Life-cycle assessment of typical Portuguese cork oak woodlands

Science of the Total Environment 452–453 (2013) 355–364

Contents lists available at SciVerse ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Life-cycle assessment of typical Portuguese cork oak woodlands Sara González-García ⁎, Ana Cláudia Dias, Luis Arroja CESAM, Department of Environment and Planning — University of Aveiro, 3810-193 Aveiro, Portugal

H I G H L I G H T S • • • • •

An environmental evaluation of Portuguese cork oak woodlands was performed. Two different scenarios were assessed in detail by means of ten impact categories. Differences were identified based on the intensity and repetition of forest activities. Cork stripping was identified as the environmental hotspot due to the cleaning and pruning processes. A sensitivity assessment of cork yield was performed since it is lower than the one used in this study.

a r t i c l e

i n f o

Article history: Received 8 October 2012 Received in revised form 18 February 2013 Accepted 18 February 2013 Available online 25 March 2013 Keywords: Environmental impact Montado Portugal Quercus suber L. Reproduction cork

a b s t r a c t Cork forest systems are responsible for making an important economic contribution to the Mediterranean region, especially Portugal where the cork oak woodlands or montados contain about 32% of the world's area. The environmental profile derived from reproduction cork production and extraction in two Portuguese regions (Tagus valley and Alentejo) representative of the Portuguese sector were assessed in detail using the Life-Cycle Assessment (LCA) methodology from a cradle-to-gate perspective. The production line was divided into four stages considering all the processes involved: stand establishment, stand management, cork stripping and field recovery. According to the environmental results, there were remarkable differences between the two production scenarios mainly due to the intensity and repetition of forest activities even though the cork yield was reported to be the same. The management system in the Alentejo region presented the worse environmental profile in almost all the impact categories under assessment, mainly due to the shorter cycle duration of the mechanical cleaning and pruning processes. Cork stripping was identified in both scenarios as the production stage with the highest contribution to the environmental profile due to the cleaning and pruning processes. A sensitivity assessment concerning the cork yield was performed since the average production yields in the Portuguese montados are lower than the ones used in this study. Thus, if the cork yield is reduced, the environmental profile in both scenarios gets worse since almost all the forest activities involved are the same. © 2013 Elsevier B.V. All rights reserved.

1. Introduction The cork oak (Quercus suber L.) belongs to the Fagaceae family and it is a perennial oak, which can achieve an average height of 15–20 m (Pereira, 2007). It is a long-lived species (200–250 years) and it is characterised by the presence of a conspicuous thick and furrowed bark with a continuous layer of cork in its outer part. This cork bark gives the cork oak its economic and environmental relevance since this is a unique tree in the world (Pereira, 2007), and cork production is the most important source of revenues in cork oak agroforestry (Borges et al., 1997). The outer bark is formed by elastic, impermeable and good thermal insulating tissue (the cork) composed of dead cells

⁎ Corresponding author. Tel.: +351 234 370 387. E-mail address: [email protected] (S. González-García). 0048-9697/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.scitotenv.2013.02.053

with walls that are impermeable due to a chemical compound named suberine (Pereira, 2007). The extraction of cork (the process known as stripping and usually carried out once every nine to fourteen years, depending on the region) does not damage the tree and after that, a new bark begins to form behind the newly exposed trunk surface. The cork oak forests are spread out in the Mediterranean climate zone of Western Europe (Portugal, Spain, France and Italy) and North Africa (Morocco, Northern Algeria and Tunisia) (Pereira, 2007; Rives et al., 2011). This tree is well adapted to hot and dry summers, and to soils of low fertility. In the Iberian Peninsula, cork oaks usually constitute stands of single species although stands with mixed species (e.g. holm oaks and other deciduous oaks) are also common (Pereira, 2007). According to the last National Forest Inventory (AFN, 2010), Portuguese cork oak pure and dominant mixed stands occupy almost 716 thousand hectares (~ 23% of national forest land), representing

356

S. González-García et al. / Science of the Total Environment 452–453 (2013) 355–364

the third most important tree species after maritime pine and eucalyptus. In Portugal, the cork oak woodlands are called montados. The most productive regions are the Alentejo and Tagus valley regions, which represent the 72% and 21% respectively of total cork oak stands (APCOR, 2011). Although the annual cork production figures fluctuate since they depend on different environmental factors and climate, nearly 300-thousand tonnes of cork are annually extracted in the world and Portugal produces about 50% of it (APCOR, 2011). There are two main cork oak forest products: i) reproduction cork, which is obtained from the third extraction onwards and presents the best quality; and, ii) cork from the first and second extraction (virgin and secondary cork respectively) with low quality. Remaining fragments of cork could be considered as forest waste. The cork with the best quality for industrial applications and with the highest market value is the reproduction cork. The Portuguese cork industry is highly developed, and is mainly located in the Aveiro and Setúbal districts (which provide 75% and 12%, respectively, of the cork industries employment) and is made up of 700 companies, which produce around 40-million cork stoppers per day, among other products, and employ about 12,000 workers (APCOR, 2011). The predominant use of the reproduction cork is as raw material for the cork stopper production since it is the most profitable product (Rives et al., 2011). This is also the most important application in Portugal and the wine industry absorbs 66% of the total cork stopper product in Portugal (APCOR, 2009). The Life-Cycle Assessment (LCA) methodology has proved to be a useful tool for the identification of the environmental impacts in numerous forest-related studies focused not only in the forest management activities (González-García et al., 2012a, 2012b; Dias and Arroja, 2012) but also in the wood products manufacture (Dias et al., 2007; González-García et al., 2012c). Concerning the cork sector, there are a few studies where the environmental performance of cork-based products manufacture have been evaluated: natural cork stoppers (Rives et al., 2011), champagne cork stoppers (Rives et al., 2012a) and cork granulate (Rives et al., 2012b) as well as its comparison with alternative materials (PricewaterhouseCoopers, 2008). However, in all of them, forest management activities were excluded from the system boundaries. So far, no detailed environmental studies have been performed for the cork produced in Portugal. The aim of this study was to assess the environmental impacts by means of the LCA methodology of cork production and extraction in Portuguese cork oak woodlands. The general approach of this study was gathering inventory data corresponding to cork production and extraction, quantifying their environmental impacts and identifying the environmental hotspots of cork oak woodlands. 2. Materials and methods LCA is a standardised framework (ISO, 14040, 2006) to evaluate the environmental burdens associated with a product, process or activity throughout its life cycle (Guinée et al., 2001). In this study, the guidelines proposed in the ISO 14040 (2006) were followed in order to assess the production and extraction of raw cork in Portuguese cork oak woodlands from a cradle-to-gate perspective excluding from the analysis the subsequent uses of the cork. The functional unit, i.e. the reference flow to which all the flows are assigned, was one tonne of fully equilibrated reproduction cork (6–10% moisture content). 2.1. System description Although cork oak stands may occur spontaneously in Portugal, the current systems are not natural systems but they are the result of centuries of oriented silvicultural practices that have shaped the trees and the cork oak woodland. Cork oak stands are complex since the products extracted today are the result of earlier management activities and the management

activities also depend on the type of cultivation model (Pereira, 2007). Cork oak systems can be classified in two different silvicultural systems (Pereira, 2007; Rives et al., 2012c): a) Cork oak forests: these are common in Catalonia (North West Spain) and have vertical and horizontal diversity with an average density of 400 trees∙ha−1. The cork extraction is undertaken every 12–14 years. b) Cork oak woodlands: these are the Portuguese montado or Spanish dehesa (Southern Spain), and they form heterogeneous habitats with varying age and height, interspersed with grass and, in some cases, cereal crops. The tree density varies from 50 to 150 trees∙ha−1. The cork extraction is undertaken every 9 years due to the better climatic and yield conditions. Moreover, considerable differences can be found in the raw cork yield between both systems: 150 kg ha−1∙year and 200 kg ha−1∙year for forest and woodlands respectively (Rives et al., 2012c). In this study, two Portuguese montados were assessed with an average lifespan of 170 years in both. They are representative of the management practices carried out in cork oak woodlands located in two of the most important cork-producing regions: Tagus valley (Scenario A) and Alentejo (Scenario B). About 93% of the cork produced in Portugal comes from these two regions (APCOR, 2011). The production systems were divided in four main stages: i) stand establishment, ii) stand management, iii) cork stripping and, iv) field recovery. All these stages included different processes that have been identified in the assessed plantations and are described in detail below. Fig. 1 shows the scheme of the system boundaries with the corresponding processes involved in each stage. In addition, Table 1 displays a detailed description of all the silvicultural processes that have been identified and considered throughout the life cycle of the montados. Although the stands under assessment are located in two different regions, there are only some minor differences between them that will be explained below. 2.1.1. Stand establishment stage The stand establishment stage involves two different processes: field preparation and cork oak plantation. Field preparation includes activities driven to improve the hydric equilibrium of the soil without affecting its physical stability. The first activity involved is the elimination of the spontaneous vegetation (cut-over clearing), which commonly takes place in the spring or autumn. Different techniques can be used in this process depending on the site conditions: it can be mechanised with a tractor and a specific implement such as discs or even manual. In the plantations under assessment, it is carried out with a disc harrow. The second activity should be soil mobilisation or scarification in order to improve the soil quality, facilitating airing and water circulation. There are also different techniques depending on the site characteristics: ploughing, sub soiling and ripping are some examples. Although these activities are normally mechanised, they could also be manual. In the stands under assessment, they are only performed in Scenario A using a ripper pulled by a tractor. The stand located in the Alentejo region (Scenario B) does not employ these activities. The cork oak plantation is performed in both scenarios with plants directly transported by truck from a nursery located 50 km away. Due to the lack of information, the nursery-related activities were excluded from the system boundaries. Both stands are managed with a double plough, which passes leaving furrow-hillock surface soil. After that, the plants are manually planted (commonly between October and February) and fertiliser is applied to each plant. An initial density of 400 plants/ha was considered in both scenarios, and the applied dose and type of fertiliser are dependent on the considered region. In Scenario A, 125 g of fertiliser NPK (7-21-21) was applied per plant, while in Scenario B, 80–100 g of fertiliser PK (22.5–5) was applied per seedling because the soil is

S. González-García et al. / Science of the Total Environment 452–453 (2013) 355–364

357

Stand establishment stage

D

P

Cut-ov er clearing

M

M

D

Planting

plants

D

Ripping

Furrowhillocking

Fertilizing

Dead plants substituion

M

plants

P

fertilizer N,P,K F Stand management stage

D

M

M

D

Cleaning

M

D

Thinning

Pruning

D

REPRODUCTION CORK

M

Cleaning

Thinning operations (x 2)

VIRGIN + SECONDARY CORK

Cork stripping stage

D

M

cork slabs

M

D

M

D

Pruning

Cleaning

Stripping

D

Cork loading + storage

M

Repeated every 9 years

Field recovery stage (170th year)

Workers transport D

Final f elling

M

SYSTEM BOUNDARIES

KEY P = cork plants production in a nursery . It has been excluded f rom assessment D = diesel/petrol production M = f orest machinery production (tractors and implements, chainsaws) F = mineral f ertilizer production; NPK 7-21-21 f or Scenario A and PK 22.5-5 f or Scenario B Fig. 1. Flowchart for the Portuguese cork oak production systems located in the Tagus valley (Scenario A) and Alentejo (Scenario B).

rich in nitrogen. It is important to note that 1 year after the plantation, it is usual to substitute dead plants with new ones. The mortality of plants was 20% in Scenario A and 40% in Scenario B mainly due to meteorological and geo-morphological conditions. 2.1.2. Stand management stage Throughout the life cycle of the montados, it is required to monitor the plantation. It starts with the cleaning carried out in the 2nd or 3rd year after the planting step and at the end of the spring in order to pull up the spontaneous vegetation since it may contribute to the mortality of the cork oak plants. This step is performed in both scenarios with a disc harrow. The vegetation removed is spread over the field to protect the soil and this step is repeated every 3 or 4 years depending on the field fertility. The next step is the pruning (5th–7th year) of the stems in order to favour the growth of the tree, correct specific malformations and increase the biomass production. Pruning can be manual or mechanised. In this study, the first pruning is manual using scissors. However, pruning steps will be repeated (before the first stripping step) every 3 or 4 years in Scenario A and every 8 years in Scenario B (Table 1).

Another silvicultural step is the thinning performed with a chainsaw and required to avoid the competition between trees. The first thinning step is carried out in the 10th–13th year and 50% of the trees are thinned. The second thinning is performed in the 15th– 20th year and once again, 50% of the trees are thinned. Thus, the final density of trees in both stands is 100 trees per ha. After each thinning, there is a cleaning step with a disc harrow. 2.1.3. Cork stripping stage This stage involves the stripping steps in order to extract the cork from the trees. This process is not harmful to the cork oak normal development. Nowadays, stripping is carried out manually with an axe in both scenarios under assessment. The first harvesting of cork or stripping takes place when the tree reaches 20–30 years of age depending on the scenario and its growth conditions. The first cork extracted is called virgin cork and it does not meet good quality requirements for industrial use. The second stripping is performed around 9 years after the first stripping and the cork extracted (secondary cork) presents deep fractures and, therefore, also does not meet the quality standards for industrial transformation. Only after the third stripping, when the cork oak is around 35–45 years of age, does the

358

S. González-García et al. / Science of the Total Environment 452–453 (2013) 355–364

Table 1 Management sequence and processes definition for Portuguese cork oak production (lifespan 170 years). All the data correspond to field data. Time (year)

Operation

Tractor

Implement

Power (hp) 0

Soil preparation

0

Planting of seedlings

1

Cut-over clearing Ripping Furrow-hillock Planting Fertilising Dead plants substitution

2–3

Cleaning

5–7 8–10 10–13

1st Pruning 2nd Pruning Thinning

Chainsaw

Cleaning Pruningc

Chain saw

Thinning

Chain saw

Cleaning Pruningd

Chain saw

15–20

20–30

29–39

38–48

>47

e

170

b c d e f g h i j

Disc harrow

1.5–2

1–1.5

Disc harrow

Chain saw

2nd Strippingg

Manual

Cleaning Pruning

Chain saw

3rd Strippingh

Manual

Cleaning Pruning

Chain saw

1.5–2

1–1.5

Disc harrow

1.5–2

1–1.5 20–40 trees/worker∙day 8 h/day∙worker 80 trees/worker∙day 8 h/day∙worker

110–180

Disc harrow

1.5–2

1–1.5 20–40 trees/worker∙day 8 h/day∙worker 80 trees/worker∙day 8 h/day∙worker

110–180

Disc harrow

1.5–2

1–1.5 20–40 trees/worker∙day 8 h/day∙worker 80 trees/worker∙day 8 h/day∙worker

110–180

Disc harrow

1.5–2

1–1.5

Manual 110–180

Pruningj

Chain saw

Final felling

Chain saw

400 plants/ha 125 g NPK/planta 80–100 g PK/plantb 80 plants/haa 160 plants/hab Repeated every 3–4 yearsa Repeated every 3 yearsb Only in the Scenario A Removal of 200 trees/ha 120–200 trees/worker∙day 8 h/day∙worker

Manual

Cleaning Pruningf

Stripping

1–1.5 – 1.5–2

Comments

20–40 trees/worker∙day 8 h/day∙worker Removal of 100 trees/ha 120–200 trees/worker∙day 8 h/day∙worker 110–180

i

2–5.5 2–2.5 1.5–4.5

Manual Manual

110–180

1st Stripping

Disc harrow Ripper Double plough Manual Manual

Scenario B

Manual 110–180

Cleaning

a

90 140 80

Work time (h/ha) Scenario A

Disc harrow

1.5–2

1–1.5

20–40 trees/worker∙day 8 h/day∙worker 80 trees/worker∙day 8 h/day∙worker Repeated every 5 yearsa Repeated every 3 yearsb 20–40 trees/worker∙day 8 h/day∙worker 3 trees/h 8 h/day∙worker

Scenario A: stand located in Tagus valley. Scenario B: stand located in Alentejo. This is the 2nd pruning for the scenario B and the 3rd pruning for the Scenario A. This pruning is specific for the Scenario A. 1st Stripping: extraction of virgin cork. There is one pruning step per stripping. 2nd Stripping: extraction of secondary cork. 3rd Stripping: extraction of reproduction cork 15 kg cork∙tree−1. Stripping processes: extraction of reproduction cork every 9 years. Pruning steps are repeated every 25 years in the Scenario A and 8–9 years in the Scenario B.

extracted cork obtain the indispensable quality requirements for the cork stopper sector (the main product of the Portuguese cork sector). This cork is known as reproduction cork and onwards, is also extracted every 9 years. The cork extracted in each stripping step is cut into sheets of an appropriate size. The cork is loaded into tractors and transported around 3 km between the field and the storage place. Between the stripping steps, there are cleaning and pruning steps with a disc harrow and a chain saw, respectively, approximately every 9 years and the biomass extracted is left over the soil for improvement of soil quality. The average production is 2–5 kg∙tree −1 of virgin cork, 10 kg∙tree −1 of secondary cork and 15 kg∙tree −1 of reproduction cork in the third stripping. In the subsequent strippings, the amount of reproduction

cork presents an average value of 30 kg tree −1 during the remaining lifespan (45–170 years); these values correspond to green cork just extracted from the tree with an average moisture content of 22–25% (Pereira, 2007). However, the water content in the cork is variable since the cork starts losing water after its harvesting. Thus, it is estimated that the cork with commercial properties is fully equilibrated under ambient air conditions when its moisture content is between 6 and 10%, which occurs around 21 days after the stripping. 2.1.4. Field recovery stage At the end of the life of the tree (approximately 170 years), it is felled with a chain saw. The stumps are not removed and remain on the land.

S. González-García et al. / Science of the Total Environment 452–453 (2013) 355–364

2.2. Inventory data Primary and site-specific data (information concerning tractors and implements, work times and input rates) were mainly supplied by associations of cork producers located in both regions under assessment by means of surveys and interviews (Table 1). Average values for these specific data have been considered although a sensitivity assessment has been proposed in order to identify possible differences. Secondary data concerning the production of the different agricultural inputs, such as fertilisers, fuel and machinery were taken from the Ecoinvent database (Althaus et al., 2007). The use of mineral fertilisers is an important source of nutrient-related emissions with an important contribution to several impact categories such as global warming, acidification and eutrophication. Moreover, the nutrient emission rates are variable due to the influence of soil type, climatic conditions and agricultural management practices (Brentrup et al., 2000). Nitrogen-based diffuse emissions (ammonia, nitrous oxide, nitrogen and nitrates) were calculated following the emission factors proposed by Audsley et al. (1997), Brentrup et al. (2000) and EMEP/ CORINAIR (2006). The phosphate emissions were estimated following the emission rate proposed by Rossier (1998), i.e. 0.01 kg kg −1 of applied P from a phosphate-based fertiliser application. Combustion emissions from diesel and petrol engines used in the field machinery (disc harrowing, ripping, furrow-hillocking, cleaning, thinning, pruning and final felling), as well as from agricultural inputs delivery (truck) and workers transportation to the plantations (passenger car) have also been taken into account and taken from the Ecoinvent database (Nemecek and Käggi, 2007; Spielmann et al., 2007). Changes on the overall soil carbon content were excluded in this study due to its specific dependency on factors such as management practices, climate, previous cropping regime and the limitation of available data in Portuguese conditions. 2.3. Allocation procedure An allocation procedure was considered since different types of cork, with variable qualities and uses, are stripped throughout the whole life cycle of the cork oak tree. The reproduction cork, i.e. the cork extracted from the 3rd stripping and onwards, is the product with the highest market price — approximately up to 5.6 times higher than the remaining cork biomass (APCOR, 2009). Therefore, an economic allocation was proposed based on the selling prices for the different types of piled cork. The selling prices for the virgin cork and reproduction cork were estimated as 0.43 € kg −1 and 2.07 € kg −1 respectively (APCOR, 2009). It was not easy to find an estimation of selling price for the secondary cork. Thus, the same price as the virgin cork was assumed based on their similar applications (APCOR, 2009). Table 2 shows market prices, quantity of stripped cork and allocation factors for both products. It is important to mention that besides cork, wood is also obtained from activities such as pruning, thinning and final harvesting of cork oak trees at the end of their life cycle. This wood is commonly burned in house boilers for heating purposes. However, in this study, it was decided to attribute all the impacts of the woodland management system to the total cork extracted (virgin, secondary and reproduction cork) because there is no information concerning the total Table 2 Prices and allocation factors for the cork based products for both scenarios under assessment. Prices based on APCOR (2009). Material

Quantity (t)

Price (€/kg)

Allocation factor

Reproduction cork Virgin + secondary cork

37.42 1.25

2.07 0.43

99% 1%

359

amount of wood biomass produced and also because cork is the resource that articulates the montados and supplies the main revenues (Rives et al., 2012c). 2.4. Methodology Among the steps defined within the life cycle impact assessment stage of the standardised LCA methodology, only classification and characterisation stages were undertaken for the assessment of the Portuguese cork oak montado (ISO, 14040, 2006). The characterisation factors reported by the Centre of Environmental Science of Leiden University (CML 2001 method) were used (Guinée et al., 2001). The impact potentials (or impact categories) evaluated according to the CML method were: abiotic depletion (ADP), acidification (AP), eutrophication (EP), global warming (GWP), ozone layer depletion (ODP) and photochemical oxidant formation (POFP). In addition, impact categories for toxicological potential (human toxicity (HTP), fresh water aquatic ecotoxicity (FEP), marine aquatic ecotoxicity (MEP) and terrestrial ecotoxicity (TEP)) were also analysed. The software SimaPro 7.3 was used for the computational implementation of the inventories (Pré Consultants, 2012). 3. Environmental results 3.1. Comparative environmental results Table 3 summarises the environmental comparative results between both scenarios under assessment for the production of reproduction cork. These environmental impacts correspond to average data from both plantations, which are representative of the corresponding production areas. According to the results shown in Table 3, Scenario B located in the Alentejo region presented the worst environmental results in almost all the categories. These environmental differences are mainly associated with differences on the frequency of processes such as pruning and cleaning. Cleaning processes are repeated every 3 years in Scenario B, while in Scenario A they are executed every 3 or 5 years depending on the stand's age. Concerning the pruning processes, they are repeated approximately every 8–9 years in Scenario B and every 25 years in Scenario A. Differences in the stripped cork production yield (including virgin, secondary and reproduction cork) have not been identified according to the consulted sources and it was estimated as 38.7 t ha −1 or 228 kg ha −1 yr −1, being 220 kg ha −1 yr −1 of reproduction cork. These values correspond to cork after the stripping and being fully equilibrated (water content of around 8% (Pereira, 2007)). Thus, it could be assumed a yield of 200 kg ha −1 yr −1 of dry reproduction cork (8 kg ha −1 yr −1 of virgin and secondary cork). Fig. 2 shows the contributions from the cork oak woodland stages (stand establishment, stand management, cork stripping and field recovery) to the environmental impact categories under assessment for both cork oak systems. The different Table 3 Comparative environmental results per functional unit (1 tonne of reproduction cork fully equilibrated) between both Montados scenarios under assessment: Scenario A located in Tagus valley and Scenario B located in Alentejo. Impact category

Unit

Scenario A

Scenario B

ADP AP EP GWP ODP HTP FEP MEP TEP POFP

kg Sbeq kg SO2eq kg PO43−eq kg CO2eq mg CFC-11eq kg 1,4-DBeq kg 1,4-DBeq kg 1,4-DBeq g 1,4-DBeq kg C2H4eq

1.99 1.44 0.42 280.00 34.00 230.00 51.00 32.00 33.00 0.32

2.27 1.30 0.43 304.00 38.00 180.00 56.00 30.00 71.00 0.73

360

S. González-García et al. / Science of the Total Environment 452–453 (2013) 355–364

4 Field recovery stage Cork stripping stage Stand management stage

kg/t reproduction cork

Stand establishement stage

3

2

1

0 A

B

A

B

ADP (kg Sbeq) AP (kg SO2eq)

A

B

EP (kg PO43eq)

A

B

A

B

A

B

A

B

GWP (kg ODP (kg CFCHTP (kg FEP (kg CO2eq*500) 11eq/20000) 1,4DBeq*150) 1,4DBeq*50)

A

B

MEP (kg 1,4DBeq*50)

A

B

TEP (kg 1,4DBeq/50)

A

B

POFP (kg 1,4DBeq)

Fig. 2. Contributions from the different stages to the environmental profile of the cork oak production systems located in the Tagus valley (Scenario A) and Alentejo (Scenario B).

contributions to the environmental impact categories were analysed as follows: 3.1.1. Abiotic depletion potential (ADP) This impact category relates to extraction of minerals and fossil fuels. The contributions in both scenarios are dominated by the cork stripping stage with contributions of 81% and 85% respectively for Scenario A and Scenario B since numerous processes take place in this stage – and for approximately 150 years – including pruning and cleaning steps as well as the cork transportation up to its storage. All these processes are mechanical using tractors and/or chain saws. The extraction of the cork (stripping) is manual and does not show any contribution except the transportation of workers to the forest. 3.1.2. Acidification potential (AP) Scenario A presented the worst result with up to 9% higher AP than Scenario B. The cork stripping stage is once again mainly responsible for the contributions to this category (82% and 86% of the total for Scenario A and B respectively). The requirement of a higher amount of fertiliser in Scenario A and the absence of N-based fertiliser application in Scenario B considerably influenced the results. Moreover, Scenario A requires a ripping stage and the effective working times for the different processes involved are higher for Scenario A than for Scenario B (Table 1), which involves higher diesel requirements. Acidifying emissions are dominated by the emission of nitrogen oxides and sulphur dioxide in both scenarios. 3.1.3. Eutrophication potential (EP) The difference in the emissions of eutrophying substances is almost negligible in both scenarios. This impact category is commonly affected in agricultural and forest systems by the application of fertilisers and on-field-derived emissions. Cork oak stands require low amounts of fertiliser in the establishment of the stands and there is no extra fertilisation during the tree growing period. Moreover, Scenario B does not require nitrogen-based fertiliser application since the soil is rich in this nutrient. Thus, the contribution from the fertilising step to this impact category is lower in Scenario B. The cork stripping stage is the main contributor with contributions of 81% and 84% for Scenario A and B respectively. The cleaning steps in

Scenario A and the pruning steps in Scenario B are mainly responsible for this result. The most important contributing eutrophication substances were nitrogen oxides, nitrate and phosphate regardless of the scenario, derived from the processes involved and forest machinery production and use. 3.1.4. Global warming potential (GWP) This impact category is around 9% higher for Scenario B than Scenario A. The average values were estimated to be 233 kg CO2eq∙ t −1 and 253 kg CO2eq∙ t −1 for Scenario A and B respectively. The cork stripping stage is the dominant stage of the contributions to GWP (81% and 86% respectively for Scenario A and B). The main greenhouse gas emissions were fossil CO2 emissions derived from the diesel combustion in the forest machines. 3.1.5. Ozone layer depletion (ODP) In this impact category, Scenario B presents a worse profile being around 13% higher than for Scenario A. The cork stripping stage is the dominant stage with contributions that add up to 81% and 85% of the total for Scenario A and Scenario B respectively. Halon 1301 emissions derived from diesel and petrol production required in the forest machines and worker transportation to the forest were mainly responsible for these emissions in both scenarios. As expected, the processes with high energy requirements presented the highest effects on this impact category. 3.1.6. Human toxicity potential (HT) In this impact category, Scenario B presents a better environmental profile than Scenario A, which is up to 22% lower. Once again, the cork stripping stage is the main area of concern (83% and 88% of total contributions respectively for Scenario A and B) due to the cleaning steps. In this impact category, the airborne emissions of chromium VI and polycyclic aromatic hydrocarbons mainly derived from forest machinery production (tractor and implements) and considerably dominated the results. It is important to remark that in the cork stripping stage, the steps are continuously repeated (e.g. pruning or cleaning), which involves high requirements of machinery with its corresponding maintenance.

S. González-García et al. / Science of the Total Environment 452–453 (2013) 355–364

3.1.7. Ecotoxicity potentials This topic includes the contributions to all the categories of ecotoxicity potentials: fresh water aquatic ecotoxicity (FEP), marine aquatic ecotoxicity (MEP) and terrestrial ecotoxicity (TEP). Scenario B presents the worse profile in FEP and TEP (9% and 117% higher than Scenario A). In contrast, in terms of MEP, Scenario B shows the best profile (4% lower than Scenario A). The cork stripping stage is once again mainly responsible for the contributions to these impact categories with contributions ranging from 75% to 82% in Scenario A and from 79% to 87% in Scenario B. As in the HTP, these impact categories are affected by the forest machinery production and use. Contributing emissions to FEP and MEP were mainly nickel and cobalt emissions. Contributing substances to TEP were cypermethrin and mercury. Cypermethrin derives from soya oil required by the chain saws. 3.1.8. Photochemical oxidant formation potential (POFP) The results obtained in this category for Scenario B are 131% higher than for Scenario A. This impact category is also mainly affected by the cork stripping stage, representing 74% and 79% of total contributions to POFP. In terms of contributing substances, POFP was mainly caused by carbon monoxide emissions in both scenarios principally derived from the petrol combustion in the chainsaws. 3.2. Environmental assessment per stage As previously mentioned, the cork oak forest systems were divided into four stages (Fig. 1): stand establishment, stand management, cork stripping and field recovery. All these stages involve several processes or operations. Moreover, the transportation of workers to the forest in the different stages was included in the assessment. Next, the contributions to each impact category from the different activities/steps involved in each cork oak woodland stage will be assessed. 3.2.1. Stand establishment stage According to Fig. 1, this stage involves the following activities: cut-over clearing, ripping (only in Scenario A), furrow-hillocking, planting, fertilising, substitution of dead plants and transport of

361

workers for the different activities. These activities take place in the years 0 and 1 of the plantations. Some of these activities i.e. planting, fertilising and substitution of dead plants are manual so they did not have any direct contribution to the environmental profile. However, it must not be forgotten that the fertilisation involves the production and application of fertilisers as well as the field emissions derived from fertilisers and they have been considered separately. The contributions from the different activities of the stand establishment stage to the environmental impact categories are shown in Fig. 3 for both scenarios. According to Fig. 3, the cut-over clearing activity is mainly responsible for the contributions to the different impact categories with contributions ranging from 32% to 41%. It is followed by the ripping and furrow-hillocking activities requiring higher work hours (h/ha) than in Scenario B (Table 1), which involves higher diesel requirements. Concerning the Scenario B, there is a clear dominance of the furrow-hillocking activity with contributions ranging from 31% to 52%, depending on the impact category. The second most important factor in this stage should be the cut-over clearing. Although both steps have similar diesel consumption in L/h, the effective work time is higher for the furrow-hillocking step (~ 40% higher), which is reflected in the results. Also, transport-related activities for the supply of fertilisers and workers transportation are an important hotspot specifically in terms of ADP, GWP, ODP, TEP and POFP, i.e. categories affected by the diesel production and combustion-derived emissions. 3.2.2. Stand management stage As shown in Fig. 1, this stage involves the following activities/steps: several cleaning steps of the field, pruning steps, two thinning steps and the transport of corresponding workers to the forest. The number of cleaning and pruning operations in this stage depends on the scenario. Thus, Scenario A commonly carries out two cleaning steps before the first thinning, one more after each thinning step and one more cleaning before the first stripping. Concerning the pruning, there are two manual operations with scissors and only one mechanised with a chainsaw in this period of time (2nd–20th years). Regarding Scenario B, there are some differences. There are several cleaning steps although the frequency is higher than in Scenario A (approximately every

1,8E-01

Transport activities

kg/t reproduction cork

1,6E-01

Fertilizers production Furrow-hillocking

1,4E-01

Ripping Cut-over clearing

1,2E-01

Field emissions 1,0E-01 8,0E-02 6,0E-02 4,0E-02 2,0E-02 0,0E+00

A

B

A

B

A

B

A

B

A

B

A

B

A

B

A

B

A

B

A

B

ADP (kg Sbeq)AP (kg SO2eq) EP (kg PO4GWP (kg ODP (kg CFCHTP (kg FEP (kg MEP (kg TEP (kg POFP (kg 3eq) CO2eq*500) 11eq/100000) 1,4DBeq*500) 1,4DBeq*100) 1,4DBeq*25) 1,4DBeq/100) 1,4DBeq/10) Fig. 3. Contributions from the different activities involved in the stand establishment stage to the environmental profile in the cork oak production systems located in the Tagus valley (Scenario A) and Alentejo (Scenario B).

362

S. González-García et al. / Science of the Total Environment 452–453 (2013) 355–364

1,25 Transport activities Pruning Thinning Cleaning

kg/t reproduction cork

1,00

0,75

0,50

0,25

0,00

A

B

A

B

A

B

ADP (kg Sbeq) AP (kg SO2eq) EP (kg PO43eq/5)

A

B

A

B

A

B

A

B

GWP (kg ODP (kg CFCHTP (kg FEP (kg CO2eq*100) 11eq/100000) 1,4DBeq*100) 1,4DBeq*25)

A

B

A

B

MEP (kg TEP (kg 1,4DBeq*25) 1,4DBeq/100)

A

B

POFP (kg 1,4DBeq/10)

Fig. 4. Contributions from the different activities involved in the stand management stage to the environmental profile in the cork oak production systems located in the Tagus valley (Scenario A) and Alentejo (Scenario B).

3 years). Thus, in the same period of time there could be up to eight cleaning operations, all of them mechanised. Concerning the pruning activity, the first one is manual (5th–7th year) and the following (every 8 years) are mechanised. In both scenarios there are two thinning processes where 50% of the trees are removed. They are carried out in the 10th–13th year and in the 20th year. These differences convert this stage for Scenario B on a more intensive stage than in the Scenario A, which is reflected in categories such as ADP (fuel requirements) and POFP (derived emissions). The contributions from

the involved activities in this stage to the environmental impact categories are shown in Fig. 4 for both scenarios. According to Fig. 4, the cleaning step in Scenario A is the main hotspot in almost all impact categories except TEP and POFP, where the pruning steps are the main contributors. Also, transport-related activities are important in categories such as ADP, GWP, ODP and MEP. Fig. 4 also displays the contributions from activities involved in Scenario B. The pruning is the main hotspot followed by the cleaning steps. The high contribution from the pruning step is due

6 Cork loading & storage Transport activities

5

Pruning

kg/t reproduction cork

Cleaning

4

3

2

1

0

A

B

A

B

ADP (kg Sbeq)AP (kg SO2eq)

A

B

EP (kg PO43eq/5)

A

B

GWP (kg CO2eq*100)

A

B

A

B

A

B

ODP (kg CFCHTP (kg FEP (kg 11eq/50000) 1,4DBeq*100) 1,4DBeq*50)

A

B

A

B

MEP (kg TEP (kg 1,4DBeq*25) 1,4DBeq/100)

A

B

POFP (kg 1,4DBeq/10)

Fig. 5. Contributions from the different activities involved in the cork stripping stage to the environmental profile in the cork oak production systems located in the Tagus valley (Scenario A) and Alentejo (Scenario B).

S. González-García et al. / Science of the Total Environment 452–453 (2013) 355–364

to the fact that only the first one is manual with scissors and the remaining ones are mechanised. 3.2.3. Cork stripping stage This stage involves the following activities: stripping (first extraction of the virgin cork, second extraction of the secondary cork and upcoming extractions of the reproduction cork — in total around 15 extractions of reproduction cork), cleaning (every 3–4 years in both scenarios), mechanised pruning (every 25 years in Scenario A and every 8–9 years in Scenario B) and cork loading and storage. These last activities are performed with a tractor and the biomass is transported between the forest and the storage place. Moreover, the transport of workers has also been included. Scenario B presents a worse profile in comparison with Scenario A due to the highest petrol requirements for the pruning stages. Fig. 5 displays the contributions from these activities in each impact category to the total contribution from the cork stripping stage. According to these figures, there are two clear hotspots in both scenarios: the cleaning and pruning steps. Cleaning processes are the hotspot in Scenario A and pruning processes in Scenario B. 3.2.4. Field recovery stage In this stage, only the final felling of the trees at the end of their life cycle (170 years) is considered, which is a mechanised operation with chainsaws in both scenarios. The stumps are not removed from the cork oak woodland according to the consulted cork producers. This process is affected by three factors: the number of trees to be cut per ha, the effective work and the petrol consumption. All of them are the same for both scenarios. Thus, the contributing emissions to the different impact categories are the same regardless of the forest system. 4. Sensitivity assessment of cork yield in Portuguese montados The raw cork and reproduction cork production in the montados under assessment are considerable. Under the Portuguese conditions,

363

the yields of reproduction cork were reported to be around 200 kg ha−1 yr −1 under commercial conditions (Rives et al., 2012c), i.e. ca. 184 kg ha−1 yr−1 of dry reproduction cork. However, the last National Forest Inventory reported for 2005 values of around 125.5 kg ha−1 yr−1 of dry reproduction cork for Portugal being 133 kg ha−1 yr −1 and 138.5 kg ha−1 yr−1 in pure stands in the Alentejo and Tagus valley respectively (AFN, 2010). The previous National Forest Inventory for 1995 reported values (on a dry basis) of 170 kg ha−1 yr −1, 177 kg ha−1 yr−1 and 179 kg ha−1 yr−1 for Portugal, Alentejo and Tagus valley respectively (DGF, 2001). Cork yield per hectare depends on the density of trees in the plantations, on the tree/cork growth, on the type of stands (pure or dominant), soil properties, climate, water availability and on the forestmanagement conditions (good practices or not). Thus, a sensitivity assessment was performed in order to analyse the effect of cork yield on the environmental profile of the montados. The alternative scenarios proposed were the following based on the cork production yields reported in the last two National Forest Inventories: • Scenario A1: alternative to Scenario A considering reproduction cork yield of 138.5 kg ha −1 yr −1. • Scenario A2: alternative to Scenario A considering reproduction cork yield of 179 kg ha −1 yr −1. • Scenario B1: alternative to Scenario B considering reproduction cork yield of 133 kg ha −1 yr −1. • Scenario B2: alternative to Scenario B considering reproduction cork yield of 177 kg ha −1 yr −1.

an average dry an average dry an average dry an average dry

The assessment was performed considering 1 tonne of dry reproduction cork as the functional unit and the same allocation approach (allocation factor of 99%) as in the base scenarios (Scenarios A and B). Table 4 displays the comparative results for the different alternative scenarios proposed in this sensitivity assessment. As expected, the environmental profile got worse when the cork yield was reduced since all forest activities involved in a hectare were the same regardless of the cork yield. The consideration of more than 200 kg ha−1 yr−1 of dry reproduction cork in the sensitivity

Table 4 Comparative environmental results per functional unit (1 tonne of dry reproduction cork) between the scenarios proposed in the sensitivity assessment based on the effect of the cork yield. Tagus valley Impact category

Unit

138.5a kg ha−1 yr−1 (Scenario A1)

179a kg ha−1 yr−1 (Scenario A2)

200a kg ha−1 yr−1 (Scenario A)

ADP AP EP GWP ODP HTP FEP MEP TEP POFP

kg Sbeq kg SO2eq kg PO43−eq kg CO2eq mg CFC-11eq kg 1,4-DBeq kg 1,4-DBeq kg 1,4-DBeq g 1,4-DBeq kg C2H4eq

3.15 2.28 0.67 445.00 54.00 364.00 81.00 50.00 52.00 0.51

2.44 1.77 0.52 344.00 41.00 282.00 63.00 39.00 40.00 0.39

2.15 1.56 0.46 304.00 37.00 250.00 56.00 34.00 36.00 0.35

Alentejo Impact category

Unit

133a kg ha−1 yr−1 (Scenario B1)

177a kg ha−1 yr−1 (Scenario B2)

200a kg ha−1 yr−1 (Scenario B)

ADP AP EP GWP ODP HTP FEP MEP TEP POFP

kg Sbeq kg SO2eq kg PO43−eq kg CO2eq mg CFC-11eq kg 1,4-DBeq kg 1,4-DBeq kg 1,4-DBeq g 1,4-DBeq kg C2H4eq

3.74 2.15 0.71 501.00 63.00 295.00 93.00 50.00 118.00 1.21

2.82 1.62 0.53 377.00 47.00 222.00 70.00 38.00 88.00 0.91

2.46 1.41 0.47 329.00 41.00 195.00 61.00 33.00 77.00 0.08

a

Production in terms of dry reproduction cork.

364

S. González-García et al. / Science of the Total Environment 452–453 (2013) 355–364

assessment was not taken into account here since it would be an unrealistic scenario and no references can be found in the literature not only for cork production in Portugal but also for other cork producing countries. According to the results, the highest cork production yield (under the same conditions) should tend to involve a lower environmental impact. Some cork products stay in use for long periods, storing carbon and delaying its return to the atmosphere. Even cork products with a relatively short lifetime may store carbon for long periods if they are disposed of in landfills at the end of their life, because under anaerobic conditions their decay is slow and incomplete mainly due to the presence of suberin and lignin, the major cork components (Pereira, 2007; Silva, 2009). Despite the potential of cork products to accumulate carbon and thus, to contribute to climate change mitigation, the amount of carbon stored in these products is currently unknown. 5. Conclusions Cork forest and cork oak woodland systems are responsible for making an important economic contribution to the Mediterranean region especially in Portugal, where the montados contain about 32% of the world's area. Different types of cork are stripped in the montados with different applications and prices, being the reproduction cork the main source of revenue. Although the montados are found all over the country, in this study special attention was paid to two different scenarios located in the Tagus valley and Alentejo. The environmental results show that there are important differences between the two production scenarios, mainly due to the intensity and repetition of forest activities, although the cork yield is reported to be the same. Management systems in the Alentejo presented the worse environmental profile in almost all the impact categories under assessment. It is mainly due to the shorter period of repetition of the cleaning and pruning processes (environmental hot spots), both of which are mechanised. Moreover, a sensitivity assessment concerning the cork yield was performed since the average production yields in the Portuguese montados vary a lot. As expected, the environmental profile in both scenarios got worse if the cork yield decreased since almost all the forest activities involved were the same regardless the yield (except the stripping processes and the cork loading and storage, which effect on the environmental profile is almost negligible). Future studies should pay attention to the quantification of other sources of biomass produced in the cork oak woodlands in order to obtain an overview of this forest sector. Acknowledgments Dr. S. González-García would like to express her gratitude to the Spanish Ministry of Education for financial support (Grant reference: EX2009-0740) for a Postdoctoral Research Fellowship taken at CESAM. Dr. A.C. Dias would like to thank FCT (Science and Technology Foundation — Portugal) for the Postdoctoral Fellowship (SFRH/BPD/ 20363/2004). Thanks are also due to the project ECOTECH SUDOE — International Network on LCA and Ecodesign for Eco-innovation (SOE2/P2/E377) funded by the EU Interreg IV B Sudoe Programme.

References AFN. 5th National Forest Inventory (5° Inventário Florestal Nacional). Lisboa: Autoridade Florestal Nacional; 2010. Althaus HJ, Chudacoff M, Hellweg S, Hischier R, Jungbluth N, Osses M, et al. Life cycle inventories of chemicals. Ecoinvent report No. 8. Dübendorf: Swiss Centre for Life Cycle Inventories; 2007. APCOR. Cork yearbook 2009. Available at:http://www.apcor.pt/2009. [accessed 01.08.2012]. APCOR. Cork yearbook 2011. Available at:http://www.apcor.pt/userfiles/File/Publicacoes/ anuario2011.pdf2011. [accessed 01.08.2012]. Audsley E (Coord.), Alber S, Clift R, Cowell S, Crettaz P, Gaillard G, et al. Harmonisation of Environmental Life Cycle Assessment for Agriculture. Final Report. Concerted Action AIR3-CT94-2028. European Commission. DG VI Agriculture. SRI, Silsoe, UK; 1997. Borges JG, Oliveira AC, Costa MA. A quantitative approach to cork oak forest management. For Ecol Manage 1997;97:223–9. Brentrup F, Küsters J, Lammel J, Kuhlmann H. Methods to estimate on-field nitrogen emissions from crop production as an input to LCA studies in the agricultural sector. Int J Life Cycle Assess 2000;5:349–57. DGF. National Forest Inventory — 3rd Revision (Inventário Florestal Nacional — 3ª Revisão). Lisboa: Direcção-Geral das Florestas; 2001. Dias AC, Arroja L. Environmental impacts of eucalypt and maritime pine wood production in Portugal. J Clean Prod 2012. http://dx.doi.org/10.1016/j.jclepro.2012.07.056. Dias A, Arroja L, Capela I. Life cycle assessment of printing and writing paper produced in Portugal. Int J Life Cycle Assess 2007;12(7):521–8. EMEP/CORINAIR. Atmospheric emission inventory guidebook. Technical report, No. 11. Copenhagen, Denmark: European Environment Agency; 2006. González-García S, Mola-Yudego B, Dimitriou I, Aronsson P, Murphy RJ. Environmental assessment of energy production based on long term commercial willow plantations in Sweden. Sci Total Environ 2012a;421–422:210–9. González-García S, Bacenetti J, Fiala M, Murphy RJ. Present and future environmental impact of poplar cultivation in Po Valley (Italy) under different crop management systems. J Clean Prod 2012b;26:56–66. González-García S, García-Lozano R, González M, Castilla Pascual R, Gabarrell X, Rieradevall J, et al. Environmental assessment and improvement alternatives of a ventilated wooden wall from LCA and DfE perspective. Int J Life Cycle Assess 2012c;17:432–43. Guinée JB, Gorrée M, Heijungs R, Huppes G, Kleijn R, de Koning A, et al. Life Cycle Assessment. An operational guide to the ISO standards. Leiden (The Netherlands): Centre of Environmental Science; 2001. ISO 14040. Environmental management – life cycle assessment – principles and framework; 2006. Nemecek T, Käggi T. Life cycle inventories of agricultural production systems. Final report ecoinvent v2.0 No. 15a. Agroscope FAL Reckenholz and FAT Taenikon. Zurich and Dübendorf, Switzerland: Swiss Centre for Life Cycle Inventories; 2007. Pereira H. Cork biology, production and uses. Elsevier Science Ltd.; 2007 [ISBN-10: 0444529675. ISBN-13: 9780444529671]. PRé Consultants. . Available at:http://www.pre.nl/2012. [accessed 01.08.2012]. PricewaterhouseCoopers. Evaluation of the environmental impacts of cork stoppers versus aluminium and plastic closures. Analysis of the life cycle of cork, aluminium and plastic wine closures. Corticeira Amorim, SGPS, SA; 2008. Rives J, Fernandez-Rodriguez I, Rieradevall J, Gabarrell X. Environmental analysis of the production of natural cork stoppers in Southern Europe (Catalonia — Spain). J Clean Prod 2011;30:949–57. Rives J, Fernandez-Rodriguez I, Rieradevall J, Gabarrell X. Environmental analysis of the production of champagne cork stoppers. J Clean Prod 2012a;25:1-13. Rives J, Fernandez-Rodriguez I, Gabarrell X, Rieradevall J. Environmental analysis of cork granulate production in Catalonia — Northern Spain. Resour Conserv Recycl 2012b;58:132–42. Rives J, Fernández-Rodríguez I, Rieradevall J, Gabarell X. Environmental analysis of raw cork extraction in cork oak forests in southern Europe (Catalonia — Spain). J Environ Manage 2012c;110:236–45. Rossier D. Adaptation de la method ecobilan pour la gestion environnementale de léxploitation agricole. Lausanne, Switzerland: Service Romand de Vulgarisation Agricole; 199849. Silva RPM. Avaliação do ciclo de vida da rolha de cortiça natural. Master thesis. Porto: Faculdade de Engenharia — Universidade do Porto; 2009. Spielmann M, Bauer C, Dones R, Tuchschmid M. Transport services. Ecoinvent report No. 14. Dübendorf, Switzerland: Swiss Centre for Life Cycle Inventories; 2007.