JUNIPER ISLANDS AND PLANT DIVERSITY · JUNIPER ISLANDS AND PLANT DIVERSITY Salah Ahmed EL Mahi March, 2003 . JUNIPER ISLANDS AND PLANT DIVERSITY A case Study with Remote Sensing and

A case Study with Remote Sensing and GIS in Karaj, Iran

JUNIPER ISLANDS AND PLANT DIVERSITY

Salah Ahmed EL Mahi March, 2003

JUNIPER ISLANDS AND PLANT DIVERSITY

A case Study with Remote Sensing and GIS in Karaj, Iran

By

Salah Ahmed EL Mahi

ITC First Supervisor: Dr. H.A.M.J. van Gils ITC Second Supervisor: Dr. J. de Leeuw

Iranian Supervisor: Dr. J. Ghouddossi Iranian Advisor: Mr. A.A. Noroozi

Thesis submitted to the International Institute for Geo-information Science and Earth Observation in partial fulfilment of the requirements for the degree of Master of Science in Rangeland and Agricul-tural Management Degree Assessment Board Dr. J. de Leeuw (Chairman and Second Supervisor), NRS DEPT., ITC Dr. D. Rugege (External Examiner), University of Natal, RSA. Dr. H.A.M.J. van Gils (First Supervisor), NRS DEPT., ITC Drs. E. Westinga (member), NRS DEPT., ITC

INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION ENSCHEDE, THE NETHERLANDS

Disclaimer This document describes work undertaken as part of a programme of study at the International Institute for Geo-information Science and Earth Observation. All views and opinions expressed therein remain the sole responsibility of the author, and do not necessarily represent those of the institute.

Table of Contents Acknowledgements…………………………………………………………………………..I Abstract…………………………………………………………………………………….. II List of abbreviations…………………………………………………………………….... III List of figures…………………………………………………………………………… IV-V List of tables……………………………………………………………………………….. VI List of plates…………………………………………………………………...…………...VII List of appendices…………………………………………………………………………VIII

1. INTRODUCTION........................................................................................................................1

1.1. GENERAL INTRODUCTION.......................................................................................................1 1.1.1. Forest globally ...............................................................................................................1 1.1.2. Fragmentation of forest..................................................................................................4 1.1.3. Island biogeography theory ...........................................................................................6 1.1.4. Habitat of Juniper excelsa subsp. polycarpos ...............................................................7 1.1.5. Taxonomy of Juniperus excelsa subsp. polycarpos........................................................8

1.2. RESEARCH OBJECTIVES ..........................................................................................................8 1.2.1. General objective ...........................................................................................................8 1.2.2. Specific objectives ..........................................................................................................8

1.3. RESEARCH QUESTIONS ...........................................................................................................8 1.4. RESEARCH HYPOTHESIS..........................................................................................................8

2. CHARACTERISTIC OF THE STUDY AREA.......................................................................10

2.1. GENERAL INFORMATION.......................................................................................................10 2.2. LOCATION OF THE STUDY AREA ...........................................................................................11 2.3. CLIMATE ...............................................................................................................................11 2.4. GEOLOGY..............................................................................................................................17 2.5. SOIL ......................................................................................................................................17 2.6. VEGETATION.........................................................................................................................18 2.7. LAND USE .............................................................................................................................18

3. METHODS .................................................................................................................................19

3.1. PRE-FIELDWORK STAGE ........................................................................................................19 3.1.1. Image processing..........................................................................................................19 3.1.2. TCC ..............................................................................................................................19 3.1.3. Image classification .....................................................................................................19 3.1.4. NDVI.............................................................................................................................22 3.1.5. Altitude .........................................................................................................................22 3.1.6. Aspect ...........................................................................................................................25 3.1.7. Slope .............................................................................................................................26 3.1.8. Sampling Technique .....................................................................................................27 3.1.9. Study materials.............................................................................................................27

3.2. FIELDWORK STAGE ...............................................................................................................28 3.3. DATA INTEGRATION AND ANALYSIS .....................................................................................29

4. RESULT......................................................................................................................................30

4.1. EXPLORATORY ANALYSIS.....................................................................................................30 4.2. CORRELATIONS ANALYSIS ....................................................................................................32 4.3. DESCRIPTIVE STATISTICS......................................................................................................32

4.3.1. Juniper density versus number of plant species...........................................................32 4.3.2. Juniper canopy cover versus number of plant species.................................................33

4.4. ENVIRONMENTAL FACTORS AND THE DENSITY OF JUNIPER .................................................34 4.4.1. Altitude .........................................................................................................................34 4.4.2. Slope .............................................................................................................................34 4.4.3. Aspect ...........................................................................................................................35 4.4.4. Soil pH..........................................................................................................................35 4.4.5. Soil EC and soil texture................................................................................................36 4.4.6. Soil moisture content....................................................................................................36

4.5. ENVIRONMENTAL FACTORS AND NUMBER OF PLANT SPECIES .............................................37 4.5.1. Altitude .........................................................................................................................37 4.5.2. Slope .............................................................................................................................37 4.5.3. Aspect ...........................................................................................................................38 4.5.4. Soil texture ...................................................................................................................38 4.5.5. Soil depth......................................................................................................................39 4.5.6. Soil pH and CEC ..........................................................................................................39

4.6. JUNIPER CANOPY COVER AND THE ENVIRONMENTAL FACTORS ...........................................40 4.6.1. Altitude .........................................................................................................................40 4.6.2. Slope .............................................................................................................................40 4.6.3. Aspect ...........................................................................................................................41 4.6.4. Soil pH..........................................................................................................................41 4.6.5. Soil texture and soil moisture content..........................................................................42 4.6.6. Soil depth......................................................................................................................42

4.7. NUMBER OF PLANT SPECIES VERSUS NDVI .........................................................................43 4.8. JUNIPER DENSITY AND CANOPY COVER VERSUS NDVI........................................................43 4.9. NDVI VERSUS ALTITUDE .....................................................................................................44 4.10. MULTIPLE REGRESSION ANALYSIS ..................................................................................45

4.10.1. Juniper density as a response ......................................................................................45 4.10.2. Juniper canopy cover as a response ............................................................................46 4.10.3. Number of plant species as a response ........................................................................47

4.11. IMAGE CLASSIFICATION AND EVALUATION ......................................................................48 4.12. STATISTICAL ANALYSIS OF SIZES AND ISOLATION OF FOREST ISLANDS ...........................50 4.13. RESULT SUMMARY............................................................................................................51

5. DISCUSSION .............................................................................................................................55

5.1. ENVIRONMENTAL FACTORS VS. JUNIPER .............................................................................55 5.2. PLANT SPECIES RICHNESS VS ENVIRONMENTAL FACTORS ...................................................56 5.3. FOREST ISLANDS SIZES VS PLANT SPECIES RICHNESS...........................................................56 5.4. ISOLATION OF FOREST ISLANDS VS PLANT SPECIES RICHNESS .............................................58

6. CONCLUSIONS & RECOMMENDATIONS ........................................................................59

6.1. CONCLUSIONS.......................................................................................................................59 6.2. RECOMMENDATIONS ............................................................................................................60

References.……………………………………………………………………………….61- 63 Plates……………………………………………………………………………………..64- 66 Appendices……………………………………………………………………………….67- 90

Acknowledgement I would like to express my sincere and heartfelt gratitude to the Iran Government through the Ministry of Agriculture and Khagi Nasir Toosi University and to the Netherlands Government representing in the Netherlands Fellowship Program (NFP) for their fellowship to study for a Master of Science de-gree without which it would have been futile to realize my ambition to study at K.N.T University and at ITC. I am grateful to my employer, Forest National Corporation through the General Director Dr. A/Azim complemented my efforts by giving me a grant to fulfil my ambition. I express my deepest gratitude to my supervisor Dr. Hein van Gils for his esteemed, constructive and consistent follow-ups and guidance throughout the work from I came to ITC, words are inadequate to express the assistance and attention he gave to my work. I am very grateful to Dr. Jan de Leeuw for his fruitful and professional advice. My thanks go to all the staff of JIK through Dr. B. Aminipouri and Dr. A. Abkar for the support and guidance throughout the modules and thesis preparation. Special thanks go to my Iranian supervisor Dr. J. Ghouddoossi and to my advisor Mr. A.A. Noroozi for their consultancy during the fieldwork. I am indebted to Dr. M. Assadi and Dr. V. Mozaffarian for their guidance. I would like to thank Dr. A. Sanadgol and Dr. M. Farahpoor for their helpfulness. Also my acknowledgements go to the staff of the laboratory of the soil and watershed research centre in Tehran. I would like to extend my gratitude to Dr. Huizing and Dr. Toxopeus for their advices during the fieldwork. My thanks go Mr. Lieshout for his continuous cooperation. Moreover I express my appre-ciation to ITC library staff for their cooperation. Last but not least I would like to thank my family for always been there, patience, supportive, encour-age, and missing me during my study, to my wife Marya and my daughters Reem, Reyan, and Reyhan as well as my son Ahmed. To my brothers Yousif, Rahama, and Yasa. To my sisters Niema, Haram, and Suaad as well as my relatives. Above all my thanks to my GOD the most gracious and merciful for being on my way I

Abstract This study was conducted in the northern part of Tehran province (Iran) at Sad amir kabir catchments area. The aim of this research was to investigate the habitat of �� An ad-ditional objective to test the relation between Juniper forest islands sizes and isolation with the plant species richness. The distribution and habitat of �� were determined in the mountains area by using the thematic mapper satellite image (ETM) and GIS technique. Through the supervised clas-sification coupled with the function of GIS, the area and the isolation of Juniper forest islands were determined. The study applied the visual interpretation of the Landsat satellite image for identifying the Juniper forest islands. Juniper forest islands were distributed in a range of altitude of 1896-2775m. Juniper is the dominant woody species at this range of elevation, where it generally forms open woodland with canopy cover range from 0 percent to 62.8 percent with the average of 17.7 percent. The slope direction (aspect) does not affect on the distribution of the Juniper trees. The Juniper forest islands mostly were existed on the wet substrates. Eleven-forest islands of Juniper were studied to investigate the research hy-potheses through the statistical analysis. The relations between the environmental factors (topography and soil) were analysed using the stepwise regressions. The density and the canopy cover of Juniper and the plant species richness were high at soil pH of (6.25-6.5). The research indicated that the coef-ficient of the regressions of the environmental factors with the plant species richness was very low. Moreover the study obtained that both the forest islands sizes and the isolation of the forest islands were highly significantly related to the plant species richness according to the hypothesis of the island biogeography theory. II

List of abbreviations CEC Cation Exchange Capacity

DEM Digital Elevation Model

EC Electrical Conductivity

ETM Thematic Mapper Landsat Image (8 bands)

FAO Food and Agricultural Organization

FRA Forest Resources Assessment

GIS Geographical Information Systems

GPS Global Positioning Systems

ILWIS Integrated Land and Water Information System

INH Iran National Herbarium

LATLON Latitudinal and Longitudinal coordinate

M1.1 Mountainous with high hills and very steep slopes Landscape

NDVI Normalized Difference Vegetation Index

NIR Near Infra-Red Wave

pH Negative Logarithm of Hydrogene Ion activity

RS Remote Sensing

SCWSRC Soil Conservation and Watershed Management Research Centre

TCC True Colour Composite

UTM Universal Transverse Mercator Coordinate

III

List of figures 1.1 Forest area by ecological zones............................................................................................2 1.2 Forest areas by regions.........................................................................................................2 1.3 Wood volume by region.......................................................................................................3 1.4 Above-ground biomass in forest by region..........................................................................3 2.1 Location of the study area...................................................................................................10 2.2 Average precipitation and mean temperature (1986-1994) Asara area..............................12 2.3 Ombrothermic diagram for Asara (1986-1994)...................................................................12 2.4 Average precipitation and mean temperature (1986-1994) Nisaa area...............................13 2.5 Ombrothermic diagram for Nisaa (1986-1994)...................................................................14 2.6 Average precipitation and mean temperature (1986-1994) Shahristanak area....................15 2.7 Ombrothermic diagram for Shahristanak (1986-1994).......................................................15 2.8 Average precipitation and mean temperature (1986-1994) Sad amir kabir area................16 2.9 Ombrothermic diagram for Sad amir kabir (1986-1994)....................................................17 3.1 Flow chart of study activities..............................................................................................20 3.2 Vegetation map 1993……………………………………………………………………..21 3.3 TCC map & sample points..................................................................................................23 3.4 NDVI map of the study area...............................................................................................23 3.5 Altitude map & sample points............................................................................................24 3.6 Aspect map & sample points..............................................................................................25 3.7 Slope map & sample points................................................................................................26 4.1 Normality test (A) histogram of the original data (B) probability plot of the original data tested by Kolmogrov-Smirnov of plant species number.............................................30 4.2 Normality test of Juniper canopy cover before and after data transformation....................31 4.3 Number of plant species versus number of Juniper trees....................................................33 4.4 Juniper canopy cover% versus number of plant species.....................................................33 4.5 Altitude versus number of Juniper trees.............................................................................34 4.6 Slope % versus number of Juniper trees.............................................................................34 4.7 (A) number of Juniper trees versus aspect (B) Number of Juniper trees versus soil pH.......................................................................................................................................35 4.8 (A) Number of Juniper trees versus soil EC (B) Number of Juniper trees against soil texture...........................................................................................................................36 4.9 (A) Number of Juniper trees versus soil moisture % (B) Number of Juniper trees versus soil depth..................................................................................................................36 4.10 Altitude number of plant species........................................................................................37 4.11 Slope % versus number of plant species............................................................................37 4.12 (A) Aspect versus number of plant species (B) Soil texture versus number of plant species.................................................................................................................................38 4.13 (A) Soil depth versus number of plant species (B) Soil moisture % versus plant species. 39 IV

4.14 (A) Soil CEC versus number of plant species (B) Soil pH versus number of plant species.................................................................................................................................39 4.15 Altitude versus Juniper canopy cover %.............................................................................40 4.16 Slope % versus Juniper canopy cover%.............................................................................40 4.17 (A) Aspect versus Juniper canopy cover % (B) Soil texture versus Juniper canopy cover %...............................................................................................................................41 4.18 (A) Soil texture versus Juniper canopy cover % (B) Soil moisture % versus Juniper canopy cover %..................................................................................................................42 4.19 (A) Soil depth versus Juniper canopy cover % (B) Soil CEC versus Juniper canopy cover % .............................................................................................................................42 4.20 NDVI versus number of plant species................................................................................43 4.21 (A) Number of Juniper trees versus NDVI (B) Juniper canopy cover % versus NDVI…43 4.22 NDVI versus altitude..........................................................................................................44 4.23 The classified maps.............................................................................................................49 4.24 (A) Total number of plant species as a function of the forest islands sizes

(B) Total number of plant species as a function of the average of edge-to-edge distance...........................................................................................................................…50

4.25 Juniper density map............................................................................................................52 4.26 Juniper canopy cover map..................................................................................................53 4.27 Plant species richness map.................................................................................................54 V

List of tables 1.1 Annual gross at net changes in forest areas 1990-2000.......................................................1

1.2 Forest area volume and aboveground biomass by region....................................................1 2.1 Summary of average monthly precipitation and temperature of Asara metrological station.................................................................................................................................11 2.2 Summary of average monthly precipitation and temperature of Nisaa metrological station.................................................................................................................................13 2.3 Summary of average monthly precipitation and temperature of Shahristanak metrological station...........................................................................................................14

2.4 Summary of average monthly precipitation and temperature of Sad amir kabir station...16 3.1 The altitude classes............................................................................................................22 3.2 The main aspect classes.....................................................................................................25 3.3 The slope classes................................................................................................................26 4.1 Pearson correlation matrix.................................................................................................32 4.2 Stepwise regression analysis using two explanatory variables for Juniper density...........45 4.3 Analysis of variance...........................................................................................................45 4.4 Regression analysis of variance.........................................................................................45 4.5 Stepwise regression analysis using three explanatory variables for Juniper canopy cover...................................................................................................................................46 4.6 Analysis of variance...........................................................................................................46 4.7 Regression analysis of variance.........................................................................................46 4.8 Stepwise regression analysis using three explanatory variables for Juniper canopy cover...................................................................................................................................47 4.9 Analysis of variance...........................................................................................................47 4.10 Regression analysis of variance.........................................................................................47 4.11 Confusion matrix for accuracy evaluation on the land cover map 2000............................48 4.12 Accuracy of the 2000’s image classification......................................................................48

4.13 Forest Island sizes distances and plant species richness of 11 Forest Islands...................50 4.14 Analysis of variance...........................................................................................................51 4.15 Regression analysis of variance.........................................................................................51

VI

List of plates Plate 1: Juniper trees with Cotoneaster sp. on sheltered wadis habitat

Plate 2: Juniperus excelsa subsp. polycarpos in Sirachal Forest Island

Plate 3: Three life forms: Juniper trees, shrubs and grasses

Plate 4: Juniperus sp. on steep slope

Plate 5: J. excelsa subsp. polycarpos and Rosa canina shrub with many grasses

Plate 6: Field worker with soil tools VII

List of appendices Appendix 1-1 Field data form

Appendix 1-2 Plant species and general information

Appendix 2(1-8) Standard relevee sheet of field data

Appendix 3(1-2) List of inventoried plant species

Appendix 4-1 Plant species of Sirachal forest island

Appendix 4-2 Plant species of Kalvan Forest Islands

Appendix 4-3 Plant species of Sercheh forest islands

Appendix 4-4 Plant species of Lanyz 1 Forest Island

Appendix 4-5 Plant species of Lanyz 2 Forest Island

Appendix 4-6 Plant species of Rayzamin Forest Island

Appendix 4-7 Plant species of Hamega Forest Island

Appendix 5 Summary of area calculated from the relief map

VIII

JUNIPER ISLANDS AND PLANT DIVERSITY: A CASE STUDY WITH RS AND GIS IN KARAJ, IRAN

1

1. Introduction

1.1. General Introduction

1.1.1. Forest globally

Using the FRA of FAO (2000) global definition of forest (minimum 10% threshold crown cover) and new baseline information, it was estimated that the world’s forest cover at the year 2000 was about 3869 million hectares (representing 30 percent of the world’s land area). About 95 percent of the for-est cover was in natural forest and 5 percent in forest plantation. Using a combination of new global maps and statistical data, FRA 2000 also estimated the distribution of forest area by ecological zones; 47 percent is in the tropic, 33 percent in the boreal zone, 11 percent in temperate areas and 9 percent in the subtropics FAO (2000). See figure (1.1) According to FAO (2000), deforestation in the 1990s was estimated at 14.6 millions ha per year. The worldwide gains in forest cover totalled 5.2 millions ha per year. Thus the net global change in the forest area between 1990 and 2000 was estimated as –9.4 millions ha per year, represent an area about the size of Portugal. The estimated net loss of forest for 1990s as a whole was 94 millions ha. See ta-ble (1.1) Table 1.1 Annual gross at net changes in forest areas 1990-2000 (Million hectares per year)

Domain Deforestation Increase in forest area Net change in forest area Tropics -14.2 +1.9 -12.3 Non-tropics -0.4 +3.3 +2.9 World -14.6 +5.2 -9.4 Source: FAO (2000).

The assumption and extrapolations had to be used for estimate wood volume and woody biomass. The year 2000 estimate for the global wood volume of forest was 386 thousand millions cubic meters and the estimate worldwide aboveground woody biomass was 422 thousand millions tons. (See table 1.2 below) Table 1.2 Forest area volume and aboveground biomass by region

Forest area Volume Biomass By area Total By area Total

Region Million ha Cubic m/ha G cubic m Ton/ha Gt Africa 650 72 46 109 71 Asia 548 63 35 82 45 Oceania 198 55 11 64 13 Europe 1038 112 116 59 61 North& Central America 549 123 67 59 52 South America 886 125 111 203 180 Total 3869 100 386 109 422

Source: FAO (2000).


2

Forest areas by ecological zones

Fig 1.1 Source: FAO (2000)

Forest areas by regions Fig 1.2

Source: FAO (2000)


3

Wood volume by regions Fig 1.3

Source: FAO (2000)

Above-ground biomass in forest by regions Fig 1.4

Source: FAO (2000)


4

The total land area of the west Asian sub region countries is about 5.4 percent of the global land area. In general those countries have poor forests with only 3.2 percent of the total area, and less than one percent of the world’s forest cover. For Iran (Islamic republic of Iran), a survey base on satellite im-age, aerial photos and field survey was carried out for the Caspian Sea forests and central Zagros in 1999. For the other parts of the country a sample inventory was used. According to the country data, the land area of Iran is about 162201 thousands ha, the total area of forests in the year 2000 was about 7299 thousand ha representing 4.5 percent of land area (FAO 2000). In the matter of fact, the total volume of wood and biomass in forest estimated were 631 millions cubic meter and 1089 millions tons respectively.

1.1.2. Fragmentation of forest

Fragmentation is considered as the first step of degradation in forest area, and is followed by the com-plete disappearance of remain of forests. Fragmentation, as a concept, is the process that occurs when a habitat or land cover type is subdivided either by a natural disturbance or by human activities (Dale and Person 1996, in Diaz, 2000). Forest fragmentation causes many physical and biological changes as a result of habitat loss and insu-larization (Lovejoy, 1980; Laurance and Richard, 1997). As forest landscapes become increasingly fragmented, populations of forest species are reduced, dispersal and migration patterns are inter-rupted, and ecosystem inputs and outputs are altered and previously isolated core habitats become ex-posed to external conditions, all of which result in a progressive erosion of biological diversity (Til-lman 1994, in Caballero, 2001). At the scale of the individual forest patch, forest loss and fragmentation can have a broad range effect on population survival, ecological interactions and biodiversity (Fahrig & Grez 1996). As patches be-come smaller, populations tend to be more vulnerable to extinction because of demographic, environ-mental or genetic risks (Gilpin 1987, Goodman 1987). When patches become isolated, with no con-nections between them, migration of organisms may be precluded (Kareiva 1987). Small patches also have a higher edge/interior ratio. For forest interior species, this represents a loss of habitat greater than predicted by the reduction in patch size alone (Wilcove et al. 1986, Williams-Linera 1990). The magnitude of such effects depends on the landscape-scale pattern of forest loss, which will determine the number of the remaining patches, their sizes and shapes, the distances among them, and the nature of the surrounding “matrix” habitat. To date there are few empirical studies that connect pattern of forest loss and fragmentation with ecological processes at the landscape scale (Groom & Schumaker 1993, in Grez et al., 2000). Anthropogenic disturbance coupled with fragmentation had a stronger and more immediate effect in reducing native species richness and increasing exotic species richness than did fragmentation alone. In the absence of major disturbance, small fragments had fewer native species than larger size classes, but only after 10 or more years since fragmentation, confirming the importance of controlling for age of fragments when examining species-area relationships (Fox et al., 2002). Fragmentation severely alters physical conditions in forest understories, but few studies have con-nected these changes to demographic impacts on forest species using detailed experimental examina-tion at the individual and population levels (Harrison et al., 2002). A common feature of habitat frag-mentation is a sharp increase in the amount of induced habitat edge. Consequently, plant and animal populations in fragmented habitats are not only reduced and subdivided, they are increasingly exposed to ecological changes associated with induced edges (Wilcove et al. 1986, in Laurance et al., 1991).


5

Due to the increase of population in the current decades and also the increase of its activities, the bio-diversity has been affected provoking destruction, degradation and habitat fragmentation until it causes the declination and total loss of biodiversity in the different areas of the globe. Thompson and Jones (1999) for example found positive correlation between population density and local plants spe-cies extinction (Caballero, 2001). An important aspect of forest fragmentation is the ensuring change in forest area, and the impact of this has on species number and composition. Quantifying this is an important step in prioritising forest fragmentation for biodiversity conservation. Species-area curves from isolated forest fragmentation in Ghana (west Africa) showed that large forests contain the greatest number of trees species (Hill & Curran, 2001). Anderson & Wait present a new hypothesis for predicting and describing patterns of species diversity on small-scale island and habitat fragments. They modified the traditional island biogeography equi-librium theory to incorporate the influence of spatial subsidies from the surrounding matrix which vary between island and habitat fragments on species diversity (Anderson & Wait, 2001). In Caicaros (Brazil) Rossalo & Lietao (1999) rely on medicinal plants following the prediction of is-land biogeography theory, a lower diversity of medicinal plants cited in island was found compared with continental communities. According to Shigeo (1995), forest fragmentation in all study sites in Japan has increased in the last 25 years. In general, the forested area decreased, and the forest patches became smaller and more dis-tant from each other. Species diversity was affected both by forest fragmentation and human man-agement of the coppice forest. The diversity index (H) for trees had a significant partial correlation with factors relating to forest fragmentation (forest patch density, decreasing ratio of forested area and stand size). The relative change in the percentage of forested land had a significant positive correla-tion with the proportion of forestland itself. It implies that sites with a larger proportion of forested area are few from large cities and that their forested land decreased less than the site near large cities (Shigeo Iida, 1995). For the study of land cover changes processes one of the most useful techniques over extensive areas is RS integrated with GIS. Remotely sensed images are a cost-effective way to acquire land cover in-formation for a large geographical area (Wickland, 1991; Lillesand and Keifer, 1994). RS instruments typically measure electromagnetic radiation reflected and emitted by earth’s surface, which can be collected at multiple scales and times (Wickland, 1991). Satellite images provide information at scale appropriate for mapping landscape-level vegetation patterns. The advent of satellite remote sensing and sensor related technological advancement has made it possible to perform precise vegetation clas-sification and mapping (Prasad 1994, in Caballero, 2001). The fragmentation of forest areas can be assessed based on the amount of patches from the classifica-tion result and the change in different indicators on landscape indices. Some research assessed spatial patterns and rates of forest fragmentation using digital remote sensing data for a region in southern New England (Vogelmann; 1995). A forest continuity index was calculated as a measure for the com-plexity of forest area and the relation in distance between patches. The values found show a decrease with increasing population density until 200 individuals per square kilometre. They concluded that satellite derived estimates of forest fragmentations are closely correlated with population density and the continuity index may be valuable for assessing patterns and rates of forests at continental levels (Diaz, 2000).


6

Rudis and Ek extended Crurtis’ descriptions of forest fragmentation in Wisconsin (1956) using spatial analysis to show that in southeastern Wisconsin, small and large forest islands are interspersed not clustered. Thus, small forest islands are likely to have larger neighbours, which may possess greater biotic diversity and serve as source of colonizing flora and fauna. However, in this region, the more mature communities are clustered, while younger stands are randomly distributed (Robert et al. and Sharpe et al., 1981)

1.1.3. Island biogeography theory

The island biogeography theory is the brain of Robert MacArthur & E.O. Wilson. The basis of Island Theory supposes there is a large source area of species and surrounding sizes and distances from the source area. Species disperse from the source area to the island (R.MacArthur & E.O.Wilson, 1963). In general, the patches that have the largest variety of habitat also support the most plants species. This because a large number of habitats promote successful dispersal and speciation. The theory has been found to be a satisfactory explanation for a wide variety of observed species dis-tribution (Galli et al. 1976, Opler 1974, Sepkoski and Rex 1974, Simpson 1974, Vuilleumier 1970, Willis 1974, inter alios), and at least a few experimental observations (Rex 1975, Simberloff and Willis 1969). Commonly, the relationship between the number of species (S) and the area of islands (A) has been expressed as S = b1 A b2 Where b1 and b2 are constants. Indices other than area have been used to explain the variation in spe-cies among islands with sometimes-greater statistical success. Some of these are a subjectively ranked index of ecological diversity (Opler 1974), foliage height diversity (MacArthur et al. 1966), elevation and distance-to-nearest-island (Mauriello and Roskoski 1974), and area distance measures in past ages (Simpson 1974, in Rudis et al. 1981). MacArthur & Wilson equilibrium theory revolutionized the island biogeography and to large degree ecology as well. The theory of island biogeography has changed little over the past three decades. It has not kept pace with relevant theory and their growing appreciation for the complexity of nature, especially with empirical findings that species diversity on many islands: 1) is not in equilibrium; 2) is influenced by different in speciation, colonization, and extinction among taxa; and 3) is influenced by differences among islands in characteristics other than area and isolation (Anderson & Wait, 2001). Species richness for trees and woody plants generally increased with island size to approximately 2.3 ha. Islands smaller than 2.3 ha functioned essentially as edge communities composed of a mix of in-tolerant species. Islands larger than 2.3 ha showed a general decline in species richness as interior, mesic conditions became established with only tolerant species persisting. Species richness ceased to decline at approximately 3.8 ha, suggesting that only the shade-tolerant, mesophytic species remained (Levenson, 1981). Ranney et al. drew a similar conclusion from analysing the impact of forest island formation on the creation of an edge community, and the subsequent effect of the edge on the functioning of the forest island. In southeastern Wisconsin, edge is definable in term of the continuum index, which incorpo-rates both importance and adaptation values of the constituent species. Continuum index increases inward from the edge. Species with high adaptation values (shade-tolerant) significantly increase away from the edge, while those with low adaptation values (shade-intolerant species) decline. Impor-tance of some species seemed to be independent of position vis-à-vis the edge. This analysis con-cluded that the edges on the east and south sides of forest islands extend 10-15 m into the forest is-


7

land, and up to 30 m on the west side. Thus, the size of a forest island must be at least twice these di-mensions in order to contain mesic species (Robert et al. and Sharpe et al., 1981) In general, larger patches of habitat contain more species and often a greater number of individuals than smaller patches of the same habitat. This occurs for several reasons. First, the larger the habitat patch, the more local environmental variability is contained within it; for example, differences in mi-croclimate, structural variation in the plants, and diversity of topographic positions. This variability provides more opportunities for organisms with different requirements and tolerances to find suitable sites within the patch (Turner, 1998). An area of fragmented forest occurs in the moderate higher part of Karaj district, where patches of Juniper forest exist within a rangelands environment. The plants species occurring within these forests differ from the species in the surrounding area. Hence, for these plants species, these forest fragments will be like “Islands” in the widest of an “Ocean” of non-suitable rangelands. We hypothesize that the same processes as described by MacArthur & Wilson determine species richness in these Juniper for-ests. The area therefore could be an ideal open-air laboratory to study species diversity in a frag-mented forest environment.

1.1.4. Habitat of Juniper excelsa subsp. polycarpos

The species of Juniper existed in the study area, is �� M.Bieb �� (K. Khoch). �� (K.Khoch) species was found in mountain area extends from Turkey through to India and as an isolated population on Jebel Akhdar in the northern mountain of Oman. This species is much more a continental taxon than the typical subspecies. It occurs in the Western Asiatic Subregion of the Irano-Turanian Region, and especially in the Armino-Iranian Prov-ince (Takhtajan, 1986: 144), characterized by mountains separated by vast steppes and deserts. The Jabal-al-Akhdar in Oman forms a disjunt enclave of this Province (Takhtajan, (1986: 144), which has its western limit in the part of Turkey, roughly from Gumushane to Maras. West of this line �� does not occur. It is possibly sympatric with the typical subspecies only in Armenia (including Turkish Arminia), and along the mountain chains from Azerbaijan eastward around the southern end of the Caspian Sea, where the typical subspecies becomes increasingly rare. Precipita-tion is still relatively high in these mountains (usually well above 500 mm annually). Often this spe-cies occurs abundantly farther east in Afghanistan and in Tadzhikistan and Uzbekistan. Precipitation is less abundant; much of it comes as snow in winter. It is generally a higher altitude taxon and the further it reaches eastward, the higher altitude it attains (see also Browicz & Zielinski, 1982); it oc-curs from 500-3800m (Kitamura, 1960), but generally between 1200-3000m. It grows exceedingly slow and in many areas groves of very old trees occur, with little or no rejuvenation. It is much more resistant to drought and radiation (heat) than the typical subspecies and strongly heliophilous, but it can tolerate winter cold equally well. An exceptionally thick cuticle prevents dehydration. It occurs on stony, rocky slopes, often spaced wide apart, sometimes mixed with �� , with which, it is said to have formed hybrid complexes (Browicz & Zielinski, 1982). No herbarium material seen for this study has any characters supporting that suggestion; the two taxa seem to be well separated mor-phologically. Above 2400 m altitude, the Juniperus excelsa woodlands, generally appear to be regen-erating and on good condition, both on exposed slopes and in wadis and sheltered gullies, where as below 2400 m, the juniper trees are the main several centuries old and most of the trunks of most stands are in poor condition and exhibit few signs of regeneration. In Balochistan trunk and crown are in poor condition. Younger trees are rare. Even in woodland plots fenced against grazing or under nursery conditions the germination rate and subsequent growth rates


8

are extremely low even on favourable sites, to non-existent in most of woodland. Mistletoe infection is common (Gils & Baig, 1992). In Khorasan altitude range from1500 to 2200 m, such as Samnan and Mashad. Moreover juniper trees are occurred in Alborz region (Tehran) between Damavound, Tar lake Shahrestanak, Polar, and Karaj catchments area (Ahmed & Zienalabedin, 1978).

1.1.5. Taxonomy of Juniperus excelsa subsp. polycarpos

In Iran �� M. Bieb �� (K. Khoch) is a tree to 10 meters tall, green head, bark dark grey, scaly, peeling off. Young branches brownish-red, rounded-tetragonal, very slender, leaves on branchlets, divergent at the tips ovate, acuminate, with a long spiny point, the dorsal gland oblong; trees occur with strict acicular leaves; leaves of young branchlets are very small, glaucous, imbricated, oblong or ovate with ovate to sub orbicular gland. Fruits solitary, globose, black, pruinose, 9-12 mm in diameter. Seeds 5-8 rarely 3 or 4 oblong-ovate, the upper part of abaxial face wrinkled. The tree is monoecious. PL. 150 cupressaceae (Ahmed & Zienalabedin, 1978).

1.2. Research objectives

1.2.1. General objective

The aim of this study is to test the hypothesis of Island Biogeography Theory and the diversity of plants species in the Juniperus excelsa fragments.

1.2.2. Specific objectives

- To test the relation between Juniper densities, Juniper canopy cover with number of plants species.

- To determine, what are the main factors affecting Juniper forest fragments and plant species. - To map the physical factors and the Juniper fragments of the study area, using GIS and TM

image as well as topography map.

1.3. Research questions

To meet the above objectives, this research will address the below specific questions • Why is the Juniper forest fragmented? What are the main factors affecting their fragmenta-

tion? • Is there any relation between the densities of Juniper with the number of plant species? • Is the number of plant species related to patch size and the isolation among them?

1.4. Research hypothesis

A number of assumptions or hypothesis have been formulated to guide the research. These can be stipulated as follows:

• There is no difference in the Juniper density in areas with the number of plant species versus there is difference in Juniper density with the number of plant species.

Ho: ��=��

H�: �1��2 Where �1 = the median of Juniper density per ha and �2 = the number of plant species.


9

• There is no significant difference in number of plant species in areas with different sizes of patches versus there is significant difference in the number of plant species in areas with dif-ferent sizes. Ho: �1= �2 H�: �1� �2 Where �1 = the number of plant species and �2 = difference sizes of patched.

• The isolation in forest patches doesn’t influence the number of plant species versus the isola-tion in forest patches influences the number of plant species. Ho: �1= �2

H�: �1� �2 Where �1 = the isolation in forest patches and �2 = the number of plant species. • There is no correlation between the density of Juniper and the selected physical and soil fac-

tors such as altitude, slope, aspect, soil depth, soil moisture, soil EC, soil CEC, and soil pH versus there is correlation in Juniper density with the physical and soil factors.

Ho: r = 0 H�: r � 0 Where r is the correlation coefficient. • There is no correlation between the number of plant species and the selected physical and soil

factors such as altitude, slope, aspect, soil depth, soil moisture, soil EC, soil CEC, and soil pH versus there is correlation in the number of plant species with the physical and soil factors.

Ho: r = 0 H�: r � 0

Where r is the correlation coefficient.


10

2. Characteristic of the study area

2.1. General information

The Islamic Republic of Iran is one of the West Asian Sub region countries. It is bordered to the north by the states of Armenia, Azerbaijan and Turkmenistan (all formerly of the USSR) and the Caspian Sea; to the east by Afghanistan and Pakistan; to the south by the Gulf of Oman and the Persian Gulf; and to the west by Iraq and Turkey. According to the country data, the total land area of Iran is about 162201 thousands ha, the total area of forests in year 2000 was about 7299 thousands ha representing 4.5 percent of land area (FAO, 2000). In the matter of fact, the total volume of stand wood and bio-mass in forest estimated were 631 millions cubic meter and 1089 millions tons respectively. The country is dominated by three mountain ranges: the fertile, volcanic Sabalan and Talesh ranges in the north-west; the very long, Jurassic-era Zagros range, down the western border; and the dominant Al-borz range, home of Iran's highest mountain, the permanently snowcapped Damavand (5670m), to the north of Tehran. The two great Iranian deserts, the Dasht-é Kavir (more than 200,000 sq km) and the Dasht-é Lut (more than 166,000 sq km), occupy most of the northeast and east of the central plain. Fig 2.1:Location of the study area


11

2.2. Location of the study area

This area names as Sad Amir Kabir or Karaj Dam catchments area. The study area is located in Te-hran province, and it lies in distance of 40 kilometres northwest of Tehran city. The total area is about 82885 hectares, with latitude range between 35˚:53�:03� N to 36˚:10�:37� N and longitude range be-tween (51˚:03�:15� E to 51˚:36�:26� E). The catchments area is surrounded by Alborz Mountain in the north, Pahn Hesar in the south, Rarian area and Dam Lake in the east, and Karaj urban area in the west. The topography of the study area, has different relief varies from high elevation (4500 m) at Kahar mountain to low elevation (1500 m) around the dam area. (Golrang: study report, 1995).

2.3. Climate

The study area has a semi-arid climate. According to the study of metrological data of (4) station dur-ing the period 1986-1994, based on the Ombrothermic diagram for each station, the dry period starts from June and ends at November (see tables and figures below). The average monthly temperature, range from –6.3 in Jan. to 24.8 degree centigrade in July. The average precipitation is about 406 mm in the south and 676 mm in the north, distributed among the seasons as follows; 30% falls in autumn, 45% in winter, 23% in spring, and only 2% in summer (Golrang: study report, 1995). Table 2.1 Summary of average monthly precipitation and temperature of Asara station:

Month

Average Precipitation (Mm)

Average Of mean tem-perature (º C)

Average Of max. tem-perature (º C)

Average Of min. tem-perature(º C)

Jan 53.54 -3.22 1.69 -8.15Feb 54.12 -3.57 2.25 -9.37Mar 70.76 1 5.24 -3.24Apr 59.96 3.94 10.28 -0.01May 51.81 8.04 15.48 3.69Jun 13.12 12.24 21.89 6.07Jul 5.55 15.91 26.66 9.19Aug 2.81 18.92 26.24 11.34Sep 2.94 16.67 24.44 8.9Oct 16.24 10.52 17.41 3.64Nov 64.17 7.06 14.03 0.06Dec 80.66 1.39 7.38 -4.61 Year 475.68

Source: Metrological organization, Tehran.


12

Average precipitation and mean temperature for Asara area (1986-1994) Fig 2.2

Ombrothermic diagram for Asara area (1986-1994) Fig 2.3

-100

102030405060708090

Jan

Feb

Mar

Apr

May Ju

n

Jul

Aug

Sep Oct

Nov

Dec Aver rain

Aver of mean temp

0102030405060708090

Jan

Feb Mar AprMay Ju

n Jul

Aug Sep Oct Nov Dec

Months/year

Ave

r. r

ain/

mm

-5

0

5

10

15

20

25A

ver.

tem

/cen

t.deg

.

Aver rain

Aver of mean temp


13

Table 2.2 Summary of average monthly precipitation and temperature of Nisaa station:

Month


Average Of mean tem-perature(º C)


Average Of min. tem-perature (º C)

Jan 69.38 -4.81 1.51 -11.11Feb 80.73 -6.16 0.91 -13.23Mar 101.63 -1.19 5.5 -7.86Apr 79.05 4.88 11.2 -1.46May 89.05 10.51 17.29 3.69Jun 23.06 15.5 23.94 7.04Jul 8.61 19.96 28.8 11.1Aug 6.19 20.35 29.1 11.59Sep 5.63 17.21 26.36 8.06Oct 22.1 11.38 19.64 3.08Nov 81.58 6.48 13.55 -0.65Dec 108.8 0.38 6.21 -5.93 Year 675.81


Average precipitation and mean temperature for Nisaa area (1986-1994) Fig 2.4

-20

0

20

40

60

80

100

120

Jan

Feb

Mar

Apr

May Jun

Jul

Aug

Sep Oct

Nov

Dec Aver rain

Aver of mean temp


14

Ombrothermic diagram for Nisaa area (1986-1994) Fig 2.5 Table 2.3 Summary of average monthly precipitation and temperature of Shahristanak station:

Month





Jan 57.63 -5.5 1.95 -12.91Feb 66.06 -6.3 1.49 -14.06Mar 87.69 0.2 5.85 -7.15Apr 71.81 3.84 11.76 -2.48May 75.06 10.2 17.56 2.54Jun 24.5 12.56 23.6 4.36Jul 5.81 18.63 28.29 8.94Aug 5.75 18.9 28.56 9.23Sep 3.81 18.75 25.6 5.64Oct 18.83 9.45 18.81 0.26Nov 63.06 5.53 12.7 -1.66Dec 77.63 -0.71 5.95 -7.38 Year 557.64


0

20

40

60

80

100

120

Jan

Feb Mar AprMay Ju

n Jul

Aug Sep Oct Nov Dec

Months/year

Ave

r. r

ain/

mm

-10

-5

0

5

10

15

20

25

Ave

r. te

m/c

ent.d

eg.

Aver rain

Aver of mean temp


15

Average precipitation and mean temperature for Shahristanak area (1986-1994) Fig 2.6

Ombrothermic diagram for Shahristanak area (1986-1994) Fig 2.7

-20

0

20

40

60

80

100

Jan

Feb

Mar

Apr

May Ju

n

Jul

Aug

Sep Oct

Nov

Dec Aver rain

Aver of mean temp

0102030405060708090

100

Jan

Feb Mar AprMay Ju

n JulAug Sep Oct Nov Dec

Months/year

Ave

r. r

ain/

mm

-10

-5

0

5

10

15

20

25

Ave

r. te

m/c

ent.d

eg.

Aver rainAver of mean temp


16

Table 2.4 Summary of average monthly precipitation and temperature of Sad Amir Kabir station:

Month

Average precipitation (Mm)




Jan 46.11 1.91 5.94 -2.16Feb 51.18 1.08 5.45 -3.34Mar 58 5.35 10.03 0.59Apr 62 10.9 16.16 5.6May 52.09 15.44 21.35 9.48Jun 14.65 20.89 27.96 13.78Jul 2.85 24.65 32.05 17.21Aug 0.86 24.84 32.29 17.38Sep 1.51 21.8 29.34 14.2Oct 9.84 16.76 22.55 9.69Nov 41.85 11.7 17.03 6.35Dec 65.21 6.08 10.21 1.9 Year 406.15


Average precipitation and mean temperature for Sad amir kabir area (1986-1994) Fig 2.8

0

10

20

30

40

50

60

70

Jan

Feb

Mar

Apr

May Ju

n

Jul

Aug

Sep Oct

Nov

Dec

Aver rain

Aver of mean temp


17

Ombrothermic diagram for Sad amir kabir area (1986-1994) Fig 2.9

2.4. Geology

This area belongs to Alborz Mountains, which are tectonically active. Geographically, the area is di-vided into three segments: 1) Southern tuff; these layers are folded and faulted which leads to gentle slope with high wide spread. 2) Central mountains are formed and represented by a sequence of lime-stone and dolomite, with high steep slope and lacking of vegetation cover, and 3) Northern tuff; this part of the area is covered by uniform distributed tuff with high crests and gentle slope. The main part of these layers is calcareous tuff. In general these layers are extensively folded and weathered. Ac-cording to the continuous tectonic movement, the area is encountered with high folding which leads to east-west orientation of crests (Golrang: study report, 1995).

2.5. Soil

Regarding the morphology, the soils of the study area have ochric and mollic surface horizons and cambic-calcic subsurface horizons. Some of the soils have no diagnostic horizon (c-horizon); gypsum accumulation and little salinity are common observed. These soils have variable thickness, and includ-ing young not developed soil (entisols) and moderate developed soils (inceptisols) with different par-ent material. The soil texture is range from very heavy (clay) to very light texture (loamy sand). The pH varies from 5.3 to 8.2. The organic matter content is limited (3%-7%) (Golrang: study report, 1995).

0

10

20

30

40

50

60

70

Jan

Feb Mar AprMay Ju

n Jul

Aug Sep Oct Nov Dec

Months/year

Ave

r. r

ain/

mm

0

5

10

15

20

25

30

Ave

r. te

m/c

ent.d

eg.

Aver rainAver of mean temp


18

2.6. Vegetation

According the study that was carried out during 1993, this area consists of more than 17 types of vegetation distributed all over the area. The vegetation types show the different life form such as tree, shrubs and forbs and grasses. The life from of the grasses was the most dominant life form in the area. The crown cover of the grasses was range from 27% to 40%. About 70 plant species of forbs and grasses were inventoried, but for the shrubs and trees life form, the plant species were limited and oc-curred along valleys (Golrang: study report, 1995).

2.7. Land use

According to FAO method of soil classification, generally this area has shallow soil cover and very steep slope with severe erosion. The land capability is limited mostly cover by bare land with some scattered orchards and tree plantations along riversides and the permanent springs. There are some grassland interspaced the mountain used traditionally seasonal pastures area for grazing only sheep and goats of the local inhabitants of the region. The land tenure of the orchards is private. However some part of this area uses for tourism purposes and sport sites during the winter. Moreover the reser-voirs of the dam lake were used for fish production (fish ponds) (Golrang: study report, 1995).


19

3. Methods

From a methodological point of view, the study has been preceded by the reconnaissance survey dur-ing the field visit based on the existing vegetation map of the area and preliminary aerial photo inter-pretation with the subsequent verification in the field for Juniper delineation. Moreover for handling raster and vector data, the ILWIS 3.11 program, which is GIS software, was used. It integrates image processing capability a tabular database and conventional GIS characteristic. According to this three stages are recognized for this study (Figure 3.1).

3.1. Pre-fieldwork stage

3.1.1. Image processing

Raw digital images usually contain geometric distortions so significant that they cannot be used as maps (Thomas and Kiefer, 1994). The geometric correction was done for the geometric distortions of satellite image data that causes due to sensor geometry, scanner and platform instabilities, earth rota-tion and earth curvature. This correction was done for geometric distortion of satellite image data. The correction process employs geographic features on the image called Ground Control Points (GPCs) whose position are known and cleared on the image using the topography map (scale; 1:50 000, 1997). Affine transformation using the nearest neighbour method as a resampling procedure was applied.

3.1.2. TCC

Before creating a colour composite, a correlation matrix for all bands of the Thematic Mapper (ETM, 2000) Landsat image was calculated. Based on the spectral characteristics and the result of the corre-lation, bands (7, 4 and1) were selected for a colour composite. A true colour composite was created using band 7, band 4(Near infra-red) and band 1 assigning each band to one of the basic colour; Red, Green, and Blue respectively (Figure 3.3).

3.1.3. Image classification

Classification is the process, by which pixels which have similar spectral characteristics, and which are consequently assumed to belong to the same class are identified and assigned a unique colour (Paul and Clare, 2000). The overall objective of image classification procedures is to automatically categorize all pixels in an image into land cover classes or themes (Thomas and Kiefer, 1994). There are two main forms of classification: supervised classification and unsupervised classification. Meanwhile supervised classification was used. Based on spectral characteristics, six lands cover classes namely; Juniper, Grassland, Bare soil, Orchards, Rocks and Water were identified and sam-pled. Using the sample set, a maximum likelihood classification was done. The maximum likelihood classifiers assume that the feature vectors of each class are statistically distributed according to a mul-tivariate normal probability density function. The training samples are used to estimate the parameters of the distributions. The boundaries between the different partions in the feature space are placed where the decision changes from one class to another.


20

Flow chart of study activities Fig 3.1

Interpretation Processing Rasterizing

Topography Map Veg. Map (Polygon) TM Image Aerial Photos

Slope Altitude Aspect

Select Samples

Interpolation

Dem Map Land Cover Map

Forest Patches Map

Patches Size

Patches Isolation

Jun. Density Jun. Canopy Cover Plant Species Number

Fieldwork Data

Physical & Soil Factors

Statistical Analysis

Veg. Map (Raster)

Juniper Canopy Cover Model Juniper Density Model Forest Island Model


21

Fig 3.2: Vegetation map 1993


22

The maximum likelihood classifier is generally considered to be the most powerful but is also consid-ered the most computers intensive (Paul and Clare, 2000). Thus according to the differentiation of the different land cover, the Juniper patches were distinguished in the study area. After classification method was carried out, a classified image may noisy with isolated pixels of one class surrounded by pixels of another class. A mode filter can be applied to the thematic image that replaces the isolated pixel by the most frequently occurring class within the filter window. In this method a majority filter was used to refine the thematic image. Taking samples of ground truth during the fieldwork as check-points and using these points for checking their classes against the classified pixels, the accuracy of the classification results was determined.

3.1.4. NDVI

The Normalized Difference Vegetation Index (NDVI) was used to correct the confounding effect of soil-vegetation interaction, (Franklin et al., 1991, in Farhang, 1997). NDVI is defined as the ratio of difference between the near infrared (NIR) and red (R) reflectance to their sum or (NIR-R)/(NIR+R). The NDVI is dimensionless and can take values from 0-1. 127 multiply the product in order to avoid fractional values (less than 1) and a constant (128) is added to avoid negative values. The mathemati-cal expression of the created NDVI map (Fig 3.4) is: NDVI= ((band4-band3)/(band4+band3))* 127 + 128 Band4 = the fourth band of ETM image between 0.78 - 0.9 um Band3 = the third band of ETM image between 0.63 - 0.69 um The NDVI was used as one variable of the raw data. This variable was very necessary for detecting the number of plant species as well as the Juniper density and canopy cover.

3.1.5. Altitude

The topography maps (scale; 1:50 000) were digitised and rasterized. A digital elevation model (DEM) was created by interpolation of the digitised topography map of the study area. In this map each raster cell (pixel) has an attribute value representing the topographic elevation. The altitude map (Figure 3.5) was produced from the classification of the DEM map into four classes. The altitude map is one of the base maps used for the ground observation program. Table 3.1 shows the range of each altitude class including the number of ground observations. Table 3.1 the altitude classes:

Class Altitude range (m) Number of ground observations 1 <2000 15 2 2001-2200 23 3 2201-2400 35 4 >2400 15


23

Fig 3.3: TCC Map and sample points

Fig 3.4: NDVI Map of the study area


24

Fig 3.5: Altitude Map and sample points


25

3.1.6. Aspect

The aspect map was created indirect from DEM map using the following formula to produce the as-pect map in degrees. ASPECTD = RADDEG (ATAN2 (DX, DY) + PI) ASPECTD = Aspect in degrees. RADDEG & ATAN2 are internal map/table calculation functions. DX & DY are the output maps from the linear filter dfdx, dfdy respectively. PI is the pixel size of DEM raster map. The aspect map in degrees was further classified into four classes using the slicing operation (Fig.3.6). Table 3.2 shows the main aspect classes including the number of ground observations of each aspect within the study area. Table 3.2 the main aspect classes:

Aspect (0-45) & (316-360) 46-135 136-225 226-315 Class N E S W Nr. Of ground observations 29 23 16 20

Fig 3.6: Aspect Map and sample points


26

3.1.7. Slope

The slope map (Fig. 3.7) was created from DEM map using the following formula; SLOPEPCT = 100*HYP (DX, DY)/PIXEL SIZE SLOPEPCT = slope in percentage. HYP is an internal map/table calculation function. DX and DY are the output maps from the linear filter DFDX, DFDY respectively. PIXEL SIZE is a pixel size of a raster map (DEM). The output slope map was classified into four classes using the slicing operation. The table 3.3 shows the classes of slope including the ground observations. Table 3.3 the slope classes.

Slope range in % 0-20 20-40 40-60 >60 Class 1 11 111 1V Nr. Of ground observations 14 28 26 20

Fig 3.7: Slope Map and sample points


27

3.1.8. Sampling Technique

According to preliminary interpretation of aerial photograph and the vegetation map (1993) of the study area, six forest islands existed in the study area. The forest islands lie in the middle of the catchments area (North of Karaj city). The pattern distribution of individuals within a species, and pattern of species, are of major concern in techniques selection and used. If all distribution were ran-dom, or even nearly so, then the measurement would not yield biased result in most areas. Random distributions imply that no pattern is present in the measure obtained which is already obtained from a random sampling process to provide unbiased estimates in the measure (Charles D. Bonham, 1988). According to the objectives of the study, all life forms of vegetation types, which include shrub or tree form, will have more variation in size over species than will grass or forbs in vegetation type. The sample size 10*10 meter was used to cover all life forms, grasses, shrubs and trees (Ecology; science and practice, 2001). Usually these plots are nested; that 1 square meter is placed in the corner of the 16 square meter plot and the latter is placed in the corner of the 100 square meter to measure grass form, shrub and tree forms respectively (Charles D. Bonham, 1988). Sample sites were selected according to mapping training units resulted from two levels stratification. First altitude map was stratified with forest patches map produced from the pre-classification of satel-lite image ETM (2000). In the second level the output map from the above stratification was further stratified with the slope map of the study area. Depending on the final findings 96 sites were pre-selected for ground observation points. Due to difficult accessibility only 88 samples were observed. Field data was generated in Excel data based and imported into ILWIS 3.11 as a point map, which was then overlaid on the classified land cover map. Based on the juniper patches, a total of 88 sampling points were selected for field observations.

3.1.9. Study materials

To achieve the objectives of the research, the following materials were used: • Aerial photographs with scale 1:50 000 dated Aug. and Sep 1955 obtained from the army

geographical centre of Tehran (Iran). • Topographic maps number 6262 (1, 11,111,1V) at scale 1:50 000 dated 1997 obtained from

the military geographic centre of Tehran (Iran). The datum is European 1924 (Iran), coordi-nate system is Latlon and the projection is Transverse Mercator (Zone 39).

• Scanned and digitised vegetation and lithology maps from SCWSRC Ministry of Agriculture. • Landsat 7 TM satellite image of July 2000. • Hard copy of soil map with scale of 1:250 000 (1992) obtained from soil research centre. • Other materials for field observations included; GPS (Magellan 315), augur, hammer, knife,

30 meters measuring tape, 40 meters rope and sample ploy bags.


28

3.2. Fieldwork stage

During the field observations, the topography and the vegetation maps were used to find the location of ground observation points. The relevee sheets (appendix 1) were used to record the characteristics of ground observation sites and collected data. The general information of the study area and secondary data (Topographic map, climatic data, and scanned and digitised geology and vegetation maps) was collected from Jihad Or-ganization. All data collected per sample were entered on standard relevee sheet (appendix 2). These data included the following: 1) Terrain characteristics: GPS was used to navigate to each the selected sample point recording the coordinate and the altitude. GPS was adjusted using the European Datum 1924 (Iran) and Transverse Mercator projection. The Latlong and UTM coordinate systems were read and recorded. As such the aspect readings like slope and slope type as well as relief type were extracted from the aspect, slope and relief maps digitised from the topographic map. 2) Soil characteristics: According to FAO method of soil classification, generally this area is considered as (M1.1) moun-tainous with high elevation without any soil cover or some part with very shallow depth and stoniness as well as with medium to heavy texture. The soil was augured to a depth up to 70 cm which is the maximum depth existed in the ground observation points, and soil samples were collected from the mixture of the pit soil. The soil samples were tested in the soil laboratory of Soil Conservation and Watershed Management Research Centre in Iran. Only the soil depth and surface stoniness was meas-ured in the field, but the other parameter, which includes soil texture, soil moisture, soil pH, soil EC and soil CEC were tested in the laboratory (SCWSRC). 3) Vegetation data: All plant species in the field sample plots were recorded by their botanical names as identified by the Botanists from the National Herbarium of Iran (Dr. Assadi & Dr. Mozaffarian). The plant species were collected and given a code and nickname. These specimens were named in the Ira-nian National Herbarium (INH). A total of 56 plant species were inventoried in the field observation points (see appendix 3). The other variables including density of juniper trees, number of plant spe-cies, juniper canopy cover and total canopy cover were recorded. Crown cover was estimated by di-rectly measure of the crown diameter of the plant species individuals. The measurements are usually taken in two directions (longest and shortest dimensions) and then averaged the diameter followed by the computation of the crown area for each species ((D/2) 2 * �� (Dawkins, 1963, in Charles D. Bon-ham, 1988). Density is defined as the number of individuals stem or plants occur per unit ground area. Density is generally considered to be readily obtainable and easily understood characteristics of vege-tation (Pieper, 1973, in Charles D. Bonham, 1988). Juniper density was estimated according to Fonda (1974, in Charles D. Bonham, 1988) considered all trees greater than 2.5 cm in diameter were counted by and expressed as the number of trees per hectare.


29

3.3. Data integration and analysis

The field data was entered into a database in MS Excel. The ground observation points were placed in column of spreadsheet and the other variables were placed in rows. Statistical software SPSS 10.0.5 (SPSS Inc.1999) was used to test the correlation of the available variables. Minitab 13.1(Minitab Inc. 2000) was used to carry out the normality test of raw data and also used to plot regression and box diagrams, where as SYSTAT 7.0 was used for regression analysis. To answer the fore stated research questions and addressing the specific research hypothesis a number of explanatory statistics were used. Prior to any statistical analysis, the nature of the distribution of the data was explored to test the as-sumption that the data is normal versus the data distribution is not normal. Anderson-Darling, Ryan-Joiner (Similar to Shapiro-Wilk) and Kolmogorov-Smirnov normally test were used. To test the influence of juniper density with the number of plant species, the influence of juniper patches sizes and isolation with plants diversity, and the categories of different physical and soil fac-tors as stated in the previous hypothesis, box plot methods were used. The possible relationships between number of plant species, the juniper density and canopy cover, physical and soil factors, and the different bands of the Landsat 7 image as well as NDVI were deter-mined by correlation analysis. Moreover regression analysis was carried out using stepwise forward multiple linear regressions. The model was run to all parameters.


30

4. Result

4.1. Exploratory analysis The field observation data was imported to Minitab 13.1 software. The data showed that the Juniper density range from zero to 700 trees per hectare with the average density of 230 trees per hectare while the Juniper canopy cover range from zero to 62.8 percent with average of 17.7 percent. Minitab 13.1 software was used to test the normality of raw data. The normality test showed that fif-teen variables were normal without data transformation. The thermal bands of TM image (B61, B62) were normal after improving the data by the square root but the Juniper canopy cover was improved by square root + 0.5 due to the occurrence of zero value among the data (Bartlett, 1947, in Charles D. Bonham, 1988). Different methods of data transformation such as the square root, reciprocal and re-ciprocal of square root were attempted for the remaining variables respectively but the result showed no improvement towards a normal distribution of data. Further analysis with Anderson-Darling, Wilk test and Kolmogorov-Smirnov normality tests were used to test the normally distributed. The four variables including aspect, soil EC, soil pH and soil moisture were not normally distributed (P-value <0.01). Although the logarithmic transformation data could not be used because it removes all the ob-servations with zero value and changes those with values less than 0.5 into negative, which does not make sense such as aspect and EC can not be negative.

(A) (B) Fig 4.1: Normality test

(A) Histogram of the original data, (B) Probability plot of the original data tested by Kolmo-gorov-Smirnov (P-value >0.15 N=88).


31

(A) (B)

(C) (D)

Jun. c. cov% = Juniper canopy cover %

Fig 4.2: Normality test (A) & (B) Histogram and probability plot of the original data (P-value <0.01). (C) & (D) Histogram and probability plot of the transformed data by square root +0.5 and tested by Anderson-Darling, Wilk test and Kolmogorov-Smirnov (P-value: 0.099, P-value>0.1 and P-value >0.15) respectively.


32

4.2. Correlations analysis

Using statistical software SPSS 10.0.5 (SPSS Inc.1999), the correlation analysis of the data was achieved. Thus simple correlation analysis gives a reasonably clear picture of the overall structure of the data set for the 88 sample points as shown in the table below. The underlined figure shows the significant correlations. Table 4.1 Pearson correlation matrix.

Variable Altitude Slop e Species number

Jun. density

Soil depth

Soil moisture

Soil pH

Soil CEC

NDVI

Altitude

Slope

.338

Species number

-.306 -.138

Juniper density

-.250 -.276 .310

Soil depth

-.450 -.321 .293 .198

Soil mo isture

-.454 -.027 .314 .329 .433

Soil pH

.196 .204 -.203 -.315 -.147 -.318

Soil CEC

-.446 -.031 .456 .209 .284 .519 -.164

NDVI

-.039 .124 .080 .029 .134 .169 .048 .177

4.3. Descriptive statistics

4.3.1. Juniper density versus number of plant species

The result of the analysis of the density of Juniper with the number of plant species showed that there was a significant difference in the median of Juniper density against the number of plant species. Sta-tistical analysis (Mann-Whitney test) showed that the densities of Juniper against the different number of plant species are significantly different (Test of ETA1=ETA2 vs. ETA1not=ETA2 is significant at 0.000). Thus the median of Juniper density is not equal the median of the number of species. The me-dian of Juniper density and the median of number of plant species are 2 and 5 respectively (Fig 4.3).


33

Fig 4.3: Number of plant species versus number of Juniper trees

4.3.2. Juniper canopy cover versus number of plant species

The result of the analysis of Juniper canopy cover with the number of plant species showed a signifi-cant difference in the median of Juniper canopy cover. Statistical analysis (Kruskal-Wallis test: R² =0.32, df = 11, P-value0.001) showed that the canopy covers of Juniper against the number of plant species are significantly different. Thus the median of Juniper canopy cover is not equal the median of number of plant species, but as in (Fig 4.4) the median of Juniper canopy cover are similar at number of plant species of 5, 6 and 9.

Fig 4.4: Juniper canopy cover % versus number of plant species


34

(A) (B) Class 1 = <2000 2 = 2001-2200 2 = 2201-2400 4 = >2400 Fig 4.5: Altitude versus number of Juniper trees

4.4. Environmental factors and the density of Juniper

4.4.1. Altitude

In figure (4.5.A) the non-linear regression of Juniper density with altitude showed the extremely low negative regression of the density of Juniper with altitude is significant association (R² =0.08, P-value=0.019, df = 3). Thus as altitude increases from class1 to class 2 the Juniper density decreases, but after that as in box diagram (Fig 4.5.B) the density of Juniper was fixed as the altitude increases.

4.4.2. Slope

The result of linear regression of density of Juniper with slope showed that the relationship is signifi-cant (R² = 0.08, P-value=0.009, confidence 95%) with inverse action, as slope increases the density of Juniper decreases, but the regression is extremely low (Fig 4.6).

Fig 4.6: Slope % versus number of Juniper trees


35

(A) (B)

Fig 4.7: (A) Number of Juniper trees versus aspect (B) Number of Juniper trees versus soil pH classes

4.4.3. Aspect

In (Fig 4.7.A) according to the analysis, it was found that there was no significant influence of the aspect on the density of Juniper. Statistical analysis (Kruskal-Wallis test) showed that the Juniper densities in different aspects are not significantly (R² =0.01, df=3, P-value=0.778).

4.4.4. Soil pH

The result of statistical analysis of Juniper density versus the soil pH showed that, there was a very week significant relationship between soil pH and the density of Juniper (Kruskal-Wallis test: R²=0.12, df=2, P-value=0.002). The density of Juniper was high in acidic soil, but it was the same in neutral and alkaline soil (Fig 4.7.B).


36

(A) (B) Fig 4.8: (A) Number of Juniper trees versus soil EC (B) Number of Juniper trees against soil texture c=clay c-l= clay loam sa-c-l= sandy clay loam l= loam si-c= silty clay si-c-l= silty clay loam

4.4.5. Soil EC and soil texture

The relationship of Juniper density with soil EC is not significant (R²= 0.015, P-value= 0.748, � =0.05) (Fig 4.8.A), while the result of analysis of density of Juniper with the soil texture showed that was not significant. Statistical analysis (Kruskal-Wallis test) showed that the soil texture was ex-tremely week influence on the density of Juniper (R² = 0.07, P-value=0.183, df = 5). As in the box diagram, the median of Juniper density versus soil texture is similar except of loamy and sandy clay loam (Fig 4.8.B).

4.4.6. Soil moisture content

In figure (4.9.A) it was found that there was a significant influence of the moisture on the density of Juniper trees. Statistical analysis (linear regression) showed that the density of Juniper trees was mar-ginally influenced by the increase of soil moisture (R²=0.108, P-value=0.002, df = 1, � = 0.05). The non-linear regression was run to test the relationship between the density of Juniper and soil depth. The result showed that there was not significant relationship (R²=0.054, P-value=0.064, df = 1, � =0.05). The result showed that as the soil depth increases the density of Juniper increases up to 50 cm depth and then declines when the soil depth increases (Fig 4.9.B).

(A) (B) Fig 4.9: (A) Number of Juniper trees versus soil moisture% (B) Juniper trees versus soil depth


37

4.5. Environmental factors and number of plant species

4.5.1. Altitude

An inverse relation was observed between the altitude and the number of plant species (Fig 4.10.A). Statistical analysis showed the too low significant relation at confidence level of 95% (R² = 0.108, P-value=0.002, df = 1). The Box and Whisker Diagram showed that the number of plant species at alti-tude class 1,class 2 and class 3 was the same (Fig 4.10.B).

(A) (B) Fig 4.10: Altitude versus number of plant species

Fig 4.11: Slope % versus number of plant species Class 1 = 0-20 % 2 = 20-40 % 3 =40-60 % 4 = > 60 %

4.5.2. Slope

The regression plot of slope versus the number of plant species showed that the number of plant spe-cies decreases as the slope steepness increases. Statistical analysis showed that there was no signifi-cant influence by slope steepness on the number of plant species (R²=0.023, P-value=0.159, df = 1, � = 0.05) (Fig 4.11).


38

4.5.3. Aspect

Statistical analysis (Kruskal-Wallis test) of the number of plant species versus the aspect showed that there is difference in the median of number of plant species against the aspect but was not significant (R²=0.058, P-value=0.121, df = 3). The Box diagram (Fig 4.12.A) showed that in the east and the north the median of the number of plant species was equal.

(A) (B) Fig 4.12: (A) Aspect versus number of plant species (B) Soil texture versus number of plant spp.

4.5.4. Soil texture

A comparison of the plant species in the different soil texture (clay, clay loam, loam, sandy clay loam, silt clay and silt clay loam) showed that there is no significant difference in the median of number of plant species on the clay and clay loam. The two soil textures however, are significantly different from that on the other soil texture (Fig 4.12.B). Detailed statistical analysis succeeded to test a week significant difference in the number of plant species on the different soil textures (Kruskal-Wallis test: R²=0.24, P-value=0.000, df = 5).


39

(A) (B) Fig 4.13: (A) Soil depth versus number of plant spp. (B) soil moisture % versus plant spp.

4.5.5. Soil depth

The regression analysis of number of plant species and the soil depth showed that there is a significant relation (Fig 4.13.A). Statistical analysis by using the linear stepwise forward regression described that the relation was significant (R²=0.105, P-value=0.002, df = 1, � = 0.05). While the statistical analysis (non-linear very week regression) showed that there was significant association between soil moisture and the number of plant species (R²=0.124, P-value=0.003, df = 1, � = 0.05). Thus as the soil moisture increases the number of plant species increases (Fig 4.13.B).

(A) (B) Fig 4.14: (A) Soil CEC versus number of plant spp. (B) Soil pH versus number of plant spp. Acidic pH < 6.5 Alkaline pH > 7.5 Neutral pH 6.5-7.5

4.5.6. Soil pH and CEC

The regression plot showed that there is significant influence by soil CEC on the number of plant spe-cies (Fig 4.14.A). The linear stepwise forwards regression analysis described a week significant rela-tion between the number of plant species and the soil CEC (R²=0.191, P-value=0.000, df = 1, � = 0.05). The soil pH and the number of plant species were analyzed by Kruskal-Wallis test. The result showed that there was not significant in the median of number of plant species on the different soil pH (R²=0.531, P-value=0.080, df = 40). The higher median of the number of plant species was observed on the acidic soil while intermediate median occurred on the soil with the neutral medium (Fig 4.14.B).


40

4.6. Juniper canopy cover and the environmental factors

4.6.1. Altitude

The linear stepwise regression analysis of the canopy cover of the Juniper and the altitude was per-formed (Fig 4.15.A). Statistical analysis showed that there was very low significant influence of the elevation on the Juniper canopy cover (R²=0.192, P-value=0.000, df = 1, � = 0.05). The box plot showed no significant difference in the median of Juniper canopy covers on class two and three of the altitude (Fig 4.15.B). The two however, are significantly different from that on the class one and class 4 of altitude.

(A) (B) Fig 4.15: Altitude versus Juniper canopy cover %

4.6.2. Slope

Statistical analysis using the linear stepwise regression showed that there was significant relation be-tween the canopy cover of Juniper and the slope (Fig 4.16.A). The detailed analysis described ex-tremely week significant relation (R²=0.071, P-value=0.012, df = 1, � = 0.05), but on the box plot, the median of the Juniper canopy cover was not significant on slope class two and slope class three (Fig.4.16.B).

(A) (B) Fig 4.16: Slope % versus Juniper canopy cover %


41

(A) (B) Fig 4.17: (A) Aspect versus Juniper canopy cover % (B) Soil pH versus Juniper canopy cover %

4.6.3. Aspect

A comparison of the Juniper canopy cover in the different aspect showed that there is almost no sig-nificant difference in the median of Juniper canopy cover on east, north and west aspect. The three however, are significantly different from that on the south aspect (Fig 4.17.A). Detailed statistical analysis failed to test any significant difference in the Juniper canopy cover on the different aspect (Kruskal-Wallis test: R²=0.013, P-value=0.714, df = 3, � = 0.05).

4.6.4. Soil pH

Statistical analysis (Kruskal-Wallis test) showed that there was significant difference in the median Juniper canopy cover versus pH classes, but it was very low (R²=0.109, P-value=0.012, � = 0.05). The intermediate median was observed on the neutral soil (Fig 4.17.B).


42

4.6.5. Soil texture and soil moisture content

It was found that there was significant influence of different soil textures on the Juniper canopy cover (Fig 4.18.A). The median of the canopy cover of Juniper is highest in silt clay soil texture while an intermediate median of Juniper canopy cover was not significantly on clay, clay loam and silt clay loam soil. Statistical analysis (Kruskal-Wallis test) showed too low significant influence in different soil texture (R²=0.154, P-value=0.009, � = 0.05). The result of statistical analysis (linear regression) of the canopy cover of Juniper versus the soil moisture showed that there was very week significant (R²= 0.157, P-value=0.000, df = 1, �= 0.05) (Fig .4.18.B).

(A) (B) Fig 4.18: (A) Soil texture versus Juniper canopy cover % (B) Soil moisture % versus Juniper canopy cover %

4.6.6. Soil depth

The regression analysis as in the regression plot (Fig 4.19.A) showed that there was a marginal rela-tionship between the canopy cover of Juniper and the soil depth (R²=0.113, P-value=0.001, df = 1, � = 0.05). Likewise the statistical analysis of Juniper canopy cover and the soil CEC using the linear stepwise forwards regression showed that there was too low significant association between Juniper canopy cover and the soil CEC (R²=0.103, P-value=0.002, df = 1, � = 0.05). Thus the result showed as soil CEC increases the canopy cover of Juniper increases (Fig 4.19.B).

(A) (B) Fig 4.19: (A) Soil depth versus Juniper canopy cover % (B) Soil CEC versus Juniper canopy cover %


43

(A) (B) Fig 4.20: NDVI versus number of plant species

4.7. Number of plant species versus NDVI

The result of non-linear regression analysis showed that there was no significant relationship between NDVI and the number of plant species (R²=0.015, P-value=0.763, df = 1, � = 0.05) (Fig 4.20.A). The box plot showed that the maximum number of plant species in class 2 of NDVI, but after that when NDVI values increase as in class 3 and class 4, the number of plant species decreases (Fig 4.20.B).

4.8. Juniper density and canopy cover versus NDVI

The result of non-linear regression analysis of NDVI with the Juniper density as presented in (Fig 4.21.A). Statistical analysis showed that there was no significant relationship (R²=0.052, P-value=0.786, � = 0.05). The regression plot showed that the density of Juniper increases as NDVI val-ues increase up to 102 after that the density decreases as the NDVI values increase. Meanwhile the statistical analysis (Kruskal-Wallis test) of the Juniper canopy covers versus the NDVI was not sig-nificant (R²=0.075, P-value=0.056, df = 3). The Box plot in (Fig 4.21.B) showed that there was no significant difference in the median of Juniper canopy cover at class 2, class 3 and class 4 of NDVI.

(A) (B) Fig 4.21: (A) Number of Juniper trees versus NDVI (B) Juniper canopy cover % versus NDVI


44

4.9. NDVI versus Altitude

Figure (4.22.A) shows the relationship between NDVI and altitude. The non-linear curve showed that the NDVI values increase with altitude to 2200 m, but after this elevation the NDVI values decrease with increasing the altitude. The statistical analysis described non-significant relationship (R²=0.019, P-value=0.721, df = 1, � = 0.05). In the Box plot can be seen in the altitude class 3 the NDVI values decrease (Fig 4.22.B).

(A) (B) Fig 4.22: NDVI versus altitude


45

4.10. Multiple Regression Analysis

4.10.1. Juniper density as a response

To test a combined effect of the group of independent variables (soil and environmental factors) on the dependent variable here in this case the density of the Juniper, a multiple linear regression analysis was applied. Only two explanatory variables were selected upon stepwise forward multiple regression analysis at confidence level 95% (� = 0.05). The result (Table 4.2) showed that, the combination of the soil moisture percent and the slope were significant effect on the Juniper density at R² = 0.179, P-value = 0.000 (Table 4.3). The model shows that (Table 4.4) the density of Juniper is high in the moist soil and gentle slope location, but the model is very week. Table 4.2: Stepwise regression analysis using two explanatory variables for Juniper density

Step 1 Step 2 Constant 1.605 2.383 Soil moisture % 0.193 0.189 T-value 3.226 3.268 P-value 0.002 0.002 Slope -0.018 T-value -2.714 P-value 0.008 S 1.464 1.413 R² 0.108 0.179 R² (adj) 0.098 0.160

Table 4.3: Analysis of variance Source DF SS MS F P Regression 2 37.025 18.512 9.273 0.000 Residual 85 169.691 1.996 Total 87 206.716

Table 4.4: Regression analysis of variance

Effect (Predictor) Coefficient Std Coeff. T P Constant 2.383 0.0 6.177 0.000 Soil moisture % 0.189 0.321 3.268 0.002 Slope -0.018 -0.267 -2.714 0.008

Juniper density = 2.383 + 0.189 * soil moisture % - 0.018 * slope %


46

4.10.2. Juniper canopy cover as a response

To establish combined effect of the group of independent variables (soil and environmental factors) on the response variable here in this case the Juniper canopy cover, a multiple linear regression analy-sis was carried out. Only three explanatory variables were selected upon stepwise forward multiple regression analysis at confidence level 95% (� = 0.05). The result (Table 4.5) showed that, the combi-nation of the altitude, soil pH and number of plant species were significant effect on the Juniper can-opy cover at R² = 0.281, P-value = 0.000 (Table 4.6). The model shows that (Table 4.7) the canopy cover percent of Juniper is high in the low elevation and acidic soil. In fact the model was very week, and the regression of step 2 was selected to build the model due to the significant correlation between altitude and number of plant species in step 3. Table 4.5: Stepwise regression analysis using three explanatory variables for Juniper canopy cover

Step 1 Step 2 Step 3 Constant 87.836 132.440 108.794 Altitude -0.031 -0.028 -0.023 T-value -4.335 -3.886 -3.170 P-value 0.000 0.000 0.002 pH -7.255 -6.374 T-value -2.556 -2.275 P-value 0.012 0.028 Number of plant species 1.330 T-value 2.237 P-value 0.025 S 13.305 12.897 12.603 R² 0.179 0.238 0.281 R² (adj) 0.170 0.220 0.225

Table 4.6: Analysis of variance Source DF SS MS F P Regression 2 4412.667 2206.353 13.265 0.000 Residual 85 14137.485 166.323 Total 87 18550.152

Table 4.7: Regression analysis of variance Effect (Predictor) Coefficient Std Coeff. T P Constant 132.440 0.0 5.637 0.000 Altitude -0.028 -0.375 -3.375 0.000 pH -7.255 -0.247 -2.556 0.012

Juniper canopy cover % = 132.44– 0.028 * Altitude (m) – 7.255 * pH


47

4.10.3. Number of plant species as a response

To test a combined effect of the group of independent variables (soil and environmental factors) on the dependent variable here in this case the number of plant species, a multiple linear regression analysis was applied. Only two explanatory variables were selected upon stepwise forward multiple regression analysis at confidence level 95% (� = 0.05). The result (Table 4.8) showed that, the combi-nation of the soil moisture percent and the slope has very low effect on the number of plant species at R² = 0.244, P-value = 0.000 (Table 4.9). The model shows that (Table 4.10) the number of plant spe-cies is high in soil with high CEC and among the Juniper community. Table 4.8: Stepwise regression analysis using two explanatory variables for the number of plant spp.

Step 1 Step 2 Constant 2.243 2.044 CEC 0.110 0.090 T-value 4.503 3.598 P-value 0.000 0.001 Juniper canopy cover % 0.040 T-value 2.455 P-value 0.016 S 2.187 2.126 R² 0.191 0.244 R² (adj) 0.181 0.227

Table 4.9: Analysis of variance

Source DF SS MS F P Regression 1 96.986 96.986 20.275 0.000 Residual 86 411.378 4.783 Total 87 508.364

Table 4.10: Regression analysis of variance

Effect (Predictor) Coefficient Std Coeff. T P Constant 2.243 0.684 3.282 0.001 CEC 0.110 0.024 4.503 0.000

Number of plant species = 2.243 + 0.110* CEC


48

4.11. Image classification and evaluation

The accuracy of image classification result was determined. The confusion matrix (table 4.11) shows the number of pixels, which falls in different classes. The columns show the ground truth or known cover types and the rows show the pixels actually classified into each land cover category (class). Kappa statistical coefficient was calculated for the classification image obtaining the value of 87.5%, which is considered as an overall accuracy according to Congalton and Mead (1983). Moreover the omission and commission errors were also calculated (table 4.12). However, producer’s accuracies range from 70 percent (Juniper) to 100 percent (water, bare land and orchard) and the user’s accura-cies range from 71.43 percent (range) to 100 percent (water, rock and orchard). Table 4.11: Confusion matrix for accuracy evaluation on the land cover map 2000

Class Water Bare land Juniper Rock Orchard Range Total Water 7 0 0 0 0 0 7 Bare land 0 8 1 0 0 2 11 Juniper 0 0 7 1 0 0 8 Rock 0 0 0 6 0 0 6 Orchard 0 0 0 0 9 0 9 Range 0 0 2 0 0 5 7 Total 7 8 10 7 9 7 48

Table 4.12: Accuracy of the 2000’s classification

Class Producer’s Accuracy (%) User’s Accuracy (%)

Error of Omission (%)

Error of Commission (%)

Water 100.00 100.00 00.00 00.00 Bare land 100.00 72.72 00.00 27.28 Juniper 70.00 87.50 30.00 12.50 Rock 85.71 100.00 14.29 00.00 Orchard 100.00 100.00 00.00 00.00 Range 71.43 71.43 28.57 28.57

Overall accuracy = (7 + 8 + 7 + 6 + 9 + 5)/48 = 87.5%


49

Fig 4.23 The classified maps

The difference between the two maps some pixels of the rock class classified as range and Juniper as you see in the above map.


50

Table 4.13 Forest Island sizes distances and plant species richness of 11 Forest Islands

Island no. Island name Size (ha) Total no. of plant species Distance (Km) 1 Sercheh1 263.43 9 6.007 2 Sercheh2 888.84 14 6.692 3 Kalvan1 19.35 4 12.504 4 Kalvan2 14.13 5 11.682 5 Kalvan3 44.28 6 11.863 6 Hamega 292.59 14 3.83 7 Kalvan4 35.1 5 11.0 8 Rayzamin 139.05 20 5.839 9 Sirachal 175.41 37 10.475 10 Lanyz1 310.14 33 0.0 (source) 11 Lanyz2 24.84 12 0.36

4.12. Statistical analysis of sizes and isolation of forest islands

All the three variables (Table 4.130 plant species and forest island distances and sizes, however, there were normal distributed (P-value > 0.15, P-value > 0.15, and P-value: 0.069) respectively. The manner of selection the source forest island, however, it naturally protected and has a rich flora compared by the other ones. The non-linear (quadratic) regression was performed to test the correlation between the plant species richness (Log10) as a response and patches sizes as an independent. Statistical analy-sis (nonlinear regression) showed that the log10 of plant species richness was significantly influenced by the increase of patches sizes (R² = 70.5%, P-value = 0.026, df = 2, � = 0.05, R² (adj) = 60.6%, n = 9). Furthermore the linear regression between plant species richness and the patches distances was also carried out, the result showed that the correlation was significant (R² = 80.9%, P-value = 0.002, df = 1, � = 0.05, R² (adj) = 77.8%, n = 8).

(A) (B)

Fig 4.24: (A) Total number of plant species as a function of the forest island size (B) Total number of plant species as a function of the average of edge-to-edge distance


51

In this case study, it was found that as the area of the forest island increases (up to 220 ha) the log 10 of the number of plant species increases and then after that the curve goes down when the area of the forest island increases (Fig 4.24. A). For the (Fig 4.24.B) the regression plot of the distance vs. log plant species shows the positive relationship. Finally the multiple stepwise regressions were run but only the variable of log distance was significant as in table (4.14) (R = 80.93%, R – adj = 77.76% R –sq (pred) = 66.07%, df = 1, P-value = 0.002). Table 4.14: Analysis of variance

Source DF SS MS F P Regression 1 0.378 0.378 25.470 0.002 Residual 6 0.089 0.015 Total 7 0.467

Table 4.15: regression analysis of variance Effect (Predictor) Coefficient Std Coeff. T P Constant 1.501 0.0 12.086 0.000 Distance Km -0.068 -0.900 -5.047 0.002

Log (10) of plant species number = 1.501 – 0.068 * Distance (Km)

4.13. Result summary

The result of the multiple regression analysis showed a very week association between the Juniper densities, Juniper canopy cover and the number of plant species on one hand and the environmental factors on the other (R = 17.9%, R = 28.1%, and R = 24.4%) respectively. Furthermore the model of log 10 of the plant species with the forest islands size was significantly to apply for creating the map of plant species richness within the Juniper forest islands.


52

Hamega and Sercheh Forest Islands

Fig 4.25: Juniper density map


53

Fig 4.26: Juniper canopy cover map

Lanyz Forest Islands


54

Fig 4.27: Plant species richness map

Kalvan Forest Islands

Sirachal Forest Island Lanyz Forest Islands Rayzamin Forest Island


55

5. Discussion

5.1. Environmental factors vs. Juniper

The occurrence of �� totally appeared at altitude of 1896-2775m, with no upper tree line. However this range is reasonable, because all the previous studies found that the dis-tribution of this species allover Iran within altitude range of 1500-3300m (Ahmed and Zienolabedin, 1978). In Oman this species was distributed in the Hajar Mountains of altitude range of 1375-3009m (Gardner and Fisher, 1996). It is generally a higher altitude species; it occurs from 500-3800m, but mostly between 1200-3000m Kitamura (1960, in Farjon et al. 1992). Also it was noticed that the den-sity and canopy cover percent were become low at high elevation because here most of the Juniper occurred in Wadis sheltered by the mountains i.e. on flat terrain and convex relief (ca. 85.88%). This finding was illustrated by Gardner and Fisher (1995) during their study of �� in the northern mountains of Oman. They found that in general, taller trees in relatively better condition are growing in and around the wadi habitat and smaller trees in relatively poorer condition are grow-ing on the slightly higher ground outside the wadis and on the higher, northern side of the valley. Furthermore the dense Juniper trees occurred in the moist substrates. This suggests that the high den-sity and a better growth of Juniper trees could be due to soil moisture content caused by topography and microclimate. However, low density and relatively low canopy cover describe trees on high eleva-tion. According to the analysis, the density of �� was not significantly affected by the slope direction i.e. the density of Juniper is similar in all aspects. However, this sub-species is much more resistant to drought and radiation than the typical subspecies due to thick cuticle layer Browicz & Zielinski (1982, in Farjon et al. 1992). The other site where this species of Juniperus has been studied in some detail is in Baluchistan, Paki-stan (Ahmed, Ahmed & Anjum 1989; Ahmed, Nagi & Wang 1990; Ahmed et al. (1990, in Fisher and Gardner, 1995). In contrast to Hayl Juwari, the Juniperus forests of Baluchistan has been attributed to anthropogenic disturbance due to overgrazing and heavy logging while there appears to be little in Hayl Juwari. Differences in stand structure between the Juniper of Baluchistan and Hayl Juwari could be at least partly explained by climatic differences (Fisher and Gardner, 1995). Though the forest of Baluchistan receive a similar total annual rainfall to Jebel Akhadar, the maximum rainfall occurs in July, the hottest month, and snow falls in the winter Ahmed et al. (1990, in Fisher and Gardner, 1995). For the soil factors we investigated the effect of pH value, the study found that both the Juniper den-sity and canopy cover percent were relatively high at low pH value for instant at soil pH of 6.3 - 6.5, in general the study pointed out that the Juniper tree existed within the range of 6.25-7.98 this value is a neutral range which is most suitable for most of the plant species even for the commercial crops (Richard, 1984). The pH values usually integrate with soil cation exchange capacity measurements. Despite the above factors, the distribution of �� species may be due to occurrence of some dis-persal mechanisms, which were not covered by this research. In southeast Spain �� was studied regarding recruitment on Mediterranean mountain (Garcia et al., 2001). He indicated that birds (Thrushes) dispersed a large proportion of the Juniper seeds every year. Almost exclusively mi-grant Thrushes, mainly �� , which over winters in these areas from October to April, and �� , which visits Juniper shrub lands from August to October (Jordano 1993; Garcia


56

1998b; Garcia et al. 1999b, in Garcia, 2001). He also found that the distribution of �� related to the density of rodents �� above the treeline (D. Garcia pers. Obs.), but also to the different methodology adopted in different studies Alcantara et al. (2000b, in Garcia et al. 2001). Seed predation was concentrated under stones and Junipers, the probability of escaping from rodents being higher in stones, wet meadows and open ground (Garcia, 2001). The germination and seedling emergence of �� on the mountain in SE Spain was stud-ied (Garcia, 2001). He found that the seedling density was highest in wet meadows and lowest in open ground. Emergence in wet meadows was five-fold higher than in the remaining microhabitats (under stone, Juniper, stone and open ground). Generally, Juniper seed germination was found to be favored in wet substrates, which allow a complete imbibition of the seed coat (Young et al. 1988; Chambers et al. 1999, in Garcia et al. 2001).

5.2. Plant species richness vs environmental factors

In this case study we concentrated on topographic factors such as altitude, slope steepness and slope direction as well as soil factors considering that the environmental factors especially rainfall may also be involved in determining species (Nunez-Olivera et al. 1995, in Pausas & Austin, 2001). So this precipitation was also affected by topography. Harrison et al. (1992, in Pausas & Austin, 2001) avoided the analysis of species-environment relation-ships as too problematic in their study of beta diversity gradients in Britain, preferring to concentrate on distance and dispersal. Richardson et al (1995, in Pausas & Austin, 2001) tested several hypotheses based on biotic interactions to explain Banksia species richness in the south of Western Australia. They did not test for any relationship between species richness and environmental parameters; how-ever they concluded that topographic and soil factors may be an explanation for the pattern of coexis-tence, but also they commented that different regions (i.e. different environments) do show different pattern of coexistence. For studying the topography factors, the research found that the lower plant species richness at high elevation this might be due to the favorable hydrological conditions in the low land (wadis) compared to the high elevation sites. At slightly steepness slope the plant species richness the infiltration be-come more consequently the soil become moist, which is favor a suitable environment for the plant species, but the plant species was not affected by slope direction. A negative relationship between altitude and woody species richness has been reported in different ecosystems (e.g., in coniferous forests by Pausas 1994 and Rey Benayas 1995; in Alaska, Tennessee and Costa Rica reviewed by Stevens 1992, in Pausas & Austin, 2001). Furthermore the soil pH was tested; however, the research indicated that the maximum plant species richness obtained at pH value of 6.25-6.5. If the soil reaction is held within a soil pH range of 6-7, the toxicity of the Aluminum, Iron and Manganese and the deficiency of Iron and Manganese may be avoided (Brady, 1984). Grime (1973) showed that the maximum number of species in unmanaged grassland occurs at a soil pH of 6.1-6.5 (Gausas & Austin, 2001).

5.3. Forest islands sizes vs plant species richness

The core of this research concerned the application of the biogeographically island theory for whether this biogeographically spatial pattern is useful to be applied on the forest islands. During the study the research found that the plant species richness was influenced by the size of the forest island according to the species-area relationship of Arrhenius (1921); Person (1960); and MacArthur and Wilson


57

(1967), in Levenson (1981). In fact the species-area curve was applying when the data collected from a limited range of geographic area. Meanwhile the variation of climate was ignored. The variation of the climate was much more related to the topography as mentioned previously. Several authors have used altitude (an indirect environmental variable) as a surrogate variable for temperature; however, this parameter is complex and may covary with other climatic factors (e.g., rainfall cloud cover wind) and with the degree of isolation on the top of mountains (Pausas & Austin, 2001). The hypothesis of the relationship between the number of plant species and the size of the forest is-land was examined in this study. However the research obtained that the plant species richness created mostly by the increase of microhabitat variety (Wiens 1962, in Levenson, 1981). This suggests that the forest island with more mapping units has more microclimates. The variation of the mapping units totally induced from the variation of the topography of the forest island. Despite this suggestion, some forest islands, especially the largest one reported here show that an increase in area (more mapping units) that added more of the same habitats, but no additional new habitats and this may be due to the orientation of this forest island. Johnson et al. (1968) found the island area extent to be the best single predictor of plant species diver-sity, but conducted that environmental richness was also significant (Levenson, 1981). In this area qualitative different environments microhabitats resulted from increasing topography variation may be lead to variation of the edaphic factors (Heatwole and Levins 1973, in Levenson, 1981). Levenson (1980) found that in southeastern Wisconsin, the woody species composition of the forest woodlots was restricted by similar factors. An increase in area very often added more of the same habitat. It was not until variation in soil, relief, or distances were present that changes in species vari-ety were realized. Here the forest islands within the range of 14.13 ha and the larger forest island of 888.84 ha. The in-crease of plant species diversity with an increase of size of the forest island was observed up to 200 ha and then the number of plant species decreases as the forest island increases. In fact as we observed many factors may cause the decline of the plant species richness. This because increasing in area cov-ered by rock does not add more to the flora richness and this was clearly observed in the larger forest islands (Niering’s 1963, in Levenson, 1981). Again, the disturbance whether naturally or man-induced is a major variable controlling plant species richness in this area. The naturally disturbance was caused by the drought that was taken place in the last two decades, but the man-induced disturbance is due to the animals grazing mainly sheep, goats and donkeys. Tramer and Suhrweir (1975, in Levenson, 1981) conducted that human interference was a major vari-able in affecting species richness of elm-ash-maple woodlots. Slack et al, (1975, in Levenson, 1981) also indicated that increased disturbance contributed to an increased number of exotic species as well as to an increased overall diversity of vascular plants and bryophytes on 56 islands in Lake George. Previously, the practical reasons were discussed due to the naturally and man-induced disturbances. Theoretically, the number of plant species, which could exist in an edge environment, increases to become asymptotic to the number of plant species of the region, as indicated in southeastern Wiscon-sin. The function increases rapidly since the major constraint is adequate physical space to accommo-date an individual of a new species (Levenson, 1981). After that with formation of interior condition e.g. in this case the interior environment was appeared at 200 ha as in figure (4.24.A). The relation-ship of plant species in the interior forest island then follows the general form of a depletion curve with increasing island size. The curve approaches the number of shade-tolerant plant species of the region.


58

In this study the decline of the curve may be to the competition of the plant species and/or the orienta-tion of the forest islands, which are not covered by the study. Generally there are different reports of forest island size. Levenson (1981) reported 2.3 ha, Vestal & Heerman (1945) mentioned 1.6 ha and Locuks (1970) found an area of 3.8 ha (Levenson, 1981). However, this research reported that plant species richness generally increase with island size to approximately 200 ha suggesting that there are many gaps through these forest islands that might create xeric habitats. Cornell & Lawton (1992) have proposed that species richness is determined by local biotic interactions such as competition and pre-dation, and regional or historical processes like dispersal and speciation (Pausas & Austin, 2001). Several authors have explored the backbone of island biogeography theory that species richness of habitat forest islands is a direct function of their sizes, which was not found by Levenson for trees and shrubs in the interior of 43 forest islands in southeastern Wisconsin. Also Hoehne indicated the same conclusion concerning the ground layer (Burgess and Sharpe, 1981). Much island biogeography re-search has assumed equilibrium conditions resulting from a balance between invasions and extinc-tions, but also the equilibrium should has first to take place within the forest island between the bal-ance of edge field and interior field plant species (Burgess and Sharpe, 1981).

5.4. Isolation of forest islands vs plant species richness

The isolation of forest islands through the study area was tested and the research indicated that the relationship between the plant species richness and the isolation of the forest islands was negatively significant i.e. the number of plant species decreases as the degree of isolation increases. The degrees of the isolation of these forest islands are within the range of 0.36-11.863 km from the source forest island reported a high coefficient than the forest island size. This finding of the research was served as the prediction of the hypothesis of the biogeography island theory. In this case we considered that the size of the forest island was much more trustable than the forest island isolation. The effect of isolation on the plant species richness depends on seed dispersal. The latter depends mainly on many factors such as seeds quantity, time of seeds release the dispersal mechanisms (wind and animals) as well as the dispersal distance from seed source (Johnson et al., 1981). Only the dispersal distance has been considered in this study in order to test the hypothesis of island biogeography theory. The other components of isolation are ignored in this study such as time since isolation, the distance between the other forest islands, and the degree of connectivity between them are all important determinants of the biotic response to fragmentation of forest islands (Saunders et al., 1991). Seed dispersal by wind thus involves a complex of processes whose outcome is difficult to predict. The animal dispersal model traces the movement of individual animals assumed to be carrying seeds from the sources. For the dispersal by wind, wind direction and velocity are very important while for the dispersal by animal the spatial direction selected randomly but biased by the animals behavioral interaction with the environment (Johnson et al., 1981).


59

6. Conclusions & Recommendations

6.1. Conclusions

• Species richness of grass, shrubs, and tree forms generally increased with island size to ap-proximately 200 ha. Forest islands smaller than 200 ha are considered as xeric habitat of communities composed of a mix of intolerant and residual tolerant shade species.

• Floristic richness has also been shown a negatively correlated with the degree of isolation of the forest islands, suggesting that the dispersal of some plant species is inadequate to bridge interfiled distances.

• The forest islands of Juniper are extremely open woodland and distributed within a high alti-tude and a rugged topography area, however, the visual interpretation of Landsat thematic mapper image was not much useful in mapping the Juniper forest islands.

• The research concluded that the Juniper density and canopy cover as well as plant species richness are slightly (negatively) related to the soil pH value. However both Juniper trees and plant species prefer the acidic soil (6.25-6.5). In fact pH is an environmental parameter related to nutrient and toxic element availability.

• Generally the relationship between the environmental factors (topography and soil) and the plant species richness including the Juniper are very weak, because the data set is limited and restricted only on the Juniper forest islands except a few samples outside.

• The case study indicated that the plant species richness was related to the different mapping units, which lead to different microhabitats. Also the plant species richness was affected by the orientation of the forest islands, especially in the largest forest island, despite that it has more mapping units and at the same times a low diversity of plant species.

• The research found that the concept of the Biogeographically Island Theory that plant species richness of habitat “terrestrial” forest islands is a direct function of their sizes. Because the ef-fect of isolation on the plant species richness depends on seed dispersal. Seed dispersal by wind thus involves a complex of processes whose outcome is difficult to predict.

• The overall indices of fragmentation were not tested here assuming that the forest islands were opened and interspaced as well as were not studied before. In contrasts the calculation of fragmentation indices is based on quantitative analyses that measure the spatial pattern of the landscape. The value of those indices describes where and how the fragmentation changes and it depends on the historical land use.

• The occurrence of �� totally appeared at altitude of 1900-2700m, with no upper tree line. The distribution of this species allover Iran within altitude ranges of 1500-3300m. It is generally a higher altitude species; it occurs from 500-3800m. The wet sub-strates (Wadis) showed a relatively high density and a good growth. Death of Juniper trees was not observed.


60

6.2. Recommendations

• Long-term studies of the individual plants and animals species in the forest islands of the dif-ferent sizes would be extremely useful in supporting the result of this study coupled with the history of the land use.

• A higher spatial resolution satellite image and/or recent aerial photographs are needed for studying the forest islands to obtain reliable results, due to the rugged topography.

• Study other factors (e.g. social and economical) that may have strong influence on forest fragmentation.

• In light of this result, I recommend in future the study of the plant communities, succession, interaction, and the seed dispersal mechanisms should be considered.

• Taxonomical studies should be undertaken for �� because some authors consider this species as �� .

• The detail study of the distribution of �� should be done regard-ing the geological substrate, pre-seed dispersal and post-dispersal seed predation.

• For application the island biogeography concept the large number of forest islands with dif-ferent sizes and isolations should be taken into account.

• The Island Biogeographically aspect should be considered in regional planning especially in the developing countries.


61

References Andrea Perlis, 2001. Rome, Italy. Global forest resources assessment FAO (2000) main report. FAO Forestry Paper, 140. Leicet Diaz Varona, 2000. Applying remote sensing and GIS for monitoring of fragmentation in an area of Atlantic dry forest of southeastern Brazil. MSc. Thesis Wageningen Agricultural University. Liliana Caballero Landin, 2001. Monitoring Forest Fragmentation using GIS and RS Techniques, Study Case of Makwanpur District, Hetauda, Nepal. MSc. thesis ITC. Grez Audrey A., Ramiro O., 2000. Bustamante, Jaavier A. Simonetti & Lenore Fahrig. Landscape Ecology, Deforestation, and Forest Fragmentation: The Case of Ruil Forest in Chile. Ross K.A., Fox B.J. and Fox M.D., 2002. Changes to plant species richness in forest fragments: Frag-ment age, disturbance and fire history may be as important as area. Journal of Biogeography 29 (5-6): 749-765. Bruna E.M., Nardy O., Strauss S.Y. and Harrison S., 2002. Experimental assessment of �� growth in a fragmented Amizonian landscape. Journal of Ecology 90(4): 639-649. William F. Laurance and Eric Yensen, 1991.Predicting the impacts of edge effects in fragmented habitats. Biological Conservation 55: 77-92. Hill J.L. and Curran P.J., 2001. Species composition in fragmented forest: conservation implications of changing forest area. Applied Geography 21: 157-174. Anderson W.B. and Wait A.A., 2001. Subsidized island biogeography hypothesis: another new twist on an old theory, Ecology Letter Journal Article. Rosallo S.C., Leitao Filha H. de F. and Begossi A., 1999.Ethnobotany of Caicaras of the atlantics for-est coast (Brazil), Economic Botany, 53 (4): 387-395. Shigeo Iida and Tohru Nakashizuka, 1995. Forest fragmentation and its effect on species diversity in sub-urban coppice forest in Japan. Forest Ecology and Management 73: 197-210. Robert L. Burgess and David M. Sharpe. 1981. Summary and conclusions. Forest Island Dynamics in Man-Dominated Landscapes. Ed. by Robert L. Burgess & David M. Sharpe. Ecological Studies (41): 267-272 Springer-Verlag. Rudis V.A. and Ek A.R., 1981.Optimization of forest island special patterns: Methodology of analysis of landscape pattern. Forest Island Dynamics in Man-Dominated Landscapes. Ed. by Robert L. Bur-gess & David M. Sharpe. Ecological Studies (41): 241-256. Springer-Verlag.


62

Levenson James B., 1981.Wood as biogeographic islands in southern Wisconsin. Forest Island Dy-namics in Man-Dominated Landscapes. Ed. by Robert L. Burgess & David M. Sharpe. Ecological Studies (41): 13-39. Springer-Verlag. Turner Monica G., 1998. Landscape ecology: Living in mosaic. Ecology. Ed. Stanley I. Dodson. Ox-ford. Farjon A., 1993.The Taxonomy of multiseed Junipers (�� sect. � �� ) in southwest Asia and east Africa (Taxonomic notes on Cupressaceae 1). Edinburgh Journal of Botany vol. 49 No. (3): 251-283. Ed. Dr.Bj. Coppins. Collection from Research (Herbarium and living collection or archives). Parsa A. and Maleki Z., 1978. Flora of Iran, vol. (1): 260-276. Van Gils Hein and Baig M. Shabbir, 1992. Environmental Profile Balochistan, Pakistan. Golrang B. M., 1995. Evaluation of changing canopy cover in Sad amir kabir catchments area (1973-1993). Study report: watershed management organization. Ministry of Agriculture, Tehran-Iran. Paul J. Gibson and Clare H. Power, 2000. Introductory Remote Sensing digital image processing and applications. Rontledge 11 New Fetler Lane, London. Lillesand, Thomas M. and Ralph W. Kiefer, 1994. Remote Sensing and image interpretation, 3rd edi-tion. Ghazani M.F. and Movahhed F.B., 1997. Rangeland Biomass Modelling and Mapping in Zanjan Mountains, Iran. A GIS and RS, Case Study. MSc. Thesis Wageningen Agricultural University. Charles D. Bonham; 1988. Measurement for Terrestrial Vegetation. A Wiley-Interscience Publication. John Wiley & Sons, New York. Clude Faurie, Christian Ferra, Paul Medori and Jean Devaux, 2001. Ecology; Science and Practice: A. A. Bulkema, Abingdon, Exton. Assadi M., Babakhanlou P., Jamzad Z. and Maassoumi A.A. Ramak, 1988. The Iranian Journal of Botany: Research Organization of Agriculture and Natural Resources, Research Institute of Forests and Rangelands (Iran). Gardner A.S. and Fisher M., 1996.The distribution and the status of the montane Juniper woodlands of Oman. Journal of Biogeography vol. 23 (6): 791-803. Fisher M. and Gardner A. S., 1995. The status and ecology of �� wood-land in the northern mountains of Oman. Vegetatio 119: 33-51. Garcia D., 2001. Effects of seed dispersal on �� recruitment on a Mediterranean moun-tain. Journal of Vegetation of Science 12: 839-848.


63

Landon J.R., 1984. Booker tropical soil manual, a handbook for soil survey and agricultural land evaluation in the tropics and subtropics. Pitman press limited. Pausas, Juli G. and Austin, Mike P., 2001. Patterns of plant species richness in relation to different environments: An appraisal. Journal of Vegetation of Science 12: 153-166. Nyle C. Brady, 1984. The nature and properties of soil. Macmillan, New York USA. Johnson W. Carter, David M. Sharpe, Donald L. DeAngelis, David E. Fields and Richard J. Olson, 1981. Modelling Seed Dispersal and Forest Island Dynamics. Forest Island Dynamics in Man Domi-nated Landscapes. Ed. by Robert L. Burgess & David M. Sharpe. Ecological Studies (41): 215-239. Springer-Verlag. Saunders Denis A., Richard J. Hobbs and Chris R. Margules, 1991. Biological Consequences of Ecosystem Fragmentation: A Review. Conservation Biology Vol. 5 (1): 18-31.


64

Plate 1: Juniper trees with �� . On sheltered wadis habitat Jul. 31. 2002

Plate 2: �� in Sirachal forest island Jul. 31.2002


65

Plate 3: Three life forms: Juniper trees, shrubs and grasses Aug. 2.2002

Plate 4: ��. On steep slope Aug. 2.2002


66

Plate 5�� !�� shrub with many grasses Aug.2.2002

Plate 6: Field worker with soil tools Aug.2.2002


67

Data form of fieldwork

Date: Day. Month. Year

Altitude (m):

Photo No.: (run. Type. Date) Forest patch No.: Area/Country : Distance from source patch: Sample No. : Sample size :

Observer(s):

Terrain data Relief type: Site/location: Very flat Geology type: Almost flat <2% Mapping unit: Undulated 2-7% Landform : Rolling 8-13% Slope type : Hilly 14-20% Slope : Steeply dissected 21-55% Aspect (exposure): Mountainous >55% Soil data Hor.Sy. Depth Texture Color PH Sur.stonyn. % Preliminary soil classification: Land cover/use data (semi natural or planted) Strata No./sample Height Cover % Dominant species No. of Jun. trees Trees Shrubs Herbs Total real cover %: Land use type Appendix 1-1: Field data form


68

Repeated No. List of species Total No. of species

General observations:

- Semi natural: - Burning: - Fuel wood collection: - Human activities: - Range condition: - Grazing traces: - Fencing: - Animal types:

Appendix 1-2: Plant species and general information


69

Site Latitude Longitude Altitude (m) Slope% Aspect Aspect (De-

gree) No.of plant species No.of juniper trees Total canopy cover% sir 1 3986447 514805 1897 3.8 n 330 17 2 64 sir 2 3986281 514924 1927 28 n 345 11 2 40 sir 3 3986122 515074 1976 26.5 n 340 11 2 52.3 sir 4 3985975 515209 2039 40.5 n 339 10 2 50.5 sir 5 3985890 515542 2120 33.7 w 273 8 5 75.8 sir 6 3986026 515393 2048 46.1 w 295 7 2 48.2 sir 7 3986198 515290 1987 26.9 w 314 6 3 65.7 sir 8 3986358 515171 1924 35.8 w 306 6 6 77.5 sir 9 3986689 515200 1895 3.9 s 180 7 7 77

sir 10 3986526 515316 1899 2.4 w 306 7 4 85 sir 11 3986361 515429 1971 46.4 n 0 9 2 45.3 sir 12 3986202 515555 2049 39.6 n 348 9 2 42.1 sir 13 3986039 515681 2128 41.1 w 308 9 4 76.8 sir 14 3985893 515820 2200 50.2 w 300 7 2 39.7 sir 15 3985754 515963 2288 40 w 276 8 2 36.4 sir 16 3985611 516102 2364 49.1 n 316 6 2 34.2 sir 17 3986896 515515 1914 40.3 s 138 5 3 55.5 sir 18 3986720 515618 1896 0 s 170 5 4 55.3 sir 19 3986541 515724 1917 58.4 n 324 5 4 66.3 sir 20 3986372 515826 2059 90.8 n 323 9 5 68.7 sir 21 3986203 515936 2216 49.8 w 309 7 2 29.3 sir 22 3986037 516045 2300 65.1 w 309 11 2 40.2 sir 23 3986180 514742 1954 42.5 n 357 2 0 5 sir 24 3986004 514905 2000 37 n 0 5 0 16 sir 25 3985737 515431 2119 37.5 n 32 2 0 7

Appendix 2-1: Field data


70

Site Juniper canopy cover% Soil depth (cm) Soil texture Soil moisture % Soil EC(ms) Soil pH Soil CEC(me/l) sir 1 29 45 si-c-l 2.56 0.5 7.54 29.83 sir 2 18 47 si-c 3.25 0.29 7.8 27.13 sir 3 16.2 50 si-c-l 6.61 0.26 6.67 29.22 sir 4 14.2 50 c 6.98 0.5 7.44 34.78 sir 5 47.4 46 c 8.52 0.77 7.52 40.52 sir 6 26.4 45 si-c-l 7.73 0.33 6.33 35.83 sir 7 37.7 45 sa-c-l 2.88 0.31 7.33 8 sir 8 54.5 46 si-c-l 6.32 0.34 6.37 36.17 sir 9 60.5 48 c-l 4.88 0.38 6.37 33.04

sir 10 62 55 si-c 1.44 0.3 6.25 31.91 sir 11 29.5 40 si-c 3.75 0.34 7.74 30.87 sir 12 25.1 40 si-c 5.52 0.27 6.61 31.3 sir 13 32.8 45 si-c 5.64 0.27 6.37 26.09 sir 14 18.2 45 si-c 5.64 0.27 6.37 26.09 sir 15 15.7 45 c 6.24 0.31 7.53 36.78 sir 16 10 55 c 6.24 0.31 7.53 36.78 sir 17 34 40 c 9.48 0.42 6.54 41.39 sir 18 33.8 40 c 9.48 0.42 6.54 41.39 sir 19 50.5 50 c 5.99 0.29 7.78 26.35 sir 20 27.5 40 si-c 10.52 0.47 7.24 38.26 sir 21 12.8 50 c 11.34 0.38 6.77 38.78 sir 22 19.2 25 c-l 1.32 0.34 7.74 30.87 sir 23 0 30 sa-c-l 1.74 0.49 7.7 17.39 sir 24 0 40 c-l 1.99 0.32 6.86 24.87 sir 25 0 55 sa-c-l 2.49 0.26 7.58 15.65



71

Site Band1 Band2 Band3 Band4 Band5 Band6.1 Band6.2 Band7 Band8 NDVI sir 1 88 84 99 63 123 166 213 99 64 100 sir 2 103 100 114 63 135 162 205 120 70 91 sir 3 89 82 99 65 131 162 207 106 66 102 sir 4 81 72 79 54 96 151 186 73 53 104 sir 5 85 79 95 64 120 162 206 91 65 103 sir 6 80 74 81 64 102 156 195 72 58 113 sir 7 89 83 96 56 117 155 194 104 59 95 sir 8 88 83 98 63 125 168 216 95 64 100 sir 9 92 90 110 64 131 163 208 107 67 94

sir 10 91 81 95 59 113 163 207 91 60 98 sir 11 87 76 86 54 98 156 194 77 55 99 sir 12 89 86 112 74 135 160 202 100 69 102 sir 13 82 76 89 59 108 156 194 83 60 102 sir 14 79 75 89 58 105 157 197 78 57 101 sir 15 82 75 85 61 103 151 186 76 57 107 sir 16 79 72 76 69 102 151 186 71 61 122 sir 17 89 85 96 59 110 157 197 89 61 98 sir 18 89 81 93 61 114 154 190 92 64 102 sir 19 85 77 88 58 107 158 199 89 57 102 sir 20 84 77 90 57 108 151 185 90 56 99 sir 21 84 79 94 63 114 154 192 88 63 103 sir 22 85 76 88 62 109 152 187 83 61 106 sir 23 105 101 111 61 134 159 199 118 68 91 sir 24 107 106 122 65 143 160 202 131 70 89 sir 25 111 106 118 59 145 159 200 141 68 86



72


gree) No.of plant species No.of juniper trees Total canopy cover% ham 1 3987508 525124 2361 31.5 w 252 6 2 29.8 ham 2 3987367 525267 2418 39 w 227 6 4 35.3 ham 3 3987213 525397 2413 41.8 s 225 5 6 59.3 ham 4 3987322 525047 2292 27.8 w 272 6 2 42.5 ham 5 3987153 525156 2300 43.3 n 6 6 1 30 ham 6 3987008 525299 2426 87 w 315 2 0 5

serch 1 3990151 529654 2600 83.3 w 270 7 1 24.7 serch 2 3990086 529452 2477 23.7 s 183 7 2 70.6 serch 3 3990281 529378 2581 54.9 s 217 4 1 19.5 serch 4 3990455 529307 2700 73.6 s 175 2 2 10 serch 5 3990519 529499 2775 89.2 s 180 3 2 36.3 serch 6 3990323 529548 2644 66 s 193 3 1 19.6 serch 7 3989580 527734 2300 69.4 s 145 7 1 40 serch 8 3989448 527583 2250 54.8 e 131 3 2 8.3 serch 9 3989323 527426 2270 16.7 e 96 6 3 37

serch 10 3989209 527263 2292 13.2 e 83 7 5 42.7 serch 11 3989105 527095 2353 19.7 e 51 6 6 40 serch 12 3989318 529222 2474 81.5 n 320 5 1 28.8 serch 13 3989166 529381 2520 86 s 189 6 2 27.2 serch 14 3989850 528895 2400 82.2 w 293 2 0 10 serch 15 3989722 527621 2281 65.4 n 39 0 0 0 serch 16 3991756 527915 2185 5.5 s 196 0 0 0



73

Site Juniper canopy cover% Soil depth (cm) Soil texture Soil moisture % Soil EC(ms) Soil pH Soil CEC(me/l) ham 1 12.3 40 c-l 0.97 0.39 7.76 15.91 ham 2 12.6 35 c-l 0.97 0.39 7.76 15.91 ham 3 34.3 30 c-l 0.97 0.39 7.76 15.91 ham 4 9.8 25 c 1.13 0.51 7.59 40 ham 5 3.14 20 c 1.13 0.51 7.59 40 ham 6 0 40 c 1.22 0.31 7.66 31.83

serch 1 6.2 30 c-l 0.78 0.34 7.27 17.65 serch 2 10.2 20 c-l 0.78 0.34 7.27 17.65 serch 3 7.1 15 l 0.9 0.37 7.46 9.22 serch 4 6.3 15 l 0.9 0.37 7.46 9.22 serch 5 6.3 30 c-l 0.78 0.34 7.27 17.65 serch 6 12.6 10 c-l 0.93 0.38 7.71 12.61 serch 7 9.6 46 c 6.03 0.28 7.56 30.09 serch 8 6.3 10 sa-c-l 2.44 0.27 7.98 9.22 serch 9 14.7 40 c-l 3.91 0.41 6.99 15.91

serch 10 15.7 40 c-l 3.91 0.41 6.99 15.91 serch 11 18.8 35 c-l 3.91 0.41 6.99 15.91 serch 12 5 10 l 0.64 0.38 7.88 12.35 serch 13 14.1 10 c-l 3.91 0.41 6.99 15.91 serch 14 0 35 sa-c-l 0.37 0.95 7.6 8 serch 15 0 20 sa-c-l 2.88 0.31 7.33 8 serch 16 0 30 c-l 3.43 0.39 7.77 12.61



74

Site Band1 Band2 Band3 Band4 Band5 Band6.1 Band6.2 Band7 Band8 NDVI ham 1 99 99 123 73 136 165 211 114 74 96 ham 2 93 94 115 65 134 153 189 111 70 93 ham 3 84 82 100 60 116 149 181 91 68 96 ham 4 87 84 100 64 114 152 187 88 68 100 ham 5 92 89 106 62 123 151 185 102 66 95 ham 6 87 82 98 58 107 147 177 87 65 95

serch 1 76 68 83 53 75 142 170 66 56 100 serch 2 92 90 118 67 105 157 196 96 71 93 serch 3 89 84 108 68 99 155 193 82 68 99 serch 4 85 80 101 64 94 146 175 81 63 100 serch 5 102 110 147 89 183 160 200 164 89 97 serch 6 93 92 121 76 115 162 206 101 75 99 serch 7 82 75 91 61 96 158 199 77 63 103 serch 8 81 74 91 66 110 153 191 86 61 108 serch 9 81 76 93 63 107 153 189 86 58 104

serch 10 88 85 107 70 122 149 183 101 66 101 serch 11 79 72 88 54 86 149 182 73 65 98 serch 12 81 73 85 53 83 150 183 68 64 99 serch 13 86 86 103 69 112 165 210 97 71 103 serch 14 79 74 89 70 88 157 196 69 56 113 serch 15 86 78 98 53 73 174 227 64 56 90 serch 16 81 74 89 60 93 144 172 78 52 103



75

Site Latitude Longitude Altitude (m) Slope% Aspect Aspect (Deg) No.of plant species No.of juniper trees Total canopycover% lan 1 3982139 524868 2168 15 e 93 6 5 77.3 lan 2 3982174 524671 2244 66 e 100 6 1 45.1 lan 3 3982218 524480 2284 54.3 n 4 7 1 26.3 lan 4 3982257 524292 2300 53.1 e 55 6 1 17 lan 5 3982286 524077 2323 76.1 n 5 7 2 56.3 lan 6 3982181 523865 2429 24.7 n 14 7 2 45.5 lan 7 3982120 524055 2378 26.7 e 91 6 2 22.8 lan 8 3982059 524242 2319 31.7 e 93 6 2 32 lan 9 3981981 524423 2272 33.1 e 91 5 2 37.3

lan 10 3981914 524611 2211 32.9 e 130 6 2 32.3 lan 11 3981850 524811 2187 12.7 w 292 5 2 51 lan 12 3982049 523799 2463 26.9 e 47 5 1 39.6 lan 13 3981932 523943 2400 39.6 e 56 5 1 23.8 lan 14 3981829 524097 2373 33.7 e 99 8 2 37.2 lan 15 3981716 524259 2340 28.4 e 64 8 4 46.8 lan 16 3981617 524436 2300 45.2 e 62 7 6 63.7 lan 17 3981517 524603 2231 48 s 138 9 2 74.2 lan 18 3981396 524757 2194 11.3 w 298 4 1 7.3 lan 19 3981270 524625 2193 2.4 n 324 5 2 38.1 lan 20 3981392 524465 2227 65.5 e 127 5 3 37.8 lan 21 3981516 524307 2320 32.2 e 103 7 2 63.6 lan 22 3981630 524144 2381 32.4 e 81 5 4 36 lan 23 3981746 523969 2418 27.6 e 69 7 2 38.3 lan 24 3981840 525047 2234 52.5 s 205 6 2 52.5 lan 25 3981782 525240 2200 95.6 s 212 8 1 18.6 lan 26 3981720 524670 2196 8.9 e 90 2 0 7



76

Site Juniper canopy cover% Soil depth (cm) Soil texture Soil moisture % Soil EC(ms) Soil pH Soil CEC(me/l) lan 1 62.8 65 c 8.29 0.34 7.72 28.09 lan 2 20.4 50 c 6.2 0.53 6.88 21.91 lan 3 5 20 si-c-l 4.22 0.35 6.99 26.09 lan 4 5 20 si-c-l 4.22 0.35 6.99 26.09 lan 5 39.3 15 c 0.97 0.6 7.72 31.3 lan 6 14.1 50 c 2.09 0.3 6.5 31.3 lan 7 6.3 15 c 2.09 0.3 6.5 31.3 lan 8 22.6 20 c 0.97 0.6 7.72 31.3 lan 9 31.8 20 c-l 1.27 0.47 6.57 25.39

lan 10 23.7 20 c-l 3.43 0.31 6.3 22.44 lan 11 25 50 c 0.97 0.6 7.72 31.3 lan 12 9.6 50 c 0.97 0.6 7.72 31.3 lan 13 7 10 c 0.97 0.6 7.72 31.3 lan 14 23.7 40 c-l 3.2 0.37 7.23 31.3 lan 15 11.8 38 c 5.8 0.32 6.38 26.35 lan 16 36.7 25 c 5.8 0.32 6.38 26.35 lan 17 38.9 60 c 8.43 0.43 6.92 27.13 lan 18 2.3 70 c-l 4.43 0.42 7.68 17.91 lan 19 14.1 20 c-l 3.86 0.31 7.66 21.13 lan 20 21.2 20 c-l 3.86 0.31 7.66 21.13 lan 21 14.1 30 sa-c-l 0.49 0.77 6.39 17.39 lan 22 19.6 15 sa-c-l 0.49 0.77 6.39 17.39 lan 23 14.1 10 c 5.25 0.32 7.76 29.83 lan 24 14.1 15 c 3.8 0.27 7.61 34.78 lan 25 9.6 10 c 3.8 0.27 7.61 34.78 lan 26 0 40 si-c-l 0.65 0.37 7.52 13.74



77

Site Band1 Band2 Band3 Band4 Band5 Band6.1 Band6.2 Band7 Band8 NDVI lan 1 100 95 112 77 144 168 217 120 75 104 lan 2 99 96 114 74 142 166 213 112 75 101 lan 3 100 98 116 78 158 172 224 130 75 103 lan 4 100 99 122 75 160 170 221 135 78 98 lan 5 101 100 123 77 151 174 227 120 77 99 lan 6 102 100 128 81 171 173 226 142 74 99 lan 7 94 95 124 76 159 175 231 135 73 98 lan 8 93 92 120 82 155 173 226 115 79 104 lan 9 88 85 99 68 125 171 221 99 69 104

lan 10 89 88 110 75 144 174 226 109 74 104 lan 11 95 95 112 76 143 176 233 121 73 104 lan 12 97 95 124 78 163 175 228 140 77 99 lan 13 100 99 124 76 166 174 229 133 76 98 lan 14 108 107 137 76 167 173 227 145 83 92 lan 15 99 99 129 74 165 176 231 144 77 94 lan 16 96 92 116 71 142 172 224 115 70 97 lan 17 108 112 145 83 184 180 238 157 90 93 lan 18 98 103 127 82 169 175 229 149 76 101 lan 19 92 90 110 73 149 173 226 120 73 102 lan 20 95 98 109 66 121 172 223 106 70 97 lan 21 102 102 131 76 173 172 224 148 77 94 lan 22 105 104 131 72 152 177 234 136 77 91 lan 23 97 99 125 72 154 171 222 134 73 94 lan 24 97 95 122 71 160 172 224 139 72 94 lan 25 93 92 113 72 153 176 230 130 73 100 lan 26 101 95 118 63 123 186 249 109 65 89



78


gree) No.of plant species No.of juniper trees Total canopy cover% ray 1 3986954 516152 1921 45.4 n 316 8 4 39.1 ray 2 3986794 516273 2014 43.5 n 324 7 4 37.7 ray 3 3986629 516385 2087 37.5 n 0 8 3 39.2 ray 4 3986464 516499 2200 61.9 n 0 7 1 42.6 ray 5 3986302 516611 2283 53 n 0 6 3 46.9 ray 6 3986589 516859 2244 49.5 w 249 8 2 48.8 ray 7 3986412 516953 2272 28.1 n 1 4 2 30 ray 8 3986259 517086 2373 55.2 n 353 6 2 35.9 ray 9 3986757 516750 2224 49.8 w 254 8 2 59

kalv 1 3990120 513124 1992 1.7 n 45 4 2 22.3 kalv 2 3989806 513686 2046 64.4 s 210 5 2 31.1 kalv 3 3989641 513149 2100 89.9 e 51 4 2 26.3 kalv 4 3989434 513306 2076 36.5 n 31 5 3 30.4 kalv 5 3989095 513686 2066 60 e 78 4 1 23.6 kalv 6 3988848 513900 2033 66.6 n 45 4 2 35.3



79

Site Juniper canopy cover% Soil depth (cm) Soil texture Soil moisture % Soil EC(ms) Soil pH Soil CEC(me/l) ray 1 16.7 46 c 2.24 0.63 7.34 40.26 ray 2 23 46 c 2.24 0.63 7.34 40.26 ray 3 21.2 50 c 6.97 0.32 7.33 36.78 ray 4 7.1 50 c-l 2.7 0.32 6.74 24.35 ray 5 10.6 35 c 2.78 0.34 7.67 27.13 ray 6 7.5 60 c 2.78 0.34 7.67 27.13 ray 7 10 40 c 2.78 0.34 7.67 27.13 ray 8 12.8 35 c-l 2.44 0.38 7.26 22.87 ray 9 25 43 c 6.13 0.31 7.2 42.96

kalv 1 6.3 20 l 0.55 0.61 7.2 14.17 kalv 2 9.8 25 c 6.13 0.31 7.2 42.96 kalv 3 9.8 20 c 2.78 0.34 7.67 27.13 kalv 4 9.4 30 c 2.78 0.34 7.67 27.13 kalv 5 7.1 35 c 6.13 0.31 7.2 42.96 kalv 6 14.1 25 c-l 2.44 0.38 7.26 22.87



80

Site Band1 Band2 Band3 Band4 Band5 Band6.1 Band6.2 Band7 Band8 NDVI ray 1 82 72 79 53 92 145 175 72 50 103 ray 2 79 70 80 60 91 146 177 64 56 110 ray 3 91 84 95 60 113 145 175 95 59 99 ray 4 77 67 77 55 92 146 178 68 53 107 ray 5 78 68 78 54 94 146 177 70 52 105 ray 6 79 69 80 61 97 150 183 68 59 111 ray 7 84 79 94 64 118 154 190 91 65 104 ray 8 84 77 90 63 117 159 199 91 60 106 ray 9 93 84 95 53 104 143 172 95 64 92

kalv 1 90 85 103 73 126 157 195 99 73 106 kalv 2 103 103 123 75 149 166 213 131 82 97 kalv 3 91 90 116 77 138 165 211 109 74 102 kalv 4 101 103 124 84 160 163 207 135 83 104 kalv 5 94 91 106 76 139 162 205 109 72 107 kalv 6 89 85 100 73 129 169 217 100 70 108



81

Appendix 3-1:List of inventoried plant species No. Name of species Family 1 �� Cupressaceae

2 �� Gramineae

3 "�� Gramineae

4 ��#��# �� Compositae

5 �� # � Compositae

6 �� Leguminosae

7 �#� �� Cruciferae

8 $ �� Gramineae

9 "�%%�� Caryophyllaceae

10 �#��&��#� �� Labiatae

11 '��#�� Euphorbiaceae

12 �� Guttiferae

13 �� Compositae


15 �� Rosaceae

16 �#�� Labiatae

17 �� Gramineae

18 �� #�� Plumbaginaceae


20 �� Labiatae

21 �� Rosaceae

22 (�� #�� #�%�� Scrophulariaceae

23 )�� #��#�� Gramineae


25 �� * �� Gramineae

26 +�� Umbelliferae


28 �� cruciferae

29 �� #�� Caryophyllaceae

30 ,� ��&��#� � Cruciferae

31 �� #�� Umbellifetae

32 �� # �� Malvaceae

33 �� #�� Compositae

34 $� ��#�� Caryophyllaceae

35 )�� Rosaceae

36 � � � �� Compositae

37 !�� Rosaceae

38 !�� Polygonaceae


82

Appendix 3-2:List of inventoried plant species No. Name of species Family 39 ��#�� Boraginaceae

40 '�#�� Compositae

41 "��% �� Umbilliferae

42 �� #�� #��%� �� Gramineae

43 -�*��#�� Labiatae

44 �#�� Labiatae

45 �� Gramineae

46 "�� Berberidaceae

47 �� Gramineae

48 '�#�� .�� Ephedraceae

49 � �� Salicaceae

50 �� # �� Polygonaceae

51 !# �� Rhamnaceae

52 �� Rosaceae

53 /�� Gramineae

54 /�� Primulaceae

55 � �� Caryophyllaceae

56 ) �� %�� Compositae


83

Appendix 4-1: Plant species of Sirachal forest island

��

��

"��

��#��# ��

�� # �

��

�#� ��

$ ��

"�%%��

�#��&��#� ��

'��#��

��

��

��

��

�#��

��

�� #��

��

��

��

(�� #�� #�%��

)�� #��#��

��

�� * ��

+��

��

��

�� #��

,� ��&��#� �

�� #��

�� # ��

�� #��

$� ��#��

)��

� � � ��

!��


84

Appendix 4-2: Plant species of Kalvn Forest islands Forest island No. 3 Kalvan 1

��

��

��

�#��&��#� ��

Forest island No.4 Kalvan 2

��

�� #��

��

��

��


��

��

��

��

�#��&��#� ��

�� #��


��

��

��

��

'��#��


85

Appendix 4-3: Plant species of Sercheh forest islands

Forest island No. 1 Sercheh 1

��

��

�� &

"��

)�� #��#��

(�� #�� #�%��

�#��&��#� ��

'�#��

��

Forest island No. 2 Sercheh 2

��

�� &

��

�#��&��#� ��

"��

'�#��

(�� #�� #�%��

)�� #��#��

�� * ��

��

"��

��

��

��


86

Appendix 4-4: Plant species of Lanyz 1 forest island

Forest island No. 10 Lanyz 1

��

��

"��

'�#��

��

��

��#��

$ ��

!��

�#��&��#� ��

'��#��

��

) �� %��

� ��

��

�#��

��

�� #��

��

��

/��

(�� #�� #�%��

)�� #��#��

/��

�� * ��

"��% ��

��

"��

!# ��

�� # ��

�� #��

� ��

'�#�� .��


87

Appendix 4-5: Plant species of Lanyz 2 forest islands

Forest island No. 11 Lanyz 2

��

�#��&��#� ��

)�� #��#��

��

��

��

�� #��

��

!��

�� #��

!# ��

�#��


88

Appendix 4-6: Plant species of Rayzamin

Forest island No. 8 Rayzamin forest island

��

"��

�#��&��#� ��

��

��

�� #��

��

�� #��

��

�� # �

��

��

��

$ ��

!��

��#��# ��

�� * ��

!��

'��#��

(�� #�� #�%��


89

Appendix 4-7: Plant species of Hamega forest island

Forest island No. 6 Hamega

��

��

"��% ��

)�� #��#��

'�#��

��#��

!��

��

�� #�� #��%� ��

(�� #�� #�%��

-�*��#��

��

�#��

��


90

Appendix 5: Summary of area calculated from relief map Convex Flat Concave 457200 1736100 388800 1599300 5753700 1422900 24300 121500 36000 3600 100800 35100 55800 323100 62100 519300 1972800 406800 52200 251100 45900 235800 985500 169200 153900 1495800 103500 355500 2381400 351000 55800 124200 62100

Total 3512700 15246000 3083400

Convex Flat Concave Total Sq. meter 3512700 15246000 3083400 21842100 Hectare 351.27 1524.6 308.34 2184.21

Convex Flat Total 351.27 1524.6 1875.87

Percentage of flat and convex area 85.8832

JUNIPER ISLANDS AND PLANT DIVERSITY · JUNIPER ISLANDS AND PLANT DIVERSITY Salah Ahmed EL Mahi March, 2003 . JUNIPER ISLANDS AND PLANT DIVERSITY A case Study with Remote Sensing and

Documents