land cover change on ecosystem services in the central highlands of Ethiopia

Ecosystem Services 23 (2017) 47–54

Contents lists available at ScienceDirect

Ecosystem Services journal homepage: www.elsevier.com/locate/ecoser

The impact of land use/land cover change on ecosystem services in the central highlands of Ethiopia

crossmark

⁎

Terefe Tolessaa,b, , Feyera Senbetaa, Moges Kidaneb a b

Center for Environment and Development, College of Development Studies, Addis Ababa University, Ethiopia Department of Natural Resource Management, College of Agriculture and Veterinary Science, Ambo University, Ethiopia

A R T I C L E I N F O

A BS T RAC T

Keywords: Ecosystem service Landscape change Payment for ecosystem services Land use/land cover Chillimo Livelihood

Ecosystems provide a wide range of services that are important for human-well being. Estimating the multiple services obtained from ecosystems is vital to support decision-making processes at different levels. This study analyzes land use/land cover (LU/LC) dynamics over four decades (i.e., 1973, 1986, 2001, 2015) to assess its impact on ecosystem services. Ecosystem Service Values (ESV) was determined using LU/LC analysis and established global data base. LU/LC analysis showed that forest cover reduced by 54.2% during study period; and settlement, bare land, shrub land and cultivated land increased considerably. The study indicates that due to forest cover change from 1973 to 2015, approximately US$ 3.69 million of ecosystem services values was lost. Among the ecosystem services reduced were: nutrient cycling, provision of raw material and erosion control. The use of LU/LC data along with established global ESV data sets reduce the costs of ground data collection, and help in tracking of past environmental changes and acquisition of quick and reliable results that can be used for decision making processes. We believe that the results obtained can be helpful in designing payment for environmental services and rural development policies.

1. Introduction Ecosystems provide a wide range of multiple services that vary in quantity and quality depending on the type of ecosystems and their status (MA, 2005). For example, grass land was found to be quite different in service provision compared to tropical forests (Costanza et al., 1997, 2014; de Groot et al., 2012), but each one of them provides a unique service that cannot be replaced by others. Certain services are local specific (pollination of agricultural crops) and others are global in their nature (mitigation of global climatic change). Many of these services are important for sustaining life on earth and maintaining the integrity of the ecosystem. These services are, nevertheless, currently under great pressure due to anthropogenic activities and climate change. Among the human activities that reduce ecosystem services include land use/land cover (LU/LC) change in a given area driven by agricultural activities, settlements, built up areas and mining (Li et al., 2007; de Groot et al., 2010; Haines-Young et al., 2012; Kindu et al., 2016). The impacts of LU/LC change on ecosystem services vary across space and time (Costanza et al., 1997, 2014; de Marko and Coelho, 2004; Hu et al., 2008; de Groot et al., 2012; HainesYoung et al., 2012; Bryan, 2013). Expressing ecosystem services in monetary values is becoming a

⁎

common practice to create awareness among users, provide evidence for decision/policy makers, help to know the opportunity costs of restoration and assist in payments for ecosystem service (Costanza et al., 1997; de Marko and Coelho, 2004; Nelson et al., 2009; de Groot et al., 2012; Alarcon et al., 2016). Since the completion of the Millennium Ecosystem Assessment (MA) (2005), research revealed the possibilities to quantify ecosystem services (Nelson et al., 2009; de Bello et al., 2010; de Groot et al., 2012; Ango et al., 2014; Kindu et al., 2016). However, quantification of ecosystem services has been a concern because of the debates surrounding the methodologies used, the type of service measured and the results obtained. In addition, ecosystem service valuation was limited to specific service and measurements are not comprehensive across the World (Costanza et al., 1997, 2014; Nelson et al., 2009; de Groot et al., 2012; Summers et al., 2012; Satz et al., 2013; Kindu et al., 2016). Despite these limitations attempts to estimate ecosystem service values are undertaking and improve our knowledge, experience and skills to refine the drawbacks. For the past two decades much effort has been made to come up with encouraging results, although much is needed for the future to incorporate wide ecological regions and services (de Bello et al., 2010; Satz et al., 2013; Tadesse et al., 2014b). The highlands of Ethiopia ( > 1500 m above sea level (masl))

Corresponding author at: Center for Environment and Development, College of Development Studies, Addis Ababa University, Ethiopia, P.O. Box 1176, Addis Ababa, Ethiopia. E-mail address: [email protected] (T. Tolessa).

http://dx.doi.org/10.1016/j.ecoser.2016.11.010 Received 12 June 2016; Received in revised form 7 November 2016; Accepted 22 November 2016 2212-0416/ © 2016 Elsevier B.V. All rights reserved.

Ecosystem Services 23 (2017) 47–54

T. Tolessa et al.

harbors some 88% of the population of the country, over 95% of the total cultivated land, about 75% of livestock population with an estimated area coverage of 44% of the land mass of the country (Hurni et al., 2005; Slaymaker, 2010). These highlands have been under intense pressure because of favorable climatic conditions for agriculture, human and animal health in comparison to the lowlands, leading to environmental resource degradation (Feoli et al., 2002; Nyssen et al., 2004; Lemenih and Teketay, 2005; Kidane et al., 2012). This environmental resource degradation causes the reduction in ecosystem services (Feoli et al., 2002; Slaymaker, 2010; Kindu et al., 2016). Our study forest, Chillimo, lies within the highlands of Ethiopia. The forest represents a vital ecological space for birds, mammal species, and water supply. It is the source of several large rivers, including Awash River, which is the major important water-way for irrigation of sugar plantations in the Rift Valley. In Ethiopia, LU/LC changes are pervasive and common phenomenon where agricultural activities and settlements dominate rural landscapes affecting ecosystem services. Combining LU/LC and ecosystem service valuation data can help identify the area most vulnerable to changes in ecosystem services at landscape level and provide an entry point for land management opportunities in the future. Furthermore, studies conducted on LU/LC changes in Ethiopia focus on the dynamics of cover changes and their causes (Reid et al., 2000; Tsegaye et al., 2010; Meshesha et al., 2014) with little attention to address the impacts of such changes on ecosystem services aspect (Kindu et al., 2016; Tolessa et al., 2016). In this study we hypothesize that LU/LC, especially forest cover change, is the reason for reducing ecosystem services valuation (ESV). Thus, we first assessed LU/LC dynamics in central highlands of Ethiopia and used global data base developed for different LU/LC as an input for ESV of different land use types to estimate the amount of services gained/ lost due to land cover changes over spatial and temporal scales in the study landscapes. We also carried out sensitivity analysis to explore the robustness of our results by 50% adjustment of value coefficients.

Fig. 1. Map of the study area.

about 12 FUGs who reside in and around the forest. On the base of the agreement set with the government, FUGs are currently extracting different forest products such firewood collection for subsistence and sell to the nearby town, harvesting of poles for sell and charcoal making. On behalf of the government, the Oromia Forest and Wildlife Enterprise is mandated to oversee and share the benefits that arise from forest product harvests. The enterprise regularly monitors forest condition in collaboration with the representatives of FUGs, including income distribution from forest income, as well as conduct training and renewing/giving licences.

2. Study area and methods 2.1. Study area

2.2. Satellite data pre-processing and land use classification The study was conducted in Chillimo Forest; which is located in Dendi district of Oromia National Regional State (Fig. 1). Its geographical location is 38°10’E and 9°05’N; with an altitudinal range between 2400 and 2900 m.a.s.l. (Tamrat, 1993). The forest is a typical dry Afromontane forest vegetation of the country with Juniperus procera, the most abundant and dominant tree species. The largest diameter classes dominate most of the forest structure with low regeneration capacity due to long years of exploitation and open access for livestock grazing (Ameha et al., 2014). Chillimo forest has been harvested for commercial timber production through selective cutting of matured trees for long period of time. The main rock type in the area is basalt, and some areas are covered with other volcanic rocks of more recent formation (Tesfaye et al., 2015). The soils are reddish brown, gravelly and shallow at higher altitudes, while at lower sites they tend to become dark-grey and deep. The soils in the surrounding low plains are vertisols, black soils with characteristic high clay content. The mean annual rainfall is 1264 mm, with a bimodal rainfall distribution of lower precipitation from November to January and there are five rainy months, May-September, with a peak in July. The mean annual temperature ranges between 15 and 20 °C (Tamrat, 1993; Mesfin, 1998). Since 1996 the forest has been under a Participatory Forest Management (PFM) scheme to improve the empowerment the local community concerning management and use of the forest. After an agreement was made between community and government, the forest was handed over to forest user groups (FUGs). Currently there are

In this study, time series data of LU/LC were produced from multispectral Landsat imagery (Land sat MSS, TM, ETM+ and Land OLS), which were acquired on four separate dates: 1973, 1986, 2001, and 2015 (Table 1). All of the raw images were taken in the same season and nearly free of cloud since they were taken during the dry season. Prior to interpretation, atmospheric correction and geometrical rectification were performed including the resembling of a 1973 satellite image to match the pixel resolution. The dates selected for processing of LU/LC was mainly dependent on the availability of images, important dates in the change of government and policies related to land and land related resources. Remote sensing image data were preprocessed and processed using ERDAS imagine.10 software by applying the basic image preprocessing techniques starting from image rectification, restoration, enhancement, image classification and accuracy assessment. To assist the supervised image classification, ground control points (GCPs) were collected from Table 1 Description of imagery data used for land cover change study in Chillimo forest.

48

Imagery date

Imagery type

Resolution

Path and raw

Source

01/31/1973 02/07/1986 01/31/2001 01/26/2015

Landsat Landsat Landsat Landsat

57 30 30 30

181/54 169/54 169/54 169/54

USGS USGS USGS USGS

MSS TM ETM+ OLS

* * * *

57 m 30 m 30 m 30 m

Ecosystem Services 23 (2017) 47–54

T. Tolessa et al.

in spatial extent. Bare land, as one kind of land use, was approximated with urban land use, because in our case this bare land consists of roads, rocky outcrops and degraded lands, where the area is dominated by regular movement of trucks, quarry for road construction and abandoned lands as a result of gully formation. Similarly, shrub lands are not identical with forests, where shrub land lack the desired level of canopy cover to be equivalent with forests, however shrub land provide ecosystem services relatively similar to forests described in Costanza et al. (1997). In addition, tropical forests are used as proxies for forest land in our case, where the forest is regarded as dry afromontane forests. It is one of the last biodiversity resorts found on the mountainous escarpments. Lastly, cultivated land was estimated as crop land; as the purpose of which is used for crop production by smallholder farmers, although cultivated land use is relatively different in terms of size, soil fertility, use of input and other important variables. The total value of ecosystem service in the study landscape for 1973, 1986, 2001 and 2015 were obtained following the methodology used by Li et al. (2007) and Hu et al. (2008):

Table 2 Descriptions of land cover types in the study area. Land cover

Description

Settlement/Built up areas Cultivated land Shrub Land

Land dominated with houses and huts

Bare Land Forest Land

Land under cultivation Land with > 20% bush or shrub cover with < 20% tree cover ( < 5 m in height). Non vegetated area dominated by rock out crops, roads, eroded and degraded lands Land dominated by trees with greater than 80% canopy cover

each land use as a classification training site and image space signature separability was computed. The satellite images were classified in to five land uses (settlement, cultivated land, bare land, shrub land and forest land (Table 2)) using maximum likelihood classification algorithm. GCPs were also collected using a GPS device to perform classification accuracy assessment for each land use class for the 2015 classification year. A total of 150 GCPs were collected for all land use/land cover classes. The number of GCPs for each class was determined by area proportion of the land uses. The overall producer's accuracy of LU/LC maps over the study period was 82.1%, overall user's accuracy of LU/LC maps over the study period was 87.9% and the overall Kappa statistics was 0.84. This met the recommended value suggested by Janssen et al. (1994). Thus, the data was analyzed for further estimation of ESV for different LU/LC types.

ESV= ∑ (AK XVCk )

Where ESV is the estimated ecosystem service value, AK is the area (ha) and VCk the value coefficient (US $ ha−1yr −1) for LU/LC category k (Table 3). The change in ecosystem service value was estimated by calculating the differences between the estimated values for each LU/ LC category in 1973, 1986, 2001 and 2015. In addition to estimating LU/LC change effects on the total value of ecosystem services, we also estimated the impacts of such changes on 17 individual ecosystem services in the study landscape (TEEB, 2010). The values of services provided by individual ecosystem were calculated using the following equation (Hu et al., 2008):

2.3. Measurement of landscape ecosystem services

ESVf = ∑ (AK XVCfk )

LU/LC data was analyzed using ArcGIS and ESV for different biomes (de Groot et al., 2012) following the methods used by Li et al. (2007) and Hu et al. (2008). The ecosystem service valuation model for 16 biomes was used for the five LU/LC categories to determine ESV (Costanza et al., 1997). Although the value coefficient proposed by Costanza et al. (1997) was criticized because of uncertainties (Nelson et al., 2009), due to the lack of local level data we resorted to use it as an alternative way to estimate ESV, but to reduce the uncertainties we used sensitivity analysis as discusses below. The most representative biome was used as a proxy for each LU/LC category including: (1) cropland for cultivated land, (2) forests for Shrub land (3) tropical forest for forested area (4) Urban for settlement (5) Urban for Bare land (Table 3). The biomes used as proxies for the LU/LC categories were not perfect matches. Specifically, settlement differed from Costanza et al. (1997) biomes of urban land use. The ecosystem services of rural settlements are by far better in supplying organic waste materials to the nearby gardens and are not impervious like urban landscapes, which tend to create massive erosion. In terms of heat generation, urban land use generates large amounts of heat undesirable for both health of people and the environment. But both serve as shelter for human beings occupying land for other alternative uses, although the two vary

Equivalent biome

Cultivated Land Shrub land Forest land Settlement Bare land

Crop land Forests Tropical forests Urban Urban

(2)

Where ESVf is the estimated ecosystem service value of function f, Ak is the area (ha) and VCfk the value coefficient of function f (US $ha−1yr −1) for LU/LC category k. The value coefficients can be obtained from Costanza et al. (1997). Taking in to account that uncertainties in the value coefficients used from Costanza et al. (1997) and the biomes used as proxies for LU/LC types are not perfect matches, sensitivity analyses were conducted to determine the percentage change in ESVs for a given value coefficient following standard economic method. The value coefficients for cultivated land, shrub land and forest land were adjusted by 50% and the corresponding coefficient of sensitivity (CS) was calculated using Eq. (3) (Li et al., 2007; Hu et al., 2008). This method helped us to determine the robustness and reasonability of our estimation of ESV (Li et al., 2007; Hu et al., 2008).

CS=

(ESVj − ESVi)/ ESVi (VCjk − VC ik)/ VC ik

(3)

Where CS is Coefficient of Sensitivity, ESVi and ESVj are initial and adjusted total estimated ecosystem service values respectively, and VCik and VCjk = initial and adjusted value coefficients (US $ ha−1yr −1) for LU/LC type ‘k’.

Table 3 Biome equivalents for land use categories, and corresponding ecosystem service coefficients. (Source: Costanza et al., 1997) Land use and land cover categories

(1)

3. Results 3.1. Analysis of land use/land cover dynamics

Ecosystem service Coefficient

Land cover classification identified from our analysis across the study period indicated the conversion of forests to the rest of other land uses (Table 4, Fig. 2). The dominant land uses that increased progressively over the study period were settlement, bare land and shrub land. For example, settlement increased by 271.1% from 1973 to 1986 and by 967.4%, from 1986 to 2001. The overall increment of settlement was 6273.9%. The higher percentage increase for settlement

(US $ ha−1yr −1) 92 969 2007 0 0

49

Ecosystem Services 23 (2017) 47–54

T. Tolessa et al.

Table 4 Land use/land cover changes of the study landscape, 1973–2015.. LULC Class

Settlement Cultivated land Shrub Land Bare Land Forest Land

Absolute area coverage (ha)

Cover change between periods (%)

1973

1986

2001

2015

1973–1986

1986–2001

2001–2015

1973–2015

5.36 3202.78 216 0 4263.12 7687.26

19.89 3348.9 858.33 212.94 3247.2 7687.26

212.31 3512.07 1124.19 358.65 2480.04 7687.26

341.64 3493.17 1161.36 739.08 1952.01 7687.26

271.1 4.56 297.4 212.94 −23.8

967.4 4.87 30.97 68.4 −41.83

60.9 −0.54 3.31 106.1 −21.3

6273.9 9.1 437.7 739.08 −54.2

3.2. Landscape ecosystem services

does not mean a large area of the land was covered by housing; rather, it refers to the proportion of increment observed during the study period. Bare land increased from zero percent in 1973 to 739.08% in 2015, due to road construction, that dissected the forest. We identified a large quarry area. In addition, shrub land increased by 437.7% (1973–2015). On the other hand, forest cover decreased by 23.8%, 41.83% for 1973–1986 and 1986–2001 respectively (Table 4). Generally, forest land experienced the least persistent, whereas settlement was the most persistent cover type (Table 5). The net change to persistence ratio was large for forest land (negative), cultivated land (positive), shrub land (positive) and settlement land (positive). Overall, 4378.34 ha (i.e., sum of diagonal elements) of the total landscape remains unchanged (Table 5). The largest area of change matrix was observed from forest land to bare land, as compared to other land uses.

Across the study period forest land use ecosystem service values decreased, while others increased with the net value variations (Table 6). Considering ecosystem service value changes across the different intervals, 1973–1986, 1986–2001, 2001–2015 and 1973– 2015, forest land ecosystem services decreased whereas shrub land use ecosystem service increased. For cultivated land use the values are mixed with increase in value at an interval of 1973–2001, then a decrease in the value for 2001–2015 but the overall value showed increment. Although there is a general trend of reduction in ESV of forest land, it still has the highest value compared to other land uses. The net ecosystem service value was drastically reduced over the study period. The change in values estimated to be 40.7% during the study period with reduction in US$ 3.69 million from US$ 9.06 million in

Fig. 2. Land use/land cover change (1973–2015) for Chillimo forest.

50

Ecosystem Services 23 (2017) 47–54

T. Tolessa et al.

Table 5 Land use/land cover transition matrix of major changes in the landscape (ha), Central highlands of Ethiopia, 1973–2015. To final state (2015)

From initial state (1973) Settlement Cultivated land Bare land Shrub land Forest land Total 2015 Gain Net changeb Net persistence (NP)c

Settlement

Cultivated land

Bare land

Shrub land

Forest land

Total 1973

Loss

5.27 284.49 0 6.83 45.5

0.09 2302.94 0 75.97 1114.17

0 311.39 0 10.35 417.34

0 303.39 0 120.77 737.2

0 0.57 0 2.08 1949.36

5.36 3202.78 0 216 4263.12 4378.34a

0.09 899.84 0 95.23 2313.76

341.64 336.37 336.28 63.81

3493.17 1190.23 290.39 0.13

739.08 739.08 739.08 0

1161.36 1040.59 945.36 7.83

1952.01 2.65 −2311.2 −1.19

Bolded diagonal elements represent proportions of each land-use/cover class that were static (persisted) between 1973 and 2015. The loss column and gain row indicate the proportion of the landscape that experienced gross loss and gain in each class, respectively. All the figures in the table are in percent except Np, which is a ratio. a The bolded figure is the sum of diagonals and represents the overall persistence (i.e., the landscape that did not change). b Net change=gain–loss. c Np refers to net change to persistence ratio (i.e., net change/diagonals of each class). Table 6 Total ecosystem service values estimated for each land use and land cover category and changes from 1973 to 2015 in the study area. LULC Class

Settlement Cultivated land Shrub Land Bare Land Forest Land Sum

ESV (US$ million)

ESV (US$ million) change

1973

1986

2001

2015

1973–1986

1986–2001

2001–2015

1973–2015

0 0.29 0.21 0 8.56 9.06

0 0.31 0.83 0 6.52 7.66

0 0.323 1.1 0 4.98 6.40

0 0.321 1.13 0 3.92 5.37

0 0.02(6.9) 0.62(295.2) 0 −2.04(−23.8) −1.40(−15.5)

0 0.01(4.2) 0.27(32.5) 0 −1.54(−23.6) −1.3(−16.4)

0 −0.02(−0.62) 0.03(2.73) 0 −1.06(−21.3) −1.030(−16.1)

0 0.031(10.7) 0.92(438.1) 0 −4.64(−54.2) −3.69(−40.7)

Figures in the parenthesis refers to the percentage of change Table 7 Total ecosystem service values estimated for each land use and land cover category and changes of 1973 and 2015 in the study area. Land use

Settlement Cultivated Land Shrub land Bare land Forest land Sum

ESV (US$ million)

Change

Year (1973)

Year (2015)

ESV

CCk (%)

0 0.29 0.21 0 8.56 9.06

0 0.321 1.13 0 3.92 5.37

0 0.031 0.92 0 −4.64 −3.69

0 0.34 10.2 0 −51.2 −40.7

Table 8 Estimated annual value of each ecosystem services (ESVf in US $ million per year).

1973 to US$ 5.37 million in 2015 (Tables 6 and 7). The ecosystem service value for settlement was not possible to estimate, although rural houses are an integral part of the landscape matrix and have some impact in ecosystem service provision. In general, the contributions of other land uses to ecosystem service are minimal with significant alteration of impacts as a whole. Table 8 estimated the annual value of 17 ecosystem services. In terms of the estimated ESV, nutrient cycling (US$ 1.78 million), raw material (US$ 0.595 million), erosion control (US$ 0.481 million) and climate regulation (US$ 0.381 million) were reduced, while the only gain in ESV observed, were pollution control (US$ 0.004 million) however, the overall service were reduced by US$ 3.69 million. One very interesting result from ESV is all values were lost except pollution control. The effect of using alternative coefficients to evaluate total ESV in the study area is presented in Table 9. The coefficients of sensitivity (CS) of these analyses were less than one in all cases. CS ranged from a

Ecosystem service

ESVf1973

ESVf2015

Change

CCf (%)

Gas regulation Climate regulation Disturbance regulation Water regulation Water supply Erosion control Soil formation Nutrient cycling Waste treatment Biological control Food production Raw material Genetic resource Recreation Cultural Pollution control Habitat/refugia Sum

– 0.98 0.02 0.026 0.035 1.07 0.045 4.01 0.389 0.0944 0.319 1.37 0.178 0.492 0.009 0.045 – 9.06

– 0.599 0.01 0.014 0.019 0.589 0.031 2.23 0.271 0.0939 0.301 0.775 0.099 0.295 0.006 0.049 – 5.37

– −0.381 −0.01 −0.012 −0.016 −0.481 −0.014 −1.78 −0.118 −0.0005 −0.018 −0.595 −0.079 −0.197 −0.003 0.004 – −3.69

– −4.2 −0.11 −0.13 −0.18 −5.3 −0.15 −19.6 −1.3 −0.01 −0.19 −0.07 −0.87 −2.17 −0.03 0.04 – −40.7

low of 0.02–0.21 for shrub land, to a high of 0.73–0.94 for forest land, when the value coefficients for these land use categories were adjusted by 50% (Table 9). CS for forest land was the highest because of the highest service value coefficient and relatively larger area coverage. If CS (the ratio of the percentage change in the estimated total ecosystem value and the percentage change in the adjusted valuation coefficient) is greater than one, then the estimated ecosystem value is elastic with respect to that coefficient. However, if the ratio is less than one, then the estimated ecosystem value is regarded as inelastic. The greater the proportional change in the ecosystem service value relative to the 51

Ecosystem Services 23 (2017) 47–54

T. Tolessa et al.

Table 9 Change in total estimated ecosystem services and coefficient of sensitivity (CS) after adjusting ecosystem services valuation coefficients (VC) in the central highlands of Ethiopia. 1973

Cultivated land VC ± 50% Shrub land VC ± 50% Forest land VC ± 50%

1986

2001

2015

Percent

CS

Percent

CS

Percent

CS

Percent

CS

1.71 1.16 47.1

0.03 0.02 0.94

2.09 5.41 42.5

0.04 0.11 0.85

2.65 8.57 38.79

0.05 0.17 0.78

3.15 8.8 36.37

0.06 0.21 0.73

in the area, which serve as buffer zone for the natural forest of Chillimo. The increased trend in shrub land might be attributed to the effort made by the office of Agricultural development.

proportional change in the valuation coefficient, the more critical the use of an ecosystem value coefficient. Our results of all analyses indicate that the total ecosystem values estimated for the study area are relatively inelastic with respect to the ecosystem value coefficients, which also suggests that our ecosystem value evaluation is reasonable and robust.

4.2. Landscape ecosystem services LU/LC change was identified as one important driver of change of ecosystems and their services. The overall ecosystem services of the study area was found to be highly reduced due to the reduction of forest land, which is in line with other findings elsewhere (Li et al., 2007; Hu et al., 2008; Martínez et al., 2009; Cai et al., 2013; Leh et al., 2013; Kindu et al., 2016). The small decrease in ecosystem service values of cultivated land use 2001–2015 may be attributed to the abandonment of this land as fallow. When disaggregated on the basis of ecosystem services, such as raw material, recreation and cultural services were also reduced in our study area, which confirms with other findings (Hu et al., 2008; Hao et al., 2012; Nahuelhual et al., 2014). Sensitivity analysis undertaken to test the robustness of our estimation confirmed that our ESV had the power to predict and it was found to be reliable, which was similar to the findings of Li et al. (2007), Hu et al. (2008) and Hao et al. (2012). However, we believe that the ecosystem service values developed by Costanza et al. (1997) may underestimated the current real services of the land use types, because some of the land uses were estimated through proxies, which, in turn, may not show the amount of values received. In addition, ecosystem service value coefficients assume spatial homogeneity of services within LU/LC. Hence, developing ecosystem service values coefficients for local specific conditions that take in to account physical characteristics of the land is also important, but given the scarcity of data on the ground, the use of existing data sets help to formulate policies related to land management practices and its resultant impacts. Although estimation of ESV is mainly focusing on (semi) natural environments, cultivated land was considered, because it provides an important source of livelihood in Ethiopia and the major policy focus is on improving the productivity of the agricultural sector. For example, Tadesse et al. (2014a) demonstrated that coffee farms sustained about 40% of the main marketed forest based ecosystem services, 67% of provisioning and < 50% of cultural and regulating services found in forest fragments. In addition, farmers deliberately retain some important tree species either on their farms or boundary of the farm for ecosystem services to compensate for tradeoffs (Nelson et al., 2009; Ango et al., 2014), and such practices are common in the Ethiopian highlands. This ecosystem service valuation is comprehensive, since it incorporates the biophysical and social components of the social-ecological system, in which both the service and the beneficiaries were identified. This means that the ecosystem services identified (Costanza et al., 1997; MA, 2005; de Groot et al., 2012) were incorporated for different land use types. In addition, it can be observed from our results that some ecosystem services are local specific (such as cultural services), but other have a global implication (for example, climate regulation), which calls for an appropriate distribution of burden due to loss of ecosystem services, which calls for institutional arrangements (Lopes et al., 2015) and calls for both local and global initiatives that encourages conservation of forests as well as payment for ecosystem services (Martínez et al., 2009).

4. Discussion 4.1. Analysis of land use dynamics Our LU/LC analysis showed that forest cover decreased during the study period (1973–2015). This is in conformity with many LU/LC studies conducted in Ethiopia (Reid et al., 2000; Dessie and Kleman, 2007; Tsegaye et al., 2010; Minale and Rao, 2012; Gebrehiwot et al., 2014; Meshesha et al., 2014) and other tropical areas (Lira et al., 2012; Nahuelhual et al., 2014). LU/LC changes are dynamic and non linear, that is, the conversion from one land use to the other does not follow a similar pattern, due to natural or anthropogenic factors like policy change, population growth and decrease in the productivity of land (Dessie and Kleman, 2007; Meshesha et al., 2014). In 1973 and subsequent years, in Ethiopia, there was a change in the feudal regime to a military regime (Berisso, 1995; Reid et al., 2000; Dessie and Kleman, 2007). The military regime proclaimed the nationalization of all rural land by abolishing private and common property of the land, thereby giving a usufruct rights for all. This policy made land and land related resources absolutely owned by the state, which in turn was unable to monitor and enforce laws. As a result, forest land was converted to settlements, agricultural land and highly degraded because of low level of land management practices. The current government, Ethiopian People Revolutionary Democratic Front (EPRDF), also maintains the same status, where land is the property of the nations, nationalities and people of Ethiopia according to Article 40(3) of the constitution endorsed in 1994 (FDRE, 1994). This makes land related resources to be easily convertible to agricultural land. Accordingly, mainly small-holder farmers convert forest land to agriculture. Expansion of cultivated land is driven by market-oriented production of high value crops for national and international markets as a raw material for agro-processing industries (Watson, 2007; Dejene et al., 2013). Such conversion of forest land into crop production in relation to government and policy changes resulted in the reduction of ecosystem services of forests. Many studies in tropical countries confirm the impacts of government policy on forest due to extensive agricultural policy approach (Li et al., 2007; Hu et al., 2008; Lira et al., 2012; Oestreicher et al., 2014). For example, in Ethiopia, farmers have more legal rights on the land if they convert forest land in to farm land, as the law stipulates natural forests as the property of the government. This makes farmers to legally or illegally convert forests in to farmland, as this guarantees them the use of the land for an indefinite period of time. The general trend observed in the study area implies a loss of forestland and an increase in cultivated areas and settlements. Continued LU/LC change, coupled with uncertain climatic conditions, greatly affects people's livelihoods and puts the agropastoral production system under stress, given the existing demographic changes in the country. There have also been some trends in the increase of shrub land 52

Ecosystem Services 23 (2017) 47–54

T. Tolessa et al.

Acknowledgements

In contrast to the improvement of the ecosystem services in developed nations (Haines-Young et al., 2012), where land under crop production is converted to forests, in most developing countries like Ethiopia, certain ecosystem services are at the highest rate of loss. This can be attributed to a number of underlying and proximate causes of land use and land cover changes. At the rural landscapes in Ethiopia, where we can find highly modified cover types, considering each landscape function in the quantification and monitoring will help land managers to optimize the benefits and reducing the losses (Nelson et al., 2009; Leh et al., 2013). Furthermore, local specific studies of ecosystem service values are rare for our context and hence made the estimation very different. We believe that future ESV data base for different LU/LC is very important to quantify the overall ecosystem services in Ethiopia. Given the location of our study area within the highland of the country, where there is diversity of ecosystem services expected due to the heterogeneity of land use characteristics, as studies such as Grết-Regamey et al. (2012) have shown that local level ESV is of having a paramount importance in Ethiopian landscapes (Tadesse et al., 2014a). Such estimation of ESV, using LU/LC and established ecosystem service value coefficient, is extremely important in Ethiopia where ground data collection is expensive, there is scarce data on the historical land uses of rural areas and land degradation is pervasive. It provides alternatives and robust information for decision making processes at landscape level and similar works can be conducted in other parts of the country. Of the LU/LC within the landscape, our results indicate that forests provide a higher level of ecosystem service than others due to the relatively larger area coverage and higher value coefficients assigned and hence a reduction in this particular component of the ecosystem hampers a balanced flow of services from the landscape.

This work was supported by Addis Ababa University research grant for thematic area related to forest fragmentation and ecosystem service research. We appreciate the anonymous reviewers of the paper and editors of the journal for their useful and constructive comments. We also would like to thank Margareta Langbacka Walker for her assistance in edition of the manuscript. The first author thanks Ambo University for granting scholarship. References Alarcon, G.G., de Freitas, L.A., dos, S., da Fountoura, G.O., de Macedo, C.X., Ribeiro, D.C., 2016. The challenges of implementing a legal framework for payment for ecosystem services in Santa Catarina. Braz. J. Nat. Conserv.. http://dx.doi.org/ 10.1016/j.ncon.2016.05.003. Ameha, A., Larson, H.O., Lemenih, M., 2014. Participatory forest management in Ethiopia: learning from pilot projects. Environ. Manag. 53, 838–854. Ango, T.G., Börjeson, L., Senbeta, F., Hylander, K., 2014. Balancing ecosystem services and disservices: smallholder farmers’ use and management of forest and trees in an agricultural landscape in southwestern Ethiopia. Ecol. Soc. 19 (1), 30. http:// dx.doi.org/10.5751/ES-06279-190130. Berisso, T., 1995. Deforestation and environmental degradation in Ethiopia: the case of Jam Jam Province. Northeast Afr. Stud. 2 (2), 139–155. Bryan, B.A., 2013. Incentives, land use, and ecosystem services: synthesizing complex linkages. Environ. Sci. Policy 27, 124–134. Cai, Y.-B., Zhang, H., Pan, W.-B., Chen, Y.-H., Wang, X.-R., 2013. Land use pattern, socio-economic development, and assessment of their impacts on ecosystem service value: study on natural wetlands distribution area (NWDA) in Fuzhou city, southeastern China. Environ. Monit. Assess. 185, 5111–5123. Costanza, R., d’Arge, R., de Groot, R., Farberk, S., Grasso, M., Hannon, B., Limburg, K., Naeem, S., O’Neill, R.V., Paruelo, J., Raskin, R.G., Suttonkk, P., van den Belt, M., 1997. The value of the world's ecosystem services and natural capital. Nature 387, 253–260. Costanza, R., de Groot, R., Sutton, P., van der Ploeg, S., Anderson, S.J., Kubiszewski, I., Farber, S., Turner, R.K., 2014. Changes in the global value of ecosystem services. Glob. Environ. Change 26, 152–158. de Bello, F., Lavorel, S., Dı´az, S., Harrington, R., Cornelissen, J.H.C., Bardgett, D., Berg, M.P., Cipriotti, P., Feld, C.K., Hering, D., da Silva, P.M., Potts, S.G., Sandin, L., Sousa, J.P., Storkey, J., Wardle, D.A., Harrison, P.A., 2010. Towards an assessment of multiple ecosystem processes and services via functional traits. Biodivers. Conserv. 19, 2873–2893. de Groot, R.S., Alkemade, R., Braat, L., Hein, L., Willemen, L., 2010. Challenges in integrating the concept of ecosystem services and values in landscape planning, management and decision making. Ecol. Complex. 7, 260–272. de Groot, R., Brander, L., Ploeg, S., Costanza, R., Bernard, F., Braat, L., Christie, M., Crossman, N., Ghermandi, A., Hein, L., Hussain, S., Kumar, P., McVittie, A., Portela, R., Rodriguez, L.C., Brink, P., van Beukering, P., 2012. Global estimates of the value of ecosystems and their services in monetary units. Ecosyst. Serv. 1, 50–61. de Marko, P.J.R., Coelho, F.M., 2004. Services performed by the ecosystem: forest remnants influence agricultural cultures’ pollination and production. Biodivers. Conserv. 13, 1245–1255. Dejene, T., Lemenih, M., Bongers, F., 2013. Manage or convert Boswellia woodlands? Can frankincense production payoff?. J. Arid Environ. 89, 77–83. Dessie, G., Kleman, J., 2007. Pattern and magnitude of deforestation in the south central rift valley region of Ethiopia. Mt. Res. Dev. 27 (2), 162–168. Federal Democratic Republic of Ethiopia (FDRE), 1994, Constitution of the Federal Democratic Republic of Ethiopia, Addis Ababa, Ethiopia, pp. 50. Feoli, E., Vuerich, L.G., Woldu, Z., 2002. Processes of environmental degradation and opportunities for rehabilitation in Adwa, Northern Ethiopia. Landsc. Ecol. 17, 315–325. Gebrehiwot, S.G., Bewket, W., Gardenas, A.I., Bishop, K., 2014. Forest cover change over four decades in the Blue Nile Basin, Ethiopia: comparison of three watersheds. Reg. Environ. Change 14 (1), 253–266. http://dx.doi.org/10.1007/s10113-013-0483-x. Grết-Regamey, A., Brunner, S.H., Kienast, F., 2012. Mountain ecosystem services: who cares? Mt. Res. Dev. 32 (S1), S23–S34. Hao, F., Lai, X., Ouyang, W., Xu, Y., Wei, X., Song, K., 2012. Effects of land use changes on the ecosystem service values of a reclamation farm in Northeast China. Environ. Manag. 50 (5), 888–899, http://dx.doi. 1007/s00267-012-9923-5. Haines-Young, R., Potschin, M., Kienast, F., 2012. Indicators of ecosystem service potential at European scales: mapping marginal changes and trade-offs. Ecol. Indic. 21, 39–53. Hu, H., Liu, W., Cao, M., 2008. Impact of land use and land cover changes on ecosystem services in Menglun, Xishuangbanna, Southwest China. Environ. Monit. Assess. 146, 147–156. Hurni, H., Tato, K., Zeleke, G., 2005. The implications of changes in population, land use, and land management for surface runoff in the upper Nile basin area of Ethiopia. Mt. Res. Dev. 25, 147–154. Janssen, L.I.F., Frans, J.M., Wel, V.D., 1994. Accuracy assessment of satellite derived land cover data: a review. Photogramm. Eng. Remote Sen. 60, 410–432. Kidane, Y., Stahlmann, R., Beierkuhnlein, C., 2012. Vegetation dynamics, and land use and land cover change in the Bale Mountains, Ethiopia. Environ. Monit. Assess. 184,

5. Conclusion Ecosystem service value estimation based on LU/LC analysis is very vital to indicate how much service changed through human activities on spatial and temporal scales. Such estimation for local, regional and global levels are utmost importance in influencing decision making processes through modifying national accounting systems to reflect the true values of ecosystem services so that it will be ultimately be used as a basis for sustainable development. Another benefit could be related to designing alternative development projects where benefits must be in line with the services aspired. Landscape influenced by a mosaic of land uses provide optimal ecosystem services, rather than monoculture, and hence maintaining an appropriate proportion of each is very essential in rural landscapes where farmers tend to diversify production to reduce risks. We found that forest land significantly reduced to other land uses by 54.2% and an increase in other land uses with varying proportions. Land use matrix also shows a significant reduction of forest land to agricultural land use by 1117.17 ha. In addition, the overall ESV decreased by 40.7%. Of the ecosystem services identified pollution control (0.04%) had positive value which is very small as compared to other ESVs, while other services had negative values indicating a decrease in those values. This result shows the higher level of overall ecosystem service value loss. CS of ESV shows our estimation is robust concerning all the values for each land use less than one. CS for forest land use is the highest among all land uses for all study period indicating the largest size and higher ESV for forest land use. Payment for ecological services, such as carbon trading for forest conservation, should be established on the basis of a rigorous valuation of these ecosystem services to support community livelihoods which are at the greatest risk due to land use change and its associated impacts. We believe that these findings will be an important milestone for future research and policy formulation in Ethiopia. 53

Ecosystem Services 23 (2017) 47–54

T. Tolessa et al.

impact on the environment in the Ethiopian and Eritrean highlands: a state of the art. Earth Sci. Rev. 64, 273–320. Oestreicher, J.S., Farella, N., Paquet, S., Davidson, R., Lucotte, M., Mertens, F., SaintCharles, J., 2014. Livelihood activities and land-use at a riparian frontier of the Brazilian Amazon: quantitative characterization and qualitative insights into the influence of knowledge, values, and beliefs. Hum. Ecol. 42, 521–540. Reid, R.S., Kruska, R.L., Muthui, N., Taye, A., Wotton, S., Wilson, C.J., Mulatu, W., 2000. Land-use and land-cover dynamics in response to changes in climatic, biological and socio-political forces: the case of southwestern Ethiopia. Landsc. Ecol. 15, 339–355. Satz, D., Gould, R.K., Chan, K.M.A., Guerry, A., Norton, B., Satterfield, T., Halpern, B.S., Levine, J., Woodside, U., Hannahs, N., Basurto, X., Klain, S., 2013. The challenges of Incorporating cultural ecosystem services into environmental assessment. AMBIO 42, 675–684. Slaymaker, O., 2010. Drivers of mountain landscape change during the twenty-first century. J. Soils Sediment. 10, 597–610. Summers, J.K., Smith, L.M., Case, J.L., Linthurst, R.A., 2012. A review of the elements of human well-being with an emphasis on the contribution of ecosystem services. AMBIO 41, 327–340. Tadesse, G., Zavaleta, E., Shennan, C., FitzSimmons, M., 2014a. Prospects for forestbased ecosystem services in forest-coffee mosaics as forest loss continues in southwestern Ethiopia. Appl. Geogr. 50, 144–151. Tadesse, G., Zavaleta, E., Shennan, C., FitzSim, M., 2014b. Local ecosystem service use and assessment vary with socio-ecological conditions: a case of native coffee-forests in southwestern Ethiopia. Hum. Ecol. 42 (6), 873–883. http://dx.doi.org/10.1007/ s10745-014-9704-2. Tamrat, B., 1993. Vegetation ecology of remnant afromontane forests on the central plateau of Shewa, Ethiopia. Acta Phytogeogr. Suec. 79, 1–64. TEEB, 2010. The Economics of Ecosystems and Biodiversity: Ecological and Economic Foundation. Earthscan, Cambridge, Washinton. Tesfaye, M.A., Bravo-Oviedo, A., Bravo, F., Ruiz-Peinado, R., 2015. Aboveground biomass equations for sustainable production of fuel wood in a native dry tropical afro-montane forest of Ethiopia. Ann. For. Sci. 72 (8), 1–12. http://dx.doi.org/ 10.1007/s13595-015-0533-2. Tolessa T., Senbeta S. and Abebe T., 2016, Land use/land cover analysis and ecosystem services valuation in the central highlands of Ethiopia, Forests, Trees and Livelihoods, 1-13, http://dx.doi:10.1080/14728028.2016.1221780. Tsegaye, D., Moe, S.R., Vedeld, P., Aynekulu, E., 2010. Land-use/cover dynamics in Northern Afar rangelands, Ethiopia. Agric. Ecosyst. Environ. 139, 174–180. Watson, C., 2007. Direct Consumptive Use Valuation of the Ecosystem Goods and Services in the Bale Mountains Eco-Region, Ethiopia. Imperial College London, Faculty of Natural Sciences, Centre of Environmental Policy, UK, 106.

7473–7489. Kindu, M., Schneider, T., Teketay, D., Knoke, T., 2016. Changes of ecosystem service values in response to land use/land cover dynamics in Munessa–Shashemene landscape of the Ethiopian highlands. Sci. Total Environ. 547, 137–147. Leh, M.D.K., Matlocka, M.D., Cummingsa, E.C., Nalley, L.L., 2013. Quantifying and mapping multiple ecosystem services change in West Africa. Agric. Ecosyst. Environ. 165, 6–18. Lemenih, M., Teketay, D., 2005. Effect of prior land use on the recolonization of native woody species under plantation forests in the highlands of Ethiopia. For. Ecol. Manag. 218, 60–73. Lira, P.K., Tambosi, L.R., Ewers, R.M., Metzger, J.P., 2012. Land-use and land-cover change in Atlantic forest landscapes. For. Ecol. Manag. 278, 80–89. Li, R.-Q., Dong, M., Cui, J.-Y., Zhang, L.-L., Cui, Q.-G., He, W.-M., 2007. Quantification of the impact of land-use changes on ecosystem services: a case study in Pingbian County, China. Environ. Monit. Assess. 128, 503–510. Lopes, L.F.G., Bentoa, J.M.R. d.-S., Cristovao, A.F.A.C., Baptista, F.O., 2015. Exploring the effect of land use on ecosystem services: the distributive issues. Land Use Policy 45, 141–149. Martínez, M.L., Pérez-Maqueo, O., Vázquez, G., Castillo-Campos, G., García-Franco, J., Mehltreter, K., Equihua, M., Landgrave, R., 2009. Effects of land use change on biodiversity and ecosystem services in tropical montane cloud forests of Mexico. For. Ecol. Manag. 258, 1856–1863. Meshesha, D.T., Tsunekawa, A., Tsubo, M., Ali, S.A., Haregeweyn, N., 2014. Land-use change and its socio-environmental impact in Eastern Ethiopia's highland. Reg. Environ. Change 14 (2), 757–768. http://dx.doi.org/10.1007/s10113-013-0535-2. Mesfin, A., 1998. Nature and Management of Ethiopian Soils. Alemaya University of Agriculture, Ethiopia, 272. Millennium Ecosystem Assessment (MA), 2005. Ecosystems and Human Well-being: A Framework For Assessment. Island Press, Washington. Minale, A.S., Rao, K.K., 2012. Impacts of land cover/use dynamics of Gilgel Abbay catchment of Lake Tana on climate variability, Northwestern Ethiopia. Appl. Geomat. 4, 155–162. Nahuelhual, L., Carmona, A., Aguayo, M., Echeverria, C., 2014. Land use change and ecosystem services provision: a case study of recreation and ecotourism opportunities in southern Chile. Landsc. Ecol. 29 (2), 239–344. http://dx.doi.org/ 10.1007/s10980-013-9958-x. Nelson, E., Mendoza, G., Regetz, J., Polasky, S., Tallis, H., Cameron, D.R., Chan, K.M.A., Daily, G.C., Goldstein, J., Kareiva, P.M., Lonsdorf, E., Naidoo, R., Ricketts, T.H., Shaw, M.R., 2009. Modeling multiple ecosystem services, biodiversity conservation, commodity production, and tradeoffs at landscape scales. Front. Ecol. Environ. 7 (1), 4–11. http://dx.doi.org/10.1890/080023. Nyssen, J., Poesen, J., Moeyersons, J., Deckers, J., Haile, M., Lang, A., 2004. Human

54