Tradeoffs of Small Scale Irrigation and Its Contribution to Land Use and Land Cover Change in Mai-Dimu, Tahtay Koraro Wereda, North-Western Tigray, North Ethiopia

Improving and understanding of land use and land cover change (LULC) can help in projecting future land use dynamics and provide appropriate interventions for achieving better land management. The aim of this study is to evaluate the status of induced small scale irrigation practices that affect the different land use changes over time in mai-dimu Kebele, Tigray, northern Ethiopia. Remote Sensing (RS), Geographic Information System (GIS) were used to determine the LULC dynamics with its land cover changes (19952015) by dividing in to three decades. In analyzing the accuracy assessment, the Kappa coefficient was found strong agreement between classified land cover classes and observed land cover/use with greater than 80% values. The coverage of cultivated land has high land use map cover with 28.45%, 31.83% and 27.74% in 1995, 2005 and 2015 respectively. No irrigation practice was observed in 1995 and 2005. However, in 2015 it was covered with 1.65 % of irrigated land. While the overall change difference from the year 1995 to 2015, was also by enlarge attributed to expansion of settlement, dam, cultivated land and irrigated lands increased positively with 700.20 ha, 124.02 ha, 33.48 ha and 181.98 ha respectively which subsequently decrease the land use of grass land (−336.48 ha), bush land (−561.52 ha), bare or rocky land (−68.94 ha) and forest land (−343.03 ha). Hence, inducing the irrigation practices could be additional yield production under dry season which later helps in improving the lively hood of the community.


Introduction
Encouraging small scale irrigation (SSI) agriculture is vital to enhance production and attain food self sufficiency in Ethiopia. The success of soil and irrigation water management to maintain soil quality depends on the understanding of how soils respond to agricultural use and practices over a certain time [1].
Considering the rapid growth of the world's populations, which is in its turn a limiting factor to the arable lands around the world, the need for effective and efficient application of the crop lands have been felt more than ever [2] [3].
Hence, much attention is given to sustainable land use and improve technology, where a given land use of an area is the best in increasing crop production through supplemental irrigation [4]. Therefore, for appropriate land use and water management in irrigated area, knowledge of the chemical composition of the soil characteristics, water, climate, drainage condition and irrigation methods should be evaluated for irrigation expansion [5].
Irrigation projects mainly small scale irrigations schemes can have several environmental and social importance that may lead to the sustainable production of agricultural goods, which is of major importance and interest in the development of Ethiopia since it contributes 44% to Ethiopia's GDP, employs 80% of the labor force, and provides a livelihood to 85% of the nearly 80 million, population [6].
For different parts of Ethiopia, land use and land cover changes were studied from small scale to large scale. For example, (West Ethiopia) [7]; (North-western Ethiopia) [8]; (North Ethiopia) [9] [10] [11] [12]. All these studies show that agricultural land has expanded at the expense of natural vegetation, including forests, grazing land and shrub lands. But there is no study conducted that revealed the expansion of irrigation and its effect on the land use changes over time. Therefore, the objective of this study was to evaluate the impact small scale irrigation practices that affect the different land use changes over time in Maidimu, Tahtay-koraro Wereda, North-western Tigray.

Description of the Study Area
Mai-dimu micro dam ( Figure 1) is found in Tabia Mai-dimu, Tahtay-koraro Wereda, North-western Tigray. It is located at 15 km west of Shire at 14˚15'40" -14˚59'00" latitude and 38˚10'35" -13˚15'40" longitude with an altitude of 2010 masl [13]. The total area of the study area (Mai-dimu and Adi-gebro) cover about 11,201.49 ha. The area lies in dry wena dega agro-climatic with diverse topographic conditions characterized with undulating terrain having step and gentle slopes. Soils are predominantly dark brown in the middle highland area and light brown and grayish color in the low land area. The mean annual air temperature was 28˚C, the maximum temperature reaches its pick during the month of April and May and annual precipitation ranges varies from 600 mm to 900 mm however there was 520.25 mm [14]. The rainfall season is three to four S. Fissaha et al. months from June to September; and more than 85% of the total precipitation rains within these months. The rainfall is erratic in nature which imposes early, mid or late season droughts.

Data Collection
The location of each data points were obtained by field surveys and actual measurements using global positioning system (GPS) and camera. Secondary data were also be obtained from previous CoSAERT study reports and the nearby wereda administrative bureau for examining the history of the farming practices, type of crop grown and the area covered for irrigation after the project intervention (dam construction). In addition, various sources such as aerial photographs, topographic maps, and satellite imageries were used to generate additional data.
In addition, for collecting high quality geographic data for input to GIS, topographic maps of the scale 1:50,000 of the study areas were purchased from the Ethiopian Mapping Agency (EMA) and different satellite maps of the year 1995 using TM, 2005 using ETM + and 2015 using land sat 8 OLI sensors with 159 path and 50 raw having 7 bands and a pixel size of 30 m × 30 m resolution imageries were down loaded from www.earthexplorer.usgs.org.

Digital Image Processing
Detailed survey were conducted in order to obtain accurate location point data for each land use change classes using the ground control points (GCP) in order to obtain the qualified and well sounded information from satellite image data for appropriate land use classification.
During data processing both data processing and interpretation were made systematically. Image interpretation phases were preceded by establishing preliminary legend [15]. While, image processing was done by two techniques called image rectification and restoration and image enhancement depending on the required correction of radiometric distortions, geometric distortion and noise [16] [17].

Image Classification
Image classifications were used in converting image data into thematic data [18].
For this study, both types of image classification systems were used named as the unsupervised classification before field visit and the supervised classification after field survey [19]. Accordingly, representative points that represent the various land cover classes were marked using GPS during field visit. These points were used to sample representative signatures for the various land cover types as (Table 1) identified during field visit and also helped for determining level of accuracy assessment.

Accuracy Assessment Matrix
Accuracy assessment matrix was employed to evaluate the accuracy of the classification. Based on a Rule of thumb devised by [20], we have taken 30 sample points for each land use for calculating the error matrix and these points were collected from the ground truth using the GPS and cross checked with the Google earth in acquiring the reliability of the produced map.

Data Processing and Method of Producing Temporal Mapping Techniques
The approach was done based on a combination of digital classification and visual interpretation of the images. The land use dynamics were analyzed with the continuous image analysis starting from 1995 to 2015 considering month of March or April because these months are good representative for the dry season (irrigation condition). These representative months were also supported using the ERDAS 9.2 approach which this is summarized in schematic diagram in Figure 2 shown below.

Data Analysis
The digital remote sensing data were analyzed, processed and geo-referenced in ERDAS imagine 9.2 software and also used for image processing to develop land use maps. Arc-GIS 10.1 was used for creating different map layers.

Accuracy Assessment for Mai-Dimu and Adi-Gebro Kebeles
The necessary elements mainly the producer's, user's, overall accuracy and the Kappa statistics were computed. In general, the maps of the study kebeles met more than 85 percent over all accuracy (Table 2). This agreed with [21] and [22] that stated as all the output maps produced have to meet the minimum 85% overall accuracy. Moreover, the kappa coefficient of the Mai-dimu kebele (combination of Maidimu and Adi gebro kebeles) was found to be 0.81, 0.84 and 0.86 for the years of 1995, 2005 and 2015 respectively. Hence, based on these kappa coefficient results, the study kebeles have strong agreement that determines the usefulness of the map. In which this statement is supported by [23] that stated as land cover accuracy is commonly defined as the degree to which the derived classification agrees with reality and the accuracy of the map in a larger part determines the usefulness of the map.

SSI Growth Trend Analysis and LULCC in Mai-Dimu and Adi-Gebro
Eventhough the study area was initially focused on the mai-dimu kebele, we have faced one additional Kebele namely Adi-gebro ( Figure 3). Due to, Maidimu Kebele is the source of the dam called Mai-dimu earthen dam while Adigebro Kebele is the beneficiary of the water for irrigation purposes. Therefore, the land use maps that temporarily produced were of Adi-gebro and Mai-dimu Kebeles.
Comparing the result of the study with the land cover map (Figure 3 and Ta     Overall ( Figure 3 and Table 3) illustrated as, there is a reduction in bush land and forest land across the three mention period of land use land cover changes.
This revealed that land degradation and deforestation were seriously affecting the two kebeles. As a result, the level of land productivity may decline at a faster rate and these kebeles may not in a position to sustain the annual food demand of the people. This idea strengthen by [25] that clarified as, land is also severely degraded due to unwise utilization of land resources, soil erosion, soil nutrient depletion, and soil moisture stress are the major land degradation problems that directly affect the livelihood of the society. The land use that is covered with settlement lands increased across the three mentioned years. Therefore, the population number is still increasing from time to time which may affect the different land use in changing from one land use to the other. This concedes with [27] that stated as land cover changes are caused by a number of natural and human driving forces. Moreover, change made in LULC by population pressure can also affect biodiversity, contribute to forest fragmentation, lead to soil erosion, alter ecosystem services, and increase natural disasters such as flooding [28].

LULC Detection Matrix and Its Conversion in Adi-Gebro and Mai-Dimu Kebeles
As evident from ( In Table 5 shown below, land use detection matrices have also been no  to 2015 which irrigation practices were implemented in these 10 years difference.
Overall land use change matrices ( where it needs more attention for using sustainably for long period of time. According to Bureau of water resource, Tigray regional office, report showed that Mai-dimu earthen dam was constructed in 2008.

Contribution of SSI to LULCC Difference in Mai-Dimu and Adi-Gebro
According to Figure 4, it revealed the amount of hectares that came due to the change difference from the year 1995 to 2005. Hence, the change that was difference in LULC in the study area was by enlarge attributed to expansion of settlement and cultivated land increases positively with 322.56 ha and 333.72 ha respectively. Which this implies, agricultural expansion and the increment of

Conclusion
Land use changes and their associated management can influence soil properties and their agricultural productivity, though the amount of changes could be varied depending upon the extent of human management in all land use areas. Over all, expansion of cultivated and settlement land were the major land use and cover changes observed. On the other hand, irrigation expansion was occurred in the years of 2005-2015 land use change time frame which can show as the irrigation expansion in this study site was found at its infant stage. This irrigation expansion was held as the expense of cultivated land, bush land and forest land. The consequence of these conversion and modification processes of the land use and land cover to irrigation land and dam construction could help in increasing the additional yield production under dry season which later helps in improving the lively hood of the community living around the study sites.