Assessment of Damage by the Tree Locust (Anacridium Melanorhodon Melanorhodon ) on Hashab Tree (Acacia Senegal) Using Ground Surveys and Remote Sensing Data

Assessment of Damage by the Tree Locust (Anacridium melanorhodon melanorhodon ) on Hashab Tree (Acacia senegal) using Ground Surveys and Remote Sensing Data

By Ahmed Ismail Ahmed Safi

B.sc. Agric.(Honours) Plant Protection Faculty of Agriculture University of Khartoum (1983) M.Sc. Faculty of Natural Resources and Environmental Studies University of Kordofan (2005) A thesis submitted to the Graduate College, University of Khartoum in fulfillment of the requirements for the degree of Doctor of Philosophy in Agriculture

Supervisor: Prof. Dr. El Sayed El Bashir Mohamed Co-supervisor: Dr. Amna Ahmed Hamid

Department of Crop Protection Faculty of Agriculture University of Khartoum May 2011

Members of Supervisory Committee: Signature

Dr. Saied Elbashier, Assoc. Professor, ………………….. Faculty of Agriculture. University of Khartoum, Sudan. (Supervisor)

Dr. Ammna Ahmed Hamid ………………

Remote Sensing Authority Director, Khartoum

(Co- Supervisor)

Examiners Committee: Signature

Dr. ……………………………, ……………………………. Assoc. Professor, University of Khartoum (External examiner)

Dr. ………………………………., ……………………………… Assist. Professor, University of Khartoum (Internal examiner)

Dr. ………………………………., ……………………………….. Assoc. Professor, Faculty of Agriculture. University of Khartoum, Sudan. (Internal examiner)

DEDICATION

I dedicate this study to all members of my family and the people who promote the following values: humanity, honesty, efficiency, punctuality, accuracy, liberty, equality and fraternity.

ahmed

ACKNOWLEDGEMENTS

The positive contribution provided by many individuals and several institutions in the completion of this study is highly appreciated and accepted as well.

First of all and last I would like to thank Allah, for giving me the health, strength and patience that supported me throughout the study period.

I would like to express my special thanks and gratitude to my supervisor Prof. Dr. El Sayed El Bashir Mohamed, my co-supervisor Dr. Amna Ahmed Hamed (Director of remote sensing authority) for their advice and guide.

I owe a lot of thanks to Prof. Emam Elkhidir and Prof. Ahmed Hashim for their valuable discussion and constructive comments in the seminar presented in Crop Protection Department, Faculty of Agriculture, University of Khartoum.

The generous financial support from Elshiekh Zaki Hamdan, Director of Sheikan Insurance and Re-insurance Company Khartoum, is very much appreciated.

Really, I appreciate very much the limitless help and the family atmosphere offered by the staff of Planet Action, Spot, France with special emphasis and deepest appreciation to Miss Patrica Dankha and Dr. Algis for their continuous help, providing free Spot imagery data of the study area.

Really, I appreciate very much the limitless help and the family atmosphere offered by the staff of Remote Sensing Authority, Khartoum and I would like to emphasize my deepest appreciation to Dr. Moayad and Mr. Abdelrahman Ahmed Khatir for their continuous help and support in data analysis.

Deep thanks are due to Acacia Company staff; Fisal, Rabie and their colleagues for their help, providing the study area and logistic support.

Thanks extend to my colleagues in Gum Arabic Research Centre, University of Kordofan.

Last and not least I would like to express my deep gratitude to my family for their limitless patience and support.

Abstract

In the present study, a field survey, laboratory experiments and remote sensing were integrated and carried out for three successive seasons. The objectives were to access the role of tree locust infestation on quantity and quality of the produced gum arabic and to develop an alternative method for field survey of this pest. The field and remote sensing trials were conducted in an area of 28000 feddans (11331 hectares) of Acacia senegal plantation of the Acacia Project (Nawa and Elrahad locations), 37 km south east of El Obeid city, North Kordofan State. The results showed that adults were found throughout the year except in February, March and April and their number decreased with increase in rainfall and relative humidity. Hoppers appeared in July, August, September and October and their number increased with increase in rainfall and relative humidity. Experiments on quantitative effect on gum production due to natural and artificial defoliation of hashab (Acacia senegal) trees were carried out, where four blocks were chosen randomly, and the following treatments were arranged in a randomized complete block design: control (no defoliation), light natural defoliation, moderate natural defoliation, high natural defoliation, light artificial defoliation, moderate artificial defoliation and high artificial defoliation. The results revealed that tree locust infestation and the artificial defoliation of trees severely reduced gum production, and the reduction was highly significant (P≤0.001) between means of all treatments except between high natural defoliation and high artificial defoliation. There was a negative correlation between gum production and levels of defoliation. Laboratory experiments showed that defoliation reduced gum viscosity, and the reduction was significant between all means except between light natural defoliation and light artificial defoliation. Defoliation also reduced gum optical rotation, but the reduction was not statistically significant.

III

The developmental period of nymphs fed on young leaves was shorter than that of nymphs fed on old leaves. Both adults and hoppers daily consumed food equivalent to their body weight. Remotely sensed data, used in this study were collected from the American Landsat satellite (with sensors; ETM+) images dated August and October 2008, and the French Spot Satellite (with the sensors; XS) images dated August and November 2009. Visual interpretation of images showed that non- defoliated hashab trees appeared as red and defoliated ones appeared as grey. The Normalized Difference Vegetation Index (NDVI) differentiated between non-defoliated hashab trees and the three levels of natural defoliations. There was a positive relationship between NDVI value, tree greenness and gum production Supervised classification showed five major classes; non-defoliated, light, moderate and high defoliated hashab trees as well as areas of tree locust swarms. Supervised classes matched with NDVI classes.

اﻟﻤﺴﺘﺨﻠﺺ ﺗﻢ إﺳﺘﺨﺪام اﻟﻤﺴﺢ اﻟﺤﻘﻠﻰ واﻟﺘﺠﺎرب اﻟﻤﻌﻤﻠﻴﺔ واﻹﺳﺘﺸﻌﺎرﻋﻦ ﺑﻌﺪ ﻟﺜﻼﺛﺔ أﻋﻮام ﻣﺘﺘﺎﻟﻴﺔ، ﺑﻬﺪف ﺗﻘﻴﻴﻢ دور إﺻﺎﺑﺔ اﻟﺠﺮاد ﺳﺎرى اﻟﻠﻴﻞ ﻷﺷﺠﺎراﻟﻬﺸﺎب ﻓﻰ آﻤﻴﺔ اﻟﺼﻤﻎ اﻟﻤﻨﺘﺞ وﻧﻮﻋﻴﺘﻪ و آﺬاﻟﻚ ﺗﻄﻮﻳﺮ ﻃﺮﻳﻘﺔ ﺑﺪﻳﻠﺔ ﻟﺤﺼﺮ هﺬﻩ اﻷﻓﺔ ﻣﻴﺪاﻧﻴﺎً. أُﺟﺮﻳﺖ اﻟﺘﺠﺎرب اﻟﺤﻘﻠﻴﺔ واﻟﻤﻌﻤﻠﻴﺔ ﻓﻲ ﻣﺴﺎﺣﺔ 28000 ﻓﺪان (11331هﻜﺘﺎر) ﺑﻐﺎﺑﺔ اﻟﻬﺸﺎب (Acacia senegal) اﻟﻤﺴﺘﺰرﻋﺔ ﺑﻤﺸﺮوع أآﻴﺸﻴﺎ اﻟﺰراﻋﻰ (ﻣﻮﻗﻌﻰ ﻧﻮى واﻟﺮهﺪ )، 35 آﻠﻢ ﺟﻨﻮب ﺷﺮق ﻣﺪﻳﻨﺔ اﻷﺑﻴﺾ، وﻻﻳﺔ ﺷﻤﺎل آﺮدﻓﺎن. أوﺿﺤﺖ اﻟﻨﺘﺎﺋﺞ وﺟﻮد اﻟﺤﺸﺮات اﻟﻜﺎﻣﻠﺔ ﻋﻠﻰ ﻣﺪاراﻟﺴﻨﺔ ﻣﺎﻋﺪا ﻓﺒﺮاﻳﺮ و ﻣﺎرس وأﺑﺮﻳﻞ وﻗﻠﺖ أﻋﺪادهﺎ ﻣﻊ زﻳﺎدة آﻤﻴﺔ اﻷﻣﻄﺎر واﻟﺮﻃﻮﺑﺔ اﻟﻨﺴﺒﻴﺔ. ﻇﻬﺮت اﻟﺤﻮرﻳﺎت ﻓﻲ ﻳﻮﻟﻴﻮ وأﻏﺴﻄﺲ و ﺳﺒﺘﻤﺒﺮ وأآﺘﻮﺑﺮوزادت أﻋﺪادهﺎ ﺑﺰﻳﺎدة آﻤﻴﺔ اﻷﻣﻄﺎر واﻟﺮﻃﻮﺑﺔ اﻟﻨﺴﺒﻴﺔ. أُﺟﺮﻳﺖ ﺗﺠﺎرب ﻟﺘﺤﺪﻳﺪ اﻟﺘﺄﺛﻴﺮاﻟﻜﻤﻲ ﻋﻠﻰ إﻧﺘﺎج اﻟﺼﻤﻎ ﺑﺴﺒﺐ إزاﻟﺔ أوراق ﺷﺠﺮة اﻟﻬﺸﺎب ﻧﺘﻴﺠﺔ ﻟﻬﺠﻮم اﻟﺠﺮاد ﺳﺎرى اﻟﻠﻴﻞ واﻹزاﻟﺔ اﻻﺻﻄﻨﺎﻋﻴﺔ. أُﺧﺘﻴﺮت أُرﺑﻌﺔ ﻣﺮﺑﻌﺎت ﺑﺸﻜﻞ ﻋﺸﻮاﺋﻲ، ووزﻋﺖ اﻟﻤﻌﺎﻣﻼت اﻟﺘﺎﻟﻴﺔ ﺑﻄﺮﻳﻘﺔ اﻟﻘﻄﺎﻋﺎت آﺎﻣﻠﺔ اﻟﻌﺸﻮاﺋﻴﺔ: اﻟﺸﺎهﺪ وإﺻﺎﺑﺔ ﻃﺒﻴﻌﻴﺔ ﺧﻔﻴﻔﺔ وإﺻﺎﺑﺔ ﻃﺒﻴﻌﻴﺔ ﻣﺘﻮﺳﻄﺔ وإﺻﺎﺑﺔ ﻃﺒﻴﻌﻴﺔ ﻋﺎﻟﻴﺔ وإﺻﺎﺑﺔ ﻣﺴﺘﺤﺪﺛﺔ ﺧﻔﻴﻔﺔ وإﺻﺎﺑﺔ ﻣﺴﺘﺤﺪﺛﺔ ﻣﺘﻮﺳﻄﺔ وإﺻﺎﺑﺔ ﻣﺴﺘﺤﺪﺛﺔ ﻋﺎﻟﻴﺔ. أﻇﻬﺮت اﻟﻨﺘﺎﺋﺞ أن اﻹﺻﺎﺑﺔ ﺑﺠﺮاد اﻟﺸﺠﺮﺗﻘﻠﻞ ﺑﺪرﺟﺔ آﺒﻴﺮة إﻧﺘﺎج اﻟﺼﻤﻎ وآﺎن هﻨﺎك ﻓﺮق ﻣﻌﻨﻮى ﻋﺎﻟﻲ (P<0.001) ﺑﻴﻦ ﺟﻤﻴﻊ اﻟﻤﺘﻮﺳﻄﺎت ﻣﺎﻋﺪا ﺑﻴﻦ اﻹﺻﺎﺑﺔ اﻟﻄﺒﻴﻌﻴﺔ اﻟﻌﺎﻟﻴﺔ واﻹﺻﺎﺑﺔ اﻟﻤﺴﺘﺤﺪﺛﺔ اﻟﻌﺎﻟﻴﺔ وإرﺗﺒﺎط ﺳﻠﺒﻰ ﺑﻴﻦ إﻧﺘﺎج اﻟﺼﻤﻎ وﻣﺴﺘﻮى اﻹﺻﺎﺑﺔ. أﺛﺒﺘﺖ اﻟﺘﺠﺎرب ﻓﻰ اﻟﻤﺨﺘﺒﺮ أن إزاﻟﺔ أُوراق اﻷُﺷﺠﺎر ﻳﺨﻔﺾ درﺟﺔ ﻟﺰوﺟﺔ اﻟﺼﻤﻎ وآﺎن اﻟﺘﺨﻔﻴﺾ ﻣﻌﻨﻮﻳﺎً (P<0.001) ﺑﻴﻦ آﻞ اﻟﻤﺘﻮﺳﻄﺎت ﻣﺎﻋﺪا ﺑﻴﻦ اﻹﺻﺎﺑﺔ اﻟﻄﺒﻴﻌﻴﺔ اﻟﺨﻔﻴﻔﺔ واﻹﺻﺎﺑﺔ اﻟﻤﺴﺘﺤﺪﺛﺔ اﻟﺨﻔﻴﻔﺔ. أ ﻳ ﻀ ﺎً اﻇﻬﺮت اﻟﺘﺠﺮﺑﺔ أن إﺻﺎﺑﺔ اﻟﺠﺮاد ﺗﺆﺛﺮ ﻋﻠﻰ اﻟﺪوران اﻟﻨﻮﻋﻰ ﻟﻜﻦ ﻟﻢ ﺗﻜﻦ هﻨﺎك ﻓﺮوﻗﺎت ﻣﻌﻨﻮﻳﺔ ﺑﻴﻦ آﻞ اﻟﻤﺘﻮﺳﻄﺎت. إﺳﺘﻐﺮﻗﺖ اﻟﺤﻮرﻳﺎت اﻟﺘﻰ ﺗﻤﺖ ﺗﻐﺬﻳﺘﻬﺎ ﻋﻠﻰ اﻷوراق اﻟﺼﻐﻴﺮة اﻟﻌﻤﺮ ﻣﺪة ﺗﻄﻮرأﻗﺼﺮﻣﻦ اﻟﺤﻮرﻳﺎت اﻟﺘﻰ ﺗﻤﺖ ﺗﻐﺬﻳﺘﻬﺎ ﻋﻠﻰ اﻷوراق اﻟﻜﺒﻴﺮة اﻟﻌﻤﺮ. إﺳﺘﻬﻠﻜﺖ آ ﻞِ ﻣﻦ اﻟﺤﺸﺮات اﻟﻜﺎﻣﻠﺔ واﻟﺤﻮرﻳﺎت ﻏﺬاء ﻳﻮﻣﻲ ﻳﻜﺎﻓﺊ وزﻧﻬﺎ. ﻓﻰ ﻣﺠﺎل اﻹﺳﺘﺸﻌﺎرﻋﻦ ﺑﻌﺪ ﺗﻢ اﻟﺤﺼﻮل ﻋﻠﻰ ﺻﻮرأﻗﻤﺎرأﻣﺮﻳﻜﻴﺔ Landsat (8/2008 و 10/2008) وﻓﺮﻧﺴﻴﺔ Spot (2009/8 و11/2009). أوﺿﺢ اﻟﺘﻔﺴﻴﺮاﻟﺒﺼﺮى ﻟﺼﻮراﻷﻗﻤﺎر وﺟﻮد أﺷﺠﺎرهﺸﺎب ﻏﻴﺮ ﻣﺼﺎﺑﺔ ذات أﻟﻮان ﺣﻤﺮاء وأﺷﺠﺎر ﻣﺼﺎﺑﺔ ذات أﻟﻮان رﻣﺎدﻳﺔ. أوﺿﺢ ﻣﺆﺷﺮاﻹﺧﻀﺮار اﻷﺷﺠﺎرﻏﻴﺮاﻟﻤﺼﺎﺑﺔ وﺛﻼﺛﺔ ﻣﺴﺘﻮﻳﺎت ﻟﻺﺻﺎﺑﺔ اﻟﻄﺒﻴﻌﻴﺔ وﺗﻨﺎﺳﺐ ﺗ ﻨ ﺎ ﺳ ﺒ ﺎً ﻃﺮدﻳﺎً ﻣﻊ ﺣﺎﻟﺔ إﺧﻀﺮاراﻟﺸﺠﺮة وإﻧﺘﺎج اﻟﺼﻤﻎ ودل اﻟﺘﺼﻨﻴﻒ اﻟﻤُﺮاﻗﺐ ﻋﻠﻰ أن هﻨﺎﻟﻚ ﺧﻤﺲ ﻣﺠﻤﻮﻋﺎت

رﺋﻴﺴﻴﺔ: أﺷﺠﺎرﻏﻴﺮ ﻣﺼﺎﺑﺔ وأﺷﺠﺎر ذات إﺻﺎﺑﺔ ﺧﻔﻴﻔﺔ و أﺷﺠﺎر ذات إﺻﺎﺑﺔ ﻣﺘﻮﺳﻄﺔ وأﺷﺠﺎر ذات إﺻﺎﺑﺔ ﻋﺎﻟﻴﺔ وﻣﻨﺎﻃﻖ أﺳﺮاب ﺟﺮاد اﻟﺸﺠﺮ. وﺟﺪ ﺗﻮاﻓﻖ ﺑﻴﻦ اﻟﺘﺼﻨﻴﻒ اﻟﻤُﺮاﻗﺐ وﻣﻮًﺷﺮاﻹﺧﻀﺮار.

LIST OF CONTENTS

Subject Page DEDICATION I ACKNOWLEDGEMENTS I I ABSTRACT I I I ARABIC ABSTRACT V LIST OF CONTENTS VII LIST OF TABLES XI LIST OF FIGURES XIII LIST OF TEMPLATES XV LIST OF IMAGES XVI CHAPTER ONE: INTRODUCTION 1 1.1 Objectives 3 CHAPTER TWO: LITERATURE REVIEW 5 2.1 The tree locust Anacridium melanorhodon melanorhodon 5 2.1.1 Taxonomy and nomenclature 5 2.1.2 Geographical distribution 6 2.1.3 General morphology 8 2.1.4 Economic importance and damage 10 2.1.4.1 Host plants and feeding habits A.m.melanorhodon 14 2.1.5 Ecology 17 2.1.6 Biology 18 2.1.7 Natural enemies 20 2.1.8 Control 21 2.1.8.1Chemical control 21 2.1.8.2 Mycopesticides 22 2.1.8.3 Botanicals 22

VII

2.1.8.4 Cultural control of grasshoppers and locusts 23 2.2 Remote sensing 24 2.2.1 Why is remote sensing 25 2.2.2 Applications of remote sensing in Sudan 25 2.2.3 Digital image processing 27 2.2.4 Images interpretation and analysis 28 2.2.5 Image classification 29 2.3 Geographic Information System (GIS) 31 2.4 Global Positioning System (GPS) 31 2.5 Crop monitoring & damage assessment 32 2.6 NDVI versus other indices 33 2.6.1 Why NDVI? 33 CHAPTER THREE: MATERIALS AND METHODS 36 3.1 Study area 36 3.1.1 Criteria for selection of study area 36 3.2 Field survey 38 3.2.1 Seasonal occurrence of adult and hopper tree locusts 38 3.2.1.1 Counting of the 1st and 2nd instar hoppers 38 3.2.1.2 Counting of the 3rd, 4th, 5th, and 6th instar hoppers on 38 Trees 3.2.1.3 Counting of the adult tree locusts on trees 38 3.2.2 Quantitative effect of natural and artificial defoliation on 40 gum Production 3.2.2.1 Collection and preparation of the neem seed 40 3.2.2.2 Preparation of neem seed aqueous extracts (Aq-extr) 44 3.2.2.3 Tapping and gum collection 44 3.3 Laboratory experiments 44 3.3.1 Preparation of gum samples 44

VIII

3.3.2 Effect of natural and artificial defoliation on gum quality 44 3.3.2.1 Gum viscosity 44 3.3.2.2 Specific optical rotation 46 3.3.3 Assessment of the daily food intake and the effect of 46 hashab leaf age on the adult and hopper development 3.3.4 Soil analysis 48 3.3.4.1 Soil moisture determination 48 3.3.4.2 Soil texture 48 3.3.4.3 PH determination 48 3.3.4.4 Nitrogen determination 49 3.4 Remotely sensed data 49 3.4.1 Ancillary data 50 3.4.2 Software 50 3.4.3 Hardware 50 3.4.4 Visual interpretation 50 3.4.5 Normalized Difference Vegetation Index (NDVI) 51 Analysis 3.4.6 Supervised classification 51 3.4.7 Data analysis 52 3.4.8 Layout and mapping 52 CHAPTER FOUR: RESULTS 53 4.1 Field survey 53 4.1.1 Seasonal occurrence of tree locust adults and hoppers in 53 relation to rainfall and relative humidity 4.1.2 Quantitative effect of natural and artificial defoliation on 53 gum production 4.2 Laboratory experiments 57 4.2.1 Effect of natural and artificial defoliation on gum quality 57

4.2. 1.1 Gum viscosity (in centistokes) 57 4.2. 1.2 Gum specific optical rotation 60 4.2.2 The daily food intake and the effect of hashab leaf age 60 on hopper development 4.2.3 Soil analysis 60 4.2.3.1 Soil moisture determination 60 4.2.3.2 Soil texture 63 4.2.3.3 PH determination 63

4.2.3.4 N2 determination 63 4.3 Remote sensing interpretation and analysis 63 4.3.1 Visual interpretation of image 63 4.3.2 Normalized Difference Vegetation Index (NDVI) 69 Analysis 4.3.3 Supervised classification 82 4.3.4 Layout and mapping 84 CHAPTER FIVE: DISCUSSION 89 CONCLUSIONS AND RECOMMENDATIONS 94 REFERENCES 96 APPENDECES 107

LIST OF TABLES

Subject Page Table 1. Characteristics of tree locust, A. m. melanorhodon 9 Table 2. Gum hashab production by regions in Sudan, 1990- 13 2010 (quantities in metric tons) Table 3. Host plants of A. m. melanorhodon and the degree of 15 preference (or damage) Table 4. Mean weight (g) of gum produced / hashab tree at 55 different blocks and treatments at Nawa and Elrahad locations (seasons 2007/2008, 2008/2009 and 2009/2010) Table 5. Gum viscosity (%) at different blocks and treatments 58 at Nawa and Elrahad locations (seasons 2007/2008, 2008/2009 and 2009/2010) Table 6.The effect of leaves age on hoppers development 62 (seasons 2007/2008, 2008/2009 and2009/2010) Table 7. Mean of NDVI values at different blocks and 72 treatments at Nawa location, season 2008/2009 Table 8. Correlation between NDVI and gum production at 72 Nawa location, season 2008/2009 Table 9. Analysis of variance of NDVI values at Nawa 72 location (season2008/2009) Table 10. Mean of NDVI values at different blocks and 75 treatments at Nawa location (season 2009/2010) Table 11. Correlation between NDVI and gum production at 75 Nawa location (season 2009/2010) Table 12. Analysis of variance of NDVI values at Nawa 75

XI location (season 2009/2010)

Table 13. Mean of NDVI values at different blocks and 78 treatments at Elrahad location (season 2008/2009) Table 14. Correlation between NDVI and gum production at 78 Elrahad location(season 2008/2009) Table 15. Analysis of variance of NDVI values at Elrahad 78 location (season2008/2009) Table 16. Mean of NDVI values at different blocks and 81 treatments at Elrahad location (season 2009/2010) Table 17. Correlation between NDVI and gum production at 81 Elrahad location (season 2009/2010) Table 18. Analysis of variance of NDVI values at Elrahad 81 location (season2009/2010)

XII

LIST OF FIGURES

Subject Page Figure 1. Distribution map of tree locust in Africa 7 Figure 2. Gum arabic production distribution in Sudan 12 Figure 3. Relations between gum producer, Acacia senegal tree 13 and tree locust Figure 4. Life cycle of tree locust A.m. melanorhodon 19 Figure 5. Spectral resolution of vegetation 34 Figure 6. Reflectance of vegetated land areas 34

Figure 7. Reflectance of non-vegetated land areas 35 Figure 8. NDVI calculation 35 Figure 9. Location map of the study area 37 Figure 10. Layout plan showing blocks (B) and estimation points 39 (o) at Nawa and Elrahad locations Figure 11. Seasonal occurrence and mean number of adults and 54 hoppers/tree at Nawa and Elrahad locations in relation to relative humidity (seasons 2007/2008, 2008/2009 and 2009/2010) Figure 12. Seasonal occurrence and mean number of adults and 54 hoppers/tree at Nawa and Elrahad locations in relation to rainfall (seasons 2007/2008, 2008/2009 and 2009/2010) Figure 13. Mean weight (g) of gum produced / hashab tree at 56 different treatments at Nawa and Elrahad locations (seasons 2007/2008, 2008/2009 and 2009/2010)

XIII

Figure 14. Mean of gum viscosity (%) at different treatments at 59 Nawa and Elrahad locations (seasons 2007/2008, 2008/2009 and 2009/2010) Figure 15. Mean of gum specific optical rotation at different 61 treatments at Nawa and Elrahad locations, seasons 2007/2008, 2008/2009 and 2009/2010 Figure 16. Area of supervised classes of Landsat image at Nawa 85 location (August and October 2008)

Figure 17. Area of supervised classes of Spot image at Nawa 86 location (August and November2009) Figure 18. Area of Supervised classes of Landsat image at 87 Elrahad location (August and October 2008) Figure 19. Area of supervised classes of Spot image at Elrahad 88 location (August and November 2009)

XIV

LIST OF TEMPLATES

Subject Page Template 1. Morphological difference between male and female 9 of tree locust A. m. melanorhodon Template 2. A.m.melanorhodon 4th instar on Acacia senegal 16 Shoot Template 3. A. m. melanorhodon 6th instar on Grewia tenax shoot 16 Template 4. A. m. melanorhodon adult on Balanites aegyptiaca 16 Shoot Template 5. Caged tree + tree locusts before defoliation 41 Template 6. Caged tree after defoliation 41 Template 7. Tapping method using the Sunki tool 45 Template 8. Rearing and breeding of tree locust 47 Template 9. Cage+ hoppers + leaves of hashab 47

LIST OF IMAGES

Subject Page Image 1. Spot image showing the different blocks and treatments 42 at Nawa location 2009 Image 2. Spot image showing the different blocks and treatments 43 at Elrahad location2009 Image 3. Landsat image of Nawa location dated August 2008 65

Image 4. Landsat image of Nawa location dated October 2008 65 Image 5. Spot subset image of Nawa location dated August 2009 66 Image 6. Spot subset image of Nawa location dated November 66 2009 Image 7. Landsat image of Elrahad location dated August 2008 67 Image 8. Landsat image of Elrahad location dated October 2008 67 Image 9. Spot subset image of Elrahad location dated August 68 2009 Image 10. Spot subset image of Elrahad location dated November 68 2009 Image 11. Landsat NDVI image of Nawa location dated 70 August 2008 Image 12. Landsat NDVI image of Nawa location dated October 71 2008 Image 13. Spot NDVI image of Nawa location dated August 73 2009 Image 14. Spot NDVI image of Nawa location dated November 74 2009 Image 15. Landsat NDVI image of Elrahad location dated August 76 2008

XVI

Image 16. Landsat NDVI image of Elrahad location dated 77 October 2008 Image 17. Spot NDVI image of Elrahad location dated August 79 2009 Image 18. Spot NDVI image of Elrahad location dated November 80 2009 Image 19. Supervised classification of Landsat image of Nawa 85 location dated August2008 Image 20. Supervised classification of Landsat image of Nawa 85 location dated October 2008 Image 21. Supervised classification of Spot image of Nawa 86 location dated August 2009 Image 22. Supervised classification of Spot image of Nawa 86 location dated November 2009 Image 23. Supervised classification of Landsat image of Elrahad 87 location dated August 2008 Image 24. Supervised classification of Landsat image of Elrahad 87 location dated October2008 Image 25. Supervised classification of Spot image of Elrahad 88 location dated August 2009 Image26. Supervised classification of Spot image of Elrahad 88 location dated November 2009

XVII CHAPTER ONE INTRODUCTION

North Kordofan State extends between latitudes 12ْ 14ً N and 16ْ 38ً N and longitudes 26ْ 56 ًE and 32ْ 22ً E with total area of 190,840 km2. It lies in low rainfall savannah zone within the semi- arid climate where the annual rainfall ranged from 200-400mm (Kevie, 1973, Abdulla, 2006 and Khiry, 2007). In the northern part, the soils are classified as Cambric Arenosols (according to FAO system of soil classification) and are locally named Goz. They are coarse textured sandy soils of Aeolian origin. They have high infiltration rates and inherent low fertility. In the southern part, the soils are alluvial in origin with a silty clay texture. They are non-cracking clay soils mixed with Aeolian sand and locally named Gardud (FAO, 1997). The State is sparsely vegetated, vegetation is exposed to extreme conditions and must survive drought, which can stretch over several years with little or no rain at all. In the semi-arid ecosystems with a single rainy season there is usually a short growth period followed by a long dry season with great reduction in the amount of green plant material (Schmidt and Karnieli, 2000 and Dafalla, 2006,). The extremes of air temperature are 10° C and 40° C, the mean relative humidity ranges from 21% in the dry season to about 75% during the rainy season, the wind speed is usually less than 8 Km/hr (Eldukheiri, 1997). Acacia senegal (L.) Willd, is a tree, approaching 8 m in height, locally known as hashab tree (Elamin, 1990 and Vogt, 1995). In Africa, it has a wide geographical distribution in the savannah belt. In Sudan it is found in the so-called, gum arabic belt, which lies within the low-rainfall savannah zone (approximately between latitudes 10-14°N). It produces the best quality of gum arabic, good fodder for livestock, fuel wood, charcoal, poles for building purposes and fence posts. It is useful for environmental protection and conservation especially in sandy soils (Goz) because it fixes

1 sand dunes and reduces wind and water erosion. In addition, it fixes nitrogen and the annual nitrogen fixation estimated to be less than 20 Kg nitrogen per ha (Ballal, 2002 and Taha, 2000). Gum arabic is a non-timber forest product provided by Acacia spp. namely A. senegal (GAC, 2000). Gum arabic contributes significantly to the revenue of local communities and to Sudan exports where it constitutes a major source of foreign exchange earnings. It has multiple uses in industrial products such as foodstuffs, pharmaceuticals, beverages, ink, textile, paints, lithography, adhesives and a wide range of additional industrial products (Ballal, 2002). Gum arabic provides an average of 12% of the gross domestic products (GDP) of the country and accounts for about 15.3 % and 10% of the household income of the gum producers and other farmers in the gum belt across the Sudan, respectively. It is often the main source of revenue for semi-nomadic Africans who gather it from wild, untended plants (Awouda, 1973, 1974, Taha, 2000 and Ballal, 2002). There are principal factors governing gum production as summarized by Chikamai (1996) and Taha (2006). All these factors are damaging to the tree and reduce gum production, especially insects. Insect-induced defoliation causes significant timber and carbon losses in hashab tree. The resulting loss may have a substantial impact on wood supply and the terrestrial carbon balance (Hall and Moody, 1994 and Fleming and Volney, 1995). The family Acrididae, including the familiar grasshoppers and the destructive locusts are polyphagous plant feeders. Insect infestation is among the most challenging environmental hazards that gum arabic producers face, particularly tree locust Anacridium melanorhodon melanorhodon Wlk, which is an important defoliator causing serious damage on hashab tree if left unchecked. Infestation starts early in September and persists till complete defoliation, especially during years of outbreak. Defoliation decreases the photosynthetic capability of the gum tree

2 and as a result the quantity of stored carbohydrates decreases. Finally, gum yield and quality are negatively affected (Elbashier, 1994). Since locusts can invade large areas, the use of traditional methods (field survey) for assessing the damage is time consuming. The use of remote sensing method (R.S) integrated with Geographic Information System (GIS) and Global Positioning System (GPS), permitting the rapid monitoring of potentially endangered regions readily suggests itself. Remote sensing can only be employed directly in the case of large-scale damage and not for assessing the damage to single plants. GPS supplemented with ground surveys, currently provides the main source of information on the extent and severity of insect defoliation (Nelson, 1983, Leckie and Ostaff, 1988, Franklin and Raske, 1994, Radeloff et al., 1999, Simpson and Coy, 1999, Heikkila et al., 2002 and Hall et al., 2003). Shank (2007) stated that satellite imaging data can be a valuable tool in forecasting, monitoring and detecting potential breeding grounds thus helping locust control team finds possible targets. In this study some questions are raised to be answered: • What are the impacts of tree locust (Anacridium m. melanorhodon) on Gum arabic production? • Which is the most effective method of remote sensing capable of recognition of tree locust defoliation? • To what extent can remote sensing be a complete substituent to ground surveys? 1.1 Objectives • Assessment of quantitative and qualitative impacts of natural and artificial defoliation on gum production. • Reviewing and evaluating the role of remote sensing and ground survey in identifying areas of tree locust swarm formation and early detection of hashab tree defoliation by tree locust with emphasis on future prospects for

better management of the tree locusts. • Study the effect of leaf age on hopper development. • Developing recommendations on the basis of the results obtained.

CHAPTER TWO LITERATURE REVIEW

2.1 The tree locust Anacridium melanorhodon melanorhodon 2.1.1 Taxonomy and nomenclature Locusts belong to the phylum Arthropoda, the largest and most diverse animal phylum. Arthropods include arachnids, crustaceans, insects, centipedes, millipedes, symphylans, pauropodans, and the extinct trilobites. Locust is a migratory member of the short-horned grasshopper, family (Acrididae), name applied to almost 9,000 different species of singing, jumping insects. Phylum:Arthropoda Class:Insecta Order:Orthoptera Suborder:Caelifera Family:Acrididae Sub-family: Cyrtacanthacridinae Scientific name: Anacridium melanorhodon melanorhodon English name: Tree locust or Sahelian tree locust Arabic name: Sari al leil (night wonderer) The name “Sari al leil” which means “The nocturnal vagabond” or the “night wonderer was given because tree locust adults can fly throughout the night” (http://pestinfo.org/ literature/litout.php3 , viewed on 25/10/ 2008). The genus Anacridium was first established by Uvarov in 1923. The species A. melanorhodon (Walker, 1870) is one of twelve species recognized in this genus by Dirsh and Uvarov (1953) with two subspecies. Popov and Ratcliffe (1968) stated that A. melanorhodon was first described by Walker, under this name from the Cape Verde Island in 1870. It was transferred as a subspecies of Anacridium moestum (Serville, 1938) when the genus Anacridium was

5 erected by Uvarov in 1923. A. moestum has been subdivided into two subspecies A. moestum moestum (Serville, 1938), and A. moestum melanorhodon Walker (Johnston, 1932). According to Dirsh and Uvarov (1953), A. melanorhodon has already been removed to a species and quite separate from A. moestum. A. melanorhodon was divided into two subspecies, A. m. melanorhodon (Walker, 1870) and A. m. arabafrum (Dirsh, 1953). The other synonyms of A. melanorhodon (Walker, 1870) are Orthacanthacris wernerella var. sphalera (Karny, 1907) and Acridium aethiopicum Finot 1907 (Anon., 1982). The other identified species of the genus Anacridium according to Dirsh and Uvarov (1953) are: A.wernerellum (Karny, 1907), A. aegyptium (Linnaeus, 1764), A. moestum (Serville, 1838), A. incisium (Renh, 1942), A. burri (Dirsh, 1953), A. rehn (Dirsh, 1953), A. eximium (Sjostedt, 1918), A. illustrissimum ( Karsch, 1896), A. rubrispinum (Bei Bienko, 1948), A. flavescens (F.), A. arabicum (Uvarov, 1953).

2.1.2 Geographical distribution The economically important species include; Anacridium aegyptium (Linné), the Egyptian tree locust, distributed from the Mediterranean zone to the Middle East, Anacridium melanorhodon (Walker) and Anacridium wernerellum (Karny) which inhabit the Sahelian zone. They occur in Cape Verde Island, Morocco, Mauritania, Mali, Niger, Chad, Nigeria, Ethiopia and Sudan (Meinzingen, 1993). The distribution of Anacridium according to Dirsh and Uvarov (1953) was between latitudes 4°N to 20°N, while A. m. melanorhodon was confined between lat. 12°N to 20°N (Fig. 1). In the Sudan, A. melanorhodon and A. wernerellum have a wide range of distribution, in the east at the Red Sea maritime plain and Red Sea hills. From there it extends south as far as Eritrea and on the west through Kassala to the Nile valley. It appears to be most abundant in the south- west (Darfur and

Legend Distribution of tree locust

African migratory locust outbreak area

African migratory locust invasion area

Figure 1. Distribution map of tree locust in Africa Source:http://ipmworld.umn.edu/chapters/showler.htm viewed on 25/10/ 2008

Kordofan). From Khartoum it extends up the Blue Nile and Gezira district, between the Blue and White Niles (Schmutterer, 1969). Anacridium arabafrum (Dirsh) occurs in the East through Arabia to Iran, and the two subspecies (A. arabafrum and A. melanorhodon) overlap in eastern Sudan and western Eritrea (Anon., 1976). Anacridium moestum (Serville) occurs throughout South Africa, the open forest land in Central Africa, Cape Verde Islands and Central Sahara (Johnston, 1932 and Anon., 1976).

2.1.3 General morphology There are slight morphological and colour differences between these species, but generally, they look the same. The female is 7.5-9.5 cm and the male is 6.5-8cm long. The body is divided into three distinct parts; the head, thorax and abdomen. The head bears fine black antennae, brown eyes and chewing mouthparts. The thorax bears two pairs of wings and three pairs of legs. The abdomen consists of eleven segments and bears the genital organs on its extremity (Table1 and Template 1). Newly laid eggs are bright yellow but shortly change to brown (Johnston, 1932, Schmutterer, 1969 and Cirad, 2006). The first stage hopper is grass- green in colour, 5-6 mm. long, basal joints of the antennae dark green, apical six joints brown, eyes black or very dark brown, pronotum distinctly keeled, brown spots cover the whole body on the dorsal surface. The second hopper stage (swarming phase) is 10-12 mm. long; head pale green covered with small black spots, mouthparts green and basal two joints of the antennae green, remainder part black, except the apical four or five joints that are light brown. In the second hopper stage (solitary phase) there is a reduction of the black markings on the body. The third hopper stage (swarming phase) is 15-18 mm. long; colour yellowish green, there is a notable change in the shape of the hopper at this stage, owing to the keel becoming more convex and thus giving a deeper appearance to the pronotum region. In the third hopper stage (solitary phase), the colour is pure green

Table 1. Characteristics of the tree locust, A. m. melanorhodon.

Type of Sex Elytra Femur Width of E/F population length (mm) length (mm) head (mm) Solitarious Male 64.2 27.5 7.20 2.33 Female 74.0 32.5 8.00 2.27 Gregarious Male 57.0 29.0 6.50 2.24 Female 60.5 34.0 6.64 2.29

Source: Popov et al., (1968)

Male - tip of abdomen smooth and rounded

Female - tip of abdomen jagged

Template 1. Morphological difference between male and female of tree locust A. m. melanorhodon

9 the black markings on the body almost totally disappear, with the exception of the black streak under the eyes. The fourth hopper stage (swarming phase) is 32 mm. long; colour greenish yellow, antennae black, except the tip that is yellow, fronts and mouthparts diffusely black spotted. In the fourth hopper stage (solitary phase), the colour is bright leaf green with the exception of a yellow area between the eyes and hind margin of the head and the blue streaks beneath the eyes. The fifth hopper stage (swarming phase) is very similar to the preceding stage except that there is a great increase in size. The fifth hoppers stage (solitary) like the hoppers of the fourth solitary stage (isolated hoppers are green or brown and become bright yellow with some dark colour when grouped). Imagoes are large and slender. They are grey or brown, never green, the elytra are long, with a pinkish tinge at the base of the wings which are quite dark at the ends (Johnston, 1932, Schmutterer, 1969, Pastre et al., 1988 and Cirad, 2006).

2.1.4 Economic importance and damage Tree locusts are important occasional pests of fruit trees, rice, sorghum, and cotton. In tropical regions, gum arabic producing acacias are the most severely affected trees. Swarms of migrating imagoes (gregarious phase) chronically attack them in the dry season thus reducing gum production. (Pastre et al., 1988,Taha, 2000 and Cirad, 2006). Pastre et al., (1988) reported that a winged locust weighing 2 grams, during its gregarious phase, can eat an amount of fresh leaf equivalent to its body weight per day, and a swarm covering 1Km2 (50 million locusts) can consume a hundred ton of vegetal material. In the Sudan, Abdulla (1990) reported that the infested area in 1987 was estimated at 20000 ha, increasing to 224000 ha the following season, reaching an unprecedented level in 1989 of 1.366.762 ha. It was the most serious outbreak of tree locust, because the entire gum arabic belt in western, central and eastern regions of the country were infested and gum production

10 decreased. Damage was reported mainly on A. senegal in addition to other Acacia species, B. aegyptiaca, Z. spina christi, fruit trees and sorghum in the milky stage. In mid-60s around 50,000 tons of gum were annually exported, but that decreased steadily to 18.000 tons during the 1990’s (Eldukheiri, 1997). Over the last two decades and since the inception of the drought years during the 1970’s coupled with pest outbreaks, gum arabic production has decreased significantly (Fig. 2). The continuous decrease in production has been attributed to low rainfall, the reduction of hashab tree populations and tree locust outbreak. The reduction in hashab stock is caused by cutting, selling the tree as wood or charcoal and the expansion of areas of field crops at the expense of forestlands. As shown in Table 2 there was a great fluctuation in gum arabic production during the period 1990-2010.This fluctuation attributed to tree locust attack especially when hoppers appear before August and partially to rainfall and marketing policy (Ministry of Agriculture, Khartoum). Fig. 3 shows the relationship between the Acacia senegal tree, gum producers and the tree locust. The gum producers obtain gum from the tree by tapping. Tree locust defoliates the Acacia senegal tree as a means of obtaining food hence the quantity of gum produced will decrease (Taha, 2006).Tree locust can be considered as a beneficial insect i.e. in certain parts of the country the adult locusts are considered a rich source of food (Johnston, 1932).Trees manufacture carbohydrates after good rains and store these carbohydrates in the parenchyma cells of the cambium. When the trees are tapped, they produce gum, which is formed in cysts in the inner bark of the branches and not in the wood. The gum cysts are formed first in the parenchyma of the phloem. In case of wounding, the gum is transported to the wounded site via new channels formed by lysis of cells. When A. senegal is defoliated prematurely by tree locusts, trees use their stored carbohydrates to generate a new generation of tears instead of producing gum when tapped. (Deans et al., 1999).

Figure 2. Gum Arabic production distribution in Sudan Source: Taha, 2006

Table 2. Gum hashab production by regions in Sudan, 1990-2010 (quantities in metrics tons). Year Regions Kordofan Darfur Eastern Central Total Sudan 1990 13445 2241 1568 5154 22408 1991 4062 1032 1362 5330 11756 1992 5715 330 344 1160 7439 1993 3697 2415 1096 3970 11410 1994 5396 2913 1544 10959 22178 1995 27094 11196 5955 1319 45564 1996 10275 2484 2100 2897 18388 1997 9000 2200 2175 2201 15576 1998 10300 3200 2100 507 16107 1999 10000 2837 1676 506 15019 2000 5191 2708 1339 10754 21358 2001 10416 2626 2241 3038 18953 2002 5571 3088 1719 11134 22878 2003 31164 2004 27273 2005 29213 2006 20608 2007 30875 2008 32216 2009 47854 2010 51813 Source: Custom Duty Corporation (2009), Khartoum, Sudan.

Acacia senegal tree

Defoliation Food Gum Tapping

Gum reduction

Control Tree locust Gum arabic producer

Figure 3. Relationship between gum producer, Acacia senegal tree and tree locust (Source: Taha, 2006)

2.1.4.1 Host plants and feeding habits of A.m.melanorhodon As shown in Table3, the host-plants of A. m. melanorhodon comprise various trees and shrubs. Generally tree locusts feed on leaves especially of Acacia senegal (Template 2), and Zizyphus spina-christi (Siddir), in addition to a number of cultivated fruit and shade trees, and others such as Grewia tenax (Template 3) and Balanites aegyptiaca (Template 4). Cotton is occasionally attacked, mainly by moving swarms of adults. Tree locusts feed by night, but a hungry locust will feed at any time. Hoppers seem to move upwards for feeding, with the result that the upper and younger portions of trees are defoliated in a very characteristic manner i.e. leaf defoliation starts regularly from bottom to top. Barking of young Acacia trees by hoppers has been observed on several occasions. Hoppers show a certain degree of preference when trees of Acacia chrenbergiana and Zizyphus spina-chrsti were growing together, only the latter was attacked, whereas in other localities the former was also eaten and the leaves were usually attacked from the edge (Cirad, 2006). Young individuals often puncture holes and eat round within the circumference (Johnston, 1932). There are many plants species recorded as hosts of A. m. melanorhodon (Popov and Ratcliffe, 1968). Table3 shows the list of host plants of A. m. melanorhodon as recorded by Chevalier (1932), Kaatz (1963), Tigani (1965) and Popov (unpublished observations): cited by Haroon, (2008). In general, there are enough trees for roosting, but when trees are scarce or absent, tree locusts may spread out on shrubs and crops (Popov & Ratcliffe, 1968). Leaves of Acacia species have been found to contain homoarginine, pipecolic acid and 4-hydroxy-pipecolic acid; these amino acids were more concentrated in young leaves than in old leaves especially Acacia senegal leaves as cited by Evans and Bell, (1979). The latter reported that Anacridium melanorhodon preferred feeding on palatable media (young leaves) where the concentration of these amino acids was greater than in old leaves.

Table 3. Host plants of A. m. melanorhodon and the degree of preference (or damage).

(2) (3) (4) Wild plants Arabic or English name x(1)

Acacia spp Shukiat X Grewia tenax Geddeim X M Prosopis chilensis Mesquite X Sclerocarya birrea Humeid X Acacia flava Seyal M A. mellifera Kitir H H M A. nubica (orfota) Laot H H H A.senegal Hashab H H H A. tortilis Seyal M M AIbizia lebbeck Labukh X A. seriocephala Arad X Balanites aegyptiaca Higeelig (fruit laloub) X M M H Bauhinia rufescence Kul kul X S Boscia senegalensis Mukheit X X S Capparis deciduas Thundub H Commiphora Africana Gafal X M C. quadricincta Salualua S Cocculus sp. Kalia X Cordia gharaf Gimbeil Elzaraf X Dalbergia melanoxylon Babanus X Ficus thinningii Gommeiz S Ficus spp Gommeiz X X

Macrua crassifolia Surh X

Pithecellobium duluce Tumr hindi X Leptadenia lancifolia Sha lub X Ziziphus lotus Nabag X Ziziphus spina –christi Sider, Nabag H H Ziziphus sp Nabag karow H H Butyrospermum parkii Lo Lo, Shea Butter X

Cultivated Plants

Cotton Gotton X Citrus Himdiat X Guava Guava X Mango Munga X X Melons Butiekh X Sorghum Zora X X (1) = Chevalier 1932, (2) = Kaatz 1963, (3) = Tigani 1965 and (4) = Popov (unpublished observations). The degree of preference (or damage) is indicated as: s= slight; m=moderate; h= high; x= unspecified.

Template 2. A.m.melanorhodon 4th instar nymph on Acacia senegal shoot (photo by Luong, CIRAD).

Template 3. A.m. melanorhodon 6th instar nymph on Grewia tenax shoot (photo by Luong, CIRAD).

Template 4. A.m.melanorhodon adult on Balanites aegyptiaca shoot (photo by Luong, CIRAD).

2.1.5 Ecology According to Johnston (1932) and Cirad (2006) the ecological requirements of the different Anacridium species vary according to their respective eco- climatic zones and ranges. Generally, A. melanorhodon is an arboreal locust that roosts on Acacia spp., Balanites aegyptiaca thickets, which grow in low grasslands in the Sahel and Zizyphus spina christi. This acridian is sometimes found in mesoxerophilous fields. It may be in gregarious or solitarious form; gregarious behavior is more noted in imagoes. In the dry season, imagoes in the resting maturity stage fly where the dominant wind will take them until the first rain. Swarms are often small, but they sometimes reach a few kilometers in size. In Chad and Sudan, during the day, they aggregate in large numbers and roost on trees. At sunset, they fly from one woody zone to another most of the night so they have been named “sari al leil” or the nocturnal “vagabond” in Arabic. During the day, the insects usually rest on the branches. They are very wary and when approached they move round the stem away from the observer. Sometimes they drop down into the undergrowth for concealment. They do not fly as far as other large migratory locust species do. They can migrate tens to hundreds of kilometers. The swarms often land on fruit orchards or fields of millet, cotton and other crops. During the oviposition period in the rainy season, a swarm breaks up, thus separating the imagoes, first instar hoppers live on the ground or roost on low branches in loose groups, their diet is flexible and miscellaneous eating different forbs (non- graminaceous diet). They are geophilic negatively geotropic and strongly positively phototropic. Their arboreal behavior is more obvious from the third instars. They climb on Acacia sp and Balanites sp trees and eat their leaves. When hoppers populations are dense, with few perches, they swarm but do not move in cohesive groups.

2.1.6 Biology The tree locust, A.m. melanorhodon, is univoltine with a dry season imaginal diapause. The females lay eggs in June-July, hoppers develop in August and September and imagoes emerge at the end of the rainy season and the onset of the dry season (November).The Sudanese tree locust, Anacridium wernerellum, has one generation per year and one imaginal diapause in the dry season. The tree locust, Anacridium arabafrum is univoltine, the life cycle of this species is similar to that of A.m. melanorhodon, but little is known about Anacridium moestum. Generally sexual dimorphism is more visible in solitarious forms than in gregarious forms (Pastre et al., 1988 and Cirad, 2006). According to Luong and Popov (1997) there are five to six instars for male of A. m melanorhodon and six or seven for female (Fig. 4). Elbashier (1994) recorded six instars for both male and female, while only five nymphal instars were recorded by Johnston (1932) and Popov & Ratcliffe (1968) under laboratory conditions. Copulation takes place after the first rains and copulating pairs stay on trees until eggs lying begins (June-July). Eggs are laid in damp soil during the rainy season 10-20 days after copulation. A female lays 1-3 egg pods containing 150-200 eggs. The egg-pod is surrounded by a moderately firm cemented coat of sand and topped by a foam plug which measures about 75 mm. in length. Each egg is approximately 6mm. long, somewhat curved. The incubation period takes two weeks but no evidence of diapause has been found. The eggs hatch to give hoppers, the shed skins of the earliest or intermediate molt are cast on emergence from the egg, The development of the hopper stages takes 48-69 days, the first fledglings appear in September-October (Johnston, 1932, Meinzingen, 1993 and Luong and Popov, 1997). Imagoes remain in a resting maturity stage until the first rain of the following year in May- June (Johnston, 1932 and Cirad, 2006).

Fledgling

Figure 4. Life cycle of the tree locust, A. melanorhodon melanorhodon http://www.daff.gov.au/animal-plant-health/locusts/about/life-cycle (viewed on 24.10. 2008)

2.1.7 Natural enemies Natural enemies cannot prevent locust outbreaks but can provide a useful constraint on the rate of population increase and may hasten the collapse of plagues (Cirad, 2006). A large number of vertebrate and invertebrate predators prey on the hoppers and adults, also the eggs if being exposed through traditional methods of land cultivation. Gregarious locust populations are particularly attractive concentrations of food; both swarms and hopper bands are often followed by predators. Sarcophagid flies (hunting wasps) attack both larvae and adults and place their larvae on the host, which enter through an intersegmental membrane. One of the commonest of these is a Sphecid (Sphex aegyptiacus) which attacks the adults. It may be often seen carrying its prey along the ground by riding on the pronotum of the locust and propelling it by the use of the long hind legs. In some instances dipterous insects lurk around presumably waiting to lay their eggs on the victim (Johnston, 1932). Scorpions and insects, such as ground beetles have been reported as predators but there is no evidence that they cause significant mortality. The large spiders that inhabit trees in autumn catch a certain number of larvae (Johnston, 1932 and Lewis and Eve, 1965). Birds are probably the chief predator of both hoppers and adults, such as kites and other species of falconidae, hornbills, the buff-backed egret (Ardea ibis) and the abdim bey stork (Ciconia abdimii). The latter two frequently hunt together in companies and destroy large numbers of tree locust population. Moreover, they fallow the swarms, harrying them constantly. Tree locust from their habit of living in trees armed with spines are probably to some extent protected from enemies, both hoppers and adults tend to retreat to the inner and thicker parts when disturbed (Cirad, 2006). A wide range of pathogens has been isolated from Acridoidea, but there is little literature indicating their importance as mortality factors (Popov & Ratcliffe, 1968). There are few records of virus infections from Acridoidea (Johnston, 1932).

2.1.8 Control 2.1.8.1Chemical control In Sudan tree locust swarms are regularly controlled in the dry and rainy seasons by pesticides. Hoppers, when in low densities, are difficult to locate and control. Aerial treatments are unsuitable when this species overrun small areas. Poisoned baits are inefficient, thus treatments of roosting swarms should be carried out with motorized handheld sprayers during the warm period of the day. Good results were obtained with organophosphates such as fenitrothion, but many imagoes escaped during the spray process (Bashir, 1997). The effect of fipronil has been assessed in Kenya on caged imagoes, and on swarms roosting on acacias. In cages, all insects died within 24 hrs post-treatment with a dose of 2 g a.i /ha. Similar results were obtained with fenitrothion at a dose of 400 g a.i /ha. Despite difficulties in spraying swarms roosting on trees, treatments with fenitrothion (400g a.i /ha) and fipronil (2g a.i /ha) were carried out. Excellent results were obtained 24 hrs post- treatments (Cirad, 2006). According to Bashir (1997), chemical insecticides are annually used to control tree locust infestation on acacia trees, cultivated crops, fruit trees, other trees and shrubs either by dusting or aerial and/or ground spraying. The main insecticides used are: Sevin 85%, Propoxur 2%, 75%, Bendiocarb 1%, Malathion 57%, 96%, Diazinon 60%, 90%, fenitrothion 50%, 97%, Dursban 45%, Dimethoate 32%, Lindane 7%, Endosulphan 50%, Torbidan 20%, Decis 50% and Kafil super. Chemical insecticides are relatively fast and reliable, but most of them are relatively hazardous to operators, natural enemies and the environment (Dobson, 2000). Butrous (1995) and Bashir (1997) mentioned that, to avoid contamination of gum arabic with insecticides, the chemical control operations against tree locusts on A. senegal and A. seyal should be suspended for sometime until gum collection is completed.

2.1.8.2 Mycopesticides Control of locusts has traditionally relied on synthetic insecticides, and this is unlikely to change soon. However, a growing awareness of the environmental issues associated with acridid control as well as the high cost of emergency control is initiating the demand for biological control. In particular, preventive, integrated control strategies with early interventions which will reduce the financial costs and environmental hazard associated with large-scale plague treatments. The recent development of effective oil formulations of Metarhizium anisopliae spores in Africa, Australia, and Brazil opens new possibilities for environmentally safe control operations. Metarhizium bio-pesticide kills 70% and 90% of treated locusts within 14-20 days, with no measurable impact on non-target organisms. An integrated pest management strategy, with emphasis on the use of Metarhizium, that incorporates rational use of chemical pesticides with biological options such as the microsporidian Nosema locustae and the hymenopterans egg parasitoids Scelio spp., has become a realistic option (Lomer et al., 2001). Spores of the fungus Metarhizium flavoviride were applied against A. m. melanorhodon in the field for the first time. The treated areas covered 55 ha near Tendelti, the White Nile State at a dose of 100g of spores/ha. The application led to a reduction in population density of 66 - 76% after 18 days. Correction of the data with the Henderson/Tilton formula gave an overall control efficacy of 68 % (Kooyman, and Abdullah, 1998).

2.1.8.3 Botanicals Some botanicals (substances extracted from plants) can be alternatives for the neurotropic insecticides. The neem tree, Azadirachta indica (A. Juss) and Melia volkensii, both are rich sources of insect growth inhibitors and widely distributed in East Africa (Krall & Wilps, 1994). Many parts of neem tree (leaves, bark, seeds) are used for plant protection purposes in many parts of

22 the tropical world (Rembold, 1994). According to Morgan, (1987): Cited by Siddig, (1991) neem seed kernels contain a number of chemical compounds with azadirachtin and salannin as the most important. According to Parmar (1987), neem has repellent, antifeedant, insect growth regulator and insecticidal effects on various insect pest species including Coleoptera, Lepidoptera, Hemiptera and Orthoptera. No measurable effect on natural enemies was observed (Schmutterer, 1983). Some botanicals have growth regulatory effect on S. gregaria (Wilps and Nasseh, 1994) and on Locusta migratoria (Rembold and Muller 1986). They also have antifeedant effect as in the case of neem seed kernel, seed coat and leaves on S. gregaria (Singh, 1986). Insecticidal effect on S. gregaria was reported by Wilps and Nasseh (1994). Recently Abdalla (2004) reported insecticidal properties of the root extract of Mucuna pruriens and neem seed kernel (water and water/ethanol extracts) against the Desert Locust and the Migratory Locust.

2.1.8.4 Cultural control of grasshoppers and locust According to Duranton et al., (1979) the agricultural practices have complex effects on the pest status of Acrididae. Traditional methods of land cultivation do not destroy the eggs of certain grasshopper species, such as Oedaleus senegalensis. Mechanization may create substrates which are more suitable for oviposition for a greater number of grasshopper species including the migratory Catantops haemorrhoidalis and Ornithacris turbida. More efficient weeding may make cultivated land unsuitable for species, which feed on wild plants, such as Chrotogonus senegalensis, but may also concentrate the feeding of other species on crop plants. Changes from mixed farming to monocultures reduce the number of pest species, but favour the rapid proliferation of one or two species. Crop rotation is usually beneficial but needs to take into account acridid population dynamics. The introduction of

23 new crop varieties and species may cause new pest problems, as for example Kraussaria angulifera feeding on soya bean. Irrigation stabilizes the soil humidity to the benefit of certain species, such as Oedaleus nigeriensis and Zonocerus variegatus, and silting changes the nature of the soil in favour of species, which prefer compact clay soils, such as Aiolopus simulator. Areas of irrigated land can form oases enabling resident species, such as Acrotylus patruelis, to go through an extra generation, or they may attract acridids, which are nomadic in the dry season, such as Anacridium wernerellum and A. melanorhodon. Crop growing on land after flood retreat, brings man into direct competition with Locusta migratoria migratorioides, as has happened in lake Chad basin.

2.2 Remote sensing Remote sensing is the science (and to some extent, art) of acquiring information about the earth surface without actually being in contact with it. This is done by sensing and recording reflected or emitted energy, processing, analyzing, and applying that information. (Werle, 1988, 1992 and Wewester et al., 1993). This is exemplified by the use of imaging systems where the following seven elements are involved: Energy source or illumination - the first requirement for remote sensing is to have an energy source, which illuminates or provides electromagnetic energy to the target of interest. . Radiation and the atmosphere - as the energy travels from its source to the target, it will come in contact and interact with the atmosphere it passes through. Interaction with the target- once the energy makes its way to the target through the atmosphere; it interacts with the target depending on the properties of both the target and the radiation.

Recording of energy by the sensor- a sensor collects and records reflected energy. Transmission, reception, and processing- the energy recorded by the sensor has to be transmitted, often in electronic form, to a receiving and processing station where the data are processed into an image. Interpretation and analysis- the processed image is interpreted, visually and/or digitally or electronically, to extract information about the target which was illuminated. Application- the final element of the remote sensing process is achieved when the information about the target being extracted from the imagery is applied to assist in solving a particular problem. These seven elements comprise the remote sensing process from beginning to end (Barton and Leonov, 1997 and CCRS, 1999).

2.2.1 Why remote sensing? Remote sensing offers an efficient, powerful and reliable means of collecting the information required about acreage, crop type, structure and health due to its good characteristics such as large coverage, good spatial resolution, accessibility to harsh areas. However disadvantages like failing to distinguish some types of land use and lack of the horizontal perspective are also expected (Jain, 1989 and Simpson and Coy, 1999).

2.2.2 Applications of remote sensing in Sudan In 1972, the Sudan remote sensing technology was applied in many studies such as land degradation, land use/land cover classifications, mapping, planning and development of natural resources (FAO). Lampery (1975) investigated vegetation change and concluded that the desert was creeping southwards at the rate of 5-6 km per year. However Hellden (1978) proved that there was no systematic desert encroachment and criticized the findings

25 of Lampery (1975) as he misinterpreted the application of the vegetation map compiled by Harrison and Jackson (1958), which depended mainly on the 100mm rainfall isohyets. Doka (1980) used remote sensing for monitoring soil resources and areas affected by desertification in central Sudan. Olsson (1985, cited by Dafalla, 2006) studied availability of fuel wood in North Kordofan using remote sensing. Hielkema et al., (1986) used NOAA-AVHRR (Advanced Very High Resolution Radiometer) data to monitor vegetation and rainfall relationship in Savanna zone. He concluded that NDVI values can be used to monitor effective rainfall in the Savanna zone of the Sudan. In 1987 remotely sensed data was used in Resource Assessment and Development (SRAAD) project with the aim of forestry inventory and rehabilitation and vegetation map production. Yagoub et al., (1994, cited by Dafalla, 2006) assessed biomass and soil potential in northern Kordofan using the NDVI indices. They concluded that the land degradation and ecological imbalance in this region was associated with the combined adverse effects of rainfall and mismanagement of land. In 1995 the Africover Project was started using remote sensing and geographic information system technologies for the monitoring and promotion of sustainable use of natural resources. Ali (1996, cited by Dafalla, 2006) assessed and mapped desertification in the western part of the Sudan using NDVI images created from AVHRR-NOAA sensor and also applied GIS. He stated that remotely sensed data provided good indicators of vegetation degradation throughout the period 1982-1994 in the form of the image maps. Kassa (1999) used NDVI based on NOAA-AVHRR and rainfall data to monitor drought risk for the Sudan and to produce drought risk map based on NDVI. This study concluded that NDVI-based map enables decision-makers to have a basic overview of areas at risk of drought in the Sudan. Eklundh and Sjöström (2002, cited by Dafalla, 2006) analyzed vegetation changes in the Sahel using imagery of Landsat and NOAA. They showed that the NDVI values during the period 1982-2002 were increased,

26 and areas of positive change showed a transition from barren or sparse vegetation to a dense vegetation cover. Elmqvist (2004) studied land use change in northern Kordofan for the period 1969-2002 by using recent high resolution earth observation satellite data such as Corona and IKONOS. Hinderson (2004) analyzed environmental changes in semi-arid areas in Kordofan during 1982-1999 using NOAA-AVHRR and Landsat imagery. The research analyzed the observed NDVI changes. Khiry (2003) and Suliman (2003) used remote sensing methods to investigate land use/land cover changes in Khartoum State, and Darfour State, respectively. They stated that vegetation cover change could be significantly detected using remote sensing analysis methods. El Haja (2005) used remote sensing to study sand encroachment in North Kordofan State and concluded that remote sensing was efficient in determining areas affected by sand encroachment. Dafalla and Csaplovics (2005, cited by Dafalla,2006) assessed the dominant land use/land cover types for the North Kordofan State by means of high resolution Landsat ETM+ imagery. The study revealed that remote sensing methods could be used with a satisfactory level of significance in land use/land cover classification. Herrmann et al., (2005 cited by Dafalla, 2006) explored the relationship between rainfall and vegetation dynamics in the Sahel region using coarse resolution satellite data. They confirmed the general positive trend of NDVI and rainfall over the period 1982-2003. In addition they concluded that rainfall emerges as the dominant causative factor in the dynamics of vegetation greenness in the Sahel region.

2.2.3 Digital image processing Digital image processing is done entirely by computer to facilitate better visual interpretation. Digital image processing can be categorized into the following four categories: Image preprocessing, image enhancement, image transformation and image classification and analysis (Roger, 2005).

Preprocessing functions involve those operations that are normally required prior to the main data analysis and extraction of information, and are generally grouped as radiometric or geometric corrections. Radiometric corrections include correcting the data for sensor irregularities and unwanted sensor or atmospheric noise, and converting the data so they accurately represent the reflected or emitted radiation measured by the sensor. Geometric corrections include correcting for geometric distortions due to sensor-earth geometry variations, and conversion of the data to real world coordinates (i.e. latitude and longitude) on the earth's surface, to represent a real world feature, the data must be referenced to the correct location on the earth’s surface, this is known as geo-referencing which is achieved through using a coordinate system (Roger, 2005). Image enhancement involves the improving of the appearance of the imagery to assist in visual interpretation and analysis. Examples of enhancement functions include contrast stretching to increase the tonal distinction between various features in a scene, and spatial filtering to enhance (or suppress) specific spatial patterns in an image (CCRS, 1999). Image transformations are operations similar in concept to those for image enhancement. However, unlike image enhancement operations which are normally applied only to a single channel of data at a time, image transformations usually involve combined processing of data from multiple spectral bands. Arithmetic operations (i.e. subtraction, addition, multiplication, division) are performed to combine and transform the original bands into "new" images which better display or highlight certain features in the scene (DeFries et al., 1999).

2.2.4 Images interpretation and analysis Interpretation and analysis of remote sensing imagery involves the identification and/or measurement of various targets in an image in order to extract useful information about them. Most interpretation and identification

28 of targets in remote sensing imagery is performed manually or visually, i.e. by a human interpreter. In many cases this is done using imagery displayed in a pictorial or photograph-type format, independent of what type of sensor was used to collect the data and how the data were collected. In this case the data is referred to as being in analogue format. Remote sensing images can also be represented in a computer as arrays of pixels, with each pixel corresponding to a digital number (DN), representing the brightness level of that pixel in the image. In this case, the data are in a digital format. Visual interpretation may also be performed by examining digital imagery displayed on a computer screen. Both analogue and digital imagery can be displayed as black and white (also called monochrome) images, or as colour images by combining different channels or bands representing different wavelengths (Jain, 1989).

2.2.5 Image classification Image classification is defined as the process of creating thematic maps from satellite imagery (process of sorting pixels into a finite number of individual classes, or categories of data based on their data file values. In general, digital classification techniques may be categorized by the training process into supervised or unsupervised classification (DeFries et al., 1999). Supervised classification procedures require considerable interaction with the analyst, who must guide the classification by identifying areas on the image which are known to belong to each category of interest. These areas are referred to as training sites. In supervised classification, any individual pixel is compared to each discrete cluster to select the one which it is closest in terms of spectral values. Generally, supervised classification methods have many advantages. Firstly, the analyst has control of a selected menu of informational categories adapted to a specific purpose and to geographic region. Secondly, supervised classification is associated with specific areas of known identity. Finally, serious classification errors are detectable by field

29 verification to determine whether they have been correctly classified. On the other hand, supervised classification has numerous disadvantages. The analyst imposes a classification structure upon the data based on predefined classes instead of finding natural classes in an image. Furthermore, the defined classes may not match the classes which may exist in the data. In supervised classification, training sites and classes are based primarily on the information categories and only secondarily on spectral properties. Another source of error is the selection of training data, since these samples of pixels may not be representative of conditions encountered throughout the image. Moreover, supervised classification is not able to recognize the specific or unique categories which are not represented in training data due to the small areas they occupy on the image or simply because they are not known to the analyst. There are two methods of supervised classification; soft classification and hard classification (Zadeh, 1965, Campbell, 1996, Bastin, 1997, and Hedge, 2003). In the case of crisp or hard classification, each pixel is assigned to only one class i.e. each pixel represents a homogeneous area on the ground and show as containing only one land cover. However, in fuzzy or soft classification, such as in spectral mixture analysis (SMA), a pixel is associated with many land cover classes i.e. there may be more than one class in mixed pixel (Markham and Townsend, 1981, Mather, 1987 and Lillesand et al., 2000). Unsupervised classification involves the process of automatically segmenting an image into spectral classes based on the natural groupings found within the data set. The objective is to group multi-band spectral response patterns into clusters which are statistically separable. The two most frequently used grouping algorithms are K-means and ISODATA cluster algorithms. These two statistical routines for grouping similar pixels together are iterative procedures (CCRS, 1999).

Advantages of unsupervised classification can be summarized into three key points. Firstly, no extensive prior knowledge of the region of interest is required. Secondly, the opportunity of human error is minimized. Finally, unique classes are recognized as distinct units in unsupervised classification. Since unsupervised classification identifies spectrally homogenous classes within the data, such classes do not necessarily correspond to the informational categories which are of interest to the analyst (Key et al., 1989, CCRS, 1999 and Foody and Atkinson, 2002). Performing an unsupervised classification is simpler than a supervised classification, because the signatures are automatically generated by the ISODATA algorithm (Jain, 1989 and CCRS, 1999).

2.3 Geographic Information System (GIS) GIS is a computer-based tool for mapping and analyzing data (collection of computer hardware, software, and geographic data for capturing, managing, analyzing, and displaying all forms of geographically referenced information). GIS technology integrates common database operations such as query and statistical analysis with the unique visualization and geographic analysis benefits offered by maps (Foody and Atkinson, 2002).

2.4 Global Positioning System (GPS) GPS is a satellite-based navigation system made up of a network of 24 satellites placed into orbit by the U.S. Department of Defense. GPS was originally intended for military applications, but in the 1980s, the USA government made the system available for civilian use. GPS works in any weather conditions, anywhere in the world, 24 hours a day. There are no subscription fees or setup charges to use GPS. GPS satellites circle the earth twice a day in a very precise orbit and transmit signal information to earth. GPS receivers take this information and use triangulation to calculate the

31 user's exact location. Essentially, the GPS receiver compares the time a signal was transmitted by a satellite with the time it was received. Now, with distance measurements from a few more satellites, the receiver can determine the user's position and display it on the unit's electronic map. The 24 satellites that make up the GPS space segment are orbiting the earth about 12,000 miles above the earth surface. These satellites are traveling at speeds of roughly 7,000 miles an hour (RSA, 2008).

2.5 Crop monitoring and damage assessment Insect infestation is among the most challenging environmental hazards that farmers face, successful mitigation depends on early detection and tracking. The use of hyper spectral data gives the required spatial overview of the land that can enhance early detection of plant defoliation, and allow the farmer to make timely decisions about managing the pests (Tucker and Sellers,1986,Campbell, 1987, Dougherty et al., 1987 and Miller et al., 2002). Healthy vegetation contains large quantities of chlorophyll. Hence reflectance in the blue and red parts of the spectrum is low since chlorophyll absorbs this energy. In contrast, reflectance in the green and near-infrared spectral regions is high (Weweter et al., 1993, CCRS, 1999, Song et al., 2001, and Wylie et al., 2002). Stressed or damaged crops experience a decrease in chlorophyll content and changes to the internal leaf structure. The reduction in chlorophyll results in low reflectance in the green region; internal leaf damage results in low reflectance in near-infrared. Examining the ratio of reflected infrared to red wavelengths is an excellent measure of vegetation health; this is the premise behind some vegetation indices, such as the “Normalized Difference Vegetation Index” (NDVI). The NDVI is a calculation, based on several spectral bands, of the photosynthetic output in a pixel in a satellite image. It measures the amount of green vegetation in an area. Healthy plants have a high NDVI value because of their high reflectance of infrared light, and

32 relatively low reflectance of red light (Figures 5, 6, 7 and 8). Phenology and vigor are the main factors affecting NDVI value. An excellent example is the difference between irrigated crops and non-irrigated land. In a real-color simulated image, the irrigated crops appear bright green while the dry rangelands are dark. In a false-color simulated image, where infrared reflectance is displayed in red, the healthy vegetation appears bright red, while the rangeland remains quite low in reflectance (Jensen, 1986, Lillesand et al., 1994, Russ and John, 1995 and CCRS, 1999). NDVI value is affected by many factors; Atmospheric effects, clouds, soil effects, anisotropic effects and spectral effects.

2.6 NDVI versus other indices The vegetation indices can be broadly divided into two basic categories: ratios and orthogonal indices. The ratio-based indices include the Ratio Vegetation Index (RVI) and the Normalized Difference Vegetation Index (NDVI). Orthogonal indices include Perpendicular Vegetation Index (PVI) and the Difference Vegetation Index (DVI). More recently a hybrid set of vegetation indices have emerged, such as Soil Adjusted Vegetation Index (SAVI) (Kaufman and Tanre, 1992).

2.6.1 Why NDVI ? NDVI is the simplest form of vegetation indices i.e. ratio between near- infrared and red reflectance .The healthy vegetation reflects very well in the near infrared part of the spectrum. Green leaves have a reflectance of 20 percent or less in the 0.5 to 0.7 micron range (green to red) and about 60 percent in the 0.7 to 1.3 micron range (near infra-red). The visible channel gives some degree of atmospheric correction. The value is then normalized to the range -1<=NDVI<=1 to partially account for differences in illumination and surface slope (Kaufman and Tanre, 1992).

اﻟﺘﻤﻴﻴﺰ اﻟﻄﻴﻔﻲ ﻟﻠﻐﻄﺎء اﻟﻨﺒﺎﺗﻲ Spectral resolution

Figure 5. Spectral resolution of vegetation (Source: CCRS, 1999)

Figure 6. Reflectance of vegetated land areas (source: CCRS, 1999)

اﻷرض ﻏﻴﺮاﻟﻤﻐﻄﺎﻩ ﺑﺎﻟﻨﺒﺎﺗﺎت

Concrete

Bare soil Gravel Shingles' Asphalt

Figure 7. Reflectance of non-vegetated land areas (source: CCRS, 1999)

Figure 8. NDVI calculation (illustration by Robert Simmon, 1999, from Wikiagro.com)

CHAPTER THREE

MATERIALS AND METHODS 3.1 Study area Laboratory work was conducted at the Gum Arabic Research Centre (GARC) at Elobeid. Field work was conducted during three successive seasons, 2007/2008, 2008/2009 and 2009/2010 at Acacia Project (Nawa and Elrahad locations), 37 Km south east Elobeid, Sheikan locality, North Kordofan State (Fig. 9). The total area of the project is 28000 feddans which was planted with gum arabic trees in 1997. The inter-row spacing is 3m and the intra-row spacing is 5m. Hashab trees were evenly grown, approximately about 280 hashab tree/ feddan (Acacia Project Reports, 2007). Hashab plantations were more or less stunted possibly due to soil compaction (Taha, 2006). The soil is hard crust; generally flat, non-cracking clay mixed with Aeolian sand and with low infiltration rates locally named Gardud (FAO, 1997). The climate is semi-arid with annual rainfall ranging from less than 200 mm in the north to about 350 mm in the south; the temperature is highest during July ranging 30 - 40° C (Abdulla, 2006).

3.1.1 Criteria for selection of study area: The following criteria were considered in the selection of the study area; North Kordofan State is considered as the leader state in gum arabic production, the accessibility of the area is good and the area represents a breeding area for tree locusts.

Figure 9. Location map of the study area Source: GAC 2000

3.2 Field Survey 3.2.1 Seasonal occurrence of adult and hopper tree locusts 3.2.1.1 Counting of the 1st and 2nd instar hoppers In each location (Elrahad and Nawa) four blocks were selected systematically and randomly, the area of each block (B) was 50 feddans (210 X 1000 meters), in each block, 10 estimation points (P) at intervals of 100m were determined systematically and randomly for counting hoppers (Fig.10). Counting on the ground was done weekly early in the morning at 8 O’clock by using a wooden frame covered with mosquito net mesh. The trapped hoppers were handled with a fork-like instrument.

3.2.1.2 Counting of the 3rd, 4th, 5th, and 6th instar hoppers on trees At the same quarters, blocks, and points (Fig. 10) visual counting of the 3rd, 4th, 5th, and 6th instar hoppers roosting on 40 estimation points (trees) was done weekly early in the morning(8 O’clock ) without much disturbance.

3.2.1.3 Counting of the adult tree locusts on trees The same method used for counting the 3rd, 4th, 5th, and 6th instar hoppers on estimation points (trees) was applied for counting the adult tree locusts. Visual counting was carried out early in the morning (8 O’clock) at weekly intervals. Counting for all tree locust stages was repeated during three successive seasons (2007/2008, 2008/2009 and 2009/2010) throughout the year and also the total number for each stage per tree was computed and tabulated for analysis. Rainfall and atmospheric RH% were recorded during each counting session. Relationship between all tree locust stages and rainfall and atmospheric RH% was studied.

1000m 1000m

B1 B2

210m 210 m

B3 B4

210 m 210 m

1000 m 1000m

Figure 10. Layout showing blocks (B) and estimation points (o) at Nawa and

Elrahad locations

3.2.2 Quantitative effect of natural and artificial defoliation on gum production Four blocks in each location were selected randomly. Each block was divided into 7 plots (experimental units). In each plot (treatment) there were 15 hashab trees. The treatments were arranged in a randomized complete block design (RCBD). the treatments were: control, light natural defoliation (5 of 15 trees were caged with tree locusts to be completely defoliated i.e. 33% naturally defoliated), moderate natural defoliation (10 of 15 trees were caged with tree locust for complete defoliation i.e. 66% naturally defoliated), high natural defoliation (all the 15 trees were caged with tree locusts and completely defoliated i.e. 100% naturally defoliated), light artificial defoliation (5 of 15 trees were completely defoliated with 1% diluted ethereal defoliant spray i.e. 33% artificially defoliated), moderate artificial defoliation (10 of 15 trees were completely defoliated with ethereal defoliant i.e. 66% artificially defoliated) and high artificial defoliation (all the 15 trees were completely defoliated with ethereal defoliant i.e. 100% artificially defoliated) (Templates 5 and 6). Both artificially and naturally defoliated trees were selected randomly. The control and the trees in other treatments which were not naturally or artificially defoliated were sprayed weekly using neem seed water solution extract 1% as recommended by Siddig (1991) as antifedant and repellent for tree locust and other pests. Each treatment was registered by GPS to allow for further integration with the spatial data in a geographic information system (GIS) and image classification system (Image 1and 2).

3.2.2.1 Collection and preparation of the neem seeds Fruits of neem tree (Azadirachta indica) were collected from Elobeid area, washed and left to dry. The seeds were then decorticated and the kernels were ground into a fine powder, stored in a glass jar tightly closed and kept at room temperature.

Template 5. Caged tree + tree locusts before defoliation

Template 6. Caged tree after defoliation

Image 1. Spot image showing the different blocks and treatments at Nawa location 2009

Image 2.Spot image showing the different blocks and treatments at Elrahad location2009

3.2.2.2 Preparation of neem seed aqueous extracts (Aq-extr) One kilogram of the prepared neem seed powder (N.S.P) was mixed with 40 litres of water as recommended by Siddig (1991) in a bucket and stirred vigorously using a wooden stick and left overnight. The mixture was stirred again before it was filtered using a piece of cloth. The filtrate was mixed with liquid soap as an emulsifier at a ratio of 4:1 (v/v), 1% gum arabic solution as a sticker and anti-oxidant and 1% molasses as uv-light protectant.

3.2.2.3 Tapping and gum collection Trees in all treatments were tapped during the period from mid-October to end of November using a tapping tool (sunki) made of steel with a long wooden handle (Template 7); the first picking of gum began 40 days after tapping followed by a series of subsequent pickings up to seven pickings at 14 days interval. Gum in each treatment was collected, mixed and weighed separately. The percentage of loss in gum produced is calculated in all levels of defoliation with respect to gum produced in the control.

3.3 Laboratory experiments 3.3.1 Preparation of gum samples Gum nodules in each treatment were dried at room temperature, and then cleaned by hand to ensure a quality free from sand, dust and other impurities, then ground using a mortar and pestle, sieved and kept in labeled containers for qualitative analysis.

3.3.2 Effect of natural and artificial defoliation on gum quality 3.3.2.1 Gum viscosity Gum viscosity is the resistance of gum liquid to shear forces and hence to flow. In each treatment, a sample of 25g of gum powder was taken and dissolved in100 ml of distilled water. The viscosity of gum solution was measured using a viscometer (model DV.II+ England).

Template 7. Tapping method using the Sunki tool

3.3.2.2 Specific optical rotation The specific optical rotation of gum is a measure (finger print) by which different gum types can be sorted and identified and it has negative values in the case of hashab gum (-22 to -35) and positive for other acacia gums (FAO, 1991). Solution of gum powder 1% (on dry weight basis) in each treatment was prepared and measured at room temperature using an optical activity polarimeter (type AA-10 automatic polarimeter, England). The solution was passed through filter paper before carrying out measurements; triplicate readings were taken and averaged.

3.3.3 Assessment of the daily food intake and the effect of hashab leaf age on the adult and hopper development Adult tree locusts were reared in cages (2 X 2 X 2m) at the Gum Arabic Research Centre yard (Template 8). Six iron cages (60 X 60 X 60cm), covered with fine wire mesh, were prepared with sandy soil at the bottom and arranged in two sets. In each cage 20 1st instar hoppers were introduced. Old and young hashab leaves weighing about 50g were daily added to each cage in set1 and set2, respectively (Template 9). That is old leaves +hoppers in set1 and young leaves +hoppers in set2 to check development and daily food-intake. 50g of each old and young leaves were put in a separate petridish in the lab at room temperature for 24 hours, and then weighed again. The difference in weight represents the daily loss on dry-basis. The actual daily amount of fresh hashab leaves consumed will be equal to the daily amount of fresh hashab leaves consumed minus the daily loss on dry-basis. The following parameters in each cage were checked and recorded daily; average weight of hoppers, amount of fresh hashab leaves added, amount of fresh hashab leaves consumed and mortality of hoppers.

Template 8. Rearing and breeding of tree locust

Template 9. Cage+ hoppers + leaves of hashab

3.3.4 Soil analysis Auger-observation soil samples were taken at different depths (0, 50 and 100cm deep) in each location. Then the soil samples in each location were mixed to make a compound sample for measuring moisture, texture, PH and nitrogen content.

3.3.4.1 Soil moisture determination An empty can with its cover were weighed using a sensitive balance (A). The can with its cover plus the soil sample (100g) in each location were weighed (B). The covered can + the soil sample were placed inside an oven at 150oc for 24 hrs then they were weighed again(C). Moisture % = (B-C ⁄ B-A) X100).

3.3.4.2 Soil texture 40 g were taken from the compound soil sample in each location in a separate beaker, 50 ml of calgon was added to disperse the soil particles, and

10 ml of H2O2 (hydrogen peroxide) were added to digest the organic matter,

15 ml diluted HCl were added to neutralize CaCo3. The solution was stirred very well and the contents were made up to 1 litre with distilled water in a measuring cylinder. The solution was stirred very well, the first reading (silt, clay) was taken after 40 seconds using a hydrometer type ASTM 412- England. The second reading (clay) was taken after 2 hours.

3.3.4.3 PH determination. 25g dry soil from the compound soil sample in each location was taken in a separate measuring cylinder. Distilled water was added to form a paste which had the following properties; seemed to glitter and no water remains on the surface. The paste was left for 24 hours. An electrode was inserted inside the paste for readings using the PH meter; model 3320 Jenway, England.

3.3.4.4 Nitrogen determination

For soil nitrogen (N2) determination, the Kjeldahl method was used. The standard Kjeldahl method involves a two-step process. 50g from the compound soil sample in each location was taken and digested with sulfuric acid to convert N2 compounds to NH4+. The converted NH4+ along with any

NH4+ that was originally present was further converted to NH3 in an alkali distillation process. The NH3 liberated in this process was then quantified to determine the total soil nitrogen.

3.4 Remotely sensed data Remotely sensed data used for this study was collected from two sensors; Landsat satellite enhanced thematic mapper with sensors; ETM+. The two Landsat ETM+ scenes covering the study area were acquired in August 2008 during the rainy season in the study area when hashab trees were very green and in October 2008 at the end of the rainy season. The scenes were acquired in seven bands covering the visible, near and middle infrared regions of the electromagnetic spectrum. The spatial resolution was 15m (after merging panchromatic with multispectral bands), band widths (band1; 0.45-0.52 (blue), band2; 0.52.0.60 (green), band3; 0.63-0.69 (red), band 4; 0.76-0.90 (near- infrared), band5; 1.55-1.75 (mid-infrared), band 6; 10-4-12.5 (thermal), band7; 2.08- 2.35 (mid-infrared), Path / row No was 174/51and Dynamic range was 8 bit. The two scenes of Spot5 with the sensor XS covering the study area were acquired in August 2009 in the rainy season in the study area when hashab trees were very green and in November 2008 at the end of the rainy season with spatial ground resolution 2.5m. The sensor provided data in green (0.50–0.59 µm), red (0.61–0.68 µm), near-infrared (NIR; 0.78–0.89 µm), and short-wave infrared (SWIR; 1.58–1.75 µm) wavelengths at 2.5m resolution. Dynamic range was 8 bit.

3.4.1 Ancillary data Ancillary data about the study area were collected from records in agricultural statistics regarding tree types, annual rainfall and relative humidity % from 2007-2009.

3.4.2 Software Remote sensing; ERDAS IMAGINE 8.5 software

GIS; ESRI Arc GIS 9.3

Arc GIS extension for GPS Others;

Microsoft Office 2007

3.4.3 Hardware Computer (PCs), laptops, scanner, GPS and digital cameras. Three methods of image interpretation and analysis were applied to extract meaningful information from remote sensing imagery data:

3.4.4 Visual interpretation Prior to visual interpretation, initial processing on the raw data was carried out to compensate for any distortion due to the characteristic of the imaging system and imaging condition. These initial processing to facilitate image interpretation and analysis which include radiometric and geometric correction, image enhancement and image georeferecing by using ground control points (GCPs) to register the image to precise map (CCRS, 1999). A false colour images composite were created, then a subset image from each false colour image composite was visually interpreted. Identification and recognition of the various features (defoliated and non-defoliated hashab trees) was done on the basis of visual elements, tone, shape, size, pattern, texture, shadow and association (Lillesand and Kiefer, 1994).

3.4.5 Normalized Difference Vegetation Index (NDVI) analysis The study area was investigated with the Normalized Difference Vegetation Index (NDVI) images created from the Landsat (2008) and Spot (2009) imagery i.e. through using multispectral ability to create new channels in an image to display the NDVI for an image with the help of ERDAS IMAGINE 8.5 software (Kaufman and Tanre, 1992). Normalized Difference Vegetation Index (NDVI) composited products were nominally corrected for atmospheric effects to provide surface reflectance. According to Larry, (1997) and Fernandes et al., (2003) a vegetation index that is sensitive to leaf area index was examined; NIR (0.76–0. 9 µm) – Red (0.63–0.69 µm)/NIR+ Red I.e. band4 –band3 / band 4 + band 3 for Landsat and NIR (0.78–0.89 µm) – Red (0.61–0.68 µm)/NIR+ Red i.e. band3 –band2 / band 3 + band 2 for Spot5. Non- defoliated and defoliated hashab trees were identified in the produced NDVI images. The created NDVI images were classified with the help of ARC map 9.3 to produce coloured images and compared with the supervised classes.

3.4.6 Supervised classification Supervised classification was done for sorting pixels into a finite number of individual classes, or categories of data based on their data file values (spectral reflectance), where it is a decision phase. In the image, homogeneous representative samples of the different levels of tree locust defoliation of interest were identified. These samples (plots) are referred to as training areas each plot was registered by using GPS technology to allow for further integration with the spatial data in a geographic information system (GIS) The numerical information in all spectral bands for the pixels comprising these areas was used to "train" the computer to recognize spectrally similar areas for each class. A special program or algorithm was used to determine the numerical "signatures" for each training class. Once the signatures for each

51 class had been determined, each pixel in the image is compared to these signatures and labeled as the class it most closely "resembles" digitally (Bastin, 1997 and DeFries et al., 1999). Thus, in a supervised classification the information classes were identified first (healthy or non-defoliated trees, lightly defoliated trees, moderately defoliated trees, highly defoliated trees and areas of tree locust swarm) then used to determine the spectral classes which represent them. The objective is to match the spectral classes in the data of interest to the information classes (Erdas, 2003). Classification accuracy assessment was carried out by visual interpretation of images maintained by collecting reference samples by stratified random sampling technique. Depending on visual interpretation on screen, the different classes were assigned to these reference samples. User accuracy, producer accuracy, overall accuracy were checked (Erdas, 2003).

3.4.7 Data analysis The satellite images were processed and analysed using ERDAS imagine 8.5 and Arc GIS. Also Sigma-Stat and Excel were used for analysing the data. Visual image interpretation, NDVI was computed and supervised classification was done. Both remotely sensed data and field work data were integrated with GIS techniques to provide information required to fulfil the objectives of the study.

3.4.8 Layout and mapping The final findings of the study of both field survey and remote sensing methods were integrated with Arc GIS methods to produce maps and graphs.

CHAPTER FOUR RESULTS

4.1 Field survey 4.1.1 Seasonal occurrence of tree locust adults and hoppers in relation to rainfall and relative humidity At Nawa and Elrahad locations during three seasons (2007/2008, 2008/2009 and 2009/2010), there was periodical appearance of adults throughout the year except in February, March and April. Hoppers appeared in July, August, September and October (Figure 11 and Appendix 1). Figure 11 shows that the mean number of adults/tree decreased with the increase of RH%, and then increased with the decrease of RH%, on the other hand the mean number of hoppers/tree increased with the increase of RH%. Figure 12 also shows that the mean number of adults/tree decreased with the increase of rainfall, then increased with the decrease of rainfall, while the mean number of hoppers/tree increased with the increase of rainfall.

4.1.2 Quantitative effect of natural and artificial defoliation on gum production At Nawa and Elrahad locations, seasons 2007/2008, 2008/2009 and 2009/2010 the gum production data is illustrated in Table 4, Fig.13 and Appendices 2, 3, and 4. Defoliation reduced gum production and the reduction was statistically significant (P≤0.001) between means of all treatments except between high natural defoliation and high artificial defoliation.

Figure 11. Seasonal occurrence and mean number of adults and hoppers/tree at Nawa and Elrahad locations in relation to relative humidity (seasons 2007/2008, 2008/2009 and 2009/2010).

Figure 12. Seasonal occurrence and mean number of adults and hoppers/tree at Nawa and Elrahad locations in relation to rainfall (seasons (2007/2008, 2008/2009 and 2009/2010).

Table 4. Mean weight (g) of gum produced / hashab tree at different blocks and treatments at Nawa and Elrahad locations (seasons 2007/2008, 2008/2009 and 2009/2010) Block gum production(g) No. Control LND MND HND LAD MAD HAD

1 4503.41088.8 559.55226.15 1438 323.9 323.9

714.6 2 4572.3 1118.6 580.8 233.3 1527 784.2 300 3 4511 1098 595.2 231.5 1500.5769.9 299.3 4 4606.41116.6 594.6 230.6 1526.5774.9 317.9 Total 18193 4422 2330.2 921.6 5992 3043.6 1241.1 a b c d e f d Mean 303. 2 73. 7 38.8 15.4 100. 0 50. 7 20. 7 ± SE ± 79.3 ± 30.1 ±16.8 ± 5.3 ± 31.1 ± 37.5 ±17.7 Mean values of gum production in each column with same superscript letter showed no significant difference at p = 0.001 as separated by Tukey test.

Figure 13. Mean weight (g) of gum produced / hashab tree at different treatments at Nawa and Elrahad locations (seasons 2007/2008, 2008/2009 and2009/2010)

The percentage of losses in gum production calculated at all levels of defoliation with respect to gum produced in the control were; 75.7%, 87.2%, 94.9%, 67%, 83.7% and 93.2% for light natural defoliation, moderate natural defoliation, high natural defoliation, light artificial defoliation, moderate artificial defoliation and high artificial defoliation, respectively. Results showed that there was a negative correlation between gum production and the different levels of natural and artificial defoliation in all seasons. The correlation was very high, ranging between -0.994 -0.995 for all levels of natural and artificial defoliation (Appendix 5). It can be concluded that both natural and artificial defoliation severely reduced gum production/tree in all seasons, but the reduction was higher in natural defoliation.

4.2 Laboratory experiments 4.2.1 Effect of natural and artificial defoliation on gum quality 4.2.1.1 Gum viscosity (in centistokes) Table 5, Fig.14 and Appendices 6, 7 and 8 show that the mean values of gum viscosity at different blocks and treatments for all seasons decreased at all levels of defoliation compared to the control. In all levels of defoliation, gum viscosity was higher in artificial defoliation than in natural defoliation. Defoliation reduced gum viscosity and the reduction was statistically significant (P≤0.001) between all means except between light natural defoliation and light artificial defoliation. The correlation between all levels of defoliation and gum viscosity is shown in Appendix 9. There was a negative correlation between all levels of defoliation and gum viscosity which arranged from -0.94 to -0.99 for natural and artificial levels of defoliation, respectively.

Table 5. Gum viscosity (%) at different blocks and treatments at Nawa and Elrahad locations (seasons 2007/2008, 2008/2009 and 2009/2010) Block Gum viscosity % No. Control LND MND HND LAD MAD HAD

1 35.5 34.5 32.1 30.9 34.6 32.7 31.5 2 34.6 34.5 32.1 30.9 34.6 32.7 31.5 3 34.0 34.5 32.1 30.9 34.6 33.4 31.6 4 34.0 34.5 32.0 30.9 34.6 32.6 31.4 Total 138.1 138 128.3 123.6 138.4 131.4 126 Mean 34.5a 34.5b 32.1c 30.9d 34.6b 32.9f 31.5g ± SE ± 0.09 ± 0.03 ± 0.03 ± 0.02 ± 0.2 ± 0.1 ± 0.1

Mean ± SE Mean values of gum viscosity in each column with same superscript letter showed no significant difference at p = 0.001 as separated by Tukey test.

Figure 14. Mean of gum viscosity (%) at different treatments at Nawa and Elrahad locations (seasons 2007/2008, 2008/2009 and 2009/2010)

4.2. 1.2 Gum specific optical rotation The gum specific optical rotation data are illustrated in Fig.15 and Appendices 10, 11and 12. Results show that the means of gum specific optical rotation at different blocks and treatments were negative values. Defoliation reduced gum specific optical rotation but there was no significant difference between means in all treatments.

4.2.2 The daily food intake and the effect of hashab leaf age on hopper development For each of the six different nymphal instars and fledglings, three parameters were measured (daily food intake, weight and developmental period of hoppers) during three successive seasons (Table 6). The daily food consumed by the different stages of tree locust fed on old and young leaves were approximately equivalent to their body weight which ranged between 23mg- 2000mg for the first nymphal instar and adult, respectively. The total development period varied; for those nymphal instars which were fed on old leaves the duration was 68 days, but for those which were fed on young leaves the duration was 59 days, the sixth nymphal instar had longest duration while the first nymphal instar had shortest duration when fed on old or young leaves.

4.2.3 Soil analysis 4.2.3.1 Soil moisture determination Results show that the soil moisture decreased slightly with increase in depth (Appendices 13 and 14).

Figure 15. Mean of gum specific optical rotation at different treatments at Nawa and Elrahad locations (seasons 2007/2008, 2008/2009 and 2009/2010)

Table 6.The effect of leaves age on hoppers development (seasons 2007/2008, 2008/2009 and 2009/2010) Old leaves Hopper stages Average weight Developmental Daily food of hopers(mg) period of consumed hoppers by hoppers (day) (mg) 1st instar hopper 24-26 6 24-25.3 2ndinstar hopper 26.3-41.6 7 25.6-40 3rdinstar hopper 45-200 7 41-198.3 4thinstar hopper 210-553.1 8 205-552.3 5thinstar hopper 593-800 12 595-793.3 6thinstar hopper 1000-1500 21 980-1500 fledgling 2000 7 2000 Total nymph duration 68 Young leaves 1st instar hopper 24.3-26 5 23-25 2ndinstar hopper 26.3-43 5 25.6-40 3rdinstar hopper 45-203.1 6 41-200 4thinstar hopper 208.6-556.2 7 205-555 5th instar hopper 602-823.1 9 595-813.6 6th instar hopper 1000-1800 19 980-1700 fledgling 2000 8 2000 Total nymph duration 59

4.2.3.2 Soil texture Soil texture was sandy loam at the different depths; sand represents 71.3 -77.3 %, but it decreased with increase in depth, while clay was about 12.6 -18.1% and it increased with increase in depth; silt was about 8-10.7% (Appendices 13 and 14).

4.2.3.3 PH determination Results show that at Nawa and Elrahad the soil PH was alkaline at all depths (Appendices 15 and 16).

4.2.3.4 N2 determination The soil nitrogen at Nawa and Elrahad ranged between 0.022 and 0.026 % at all depths during three seasons (Appendices 15 and 16).

4.3 Remote sensing interpretation and analysis 4.3.1 Visual interpretation of image The visual interpretation of false colour image composite (Green, red, infrared bands) of landsat image of Nawa dated August 2008 (Image 3) indicated that there were some areas of non- defoliated hashab trees (dark pink colour) others of defoliated hashab trees (dark colour). Similarly the image of the same area dated October 2008 exhibited non- defoliated and defoliated hashab trees (Image 4). In Nawa, visualization of the false colour composite of Spot subset image dated August 2009 (Image 5), demonstrated that some areas covered by highly dense non-defoliated (red colour) and sparse non-defoliated hashab trees (red colour) were larger than that areas covered by highly dense non-defoliated and sparse hashab trees of the same image captured in November 2009 (Image 6).

Both defoliated and non-defoliated trees appeared in regular shape and evenly spaced and systematically arranged. At Elrahad Landsat image dated August 2008 (Image 7) showed similar results when compared to the image of the same area taken in October 2008 (Image 8). In false colour composite of Spot image dated August 2009 (Image 9) and November 2009 (Image 10) of the same location some areas were dominated with non-defoliated and defoliated hashab trees at different levels. In Spot images the recognition of hashab trees whether dense non- defoliated, sparse or defoliated at different levels was more easily, obvious and clear than in Landsat images. Visual recognition and identification of the area of tree locust swarms was not possible in both Landsat and Spot images. Water ponds are shown in August images 2008 and 2009. The result of visual image interpretation was validated through field work and it was satisfactory.

Non-defoliated hashab trees

Defoliated hashab trees

Image 3. Landsat image of Nawa location dated August 2008

Non-defoliated hashab trees

Defoliated hashab trees

Image 4. Landsat image of Nawa location dated October 2008

Water

LDHT

MDHT

DNDHT

SNDHT HDHT Image 5. Spot subset image of Nawa location dated August 2009

DNDHT

SNDHT

LDHT

MDHT

HDHT

Image 6. Spot subset image of Nawa location dated November 2009

W: Water DNDHT: Dense non-defoliated hashab trees SNDHT: Sparse non- defoliated hashab trees LDHT: Light defoliated hashab trees MDHT: Moderate defoliated hashab trees HDHT: High defoliated hashab trees

Defoliated hashab trees

Non-defoliated hashab trees

Image 7. Landsat image of Elrahad location dated August 2008

Non-defoliated hashab Defoliated hashab trees t

Image 8. Landsat image of Elrahad location dated October 2008

LDHT MDHT HDHT

SNDHT

DNDHT

Image 9. Spot subset image of Elrahad location dated August 2009

HDHT MDHT LDHT

NDHT

Image 10. Spot subset image of Elrahad location dated November 2009

4.3.2 Normalized Difference Vegetation Index (NDVI) analysis At Nawa Landsat images dated August and October 2008(Image 11, 12) four NDVI classes that were used showed; non defoliated, light, moderate and high natural defoliated hashab trees). The non defoliated hashab trees showed the highest NDVI value and high natural defoliation showed the lowest one (Table 7 and Appendices 17). There was a positive correlation (r= 0.97) between NDVI values and gum production/tree in all treatments (Table 8). Defoliation reduced NDVI value and the reduction was highly significant (P≤0.001) between all treatments (Table 9). Nawa Spot images dated August and November 2009 (Images 13, 14) were classified into the same four NDVI classes; the non-defoliated trees showed the highest NDVI value and high natural defoliation showed the lowest one (Table 10 and Appendix 18). NDVI values and gum production/tree were positively correlated (r=0.85) in all treatments (Table 11). Defoliation decreased the NDVI value and the decrease was highly significant (P≤0.001) between all treatments (Table 12). Also the same classes were depicted at Elrahad Landsat images dated August and October 2008 (Images 15, 16). NDVI value was higher in the non-defoliated trees and lower in the high natural defoliated ones (Table 13 and Appendix 19). The correlation between NDVI values and gum production/tree was positive (r=0.91) in all treatments (Table 14). Defoliation reduced the NDVI value and the reduction was highly significant (P≤0.001) between all treatments (Table 15). The same four NDVI classes were exhibited at Elrahad Spot images dated August and November 2009 (Images 17, 18). NDVI value was higher in non-defoliated trees and lower in high natural defoliated trees (Table 16 and Appendices 20). NDVI values and gum production/tree were positively correlated (r=0.88) in all treatments (Table 17). Defoliation decreased the NDVI value and the decrease was highly significant (P≤0.001) between all treatments (Table 18).

Image 11. Landsat NDVI image of Nawa location dated August 2008

Image 12. Landsat NDVI image of Nawa location dated October 2008

Table 7. Mean of NDVI values at different blocks and treatments at Nawa location (season 2008/2009) Block NDVI No. Control LND MND HND LAD MAD HAD

1 0.590 0.167 0.147 0.090 0.172 0.151 0.097 2 0.368 0.162 0.139 0.093 0.185 0.141 0.103 3 0.566 0.162 0.147 0.091 0.179 0.147 0.092 4 0.583 0.172 0.134 0.072 0.173 0.159 0.088 Total 2.107 0.663 0.567 0.346 0.709 0.598 0.380 a b c d e f g Mean 0.53 0.17 0.14 0.09 0.18 0.15 0.1 ± SE ±.0.05 ±.0.002 ± .0.003 ± .0.004 ±.003 ±.0.004 ± 0.003

Mean ± SE Mean values of NDVI in each column with same superscript letter showed no significant difference at p = 0.001 as separated by Tukey test.

Table 8. Correlation between NDVI and gum production at Nawa location, (season 2008/2009) NDVI Gum production X Y XY X 2 Y 2

∑ X = 5.37 ∑Y = 2456.2 ∑ XY = 844.85 ∑ X 2 = 1.62 ∑Y 2 = 469750.1

− − X = 0.37 Y = 169.4

R = 0.97 y= -67.4+ 640X

Table 9. Analysis of variance of NDVI values at Nawa location (season 2008/2009) Source of Variation DF SS MS F P Between Treatments 6 0.551 0.0919 55.468 P<0.001 Residual 21 0.0348 0.00166 Total 27 0.586

The differences in the mean values among the treatment groups are greater than would be expected by chance; there is a statistically significant difference (P = <0.001).

Image 13. Spot NDVI image of Nawa location dated August 2009

Image 14. Spot NDVI image of Nawa location dated November 2009

Table 10. Mean of NDVI values at different blocks and treatments at Nawa location (season 2009/2010). Block NDVI No. Control LND MND HND LAD MAD HAD

1 0.243 0.148 0.103 0.069 0.158 0.107 0.082 2 0.239 0.153 0.133 0.084 0.157 0.135 0.099 3 0.224 0.151 0.126 0.067 0.154 0.127 0.089 4 0.230 0.153 0.126 0.091 0.156 0.127 0.098 Total 0.936 0.605 0.488 0.311 0.625 0.496 0.368 a b c d e f g Mean 0.234 0.15 0.12 0.08 0.16 0.12 0.1 ± SE ± .004 ± .001 ± .007 ± .006 ± .001 ± .006 ± .004 Mean ± SE Mean values of gum production in each column with same superscript letter showed no significant difference at p = 0.001 as separated by Tukey test.

Table 11. Correlation between NDVI and gum production at Nawa location (season 2009/2010) NDVI Gum production X Y XY X 2 Y 2

∑ X = 3.829 ∑Y = 1994.06 ∑ XY = 370.6068 ∑ X 2 = 0.589053 ∑Y 2 = 307099

− − X = 0.26 Y = 137.5

R = 0.85 y= -170.9+ 1186X

Table 12. Analysis of variance of NDVI values at Nawa location (season 2009/2010) Source of Variation DF SS MS F P Between Treatments 6 0.0636 0.0106 124.432 <0.001 Residual 21 0.00179 0.0000853 Total 27 0.0654 The differences in the mean values among the treatment groups are greater than would be expected by chance; there is a statistically significant difference (P = <0.001).

Image 15. Landsat NDVI image of Elrahad location dated

August 2008

Image 16. Landsat NDVI image of Elrahad location dated October 2008

Table 13. Mean of NDVI values at different blocks and treatments at Elrahad location (season 2008/2009) Block NDVI No. Control LND MND HND LAD MAD HAD

1 0.325 0.172 0.163 0.120 0.189 0.167 0.148 2 0.307 0.160 0.089 0.027 0.168 0.092 0.045 3 0.271 0.157 0.088 0.040 0.177 0.099 0.044 4 0.259 0.139 0.081 0.053 0.147 0.087 0.054 Total 1.162 0.628 0.421 0.24 0.681 0.445 0.291 a b c d e f g Mean 0.30 0. 16 0.11 0.06 0.17 0.11 0.07 ± SE ±.02 ±.01 ±.02 ±.02 ±.01 ±.02 ±.02 Mean ± SE Mean values of gum production in each column with same superscript letter showed no significant difference at p = 0.001 as separated by Tukey test.

Table 14. Correlation between NDVI and gum production at Elrahad location (season 2008/2009) NDVI Gum production X Y XY X 2 Y 2 ∑ X = 3.87 ∑Y = 2785.7 ∑ XY = 598.14 ∑ X 2 = 0.7072 ∑Y 2 = 600363.4 − − X = 0.267 Y = 192.12

R = 0.91 Y = -149.84+1282.2X

Table 15. Analysis of variance of NDVI values at Elrahad location (season 2008/2009) Source of Variation DF SS MS F P Between Treatments 6 0.147 0.0245 19.973 P< 0.001 Residual 21 0.0258 0.00123 Total 27 0.173 The differences in the mean values among the treatment groups are greater than would be expected by chance; there is a statistically significant difference (P = <0.001).

Image 17. Spot NDVI image of Elrahad location dated August 2009

Image 18. Spot NDVI image of Elrahad location dated November 2009

Table 16. Mean of NDVI values at different blocks and treatments at Elrahad location (season 2009/2010) Block NDVI No. Control LND MND HND LAD MAD HAD

1 0.243 0.156 0.123 0.083 0.157 0.128 0.095 2 0.207 0.153 0.124 0.085 0.153 0.126 0.086 3 0.215 0.151 0.126 0.081 0.153 0.127 0.089 4 0.200 0.152 0.132 0.083 0.154 0.136 0.084 Total 0.865 0.612 0.505 0.332 0.617 0.517 0.374 a b c d e f g Mean 0.22 0.15 0.13 0.1 0.15 0.13 0.1 ± SE ± .009 ± .001 ± .002 ± .001 ± .001 ± .002 ± .002 Mean ± SE Mean values of gum production in each column with same superscript letter showed no significant difference at p = 0.001 as separated by Tukey test.

Table 17. Correlation between NDVI and gum production at Elrahad location, (season 2009/2010) NDVI Gum production X Y XY X 2 Y 2 ∑ X = 3.80 ∑Y = 2205.51 ∑ XY = 390.396 ∑ X 2 = 0.566608 ∑Y 2 = 372427.7 − − X = 0.14 Y = 78. 8

R = 0.88 Y = -72.5+1080.6X

Table 18. Analysis of variance of NDVI values at Elrahad location (season 2009/2010) Source of Variation DF SS MS F P Between Treatments 6 0.0491 0.00818 134.178 <0.001 Residual 21 0.00128 0.0000610 Total 27 0.0504 The differences in the mean values among the treatment groups are greater than would be expected by chance; there is a statistically significant difference (P = <0.001).

4.3.3 Supervised classification The classified Landsat images of Nawa dated August and October 2008 (Images19 and20) highlighted five classes of different areas; In August image the area covered by non-defoliated hashab trees and moderate light defoliated hashab was larger while the area dominated by light and high defoliated hashab was smaller when compared to October image (Fig.16). The overall classification accuracy of August classified image was 80%, the producer accuracy ranged from100% for non-defoliated and light defoliated hashab trees to 66.67% for high defoliated hashab trees, the user accuracy ranged from 87.5-66.67% for moderate and high defoliated hashab trees, respectively. The overall Kappa Statistics = 0.75. The overall classification accuracy of October classified image was 80.65%, the producer accuracy ranged from100-60% for non-defoliated and high defoliated hashab trees, respectively. The user accuracy ranged from 85% for non-defoliated hashab trees to 75% for light and high defoliated hashab trees. The overall Kappa Statistics = 0.75. The classified Spot images of Nawa dated August and November 2009 (Image21and22) highlighted the same five classes with different area coverage; the area covered by non-defoliated, moderate and high defoliated hashab was approximately the same in the two images, while the area covered by light defoliated hashab was larger in November image than in August image (Fig.17). The overall classification accuracy of August classified image was 86.67%, the producer accuracy ranged from100% for non-defoliated and light defoliated hashab trees to 71.43% for high defoliated hashab trees, and the user accuracy raged from100% for non-defoliated hashab trees to 83.33% for high defoliated hashab trees and swarm. The overall Kappa Statistics = 0.82. The overall classification accuracy of November classified image was 87.88%, the producer accuracy ranged from100% for non-defoliated hashab

82 trees and swarm to 33.33% for high defoliated hashab trees, the user accuracy raged from100% for high defoliated hashab trees to 80% for moderate and non- defoliated hashab trees and swarm. The overall Kappa Statistics = 0.84. The classified Landsat images of Elrahad dated August and October 2008 (Images23and24) displayed the five classes as follows; in August image both non-defoliated and light defoliated hashab covered larger areas, while moderate defoliated hashab covered less area when compared to that of October image (Fig.18). The overall classification accuracy of the classified August image was 83.33%, the producer accuracy ranged from100% for light and non-defoliated hashab trees to 33.33% for high defoliated hashab trees, the user accuracy ranged from 100% for high and non-defoliated hashab trees to 71.43% for swarm. The overall Kappa Statistics = 0.77. The overall classification accuracy of the classified October image was 80%. The producer accuracy ranged from100% for light and non-defoliated hashab trees to 50% for high defoliated hashab trees, the user accuracy ranged from 100% for light defoliated hashab trees to 66.67% for high defoliated hashab trees and swarm. The overall Kappa Statistics = 0.74. The classified Spot images of Elrahad dated August and November 2009 (Images25and26) exhibited the same five classes where the area covered by the three levels of defoliation (light, moderate and high) was higher in August image when compared to that of November image (Fig.19). The overall classification accuracy of August classified image was 86.36%, the producer accuracy ranged from100% for moderate and non-defoliated to 50% for swarm, the user accuracy ranged from 100% for light defoliated hashab trees and swarm to 66.67% for non-defoliated hashab trees. The overall Kappa Statistics = 0.81. The overall classification accuracy of November classified image was 83.33%, the producer accuracy ranged from100% for non-defoliated to

77.78% for swarm, the user accuracy ranged from 87.5% for light defoliated hashab trees and swarm to 66.67% for non-defoliated hashab trees. The overall Kappa Statistics = 0.78. Swarms of tree locust almost confined to hashab trees surrounding khors area. The classes from the supervised method of classification matched with NDVI classes.

4.3.4 Layout and mapping The final findings of the study of both ground data and imagery data were integrated with GIS methods to produce maps and graphs (Images, 19, 20, 21, 22, 23, 24, 25 and 26; Figs. 16, 17, 18 and 19).

It can be summarized that the early appearance of hoppers in last July coupled with the increase in their numbers and their high rate of feeding, (individual tree locust can eats an amount of fresh leaf equivalent to its body weight per day) resulted in early infestation. Hence hashab trees prematurely defoliated at different levels. The area of each level varied for the same area at the two different dates, therefore gum quantity and quality in addition to NDVI values were negatively affected by defoliation.

Image19. Supervised classification of Image20. Supervised classification of Landsat image of Nawa location dated Landsat image of Nawa location dated August 2008 October 2008

Figure 16. Area of supervised classes of Landsat image at Nawa location (August and October 2008)

NDHT : Non-defoliated hashab trees, LDHT : Light-defoliated hashab trees,MDHT : Moderate-defoliated hashab trees, HDHT : High-defoliated hashab trees and STL : Swarm of tree locust.

Image 21. Supervised classification of Image 22. Supervised classification of Spot Spot image of Nawa location in August image of Nawa location in November2009 2009

Figure 17. Area of Supervised classes of Spot image at Nawa location(August and November 2009).

NDHT:Non-defoliated hashab trees, LDHT:Light-defoliated hashab trees,MDHT:Moderate-defoliated hashab trees, HDHT:High-defoliated hashab trees and STL:Swarm of tree locust. 86

Image 23. Supervised classification of Image 24. Supervised classification of Landsat image of Elrahad location in Landsat image of Elrahad location in August2008 October 2008

Figure 18. Area of supervised classes of Landsat image at Elrahad location (August and October 2008)

DHT:Non-defoliated hashab trees, LDHT:Light-defoliated hashabtrees,MDHT:Moderate-defoliated hashab trees, HDHT:High-defoliated hashab trees and STL:Swarm of tree locust

Image 25. Supervised classification of Spot Image 26 Supervised classification of Spot image of Elrahad location dated August 2009 image of Elrahad location dated November 2009

Figure 19. Area of supervised classes of Spot image at Elrahad location(August and November 2009)

NDHT:Non-defoliated hashab trees, LDHT:Light-defoliated hashab trees,MDHT:Moderate-defoliated hashab trees, HDHT:High-defoliated hashab trees and STL:Swarm of tree locust

CHAPTER FIVE DISCUSSION

Periodical occurrence of adult tree locusts at Nawa and Elrahad was noted throughout the year except in February, March and April. This could be attributed to; the unavailability of food during these months because all hashab trees were naturally completely defoliated and adult locusts probably migrated to the adjacent fruit orchards and other alternative hosts. Periodical appearance of hoppers occurred only in July, August, September and October reaching the peak in September. The results agreed with findings of other workers. Johnston (1932) and Meinzingen (1993) reported that the Sahelian tree locust, A.m. melanorhodon, is univoltine with dry season imaginal diapause, the female lays eggs during June-July, hoppers develop in August and September and adults emerge at the end of the rainy season and the onset of the dry season (October and November). Luong and Popov (1997) and CIRAD (2006) reported that imagoes remain in a resting maturation stage until the first rain of the following year in May- June. Then they take four weeks before copulation, after that females lay eggs which hatch to give hoppers during the period from July to October. There was a negative correlation between number of adults/tree and rainfall and relative humidity, but a positive correlation between number of hoppers/tree and rainfall and relative humidity. This could be attributed to; the break-up of swarm during egg laying and the subsequent death of adults following egg laying at the beginning of the rainy season, where rainfall has the greatest influence on the breeding and behavior of locusts, supplying the moisture needed both for hoppers development and for the growth of vegetation to sustain a population (Luong and Popov, 1997). Defoliation has significantly (P≤0.001) reduced gum production and the reduction was statistically significant between means except between high natural defoliation and high artificial defoliation. There was a negative correlation between gum production and the different levels of

89 defoliation. The tree locust A. melanorhodon melanorhodon decreased hashab trees production through chewing the foliage (the leaf surface is the main source of food to tree locust) and hence a large portion of tree crown is consumed and both leaf weight ratio (LWR) and leaf area duration (LAD) will be reduced. The reduction in (LWR) and (LAD) results in decreased photosynthesis and hence the amount of carbohydrate produced. Generally leaf defoliators reduce photosynthesis, respiration, sap translocation, growth and gum production (Salih et al., 1981cited by Elbashier, 1994). It was found that the premature removal of a portion of leaves of a tree results in compensatory growth of leaves at the expense of gum production (Elbashier, 1994 and Taha, 2006). Tree locust feeds on hashab tree thus decreasing the quantity of produced gum (Taha, 2002). In Sudan, Abdulla (1990) reported that the infested hashab area in 1987 was estimated at 20000 ha, which increased to 224000 ha in 1989 and reached an unprecedented level of 1.366.762 ha the following season, where the gum produced in 1992 was around 7000 metric tons, the lowest in contemporary history due to the cosmopolitan plague of swarm in 1990. It was the most serious outbreak of tree locust, because the entire gum arabic belt in western, central and eastern regions of the country was infested and gum production was thus decreased (Table 2). In locations where there was no significant reduction in gum production, the compensatory growth could be the responsible factor. However the compensatory growth after the tree was subjected to stress usually resulted in reduced quantity of gum. Defoliation reduced gum viscosity and the reduction was statistically significant between means except between light natural defoliation and light artificial defoliation. There was a negative correlation between gum viscosity and the different levels of defoliation. Defoliation also reduced gum specific optical rotation but the reduction was not significant between means at all seasons. These qualitative effects (gum viscosity and

90 specific optical rotation) could be attributed to physiological changes caused by tree locust infestation, because hashab tree absorbed more water through the root system in order to compensate for the loss caused by defoliation. Therefore the tree sap becomes more diluted thus affecting gum viscosity. Elbashier (1994) reported that; in season 1991, gum viscosity was slightly reduced by defoliation although the reduction was not statistically significant as compared to the control, but in season 1992 the gum viscosity was reduced by defoliation and the reduction was highly significant as compared to the control. He attributed this to the internal physiological changes occurred in hashab tree that were caused by tree locust. The total developmental period of hoppers varied; for those nymphal instars which were fed on old leaves the duration was 68 days, but for those which were fed on young leaves the duration was 59 days, the sixth nymphal instar had longest duration while the first nymphal instar had shortest duration when fed on old or young leaves. This is because young leaves contain more amino acids (Evans and Bell, 1979). The latter reported that leaves of Acacia species contained homoarginine, pipecolic acid and 4-hydroxy-pipecolic acid which are preferred by tree locust and these substances are more concentrated in Acacia senegal leaves than in leaves of other Acacia species. Johnston (1932), Meinzingen (1993) and Luong and Popov (1997) reported that the development of the hopper stages takes 48-69 days. An adult locust daily consumed an amount of green leaves of hashab tree equivalent to its body weight. Pastre et al., (1988) reported that a winged locust weighs about 2g, during its gregarious phase, eats the equivalent of its own body weight per day. The visual interpretation of false colour image composite (green, red, infrared bands) of landsat ETM+ and Spot at Nawa and Elrahad indicated that there were areas of non- defoliated hashab trees(red colour) and defoliated hashab trees ( grey colour). Phenology and vigor were the main factors affecting reflectance value, thus in infrared simulated images, the infrared

91 reflectance is displayed as red. Therefore the non-defoliated hashab trees appeared bright red, while the defoliated hashab trees remained grey in colour (Jain, 1989 and CCRS, 1999). The visual interpretation showed light and moderate levels of defoliation only in Spot images, this is because the spatial resolution of Spot image was very high (2.5m) and that of Landsat image was very low (15m). The four classes of hashab trees that were identified based on NDVI analysis are; non-defoliated class which represents the high NDVI value, the light defoliated class that shows NDVI value less than the former class, the moderate defoliated class has NDVI value lower than the second class and highly defoliated class which represents the lowest NDVI value. This could be attributed to the fact that NDVI is calculated from the visible and near infrared light reflected by hashab tree (NDVI= reflection in infrared light minus reflection in visible light over reflection in infrared light plus reflection in visible light), non-defoliated trees reflect less visible light and more infrared light, while defoliated hashab trees show the opposite (Shank, 2007). There was a positive relationship between NDVI values and gum production/tree in all treatments. Because the non-defoliated hashab trees have a high NDVI value and contain large quantities of chlorophyll which increases photosynthesis hence the amount of carbohydrates (Jensen, 1986, Lillesand et al., 1994, Russ and John, 1995 and CCRS, 1999).The five classes of hashab trees that were identified based on the supervised method of classification were achieved depending on the background knowledge of the study area and the selected training sample of known identity (treatments in all blocks), which are identified by the researcher. Hence the image was classified into the Non-defoliated, light defoliated moderate defoliated and high defoliated hashab trees in addition to swarms of tree locust. The classes from the supervised method were found to be matched with NDVI classes. Variation in areas of defoliation between locations could be attributed to the feeding habit, preference, behavior and population dynamics of tree locust within the study

92 area. Swarms of tree locust always roosted on hashab trees within the vicinity of khors, because hashab trees surrounding khors area were green and highly vegetative and attractive to tree locust population.

CONCLUSIONS AND RECOMMENDATIONS

Three methods (visual interpretation, NDVI analysis and supervised classification) were used to assess the damage to gum arabic trees caused by tree locust using Landsat and Spot images. Visual interpretation of the images identified both non-defoliated and defoliated hashab areas in the study area. NDVI analysis also showed some areas of non-defoliated hashab trees and other defoliated areas at different levels. There was a positive correlation between NDVI values, level of tree greenness and gum production. Supervised classification revealed five major classes of images namely: non- defoliated hashab trees, light, moderate, and high defoliated hashab trees and a swarm of tree locust. Generally adult tree locusts occurred throughout the year in the field except in February, March and April. Hoppers occurred in July, August, September and October. The number of adults/tree decreased with rainfall and increased relative humidity, while the number of hoppers/tree increased with the increase of rainfall and relative humidity. Both natural and artificial defoliation severely reduced gum production, but natural defoliation had more effect than artificial defoliation. There was a negative correlation between gum production and defoliation. Defoliation also significantly reduced gum viscosity and gum optical rotation but the reduction was not significant. The daily food intake of the different stages was approximately equivalent to their body weight, which ranged between 24mg for the first instar and 2g for the adult. Developmental period of the hopper stages varied according to the food quality. When nymphs fed on old leaves the duration was 68 days and when fed on young leaves it was about 59 days. Soil physical and chemical analysis showed no difference between Nawa and Elrahad locations with respect to moisture, texture, PH and nitrogen content. The study concluded that;

The study area represented suitable media for tree locusts roosting and landing most of the year. Tree locust was a dangerous hashab tree feeder that can daily eat amount equal to its body weight and severely affects gum production. Visual interpretation was incapable of differentiating between the different levels of tree defoliation at low resolution. NDVI was a useful index to generate an easy and practical method for classifying tree greenness. Supervised classification is a powerful tool for recognition of the different levels of hashab tree defoliation as well as swarms of tree locust. The limitations experienced by this study can be summarized as fallows: • Difficulties of acquiring Satellite imagery data in time. • Lack of previous studies on monitoring of hashab tree defoliation and absence of a spectral library. • Inaccessibility of some locations of the study area for field surveys on rainy days.

RECOMMENDATIONS Based on the findings and the above mentioned limitations the study poses the following recommendations: • Remote sensing was not a complete substitute for field survey. Both methods should be integrated in order to complement one another. • Further studies should be conducted on application of remote sensing to promote NDVI and supervised classification methods for assessment of tree locust damage in the study area. • Establishment of more extensive regional monitoring network to collect baseline data relevant to all aspects of tree locust to help in locust forecasting. • Use of remote sensing, GIS and GPS in crop assessment and gum production forecasting. • Use of neem extracts on wider scale in crucial gum production areas.

REFFERENCES

Abdulla, A. A. (1990). Review of the (1989-1990) Situation of Grasshoppers, Tree Locust and African Migratory Locust in the Sudan. And outlook for (1990 -91). Ministry of Agriculture. PPD- Khartoum North- Sudan. Abdalla, A. M. (2004). Studies on the Insecticidal Properties of Roots Extracts of Mucuna pruriens (Fabaceae) against the Migratory Locust Locusta migratoria and the Desert Locust Schistocerca gregaria. Ph.D. Thesis, University of Khartoum, Sudan 2004. Abdulla, N. A. (2006). Effect of Tillage on Soil Moisture Storage of Three Gardud Soils of North Kordofan State. Ph.D. Thesis, University of Kordofan- Sudan, September 2006. Anon. (1976). Pest: Anacridium m. melanorhodon (Wlk. 1870) (including A. m.arabafrum Dirsh) (Rosholt., Acrididae) (Sahelian tree locust) Distrib. Maps Insect Pests Commonwealth Institute of Entomology, 355, pp. Anon. (1982). The Locust and Grasshopper Agricultural Manual Published by the Centre Overseas Pest Research. ISBN : 085135. London. Encyclopedia, 511 p. Artec House, Norwood, MA, USA. ISBN: 0- 89006-893-3. Awouda, E. M. (1973). Social and Economic Problems of the Gum Arabic Industry. Oxford University, UK. Awouda, E. M. (1974). Production and Supply of Gum Arabic. Gum Arabic Company, Khartoum, Sudan. Ballal, M. E. (2002). Yield trends of gum arabic from Acacia senegal as related to some environmental and managerial factors. Ph.D. Thesis, University of Khartoum.

Barton, D. and Leonov, S. (1997). Radar technology encyclopedia, 511 p. Artec House, Norwood, MA, USA, ISBN 0-89006-893-3. Bashir, El. M. (1997). Study on the Biology, Seasonal Occurrence and Control of the Sahelian Tree Locust Anacridium m. melanorhodon ( Walker ), (Orthoptera: Acrididae ). M.Sc. Thesis, University of Khartoum, Sudan 1997. Bastin, L. (1997). Comparison of fuzzy c-means classification, linear mixture modelling and MLC probabilities as tools for unmixing coarse pixels, Int. J. Remote Sensing, 1997, vol.18, no.17, 3629-3648. Butrous, M. G. (1995). Radio Message No. 384 - Ministry of Agric. PPD Khartoum- North, Sudan. Campbell, J.B. (1987). Introduction to Remote Sensing. The Guilford Press, NewYork. Canada Centre for Remote Sensing / Natural Resources Canada (1997). Campbell, J.B. (1996). Introduction to Remote Sensing (Second Edition). Taylor and Francis, London. CCRS, (1999). Fundamental of Remote Sensing by Canada centre of remote sensinghttp://www.uni-de:8080/castle/ch5/s414p010.htm. Chikamai, B. N. (1996). A Review of production, Markets and Quality Control of Gum Arabic in Africa. Prepared from Mission Reports by: E. Casadei, J. J. W. Coppen, H. O. Abdel Nour and D. Cesarreo. FAO-Rome, 1996. Cirad (2006): Anacridium spp • Contact: [email protected]. Dafalla, M. Salih, (2006). Mapping and Assessment of Land Use / Land Cover Using Remote Sensing and GIS in North Kordofan State, Sudan. Ph.D. Thesis TU- Dresden, Germany. Deans, J. D., Diagne, O., Lindley, D. K., Dione, M. and Parkinson, J. A. (1999). Nutrient and organic-matter accumulation in Acacia senegal fallows over 18 years. Forest Ecology and Management 124: 153–167.

DeFries, R.S., J. R. G, Townshend and M. C. Hansen (1999). Continuous field of vegetation characteristics at the global scale at 1 km resolution. Journal of Geophysical Research, 104:16911-16925. Dirsh, V. M. and Uvarov, B. P. (1953). Tree locusts of the genus Anacridium (Orth. Acrididae). (Revista Española de Entomología) XXIX (1), 7-69. Madrid. Ref Type: Magazine Article. Dobson, Hans (2000). A Review of the Current Knowledge on Pesticide Application Techniques Related to Desert Locust Control. Nouakchott, Mauritania, and other countries of the EMPRES. Executed by FAO GCP/INT/651/NOR Improving Pesticide Application Techniques for Desert Locust Control. Doka, A. M. A. (1980). Remote Sensing for monitoring soil resources and areas affected by desertification in central Sudan. Proceeding of Sudan Symposium and Workshop on Remote Sensing. Vol. 2, October 1980. Visiting international scientist programm, Remote Sensing Institute. SDSU. USA. Dougherty, Edward, R. and Charles, R. Giardina (1987). Matrix Structured Image. Prentice-Hall, NewJersey. Duranton, J. F.; and Launois M. (1979). The Acridological Consequences of Agricultural Improvement in Sahel (origin language French) http//pestinfo.org/literature/litout.php3. Elamin, H. M. (1990). Trees and shrubs of the Sudan. Extracted from ´´Trees and Shrubs of The Sudan`` Ph.D.Thesis, University of Khartoum- Sudan. Ithaca Press Exeter 1990 (50,163). Elbashier, E. M, (1994). The Impact of Defoliation by the Tree locust ( Anacridium melanorhdon melanorhodon Wlk.) on the Gum Arabic Production by Hashab Tree ( Acacia senegal) M.Sc. Thesis, Faculty of Agriculture, University of Khartoum.

Eldukheiri, I. (1997). Pest change of future prospects of traditional rainfall farming in North Kordofan, Sudan. Ph.D. Dissertation, Technische Universtität München, Germany. El Haja, M.E. (2005). Study of Desertification Based upon Remote Sensing and GIS Techniques (Khor Abu Hbil Area). M.Sc. Thesis. University of Kordofan,Sudan. Elmqvist, B. (2004). Land Use Assessment in Dry Lands of Sudan Using Historical and Recent High Resolution Satellite Data. Ph.D. Thesis Lund University, Centre of Sustainability. Erdas (2003). ERDAS Field Guide™, Seventh Edition. Leica Geosystems GIS & Mapping, LLC. Evans, C.S. and Bell, E. (1979). Non-protein amino acids of Acacia species and their effect on the feeding of the acridids A.melanorhodon. The 2nd ed., Oxford University Press, Oxford. FAO (1991). Food and Nutrition, Paper 49 Specifications for Identity and Purity of Certain Food Additives. FAO (1997). Food and Agriculture Organization. Prevention and Disposal of Obsolete Pesticides In: Annu. Rev. Entomol. 2001.46:671 Downloaded from arjouenals. annualreviews. org by CIRAD- DIST- UNIT BIBLIOTHEQUE Fernandes, R., Butso, C., Leblanc, S. and Latifovic, R. (2003). Landsat-5 TM and Landsat-7 ETM+ based accuracy assessment of leaf area index products for Canada derived from Spot-4 vegetation data. Canadian Journal of Remote Sensing, 29, pp. 241–258. Fleming, R. A. and Volney, W. J. A. (1995). Effects of climate change on insect defoliator population processes in Canada’s boreal forest: some plausible scenarios. Water, Air, and Soil Pollution, 82, pp. 445– 454.

Foody, G. M. and Atkinson, P. M. (2002). Uncertainty in Remote Sensing and GIS, John Wiley & Sons, London. Franklin, S.E. and Raske, A.G. (1994). Satellite remote sensing of spruce Budworm forest defoliation in Western Newfoundland. Canadian Journal of Remote Sensing, 20, pp. 37–48. GAC (2000). Analysis of gum Arabic supply dimensions, paper (Arabic issued) presented by the administration research and development, Gum arabic company, in February 2000. Khartoum, Sudan. Hall, J.P. and Moody, B.H. (1994). Forest depletions caused by insects and diseases in Canada, 1982–1987. Forest Insect and Disease Survey Information Report ST-X-8, Natural Resources Canada, Canadian Forest Service, Ottawa, Canada, 14 pp. Hall, R.J., Fernandes, R.A., Butson, C., Hogg, E.H., Brandt, J.P., Case, B.S. and Leblanc, S.G. (2003). Relating aspen defoliation to changes in leaf area from field and satellite remote sensing perspectives. Canadian Journal of Remote Sensing, 29, pp. 299–313. Haroon, W. M. (2008). Integration of some Bio-pesticides in control of the Tree Locust, Anacridium melanorhodon melanorhodon (Walker, 1870), Sudan. M.Sc. Thesis, TUD, Germany. Harrison, M. N. and Jackson, J. K. (1958). Ecological Classification of the Vegetation of the Sudan, Forestry bulletin No.2, Khartoum. Hedge, S. (2003). Modelling land cover change: A Fuzzy approach, M.Sc. Thesis, International Institute for Geo-information Science and Earth Observation, Enschede, The Netherlands. Heikkila, J., Nevalainen, S. and Tokola, T. (2002). Estimating defoliation in aboreal coniferous forests by combining Landsat TM aerial photographs and field data. Forest Ecology and Management, 158, pp. 9–23.

100

Hellden, U. (1978). Evaluation of Landsat-2 imagery for desertification Studies in Northern Kordofan, Sudan. Lund University. Naturgegrafiska Institution, Raporter Och Notiser Nr 38. Hielkema, J.U., Prince, S. D., Astle, W. L. (1986). Rainfall and vegetation monitoring in the savanna zone of the Democratic-Republic of Sudan using the NOAA advanced very high-resolution radiometer. International Journal of Remote Sensing 7, 1499–1513. Hinderson, T. (2004). Analysing Environmental Change in Semi-Arid Areas in Kordofan, Sudan. Ph.D. Thesis, Lund University, Geobisphere Seience Center. Jain, A.K. (1989). Fundamentals of Digital Image Processing. Prentice- Hall, New Jersey. Jensen, John, R. (1986). Introductory Digital Image Processing. Prentice- Hall, New Jersey. Johnston, H. B. (1932). Notes on two locusts of minor economic importance in the Sudan. Bulletin of Entomological Research 23(1), pages 49-64. Londres. Kassa,A.(1999). Drought Risk Monitoring for the Sudan. Thesis/dissertation. SOAS Water Issues Study Group, University College London. Kaufman, Y. J. and Tanre.D. (1992). Atmospherically resistant vegetation index (ARVI) for EOS-MODIS', in 'Proc. IEEE Int. Geosci. and Remote Sensing Symp. '92, IEEE, New York, 261-270. Kevie, W. (1973). Climatic zone in Sudan. SSA/Wad Medani, Sudan. 8: 277- 287. Key, J. R., Masalanik, J. A. and Barry R. G. (1989). Cloud classification from satellite data using a fuzzy sets algorithm: A polar example, international journal of remote sensing , vol. 10, 1823 - 1842.

101

Khiry, M. A. (2003). Monitoring and Evaluation of Vegetation Cover Changes in Semi-arid Areas: A case Study of Khartoum Forest Sub- sector, Sudan. M.Sc. Thesis, TUD, Germany. Khiry,M.A. (2007). Spectral Mixture Analysis for Monitoring and Mapping Desertification Process in Semi-arid Areas in North Kordofan State, Sudan. Ph.D. Thesis TU- Dresden, Germany. Kooyman, C. and Abdulla, O. M. (1998). Application of Metarhizium flavoviride ( Deuteromycotina, Hyphomycetes) spores against the tree locust, Anacridium melanorhodon (Orthoptera, Acrididae), in Sudan Biocontrol Science and Technology 8 (2), 215-219. Krall, S. and Wilps, H. (1994). Forward: New Trends in Locust Control Ecotoxicology, Botanicals, Pathogens, Attractants, Hormones, Pheromones, Remot Sensing. ISBN 3-88085-504-8. GTZ, Germany. Lampery, H.F. (1975). Report on the desert encroachment reconnaissance in Northern Sudan, 21 Oct to 10Nov, 1975. UNESCO/UNEP.16p. Larry, R. (1997). Creating a Normalized Difference Vegetation Index (NDVI) Image, University of New Hampshire Durham, NH 03824. Leckie, D.G. and Ostaff, D.P.(1988). Classification of airborne multispectral scanner data for mapping current defoliation caused by the spruce budworm. Forest Science, 34, pp. 259–275. Lewis, J.G. E. and Eve, A. (1965). Observations on the biology of a spider of the genus Argiope in the Sudan. The Entomologist 98 (1221), 34-37. Lillesand, T. M. and Kiefer, R. W. (1994). Remote Sensing and Image Interpretation. John Wiley and Sons Inc., New York. Lillesand, T. M. and Kiefer, R. W. (2000). Remote Sensing and Image Interpretation. (4th edition). John Wiley & Sons, New York. Lomer, C. J. R. P., Bateman, D. L., Johnson, J. Langewald, and M. Thomas (2001). Biological Control of Locusts and Grasshoppers Annual Review of Entomology 46, 667-702.

102

Luong,M.H.,Launois and Popov,G.B.(1997). Anacridium m. melanorhodon (Walker, 1870) Acrididae–Cyrtacanthacridinae. Pamphlet, collaborative work Cirad, DLCO, OCLALAV, Rhone- Poulenc agro. isbn: 2-87614- 289-9. Markham, B. L. and Townshend, J. R. G. (1981). Land cover classification accuracy as a function of sensor spatial resolution. Proceedings of the 15th International Symposium on Remote Sensing of Environment, Ann Arbor, MI (Ann Arbor, MI: ERIM), pp. 1075-1090). Mather, P.M. (1987). Computer Processing of Remotely Sensed Images. An Introduction, Chichester: Wiley. Meinzingen, W. F. (1993). A guide To Migrant Pest Management in Africa. FAO. Rome, Italy. Miller, J. R. H., Mcnairy, E., Cloutis, E., Pattey, N. and Tremblay (2002). " User Requirements Report for Agriculture, " Canadian Hyperspectral Users and Science Team, Canadian Space Agency. Nelson, R.F. (1983). Detecting forest canopy change due to insect activity using Landsat MSS. Photogrammetric Engineering and Remote Sensing, 49, pp. 1303–1314. Parmar, B. S. (1987). An overview of Neem research and use in India during the years 1983-1986. In: Natural Pesticides from the Neem Tree and Other Tropical Plants (Eds.: Schmutterer, H., Ascher, K. R. S.), Eschborn, pp. 55-80. Pastre, P. S., Samolikowski, G. and Thewys, E. (1988). Locusts and grasshoppers control: deltamethrin file. ROUSSEL OCLALAV- DIVISION AGRO. Paris, France. Popov, G. and Ratcliffe, M. (1968). The Sahelian Tree Locust Anacridium melanorhodon (Walker); Anti-Locust Memoir 9. Ministry of Overseas Development. Anti-Locust Research Centre, College House, Wrights Lane, London, W.8 No 9:48 pp.

103

Radeloff, V. C., Mladenoff, D. J. and Boye, M. S. (1999). Detecting jack pine budworm defoliation using spectral mixture analysis: separating effects from determinants. Remote Sensing of Environment, 69, pp. 156–169. Rembold, H. (1994). Controlling locusts with plant chemicals pages 41-49. In: New Trends in Locust Control Ecotoxicology,Botanicals, Pathogens, Attractants, Hormones, Pheromones Remot Sensing. ISBN 3-88085- 504-8. GTZ, Germany. Rembold, H., UHL, M. and Muller, TH. (1986). Effect of Azadirachtin on Hormones titters during the Gonadotrophic cycle of Locusta migratoria pages 289-298 in Proceedings of the 3rd International Neem conf.Nairobi. Roger, M. Mccopy (2005). Field Methods in Remote Sensing, the Guilford Press New York London. RSA (2008). Remote Sensing Authority Reports, Khartoum, Sudan. Russ and John, C. (1995). The Image Processing Handbook. 2nd edition. CRC Press, Baca Raton. Schmidt, H. and Karnieli, A. (2000). Remote sensing of the seasonal variability of vegetation in a semi-arid environment. Journal of Arid Environments, 45, 43-59. Schmutterer, H. (1969). Pests of crops in Northeast and Central Africa. Gustarv Fisher Verlag. Stuttgard, Portland, USA. Schmuterer, H. (1983). Cited in: Saxena, R. C. (1987). Neem seed derivatives for management of rice insect pest.A review of recent studies Proc. 3rd Int. Neem Conf., Nairobi, 1986, pp. 81 - 93. Shank, M. (2007). Using Remote Sensing to Map Vegetation Density on a Reclaimed Surface Mine. Presented at the Incorporating Geospatial Technologies into SMCRA Business Processes Conference (Atlanta Ga., March 25-27 2008).

104

Siddig, S.A. (1991). Evaluation of neem seed and leaf water extracts and powders for the control of insect pests in the Sudan. Tech. Bull. No.6 Shambat Research Station, Khartoum, Sudan. Simpson, R. and Coy, D. (1999). An ecological atlas of forest insect defoliation in Canada: 1980–1996. Information Report M-X-206E, Natural Resources Canada, Canadian Forest Service, Fredericton, Canada, 15 pp. Song, C., Woodcock, C. E., Seto, K. C., Pax-Lenney, M., and Macomber, S. A. (2001). Classification and change detection using Landsat TM data: when and how to correct atmospheric effects. Remote Sensing of Environment, 75, 230– 244. Suliman, M.M. (2003). Assessing and Mapping Land Use / Land Cover Change Using Remote Sensing and GIS: A case study of El Amud Al Akhdar Settlement, Southern Darfur-Sudan. M.Sc. Thesis. TUD, Germany. Taha, M. E. (2000). The Socio-economic Role of Acacia senegal in Stainable Development of the Rural Areas in the Gum Belt of the Sudan. Ph.D. Thesis Dresden University of Technology, Germany Institute: Institute of International Forestry and Forest Products, Tharandt, Germany. Taha, M. E. (2006). The Socio-economic Role of Acacia senegal in Sustainable Development of Rural Areas in the Gum Belt of the Sudan. ISBN3 - 9809816-4-9. Dresden University of Technology, Germany Institute: Institute of International Forestry and Forest Products, Tharandt, Germany. Tigani, M. H. (1965). Biology of Poekiloccrus hieroglyphicus (Klug) and Tree Locusts of the genus Anacridium in the Sudan. M.Sc. Thesis, Khartoum, Sudan.

105

Tucker, C. T. and Sellers, P. J., (1986). Satellite Remote Sensing of Primary Production. International Journal of Remote Sensing. Vol. 7, No. 11, pp.1395-1416. Vogt, K. (1995). A Field Workers Guide to the Identification, Propagation and Uses of Common Trees and Shrubs of Dry land Sudan. SOS Sahel International, London. Khartoum, Sudan. Werle, D. (1988 and 1992). Radar Remote Sensing - A Training Manual, 193p, 7535 mm slides, Dendron Resource Surveys Ltd, Ottawa, Ontario, Canada, ISBN 0-9693733-0-9. Wewester, H., Krail, S.and Schulz, F. A. (1993). Methods for the assessment of Crop Losses due to Grasshoppers and Locusts. Technical Corporation Federal Republic of Germany. EFBHEOM-GTZ, Germany. Wilps, H. and Nasseh, O. (1994). Field trial with botanicals, micocides and chitin synthesis inhibitors pages 51-79 In: New Trends in Locust Control Ecotoxicology, Botanicals, Pathogens, Attractants, Hormones, Pheromones, Remot Sensing. ISBN 3-88085-504-8. GTZ, Germany. Wylie, B.K., Meyer D. J., Tieszen L.L .and Mannel,S. (2002). “Satellite mapping of surface biophysical parameters at the biome scale over the North American grasslands A case study” Remote Sensing of Environment 79, 266– 278. Zadeh, L.A. (1965). Fuzzy sets: Information and Control.Vol. 8, 338-353.

106

APPENDICES

Appendix 1. Seasonal occurrence and number of adults and hoppers/tree at Nawa and Elrahad locations in relation to rainfall and relative humidity (seasons 2007/2008, 2008/2009 and 2009/2010).

Months Mean number Mean number Rainfall Relative of adults/tree of hoppers/tree (mm) humidity% January 1 0 0 24 February 0 0 0 16 March 0 0 0 13 April 0 0 0 11 May 10 0 22 40 June 6 0 42 57 July 5 5 63 71 August 3 14 123 83 September 2 32 184 92 October 9 5 61 48 November 15 0 0 26 December 20 0 0 24

Appendix 2. Mean weight (g) of gum produced / hashab tree at different blocks and treatments at Nawa location (seasons 2007/2008, 2008/2009 and 2009/2010)

Block gum production (g) No. Control LND MND HND LAD MAD HAD

1 4429.4 1078 548.3 212 1452 764.4 286.6 2 4190.5 1026.9 538.0 215.9 1400.7 727.9 271 3 4127.9 1001.2 557.5 208.8 1387.6 707.5 260.8 4 4272.2 1011.8 542.9 204.6 1410.7 703.2 267.1 Total 17020.0 4118.5 2139.7 861.3 5651 2903 1085.4 a b c d e f d Mean 283.7 69.2 35.7 14.4 94.2 48.4 18.1 ± SE ± 63.2 ± 24.2 ±11.7 ± 5.4 ±12.5 ± 21.8 ±12.1 Mean values of gum production in each column with same superscript letter showed no significant difference at p = 0.001 as separated by Tukey test.

107

Appendix 3. Mean weight (g) of gum produced/ hashab tree at different blocks and treatments at Elrahad location (seasons 2007/2008, 2008/2009 and 2009/2010

Block gum production(g) No. Control LND MND HND LAD MAD HAD

1 4577.3 1099.6 570.8 240.3 1424.4 664.7 361.2 2 4954 1210.3 623.6 250.7 1653.3 840.2 328.9 3 4894 1194.8 632.9 254.2 1613.3 832.2 337.8 4 4940.6 1221.3 646.2 256.5 1642.3 846.6 368.7 Total 19366 4969.5 2473.5 1010.7 4939.3 3183.7 1396.5 a b c d e c d Mean 322. 8 78. 8 41. 2 16.7 105. 6 53. 1 25. 2 ± SE ± 95.3 ± 36 ± 21.9 ± 5.1 ± 49.7 ± 33.2 ± 23.3 Mean ± SE Mean values of gum optical rotation in each column with same superscript letter showed no significant difference at p = 0.001 as separated by Tukey test.

Appendix 4. Analysis of variance for gum production at Nawa and Elrahad location (season 2007/2008) Source of DF SS MS F P Variation Between 6 54130734.721 9021789.120 3349.570 <0.001 Treatments Residual 21 56561.764 2693.417 Total 27 54187296.485 The differences in the mean values among the treatment groups are greater than would be expected by chance; there is a statistically significant difference (P ≤0.001).

108

Appendix 5. Correlation between gum production/tree and natural and artificial defoliation at Nawa and Elrahad locations (seasons 2007/2008, 2008/2009 and 2009/2019) Level of Intensity of Gum defoliation defoliation production X Y(g) XY X2 Y2 LND 5 73.7 368.5 25 5431.7 MND 10 38.8 388 100 1505.4 HND 15 15.4 231 225 237.2 ∑ 30 127.9 987.5 350 7174.3 Mean 10 39.8 R = -0.994 Y = 103.2 – 5.92X

LAD 5 100 500 25 10000 MAD 10 50.7 507 100 2570.5 HAD 15 20.7 310.5 225 428.5 ∑ 30 160.7 1317.5 350 12999 Mean 10 53.6 R = -0.996 Y = 140.5 – 7.45X

Appendix 6. Gum viscosity (%) at different blocks and treatments at Nawa location (seasons2007/2008, 2008/2009 and 2009/2010) Block Gum viscosity % No. Control LND MND HND LAD MAD HAD

1 35.3 34.4 32.1 30.8 34.6 32.7 31.3 2 35.4 34.3 32.1 30.8 34.5 32.6 31.2 3 35.4 34.4 32.0 30.8 34.5 34.1 31.3 4 35.4 34.4 31.9 30.8 34.6 32.6 31.2 Total 141.5 137.5 128.1 123.2 138.2 132.0 125.0 Mean 35.4a 34.4b 32.0c 30.8d 34.6 b 33.0f 31.3g ± SE ± 0.08 ± 0.02 ± 0.03 ± 0.0 ± 0.01 ± 0.2 ± 0.07 Mean ± SE Mean values of gum viscosity in each column with same superscript letter showed no significant difference at p = 0.001 as separated by Tukey test.

109

Appendix 7. Gum viscosity (%) at different blocks and treatments at Elrahad location (seasons2007/2008, 2008/2009 and 2009/2010) Block Gum viscosity % No. Control LND MND HND LAD MAD HAD

1 35.6 34.5 32.0 30.9 34.6 32.6 31.7 2 35.7 34.6 32.1 31.0 34.6 32.7 31.7 3 35.6 34.5 32.1 31.0 34.7 32.7 31.8 4 35.6 34.5 32.0 30.9 34.6 32.6 31.6 Total 142.5 138.1 128.2 123.8 138.5 130.6 126.8 Mean 35.6a 34.5b 32.1c 31.0d 34.7b 32.7f 31.7g ± SE ± 0.1 ± 0.03 ± 0.03 ± 0.03 ± 0.3 ± 0.03 ± 0.04

Mean ± SE Mean values of gum viscosity in each column with same superscript letter showed no significant difference at p = 0.001 as separated by Tukey test.

Appendix 8. Analysis of variance for gum viscosity at Nawa and Elrahad location (season2007/2008) Source of DF SS MS F P Variation Between 6 74.379 12.397 585.006 <0.001 Treatments Residual 21 0.445 0.0212 Total 27 74.824 The differences in the mean values among the treatment groups are greater than would be expected by chance; there is a statistically significant difference (P ≤ 0.001).

110

Appendix 9. Correlation between natural and artificial defoliation and gum viscosity (%) at Nawa and Elrahad locations( seasons 2007/2008, 2008/2009 and 2009/2010 Level of Intensity Gum infestation of viscosity infestation % XY X2 Y2 X Y LND 5 34.5 172.5 25 1190.3 MND 10 32.1 321 100 1030.4 HND 15 30.6 459 225 936.4 ∑ 30 97.2 952.5 350 3157. 1 Mean 10 32.4 R = -0.94 Y = 36 – 0.36 X LAD 5 34.6 173 25 1197.2 MAD 10 32.9 329 100 1082.4 HAD 15 31.5 472.5 225 992.3 ∑ 30 98.9 974.5 350 3271.9 Mean 10 32.9 R = -0.99 Y = 36.1 – 0.32 X

Appendix 10. Gum specific optical rotation at different blocks and treatments at Nawa location (seasons 2007/2008, 2008/2009 and 2009/2010) Block Gum specific optical rotation No. Control LND MND HND LAD MAD HAD

1 -22.8 -22.7 -22.4 -22.3 -22.7 -22.6 -22.5 2 -22.7 -22.5 -22.5 -22.4 -22.5 -22.4 -22.4 3 -22.6 -22.6 -22.4 -22.5 -22.5 -22.5 -22.6 4 -22.8 -22.5 -22.5 -22.4 -22.4 -22.6 -22.5 Total -90.9 -90.3 -89.8 -89.6 -90.1 -90.0 -90.0 Mean -22.7a -22.6a -22.5a -22.4a -22.5a -22.5a -22.5a ± SE ± 0.04 ± 0.06 ± 0.05 ± 0.05 ± 0.04 ± 0.06 ± 0.03

Mean ± SE Mean values of gum optical rotation in each column share same superscript letter show no significant differences at p = 0.001 as estimated by Tukey Test.

111

Appendix 11. Gum specific optical rotation at different blocks and treatments at Elrahad location, seasons 2007/2008, 2008/2009 and 2009/2010 Gum specific optical rotation Block Control LND MND HND LAD MAD HAD No. 1 -22.9 -22.6 -22.4 -22.3 -22.6 -22.4 -22.4 2 -22.8 -22.5 -22.5 -22.4 -22.5 -22.6 -22.4 3 -22.8 -22.4 -22.5 -22.4 -22.5 -22.6 -22.5 4 -22.6 -22.6 -22.4 -22.5 -22.5 -22.5 -22.5 Total -91.1 -90.1 -89.8 -89.6 -90.1 -90.1 -89.8 a a a a a a a Mean -22. 8 -22. 5 -22.5 -22.4 -22.5 -22.5 -22.5 ± SE ± 0.03 ± 0.04 ± 0.04 ± 0.04 ± 0.04 ± 0.04 ± 0.05 Mean ± SE Mean values of gum specific optical rotation in each column with same superscript letter showed no significant difference at p = 0.001 as separated by Tukey test

Appendix 12. Gum specific optical rotation at different blocks and treatments at Nawa and Elrahad locations, seasons 2007/2008, 2008/2009 and 2009/2010 Gum specific optical rotation Block Control LND MND HND LAD MAD HAD No. 1 -22.9 -22.7 -22.4 -22.3 -22.7 -22.5 -22.5 2 -22.8 -22.5 -22.5 -22.4 -22.5 -22.5 -22.4 3 -22.7 -22.5 -22.5 -22.5 -22.5 -22.6 -22.6 4 -22.7 -22.6 -22.4 -22.5 -22.5 -22.6 -22.5 Total -91.1 -89.3 -89.8 -89.7 -90.2 -90.2 -90.0 a a a a a a a Mean -22. 8 -22. 3 -22.5 -22.4 -22.6 -22.6 -22.5 ± SE ± 0.03 ± 0.04 ± 0.04 ± 0.04 ± 0.04 ± 0.04 ± 0.05 Mean values of gum specific optical rotation in each column with same superscript letter showed no significant difference at p = 0.001 as separated by Tukey test

112

Appendix 13. Soil moisture determination and soil texture at Nawa location (seasons 2007/2008, 2008/2009 and 2009/2010)

Season2007/2008 Soil depth(cm) Textural class Mechanical analysis Soil Silt Sand Clay moisture 0 Sandy loam 10.0 77.3 12.7 5.40 0-50 Sandy loam 8.2 73.9 17.9 5.39 50-100 Sandy loam 10.6 71.5 17.9 5.35 Season 2008/2009 0 Sandy loam 10.1 77.3 12.6 5.41 0-50 Sandy loam 8.1 73.9 18.0 5.39 50-100 Sandy loam 10.4 71.5 18.1 5.36 Season 2009/2010 0 Sandy loam 10.1 77.5 12.9 5.42 0-50 Sandy loam 8.4 74.9 17.9 5.39 50-100 Sandy loam 10.5 71.3 17.9 5.3

Appendix14. Soil moisture determination and soil texture at Elrahad location, (seasons2007/2008,2008/2009 and 2009/2010)

Season 2007/2008 Soil Textural class Mechanical analysis Soil depth(cm) Silt Sand Clay moisture 0 Sandy loam 10.7 76.4 12.9 5.50 0-50 Sandy loam 8.2 73.9 17.9 5.40 50-100 Sandy loam 10.7 71.3 18.0 5.20 Season 2008/2009 0 Sandy loam 10.7 76.3 13.0 5.51 0-50 Sandy loam 8.2 73.9 17.9 5.40 50-100 Sandy loam 10.6 71.4 18.0 5.22 Season 2009/2010 0 Sandy loam 10.6 76.4 12.9 5.60 0-50 Sandy loam 8.0 73.9 17.9 5.30 50-100 Sandy loam 10.7 71.3 18.0 5.00

113

Appendix 15. PH and N2 determination at Nawa location (season2007/2008, 2008/2009 and 2009/2010)

Season 2007/2008 Soil depth(cm) PH CaCO3 Nitrogen Organic carbon 0 7.5 0 0.025 0.24 50 7.9 2 0.026 0.23 100 8.2 3 0.023 0.21 Season 2008/2009 0 7.6 0 0.026 0.25 50 7.9 2 0.026 0.23 100 8.3 3 0.022 0.22 Season 2009/2010 0 7.4 0 0.025 0.23 50 7.7 2 0.026 0.23 100 8.2 3 0.023 0.22

Appendix 16. PH and N2 determination at Elrahad location (seasons 2007/2008, 2008/2009 and 2009/2010)

Season 2007/2008 Soil depth(cm) PH CaCO3 Nitrogen Organic carbon 0 7.5 0 0.026 0.25 50 7.8 1 0.026 0.23 100 8.2 3 0.024 0.22 Season 2008/2009 0 7.6 0 0.026 0.25 50 7.8 1 0.025 0.24 100 8.2 3 0.024 0.22 Season 2009/2010 0 7.2 0 0.024 0.25 50 7.6 3 0.026 0.21 100 8.1 3 0.024 0.22

114

Appendix17. The relationship between different levels of defoliation, NDVI and gum production/tree at Nawa location (season 2008/2009)

Treatment X-coordinate Y-coordinate NDVI GP/tree (g) C1 223934.10 1433697.50 0.590 321.1 C2 221458.95 1429053.95 0.368 301.1 C3 222904.65 1434636.68 0.566 307.1 C4 227709.34 1434108.77 0.583 312.1 LND1 223938.35 1433063.48 0.167 74.7 LND2 221830.64 1429519.01 0.162 72.4 LND3 223647.88 1434923.36 0.162 72.2 LND4 227418.53 1433044.47 0.172 77.4 MND1 223721.22 1432742.13 0.147 39.8 MND2 222572.25 1430333.80 0.139 39 MND3 222357.84 1434387.17 0.147 39.6 MND4 227295.62 1433356.27 0.134 38 HND1 223647.40 1431691.23 0.090 14.6 HND2 222245.36 1430031.30 0.093 15.5 HND3 223769.98 1435215.34 0.091 14.8 HND4 226764.97 1432573.80 0.072 14.2 LAD1 223751.62 1433367.46 0.172 102.7 LAD2 221830.64 1429177.48 0.185 104.6 LAD3 224162.83 1435141.02 0.179 102.7 LAD4 227604.41 1433518.16 0.173 101.5 MAD1 223569.23 1432038.64 0.151 54.1 MAD2 222245.36 1429709.28 0.141 51.1 MAD3 222034.01 1434387.17 0.147 52.1 MAD4 227445.52 1433835.95 0.159 56.1 HAD1 223864.53 1432355.65 0.097 20.5 HAD2 222616.16 1430002.02 0.103 19.6 HAD3 223228.48 1434912.74 0.092 18.6 HAD4 227055.77 1432840.61 0.088 19.03

115

Appendix18. The relationship between different levels of defoliation, NDVI and gum production/tree at Nawa location (season 2009/2010)

Treatment X-coordinate Y-coordinate NDVI GP/tree (g) C1 226070.879 1433462.147 0. 243 263.7 C2 225916.073 1434532.313 0. 239 250.3 C3 224397.43 1433360.984 0. 224 243.4 C4 222617.746 1434721.506 0. 230 243.9 LND1 225830.538 1434194.615 0. 148 60.84 LND2 225793.492 1434875.086 0. 153 65.90 LND3 225033.040 1432752.620 0. 151 61.00 LND4 222901.129 1434438.123 0. 153 62.56 MND1 226025.100 1432821.237 0. 103 31.28 MND2 225126.106 1435108.898 0. 133 32.96 MND3 224742.473 1432906.980 0. 126 30.49 MND4 223040.497 1435460.160 0. 126 30.42 HND1 226128.103 1433908.494 0. 069 12.51 HND2 226552.036 1434480.103 0. 084 12.19 HND3 225378.079 1432371.255 0. 067 12.17 HND4 222854.672 1435785.353 0. 091 13.19 LAD1 226311.220 1433187.471 0. 158 87.90 LAD2 225475.689 1434752.505 0. 157 83.41 LAD3 224224.909 1433034.101 0. 154 81.30 LAD4 222924.375 1434990.952 0. 156 81.12 MAD1 226265.441 1432489.338 0.107 40.56 MAD2 225477.959 1435015828 0.135 43.95 MAD3 224660.750 1432580.100 0.127 40.65 MAD4 222696.721 1435265.044 0.127 41.71 HAD1 226025.100 1432157.438 0.082 16.23 HAD2 224744.743 1435453.941 0.099 17.58 HAD3 225605.080 1432725.378 0.089 16.26 HAD4 222970.813 1436119.838 0.098 16.58

116

Appendix19. The relationship between different levels of defoliation, NDVI and gum production/tree at Elrahad location (season 2008/2009)

Treatment X-coordinate Y-coordinate NDVI GP/tree (g) C1 239223.29 1427355.42 0.325 362.1 C2 232739.21 1431843.93 0.307 361.1 C3 238363.73 1430419.05 0.271 359.1 C4 239785.53 1427213.57 0.259 321.1 LND1 239456.90 1427107.21 0.172 88.9 LND2 233157.27 1431283.34 0.160 87.7 LND3 238116.32 1429843.33 0.157 86.5 LND4 239323.59 1427123.51 0.139 75.8 MND1 240000.06 1426894.03 0.163 48 MND2 233818.34 1430529.72 0.089 47 MND3 237278.91 1428748.99 0.088 45.7 MND4 240537.99 1427701.36 0.081 40 HND1 240359.24 1427180.21 0.120 19 HND2 233104.38 1430430.56 0.027 17.2 HND3 237735.68 1429472.21 0.040 18.4 HND4 240718.12 1428201.36 0.053 18.9 LAD1 238940.03 1427241.54 0.189 120.9 LAD2 232912.67 1430985.86 0.168 117.9 LAD3 238092.53 1430176.39 0.177 119.9 LAD4 239974.37 1427692.94 0.147 96.9 MAD1 239810.25 1427224.01 0.167 62.5 MAD2 232727.57 1431428.78 0.092 61.5 MAD3 237635.76 1429072.53 0.099 61.8 MAD4 241270.12 1428207.17 0.087 43.7 HAD1 240534.45 1426952.44 0.148 28 HAD2 233388.64 1430899.92 0.045 25.1 HAD3 237312.21 1428335.04 0.044 24 HAD4 241252.68 1428907.34 0.054 27

117

Appendix20. The relationship between different levels of defoliation, NDVI and gum production/tree at Elrahad location (season 2009/2010)

Treatment X-coordinate Y-coordinate NDVI GP/tree (g) C1 235379.308 1433462.147 0.243 263.70 C2 238602.222 1427495.910 0.207 278.60 C3 239596.731 1428202.202 0.215 285.26 C4 240551.810 1428734.750 0.200 272.60 LND1 235064.806 1429097.815 0.156 71.32 LND2 237062.135 1428080.228 0.153 69.7 LND3 239233.794 1428475.735 0.151 68.2 LND4 240367.962 1429057.424 0.152 69.5 MND1 233655.951 1429669.440 0.123 34.1 MND2 237964.688 1427684.297 0.124 34.74 MND3 237990.656 1428961.070 0.126 34.80 MND4 239903.708 1430057.358 0.132 35.70 HND1 233777.205 1429190.198 0.083 13.9 HND2 238232.899 1427373.512 0.085 14.3 HND3 238024.714 1428586.425 0.081 13.6 HND4 240339.393 1430057.358 0.083 13.9 LAD1 235826.973 1428780.245 0.157 95.10 LAD2 237449.552 1428050.427 0.153 92.67 LAD3 238484.505 1428680.086 0.153 90.90 LAD4 240617.946 1429443.113 0.154 92.90 MAD1 234666.400 1428993.883 0.128 46.4 MAD2 237573.014 1427701.326 0.126 45.4 MAD3 238825.091 1428305.442 0.127 46.32 MAD4 239675.151 1430335.911 0.136 47.6 HAD1 234360.378 1429299.904 0.095 19 HAD2 238807.637 1427237.278 0.086 18.5 HAD3 239889.422 1428407.618 0.089 18.6 HAD4 240253.684 1429707.381 0.084 18.2

118