3.1 Land use and land cover (LULC) change Trends and their conversion matrix
Analysis of LULC patterns and the spatial distribution in the study watershed are presented in Tables 2 and 3 and Fig. 4.
Cultivated land and Rural settlement area (CL and RS)
Results show that cultivated land and Rural settlement areas have expanded from 742 ha (19.9%) in 1988, to 1405 ha (37.7%) in 1998, to 1423 ha (38.1%) in 2008 and 1435 ha (38.5%) in 2018 (Fig. 4; Table 2). The rate of change between 1988–1998, 1998–2008, 2008–2018 and also overall 1988–2018 periods, increased by 89.4, 1.3, 0.8 and 93.4%, respectively. The corresponding rates of increment for these study periods were 66.3, 1.8, 1.2 and 23.1 ha/year, respectively (Table 2).
Table 3 shows that during 1988–2018 periods, 333, 224 and 550 ha of CL & RS were gained from forest land, bush land and grassland, respectively. In this period, the highest gain was observed in CL and RS area 1107 ha (38.3%) of which the highest area was from grassland 550 ha (49.7%) followed by forest land 333 ha (30.1%). It indicates that it might be occurred the scarcity of grass land for live stocks and cutting more trees for fuel and construction materials which could associate with overgrazing and deforestation that may lead to environmental degradation in the study watershed. In reverse, 329, 27, 39, 8 and 11 ha of CL and RS area were transformed to forest land, bush land, grass land, built up area and bare land, respectively and the total loss from CL and RS to other land types was 414 ha (14.3%). Majority of CL and RS was converted to forest land 329 ha (79.5%) followed by grass land 39 ha (9.4%) during 1988–2018 (Table 3). This revealed that CL and RS area used for afforestation at the specified periods in the study watershed including majority area for eucalyptus trees.
From these findings, CL and RS is the dominant LULC type with a consistent increase at the expense of other LULCs in the catchment even though its growth was very slow during 1998–2008 and 2008–2018 (Table 2). In these periods, a regional resettlement programme occurred between1998-2005 for seven continuous years that framers from densely populated highland areas including this watershed moved to different other less populated lowland areas in southern Ethiopia. This government resettlement police implementation might have slowed down the land conversion in the second and third periods. Also the rest land may not be suitable for farming system due to landscape position. The rapid expansion of CL and RS was during 1988–1998 periods leaping from 19.9–37.7% which is almost 50% increment. According to FGD and KI interview the driving force for expansion of CL and RS was associated to an increment of population size demanding for food, construction and fuel production. Additionally, they said that the study watershed community faced food insecurity problems for long periods of years due to low crop productivity and environmental impact of eucalyptus trees expansion. Environmental impact of eucalyptus tree is negatively known as climate change due to its releasing halophytic chemicals and positively as income source for household farmers. The population in study area was increased by 59.4% from 1994 to 2018 (CSA, 2019, Fig. 3). It might be related to the 1991 change of government that resulted in the disbanding of the dergue army many of whom returned to their villages and given farm lands in the study watershed. Expansion of cultivated land may lead to degrading environment in one or other ways because expansion of agricultural activities could be associated with soil erosion, biodiversity loss, deforestation, climate change and hydrological disturbance. For example, when cultivated land increased extensively without appropriate management, nitrogen and phosphorus fertilizers unused by plants could be taken by runoff to water bodies which degrading water quality by eutrophication.
Forest Land
The forest area reduced from 34.6% in 1988 to 21.2% in 1998 and from 24.6% in 2008 to 23.0% in 2018, with annual diminishing rate of 49.9, 5.8 and 14.4 ha/year during 1988–1998, 2008–2018 and overall 1988–2018 periods, respectively (Fig. 4; Table 2). Conversely, forest land increased from 21.2% in 1998 to 24.6% in 2008 with the increasing rate 12.5 ha/year between 1998–2008 (Table 2).
In Table 3, areas of forest land were transformed from bush land (104 ha), CL and RS (329 ha) and grassland (183 ha). Forest land gained much land from CL and RS (53.4%) followed by grass land (29.7%) and the total gain of forest land is 616 ha (21.3%). This indicates that the increasing trend during 1998–2008 in forest land is associated with the adoption of sustainable land management (SLM) approach based on the participatory watershed rehabilitation scheme of the government that went to operation between 2002–2008 periods and mainly due to the expansion of Eucalyptus plantations in the study watershed. Divergently, 473, 333, 197, 6, 20 and 19 ha of forest land were converted to bush land, CL and RS, grassland, built up area, bare land and water body, respectively (Table 3). In these periods, the highest loss of forest land was observed in the study watershed which was 1048 ha (36.2%), which is a large reduction that can significantly contribute to eco-environment degradation. FGD and KI interview made clear that this reduction of forest land due to expansion of cultivated land for food production and high deforestation occurred for fuel, charcoal and construction. Whereas, a simultaneous shrinkage was observed in the area of forest land, even though considerable area of land added to it from different land categories. This indicates that Coka watershed forest land is under considerable pressure due to a rapidly increasing population, expanding agricultural activities and increasing deforestation.
Table 2
Summary of the classified LULC and their changes during 1988–2018 in Coka watershed.
LULC Types
|
LULC area coverage in ( Ha) and (%) during 1988–2018
|
LULC change in (Ha) and (%) during 1988–2018
|
1988
|
1998
|
2008
|
2018
|
1988–1998
|
1998–2008
|
2008–2018
|
1988–2018
|
Ha
|
%
|
Ha
|
%
|
Ha
|
%
|
Ha
|
%
|
Ha
|
%
|
Ha
|
%
|
Ha
|
%
|
Ha
|
%
|
Forest land
|
1291
|
34.6
|
792
|
21.2
|
917
|
24.6
|
859
|
23
|
-49.9
|
-38.7
|
12.5
|
15.8
|
-5.8
|
-6.3
|
-14.4
|
-33.5
|
Bush land
|
807
|
21.6
|
956
|
25.6
|
750
|
20.1
|
685
|
18.3
|
14.9
|
18.5
|
-20.6
|
-21.5
|
-6.5
|
-8.7
|
-4.1
|
-15.1
|
CL & RS
|
742
|
19.9
|
1405
|
37.7
|
1423
|
38.1
|
1435
|
38.5
|
66.3
|
89.4
|
1.8
|
1.3
|
1.2
|
0.8
|
23.1
|
93.4
|
Grass land
|
891
|
23.9
|
578
|
15.5
|
641
|
17.2
|
631
|
16.9
|
-31.3
|
-35.1
|
6.3
|
10.9
|
-1
|
-1.6
|
-8.7
|
-29.2
|
Built up
|
0
|
0
|
0
|
0
|
0
|
0
|
23
|
0.6
|
0
|
0
|
0
|
0
|
2.3
|
|
0.8
|
|
Bare land
|
0
|
0
|
0
|
0
|
0
|
0
|
66
|
1.8
|
0
|
0
|
0
|
0
|
6.6
|
|
2.2
|
|
Water body
|
0
|
0
|
0
|
0
|
0
|
0
|
32
|
0.9
|
0
|
0
|
0
|
0
|
3.2
|
|
1.1
|
|
Total
|
3731
|
100
|
3731
|
100
|
3731
|
100
|
3731
|
100
|
A
|
|
B
|
|
C
|
|
D
|
|
Bush Land
The area of bush land increased from 21.6% in 1988 to 25.6% in 1998 but diminished from 25.6% in 1998 to 20.1% in 2008 and to 18.5% in 2018 (Fig. 4; Table 2). The rate of change was increased by 14.9 ha/year during 1988 − 1998 but decreased by 20.6, 6.5, and 4.1 ha/year during 1998–2008, 2008–2018 and 1988–2018, respectively (Table 2).
During 1988–2018 (Table 3), bush land expanded in expense of CL & RS (27 ha), grassland (22 ha) & forest land (473 ha) and the major gain was from forest land (90.6%) with total gain of 522 ha (18.0%). This reveals that conversion of more forest land to bush land could be associated with CO2 emission which might lead to climate change in the study watershed. In the same periods, 104, 224, 290, 3, 14, and 9 ha of bush land were converted to forest land, CL and RS, grassland, built up area, bare land, and water body, respectively. The highest bush land converted was to grass land (45%) followed by cultivated land (34.8%) with total loss of 644 ha (22.3%) during the mentioned period (Table 3). The result shows that transformation of more bush land to grass land is one kind of environmental degradation which may come from overgrazing. According to FGD and KI interview the reduction of bush land was linked to an increment of number of animals and due to high demand of grazing area and expansion of CL & RS in Coka watershed.
Grassland
The grassland area coverage of the watershed was 23.9% in 1988, 15.5% in 1998, 17.2% in 2008 and 16.9% in 2018 (Fig. 4; Table 2). During 1988–1998, 2008–2018 and 1988–2018 periods, grass land decreased by 31.3, 1.0 and 8.7 ha/year, respectively.
Generally, grassland gained from forest land (197 ha), bush land (290 ha) and CL & RS (39 ha) during 1988–2018 periods (Table 3). The highest gain occurred was from bush land 290 ha (55.1%) followed by forest land 197 ha (37.5%) whose total gain is 526 ha (18.2) in the same periods. This showed that it increased by 6.3 ha/year between 1998–2008 period due to clearing forest land and bush land for animal grazing (Table 3). In reverse, grassland converted to forest land (183 ha), bush land (22), CL and RS (550 ha), built up area (6 ha), bare land (21 ha) and water body (4 ha). The largest area of grassland 550 ha (70%) transformed to CL & RS followed by forest land 183 ha (23.3%) and the total loss from the grass land is 786 ha (27.2%) during 1988–2018 periods. This result showed that more grassland used for agricultural activities during over all periods (1988–2018) that might be related to increasing population size bearing expansion of CL and RS in the study area. FGD and local experts confirmed that the major driving force for decrement of grassland is expansion of agricultural activities which come from demand of food production in study watershed.
Bare Land
The area occupied by bare land was 66 ha (1.8%) in 2018 (Fig. 4; Table 2). Its rate of increment during 2008–2018 and overall 1988–2018 periods was 6.6 ha/year and 2.2 ha/year, respectively (Table 2). The 1988, 1998 and 2008 LULC analysis divulges that bare land was not occurred in the Coka watershed.
In Table 3 (1988–2018 periods), bare land transformed from forest land (20 ha), bush land (14 ha), cultivated land (11 ha) and grassland (21 ha). Bare land gains the highest amount of land from grass land (32%) with total gain of 66 ha (2.3%). As FGD and KI interview put the reason of bare land expansion at third period was mismanagement of LULCs systems which results the degradation of cultivated land, forest land and grassland which could be associated with depletion of fertile land, loss of biological and economic productivity, soil erosion, habitat fragmentation and reduced ecosystem services in the study area.
Built up Area
The 1988, 1998 and 2008 LULC analysis reveals that the study watershed was not used by built up areas within the Tembaro woreda Mudulla Twon and Semene ambukuna kebele center (Fig.; Table 2). Built up area expanded 23 ha (0.6%) in 2018. Its rate of increment during 2008–2018 and 1988–2018 periods, 2.3 ha/year and 0.8 ha/year, respectively (Table 2).
As indicated in Table 3 (1988–2018 periods), built up area was converted from forest land (6 ha), bush land (3 ha), cultivated land (8 ha) and grassland (6 ha). In the same period the least gain was observed in built up area 23 ha (0.8%) and which the highest conversion from cultivated land (34.8%) and the total gain of built up area is 23 ha (0.8%) (Table 3). The expansion of built up area was occurred in the expense of cultivated land, forest land, bush land and grass land. As indicated in Table, 33 (49.3%) of farmers and local experts involved in FGD and KII elucidated that major cause for the expansion of built up area due to increment of land price, cutting more forest for construction and migration of people from rural to mudula town some land use types converted from cultivated land, grass land and bush land to built up area from Coka watershed in the north-east side of mudula town. This finding shows that the occurrence of expansion of urbanization toward the study watershed which may lead to habitat fragmentation and loss of natural resources which may provide remarkable wastes and causes environmental pollution for surrounding areas.
Water Body
The area coverage of water body was 32 ha (0.9%) in 2018 (Fig. 4; Table 2). Its rate of increment during 2008–2018 and 1988–2018 periods 3.2 ha/year and 1.1 ha/year, respectively (Table 2). The 1988, 1998 and 2008 LULC analysis divulges that the back flow of gilgel gibe 3 dam water was not happened in the Coka watershed.
During 1988–2018 periods, water body was gained from bush (9 ha), grassland (3 ha) and forest land (19 ha) (Table 3). This result indicates that expansion of water body was occurred in the expense of cultivated land, forest land, bush land and grass land at the out let of Coka watershed. FGD and KI interview confirmed that the water body was increased during this period because of back flow of Gilgal gibe 3 dam accumulations at the bottom (outlet point) part of the study watershed.
Table 3
LULC change matrix, net gain and net loss during 1988–2018 in Coka watershed
Year 1
(1988)
|
LULC types in Coka watershed in Year 2 (2018)
|
Year1 Total
(Ha)
|
Loss in1988 (Ha)
|
Loss in1988
(%)
|
Forest land
|
Bush land
|
CL & RS
|
Grass land
|
Built up
|
Bare land
|
Water body
|
Forest land
|
243
|
473
|
333
|
197
|
6
|
20
|
19
|
1291
|
1048
|
36.2
|
Bush land
|
104
|
163
|
224
|
290
|
3
|
14
|
9
|
807
|
644
|
22.3
|
CL & RS
|
329
|
27
|
328
|
39
|
8
|
11
|
0
|
742
|
414
|
14.3
|
Grass land
|
183
|
22
|
550
|
105
|
6
|
21
|
4
|
891
|
786
|
27.2
|
Built up area
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
|
|
Bare land
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
|
|
Water body
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
|
|
Year 2 Total(Ha)
|
859
|
685
|
1435
|
631
|
23
|
66
|
32
|
3731
|
|
|
Gain(Ha) in2018
|
616
|
522
|
1107
|
526
|
23
|
66
|
32
|
|
2892
|
|
Gain(%) in 2018
|
21.3
|
18.0
|
38.3
|
18.2
|
0.8
|
2.3
|
1.1
|
|
|
100
|
3.2. LULC change along slope gradients in Coka watershed
During 1988–2018, LULC change along slope gradient categorized into cultivated land (cultivated land & rural settlement, bare land and built up area) and non-cultivated land (forest land, bush land, grass land and water body) (Fig. 5). Cultivated land along slope classes (< 30% & >30%) were expanded by 13.4 & 7.3% in 1988 to 21.9 & 16.1% in 1998 to 22 & 17.2% in 2008 and to 22.8 & 17.6% in 2018, respectively. Its annual rate of expansion along stope gradients (< 30% & >30%) was 31.9 & 32.6 ha/year, 0.2 & 4.2 ha/year, 3 & 1.6 ha/year and 11.7 & 12.8 ha/year during 1988–1998, 1998–2008, 2008–2018 and 1988–2018, respectively (Fig. 5).
As indicated in Fig. 5, large area coverage of cultivated land was mainly on gentle slopes (< 30%) than steep slope (> 30%) in 1988 but a dramatic increase occurred from 1988 to 1998 on both gentle (< 30%) and steep (> 30%) slopes might be due to population growth and disbanded durgue arm settlement in the study watershed. Conversely, a less dramatic increase from 2008 to 2018 on both gentle (< 30%) and steep (> 30%) slopes because almost every gentle and steep slopes had already been converted to cultivated land. The areas remained unchanged to cultivated land might be rocky exposure, cliffs, some very marginal bushy area, forest land and grass land around rivers. Under normal situations, steep lands should be covered by forest or used for perennial crops but not for annual crop cultivation. The increasing demand for agriculture activities over decades had brought changes in the LULC system and deforestation of higher slope areas. This steep slope cultivation could result accelerating soil erosion, deforestation, biodiversity loss, climate change and hydrological disturbance in the study watershed.
On the other hand, non-cultivated land on slope classes (< 30% & >30%) reduced from 34.8 & 44.5% in 1988 to 26.2 & 35.8% in 1998 to 25.7 & 35.1% in 2008 and 25.5 & 34% in 2018, respectively. Its annual rate of decrement was 32.1 & 32.4 ha/year, 1.8 & 2.6 ha/year, 0.7 & 3.9 ha/year and 11.5 & 13 ha/year along slope classes (< 30% & >30%) during 1988–1998, 1998–2008, 2008–2018 and 1988–2018 p, respectively (Fig. 5). This reveals that non-cultivated land diminished during 1988–2018 periods might be due to expansion of cultivated land to satisfy growing population at the expense of decreasing forest, bush and grass land in the study area. The decrement of non-cultivated land could be linked with environmental degradation in terms of deforestation and land degradation.
3.3. Projection of future LULC area and conversion matrix in Coka watershed
Cultivated land and rural settlement area (CL & RS)
A continuous raise of CL & RS is projected to be 43.3% by 2040 and 52.8% by 2060 (Fig. 6; Table 4). Its rate of increment during 2018–2040, 2040–2060 and 2018–2060 periods will be 8.1, 17.9 and 12.2 ha/year, respectively, which indicates high expansion of CL & RS for projection periods. During 2018–2060 periods, 360 ha of bush land, 301 ha of grassland and 466 ha of forest land is predicted to be transformed to CL & RS (Table 5). Among which majority of its gain is from forest land (41.4%) followed by bush land (31.9%) with total gain of 1127 ha (53%). This projection reveals that expansion of CL & RS might lead to land degradation if inappropriate land use is common. In reverse, 39, 123, 342, 36 and 51 ha of CL & RS will be converted to bush land, grassland, forest land, built up area and bare land, respectively (Table 5). Its’ total loss will be 591 ha (27.8%) (Table 5).
Forest land
The projected forest land showed that a decrement trend from 859 ha (23%) in 2018 to 840 ha (22.5%) in 2040 and to 728 ha (19.5 %) in 2060 (Fig. 6; Table 4). The rate of change reveals that forest land will decrease by 0.9, 5.6 and 3.1 ha/year during 2018–2040, 2040–2060 and 2018–2060 periods, respectively (Table 4). As shown in Table 6, it is projected that 65 ha of bush land, 12 ha of grassland and 342 ha of CL & RS with total of 419 ha (19.7%) will be transformed to forest land during 2018–2060 periods. Conversely, 51, 466, 10, 4 and 19 ha of forest land will be converted to bush land, CL & RS, built up area, bare land and water body respectively with total loss of 550 ha (25.9%) in the same periods which projected decrement of forest could be directly connected with CO2 emission (climate change) that may contribute to environmental degradation in study watershed.
Table 4
The predicted LULC area between 2018 and 2060 in Coka watershed
|
Predicted LULC area (2018–2060)
|
Predicted LULC change rate (2018–2060)
|
LULC Types
|
2018
|
2040
|
2060
|
2018–2040
|
2040–2060
|
2018–2060
|
|
Ha
|
%
|
Ha
|
%
|
Ha
|
%
|
Ha
|
%
|
Ha
|
%
|
Ha
|
%
|
Bush land
|
685
|
18.3
|
495
|
13.3
|
261
|
7.0
|
-8.6
|
-27.7
|
-11.7
|
-47.3
|
-10.1
|
-61.9
|
Grass land
|
631
|
16.9
|
598
|
16.0
|
470
|
12.6
|
-1.5
|
-5.2
|
-6.4
|
-21.4
|
-3.8
|
-25.5
|
Forest land
|
859
|
23
|
840
|
22.5
|
728
|
19.5
|
-0.9
|
-2.2
|
-5.6
|
-13.3
|
-3.1
|
-15.3
|
CL & RS
|
1435
|
38.5
|
1614
|
43.3
|
1971
|
52.8
|
8.1
|
12.5
|
17.9
|
22.1
|
12.8
|
37.4
|
Built Up Area
|
23
|
0.6
|
42
|
1.1
|
80
|
2.1
|
0.9
|
82.6
|
1.9
|
90.5
|
1.4
|
247.8
|
Bare land
|
66
|
1.8
|
93
|
2.5
|
139
|
3.7
|
1.2
|
40.9
|
2.3
|
49.5
|
1.7
|
110.6
|
Waterbody
|
32
|
0.9
|
49
|
1.3
|
82
|
2.2
|
0.8
|
53.1
|
1.7
|
67.3
|
1.2
|
156.3
|
Total
|
3730
|
100
|
3731
|
100
|
3731
|
100
|
|
|
|
|
|
|
Bush land
As indicated (Fig. 6; Table 4), bush land will be transformed from 18.3% in 2018 to 13.3% in 2040 and to 7.0% in 2060. Its predicted rate of decrement during 2018–2040, 2040–2060 and overall 2018–2060 is 27.7, 47.3 and 61.9%, respectively. However, 63 ha of grassland, 51 ha of forest land and 39 ha of CL & RS will be converted to bush land with total gain of 153 ha (7.2%). Reversely, among 577 ha (27.1%) total loss of bush land,125 ha to grassland, 65 ha to forest land, 360 ha to CL & RS, 5 ha to built up area, 6 ha to bare land and 16 ha to water body will be transformed during 2018–2060 (Table 5). This conversion shows that human activities in bush land area are high.
Grassland
A decrement of grass land is predicted to be observed from 631 ha (16.9%) in 2018 to 598 ha (16.0%) in 2040 and to 470 ha (12.6%) in 2060 (Fig. 6; Table 4). The rate of decreasing pattern between 2018–2040, 2040–2060 and 2018–2060 periods is projected to be by 1.5, 6.4 and 3.8 ha/year, respectively. As indicated in Table 5, transformation of grassland will be 125 ha from bushland and 123 ha from CL & RS with total gain of 248 ha (11.7%). In opposite, 63 ha, 12 ha, 301 ha, 6 ha, 12 ha and 15 ha of grass land will be converted to bush land, forest land, CL & RS, built up area, bare land and water body, respectively, with the total loss of 409 ha (19.2%) in the mentioned periods. 301 ha (73.6%) of grass land which is the highest, will be transformed to CL & RS in projected land map. This reveals that expansion of CL & RS will be high at the expense of grassland if not protected by rules and regulations for the projected periods in the study watershed.
Table 5
LULC conversion matrix between 2018 and 2060 in Coka watershed
LULC in 2018(Ha)
|
LULC conversion matrix in Coka watershed in 2060(Ha)
|
|
Bush land
|
Grass land
|
Forest land
|
CL &RS
|
Built Up
|
Bare land
|
Water body
|
Total in 2018
|
Loss (Ha)
|
Loss (%) in 2018
|
Bush land
|
108
|
125
|
65
|
360
|
5
|
6
|
16
|
685
|
577
|
27.1
|
Grass land
|
63
|
222
|
12
|
301
|
6
|
12
|
15
|
631
|
409
|
19.2
|
Forest land
|
51
|
0
|
309
|
466
|
10
|
4
|
19
|
859
|
550
|
25.9
|
CL & RS
|
39
|
123
|
342
|
844
|
36
|
51
|
0
|
1435
|
591
|
27.8
|
Built Up
|
0
|
0
|
0
|
0
|
23
|
0
|
0
|
23
|
0
|
0
|
Bare land
|
0
|
0
|
0
|
0
|
0
|
66
|
0
|
66
|
0
|
0
|
Waterbody
|
0
|
0
|
0
|
0
|
0
|
0
|
32
|
32
|
0
|
0
|
Total in2060
|
261
|
470
|
728
|
1971
|
80
|
139
|
82
|
3731
|
2127
|
100
|
Gain in 2060
|
153
|
248
|
419
|
1127
|
57
|
73
|
50
|
|
Gain(%)
|
7.2
|
11.7
|
19.7
|
53.0
|
2.7
|
3.4
|
2.3
|
|
|
Built up area
Built up area will increase from 23 ha (0.6%) in 2018 to 42 ha (1.1%) in 2040 and to 80 ha (2.1%) in 2060 by rate of increment 0.9, 1.9 and 1.4 ha/year during 2018–2040, 2040–2060 and 2018–2060, respectively (Fig. 6; Table 4). During 2018–2060, 5 ha of bush land, 6 ha of grassland, 10 ha of forest land and 36 ha of CL & RS with total of 57 ha (2.7%) will be transformed to built up area (Table 5). This divulges that expansion of built up area (urbanization) will be occurred at the expense of other land use types during 2018 − 260.
Bare land
It is predicted that bare land will increase from 66 ha in 2018 to 93 ha in 2040 and to 139 ha in 2060 (Fig. 6; Table 4). The rate of increment during 2018–2040, 2040–2060 and 2018–2060 periods will be 1.2, 2.3 and 1.7 ha/year, respectively. As shown in Table 5, 6 ha of bush land, 12 ha of grassland, 4 ha of forest land and 51 ha of CL & RS with total of 73 ha (3.4%) will be transformed to bare land. This projection indicates that expansion of bare land will be happened during 2018–2060 which could be associated to land degradation, low crop productivity, climate change and soil fertility deterioration if management intervention is slow.
Water body
An increment in water body from 32 ha (0.9%) in 2018 to 49 ha (1.3%) in 2040 and to 82 ha (2.2%) in 2060 with changing rate of 0.8, 1.7 and 1.2 ha/year during 2018–2040, 2040–2060 and 2018–2060 periods respectively, is projected (Fig. 6; Table 4). Thus, during the mentioned periods, 50 ha (2.3%) of other land categories will be changed to water body. In this regard, 16 ha of bush land, 15 ha of grassland and 19 ha of forest land will be transformed to water body (Table 5). This finding divulges that expansion of water body will be happened during 2018–2060 periods which could be opportunity for local community to use water for crop irrigation and harvest fishes and also it could cause for biodiversity loss due to displacement of wild animals from their homestead and loss of medicinal and indigenous plants in the Coka watershed.
3.4. Accuracy assessments
The accuracy and kappa statistics result of the three decades classified images are presented in Table 2. An overall accuracy of 87, 85, 88 and 91% were obtained for the classified images of 1988, 1998, 2008 and 2018 years, respectively (Table 7). Producer’s accuracy (PA) of the individual types of the four classified maps varied from 77% (bush land) in 2008 to 100% (water body) in 2018 and User’s accuracy (UA) was highest for the water body (95%) in 2018 and lowest for the grassland (79%) in 2008. A kappa coefficient of 0.84, 0.83, 0.87 and 0.89 were resulted for 1988, 1998, 2008 and 2018 LULC maps, respectively. Hence, the results indicated strong agreement of the classified image with ground truths.
3.5. Discussion
Generally, during 1988–2018 periods, cultivated and rural settlement land, built up area, bare land and water body have revealed an increment but forest land, grassland and bush land have shown a decrement in size in the Coka watershed (Table 2). According to Kindu et al. (2013) and Solomon et al. (2018) accuracy rating, accuracy of the study watershed result is reasonably high range of LULC changes classification accuracies with ground truth. The finding of this study is consistent with other studies reported by Temesgen et al. (2014) in dera District of north westen Ethiopia; Solomon et al. (2014) in Birr and Upper-Didesa watershed of Blue Nile basin, Gessesse and Kleman (2007) in South Central Rift Valley Region of Ehiopia; Rientjes et al. (2011) in Upper Gilgel Abbay catchment of Blue Nile basin; Gebremicael et al. (2013) in Blue Nile basin; where the cultivated and rural settlement land, built up area and bare land dramatically increased whereas forest land, bush land and grass land were shrinking in the Coka watershed. Ebrahium and Mohamed (2017) in Geleda catchment and Solomon et al. (2014) in Blue Nile basin also reported the growth of cultivated land at the reduction of forest cover in the respective study periods. In the same way, grassland declined by 32% between 1985–2010 periods in Metema watershed, Ethiopia (Binyam et al., 2015). Similarly, forest land showed decrement in the present study area which was agreed with the result of kindu et al. (2013) and Girmay et al. (2010) who reported a fast decline of forest cover and increment of other land covers in Munessa-shashemene and Gum selassa, respectively.
The projected LULC change showed that cultivated and rural settlement land, built up area, bare land and water body will be increased but forest land, grass land and bush land will be decreased in different proportions in Coka watershed (Table 4). On top of this, Temesgen et al. (2017) reported that projected cultivated land and built up area will be increased during 2030–2045 at the expense of forest land, bush land and grass land, Ethiopia. As reported by Sinan et al. (2014), bare land will be raised in 2030 by 31.75% in Tikrit, Iraq.
The conversions of one LULC category to another during 1988–2018 and projected during 2018–2060 period are presented in Tables 3 & 5, respectively. Transformations of LULC from one category to another are clearly occurred in the Coka watershed at the specified three periods and projected four periods. Generally, during the three periods the transformations of cultivated and rural settlement land from bush land and grass land are higher than any other categories conversions in Coka watershed. In agreement with this study, Hailemariam et al. (2016) showed that, 92, 577 ha of bush land converted into cultivated land between 1985 and 2016 in the Bale Mountain Eco-Region of Ethiopia. Similarly, cultivated land, built up area and bare land increased by 19.9%, 35.1% and 28.6% in expense of forest land, grassland and bush land during 1973–2007 periods in the Modjo watershed (Berhan et al., 2014) and by 12.5%, 1677.8% and 50% during 1957–2017 in Wanka watershed (Wondwosen et al. 2018). However, such trend of forest land, grass land and bush land loss is not uniform throughout the country indicating land use and land cover dynamics are socioeconomic, culture and site dependent (Girmay et al., 2010). The projected cultivated land, built up area and water body will be expanded at the expense of forest land, grass land and bush land in the Coka watershed which is in agreement findings reported by Temesgen et al. (2017) in Ethiopia and Muhammad and Han (2019) in Indonesia.
The expansion of cultivated land over non cultivated land observed in past and present land use even into the marginal and steep slope land of Coka watershed during 1988–2018 period (Fig. 5). There was mismanagement of natural resource such as overgrazing, deforestation and expansion of cultivated land into the marginal land as well as steep slope for the survival of their livelihoods which is similar with reported by (Minale and Rao 2012a, b; Amare 2013; Asres et al. 2016). Despite the expansion of cultivation from sloping into steeper slope with inappropriate soil and water conservation measures, crop production is still lagging behind by 2.67 % annual population growth rate (Asres et al. 2016). Population growth is one of major driving force for land use and land cover change in the Coka watershed. Similarly, Population growth exerts massive pressure on land use and land cover change, agricultural productivity, and the use of ecosystem in Ethiopia (Minale and Rao 2011). Almost all steep slopes, even that > 60% are under cultivation (Fig. 5). No special protection mechanisms exist except some traditional ditches, which in most cases aggravate runoff, especially on steep slopes. Because of these processes, the area is heavily threatened by land degradation. Amare and Rao (2012) have shown that in Ethiopian highlands, rural poor households have caused land degradation.
According to FGD and KII respondents, major driving forces for LULC changes in Coka watershed are population growth, expansion of cultivated land, fuel and construction material need, socioeconomic factors, police and instructional changes and expansion of built up area (urbanization) which is reliable with the findings reported by (Yesuf et al. 2015 and Warra, et al. 2013). This may become with very severe environmental degradation in the Coka watershed in near future. Communities have been under food security problem for many decades and have tried to adapt to changing conditions. However, since their efforts have not been systematically supported by stable institutions and policies, the farming system currently appears to be in a very critical condition.