Life cycle assessment of peat for growing media and evaluation of the suitability of using the Product Environmental Footprint methodology for peat

Peat extraction rapidly removes carbon from the peatland carbon store and furthermore leads to substantial losses of carbon from the extraction site by stimulating decomposition and erosion. Aim of this study is to evaluate whether the Product Environmental Footprint (PEF) approach is suitable for assessing the environmental impacts of white and black peat used in growing media as well as to provide generic data collected from growing media producers and the scientific literature. It is not the aim of this study to compare different land use options for peatlands. PEF is developed in order to make environmental product declaration claims more reliable, comparable and verifiable across the EU, and to increase consumer confidence in eco-labels and environmental impact information. For PEF-compliant studies all land use activities must be considered. For peat extraction either pristine peatland or previously drained peatland used for forestry or agriculture has to be transformed. Hence, the suitability of land use-related PEF indicators is also investigated. Diesel consumed for peat extraction, electricity used for peat processing and transport are the main contributors to acidification. Fuel production and consumption are the main contributors to human toxicity, with heavy metals to air and freshwater the contributing emissions. Ionising radiation, ozone depletion and resource depletion of minerals and metals are mainly caused by the electricity used. Climate change increased from 26 kg CO2eq per m3 processed white peat to 51 kg CO2eq per m3 processed black peat. The use of peat causes substantially higher GHG than the previous life cycle stages combined; white peat causes approximately 183 kg CO2eq per m3 and black peat 257 kg CO2eq per m3. Environmental impacts caused by peat are variable and depend on a number of spatial and temporal factors. Although most indicators used for PEF are suitable for assessing peat systems, that does not apply for the land use indicator and is at least questionable for the water use indicator, respectively, its weighting factor. Consequently, it is neither possible to identify the most relevant impact categories based on normalised and weighted results nor to calculate an overall single score for peat containing growing media. Since normalisation and weighting are mandatory steps for PEF conform impact assessment, the current PEF approach is not suitable to assess peat as intermediate product without adaptation of the land use indicator.


Introduction
Land provides the principal basis for human livelihoods and well-being including the supply of food, freshwater and multiple other ecosystem services (Shukla et al. 2019). Amongst the most vulnerable ecosystems to degradation are high-carbon-stock wetlands (including peatlands). Peatland ecosystems contain disproportionally more organic carbon than other terrestrial ecosystems (Beaulne et al. 2021). Peatlands cover only 3% of the world's land area and contain 550 gigatons (Gt) of carbon (Joosten et al. 2012;Laine et al. 2009), but also 12-21% of global soil nitrogen (Limpens et al. 2006). Natural peatlands may shift between being carbon and greenhouse gas (GHG) sinks and GHG sources on seasonal and inter-annual time scales and show variability in carbon accumulation rates. Drainage of peatland for various purposes leads to high GHG emissions (Strack and Waddington 2008;Limpens et al. 2008;Murdiyarso et al. 2015;Evans et al. 2019). Drained peatlands are usually considered to be negligible methane sources, but they emit CH 4 under wet weather conditions and from drainage ditches. Whilst ditches cover only a small percentage of the drained area, emissions can be sufficiently high. Drained peatlands emit substantially more CH 4 as undrained ones (Wilson et al. 2016). Drainage causes also N 2 O emissions; the total amount depends on the nutrient content of the peat (Salm et al. 2012). Hence nutrient-rich peat causes higher N 2 O emissions than nutrient-poor peat Today there is < 15% of the original wetland area globally than was present 300 years ago (Davidson 2014). Around 65,000 km 2 or 10% of the European peatland area has been lost and 44% of the remaining European peatlands are degraded (Tanneberger et al. 2017). Peat is used for energy generation, horticulture, landscaping and other purposes. Peat extraction rapidly removes carbon from the peatland carbon store and furthermore leads to substantial losses of carbon from the extraction site by stimulating decomposition and erosion Peat extraction can be sub-divided into four steps: peatland preparation, peat extraction or excavation, peat processing and peatland restoration. Peatland conversion involves the initial drainage of the peatland. Backhoes and bulldozers cut drainage ditches to lower the water table in the peatland (Waddington et al. 2009). The affected area by drainage was assumed to be twice of the extraction area; the area not used for extraction is called surrounding area (Uppenberg et al. 2001;Zetterberg et al. 2004). For horticultural peat two extraction methods are commonly used. For sod peat blocks of peat are excavated and left on the extraction site to dry. For milled peat the upper peat layer is milled before stockpiling in order to enhance drying Exhausted peatlands cannot return to functional peatland ecosystems because the viable seed bank has been removed during the extraction process. Reducing the GHG emissions from peat soils is a space-and cost-effective climate change mitigation option within the land use and agricultural sectors (Wilson et al. 2015;Ojanen et al. 2010; Barbier and Burgees 2021;Günther et al. 2020). Soil carbon sequestration and avoidable emissions through peatland restoration are both strategies to tackle climate change. However, restoring peatlands is 3.4 times less nitrogen costly and involves a much smaller land area demand than mineral soil carbon sequestration (Leifeld and Menichetti 2018) In order to return the exhausted peatland to a net carbon sink again, active restoration is needed. The most common restoration method is rewetting. For that drainage ditches are blocked in order to increase the water table. Abandoned peat extraction sites that are not rewetted remain important sources of carbon emissions. Several studies have demonstrated the co-benefits of rewetting degraded peatlands for biodiversity, and improvement of water storage and quality (Martin-Ortega et al. 2014) with beneficial consequences for human well-being (Bonn et al. 2014;Parry et al. 2014) Emissions from restoration will likely decrease through time; however, Schaller et al. have shown that even after 18 years of rewetting, peatland was still a GHG source (Schaller, Hofer et al. 2022). In order to return to a fully functioning ecosystem, it can take several decades (Lunt et al. 2010;Alm et al. 2007;Strobl et al. 2019) and it might take up to 2000 years to restore the carbon lost due to peat extraction (Clearly et al. 2005) Peat extraction leads to substantial carbon losses through vegetation removal during site preparation, drainage of the extraction site and its surroundings and the peat collection process (e.g. milling which increases aeration and oxidation of the upper peat layer) (Barthelmes 2018; Waddington et al. 2009;Sarkkola, eds. Sarkkola 2007). When the sod or milled peat is stored in piles the decomposition of peat continues, which can lead to emissions of particulate matter and can cause methane emissions (Silvan et al. 2012); methane emission from stock piles during winter time are substantially higher than in summer (Nykänen et al. 1996). In addition to particulate emissions from piles, the bare dark and lightweight soils of the excavation sites are easily warmed and susceptible to wind and water erosion (Holden et al. 2006;Li et al. 2018;Tuukkanen et al. 2014). The milling process causes also emissions of particulate matter (Tissari et al. 2006) Organic matter decomposition is a well-known phenomenon in wetlands when the water table is lowered and the oxygen content in the upper soil or peat layer increases (Grant et al. 2012). Increased oxygen availability initiates a cascade of organic matter breakdown that culminates in peat decomposition (Fenner and Freeman 2011). That causes also the release of nutrients (Wind-Mulder et al. 1996;Nieminen 2004;Tuukkanen et al. 2017) and methane emissions from ditches, which can have substantial impact on the overall GHG emissions from drained peatlands (Sundh et al. 2000;Uppenberg et al. 2001;Zetterberg et al. 2004;Parish et al. 2008) Peat extraction shows adverse effect on biodiversity and water quality in Ireland (MacDonald 2020). Nutrient export from drained peatland has significant impacts on aquatic environments in the northern hemisphere (Laine et al. 2013). It is difficult to predict the change in concentration of certain nitrogen compounds in peat pore water since different processes of N cycle are functioning simultaneously (Vassiljev et al. 2018(Vassiljev et al. , 2019. In many peat land areas in The Netherlands target concentrations for nitrogen (N) and phosphorus (P) in surface water of peatland catchments are exceeded (van Beek et al. 2007); also elevated dissolved organic carbon (DOC) concentration in Scottish peatland catchments affects drinking water quality (Ferretto et al. 2021). Leaching of nitrate was identified at 16 of the 20 United Kingdom Acid Waters Monitoring Network sites during [2005][2006]. Headwaters draining eroded South Pennine (UK) peatlands are nitrogen saturated (Daniels et al. 2012). Draining peatland for agriculture in Germany can release 1.000 kg N·ha −1 ·yr −1 from the top layer (upper 30 cm) (Gäth et al. 1997). In Finland concentrations of nutrients and suspended solids were relatively high in catchments containing drained peatland (Koskinen et al. 2017;Marttila et al. 2018). In Estonia the release of nutrients from peat excavation during the drainage construction and drainage is substantially higher than from natural bogs. 1 The release of nutrients from peat excavation areas causes eutrophication in rivers and may have a mortal effect to river's fish stock. Large quantities of nitrogen are released to freshwater bodies, particularly during storm events, subsequently enter coastal estuarine and marine ecosystems with negative consequences (Wang et al. 2016) Studies of environmental impacts of peat have mainly focused on greenhouse gases (Zetterberg et al. 2004;Hagberg and Holgrem 2008;Kirkinen et al. 2008;Seppälä eta l. 2010;Grönroos et al. 2013). Different studies generate different results depending on the choice of peatland, surrounding area, production methods and the after-extraction alternatives (Höglund and Martinsson 2013;Murphy et al. 2015). Few LCA studies of peat utilisation (Peano and Loenrincik 2012;Eymann et al. 2015;Stucki et al. 2019) take other than GHG emissions into account Environmental impacts from land use change such as land preparation before peat extraction and restoration after peat extraction are not always considered by studies mentioned above, despite restoration/renaturation is sometimes required by peat extraction permits. Moreover, most LCA studies of peat utilisation do not take nutrient release and particulate matter emissions from peat extraction areas into account. The reasoning for excluding those emissions is mostly not provided The Product Environmental Footprint (PEF) approach is developed in order to make environmental product declaration claims more reliable, comparable and verifiable across the EU, and to increase consumer confidence in ecolabels and environmental impact information. 2 During the development and testing of the approach, the approach has been subject of serious critique from the scientific community (Pedersen and Remmen 2022). The PEF method is continuously updated. The latest update of the PEF method was published in 2019 (Zampori and Pant 2019) and used as Annex of the "COMMISSION RECOMMENDATION on the use of the Environmental Footprint methods to measure and communicate the life cycle environmental performance of products and organisations" (EC 2021) Several product systems are examined from various industrial sectors and relevant data provided in PEF-compliant format. Peat as intermediate product is not covered so far; moreover product category rules for growing media are under development but not published yet Aim of this study is to evaluate whether the PEF approach is suitable for assessing the environmental impacts of peat used in growing media as well as to provide generic data collected from growing media producers and the scientific literature. It is not the aim of this study to compare different land use options for peatlands

Methodology
In general, the precautionary principle is applied for this study in order not to underestimate environmental impacts. A life cycle approach following the most recent PEF methodology is used considering all emissions associated with the extraction and processing of peat. According to PEF all known emissions shall be considered (Zampori and Pant 2019). For peat extraction either pristine peatland or previously drained peatland used for forestry or agriculture has to be transformed. Whilst peat is extracted the land is occupied and after peat extraction the area is frequently rewetted, i.e. re-transformed. All land use activities from area preparing to area restoration must be taken into account for PEF-compliant studies (Zampori and Pant 2019).

a. Goal and scope
Goal of this study is to provide information on environmental impacts of peat as intermediate product used for the production of growing media. In addition, the suitability of relevant indicators used for PEF impact categories is assessed for peat as an intermediate product.
In the PEF method, it is required that the functional unit (FU) be defined based on the four elements "what," "how much," "how well," and "how long." Since peat is quantitatively the most important but not the only ingredient in growing media, the latter two questions are not considered here. The FU is 1-m 3 processed white or black peat ready to be used as growing media ingredient. The reference flow equals the FU. The use of growing media is not subject of this study, although GHG emissions from peat are calculated according to PAS2050-1 (BSI 2012) and discussed. Different growing media applications require different peat fractions/qualities. The peat fractions/qualities are producer dependent and related data is protected by confidentiality agreements. For sake of simplicity, an average energy consumption for processing 1-m 3 harvested, processed and stored black and, respectively, white peat is used here.
The system boundary includes peatland preparation and restoration, peat extraction, storage, and processing and is shown in Fig. 1; the system boundary is in agreement with Grönoos (Grönroos et al. 2013). It corresponds to a cradleto-gate analysis. Hence, neither the production of growing media nor its utilisation is part of this analysis apart from GHG emissions from peat, which are separately reported.
Peatland drainage and re-wetting may occur in many different ways over different timescales and their effect cannot simply be attributed to lowering and raising the water table. The variability of peat extraction systems is high (Grönroos et al. 2013). Relevant parameters/assumptions determining emissions particularly of peat extraction are area demand, extraction yield, emission factors for extraction and renaturation/restoration, peat density and moisture, etc. The annual peat yield can vary by a factor of 4 depending on weather conditions, diesel consumption by a factor of 3, the surrounding area by a factor of 2 and also drainage efficiency (Uppenberg et al. 2001, Hagberg andHolgrem 2008). Particulate matter emissions occur from peatland erosion, peat storage and applied extraction technique. For the latter measurements are available (Tissari et al. 2006) and already used in LCA (Eymann et al. 2015), particulate matter emissions from erosion are highly variable and not considered here. Leaching from drained peatland causes various emissions, such as nutrients, dissolved carbon and suspended solids. In a nutshell, emissions from land use, from peat extraction, from storage and peat processing depend on a number of factors and modelling assumptions. Therefore, it is paramount to describe the data limitation and modelling assumptions, if transparency as well as reliable, reproducible and verifiable results are the final aim.

b. Important assumptions and limitations
-GHG emissions during drainage and area preparation are assumed to be equal as during peat extraction, whilst nutrient emissions are approximately three times higher during drainage than during the extraction. 3 The drainage period can vary between 2 and 5 years (Uppenberg et al. 2001;Hagberg and Holgrem 2008).
-Land previously used for forestry or agricultural is considered to be burden free, because emissions from drainage are allocated to the previous use (Stichnothe 2021). -Every ha used for peat extraction causes another ha of "surrounding" area, which has also been drained. -GHG emissions from land use and land occupation are calculated based on emission factors (EFs) as shown in the Annex. For this study peat extraction in temperate climate is assumed. -The fossil carbon in peat containing products applied at cultivation in open fields or glasshouses is assumed to be completely oxidised in line with PAS2050-1 and ISO14067 (BSI 2012; ISO 2018). -Leaching emissions such as nutrients, suspended solids and dissolved organic carbon are taken from Kløve (2001). Nutrient leaching during the restoration period is not known. Therefore, it is assumed that the nutrient emissions correspond with DOC emissions; hence the ratio of DOC leaching during peat extraction and DOC leaching during rewetting is used to calculate nutrient leaching from rewetting. -Particulate matter (PM) emissions and particle distribution from harvesting are taken from Tissari et al. (2006). PM emissions from wind erosion of the drained area and the stockpiles are not considered. -Diesel consumption for sod-und milled-peat extraction, on-site transport and processing from German growing media producers is based on an annual average from different sites. It results in approximately 770 l·ha −1 ·a −1 for black peat, which is broadly in line with Pakere, who reports 7.67 kWh per m 3 peat (Pakere and Blumberga 2017). -The time period for drainage is 3.5 years, for area preparation of previously used peatland 1 year, for peat extraction 20 years and for restoration 20 years. Areabased emissions are calculated and then recalculated on a per m 3 basis using the annual average extraction yield. For the drainage period that means emissions per ha and year during drainage multiplied by the drainage period (3.5 years) and divided by the extraction period (15 years), → emissions per ha divided by the annual yield that results in emissions from drainage per m 3 . -Emissions from peat storage at the extraction site and at the growing media producer's site are not exactly known. Solbikova reports that 5% of peat-carbon is decomposed per year (Stolbikova and Chertkova 2021), which in agreement with Clearly et al. (2005). The storage period during stockpiling can vary from weeks to years and the onsite storage lasts frequently several months. For this analysis, it is assumed that 95% of decomposed peat-C is emitted as CO 2 and 5% as CH 4 . The storage period for stockpiling peat is assumed to be 1 year and on-site storage of processed peat 3 months.
-Depending on the occurrence of weed seeds peat sometimes has to be sanitised by steaming before it can be used for growing media (Grießer 2016). Sanitation requires fossil resources, but it is not considered in this study. -The choice of post-harvest treatment depends on many factors. Here, only rewetting is assumed to be used as post-extraction treatment. Peatland restoration emissions were assumed to be constant over the restoration period, although that is a simplification. The diesel consumption for blocking ditches, etc. is not specified; it was not separately reported by peat producers but included in the figures for peat extraction. -A transport distance of 20 km in rural areas with a truck 5t of payload capacity is assumed. This is a rather arbitrary choice but can easily be adapted when data on transport means and distances are available. -For peat long-distance transport, the route from the Baltic to Germany is assumed, which is 3.200-km ocean transport in bulk carrier and 200-km road transport by lorry.
Relevant parameters used in this study are shown in Table 1 and Table 2 together with minimum and maximum values from scientific literature.

Life Cycle Inventory
An extended literature search is conducted in order to gain relevant data that is complemented by a survey amongst German growing media producers. The latter data is provided on basis of confidential agreement and for this study averaged and anonymised. In general, data selection is based on the hierarchy outlined in the FAO (2016). The life cycle assessment is carried out in Gabi version 10.6.0.110 with background data from the ILCD-database, ecoinvent 3.8 4 (cut-off, capital goods excluded) and GABI in that order. European average datasets (EU-28) of ILCD are used whenever possible.

Modelling emissions from land transformation and occupation, or peat extraction, respectively
Pristine peatland must be drained and the top layer removed, whilst for land previously used for forestry or agriculture just the top layer have to be removed. The material is frequently used to refill exhausted peat extraction areas and/or to block ditches for rewetting. Emissions from land transformation are calculated using area-based emissions from peat. Area-based GHG EFs are taken from the latest IPCC Guidelines for National GHG Inventories from Wetlands (Hiraishi et al. 2014). EFs for rewetting are taken from Wilson et al. (2016). A similar approach is used for nutrient emissions occurring during land transformation and land occupation. If more information on the location of peat extraction is available the decision tree provided in Annex 1 can be used to select specific national EF.
Fuel consumption data for white and black peat are collected from peat producers in Germany. Moreover, plausibility tests are conducted and the variation of some parameters summarised, which can be used for sensitivity analysis. However, the latter is not part of this paper. Relevant data and references for peat extraction are provided in Table 1 and for leaching in Table 2.

Life Cycle Impact Assessment (LCIA)
The PEF development and amendments are well documented (Sala et al. 2017(Sala et al. , 2018Fazio et al. 2018, b;Saouter et al. 2018). First, we screen the suitability of PEF's environmental impact categories, and second, we will discuss those EF impact categories that are dominated by the foreground activities and not by background datasets. Environmental Footprint 3.0 version (EF3.0) is used for this study. The suitability of relevant PEF indicators for assessing peat extraction is discussed below.

Land use
The recommended midpoint model for land use is a dimensionless soil quality index (SQI) based on LANCA (Bos et al. 2016). The soil quality index builds upon the aggregation of four indicators from LANCA ). PEF's approach for land use approach does not provide elementary flows for the transformation from or to (drained) peatland; hence land use impacts from peat extraction cannot be properly addressed. Moreover, the LANCA indicator mechanical filtration, used to calculate the soil quality index, is not available for peatland. In the LANCA documentation it is verbatim stated "The soil class X displays peatland. The kf-value of peatlands is dependent on the degree of decomposition and has to be determined specifically. Therefore, this class has not been included into LANCA ® " (Beck et al. 2010). LANCA ® characterisation factors for mechanical filtration in peatland are not provided (Bos et al. 2016). 6 Hence, the soil quality indicator of PEF is not suitable for peat extraction and will therefore not used in this study.

Water footprint
7 Using the AWARE approach (UNEP 2016) is challenging for peat, because drainage of peatland does hardly affect water availability or consumption at the watershed level. The use of water can be specified for all technical processes involved. However, freshwater is drained from peatland and will sooner or later enter coastal estuarine and marine ecosystems (Wang et al. 2016), but the amount varies between years and is usually not exactly known. This applies also for rewetting exhausted peat extraction sites; the amount of water additionally stored in rewetted peatlands is not exactly known. Hence, neither the amount of drained freshwater from peatland nor the amount for rewetting is included in the results shown in Table 3. 8 Water stress or water scarcity is usually not an issue in peatland regions. Therefore, this indicator is of minor relevance for peat and peat-containing products. Hence, it is questionable whether the corresponding footprint weighting factor is appropriate when peat systems are assessed.

Eutrophication
Eutrophication in EF3.0 is expressed for three different environmental compartments, namely, eutrophication of freshwater, marine water and terrestrial systems (Zampori and Pant 2019). The latter is relevant for the utilisation of growing media when nutrients are added, but it is not for peat extraction; moreover peat layers are constantly removed. For freshwater and marine eutrophication, the respective REC-IPE methods are used. RECIPE considers for freshwater 12 elementary flows (only N) and for marine eutrophication 17 elementary flows (only P) (Bach and Finkbeiner 2017). RECIPE does not consider impacts from biological material (DOC, BOD, or COD 5 ) for eutrophication, which is suitable in most peat-related cases because peat is nutrient poor and the eutrophication in respective areas is determined by the nutrient content rather than released organic matter.

Results and discussion
Results for white peat from drained peatland, previously used peatland and white peat transported over long distance are shown in Table 3. Table 4 shows 9 disaggregated results for peat extraction from previously used peatland, i.e. without environmental impacts from drainage and long-distance transport.
A number of environmental impacts are associated with the energy used for peat extraction, transport and processing and consequently determined by the background data. The right column in Table 3 shows the contribution of energy production and consumption to the environmental impact categories. Diesel consumed for peat extraction, electricity used for peat processing and transport are the main contributors to acidification, with ocean transport contributing the most. Ocean transport and diesel production are the main contributors to HT-cancer and HT-non-cancer, with heavy metals to air and freshwater the contributing emissions. The picture looks similar for freshwater ecotoxicity where more than 40% is caused by emissions from transport and diesel consumption. Ion-HH, OD and Res M&M are mainly caused by the electricity used for peat processing.
Land use and peat extraction activities contribute the most to CC, Eutro, PM and the fossil resources. White peat extraction and processing causes 31 kg CO 2eq per m 3 when transported over long distances, and 26 kg CO 2eq per m 3 when extracted close to the growing media producer. During the use phase peat is completely oxidised, which results in 183 kg CO 2eq per m 3 and corresponds  et al. 2001). Peat is a fossil resource and extracted; thus, all GHG emissions are aggregated rather than separately reported as fossil, biogenic and from land use. If required, it can be easily calculated with the information provided. GHG emissions from the utilisation of white peat are substantially higher than GHG emissions from extraction and processing. The same applies for the use of fossil resources as peat is a fossil resource. PM is mainly caused by peat extraction and ocean transport; if the latter is not required more than 95% is caused by peat extraction. Eutro marine is caused by emissions of various nitrogen species, approximately 70% from nutrient emissions and 30% due to energy use; in summary more than 90% is caused by land preparation and peat extraction. Phosphate emissions from land preparation and occupation cause approximately 90% of Eutro fw.
Results for black peat are shown in Table 5. The system is separated into five life cycle stages. Additionally, to the life cycle stages shown in Table 4, 10 GHG emissions from stockpiling for 1 year at the extraction site are included 11 .
In contrast to white peat GHG emissions from stockpiling are singled out, because black peat is usually stored to dry at the extraction site. In peat extraction areas approximately 2% of the area is used for stockpiles. Similar to white peat the diesel consumed for peat extraction and electricity used for peat processing are the main contributors to acidification; ocean transport is not part of the system but would have a similar effect on acidification as for white peat.
Diesel production is the main contributor to HT-cancer and HT-non-cancer, with heavy metals to air and freshwater the most relevant emissions. The picture looks similar for freshwater ecotoxicity where emissions from diesel used for peat extraction, transport and processing contribute the most. Ion-HH, OD and Res M&M are mainly caused by the electricity (EU-28-mix) used for peat processing. That of course depends on the electricity mix, and may change when national electricity mixes are used. The environmental impacts discussed so far are determined by the background datasets.
The contributors to the environmental impacts, e.g. fossil resource or acidification, are the same as for white peat, although the figures change. The main difference for black peat compared to white peat is the consideration of stockpile emissions. Methane emissions occur during stockpiling (Nykänen et al. 1996), although very few long-term measurements have been conducted. CC increased from 26 kg CO 2eq per m 3 processed white peat to 51 kg CO 2eq per m 3 processed black peat, whereat 23 kg CO 2eq coming from stockpiling. Both GHG emissions from using black peat as well as fossil resource are substantially higher because of the elevated C-content of black peat compared to white peat. The same applies to the GHG emissions from using peat; white peat causes approximately 183 kg CO 2eq per m 3 and black peat 257 kg CO 2eq per m 3 . Nutrient emissions from drainage and peat extraction sites are assumed to be the same for black and white on a per ha basis. This is reasonable as both are frequently different layers of the peat profile at the extraction site. Site-specific information on nutrient leaching from peat extraction sites are scarce.

Conclusion
Environmental impacts caused by peat extraction, processing and transport are variable and depend on a number of spatial and temporal factors. Peatland restoration might have potential to mitigate the negative climate impact of peat extraction; however, information about the dynamic of annual carbon balances and GHG emissions is limited (Järveoja et al. 2016). In a nutshell, emissions from land use, from peat extraction, from storage and peat processing depend on a number of factors and modelling assumptions. Therefore, it is paramount to describe the data limitation and modelling assumptions, if transparency as well as reliable, reproducible and verifiable results are the final aim. This paper provides generic values for black and white peat used for horticulture. Data and assumptions are presented in a clear and transparent manner, so that data can be easily adapted to site-specific and/or company-specific conditions.
The PEF approach is used for this study and the corresponding requirements, e.g. for cut-off emissions and complete inventories tested. All land use activities, including land preparation and restoration, must be included as well as the corresponding emissions, i.e. GHG emissions and nutrient emissions. The same applied for emissions due to peat extraction, i.e. emissions from diesel consumption and particulate matter. Although most indicators used for PEF are suitable for assessing peat systems, that does not apply for the land use indicator and it is at least questionable for the water use indicator. The LANCA ® characterisation factors for mechanical filtration in peatland are not available. Hence, the aggregated soil quality indicator of PEF is not suitable for peat extraction. The use of water can be specified for all technical processes involved. However, freshwater is constantly drained from peatland, but the amount varies between years and is usually not exactly known. Water scarcity is usually not an issue in regions where peat is extracted; therefore, it is questionable whether the corresponding PEF weighting factor is appropriate. Consequently, it is neither possible to identify the most relevant impact categories based on normalised and weighted results nor to calculate an overall single score.
When land use-related environmental impacts are of interest then climate change, freshwater eutrophication and particulate matter are the most relevant environmental impact categories, with land occupation (peat extraction) as the main contributor. For climate change using peat is much more relevant than its extraction and processing. This is a cradle-to-gate study that can be used as data source for comparing growing medias of different compositions. However, in that case it must be ensured that the compared systems provide the same function. Solely comparing single compounds of growing media does not provide meaningful information.
Supplementary information.
Acknowledgements I thank all anonymous German growing media producers and experts in the field for their valuable input and for providing data and feedback on the assumptions taken for this study. I also thank the reviewers for their constructive comments that helped to amend the paper.
Funding Open Access funding enabled and organized by Projekt DEAL.
Data availability All data and data sources used for this study are presented in Tables 1 and 2. Raw data collected from growing media producers are subject of confidentiality agreements and cannot be shared.

Conflict of interest The author declares no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.