Background

In the boreal zone, harvesting and management of forests on an industrial scale is the most important factor driving habitat change and degradation [1]. In addition to regular harvesting, removal of stumps and harvesting residues for energy wood has increased in recent decades because of the current climate policies in Europe [2] but also in North America [3]. Stump removal increases the degree of disturbance by decreasing the amount of dead wood dramatically [4, 5] and may therefore lead to stronger responses of the dead-wood dependent species. There are lots of studies on the impacts of forest harvesting on different species groups in Fennoscandia, and long-term monitoring shows declines of biological communities, for example in many forest bird populations [6, 7]. Also, species assemblages may not be maintained in protected areas if they are embedded in heavily managed landscape [8, 9]. Under these circumstances the management of productive forests is a key aspect for maintaining biodiversity.

A common forest management regime in the whole boreal zone has long been even-aged management (Table 1) [10]. In Finland, for example, even-aged management was the primary management regime in forestry dictated by law for more than 60 years, until the year 2014, when uneven-aged management regime was enabled again [11]. As boreal forests in their natural state are usually heterogenic with trees and stands of different species, ages and size, even-aged management simplifies the forest structure which is noticed to have negative impacts on biodiversity [12].

Table 1 Definitions of different forest management regimes. Also, common synonyms for the main term are given 
Full size table

To slow down the decline of forest species diversity alternative management regimes have been taken into practice. Retention forestry [13] is usually similar to even-aged management, but some old trees, dead or living, or small stands of trees are retained in harvest to create structural diversity [14, 15]. In uneven-aged management, i.e. continuous-cover forestry, harvested patches are usually smaller than in even-aged management, mature trees or tree groups are selected for harvesting, and younger trees are left to grow [10]. Uneven-aged forest management aims for more heterogeneous stand structures, assumed to be less damaging to forest biodiversity than clear-cuts [16].

Even though even-aged management is still the most common harvesting method in Finland, Sweden and Norway, the public interest towards uneven-aged forest management has increased in the last decades [17,18,19,20,21]. In European Russia, most final feellings are clear-cuts and continuous cover forestry with selective logging method is more widely used only in western parts of the country in Murmansk and Leningrad regions [22].

Despite the rising public interest, the adoption of uneven-aged forest management in Finland, Sweden and Norway has been relatively modest. There is a strong, ongoing debate on economic profitability of uneven-aged management, with studies showing that at least in some forest types uneven-aged management would be more profitable than even-aged management [23, 24]. Similarly, recent scientific studies suggest that uneven-aged management provides higher values for some biodiversity aspects but even-aged management for others [10, 25]. Therefore, it is not surprising that there is no strong consensus between stakeholders on the impacts of these two forest management regimes on biodiversity [26].

The topic for the systematic review arose from the discussions with the Finnish forest industry on their evidence needs related to sustainability issues. In Finland 91% of forests are in commercial timber production [27] and hence, forest industry has large influence on forest biodiversity. Specifically, three forestry companies (Metsä Group, Stora Enso Oyj, and UPM-Kymmene Oyj), industry representative group Finnish Forest Industries, and the state-owned enterprise Metsähallitus, which is responsible for the management of one-third of Finland’s surface area, participated in the discussions that lead to the broad definition of the topic. The topic was further defined in a stakeholder workshop that was held 23 August 2018. Purposive selection based on known contacts, snowballing, and internet search were used to compile a list of stakeholders following recommendations by Haddaway et al. [28]. An open invitation to participate in the workshop was published on the website of the Evidence-Based Forestry in Finland initiative 24 July 2018 and sent by email to 35 stakeholder organisations (Additional file 1) with a notice that it can be further shared with interested individuals and organisations. A reminder email was sent 2 weeks later to those individuals and organisations that had not responded.

In the end, 13 individuals from 12 stakeholder organisations participated in the workshop (Additional file 1). At the workshop, the topic, research questions, initial theory of change, PECO-based search terms, and factors creating heterogeneity were presented and discussed. Based on the discussion with the participants the topic was narrowed to the two research questions that are presented in this protocol. The participants also suggested sources of grey literature and potential sources for unpublished data. Comments and suggestions of the participants related to other elements discussed have been integrated into the protocol.

Objective of the review

The objective of the review is to systematically review and synthesise results of the studies on the impacts of even-aged and uneven-aged forest management on biodiversity, specifically species of different taxa, at different spatial and time scales. Figure 1 shows a simple theory of change of the impacts of forest harvesting on biodiversity. The review will focus on Fennoscandia (Finland, Sweden and Norway) and European Russia. The geographical scope of the review was discussed at the stakeholder workshop and is based on the similarity of the forests, including tree species, and their management.

Fig. 1
figure 1

Simplified depiction of impact pathways influencing species assemblages over time. Clear-cut is a typical management practice for even-aged forest management whereas uneven-aged forest management retains continuous forest cover. In practice, the dichotomy between even-aged and uneven-aged forest management regimes is less distinct than depicted here due to effect modifiers such as the number of retention trees, gap size, and felling style

Full size image

The review has two key research questions:

  1. 1.

    What are the stand-level effects of even-aged and uneven-aged forest management on boreal forest biodiversity in Fennoscandia and European Russia?

  2. 2.

    What is the effect of forest management on boreal forest biodiversity in Fennoscandia and European Russia at landscape level?

The impacts of even-aged and uneven-aged forest management regimes will be compared to each other as well as to forest areas where no intervention has taken place to give a full picture of the impacts. The first question focuses on the impact of different forest management regimes on alpha diversity whereas the second question focuses on gamma diversity. Most of the biodiversity studies are stand-level studies focused on species richness and other alpha diversity measures, which commonly decrease in response to harvesting. However, it is important to know whether the stand-level habitat loss means an overall loss of biodiversity at a larger scale or whether biodiversity is maintained despite a temporary loss of a habitat in one or more places in the landscape that is formed by the different forest stands. The question components are outlined in Table 2.

Table 2 Components of the review questions
Full size table

Methods

This protocol follows the guidelines of Collaboration for Environmental Evidence and complies with the ROSES reporting standards. The ROSES form is included as an Additional file 2.

Searching for articles

Search string

A list of search terms relating to the PECO components was proposed at the stakeholder meeting and validated by the participants (Table 3).

Table 3 Search terms proposed at the stakeholder meeting
Full size table

Based on the five categories, a search string was formulated using Boolean operators ‘OR’, ‘AND’ and ‘NEAR’. The performance of the search string was tested using a test list of 20 articles collected from previous reviews and from experts (Additional file 3). The testing was conducted primarily in the Web of Science (Core Collection) but also included testing in Scopus and CAB Abstracts (Additional file 4). After scoping and taking reviewer suggestions into account, the search string was modified into its current form. The final search string in English is:

#1 TS = ((Boreal NEAR/5 (forest* OR zone OR tree*)) OR taiga OR spruce* OR picea OR pine* OR pinus OR birch* OR aspen* OR populus).

#2 TS = (Finland OR Finnish OR Swed* OR Norw* OR Russia* OR Fennoscan* OR Scandin* OR “north* europ*” OR “nord* countr*”) and TS = (forest* OR tree*).

#3 TS = (clear-cut* OR clearcut* OR clearfell* OR clear-fell* OR “clear fell*” OR even-aged OR uneven-aged).

#4 TS = (forest* NEAR/5(“continu* cover*” OR “natural* regenerat*” OR multiage* OR alternativ* OR “common* sens*” OR unmanaged OR managed OR sustainabl*)).

#5 TS = (silvicult* NEAR/5(“continu* cover*” OR “natural* regenerat*” OR multiage* OR alternativ* OR “common* sens*” OR unmanaged OR managed OR sustainabl*)).

#6 TS = (Regenerat* NEAR/5 (cut* OR fell* OR harvest* OR log*)) OR TS = (select* NEAR/5 (cut* OR fell* OR harvest* OR log*)) OR TS = (partial* NEAR/5 (cut* OR fell* OR harvest* OR log*)) OR TS = (alternat* NEAR/5 (cut* OR fell* OR harvest* OR log*)) OR TS = (retent* NEAR/5 (cut* OR fell* OR harvest* OR log*)) OR TS = (conserv* NEAR/5 (cut* OR fell* OR harvest* OR log*)) OR TS = (gap* NEAR/5 (cut* OR fell* OR harvest* OR log*)) OR TS = (patch* NEAR/5 (cut* OR fell* OR harvest* OR log*)) OR TS = (dispers* NEAR/5 (cut* OR fell* OR harvest* OR log*)).

#7 TS = (biodiversi* OR fauna OR flora OR fungi OR eukaryot* OR vertebrat* OR invertebrat* OR animal* OR plant* OR arthropod* OR lichen* OR insect* OR bird* OR mammal* OR vegetat* OR bryophyte* OR amphibian* OR reptile*).

#8 TS = (species NEAR/5 (divers* OR rich* OR assemb* OR abund*)).

#9 #2 OR #1.

#10 #6 OR #5 OR #4 OR #3.

#11 #8 OR #7.

#12 #11 AND #10 AND #9.

The search string will be translated to other search languages. It will be simplified by reducing the number of search terms to search organizational websites and to conduct internet searches where the search interface often has more limited capacity regarding search strings. Boolean operators will be used to combine main search terms whenever the search engine allows it. The used search strings will be recorded and published as additional information in the review report.

A search alert will be set in bibliographic databases to screen articles that are published before the data synthesis commences. The number of articles retrieved through the search alerts will be reported in the review report.

Languages

The systematic review will include studies published in English, Finnish, Swedish, and Russian. The selection of languages is based on the geographical scope of the systematic review and limited by the language skills of the review team. Organisational websites will be searched in the primary language the website is published except the websites in Norwegian, which will be searched in English. In addition, if the publications section includes studies published in other of the review languages (e.g. main website language is Swedish but there are also unique publications in English), the search will be conducted in those languages as well.

Bibliographic searches

The following bibliographic searches will be conducted:

  • CAB Abstracts (https://www.cabi.org/); Keyword search from 1973 onwards.

  • Directory of Open Access Repositories (https://doaj.org/); ‘Search all’ field will be used with not further limitations.

  • Digital Dissertations Library of Russian State Library (http://diss.rsl.ru/).

  • Doria (https://www.doria.fi/).

  • Helka—University of Helsinki Catalogue (https://helka.finna.fi/); All fields will be searched with no further limitations.

  • Jultika—University of Oulu repository; All fields will be searched with no further limitations.

  • JYX—Publication archive of the University of Jyväskylä.

  • Russian Science Citation Index on the Web of Science (https://clarivate.com/); Topic search, access from 2005 onwards.

  • Russian Scientific Electronic Library (https://elibrary.ru/).

  • Scopus (https://www.scopus.com/home.uri); Title, abstract, and keyword search.

  • Swedish University Dissertations (http://www.avhandlingar.se/).

  • UTUPub—University of Turku repository.

  • Web of Science Core collection (https://clarivate.com/); Topic search covering all years within Science Citation Index Expanded (1945-present), Social Sciences Citation Index (1956-present), Arts & Humanities Citation Index (1975-present), Conference Proceedings Citation Index-Science (1990-present), Conference Proceedings Citation Index-Social Science & Humanities (1990-present), Emerging Sources Citation Index (2015-present).

Search engines

The internet searches will be conducted in ‘private’ mode to prevent the influence of previous browsing history and location on search results. The results will be organised by relevance. After the first 50 hits, results will be checked until relevant articles are no longer retrieved as advised in Livoreil et al. [29]. The date and number of hits received and searched will be recorded and included in the review report.

Organisational websites

The websites of the specialist organisations listed below will be searched. The Russian websites will be searched manually due to the low performance of the “search” function to find relevant hits based on scoping of the organisational websites. If the organisation is publishing a journal, the site of the journal will also be searched if the journal is not already included in some of the bibliographic databases searched.

Supplementary searches

Citation chasing will be undertaken to supplement the search. A call for unpublished data will be published on the website of the Evidence-Based Forestry in Finland project (http://npmetsa.fi/en/frontpage/) and sent directly to stakeholder organisations that may have unpublished data on the topic. Also, data will be asked from individuals suggested at the stakeholder workshop.

Search record database

The search results will be exported into separate files using a reference management software. If the document cannot be exported into a reference management software, a record will be created manually into a separate file. Once all the searches have been conducted, the reference files will be merged, and duplicates will be removed before commencing article screening.

Article screening and study eligibility criteria

Screening process

Articles will be screened by three people at the title, abstract, and full text level. At the title stage, a random set of 100 articles will be independently screened by all three screeners. If their screening decisions are in agreement, i.e. they would include/exclude the same articles, rest of the articles will be divided among the screeners. If their screening decisions differ, discrepancies in inclusion decisions are discussed to facilitate consistency before another 100 articles will be independently screened. The process will be repeated until 95 to 100% screener agreement is achieved. The process will be repeated at the abstract stage with a random set of 50 articles. If a screener is unsure whether to include an article, it will be moved to the next stage. Articles at the full text stage will be screened by all three screeners except for studies in Russian that will be screened by only one person at all stages. To check that inclusion criteria is used consistently, the Russian speaker will talk the other screeners through the decision process on a random set of 20 articles at each of the screening stages.

The review may include articles published by the authors of the review. Their inclusion in the review at the screening and critical appraisal stage will be jointly determined by the other authors in accordance with the eligibility and appraisal criteria.

Eligibility criteria

The eligibility criteria are based on the PECO components, study design and geographical location of the studies (Tables 4 and 5). Only studies conducted in Finland, Sweden, Norway and European Russia will be included.

Table 4 The eligibility criteria for article screening for the study question 1
Full size table
Table 5 The eligibility criteria for article screening for the study question 2
Full size table

At each stage of the screening a separate file will be created of the excluded articles. At the full text stage, a reason for exclusion will be recorded, and a list of the excluded articles with the reason for rejection will be included as additional information in the review report.

If there are multiple articles from one study site (i.e. linked articles), they will be appraised as a group to avoid inclusion of duplicate data following Frampton et al. [30]. True duplicate studies will be removed, and the rest will be screened as a single unit to consider all available data pertinent to the study when making eligibility decisions.

Study validity assessment

All studies included in the full text stage will be critically appraised and categorized as ‘low’, ‘medium’ or ‘high’ risk of bias. The assessment is based on the following factors (Table 6):

  • Study design.

  • Sampling.

  • Accounting for potential effect modifiers and heterogeneity.

  • Data analysis methods.

Table 6 Critical appraisal criteria to assess studies in the full text stage
Full size table

Studies that fulfil any one of the criteria in the category ‘high’ will be excluded. Also, studies with insufficient methodological description will be excluded if sufficient clarifying details are not received by contacting the author of the study. All the studies will be assessed by two persons, and any inconsistencies or uncertainties discussed with other research group members.

Data coding and extraction strategy

Data from included studies will be extracted and recorded in an Excel spreadsheet (Additional file 5) and will be made available as supplementary information of the systematic review. Data will include study meta-data (study characteristics) and data on outcomes, e.g. sample size, mean, standard deviation (SD), and standard error (SE) (see Additional file 5 for a full list). Data on test statistics that can be converted into effect size metrics will be collected in case data on outcome mean, SD or SE is not available. Also, data on effect modifiers and potential sources of heterogeneity will be extracted to enable statistical exploration of the relationship between outcomes and sources of heterogeneity at the data analysis stage. All the extracted data will be published as supplementary information of the review. If an article contains independent results from more than one study, these will be treated as separate studies in data extraction. Authors of the studies will be contacted to retrieve any missing information or data.

Data will be extracted by more than one person. Hence, a set of five studies will be coded together to ensure consistency. If there are conflicting decisions on what data to extract, the decisions will be discussed among the group. Also, any uncertainties regarding data extraction will be discussed among the group. Data from the studies in Russian will be extracted by one person only. The person will discuss any uncertain decisions with the research group members.

Potential effect modifiers and reasons for heterogeneity

To understand possible variation in the effects of studies better, possible effect modifiers will be extracted from the studies. As the studies included in this systematic review may have been completed in a relatively large area, there are several factors that may cause heterogeneity among studies, such as climatic conditions and geographic location of the study site. Also, temporal variation is expected. The year a study was conducted may influence the results as forest management has changed over the years. Also, time passed since intervention was started may cause variation depending on the timing and nature of harvests as well as natural succession of vegetation after harvests. Energy wood harvesting, i.e. removing stumps and branches beyond regular harvesting may have various impacts on biodiversity mainly by reducing the amount of dead wood [2, 31]. Habitat connectivity can be an important factor in the dispersal of species across the landscape, and there are several estimates that can be used to quantify connectivity. In case the used connectivity estimates differ between the included studies, we will instead use categories low, medium or high in the data analysis.

Below is a non-comprehensive list of potential effect modifiers and sources of heterogeneity. Additional effect modifiers and sources of heterogeneity may be identified from the studies included in the review. The list was compiled based on the authors’ experience and consultation at the stakeholder meeting.

  • Geographic location

  • Climatic conditions

  • The year(s) the study was conducted

  • Time since the intervention started

  • The length of the study

  • Size and extent of sampling area

  • Forest type and soil type

  • Humidity (drained vs. non-drained)

  • Connectivity of the study site(s)

  • Differences in management type (for example, single tree selection and small patch selection felling are both considered as continuous cover forestry)

  • Certification

  • Owner of the study site(s) (private, company, state)

  • Harvesting of energy wood (stumps, branches)

Data synthesis and presentation

A narrative synthesis of data from all the included studies will be produced. The narrative synthesis will describe the evidence-base with tables and figures, including description of interventions and comparators, study locations and designs, length of the studies, and studied taxa. It will also describe the effects of the interventions on biodiversity outcomes.

If enough quantitative data can be extracted from the included studies, a meta-analysis will be conducted to assess the effects of forest management on biodiversity outcomes at stand-level. If data allows, sub-group analyses on different taxa will be undertaken. Furthermore, heterogeneity in the results will be explored using meta-regression if there is enough data to conduct the analysis. In case there are several outcomes from the same experimental setup, their treatment will be considered prior to the statistical analyses to avoid the risk of false-positive results. Also, only data from comparable settings will be included in the same analysis. For example, before-after data from pforest that has been previously harvested will not be compared with data from control-intervention design that compares (unharvested) national park with production forest. If enough data for meta-analysis cannot be extracted, other analytical methods will be considered alongside narrative synthesis.

To assess landscape level diversity (gamma diversity) a framework developed by Chao et al. [32] will be used. It uses effective number of species (Hill numbers) incorporating relative abundance, which makes it suitable for landscape level comparisons of species assemblages. Hill numbers (qD) quantify diversity in units of equivalent numbers of equally abundant species by increasingly weighting abundance with the order of diversity q. If there is missing or incomplete information in the article, and that information cannot be retrieved by contacting the authors, the study will be excluded from the analyses.

When the systematic review is conducted steps are taken to minimise bias in the results, for example by searching extensively both peer-reviewed and grey literature and by excluding articles that have high risk of bias. To test the effect of the validity assessment (i.e. exclusion of articles) and the robustness of the studied outcomes, a sensitivity analysis will be conducted. This will be done by conducting analyses including and excluding studies with high risk of bias. Not all sources of bias can be excluded, such as publication bias which stems from the practice that studies showing statistically significant effect are more likely to be published than those that do not. Therefore, the presence of publication bias will be evaluated visually by producing funnel plots. If publication bias is detected visually, ‘trim and fill’ method will be used to adjust the results for publication bias if enough data is available to do this.