sweeping generalization that wood is a CO₂-neutral energy source requires detailed analysis. Our findings are then placed in the context of the role of wood in implementing political goals for climate protection.

Finally, in the synopsis of the findings of both parts of the article, we will draw conclusions for how to bring objectivity to the dispute over the value of primeval and natural forests.

Conclusion for practice

• In conceptual terms, primeval forest and natural forest must be clearly separated. Forests that have only recently been set aside are not (yet) a sufficient reference for the assessment of biodiversity and climate protection performance and for the goals of natural process conservation, because they require very long periods of time to begin to resemble primeval forests. 
• The imperative to protect primeval forests rests on ethical, scientific, and anthropocentric grounds: they preserve genetic diversity and thus the potential for climate adaptation, act as refugial and source biotopes for highly specialized species, provide learning sites for forestry strategies, and serve as research laboratories for effects of long-term environmental change.
• Critical studies on the biodiversity performance of unmanaged forests are typically based on faulty assumptions or are usually limited to the consideration of vascular plants alone. Highly specialized species of animals, fungi, lichens, and mosses in particular comprise the unique value of forests that have been unmanaged for a very long time. Conservation assessments require an objective analysis of a much wider variety of criteria than mere species numbers. Our findings clearly indicate a very significant increase in the ecological value of unused forests with a growing primeval forest character.
• The carbon storage capacity of forests is decreasing in overall terms due to increasing use of wood for energy along with climate change impacts. However, such issues are very complex. C storage is significantly greater in primeval forests at the landscape level than in managed forests, while the C sink function may be higher in intensively managed (thinned) forests if thinning strengthens the growth of productive dominant trees (target trees) at the expense of weaker-growing or silviculturally undesirable individuals of inferior quality. Here, too, the data used in studies must be critically questioned. The argument that only a utilized forest is a good forest for climate protection cannot be substantiated by the facts.

---

Contact

• Prof. Dr. Rainer Luick, University of Applied Forest Sciences > luick@hs-rottenburg.de
• Dr. Klaus Hennenberg , Öko-Institut e.V. in Darmstadt > k.hennenberg@oeko.de
• Prof. Dr. Christoph Leuschner, Georg-August-Universität Götttingen, Dept. of Plant Ecology in the Albrecht von Haller Institute for Plant Sciences > cleusch@gwdg.de
• Dipl.-Ing. Manfred Grossmann, Head of Hainich National Park, Bad Langensalza > manfred.grossmann@nnl.thueringen.de
• Prof. Dr. Eckhard Jedicke, Geisenheim University, Competence Center Cultural Landscape, Chair of Landscape Development > eckhard.jedicke@hs-gm.de
• Dr. Nicolas Schoof, Albert-Ludwigs-Universität Freiburg, Chair of Site Classification and Vegetation Science > nicolas.schoof@waldbau.uni-freiburg.de
• Dr. Thomas Waldenspuhl, Head of the Black Forest National Park, Seebach > thomas.waldenspuhl@nlp.bwl.de