Accepted Manuscript
Geospatial distribution of ecosystem services and biomass energy potential in eastern
Japan
Makoto Ooba, Minoru Fujii, Kiichiro Hayashi
PII:

S0959-6526(16)00100-1

DOI:

10.1016/j.jclepro.2016.01.065

Reference:

JCLP 6653

To appear in:

Journal of Cleaner Production

Received Date: 1 April 2015
Revised Date:

20 January 2016

Accepted Date: 24 January 2016

Please cite this article as: Ooba M, Fujii M, Hayashi K, Geospatial distribution of ecosystem services
and biomass energy potential in eastern Japan, Journal of Cleaner Production (2016), doi: 10.1016/
j.jclepro.2016.01.065.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to

our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.


ACCEPTED MANUSCRIPT
1

Geospatial distribution of ecosystem services and biomass energy potential in eastern Japan

2
3

Makoto Ooba*1, Minoru Fujii1, Kiichiro Hayashi2

5

1

National Institute for Environmental Studies, Tsukuba, Japan

6

2

EcoTopia Science Institute, Nagoya University, Nagoya, Japan

7


*Corresponding Author:

8
Abstract

SC

9

RI
PT

4

Detailed assessments of the effects of biomass production on ecosystems were carried out in the

11

eastern region of Japan using geographical statistics and statistical methods. Ecosystems that

12

might be used as a source of energy-related biomass already provide a variety of goods and

13

services for humans widely known as ecosystem services. Various indices were mapped to

14


describe the potential supply of biomass energy and the proxy variables for ecosystem services

15

provided in the region. These indices were analyzed using a multivariate statistical technique to

16

identify specific key factors for the use of biomass and ecosystem services. Finally, using

17

zoning software, priority areas of potential supply of biomass energy and ecosystem services

18

were indicated and the conflict between them analyzed. Biomass energy was clearly

19

distinguished into three axes, suggesting that biomass is strongly related to the location and

20

ecosystem, while the distribution of the types of ecosystem services in the studied areas was not

21

separated clearly. The relative priority ranks of bioenergy and ecosystem services were


22

complementarily distributed; however, parts of the studied area had high-ranking areas. The results

23

suggested that a more detailed zoning information is needed for promoting energy-related

24

biomass production considering the high supply of ecosystem services.

TE
D

EP

AC
C

25

M
AN
U

10

26


Keywords: Biomass, Eastern Japan, Ecosystem services, Geographic information system,

27

Spatial analysis, Zonation

28

1


ACCEPTED MANUSCRIPT
1

1. Introduction
Biomass is a crucial energy resource for creating a sustainable society because of its

3

renewability, low or no carbon emission, and low environmental impact. However, the use of

4

biomass has many disadvantages, as do other renewable resources such as solar and wind power.

5

Transport of the biomass is relatively difficult due to its high moisture content, and its gross

6


heating value is lower than that of other energy resources. Additionally, intensive production of

7

biomass can lead to competition with forest conservation and cultivation for food supply.

8

Harvesting the biomass from forests and agricultural ecosystems has also some effects on these

9

and the neighboring ecosystems, and the growth of biomass requires the use of land for this

SC

10

RI
PT

2

purpose over relatively long intervals.

Surveys of potential biomass were conducted from the regional to global level. Hoogwijk et

12


al. (2005) estimated the timeline of the production and consumption of biomass and land use at

13

the global level using their model (IMAGE mode, Hoogwijk et al., 2003) and considering

14

several Intergovernmental Panel for Climate Change (IPCC) scenarios. Moreover, Hoogwijk et

15

al. (2009), applying the economic cost-supply analysis, indicated a region at global level that is

16

of interest for its low production cost and high potential of biomass energy. Ericsson and

17

Nilsson (2006) analyzed the potential biomass supply in 15 EU countries. Henry (2010)

18

discussed a possibility for replacement of fossil fuel by biofuel using high-yielding crop and

19

biotechnology at a global level.


TE
D

M
AN
U

11

Previous studies have also suggested that the analysis of biomass supply may be conducted

21

at a small scale as well as at a country level using geospatial analysis. This is because

22

management and production costs affecting the potential (or available) amount of target biomass

23

also depend on geospatial conditions including ecosystem distribution, access roads, distance to

24

the production factory, and location of demand for the biomass; the cost should also include the

25

disadvantages of using the biomass as energy resource. Sacchelli et al. (2014) analyzed the


26

socio-economic and environmental effects of multiple factors on wood residue energy, including

27

geographical conditions on a local scale. They concluded that both the implementation of

28

advanced technology and environmental parameters related to allocation of sources and

29

demands were important. Delivand et al. (2015) also carried out geographical analysis, and the

30

effects of logistics costs and greenhouse gas emissions were discussed. The availability of land

AC
C

EP

20

2



ACCEPTED MANUSCRIPT
for bioenergy crops in Mozambique, in the timeframe 2005–2030, was modeled by van der Hilst

2

(2012). From geographical detailed analysis, the most suitable locations for bioenergy

3

production were determined based on agro-ecological suitability and accessibility and partly

4

based on the most suitable locations for current agricultural practices. Ooba et al. (2012, 2015)

5

described the relationship between the cost of woody biomass production and the geographical

6

location of forests in two different regions in Japan.

RI
PT

1

Detailed assessments of the effects of biomass production and consumption on ecosystems


8

and social systems, taking into account ecological processes and regional characteristics, have

9

not yet been conducted in Japan. After the great earthquake and nuclear accident in 2011,

10

renewable energy received more attention compared to the period before these disasters. The

11

Japanese government developed a new policy to promote the use of biomass (Ministry of

12

Economy, Trade and Industry of Japan 2014), and local governments, especially in the areas

13

damaged by the earthquake, began planning the development of biomass boilers and electric

14

generators (Kaji et al., 2013). Japan introduced a feed-in tariff (FIT) scheme for renewable

15


energy in 2012 to promote the use of these energy sources; hence, the demand for biomass is

16

now higher than it was before the earthquake. Under such conditions, more changes in land use

17

(e.g., conversion from forest to cropland) and development (e.g., conversion from natural forest

18

to plantation forest) may occur to enhance biomass production.

TE
D

M
AN
U

SC

7

Ecosystems that might be used as a source of energy-related biomass already provide a

20


variety of goods and services to humans widely known as ecosystem services (ES; Millennium

21

Ecosystem Assessment, 2005). Many studies have stressed the negative effects on ecosystems

22

due to production of biomass without considering the ecosystem services and biodiversity.

23

Several studies have indicated that the assessment of environmental impact of biomass production

24

for energy must consider the existing ecosystems and biodiversity of the potential production areas.

25

Hanafiah et al. (2012) found that inclusion of the impacts on biodiversity is needed for calculating

26

the production footprint by comparing the ecological footprint and biodiversity footprint.

27

Myllyviita et al. (2012) mentioned less impact of imported biomass compared to local biomass


28

production in Finland, as inferred from the life cycle assessment and the multi-criteria decision

29

analysis. Cao et al. (2015) performed an impact assessment of land use based on economic values

30

and ecosystem services at country level. The distribution of ecosystems appropriate for

AC
C

EP

19

3


ACCEPTED MANUSCRIPT
1

production of biomass is neither uniform nor coherent with the current land use. Ecological

2

impact assessment is also needed in relation to the development of biomass production, as


3

already pointed out in previous studies on biomass potential (e.g., Hoogwijk et al., 2005).
To suggest a conservation or development in specific areas, geospatial analysis may be

5

needed at a local scale. Many geographical software programs have been used in conservation

6

planning of the biodiversity in ecosystems under particular socioeconomic constraints (e.g.,

7

Moilanen et al., 2012). They can indicate hot spots and cold spots under certain conditions and

8

constraints (e.g., management cost, cost effectiveness, and subjective weight of various

9

services). These tools are also used to determine geographical priority in terms of biomass

SC

10


RI
PT

4

development and to resolve conflicts between development and conservation.

The objective of this study was to assess the impact of biomass production on ecosystem

12

services in the eastern region of Japan using geographical information system (GIS). Data sets

13

were collected and indices created by which to estimate the geographical distribution of

14

energy-related biomass and the current state of ecosystem services. Various indices were used to

15

map potential supplies of biomass energy and proxy variables for ecosystem services provided

16

in the region. These indices were analyzed using a multivariate statistical technique to identify

17


specific key factors for the use of biomass and ecosystem services. To detect potential hot spots

18

of these resources and areas of conflict with the current ecosystem, the potential supplies of

19

biomass energy and ecosystem services were assigned a rank using Zonation software

20

(Moilanen et al., 2012). A comparison of ranks with or without the FIT weighted prices was also

21

carried out to estimate the effect of the FIT system on ecosystem services. The results provided

22

useful planning and zoning information for promoting the production of biomass and

23

conservation of ecosystem.

AC
C


EP

TE
D

M
AN
U

11

24

In this study, the potential for biomass energy was evaluated for two energy-producing

25

processes: direct combustion of biomass, and fermentation of biomass to produce methane.

26

These methods are not the latest technology (Naik, et al., 2010), but they are relatively common

27

in Japan.

28
29


2. Models and study site

30

2.1. Study area

4


ACCEPTED MANUSCRIPT
1

The eastern part of Japan that was selected as the study area includes Kanoto, Tohoku, and

2

Jouetsu regions (14 prefectures; area: 110,000 km2). The islands located far from the Tokyo

3

metropolitan area were omitted from this study because of the difficulty in transporting the

4

biomass produced on these islands.

6

RI
PT


5
2.2. Data sources

Biomass data: A comprehensive biomass dataset from 2011 used in this study was provided

8

by the New Energy and Industrial Technology Development Organization (NEDO). This dataset

9

was initially developed by Iuchi (2003) for 15 types of biomass. Table data at the municipality

10

level published by the NEDO were downloaded and subdivided as follows (Table 1): wood

11

residual from forest (wf, two types); wood residual from other ecosystems (we, two types);

12

agricultural residual (aa, four types); grassland residual (ae, two types); livestock manure (ma,

13

five types); sludge (sl, three types); and food processing waste (fw, three types). The biomass


14

dataset provided the annual maximum potential, available amount (dry weight), and heat energy

15

(GJ/y). In this study, the available heat energy of the biomass was used for realistic estimation.

16

For livestock manure, sludge, and food processing waste, heat energy was generated by methane

17

fermentation, and for other types of biomass, heat energy was calculated using their lower

18

calorific value.

TE
D

M
AN
U

SC

7


These biomass data were represented in thermal units, on the assumption that they would be

20

used for combustion in boilers and in methane fermentation (NEDO, 2011). Woody and

21

agricultural biomass was combusted in biomass or multi-fuel combustion boilers with

22

combustion efficiency set to 1.0. Manure and food processing waste were consumed in a

23

methane fermentation plant; the detailed conditions of the fermentation are given in Table 1.

24

The energy of biomass-derived methane was used as heat. These assumptions were not fully

25

realistic, but they were effective for estimating the maximum potential amount of biomass in the

26

local areas considered.


27
28

AC
C

EP

19

Data on natural and social parameters about the study area were also obtained and used as
variables in calculations of biomass energy and ecosystem services (Table 2).

29

Biological data (Table 2): Vegetation maps (Vg) and data on the occurrence of mammalian

30

species (Sp) were obtained from the Biodiversity Center of Japan (2014). The distribution of the

5


ACCEPTED MANUSCRIPT
1

plant community and the degree of artificial disturbance (10 levels) were indicated on the Vg.


2

Sp data indicated the distribution of eight mammal species (Macaca fuscata, Cervus nippon,

3

Capricornis crispus, Ursus thibetanus, Sus scrofa, Vulpes vulpes japonica, Nyctereutes

4

procyonoides, and Meles meles).

6

Climatic data: Annual precipitation (Cp) and mean temperature (Ct) with a 1-km mesh

RI
PT

5

(30-year averages) were used (Ministry of Land, Infrastructure, Transport and Tourism, 2014).
Agricultural data: Areas of agricultural land use (Aa) and annual gross agricultural

8

production (Ag) were obtained from the World Census of Agriculture and Forestry in Japan and

9


from the Statistics of Agricultural Production and Income (Ministry of Internal Affairs and

10

SC

7

Communications, 2014), respectively. Ag statistics data were collected at prefectural levels.
Social data: Population data at a municipal level were obtained from the Population Census

12

(Statistics Bureau, Population Census 2014). Tourism spots (Tr) were listed according to the

13

Ministry of Land, Infrastructure, Transport and Tourism (2014), and domestic tourism statistics

14

(Tp) data were obtained from the Domestic Tourism Consumption Trend Survey for Tourism

15

(Japan Tourism Agency, 2014). Tr indicated locations of both natural spots and leisure facilities.

M
AN
U


11

TE
D

16
17

2.3. Data processing

18

2.3.1 Target year and processing software

The above-mentioned data for natural and social parameters were converted into raster (cell)

20

data with a 5-km mesh. The sampling year selected was 2010 because after the 2011 earthquake

21

in the northeast region of the country, the population and land use data were not as well

22

developed as before the earthquake. Some of the data were older than 2010 because of data

23


availability constraints.

AC
C

EP

19

24

The data were analyzed using Microsoft Excel 2013 statistical software (Excel statistics,

25

Social Survey Research Information, Japan). Geographical processing was carried out using

26

ArcGIS 10.2 (ESRI Japan).

27
28

2.3.2 Ecosystem services

29

The following proxy variables, which were categorized based on the MEA (2005) and The


30

Economics of Ecosystems and Biodiversity (TEEB 2010) methodologies, were selected for

6


ACCEPTED MANUSCRIPT
1

estimating potential supply of ecosystem services (Table 3) using the methods of Ooba et al.

2

(2014). For the purposes of this work, conservation of both habitat and biodiversity was

3

considered a supporting service.
Provisioning services: Annual economic output from the gross agricultural production (Ap,

5

JPY/y), including rice, other vegetables, and orchard tree fruits, was used as a proxy for the

6

provisioning services of an agricultural ecosystem. Data source Ag was divided into


7

municipal-level values according to agricultural areas in the municipalities (Aa) and converted

8

to a 5-km mesh. Precipitation that occurs in urban areas does not infiltrate into the soil.

9

Therefore, annual potential water resources (Wr, mm/y) in an area were estimated using annual

10

rainfall (Cr, mm/y) and the ratio of non-urban areas to total area. The non-urban area was

11

estimated from the index for degree of artificial disturbance given in Vg.

M
AN
U

SC

RI
PT

4


12

Regulation services: A simpler method was used in this study to estimate carbon

13

sequestration rate (Sc) according to ecosystem type (Vg) and climate condition (Ct). This

14

method has been outlined in the forest monitoring research (Hirata et al., 2008) and described in

15

detail in the Appendix.

Supporting services: Ecosystem continuity (Vc) was calculated using focal statistics in

17

ArcGIS (with radius set at 10 km) to assess fragmentation of vegetation, which disturbs

18

biological and ecosystem processes. A natural ecosystem was assumed to be an ecosystem

19

without urban and agricultural land uses, while biodiversity was assumed to be represented by


20

the number of species indicated in the data from a survey of domestic mammals (Sp).

EP

TE
D

16

Cultural services: Ecosystems provide cultural services for human psychological and

22

recreational activities. Indirect values (option value, the value of maintaining ecosystems for

23

future generations, and existence value) are also important services for humans, but they are not

24

easy to measure. Thus, herein, we used the value of recreation services from ecosystems as a

25

more direct value. The number of individuals visiting a natural ecosystem for sightseeing (Pd)


26

was used as a proxy of cultural services. The method used to estimate Pd from tourism locations

27

and statistics (Tr and Tp) is described in the Appendix.

AC
C

21

28

These proxy variables have units different from the physical units (Sc, Mg-C/(ha y)) to a

29

social unit (person-day/y). For the statistical and zoning analyses, proxies were converted to

30

relative values using maximum and minimum values, due to differences in units.

7


ACCEPTED MANUSCRIPT
1

2
3
4

2.4. Statistical analysis
The variables for ES and biomass were analyzed using principal component analysis (PCA)
to identify potential factors and to classify the distribution of these variables.
The principal components (PCs) represent potential factors, and PCs with high order have

6

relatively strong descriptive power in relation to the given dataset. Scatter plots representing the

7

scores of high-order PCs indicate variance of the multi-dimensional dataset in low (e.g., two)

8

dimensions. The distribution of points in the scatter plot was divided into sub-classes to clarify

9

key factors.

SC

RI
PT


5

10
2.5 Hotspot and conflict analysis

M
AN
U

11
12

The conservation planning software Zonation (Moilanen et al., 2012) provides several

13

algorithms to determine conservation priorities (e.g., core-area zonation) when calculating the

14

rank of potential supplies of ES and biomass energy. Version 4.0.0b26 was used in this study.
Zonation software can calculate conservation priority as value order according to an

16

evaluation function. For example, Moilanen et al. (2011) researched the competing land uses (in

17

terms of their biodiversity, carbon storage, agricultural production, and urban area) in Great


18

Britain using Zonation. It is likely that the application of Zonation for biomass supply may

19

reveal valuable areas from geospatial analysis.

TE
D

15

The authors chose to use the simplest algorithm, the additive benefit function, which

21

calculates the sum of all calculated values of ecosystem services for each mesh cell and

22

produces a mesh map of the sums. In this study, the input values were the absolute values of

23

biomass energy and the relative values of ecosystem services.

AC
C


EP

20

24

The additive benefit function was used with equal weights (1.0) for every proxy of the

25

ecosystem services. The priority was also calculated for biomass energy (absolute values) using

26

two methods for weighting—equal weights (as was used for ecosystem services) and weighting

27

biomass types—according to their corresponding prices (in 2015) in the FIT scheme for renewable

28

energy in Japan (Table 1). For each variable, the default value (1.0) was assumed to be the cost.

29
30

3. Results and discussion


8


ACCEPTED MANUSCRIPT
1

3.1 Geographical distribution of biomass energy and ecosystem services

2

Various maps of the geographical distribution of the biomass potential and ecosystem services

3

before 2010 are shown in Figs. 2 and 3. These maps show a 5-km mesh grid in the eastern part of

4

Japan.
The potentials of biomass energy are indicated as absolute values (TJ/(y 5-km mesh), Fig. 2).

6

High potentials were observed for mountainous and agricultural areas in cases of forest-origin

7

biomass (wf and we; Table 1, Nomenclature) and biomass originated from agricultural land (aa, ae,

8


and ma). The distributions of these two biomass types were relatively distinct, possibly owing to

9

land use patterns. Urban-origin biomass (ww and fw) was widely distributed and related to the

SC

10

RI
PT

5

population distribution (Fig. 1c).

Proxy variables for the potential supply of ecosystem services are indicated as relative values

12

that were transformed to fall within the range 0 to 1 for the minimum and maximum values,

13

respectively, due to the difficulty in assigning values to ecosystem services (Fig. 3). Provisioning

14


services classified as agricultural production (Ap) used proxies that were different from the

15

agricultural biomass (aa), because Ap represented the economic value of agricultural products from

16

agricultural fields (ecosystem) and had a weak relationship to the amount of agricultural products.

17

The highest carbon sequestration rates were indicated in mountainous areas, and this distribution

18

resembled that of supporting services (Sp and Vc). In the western region of eastern Japan, Wr was

19

relatively high. This was explained by the difference between the annual precipitation on the

20

coastline of the Sea of Japan and that of the Pacific Ocean. The Kanto Region, inhabited by

21

approximately 66% of the population of eastern Japan, and the areas around the concentrated cities


22

are generally used for agriculture and are partly mountainous. Kanto Region provides substantial

23

cultural services (Pd), which may reflect the accessibility of these sites from areas of high

24

population concentration.

TE
D

EP

AC
C

25

M
AN
U

11

26


3.2 Geospatial analysis

27

3.2.1 Biomass energy

28

The correlation matrix was calculated for biomass energy variables (data not shown). High

29

correlation coefficients were detected between the following biomass resources: wf and we (0.87),

30

aa and ae (0.99), and ww and fw (0.69).

9


ACCEPTED MANUSCRIPT
The cumulative contribution of these eight variables of biomass energy inferred from the PCA

2

was 71.7% for the three components. The loading vectors for the PCA analysis indicated that the

3


first PC was explained by the total power of all the variables, and the second and third PC were

4

related to the biomass type (forest-, agricultural land-, and urban-origin biomass). A scatter plot of

5

the scores for the second and third PC showed a clear distinction between categories on the three

6

axes (Fig. 4a, c), suggesting that biomass is strongly related to location and ecosystem. Therefore,

7

this analysis confirmed that land use strongly determines the available biomass types.

RI
PT

1

From the PCA results pertaining to biomass energy, regional supply potential of biomass energy

9

was classified into three types as follows: urban biomass (uw, ww, and sl), agricultural biomass (aa,

10


ae, and ma), and forest biomass (wf and we). Fig. 5a indicates the classification using three primary

11

colors and power (the first PC) by brightness. The map (Fig. 5a) clarifies land use in the eastern

12

part of Japan. Large sections of the southern part of the study area (Kanto Region) have core,

13

concentrated-population, and related agricultural areas. Regional-core areas with relatively high

14

population density were dispersed. Other (see also Fig. 1b) areas have relatively high agricultural

15

biomass potential, and mountainous areas have potential for woody biomass, but at a relatively low

16

power.

17
18


3.2.2 Ecosystem services

TE
D

M
AN
U

SC

8

A correlation matrix for the proxy variables of ecosystem services was also calculated (data not

20

shown). The correlations between specific pairs of proxies (Sc and Sp, Sc and Vc, and Vc and Wr)

21

that were related to the distribution of forest ecosystems were fairly high (0.55–0.64). For other

22

pairs of variables, the correlation coefficients were less than 0.43. According to the PCA of these

23

ES variables, the cumulative contributions were 76.0% for the three PCs. The first PC was


24

explained by all proxy variables except Ap. The second PC was strongly correlated to Ap and

25

secondarily to Wr, and the third PC represented the difference between cultural services (Pd) and

26

supporting services (Sc and Sp). Pd is related to the number of visitors to a natural ecosystem,

27

which may be affected by the population distribution; thus, Pd has the opposite scale of Sc and

28

Sp. A scatter plot of scores for the second and third PC revealed a continuous distribution that

29

represented various degrees and combinations of ecosystem services with respect to location (Fig.

30

4b, d).

AC

C

EP

19

10


ACCEPTED MANUSCRIPT
The distribution of ES type in the studied areas was not separated clearly compared to the type

2

of biomass energy. Using the results of the PCA, Fig. 5b indicates the relative values of the second

3

and third PC by two colors. The second component was linearly related to the absolute amount of

4

agricultural products (Ap). High positive value of the third PC means high cultural services from

5

the ecosystem (Pd), and low negative value means high regulation and supporting services (Sc and

6


Sp). In Fig. 5b, the areas intermediate between Kanto and Jouetsu have a high Pd potential (dark

7

area), which has a different meaning from the dark area of Kanto (low potential for both

8

agricultural and forest ES). Other areas show blue or green colors indicating high Ap or Sc.

9

However, the center of the Tohoku Region is cyan, which means high Ap and Sc.

SC

RI
PT

1

The relationship between the types of ecosystem services and the type of land use indicated in a

11

previous study (Ooba et al., 2014) at the municipality level was confirmed at the 5-km-mesh level

12

in this study.


13
14

3.3 Hotspot and conflict analysis

M
AN
U

10

The rank of priorities was calculated from the variables of biomass energy and ecosystem

16

services using the Zonation software (Moilanen et al., 2012). The distributions of the rank ranged

17

from 0 (low) to 1 (high) and they are indicated in Fig. 6.

TE
D

15

It is remarkable that the priority rank, Rbio, under the priority calculation with equal weight for

19


all biomass, was high in highly populated areas and it was followed by that in agricultural areas

20

(Fig. 6a), as already indicated by the PCA analysis. Regarding the transportation of biomass, the

21

most available biomass is the abundant biomass near the areas of concentrated population.

EP

18

The priority of ecosystem services (conservation priority), Reco, was high in rural and

23

mountainous areas (Fig. 6b), which was complementary with Rbio. The distribution of Reco was a

24

composite of high supply areas of the ecosystem services (i.e., the northern area of the Kanto

25

Region [Pd] and middle area of the Tohoku Region [Ap]). In general, Rbio and Reco were

26


complementary; however, this was not true for all regions. High-ranking areas in which both Rbio

27

and Reco were larger than 0.7 are indicated in the classified maps (Fig. 5). These indicate a potential

28

area of conflict with a high supply of ecosystem services and biomass supply potential. Potential

29

areas of conflict were distributed in the areas surrounding the Kanoto Region and on the border of

30

the prefectures of the Tohoku and Jouetsu regions. In these areas, intensive biomass production

AC
C

22

11


ACCEPTED MANUSCRIPT
1


may cause ecosystem degradation and a related decrease in ecosystem services.
After weighting the woody biomass according to the actual Japanese FIT condition, the rank

3

decreased in urban areas, especially in local urban areas (Fig. 6c). In these regions, Rbio was of

4

lower priority than before accounting for price. In areas surrounding the highly populated areas, Rbio

5

was relatively high due to the large amount of agricultural biomass (aa and ae) available. This was

6

especially true in regional urban areas excluding Kanto. The differences in ranks (Fig. 6d) indicated

7

that the maximum decrease of Rbio in urban areas reached 0.2. The trend of increasing rank was also

8

indicated in mountainous regions. In the central area of the Tohoku Region, Rbio indicated a higher

9

value than that under the priority calculation with equal weight. Parts of rural and mountainous


10

areas also had high Rbio. In the analysis weighted by FIT (Table 1), the woody residuals of

11

harvesting and thinning (wf) were assigned values that were two or three times higher than those of

12

the urban-origin biomass, waste wood, sludge, and food processing waste (ww, sl, and fw).

M
AN
U

SC

RI
PT

2

The current situation in Japan indicates that the cropland can be relatively easily switched to the

14

production of energy crop if the economic competitive conditions of biomass are better than those


15

of fossil or renewable energy. A serious conflict of agricultural biomass between energy and food

16

may not occur, because there is a trend to compensate the deficiency of food from domestic

17

agricultural sectors in Japan by imported food. This shift to energy crop may have less impact on

18

cropland and the neighboring ecosystems. However, extreme exploitation of forest biomass due to

19

the FIT incentive may cause conflict between biomass use and ecosystem because forest

20

plantations in Japan have not been well-managed, as the authors have repeatedly pointed out (Ooba

21

et al., 2012, 2014). Sudden harvesting from these unmanaged forests may cause serious problems

22


to cycling of the material inside the forest ecosystem, ultimately affecting forest ecosystem services

23

such as soil erosion, flood control, biodiversity, and so on.

AC
C

EP

TE
D

13

24

In the previous discussion on biomass use and design of the FIT in Japan, the total potential

25

amount of biomass in Japan, the conflicts between biomass and other renewable energy (solar

26

power, wind power, geothermal), the coordination of grid-based power, and the project cost and

27


profit analyses of renewable energy were central issues. In addition to these issues, it has been

28

suggested that geographical distribution and conflicts between local socioeconomic factors and the

29

ecosystem are important problems.

30

Figure 5 clearly reveals areas that are sensitive with respect to biomass production in the context

12


ACCEPTED MANUSCRIPT
1

of maintaining ecosystem services at a local scale. This study showed that the sensitivity of biomass

2

prices also affected the distribution of fragile areas (Fig. 5). Accordingly, the future planning for

3

promotion of biomass production, such as the FIT, must consider ecosystem impacts of biomass


4

use based on geographical analysis at a local scale.

6

RI
PT

5
3.4 Outlook

Sacchelli et al. (2014) suggested that a geographic visualization of biomass allocation is the

8

absence of a qualitative analysis between supply and demand. Improvements may be needed to

9

estimate biomass, especially woody and agricultural biomass in medium to long term. Forest and

10

agricultural management has large effects on biomass productivity. The NEDO biomass database

11

contains only residuals of harvested wood and agricultural food; however, biomass for energy use


12

is the main product of forest plantations and agricultural land. Detailed assessments of

13

environmental and ecological impacts on the use of biomass and other renewable energy sources

14

are also needed for more localized situations. Machado et al. (2013) developed a dynamic system

15

model for forest growth and carbon stock. Ooba et al. (2015) indicated long-term effects on forest

16

productivity of the forest management scenario using an ecological process-based model. Schmidt

17

et al. (2015) developed a land-use model that is generally applicable to all land-use types and

18

indicated the dynamics of land use caused by economic reasons. This may be helpful for

19


considering long-term temporal analysis by these modeling approaches.

TE
D

M
AN
U

SC

7

In addition, an understanding of the demands for biomass energy and ecosystem services is also

21

important to match the supply from ecosystems. Biomass energy and heat generated from biomass

22

boiler has a disadvantage with respect to its transportation requirements, while electric power can

23

easily be conducted to sites demonstrating demand. Spatial planning of biomass transport and the

24

location of electric power generators can rely on a simultaneous evaluation of supply and demand.


25

Sacchelli et al. (2014) and Delivand et al. (2015) studied optimal locations for bioenergy facilities

26

by using the multi-criteria analysis with geographical information. The spatial ecosystem impact

27

assessment indicated in this study is useful for the multi-aspect decision making.

AC
C

EP

20

28

It may effective that regional exploitation of biomass production was zoning according to the

29

results of geographical ecological impact assessment, as indicated in Fig. 5. The zoning already

30


exists for agricultural and forestry land use in Japan. For example, the Agency of Forestry installed

13


ACCEPTED MANUSCRIPT
a five-zoning system in Japan. While the next study needs more detailed analysis from both

2

economic and ecological aspects, rough mapping of conflicting areas between biomass production

3

and ecosystem services may be useful. An accurate and comprehensive assessment of the supply of

4

ecosystem services provides policy recommendations about a more eco-healthy biomass

5

production.

RI
PT

1

6

7

4. Conclusions

This study determined the spatial distributions of the potential supply of ecosystem services

9

and biomass energy within the eastern region of Japan using 14 variables estimated from

10

SC

8

various data sources and the NEDO biomass database on a 5-km-mesh scale.

PCA analysis was conducted for three biomass types related to forests, agricultural land, and

12

urban areas, clearly (Fig. 4c). From the same analysis, the ecosystem services were classified

13

continuously by two axis (agricultural supporting services and other services).

M
AN

U

11

Considering the ranks, the potential distribution of biomass and ecosystem services were

15

complementary. However, both ranks were high in the area surrounding the Kanto Region and

16

the middle of the Tohoku Region, and intensive biomass production in these areas affected the

17

regional ecosystems that provide ecosystem services with high conservation priority. Future

18

planning for promoting biomass production such as FIT needs should be based on a geographical

19

analysis. Considering the ecosystem services, zoning exploitation and visualization of

20

conflicting areas may be useful.


EP

TE
D

14

21

Acknowledgments

23

This study was supported by the Environment Research and Technology Development Fund

24

(2-1404, MOE, Japan) and the joint research program of the EcoTopia Science Institute, Nagoya

25

University. We also thank Y. Hasegawa, H. Sumi, and T. Suzuki for helping with the data

26

processing for the study.

AC
C


22

27
28

Nomenclature
Aa

Area of agricultural land use

aa

Agricultural residual

14


Ag

Agricultural production (prefecture level)

Ap

Agricultural production

Ba

Agricultural land area

Cp


Annual precipitation

Ct

Mean annual air temperature

FIT

Feed-in tariff scheme for renewable energy

fw

Food-processing waste

ma

Manure slurry

PCA

Principal component analysis

Pd

Tourists in natural ecosystems

Po

Population number


Sc

Carbon sequestration rate

sl

Sludge

Sp

Occurrence of mammals

Vc

Index of continuity of natural ecosystems

Vg

Vegetation map

we

Wood residual from other ecosystems

wf

Wood residual from forest

Wr


Effective precipitation

ww

Wood waste

EP

TE
D

M
AN
U

SC

Grassland residual

AC
C

1

ae

RI
PT


ACCEPTED MANUSCRIPT

2

Appendix

3

Carbon sequestration rate (Sc): Gross primary production (GPP) and ecosystem respiration

4

(RE) of a forest ecosystem change with annual mean temperature (Ta, °C) as indicated by

5

meta-analysis of flux measurement observations (Hirata et al. 2008). GPP and RE are calculated

6

using the following equations:

7
8

GPP = a Ta + b,

RE = RE0 exp(cT)

(A1),


9

15


ACCEPTED MANUSCRIPT
1

where a, b, c, and RE0 are constants (0.97 Mg-C/(ha y °C), 8.4 Mg-C/(ha y), 207.8 °C, 14.47

2

Mg-C/(ha y)). T is defined as follows:

3
4

T = 1/(Tk + Tref − T0) − 1/(Tk + Ta − T0)

(A2),

6

RI
PT

5
where Tk, Tref, and T0 are constants (273.15 °C, 10 °C, 227.13 °C). For forest ecosystems,


7
8

Sc = GPP − RE

(A3).

SC

9

Agricultural and other ecosystems and urban areas were assumed to provide no carbon

11

sequestration (Sc = 0). Finally, the spatial average, Sc (Mg-C/(ha y)), was calculated for each

12

municipality.

M
AN
U

10

13

Number of visitors to a natural ecosystem for sightseeing (Pd): The number of tourism spots


15

(for natural ecosystems only), n, was determined from data of tourism spots (Tp) that received

16

cultural services from the natural ecosystem. The rate of n to total tourism spots (for both

17

natural and urban spots) for each prefecture was obtained (r) and ranged from 0.11 (Chiba

18

Prefecture) to 0.84 (Iwate Prefecture).

19

Domestic tourism statistics (Td) included the annual number of tourists, the purpose of visit

20

(sightseeing or not), and destination (at prefecture level). The total stay (in days) per year (PD1)

21

in a given prefecture for sightseeing purposes was estimated from the sum of the values of

22


domestic one-day trips (d1) and domestic overnight trips (d2),

24
25

EP

AC
C

23

TE
D

14

PD1 = d1 + m d2

(A4),

26

where m is the mean length of an overnight domestic trip (= 2.3, according to tourism statistics

27

from 2010). The total number of days of stay per year (PD2) for sightseeing within a natural


28

ecosystem inside a prefecture were estimated from PD1 and r.

29
30

PD2 = r PD1

(A5)

16


ACCEPTED MANUSCRIPT
1
Finally, a number of person-days (Pd, person-days/y) for sightseeing a natural ecosystem at the

3

municipality level was estimated from PD2 and the ratio of the area of natural ecosystems in the

4

municipality (Amun) to the area of natural ecosystems in the prefecture(Apref), which was

5

calculated from Vg,


RI
PT

2

6
7

Pd = PD2 (Amun/Apref).

(A6).

8

11
12

SC

10

References

Biodiversity Canter of Japan, 2014. Japan Integrated Biodiversity Information System.
Available at: . (In Japanese with English summary)

M
AN
U


9

Cao, V., Margni, M., Favis, B, D., Deschênes,L., 2015. Aggregated indicator to assess land use

13

impacts in life cycle assessment (LCA) based on the economic value of ecosystem services.

14

J. Clean. Prod. 94, 56-66.

15

Delivand, M, K., Rita, A., Cammerino, B., Garofalo, P., Monteleone, M., 2015. Optimal
locations of bioenergy facilities, biomass spatial availability, logistics costs and GHG

17

(greenhouse gas) emissions: a case study on electricity productions in South Italy. J. Clean.

18

Prod. 99, 129-139.

21
22
23
24
25


resource-focused approach. Biomass and Bioenergy 30(1), 1-15.

EP

20

Ericsson, K., Nilsson, L., 2006. Assessment of the potential biomass supply in Europe using a

Hanafiah, M, M., Hendriks, A, J., Huijbregts, M, A, J., 2012. Comparing the ecological
footprint with the biodiversity footprint of products. J. Clean. Prod. 37, 107-114.

AC
C

19

TE
D

16

Henry, R, J., 2010. Evaluation of plant biomass resources available for replacement of fossil oil.
Plant Biotech. J. 8, 288–293.

van der Hilst, F., Verstegen, J. A., Karssenberg, D., Faaij, A. P. C., 2012. Spatiotemporal land

26

use modelling to assess land availability for energy crops ‐ illustrated for Mozambique.


27

Glob. Change Biol. Bioenergy, 4, 859‐874.

28

Hirata, R., Saigusa, N., Yamamoto, S., Ohtani, Y., Ide, R., Asanuma, J., Gamo, M., Hirano, T.,

29

Kondo, H., Kosugi, Y., Li, S, G., Nakai, Y., Takagi, K., Tani, M., Wang, H., 2008. Spatial

30

Distribution of Carbon Balance in Forest Ecosystems Across East Asia. Agric. Forest

17


ACCEPTED MANUSCRIPT
1
2

Meteorol. 148, 761-775.
Hoogwijk, M., Faaij, A., Broek, R., Berndes, G., Gielen, D., Turkenburg, W., 2003. Exploration

3

of the ranges of the global potential of biomass for energy. Biomass Bioenergy, 25,


4

119-133.
Hoogwijk, M., Faaij, A., Eickhout, B., de Vries, B., Turkenburg, W., 2005. Potential of biomass

6

energy out to 2100, for four IPCC SRES land-use scenarios, Biomass Bioenergy, 29,

7

225-257.

RI
PT

5

Hoogwijk, M., Faaij, A., de Vries, B., Turkenburg, W., 2009. Exploration of regional and global

9

cost‐supply curves of biomass energy from short‐rotation crops at abandoned cropland

10

and rest land under four IPCC SRES land‐use scenarios. Biomass Bioenergy, 33, 26-43.
Iuchi, M., 2004. Development of system that supporting use planning of biomass energy:


M
AN
U

11

SC

8

12

modeling of storage database and cost of collecting. Central Research Institute of Electric

13

Power Industry, Research Report Y03023, 1-26. (In Japanese with English Summary)

14

Japan Tourism Agency, 2014. Consumption Trend Survey for Tourism. Available at:
(in Japanese with

16

English summary)

17

TE

D

15

Kaji, K., Tanaka, K., Nanno, M., Miyamura, Y., Shibata, K., Zhang, J., Miyata, H., 2013.

18

Specification Design of Renewable Energy Management System for Recovery Planning of

19

Japanese Coastal Community After Tsunami Disaster. Springer, pp. 51-62.
Machado, R, R., Conceição, S, V., Leite, H, G., Souza, A, L., Wolff, E., 2013. Evaluation of

EP

20

forest growth and carbon stock in forestry projects by system dynamics. J. Clean. Prod. 96,

22

520-530.

23
24
25
26
27

28

AC
C

21

Ministry of Land, Infrastructure, Transport and Tourism, 2014. National Land Numerical
Information download service. Available at: (In

Japanese)

Ministry of Internal Affairs and Communications, 2014. Portal Site of Official Statistics of
Japan. Available at: (In Japanese)
Ministry of Economy, Trade and Industry of Japan, 2014. Cabinet Decision on the New

29

Strategic Energy Plan. Available at:

30

(In Japanese)

18


ACCEPTED MANUSCRIPT
1
2

3

Millennium Ecosystem Assessment, 2005. Ecosystems and Human Well Being: Synthesis.
Island Press, Washington, 137 pp.
Moilanen, A., Meller, L., Leppanen, J., Pouzols, F.M., Arponen, A., Kujala, A., 2012. Zonation
version 3.1 - user manual. Biodiversity Conservation Informatics Group, Department of

5

Biosciences, University of Helsinki, Finland. Available at:

6

/>
7

RI
PT

4

Moilanen, A., Anderson, B. J., Eigenbrod, F., Heinemeyer, A., Roy, D. B., Gillings, S.,

Armsworth, P. R., Gaston, K. J., Thomas, C. D., 2011. Balancing alternative land uses in

9

conservation prioritization. Ecol. Appl., 21, 1419-1426.

10


SC

8

Myllyviita, T., Holma, A., Antikainen, R., Lähtinen, K., Leskinen,P., 2012. Assessing
environmental impacts of biomass production chains - application of life cycle assessment

12

(LCA) and multi-criteria decision analysis (MCDA). J. Clean. Prod. 29-30, 238-245.

13

M
AN
U

11

Naik, S. N., Goud, V. V., Rout, P. K., & Dalai, A. K. (2010). Production of first and second

14

generation biofuels: A comprehensive review. Renewable & Sustainable Energy Reviews,

15

14(2), 578‐597.


17
18

NEDO, Estimation about Potential and Available Biomass. Available at:

TE
D

16

(In Japanese)
Ooba, M., Hayashi, K., Fujii, M., Fujita, T., Machimura, T., Matsui, T., 2015. A long-term
assessment of ecological-economic sustainability of woody biomass production in Japan. J.

20

Clean. Prod., 88, 318-325.

EP

19

Ooba, M., Fujita, T., Togawa, T, Hirano, Y., Fujii, M., Hayashi K., 2014. Geospatial distribution

22

of ecosystem services and renewable energy potential within eastern Japan. Proceedings of

23
24

25

AC
C

21

the 9th Conference on Sustainable Development of Energy, Water and Environment
Systems, SDEWES2014.0098, 1-13.

Ooba, M., Fujita, T., Mizuochi, M., Fuji, M., Machimura, T., Matsui, T., 2012. Sustainable Use

26

of Regional Wood Biomass in Kushida River Basin. Waste Biomass Valorization 3,

27

425-433.

28

Sacchelli, S., Bernetti, I., Meo, I., Fiori, L., Paletto, A., Zambelli, P., Ciolli, M., 2014. Matching

29

socio-economic and environmental efficiency of wood-residues energy chain: a partial

30


equilibrium model for a case study in Alpine area. J. Clean. Prod. 66, 431-442.

19


ACCEPTED MANUSCRIPT

2
3
4
5

Schmidt, J, H., Weidema, B, P., Brandão, M., 2015. A framework for modelling indirect land
use changes in Life Cycle Assessment. J. Clean. Prod. 99, 230-238.
Statistics Bureau, 2014. Population Census of Japan. Available at:
/>TEEB, The Economics of Ecosystems and Biodiversity, 2014. Mainstreaming the Economics of

RI
PT

1

Nature: A Synthesis of the Approach, Conclusions and Recommendations of TEEB.

7

Available at: />
AC
C


EP

TE
D

M
AN
U

SC

6

20


ACCEPTED MANUSCRIPT

Table 1 Biomass data set
Parameters for biomass use

Type and detail of biomass

Weight b

Wood residual from forest
wf

Combustion, Lower Heating


(Harvest and thinning of plantation

Value(LHV): 18.1 GJ/t

0.202

we

RI
PT

forest)
Wood residual from other ecosystems

Combustion, LHV: 11.5-12.5

(Orchard forest, and bamboo)

GJ/t

Agricultural residual

Combustion, LHV: 13.6-14.2

0.121

GJ/t (for crop straw), Methane

(Rice husk and straw, and wheat husk )


fermentation a: VS/TS=0.75,
3

SC

aa

0.121

GR=400 m /t (for other

Grassland residual
ae

(Bamboo grass and Japanese silver
grass)
Wood waste

fw

a

construction waste) and 11.5 GJ/t

demolition debris, construction debris,

(for pruned residual from park

and pruning branch from public parks)


forest)

Livestock manure

Methane fermentation a

(Dairy cattle, beef cattle, swine, layer

VS/TS=0.8-0.83, VS=0.4,
GR=500-650 m /t

Sludge

Methane fermentation a

(Two types of sewage sludge, and

VS/TS=0.75-0.77, VS=0.46-0.52,

human waste sludge)

GR=620-780

0.066

0.197

Food-processing waste

Methane fermentation a


0.086

m3/t

VS/TS=0.2, VS=0.80, GR=500

(Waste from food-processing factory,

(for food processing waste),

kitchen, food vendor waste)

VS/TS=0.84, VS=0.84, GR=808
m3/t

0.086

(for other)

Lower Heating Value (LHV) for methane: 0.036 GJ/m3, VS/TS: Ratio of volatile solid to total

solid, VS: decomposed rate of volatile solid, GR: gas production rate.
b

0.121

3

chicken, and broiler chicken)


AC
C

sl

(Residual of lumber sawing,

TE
D

ma

Combustion, LHV: 13.6 GJ/t

Combustion, LHV: 18.1 GJ/t (for

EP

ww

M
AN
U

agricultural residual)

FIT weight, see sections 2.5 and 3.3, and Fig. 6 for the zoning analysis.

21



ACCEPTED MANUSCRIPT

Table 2 Data sources for assessment of biomass energy and ecosystem services
Name

Cp,
Ct

Animal distribution
survey map a
Climate map b
World census of

Aa

agriculture and forestry in

Plant community,
Degree of disturbance
Occurrence of mammals
Annual precipitation and
mean air temperature
Area of agricultural land
use

Japan c

1 km mesh


Municipality
level table

Gross agricultural

Prefecture

production and income c

production

level table

Population number and

Municipality

density

level table

Po

Population census d

Tr

Map of tourism spots d


Tp

5 km mesh

Statistics of agricultural

Consumption trend survey
for tourism e

M
AN
U

Ag

1/50,000 map

Point map

Number of people and

Municipality

days for domestic tourism

level table

Biodiversity Canter of Japan, 2014

b


Ministry of Land, Infrastructure, Transport and Tourism, 2014

c

Ministry of Internal Affairs and Communications, 2014

d

Statistics Bureau, 2014

e

Japan Tourism Agency, 2014

AC
C

EP

TE
D

a

22

Year
1979–
1998

2000–

RI
PT

Sp

Vegetation survey map a

Data Type

SC

Vg

Details

2004
1982

2010

2010
2010
1999
2010


ACCEPTED MANUSCRIPT


Table 3 Proxy variables for ecosystem services
Category

Unit

Sources

Wr

Effective precipitation

mm/y

Vg, Cp

Ap

Economic gross agricultural production

JPY/y

Ag, Aa

Regulation

Sc

Carbon sequestration rate

Mg-C/(ha y)


Vg, Ct

Supporting

Vc

Index of continuity of natural ecosystem

Sp

Species number

Pd

Tourists in natural ecosystems

Vg
Sp

person-day/y

AC
C

EP

TE
D


M
AN
U

SC

Cultural

RI
PT

Provisioning

Proxy variable

23

Vg, Tr, Tp


ACCEPTED MANUSCRIPT

Figure Captions

RI
PT

Figure 1 Areas studied (Tohoku, Kanto, and Jouetsu regions). (a) Location of study area,
(b) Altitude distribution (m), (c) Population density distribution (people/km2).


Figure 2 Potential supply distribution of biomass energy (TJ/y in 5-km grid squares)
within eastern Japan: (a) wf: Wood residual from forest, (b) we: Wood residual from other

SC

ecosystems, (c) aa: Agricultural residual, (d) ae: Grassland residual, (e) ww: Wood waste,
(f) ma: Livestock manure, (g) sl: Sludge, (h) fw: Food-processing waste.
Figure 3 Potential supply distribution of ecosystem services (relative values): (a) Wr:

M
AN
U

Effective precipitation, (b) Ap: Economic gross agricultural production, (c) Sc: Carbon
sequestration rate, (d) Vc: Index of continuity of natural ecosystem, (e) Sp: Species
number, (f) Pd: Tourists in natural ecosystems.

Figure 4 Loading vectors and scatter plots of the principle component scores (2nd and 3rd
principle components) from the principle component analysis (PCA) for 8 variables of

respectively.

TE
D

biomass energy, (a) and (c), and 6 variables of ecosystem services, (b) and (d),

Figure 5 Classification for (a) biomass energy and (b) ecosystem services. (a) shows the
classification of urban (red), agricultural (green), and forest biomass (blue). Brightness


EP

indicates the relative amount of the potential supply (1st principle component, PC, see
text). (b) shows the relative values of the second and third PCs in green (related to

AC
C

provisioning services) and blue (related to other services). Grid cells in (a) white and (b)
red represented high-ranking areas for both biomass and ecosystem services.
Figure 6 Relative ranking of priority areas for biomass energy and ecosystem services. (a) Rank of
biomass energy with equal weighting, (b) Rank of ecosystem services, (c) Rank of biomass
energy with the FIT weighting (see Table 1), and (d) Difference between (a) and (c).