Für neue Kunden:
Für bereits registrierte Kunden:
List of Tables
List of Figures
1.2 Statement of the Problem
1.3. Purpose of the Study
Chapter II: Literature Review
2.1 Invasive Alien Species
2.2 P. juliflora
2.3 Concepts of Remote Sensing and Geographic Information Systems
Chapter III: Materials and Methods
3.1 Description of the Study Site
3.2 Material Used
3.4 Limitation of the Study
Chapter IV: Result
4.1 Accuracy Assessment of LULC Classification
4.2 Land Use/Land Cover Change Detection
4.3 Ecological and Socio-Economic Impacts
4.4 Management of P. juliflora Invasion
Chapter V: Discussion
5.1 Land Use/Land Cover Change Detection
5.2 Ecological and Socio-Economic Impacts
5.3 Management of P. juliflora Invasion
Chapter VI: Conclusions and Recommendations
P. juliflora is a powerful invader equipped with a number of special features that promote its rapid invasion of new areas. This invasive species is currently becoming a problematic weed in Ethiopia, especially in Afar Regional State. Land use land cover changes (LULUC) in relation to P. juliflora invasion that occurred from 1973 – 2004 in about 527.90 km2 areas in Amibara District of Afar regional state were analyzed. The objective of the study was to analyse the invasion rate of P. juliflora through detection LULUC using multi-temporal Landsat imagery data obtained in 1973, 1987, 1999 and 2004, and field survey were conducted to analyse impacts of invasion, to identify management techniques which reduce the invasion and re-invasion of P. juliflora and to fill the information gap that arise from the imagery analysis.This study then reveals, over the period of 31 years, LULCC in relation to P. juliflora invasion was changed significantly. The annual invasion rate of P. juliflora during 1973-1987,1987-1999 and 1999-2004 was 5.61 km2/annum ,0.80 km2/annum and 3.95 km2/annum respectively. The average annual invasion rate was 3.45km2/annum. This study also indicated that, 121.66 km2 (23.04%), 128.56 km2(24.35%),145.81 km2(27.62%) and 163.064 km2(30.89%) of the study area would be under P. juliflora by the year 2008, 2010, 2015 and 2020 respectively, if proper controlling measures are not taken without delay. Invasion of P. juliflora causes loss of biodiversity and ecological service, affect livelihood of the pastoralist and agro-pastoralist, crop production and human health. If properly managed it shows positive implication on the soil amelioration, micro-climate and income generation for charcoal makers. But the pastoralist and agro-pastoralist call for its eradication because under the current management system the disadvantages outweigh the advantages. Management techniques explored was clearing and digging out of the root system below 10-30cm from the ground ,seedling uprooting with continuous follow-up during wet season, utilization of P. juliflora as a resource by making charcoal, pod collection and crushing; since livestock is the main dispersal agent of the species. To ensure the successfulness and sustainability of management techniques, commitment, participation and continuous follow up of the community and different stakeholders should be given more focus to control the invasion and re-invasion of P. juliflora.
Keywords: P. juliflora; Land use/land cover change; Satellite Imagery; GIS; Amibara; Ethiopia.
First of all Glory be to GOD for making it all happen. I would like to extend my sincere gratitude to Dr. Nigussie Haregeweyn and Dr. Kindeya Gebrehiwot who guided and supported me to successfully complete the present research work. My deepest gratitude is extended to Alemayehu Eshetie for his invaluable support and guidance through every stage of my thesis. A special thanks is extended to Ashebir Kiflie for his guidance and material support.
My acknowledgement will not be complete without mentioning the comfort and care I procured from colleagues and friends on my sadness when I and my family were hit by a sudden loss of my lovely mother in the middle of the thesis research work.
I also extend my heartfelt thanks and appreciations to my brother Melaku Negash for his spiritual and moral support during all my life time. I am highly indebted to Ato Tibebu Kassawmar for his continuous help in finalizing the GIS work and for his material support and for providing satellite imagery. I like to express my profound gratitude to Worer Agriculture Research Institute and Awash Basin Authority for their valuable information and transport support. My sincere thanks and respect also goes to all people and institutions who in one way or another support or helped me during my study.
Hana Alex, Solomon Zewdu, Mohammed Amin, Muktar Siraje, Ferid Shewarega, Frew Abebe, Andinet Getachew, Dr. Zerihun Yemanebirhan, Kassaye Dejene and Negash Tamiru; I thank you for the consistent moral support and brotherly gesture which I received during the study period.
Last but not least is the heartfelt thanks to my lovely mother the late Tenagnework Tessema and my sister Haregeweyn Negash for moral, spiritual support and care during my childhood up to now without them I could have not reached to this level; may GOD pay them back!
ABA Awash Basin Authority
EMR Electro Magnetic Radiation
ENVI Environmental for Visualizing System
ERDAS Earth Resources Data Analysis System
ETM+ Enhanced Thematic Mapper Plus
FARM-Africa Food and Agricultural Research Management-Africa, UK based International Non-Governmental Organization
FAO Food and Agricultural Organization
FCC False Colour Combination
GCPs Ground Control Points
GIS Geographic Information System
GOs Governmental organizations
GPS Global Positioning System
IAS Invasive Alien Species
IR Infra Red
Kebele the smallest administrative unit in the government structure which covers several villages
LULC Land Use/Land Cover
LULCC Land Use/Land Cover Change
m a.s.l meter above sea level
MOA Ministry Of Agriculture
MSS Multi Spectral Scanner
NDVI Normalized Difference Vegetation Index
NGOs Non Governmental Organizations
NIR Near Infra Red
TM Thematic Mapper
RS Remote Sensing
SPSS Statistical Package for Social Science
WARI Worer Agricultural Research Institute
Table 1. Description of satellite image used in change detection in relation to P. juliflora
Table 2. The description of LULC classes used for change detection in the study area
Table 3. Error/confusion matrix for 1973 LULC classification of the study area
Table 4. Error/confusion matrix for 1987 LULC classification of the study area
Table 5. Error/confusion matrix for 1999 LULC classification of the study area
Table 6. Error/confusion matrix for 2004 LULC classification of the study area
Table 7. LULC classes, their corresponding areas and change of 1973 – 1987 in Amibara
Table 8. LULC classes, their corresponding areas and change of 1987 – 1999 in Amibara
Table 9. LULC classes, their corresponding areas and change of 1999 – 2004 in Amibara
Table 10. Major and supplementary source of income by main occupation in Amibara
Table 11. Positive contribution of P. juliflora for selected land rehabilitation parameters
Table 12. Types of disease/injury caused by P. juliflora in Amibara
Table 13. Other impact of P. juliflora on community in Amibara
Figure 1. Map of the study area
Figure 2. Over all methodology of the study
Figure 3. False color composite image of the study area
Figure 4. Amibara’s NDVI images of 1973, 1986, 1999 and 2004
Figure 5. Amibara’s land use/Land cover map of the period 1973
Figure 6. Amibara’s Land use/Land cover map of the period 1987
Figure 7. Amibara’s land use/Land cover map of the period 1999
Figure 8. Amibara’s Land use/Land cover map of the period 2004
Figure 9. P. juliflora reduce soil erosion by well developed root system in Amibara
Figure 10. Some pastoralists shifts livestock production to crop production as thier main livelihood activity at the study area
Figure 11. In Amibara the shade and fences from P. juliflora
Figure 12. Pruned stand of P. juliflora at Ginnery Amibara
Figure 13. The cattle stands, nothing to graze because of P. juliflora invasion in Amibara
Figure 14. Plastic roof covered Afar's hut at the study area
Figure 15. P. juliflora invasion induce land degradation in Amibara
Invasion by exotic species is one of the main threats to the conservation of biodiversity and habitat. Once an invasive species becomes firmly established, its control can often be difficult and eradication is usually impossible. In addition, the impact on natural communities, ecosystem processes and socio-economic conditions can be very lethal (Stefan, 2005).
Invasive species alter the biogeochemical cycles and act as competitors, predators, parasites or pathogens of the native species placing their survival at risk (Raghubanshi et al., 2005). According to Lonsdale (1999), explanation, the actual invasion of an environment by new species is influenced by three factors: (1) the number of propagules entering the new environment (propagules pressure), (2) the characteristics of the new species, and (3) the susceptibility of the environment to invasion by new species (invisibility), which is depending on either disturbance or unused resources by the resident species or over supply of nutrients due to other effects. Invisibility is an emergent property of an environment, the outcome of several factors, including the regional climate, the environments disturbance regime, and the competitive abilities of the resident species (Lonsdale, 1999).
Biophysical materials like forests are often subject to rapid change when subject to invasive species. To fully understand the changes taking place, it is important that such changes be detected and quantified accurately. Accurate and up-to date data describing land use and land cover changes support these studies in a number of ways. Among tree species that have become widespread weeds in Ethiopia is the Prosopis juliflora (Swartz) DC. here after called as P. juliflora.
Because of the impacts of biological invasions and the difficulty of eradicating an exotic species once it has established, it is important to develop prospective work that allows the detection of invasions in their initial stages. A fundamental component of this strategy consists in identifying those areas that are more prone to be colonized, in order to optimize monitoring actions and then mitigating its uncontrolled invasion (Reichard and Hamilton, 1997).Thus, the susceptible area for exotic invasive species (P. juliflora, in this case) in Ethiopia is now the Afar Regional State.
According to Alemayehu (2006), P. juliflora was introduced by Awash Valley Development Authority in middle Awash to control desertification and the high dust wind in area during 1970s. Currently, P. juliflora occupies prime dry season grazing lands in Afar Region creating disappointment among Afar pastoralists.
P. juliflora is a powerful invader equipped with a number of special features that promote its rapid invasion of new areas (Shiferaw et al, 2004). This invasive species is currently becoming a problematic weed in Ethiopia, especially in Afar Regional State. P. juliflora is expanding at an alarming rate, and it is invading large scale farms, rangelands, riverbanks (Kassahun, 1999). Due to its invasive nature and suppression of the herbaceous layer, this woody species is viewed by all small scale irrigation practitioner, investor, researchers and pastoralists or landowners as a serious threat (Alemayehu,2006).
In the study area, there are no research document as to its invasion status, on ecological and socio-economic impacts and management techniques.
Therefore, this study attempts to fill this gap in Amibara District of Afar Region, which may contribute to the sustainability of the area and hence the betterment of the livelihoods of communities in the study area.
Resource managers require spatial and temporal information in order to make decisions that would be sound in managing this invasive species. Hence, it is essential to investigate the invasion rate, its impact and management options of P. juliflora so that the information generated may be used as a starting point for future interventions.
Therefore , the result is expected to be used by researchers, policy and decision makers, development agents and institutions who are concerned with natural resources management in pastoral areas.
The hypothesis is that, the invasion rate of P. juliflora is causing significant negative ecological and socio-economic impacts in Amibara District.
The general objective of the study was to analyse the invasion rate of P. juliflora and its associated impact at Amibara District.
The specific objectives of this research were:
- To analyse current invasion rate and future trends of P. juliflora.
- To analyse associated impacts of P. juliflora invasion.
- To identify possible management options in order to minimize expansion of P. juliflora.
Invasive Alien Species (IAS) are non-native or exotic organisms that occur outside their natural adapted ranges and dispersal potential (Raghubanshi et al., 2005). Many alien species support our farming and forestry systems in a big way. However, some of the alien species (P.juliflora in the case of Afar) become invasive when they are introduced deliberately or unintentionally outside their natural habitats into new areas where they express the capability to establish, invade and out-compete native species. The threat to biodiversity due to invasive alien species is considered second only to that of habitat destruction. Invasive species cause loss of biodiversity including species extinctions, and changes in hydrology and ecosystem function. Differences between native and exotic plant species in their requirements and modes of resource acquisition and consumption may cause a change in soil structure, its profile, decomposition, nutrient content of soil, moisture availability, etc. Invasive species are thus a serious hindrance to conservation and sustainable use of biodiversity, with significant undesirable impacts on the goods and services provided by ecosystems. Biological invasions now operate on a global scale and will undergo rapid increase in this century due to interaction with other changes such as increasing globalization of markets, rise in global trade, travel and tourism (Raghubanshi et al., 2005).
Geesing et al. (2004) added that IAS are species that are non-native to a particular ecosystem and whose introduction causes, or is likely to cause, economic or environmental harm, and characterized by rapid growth rates, extensive dispersal capabilities, large and rapid reproductive output and broad environmental tolerance.
P. juliflora is an evergreen tree native to northern South America, Central America and the Caribbean (Pasiecznik et al., 2001). It is spiny, prickly or armed shrub/tree, fast growing and has the ability to develop extensive and deep root systems, sometimes exceeding 20-25m (Jorn, 2007). Mature P. juliflora is a medium size shrub or spreading, short-trunk trees, which in favorable growing conditions can develop into a tree 20-m in height and more than 1-m diameter (Jorn, 2007; Pasiecznik et al., 2001). Mature trees have been found to radiate to 10-m elsewhere-lateral roots from the trunk (Agrawal, 1996).
Due to their wide ecological spectrum, P.juliflora occurs on a large variety of soils and over a wide range of altitudes. They endure temperature as high as 500C, and resist occasional frost of –120C (Mohamed, 1997). The ecological adaptation to the hard conditions of dry land is indicated by several mechanisms. Of these, drought adaptation including deep rooting systems, stomatal control of water loss (Nilsen et al., 1983).Many species of Prosopis also exhibit a high degree of salt tolerance, as high as 3.3% NaCl. It also grows luxuriantly up to saline of soils having electrical conductivity of 15 d S/m, which is a concentration of more than 10 times greater than most annual crops can resist (Silva, 1990).
P. juliflora has the ability to thrive in hostile climatic and edaphic conditions. It grows even in wasteland and rocky terrain, requires little water, fixes nitrogen and then has fast growth rate and becomes successful invaders in dry areas. The uncontrollable spread of the tree may be due to the prolific production of seeds, high germination potential of the scarified seeds through feeding of pods by livestock and wild animals (Shiferaw et al., 2004).
Many species of Prosopis are known to nodulate and fixing nitrogen in an endo-symbiotic relationship with root nodule bacteria (Silva, 1990). Pasiecznik et al. (2001) added that the ameliorating effects involve reducing soil salinity, neutralizing alkaline soils and improving soil nutritional status and physical properties. These are primarily due to complex interactions between the effects of nitrogen fixation, incorporation of leaf litter, changes in microclimate, and changes in the floral, soil faunal and soil microbial populations. P. juliflora considerably improved soil properties by decreasing soil pH, electrical conductivity and exchangeable Na levels, and increasing organic matter, total N and available P (Mendes, 1990; Abebe et al., 2006).
In areas where P.juliflora emerges, grasses, trees and shrubs will disappear and it naturalized itself instead (Pasiecznik et al., 2001). It is also reported that P.juliflora has been shown inhibit the germination or growth of many plant species growing in its vicinity through allelopathic substance exuded from its leaves, root and fruits. The exuded allelopathic (L-tryptophan) substance from P.juliflora leaves plays an important role on the experiments conducted on Bermuda-grass (Cyanodon dactylon), which suppresses its development (Nakano et al., 2003).
Ingestion over long periods of time would result in death of cattle. If P.juliflora pods are the sole food source for cattle, ca. 1% become sick, and some die with compacted pod ball in the rumen. Death is attributed to high sugar content repressing the rumen-bacterial cellulose activity. Moreover, the pollen may cause allergic rhinitis, bronchial asthma and/or hypersensitivity pneumonitis. Its effect on livestock, when cattle feed unripe fruit, their neck becomes twisted and some will die. It makes undigested ball matter inside their stomach and some seeds have observed germinating in their rumen when the cattle die with this effect (Nakano et al., 2001; Shiferaw et al., 2004). P. juliflora leaves fed to goats were reported to suppress rumen microbes and increase levels of rumen ammonia and blood urea (Pasiecznik et al., 2001). The spine attacks camel, goat, cattle, and sheep’s hooks and which prevent their movement in search of food that in turn reduces milk production and sickness of the livestock, and their wound do not heal soon. The results of P. juliflora invasion in the Ng’ambo area of Kenya resulted in livestock moving long distances in search of pasture, sometimes up to 50 kilometers from their original grazing areas (Esther and Brent, 2005).
Pasiecznik et al. (2001) show that, the roadsides also become narrow, irrigation canals and ditches, rangelands, farmlands and other infrastructures become jeopardize as a result of P. juliflora invasion for instance the other impact, to the other side of people is the cost they pay to the daily laborer and the bulldozer to clear their farmlands every sowing seasons. On using bulldozer, the spines penetrate the tires and affect its efficacy and this also exposes the farmland owners to extra costs. Wildlife stocks like warthogs, hyenas, etc are reproduced inside the thickets of P. juliflora and they compete on grasses and water sources, and even they attack and hunt domestic animals. (Shiferaw et al., 2004)
The most prominent benefit of P. juliflora to the communities, in addition to wind speed decrement, is provision of wood fuel for the households, charcoal making and fence. Selling P. juliflora firewood and charcoal is an important economic activity. Records of commercial production of charcoal and firewood in 1996/97 from Gash and Atbara rivers were 600,000 sacks and 135,000 m3, respectively (Elsidig et al., 1998 cited in Jorn, 2007).
The wood is also used as an industrial fuel for oven and kilns in small scale industries. Rather than firewood, the wood is also used as charcoal, reducing weight and increasing the energy and economic value of the product, compared to firewood. Bigger trunks and upper root stocks of P. juliflora are burnt to make charcoal. The wood of P. juliflora i s also used as timber. Fence posts, poles, particle boards and cardboard are manufactured using wood of P. juliflora. Trees with bigger boles are good for making furniture. Pods of P. juliflora have been used as animal feed for cattle, sheep, goats, Camels and horses. The pods are generally consumed raw by animals. Average annual pod production P. juliflora in India was assessed to be 20kg/tree (Tewari et al., 2000).
The pods serve as a nutritive source of animal feed with 10% protein, 14% fiber, 55% soluble carbohydrates, 0.20% Calcium and 0.15% Phosphorus. The pods are also processed in a disc-mill to prepare for flour inclusion in livestock ration (Pasiecznik et al., 2001). Superior honey could be obtained since it provides good bee pasture and nectar as it is abundant and occurs in the dry season of the year when almost none of the native vegetation has any flower. Due to this and other characteristics, it is one of the prime plants for supporting apiculture in semiarid zone (Mendes, 1990).The gum exuded from the sapwood is used in paper and cosmetic industry (Wick et al., 2000).
The increment of P.juliflora invasion is the result of several factors. A major reason for the spread of P.juliflora is that livestock eat the palatable mature pods. The presence of animals is usually important for P.juliflora to be dispersed over long distance and to germinate (Geesing et al., 2004). Shiferaw et al. (2004) reported that one kg of goat and cattle droppings contained, on average, 760 and 2,833 P. juliflora seeds, respectively, thus suggesting that cattle are the primary pod-eaters and thus dispersers of P. juliflora seeds. The passage through the digestive tract facilitates germination of the seed, which are later deposited with the faeces some distance away from the parent plant. The faeces might also serves as fertilizer to seedlings in an initial stage of establishment.
Another reason for fast spread is that P.juliflora produces many and small seeds that can form dormant seed banks. The seed can then germinate when favourable conditions appear through disturbances such as flooding or rainfall (Pasieczink et al., 2001; Geesing et al., 2004; Shiferaw et al., 2004). Mohammed (1997) indicated that in India P. juliflora produce 37 to 75 Kg/tree/yr. Also Alves et al. (1990), showed that around 10,000-20,000 seeds was measured a kg of P. juliflora seeds. Shiferaw et al. (2004) added that in Ethiopia 32,500-33,000 seeds measure a kg of P. juliflora seeds. This may be due to different in pod size, seed size and even plants’ variety. Flood is also an important dispersal agent for P.juliflora seeds. Heavy rain make seed fall from the trees and the following flooding ensure widespread dissemination of the seeds. Since P.juliflora has the ability to survive cutting and resprout with fast coppice growth, the species becomes a very strong invader. As the height of cut increases, the number of resprouts also increases and 50cm-cutting height produces higher number of coppices and also high in lengths of these coppices. When the number of stems increases the number of coppices increases (Shiferaw et al., 2004). This will help to derive some management scheme in mitigating the dispersal of this invading species in the area.
Prevention is a first priority option in the management of invasive P.juliflora, and the success is then measured in terms of invasions that did not happen and the costs involved. The next option is then to organize an early detection system, which has a vital role in the reduction of costs from invasion (Jorn, 2007).
In case an early detection fails and P. juliflora becomes established in wrong places, then eradication will often cease to be a viable economic option. P. juliflora is not always invasive in an environment, and therefore it is useful to have sufficient knowledge from other similar sites to decide on the intervention intensity in any given situation. A further option would be the technical and financial planning and the actual management of P. juliflora control, as a specific project. In case it is known that P. juliflora will create higher costs than benefits and the costs would be high also in monetary terms, eradication should be considered as an alternative. This should be attempted only if it has a chance to be successful. If eradication is not feasible, then the next step is containment of P. juliflora in the infested area by monitoring that it does not spread to a larger area (Geesing et al., 2004).
The size of the trees is an important factor when determining the options. With young seedlings it may be enough to spray them with herbicides or just burn them by using containerized gas equipment. Older trees need to be mechanically felled and cleared away before digging up the stumps with an excavator or human labour. The stumps need to be lifted up from the soil so that 30 cm of the root system below the ground level is cleared away, as the root collar has dormant buds which would otherwise sprout. Fairly successful killing of stumps has also been achieved through soaking the stumps in kerosene and burning them on the spot without lifting. Biological control has been tried in some cases, as there are several shoot borer beetles which are able to at least weaken the tree. A successful option to control P. juliflora is also to promote its heavy utilization (Geesing et al., 2004). Shiferaw et al. (2004), indicated that the cutting height of 10cm deep down to the ground do not produce any coppices until 3 months. This could be the reason that (1) underground cut affects the lateral root systems and the plant itself could not survive, and forget the resprouts, (2) factors of geothermal/or soil temperature (high-staffed temperature affect the young resprouts). Also Yirgalem (2001) comes up with the same result in that 5-10 cm deep cutting from the surfaces do not have any revegetation.
Better management of P. juliflora, different land use strategies and the exploitation of P. juliflora as a resource may reduce its invasiveness in some regions as well as improving local economies (Pasiecznik, 2002; Waweru, 2006). As obviously the simplest options to constrain its spreading would be through its utilization. It would be important to maintain a large scale utilization of P. juliflora that can control the stock to such an extent that an optimum amount of benefit would accrue from the species (Pasiecznik et al., 2001).
Weedy stands thinned to 100-400 trees per hectare, in stages. Broad strips cleared and cut stumps are removed manually/mechanically, treated by stripping the bark or treated with used motor oil/triclopyr and diesel mixture. Animals may re-enter immediately as these chemicals have little mammalian toxicity. Selected trees in the remaining rows are pruned to single stems at final spacing of 5 by 5 meters to 10 by 10 meters. The cost of the operation should be at least covered by the charcoal, wood chips and small timber obtained from operation (Pasiecznik et al., 2001).
Pruning appears to be the single most important technique in improving tree and understorey yields. Weedy shrubs are turned into valuable, productive trees by removal of side branches. Regularly pruned trees are found to have smaller root systems, use soil water more efficiently and compete less with neighbouring crops and grasses (Pasiecznik, 2002; Waweru, 2006). Pruning of P. juliflora is strongly recommended to increase tree growth and also to minimize negative influences on adjacent agricultural crops, maximizing complimentarity between tree and crop components of Agroforestry (Pasiecznik et al.,2001).Re-invasion can be minimized by maintaining a high-pruned tree canopy and improved understorey management. A reduction in stocking rates, for example, can encourage good grass cover, preventing P.juliflora seedling establishment because P.juliflora seedlings rarely establishing under mature trees or in tall grass (Esther and Brent, 2005). Destroying seed, by the collection and use of pods for stall feeding or processing, reduces re-invasion and a change of livestock from cattle to sheep or pigs (which kill most to all of seeds ingested) also limits P.juliflora spread (Pasiecznik, 2002). Also Geesing et al. (2004) indicated that crushing makes the protein in the seeds more available and at the same time destroys the seeds, preventing germination of new plants and thus contributing to containment of the P. juliflora invasion.
Remote sensing is defined as the use of electromagnetic radiation(EMR) sensor to record images of the environment which can be interpreted to yield useful information (Paul, 1985). A Remote Sensing Systems offer four basic components to measure and record data about an area from a distance by using EMR, which is crucial medium required to transmit information from the target to the sensor. These components include the energy source, the transmission path, the target and the satellite sensor (http://www.ucalgary.ca/geog/Virtual/remoteintro.html).
The amount and characteristics of radiation emitted or reflected from the Earth’s surface is dependent and different based upon the characteristics of the objects on the Earth’s surface. So that, different objects on the Earth interact with radiation at distinct way and knowledge of this interaction is fundamental issue on classification of satellite image. Therefore, it is possible to distinguish and classify based on the reflectance variation of Earth surface objects. The classification of a satellite image can be achieved by supervised or unsupervised procedures. A supervised approach relies on the prior specification of training areas by the analyst, in which major land cover types are delimited manually as a key for electronically classifying the image. It needs the knowledge of study area in advance. In contrast, no such visual interpretation is involved in an unsupervised method. It uses automated methods to cluster reflectance values in order to derive a required number of land classes and their associated spectral signatures (Lillesand and Kiefer, 1994).
Geographic information systems refer to a set of tools for collecting, storing, retrieving, transforming, and displaying spatial data from the real world for a particular set of purpose (Burrough et al., 1998 cited in Behailu 2006). By allowing data to be organized and viewed efficiently, by integrating them with other data for analysis and by creation of new data that can be operated in turn, GIS provides an added value to spatial data set. For use in GIS, remote sensing provides a spatial data such as aerial photographs and satellite images (Heywood et al., 1998 cited in Behailu 2006)
Also the merging of remote sensing, GIS and visualization techniques was applied to demonstrate the potential for realistic computer visualizations depicting the dynamic nature of forested environments. Scientific visualizations can aid in environmental and forest management decision making as a support tool and in landscape ecology to relay the findings of studies (Matt et al., 2004). One important method of understanding ecological dynamics, such as natural and human disturbances, ecological succession and recovery from previous disturbances, is the analysis of changing land cover patterns (Turner, 1990). Satellite imagery provides an excellent source of data for performing structural studies of a landscape (Forman, 1995).
For design of meaningful conservation strategies, comprehensive information on the distribution of species, as well as information on changes in distribution with time, is required. It is nearly impossible to acquire such information purely on the basis of field assessment and monitoring. Remote sensing (RS) provides a systematic, synoptic view of earth cover at regular time intervals, and has been useful for this purpose (Nagendra, 2001). The integration of these tools can increase the accuracy of forest analysis at different spatial scale.
Information extractions from digital images involve both visual interpretation and computer assisted techniques. Digital processing of data acquired from a remote sensor system basically falls in two classes: image restoration (Preprocessing) and information extraction techniques. The most common undertakings in image preprocessing include restoring missing scan lines, and correction for geometric and radiometric distortions (Mather, 2003 cited in Behailu, 2006). Computer assisted digital image processing to extract information involve image enhancement, image classification, and accuracy assessment. Digital image classification tends to fall in to one of the two operational classes; supervised classification and unsupervised classification (Kelley, 1983 cited in Behailu, 2006).
Supervised classification, to do this representative sample sites of known cover types, called training areas are used to compile a numerical “interpretation key” that describes the spectral attributes for each features types of interest. Each pixel in the data set is then compared numerically to each category it “looks most like”. There are a number of numerical strategies that can be employed to make this comparison between unknown pixels and training set pixel. There are three basic steps in supervised classification; in the training stage the analyst identifies representative training areas and develops a numerical description of the spectral attributes of each land cover types of interest in the scene, in the classification stage compare each unknown pixel to spectral patterns; assign to most similar category and in output stage present results as maps, tables of area data and GIS data files (Lillesand and Kiefer, 1994). Supervised classification requires knowledge of the area at hand. If this knowledge is not sufficient available or the classes of interest are not yet defined, an unsupervised classification can be applied.
Unsupervised classification, the image data are first classified by aggregating them in to the natural spectral grouping, or clusters, present in the scene. Then the image analyst determines the land cover identity of these spectral groups by comparing the classified image data to ground reference data. The main purpose is to produce spectral grouping based on certain similarities (Lillesand and Kiefer, 1994). But unlike the case of supervised classification which deals with thematic classes, unsupervised classification deals with spectral classes. Unsupervised classification resulting in clusters requires to be given some physical meaning and can be used as a mechanism to help identify representative pixel for supervised classification and are not commonly used as a sole means of classifying images for a specified application (Wilkinson, 1991).
FAO (1999) illustrates the Land use Land Cover dynamics by characterizing land use as the arrangements, activities and inputs people undertake in a certain land cover type to produce, change or maintain it, while land cover is the observed biophysical cover on the earth’s surface. Therefore, land use defined in this way establishes a direct link between land cover and the actions of people in their environment.
One concept that has much merit is that land use refers to, man's activities on land which are directly related to the land. On the other hand, land cover describes, the vegetation and artificial constructions covering the land surface (James et al., 2001).Therefore, land use change is the proximate cause of land-cover change. In order to understand the various implications of land cover change, understanding of land-use change is essential. According to Turner et al. (1993), the main actors between land use and land cover changes are human beings despite the slow natural processes.
Land cover has gone under continuous change for millennia. This change has occurred through the use of fire for game hunting and clearance of patches of land for agriculture and livestock production, since the advent of plant and animal domestication. This is because human’s production demands cannot be fulfilled without modification and/or conversion of land covers (de Sherbinin, 2002 cited in Kahsay, 2004). According to McClelland (1998), the alterations of ecosystem services, due to changes in LULC, in turn negatively affect the ability of the biological systems to support the human needs.
Land use/Land cover changes may occur due to various factors, which may be broadly divided into natural and human induced or anthropogenic causes. United Nations Environmental Protection (UNEP, 2002), identified the general causes of land use and land cover changes, which are: (1) natural processes, such as climate and atmospheric changes, wildfire, and pest infestation; (2) direct effects of human activity, such as deforestation and road-building; and (3) indirect effects of human activity, such as water diversion leading to lowering of the water table. Even though, natural processes may also contribute to changes in land cover, the major driving force is human induced land uses (Allen and Barnes, 1985 cited in Kahsay, 2004). These human induced causes of land cover change, which are critical and currently increasing in alarming rate; and can be categorized into two broad divisions: proximate and driving causes. The proximate causes are causes which results immediate land cover change; while driving causes are causes which drives behind the immediate causes.
Change detection is an important process in monitoring and managing natural resources and urban development because it provides quantitative analysis of the spatial distribution of the population of interest (Zubair, 2006). Fung (1990) indicated the importance of techniques and methods of using satellite imageries as data sources have been developed and successfully applied for land use classification and change detection in various environments including rural, urban, and urban fringes. Turner et al. (1993) also showed the use of remote sensing for biodiversity monitoring through direct and indirect approaches that basis on individual organisms and through reliance on environmental parameters as proxies respectively. Therefore, the use of remote sensing satellite data for land use land cover change detection and monitoring is widely applying throughout the world with the aid of technological improvement that provides high resolution images.
An image differencing technique has been implemented in this change detection study. According to Campbell,1987(cited in Efrem,2008), image differencing appears to perform generally better than other methods of change detection; and such monitoring techniques based on multispectral satellite data have demonstrated potential as a means to detect, identify, and map changes in forest cover. Image differencing is probably the most widely applied change detection algorithm for a variety of geographical environments (Singh, 1989 cited in Efrem, 2008). It involves subtracting one date of imagery from a second date that has been precisely registered to the first.
The study was conducted in Amibara District which is located in the Afar Regional State, with geographic co-ordinates of 09013'5” - 09030'47”N and 40005' - 40025'E and altitude ranging from 665 to 815m.a.s.l.(Figure 1). Amibara district where P. juliflora was presumed to be initially planted and the invasion has become seriously pronounced. Out of the total area of the Region, the study area constitutes 52790 ha and it is located in the south eastern part of the region and 285 kms far from that of Addis Ababa, the capital of Ethiopia (Shiferaw et al., 2004). The district is divided in to 16 Kebeles.
Editorial Note: Figure 1 was removed due to copyright issues.
Figure 1. Map of the study area
The physiography is dominated by plains and undulating side-slopes with 0-8% slopes (MOA, 1997). The soil types are Eutric Fluvisols, Eutric Cambisols, Vertic Cambisols and some part is designated as Vertisols, Orthic Solonchaks, Eutric Regosols, Lithosols with moderately deep to deep, loam to loamy sand texture and moderately well drainage class (MoA, 1997). Basins and depressions are the major physiographic units and the soils have become drainage problem.
Among the factors responsible for the nature and distribution of soils in the region, climate, geology and topography are the dominant ones (MoA, 1997). Owing to the low- lying topography a number of both perennial and intermittent streams drain into the region, Awash and Mille, which carry the top-fertile soil materials from the highlands and deposit in the region.
The Afar Region is both the hottest and driest part of the country with two agro ecological zones, the semi arid and arid. The semi arid agro ecological zone covers areas between 500 and 1500 masl. This is about 20% of the total area. The major part of the region falls within the arid agro ecological zone below 500masl.While in the Dallol depression; there are parts as low as 110m below sea level. The mean annual temperature of the region as a whole is 35OC (MIC, 2000). Rainfall is rather sparse and erratic. The mean annual rainfall varies from 500mm in the south west to less than 200 mm in the north eastern part of the region. The coolest and wettest parts of the region are those in the root hills of escarpment on the banks of the rivers. There are variations to this agro ecology in different parts of the region (AFAP, 1998). Amibara District has an average annual rainfall of 560mm. May and June is the direst season while July – September is the main rainy season. The maximum and minimum temperatures vary from 25 - 420c and 15.2 - 23.50c respectively (WAS, 2008).
Amibara is the most lucky area in natural resource most of the wood land cover follow the awash river bank and among the tree species Acacia nilotica, Acacia seyal, Acacia oerfota, Acacia mellifera, Salvadora persica and Dobera glabra are present in scattered manner, far away from the river bank, majority of the area is covered with small sized acacia & small bush shrub and scrub of different species. But now significant areas of the district were covered with P. juliflora. Different wild animal like, Grevys Zebra, Dik-Dik, Gazella soemmerringi, Lepus habessinicus, lesser kudu all residing in the forest area of the district (Shiferaw et al., 2004). Natural vegetation continues to play an essential role in maintaining the district ecology and economy. This is because natural vegetation provides food, fodder, fuel, building materials and protects the soil from erosion and fertility in addition to stabilizing and improving the environmental condition of the area (MoA, 1997).
Transhumance pastoralism is the major production system in the study areas where cattle, camel, goats and sheep are dominant animals reared. Livestock were kept primarily for their products (milk, milk products and meat) and income (Abule et al., 2005). Livestock for the community are the back bone of their economy. Over 90% of the population depends largely on milk from cattle, sheep, goat and camels as their staple food. The cattle population of the Afar is about (1,620,147).Out of the total population, Amibara constitutes (131,832). The Afar cattle are adaptable to the arid and semi arid ecological zones can survive and produce with the meager feed and water resources, are to some extent tolerant to drought, heat and disease (Shiferaw et al., 2004).
Amibara District has large fertile cultivable land and is very suitable for irrigation agriculture. The dominant crops currently grown (mainly by state and private farms) is cotton and following horticultural crops, the second most grown by small scale farmers. The area has promising potential for producing other food crops like maize, sorghum, groundnut (Shiferaw et al., 2004).
Remote sensed data used for this study were multi-temporal and multi-sensor Landsat images; Landsat Multi Spectral Scanner(MSS) of 1973,Landsat Thematic Mapper(TM) of 1987 and 1999,and Landsat Enhanced Thematic Mapper(ETM+) of 2004(Table 1).
Table 1. Description of satellite image used in change detection in relation to P. juliflora
Abbildung in dieser Leseprobe nicht enthalten
The ancillary data consist of topographic maps and old maps of the district were used for boundary delineation, navigation purpose and supported ground truth and training site establishment activities.
Erath Resources Data Analysis System(ERDAS) Imagine version 8.6, Environmental for Visualizing System(ENVI) version 4.2, and ArcGIS version 9.2 were used in image analysis, LULC change detection, mapping and matrix analysis. In addition, SPSS were used for statistical data analysis. Magellan explorist 600 Model of Global Positioning System (GPS) receiver with an accuracy of + or – 6 meter was used in order to obtain accurate location point data for each land use classes.
The study has used: satellite image analysis and socio-economic data collection and analysis in order to address its objectives (Figure 2).
Information extractions from digital images involve both visual interpretation and computer assisted techniques. Digital processing of data acquired from a remote sensor system basically falls in to two classes: image restoration (Pre-processing) and image analysis (Behailu, 2006).
(i) Delineation and Subseting of the Study Area
Delineation of the study area was performed using ArcGIS 9.2. Due to the large area of the entire Landsat satellite scenes acquired for all years, it was necessary to subset only the study area of interest based on the delineated area using ERDAS version 8.6, which covered 527.90 km2, before any analysis was conducted. The most common undertakings in image preprocessing include restoring missing scan lines, and correction for geometric and radiometric distortions. The satellite imageries were utilized in this research are already radiometrically and geometrically corrected, restored missing scan lines and are geo referenced when obtained that there was no need to work on these aspects of the images.
Editorial Note: Figure2 was removed due to copyright issues.
Figure 2. Over all methodology of the study
Abbildung in dieser Leseprobe nicht enthalten