Studies on the taxonomy of honeybees in the Sudan

By Elhadi Adam Omer BSc. (Science) University of El Azhar, Cairo, Egypt. MSc. (Agric.) University of Khartoum.

A thesis submitted in fulfillment of the requirements for the degree of Doctor of philosophy

Supervisor Professor Mohamed Abd El Halim Siddig Co supervisor Professor Mohamed Saeed Ali El Sarrag Department of Crop Protection Faculty of Agriculture University of Khartoum.

September 2007

I

ﺑِ ﺴْ ﻢِ اﻟﻠّﻪِ ا ﻟ ﺮﱠ ﺣْ ﻤَ ﻦِ اﻟﺮﱠﺣِﻴﻢ {37} وَﻣَﺎ ﻣِﻦ دَ ﺁ ﺑﱠ ﺔٍ ﻓِﻲ ا ﻷَ رْ ضِ وَ ﻻَ ﻃَ ﺎ ﺋِ ﺮٍ ﻳَ ﻄِ ﻴ ﺮُ ﺑِ ﺠَ ﻨَ ﺎ ﺣَ ﻴْ ﻪِ إِ ﻻﱠ أُ ﻣَ ﻢٌ أَ ﻣْ ﺜَ ﺎ ﻟُ ﻜُ ﻢ ﻣﱠﺎ ﻓَﺮﱠﻃْﻨَﺎ ﻓِﻲ ا ﻟ ﻜِ ﺘَ ﺎ بِ ﻣِﻦ ﺷَ ﻲْ ءٍ ﺛُ ﻢﱠ إِﻟَﻰ رَ ﺑﱢ ﻬِ ﻢْ ﻳُ ﺤْ ﺸَ ﺮُ و نَ {38} ﺳﻮرة اﻷﻧﻌﺎم.

II

DEDICATION I dedicate this work to those whom I love and admire particularly, to the unassuming person, who lightened and still light many dark corners in this life, to my mother and father for their relentless efforts towards my education in spite of their very meager resources. To my wife to whom I am greatly indebted for her patience towards my kids while, I was outside country during the practical course of the study, I dedicate this work to her with my cordial appreciation. To my kids, brothers and sisters with great gratitude and love.

III ACKNOWLEDGEMENT My sincere gratitude is devoted to my supervisor professor Mohamed Abd Elhalim Siddig for his professional guidance and sincere interest in this study. Thanks are extended to my co-supervisor professor Mohamed S. A. Elssarag for his advices and grateful helps. I am greatly indebted to Dr. S. Fuchs for his deep interest, keen supervision, patience and continuous advices and encouragement throughout the practical and analysis part of the morphometric section of this study. My gratitude’s are equally due to professor N. Koeniger (director) for availing me every facilitates in the institute [Institute fuer Bienenkunde (Polytechnische Gesellschaft) Fachbereich Biologie der J. W. Goethe- Universitaet Frankfurt am Main Karl-von-Frisch-Weg 2, D-61440 Oberursel, Germany, where I finished the biometric analysis]. Thus my deep thanks are converted to all the staff members of the institute. I feel pleased to express my deep thanks and sincere gratitude to professor Dr. Moritz and professor Hans (Institut für Zoologie Martin- Luther-Universitut Halle-Wittenberg Hoher Weg 4, D 06099 Halle/Saale, Germany.) for their grateful supervision and criticism during the molecular genetic practical part of this study at their institute. My thanks and wishes are offered to Dr. Marina Meixner (Research Associate Department of Entomology Washington State University Pullman, WA 99164-6382. USA.) for her considerable assistance and advices during the molecular genetic analysis part of this work. At last and not least I would like to express my thanks to my colleges in the Faculty of Natural Resources and Environmental Studies, University of Juba and the department of crop protection, Faculty of Agriculture, University of Khartoum for sparing good atmosphere to complete this study.

IV ABSTRACT

A thorough morphometrical and some molecular genetic studies (Mitochondrial DNA) were carried out on the most common honeybees in the Sudan. These so far contribute in the identification of the Sudanese honeybees. Nineteen samples of honeybee workers Apis mellifera L. were collected from four different geographical zones of the Sudan. Four samples of the small Asian bee workers Apis florea obtained from Gerry, Khartoum, Madani and El-Dender were also included in the study. Biometric measurements and analysis were performed for all the samples. The 19 colonies were subjected to morphometric measurements plus another 8 different samples of Apis mellifera L. were further subjected to Mitochondrial DNA investigation and analysis. Results were compared with those of the biometric study. The morphometric statistical analysis of the nineteen samples revealed a wide range of differences in most discriminant characters among the samples. In the principal component analysis (PCA), three clusters were graphically formed. Furthermore, the presence of these three clusters was confirmed by some modern discriminant analysis methods, and they were geographically correlated. The cluster with the smallest measurements of some discriminant characters originated from the forest zone. Its average measurements were as follows: forewing length 8.23 mm., width 2.82 mm.; proboscis length 5.55 mm.; hind-leg length 6.83 mm.; body size (T3+ T4) 3.88 mm., and cubital index 1.85 mm. The second cluster with medium measurements of some discriminant characters, originated from the semi-desert zone. Its mean average measurements were as follows: forewing length 8.27 mm., width 2.88 mm.;

V proboscis length 5.63 mm.; hind-leg length 7.00 mm.; body size (T3+ T4) 3.88 mm.; and cubital index 2.04 mm. The third cluster, with the highest measurements of some discriminant characters, originated from the savannah zone; mainly towords the border with Ethiopia. Its average measurements were as follows: forewing length 8.45 mm., width 2.95 mm.; proboscis length 5.59 mm.; hind-leg 7.05 mm.; body size (T3+ T4) 4.00 mm., and comparatively the highest cubital index of 2.24 mm. Comparison between the 19 Sudanese honeybees samples and 242 banck samples (data banck, Institute für Bienenkunde, Oberursel, Germany from a neighbouring countries) was done using PCA. The three clusters of the Sudanese bees were like-wise distinguishable as subclusters. The same results were also confirmed by the discriminant analysis. Therefore, the smallest bees of Sudan were identified as Apis mellifera sudanesis instead of Apis mellifera yemenitica which represent the bees of the forest zone. The medium sized bees were identified as Apis mellifera yemenitica instead of sudanesis., representing the semi-desert zone bees, while the bigest bees retained the name Apis mellifera bandasii., representing the Savannah zone bees. The measurement of genetic variation in the Sudanese honeybees Apis mellifera L., at the mitochondrial DNA level of the 27 samples revealed the present of sex different haplotypes. The cluster with the smallest measurements (forest zone colonies) had only haplotype A1 representing 100% of the whole measured colonies; the medium cluster (semi-desert zone colonies) posses two different haplotypes O1 and Y2 with percentages 75% and 25% respectively from the whole measured colonies of the zone, while the cluster of highest measurements (savannah zone) showed four different haplotypes, O1, O1`, A2 and A4, representing 54%, 13%, 13% and 20%

VI respectively. These results partially confirmed the biometric measurements of the PCA and discriminant analysis. The current study represent the first record on the classification of the Sudanese honeybees according to mitochondrial DNA variability. The present study suggest that, the presence of the gene flow among the Sudanese bees in the southern part of the semi-desert zone and almost all the savannah zone of the Sudan is a result of heterogeneous blood mixture between the Sudanese bees and the Ethiopian bees in the border between the two countries and the gene flow direction might be from the low land of the savannah zone of Ethiopia towards the western part of the Sudan in the area between latitudes 9º N and 15º N. Also this study suggests that the origin haplotype of the Sudanese bees is A1 and the pure Sudanese bees might be the south Sudan race (A. m. sudanesis). The four Apis florea samples were also treated by PCA and discriminant analysis, the results obtained so far revealed that, colonies are not very distinct indicating that, all of these colonies were similar and originally they were descendent of the first recorded colony of Apis florea in Khartoum in 1985. Treatment of the four Sudanese Florea samples together with 6 Florea colonies of different origins [2 from Sudan “Moggas ones” and 4 from the data bank, Institute fur Bienenkunde-Oberursel-Germany (Mogga 1988)], by cluster column analysis (which compare values across categories); revealed that, the four target Florea samples of Sudan might be brought from Pakistan or South Iran.

VII ﻣﻠﺨﺺ اﻷﻃﺮوﺣﺔ

ﺃﺠﺭﻴﺕ ﺩﺭﺍﺴﺔ ﺸﺎﻤﻠﺔ ﻓﻲ ﺍﻟﻘﻴﺎﺴـﺎﺕ ﺍﻟﺒﻴﻭﻟﻭﺠﻴـ ﺔ (Biomorphometrics) ﻭﺍﻟﻭﺭﺍﺜـﺔ ﺍﻟﺠﺯﻴﺌﻴﺔ (Molecular Biology) ﻷﻜﺜﺭ ﺃﻨﻭﺍﻉ ﻨﺤل ﺍﻟﻌﺴل ﺇﻨﺘﺸﺎﺭﺍﹰ ﻓﻰ ﺍﻟﺴﻭﺩﺍﻥ. ﻤﻤﺎ ﻴـﺴﺎﻋﺩ ﻓﻲ ﺘﻌﺭﻴﻑ ﻨﺤل ﺍﻟﻌﺴل ﺍﻟﺴﻭﺩﺍﻨﻲ. ﺠﻤﻌﺕ ﺘﺴﻌﺔ ﻋﺸﺭ ﻋﻴﻨﺔ ﻤﻥ ﻨﺤل ﺍﻟﻌﺴل ﺍﻟﺴﻭﺩﺍﻨﻲ Apis mellifera l ﻤﻥ ﺃﺭﺒﻊ ﻤﻨـﺎﻁﻕ ﺠﻐﺭﺍﻓﻴﺔ ﻤﺨﺘﻠﻔﺔ ﻓﻲ ﺍﻟﺴﻭﺩﺍﻥ . ﻜﻤﺎ ﺍﺸﺘﻤﻠﺕ ﺃﻴﻀﺎ ﺍﻟﺩﺭﺍﺴﺔ ﺃﺭﺒﻊ ﻋﻴﻨـﺎﺕ ﻤـﻥ ﺍﻟﻨﺤـل ﺍﻵﺴـﻴﻭﻱ ﺍﻟﺼﻐﻴﺭApis florea ﺍﻟﻤﻭﺠﻭﺩﺓ ﻓﻲ ﺍﻟﺴﻭﺩﺍﻥ ﺠﻤﻌﺕ ﻤﻥ ﻜ لٍ ﻤﻥ ﻗـﺭﻯ ، ﺍﻟﺨﺭﻁـﻭﻡ ، ﻤـﺩﻨﻲ ﻭ ﺍﻟﺩﻨﺩﺭ. ﻜل ﻫﺫﻩ ﺍﻟﻌﻴﻨﺎﺕ ﺃﺨﻀﻌﺕ ﻟﻠﻘﻴﺎﺴﺎﺕ ﺍﻟﺒﻴﻭﻟﻭﺠﻴﺔ ﻜﻤﺎ ﺤﻠﻠﺕ ﻨﺘﺎﺌﺠﻬﺎ. ﺴﺒﻌﺔ ﻭﻋﺸﺭﻭﻥ ﻋﻴﻨﺔ ﻤﻥ ﻨﺤل ﺍﻟﻌﺴل ﺍﻟﺴﻭﺩﺍﻨﻲ ﺃﺠﺭﻴﺕ ﻟﻬﺎ ﺩﺭﺍﺴﺎﺕ ﻓﻲ ﺍﻟﻭﺭﺍﺜﺔ ﺍﻟﺠﺯﻴﺌﻴـﺔ Mitochondrial DNA (ﺘﺴﻌﺔ ﻋﺸﺭ ﻋﻴﻨﺔ ﻤﻥ ﺍﻟﻤﻘﺎﺴﺔ ﺒﻴﻭﻟﻭﺠﻴﺎ ﺒﺎﻹﻀﺎﻓﺔ ﺁﻟﻲ ﺜﻤـﺎﻨﻲ ﻋﻴﻨـﺎﺕ ﺃﺨﺭﻯ ). ﻫﺫﺍ ﻭﻗﺩ ﻗﻭﺭﻨﺕ ﻨﺘﺎﺌﺞ ﺩﺭﺍﺴﺔ ﺍﻟﻭﺭﺍﺜﺔ ﺍﻟﺠﺯﻴﺌﻴﺔ ﺒﻨﺘﺎﺌﺞ ﺍﻟﻘﻴﺎﺴﺎﺕ ﺍﻟﺒﻴﻭﻟﻭﺠﻴﺔ. ﺃﻅﻬﺭﺕ ﺍﻟﺘﺤﺎﻟﻴل ﺍﻹﺤﺼﺎﺌﻴﺔ ﻟﻠﺘﺴﻌﺔ ﻋﺸﺭ ﻋﻴﻨﺔ ﺍ ﺨ ﺘ ﻼ ﻓ ﺎﹰ ﻭ ﺍ ﻀ ﺤ ﺎﹰ ﻓﻲ ﻤﻌﻅﻡ ﺍﻟﺼﻔﺎﺕ ﺍﻟﻤﻤﻴﺯﺓ ﺒﻴﻥ ﺍﻟﻌﻴﻨﺎﺕ . ﻭ ﺒﺘﺤﻠﻴل ﺍﻟﻤﻜﻭﻨﺎﺕ ﺍﻷﺴﺎﺴﻴﺔ (PCA) ﺘﻜﻭﻨﺕ ﺒﻴﺎﻨﺎﺕ ﺜﻼﺜﺔ ﺘﺠ ﻤﻌﺎﺕ ﻤﺘﻨﺎﺴﺒﺔ ﺠﻐﺭﺍﻓﻴﺎ . ﻭﻜﺫﻟﻙ ﺒﺘﺤﻠﻴل ﻨﺘﻴﺠﺔ ﺍﻟﻤﻜﻭﻨﺎﺕ ﺍﻷﺴﺎﺴﻴﺔ ﺒﻭﺍﺴـﻁﺔ ﺍﻟﺘﺤﻠﻴـل ﺍﻟﻤﻤﻴـﺯ ﺍﻟﺤـﺩﻴﺙ Discriminant analysis ﺘﻡ ﺍﻟﺘﺄﻜﺩ ﻤﻥ ﻭﺠﻭﺩ ﺘﻠﻙ ﺍﻟﺘﺠﻤﻌﺎﺕ ﺍﻟﺜﻼﺜﺔ ﻭﻫﻰ ﻜﺎﻻﺘﻰ : ﺃ- ﺍﻟﻤﺠﻤﻭﻋﺔ ﺫﺍﺕ ﺍﻟﻘﻴﺎﺴﺎﺕ ﺍﻟﺼﻐﺭﻯ ﻟﺒﻌﺽ ﺍﻟﺼﻔﺎﺕ ﺍﻟﻤﻤﻴﺯﺓ ﻭﻗﺩ ﻨﺸﺄﺕ ﻫـﺫﻩ ﺍﻟﻤﺠﻤﻭﻋـﺔ ﻤـﻥ ﻤﻨﺎﻁﻕ ﺍﻟﻐﺎﺒﺎﺕ ﻭﻜﺎﻥ ﻤﺘﻭﺴﻁ ﻗﻴﺎﺴﺎﺘﻬﺎ ﻜﺂﻻﺘﻲ :- (ﺃ) ﻁﻭل ﺍﻟﺠﻨﺎﺡ ﺍﻷﻤﺎﻤﻲ 8.23 ﻤﻠﻡ (ﺏ) ﻋﺭﺽ ﺍﻟﺠﻨﺎﺡ ﺍﻷﻤﺎﻤﻲ 2.82 ﻤﻠﻡ (ﺝ) ﻁﻭل ﺍﻟﺨﺭﻁﻭﻡ 5.55 ﻤﻠﻡ (ﺩ) ﻁﻭل ﺍﻟﺭﺠل ﺍﻟﺨﻠﻔﻴﺔ 6.83 ﻤﻠﻡ (ﻩ) ﻁﻭل ﺍﻟﺼﻔﺎﺌﺢ ﺍﻟﻅﻬﺭﻴﺔ ﺍﻟﺜﺎﻟﺜﺔ ﻭﺍﻟﺭﺍﺒﻌﺔ (ﺤﺠﻡ ﺠﺴﻡ) 3.88 ﻤﻠﻡ (ﻯ) ﻤﻌﺎﻤل ﺍﻟﺠﻨﺎﺡ ﺍﻷﻤﺎﻤﻲ (Cubital index) 1.85 ﻤﻠﻡ ﺏ- ﺍﻟﻤﺠﻤﻭﻋﺔ ﺫﺍﺕ ﺍﻟﻘﻴﺎﺴﺎﺕ ﺍﻟﻤﺘﻭﺴﻁﺔ ﻟﺒﻌﺽ ﺍﻟﺼﻔﺎﺕ ﺍﻟﻤﻤﻴﺯﺓ ﻭﻗﺩ ﻨﺸﺄﺕ ﻫﺫﻩ ﺍﻟﻤﺠﻤﻭﻋـﺔ ﻤـﻥ ﻤﻨﻁﻘﺔ ﺸﺒﻪ ﺍﻟﺼﺤﺭﺍﺀ. ﻭﻜﺎﻥ ﻤﺘﻭﺴﻁ ﻗﻴﺎﺴﺎﺘﻬﺎ ﻜﺂﻻﺘﻲ :- (ﺃ) ﻁﻭل ﺍﻟﺠﻨﺎﺡ ﺍﻷﻤﺎﻤﻲ 8.27 ﻤﻠﻡ (ﺏ) ﻋﺭﺽ ﺍﻟﺠﻨﺎﺡ ﺍﻷﻤﺎﻤﻲ 2.88 ﻤﻠﻡ (ﺝ) ﻁﻭل ﺍﻟﺨﺭﻁﻭﻡ 5.63 ﻤﻠﻡ

VIII (ﺩ) ﻁﻭل ﺍﻟﺭﺠل ﺍﻟﺨﻠﻔﻴﺔ 7.00 ﻤﻠﻡ (ﻩ) ﻁﻭل ﺍﻟﺼﻔﺎﺌﺢ ﺍﻟﻅﻬﺭﻴﺔ ﺍﻟﺜﺎﻟﺜﺔ ﻭﺍﻟﺭﺍﺒﻌﺔ (ﺤﺠﻡ ﺠﺴﻡ) 3.88 ﻤﻠﻡ (ﻯ) ﻤﻌﺎﻤل ﺍﻟﺠﻨﺎﺡ ﺍﻷﻤﺎﻤﻲ (Cubital index) 2.04 ﻤﻠﻡ ﺝ- ﺃﻤﺎ ﺍﻟﻤﺠﻤﻭﻋﺔ ﺫﺍﺕ ﺍﻟﻘﻴﺎﺴﺎﺕ ﺍﻟﻜﺒﺭﻯ ﻟﺒﻌﺽ ﺍﻟﺼﻔﺎﺕ ﺍﻟﻤﻤﻴﺯﺓ . ﻨﺸﺄﺕ ﻫﺫﻩ ﺍﻟﻤﺠﻤﻭﻋـﺔ ﻓـﻲ ﻤﻨﻁﻘﺔ ﺍﻟﺴﺎﻓﻨﺎ ﻭ ﻏﺎﻟﺒﺎ ﻤﻊ ﺍﻟﺤﺩﻭﺩ ﺍﻷﺜﻴﻭﺒﻴﺔ. ﻭﻜﺎﻥ ﻤﺘﻭﺴﻁ ﻗﻴﺎﺴﺎﺘﻬﺎ ﻜﺂﻻﺘﻲ :- (ﺃ) ﻁﻭل ﺍﻟﺠﻨﺎﺡ ﺍﻷﻤﺎﻤﻲ 8.45 ﻤﻠﻡ (ﺏ) ﻋﺭﺽ ﺍﻟﺠﻨﺎﺡ ﺍﻷﻤﺎﻤﻲ 2.95 ﻤﻠﻡ (ﺝ) ﻁﻭل ﺍﻟﺨﺭﻁﻭﻡ 5.59 ﻤﻠﻡ (ﺩ) ﻁﻭل ﺍﻟﺭﺠل ﺍﻟﺨﻠﻔﻴﺔ 7.05 ﻤﻠﻡ (ﻩ) ﻁﻭل ﺍﻟﺼﻔﺎﺌﺢ ﺍﻟﻅﻬﺭﻴﺔ ﺍﻟﺜﺎﻟﺜﺔ ﻭﺍﻟﺭﺍﺒﻌﺔ (ﺤﺠﻡ ﺠﺴﻡ) 4.00 ﻤﻠﻡ (ﻯ) ﻤﻌﺎﻤل ﺍﻟﺠﻨﺎﺡ ﺍﻷﻤﺎﻤﻲ (Cubital index) 2.24 ﻤﻠﻡ ﺘﻤﺕ ﺃﻴﻀﺎ ﻤﻘﺎﺭﻨﺔ ﺘﺴﻌﺔ ﻋﺸﺭ ﻋﻴﻨﺔ ﻤﻥ ﺍﻟﻨﺤل ﺍﻟﺴﻭﺩﺍﻨﻲ ﻤﻊ ﻤﺌﺎﺘﻴﻥ ﺍﺜﻨﺎﻥ ﻭ ﺃﺭﺒﻌﻭﻥ ﻋﻴﻨـﺔ ﺒﻨﻜﻴﺔ ( ﺒﻨﻙ ﺍﻟﻤﻌﻠﻭﻤﺎﺕ ﺒﻤﻌﻬﺩ ﻋﻠﻡ ﻨﺤل ﺍﻟﻌﺴل ﺒﻤﺩﻴﻨﺔ ﺍﻭﺒﺭﺍﺴﻭل ﺒﺎﻟﻤﺎﻨﻴـﺎ ، ﻋﻴﻨـﺎﺕ ﻤـ ﻥ ﺍﻟـﺩﻭل ﺍﻟﻤﺠﺎﻭﺭﺓ) ﺒﺎﺴﺘﻌﻤﺎل ﺘﺤﻠﻴل ﺍﻟﻤﻜﻭﻨﺎﺕ ﺍﻷﺴﺎﺴﻴﺔ ﻭ ﺍﻟﺘﺤﻠﻴل ﺍﻟﻤﻤﻴﺯ ﺍﻟﺤـﺩﻴﺙ ﺃﻤﻜـﻥ ﺘﻤﻴﻴـﺯ ﺍﻟﺜﻼﺜـﺔ ﺘﺠﻤﻌﺎﺕ ﺍﻟﺴﻭﺩﺍﻨﻴﺔ ﻜﺘﺤﺕ ﻤﺠﻤﻭﻋﺎﺕ . ﻴﺴﺘﺨﻠﺹ ﻤﻥ ﺫﻟـﻙ ﺃﻥ ﺍﻟﻨﺤـل ﺍﻟـﺴﻭﺩﺍﻨﻲ ﺍﻟـﺼﻐﻴﺭ ﺭﺒﻤـﺎ ﻴﻜـﻭﻥ Apis mellifera sudanesis ﺒﺩﻻ ﻋﻥ A. m. yementica ﻭﻴﻤﺜل ﻨﺤل ﺇﻗﻠﻴﻡ ﺍﻟﻐﺎﺒﺎﺕ. ﺍﻟﻨﺤل ﺍﻟﺴﻭﺩﺍﻨﻲ ﺍﻟﻤﺘﻭﺴﻁ ﺭﺒﻤﺎ ﻴﻜﻭﻥ A. m. yementica ﺒﺩﻻ ﻋـﻥ A. m. sudanesis ﻭﻴﻤﺜـل ﻨﺤـل ﺍﻹﻗﻠـﻴﻡ ﺸـﺒﻪ ﺍﻟﺼﺤﺭﺍﻭﻱ . ﺒﻴﻨﻤﺎ ﻴﺤﺘﻔﻅ ﺍﻟﻨﺤل ﺍﻟﺴﻭﺩﺍﻨﻲ ﺍﻟﻜﺒﻴﺭ ﺒﺎﺴﻡ A. m. bandasii ﻭﻴﻤﺜل ﻨﺤـل ﺇﻗﻠـﻴﻡ ﺍﻟﺴﺎﻓﻨﺎ . ﺘﻤﺨﻀﺕ ﺩﺭﺍﺴﺔ ﺍﻟﻔﺭﻭ ﻗﺎ ﺕ ﺍﻟﻭﺭﺍﺜﻴﺔ ﺒﺎﺴﺘﺨﺩﺍﻡ ﺠﻴﻥ mtDNA ﻓﻰ ﺍﻟﺴﺒﻌﺔ ﻭﻋﺸﺭ ﻭﻥ ﻋﻴﻨﺔ ﻤـﻥ ﻤﺠﻤﻭﻋﺎﺕ ﺍﻟﻨﺤل ﺍﻟﺴﻭﺩﺍﻨﻲ ﻋﻥ ﻅﻬﻭﺭ ﺴﺘﺔ ﻋﻴﻨﺎﺕ ﻤﺨﺘﻠﻔﺔ ﻤﻥ ﺤﻴـﺙ ﺍﻟﻨـﺴﺏ . ﺍﻟﻤﺠﻤﻭﻋـﺔ ﺫﺍﺕ ﺍﻟﻘﻴﺎﺴﺎﺕ ﺍﻟﺼﻐﺭﻯ ﻟﺒﻌﺽ ﺍﻟﺼﻔﺎﺕ ﺍﻟﻤﻤﻴﺯﺓ ﺤﺎﺯﺕ ﻋﻠﻰ ﺍﻟﻤﺼﻨﻔﺔ ﺍﻟﺠﻴﻨﻴﺔ A1 ﺒﻨﺴﺒﺔ %100 ﻤـﻥ ﻜل ﺍﻟﻌﻴﻨﺎﺕ ﺍﻟﻤﺩﺭﻭﺴﺔ ﻓﻲ ﺍﻹﻗﻠﻴﻡ ﺃﻤﺎ ﺍﻟﻤﺠﻤﻭﻋﺔ ﺫﺍﺕ ﺍﻟﻘﻴﺎﺴﺎﺕ ﺍﻟﻤﺘﻭﺴﻁﺔ ﻟﺒﻌﺽ ﺍﻟﺼﻔﺎﺕ ﺍﻟﻭﺭﺍﺜﻴـﺔ ﺤﺎﺯﺕ ﻋﻠﻰ ﻨﻭﻋﻴﻥ ﻤﻥ ﺍﻟﻤﺼﻨﻔﺎﺕ ﺍﻟﺠﻴﻨﻴﺔ O1 ﻭ Y2 ﺒﻨﺴﺒﺔ 75 ﺇﻟﻰ %25 ﻋﻠـﻰ ﺍﻟﺘـﻭﺍﻟﻲ ﻤـﻥ ﺍﻟﻌﻴﻨﺎﺕ ﺍﻟﻤﺩﺭﻭﺴﺔ ﻓﻲ ﺍﻹﻗﻠﻴﻡ . ﺒﻴﻨﻤﺎ ﺍﻟﻤﺠﻤﻭﻋﺔ ﺫﺍﺕ ﺍﻟﻘﻴﺎﺴﺎﺕ ﺍﻟﻜﺒﺭﻯ ﻟـﺒﻌﺽ ﺍﻟـﺼﻔﺎﺕ ﺍﻟﻤﻤﻴـﺯﺓ - (ﻤﺠﻤﻭﻋﺔ ﺇﻗﻠﻴﻡ ﺍﻟﺴﺎﻓﻨﺎ) ﺤﺎﺯﺕ ﻋﻠﻲ ﺃﺭﺒﻌﺔ ﻤﺼﻨﻔﺎﺕ ﺠﻴﻨﻴﺔ ﻤﺨﺘﻠﻔﺔ A4 ، A2 ، O1 ، O1 ﺒﻨﺴﺏ %54 ، %13 ، %13 ، %20 ﻋﻠﻰ ﺍﻟﺘﻭﺍﻟﻲ ﻤﻥ ﺍﻟﻌﻴﻨﺎﺕ ﺍﻟﻤﺩﺭﻭﺴﺔ ﻓﻲ ﺍﻹﻗﻠﻴﻡ. ﺒﺭﻫﻨﺕ ﻫﺫﻩ ﺍﻟﻨﺘﺎﻴﺞ

IX ﻨ ﻭ ﻋ ﺎﹰ ﻤﺎ ﻨﺘﺎﺌﺞ ﺍﻟﻘﻴﺎﺴﺎﺕ ﺍﻟﺒﻴﻭﻟﻭﺠﻴﺔ (ﺘﺤﻠﻴل ﺍﻟﻤﻜﻭﻨﺎﺕ ﺍﻷﺴﺎﺴﻴﺔ PCA ﻭﺍﻟﺘﺤﻠﻴل ﺍﻟﻤﻤﻴﺯ ﺍﻟﺤـﺩﻴﺙ ) ﺍﻟﺴﺎﺒﻘﻴﻥ. ﻜﺫﻟﻙ ﺘﻘﺘﺭﺡ ﻫﺫﻩ ﺍﻟﺩﺭﺍﺴﺔ ﺃﻥ ﻭﺠﻭﺩ ﺍﻹﻨﺴﻴﺎﺏ ﺍﻟﺠﻴﻨﻰ ﺒﻴﻥ ﺍﻟﻨﺤل ﺍﻟﺴﻭﺩﺍﻨﻰ ﻓﻰ ﺍﻟﺠﺯﺀ ﺍﻟﺠﻨﻭﺒﻰ ﻤﻥ ﺍﻹﻗﻠﻴﻡ ﺸﺒﻪ ﺍﻟﺼﺤﺭﺍﻭﻯ ﻭﻜل ﺇﻗﻠﻴﻡ ﺍﻟﺴﺎﻓﻨﺎ ﻴﻌﺯﻯ ﻟﻠﺘﺩﺍﺨل ﻭﺍﻟﺘﺯﺍﻭﺝ ﺒﻴﻥ ﺍﻟﻨﺤل ﺍﻟﺴﻭﺩﺍﻨﻰ ﻭﺍﻟﻨﺤل ﺍﻻﺜﻴﻭﺒﻰ ﻓﻰ ﺍﻟﺤﺩﻭﺩ ﺒﻴﻥ ﺍﻟﺒﻠﺩﻴﻥ ﻭﻜﺫﻟﻙ ﺭﺒﻤﺎ ﻴﻜﻭﻥ ﻤﺩﻯ ﻫﺫﺍ ﺍﻹﻨﺴﻴﺎﺏ ﺍﻟﺠﻴﻨـﻰ ﻴـﺸﻤل ﺍﻟﻤﻨﻁﻘـﺔ ﺍﻟﻤﻨﺨﻔﻀﺔ ﻤﻥ ﺍﻟﺴﺎﻓﻨﺎ ﺍﻻﺜﻴﻭﺒﻴﺔ ﺤﺘﻰ ﻏﺭﺏ ﺍﻟﺴﻭﺩﺍﻥ ﺒﻴﻥ ﺨﻁﻰ ﻋﺭﺽ N°9 ﻭ 15°N. ﺃ ﻴ ﻀ ﺎﹰ ﺘﻘﺘﺭﺡ ﻫﺫﻩ ﺍﻟﺩﺭﺍﺴﺔ ﺃﻥ ﺍﻟﻨﺤل ﺍﻟﺴﻭﺩﺍﻨﻰ ﺍﻟﻨﻘﻰ ﺭﺒﻤﺎ ﻴﻜﻭﻥ ﻨﺤل ﺠﻨﻭﺏ ﺍﻟـﺴﻭﺩﺍﻥ .A. m sudanesis ﻜﺫﻟﻙ ﺃﻥ ﺍﻟﻤﺼﻨﻑ ﺍﻟﺠﻴﻨﻰ ﺍﻟﻨﻘﻰ ﻟﻠﻨﺤل ﺍﻟﺴﻭﺩﺍﻨﻰ ﻗﺩ ﻴﻜﻭﻥ A1 ﻭﻫﺫﻩ ﺍﻟﺩﺭﺍﺴﺔ ﺘﻤﺜل ﺃﻭل ﺘﻘﺭﻴﺭ ﻓﻲ ﺩﺭﺍﺴﺔ ﺘﻘﺴﻴﻡ ﺍﻟﻨﺤل ﺍﻟﺴﻭﺩﺍﻨﻲ ﻤﻥ ﺤﻴﺙ ﺍﻟﻔﺭﻭ ﻗﺎﺕ ﺍﻟﻭﺭﺍﺜﻴﺔ ﺒﺎﺴﺘﺨﺩﺍﻡ

ﺠﻴﻥ mt DNA. ﻜﺫﻟﻙ ﺃُﺨﻀﻌﺕ ﺃﺭﺒﻌﺔ ﻋﻴﻨﺎﺕ ﻤﻥ ﺍﻟﻨﺤل ﺍﻵﺴﻴﻭﻱ ﺍﻟﺼﻐﻴﺭ Apis florea ﻟﺘﺤﻠﻴل ﺍﻟﻤﻜﻭﻨﺎﺕ ﺍﻷﺴﺎﺴﻴﺔ (PCA ) ﻭ ﺍﻟﺘﺤﻠﻴل ﺍﻟﻤﻤﻴﺯ ﺍﻟﺤﺩﻴﺙ (Discriminant analysis) ﺤﻴﺙ ﺃﻭﻀﺤﺕ ﺍﻟﻨﺘﺎﺌﺞ ﺃﻥ ﺍﻟﻤﺴﺘﻌﻤﺭﺍﺕ ﺍﻷﺭﺒﻌﺔ ﻟﻴﺴﺕ ﻭﺍﻀﺤﺔ ﺍﻟﺘﻤﻴﻴﺯ ﻤﻤﺎ ﻴﺩل ﻋﻠﻰ ﺃﻥ ﻫﺫﻩ ﺍﻟﻤﺴﺘﻌﻤﺭﺍﺕ ﻤﺘﻤﺎﺜﻠـﺔ ﻭﻓـﻰ ﺍﻷﺼل ﺘﺭﺠﻊ ﺇﻟﻰ ﻤﺼﺩﺭ ﻭﺍﺤﺩ ﻭﻫﻭ ﺃﻭل ﺨﻠﻴﺔ ﺍﻜﺘﺸﻔﺕ ﻓﻲ ﺍﻟﺨﺭﻁﻭﻡ ﺴﻨﺔ 1985ﻡ . ﺃُﺨﻀﻌﺕ ﺃﻴﻀﺎ ﺍﻷﺭﺒﻌﺔ ﻋﻴﻨﺎﺕ ﻤﻥ ﺍﻟﻨﺤل ﺍﻟﺼﻐﻴﺭ . Apis Florea ﻟﻠﻤﻘﺎﺭﻨﺔ ﻤﻊ ﺴـﺕ ﻋﻴﻨﺎﺕ ﺃﺨﺭﻯ (ﻋﻴﻨﺘﻴﻥ ﻤﻥ ﺍﻟﺴﻭﺩﺍﻥ .Mogga 1988, ﻭﺃﺭﺒﻌﺔ ﻤﻥ ﻤﻌﻬﺩ ﻋﻠﻡ ﺍﻟﻨﺤل ﺒﺄﻟﻤﺎﻨﻴﺎ ) ﻋـﻥ ﻁﺭﻴﻕ ﺘﺤﻠﻴل ﺍﻟﻌﻤﻭﺩ ﺍﻟﻌﻨﻘﻭﺩﻱ (ﻤﻘﺎﺭﻨﺔ ﺍﻟﻘﻴﻡ ﻋﺒﺭ ﺍﻷﺼﻨﺎﻑ ﺍﻟ ﻤﺨﺘﻠﻔﺔ). ﺃﻭﻀﺤﺕ ﺍﻟﻨﺘﺎﺌﺞ ﺃﻥ ﺍﻷﺭﺒﻌﺔ ﻋﻴﻨﺎﺕ ﺍﻟﺘﻰ ﺩﺭﺴﺕ ﺭﺒﻤﺎ ﻴﻜﻭﻥ ﻤﺼﺩﺭﻫﺎ ﻤﻥ ﺒﺎﻜﺴﺘﺎﻥ ﺃﻭ ﺠﻨﻭﺏ ﺇﻴﺭﺍﻥ.

X TABLE OF CONTENTS

Title Page

DEDICATION II

ACKNOWLEDGEMENT III

ABSTRACT VII

ARABIC ABSTRACT VIII

TABLE OF CONTENTS X

LIST OF TABLES XVIII

LIST OF FIGURES XXI

CHAPTER: ONE INTRODUCTION 1

1- Sudan climate and vegetation zones. 1

1- a- Climate 1

1- b- Vegetation 1

1- b- i- The desert zone 1

1- b- ii- The semi-desert zone 1

1- b- iii- The poor savannah zone 1

1- b- iv- The rich savannah zone 2

1- b- v- The forest zone 2

2- Historical background of the Sudanese honeybees (Apis mellifera L.) morphometrics. 4

XI 1- 3 Apis florea (Fabricius) from Sudan 6

1- 4- Molecular biology development 7

1- 5- The objectives of the present study 8

CHAPTER TWO: LITERATURE REVIEW 9

2- a- History of bees evolution 9

2- b- Origins of honeybees 11

2- c- Differentiation of Apinae 12

2- c- i- Genus Apis 12

2- c- ii- Distribution of the genus Apis 13

2- d- The dwarf honeybee Apis florea and Apis andrenifomis taxa 13

2- d- i- Distribution and description of Apis florea. 15

2- d- ii- Classification of Apis florea. 16

2- d- iii- Historical background of Apis florea in Africa. 17

2- e- Apis dorsata and Apis laboriosa (the giant honeybees) 19

2- f- Apis cerana and Apis mellifera L. 19

2- f- i- Races of Apis cerana: 20

2- f- i- (1) Apis cerana cerana. 20

2- f- i- (2) A. c. indica. 20

2- f- i- (3) A. c. japonica 20

2- f- i- (4) A. c. himalaya. 20

2- f- ii- Races of Apis mellifera L.: 20

XII 2- f- ii- (1) European races: 23

2- f- ii- (1)- a- Apis mellifera mellifera (German dark bees). 23

2- f- ii- (1)- b- A. m. ligustica Spin. (Italian bees). 23

2- f- ii- (1)- c- A. m. carnica Poltman (carniolin bees). 24

2- f- ii- (1)- d- A. m. caucasica Gorb (Caucasian bees). 24

2 f- ii- (2)- African races: 25

2- f- ii- (2)- a- Apis mellifera intermissa "Tellian bees" (Maa, 1953). 25

2- f- ii- (2)- b- A. m. lamarkii (Cockerel, 1906). 25

2- f- ii- (2)- c- A. m. scutellata Lepeltier (East Africa bees). 26

2- f- ii- (2)- d- A. m. adansonii (Latreille, 1804). 26

2- f- ii- (2)- e- A. m. monticolla (Smith, 1961) "Mountain bees". 26

2- f- ii- (2)- f- A. m. capensis (Escholtz, 1822). 26

2- f- ii- (3)- Oriental races: 27

2- f- ii- (4)- Northern and South America races. 28

2- g- Probability and identification of the honeybees 28

2- h- Morphometrics. 30

2- h- i- Historical background of bees morphometrics. 30

2- h- ii- Recent developed biometry of Apis mellifera L. 35 2- i- Historical background of African honeybees classification: 38

2- j- East Africa races of honeybees. 42

XIII 2- j- i- Apis mellifera scutellata. 42

2- j- ii- A. m. monticolla 42

2- j- iii- A. m. yemenitica L. 43

2- j- iv- A. m. litorea. 43

2- j- v- A. m. sudanesis 43

2- j- vi- A. m. bandasii. 43

2- k- Classification of the Sudanese honeybee races. 43

2- l- Some biologic and ecological aspects of honeybees. 46

2- l- i- Biological factors 47

2- l- i- (1)- Migration 47

2- l- i- (2)- Reproductive swarming 48

2- l- i- (3)- Seasonal cycles of honeybee colonies 50

2- l- i- (4)- Temperament 51

2- l- ii- Ecological factors 52

2- m- Genetic diversity and honeybees 54

2- m- i- Allozymes 55

2- m- ii- Nuclear DNA 60

2- m- iii- Mitochondrial DNA 64

2- n- Animal mitochondrial DNA 65

2- n- i- Size of animal mitochondrial DNA 65

2- n- ii- Apis mellifera L. mitochondrial DNA 67

XIV CHAPTER THREE: MATERIALS AND METHODS 88

3- 1- Sampling 88

3- 2- Morphometric analysis 91

3- 2- a- Preparation and measurements records of bees 94

3- 2- b- Statistical analysis 106

3- 3- Mitochondrial DNA 107

3- 3- a- DNA extraction 108

3- 3- b- PCR amplification 109

3- 3- c- DNA purification 110

3- 3- d- Size category of the fragments 111

3- 3- e- Endonuclease digestion 111

CHAPTER FOUR: RESULTS 112

4- 1- Morphometric analysis (Apis mellifera L.) 112

4- 1- i- Uni-variate analysis 112

4- 1- i- a- Proboscis and hind leg measurements (Table 3). 112

4- 1- i- b- Forewing measurements (Table 4). 112

4- 1- i- c- Forewing venation angles measurements (Table 5). 114

4- 1- i- d- Body size (tergites 3+4) and sternite 3 measurements (Table 6): 114

4- 1- i- e- Pilosity measurements (Table 7): 114

4- 1- i- f- Sternite 6 (Abdominal slenderness) measurements (Table 8). 115

XV 4- 1- i- g- Pigmentation (coloration) measurements (Table 9). 115

4- 1- i- h- Means, minimum, maximum and standard deviation of each morphometric character from the 285 individual bees measured (Table 10). 124

4- 1- i- i- Sum of squares, dF, mean square, F values and significances for each phenotypic character from the measured individuals. (Sudanese honeybee Apis mellifera L.) [Appendix H]. 124

4- 1- ii- Multi-variety analysis. 124

4- 1- ii- a- Principal Component Analysis (PCA). 124

4- 1- ii- b- Discriminant analysis. 133

4- 2- Mitochondrial DNA analysis (Apis mellifera L.). 150

4- 2- i- Amplified PCR analysis. 151

4- 2- ii- Restriction analysis. 151

4- 3- observations on some behavior and biology of the Sudanese honeybees Apis mellifera L. 165

4- 3- a- Colored colonies. 165

4- 3- b- Defensive behavior. 165

4- 3- c- Swarming and migration. 166

4- 3- d- Nesting sites. 167

4- 4- Morphometric statistical analysis of Apis florea. 168

XVI

CHAPTER FIVE: DISCUSSION, SUMMARY AND CONCLUSION 182

5- 1- DISCUSSION 182

5- 1- a- Morphometrics: Apis mellifera L. 182

5- 1- b- Mitochondrial DNA. Apis mellifera L. 198

5- 1- c- Apis florea. 203

5- 2- SUMMARY AND CONCLUSION 208

5- 2- a- Apis mellifera L. 208

5- 2- b- Apis florea. 215

CHAPTER SIX: LIST OF REFRENCES 216

CHAPTER SEVEN: APPENDIXES 263

Appendix (A): Abbreviations of the morphometric characters used in the study. 263

Appendix (B): Climatologically and Rainfall averages for at least 30 years from Climatologically stations in or/near the sample collection areas. 265

Appendix (C): Taxonomic relationships between bees in the family Apidae. 284 Appendix (D): Different species of the Genus Apini (Institute Für Bienenkunde, Oberursel, Germany). 285

Appendix (E): Natural distribution of Honeybee Species (Institute Für Bienenkunde, Oberursel, Germany). 286

Appendix (F): Geographical distribution of Genus Apis (Institute Für Bienenkunde, Oberursel, Germany). 287

XVII Appendix (G): Distribution of geographical honeybee races and mean of annual temperature (F. Ruttner, Institute Für Bienenkunde Oberursel, Germany). 288

AAppendix (H): Multivariate ANOVA Table: Sum of squares, dF, mean values and significances for each phenotypic character character from the

s. (Sudanese honeybees Apis melli Apis mellifera L.) 289

XVIII LIST OF TABLES

Table Page

1- Sampling localities, respective geographical zones, map reference numbers and coordinates of honeybee localities 89

2- List of characters used for the analysis 92

3- Means of measurements of proboscis and hind- leg of the Sudanese honeybee workers Apis mellifera (mm.) 113

4- Means of measurements of the forewing of the Sudanese honeybee workers Apis mellifera (mm.) 116

5- Means of measurements of forewing venation angles of the Sudanese honeybee workers Apis mellifera (degrees) 117

6- Means of measurements of some tergites and sternites of the Sudanese honeybee Apis mellifera (mm.). 118

7- Means of measurements of pilosity of the Sudanese honeybee workers Apis mellifera (mm.) 119

8- Means of measurements of sternite 6 of the Sudanese honeybee workers Apis mellifera (mm.) 120

9- Means of measurements of coloration of the Sudanese honeybee workers Apis mellifera. 121

10- Mean, minimum, maximum and standard deviation of each morphometric character from the 285 individual bees Apis mellifera L. measured (measurements in mm. Angels in degree) 122

12- Factor loadings in varimax rotation for each character in the principal component analysis 127

13- Means of some discriminant characters for the Sudanese honeybee workers Apis mellifera (mm.), from the semi desert region 137

XIX 14- Means of some discriminant characters for the Sudanese honeybee workers Apis mellifera (mm.), from the savannah region 138

15- Means of some discriminant characters for the Sudanese honeybee workers Apis mellifera (mm.), from the forest region 139

16- Classification matrices of colonies in cluster groups based on step-wise discriminant analysis. (Sudan samples only) 140

17- Pair-wise Group comparison a, b, c, d, e. (Sudan samples) 142

18- Discriminant analysis probability 143

19- a- Discriminant Classification results (Sudan and others) 144

19- b- Discriminant Classification results (Sudan and others) 146

20- Proximities discriminant centroid distances between the groups (Dissimilarity matrix) 147

21- a- Sudanese honeybee workers Apis mellifera L. different haplotypes 159

21- b- Distribution of the Sudanese honeybee worker haplotypes according to the three different geographical zones 160

22- Means of measurements of proboscis and hind- leg of the honeybee workers Apis florea (mm.) from Sudan 169

23- Means of measurements of forewing venation angles of the honeybee workers Apis florea (degrees.) from Sudan 170

24- Mean of measurements of coloration of the honeybee workers Apis florea from Sudan 171

25- Means of measurements of the forewing of the honeybee workers Apis florea from Sudan (mm.) 172

26- Means of measurements of some tergites and sternites of honeybee workers Apis florea (mm.) from Sudan 173

XX 27- Means of measurements of pilosity of the honeybee workers Apis florea (mm.) from Sudan 173

28- Means of measurements of sternite 6 of the honeybee workers Apis florea (mm.) from Sudan 173

29- Means of some discriminant characters for the honeybee workers Apis florea from Sudan 174

30- Means, minimum, maximum and standard deviation of each morphometric character from the 40 individual bees measured (Apis florea) [Measurements in mm. Angels in degree] 175

31- Proximities discriminant centroid distances between the groups (dissimilarity matrix). Apis florae 177

32- Some characteristics measurements of Apis florea from different origins 178

33- Means of some characters of African bees mm. (Ruttner, 1975) 183

34- The average values of measurements taken for the different characters of the Sudanese honeybee workers (El Sarrag, 1977) 184

35- The average values of measurements taken for the different characters of the Sudanese honeybee workers (Saeed 1981) 185

36- The average values of measurements taken for the different characters of the Sudanese honeybee workers (Mohamed 1982) 186

37- Average means of some discriminant characters for the Sudanese honeybee workers (Apis mellifera L.) From different geographical zones Mogga (1988) 187

XXI LIST OF FIGURES

Figure. Page.

1- Sudan map showing the collection localities. 3

2- Abdomin of the honeybee workers Apis mellifera L. 95

3- Length of the proboscis of the honeybee workers A. mellifera L. 96

4- Hind-leg of the honeybee workers A. mellifera L. 97

5- Classes of pigmentation of tergites (2 to 4) of the ho8neybee workers Apis mellifera L. 98

6- Longitudinal diameter of tergite 3 and 4 of honeybee workers A. mellifra L. 99

7- Sternite 3 of honeybee workers Apis mellifera L. 99

8- Length and width of sternite 6, of honeybee workers Apis mellifera L. 100

9- Honeybee workers Apis mellifera L. forewing. 101

10- Scutellum of the honeybee workers Apis mellfera L. 102

11- a Bee workers Apis mellfera L. Labrum. 103

11-b Labrum pigmentation of the honeybee workers Apis mellifera L. 104

12- Angles of wing venation of the honeybee workers Apis mellifera L. 105

13- Scatter plot graph of factor scores of factor 1 and factor 2 from principal components analysis of 19 colony means of all morphometric data. 128

14- Scatter plot graph of factor scores of factor 1 and factor 3 from principal components analysis of 19 colony means of all morphometric data. 129

15- Scatter plot graph of factor analysis of 239 samples of worker honeybees of different origin. 132

XXII 17- Canonical Discriminant Function (Sudan samples only). 140

18- Dendrogram using average linkage between the groups (Sudan colonies only). 141

19- Canonical Discriminant Function (Sudan and others). 148

20- Dendrogram using average linkage between the groups (Sudan and others). 149

21- a- Agarose gel (1.5%) with the amplified PCR products of the of the Sudanese honeybee worker samples. 153

21- b- Agarose gel (1.5%) with the amplified PCR products of the Sudanese honeybee worker samples. 153

22- a- Acrylamide gel (8%) showing the Dra 1 restriction patterns of the Intergenic region of the Sudanese honeybee worker samples 154

22- b- Acrylamide gel (10%) showing the Dra 1 restriction patterns of the Intergenic region of the Sudanese honeybee worker samples 155

22- c- Acrylamide gel (8%) showing the Dra 1 restriction patterns of the Intergenic region of the Sudanese honeybee worker samples. 156

22- d- Acrylamide gel (8%) showing the Dra 1 restriction patterns of the Intergenic region of the Sudanese honeybee worker samples 157

22- e- Acrylamide gel (8%) showing the Dra 1 restriction patterns of the Intergenic region of the Sudanese honeybee worker samples 158

23- a- Pie with 3rd visual effects representing the distribution percentage of the Sudanese honeybee worker haplotypes of the studied 27 colonies. 161

23- b- Cluster column with 3rd visual effects representing the distribution percentage of the Sudanese honeybee worker haplotypes of the studied 27 colonies. 161

XXIII 24- a- Pie with 3rd visual effects representing the distribution percentage 162 of semi-desert zone Sudanese honeybee worker haplotypes.

24- b- Pie with 3rd visual effects representing the distribution percentage of savannah zone Sudanese honeybee worker haplotypes. 163

24- c- Pie with 3rd visual effects representing the distribution percentage of forest zone Sudanese honeybee worker haplotypes. 164

25- Scatter plot graph of factor scores of factor 1 and factor 2 from principal components analysis of 4 colonies means of all 179 morphometric data (Apis florea).

26- Canonical Discriminant Function (Apis florea). 180

27- Graphic representation of cluster column analysis of 10 samples of Florea bees of different origins. 181

XXIV

XXV

26

27 CHAPTER ONE: INTRODUCTION 1- 1- Sudan climate and vegetation zones: The Sudan is a vast country in Africa exhibiting the heart of the continent. The total area measurements approximately 2,505, 813 square kilometers. 1- 1- a- Climate: The climate is the most important force that keeping different population in check. The temperature, humidity and rainfall affect directly or indirectly every phase of the life. 1- 1- b- Vegetation: According to Schmutterer (1969), There are main five types of vegetation regions or zones in Sudan, which in general follow each other from north to south, namely: desert, semi–desert, poor savannah, rich savannah and forest region. 1- 1- b- i- The desert zone: Represent about one third of the entire country. The rainfall is very light and irregular (0 to 50 mm. per year). Various species of Acacia are found along the Nile and in depressions not too far from the river, thus Date palms ”Dom”, palms and Halfa grass are found in plenty along the banks of the Nile river. 1- 1- b- ii- The semi - desert zone: It is much smaller than the latter; its vegetation is richer than that of the desert, which somewhat due to the high rainfall of about 50 to 300 mm. per year. Numerous shrubs and small trees grow in this area; most of them are members of Acacia genus, thus various species of short grasses are also common. 1- 1- b- iii- The poor savannah zone: This region comparatively forms narrow belt, which runs approximately in an east - west direction across the Sudan. The

1 rainfall of this vegetation belt varies from 300 to 500 mm. per annum and the dry season lasts 4 – 6 months. The dominant trees are still Acacias, which in many areas of the zone form open woodland, the variety of trees in general is greater than in the preceding belt; the same applies to herbs and grasses. 1- 1- b- iv- The rich savannah zone: Lies between the poor savannah and the forest region, and it is more- or less triangular in shape with a broad extension to the west where it runs up to the Bahr el Arab in the south. The annual rainfall amounts between 500 and 1000 mm. per year. Acacia spp. is still the predominant trees of the region, but there is an obvious change in the flora in comparison to that of the preceding vegetation belt. Broad-leaved trees of the family Combretaceae become more and more abundant, thus ‘Tebeldi’ tree Adansonia digitata may be commonly found and ‘Heglig’ Balanites aegyptiaca is also abundant in many places. 1- 1- b- v- The forest zone: This region forms the most southerly vegetation belt in the Sudan. The annual rainfall of this area is as high as 1000 to 1500 mm. per annum resulting in very thick vegetation in many places, especially along water streams and in depressions where gallery forests and dense depression forests occur. These forests lie mainly in the southwestern parts of the region and are the link to the rain forests of the neighboring Congo and Uganda.

2

Fig (1): Sudan map showing the collection localities.

3 1- 2- Historical background of the Sudanese honeybees (Apis mellifera L.) morphometrics: It was understood that the Sudanese honeybee A. mellifera are not thoroughly examined by earlier workers but the morphometric studies, which were carried out in the last years revealed significant differences among the Sudanese honeybees. Many research workers have conducted several studies of the honeybee Apis mellifera L. races based on modern biometrics. These suggested systems aimed at characterizing and classifying the different geographical types of the Sudanese honeybees. In 1975, Ruttner designated the Sudanese honeybees as a separate group and gave them the name Apis mellifera nubica (Ruttner), thus he characterize it with very small size, covered with very short hairs, most slender abdomen of all African bees, the largest area of yellow coloration on abdomen, with a tongue length of 5.36 mm and a wing length of 7.98 mm. He stated that the Sudanese honeybee is similar to litorea, but clearly belongs to another group. Later, Elsarrag (1977) indicated that, the limited numbers of samples collected from Khartoum state and described by Ruttner (1975) do not perfectly represent the Sudanese honeybees. This was confirmed in (1978) by Rashad, and Elsarrag who illustrated that, several foreign races of honeybees were introduced to Khartoum province and honeybees in Khartoum were definitely hybridized with foreign blood. Comparing the data given by Elsarrag (1977) on Khartoum bees with those obtained for their hybrids F1 (Khartoum X unknown drones), Khartoum X Carniolan drones, they’re backcross reciprocals and their F2 hybrids. Later on Ruttner (1986), found that the A. mellifera nubica is completely scattered among samples of Yemen, Oman, Saudia Arabia, Somalia and Chad. Accordingly he nominated these bees as Apis mellifera yemenitica. To concludes this

4 taxon is incorporated into the subspecies yemenitica as a Sudanese population. In 1988, on the basis of multivariate morphometric analysis Mogga was able to distinguish three morphoclusters: Yemenitica, a small bee of the semi-desert zone; Sudanesis, a medium size bee of the forest and rich- savannah; and Bandasii, a large bee along the border between Sudan and Ethiopia. Matters became somewhat more complicated with the work of Elsarrag et al., (1992), who reported only two subspecies for the region: Sudanesis distributed throughout Sudan south of the Nubian Desert, and Nubica along the Sudan Ethiopia border. The altitude may form an ecological barrier to the spread of bee races in the tropics. This is most clear in East Africa, where the costal plain and the adjacent Kilimanjaro-Meru mountains are inhabited by different bee races, namely Apis mellifera litorea (Smith) and Apis mellifera monticola (Smith), respectively. However, in the Sudan the difference is mainly climatic and vegetative. The ecological barriers resulting from higher altitude are expected at Jebel Marra mountain (3042 m) in the west and the Imatong mountains (3187 m) in the south of Sudan. While there is continuous natural crossing among the Sudanese honeybees, it is still inconceivable to expect the different ecological zones to be inhabited by a single race of honeybees, having such a high capacity for specialized adaptation. Thousands years ago, within some areas in the Sudan honeybees were found preserved for trade purpose (cylindrical and boot hives). It did not receive real intensive investigation. Early seventies professor, Dr. Mohammed Saeed Ali Elsarrag and his colleagues, explored this field and presented commendable work on Sudanese honeybees records. Bees were imported to Khartoum province as early April 1928 (European honeybee races) by King from the apiary of the Egyptian Ministry of

5 Agriculture. These races included Italian, Cyprian, Carniolan as well as F1 Carnio-Egyptian honeybees. The latest large importation of bees, which received extensive distribution to east and northern regions, was in 1967. But generally, all the pure line colonies soon lost vigor and gradually dwindled, as they may not compete with the indigenous bees. The natural variability among honeybees in general and particularly the Sudanese honeybees which showed high degree of variables morphometrics were reported by Ruttner (1975); Elsarrag (1977); Rashad and Elsarrag (1978 and 1980); Saeed, (1981); Mohammed (1982), and Mogga (1988). The present investigation attempts to clarify in a large context the geographical variability and classification of the Sudanese honeybees (morphometrically and genetically). It is thought, the results of such a study would explain the geographical distribution of the Sudanese honeybees Which would presumably help any future plan of biological study of A. mellifera and subsequent selective breading to improve beekeeping in the Sudan. 1- 3- Apis florea (Fabricius) from Sudan: Some samples of Apis florea from four different towns were included in this study. This is a native honeybee species of the Asian sub-continent. It was first discovered in Khartoum at a garden near the international airport by Lord and Nagi (1985). It was believed that the initial colony might have entered the country as part of an air cargo. By January 1987, twenty additional colonies had been found, in a distance of 12 km from the origin. In 1988, morphometrical study was carried out by Mogga and Ruttner in an attempt to distinguished the origin of this bee and they stated that this A. florea was brought to Sudan from a country of western Asia and exactly from Pakistan. The current study, re-focus on the florea, is it the same samples of Lord and Nagi (1985)? Or there were some new colonies entered the country?

6 1- 4- Molecular biology development: In the last decade techniques for the measurement of genetic variation in honeybees at the DNA level have been developed and are proving to be extremely powerful probes for the analysis of genetic variation. The general technique is to look for restriction fragment length polymorphisms (RFLPs). Which have been applied to nuclear DNA (Hall 1986, 1988; Severson et al., 1988) as well as to mitochondrial DNA (Moritz et al., 1986; Smith, 1988, 1991; Smith and Brown, 1988, 1990; Smith et al., 1989, 1991), and by sequence analysis (Cornuet, Garnery & Solignac 1991. Garnery et al., 1991. Koulianos & Crozier 1991). Besides the potential discriminating power the mitochondrial DNA may provide valuables information on the phylogenetic links between subspecies or populations of the honeybees, data on mitochondrial DNA indicate that the subspecies Apis mellifera can be grouped into three lineages, A, M, and C, that are largely congruent with that ones put forward by Ruttner (1988), on the bases of detailed morphometrical studies. In turn, each lineage includes different types of variants of mtDNA (haplotypes) discovered after sequencing and/ or digestion with endonucleases. The variable regions best studied are the intergenic spacer between the tRNAleu and the cytochrome oxidase II gene (Corneut and Garnery 1991; Garnery et al., 1992, 1993, and 1995). This region shows a length polymorphism due to a variable number of copies (1-3) of a 192-196 bp sequence (Q) and second sequence P0 of 67 bp that may be partially deleted (sequence P with 52 bp), or completely lacking. In addition, this spacer shows restriction polymorphisms with Dra1, making it possible to characterize different haplotypes with low sampling and methodological efforts (Garnery et al., 1993). Based on the above mentioned genetic techniques, mainly mtDNA, a part of this study was directed to fined differences that characterize the

7 subspecies of the Sudanese honeybee and compare the results with the morphometric ones. 1- 5- The objectives of the present study: 1- The aim of this study was to investigate and classify the Sudanese honeybees, by using the Principal Component Analysis (the modern biometric method) together with genetic techniques, mainly mtDNA variation (to study the haplotype distribution of mtDNA in Sudan honeybees). Its was particularly felt necessary to conduct this work following morphometrical studies carried out by Ruttner (1975); Elsarrag (1977); Rashad and Elsarrag (1978 and 1980); Saeed (1981); Mohammed (1982); and Mogga (1988). All there results revealed significant differences among the Sudanese honeybees. 2- Establishing the taxonomical status of the Sudanese honeybees from different climate zones, in light of Ruttner (1986) classification of bees from the Sudan. 3- Investigating the geographical variants and distribution of the Sudanese honeybees. 4- To check up the origin of Apis florea, which was done before by Mogga (1988); following biometric modern techniques.

8 CHAPTER TWO: LITERATURE REVIEW 2-a- History of bees Evolution: The ancient of insect’s dates back to Upper Carboniferous period, which is about 350 million years ago. Some changes in insect fauna were noticed in the Permian, Mesozoic, Triassic and Jurassic periods that followed. Once flowering plants become established in Cretaceous period, many insects including specoid (predatory wasps) and ants with social behaviour were found associated with the plants (Winston, 1978). Bees diverged from a wasp ancestor approximately 100 million years ago (Michener, 1974; Michener and Grimaldi, 1988), when the angiosperms were becoming the dominant vegetation. The development of characteristics such as plumose hair, broadened hind legs for pollen collection, and mouth parts capable for ingesting nectar allowed an ancestral form to abandon predatory lifestyle and make flowering plants its primary food source (Raven and Axelrod, 1974). Due to pollen-collecting structures and habits, taxonomists place bees in their own super family, Apoidea (order: Hymenoptera), Culliney, (1983); Winston (1987), with 10 or 11 families (Michener, 1979. Michener and Greenberg, 1980), 700 genera (Malyshev, 1968), and 20.000 species (Michener, 1969) described. The honeybee is a long-tongued bee, classified in the family Apidae (Apinae: Apini) (reviewed in Raven and Axelord, 1974) along with the bumble bees (Bombinae: Bombini), the Orchid bees (Bombinae: Euglossini) and the sting less bees (Meliponinae) (Winston and Michener, 1977. Kimsey, 1984), as in appendix (C). Modern honeybees belong to a single genus, Apis, which contains at least seven species: A. andreniformis, A. cerana, A. dorsata, A. florea, A. koschevnikovi, A. labortiosa, and A. mellifera, (Alexander 1990; Otis, 1990 and Michener, 1990), appendix (D).

9 Honeybees are believed to have diverged as a separate genus about 40-50 mya (Michener and Grimaldi, 1988. Kelner-Pillaut, 1969. Zeuner et al., 1976). Apis dorsata and A. florea have been considered extend of ancestry. The two are believed to have separated as far back as the Oligocene, because dorsata-type bees are present in the early Miocene (22mya) and because these two species are believed to be the most distantly related members of the genus (Cockerell, 1908). Further Apis evolution is believed to have occurred during the Pliocene (2-6 mya) or early Pleistocene (2 mya) when the ability to thermo regulates its colonies resulted in an enormous increase in Apis distribution, including the colonization of Europe and Africa. Ecological and morphological diversification at the subspecies level followed, and the group spread rapidly to various climatic zones of the New World. The sister species A. cerana and A. mellifera are believed to be still in an early stage of speciation (Ruttner, 1988; Ruttner and Maul, 1983), having split immediately before or during the Pleistocene (1-2 mya). A considerable confusion with respect to honeybee taxonomy has existed since several decades. The earliest classification listed four species within the tribe Apini: A. mellifera Linnaeus (1758: 576), A. cerana Fabricius (1793: 327), A. dorsata Fabricius (1793: 328) and A. florea Fabricius (1787: 305). Various generic subdivisions have been created and discarded (Ashmead, 1904; Buttel-Reepen, 1906; Maa, 1953). Despite, these records more recent classification systems are also ambiguous. Ruttner (1988), using morphometric analysis, favors a return to the earlier system of four species. (A. florea, A. cerana, A. mellifera, A. dorsata). In contrast, Sakagami et al., (1980) favor expansion of the taxonomy to include A. laboriosa (a close relative of Dorsata), while Starr et al., (1987) favor the addition of laboriosa as well as A. breviligula. The designation of a new genus, Micrapis, has also be proposed (Wu and Kuang,

10 1986) in which Florea would be divided into two distinct species, Micrapis florea and Micrapis andreniformis. To date however, the most accepted classification system includes Florea, Andreniformis, Dorsata, Cerana, Mellifera, and Koschevnikovi (Alexander, 1990), while Laboriosa will probably be given species status in the near future (Otis, 1990). The genus Apis has been studied using morphological (Alexander, 1990. Smith, 1991), biogeographically (Kellner-Pillault, 1969. Smith, 1991), and molecular methods (Smith, 1991. Garney et al., 1991. Shepard and Berlocher, 1989; Smith, 1990). The current accepted review of Apis (Alexander, 1990) places the A. florea, A. andreniformis line as the most ancestral, giving rise to A. dorsata followed by A. mellifera, A. cerana, and finally A. koschevnikovi. 2- b- Origins of honey bees: The earliest Apini (Apidae) fossil specimens have been found in Baltic Amber from Eocene layers approximately 40 million years old. A fossilized honeybee comb has recently been discovered in Malaysia dating from the late Tertiary or Quaternary period suggesting an earlier origin of the genus (Stauffer, 1979). More recently, in January 12, 2006 Micheael, S. Engel. Stated that, a new fossil honeybee is described and figured from middle Miocene deposits of Iki Island, Japan. Apis lithohermaea (new species) is the largest fossil honeybee discovered, rivaling in size the modern giant honeybee, A. dorsata Fabricius. Thus it is the first fossil of the Dorsata species group recorded. Although the Dorsata group does not occur farther north than Tibet and southern China and in the Philippines in the Pacific, this lineage occurred near what is today southern Korea and Japan during the Miocene. This findings and the fact that fossil honeybee specimens are generally found with individuals grouped together suggest early evolution of social behavior in the Apini (Apidae).

11 Specimens found from the Oligocene, when considerable change occurred in external morphology show rapid evolution during the next 10 million years. Also, comparative biochemical studies of extent bees have indicated a greater degree of amino acid substitution in Apis mellifera compared with other bees, and therefore, a more rapid protein evolutionary rate in the honey bee lineage (Carlson and Brosemer, 1971, 1973). On the basis of morphological evidence however, there has been relatively little change in honeybees during the last 30 million years (Culliney, 1983), and the physical resemblance of fossil forms to modern worker bees suggests that complex social behavior had already developed by the Miocene, 27 million years ago. These bees have morphological characteristics, which partially point to the present day Meliponini and partially to the Apini. Eusociality in bees originated from solitary living and later sub social living. In the evolutionary path bumblebees are next to solitary bees. Meliponidae are eusocial bees, which evolved after bumblebees. Mass provisioning is noticed in Meliponidae. 2- c- Differentiation of Apinae: Honeybees belonging to Apidae are the most evolved bees, which have sting for defense, progressive provisioning, division of labor and reproductive casts. 2- c- i- Genus Apis :- The modern honey bees (Apidae: Apini) are all classified in only one genus, Apis that includes the following species: A- Single comb open-air nest includes: Apis florea, A. andreniformis (the dwarf honeybees) A. laboriosa and A. dorsata (the giant honeybee). B- Multiple-comb cavity nest includes: A. mellifera, A. koschevnikovi (the Saban honeybee), A. nuluensis (the Bornean mountain honeybees) A. nigrocincta, A. cerana (the Asian honeybee), appendix (D). But extinct

12 apini have been classified into their own genus Electrapis. (Manning, 1960; Zeuner and Manning, 1976; Culliney, 1983). The natural geographical distribution of the genus Apis shows its greatest species diversity in India and adjacent regions, with all of the species except Apis mellifera don’t found there; appendix (E) and appendix (F). Therefore, these regions probably constitute the area of origin and early evolution of the Apini (Doediker, Thakar, and Shaw, 1959; Michener, 1974; Doediker, 1978). A. mellifera is thought to have originated in the Africa tropics or sub tropics during the Tertiary period, migrating to western Asia and colder European climates somewhat later. 2- c- ii- Distribution of the genus Apis: Until modern times Apis was not found anywhere in the western hemisphere, Australia, or the pacific except for some of the continental islands such as Japan, Formosa, Philippines, and Indonesia (Michener, 1974). But movement of bees by European settlers for beekeeping has resulted in Apis mellifera now having worldwide distribution appendix (F and G) thus, some of the other species being more widespread in Asia as the dwarf honey bee A. florea. 2- d- The dwarf honeybee Apis florea and Apis andreniformis taxa: Many years ago, several subspecies, varieties, and nations of Apis florea were described, and specimens of Apis andreniformis identified as Apis florea (Maa, 1953). The accuracy of separation between those two honeybee species is often difficult to assess, because the worker bees are more sympatric (Otis, 1997). The confusion and mixing of the characters of those two species is evident in the monograph of Ruttner (1988). However, the distinction between the two species as unequivocal biological species nowadays, has been well established based on:

13 a- Drone morphology (Ruttner, 1975; Kuang and Li, 1985; Wu and Kuang, 1986, 1987; Ruttner, 1988; Wongsiri et al., 1990; Chen, 1993). b- Nest structure (Dung et al., 1996; Rinderer et al., 1996). c- Morphometrics (Rinderer et al., 1995). d- Allozyme polymorphism (Nunamaker et al., 1984; Li et al., 1986; Gan et al., 1991). e- Mitochondrial DNA sequence divergences (Smith, 1991; Nanork et al., 2001; Takahashi et al., unpublished data, cited in H. Randall Hepburn et al., 2005). f- Differences in the timing of mating flights (Rinderer et al., 1993). g- Several of the above mentioned differences contribute to complete reproductive isolation between the two species (Koeniger and Koeniger, 1991, 2000, 2001; Dung et al., 1996). The most accurate characters used for identification of Apis florea and Apis andreniformis are: The “thumb” of the bifurcated basitarsus of the hind leg of drones of A. florea is much longer than that of A. andreniformis (Ruttner, 1988); differences in the structure of the endophallus (Lavrekhin, 1935; Wongsiri et al., 1990; Koeniger, 1991); in worker bees, the jugal-vannal ratio of the hind wing of A. florea is greater (about 75) than that of A. andreniformis (about 65); and the cubital index of A. florea (about 3) is significantly less than in A. andreniformis (about 6). Abdominal tergite 2 of A. andreniformis is deeply punctuate, that of A. florea is not. The marginal setae on the hind tibiae of A. florea are usually entirely white, those of A. andreniformis dark- brown to blackish in sclerotized individuals. Permeating the older literature is the idea that abdominal tergites 1 and 2 of A. florea are reddish and other segments at least partially reddish, while those of A. andreniformis are uniformity black.

14 An inspection of several hundred workers from each of several different colonies of each species quickly demonstrates the extreme variation in pigmentation thus precluding this as a useful distinguishing trait, a point recognized rather long ago (Drory, 1888). Finally the combs of the two species are very different (Rinderer et al., 1996). 2- d- i- Distribution and description of A. florea (Fabricius): Apis florea is widespread species and its prevalence extends some 7000 km from its eastern-most extreme in Vietnam and southeastern China, across mainland Asia along and below the southern flanks of the Himalayas, westwards to the Plateau of Iran and southerly into Oman. This constitutes some 70 degrees of longitude (40°–110° East) and nearly 30 degrees of latitude (6°–34° North). Variations in altitude range from sea level to about 2000 m. A. florea has also been reported in historical times in Saudi Arabia and Sudan, and occurred on Java, Indonesia. Phenotypic variation among A. florea is not well understood. This species refers to bees of the plains up to 500 m. and, seasonal migrations occur up to 1500 m. (Muttoo, 1956). These species appear to maintain several ancestral characteristics of the genus Apis and show aspects of descendancy of the earliest honeybees in which Workers are small, approximately 7 mm in length, and colonies construct a single comb supported from branches, frequently at sites surrounded by dense vegetation (Seeley, and Akratanakul, 1982). In spite of its small size it competes with other Apis species (Koniger, 1976). The author of this work notified that in the Sudan seldom prevalence of Mellifera colonies around areas with abundant Florea species assembly around. Moreover, in the northern parts of the Sudan Florea replaced the abundance of Mellifera for instance. Their communicative dance occur on horizontal platform built on the top of the comb, and direction to flowers is indicated by straight runs toward the food source. Colonies tend to be small, less than 5000 individuals, and the workers are relatively docile (Michener,1974).

15 2- d-ii- Classification of Apis florea: Several regional univariate morphometric analysis on Apis florea have appeared over the last two centuries, but did not affected the taxonomy of this species. In the first multivariate morphometric analysis of A. florea, Ruttner (1988) had few samples from geographically non-contiguous regions. Although the data were insufficient for a comprehensive analysis, Ruttner (1988) demonstrated geographic variability and obtained three morphoclusters for A. florea. Tahmasebi et al. (2002) analyzed the A. florea of Iran and defined two morphoclusters from a geographical continuum. Combining their data with that of Ruttner (1988) and Mogga and Ruttner (1988), the latter reported three morphoclusters for all A. florea but a lack of geographical contiguity applies to this database as well. New collections of A. florea from Myanmar, Nepal, Cambodia, Thailand, Vietnam, Iran, Iraq, Afghanistan, Sri Lanka and Saudi Arabia have greatly augmented the database of A. florea over a geographical continuum of about 7000 km. The additional, new samples fill gaps to provide a population continuum over the full range of the natural distribution of A. florea for the first time. Following the previous studies on the morphometrics, classification and biography of Apis florea, Hepburn, et al., (2005) by using multivariate morphometric analysis studied 2923 individual worker bees from 184 colonies representing 103 localities across the full distributional area of Apis florea Fabricius 1787 from Vietnam and southeastern China to Iran and Oman (~7000 km), reporting that, comparisons of geographically separated A. florea populations resulting in morphoclusters that reflect sampling artifacts. These morphoclusters change with latitude but overlap when the full database is contained in the same principal component analysis. A cluster analysis based on Euclidean distances suggests degrees of affinity between various geographic groupings of A. florea. This species occupies a large area that includes rainforests, savannas, subtropical stepes, and semi deserts. The

16 seasonality of reproductive swarming is temporally continuous allowing gene flow throughout this panmictic species. Thus, Hepburn, et al., (2005), recommended that, A. florea is a single species that comprise of three discernible morphoclusters. In the northwestern-most bees comprise a morphocluster (1) that is statistically quite distinct from that to the southeast (3); but, they are not isolated, they are joined by large areas of intermediate forms (2) resulting in a continuous cline in morphometric traits within this panmictic species. The confirmation stated by Hepburn, et al., (2005) results, in studies of variation in mitochondrial DNA, Smith (1991) established that A. florea were homogeneous in Thailand and also in southern India but had diverged between the two countries. The homogeneity of mtDNA in A. florea from Thailand was subsequently confirmed (Nanork et al., 2001). Smith also observed that different mtDNA clusters occur in A. cerana from N. and S. India, paralleling the differences observed in A. florea. More recently, Takahashi et al. (unpublished data, cited in Hepburn, et al., 2005) proposed three distinct mtDNA lineages for A. florea from eastern Asia: (1) China/Myanmar; (2) Southeast Asia: Thailand, Vietnam, Cambodia, and part of China; (3) India. While there are no inconsistencies among these three studies, available information is insufficient to apply to the whole area of A. florea distribution. Similarly, available data on enzyme polymorphism in A. florea (Li et al., 1986; Sheppard and Berlocher, 1989; Gan et al., 1991) are likewise geographically limited precluding extrapolation to the whole A. florea population. The available genetic data are too regional in nature to be informative for the species as a whole. 2- d- iii- Historical background of Apis florea in Africa: The first colony of Apis florea in Africa was discovered in November 1985 in Khartoum. Mogga and Ruttner (1986) reported that, the colony was

17 cut from a lemon tree and brought to the beekeeping unit of the university of Khartoum for determination. On November 2, 1985 Siham Kamil of the university of Khartoum, William Lord (beekeeping specialist working for the Near East Foundation in Sudan) and J. Mogga were called to see a second colony of Apis florea established in another Khartoum garden, and hence more colonies were detected. The city of Khartoum is surrounded by desert, but in the centre of the city where these Apis florea have been found, the gardens are irrigated and there is flowering vegetation-obviously sufficient to sustain colonies of Apis florea. There are no colonies of Apis mellifera present at that time in Khartoum, other than those maintained by the University Bee Unit, which have to be provided with continual supplies of food and water to ensure their survival. That is the first record of Apis florea in Africa, and it raises the interesting questions of how long these bees have been present in Khartoum, and how they arrived there. All colonies found so far have been well established and local people confirm that the colonies have been present for at least sex months. The nearest place to Khartoum where Apis florea is recorded is Oman, some 2700 km to the East. A clue to their arrival may be the fact that all colonies found so far are near to Khartoum International Airport, which is also a base for aid lorries carrying supplies to and from Port Sudan town. It is thought that accidental or deliberate, human intervention is in Apis florea‚ introductory in Sudan. Bees disseminate or spread disease and pests and rendered hard to eradiate. Introduced bees compete with the native species, which was adversely affected.

18 2- e- Apis dorsata and A. laboriosa (the giant honey bees): There are two other closely related species of honeybees that construct nests consisting of a single comb in the open. They are known as the giant honeybees. These are large, feisty bees 17-19 mm in length, with 20,000 or more workers in a colony. Their nests are constructed high in trees or suspended from open cliff faces, and nests do not need to be concealed because of the workers aggressive nature. Nests are also frequently aggregated and colonies may migrate up and down mountains to take advantage of seasonal nectar sources. Communicative dances are more advanced collective in A. florea since they occurred in the vertical comb face, and the direction to flowers has to be translated by the workers from the vertical dance angle to the direction of the sun (Michener,1974). A. laboriosa is the larger of the two species, and its large size, dark colour, and longhair coat are probably adaptations for its high altitude Himalayan habitat (Sakagami, Matsumura, and Ito, 1980). 2- f- A. cerana and A. mellifera: A. cerana and A. mellifera, are medium size bees (10-11 mm), which generally build multiple comb nests inside cavities. Colonies of A. cerana are relatively small, 6000-7000 workers (Seeley and Akratanakul, 1982), but A. mellifera colonies can reach size of 100,000 or more individuals. These species are similar in morphology and behaviour and frequently considered distant races of the same species. However, Ruttner and Maul (1983), demonstrated that, although Cerana and Mellifera queens and drowns attempt to mate with each other, no offspring result, and instrumental insemination of Mellifera and Cerana queens with heterospecific semen revealed that the hybrid fertilized eggs would cease development at the blastula stage. These results indicate that Cerana and Mellifera are indeed separate species, although are closely related.

19 2- f- i- Races of Apis cerana: The eastern cavity-nesting honeybee, Apis cerana F., is widespread over Asia and occupies a distribution range extending from Afghanistan to China and from Japan to southern Indonesia (Ruttner, 1988). This species has been grouped based on morphometric analysis (Ruttner, 1988) in four subspecies with different distribution ranges: 2- f- i- (1)- Apis cerana cerana, from Afghanistan, Pakistan, north India, China and North Vietnam. 2- f- i- (2)- A. c. indica, from south India, Sri Lanka, Bangladesh, Burma, Malaysia, Indonesia and the Philippines. 2- f- i- (3)- A. c. japonica, from Japan. 2- f- I- (4)- A. c. himalaya, from central and east Himalayan Mountains (Smith, 1991). These subspecies include many populations some of which are geographically isolated, such as those in the Philippines archipelago. As on other oceanic islands, these populations may have undergone evolutionary changes giving rise to reproductively isolated populations. 2- f- ii- Races of Apis mellifera : The concept of a widespread race of honeybees in sub-Saharan Africa that is closely related to the European races can be found in most reviews of bee classification. Before 1958, however, opinions differed as to wither sub-Saharan bees where distinct species (Smith, 1865; Goetze, 1930, 1940; Maa, 1953) or an infraspecific form of A. mellifera (Buttel-Reepen, 1906; Enderlein, 1906; Ruttner, Kerr and Laidlaw, 1956). The modern bees classification started with Kerr and Portugal-Araüjo (1958), who applied the biological species concept to the problem and cited genetic crosses among European and African races to justify association to one species. All evidence up to date

20 supports their conclusion, but partial reproductive isolation may exist between some races (Kerr and Bueno, 1970). The endemic distribution of the honeybee Apis mellifera L. is generally considered to encompass Africa, Europe and pockets of western Asia appendix (F). Across this range, variation in behavior, morphology and genetic markers supports an evolutionary history of the species that includes differentiation into several major phylogenetic lineages (Ruttner, 1988; Cornuet and Garnery, 1991; Garnery et al., 1992). Based primarily on morphological characters, more than two dozen subspecies have been described within the lineages (Ruttner, 1992; Sheppard et al., 1997). These subspecies typically exhibit reduced gene flow with other such groups due to water, mountain or desert barriers and have been called “geographic races”, to reflect their adaptation to specific geographic areas (Ruttner, 1988). Based on morphological similarities and sequence divergence among described subspecies, the speciation event that produced Apis mellifera has been estimated to occur between 0.7 to 1.3 million years ago (Ruttner, 1988; Cornuet and Garnery, 1991; Arias and Sheppard, 1996). While Apis mellifera occupies a large geographic distribution allopatric from the rest of the genus, a number of the other Apis species reside sympatrically in Asia. Western Afghanistan is considered to be the eastern limit of A. mellifera, with the closest proximity between Apis mellifera and its congeners occurring somewhere in central Afghanistan (Ruttner, 1988). Consistent with the estimated antiquity of A. mellifera and the pale climate of the region separating A. cerana and A. mellifera, Ruttner postulated that an A. mellifera ancestor reached western Asia about one million years ago (ibid). While the relative youth of a million-year old A. mellifera lineage has been supported by molecular studies, cladogenesis of A. mellifera and A. cerana appears to have occurred at a much earlier time (6–9 mya) based on allozyme and DNA sequence differences (Sheppard and Berlocher, 1989; Cornuet and Garnery,

21 1991; Arias et al., 1996). The apparent discrepancy between the age of A. mellifera subspecies and the A. mellifera /A. cerana cladogenesis suggests that alternative hypotheses for the origin of A. mellifera should be considered. These include the possibility that A. cerana and A. mellifera are not related to each other as sister taxa, rather they may be so related to as yet un described or extinct species. Unfortunately, perhaps the least studied area within the distribution of Apis is in the region between the known distribution of Apis mellifera and Apis cerana. The natural habitat of the honeybee Apis mellifera extends from the southern tip of Africa through savannah, rain forest, desert, and the mild climate of the Mediterranean before reaching the limit of its range in northern Europe and southern Scandinavia appendix (F). With such a variety of habitats, climatic conditions, and floras, it is not surprising to find numerous subspecies (races) of honey bees (Apis mellifera L.), each with distinctive characteristic adapted to each region (Louveaux, 1966), appendix (G). Still, recognition of valid races has been difficult for a number of reasons. The Most important reason has been the movement of honey bees all over the world for beekeeping, which has changed the natural range of each race and resulted in considerable hybridisation of races. Selection by beekeepers characteristics useful in management may have altered the natural genotype of races as well, particularly in areas of intensive beekeeping where many feral colonies are descended from swarms, which have escaped from hives. Further difficulty may be that scientists and beekeepers don’t always use the same criteria for determining what a “race” is. Scientists tend to use morphometric measurements of such characteristics as wing veins, mouthparts, antennal length, and the size of certain body parts (Ruttner, 1975a; Daly and Balling, 1978; Ruttner, Tassencourt, and Luveaux, 1978), whereas beekeepers prefer characters like colour and behavioural traits, such as tendency to swarm, good honey

22 production, and gentleness. Finally, even within a single race there can be tremendous variation and where to divide races and determining what is “typical” for a race have always been somewhat subjective. Some general conclusions have emerged concerning the characteristics and places of origin for many honey bee races, and these have been summarized by Ruttner (1975b); Ruttner, Tassencourt, and Louveaux, (1978). He divided honeybee races into three distinctive groups: European, Oriental (Near Eastern), and African. Little is known about the Oriental races, and studies of many African regions are based on few specimens. The European races have been relatively well studied, and there appears to be more general agreement on these than on the African races. The brief descriptions below are based on Ruttner`s conclusions. 2- f- ii- (1)- European races: 2- f- ii- (1)- a- Apis mellifera mellifera L. (German dark bees): Originated in northern Europe and west- central Russia, probably extending down into the Iberian Peninsula. They are large bees, although with relatively short tongues (5.7-6.4 mm), and their common name is derived from their brown-black colour, with only a few lighter yellow spots on the abdomen. They tend to be nervous and aggressive but winter well in severs climates. Worker population expands slowly in the spring, although these bees where once popular for export around the world, their aggressive nature, poor spring and early summer performance, and difficult in working with flowers with long corollas such as clover have resulted in diminished use of Mellifera for beekeeping. 2- f- ii- (1)- b- A. mellifera ligustica Spin. (Italian bees): Originated in Italy, has been the most popular honeybee for beekeeping throughout the world. Although, somewhat smaller than A. mellifera Ligustica have relatively long tongues (6.3-6.6) mm) and abdomens with bright yellow bands. They tend to be docile, and colony populations build up

23 quickly in the spring and remain strong throughout the summer. They over winter with strong worker populations, although with high consumption of honey that causes some difficulties in northern latitudes. They also have a reputation as rapid comb builders and seem to initiate robbing of honey from other colonies more quickly than the other European races. 2- f- ii- (1)- c- A. mellifera carnica Pollman (Carniolin bees): Originated in the southern Austrian Alps, northern Yugoslavia, and the Danube Valley. They are of a similar size to Ligustica, but tend to be grey or brown in colour. These bees have also been popular for beekeeping, particularly with hobbyists because of their gentle disposition. They over winter in small colonies with low food consumption but developed quickly in the spring. They may not maintain this high population throughout the summer and seem to swarm more readily than the Italian bees. They are also slow to construct new comb. 2- f- ii- (1)- d- A. mellifera caucasica Gorb. (Caucasian bees): Evolved in the high valleys of the central Caucasus. They appear similar to Carnica, although perhaps with a more lead-grey colour. Although their behaviour is not as well known, they are considered gentle, slow to expand in the spring but capable of reaching adequate summer populations, and poor at over wintering because of susceptibility to the adult disease Nosema. They also used propolis extensively and are reported to have only a weak disposition to swarm. There are a number of other European races, either been insufficiently studied or grouped within one of the other European groups.. The Macedonian bee A. m. cecropia Kiesw., now appears to belong to Carnica race, but the position of the Russian steppe bee A. m. acervorum and the trans Caucasian A. m. remipes are not so clear.

24 2- f- ii- (2)- African races: The species name for the African bee is Apis mellifera Linnaeus (1758). Even this name has been subject to controversy. In 1761 Linneaus changed the name to Apis mellifica because the original name means, “honey carrier” rather than “honey maker” which he preferred. According to the principal of priority (Art. 23, International Code of Zoological Nomenclature, hereafter abbreviated as ICZN; Ride et al., 1985), the oldest available name is the valid name of a taxon. In spite of this long-standing rule, the junior name still appears in some European literature. According to a recent review of the infraspecific nomenclature of Apis mellifera (Engle, 1999). 10 valid subspecies are recognized in Africa: 2- f- ii- (2)- a- Apis mellifera intermissa “Tellian bees” (Maa, 1953): It is a North African race, found north of the Sahara from Libya to Morocco. Small, dark bee it is reputedly aggressive and swarms frequently, rearing over 100 queens in each swarming period. During droughts, over 80% of colonies may die; owing to intensive swarming colony numbers rebound when conditions improve ( Louveaux, cited in Ruttner, 1975b). 2- f- ii- (2)- b- A. melliferea lamarckii (Cockerell, 1906): It is Egyptian bees, formerly named A. m. fasciata, found in northeast Africa, primarily in Egypt a long the Nile valley. Like Intermissa, they rear numerous queens, with one colony recorded as rearing 368-queen cell and producing one small swarm with 30 queens. They appear to be more closely related to the bees of Central Africa, however, based on similarities of dance dialects between Lamarckii and Adansonii (von Frisch, 1967a). Thus Ruttner (1988), showed that Lamarckii belongs statistically, to the sub-Saharan rather than the North Africa races primarily because of its yellow pigmentation and small slender body.

25 2- f- ii- (2)- c- A. mellifera scutellata Lepeletier (East Africa bees): Though much of their range, were considered to be adansonii (Smith, 1961) until Rutner (1975b) proposed that these bees from the savannahs of central and equatorial East Africa and most of South Africa were actually a separate sub species, A. m. scutellata. This has created some confusion in 1956 were thought to be Adansonii, and all of the literature concerning these bees prior to the mid-seventies refers to them as Adansonii are indeed different subspecies, and also about which sub species was introduced to Brazil. Since Ruttner´s 1975 study, is the most recent and complete taxonomic evaluation of African bees, although the status of these subspecies is currently being re-evaluated and additional evidence may result in further changes. A. m. scutellata, (Lepeletier, 1836), in south Africa; it is a small bee with a relatively short tongue, it is highly aggressive, swarms and absconds frequently ,and is able to nest in a broad range of cites from cavities to open nests. 2- f- ii- (2)- d- A. mellifera adansonii (Latreille, 1804): Are found in West Africa and are conspicuously yellow in colour. They appear to be similar to Scutellata in many of their behaviours, but have not been well studied. 2- f- ii- (2)- e- A. mellifera monticola (Smith, 1961) “Mountain bees”: Found in southeastern Africa, are of interest because of the high altitude at which they are found in Tanzania, from 1500 to 3100 m. It is a large, dark, gentle race, with longer hairs than the other African bees. 2- f- ii- (2)- f- A. mellifera capensis (Escholtz, 1822): In South Africa “Cape bees” are only found at the tip of South Africa and are unique among Apis mellifera in the common occurrence of female-

26 producing laying workers. They are morphologically similar to Scutellata, but the degree of ovariole development and ability to lay parthenogenesis females regularly separates them from the Scutellata group. Other African races are found in limited areas of Africa. These may be morphometrically distinguishable from other races. Only a few specimens of each have been examined, and their biology has not been studied well enough to reach firm conclusions about their taxonomic status. These subspecies include: A. m. major Ruttner. A. m. sahariensis (Baldensperger, 1932) in Maghreb. A. m. jementica (Ruttner, 1976), in Northeast Africa. A. m. litorea (Smith, 1961), in south-easernt Africa. A. m. unicolor (Latreille, 1804) in Madagascar. Hepburn & Radlo. (1998) Showed that these 10 subspecies have clearly separated morpho-clusters and they precisely delineated their geographical distribution. They also indicated geographical zones of high morphological variance within and between colonies, thus identifying hybrid zones among subspecies. 2- f- ii- (3)- Oriental races : A number of oriental races from Turkey west to Iran have been proposed, including A. m. syriaca, A. m. anatolia and A. m. meda, which is similar to ligostica (Ruttner, Pourasghar, and Kauhausen, 1985). The relationships between these groups have not been studied. A thorough evaluation of the systematic of oriental honey bees might be important, since presumably transition forms between temperate and tropically evolved races might be found, possibly between A. mellifera and A. cerana.

27 2- f- ii- (4)- Northern and South American races: Although honeybees are not native to Northern or South America, both European and African races have been introduced in the last few hundred years. In North America racial lines of European origin have generally been maintained, although extensive mating between races and different selective criteria by queen breeders have undoubtedly modified some of the bee’s original characteristics. In South America the introduction of African bees in 1956 has resulted in the establishment and spread of A. m. scutellata throughout much of South and Central America. These bees will be referred to as “Africanized“ to differentiate them from bees studied in Africa, but they appear to be morphologically, behaviourally, and ecologically almost identical to Scutellata, and thus don‘t constitute a separate race. 2- g- Probability and identification of the honeybees: Identifications are made with various degree of assurance. When specimens of a species have unique and clearly defined structural or other characters, then the identifications are irrefutable within the context of the current classification. For example A. mellifera is distinguished structurally from its nearest relative, A. cerana. The latter species have two vein lets extending distal from the large basal cell of the hind wing rather than one as in A. mellifera. This and other key characters have proven to be consistent and species specific. Specimens of A. mellifera, therefore, can be conclusively identified (Daly, 1988). The geographical races and other distinctive populations of A. mellifera, however usually cannot be conclusively identified. They exhibit characters that may overlap to some degree or may grade imperceptibly into adjacent populations. Based on comparison of samples known to be typical of two or more populations, one can estimate how often an identification based on certain characters is likely to be correct. If quantative characters are used, statistical

28 analysis can provide a statement about the probability that a new sample is correctly identified. In this case, the identification is probable rather than conclusive. Identification of subspecies, geographical races, genetic hybrid swarms, ecotypes, or biotypes are usually of this nature. The accuracy of probable identifications depends entirely on how representative the initial samples are with respect to the total populations to be identified. The samples unit is usually a collection of bees from a colony and identification is based on pooled extracts or averages of characters of the collection. Some procedures can identify individual bees. Thus Daly (1991) demonstrated that, probability statements of identification must be interpreted within the context of the procedure. For example, with current methods in morphometrics, the statement that a colony collection is Africanized (Brazil bees race) at 0.7 or 70% probability also indicates the sample is European at 0.3 or 30% probability. The samples could be of normal Africanized bees or normal European bees, but it is more likely to be the former based on the analysis of known Africanized or European bees. The statement doesn‘t mean that the colony is composed of 70% Africanized bees and 30% European bees or that worker have 70% Africanized genes and 30% European genes. To make such statements, the procedure must be able to distinguish individuals or be based on genetic analysis, respectively. Furthermore the statement that a new sample is Africanized at 1.0 or 100% probability is not conclusive identification; it is still a probable identification based on the initial analysis of known Africanized and European bees. Daly (1991) showed that, all probable identifications carry the risk of actual misidentification. Any method (morphometric, biochemical, behavioural, genetics) that yields a probable samples are being identified, even a small risk becomes an important consideration in terms of the numbers of samples that may be misidentified.

29 2- h- Morphometrics: Morphometrics is the measurement and analysis of form. In biology the forms measured are morphological structures of organisms and analysis is usually by statistics. It is widely used in the study of insect life history, physiology, ecology, and systematic (Daly, 1985). Morphometric analysis is very important because it deals with variations in phenotypic characters which are induced both by genetic and environmental factors (Daly, 1991). To be useful in identification, the genetically determine racial differences between taxa must be large enough to provide distinguishing characters in spite of environmentally induced and local genetic variation with each taxon. Since morphological characters generally have a higher heritability than physiological characters (Soller and Bar Cohen, 1968; Falconer, 1989), the analysis of morphological characters is an important tool in the discrimination of different populations of honeybees. 2- h- i- Historical background of bee’s morphometrics: The first man who used the name Apis mellifera was Linnaeus in 1758, (Ruttner, 1988). Morphometric of bees Apis mellifera have been extensively analysed, especially during the first third of this century by Russians scientists, when they were searching for bees with a long proboscis for the efficient pollination of red clover and when apiculturists thought bees with longer tongues that could reach nectar in flowers with long corolla tubes (Daly, 1991, 1992). In 1992, Merril stated that the early studies of honeybees gave us much information on geographic variation and inheritance of morphometrics and environmental influences on morphometrics. The early exact morphometric measurement on honeybees was started by Koshevnikov (1900) and followed by Martynov (1901) and Kulagin (1906) (Alpatov, 1929). Ruttner (1978) wrote that in 1906 H. von Buttel- Reepen, made the first attempt to organize the multiplicity of honeybee

30 types in a rational manner, by using trinomial system. First the genus and species designation; Apis mellifica ( now A. mellifera) and geographical races or variety as third name. He also wrote that between 1925 and 1940 Alpatov and Geotze, provided a more exact bases for describing bee races by introducing biometrics-exact measurements. However, the early classification of honeybees was based on individual morphometric characters without statistical analysis. Since then various investigators at different times in different places continued to improve the gradual development of morphometric analysis of honey bees. The earliest apparently advanced morphometric measurement with adequate honeybee samples for statistical analysis was done by Chochlov (1916), Michailov (1924) and Alpatov (1929) who classified honeybee races based on morphometric measurements. Beside their attempts to classify honeybees, they were also able to recognize the effects of environmental factors on geographical variation of honeybees. They demonstrate the linear relationship between the honeybee tongue and hind leg length with latitude in the area along a line from the Baltic Sea to the Caucasus Mountains. In addition to the length of tongue, Alpatov included more characters like femur, tibia, metatarsus, length and width of wing and size of wax mirror. As Alpatov (1929) took more characters, he found trends in geographical variation, opposite to tongue length, which is decrease in body size from north to south in the plains of Russia. Alpatov (1933) studied geographical variation of the honeybees. He found that bees of southern localities as compared with northern ones had smaller body size, are larger number of hooks on the hind wings, a smaller relative surface of the first wax glands, relatively broader wings and legs. They were also more yellow coloured on the tergites with longer tongues. He further wrote that, the first important difference between A. m. adansonii and A. m. unicolor was the

31 longer tongue of the former; 5.820± 0.011 and shorter one of the later 5.587 ± 0.007. In summary of all the differences observed, Alpatov concluded that the relation of the yellow and dark African bees to each other, were the same as that of A. m. mellifera, the yellow Italian bee to A. m. mellifera, the dark European bee. Ruttner et al., (1952) concluded that, the colouration of the exoskeleton, which at one time constituted one of the bases for races taxonomy, has become, at least for European bee races of little importance. They identified three specific statistically constant groups of characters: (1) Hair characteristics. (i)colour of hair. (ii)length of over hairs on the 5th . tergite. (iii) width and the thickness of the tomenta (tomentum index). (2) Wing venation where the cubital index was most useful-defined as ratio of distances ‘a‘ determined by the point at which the nervous recurrence Nr; joins the lower vein of the third cubital cell ‘c‘ of the front wing. (3) Length of the proboscis. Thus Alpatov (1929), summarized that the following conditions have pronounced effect on the body size of worker bee: (1) the season of development. (2) the temperature of the surroundings during the pupal stage. (3) the size of the cell. (4) feeding by nurse bee of different age. (5) individuality of the colony. Alpatov noted that absolute body size and changes is some proportions could be related to reduced larval feeding. Grout (1937) demonstrated that, workers reared from enlarged brood cells were significantly larger than workers from normal brood cells. Recently, Eischen et al., (1982, 1983) reared worker larvae with different

32 numbers of nurse bees, finding positive correlations between the number of nurse bees and dry weight and life span of the progeny. Moreover, in 1929, Alpatov tried to indicate the rules of genetics and environment in the geographical variation of honeybees, though he reviewed the pioneering studies of Russian scientists on the effect of genetics and environment in geographical variation in bees. By transplanting colonies to new localities in Russia and observing European races in the United State, he concluded that, the races and the geographical variants within races has specific characters, including morphometrics, that were genetically determined. During the same early period in the United States, Kellogg and Bell (1904), Casteel and Phillips (1903), and Phillips (1929) produced major papers on bee biometry and showed that drones were more variable than workers. This feature of drones was later explained by Brueckner (1976) to be the consequence of reduced developmental homeostasis that arises from their homozygous genome. The genetic basis for seven morphometric characters in bees, was first established by Roberts (1961) who estimated heritabilities at 0.28 (number of hamuli) to 0.85 (wing width and tongue length). Morphometric characters are usually used for breeding and certification in Europe (Ruttner, 1988a). Wayne and Hendrickson (1973), wrote that, the source of data in biological taxonomy is essentially always the phenotypic; defined by Dobzhansky as “ what is perceived by observation of the organisms structures and functions, or what a living being appears to be to our sense organs unaided or assisted by various devices”. Mani (1973) stated that, whether a species occurs or not in a given biotop, depends upon: (i) historical grounds: the species must have had opportunities to reach the given biotop and having reached it, it must also have actually penterated it.

33 (ii) topographical grounds: geographical barreries to dispersal were an important factor. (iii) ecological grounds: the present state of affairs that can either exclude the existance or determine the density of population. Ruttner (1988) proved the presence of gradual variations in quantative characters of honeybees with geographical latitude changes along the east coast of the Atlantic Ocean from Scandinavia to the Cape of Good Hope. Alpatov (1929) also noted the difficulty of measuring the overall body size of honeybees and substituted single parts of the abdomen (sternites and tergites), which are closely correlated to the overall size of the honeybee worker (Ruttner, 1988). The early biometrics of honeybees was initially based exclusively on characters related to size of body parts. Goetze (1930, 1940, 1964) introduced two taxonomically significant quantative characters to Alpatove‘s list like indices of venation of the forewing and length of hairs of the abdominal tergites, which proved very efficient in discriminating European races. Louis (1963) also made an extensive study on the geographical variability on wing vein crossing points. Goncalves (1972) using the number of hamuli, categorized A. mellifera castes into two; low line with hamuli ranges of 11- 13 in queens and 12- 13 in drones, and high line with hamuli ranges of 21- 22 in queens and 11- 30 in drones. He further stated that, queens and drones equally influence the phenotypic development of the characteristics in descendants. Smith (1972) stated that, geographical barriers or environmental factors limited distribution of the African honeybees. He went on to say that all colour forms might be produced by the same queen, although there is a tendency towards the darker form and higher proportion of dark workers in the more mountainous areas at high altitudes.

34 Along with the development of morphometric measurements in 1930‘s statistical methods such as mean, standard deviation, coefficient of variation, ratio of differences and correlation coefficients were introduced to compare the geographical variability between European races of honeybees. Before 1964 morphometric analysis of honeybees entirely based on sample statistics and univariate methods. For the first time Dupraw (1964, 1965) used multivariate analysis to classify honeybee races. Moreover, Dupraw (1965) used discriminant function analysis of 15 variables, based on venation of 13 forewing angles and length and wide of the forewing and was able to establish cluster group of European, African and Asian honeybees, which are very similar to Ruttner´s (1988) geographical races of honeybees. 2- h- ii- Recent developed biometry of Apis mellifera. Wafa et al., (1965) presented mean values of biometrical measurements of the Egyptian honeybee A. m. lamarckii. Tongue length 5.65 ± 0.009 mm and range of 5.31- 5.83 mm. Fore wing length 8.36 ± 0.007 mm and range of 8.18- 8.56 mm. Fore wing width 2.84 ± 0.003 mm and range of 2.80- 2.93 mm. Cubital index 2.46 ± 0.016 mm and range of 2.23- 2.62 mm. No. of hooks on hind wing 21.10 ± 0.001 and range of 20.54- 21.98. Length of basitarsus 2.21 ± 0.005 mm and range of 2.17- 2.23 mm. Width of basitarsus 1.09 ± 0.005 mm and range of 1.04- 1.11 mm. Tomentum index 0.26 ± 0.022 mm and range of 0.18- 0.28 mm. Length of first wax gland 1.30 ± 0.003 mm and range of 1.23- 1.35 mm. Width of first wax glang 1.96 ± 0.004 mm and range of 1.90- 2.07 mm. They found that most differences between the means were non-significant and the coefficients of variation for all characters except cubital and tomentum indices were relatively low. They concluded that the group of bees studied were homogenous. Daly (1975) applied univariate and multivariate statistical analysis for identification of the Africanized bees of South America. By univariate 35 analysis, a series of overlapping for each character was found without clear separation. In multivariate analysis approach, he was able to identify three distinct groups: (i) bees from Africa and their hybrids. (ii) samples from Colombia and Venezuela. (iii) mixed bees of European origin from Costa Rica, Surinam and United States. Mitev et al., (1975) stated that, three varieties of adansonii existed in Guina Republic: A. m. adansonii (Latr.), A. m. monticola (Smith.) and A. m. litorea (Smith.). They further divided these honeybees by the colour of chitin to five groups. The following morphological measurements for the Guina bees were provided: length of forewing 8.457 ± 0.018 mm. and 8.551 ± 0.02 mm. in 1971, and 8.510 ± 0.017 mm. and 8.794 ± 0.016 mm. in 1972. Cubital index, 2.035 ± 0.043 in 1971, and 1.936 ± 0.029 mm. and 2.304 ± 0.038 in 1972. length of proboscis 5.445 ± 0.015 mm. and 5.468 ± 0.020 mm. in 1971, and 5.17 ± 0.02 mm. and 5.422 ± 0.038 mm. in 1972. Cornuet et al., (1975) classified ten worker samples according to the average values of four indices: (1) width of yellow band on the second tergite. (2) Length of hairs on the 5th tergite. (3) Width of tomentum on the 4th tergite. (4) length of proboscis. Then thirty workers samples by the average values of two wing indices: (5) and (6); ‚a‚ and ‚b‚ components of cubital index. Ruttner et al., (1978) established a standard biometry of honeybees based on 40 morphometric characters, by screening the less significant ones and induced more characters than Alpatov (1929), Goetze (1930, 1940, 1964), and DuPraw (1964, 1965), as in table ( 2).

36 Based on the standard biometry, Ruttner (1988) recognized 24 distinct taxonomic groups or geographical races of Apis mellifera, 7 in the Near East, 10 in Africa and 7 in North and southeast Europe. However, by applying a step-wise discriminate analysis procedure, Ruttner et al., (1978) ; Daly and Balling (1978) showed the possibilities of discriminating one race from another using fewer numbers of selected characters based on the region under investigation. Ruttner (1988), particularly suggested the possibilities of using one third of the original selected characters to discriminate African races of honeybees, however, he emphasised the inclusion of different categories of characters such as size, hairs, colour and wing venation. Crewe et al., (1994), also showed that 10 characters are fully adequate to discriminate honeybees of the southern Africa region. Moreover, Hepburn and Radolff (1996, 1997); Radolff and Hepburn (1997a,b) using 11 morphometric characters were able to classify African honeybee populations in distinct geographical races. Along with the development of morphometric measurements, the introduction of different multivariate techniques like principal components and factor analyses were used to detect clusters of colonies within populations (Ruttner et al., 1978; Ruttner, 1988). Step-wise discriminate analysis was used to confirm the separation of clusters, to detect the most discriminatory variables and to calculate the percentage of correctly classified colonies (Ruttner, 1988; Daly, 1992). To depict the distances between clusters, dendrograms and mahalanobis distances were introduced (Tomasson and Fresanaye, 1971; Cornuet et al., 1975; Cornuet and Garnery, 1991a, b; Daly, 1992). The introduction of different multivariate techniques proven to be powerful tools in the discrimination of honeybee races, ecotypes or strains within a race and

37 between genetic lines (Louis et al., 1968) and even to the level of F1 hybrids (Rinderer et al., 1990). 2- i- Historical background of African honeybees Classification: In 1979, Pager stated that, the earliest information on African bees was found on a copper plate engraved after a drawing by the Dutch author Pierter de Marees, first published in 1602. Baldensperger (1924) wrote that in Abyssinia (Ethiopia) east or in Senegambia west of Africa, both the Adansonii and the Intermissa were found; and that there were no really pure ones of the first one or the other. He further stated that, Frere Jules in Abyssinia, found two colonies of black and yellow bees in the same hive, daughters of the same mother. Baldensperger (1926) concluded that, the bee was an outcome of natural breeding and not been produced by beekeepers in their selection. Harris (1932) writing on the bees in Tanganyika Territory (Tanzania) listed three races; Apis m. unicolor (Latr.), Apis m. unicolor intermissa (Butt-Reap.) and Apis m. unicolor adansonii ( Latr.). He further stated that, Adansonii was the common bee of the Territory, while unicolor was restricted to the highlands of the North. Intermissa was only recorded from Mountain Kilimanjaro and the highlands in the vicinity of Lake Nyasa in the Southwest. This nomenclature was later proven incorrect by the work of Smith (1961), as follow: by reviewed the development of taxonomy of African bees he listed seven races: A. m. unicolor of Madagascar, Mouritus and Reunion Islands, A. fasciata of Egypt and A. m. adansonii of Senegal, all named by Latreillein 1804. A. m. scutellata of South Africa, A. m. nigritarium of Congo and A. m. caffra, which resembled A. m. capensis, were named in 1932 by Lepeletier. The A. m. capensis of Cape of Good Hope, named by Eschscholtzin 1922. A. m. fasciata was renamed in 1906 by Cockerell. In the same publication, Smith presented biometrical data on three African bee races. A. mellifera adansonii, which he stated inhabited

38 Africa from the Sahara desert in the north to the Kalahari and Karoo desert in the south: length of forewing 8.1 to 8.7 mm. with colony average of 8.42 to 8.51 mm. and tonguth 5.8- 5.9 mm. with mean of 5.85 mm. A. m. monticola, length of forewing 8.7- 9.3 mm. with colony average 9.0- 9.09 mm. and tongue length 5.9- 6.2 mm. with mean of 6.05 mm. A. m. litorea length of forewings 7.9- 8.4 mm. with colony average 8.13- 8.23 mm. and tongue length 5.7- 5.8 mm. with mean of 5.75 mm. The presence of morphological variability within the honeybees of Africa has been recognized since the 1920´s (Rotter, 1920, 1921; Baldensperger, 1922, 1924, 1932; Rueher, 1926; Giavarini, 1937). However, classification was mainly based on colour variations. Such descriptive classification continued until 1950´s (Aurelien, 1950; de Roeck, 1950; Dubois and Collart, 1950; Alber, 1952: Hassanein and Elbanby, 1956; Kaschhef, 1959). In this period all the honeybees of the sub-Saharan regions of Africa were considered as one taxon due to the presence of common behavioural characters and uniform yellow pigmentation. Kerr and Portugal-Araujo (1958) by means of genetic crossing confirmed that the honeybees of Africa, south of Sahara belong to the same species of Apis mellifera. For the first time they recognized five morphological distinct races: A. mellifera scutellata in all area south of Sahara ( except the Cape region), A. m. capensis in southwest parts of the Cape region, A. m. lamarckii in Egypt along the Nile Valley, A. m. unicolor in Madagascar and A. m. intermissa in northwest Africa between Libya and Morocco, which are geographically separated in different regions of the continent of Africa. Smith (1961) classified the honeybees of East Africa based on univariate analysis of morphometric measurements. Besides morphometric measurements he included behavioural and ecological characters and he recognized three races, A. m. scutellata, A. m. litorea and A. m. monticola.

39 DuPraw (1964, 1965) in his multivariate methods of honeybees morphometric study, tried to discriminate the honeybees of Africa as well on the bases of size of forewing and venation of wing angles. Thus DuPraw(1965) by using non-linear classifications, identified geographical variants native to Africa south of the Sahara which included both the yellow bees of Central Africa A. m. adansonii and the dark forms from Cape of Good Hope A. m. capensis and from Madgascar A. m. unicolor. He however, found it impossible to distinguish on the basis of wing variable, between Central Africa bees and those from Cape of Good Hope. The body colour might be a useful supplementary character for this purpose. The situation was not entirely simple, Dupraw concluded. In his comprehensive multivariate study of geographical variation in African honeybees Ruttner (1975,1988) recognized 10 races of A. m. honeybees in different regions of the continent: A. m. adanosii, A. m. lamarkii, A. m. litorea, A. m. jemenitica, A. m. monticola, A. m. scutellata, A. m. sahariensis, A. m. intermissa, A. m. unicolor, and A. m. capensis. He found that climate is one of the major isolating factors for races of honeybees in tropical Africa. However, he also noted that the race of honeybees (A. m. adanosii in west Africa) occurred in distinct ecological areas over vast geographical distances. He also observed the existents of different races of honeybees (A. m. jemenitica and A. m. adanosii) without substantial ecological differences. The work of Ruttner was based on a macro level sampling at continent level (Hepburn and Radloff, 1998) and, as Ruttner (1988) stated, his morphometric studies of all Africa honeybees doesn’t achieve a complete analysis of all variability of the huge continent nor does it indicate the borders for the identified geographical races. Recently, different authors tried to classify African races of honeybees based on morphometric, DNA and pheromone analyses and got a more

40 refined pictures for different regions of the continent (Saeed,1981; Mohamed; 1982, Mogga 1988; Meixner et al., 1989, 1994; Kassaye,1990; Smith et al., 1991; Lebdi-Grissa et al., 1991b; Cornuet and Garnery, 1991b; Kerr, 1992; Elsarrag et al., 1992; Crewe et al., 1994; Moritz et al., 1994; Hepbun et al., 1994; Garnery et al., 1995). Moreover, morphometric classification of honeybees at the continental level with large sample size across five major transects of the continent with different multivariate procedures was conducted by Hepburn and Radloff (1996, 1997, 1998) Radloff and Hepburn (1997a, b), Radloff et al., (1997) and Radloff et al., (1998). Along with the morphometric data they also used pheromone analysis to discriminate the cluster groups and were able to recognized a number of variations and ecotypes, which were not detected earlier (different morphoclusters of Scutellata, Jemenitica and Monticola). They also observed variations in races across the different ecological and climatologically zones of the continent and tried to locate zones of introgression and hyperdization of natural populations in different regions. However, the classification of honeybees into well-defined sub species still remains a controversial issue (Hepburn and Radloff, 1998). Nowadays there are three major thoughts reflected in presenting the observed geographical variations between populations of honeybees.. These are: (1) as sub species or geographical races (Ruttner 1988, 1992). (2) as adaptive ecotypes derived from adjacent populations ( Kerr 1992). (3) as product of asynchronous gene fluctuations within a contiguous met population for which the term “subspecies” may not be appropriate (Hepburn and Crewe,1991: Hepburn and Radllof 1998).

41 Due to high migration, absconding and swarming behaviour and consequential genetic mixing, lower molecular differentiation is observed among African subspecies (Franck et al., 2001). Moreover, the percentage of gene flow among honeybee population; lack of coherence between the distributing of the biological traits and morphometrically defined subspecies of Africa (Hepburn and Radloff,1998) and the existent of the same subspecies in distinct ecological areas and the occur rents of differences (Ruttner,1988) make the classification of honeybees of the continent more complex. Besides these general problems of classification of African honeybee populations, certain regions of the continent like the East North of Africa in general and Sudan in particular have not yet been adequately studied. 2- j- East Africa races of honeybees: According to Hepburn, et al., 1998 the honeybees reported from the east Africa region include: 2- j- i- A. m. scutellata: At mid-altitudes between 500-2400 m. in woodland and tall grass savannah of Kenya and Tanzania (Smith, 1961). Ruttner (1988) and Hepburn and Radloff (1998) indicated the wide distribution of A. m. scutellata from Ethiopia doun to South Africa including countries Such as Rwanda, Burundi, Uganda, Malawi and Zembabwe. 2- j- ii- A. m. monticola: Reported to occur in East Africa mountains areas between altitiudes of 2400-3200 m. (Smith, Ruttner, 1988). The distribution of this subspecies is thought to be unique and consist of disjunct areas, which are isolated by ecological barriers (Ruttner, 1988). This bee is reported to occur in Tanzania, Kenya, Burundi and Ethiopia.

42 2- j- iii- A. m. jementica: Asmall yellow bee, reported from hot and arid zones of East Africa (Ruttner, 1998). However, the distribution of this bee is large, extending 4500 km from Chad (Gadbin, et al., 1979), Sudan (Ruttner, 1975; Rashad and Elsarrag, 1980, Saeed, 1981; Mohamed, 1982; Mogga 1988), Somalia and Saudia Arabia (Ruttner, 1988), Yemen (Ruttner, 1975) up to Oman (Dutton et al., 1981). 2- j- iv- A. m. litorea: Reported to occur in the warm and humid coastal plains of Kenya and Tanzania at altitudes between 0-500 m. above sea level (Smith, 1961). This bee is replaced by A. m. jemenitica in the arid coastal plain of Somalia, but it extends south wards to the costal plains of Mozambique (Ruttner, 1988). 2- j- v- A. m. sudanesis Reported to occur in Sudan (Mogga, 1988) and in Ethiopia (Radloff and Hepburn, 1997a). 2- j- vi- A. m. bandasii Reported from Sudan (Mogga, 1988) and Ethiopia (Radloff and Hepburn, 1997a). 2- k- Classification of the Sudanese honeybee races:-

The first honeybee race reported to occur in Sudan was A. mellifera nubica, by Ruttner (1976). He described eleven African honeybee races including A. m. nubica (yemenitica), the Sudanese bee race. He gave the average morphometric as follows: Length T3 +T4 3.965 mm., Tongue length 5.45 mm., hindleg length 7.2 mm., length and width of forewing 8.219 mm., and 2.88 respectively, sternite 6(L/W) 89.31 mm., colour (cT3) 8.5, colour scutellum 7.28, cubital index 2.46 and wing venation B4 100.3 and J10 52.28. He also described A. m. yemenitica of southwest Arabian Peninsula as being similar to the nubica and litorea races. 43 Elsarrag (1977), studied morphometrics of worker honeybees from four provinces in Sudan. He recorded the following averages: tongue length 5.50 ± 0.019 mm., flagellum length 2.66 ± 0.006 mm., basitarsus length 2.20 ± 0.006 mm., basitarsus width 1.11 ± 0.003 mm., number of hair rows on inner surface of basitarsus 11.90 ± 0.014. forewing length 8.60 ± 0.014 mm., and width 3.02 ± 0.008 mm., cubital index 2.37 ± 0.019, number of hamuli 21.40 ± 0.009, length of T3+4 3.70 ± 0.008 mm., slenderness (L/W) 86.00 ± 0.003 and percentage of yellow colouration on T3 71.38 ± 0.003. He stated that the results showed high significant differences between provinces for all tewelve traits studied. He concluded that, this was an indication of the bee samples not belonging to the same race.

Saeed (1980) and Mohamed (1982), both investigated morphometrics of honeybee worker of Sudan. They both reported highly significant differences between the samples in the twelve traits studied.

Dutton et al., (1981) find two widely seperated A. mellifera populations in mountains of northern and southern Oman identical with each other and with A. m. yemenitica from north Yemen. They also established morphometrical relationship between them and the bees from Sudan. They measured the following parameters: length of hairs, tergites 3+4, hindleg, tongue, length and wide of the forewing, length of sternite 6, colour of tergite 3 and scutellum and wing venation angles A4, D7 and G18 of worker bees. They also measured drones parameters.

Mogga (1988) investigated on the taxonomy and the geographical variability of the Sudanese honeybee A. mellifera from four different geographical zones, namely semi-desert, poor savannah, rich savannah and forest zones. He presented the means and average values of biometrical measurements of the Sudanese bees as follow: The mean length of proboscis 4.89 ± 5.76 mm., with an average 5.33 mm.; femur 2.14 ± 2.42 mm.

44 average 2.28 mm.; tibia 2.75 ± 3.05mm. average 2.90mm.; metatarsus 1.75 ± 1.94 mm., average 1.84 mm.; total hindleg 6.72 ± 7.41 mm., average 7.06 mm.; width metatarsus 0.98 ± 1.13 mm., average 1.05 mm.; metatarsal index 54.01 ± 59.77mm., average 56.89mm.; forewing length7.88 ± 8.64 mm., average 8.26 mm., forewing width 2.69 ± 3.00 mm., average 2.85 mm.; cubital vein “a” length 3.75 ± 4.43 mm., average 4.09 mm., cubital vein “b” 1.55 ± 2.05 mm., average 1.80mm.; the cubital index a/b 1.93 ± 2.67 average 2.30; number of hooks in the hindwing 18.80 ± 22.40, average 20.60; The angles of the forewing: A4 30.60 ± 34.40 average 32.50; B4 95.20 ± 104.33, average 100.27; D7 97.13 ± 104.80, average 100.97; E9 17.87 ± 21.20. average 19.53; G18 93.13 ± 102.07, with an average 97.60; J10 49.60 ± 55.60, average 52.60; J16 89.60 ± 98.33, average 93.97; K19 75.47 ± 86.73, average 81.10; L13 12.33 ± 15.60, average 13.97; N23 85.40 ± 92.80, average 89.10; O26 33.40 ± 43.07, average 34.07; length tergite3, 1.90 ± 2.11 mm., average 2.01 mm.; length tergite4, 1.84 ± 2.07 mm., average 1.96 mm; the body length (T3+ T4), 3.75 ± 4.19 mm., average 3.95 mm.; length sternite3,2.34 ± 2.60 mm., average 2.47 mm.; length wax mirror on S3 0.97 ± 1.26mm., average 1.12 mm.; width of wax mirror on S3 1.84 ± 2.06 mm., average 1.95 mm.; distance between waxmirror in S3 0.23 ± 0.40 mm., average 0.32 mm.; length of hair in tergite 5 0.17 ± 0.22 mm., average 0.22 mm.; width of tomentum on tergite4 0.42 ± 0.75 mm., average 0.50 mm.; width of dark stripe on tergite 4 0.19 ± 0.35 mm., average 0.28 mm.; tomentum index 1.61 ± 2.56, average 2.08; length sternite6 2.19 ± 2.41 mm., average 2.30 mm.; width sternite 6 2.51 ± 2.82 mm., average 2.67 mm.; the abdominal slenderness (L/W) 82.63 ± 87.92, average 85.28; colouration on tergite2 5.80 ± 9.00, average 7.40; colouration on tergite3, 5.60 ± 9.00, average 7.30; colouration on tergite4 2.87 ± 6.80, average 4.83;

45 colouration on scutellum 3.80 ± 8.13, average 5.97; colouration of metatargum 0.00 ± 5.40, average 2.70; colouration on labrum1, 2.53 ± 5.07, average 3.80; and labrum2, 0.20 ± 2.87, average 1.53.

2- l- Some biological and ecological aspects of honeybees: One of the most highly polytypic species of all honeybee races is Apis mellifera, with large geographical variations not only in morphology but also in behaviour and physiology. Merrill (1922), noticed the geographic variations and the inheritance of morphological characters and the influence of environment of the morphological characters of honeybees. Winston et al., (1983), Stated that, one of the most striking aspects of honeybee biology was the variability found within and between races of A. mellifera. These included behavioural variants, morphological and physiological characteristics as colour, size, tongue length, defensive behaviour, amount of propolis and burr comb used, dialects of dance language and susceptibility to diseases. The biology of honeybees is variable within and between races of Apis mellifera is a result of adaptive responses to diverse ecological conditions like climate, patterns of resource abundance and predation pressure (Ruttner, 1988; Hepburn and Radloff, 1988). As a result, temperament, during the development stages, patterns of seasonal variability like brood rearing, migration, swarming and absconding of bees vary from one race to another. Moreover, geographical variations in behaviour, like orientation, defence, tendency towards propolis collection and utilisation, robbing and drifting, vary from race to race (Lauer and Lindauer, 1971, 1973; Ruttner, 1975). Understanding the biology and ecology of honey- bees of an area is very important not only for classification purposes but also for the efficient and profitable management of honeybees according to their biological behaviour and respective ecology. 46 2- l- i- Biological factors: Research and studying the biological characteristics of bees belonging to different races were much more important from the point of view of practical beekeeper. Ex: 2- l- i- (1)- Migration: Migration of honeybee colonies in the tropics considered as an evolutionary adaptation to escape darsh periods and also as a means of exploiting resource available in different ecological habitats at different times (Chandler, 1976; Castagne, 1983; Hepburn and Radloff, 1995). In a sense of evolution, the honeybees of tropical and temperate regions develop different means of surviving to escape harsh periods. The honeybees of the tropics adapt by migrating to resource rich areas, while the temperate bees survive by means of massive hoarding (Chandler, 1976). Seasonal migration of honeybees is considered as a unique characteristic of tropical honeybees (Ruttner, 1988). Honeybees of most sub- Saharan Africa are reported to migrate on a seasonal basis, following dry periods: A. m. yemenitica (Rashad an El-Sarrag, 1978; Peterson, 1985; Sawadogo, 1993; Woyke, 1993). A. m. litorea (Ntenga, 1976); A. m. capensis (Hepburn and Radloff, 1998). A. m. adansonii (Woyke,1989 ; Adjaloo, 1991 ; Adjare, 1990 ; Mutsaers,1991) ; A. m. scutellata (Smith, 1961; Chandler, 1976; Nightingale, 1993); A. m. monticola (Smith, 1961; Ntenga, 1976). However, the tendency of migration differs from ecotype to ecotype, from one ecological zone to another and also depends on responses to different stimuli (Chandler, 1976). Lack of food and water, over heating and fire were reported to be the major causes of migration of tropical Africa (Fletcher, 1978). Hepburn and Radloff (1998), indicate that the migration of African honeybees is not a completely fixed trait and can vary within and between

47 races depending on varying environmental conditions. In Kenya the migration of A. m. scutellata is reported to be facultative depending on the availability of resources (Nnightingale, 1983). Moreover, A. m. yemenitica is reported to not migrate in north Oman and Yemen but commonly migrate in Sudan and Chad ( Hepburn and Radloff, 1998). In tropical Africa migration is a common phenomenon and generally considered as an adaptation to the type of environment where bees live, losing colonies partially or totally as a result of migration ever season could be one of the discouraging factors in the development of beekeeping programs in the region. However, in Africa as a whole literature on the type or races of bees, which commonly migrate, the nature of ecology where much migration takes place, the periods and extent of migration, where the bees migrate to and the possible causes of migration are very poor. Such information are very valuable not only from the biological point of view but also from a practical beekeeping point of view to understand the real causes and associated factors contributing to the migration of bees. The information would be very important to formulate possible recommendations to minimise the migration of bees and to develop an appropriate system to manage bees with migratory behaviour. 2- l- i- (2)- Reproductive swarming: For temperate and Africa races of bees the general factor associated with reproductive swarming are considered the same (Smith, 1961; Anderson et al., 1983; Ruttner, 1983, 1988, 1992). Paterson (1977), observed that colonies of bees showing aggressive tendencies were avoided, thus leaving the more aggressive bees to propagate. This resulted in high incidence of swarming for the bees to maintain their natural population. However, unlike temperate races, African races invest more in a reproductive swarming than in a hoarding strategy (Hepburn and Radloff,

48 1998), which is believed to be a natural mechanism of balancing the loss of myriad colonies annually due to various hazards in their environment ( Fletcher,1978; Schneider and Blyther, 1988). The high reproductive swarming potential of African honeybees is believed to be attributed to many of their adaptive characters such as fecundity, short developmental period, high foraging efficiency and small body size (Fletcher, 1978). Moreover it seems more influenced by within nest conditions than genetic factors (Fletcher, 1978). Georges (1912) indicate that bees kept in small volume hives show more inclination to swarm than those in large hives. On the other hand, it was observed that relatively weak colonies in large volume box hives also swarm (Hepburn, 1993). Regardless of the population size of a colony and volume of hives space it was also observed that in some seasons and years most colonies tend to swarm while in some other seasons and years even highly overcrowded and strong colonies don’t show signs of swarming (Nuru and Dereje, 1999). However, the tendency to swarm still differs from race to another (Ruttner, 1975). The reproductive periods can also vary between races and may also be biphasic depending on the agro climatic conditions of the localities. In African honeybees the close correspondence of the phenology of reproductive swarming with the local climates, weather and the availability of forage has been reported (Hepburn, and Radloff, 1998). Moreover, reproductive swarm time variation within the same subspecies as the result of ecological and climate variations were observed for A. m. scutellata, A. m. adansonii, and A. m. capensis ( Hepburn and Jacot Guillarmod 1991; Hepburn and Radloff, 1988). The number of queens produced by a reproductive swarm colony of African honeybee also varies from 10-200 depending on race and agro ecological conditions (Hepburn and Radloff, 1998).

49 The information about African honeybees on reproductive swarming such as the period of swarming, the reproductive swarming tendency, and the number of swarms per colony is some what very week. Moreover, the types of honeybee populations with high swarming tendencies and the ecological conditions at which honeybees must swarm, are not sufficient. Such information is essential to develop an appropriate management system for bees with a high swarming tendency and in the long run the information would be important for selecting honeybee colonies with relatively high hoarding strategy rather than emphasising extreme brood rearing and subsequent swarming. 2- l- i- (3)- Seasonal cycles of honeybee colonies :- Every honeybee race is specifically adapted to its environment of origin through long periods of natural selection (Ruttner, 1976). The development of seasonal cycles according to changes in the environmental conditions is one of the fundamental evolutionary successes of honeybee colonies (Hepburn and Radloff, 1998). In tropical Africa, unlike temperate regions, seasonal cycles of the honeybee colony are governed by dry and wet seasons and the associated flowering patterns of honey plants (Hepburn and Radloff, 1998). The close association of seasonal cycles of honeybee colonies (brood rearing, population build-up, swarming and declining of population (dearth period) with environmental conditions (rainfall pattern, flowering time and dry periods) and the existence of time shifts in reproductive swarming within and between races in different climatic zones of Africa are well established (Hepburn and Radloff,1998). They were also observed significant correlations between brood rearing and swarming and the phenology of bee plants for some of major climatologically and ecological zones of Africa. Moreover, depending on the rainfall and flowering patterns of an area the honeybee colony cycle can be monophasic, short or long, depending on the

50 duration of flowering. Honeybee populations, which adapted well to synchronise to the change in the environmental condition would have a better chance to survive in the dearth periods and are believed to be more productive (Hepburn and Radloff,1998). Information on colony seasonal cycles is very important to develop a seasonal colony management calendar for different agro-ecological zones of the country. Moreover, the information would be valuable to select the types of honeybee populations, which easily synchronise to the change of environmental situations. Strains of honeybee populations with a fast colony build-up ability and fast honey storing tendency would be important to meet the conditions of most parts of the country where flowering periods are short. 2- l- i- (4)- Temperament :- Generally, all tropical African honeybees are considered as highly defensive or aggressive. The defensive nature of tropical African honeybees is believed to have arisen due to the extreme pressures of predators and disturbance in their ecology. However, in some areas of tropical Africa, beekeeping can be done without protective clothes (Clauss, 1983), while in some other areas the bees are reported to be swift and violent (Fletcher, 1978). Moreover, the presences of variation in the degree of aggressiveness among different races of tropical Africa honeybees and its association with genetic variations have often been reported (Chandler, 1976; Fletcher, 1978; Hepburn and Radloff, 1998). Colline et al., demonstrated the presents of significant variations in sting alarm pheromone levels between genetically and behaviourally different bees. Moreover, Fletecher (1978) indicated the existence of inter and intracolonial variations among different honeybees population.

51 Besides the genetic factors, temperament is also believed to be influenced largely by climate (Castagne, 1983). Temperature is considered the most important environmental factor that lowers the threshold responses of bees (Fletcher, 1978). As a result, honeybees believed to be more aggressive at hot, low altitude areas than cool- higher ones (Corner, 1985). On the other hand the same subspecies of honeybees A. mellifera. yemenitica is reported to be docile in very hot North Oman and North Yemen, but aggressive in Sudan and in Chad ( Rashad and El-sarrag, 1980; Dutton et al., 1980; Field, 1980; Gadbin, 1976). Moreover, the presence of aggressive and docile bees within apiary and its association with colony size and its variations from season to season and within the day with the changes of weather have also been reported (Hepburn and Radloff, 1998). In 1985, Kigatiira stated that, aggressiveness of tropical honeybees seems to be positively correlated within colony size. Information of the relative defensive behaviour of different honeybee populations of different ecological areas and factors associated with temperament variation is somewhat very poor for Sudan. Along with morphometric analysis of local honeybee populations, having information on the degree of the aggressiveness of different honeybee populations would be important to supplement the morphometric and genetic classification of honeybees of the area with behavioural characters. The information would be also important to select honeybee population with relatively gentle and reasonably manageable behaviour. 2- l- ii- Ecological factors:- Knowledge of the environmental conditions where the bees lived is obviously important. Different authors have indicated that, with some exceptions, the general morphometric and behavioural characters of honeybees have been found to be influenced by environmental factors (Daly,and Balling, 1978; Spivak et al., 1988; Coenuet and Garnery, 1991;

52 Daly and Morse, 1991; Nazzi, 1992). Alpatov (1929) and Ruttner (1988) demonstrated morphological variations across latitude. Ruttner (1988) stated that subspecies are a result of an adaptation in physiology and behaviour to give types of environment, which are associated with secondary variation in the external morphological characters. Murphy (1973) and Falconer (1989) also indicated the effects of environmental influence on morphological and behavioural characteristics of honeybees. Falconer (1989) documented that phenotypic characters in a population are the result of the combined effects of genotypic variance, environmental variances and gene-environment variances. Tsurata et al., (1989) and Spivak et al., (1990) found that the colour patterns of A. mellifera queens depend on the developmental temperature at the pupal stage; at lower temperature the pigmentation becomes dark. Again Szabo and Lefkovitch (1992) showed that the colour patterns of honeybees is less than 40% heritable, while more than 60% is attributed to environmental, gene-environment variation and error. Ruttner and Kauhausen (1985) stated that, in tropical Africa, significant geographical variability in honeybees occurs in spite of the absence of physical isolating barriers. The mechanism, which brings about isolation was the selective adaptation of races of bees to certain biotopes. In tropical Africa the presence of clear correlations between climatic conditions and morphometric characters like pigmentations, hair length and body size are well recognized (Ruttner, 1988). Gradual variation in morphological characters was also observed with change in altitude (Smith, 1961). Smith observed three distinct geographical races A. m. litorea, A. m. scutallata and A. m. monticola in Tanzania across a 300 km distance involving a change of altitude of 3000 m. from the coast to the rain forest of Mt kilimanjaro. Kigatiira (1984) stated that, each group of bees in Kenya were characterized

53 by a specific geographical distribution confined by natural barriers thus: A. m. monticola the large black mountain bee was found between 2400-3100 m. altitude and A. m. scutellata occupied the Acacia savannah plains. Mbaya (1985) concluded that, the morphology and behaviour of African bees in Kenya varied according to ecological zones. He went on to say that, special biological modifications of some characteristics of honeybees were necessary if they were to become adapted to certain ecological zone of Kenya. The effects of altitude and the existence of short distance ecoclines are also well recognized ( Mattu and Verma, 1984; Ruttner, 1988; Meixner et al., 1989, 1994). In a country like Sudan with contrasting and agro climatically features, it is very important to consider the environmental factors (like altitude, temperature and rainfall), along with the morphometric characterization of honeybees of the area to fully understand the variations, from the biographic context. Today in the classification of geographical races of honeybees, holistic approaches, including all possible data such as morphometric, biological, behavioural and ecological characters have become more important to understand the interaction of various factors and to obtain clear pictures of the geographical races of ecotypes of honeybees of a given area. Therefore, in this work an attempt was made to classify the honeybee populations of the area with greater sampling distance resolution by obtaining morphometric, , ecology and genetics characters of the honeybees of the Sudan. 2- m- Genetic diversity and Honeybees: Bees (Genus: Apinae; Hymenoptera) have differentiated into numerous geographic races or subspecies. The subspecies differed in various characteristics such as morphology, behaviour, ecology, sensitivity to diseases and biochemical components. Because most of those characteristics

54 have a genetic basis, the level of genetic diversity within the species as a whole may be considered high. Thus one of the honeybee DNA research goals is to find differences that characterized the subspecies. Understanding the structure of natural populations of honeybees and identifying the nature and dynamics of the selection processes that operate on them a wide range of suitable traits of high heritability is required. Until very recently morphometrical characters were the primary means of such studies, even though environmental effects might modify the expression of the genotype thus detracting from the discriminatory power of the morphometric technique. It has long been recognized that because morphometric characters are polygenic in origin, the genotype cannot be directly established through phenotypes. Nonetheless, the usefulness of morphometric studies, coupled to multivariate techniques of analysis remains well established (Ruttner 1988, 1992). Because of the limitations of morphometric analysis in the measurement of genetic diversity, the development of more accurate techniques such as, allozymic variation, nuclear DNA (micro satellites) and mitochondrial DNA, variations are needed. 2- m- i- Allozymes: Considering the evaluation and discrimination of more precisely bee’s genetic diversity, the second approach used after morphometrics is the study of allozymic variation. Chemo taxonomic methods have been used to differentiate among subspecies of Apis mellifera L. Sylvester (1976) and Cornuet (1979) found malate dehydrogenase (MDH) to be polymorphism in Apis, while Nunamaker & Wilson (1981) and Nunamaker et al., (1984) demonstrated that this enzyme fulfills the diagnostic requirements established by Ayala &

55 Powell (1972), and can be used to identify the African honeybee, A. mellifera scutellata Lepeletier. Thus, in the last century allozyme analysis became a widely used instrument to study honeybee racial relationships (Badino et al., 1983, 1984; Cornuet, 1982; Sheppard and Berlocher, 1984, 1985; Sheppard and Mcpheron, 1986). Allozymic variation has proven more amenable to the determination of genotypes in a population (Contel et al., 1977). Breeding experiments have established that the phenotypic expression of the allozymes is Mendelian and not susceptible to environmental effects. Of the forty-odd enzyme systems that have been investigated in honeybees, only seven are known to be polymorphism. Of these, cytoplasm malate dehydrogenase (MDH) has proven to be especially useful in the study of honeybee populations (Lobo et al., 1989; Cornuet and Garnery 1991a). Ndiritu et al., 1986, demonstrated that, little experimental work has been done on allozymes of African honeybees. Most of the knowledge accrued to date comes from the studies of Africanized bees in South and Central America (Sylvester, 1982; Del Lama et al., 1988, 1990; Sheppart et al., 1991), or comparisons of these bees with colonies obtained from South Africa (Nunamaker and Wilson, 1981), but no multi-locus study on a large collections of honeybees from Africa has been conducted. Thus, the allozymic variability of the honeybee in Africa and its potential to provide an additional source of data for taxonomic studies is still very poor. Investigation of the honeybees of Southern Africa have shown that, the bees classified as Capensis and Scutellata are morphometric and homozygous for the fast form of malate dehydrogenase, MDH-100 (Nunamarker & Wilson 1981; Sylvester 1982; Nunamarker et al., 1984, 1986; Sheppard & Huettel 1988).

56 Based on the fact that, the size of adult honeybees can be influenced by certain environmental factors (Alpatov 1929), and that bees which are reared in old comb (with relatively small diameter cells) are noticeably smaller in size than bees reared in new comb (with relatively larger cells). Nunamaker et al., (1986) tried to determine: 1- whether the morphological variation that results from brood development in combs with different cell size is a companied by enzyme variability. 2- whether the malate dehydrogenase (MDH) and nonspecific esterase (EST) enzyme of A. m. capensis Escholtz are the same in specimens collected from native populations in South Africa compared with specimens reared in an altered environment within a hive in Federal Republic of Germany where natural colonies do not exist. They studied isozymic uniformity in the presence of environmentally induced morphological variation in Apis mellifera capensis, they demonstrated that, Apis mellifera capensis Escholtz specimens from South Africa and Federal Republic of Germany exhibited identical electrophoretic patterns for malate dehydrogenase and non-specific esterase. However, adults from Federal Republic of Germany (which were reared in cells of large diameter) were significantly larger than workers from native populations in South Africa, and the same in three body measurements (body length, head capsule width, and length of forewing) as A. mellifera subspecies from Laramie, Woy. Certain environmental factors appear to have at least as much influence as genetic components in determining final body size of adult worker bees in A. m. capensi. Three different studies of allozymes of the honeybees of east Africa (Kenya) have been reported. In the first account, Ndiritu et al., (1986) showed that there is natural variation in the distribution of malate dehydrogenase alleles in the region.

57

58 Unfortunately, it is not quite possible to pinpoint the relevant subspecies involved. Subsequently, Sheppard and Huettel (1988) confirmed this malate hydrogenase polymorphism and established that there is variation in aconitase and esterase as well. On the basis of the distribution of the bee samples these results would apply to both Scutellata and Monticola (Sheppard &Huettel 1988). Thus, Meixner et al., (1994) in their attempts to determine allozymic variability of honeybees in Kenya and to evaluate differences in Allozyme patterns of the morphometrically classified Apis mellifera scutellata and A. Mellifera monticola. They studied; Morphological and allozyme variability in honeybees from Kenya. 43 samples of honeybees from three different regions in Kenya were analyzed morphometrically and surveyed for electrophoresis variation ( allozyme study) at five enzyme loci; Malate dehydrogenase “ MDH-1”; Phosphoglucomutase “ PGM-1”; Malic enzyme “ME2´”; Esterase “ES-1” and Hexokinase “HK” ( those enzymes are known to be polymorphism in European bees); they found that discriminate analysis of the morphometrical measurements classified the samples in three clusters; samples from above 2000 m as Apis mellifera monticola, samples from below as A. m. scutellata and the samples which collected from Ngong region (200 m.) has intermediate between the two mentioned clusters. Also all enzyme loci in the study were polymorphism, with Est. And HK showing highest degree of polymorphism. For ME, PGM and Est., new alleles were reported, thus the analysis of allozyme data by a Distance Wagner procedure of the provost genetic distances was performed so as to visualize similarities between population groups, using the BIOS’s program of Softwood and Selander (1981), resulted in two main clusters, consisting of A. m. monticola from Mt Elgon and Mt Kenya in one cluster and all other populations in a

59 second cluster. Savannah and mountain bees from Ngong formed a separate cluster. Meixner et al., (1994) concluded that the frequency distributions of the allozymes coupled to the morphometric discrimination, support the hypothesis that the currently disjunction Afromontane populations of Monticola derive from some common ancestor different from the Scutellata of the plains and Ngong Hills below them. Also, Sheppard (1988), reported that allozyme analysis might be more useful for African bees than New World genetic studies, as true for European bees. Shifting further north of Africa in the Maghreb, several important and in depth studies of allozymes have been performed. Firstly, Cornuet (1993), established that the frequency distributions were 0.90 for MDH-100 and only0.10 for MDH-80 among Intermissa populations. The point of Cornuet’s study was to investigate the relationship of the bees of North Africa and the Iberian Peninsula. As it turns out, frequency distribution differences for the alleles of MDH differ sharply in the African ‘A’ and western Mediterranean ‘M’ lineages (Cornuet & Garnery 1991a). In extension to the above studies, Smith et al., (1991), examined numerous enzyme loci for bees of the Iberian peninsula and found that all the bees of Southern Spain possess the same MDH alleles as occurs in the Intermissa of Morocco. Those of Mellifera extending down from the Pyrenees displace the same alleles in northern Spain. The direction of gene flow from Africa into Europe is indicated by the gradual dilution and eventual elimination of the Intermissa MDH alleles as one progress through Spain (Smith & Glenn 1994). Collectively, the three regions of Africa for which honeybee allozymes have been investigated actually provide four entirely different kinds of information, (Hepburn et al., 1998). Firstly, the frequency distributions of

60 the MDH allozymes are consistent with the separation of African and European lineages previously established in morphometric and mtDNA analyses (Ruttner 1988; Cornuet and Garnery 1991a,b). In southern Africa the morphometrically defined honeybee races of this region are allozymically indistinguishable (Nunamaker & Wilson 1981; Sheppard and Huettel 1988; Hepburn 1990). Entirely different results were obtained from East Africa. Meixner et al., (1994) were able to show that frequency differences in allozymes were significantly different among the Afromontane and lowland populations that correspond with the morphometrically defined clusters, Monticola and Scutellata respectively. Finally, in the Maghreb, the direction of gene flow from Intermissa across the Mediterranean Sea to the Iberica populations is also indicated by changing enzyme loci frequencies (Smith et al., 1991). 2- m- ii- Nuclear DNA: Nuclear DNA has also been used as genetic probe in honeybees, but most investigations have been directed to analyses of patriline structure (Blanchetot 1991; Moritz et al., 1991; Fondrk et al., 1993). Micro satellites loci are regions of DNA composed of short sequences repeated in tandem that can be amplified by a polymerase chain reaction (PCR) and then separated by electrophoresis (Hughes and Queller, 1993). Since micro satellites represent an abundant class of hyper variable markers, also in the honeybee (Estoup et al., 1993), they became particularly suitable tool in the analysis of an intracolonial genetic relationship and polyandry (multiple queen mating). Recently, Estoup et al., (1994) estimated 7 to 20 patirilines in Apis mellifera, with the average of 12.40 males per queen. The results for other Apis have also been reported. A high degree of polyandry in Apis dorsata

61 queens showed the average effective number of mating at 25.5 (Moritz et al., 1995), while relatively lower value of 5.56 in A. florea was reported by Oldroyd et al., (1995). The first application of nuclear DNA analysis to the population genetics of honeybees is one based on micro satellite variation (Estoup et al., 1995). The principal conclusions reached in this study were two fold. Firstly, the micro satellite analyses confirmed the distinctness of the three honeybee lineages proposed on morphometric (Ruttner 1988) and mtDNA grounds (Smith et al., 1991; Garnery et al., 1992). Secondly, Estoup et al., (1995) showed that the average heterozygosity and numbers of alleles were greater in subspecies of the ‘A’ lineage (Intermissa. Scutellata, Capensis) than in those of the ‘C’ and ‘M’ lineages. Micro satellite loci are extremely polymorphism in African populations compared to European honeybee populations and this has been interpreted as a consequence of larger effective population sizes in Africa (Estoup et al., 1995; McMichael & Hall, 1996; Franck et al., 1998). African populations would have been less influenced by quaternary ice episodes, which are considered to be the main cause of honeybee subspecies differentiation in Europe (Ruttner, 1988). Differences in allele frequencies have been found among Apis mellifera subspecies at several polymorphism loci (e.g. Mestriner and Contel et al., 1977; Martins et al., 1977; Sylvester 1982; Del Lama et al., 1985 1988; Spivak et al., 1988). These Polymorphisms are useful in the study of honeybee biography and population biology (e. g. Cornuet 1979; Sheppard and Berlocher 1984, 1985; Cornuet et al., 1986; Sheppard and Mcpheron 1986) and in the study of Africanized bees (Nunamaker and Wilson 1981, Sylvester 1982, Nunamaker et al., 1984).

62 Lionel Garney, Pierre Franck, Emmanuelle Baudry, Dominique Vautrin, Jean- Marie Cornuet and Michel Solignac (1998), investigated the genetic variability and differentiation of the west European honeybee populations (Apis mellifera mellifera & A. m. iberica), through micro satellite loci, they postulated that, these two subspecies are characterized by a lower genetic variability than most other studied subspecies and several tests are indicative of a recent increase of the population size. Moreover, the genetic profile is rather homogenous from southern Spain to Scandinavia. French populations are more or less introgressed (a few percent up to 57%) by genes from the north Mediterranean lineage that provides most of the imported queens. The inferred percentage of introgressed nuclear genes is generally well correlated with the proportional of alien mitochondrial deoxyribonucleic acid (mtDNA) haplotypes detected in the same populations. The level of introgression is the main source of genetic distances among populations. When introgressed genes are disregarded, however, population’s cluster is two groups, which correspond to both subspecies (Iberica and Mellifera), giving full support to the taxonomy of this lineage. Thus, Lionel Garnery et al., 1998. Noticed that, the use of micro satellites for population genetics studies is expanding exponentially. While these markers are very useful for the study of polymorphism in a variety of species. According to the hypothesis that, the use of nuclear DNA (micro satellite) and mitochondrial DNA markers often display discordant patterns of differentiation in the honeybees, Franck et al., 2001, in their previously mentioned study (genetic diversity of the honeybee in Africa: micro satellite and mitochondrial data) micro satellite marker section, they demonstrated that, furthermore, eight populations from Morocco, Guinea, Malawi and

63 South Africa have been assayed with six micro satellite loci and compared to a set of eight additional populations from Europe and the Middle East. The African populations display higher genetic variability than European populations at all micro satellite loci studied thus far. This suggests that African populations have larger effective sizes than European ones. According to their micro satellite allele frequencies, the eight African populations cluster together, but are divided in two subgroups. These are the populations from Morocco and those from the other African countries. De la Ru´a1, J. Galia´n1, Serrano1 and Moritz (2002), in their study Micro satellite analysis of non-migratory colonies of Apis mellifera iberica from southeastern Spain. Demonstrated that, forty-five unmanaged honeybee colonies from the southeast of the Iberian Peninsula (Apis mellifera iberica) were selected for analyzing their genetic structure using eight micro satellite loci. These colonies were not subjected to management for queen replacement, rearing or migratory movements and previous studies showed that they had mitochondrial DNA (mtDNA) of African origin. Six of the micro satellite loci show intermediate levels of polymorphism with a total number of alleles detected per locus ranging from 4 to 10. Micro satellite data relate these Iberian populations to the African A. m. intermissa, although the presence of some alleles and the observed heterozygosity are characteristic of the European A. m. mellifera, thus corroborating the postulated hybrid origin of A. m. iberica. The results suggest that no recent introgression from Africa has happened and that the populations of A. m. iberica are differentiated in many demes. Franck et al., (1998) have studied the distribution of eight micro satellite loci along a transect of honeybee populations from France to Morocco. Their results showed that there is no gradual modification or frequency cline of alleles across the Iberian populations. The three studied Spanish populations

64 (located in the Basque Country, Castilla and Andalucia) were similar to the French populations and did not show introgression of African alleles. These authors concluded that the Iberian Peninsula does not seem to be an intergradations zone between European and African subspecies. An additional, unexpected finding in the study of Estoup et al., (1995) was the discovery of certain predominantly European alleles of the micro satellite genes in the sahariensis region of southern Morocco. This results suggests the possible introgression of ‘M’ lineage nuclear genes through introduced drones, via-`a-vis results of the mtDNA analyses of the same region (Garnery et al., 1995). This example illustrates well the advantage of combining nuclear and mtDNA probes to genetically characterize honeybee populations. Similar observations elsewhere also confirmed that when two distinct lineages or populations come into contact the distribution of the nuclear and mt DNA markers may not coincide (Sheppard et al., 1991; Rinderer et al., 1991; Oldroyd et al., 1995; Garnery et al., 1995). 2- m- iii- Mitochondrial DNA: Although allozymes provide a straightforward genetic interpretation that is not susceptible to environmental effects, it is now well established that they exhibit a relatively low level of polymorphism in honeybees (Daly 1991). This is thought to be a consequence of the haplodiploidy system in the species (Pamilo & Crozier 1981). DNA analyses overcome the limitations of morphometrics characters and the relatively low polymorphism of allozymes. The ascendancy of the polymerase chain reaction (PCR) as a tool for bypassing cloning into micro-organisms to obtain sufficient amounts of specific DNA has made population comparisons or molecular biosystematics analyses of insects much more feasible today than a few years ago. Of particular utility are primer pairs that will amplify the same genes from a

65 wide variety of insects ( and possibly other organisms). An extensive, though no longer complete, list of primers from a number of different laboratories that amplify portions of insect mitochondrial DNA has been reported (Simon et al,. 1991). 2- n- Animals mitochondrial DNA: In animals many mitochondria are formed within the cells. During egg maturation, the large number of mitochondria that accumulate give rise to those of the embryo. Sperms have relatively few mitochondria, and in the animal tested, including vertebrates and insects, no paternal contribution of these to the progeny has been detected. Compared to the biparental transmission and recombination of the nuclear DNA, maternal inheritance of mtDNA allows genetic divergence to be more easily followed. Nucleotide changes in mtDNA accumulate at a constant rate, so that the times when ancestral lineages diverged can be estimated. Animals mt DNA generally, encodes 13 proteins, two ribosomal RNAs, and 22 tRNAs and has a region controlling replication and devoid of other known functions. One exception is the mtDNA of nematodes, in which the ATPase 8 gene has been lost ( Wolstenholme et al., 1987). Nematodes also present unusual tRNAs, all of which have been inferred to lack the TψC loop (Wolstenholme et al., 1987). 2- n- i- Size of animal mitochondrial DNA: In most animals species studied so far, the size of mtDNA is remarkably homogenous. The mitochondrial DNA (mtDNA) of animals (multicultural eukaryotes) are present in the extra nuclear cytoplasm and are relatively small circular DNA molecule, usually 16 kb (1 kilo base = 1000 nucleotide base pairs = 1kb) long, and has been most commonly used to determine genetic relationship among organisms (Brown, 1985; Wilson et al., 1985; Avise, 1986; Avise et al., 1987; Moritz et al., 1987), although it ranges from

66 14 kb to 39 kb in length (Snyder et al., 1987; Wolstenholme et al., 1987). Obviously this is due first to its high conservation in gene contents, but also to the absence or shortness of intergenic sequences and to the lack of introns ( Attardi 1985; Brown 1985). Most of the large-scale size variation lies in a single region of the molecule, (the control region), which contain most of the regulatory sequences. As a rule, this variation is related to the existence of repeated sequences, the different number of units of repetition being responsible for the length variability observed within several species and between related species, (Cnemidophorus sp., Dens-More, Wright and Brown (1985); Drosophila sp., Solignac, Monnerot and Mounolou (1986); Triturus cristatus, Wallis 1987; Grylus sp., Rand and Harrison (1989); Acipenser transmontanus, Buroker et al., (1990); Oryctolagus cuniculus, Mignotte et al., (1990)”. The occurrence of repeated sequence also accounts for the unusual length of the mitochondrial genome in some animal species: 26 kb for the nematode Romanomermis culicivorax (Powers, Platzer and Hyman 1986), 42 kb for the scallop Placopecten magellanicus (Snyder et al., 1987; La Roche et al., 1990), and 36 kb for several species of the bark weevil Piossodes (Boyce, Zwick and Aquardo 1989). The mtDNA of one insect, Drosophila yakuba, has been sequenced in full (Clary and Wolstenholme 1985. ), and partial sequences are known from D. virilis (Clary and Wolstenholme 1987), D. melanogaster (de Bruijn 1983; Satta et al. 1987; Garesse 1988), the mosquito Aedes albopictus (HsuChen et al., 1984a, 19843), the locust Locusta migratoria (McCracken et al., 1987; Uhlenbusch et al., 1987), and the honeybee Apis mellifera (Vlasak et al., 1987 ) . These studies have indicated that the order of major genes has changed between phyla, that tRNA positions change between members of

67 the same insect order (Drosophila and Aedes), and that insect mtDNAs are very A+T rich. In vertebrates, mtDNA has long been regarded as evolving much more rapidly than single-copy nuclear DNA (scDNA), as judged on the basis of mtDNA sequence comparisons and scnDNA heteroduplex formation (Brown et al., 1979; Moritz et al., 1987). By contrast, studies on echinoids (Vawter and Brown 1986) and Drosophila (Powell et al., 1986) indicate roughly similar rates of divergence for scnDNA and mtDNA. Vawter and Brown ( 1986) and Moritz et al., ( 1987) attribute these relative differences in evolutionary rates between mtDNA and scnDNA to differences in scnDNA rates, suggesting that the mtDNA rates are relatively invariant. Nuclear genes have indeed been found to vary significantly between various groups in evolutionary rates ( Wu and Li 1985; B&ten 1986; Li and Tanimura 1987; Lake 1988; also see Ochman and Wilson 1987). 2- n- ii- Apis mellifera L. Mitochondrial DNA: Honeybees (Apis mellifera L.) mt DNA was found to be, between 16.5 and 17 kb length, ( Smith and Brown 1988). This range is explained by length variability in at least two regions (Smith and Brown 1990): one of them, as usual, is the control region and another one, where size differences are larger (ca. 450 bp between the longest and shortest types), is the COI- COII junction. This region has been recently sequenced by Crozier, Crozier and Mackinlay (1989)., between the tRNAleu and the COII genes, they found an unassigned sequence of 194 bp. Where only five nucleotides are present in the Drosophila yakuba mtDNA (Clary and Wolsten-Holme 1985). The sequence obtained by Crozier, Crozier and Mackinlay (1989) belongs to the shortest type found in Apis mellifra. Corneut, et al., (1991),

68 sequenced a domain encompassing this unassigned sequence in a genome of the largest size class in Apis mellifera. In addition, the corresponding domain has been sequenced in three other Apis species, and in two related species Bombus lucorum and Xylocopa violacea (Anthophoridae). Thus far, mitochondrial DNA has enjoyed most attention in the analysis of honeybees for several important reasons: mitochondrial DNA of honeybee A. mellifera L. is a small circular, double-stranded DNA molecule of between 16 500 and 17 000 base pairs ( Smith and Brown 1988) and whose sequence has been determine (Crozier & Crozier 1993). The gene content is highly conserved, and there are many identical copies of the chromosome in each cell. The mitochondrial genome is substantially smaller than the nuclear genome and can be readily studied as an entity (Smith 1991). Thus, although honeybees appear to have relatively low levels of allozyme variability (e.g. Sheppard and Berlocher 1984, 1985), the level of variation in their mtDNA is well within the rate found in other species (Avise and Lansman 1983). Second, although of fixed differences in allozymes have been found among honeybee species, preliminary studies of the mtDNAs of European and African subspecies (Smith 1988) indicate that at least some have unique cleavage site patterns. Thus, because all the offspring of a queen inherit the same mtDNA, large quantities of mtDNA from a single source can be prepared by pooling tissues from hive mates. The mitochondrial DNA molecule of honeybees has been studied within and between populations (Smith 1988; Cornuet & Gernery 1991b; Garnery et al., 1992; Moritz 1994; Moritz et al., 1994). Mitochondrial DNA is a particularly valuables probe because of its maternal inheritance, which means that all offspring of a single queen will have the identical mtDNA molecule (Meusel & Moritz 1992). Also, because

69 of the strict maternal inheritance, honeybees of hybrid origin do not carry a mixture of mtDNA's, they show only the pattern of their queen (Cornuet and Garnery 1991a,b; Smith 1991), making the molecule particularly powerful in the study of hybrid zones (Smith et al., 1989; Moritz et al., 1994, 1998). Similarly, the genetic details of swarming bees can be resolved with the mtDNA, so it is a potential means for studying colonization processes (Garnery et al., 1992). Mitochondrial DNA variability has also been successfully used in the analysis of phylogeographic variations (Smith et al.,1991; Garnery et al., 1992). Many studies have used mitochondrial DNA polymorphisms to identify mitochondrial lineages within a species, and to make inferences about the phylogeny and biography of the organisms carrying them. The honeybees “Apis” are a particularly rich group for study of both biogeography and mitochondrial DNA variation. The biography of honeybees is complex and moderately well documented (Ruttner, 1968, 1988, and 1992). In addition, the mitochondrial genomes of Apis species include an apparently unique non-coding intergenic sequence, which provides a source of rapidly evolving characters (Cornuet et al., 1991). MtDNA characters have been used to study phylogeny and biogeography of A. mellifera subspecies (Garnery et al., 1991, 1992; Smith 1991b; Smith et al., 1991), and the pattern of gene flow between introduced European and African honeybees in the New World (Hall and Muralidharan, 1989; Hall and Smith, 1991; Sheppard et al., 1991; Smith et al., 1989). The novel non – coding region of Apis mtDNA is of particular interest for studies of infra–specific biogeography and phylogeny. Early studies of restriction polymorphisms in honeybees showed a region of size variations in the mitochondrial genome. This region was small and uniform in size in

70 east European A. mellifera subspecies and larger and variable in west European and African subspecies. The size variations in the west European and African A. mellifera restriction fragments occurred in discrete unites that suggested a tandemly repeated element ( Hall and Smith, 1991; Smith, 1991a; Smith and Brown, 1988, 1990; Smith et al., 1989). Cornuet et al., 1991, subsequently sequenced this region of the mitochondrial genome in several A. mellifera subspecies and in one individual each of A. cerana, Dorsata and Florea. All four of the Apis species were examined (as well as A. koschevnikovi and A. anderniformis: Smith, unpub. Data; cited in Deborah, et al., 1996) and all have a non – coding intergenic region, which is A+T rich. This region is flanked by the cytochrome oxidase 1 (CO1) and Lucien tRNA genes on the 5' end and the cytochrome oxidase 11 gene (COII) on the 3' end. In A. mellifera this region consists of two units, a P unit of 54 bp which is 100% A+T, and a more complex “Q” unit of 196 bp (93.4% A+T), which is similar in sequence to the 3' end of the COI gene ( “Q1” ), the Lucien tRNA gene (the “Q2” portion) and the “P” sequence (“Q3”; Cornuet et al., 1991). The size class found in east Mediterranean A. mellifera corresponds to a single Q sequence; the different size classes found in west European and African A. mellifera correspond to: PQ, PQQ, and PQQQ. Cornuet et al., (1991) suggest that these repeats arose as a result of a duplication of portions of the CO1 and the adjacent Lucien tRNA gene, followed by additional tandem duplications. The P and Q sequences of A. mellifera can be folded into hairpin and cloverleaf loop structures (Cornuet et al., 1991). This intergenic sequence is smaller and less complex in A. cerana and other Apis. species. The intergenic region of all Apis species share a pair of short sequences (“stem sequences”) at the beginning and end of the non – coding region which appear to be capable of base – pairing with one another

71 (Cornuet et al., 1991); Smith, unpubl. Data; cited in Deborah, et al., 1996), and which are identical to the stem sequences of Mellifera P and Q3 sequences. Honeybees have been surveyed for mitochondrial DNA (Moritz et al., 1986 Smith 1988; Smith and Brown 1988) and nuclear (Hall 1986) RFLPs. In an attempt to extend the knowledge of honeybee mtDNA sequence, providing information useful for general evolutionary studies on insects; R. H. Oozier, Y. C. Oozier, and A. G. Mackinlay (1989), studied the CO-I and CO-II Region of Honeybee Mitochondrial DNA: Evidence for Variation in Insect Mitochondrial Evolutionary Rates. They demonstrated that, the sequence of a region of honeybee (Apis mellifera ligustica) mitochondrial DNA, which contains the genes for cytochrome C oxidase subunits I and II (CO-I and CO-II) and inferred genes for tRNA Asp, tRNA Leu UUR, tRNA Lys, and tRNA Trp, is presented. The region includes the segment previously identified as incurring a length increase in some other bee strains, including Africanized bees. The sequence information of this study and of that by Vlasak et al., shows that several shifts of tRNA genes have occurred between Apis and Drosophila, but shifts of other kinds of genes have yet to be demonstrated. The CO-I and CO-II gene sequences are both more A+T rich than the corresponding Drosophila genes. Parsimony analyses using the mouse and Xenopus sequences as out groups show significantly more amino acid substitutions on the branch to Apis ( 120) than on that to Drosophila (44) indicating a difference in the long-term evolutionary rates of Hymenopteran and Dipteran mtDNA. Thus this study indicate that, the rates of mtDNA evolution have differed greatly between these lineages. Leslie G. Willis et al., (1992), studied phylogenetic relationships in the honeybee (Genus Apis) as determined by the sequence of the cytochrome

72 oxidase II Region of Mitochondrial DNA, from which they determined the complete nucleotide sequence of the mitochondrial cytochrom oxidase II (COII) gene for five species of the honeybee (Genus: Apis): A. andreniformis, A. cerana, A. dorsata, A. florea, and A. koschevnikovi, these were then compared to the known sequence of the A. mellifera gene from Crozier et al., (1989), and the wasp Excristes roborator, Liu and Beckenbach, (1992). Phylogenetic relationships were derived using the parsimony methods DNAPARS and PROTPARS of Felsensrein (PHYLIP Manual Version 3.4. University Herbarium, Univ. Of California, Berkeley). The results suggest that A. dorsata is the most ancestral species, followed by the branching of A. florea /A. andreniformis and A. koschevnikovi, and then A. mellifera and A. cerana. This inference differs from the currently accepted view that considers the A. florea / A. andreniformis line to be the most ancestral. The genus Apis has been studied using morphological ( Alexander, 1990. Smith, 1991), biogeographically (Kellner-Pillault, 1969. Smith, 1991), and molecular methods ( Smith, 1991. Garney et al., 1991. Shepard and Berlocher, 1989. Smith, 1990). The currently accepted view of Apis (Alexander, 1990) places the A. florea, A. andreniformis line as the most ancestral, giving rise to A. dorsata followed by A. mellifera, A. cerana, and finally A. koschevnikovi. Under the following title: A simple test using restricted PCR-amplified mitochondrial DNA to study the genetic structure of Apis mellifera L., Garnery, M. Solignac, G. Celebrano and J.-M. Cornet, (1993), demonstrated that, the COI-COII intergenic region of Apis mellifera mtDNA contains an important length polymorphism based on a variable number of copies of a 192- 196 bp sequence (Q) and the complete or partial detection of 67 bp sequence (P0). This length variability has been combined with a restriction site polymorphism to produce a rapid and simple test for the characterization

73 of mtDNA haplotypes. This test includes the amplification by the polymerase chain reaction of the COI-COII region followed by Dra1, restriction of the amplified fragment. In a survey of 302 colonies belong to 12 subspecies, 21 different haplotypes have been unambiguously allocated to one of the 3-mtDNA lineages of the species. Although all colonies of lineage C exhibit the same pattern (C ), each one of lineage A and M present up to 10 different haplotypes, opening the way to studies on the genetic structure and the evolution of a large fraction of the species. This test also differentiates Southern Spanish and South Africa colonies, which can bee of great interest for the Africanized bee problem. Although they propose an even more discriminatory test which requires the amplification and restriction by a single enzyme of only one PCR fragment. This fragment contains the COI- COII intergenic region, which exhibits at least 7 length variants which can be explained by the combination of 3 related sequences P0 (67bp), P (54bp) and Q (192- 196bp): P0, P0QQ, P0QQQ, PQ, PQQ, PQQQ and Q. The three-mtDNA lineages corresponding to the aforementioned three branches are characterized by P0 (A), P (M) and neither P0 nor P (C). The examination of the available COI-COII sequences led to the conclusion that Dra1, the recognition site of which is TTTAAA, should show a significant amount of polymorphism (Grnery, et al., 1993). Genetic diversity of the west European honeybee (Apis mellifera mellifera and A. m. iberica). I. Mitochondrial DNA. under this title Lionel Garney, Pierre Franck, Emmanuelle Baudry, Dominique Vautrin, Jean- Marie Cornuet and Michel Solignac (1998), studied the variability of mitochondrial deoxyribonucleic acid ( mtDNA) in 973 colonies from 23 populations of the West European honeybees ( lineages M) using restriction profiles of a polymerase chain reaction (PCR) amplified DNA fragment of the COI- COII intergenic region. Although populations are almost always

74 introgressed by two other mtDNA lineages (A and C), results confirmed that, the original haplotypes in Western Europe are those of mtDNA lineage M. Iberian populations ( Apis mellifera iberica) are characterized by a extended cline between haplotypes A and M, the former being almost fixed in South Spain and Portugal, and the latter almost pure in northeastern populations. This introgression is most likely attributable to humans and is probably ancient. French populations (A. m. mellifera) exhibit various levels of introgression by the C mt DNA lineage. Introgression is rather low in regions with a dominance of a mature bee keeping while it reaches very high values in regions where professional beekeepers regularly import foreign queens (mainly A. m. ligustica and A. m. caucasica). When discarding introgressed haplotypes, French populations group in two clusters, one for the northeastern part of France and the other one for all other populations, including Swedish and northeastern Spanish populations. Thus, Lionel Garney, Pierre Franck, Emmanuelle Baudry, Dominique Vautrin, Jean- Marie Cornuet and Michel Solignac (1998), noted that in the original distribution area of the (Europe, Africa and Near and Middle East), honeybee populations are relatively homogeneous over wide territories corresponding to the subspecies or the geographical race. The honeybee species has thus been split into 24 races according to morphometric criteria. These 24 species have been grouped into larger sets corresponding to evolutionary branches. Three such branches have been distinguished: an Africa (A) branch including all African subspecies, a north Mediterranean (C) branch in the central and east Mediterranean and in central Europe, and a West European (M) branch, from Spain to Sweden and Poland. These three branches were first identified through morphometry ( Ruttner et al., 1978, Ruttner 1988) and were approximately confirmed through deoxyribonucleic acid (DNA) studies (Cornuet and Garnery, 1991.

75 Smith, 1991a and b. Garnery et al., 1992. Estoup et al., 1995. Arias and Sheppard, 1996). Also, Lionel Garney, Pierre Franck, Emmanuelle Baudry, Dominique Vautrin, Jean- Marie Cornuet and Michel Solignac (1998) reported that, the French indigenous subspecies, Apis mellifera mellifera, belongs to branch M. It extends widely beyond the country limits since it was naturally found everywhere in Europe north of the Alpine Arc (Ruttner, 1988). In the Iberian Peninsula, a related subspecies, A. m. iberica (Geotze, 1964), has been recognized, although it has long been considered as a local form of A. m. mellifera. Only other subspecies that are naturally present on a French boarder is the Italian A. m. ligustica, a member of branch C, with which A. m. mellifera hybridizes over the Alps and beyond (Badino et al., 1993. Sheppard and Berlocher, 1985). However, another subspecies, A. m. carnica (also belong to branch C) has been massively imported in Germany and has almost replaced the former A. m. mellifera populations ( Kauhausen-Keller and Keller, 1994. Maul and Hahnle, 1994). The consequence is that there is now a possibility of hybridization of A. m. mellifera with A. m. carnica in French areas close to the German border. Although beekeepers have long been expert in managing hives to get the best productions out of them, they generally do not control the entire biological cycle of honeybees, especially the mating of queens and drones. Hence, the current genetic biodiversity in the west part of the Old World, the original distribution area of the species, is still structured in a way that expresses its evolutionary heritage, i.e. it is determined by past demography, adaptation to local conditions, duration of isolation and natural migrations. However, one has to take into account several technical improvements more or less recently introduced in hive management and which may have

76 inferred with the natural evolution of populations. First queen breeding, when operated on a large scale, artificially reduces the effective size of populations, and can result in a loss of genetic variability. Second, importation of foreign queens can modify the genetic pool of local bees through hybridization. This genetic pollution is mainly due to professional bee-keepers who import foreign subspecies for their own qualities and/ or in the hope of producing superior hybrids with the local subspecies (Fresnaye et al., 1974. Cornuet and Fresnaye 1979). In France, A. m. ligustica and A. m. caucasica are the most in ported subspecies, mainly because their hybrids with A. m. mellifera have demonstrated superior qualities for honey yield (Fresnaye et al., 1974. Fresnaye and Lavie, 1976). Other synthetic strains as the English “buckfast” (Adam Br, 1966) or the American “midnight” and “star line” (Witherell, 1976) have also been imported, but to a lesser extent. A third factor is the practice of moving hives several times through the year to increase and diversity the honey production. According to when and where the mating season occurs, this practice may have effects similar to those of in porting queens. In addition to change in the genetic pool of populations, this can artificially increase the genetic diversity in two ways: by introduction new alleles and by increasing the effective population size. Deborah R. Smith and Robert H. Hagen (1996), studied the biography of Apis cerana as Revealed by Mitochondrial DNA Sequence Data. They reported that, the non-coding intergenic region of the Apis cerana mitochondrial DNA genome provides a rapidly evolving source of characters for study in intra-specific biography. They sequenced the non- coding intergenic region in bees from 110 colonies of Apis cerana collected over most of the species range. They found two major forms of non-coding sequence: a western form, occurring in bees from India, Sri Lanka and the

77 Andaman Islands, and an eastern form, occurring in bees from Nepal, Thailand, Malaysia, Indonesia, the Philippines, Hong Kong, Korea, Japan, and India. Thus the eastern and western haplotypes co-occur in India. Thus they found that within the eastern form, phylogenetic analysis of sequence variation indicated two well supported groups of haplotypes: a *Sundaland group* which was found in bees from peninsular Malaysia, Borneo, Java, Bali, Lombok, Timor, and Flores., and a *Philippine group* which was found in bees from Luzon, Mindanao, and Sangihe. Haplotypes from both the Sundaland group and the Philippine group were found on the island of Sulawesi, suggesting that this island was colonized independently by two groups of A. cerana. In addition, the bees of Taiwan and a third group of Sulawesi mitochondrial haplotypes characterized by absence of most of the non-coding sequence. Delarua, J. Serrano and Galian (1998), investigated the mitochondrial DNA variability in the Canary Islands honeybees (Apis mellifera L.), they reported that, the mt DNA of individuals from 79 colonies of Apis mellifera from five Canary Islands was studied using the Dra1 test based on the restriction of PCR products of the tRNAleu-COII intergenic region. Five haplotypes of the Africa (A) lineage and one of the west European (C) lineage were found. The haplotypes A14 and A15 are described for the first time. These haplotypes have a new P sequence named P1. The wide distribution and high frequency of haplotype A15 suggest that it is characteristic of the Canarian Archipelago. Thirteen haplotypes for the A lineage and their distribution have been described to date (Garnery et al., 1992, 1993, 1995. Moritz et al., 1994). One of these haplotypes A2, belonging to the African lineage, is present in southern Spain, Sicily and some Greek Islands, suggesting that its present distribution has been influenced by human activities. This haplotype was

78 also found in a few colonies from the Canary Islands, leading Garnery et al., (1993) to speculate about the Spanish origin of Canarian bee populations. The Canarian Archipelago is composed of a chain of volcanic islands originated in the Atlantic Ocean, 110 km of the northwest Africa coast. The ancestors of Canarian organisms might come from the close mainland and also from southern Iberia via the northeast trade winds (Oromi et al., 1991). The evolutionary of many of them into endemic species or subspecies and their relationships to the continental fauna is becoming clearer with the availability of DNA sequence data (Tenebrionidae: pimelia, Juan et al., (1995, 1994), which suggest a congruent pattern of colonization for different kinds of animals. P. De La Ru A, U. E. Simon, A. C. Tilde, R. F. A. Moritz & S. Fuchs, (2000), in their study MtDNA variation in Apis cerana populations from the Philippines, they reported that, The cavity-nesting honeybee Apis cerana occurs in Asia, from Afghanistan to China and from Japan to southern Indonesia. Based on morphometric values, this species can be grouped into four subspecies: A. c. cerana, A. c. indica, A. c. japonica and A. c. himalaya. In order to analyze the geographical variability of A. c. indica from the Philippine Islands, 47 colonies from different locations in three of the larger islands (Mindanao, Luzon and Palawan) and four of the Visayan Islands (Panay, Negros, Cebu and Leyte) were studied. Genetic variation was estimated by Dra1 restriction enzyme and sequence analysis of PCR- amplified fragments of the tRNAleu-COII region. They found four different haplotypes namely Ce1, Ce2, Ce3 and Ce4 that discriminate among the bee populations from different islands. The Ce1 haplotype is present in Mindanao and Visayan Islands, Ce2 is restricted to Luzon, and both Ce3 and Ce4 are only present in Palawan. Phylogenetic analysis of the sequences shows a great intraspecific variability, is in accordance with the geological

79 history of these islands and partially agrees with some previous morphological and molecular studies. A total of 738 colonies from 64 localities along the African continent have been analyzed by P. Franck et al., 2001, in their study genetic diversity of the honeybee in Africa: micro satellite and mitochondrial data. They used the restriction enzyme Dra 1 (RFLP) of the COI-COII mitochondrial region. Mitochondrial DNA of African honeybees appears to be composed of three highly divergent lineages. The African lineage previously reported (named A) is present in almost all the localities except those from northeastern Africa. In this area, two newly described lineages (called O and Y), putatively originating from the Near East, are observed in high proportion. This suggests an important differentiation of Ethiopian and Egyptian honeybees from those of other African areas. The A lineage is also present in high proportion in populations from the Iberian Peninsula and Sicily. Hiroyuki Tanaka, et al., 2001, studied the genetic differentiation among geographic groups of three honeybee species, Apis cerana, A. koschevnikovi and A. dorsata in Boroneo. They sequenced the mitochondrial cytochrome oxidase 1 (CO1) gene and carried out phylogenetic analysis in order to examine genetic relationships among geographic groups of each species (geographic genetic variation). They compared the sequence divergence- geographic distance relationships among the three species. Estimated genetic differentiation was an order of magnitude large in A. kosochevnikovi than in A. cerana and A. dorsata. Migratory nesting behavior and cold tolerance of each honeybee, and the pale climate of the Southeast Asian tropics, are discussed as factors that produced these characteristics for mitochondrial genetic markers and conservation priorities are recommended.

80 Also, in this study, they suggested that, the genetic variations within plant-pollinator assemblages clarifies not only phylogenetic concepts associated with biological species, but may also indicate evolutionary potential and adaptive diversity in mutualisms (Thompson, 1994). Walter S. Sheppard & Marina D. Meixner (2003) named a new honeybee subspecies from Central Asia (Apis mellifera pomonella), demonstrating that, endemic honeybees of the Tien Shan Mountains in Central Asia are described as a new subspecies, Apis mellifera pomonella, on the basis of morphometric analyses. Principal component and discriminant analysis of the morphological characters measured clearly place these bees into the oriental evolutionary branch of honeybees, but also show that they are distinct from the other subspecies in this lineage. The existence of this newly described honeybee subspecies extends the range of endemic A. mellifera more than 2000 km eastward than previously estimated. Sequence analysis of mitochondrial DNA places A. m. pomonella within the C mitochondrial lineage (a group that is inclusive of both C and O morphological lineages). These findings support the conclusion that A. m. pomonella has a phylogeographic history shared with subspecies from the eastern limit of the previously known range. The Iberian honeybee, Apis mellifera iberica (Goetze, 1964; Ruttner, 1988) has been subject of numerous studies about its molecular diversity (Smith et al., 1991; Garnery et al., 1995, 1998a, b; Sheppard et al., 1996; ., Franck et al., 1998; De la Rْa et al., 1999, 2002; Canovas et al ; 2002 Hernandez-Garcia et al., 2002) and morphometrical variations (Izquierdo et al., 1985; Cornuet and Fresnaye 1989; Orantes-Bermejo and Garcia- Fernandez, 1995; Padilla et al., 1998, 2001). Earlier analyses showed that two out of the five evolutionary lineages of A. mellifera coexist on the Iberian Peninsula (Smith et al., 1991; Garnery et al., 1995). In the North, the

81 Iberian honeybee populations are more similar to the West European lineage as they bear particular sequences in the intergenic tRNAleu-COII region (so called mitochondrial haplotypes), whereas in the South they show predominantly African haplotypes. These results led Smith et al., (1991) to postulate that A. mellifera iberica is the result of the hybridization between the West European A. mellifera mellifera and the North African A. mellifera intermissa. In spite of the occurrence of this generalized pattern of mitochondrial haplotypes across the Iberian Peninsula, the studies realized at the regional level have demonstrated that every region holds a peculiar composition of molecular markers. Thus, the honeybees from Murcia show a homogeneous composition of mitochondrial haplotypes (De la Rْa et al., 1999), and neither traces of recent introgression events from African subspecies (De la Rua et al., 2002), nor the effects of transhumance movements (Hernandez- Garcia et al., 2002). In Galicia (NW Spain) the gradient of lineage distribution shows a rather steep transition, as the Southern provinces of Pontevedra and Orense have predominantly (>90 percent) African haplotypes, whereas La Coruna and Lugo in the North show almost only (>90 percent) West European haplotypes (Canovas et al., 2002). In contrast, in East Spain the percentage of African haplotypes smoothly decreases northwards along the Mediterranean provinces of Valencia (De la Rua et al., submitted: cited in De la Rua P., R. et al., 2004). De la Rua P., R. Hernadez-Garcia, B.V. Pedersen, J. Galian and J. Serrano (2004), in their study molecular diversity of the honey bee Apis mellifera iberica from Western Andalusia, they tried to characterize its populations according to the variability shown by mtDNA haplotypes and micro satellite loci as previously mentioned . They analyzed the

82 mitochondrial and nuclear DNA ; mitochondrial haplotype corresponding to the intergenic region tRNAleu-COII, and six micro satellite loci has been determined in hives distributed in 24 localities of the provinces of Malaga, Seville, Cadiz and Huelva. Six different haplotypes have been found, five of the African and one of the West European evolutionary lineage. These results corroborate the hybrid nature of the subspecies Apis mellifera iberica, which has a predominant influence of the African lineage in the South, that is gradually or steeply replaced northwards by the West European lineage. The variability of the micro satellite loci is similar to that found in African populations in relation to the detected number of alleles and the values of genetic diversity. These observations show the genetic relationship between Andalusian honeybee populations and those ones from North Africa. Micro satellite data vary notably between the studied provinces. In the province of Cadiz the mitochondrial homogeneity contrasts with the micro satellite variability, what suggests a recent introgression event from African-like populations of unknown geographic origin. In regard to the African honeybees, one must first consider the results of mtDNA studies with respect to the major lineages erected by Ruttner (1988), based on the multivariate analyses morphometric characters. The combined data of several studies using restriction enzymes to obtain restriction fragment length polymorphisms, as well as actual sequence data (Arias & Sheppard 1996), clearly support the existence of three major mtDNA lineages. The African lineage (A) includes all mtDNA types from colonies of African origin,; a north Mediterranean – Caucasian lineage (C) the bees of eastern Europe and the Caucasus region; and a western Mediterranean lineage (M) for the bees of that region (Smith et al., 1989, 1991; Garnery et al., 1992, 1995).

83 Most African bees currently analyzed belong to the A mitochondrial lineage (Smith, 1991; Garnery et al., 1992, 1993; Arias & Sheppard, 1996; De la RuÁ a et al., 1998). Only two colonies from Egypt have been recognized as belonging to lineage O (Arias & Sheppard, 1996; Franck et al., 2000b), thus further more in 2001, P. Franck et al., described two newly lineages from north-eastern Africa called “O” and “Y”. Each of these lineages has unique length fragment polymorphisms as well as intergenic sequences. Garnery et al., (1992) proposed an hypothesis for the origin and evolution of the population of Apis mellifera based on their mtDNA types. These interpretations are very similar to those of Ruttner et al., (1978), but there are some slight discrepancies. In Ruttner's scheme, the north Africa intermissia were associated with iberica and mellifera; the mtDNA associates intermissia with the other African subspecies. In fact, the mtDNA data clearly indicate that the Spain is a hybrid zone, the southern half of the country dominated by the intermissa mitotypes, the northern half by mellifera mitotypes (Cornuet & Garnery 1991b; Garnery et al., 1992; Smith et al., 1991; Sheppard et al., 1996). This interpretation is supported by allozymic data (Cornuet 1982) and by more recent studies of morphometrics and phermonal variance for the region (Hepburn & Radloff 1996a). However, that the Iberian peninsula is such a transition zone is not consistent with recent micro satellite data (Frank et al., 1998). The high similarity of mtDNA profiles among the bees of northern and southern Africa is somewhat surprising (Cornuet &Garnery 1991b) that, they share a common mitotype defined by a combination of 22 polymorphism restriction sites (Smith 1991; Garnery et al., 1992). The reason for relatively low within – lineage diversity of the African lineage

84 can be parsimoniously explained as the result of the isolation of individual’s populations rather than multiple, highly polymorphism ancestral populations (Garnery et al., 1992). In two different studies, Smith (1988, 1991) analyzed restriction site poly-morphisms and fragment length variations for several European subspecies and others from Africa. She was able to separate Intermissa, Scutellata, Capensis and Unicolor as a group from the European subspecies, thus once again confirming the mitochondrial distinction of the A, C and M lineages. However if her data are subjected to cluster analysis, none of the subspecies within the Africa A lineage can be distinguished from one another. In a similar study, but using a slightly different suite of restriction endonucleases, Garney et al., (1992, 1995) also confirmed the mitochondrial distinctions between lineages. However, clusters analyses of their data do not allow the discrimination of the African Monticola, Scutellata and Capensis from one another. Thus, Moritz R. F. A. et al., (1994) in their study: Mitochondrial DNA variability in South African honeybees (Apis mellifera L), demonstrated that, the mtDNA size variability of honeybees (Apis mellifera) in a sample of 102 colonies covering the area south of the 27th parallel of latitude in Africa was analyzed using PCR. A region between the COI and COll genes revealed four different size variants one of which being a novel mitotype for honeybees not fitting the previously published repeat pattern in the region (a fragment po with 69 bp and varying number of fragment Q of 196 bp length). This region, which has been shown to be useful for the biogeography classification of Apis mellifera subspecies, only partially corresponded to the known distribution of African subspecies of honeybees based on morphometrical and physiological data. The PoQQ- type was the most common with an overall frequency of 0.76. The region,

85 which has been addressed as the hybrid zone between A. m. capensis and A. m. scuteltata showed no mitotype variability and was monomorphic for the

PoQQ type. A considerable length polymorphism was found north and east of this region with a frequency of 0.57 for PoQQ type and 0.36 for the PoQ type. Less common were the PoQQQ type (0.02) and a type not fitting the known P and Q repeat system (0.02). Digestion of the region with the Dral restriction enzyme revealed previously undetected mtDNA variability in Apis mellifera populations. More recently, Sheppard et al., (1996) performed a restriction enzyme analysis of eight subspecies (three European and five African). With the enzyme Hinf 1, twenty distinct haplotypes were obtained taxonomically useful. Lamarkii exhibited a single unique variant. Intermissa, monticolla and sahariensis each exhibited at least one haplotype not shared with one another, and other haplotypes that were shared with each other and with Scutellata. For example, while Sahariensis had one unique haplotype its other were identical to those of the neighboring Intermissa; the same being true for Monticola and its neighbor Scutellata. Similarly, Intermissa shared three of eight haplotypes the European Iberica. While such results hold promise for studies of honeybee population discrimination, it’s likely that the use of multiple probes would enhance discrimination. Whatever suite of restriction enzymes chosen as probes, the origin of the honeybees themselves must be carefully considered. In several of the above studies, a particular combination of subspecies was acquired on the basis of accessibility and typological or racial homogeneity (Meixner et al., 1994; Arias and Sheppard 1996). They actually represent point or spot samples taken from continuous populations. These kinds of classification constraints may well introduce a bias in the final interpretation of the data.

86 Thus, far there have been only two studies in which mitochondrial DNA analyses were performed on a transect basis without pre-grouping the data into sub specific groups before analysis. Garnery et al., 1995, study the mitochondrial DNA variation along a transect through Morocco and Spain using a combination of particular power: length polymorphism and restriction site variation, their conclusions is that, gene flow in the region has primarily been from south to north, which is from populations of Intermissa in the Maghreb to Iberica populations on the Iberian peninsula. Referring only to Intermissa. Garnery et al., (1995) also suggested that the bees in Morocco it self probably represent two sub lineages of the past, one coming from the northeast and other from the south , and that the contact points between them actually extend on both sides of the Atlas mountains for the length of the country. In this particular study, Garnery et al., (1995), were principally concerned with a phylogenetic analysis of the bees of the region. Their technique was essentially a cluster analysis modified to estimate phylogenetic distance between samples. However, these data (Garnery et al., 1995) can be analyzed in a very different way. In order to compare the results of this mtDNA study with the variance characteristics of morphometric and phermonal analyses for the same region, but without imposing the limits of classification into subspecies.

Hepburn and Radloff (1996, 1997) re-analysed their mtDNA data. Applying Greenacre's method of analysis (1988), they found that the mtDNA data used in the Garnery et al., (1995) study resolved into six distinct and significantly different mtDNA clusters or groups. The three mtDNA clusters of Spain matched the distributions of three iberica

87 morphometric clusters. The three mtDNA clusters obtained in Morocco did not exactly match the two morphometric clusters corresponding to Intermissa and Sahariensis (Hepburn &Radloff 1996a). In summary analysis of mtDNA unequivocsally support the three major honeybee lineages; however the technique has been less successful in defining mtDNA clusters that correspond with morphoclusters derived from multivariate analysis.

88 CHAPTER THREE: MATERIALS AND METHODS

3- 1- Sampling: In a number of survey trips (2005-2006), nineteen samples of the native Sudanese honeybee Apis mellifera workers were collected. Samples were collected within the latitude 3º N to 22º N and longitude 23º E and 38º E. and from at least three different geographical zones of the Sudan; namely, semi-desert, savannah and forest zones. Four samples were captured from the following locations at the semi-desert region: Shendi, Khartoum, El-Hissahisa and Madani. From the poor and rich savannah regions, eleven samples were captured at the following locations: Doka, El-Galabat, El-Hawata, Galla-Elnahal, Kosti, El-Dalng, Kadogli, Malakal, Doleib, Ganal and Wau-sholok. Then from the forest region, four samples were captured at the following locations: Juba, Liria, Bango and Khour-Maquire. See Table (1) and fig. (1). {Sudan map sampling localities}. The samples were collected from nests found in traditional cylindrical hives, tree cavities, resting feral swarms, established wild colonies on tree branches, in soil and rock crevices, crevice in a store wall in a farm and Apiary (Khartoum sample only). Samples were captured by using an ordinary insect catching net placed round the entrance of established colonies which were disturbed and then attacking bees were caught, or swarm clusters and exposed colonies nesting on shrub branches were shaken into the net. Sampling sites were at least over 25 miles from each other. Thus 300 to 1000 miles separated the geographical zones from each other.

89 Four samples of Apis florea from Gerry, Khartoum, Madani, and El- Dender were also captured on tree branches for investigation. The bees so collected were divided into two groups (100 to 150 bee per a group) then one group killed by hot water (so as to obtain a fully stretched proboscis) and the second one killed by chloroform. Both of the two groups were separately preserved in Ethanol (75 to 90%). Table (1): Sampling localities, respective geographical zones, map refrence numbers & coordinates of honeybee localities analysed in this study.

Sample Geographical Map No. localities zones reference Coordinates Longitude Latitude Semi- desert 1. 33°45E 17°00N Shendi zone. (1). 2. Gerry Semi- desert zone. (2). 32°48 E 16°15 N 3. Khartoum Semi- desert zone. (3). 32°30E 15°45N 4. El-Hissahisa Semi- desert zone. (4). 33°21E 14°45N 5. Madany Semi- desert zone. (5). 34°00E 14°15N 6. Erkawit Semi- desert zone. (6). Poor savannah 7. 34°45E 13°36N Gala Elnahal. zone. (7).

Poor savannah 8. 34°30E 13°30N Elhawata zone. (8).

9. Poor savannah Dokka zone. (9). 35°54E 12°45N

10. Poor savannah Glabat zone. (10). 36°6 E 13°00N Poor savannah 11. Semsim zone. (11). Poor savannah (12). 34° 45 E 12° 36 N 12. El- Dender zone. Poor savannah (13). 32°24E 13°24N 13. Kosti. zone. Poor savannah (14). 33° 95 E 13° 15 N 14. Singga zone.

90

Table (1) continued

Poor savannah (15). 34° 38 E 11° 85 N 15. Rosseris zone. 16. El- Damazin Poor savannah (16). 34° 29 E 11° 78 N zone. 17. Om- Rawaba Poor savannah (17). 31° 20 E 12° 80 N zone. 18. Eldalang Poor savannah (18). 30°00E 12°30N zone. 19. Kadogly Poor savannah (19). 29°45E 11°00N zone. 20. Wau-Sholok. Rich savannah (20). 31°48E 9° 45N zone. 21. Malakal Rich savannah (21). 31°45E 9° 30 N zone. 22. Dolaib. Rich savannah (22). 32°00E 9° 18N zone. 23. Gannal. Rich savannah (23). 31°12E 9° 15N zone. 24. Raja Forest zone. (24). 28.08° E 7.70° N

25. Shawish- Rich savannah (25). Mahadi zone. 26. Kafindabi Rich savannah (26). zone. 27. Kubum Poor savannah (27). 23° 80 E 11° zone. 80 N 28. Zalinge Poor savannah (28). 34° 29 E 11° 78 zone. N 29. Juba Forest zone. (29). 31°30E 4° 48N

30. Bango Forest zone. (30). 31°15E 4° 30N

31. Liria Forest zone. (31). 32°15E 4° 15N

32. Khour- Forest zone. (32). 31°45E 5° 15N Maquire. 33. Yei Forest zone. (33). 30° 60 E 4° 00 N

34. Latokka Forest zone (34).

91 35. Kajiko Forest zone (35).

3- 2- Morphometric analysis: Fifteen bee workers were randomly taken from each colony as representive samples for morphometric measurements at the institute für Bienekunde [Oberursel, Germany]. Thirty eight characters as listed in Table (2), consist of the characters introduced and tested by Goetze (1964) and Alpatov (1929), i.e. pilosity, size, colour, cubital venation (Nos. 1, 3-17, 21, 22, 27-30), plus some of the angles in the wing venation tested by Dupraw (1964) in collaboration with the laboratory of the institute für Bienekunde [Oberursel, Germany (Nos. 31-41)] and some newly selected characters (No. 18-20; 23-26 Ruttner, 1978). Three of these characters were eliminated, either because they added no new information (length of hairs between the facets of the eye and hind wing hamulies) or because of they proved redundancy (No. 27, 28, cubital veins left).

The above-mentioned 39 characters were considered as, quantifiable number for the taxonomic values in the institute für Bienekunde, Oberursel, Germany), Ruttner et al., (1978).

92

Table (2): List of characters used for the analysis:

93 Nu Character Fig mbe Author ure r 1 Length of hairs on tergite 5. Goetze (1964). 2 2 Width of the tomentum band on Goetze the side of tergite 4. (1964). 2 3 Width of the dark stripe between Goetze the tomentum and the posterior (1964). 2 rim of the tergite. 4 Length of the stretched proboscis Alpatov (glossa+ mentum+ submentum). (1929). 3 6- Length of the hind leg (femur Alpatov 8 No.6, tibia No. 7, metatarsus No. (1929). 4 8). 9 Width of metatarsus 3. Alpatov (1929). 4 10- Pigmentation of tergites 2, 3 &4, Goetze 12 evaluated according to a scale of (1964). 5 10 grades between the darkest (0) and brightest (9). 13, Diameter of tergites 3 and 4, Alpatov 14 longitudinal* (1929). 6 15 Sternite 3 longitudinal* Alpatov (1929). 7 16, Wax mirrior, on sternite 3, Alpatov 17 longitudinal and transversal* (1929). 7 18 Distance between wax mirrors, of Ruttner sternite 3. (1978). 7

94 19, Sternite 6, longitudinal and Ruttner 20 transversal* (1978). 8

Table (2) continued Forewing length and width. Alpatov 21 (1928). 9 , 22 Pigmentation (coloration) of the Ruttner 23 scutellum. (1978). 1 , 0 24 Pigmentation of the labrum. Ruttner 25 (1978). 1 , 1 26 - b Segment ‘a’ and ‘b’ of the cubital cell of Alpatov 27 right forewing. (1929). 9 , 28 11 angles between lines connecting cross Goetze 31 points of the venation on the forewing (1964). 1 (No.31= angle A4, 32= B4, 33= D7, 34= - 2 E9, 35= G18, 36= J10, 37= J16, 38= 41 K19, 39= L13, 40= N23, 41= O26). *The term ''longitudinal & transversal'' are used instead of ''length'' and ''width'' which may be misleading.

95 In addition to the above primary measurements, there were some secondary values calculated by summation and division. For instance, No. 2 and 3 for tomentum index, No. 6, 7 and 8 for length of the hind leg, No. 8 and 9 for metatarsus index, No. 13 and 14 for values of body size, No. 19 and 20 for index of body slenderness and No. 29 and 30 for cubital index. For the statistical analysis, only the Primary values were used, but the indices, being independent of body size were very useful in characterizing a race. Even in early times a race was characterized by broad metatarsi (Apis m. remipes, Gertstacker 1860; Buttel- Reepen 1906) or high cubital index (A. m. carnica; Goetze 1940) [the cubital index is the ratio of the distance “a” and “b”, determined by the point at which the nervous recurrent Nr joints the lower vein of the third cubital cell “C”. of the right forewing Ruttner & Mackensen, 1952. The index of slenderness is an objective measure of a slender or broad body distinguishable even with the naked eye, thus mathematically it is the ratio of the length to the width of the abdomen sternite 6. 3- 2- a- Preparation and measurements records of the bees: The preparation of the bees and recording of the measurements are organized in such a way that, all the values for each individual bee are registered en bloc. The measurements of the pilosity (length of hairs on tergite 5 and width of the tomentum band on the side of tergite 4) are made on the un dissected bee and that of other characters of the parts of the body are adjusted on a slide. The measurements of wing venation angles, cubital veins and characters of size were measured using a CCD camera combined with an on-screen measuring system (Leitz)

96 magnification 50X, (Meixner, 1994). Or for some characters (hairs, tergites, colour) on a stereomicroscope, magnification 40X. With roughly about 40 characters per bee nearly 600 data were accumulated per one sample; 11,400 data per all the samples.

97 Fig (2): - Abdomen of the honeybee workers Apis mellifera L. a. Width of tomentum of tergite 4 (No.2); b. Width of the dark stripe between tomentum and posterior rim of the tergite (No. 3); h. Length of hairs on tergite 5 (No.1).

98

Fig (3): Length of Proboscis of honeybee Workers Apis mellifera L.

99

Fig (4): Hind leg of the honeybee workers (Apis mellifera L.) Fe: length of femur (No. 6), Ti: length of tibia (No. 7), ML: length of metatarsus (No. 8), MT: width of metatarsus (No. 9).

100

Fig (5): Classes of pigmentation of tergites (2 to 4). Evaluated by 10 classes (No. 10- 12).

101

Fig (6): Longitudinal diameter of tergite 3 and 4 of honeybee workers Apis mellifera L. (No. 13, 14).

Fig. (7): Sternite 3 diameters of honeybee workers Apis mellifera L.: S3: longitudinal diameter (No. 15), WL: wax mirror, longitudinal (No. 16), WT: wax mirror transversal (No. 17), WD: distance between wax mirrors (No. 18). [Honeybee workers Apis mellifera L.]

102

Fig (8): Length and Width of Sternite 6. of honeybee workers Apis mellifera L. L6: longitudinal (No. 19); T6: transversal (No. 20).

103

Fig (9): Ho ne yb

104 ee wo rk ers Ap is me llif er a L. for ew in g; FL: length (No. 21); FB: Width (NO. 22) a: cubital vein (No. 27); b: cubital vein (No. 28).

105

Fig (10): Scutellum of the honeybee workers Apis mellfera L. Sc = Scutellum: Scale of pigmentation 0 (completely dark); 9 (yellow). B, K: Metatergum and mesotergal sclerite (scale of pigmentation 0-5). [No. 23 & 24].

106

Fig (11- a): Bee workers Apis mellifera L. Labrum.

107

Fig (11- b.): Labrum pigmentation of the honeybee workers Apis mellifera L. (No.25 &26).

108

Fig (12): Wing venation Angles of Honeybee worker Apis mellifera L.

109 3- 2- b- Statistical analysis: - First of all the crude data was recorded in perforated cards. Two cards were necessary per bee, per sample and each card was labelled with the number of the sample, number of the bees, sex, country and region. The first steps of the statistical analysis were the transformation of the data into a uniform scale and computation of the means and standard deviation. The important next step was the elimination of errors by inspection of the standard deviations. A high value than usual indicated that a mistake probably was made in writing, reading or punching. The printouts were then compared with the original records and the erroneous cards were replaced. The statistical analysis of the data was performed with SPSS for windows and Statistical computer programs. Thus this statistical analysis has so far been made on 32 characters and 285 samples of worker bees from all zones with autochthonous mellifera bees available, so the basic information subjected to statistical treatment was represented by a 285 X 32 array, where each of the 285 lines was a sample of 15 bees. The 32 components of this vector line were the means of individually measured character. Multivariate analysis allowed a general survey of these data. As it was not easy to imagine a cloud of 285 points in 32 dimensional space and in order to simplify this problem, an attempt was made to determine a few number of privileged axes: This was the aim of the Principal Component Analysis (PCA). From all the correlation coefficients between the first variable two by two, it was possible to obtain histograms on the new axes and point- projections on the planes formed by two of these axes.

110 Interpretation of the results consisted of characterizing the clusters of points. The contribution of each of the basic variables in the formation of the new ones was estimated. In this way, some little interesting characters could be discarded, since selection of the most discriminant criteria was the most important goal. So in this initial study, the basic tool was the PCA: a descriptive and statistical hypothesis-free method. The Progressive Discriminant Analysis was also used. With this technique, and using a progressive number of the first variables, it was possible to obtain a borderline between two populations to be identified for classification. The last objective was to get a picture of each bee-type collected and locate it geographically. The Factor Analysis of correspondences allows representation on the same plane, the profiles of the variables as well as those of the samples. 3- 3- Mitochondrial DNA: - In addition to the captured samples there were 16 Sudanese honeybee worker samples added to the mitochondrial DNA analysis (samples collected from localities which I am not captured bees). Those samples are: 5 samples from; Halfa- Elgadeda and Fau (semi-desert zone). El- Damazin and Om-Rawaba (savannah zone), and Raja (forest zone), collected by Mogbil (2004-2005). Eleven samples were obtained from the data bank of the institute für Bienekunde [Oberursel, Germany]; (Sudanese honeybee workers collected by Mogga 20 years ago, for morphometric studies at that time, thus they were preserved in ethanol at room temperature). Those samples are: Yei, Lwatoka, Kafindapy, Shawish- Mahadi and Kubbum (forest and rich savannah zone). Zalinge, Roseris, Singa and

111 Simsim (poor savannah zone). Salahap and Kassala (semi-desert zone). Worker bees used for the mitochondrial DNA analysis were killed by chloroform and immediately immersed in ethanol 90% for preservation at room temperature. 3- 3- a- DNA extraction: Before DNA extraction, a single bee from each preserved bees samples in ethanol 90%, was taken in a separate tube (one bee from each group representing the sample), and were carefully washed by distilled water (to remove alcohol). The sample tubes were shaken by a thermo mixture (shaking only) for more than one hour and the distilled water was changed from time to time. Each target bee mesotharax was dissected and the muscles were kept in a separate tubes (1.5 ml Eppendorf tubes) and stored at –20º C. until they were processed in the laboratory. Total DNA was extracted from thoraces using Phenol- chloroform extraction and ethanol precipitation protocols (Sambrook et al., 1989), with slight modification as follow: 1- Some distal water was added to the muscles tubes, and shaked on the thermomixture for 15 minutes. 2- Water was removed outside the tubes. 400 µL. Wilson Buffer (PH= 0,8) was added to each tube. 3- Tissues inside tubes were crashed by an special plastic stick. 4- Ten µL. of Proteinase-K, was added to each tube and gentelly shaked (so as to perform protein digestion). 5- Samples were put in the thermomixture (55º C and 6-7 cycles) for eight hours (tubes were genttely shaked from time to time).

112 6- 410 µL. of Phenol-chloroform-isoamylalkohol (P C 1), 25: 24: 1. was added to each sample tube (the same volume as in step 2 and 4), with gentelly mixing for five minutes. 7- Samples were centrifuged for 15 minutes at 13000 rpm. speed. 8- The supernatant (270 µL.) was taked to a new tube, then an equal amount of chlorofom-iso was added, with gentelly mixing for five minutes and centrifuged for 15 minutes at 13000 rpm. speed. 9- The supernatant (about 180 µL.) was transformed into a new tube. 10% from the volume of the supernatant sodium acetate (18 µL.) and 2 volumes ethanol 100% (360 µL.) were added. The sample tubes were kept inside freezer for overnight at –20º C. 10- Samples were centrifuged at 13000 rpm. speed for 15 minutes. The supernatant were carfully thrown out the tubes. Two volumes of ethanol 70%, were added and centrifuged at 13000 rpm. speed for 5 minutes. 11- The supernatant were carfully thrown out and the remaining pellets were dried at 55º C (on the thermomixture. The tubes were opened). 12- 30 µL of water or Buffer was added, then the DNA was preserved at –20º C. 3- 3- b- PCR amplification:

The mitochondrial DNA fragment including the COI- COII intergenic region (tRNAleu gene and the 5´-end of the COII gene), was amplified by PCR (Biometra thermocycler), using the primers E2 (5´- GGCAAGAATAAGTGCATTG-3´) and H2 (5´-CAATAT CATTGATGACC-3´) (Garnery et al., 1992), as follow:

The PCR reaction was performed with the corresponding 1x buffer, 0.25

µL of each dNTPs, 0.1 µL Mgcl2, 0.1 µL Promega Taq polymerase, 0.25 µL of primers E2 and H2, 1.0 µL of extracted DNA, in a total volume of 30 µL.

113 Reactions were submitted to an initial denaturation of 5 min at 96º C, 30. Cycles of 95º C for 0.5 min, 50ºC for 1.5 min and 72ºC for 1.5 min, and a final extension of 10 min at 72º C. 3- 3- c- DNA purification: The extracted DNA was purified from the remaining of the chemicals used in the PCR reaction by DNA Clean-up System (promega) following the manufacturer’s instructions as follow: 1- 5 volume of Buffer P.B was added to 1 volume of the PCR product and mixed (5× 30 µL or 150 µL to each). 2- The QLA spin quick column was placed in a provided 2 ml collection tube. 3- The mixture (of step 1.) was transformed to the QLA quick column and centrifuged for 30 to 60 seconds at 13.000 rpm. 4- The supernatant was discarded, and the QLA quick column was placed back into the same tube. 5- 750 µL Buffer P.E was added into the QLA quick column and centrifuged for 30 to 60 seconds at 13.000 rpm. 6- As in 4 plus the QLA quick column was centrifuged for an additional 1- minute. 7- The QLA column was placed in a cleaned 1.5 ml microcentrifuge tube. 8- The DNA was diluted by adding 30 µL of sterilled water to the centre of the QLA quick column membrane; and the QLA quick column was left stand for 2 minutes and centrifuged for 1 minute.

114 4- 3- d- Size category of the fragments: Five µL. of the PCR product was electrophoresed through 1.5% agarose gel and stained with ethidium bromide in order to determine the size of each product. 3- 3- e- Endonuclease digestion: Fifteen µL. of the PCR product was digested with the restriction enzyme DraI at 37º C for over night. Restriction fragments were separated on 10% and 8% acryl amide gel and stained with ethidume bromide.

115 CHAPTER FOUR: RESULTS 4- 1- Morphometric analysis (Apis mellifera L.): 4- 1- i - Univariate analyses: The results obtained in the univariate analysis of the 19 colony samples of Apis mellifera L. from at least four different geographical regions were shown in Tables (3 to 9). 4- 1- i- a- The proboscis and hind- leg measurements (mm.), {Table 3}: The mean length of the proboscis for the samples varied from 5.51 To 5.75, with an average of 5.63. Mean length of the different parts of the hind-leg for the sample varied from: Femur 2.23 To 2.46, with an average of 2.34. Tibia 2.86 to 3.16, with an average of 2.98. The metatarsus 1.78 to 2.01, with an average of 1.88. The mean total hind-leg length for the samples varied from 7.00 to 7.87, with an average of 7.26. The mean width of metatarsus for the samples varied from 1.00 to 1.13, with an average of 1.07. The samples metatarsal index varied from 55.91 to 60. 04 with an average of 57.39. 4- 1- i- b- Forewing measurements (mm.), {Table 4}: The mean length of forewing for the samples varied from 8.16 to 8.63, with an average of 8.36, while the mean width varied from 2.79 to 3.01, with an average of 2.90. The mean cubital vein “a” for the samples varied from 4.19 to 5.25, with an average of 4.67, and vein “b” varied from 1.56 to 2.12, with an average of 1.90. The mean cubital index a/b of the samples varied from 1.75 to 2.67, with an average of 2.13.

116

Table (3): Means of measurements of proboscis and hind- leg of the Sudanese honeybee workers Apis mellifera (mm.).

117 4- 1- i- c- Forewing venation angles measurements (degrees), (Table 5) :

The means of angles of the forewing venations for the samples varied

as follows:

A4: from 30.37 to 35.32 , with an average of 32.40. B4: from 97.84 to 105.70 , with an average of 101.86. D7: from 96.85 to 103.59 , with an average of 100.17. E9: from 18.17 to 20.96 , with an average of 19.73. G18: from 95.63 to 101.78 , with an average of 98.60. J10: from 49.32 to 56.13 , with an average of 52.87. J16: from 88.61 to 95.57 , with an average of 93.01. K19: from 74.58 to 83.54 , with an average of 80. 39. L13: from 12.66 to 23.60 , with an average of 15.30. N23: from 83.85 to 93.72 , with an average of 88.11. O26: from 34.07 to 43.72 , with an average of 37. 77.

4- 1- i- d- Body size (tergites 3 and 4) and sternite 3 measurements (mm.), {Table 6}:

The mean length of tergite 3 (T3) for the samples varied from 1.94 to 2.09, with an average of 2.03. Tergite 4 (T4) varied from 1.90 to 2.05, with an average of 1.97. The mean body length T3+4 varied from 3.79 to 4.06, with an average of 4.00. Mean length of sternite (S3) for the samples varied from 2.36 to 2.58, with an average of 2.48.Mean length of wax merrior on S3 varied from 1.03 to 1.16, with an average of 1.11. Mean width of the wax merrior on S3 varied from 1.90 to 2.04, with an average of 1.98. Mean distance between wax merriors on S3 varied from 0.30 to 0.38. with an average of 0.34.

4- 1- i- e- Pilosity measurements (mm.), {Table 7}:

The mean length of hairs on tergite (T5) for the samples varied from 0.17 to 0.26, with an average of 0.20. Mean width of tomentum on tergite (T4) varied from 0.59 to 1.05, with an average of 0.89. The mean width of the dark stripe

118 on T4 varied from 0.26 to 0.66, with an average of 0.44. The mean tomentum index for the samples varied from 2.21 to 5.90, with an average of 3.14.

4- 1- i- f- Sternite 6 (Abdominal slenderness) measurements (mm.), {Table 8}:

The mean length of sternite (S6) for the samples varied from 2.22 to 2.49, with an average of 2.19. While the mean width varied from 2.59 to 2.83, with an average of 2,72. The abdominal slenderness (L/W) for the samples varied from 81.31 to 87.31, with an average of 84.49.

4- 1- i- g- Pigmentation (Colouration) measurements {Table 9}:

The mean colouration on T2 for the samples varied from 5.16 to 8.94, with an average of 8.66; Mean colouration on T3 varied from 5.96, to 9.54 , with an average of 8.51.

The mean colouration on tergie 4 (T4) varied from 3.29 to 5.25, with an average of 4.47. The mean colouration of the scutellum for the samples varied from 4.48 to 7.44, with an average of 6.89; while the mean colouration of metatergum varied from 1.03 to 6.44, with an average of 3.82.

The mean colouration of the labrum for the samples varied from 1.60 to 8.10, with an average of 4.10, and from 0.13 to 3.02 with an average of 1.36.

119

Table (4): Means of measurements of the forewing of the Sudanese honeybee workers Apis mellifera L. (mm.)

Sample Length Width Cubital Cubital C. Index Vein a Vein b A/b Gala Elnahal 8.38 2.86 4.35 1.76 2.12

Madani 8.19 2.79 4.52 1.95 2.02

Khartoum 8.22 2.86 4.61 2.05 1.93 Shendi 8.19 2.94 4.64 2.03 1.96 Kadogli 8.42 2.93 4.75 1.95 2.09

Eldalang 8.45 2.95 4.66 1.96 2.04 Wau- Sholok 8.51 2.95 5.17 1.79 2.49 Malakal 8.39 2.93 5.25 1.73 2.62 Doleib. 8.63 2.99 5.15 1.90 2.34 Gannal. 8.45 3.01 4.87 1.56 2.67

Elhawata 8.54 2.98 5.20 1.83 2.49

Doka 8.16 2.87 4.63 1.78 2.24

Glabat 8.52 3.00 4.61 2.12 1.88

Kosti 8.49 2.93 4.52 2.03 1.93

Hissahisa 8.48 2.92 4.93 1.87 2.26 Juba 8.23 2.81 4.19 1.91 1.89

Bango 8.27 2.85 4.21 2.06 1.75 Liria 8.18 2.82 4.38 1.89 1.99 Khour- Maqure 8.24 2.82 4.25 2.03 1.80

120

Table (5): Means of measurements of forewing venation angles of the Sudanese honeybee workers Apis mellifera L. (degrees)

Sample Forewing venation angles

A4 B4 D7 E9 G18 J10 J16 K19 L13 N23 O26 Gala- Elnahal 35.32 97.84 102.82 20.08 97.14 49.87 94.62 83.22 12.66 89.22 39.17

Madani 33.96 98.36 97.74 18.17 96.76 49.32 95.30 78.42 15.11 87.55 35.06

Khartoum 31.38 103.62 100.39 20.56 95.86 52.24 93.00 79.49 14.80 87.33 37.94 Shendi 32.23 101.33 99.79 20.02 95.63 51.94 94.28 80.19 15.12 87.42 36.84

Kadogli 32.21 101.65 97.78 20.56 100.89 51.19 90.56 82.91 14.24 85.98 38.62

Eldalang 32.00 103.06 96.85 20.96 99.55 51.90 88.96 82.83 14.33 84.48 38.37 Wau-Sholok 31.71 100.63 99.03 19.91 100.19 53.78 91.11 83.08 16.03 83.85 36.07

Malakal 31.90 100.37 97.43 20.26 101.78 54.74 92.73 82.43 15.22 86.45 34.75 Doleib. 30.37 105.00 101.81 20.08 96.30 55.73 93.78 79.69 15.86 90.25 38.55

Gannal. 34.05 100.99 99.53 20.62 99.41 55.78 88.61 74.58 15.93 85.24 36.59

Elhawata 32.23 99.70 99.03 18.76 98.28 54.75 93.18 81.29 15.78 90.22 34.07

Doka 33.47 101.34 102.86 19.89 98.53 51.68 91.99 82.32 14.72 87.44 35.01 Glabat 31.67 103.34 102.89 20.03 99.97 50.78 92.92 83.54 13.63 88.18 37.38 Kosti 33.39 100.36 101.29 19.82 99.12 50.83 95.13 80.31 15.65 88.15 38.32

Hissahisa 31.96 101.05 98.96 19.00 97.05 53.03 91.70 80.60 15.09 88.97 35.00

Juba 31.67 105.70 100.27 19.36 99.32 53.23 95.34 78.83 14.73 91.05 43.72

Bango 31.83 104.53 100.42 18.88 99.13 56.13 95.51 74.66 23.60 93.72 40.92

Liria 32.02 102.98 100.67 18.66 99.60 53.19 95.57 79.76 14.46 89.15 40.44 Khour- Maqure 32.22 103.46 103.59 19.35 98.84 54.40 92.86 79.22 13.76 89.49 40.87

121 Table (6): Means of measurements of some tergites and sternites of the Sudanese honeybee Apis mellifera L. (mm.) Sample Length Length Length Length Width Distance T3 Length T3 +4 S3 Wax wax Between T4 Mirror Mirror W/mirror S3 S3 S3 Gala- Elnahal 2.07 2.05 4.06 2.55 1.16 2.04 0.31

Madani 1.96 1.91 3.82 2.45 1.13 2.02 0.34

Khartoum 1.94 1.9 3.79 2.4 1.11 1.96 0.32

Shendi 2.06 2.02 4.03 2.42 1.11 1.94 0.32

Kadogli 2.09 1.98 4.02 2.5 1.13 2 0.33

Eldalang 2.09 2.02 4.05 2.53 1.14 1.97 0.34

Wau-Sholok 2.03 1.99 3.97 2.48 1.11 2.03 0.34

Malakal 2.03 2 3.98 2.47 1.11 2.01 0.35

Doleib. 2.06 2.01 4.01 2.53 1.12 2.02 0.32

Gannal. 2.05 2.03 4.02 2.57 1.16 1.99 0.36

Elhawata 2.01 2 3.96 2.51 1.07 1.99 0.38

Doka 2.06 1.98 3.98 2.48 1.12 2 0.31

Glabat 2.05 2.01 4.01 2.58 1.15 2.03 0.33

Kosti 1.99 1.94 3.88 2.52 1.11 1.98 0.3

Hissahisa 1.99 1.95 3.88 2.56 1.11 2.02 0.36

Juba 2.07 1.95 3.96 2.42 1.08 1.94 0.34

Bango 1.99 1.94 3.87 2.36 1.08 1.9 0.34

Liria 1.98 1.91 3.84 2.36 1.03 1.9 0.35

Khour-Maqure 1.97 1.92 3.84 2.44 1.07 1.92 0.36

122

Table (7): Means of measurements of pilosity of the Sudanese honeybee workers Apis mellifera L. (mm.) Samples Length of Width of Width of Tomentum hair tomentum Dark Index stripe 0.21 0.88 0.45 2.91 Gala Elnahal 0.21 0.98 0.26 5.68 Madani 0.21 0.93 0.45 3.04 Khartoum 0.21 0.92 0.50 2.74 Shendi 0.26 0.87 0.53 2.44 Kadogli 0.19 0.94 0.53 2.66 Eldalang 0.18 0.79 0.45 2.62 Wau- Sholok 0.19 0.99 0.43 3.43 Malakal 0.22 0.91 0.46 2.93 Doleib.

0.20 0.87 0.41 3.10 Gannal.

0.22 0.85 0.49 2.61 Elhawata 0.18 0.96 0.51 2.82 Doka

0.17 0.98 0.66 2.21 Glabat

0.21 1.05 0.26 5.90 Kosti

0.17 0.92 0.36 3.74 Hissahisa 0.17 0.83 0.41 3.03 Juba

0.23 0.83 0.49 2.54 Bango 0.22 0.59 0.38 2.31 Liria 0.24 0.81 0.40 3.00 Khour-Maqure

123

Table (8): Means of measurements of sternite 6 of the Sudanese honeybee workers Apis mellifera L. (mm.) Samples Length of Width of Abdominal S6 S6 Slenderness L/W.

Gala Elnahal 2.37 2.76 85.20 Madani 2.23 2.72 81.31 Khartoum 2.22 2.59 84.94

Shendi 2.22 2.64 83.26 Kadogli 2.34 2.75 84.18

Eldalang 2.33 2.79 82.96

Wau- Sholok 2.37 2.77 84.66

Malakal 2.31 2.76 82.92

Doleib. 2.38 2.75 85.94 Gannal. 2.49 2.83 87.31 Elhawata 2.37 2.74 85.64 Doka 2.28 2.72 82.95 Glabat 2.37 2.77 84.82 Kosti 2.32 2.68 86.01 Hissahisa 2.36 2.75 84.80 Juba 2.27 2.66 84.89 Bango 2.26 2.67 83.80

Liria 2.24 2.61 85.16 Khour-Maqure 2.24 2.63 84.51

124

Table (9): Means of measurements of colouration of the Sudanese honeybee workers

125

Table (10): Mean, minimum, maximum and standard deviation of each morphometric character from the 285 Individual bees Apis mellifera L. measured (measurements in mm. Angels in degree).

126

Table (10) continued:

127

4-1- i- h- Mean, minimum, maximum and standard deviation of each morphometric character from the 285 individual bees measured: The minimum, maximum and standard deviation values of each morphometric character for the 285 individual worker bees indicated the presence of a wide range of variability within the honeybees collected in this study., (Table 10). 4- 1- i- i- Sum of squares, dF, mean square, F values and significances for each phenotypic character from the measured individuals. (Sudanese honeybees Apis mellifera L.). The analysis of variance means for the phenotypic characters of 19 colonies also revealed the existence of extensive variability in most of the measured characters across all localities (p ≤ 0.005). Characters which had high F values are as follow (from the highest values to the lowers ones), width of the wax mirror on sternite3 (17), width of the forewing (22), length of the sternite3 (15), length of the cubital vein a (27), wing venation of the angle O26 (41), length of the proboscis (4), width of the sternite6 (20), length of the wax mirror on sternite3 (16), length of the forewing (21), length of the tergite3 (13), width of the metatarsus (8), length of the sternite6 (19), pigmentation of the labrum 1 (25), and the wing venation of the angle G18 (35) [Numbers between brackets are the code numbers of the character measured as in Table (2) ]. But generally, they were all had high F values than the rest of the other characters, the summary of the analysis of variance of all the characters with their respective F values As shown in Appendix (H). 4- 1- ii - Multivariate analyses: 4- 1- ii- a- Principal components analysis (PCA):

128 The principal components analysis of morphometric characters of 19 colony means data was used to detect the presence of possible cluster groups of colonies. According to this analysis three factors with eigenvalues greater than 1 were exracted. The a cumulative eigenvalue of these factors was 18.179. Factor 1 had the highest eigenvalue of 12.347 followed by factors 2 and 3 with eigenvalues of 3.458 and 2.374 respectively (Table 12). Except some wing venation angles as D7, G18, K19, L13, and some body size characters as cubital vein 2 length, distance between wax mirror in sternite 3 and hair length which had absolute value of factor loadings less than 0.50; the remaining characters had absolute value of factor loading between 0.51 – 0.90. Based on varimax rotation factor loadings analysis, characters such as cubital vein 1, forewing venation angle E9, hind leg femur length, forewing

venation angle J16, forewing length, sternite3 length, sternite 6 length, tergite 3 length, tergite 4 length, hind leg metatarsus length, forewing venation angle N23, forewing venation angle O26, hind leg tibia length, forewing width, sternite 6 width and hind leg metatarsus width all had high loading values in factor 1 and 37.414 % of the variance in the data is attributed to

this factor. While characters such as forewing venation angles A4, B4, and

J10; wax mirror in sternite 3 length, tomentum 1 length and wax mirror in sternite 3 width all had high loading values in factor 2, this accounted for 10.479 % of the variance in the data. Thus colouration characters such as pigmentation on tergites 2, 3 and 4; scutellum 1 and 2 plus tomentum 2 length all had high loading values in factor 3 and accounted 7.194 % of the variance in the data. Generally, these three factors accounted for 55.087 % of the variance from the data used in this analysis. The values of factor loading of each character in the respective group of extracted factors were shown in Table 129 (12). The scatter plot graph of factor scores of factor 1 and 2 and that of factor 1 and 3, of the principal components analysis of the 19 colony means of all morphometric characters revealed the presence of three possible groups of clusters of colonies (Figure 13 & 14). Three clear scattered clusters of samples were formed in this graph figure (13). The first cluster (dark green colour) was completely separated from the others, consisted of the forest zone samples. While the middle dark orange colour cluster composed of the semi-desert zone samples and the last largest cluster (light blue colour) consisted of savannah zone samples, in which three samples; Doka No. 9, Kosti No. 13 and Galla- El Nahal No. 7 were considered to be within the semi- desert zone samples.

130 Table (12) Factor loadings in varimax rotation for each character in the principal components analysis. Component Characters Factor 1 Factor 2 Factor 3 Wing angle A4. -0,736 Wing angle B4. 0,766 Cubital vein 1 length 0,769 Cubital vein 2 length Wing angle D7. Distance between wax mirror in sternite 3. Wing angle E9. 0,610 Hind leg femur length 0,652 Wing angle G18 Hair length. Wing angle J10 0,679 Wing angle J16 -0,765 Wing angle K19 Wing angle L13 Forewing length 0,752 Sternite 3 length 0,696 Sternite 6 length 0,728 Tergite 3 length 0,544 -0,516 Tergite 4 length 0,716 Metatarsus length 0,721 Wax mirror length 0,554 -0,610 Wing angle N23. -0,587 Wing angle O26. -0,507 Tergite 2 pigment 0,754 Tergite 3 pigment 0,750 Tergite 4 pigment 0,876 Scutellum 1 pigment 0,796 Scutellum 2 pigment 0,564 Hind leg tibia length 0,667 Width of tomentum 1 -0,512 Width of tomentum 2 -0,582 Width of forewing 0,899 Width of sternite 6 0,792 Width of metatarsus 0,772 Width of wax mirror 0,559 -0,680 Explained variance 12.347 3.458 2.374 Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. A Rotation converged in 6 iterations.

131

Fig (13): Scatter plot graph of factor scores of factor 1 and factor 2 from principal components analysis of 19 colony means of all morphometric data. Horizontal axis: factor 1; Vertical axis: factor 2. (Numbers in the scatter plot graph indicate the locality code). 1- Shendi, 3- Khartoum, 4- El-Hissahisa, 5- Madani, 7- Galla- Elnahal, 8- El- Hawata, 9- Dokka, 10- El-Galabat, 13- Kosti, 18- El-Dalang, 19- Kadogli, 20- Wau- Sholok, 21- Malakal, 22- Doleib, 23- Ganal, 29- Juba, 30- Bango, 31- Lirria, 32- Kour- Maquire.

132 Fig (14): Scatter plot graph of factor scores of factor 1 and factor 3 from principal components analysis of 19 colony means of all morphometric data. Horizontal axis: factor 1; Vertical axis: factor 3. (Numbers in the scatter plot graph indicate the locality code).

1- Shendi, 3- Khartoum, 4- El-Hissahisa, 5- Madani, 7- Galla- Elnahal, 8- El- Hawata, 9- Dokka, 10- El-Galabat, 13- Kosti, 18- El-Dalang, 19- Kadogli, 20- Wau- Sholok, 21- Malakal, 22- Doleib, 23- Ganal, 29- Juba, 30- Bango, 31- Lirria, 32- Kour- Maquire.

133 From side of view of the morphometric methods of honeybees classification an attempt was made to establish the taxonomic states of the target ninteen colonies of Sudanese honeybee Apis mellifera in relation to the rest of the African Apis mellifera subspecies. The samples were analysed by PCA with 239 samples (from the data bank, University of Frankfurt, Institut für Bienenkunde, Oberursel, Germany). These included 18 samples: Ethwest (west of Ethiopia Mellifera bees), 19 samples Ethmount (Ethiopia Mountain Mellifera bees), 24 samples Lamarkii, 26 samples Jemenitica, 9 samples Litorea, 47 samples Scutellata, 27 samples Monticola, 50 samples Adansoni and 22 samples Jemenitica- Sudan (Mogga 1987). Atotal of 258 samples were processed. Graphical presentation of the results were shown in Figure (15). In figure (15), ninteen of the Sudanese honeybees samples showed three distinguishable clusters on the uper side of the scatter plot graph among the Jemenitica and Jemenitica-Sudan samples. These three clusters were composed as follows: Samples originating from the forest zone to the left uper side of the graph; samples originating from the savannah zone to the left uper side of the graph, closing to Jemenitica-Sudan and somewhat near to some Scutellata samples; while cluster originating from the semi-desert zone to the left uper side of the graph within Jemenitica-Sudan and between Jemenitica and savannah zone sampls. Lamarkii and Adansoni samples formed two different clusters but, near to each other at the right uper side of the graph. Ethmount samples formed a clear seperate clusters to the 134 right downwards side of the graph, also Ethwest samples were scattered in the downwards midle side of the graph between Ethmount and Scutellata samples. Most of Monticola samples were clustering at the left downwards side of the graph. In figure (15), most Sudanese samples were distributed among and arround Jemenitica and Jemenitica-Sudan races. However, Sudanese clusters were distinguishable. The forest and semi-desert samples were distributed in between Jemenitica and Jemenitica-Sudan clusters at the left side of the graph, thus they were near to some Monticola and Scutellata samples; while the savannah samples were distributed near to the centre at the uper side of the graph. Ethmount, Ethwest and Lamarkii samples were distributed at the downwards side of the graph, quite seperate from Apis mellifera jemenitica samples.

135

Fig (15): Scatter plot graph of factor analysis of 239 samples of worker honeybees of different origin. Horizontal axis: factor 1;

Vertical axis: factor 2.

136 4- 1- ii- b- Discriminant analysis: In a further inspection and scrutiny on the three cluster groups observed in the principal components analysis, some step-wise discriminant analyses tests were done. In this case the colony means data were also used. There are some discriminant characters for each of the three clusters which analysed as shown in Tables (13, 14 and 15). The three clusters showed some clear morphological and geographical relationship. Table (13) represented samples from the semi-desert zone, Table (14) represented samples from the savannah zone and Table (15) represented samples from the forest zone. The variables used in the discriminant analysis are: L. Hair, L. Femur, and L.Tib. L.Tar., W.Tar., L. Probo., P.T2., P.T3., P.T4., L.T3., L.T4., L.ST3., L.WM., W.WM., D.WM., L.ST6., W.ST6., L.FW.,W.FW., SCUT1., SCUT2., Cub1., Cub2. Angles A4. B4. D7., E9., G18., J10., J16., K19., L13., N23., and O26. In this analysis each colony was assumed to have a priori probability of being in any particular cluster. Based on this analysis as shown in Table (16), out of 4 colonies grouped in the semi-desert cluster all of them or 100% were correctly classified in this cluster. Forest cluster had the same results as semi-desert in which out of the 4 colonies grouped in the cluster or 100% were all correctly classified in the forest cluster. Considering savannah cluster out of 11 colonies grouped in this cluster 10 of them were correctly classified as savannah cluster. The remaining one colony from kosti was in semi-desert cluster, the PC rather thinks Kosti is semi-desert

137 with P= 0.509. Generally a 94,7% of original grouped cases correctly classified. By using the pair-wise group comparison test Table (17), the separation of cluster groups was highly significant for all cluster groups. Between the clusters semi-desert and savannah, the separations of F values obtained were: 10.435 (the highest one) and 0.618 (the lower one). While between semi-desert and forest the separations of F values obtained were: 21.812 (the highest one) and 13.521 (the lower one). Also the separation of the F value between savannah clusters and forest obtained were 45.647 (the highest one) and 22.087 (the lower one) as shown in Table (17). The discriminant analysis of the colony means data also confirmed the presence of three morphoclusters of colonies as shown in figure (17). Thus average linkage between the group centroids analysis Table (20), shows the different distances between the groups (semi-desert cluster to savannah 4.336; semi-desert to forest 6.108 and savannah to forest 5.941). In more details using the average linkage between the groups as in figure (18), colonies Juba, Lirria, Kour- Maquire were very closed to each other and Bango close to them too, (the forest cluster). Also colonies Shendi, Khartoum, Kosti and Madani appeared to be one cluster (semi-desert cluster) than the rest of the clusters. Thus colonies El-Dalang, Kadogli, Malakal, Wau- Sholok, El-Hissahisa, El- Hawata, then El-Galabat, Doleib and Ganal were all appeared to be one cluster (the savannah cluster). Only one colony

138 “Dokka” was far away from semi-desert and forest clusters though somewhat near to savannah cluster. For more clarification to the previous results obtained by the PCA analysis of the relationship between the target 19 Sudanese Apis mellifera colonies and some other African mellifera subspecies. A discriminant analysis was carried out in Table (18) demonstrates a discriminant analysis probability (regarding the individual colonies) in which colonies Gala- El nahal, El-Dalang, Wau- Sholok, Malakal, Ganal, El- Hawata, then El-Galabat, Kosti, El-Hissahisa, Juba and Lirria were all had posterior probability of 1.00 for being in cluster jemenitica- Sudan. Colony Bango had 78% posterior probability for being in cluster jemenitica-Sudan. Only one colony “Kadogly” had a low posterior probability (46%) for being in cluster jemenitica- Sudan. The previous results were summarized in table (19b) in which all the 19 colonies considered as one group (ungrouped cases), 18 colonies or 94.7% had posterior probability for being in cluster jemenitica-Sudan. One colony had a very low posterior probability (5.3%) for being in cluster Adansoni. Thus in Table (18), the probabilities without Jemenitica- Sudan (with other Mellifera subspecies); colonies Wau-Sholok, El-Hissahisa, El-Hawata and Doleib had a high posterior probability (1.00) for being in cluster Adansoni. Only one colony “Shendi” had a high posterior probability (1.00) for being in cluster Lamarkii. Dokka colony had a high posterior probability (1.00) for being in cluster Scutellata, while El- Dalang, El-Galabat, Gala-Elnahal, had 98%, 92%, and 85% posterior probability for being in Scutellata cluster respectively. 139 Thus colonies Khartoum, Lirria, Madani and Kosti had 67%, 66%, 63% and 50% respectively probability for being in cluster yemenitica. The rest probabilities were somewhat weak. The discriminant analysis probability between the data bank African mellifera subspecies (Apis mellifera jemenitica- Sudan, A. m. jemenitica, A. m. scutellata, A. m. monticola, and A. m. litorea) was shown in Table 19a. Out of 22 colonies grouped in cluster Jemenitica-Sudan 19 or 86.4% of them were correctly classified as jemenitica-Sudan cluster, the remaining three colonies one (4.5%) was in cluster Jemenitica and one (4.5%) was in cluster Litorea and the last one (4.5%) in cluster Scutellata. Out of 26 colonies grouped in cluster Jemenitica 24 or 92.3% of them were correctly classified as Jemenitica cluster whereas, the remaining two colonies were, one (4.5) was in cluster Litorea and the second (4.5) in Scutellata cluster. Out of 50 colonies grouped in cluster Scutellata 44 or 88% of them were correctly classified in cluster Scutellata while, the remaining 6 colonies were classified 2 colonies (4.0%) to the following clusters: Jemenitica, Litorea and Monticola respectively. In the same Table out of 9 colonies grouped in cluster Litorea 8 or 88.9% of them were correctly classified in cluster Litorea the rest one or 11.1% was classified in cluster Jemenitica. Also out of 27 colonies grouped in cluster Monticolla 26 or 96.3% of them were correctly classified in cluster Monticola, while the remaining one (3.7%) was in cluster Scutellata. Discriminant analysis of the distances between the group centroids (the three predicted clusters: semi-desert; savannah; 140 forest and the data bank African Mellifera subspecies) Table (20) and figure (19), in which the semi-desert cluster colonies centre was far away from the following clusters: Ethmount, Ethwest, Monticola, Lamarkii, Scutellata, Litorea, Adansoni, Jemenitica, and Jemenitica-Sudan respectively. Cluster savannah colonies centre was far away from Ethmount, Ethwest Monticola, Lamarkii, Litorea, Jemenitica, Adansoni, Scutellata, and Jemenitica-Sudan clusters respectively. While forest cluster colonies centre was far away from clusters; Ethmount, Ethwest, Monticola,Lamarkii, Scutellata, Adansoni, Litorea, Jemenitica, and Jemenitica-Sudan respectively. The discriminant analysis of the distances between the group’s centroids Table (20) and figure (20) demonstrated the independences of the different clusters.

141

Table (13): Means of some discriminant characters for the Sudanese Honeybee workers Apis mellifera (mm.). From The semi desert region.

142

Table (14): Means of some discriminant characters for The Sudanese honeybee workers Apis mellifera (mm.). From The Savannah region.

143

Table (15): Means of some discriminant characters for The Sudanese honeybee workers Apis mellifera (mm.). From The Forest region.

144 Table (16): Classification matrices of colonies in cluster groups based on step-wise discriminant analysis. (Sudan samples only). Predicted Group Total No. of Samples Membership Colonies Semi- Savan desert nah Forest Original- Semi- 4 0 0 4 count desert Savanna 1 10 0 11 Forest 0 0 4 4 % Of Semi- 100,0 ,0 , 0 100,0 classification desert Savanna 9,1 90,9 , 0 100,0

Forest , 0 , 0 100,0 100,0

a 94,7% of original grouped cases correctly classified.

Fig (17): Canonical Discriminant Function (Sudan samples only).

145

Fig (18 ): Dendrogram using average linkage

Between the groups (Sudan colonies only).

146 a,b,c,d,e Table (17): Pair-wise Group comparison . (Sudan samples).

Step Samples Semi-desert Savannah Forest 1 Semi- F , 618 15,093 desert Sig. , 443 , 001 Savannah F , 618 30,154 Sig. , 443 , 000 Forest F 15,093 30,154 Sig , 001 , 000 2 Semi-desert F 2,501 18,685 Sig , 116 , 000 Savannah F 2,501 45,647 Sig. , 116 , 000 Forest F 18,685 45,647 Sig , 000 , 000 3 Semi-desert F 7,183 13,521 Sig , 004 , 000 Savannah F 7,183 28,900 Sig , 004 , 000 Forest F 13,521 28,900 Sig , 000 , 000 4 Semi-desert F 5,823 19,799 Sig , 007 , 000 Savannah F 5,823 29,105 Sig , 007 , 000 Forest F 19,799 29,105 Sig , 000 , 000 5 Semi-desert F 10,435 21,812 Sig , 000 , 000 Savannah F 10,435 22,087 Sig , 000 , 000 Forest F 21,812 22,087 Sig , 000 , 000 a. 1, 16 degrees of freedom for step 1 b. 2, 15 degrees of freedom for step 2. c. 3, 14 degrees of freedom for step 3. d. 4, 13 degrees of freedom for step 4. e. 5, 12 degrees of freedom for step 5.

147 Table (18) Discriminant analysis probability:

With Jemenitica- Without Jemenitica Sudan. Sudan. Sample Locality Probability Assigned Probability Assigned Gala Elnahal Semi- desert 9 1.000 6 0.855 Madani Semi- desert 9 0.850 4 0.639 Khartoum Semi- desert 9 0.920 4 0.677 Shendi Semi- desert 9 0.720 3 0.999 Kadogli Savannah 9 0.460 4 0.300 Eldalang Savannah 9 0.990 6 0.988 Wau- Sholok Savannah 9 0.990 8 1.000 Malakal Savannah 9 0.990 8 0.979 Doleib. Savannah 9 0.980 8 0.992 Gannal. Savannah 9 0.990 8 0.606 Elhawata Savannah 9 0.990 8 0.996 Doka Savannah 9 0.990 6 0.991 Glabat Savannah 9 0.990 6 0.929 Kosti Savannah 9 0.990 4 0.506 Hissahisa Savannah 9 0.990 8 1.000 Juba Forest 9 1.000 8 0.862 Bango Forest 9 0.780 8 0.973 Liria Forest 9 0.990 4 0.660 Khour- Maqure Forest 9 0.850 5 0.673

3 = Lamarkii; 4 = Jemenitica; 5 = Litorea; 6 = Scutellata; 7 = Monticola; 8Adansoni; 9 = Jemenitica- Sudan.

148 Table (19a): Discriminant Classification results (Sudan and others).

149 Table (19a) continued: Discriminant Classification results (Sudan and others). A 92,1% of original grouped cases correctly classified.

150 le (19b): Discriminant Classification results (Sudan and others).

151 ble (20): Proximities discriminant centroid distances between the groups (dissimilarity matrix).

152

Fig (19): Canonical Discriminant Function (Sudan and others).

153 Fig (20 ): Dendrogram using average linkage between the groups (Sudan and others).

154 4- 2- Mitochondrial DNA analysis of Apis mellifera L.: PCR- amplified DNA can be used for direct analysis (length variation), restriction (Hall and Smith 1991; Mortiz et al., 1994) or sequence (Cornuet et al., 1991) analysis. The intergenic region analysed in this work is located between the tRNAleu gene and the COII gene. The variability of this region results from: a- The intergenic region amplified PCR products size in which The PCR product was electrophoretically separated in a 1.5% agarose gel at 100 V for 2.5 h. The gels were stained with ethidium bromide and photographed over a UV light screen. then later from the gel photo the fragment total size will estimated in comparison with the marker (added on the gel). b- The Dra 1 restriction patterns length variation of the PCR products which rune in 8% and 10% Acryl amide gel, then stained with ethidium bromide. As in the above the restricted patterns lengths was estimated from the gel photo depending on the specific marker added. c- Sequenced intergenic region PCR products; in which the variability results from the superimposition of length variation (presence / absence of the P sequence, number of reiterated Q sequences, possible small deletions in both) and nucleotide substitutions. Sequence P0 = 67 bp. (only found in African lineages); sequence Q has variable number of copies (192 – 196 bp.) about 200 bp. (Garnery et al., 1993.) Structure and part of the nucleotide changes are accessible through PCR amplification of the entire region and analysis of the restriction profiles obtained with the enzyme Dra 1 (TTTAAA). In this work restriction method was applied.

155 4- 2- i- Amplified PCR analysis: Due to size and Dra l restriction site polymorphisms, a total of six products of different size of the tRNAleu – CO II intergenic region were observed after PCR amplification as shown in Table (21a); Figure (21a) and (21b). One corresponds to the lineage A1 and the others to, A2, A4, Y2, O1, or O1` lineages. In Figure (21a) and (21b) according to Cornuet et al., (1991), the amplified PCR products demonstrated two length categories of fragments

corresponding to the following combinations; Lane 25, P0QQ (A4); Lane

24, P0Q (A1); Lane 3, P0QQ (Y2); Lane 16, P0Q (O1); Lane 11, P0QQ

(O1`); Lane 13, P0Q (O1); Lane 9, P0QQ (O1`); Lane 23, P0Q (O1); Lane 8,

P0QQ (A2); Lane 33, P0Q (A1); Lane 19, P0Q (O1); Lane 31, P0Q (A1);

Lane 17, P0QQ (A4); Lane 5, P0Q (O1) and Lane 20, P0Q (O1). [Lane M in all the gels is the molecular weight markers (50 bp ladder, Gibco BRL). The

bands at ~ 640 bp are of size P0Q, and the bands of ~ 820 bp are of size

P0QQ.]. 4- 2- ii- Restriction analysis: Dra 1 (RFLP) of the tRNAleu - COII intergenic region exhibit from 3 to 5 bands, provided a total of 6 different haplotypes among the 27 Sudanese honeybee workers colonies analysed in this study as shown in the 8% and 10% Acryl amide gels in figures (22a), (22b), (22c), (22d) and (22e), those haplotypes are: A1 (Lane 31, 24, 33, 32, 34, 29, 30, and Lane 35); A2 (Lane 8 and Lane 7); A4 (Lane 25, 17, 26 and Lane 25); O1 (Lane 19, 23, 5, 20, 13, 16, 6, 18, 21, 22, 23 and Lane 4); Y2 (Lane 3); O1` (Lane 9 and Lane 11). Lane M in all the gels is the molecular weight markers (50 bp ladder, Gibco BRL). Results of the different haplotypes obtained from the Sudanese honeybee workers were summarized in Table (21a).

156 For more scrutiny to the results obtained in Table (21a), another analysis was taken, by using the cluster column with a 3rd visual effect and pie with a 3rd visual effect analysis methods the percentage distribution of the haplotypes (O1, A1, A4, A2, O1`and Y2) of the studied colonies as a whole were detected as shown in figures (23a and 23b). In regards to the three target geographical zones individually haplotypes distribution were depicted in the semi-desert zone colonies, as shown in Table (21b) and Figure (24a). In the savannah zone colonies demonstrated in Table (21b) and figure (24b). For the forest zone colonies as shown in Table (21b) and figure (24c).

157

Fig (21a): Agarose gel (1.5%) with the amplified PCR products of the Sudanese honeybee workers (samples 25, 24, 3, 16, 11, 13, 9, 23, 8, 33, 19, and 31).

Fig (21b): Agarose gel (1.5%) with the amplified PCR products of the Sudanese honeybee workers (samples 17, 5, 20, and 25).

158

Fig ( 22a): Acrylamide gel (8%) showing the Dra 1 restriction patterns of the Intergenic region of the Sudanese honeybee worker samples. (samples 31, 19, 24, 23, 5, 20, 33, 8, 13, 16, 9, 25, 11, and 17 Big fragments).

159

Fig (22b): Acrylamide gel (10%) showing the Dra 1 restriction patterns of the Intergenic region of the Sudanese honeybee worker samples. (samples 31, 19, 24, 23, 5, 20, 33, 8, 13, 16, 9, 25, 11, and 17 Small fragments).

160

Fig (22c ): Acrylamide gel (8%) showing the Dra 1 restriction patterns of the Intergenic region of the Sudanese honeybee worker samples. (Samples 26, 25, 6, 32, and 34).

161

Fig (22d): Acrylamide gel (8%) showing the Dra 1 restriction patterns of the Intergenic region of the Sudanese honeybee worker samples. (Samples 31, 18, 21, 22, 23, 7, 5, and 3).

162

Fig (22e):Acrylamide gel (8%) showing the Dra 1 restriction patterns of the Intergenic region of the Sudanese honeybee worker samples. (Samples 29, 4, 5, 30, 23, 13, 35, and 6).

163 Table (21a): Sudanese honeybee workers Apis mellifera L. different haplotypes. Colony PCR ID. Location size Dra 1 fragment sizes (bp). Haplotypes 25 Sha- Mahadi P0QQ 47 108 193 483 A4 17 Om-Rawaba. P0QQ 47 108 193 483 A4 26 Kafindapi P0QQ 47 108 193 483 A4 8 Hawata P0QQ 47 108 670 A2 7 Galla-Elnahal P0QQ 47 108 670 A2 29 Juba P0Q 47 108 483 A1 30 Bango P0Q 47 108 483 A1 31 Lirria P0Q 47 108 483 A1 35 Kajiko P0Q 47 108 483 A1 24 Raja P0Q 47 108 483 A1 33 Yei P0Q 47 108 483 A1 Kour- 32 Maquire P0Q 47 108 483 A1 34 Latokka P0Q 47 108 483 A1 19 Kadogli P0Q 47 65 108 420 "O1" 23 Ganal P0Q 47 65 108 420 "O1" 4 El-Hissahisa P0Q 47 65 108 420 "O1" 5 Madani P0Q 47 65 108 420 "O1" 20 Wau-Sholok P0Q 47 65 108 420 "O1" 13 Kosti P0Q 47 65 108 420 "O1" 16 Damazin P0Q 47 65 108 420 "O1" 6 Erkawit P0Q 47 65 108 420 "O1" 18 El- Dalang P0Q 47 65 108 420 "O1" 21 Malakal P0Q 47 65 108 420 "O1" 22 Doleib P0Q 47 65 108 420 "O1" 9 Dokka P0QQ 47 65 108 130 420 "O1'“ 11 Semsim P0QQ 47 65 108 130 420 "O1'" 164 3 Khartoum P0QQ 47 65 92 130 482 "Y2"

165

Table (21b):Distribution of the sudanese honeybee worker haplotypes according to the three different geographical zones. Geographical Haplotypes Time zone found numbers of the haplotypes Semi-desert O1 3 zone Y2 1 Savannah A4 3 zone A2 2 O1 8 O1` 2 Forest zone A1 8

166

Fig (23a ): Pie with 3rd visual effects representing the distribution percentage of the Sudanese honeybee worker haplotypes of the studied 27 colonies.

Fig (23b): Cluster column with 3rd visual effects representing the distribution percentage of the Sudanese honeybee worker haplotypes of the studied 27 colonies. 167

Fig (24a): Pie with 3rd visual effects representing the distribution percentage of semi-desert zone of Sudanese honeybee worker haplotypes.

168

Fig (24b): Pie with 3rd visual effects representing the distribution percentage of savannah zone of Sudanese honeybee worker haplotypes.

169

Fig (24c): Pie with 3rd visual effects representing the distribution percentage of forest zone of Sudanese honeybee worker haplotypes.

170 4- 3- observations on some behaviour and biology of the Sudanese honeybees Apis mellifera L.: Understanding the biology and ecology of honeybees of an area is very important not only for classification purposes but also for the efficient and profitable management of honeybees according to their biological behaviour and respective ecology. Fore this reason, the following observations were included. 4- 3- a- Coloured colonies: In the east region of semi-desert and savannah zones, towards the Sudan- Ethiopia border, mixed colonies of black and yellow honeybees were common. The ratio of black to yellow; however from Dokka ‘ one of the studied sample areas; 100 bees were taken randomly. The numbers of black to yellow bees were counted. A ratio of 59: 41 black to yellow bees was found. This colony was one of the most defensive and nervous colonies. Worker bees started attacking about 6 to 8 meters away from the hive. By the time the hive was opened, almost 95% or the whole hive population was out knocking the intruders and each other nervously. Another colony at El- Galabat from the same region has somewhat, the same ratio of black to yellow bees. Also colonies of mixed bees were also reported by Mohamed (1982) and Mogga (1988). Thus colonies of all yellow bees were also found in this zone as in El-Hawata and Galla-Elnahal regions. 4- 3- b- Defensive behaviour: Generally all colonies from which samples were collected defended themselves effectively. Feral bee colonies of the semi-desert zones, Madani, El-Hissahisa, and Shendi, attacked at distances ranging between 150 m. to about half kilometre away from their colonies, Shendi colony was the most defensive and nervous colony in this region, we have been told from some natives that, this colony killed a man few months ago. 171 Samples from savannah zone specially that from rock crevice as in Kadogly was very defensive. A large bee population persisted followed us nearly one and half kilometre away and the natives believed that this colony is a magic colony and its 20 years old. Equally, samples from the forest zone were very defensive once disturbed. In the central region of the semi-desert zone at Khartoum, the only collection trip from a colony bees kept in Langstroth hive; were equally very defensive, with which a day before our visit we told the surrounded natives that, tomorrow they have to take care and kept inside doors. The same with Kosti colony in which we dig a half-meter under the ground so as to reach the colony combs. So with this experience, it could be concluded that, the Sudanese honeybees were very defensive and aggressive. 4- 3- c- Swarming and migration: Reproductive swarms or seasonal migration of honeybees in Sudan occurred in the different geographical zones at different times of the year, depending on the climatic conditions. Ruttner, 1988 demonstrated that, seasonal migration of honeybees is considered as a unique characteristic of tropical honeybees. Honeybees of most sub-Saharan Africa are reported to migrate on a seasonal basis, following dry periods: A. m. yemenitica (Rashad and El-Sarrag, 1978; Peterson, 1985; Sawadogo, 1993; Woyke, 1993). In the semi-desert and most of savannah regions, swarming took place between August and October. In forest zone it occurred between February and April, and to a lesser extent around June and August to September. Migration swarms seem to be common in areas devoid of water and forage during dry months of the year. This occurred in the poor savannah zone. In the rain season migration swarms move from the riverbanks and established

172 away from the Niles, while in the dry season they established on the Niles banks. 4- 3- d- Nesting sites: Honeybees in Sudan construct nests in different areas with different conditions, some times within one area you may find two different nesting sites. For instance, in Kosti (savannah region) we found a colony nest constructed in the ground crevices to a depth of about half meter or more; another colony nesting in a tree cavity. In the forest and some savannah zones they built nests mostly in tree cavities, rock crevices, and evacuated termite mounds. Beekeepers in south Kordofan (savannah region) believed there were two honeybee strains in their area. One strain nested mainly in termite mounds, while the second nested in tree cavities. The former was generally larger in size, docile and when found could be harvested during the day with relative ease. The later, nesting in cavities other than termite mounds, was smaller, very defensive and only accustomed beekeepers may harvest this strain at daytime.

173 4- 4- Morphometric statistical analysis of Apis florea. The results obtained in the analysis of the 4 colonies of Apis florea collected from four different towns in Sudan were shown in Tables (22 – 28). Thus there are some discriminant characters for each of the four samples analysed as demonstrated in Table (29) in which the four colonies show some morphometrical and geographical relationship. Table (30) represent the analysis of variance means for the phenotypic characters from the measured individuals of the four colonies, which revealed the existence of some variability in most of the measured characters across all localities. In a further treatment, the four colonies are analysed by PCA (cluster analysis), the results obtained were shown in figure (25). Also results were subjected to proximities discriminant centroid distances between the group’s analysis as shown in Table (31) and figure (26). The results of Khartoum and Madani are somewhat near to each other, and Gerry is more closed to El-Dender, than the other distances. In a further scrutiny, the average mean measurements of some characters (forewing length, forewing width, hind-leg length, tergite 3 plus 4 length and cubital index) of the four samples from Sudan were analysed by a cluster column (compares values across categories) with 6 Florea samples [2 from Sudan “Moggas ones” and 4 from the data bank, Institute fur Bienenkunde- Oberursel-Germany (Mogga 1988)]: Khartoum south, Toti, South India, Oman, South Iran and Pakistan. The results are presented in Table (32) and figure (27), which indicated that, the colonies were somewhat big in size like those from Pakistan and south Iran.

174

Table (22). Florea.

175

Table (23). Florea.

176 Table (24). Florea.

177

Table (25). Florea.

178 Tble (26 ): Means of measurements of some tergites and sternites of honeybee worker Apis florea (mm.) from Sudan

Length Length Length Length Length Width Distan- - Sample T3 T4 T3 +4 S3 Wax w a x bet Mirror S3 Mirror ween S3 Wax Mirror

El- Dender 150.7302 143.1693 293.8995 175.6699 147.6299 149.1466 7.8404

Gerry 145.3644 140.2425 285.6069 176.6256 151.6191 146.6624 8.1349 Khartoum 135.3645 132.1938 267.5583 169.5331 141.8922 141.0434 7.8131 Madani 150.2424 140.9742 291.2166 174.0833 145.327 146.7718 6.7861

Table (27): Means of measurements of pilosity of the honeybee workers Apis florea (mm.) from Sudan

Samples Length of Width of Width of Tomentum hair tomentum Dark Index stripe El- Dender 2.439 56.9208 32.691 1.7763 Gerry 2.439 55.767 18.0762 3.7941 Khartoum 2.439 51.5364 20.7684 2.8279 Madani 2.439 54.9978 21.153 3.0317

Table (28): Means of measurements of sternite 6 of the honeybee workers Apis florea (mm.) from Sudan Samples Length of Width of Abdominal Slenderness S6 S6 L/W. S6 El- Dender 147.6299 196.2996 75.2267 Gerry 151.6191 200.2218 75.7551 Khartoum 141.8922 190.4472 74.51444 Madani 145.327 193.7276 75.02733

179

Table (29). Florea.

180 Table (30). Florea.

181 Table (30) continued. Florea.

182 Table (31). Florea.

183 Table (32). Florea.

184 Fig (25) Florea.

185 Fig (26) Florea.

186 Fig (27) Florea.

187 CHAPTER FIVE: DISCUSSION, SUMMARY AND CONCLUSION. 5- 1- DISCUSSION 5- 1- a- Morphometrics (Apis mellifera L.): Along with the development of morphometric measurements, the introduction of different multivariate techniques like principal components and factor analyses were used to detect clusters of colonies within populations (Ruttner et al., 1978; Ruttner, 1988). The first attempt in the racial studies of the Sudanese honeybees Apis mellifera L. was conducted by Ruttner (1975), Table (33); El Sarrag (1977), Table (34); Saeed (1981), Table (35); Mohamed (1982), Table (36) and Mogga (1988), Table (37). In the present work, analysis of morphometric characters of the honeybees Apis mellifera L. of Sudan as in Tables (3- 10) revealed a wide variation in the fourty five characteristics measured. Fourteen of these formed the discriminant characters which include; forewing length and

width, wing venation angles G18 and O26, proboscis length, hind-leg metatarsus length and width, sternite 3 length, sternite 6 length and width, wax mirror length and width, cubital vein 1 length and labrium 1 colouration; Appendix (H) summerized the significances for each phenotypic character from the measured individuals at P ≤ 0.005. This wide variation might be taken as an indication that, these honeybee samples do not belong to one race. Elsarrag (1977), Saeed (1981), Mohamed (1982), all found highly significant differences for all the 12 traits studied and concluded that, this is an indication that the samples did not belong to one race. Similarly Rashad and Elsarrag (1978, 1980), and Rashad et al., 1984, Mogga (1988) and Elsarrag (1992), stated a high degree of regional variation of the Sudanese honeybees.

188

Table (33) Ruttner:

189

Table (34): The average values of measurements taken for the different characters of the S Characters Mean SE ± u Tongue length (mm.) 5.50 0.019 d Flagellum length (mm.) 2.66 0.006 a B. Tarsus iii length (mm.) 2.20 0.006 n B. Tarsus iii width (mm.) 1.11 0.003 e No. Hair rows 11.90 0.014 s Forewing length (mm.) 8.60 0.014 e Forewing width (mm.) 3.02 0.014

Cubital index 2.37 0.008 h No. Hamuli 21.40 0.019 o T + T length (mm.) 3.70 0.009 3n 4 Slenderness S L/I 86.00 0.008 e 6 % Color T 71.36 0.003 y 3 bee workers Apis mellifera L. (El Sarrag, 1977).

190

Table (35): The avera ge values of measurements taken for the different characters of the Sudanese honeybee workers Apis mellifera L. (Saeed 1981). Characters Mean SE ± Tongue length (mm.) 5.18 0.379 Flagellum length (mm.) 2.71 0.104 B. Tarsus iii length (mm.) 8.29 0.165 B. Tarsus iii width (mm.) 3.13 0.156 No. Hair rows 1.99 0.234 Forewing length (mm.) 20.57 0.775 Forewing width (mm.) 2.02 0.097 Cubital index 1.03 0.073 No. Hamuli 10.72 0.859

T3 + T4 length (mm.) 69.36 12.160

Slenderness S6 L/I 85.54 4.105

% Color T3 3.74 0.136

191

192

Table (36): The average values of measurements taken for the different characters of the Sudanese honeybee workers Apis mellifera L. (Mohamed 1982).

Characters Mean SE ± Tongue length (mm.) 5.46 0.025 Flagellum length (mm.) 2.70 0.086 B. Tarsus iii length 2.11 0.015 (mm.) B. Tarsus iii width (mm.) 1.11 0.001 No. Hair rows 11.96 0.098 Forewing length (mm.) 8.12 0.034 Forewing width (mm.) 2.85 0.015 Cubital index 2.34 0.070 No. Hamuli 20.84 0.329

T3 + T4 length (mm.) 3.86 0.025

Slenderness S6 L/I 87.56 0.713

% Color T3 66.90 1.526

194

Table (37): Average means of some discriminant characters for the Sudanese honeybee workers (Apis mellifera L.) From different geographical zones. Mogga (1988).

195

Similar variability was also observed for the neighbouring Ethiopia honeybees (Smith, 1961; Ruttner, 1975, 1988; Kassaye, 1990; Radloff and Hepburn, 1997a, 1988; and Hepburn and Radloff, 1998). Kenya honeybees (Meixner et al., 1989, and 1994). Winston et al., (1983) concluded that, the most striking aspects of honeybee biology were the variability found within and between races of Apis mellifera L. They mentioned that among these variations were morphological and physiological characteristics as colour, body size and tongue length. Ruttner and Kauhaser (1985) also found a very high degree of variability among honeybees of tropical countries compared with the total variability for Apis mellifera L. They concluded that for body size, the African portion of the total variability was 63% and for the length of proboscis was 55%. In Africa, in the absent of geographical barriers, the occurence of different subspecies are results of ecological factors was recognized (Ruttner, 1981, 1988; Dietz et al., 1986). Although the Sudanese honeybee samples for the present study originated from areas without geographical barries, they showed very clear geographical distribution pattern. The dark green colour samples displayed in the left side of the figures (13 and 14) orginated from forest zone. These samples presented most of the least measurements and the most slender abdomin of the two Sudanese honeybee subgroups. In 1982, Mohamed Ali Hussen had described the Sudanese honeybees originated from south Sudan, particulary Equatoria province (the same region of the forest zone samples) as having the shortest proboscis measurements among Apis

196 mellifera L. and they are very small in body size and they have the most slender abdomen compared to the other African and/or European honeybee strains. Samples from the forest zone were completely seperated in morphometrics from the samples of the semi-desert and savannah zones and even from Mogga (1988) samples but all of them aggregated among the Yemenitica samples of Ruttner as shown in figure (15). The dark orange samples distributed around the middle of the figures (13, 14) were those originating from semi-desert zone. These samples presented most of the medium measurements of the two other subgroups described in this study. In 1976, Ruttner had described the Sudanese bees originating from the semi-desert zones as having the most slender abdomen of all the African bees examined. He concluded that the samples clearly belong to another group, correctly to be considered as an independant race, and named them Apis mellifera nubica. In 1986, Ruttner further found Sudanese bees from the semi-desert morphometrically inseperated from the bees of Saudia Arabia, Yemen, Oman and Chad, thus he renamed the Sudanese bees Apis mellifera yemenitica. He then concluded that Apis mellifera jemenitica was the African bees from dry thorny-bush zone. It is occurence in Yemen, Oman, and Saudia Arabia was not astonishing since this was part of the Ethiopia zoogeographic zone. Thus, the results of Mogga (1988) agreed with Ruttner (1986) discription of the size and slenderness of the Sudanese bees of the semi-desert zone. The results of the semi-desert zone cluster of this study agreed with the previous mentioned discription of Ruttner (1986) as compared to the savannah zone cluster only, because forest zone bees were not considered in Ruttner investigation. However, samples from Shendi and Khartoum originated from the

197 semi-desert region along the Nile valley, in addition to their natural vegetation of various small Acacia sp. and other flowering shrubs, there are small irrigation native farms of vegetables and fruits, thus samples Madani and El-Hissahisa in spite of their origin on the Nile valley they possess the same mentioned natural vegetations, and they were among one of the most bigest national irrigation schemes in Sudan “El-Gazirra Scheme“ where various economic crops, field crops, vegetations and fruits were grown. Regarding most of the discriminant characters of the semi-desert zone colonies (Shendi, Khartoum, El-Hissahisa and Madani); some up normality in Khartoum colony measurements was found, in other wards it had the least proboscis value, at the same time it had the largest or medium values in the rest of the measurements. The large values of characters for the Khartoum sample from the semi-desert zone might be an infleunce of the imported European bees to Khartoum state. Elsarrag (1977) reported that bees of Khartoum province were hybridized with foreign honeybee races. In the dearth season, bees resorted to forage on flowering plant species growing along the Nile vallies. All of the four colonies of the semi-desert zone were aggregated around Moggas and Yemenitica samples as shown in figure (15). Considering savannah zone (the widest geographical vegetation zone), Nine samples out of eleven (light blue colour samples) were distributed to the right side of the graph figures (13 & 14). Most of these samples were from the eastern region, towards the border with Ethiopia. Samples from Om-Rawaba, El-Dalang and Kadogli were from the far west part of this region. Bees from the eastern zone of Sudan might be infleunced by the Ethiopia honeybees which were generally large. Most of these samples

198 were distributed to middle side of figure (15), among studied samples Scutellata, Monticola and Ethwest (Ethiopia western bees). Generally bees displayed to the right side of the graph in PCA consist of large bees, while those distributed at the left side of PCA graph comprises small bees. The largest average of samples from the savannah zone mean values for forewing length were 8.45 mm., width 2.95 mm.; hind-leg 7.05 mm.; body size (T3+ T4) 4.00 mm., and comparatively largest cubital index 2.24 mm. These results agreed with those found by Mogga (1988), he found honeybee samples originating from the north part of the savannah zone (poor savannah zone), to have largest values for the forewing length, 8.46 mm., width 2.94 mm.; hind-leg 7.26 mm.; body sixe 4.04 mm.; and cubital index 2.30 mm. Also the same results confirmed those found by Rashad et al., (1982). Who found bee samples originating from Nile province (savannah zone) to have largest values for the forewing length 8.56 mm., and width 3.16 mm. The present correlation between climatic zone and morphometric characters also agreed with the finding of Ruttner and Kauhaser (1985). who found in tropical Africa, significant geographical variability in honeybees inspite of the absence of physical isolation barriers. They stated that, the mechanism bringing about isolation was the selective adaptation of the races bees to certain biotopes. This principal implied permanent hybridization and existence of intermediates, reflecting the intermediate ecological zones. In the Sudan, the climatic conditions affecting the supporting vegetative forage for bees seemed to play a greater role. Graded long distance variations of quantative characters in bees referred to as “geocling“ by Huxley (1938), while the graded variations over a short distance on slopes of monutains as found in East Africa were termed as

199 “geocling“. Considering the present results, it is tempting to suggest that, geocling exist, among the Sudanese honeybees. Samples from the forest zone exhibited the smallest mean average values of forewing length 8.23 mm., and width 2.82 mm.; proboscis length 5.55 mm.; hind-leg length 6.83 mm.; body size 3.88 mm., and cubital index 1.85 mm. Values of these parameters were similar to the findings of Mohamed Ali Hussen (1982), proboscis length 5.46 mm., forewing length 8.12 mm., forewing width 2.85 and cubital index 2.34 mm. But larger in samples from savannah zone to the north and west. Forewing length 8.45 mm., and width 2.95 mm.; proboscis length 5.67 mm.; hind- leg length 7.05 mm.; body size 4.00 mm.; and cubital index 2.26 mm. Samples from the semi-desert zone gained medium mean average values of forewing length 8.27 mm., and width 2.88 mm.; proboscis length 5.63 mm.; hind-leg length 7.00 mm.; body size 3.88 mm.; and cubital index 2.04 mm. The smallest mean average values of the body different parameters in the forest zone samples, may be attributed to the small size of the cells, the honeybee workers build in their combs. Hence it is an adaptive mechanism that honeybees utlize the maximum area for building up their nests. Grout (1937) studied the influnce of cell size upon the size and variability of the honeybees. He found that the size of the brood cell affect the size of the adult worker bee and significantly larger bees were obtained through the use of enlarged cell foundation. Hassanein and Elbanby (1956) studied the biometrics of the Egyptian honeybees. They obtained great differences in measurements of the worker appendages for those reared in Langstroth hive compared to those reared in pipe-hive. The former gave higher figures. From these facts it can be said that, the small size of the forest zone honeybees refered to the small cell size they

200 build in their combs. The smallest mean average values of body parameters which found in the semi-desert zone bees (as compared to savannah zone bees). This phenomenon might be due to the influnce of the high temperatures and law humidity almost all year round in the semi-arid zone. This limits the amount of forage in this zone, compared to the forage found in the savannah zone. Nagi (1984) found that from April to June and from September to November, there was insufficient food in Shambat area (located in the semi-desert zone), resulting in construction of small queens. Therefore, lack of nectar and insufficient forage might play a role in the smallness size of the bees in the semi-desert zone. Kigtiira (1984), wrote that, bees in Kenya were characterized by a specific geographical distribution confined by natural barriers; namely Apis mellifera monticola the mountain bees (2400- 3100), then A. m. Scutellata and A. m. nubica (yemenitica) the Acacia savannah plain bees. Ruttner and Kauhausen (1985) like wise stated that, the variation of bee in tropical Africa showed new principle in bee taxonomy, hardly considered before: diversification and isolation by ecological factors. Thus it could be concluded from figure (13, 14 and 15) that, the alignment of the Sudanese honeybee samples was not by coincidence. In 1985, Mbaya also found that, the morphology and behavior of bees in Kenya varied according to ecological zone. He further said special biological modifications of some characteristics of honeybees were necessary if they were to become adapted to certain ecological zones of Kenya. The probability of the existance of at least three different ecotypes of the Sudanese bees as mentioned above through the PCA analysis, was confirmed by some advanced modern discriminant analysis methods. Such as step-wise discriminate analysis which was used to confirm the

201 separation of clusters, detect the most discriminatory variables and calculation of the percentage of correctly classified colonies (Ruttner, 1988; Daly, 1992). However, by applying a step-wise discriminate analysis Procedure, Ruttner et al., (1978); Daly and Balling (1978) showed the possibilities of discriminating one race from another using fewer numbers of selected characters based on the region under investigation. In the step-wise discriminant analysis Table (16), 94.7% of the original cases were correctly classified, in which all the semi-desert and forest colonies were 100% correctly classified as different clusters, but in case of the savannah colonies 10 out of 11 colonies were correcly classified as one cluster but the remaining colony (Kosti), the PC rather thinks it should be semi-desert (with P = 0.509) and this result was very clear in the PCA analysis graph figure (14); this apparently can be attributed to: Kosti colony might be migrated from the semi-desert zone during the dry periods as honeybees of sub-Saharan Africa were reported to be migrate on a seasonal basis, following dry periods: A. m. yemenitica L. (Rashad and El-sarrg, 1978; Peterson, 1985, Sawadogo, 1993; and Woyke, 1993) thus Kosti colony geographicaly is somewhat near to the border between the semi-desert and savannah zones. Based on pair-wise group comparison test (at different degrees of freedoms) the seperation of cluster groups (semi-desert, savannh and forest) was highly significant for all the clusters as shown in Table (17), and it conformed with the PCA results of of the univariate analysis Appendix (H). On the other hand considering the relationship between our target three clusters with Yemenitica Sudan and the other African mellifera subspecies used in this study, the discriminsant analysis probability Table (18), demonstrated that colonies Gala- Elnahal, El-Hawata, El-Galabat,

202 Kosti, El-Hissahisa, Juba and Lirria were all had posterior probability of 100% for being in cluster Yemenitica-Sudan. Colonies Khartoum, Doleib, and Dokka had 92, 98, and 98% respectively posterior probability for being in cluster Yemenitica-Sudan. While colonies Kour-Maquir, Bango, and Shendi had 85, 78, and 72% respectively posterior probability for being in cluster Yemenitica-Sudan. However, only one coloni (Kadogli) had less than 50% (46%) for being in cluster Yemenitica-Sudan. The discriminant analysis probabilities (Table 19b) confirmed and the above mentioned results, in which out of the target 19 colonies 18 (94.7%) colonies had posterior probabiliy for being in cluster Yemenitica-Sudan. Only one colony had a very law probability for being in cluster Adansoni. Considering the second choice of the discriminant analysis probability as demonstrated in Tables (18, 19a and 19b) without Yemenitica-Sudan cluster; there were some colonies showed a high posterior probability for being in another African mellifera subspecies clusters; for instance in Table (18) colony Dokka, had a high posterior probability of 100% for being in cluster Scutellata.this may be attributed to: Dokka was one of the coloured colonies in this study in which the percentage ratio of black to yellow colour is 59: 41 thus it is located in the border between Sudan and Ethiopia. Colonies Khartoum, Lirria, and Madani had a high posterior probability for being in cluster Yemenitica. Away from Sudan clusters, the probability between the other African mellifera clusters revealed a high posterior probabilities for many clusters to be in the other; for instance out of 50 colonies grouped in cluster Scutellata 44 colonies were correctly classified as Scutellata while the remaining 6 colonies were classified as: 2 colonies (4.0%) to each of the following clusters: Yemenitica, Litorea and Monticola respectively (Table 19a). To depict the distances between clusters, dendrograms and mahalanobis distances were introduced (Tomasson and Fresanaye, 1971;

203 Cornuet et al., 1975; Cornuet and Garnery, 1991a, b; Daly, 1992). By using the discriminant analysis, distances between the group centroids of the 3 target clusters: semi-desert, savannah and forest were depicted in Table (20) as: Semi-desert to savannah = 4.3; Semi-desert to forest = 6,1; Savannah to forest = 6,0.(And vice versa); which demonstrated: clusters semi-desert and savannah were more close to each other and both of them were far away from forest cluster by semi equal distances and these supported the independence of the forest cluster as it is very clear in the PCA figures (13, 14 and 15). These results strongly indicated the existence of at least three different honeybee ecotypes among the Sudanese bees (see Fig 17). The same results were confirmed in figure (19), in which the semi- desert cluster colonies centre was far away from the following clusters: Ethmount, Ethwest, Monticola, Lamarkii, Scutellata, Litorea, Adansoni, Jemenitica, and Jemenitica-Sudan respectively. Cluster savannah colonies centre was far away from Ethmount, Ethwest Monticola, Lamarkii, Litorea, Jemenitica, Adansoni, Scutellata, and Jemenitica- Sudan clusters respectively. While forest cluster colonies centre was far away from clusters; Ethmount, Ethwest, Monticola,Lamarkii, Scutellata, Adansoni, Litorea, Jemenitica, and Jemenitica-Sudan respectively. Using average linkage between the groups (Sudan samples and others) a dendrogram figure (20) was issued, in which all the clusters can be divided into five different groups as followed: Group 1 include: Jemenitica, Adansoni, Litorea and Scutellata clusters, in which Jemenitica and Adansoni were very closed to each other and Litorea; Scutellata were closed to them too. Group 2 include: Jemenitica-Sudan, savannah, and semi-desert clusters. In which Jemenitica-Sudan and savannah were very closed to each other and semi-desert was closed to them. Group 1 and 2 were closed to each

204 other. Group 3 include: forest cluster, which is closed to group 1 and 2 together. Group 4 include: Ethmonut, Ethwest and Monticola clusters; in which Ethmonut, Ethwest and Monticola colonies were very closed to each other. Group 5 include: Lamarkii, which was very close to group 1, 2, and 3. Group 4 was close to all the groups. All the different 5 groups (fig. 20) had a very cleared relationship with each other as one species (Apis mellifera.), thus it is strongly proved that Sudan colonies represent at least three different clusters. However, the above-mentioned relationship and the interaction between the bee’s subspecies are not surprising between the African honeybees clusters mainly Yemenitica, Scutellata and Monticola (North- East Africa races). The distribution and interaction between these subspecies is predicted with the following assumptions: The gene flow within African honeybees is very high due to high swarming and migratory behaviour (Hepburn and Radloff, 1988) and subsequent law molecular differentiation among African subspecies is well noted (Frank et al., 2001). Considering all the above-mentioned presentations, which confirmed the presence of three different clusters “ecotypes” among the Sudanese honeybees, Elsarrag (1992) suggested that, two honeybee subspecies exist in Sudan; namely A. m. sudanesis and A. m. nubica. The former was described to be distributed all over the Sudan, between latitude 3º N and

16º 20 N. The latter subspecies was described as mixed bees, distributed along the international boundaries of Sudan, Ethiopia and Uganda. In the present study, this last description corresponded in part with the samples from savannah zone. While the former description covered most of the forest and western parts of the savannah zones. Mogga (1988), by using

205 the first analysis method only (PCA) suggested that, three ecotypes of honeybees exist in the Sudan; namely: A. m. yemenitica L. (the small bee of semi-desert zone), A. m. sudanesis (the medium bee of forest and rich savannah zone) and A. m. bandasii (the large bee along the border of Sudan and Ethiopia). In the present study the description of the former bees (Yemenitica) corresponded with the cluster of bee samples originating mainly from the semi-desert zone; while Bandasii subspecies partially corresponded with the samples from the poor savannah. The race Sudanesis as in Mogga (1988), was medium in size and should be distributed in both rich savannah and forest zones but, in the present study the forest zone cluster was confirmed to be completely different from both semi-desert and savannah clusters through the PCA analysis (figures 13, 14 and 15) and the discriminant analysis (Tables 16 & 17and figures 17, 18, 19 &20) thus it gained the smallest size among all the Sudanese bees other clusters which conformed with Mohamed investigations (1982). 5- 1- b- Mitochondrial DNA (Apis mellifer L.): In the last decade techniques for the measurements of genetic variations in honeybees at the DNA level have been developed and are proving to be extremely powerful probes for the analysis of genetic variation. The mtDNA technique was based on examining restriction fragment length polymorphisms (RFLPs). Such techniques have been applied to honeybee’s mitochondrial DNA (Moritz et al., 1986, 1994; Smith, 1988, 1991; Smith and Brown, 1988, 1990; Smith et al., 1989, 1991). Thus study of mitochondrial DNA is of particular application in honeybees as it is the ideal marker of colony – since all individuals in the colony, queen, workers and drones sharing the same haplotype (excluding mutations). Therefore, studying 27 individuals obtained information on

206 27 colonies. This property, combined with the haploids of the genome and the high variability of some of its regions, confers a high power of resolution, as it enables precise detection of foreign haplotypes in populations. With the Dra 1 restriction enzyme test, it was possible to differentiate 6 different haplotypes (O1, A1, A4, A2, O1` and Y2) within the 27 colonies studied in the Sudan. (Table 21a) and figure 22 (a, b, c, d, and e), confirms the potential use of this technique as a powerful tool in population genetic studies of honeybees. Thus, Primarily these results presented strong evidence for the existence of more than one subspecies among the Sudanese honeybee populations. Moritz et al., (1994) in their study, the Mitochondrial DNA variability in South African honeybees (Apis mellifera L.), demonstrated that, the variability of mtDNA size of honeybees (Apis mellifera L.) in a sample of 102 colonies covering the area south of the 27th parallel of latitude in Africa were analysed using PCR and Dra I restriction enzyme. A region between the COI and COll genes revealed four different size variant haplotypes, which has been shown to be useful for the bio geographic classification of Apis mellifera subspecies, it is partially corresponded to the known distribution of African subspecies of honeybees based on morphometrical and physiological data. De la Rúa, et al., (2000), investigated the mtDNA variation in Apis cerana populations, 47 colonies were collected from different locations in Philippines Islands, genetic variation was estimated by restriction analysis of PCR-amplified fragments of the tRNALeu -COII region and they found four different haplotypes (Ce1, Ce2, Ce3, and Ce4) that discriminated among the bee population, from different Islands. Using the restriction enzymes HpaII and AluI, Meixner et al., (2000), had a strong correlation with mtDNA haplotype and morphology among

207 honeybees from Kenya. The different mobility of the restricted fragments in the gel photos compared to the marker is due to the skewed base-pair composition of the mtDNA. (De la Rúa, et al., 1998). Thus Lane M in all the gels is the molecular weight markers (50 bp ladder, Gibco BRL). Using the cluster column and pie with a 3rd visual effects, the percentage distribution of the haplotypes (O1, A1, A4, A2, O1`and Y2) of the studied colonies as a whole were detected as: 41% (11 colonies), 30% (8 colonies), 11% (3 colonies), 7% (2 colonies), 7% (2 colonies) and 4% (one colony) for the following haplotypes O1, A1, A4, A2, O1` and Y2 respectively; as demonstrated in Table (21a) and figures (23a ; 23b). The individual haplotype distribution in the three target geographical zones were as follows: In the semi-desert zone colonies, haplotype O1 is the common haplotype and represent 75% of the analysed colonies; while haplotype Y2 represent only 25% and all the rest haplotypes (A1, A2, A4 and O1`) were nil as shown in Table (21b) and Figure (24a). The most common haplotype in the savannah zone is O1 (54%) followed by A4 (20%), O1` (13%) and A2 (13%) respectively while haplotypes A1 and Y2 represent 0% in this region as shown in Table (21b) and figure (24b). For the forest zone haplotype A1 is the only abundant haplotype in the region, which represent 100% as, shown in Table (21b) and figure (24c) though the rest of the haplotypes (A2, A4, O1, O1` and Y2) were nil. The higher percentage of O1 haplotype in savannah zone(Table 21b and figure 24b ), can be attributed to the fact that most of the analysed colonies were collected from the far eastern part of this zone (the border of Sudan with Ethiopia). Meixner., (2006) *personal communication* , in her study about mtDNA of Ethiopia honey bees (unpublished paper), indicated that, haplotype O is relatively common on the west rim of Ethiopia dome. Rashad and Elsarrag (1978) demonstrated that, the honeybees in some parts of the

208 Sudan were hybridized with foreign honeybee races either through the several importation of particular races to the country or through the migratory swarms of honeybees from adjacent countries. Therefore, it can be concluded that honeybees from the savannah zone are mixture of many different strains and this is strongly confirmed by the morphometric PCA analysis results (figure, 15), which showed the scatter of the individuals from different subspecies particularly between the North-east Africa races (Jemenitica, Scutellata, Monticola and Adansonii). This justification can be attributed to the fact that Scutellata and Monticola were the most abundant races on the border region between Sudan and Ethiopia. The same results were confirmed by the proximities discriminant centroid distances between the groups (Table 20), (figures 19a, 19b, and 20). There were numerous publications, which reported that, A. m. scutellata was highly mobile insect that easily absconds (Hepburn and Radloff 1988; Ruttner 1976 and Ruttner 1988). However, colonies Om-Rawaba, Shawish-Mahadi and Kafindapi (savannah zone) were from the extreme western part of the zone near to the west border of Sudan. Their possession of A4 haplotype is not a surprise, according to Franck et al., (2001) except North-east Africa all the rest of the continental Mellifera races had only haplotype A. Also haplotype O1 was the dominant haplotype in semi-desert zone as shown in Table (21b) and figure (24a), and this might be due to the fact that clusters of semi-desert and savannah were more close to each other as it was confirmed in the discriminant analysis of distances between the group centroids, Table (20) and figure (17). Haplotype Y2 was found in Khartoum colony of semi-desert zone, this phenomenon can be explained by assuming that Khartoum state was invaded with other foreign imported races. Elsarrag (1977), reported that, bees of Khartoum province were hybridised with foreign honeybee races.

209 Haplotype A1 of the forest zone indicated that, the honeybee populations from this zone are homogenous in their mtDNA haplotype (all of the colonies gained A1 haplotype only), even though, the colonies were sampled from different regions of the zone, as shown in Table (21b) and figure (24c). The molecular study results of mitochondrial DNA variability is partially confirmed the previous morphometrical grouping. Undoubtedly the forest zone colonies were confirmed 100% as a completely separate cluster and this is a very strong evidence for the morphometric results of PCA and discriminant analysis. Comparing these results with Mogga (1988), this cluster should be A. m. sudanesis Rashad. In the case of savannah colonies, the presence of different haplotypes is not a big deal, as this phenomenon can be attributed to the heterogeneous mixture of the blood between the previous mentioned subspecies. This was supported by Meixner et al., (2000), who demonstrated that, some honeybee colonies collected in savannah environment of Kenya had both the morphology and the mtDNA haplotype of A. m. monticola or the morphology of one race and the mtDNA of the other. They reported that, there was a colony, which combined the morphology of A. m. monticola with the mtDNA haplotype of A. m. scutellata. Thus all samples of A. m. litorea that they analysed share the mitochondrial haplotype typical of A. m. scutellata. So it is just a matter of heterogeneous mixture of the blood between the different races of the zone. Comparing savannah colonies with the results of Mogga (1988) regarding the classification of the Sudanese honeybees, this cluster should be A. m. bandassi. While the semi-desert zone cluster as demonstrated before corresponds to A. m. yemenitica race. From the present results, such bee migration might have possibly happened across the Sudan Ethiopia border, influencing the bee population in the savannah zone of eastern region. It was also here that the mixed bees

210 of black and yellow individuals, of the same mother were found. Mixed colonies were reported by Bal densperger (1924) to have found black and yellow bees in the same hive in Ethiopia. The land also rises here in altitude from the eastern region of Sudan towards the Ethiopia high lands. Thus at higher altitudes, larger and darker bees were found as reported by Smith (1961). Evidently, the Ethiopian honeybees influence on the bees of eastern region of Sudan was most likely. It could, however be mentioned at this point that, for a complete picture of the Sudanese bee race and their distribution, more work is needed specially with regards to the geographical variation. There is also a need to investigate a possible existence of ecocline on the sloops of Imatong mountains (3187 m) in the extreme southeast and Jebel Marra (3042 m) in the western part of Sudan. Considering the molecular genetic studies, this is the first record on the classification of the Sudanese honeybees according to the mitochondrial DNA variability. Mitochondrial DNA data alone are probably insufficient to infer taxonomic and genetic status of honeybee colonies. Extending micro satellite (nuclear DNA) analysis to honeybee subspecies will be useful in the future for understanding the phylogeography of Apis mellifera and resolving relationships among the African subspecies. Also more detailed analysis of savannah cluster mitochondrial DNA are needed to better resolve the relationship of this wide cluster populations with the neighbouring subspecies. 5- 1- c- Apis florea. Analysis of morphometric characters of the honeybees A. florea from Sudan as in Tables (22- 28), revealed some morphometrical and geographical relationship among the four measured colonis. The results obtained from the principle component analysis (PCA) figure (25), indicated that colonies are not very distinct. The plot shows

211 only two axes loaded with 18.2 and 12.8% of total variance. Factor loadings seem not very clear, but apparently it is mostly sizes as factor 1 and colours as factor 2. In the results obtained in Table (31), the proximity matrix indicates colony discriminant analysis centroids as follows; close relationship between Khartoum and madani colonies and again between Gerry and El- Dender colonies, in comparison to the other distances. Also in figure (26), the discriminant analysis plot demonstrated that most bees were allocated to their correct colonies, except one of Madani group, which was allocated to Khartoum group. From the principal components and factor analysis of 20 characters of 18 samples of A. florea, Ruttner (1988) obtained three morphoclusters (1) South India and Sri Lanka, (2) Thailand and (3) Oman, Pakistan and Iran. More recently Tahmasebi et al. (2002), analyzed the A. florea of Iran and defined two morphoclusters. Combining their data with that of Ruttner (1988) and Mogga and Ruttner (1988) a same groups of countries, they also reported three morphoclusters for A. florea. However, all the individuals of the four colonies were analyzed through the PCA and factor analysis in which, colonies were not very distinct. In the discriminant analysis of the centroid distances between the groups apparently there was a clear relationship between colonies Gerry and El-Dender and Madani and Khartoum colonies respectively. These results may be due to, that all the analyzed colonies were from one geographical zone (semi-desert zone). Radloff and Hepburn (1998, 2000), and Hepburn et al. (2001b) have established empirically that the greater the sampling distances between localities the greater the likelihood that factual morphoclusters will emerge in multivariate analyses. Conversely, where between-group variation is larger than within group variation, biometric subgroups within smaller geographic

212 domains may be swamped and obscured. The average mean measurements of the forewing length, forewing width, hind-leg length, tergites 3 +4 length and cubital index, of the four colonies (El-Dender Gerry, Khartoum, and Madani) were analysed by a cluster column (compares values across categories) for 6 florea colonies from different regions [2 from Sudan “Moggas ones” and 4 from the data bank, Institute fur Bienenkunde-Oberursel-Germany (Mogga 1988)]: Khartoum south, Toti, South India, Oman, South Iran and Pakistan. As shown in Table (32) and figure (27). Samples Madani, Khartoum, Tuti and South Iran were very closed to each other in the average mean of characters: cubital index, length of tergites 3+4, hind-leg length and forewing width; thus colonies Gerry, Khartoum south, El- Dender and Pakistan were closed to each other too, mainly in the average means of the following characters: cubital index, length of tergites 3+4, hind-leg length and forewing width. This indicated that, the colonies were somewhat big in size like those from Pakistan or south Iran. Apis florea ecotypes of Pakistan and Iran, according to Ruttner (1988) were considered as one geographical race; consequently our target four-florea colonies of Sudan should be from the same origin. Mogga (1988), indicated that, the florea bees in Khartoum originated from countries of Western Asia; Pakistan, South Iran and Oman. The biogeography of A. florea as demonstrated by (Hepburn, et al., 2005) is extremely widespread, extending some 7000 km from its eastern-most extreme in Vietnam and Southeastern China, across Mainland Asia along and below the southern flanks of the Himalayas, Westwards to the Plateau of Iran and Southerly into Oman. This constitutes some 70 degrees of longitude (40°–110° East) and nearly 30 degrees of latitude (6°–34° North). Variations in altitude range from sea level to about 2000 m. A. florea has also been introduced in historical times in Saudi Arabia

213 and Sudan, and occurred on Java, Indonesia, since ~50 years ago. Evidently the Florea bees in El-Dender Gerry, and Madani, have been swarmed and migrated from Khartoum. The first discovery of Apis florea in Sudan was in Khartoum at a garden near the international airport by Lord and Nagi (1985). It was believed that the initial colony might have entered the country as part of an air cargo. By January 1987, twenty additional colonies had been found, the most distance being 12 km from the Khartoum airport. This was again a prove that no means can prevent the distribution of honeybee species. Twenty-two years ago is the age of the first Florea colony invaded Sudan so it was expect the distribution of A. florea in Sudan would be extremely widespread. Instead of the above mentioned four studied colonies from the semi-desert zone. There are some evidences [personal communications with some colleges from the field of honey beekeepers] that some colonies of Florea had been found in the far north part of Sudan (desert zone) particularly in Abu-Hamed town on the Nile valley. Also it had been found in the far eastern part of the country at El- Damazin city near the border between Sudan and Ethiopia (Savannah zone). Also in South Sudan at Malakal city (rich savannah zone, near to the forest zone), few months ago it was found in “Kosti“ near the author’s house. It was quite significant also that Apis florea accepted any site for nesting. Colonies of Florea had been found nesting in different areas which the indigenous Apis mellifera L. that seeks well protected sites away from any human activities. Florea therefore competes well for nesting sites and most probably for the forage in urban areas even outside its native habitat. This was confirmed with the rapid establishment of the Florea populations in Khartoum city and the rest of the previous mentioned towns in the Sudan. In 1984, Whitecombe stated that, Apis

214 florea were well adapted to a higher range of ambient temperature than Apis mellifera yemenitica L. The rapid establishment of Apis florea was also confirmed by Ruttner et al., (1985) who studied nesting sites for Apis florea in Iran. It could, however be mentioned that, within the few coming years Florea colonies are going to invade some African countries from Sudan, specially Ethiopia at the eastern part of Sudan, Kenya and Uganda at the southern part of the country even Egypt at the northern part border.

215 5- 2- SUMMARY AND CONCLUSION 4-2- a- Apis mellifera L. The longest North- South latitude 3°N to 22°N and the broadest East- West longitude 23°E to 30°E, gives the Sudan a contrast in climate and consequent variety of seasons and natural vegetation. This resulted in five distinct climatically zones: desert, semi-desert, poor savannah, rich savannah and forest zones. With such degree of variation, the Sudanese honeybees found wherever climatic conditions permit, differed accordingly. This study focused on the taxonomic status and geographical variations of the Sudanese honeybees using the biometrical and mitochondrial DNA variation analysis, nineteen samples were collected from four geographical zones of the country via: semi-desert, poor savannah, rich savannah and forest zones. Further 8 other samples were added in the molecular genetic (mtDNA variability) part of the study. The 19 bee samples were collected from resting swarms, feral established colonies and one hived colony. From these samples 39 morphometrical characteristics were measured. The Sudanese honeybees were classified and named as Apis mellifera nubica, by Ruttner (1975). He used samples originating from the semi-desert zone. In 1976, he added that, besides the Sudanese’s bee’s total small size, their short appendices were remarkable. Therefore, clearly belong to another group, correctly to be considered as an independent race. Rashad et al., (1984) suggested that, there were two independent honeybee subspecies in the Sudan. Firstly Apis mellifera sudanesis, distributed all over Sudan between latitudes 3° N and 16° 20` N. The second was Apis mellifera nubica Ruttner, described as mixed bees, distributed along the International boundaries of Sudan, Ethiopia and Uganda. However, in 1986, Ruttner withdrew the name “Nubica” in

216 favor of Yemenitica, a name that included bees from the Saudia Arabia, Yemen, Oman, Somalia and Chad. Mogga (1988) by using the PCA analysis method suggested that, three ecotypes of honeybees exist in the Sudan; namely: A. m. yemenitica L. (the small bee of semi-desert zone), A. m. sudanesis (the medium bee of forest and rich savannah zone) and A. m. bandasii (the large bee along the border of Sudan and Ethiopia). Importation of European honeybees to Khartoum state started since colonial period; the earliest being in 1928 by King. It went on in 1987. Owing to this, Elsarrag (1977), considered the Khartoum state bees as being hybridized. Also Rashad and Elsarrag (1978) demonstrated that, the honeybees in some parts of the Sudan were hybridized with foreign honeybee races either through the several importations of particular races to the country or through the migratory swarms of honeybees from adjacent countries. This argument may apply to sample of Khartoum from the semi-desert zone of the present study, particularly the mitochondrial DNA variation part of the study. In general this sample gained haplotype Y2, which was completely different from the haplotypes of the other studied colonies of the semi-desert zone and even, from the savannah and forest zone colonies too. Meanwhile, other samples, El-Hissahisa, Madani, and Arkawit were all gained haplotype O1. Fletcher (1978) reported that, compared to the quantity of genetic material introduced from African into Brazil where, European bees were already established, the scale of honeybee importations into Africa from the United States and European may be regarded as massive. This was particularly more so, when the Production of Italian drones by 30 or more colonies were encouraged in Pretoria over several decades. Nevertheless, the Adansonii population of Pretoria area had remained un affected by its prolonged exposure to imported genetic materials, Fletcher concluded.

217 Buco et al., (1987) also found that, the Africanized bees in South America were distinctly different and smaller than European bees. Thus, the feral bees of South America clearly show the influence of both their European and African parentage, although they were more similar to their African parents. The up normal Y2 haplotype of the Khartoum colony, despite its origin from semi-desert zone might most probably be attributed to hybridization with foreign bees through the migratory swarms of honeybees from adjacent countries. Morse et al., (1973) had also stated that, the success of imported bees with an established success in honey production and acceptable management behavior in the tropics, with its well-adapted native bee was far from reach. In the principal component analysis together with the modern discriminant analysis of our target 19-honeybee colonies (Apis mellifera L.), the results revealed the present of three clusters of the Sudanese bees and they were geographically identifiable. While with the Dra 1 restriction enzyme test, there were 6 different haplotypes had been differentiated among the 27 studied colonies? The dark green colour samples distributed to the left side of the PCA graph consisted of the smallest bees. Their average value measurements were: forewing length 8.23 mm., and width 2.82 mm.; proboscis length

5.55 mm.; hind-leg length 6.83 mm.; T 3+4 length (body size) 3.88 mm., wing venation angles G18, 99.2 and O26, 41.49 ;hind-leg metatarsus length 1.80 mm.; hind-leg metatarsus width 1.04 mm.; Sternlte 3 length 2.39 mm.; sternite 6 length 2.25 mm.; sternite 6 width 2.64 mm.; wax mirror length 1.06 mm.; wax mirror width 1.92 mm.; cubital vein a length 4.26 mm.; labrum 1 colour 3.05; and cubital index 1.85 mm. The mitochondrial DNA variation analysis of this cluster resulted the appearance of haplotype A1 in all the studied 8 colonies of the zone. This result confirmed the PCA and discriminant analysis of the biometrics,

218 that the forest zone cluster was completely seperated from both the savannah and semi-desert zone clusters. The second cluster (dark orange colour) samples distributed around the middle of the PCA graph had medium measurement values. Their mean average measurements were: forewing length 8.27 mm., and width

2.88 mm.; proboscis length 5.63 mm.; hind-leg length 7.00 mm.; T 3+4 length (body size) 3.88 mm.; wing venation angles G18, 96.33 and O26 36.21; hind-leg metatarsus length 1.91 mm.; hind-leg metatarsus width 1.09 mm.; Sternlte 3 length 2.46 mm.; sternite 6 length 2.26 mm.; sternite 6 width 2.68 mm.; wax mirror length 1.12 mm.; wax mirror width 1.99 mm.; cubital vein a length 4.68 mm.; labrum 1 colour 6.39; and cubital index 2.04 mm. Mitochondrial DNA variation analysis of this cluster colonies (4 colonies) resulted the appearance of two different haplotypes; O1 and Y2 with representing percentages 75 and 25% respectively of all the analysed colonies. The third cluster of samples (light blue colour samples) distributed to the right side of the PCA graph possessed most of the largest measurements. Their average mean measurements were: forewing length

8.45 mm., width 2.95 mm.; hind-leg 7.05 mm.; T 3+4 length (body size) 4.00 mm., wing venation angles G18 99.20 and O26 36.99; hind-leg metatarsus length 1.90 mm.; hind-leg metatarsus width 1.09 mm.; Sternlte 3 length 2.52 mm.; sternite 6 length 2.36 mm.; sternite 6 width 2.76 mm.; wax mirror length 1.13mm.; wax mirror width 2.01 mm.; cubital vein a length 4.83 mm.; labrum 1 colour 3.65 mm.; and comparatively largest cubital index 2.24 mm. While the mitochondrial DNA variation analysis of this cluster revealed the presence of four different haplotypes: O1, A4, O1` and A2. The most common haplotype in the savannah zone is O1 (54%) followed by A4 (20%), O1` (13%) and A2 (13%) respectively.

219 Franck, et al., 2001, demonstrated that, honeybees from North-eastern Africa contain three highly divergent mitochondrial lineages A, O, and Y. While in the other parts of Africa, honeybees carry only mitotypes of lineage A. Also the gene flow within African honeybees is very high due to high swarming and migratory behaviour (Hepburn and Radloff, 1988). Considering the mitochondrial DNA variation analysis of this study, It seems that, Sudanese honeybees as a whole harbours a mixture of A, O and Y haplotypes which is really not surprised from what is known about the North-east African bees. Haplotype O (O1 and O1`) represent 47%; thus, haplotype A (A1, A2, and A4) represent 47% while haplotype Y (Y2) represent only 6% of the total haplotypes analysed in this study. Which indicated that the abundant honeybee haplotypes of Sudan were haplotypes O and A. These three clusters of the Sudanese honeybees were geographically correlated. The smallest bees originated from the forest zone. The medium bees originated from the semi-desert zone. While the largest Sudanese bees originated from the savannah zone, mainly along the Sudan Ethiopia border. Where these bee samples originated in the three zones, no physical isolation barriers exist. This conformed to Ruttner and Kauhausen (1985) finding: existence of geographical variability of honeybee in spite of the absence of physical isolating barriers in tropical Africa. Thus, the present of more than one haplotype in one cluster as in the savannah cluster is not surprised and this conformed to Meixner et al., (200), they demonstrated that, some honeybee colonies that were collected in savannah zone environment of Kenya had both the morphology and the mtDNA haplotype of A. m. monticola or the morphology of one race and the mtDNA of other, thus they reported that, there was a colony in their study combined the morphology of A. m. monticola with the mtDNA haplotype of

220 A. m. scutellata. Thus all samples of A. m. litorea that they analysed shared the mitochondrial haplotype typical of A. m. scutellata. More than this Franck, et al., (2001) documented that; honeybees from North-eastern Africa contain three highly divergent mitochondrial lineages A, O, and Y. While in the other parts of Africa, honeybees carry only one mitotypes of lineage A. thus, Meixner. (2006) *Personal communication, in her study mtDNA variation of Ethiopian honey bees (unpublished paper) indicated that, haplotype O is relatively common on the west rim of Ethiopia dome, and this zone is the eastern border of the Sudan and Ethiopia, in the same time most of the savannah zone colonies were collected from this area. So it is just a matter of heterogeneous mixture of the blood between the different races of the zone. Supporting this most of this zone colonies were coloured (black and yellow). Rainey (1963) proved that insects fly in the same direction of the wind. Moreover, Papdopoulo (1975) stated that, the African honeybees frequently migrate and they can migrate up to 28- 30 km. Again, Brown et al., (1969) were able to demonstrate the influence of the Inter Tropical Continental Zone (ITCZ, moist front) on the movement of some insects. Elsarrag (1977) suggested that, ITCZ to some extent influences the movements of the migratory honeybee swarms from the south to the north with the moist south-westerly winds within the country (Sudan) as well as across the south-eastern boarders. El Sarrag (1977) suggested the possibility of migration of the African honeybee races, from the south with the moist southwesterly wind. Observation around Kosti however, indicated migration between March to May towards the Nile, when the green Vegetation and water away from the Nile dried out with the advent of the dry season. Along the Nile, the bees were able to find water and flowering plants, fruits and some vegetables. During the rainy season, reproductive swarms migrate back, away from the Nile. The above mentioned observation on the

221 bees migration around Kosti was conformed with the PCA result of Kosti colony as it appeared to be more closed to semi-desert zone than its origin savannah zone. Also Wille, H. (1979) reported that, bees swarms at Abu Naama and Damazin areas, which had spend the dry season in the banks of the Blue Nile, migrated during or shortly after the rainy season hundreds of kilometres east and west wards. As drought became stronger, they went again back to the Blue Nile in several steps. These migrations would be admitted to be a response to lack or presence of water, nectar or honeydew and pollen. Considering the mitochondrial DNA results of the study plus the presence of colored bees (black and yellow bees in one hive) of some studied colonies, it could be assumed that there was a gene flow among the honeybees of Sudan in the area between the latitudes 9º N and 15º N (the southern part of semi-desert zone and almost all the savannah zone of Sudan). The direction of the gene flow was from the low land of savannah zone of Ethiopia towards the western part of the Sudan. Also as it was confirmed that, the forest zone cluster was completely separated from the other two clusters (semi-desert and savannah zone clusters), the present of haplotype A1 in the forest zone colonies only, plus no evidences that, south Sudan was invaded by a foreign honey bee races before, it could be assume that the original Sudanese honey bees mitochondrial DNA haplotype may be is haplotype A1, thus the pure Sudanese honey bee race may be Apis mellifera sudanesis (South Sudan bees). Based on both biometric and mitochondrial DNA results of the present study which were some what conformed each other, the Sudanese honeybees, through selective adaptation to certain biotopes consisted of three ecotypes or races, showed some clear geographical distribution pattern for the character measured, considering the biometrical analysis it is typical of the tropical African bees (Ruttner and Kauhausen, 1985). While

222 regarding the mitochondrial DNA variation also it is typical of African bees haplotypes distribution Franck, et al., (2001). The medium size bees (O1 and Y2 haplotypes) distributed in the semi- desert zone were inseparable from the Yemenitica race. Thus the bees from this zone of Sudan maintained the name Apis mellifera yemenitica, Ruttner. The small size bees (A1 haplotype) of Sudan were distributed in the forest zone. These bees retained the name Apis mellifera sudanesis, Rashad. The largest Sudanese honeybees (O1, A4, O1` and A2 haplotypes) were distributed in the savannah zone. The distribution area covered most of the country’s potential region for beekeeping, between 5° N to 16° N latitude. Some bee colonies of this zone composed of black and yellow bees. These bees retained the name Apis mellifera bandasii, Mogga. Honeybee samples along Sudan Uganda and Kenya borders plus samples of mountainous areas such as Imatong 3187 m. and Jebel Marra 3042 m. could not be included in this study. Inclusion of such samples in future biometric and molecular genetics studies may clarify the honeybee races and their geographical distribution in the Sudan. 5- 2- b- Apis florea: This bee species, a native of Asian subcontinent has got established in different towns or geographical zones of the Sudan, in fairly short period. With this rapid established, it could be concluded that, Florea competed well for nesting sites and forage with the indigenous Apis mellifera yemenitica. Thus it can be concluded that, within the few coming years the Florea bees will distribute from Sudan to the neighbouring countries particularly Ethiopia, Kenya and Egypt. Concluding from the present results, the four different Florea bee colonies of Gerry, Khartoum, Madani and El-Dender originally were from the first discovered Florea colony in Sudan (Khartoum) by Lord and Nagi (1985). Thus their mother colony entered Sudan from Pakistan or South Iran.

223 CHAPTER SEVEN: LIST OF REFERENCES Adam, Brother. (1951). In search of the best strains of bees. Bee World 32: 49-52. Adam, Brother. (1954). In search of the best strains of bee: second journey. Bee World 35: 193-203, 223-244. Adam, Brother. (1961). The honeybees of the Iberian Peninsula. Bee world 42: 252-255. Adam, Brother. (1964). In search of the best strains of bee: concluding journeys. Bee World 45: 70-83, 104-118. Adam, O. B. E. Br, (1966). In Search of the Best Strains of Bees, Northern Bee Books. Walman Verlag, Zell-Weirbach. Adam, Brother. (1977). In search of the best strains of bees: supplementary journey to Asia Minor, (1973). Bee World 58: 57-66. Adjaloo, M. K., yeboah- Gyan (1991). The foraging strategies of the African honeybee Apis mellifera adansonii in the humid forest. In: West African Bee Research Seminar, Bakau, The Gambia (Ed N Bradbear), 28- 30, Bees for Development, Monmouth, UK. Adjare, S. O. (1990). Beekeeping in Africa. FAO, Rome, Italy. Alber M. A. (1952). Imkerbummel in Algerien. Deutsche Bienenewritschaft 3:204- 208. Alexander, B. (1990). A phylogenetic analysis of honeybees (Hymenoptera: Apidae: Apis). In “11th International Congress. IUSSI, India,” pp. 120-121.

Anderson, R. H., Busy B., Johannsmeire, M. F. (1983). Beekeeping in South Africa. Bulletin No. 394, Department of Agriculture, Pretoria South Africa.

Alpatov, W. W. (1925). Über die Verkleinerung der Rüssellänge der Honigbiene vom Süden nach Norden hin. Zoologische Anzeiger

224 65: 103- 111. Alpatov, W. W. (1929). Biometrical studies on variation and races of honeybee Apis mellifera L. Quar. Rev. Biol. 4 (1): 1-58. Alpatov, W. W. (1932). Some data on comparative biology of different bee races. Bee World VIII (12): 138-139. Alpatov, W. W. (1933). South Africa bees biometrically investigated. Bee World XIV (6): 62-64. Alpatov, W. W. (1933). South Africa bees biometrically investigated. South African Bee Journal 15: 10- 12. Arias M.C., Sheppard W.S. (1996) Molecular phylogenetics of honeybee subspecies (Apis mellifera L.) inferred from mitochondrial DNA sequence, Mol. Phylogenet. Evol. 5, 557–566. Arias M.C., Tingek S., Kelitu A., Sheppard WS. Koeniger and Koeniger, (1996) Apis nuluensis Tingek, and its genetic relationship with sympatric species inferred from DNA sequences, Apidologie 27. 415–422. Ashmead. W. R. (1904). Remarks on honeybees. Proc. Entomol. Soc., Wash. 6: 120- 123. Attardi, G., (1985). Animal mitochondrial DNA: an extreme example of genetic economy. Int. Rev. Cytol. 93: 93-145. Aurelien, R. F. (adapte Collart, E) (1950). Manuel d‘Apiculture. Service de I‘Agriculture, Leopoldville (Kinshasha), Zaire. Avis, J. C. (1986). Mitochondrial DNA and the evolutionary genetics of higher animals. Phil. Trans. R. Soc. London B 312: 325- 342. Avise, J C., Arnold, J. Ball, RM (1987). Intraspecific phylogeography: the mitochondrial DNA bridge between population genetics and systematic. Annul. Rev. Ecol. Syst. 1: 489-522. Avise, J. C. and R. A. Lansman. (1983). Polymorphisms of mitochondrial DNA in populations of animals, pp. 147-164. In M. Nei and R. K.

225 Koehn (eds.), Evolution of genes and proteins. Sinauer, Sunderland, Mass. Ayala, F. J. & J. R. Powell. (1972). Allozyme are diagnostic characters of sibling species of Drosophila. Proc. Natl. Acad. Sci. U. S. A. 69: 1094-1096. Badino, G., Celebrano, G., Manino, A. (1983a). Population structure and Mdh-1 locus variation in Apis mellifera ligustica. J. Hered. 74: 443-446. Badino, G., Celebrano, G., Manino, A. (1983b). Identificazione de Apis mellifera ligustica Spinola sulla base di sistemi geneenzima. Boll. Mus. Reg. Sci. Nat. Torino 1: 451-460. Badino, G., Celebrano, G., Manino, A. (1984). Population genetics of the Italian honeybee (Apis mellifera ligustica Spin.) and its relationships with neighboring subspecies. Boll. Mus. Reg. Sci. Nat. Torino 2: 571-584. Bal densperger, J. (1922). Sur l‘apiculture en orient. Proceedings of the 6th International Congress of Apiculture, Marseille, France, pp. 59- 64. Bal densperger, J. (1924). North Africa bees II. Bee World v (12): 189-190. Bal densperger, J. (1926). The various races found in Europe and outstanding characteristics of each. Bee World VIII (1): 4-5. Bal densperger , P. J. (1932). Varietes d’abeilles en Afrique du Nord. Proceedings of the 5th International Congress of Entomology, Paris France, pp. 829 – 839. Blanchetot, A. (1991). Genetic relatedness in honeybees as established by DNA fingerprinting. Journal of Heredity 82: 391- 396. Britten, R. J. (1986). Rates of DNA sequence evolution differ between taxonomic groups. Science 231: 1393-l 398. Boyce, T. M., M. E. Zwick and C. Aquadro, (1989). Mitochondrial DNA

226 in the bark weevils: site structure and heteroplasmy. Genetics 123: 825-836. Brown, W. M., M. George, and A. C. Wilson. (1979). Rapid evolution of animal mitochondrial DNA. Proc. Natl. Acad. Sci. USA 76: 1967- 197 1. Brown, W. M., M. (1985). The mitochondrial genome of animals, pp. 95- 130 in Molecular Evolutionary Genetics. Edited by R. J. Macintyre. Plenum press, New York. Brueckner, D. (1987). The influence of genetic variability on wing symmetry in honeybees (Apis mellifera). Evolution 30: 100-108. Buco, S. M., Rinderer, T. E., Sylvester, H. A., Collins, A. M., Lancaster, V. A. and Crewa, R. M. (1987). Morphometric differences between South America Africanized and South African (Apis mellifera scutellata) honeybees. Apidologie 18 (3): 217-222. Buttel- Reepen, H. (1906). Apistica. Beitrage zur Systematik, Biologie, sowie zur geschichtlichen undgeographischen Verbreitung der Honingbiene (Apis mellifera L.), ihere Varietaten und der ubrigen Apis-arten. Veroeff. Zool. Mus. Berlin, 118-120. Buroker, N. E., J. R. Brown, T. A. Gilbert, P. J. O‘hara, A. T. Beckenbarch, W. K. Thomas and M. J. Smith, (1990). Length heteroplasmy of sturgeon mitochondrial DNA: an ellegitimate elongation model. Genetics 124: 157-163. Cánovas, F., P. De la Rúْa, J. Serrano y J. Galián. 2002. Variabilidad del DNA mitocondrial en poblaciones de Apis mellifera iberica de Galicia (NW Espana). Arch. Zootec., 51: 441- 448. Carlson, C. W., and R. W. Brosemer (1971). Comparative structural properities of insect triose phosphate dehydrogenases. Biochem. 10: 2113- 2119. Carlson, C. W., and R. W. Brosemer (1973). Amino acid compositions of

227 cytochrome C from four hymenopteran species: evolutionary significance. Syst. Zool. 22:77- 83. Carlson, D. A. and A. B. Bolten. (1984). Identification of Africanized and European honeybees, using extracted hydrocarbons. Soc. Am. 30 (2): 32-35. Casteel, D. B., Phillips, E. F. (1903). Comparative variability of drones and workers of the honeybee. Biol. Bull. 6: 18-37. Castagné J. B. (1993). L‘apiculture au Congo- Brazzaville. Bulletin Technique Appicole 10 (4): 197- 208. Clary, D. O., and D. R. Wolstenholme. (1983). Nucleotide sequence of a segment of Drosophila mitochondrial DNA that contains the genes for cytochrome C oxidase subunits II and III and ATPase subunit 6. Nucleic Acids Res. 11: 42 1 l-4227. Clary, D. O., and D. R. Wolstenholme. (1985). The mitochondrial DNA molecule of Drosophila yakuba: nucleotide sequence, gene organization, and genetic code. J. Mol. Evol. 22: 252-27 1. Clary, D. O., and D. R. Wolstenholme. (1987). Drosophila mitochondrial DNA: conserved sequence in the A+T-rich region. and supporting evidence for a secondary structure model of the small ribosomal RNA. J. Mol. Evol. 25:116-125. Clauss, B. (1983). Bees and Beekeeping in Botswana. Ministry of Agriculture, Botswana. Chandler, M. T. (1976). The African honeybee- Apis mellifera adansonii: the biological basis of its management. Proceedings of the First International Congress of Apiculture in Tropical Climates, London, UK, pp. 61- 68. Chochlov, B. A. (1916). Investigations of the length of the bee tongue. Ministry of Agriculture, Petrograd, Russia, 17- 41. (cited from Alpatov 1929, in Russian).

228 Cockerell , T. D. A. (1906). New rocky mountain bees, and other notes. Can. Entomol., 38, 160-166. Cockerell, T. D. A. (1908). Descriptions and records of bees. Annu. Mag. Nat. Hist. Lond. 2: 323-334. Collins, A. M., Blum M. S. (1982). Bioassay of compounds derived from the honeybee sting. Journal of Chemical Ecology 8: 463- 470. Comer, J. (1985). Apiculture and bee management problems in African countries. Proceedings of the 3rd International Conference on Apiculture in Tropical Climates, Nairobi, Kenya. PP. 41- 44. Contel, E. P. B., Mestriner M. A,, Martins, E (1977). Genetic control and development expression of malate dehydrogenase in Apis mellifera. Biochemical Genetics 15: 859- 876. Cornuet, J. M. and Fresnaye, J. (1975). Biometry and classification of Apis mellifera. XXVth . Int. Apic. Cong: 351-355. Cornuet, J. M. and Fresnaye, J. Tassencourt L (1975). Discrimination et classification de populations d‘abeilles á partir de caractéres biometriques. Apidologie, 6, 145-187. Cornuet, J. M. (1979). The MDH system in honeybees, of Guadaloupe. J. Hered. 70: 223-229. Cornuet, J. M., Fresnaye J., (1979). Production de miel chez les hybrides interaciaux d‚abeilles (Apis mellifera L.) lors de générations successives de rétrocroiseement sur la race locale, Apidologie, 10: 3-15. Cornuet, J. M., Albisetti, J., Mallet, N., Fresnaye, J. (1982). Representation graphique de populations multinormals par des ellipses de confiance. Apidologie 13: 15- 20. Cornuet, J. M., (1983). Reproduction. Génétique et sélection de l‘abeille. Bulletin Technique Apicole 10 (1): 13- 36. Cornuet J.M., Daoudi A., Chevalet C. (1986) Genetic population and

229 number of matings in a black honey bee (Apis mellifera mellifera) population, Theor. Appl. Genet. 73, 223–227. Cornuet, J-M. et J. Fresnaye. (1989). Etude biométrique de colonies d’abeilles d’Espagne et du Portugal. Apidologie, 20: 93-101. Cornuet , J. M. , Garnery , L. and Solgnac , M. (1991). Putative origin and function of the intergenic region between COI and COII of Apis mellifera L. mitochondrial DNA. Genetics, 128, 393-403. Cornuet J.M., Garnery L. (1991a). Genetic dversity in Apis mellifera. In: Diversity in the Genus Apis (Ed DR Smith), 103- 115, Westview, Boulder, Colorado, USA. Cornuet J.M., Garnery L. (1991b). Mitochondrial DNA variability in honeybees and its phylogeographic implications. Apidologie 22, 627–642. Cornuet J.M., Luikat G., (1997). Description and power analysis of two tests for detecting recent population bottlencks from allele frequency data, Genetics 144: 2001-2014. Crewe, R. M., Hepburn, H. R., Moritz, R. F. A., (1994). Morphometric analysis of 2 southern Africa races of hoeybee. Apidologie 25: 61- 70. Crozier , R. H., Y. C. Crozier and A. G. Mackinlay, (1989). The CO-I and CO-II region of honeybee mitochondrial DNA: evidence for variation in insect mitochondrial evolutionary rates. Mol. Biol. Evol. 6:399-411. Crozier , R. H. and Crozier , C. (1993). The mitochondrial genome of the honeybee Apis mellifera: complete sequence and genome organisation. Genetics, 133, 97-117. Culliney, T. W. (1983). Origin and evolutionary history of the honeybees Apis. Bee World 64: 29-37. Daly, H. V. (1975). Identification of Africanized bees by Multivariate

230 morphometrics. XXVth. Int. Apic. Cong: 356-35. Daly, H.V. and Balling, S.S. (1978). Identification of Africanized honey bees in the Western Hemisphere by discriminant analysis. J. Kans. Entomol. Soc. 51: 857-869. Daly, H.V. (1978). Discriminant analysis of Africanized and European honeybees. In Simp. Int. Apimondia sobre apiculture en clime quente, Florianopolis. Brazil. Pp. 93-95. Bucharest: Apimondia. Daly, H.V. (1985). Insect morphometrics. Ann. Rev. Entomol. 30: 415-438. Daly, H.V. (1988). Overview of the identification of Africanized honeybees. In Africanized honeybees and bee mites, ed. G. R. Needham, R. E. Page, Jr., M. Delfiinado-Baker, C. E. Bowman, pp. 245-249. Chichester, England: Ellis Horwood Limit. Daly, H.V. (1991). Systematics and identification of Africanized honeybees. In the “African“ Honey Bee (Eds M Spivak, DJC Fletcher, MD Breed). 13- 44, Westview Press, Boulder, Colorado, USA. Daly, H.V. Morse, R. A. (1991). Abnormal sizes of worker honeybees (Apis mellifera L.) reared from drone comb (Hymenoptera, Apoidae). Journal of the Kansas Entomological society 64: 193- 196. Daly, H.V. Hoelmer K, Gambino, P. (1991). Clinical geographic variation in feral honeybees in California, USA. Apidologie 22: 591- 609. Daly, H.V. (1992). A statistical and empirical evolution of some morphometric variables of honeybee classification. In: Ordination in the Study of Morphology, Evolution and Systematics: Applications and Quantative Genetic Rationals (Eds JT Sorensen, RJ Footit), 127- 155, Elsevier, Amsterdam, The Netherlands. Davies, H. R. J. (Ed.) (1985). Natural Resources and Rural Development in

231 Arid Lands: Case Studies From Sudan. Tokyo: United Nations University Press. 94 pp. Deborah, R. Smith and Robert, H. Hagen (1996). The Biography of Apis cerana as Revealed by Mitochondrial DNA Sequence Data. Journal of the Kansas Entomological Society 69 (4) suppl., 1996, pp. 294- 310. Debruijn, M. H. L. (1983). Drosophila rndanogaster mitochondrial DNA: a novel organization and genetic code. Nature 304:234-24 1. Del Lama, M. A., Figueiredo, R. A., Sores, A. E. E., Del Lama, S. N. (1988). Hexokinase polymorphisms in Apis mellifera and its use for Africanized honeybee identification. Rev. Bras. Genet. 11:287-297. Del Lama, M.A., Lobo, J.A., Soares, A. E. E. and Del Lama, S.N. (1990). Genetic differentiation estimated by isozymic analysis of Africanized honey bee populations from Brazil and from Central America. Apidologie 21: 271-280. Del Lama, M. A., M. A. Mestriner & J. C. A. Paiva. (1985). Est-5 and PGMI, new polymorphisms in Apis mellifera, Rev. Bras. Genet. 8: 17-27. Del Lama, M. A., R. A. Figueiredo, A. E. E. Soares and S. N. Del Lama. (1988). Hexokinase polymorphism in Apis mellifera and its used for Africanized honeybee identification. Rev. Bras. Genet. II: 287-297. De la Rúa, P., J. Serrano and J. Galián. (1998) Mitochondrial DNA variability in the Canary Islands honeybees (Apis mellifera L.). Mol. Ecol., 7: 1543-1547. De La Rua, P., Simon, U. E., Tilde, A. C., Moritz, R. F. A. and Fuchs, S. (2000) MtDNA variation in Apis cerana populations from the Philippines. Heridity 84 (2000): 124- 130.

232 De la Rúْa, P., J. Galián, J. Serrano and R. F.A. Moritz. (2001a). Molecular characterization and population structure of the honeybees from the Balearic Islands .Apidologie, 32: 417- 427. De la Rúْa, P., J. Galián, J. Serrano and R. F.A. Moritz. (2001b). Genetic structure and distinctness of Apis mellifera L. populations from the Canary Islands. Mol. Ecol., 10: 1733- 1742. .(De la Rúْa, P., J. Galián, J. Serrano and R. F.A. Moritz. (2002 Microsatellite analysis of nonmigratory colonies of Apis mellifera iberica from south-eastern Spain. J. Zool. Syst. Evol. Res., 40: 1- 5. De la Rúْa, P., J. Galián, J. Serrano and R.F.A. Moritz. 2003. Genetic structure of Balearic honeybee populations based microsatellite ariation. Gen. Sel. Evol., 35: 339-350. Densmore, L. D., J. W. Wright and W. M. Brown, (1985). Length variation and heteroplasmy are frequent in mitochondrial DNA form parthenogenetic and bisexual lizards (genus Cnemidophorus). Genetics 110: 689-707. Deowanish, S., Nakamura, J., Matsuka, M. and Kimura, K. (1996). MtDNA variation among subspecies of Apis cerana using restriction fragment length polymorphisim. Apidologie, 27: 407- 413. De Roeck, H. (1950). L‘apiculture au Congo Belge du point du vue social. Institut Universitaire des Territoires d‘Outremer , Antwerp, Belgium. Doedikar, G. B. C. V. Thakar, and P. N. Shaw. (1959). Cytogenetic studies in Indian honeybees, 1. Somatic chromosome complement in Apis indica and its bearing on evolution and phylogeny. Proc. Ind. Sci. 49: 194- 206. Doedikar, G. B. (1978). Possibilities of origin and diversification of

233 angiosperms prior to continental drift. In Recent researches in geology, vol. 4, ed. K. B. Power, pp. 474- 481. Delhi, Hindustan. Drory E. (1888) From my notebook: notes on bees from a trip around the world, Bienenvater 20, 20– 30 (in German). Dubois, L. Collart, E. (1950). L‘Apiculture au Congo Belge et au Ruanda- Urundi. Ministère des Colonies. Direction de L‘Agriculture, de l‘Ēlevage et de la Colonisation, Bruxelles, Belgium. Dung P.X., Thai P.H., Trung L.Q. (1996) Some comparative characteristics between Apis florea and Apis andreniformis in Minh Hai Province, Proc. 3rd Asian Apic. Assoc. Hanoi, p. 62. Dupraw, E. J. (1964). Non- Linnean taxonomy. Nature 202: 849- 852. Dupraw, E. J. (1965a). Non- Linnean taxonomy and the systematics of honeybees. System Zool. 14 (1): 1-24. Dupraw, E. J. (1965b). The recognition and handling of honeybees specimens in Non-Linnean taxonomy. J. Apic. Res. 4 (2): 71-84. Dutton, R. W., Free, J. B. (1979). The present status of beekeeping in Oman. Bee World 60: 176- 185. Dutton, R. W., Ruttner, F.; Berkely, A. and Manley, J. D. (1981). Observation on the morphology, relationships and ecology of Apis mellifera L. of Oman. J. Apic. Res. 20 (4) : 201-214. El- Sarrag, M. S. A. (1977). Morphometrical and biological studies on Sudanese honeybees Apis mellifera L. (Hymenoptera; Apidae). Ph.D. Thesis. Cairo University. El- Sarrag, M. S. A., Saeed, A. A., Hussein, M. A. (1992). Morphometrical study on the Sudanese honeybees. Journal of King Saud University, Agricultural Sciences 4: 99- 108. Enderlein, G. (1906). Neue honigbienen und beitrage zur kenntnis der verbreitung der gattung Apis . Stett. Entomol. Zeit. 67: 331-344. Engel , M. S. (1999). The taxonomy of recent and fossil honey bees

234 (Hymenoptera: Apidae; Apis). J. Hym. Res., 8, 165-196. Eischen, F. A., Rothenbuhler, W. C., Kulincevic, J. M. (1982). Length of life and dry wieght of worker honeybees reared in colonies with different worker-larva ratios. J. Apic. Res. 21: 19-25. Eischen, F. A., Rothenbuhler, W. C., Kulincevic, J. M. (1983). Brood rearing associated with a range of worker-larva ratios in the honeybee. J. Apic. Res. 22: 163-168. Eschscholtz , J. F. (1821). Entomographien. Reimer, Berlin. Estoup, A.; Solignac, M.; Harry M., Cornuet J. M., (1993). Characterization of (GT) n and (CT) n microsatellites in two insect species. Apis mellifera and Bombus terrestris. Nucleic Acids Res. 21: 1427-1431. Estoup, A.; Solignac, M.; Cornuet, J.-M., (1994): Precise assessment of the number of patrilines and of genetic relatedness in honeybee colonies. Proc. R. Soc. London [Biol.] 258, 1–7. Estoup , A. , Grnery , L. , Solignac , M. and Cornuet , J.-M. (1995a). Microsatellite variation in honey bee (Apis mellifera L.) populations: hierarchical genetic structure and test of the infinite allele and stepwise mutation models. Genetics, 140, 679-695. Estoup , A. ,Taillez C., Cornuet J.-M., Solignac , M. (1995b). Size homoplasy and mutational processes of interrupted microsatellites in two bee species Apis mellifera and Bombus terrestris (Apidae), Mol. Biol. Evol. 12: 1074-1084. Fabricius J.C. (1787). Mantissa Insectorum, Hafniae, Proft. Vol. 1, p. 348 (cited from Maa, 1953). Falconer, D. S. (1989). Introduction to Quantative Genetics. 3rd Edition, Longman, Harlow, UK. Field, A. N. (1933). Notes from all quarters, Worcester, Cape Province. South African Bee Journal 8 (3): 5- 6.

235 Fletcher, D. J. C. (1978a). The influence of vibratory dances by worker honeybees on the activity of virgin queens. Journal of Apicultural Research 17: 3- 13. Fletcher, D. J. C. (1978b). Vibration of queen cells by worker honeybees and its relation to the issue of swarms with virgin queens. Journal of Apicultural Research 17: 14- 26. Fletcher, D. J. C. (1978c). The African bee Apis mellifera adansonii in Africa. Ann. Rev. Ent. 23: 151-171. Felsenstein , J. (1993). Phylips (Phylogeny inference package), version 3.5c. Distributed by the author. Department of Genetics, University of Washington, Seattle. Fondark, M. K., Page, R. E., Hunt, G. J. (1993). Paternity analysis of worker honeybees using random amplified polymorphic DNA. Naturwissenschaften 80: 226- 231. Franck, P., L. Garnery, M. Solignac and J.M. Cornuet. (1998). The origin of West European subspecies of honeybees (Apis mellifera): new insights from microsatellite and mitochondrial data. Evolution, 52: 1119-1134. Franck, P., Garnery, L., Solignac, M., and Cornuet, J.M., (1998). The origin of West European subspecies of honeybees (Apis mellifera): new insights from microsatellite and mitochondrial data. (in manuscript). Franck , P. , Garnery , L. , Celebrano , G. , Solignac , M. et al., (2000a). Hybrid origins of the Italian honeybees, Apis mellifera ligustica and A. m. sicula. Mol. Ecol., 9, 907-923. Franck , P. , Garnery , L. , Solignac , M. and Cornuet , J.-M. (2000b). Molecular confirmation of a fourth lineage in honeybees from Middle-East. Apidologie, 31, 167-180. Franck, P., L. Garnery, A. Loiseau, H.R. Hepburn, M. Solignac and J.-M.

236 Cornuet. (2001). Genetic diversity of the honeybee in Africa: microsatellite and mitochondrial data. Heredity, 86: 420-430. Fresnaye J., Lavie P., Boesinger E. (1974). La variabilité de la production de miel chez l‚abeille de race noire (Apis mellifera L.) et chez quelques hybrides interaciaux, Apidologie, 5: 1-20. Fresnaye J., Lavie P., (1976). Sélection et hybridation de l‚abeilles en France. Symp. Génétique, sélection et reproduction de l‚abeilles, Bull. Tech. Apic. OPIDA 3: 15-20. Frisch, K. von. (1967a). The dance language and orientation of bees. Cambridge, Mass., Harvard Univ. Press. Gadbin, C. (1978). L‘Abeille domestique Apis mellifera L. du Tchat méridional. Ses rapports avec le milieu vérgétal. Thése, l‘Universite d‘Axi- Marseille III, Marseille, France. Gadbin, C, Cornuet, J. M., Fresnaye, J. (1979). Approche biométrique de la variété local d‘Apis mellifera L. dans le sud Tchadien. Apidologie 10: 137- 148. Gan Y.Y., Otis G.W., Mardan M.B., Tan S.G. (1991) Allozyme diversity in Asian Apis, in: Smith D.R. (Ed.), Diversity in the Genus Apis, Westview Press, Boulder, Co., pp. 117–130. Garesse, R. (1988). Drosophila melanogaster mitochondrial DNA: gene organization and evo- lutionary considerations. Genetics 118:649-663. Garnery, L., Vautrin, D., Cornuet, J. M., and Silignac, M. (1991). Phylogenetic relationships in the genus Apis inferred from mitochondrial DNA sequence data. Apidologie 22: 87-92. Garnery , L. , Cornuet, J. M. and Solignac, M. (1992). Evolutionary history of the honey bee Apis mellifera inferred from mitochondrial DNA analysis. Mol. Ecol., 1, 145-154. Garnery, L., M. Solignac, G. Celebrano and J. M. Cornuet. (1993). A

237 simple test using restricted PCR-amplified mitochondrial DNA to study the genetic structures of Apis mellifera L. Experientia, 49: 1016-1021. Garnery, L., E. H. Mosshine and J. M. Cornuet. (1995). Mitochondrial DNA variation in Moroccan and Spanish honey bee populations. Mol. Ecol., 4: 465-471. Garnery, L., P. Franck, E. Baudry, D. Vautrin, J-M. Cornuet and M. Solignac. (1998a). Genetic biodiversity of the West European honeybee (Apis mellifera mellifera and A. m. iberica). I Mitochondrial DNA. Gen. Sel. Evol., 30: 31- 47. Garnery, L., P. Franck, E. Baudry, D. Vautrin, J-M. Cornuet and M. Solignac. (1998b). Genetic biodiversity of the West European honeybee (Apis mellifera mellifera and A. m. iberica). II. Microsatellite loci. Gen. Sel. Evol., 30: 49-74. Georges, J. (1912). La Tunisie Apicole. Ginle et Cie, Tunis, Tunisia. Giavarini, I. (1937). Notizie sulle api e sull apicolture Abissina. V Congresso Nazionale Apicoltora Italiana, Bergamo, Italy, pp. 1- 12. Gillespie, J. H. (1986). Variability of evolutionary rates of DNA. Genetics 112:1077-1091. Goetze, G. K. L. (1940). Die beste Biene. Liepzig: Liedloff, Loth und Michaelis. Goetze, G. K. L. (1964). Die Honigbiene in naturalicher und kunstlicher Zuchtauslese. Teil I, II. Monogr. Angew. Entomol. 19: 1-120, 20: 1-92. Goetze, G. K. L. (1964). Die Honigbiene in natürlicher und Künstlicher Zunchtauslese. Paul Parey, Hamburg. Goncalves, L. S. and Stort, A. C. (1978). Honeybee improvement through behavioural genetics. Ann. Rev. Ent. 23: 197-213.

238 Greenace, M. J. (1988). Clustering the rows and columns of a contingency table. Journal of classification 5: 39- 51. Grout, R. A. (1937). The influence of size of brood cell upon the size and variability of the honeybee (Apis mellifera L.). Research Bulletin, Agricultural Experiment Station, Iowa State College 218: 260- 280. Hall, H. G. (1986). DNA differences found between Africanized and European honeybees. Proc. Natl. Acad. Sci. USA 83:4874-4877. Hall , H. G. (1988). Charactrization of the African honeybee genotype by DNA restriction fragments. In Africanized Honey Bees and Bee Mites, ed. G. R. Needham, R. E. Page, Jr., M. Delfinado-Baker, C. E. Bowman, pp. 287-293. Chichester, England: Ellis Horwood Limited. Hall , H. G. (1992). Further characterization of nuclear DNA RFLP markers that distinguish African and European honeybees. Arch. Insect Biochem. Physiol., 19, 163-175. Hall , H. G. (1998). PCR amplification of a locus with RFLP allele specific to African honey bees. Biochem. Genet., 36, 351-361. Harris, W. V. (1931). A short note on the races of honeybees in Tanganyika Territory. Bee World XII (8): 90-91. Harris, W. V. (1932). Native beekeeping in Tanganyika. Tropical Agriculture 9 (8): 232- 235. Hassanien, M. H. and Banby, M. A. (1956). Studies on the biometrics of the Egyptian honeybee Apis mellifera fasciata Latr. Bull. Sc. Ent. Egypt XI: 127-130. Hassanien, M. H., El-Banby, M. A. (1956a). Biological studies on the Egyptian honeybees, Apis mellifera faciata Latr. (Hymenoptera: Apoidea- Apidae). Bulletin de la Sociètè de Entomologique d‘Ēgypte 40: 199- 206.

239 Hassanien, M. H., El-Banby, M. A. (1956b). Studies on the ability of the Egyptian honeybee on carrying nectar and pollen. Annals of Agriculural Science, Faculty of Agriculture, Ain Shams University, Cairo, Egypt 1 (1): 23- 36. Hepburn, H. R. (1990). What and where is the Cape bee- an interim report. In: Bees and Beekeeping in southern Africa (Eds RH Anderson, B Buys), 74- 83, Apimondia Symposium, Stellenbosch, South Africa. Hepburn, H. R., Jacot, Guillarmod, A. (1991). The Cape honeybee and the fynbos biome. South African Journal of Science 87: 70- 73. Hepburn, H. R., Crew, R. M. (1991a). Portrait of the Cape honeybee, Apis mellifera capinesis. Apidologie 22: 567- 580. Hepburn, H. R., Crew, R. M. (1991b). Geography of the Cape honeybee based on laying worker performance, South Africa Bee Journal 63: 51- 59. Hepburn, H. R. (1993). Swarming, absconding and migration in southern Africa bees. South African Bee Journal 65: 61- 66. Hepburn, H. R., Allsopp, M. H. (1994). Reproductive conflict between honeybees: usurption of Apis mellifera scutellata colonies by Apis mellifera capensis. South Africa Journal of Science 90: 247- 249. Hepburn, H. R., Radloff, S. E. (1995). First approximation to a phenology of the honeybee (Apis mellifera) and Flora of Africa. Oecologia 101: 265- 273. Hepburn, H. R., Radloff, S. E. (1996a). Morphometric and pheromonal analyses of Apis mellifera L along a transect from the Sahara to the Pyrenees. Apidologie 27, 35–45. Hepburn H. R., Radloff S.E. (1997). Biogeographical correlates of population variance in thehoneybees (Apis mellifera L) of Africa. Apidologie 28, 243–258. Hepburn H. R., Radloff, S. E., Fuchs, S. (1998). Population structure 240 and the interface between Apis mellifera capensis and Apis mellifera scutellata. Apidologie 29, 333–346. Hepburn H.R., Radloff S.E. (1998). Honeybees of Africa. Springer Verlag Heidelberg, New York. Hepburn H. R., Radloff, S.E., Oghiakhe S. (2000). Mountain bees of Africa. Apidologie 31: 205–221. .Hernandez-Garcia,R.,B. Fernández,P. De La Rْa, J. Galian and J. Serrano (2002). The effect of transhumance on the geneti variability of Apis mellifera iberica from Murcia (Southeast Spain). In: Jones, R (ed) Bees without frontiers. pp. 96-103. Hepburn H.R., Radloff S.E. (1998). Honeybees of Africa. Springer Verlag Heidelberg, New York. Hepburn H. R., Radloff, S.E., Oghiakhe S. (2000). Mountain bees of Africa. Apidologie 31: 205–221. .Hernandez-Garcia, R., B. Fernández, P. De La Rْa, J. Galian and J Serrano. (2002). The effect of transhumance on the geneti variability of Apis mellifera iberica from Murcia (Southeast Spain). In: Jones, R (ed) Bees without frontiers. pp. 96-103. H. Randall Hepburn; Sarah E. Radloff; Gard W. Otis; Stefan Fuchs; L.R. Verma; Tan Ken; T. Chaiyawong; G. Tahmasebi; R. Ebadi; Siriwat Wongsiri (2005). Apis florea: morphometrics, classification and biogeography1. Apidologie 36 (2005) 359–376. Howell, D. E., and R. L. Usinger. (1933). Observations on the flight and length of life of drone bees. Ann. Enomol. Soc. Amer. 26: 239- 246. Howell V. Daly1 (1991). African Honeybees; 2 Systematics and

identification of Africanized honeybees. Department of Entomology, University of California, Berkeley, CA, 94720, USA.

241 Hsu Chen, C.-C., and D. T. Dubin. (1984a). A cluster of four transfer RNA genes in mosquito mitochondrial DNA. Biochem. Int. 8:385-39 1. Hsu Cchen, C.-C., R. M. Kotin, and D. T. Dubin. (1984b). Sequences of the Coding and flanking regions of the large ribosomal subunit RNA Gene of mosquito mitochondria. Nucleic Acids Res. 12: 7771-7785. Huxley, J. (1938). Clines: an auxiliary taxonomic principle. Nature, 142: 219- 220. Hunt G. J. & Page R. E. (1995) Linkage map of the honey bee, Apis mellifera , based on RAPD markers. Genetics 139: 1371–1382. Itenov K, Pedersen BV (1991) Nucleotide sequence differences in the mitochondrial CO-I gene in the honeybee subspecies Apis mellifera mellifera and Apis m. ligustica. Tidsskr. Planteavl, 95, 101-103. Izquierdo, J., A. Domanguez, J. Albornoz y E. Santiago. (1985). Discriminación entre poblaciones de abejas) Apis mellifera) de Asturias y la Submeseta Norte. Bol. C. Nat. IDEA, 35: 87-101. Juan, C., Ibrahim K. M., Orami, P. Hewitt, G. M. (1995). Mitochondrial DNA variation in Moroccan and Spanish honeybee populations. Molcular Ecology. 4: 465-471. Juan, C., Orami, P. Hewitt, G. M. (1995). Mitochondrial DNA phylogeny and sequencial colonization of Canary Islands by darkling beetles of the genus Pimelia (Tenebrionidae). Proceeding of the Royal Society of London, Series B, 261: 173-180. Kaschef, A. H. (1959a). The single strain of the Egyptian honeybee, Apis mellifera fasciata Latr. Insects Sociaux 6: 243- 257. Kaschef, A. H. (1959b).The sensory physiology and behaviour of the honeybee louse Braula coeca Nitsch. (Diptera: Braulidae).

242 Insects Sociaux 6: 313- 342. Kassaye, A. (1990). The honeybees (Apis mellifera) of Ethiopia- A morphometric Study. MSc. Thesis, Agricultural University of Norway, As, Norway. Kauhausen-Keller D., Keller R. (1994). Morphometrical control of pure race breeding in the honeybee (Apis mellfera L.), Apidologie 25: 133-143. Kellogg, V. L., Bell, R. G., (1904). Studies of variation in insects. Proc. Wash. Acad. Sci. 6: 203- 232. Kelenr-Pillaut, S. (1969). Abeilles fossiles ancestres des Apides sociaux. In “Proceedings, VI Congress IUSSI.”pp. 85-93. Kerr , W. E. and Laidlow, H. H. (1956). General genetics of bees. Advan. Genet., 8: 109- 153. Kerr , W. E. (1957). Introduc E a. o de abelhas Africanas no Brasil Bras. Apic., 8, 211-213. Kerr W.E., Portugal-Araújo V., de (1958). Racas de abelhas de Africa, Garcia da Ortä 6: 53–59. Kerr , W. E. (1970). Natural crossing between Apis mellifera adansonii and Apis mellifera ligustica. Evolution 24: 145- 155. Kerr , W. E. (1992). Abejas Africanas su intriduccion y expansion en el coninente Americano. Subspecies Y ecotipas Africanos. Industria Apicola 13: 12- 21. Kimura , M. (1980). A simple model for estimating evolutionary rates of base substitutions. J. Mol. Evol., 16, 111-120. Kimsey. L. S. (1984). The re-evolution of the phylogenetic relationships in the Apidae (Hymenoptera). Syst. Entomol. 9: 435-441. Kigatiira, K. I. (1984a). The aspects of the Ecology of the African Honeybees. PhD. Thesis, Cambridge University, Cambridge, UK.

243 Kigatiira, K. I. (1984b). Bees and beekeeping in Kenya. Bee World 65 (2): 74-80. Kigatiira, K. I. (1985). Apiculture and Apiculural research in Kenya. Proc. 3rd. Int. Cong Apic. Trop. Climates, Nairobi, 1984: 33-38. King, H. H. (1920). A bee hive designed for the production of bee wax suitable for use by natives of the Southern Sudan. Welcome Tropical Res. Lab. Khartoum Sudan. King, H. H. (1932). On the use of an improved pattern of earthern ware bee hive by natives of the Sudan. Bee- file, Gezira Res. Corp.: 90-94, Wad Medani. Koeniger, N. (1976). Importing “better” bees? Bee World. 57 (2): 411- 412. Koeniger, N., Koeniger, G.; Punchihewa, R. K. W.; Fabritius, Mi. (1982). Observations and experiments on dance comunication in Apis florea in Srilanka. J. Apic. Res. 21 (1): 45-52. Koeniger, N. and Koeniger, G (1983). Observation on mites of the Asian honeybee species Apis cerana, Apis dorsata and Apis florea. Apidologie 14 (3): 197-204. Koeniger, N., G. Koeniger S. Tingek, M. Mardan, and T. E. Rinderer. (1988). Reproductive isolation by different time of drone flight between Apis cerana vechti (Maa, 1953). Apidologie 19:103- 106. Koeniger N., Koeniger G. (2001) Behavioural mating barriers among sympatric species of the genus, Apis, Proc. 7th Int. Conf. Apic. Trop. Climates, Chiang Mai, pp. 80–85. Koulianos S, Crozier RH (1991) Two ancient mitochondrial alleles in Australian honeybees. Apidologie, 22, 621-626. Koeniger N., Koeniger G. (1991) An evolutionary approach to mating behaviour and drone copulatory organs in Apis, Apidologie 22,

244 581–590. Koeniger N., Koeniger G. (2000) Reproductive isolation among species of the genus Apis, Apidologie 31, 313–339. Kuang B.Y., Li Y.C. (1985) The genus Apis in China, Chinese Beekeeping 76, 7–9 (in Chinese). Kumar , S. , Tamora , K. and Nei , M. (1993). Mega : Molecular Evolutionary Genetic Analysis, 1.0 edn. Pennsylvania State University, University Park, PA. Kulagin, N. M. (1906). Die Iane des bienenrussels. Zoologische Anzeiger 29: 711- 716 (cited from Alpatov, 1929). Lavrekhin F. A. (1935) Contribution to the study of variation of the honey bee. 2. A comparative biometric characteristic of the sexual appendages of the drones belonging to different forms of the genus Apis, Zool. Zh. 14, 655–663. La Roche, J., M. Snyder, D. I. Cook, K. Fuller and E. Zouros, (1990). Molecular characterization of a repeat element causing large- scale size variation in the mitochondrial DNA of the sea acallop placopecten magellanicus Mol. Biol. Evol. 7: 45-64. Lebdi-Grissa, K., M‘Sadda, K., Cornuet, J. M., Fresnaye, J. (1991b). Relations phylogénétiques de I‘abeille tunisieme A. m. intermisa aves les races d‘abeilles voisines africaines et ouest méditerranéennes. Revue l‘Agriculture 44: 1231- 1238. Lee, S. L., Wickneswari, R., Mahani, M. C. And Zakri, A. H. (200). Genetic diversity of a tropical tree species, Shorea leprosula Miq. (Dipterocarpaceae), in Malaysia: implications for conservation of genetic resources and tree improvement. Biotropica, 32: 213- 224. Lepeletier, A. (1836). Histoire Naturelle des Insectes. Suites a. Buffon. Hymenopteres. Roret, Paris.

245 Linnaeus, C. (1758). Systema Naturae. Tomus I. Holmiae: L. Salvii. 10th ed. Li S., Meng Y.P., Chang J.T., Li J.H., He S.Y., Kuang B.Y. (1986) A comparative study of esterase isozymes in 6 species of Apis and 9 genera of Apoidea, J. Apic. Res. 25, 129–133. LI, W.-H., and M. Tanimura. (1987). The molecular clock runs more slowly in man than in apes and monkeys. Nature 326:93-96. Liu and Beckenbach, (1992). Mol. Phylogenic. Evol., 1:41- 52. Latreille , P. A. (1804). Notice des espe. ces d’abeilles vivant en grande socieA teA, ou abeilles proprement dites, et description d’espe. ces nouvelles. Ann. Mus. Natl. Hist. Nat., 5, 161-178. Lord, W. G. and Nagi, S. K. A. (1987). Apis florea discovered in Africa. Bee World 68 (1): 39-40. Lobo, J. A., Del Lama, M. A., Mestriner, M. A. (1989). Population differentation and racial admixture in the Africanized honeybee (Apis mellifera L.). Evolution 43: 794- 802. Louveaux J., (1966). Les modalites de l‘adaptation des Abeilles (Apis mellifera L.) au milieu natural. Ann. Abeille 9: 323- 350. Louveaux J., (1969). L‘importance de la notion d [ecotype] chez l‘abeille, in 22 Conger. Int. Apic., Munich, pp. 499-501. Louis, J. Lefebvre, J. (1968). Etude quantitative de la divergence dans I‘evolution morphologique de certaines entites infraspecifiques d‘abeilles domestiques (Apis mellifera L.). C. R. Acad. Sci. Paris 266: 1131- 1133. Louis, J. Lefebvre, J., Moratille, R., Fresnaye, J. (1968). Essai de discrimination de lignees consanguines d‘abeilles domestiques (Apis mellifera mellifera L.) obtenus par insemination artificielle. C. R. Acad. Sci. Paris 267: 526- 528.

246 Maa , T. C. (1953). An inquiry into the systematics of the tribus Apidini or honeybees (Hymenoptera). Treubia, 21, 525-640. Mani, M. S. (1973). General Entomology. Calcutta, Oxford and IBH Publishing Co. India. Manning, F. J. (1952). Recent and fossil honeybees: some aspects of their cytology, phylogeny and evolution. Proc. Lim. Soc. Lond. 163: 3- 8. Manning, F. J. (1960). A new fossil bee from Baltic amber. Proc. XI Int. Congr. Entomol. (Vienna) 1: 306- 308. Malyshev, S. I. (1968). “Geneticis of the Hymenoptera,” Methuen. London. Mattu, V. K., Verma, L. R. (1984). Comparative morphometric studies on the Indian honeybee of the North- West Himalayas 2. Wigs. Journal of Apicultural Research 23: 3- 10. Mark, L. Winston (1987). The Biology of Honey Bees. 2 The Origin and Evolutionary History of Bees. pp. 4. Martins, E., M. A. Mestriner & E. P. B. Contel. (1977). Alcohol dehydrogenase polymorphisms in Apis mellifera . Biochem. Genet. 15: 357-366. Martynov, W. A., (1901). A bee colony exclusively with drone combs. Proceedings of the Mosco Agricultural Academy 7 (1) (cited from Alpatov 1929, in Russia). Maul, V., Hahnle A., (1994). Morphometric studies with pure bred stock of Apis mellifera carnica Pollmann from Hessen, Apidologie 25: 119-132. Mbaya, J. S. K. (1985). The disterbution of African honeybees in Kenya, and some aspects on their behaviour. Proc. 3rd . Int. Conf. Apic. Trop. Climate, Nairobi 1984: 56-60. McCracken, A., I. Uhlenbusch, and G. Gellissen. (1987). Structure of the

247 cloned Locusta migratoria mitochondrial genome: restriction mapping and sequence of its ND- 1 (URF- 1) gene. Curr. Genet. 11:625-630. McMichael , M. and Hall , H. G. (1996). DNA RFLPs at a hightly polymorphic locus distinguish European and African subspecies of the honey bee Apis mellifera L. and suggest geographical origins of New World honey bees. Mol. Ecol., 5, 403-416. Meixner M., Ruttner F., Koeniger N., Koeniger G. (1989). The mountain bees of the Kilimanjaro region and their relation to neighbouring bee populations, Apidologie 20 165–174. Meixner M., Sheppard W.S., Poklukar J. (1993) Asymmetrical distribution of a mitochondrial DNA polymorphism between two introgressing honey bee races, Apidologie 24, 147–153. Meixner , M. D. , Sheppard , W. S. , Dietz , A. and Krell , R. (1994). Morphological and allozyme variability in honey bees from Kenya. Apidologie, 25, 188-202. Meixner , M. D. , Arias , M. C. and Sheppard , W. S. (2000). Mitochondrial DNA polymorphism in honeybee subspecies from Kenya. Apidologie, 31, 181-190. Mestriner, M. A. & E. P. B. Contel. (1972). The P-3 and Est loci in the honeybee Apis mellifera. Genetics 72: 733-738. Merrill, J. H. (1992). The correlation between some physical characters of the bee and its honey- storing abilities. J. Econ. Entomol 15: 125- 129. Micheael, S. Engel (January 12, 2006). A Giant Honey Bee from the Middle Miocene of Japan (Hymenoptera: Apidae). TheAmerican Museum of Natural History Central Park West AT 79 TH street, New York, NY 10024, Number 3504, 12 pp.

248 Michailov, A. S., (1924). On the tongue length of the bees from Tchcherepovetz District in conection with the problem of the local differences in tongue length. Ptschelowdnoje Djelo (cited from Alpatov 1929, in Russian). Michener, C. D. (1944). Comparative external morphology, phylogeny, and a classification of the bees (Hymenoptera). Bull. Am. Mus. Nat. Hist. 82: 151-326. Michener, C. D. (1969). Comparative social behavior of bees. Ann. Rev. Entomol. 14: 299- 342. Michener, C. D. (1974). “ The Social Behavior of the bees,” Harvard Univ. Press, Cambridge. Michener, C. D., M. L. Winston, and R. Jander. (1978). Pollen manipulation and related activities and structures in the Apidae. Univ. Kans. Sci. Bull. 51: 575- 601. Michener, C. D., and Greenberg, L. (1980). Ctenoplecridae and the origin of long-tngued bees. Zool. J. Linn. Soc. 69: 183-203. Michener, C. D. and Grimaldi, D. A. (1988). The oldest fossil bee: Apoid history, evolutionary ststis, and antiquity of social behavior. Pro Natl. Acad. Sci. USA 85: 6424-6426. Michener, C. D. (1990). Classification of the Apidae. Univ. Kansas Sci. Bull. 54: 75-164. Mignotte, F., M. Gueride, A. M. Champagne and J. C. Mounolou, (1991). Direct repeats in the non-coding region of rabbit mitochondrial DNA : involvement in the generation of intra and inter-individual heterogeneity. Eur. J. Biochem. 194: 561-571. Mitev, B.; Stoilov, N. and Kolev, S. (1975). Study of the honeybee in Guinea Republic. Proc. XXVth. Int. Apic. Cong.: 355-356. Mogga, J. B., Ruttner (1986). Apis florea in Africa: source of the founder population. Bee World 69: 100- 103.

249 Mogga, J. B. (1988). The taxonomy and geographical variability of the honeybee Apis mellifera L. in Sudan. M. Sc. Thesis, Khartoum University. Mogga, J. B. (1994). Comparison between the daily flight activity of Apis mellifera yemenitica R. and Apis florea F. IN Khartoum. Apidologie 25: 490. Mohamed, M. A. H. (1982). Morphometrical studies on honeybees (Apis mellifera L., Apidae: Hymenoptera) in the Southern Sudan (Eastern and Western Equatoria Province). M. Sc. Thesis, Khartoum University. Momose, K., Yumoto, T., Nagamitus, T., Kato, M., Nagamasu, H., Sakai, S., Harrison, R. D., Itioka, T., Hamid, A. A. and Inoue, T. (1998). Pollination biology in a lowland diptercarp forest in Sarwak, Malaysia. 1. Characteristics of the plant-pollinator community in a lowland diptercoarp forest. Am. J. Bot., 85: 1477-1501. Moritz, R. F. A., C. F. Hawkins, R. H. Crozier, and A. G. Mackinlay. (1986). A mitochondrial DNA polymorphism in honeybees (Apis mellzjka L.) Experientia 42:322-324. Moritz, C., T. E. Dowling, and W. M. Brown. (1987). Evolution of animal mitochondrial DNA: relevance for population biology and systematics. Annu. Rev. Ecol. Syst. 1: 269-292. Moritz, R. F. A., Meusel, M. S., Habert, M. (1991b). Oligonucleotide DNA fingerprinting discriminates super- and half- sisters in honeybees colonies (Apis mellifera L.). Naturwissenschaften 78: 422- 424. Moritz, R. F. A., Southwick, E. E. (1992). Bees as superorganisms. Spriger-Verlag, Berlin, Germany. Moritz, R.F.A. and Meusel, M.S. (1992). Mitochondrial gene frequencies in Africanized honey bees (Apis mellifera L.): Theoretical model

250 and empirical evidence. J. Evol. Biol. 5: 71-81. Moritz , R. F. A. , Cornuet , J.-M. , Kryger , P. , GarneryA , L. et al., (1994). Mitochondrial DNA variability in South African honeybees (Apis mellifera L.). Apidologie, 25, 169-178. Moritz, R. F. A. (1994). Molecular biology of the honeybee. Advances in Insect physiology 25: 105- 149. Moritz, R. F. A., Haberl, M. (1994). Lack of meiotic recombination in thelytokous parthenogenesis of laying workers of Apis mellifera capensis (the Cape honeybee). Heredity 73: 98- 102. Moritz, R F. A., Kryger, P., koeniger, N. Estoup, A., Tingek, S. (1995): High-degree of polyandry in Apis dorsata queens detected by DNA microsatellite variability. Behavioral Ecology and Sociobiology 37: 357-363. Morse, R. A., Burgett, D. M., Ambrose, J. T., Conner, W. E. and Fell, R. D. (1972). Early Introductions of African bees into Europe and the New World. Bee World 54 (2): 57-60. Morse, R. A. (1977). The Intercontinental movement of honeybees: History advantages and dangers. Proc. Apim. Int. Symp. Apimondia: 58-62. Moss, W. W. and Hendrickson, J. A. (1973). Numerical Taxonomy. Ann. Rev. Ent. 18: 227-258. Murphy, R. M. (1973). Genetic correlations of behaviour. In: Perespective on Animal Behaviour (Ed G Bermant), 72- 101, Scott, Foresman and Co, San Francisco, CA, USA. Mutsaers, M. (1991). Bees in their natural environment in southwestern Nigeria. The Nigerian Field 56: 3- 18. Nagi, S. K. A. (1984). Studies on some factors affecting rearing of queen honeybees Apis mellifera L. (Hymenoptera: Apidae) under Shambat conditions. M. Sc. Thesis, Khartoum University.

251 Nunamaker, R. A., (1980). Subspecies determination in the honeybee (Apis mellfera L.) based on isoelectric focusing of malate dehydrogenase, Ph.D. thesis, University of Wyoming, Laramie, WO. Nanork P., Deowanish S., Wongsiri S. (2001) Mitochondrial DNA variability of dwarf honey bee [Apis florea Fabricius (1787)] in Thailand by using PCR-RFLP technique, Proc. 7th Int. Conf. Apic. Trop. Climates, Chiang Mai, pp. 341–345. Nazzi, F. (1992). Fluctuation of forewing characters in hybrid honeybees from north-eastern Italy. Journal of Apicultural Resarch 31: 27- 31. Ndiritu DW, Mutugi N, Ndungu S (1986). Variation in malate dehydrogenase allozymes among honeybee populations in Kenya. Journal of Apiculutral Research 25:234-237. Nei, M. (1973). Analysis of gene diversity in subdivided populations. Proc. Natl. Acad. Sci., USA, 70: 3321-3323. Newell, W. (1915). Inheritance in the honeybee. Science N. S. 41 (1049): 218- 219. Nightingale J.M. (1983). A Lifetime‘s Recollection of Kenia Tribal Beekeeping. IBRA, London, UK. Ntenga, G. (1976). The honeybee of Tanzania, Apis mellifera adansonii. Twenty-first International Congress of Apiculture, Apimondia, Bucharest, Romania, pp 234- 238. Nolan, W. T., (1929). Success in the artificial insemination of queen honeybees at the bee cluture laboratory. J. Econ. Entomol., 22: 544- 551. Nagi, S. K. A. (1984). Studies on some factors affecting rearing of queen honeybees Apis mellifera L. (Hymenoptera: Apidae) under Shambat conditions. M. Sc. Thesis, Khartoum University.

252 Nanork P., Deowanish S., Wongsiri S. (2001) Mitochondrial DNA variability of dwarf honey bee [Apis florea Fabricius (1787)] in Thailand by using PCR-RFLP technique, Proc. 7th Int. Conf. Apic. Trop. Climates, Chiang Mai, pp. 341–345. Nazzi, F. (1992). Fluctuation of forewing characters in hybrid honeybees from north-eastern Italy. Journal of Apicultural Resarch 31: 27- 31. Ndiritu DW, Mutugi N, Ndungu S (1986). Variation in malate dehydrogenase allozymes among honeybee populations in Kenya. Journal of Apiculutral Research 25:234-237. Nei, M. (1973). Analysis of gene diversity in subdivided populations. Proc. Natl. Acad. Sci., USA, 70: 3321-3323. Newell, W. (1915). Inheritance in the honeybee. Science N. S. 41 (1049): 218- 219. Nightingale J.M. (1983). A Lifetime‘s Recollection of Kenia Tribal Beekeeping. IBRA, London, UK. Ntenga, G. (1976). The honeybee of Tanzania, Apis mellifera adansonii. Twenty-first International Congress of Apiculture, Apimondia, Bucharest, Romania, pp 234- 238. Nolan, W. T., (1929). Success in the artificial insemination of queen honeybees at the bee cluture laboratory. J. Econ. Entomol., 22: 544- 551. Nunamaker, R. A., (1980). Subspecies determination in the honeybee (Apis mellfera L.) based on isoelectric focusing of malate dehydrogenase, Ph.D. thesis, University of Wyoming, Laramie, WO. Nunamaker, R. A., Wilson, W. T. (1981a.) Malate dehydrogenase and nonspecific estrase isoenzyme of eggs of the honeybee (Apis mellifera). Comp. Biochem. Physiol. 70B:607-609.

253 Nunamaker, R. A., Wilson, W. T. (1981b.) Comparison of MDH allozyme patterns in the Africanized honeybee (Apis mellifera adansonii L.) and the Africanized populations of Barazil. J. Kans. Entomol. Soc. 54:704-710. Nunamaker, R. A., Wilson, W. T. & B. E. Halley. (1984.) Electrophoretic detection of Africanized honeybees (Apis mellifera scutellata) in Guatemala and Mexico based on malate dehydrogenases allozyme patterns. J. Kans. Entomol. Soc. 57: 622-631. Nunamaker, R. A., Wilson, W. T., Cox, R. L. (1986). Isozymic uniformity in the presence of environmentally induced morphological variation in Apis mellifera capensis. Annals of the Entomological Society of America 79: 576- 578. Ochman, H., and A. C. Wilson. (1987). Evolution in bacteria: evidence for a universal substitution rate in cellular genomes. J. Mol. Evol. 26:74-86. Oldroyal, B. P., Cornuet, J. M., Rowe, D. Rinderer, T. E., Crozier, R. H. (1995). Racial admixture of Apis mellifera in Tasmania, Australia: similarities and differences with natural hybrid zones in Europe. Heredity 74: 315- 325. Otis, G. (1990). Diversity of Apis in Southeast Asia. In “proceedings, XI Congress, IUSSI, India,” pp. 104-105. Otis, G. (1991). A review of the diversity of the species within Apis. In D. R. Smith (edt.), Diversity in the Genus Apis, pp. 29-49. Westview Press, Boulder, CO. Otis, G. W. (1996). Disterbution of recently recognized species of honeybees (Hymenoptera: Apidae; Apis) in Asia. J. Kans. Entomol. Soc., 69 (Suppl.): 311-333. Orantes-Bermejo F.J. and P. Garcia-Fernandez. (1995). Morphological variability of Apis mellifera iberica in different apiaries of

254 southern Spain. J. Apic. Res., 34: 23-30. Ormi, P., Martin J. L., Medina, A. L., lzquierdo, I. (1991). The evolution of the hypogean fauna in the Canary Island. In. The Unity of Evolutionary Biology. Portland: Discorides Press. Padilla Alvarez, F., R. Hernandez-Fernandez, J. Reyes Lopez, F. Puerta Puerta, J.M. Flores Serrano y M. Bustos. (1998). Estudio morfolَgico de las abejas meliferas del Archipiélago Canario (Gran Canaria, Tenerife, La Palma, Gomera). Arch. Zootec., 47: 451-459. Padilla Alvarez F., M.J. Valerio Da Silva, F. Campano Cabanes, E. Jiménez Vaquero, J.M. Flores Serrano, F. Puerta Puerta y M. Bustos Ruiz. (2001). Discriminacion entre poblaciones de abejas (Apis mellifera L.) del sur de Espana, centro de Portugal y Madeira. Arch. Zootec., 50: 79-89. Pager, H. (1971). Ndedma, Akademische Druck- und Verlagsanstalt, Graz, Austria. Pager, H. (1979). The earliest European depection of African bees 1607. Bee World 57 (3): 103. Palmer , M. R. , Smith , D. R. and Kaftanoglu , O. (2000). Turkish honeybees: genetic variation and evidence for a fourth lineage of Apis mellifera mtDNA. J. Hered., 91, 42-46. Pamilo, P. S. L., Crozier, R. H. (1981). Genetic Variation in maile haploids under deterministic selection. Genetics 98: 190- 214. Papadopoulo (Mrs) (1975). Standing commissions of economy, biology and beekeeping technology. XXVth. Int. Apic. Congr. Apino., 348- 349. Paterson, P. D. (1966). The present economic status of Apis mellifera adansonii: Summary and Bibliography. Bee World 47 (4):123- 131.

255 Paterson, P. D. (1977). Southern Sudan honey production and marketing improvement project. Report and Recommondations. Regional Ministry of Agriculture Library-Juba. Paterson, P. D. (1978). Southern Sudan honey production and marketing improvement project. Second Report and Recommondations. Regional Ministry of Agriculture Library-Juba. Peng , Y. S. , Nasr , M. E. and Locke , S. J. (1989). Geographical races of Apis cerana Fabricius in China and its distribution. Review of recent Chinese publications and a preliminary statistical analysis. Apidologie, 20, 9-20. Peterson, M. (1985). African honeybees in east and west Africa, and africanized bees in Venezuela: Some observations on behaviour. Proceedings of The third International Conference on Apiculture in Tropical Climates, Nairobi, Kenia, pp 109- 111. Phillips, E. F. (1929). Variation and correlation in the appendages of the honeybee. Cornell Univ. Agr. Expt. Sta. Memoir 121: 52 pp. Powell, J. R., A. Caccone, G. D. Amato, and C. Yoon. (1986). Rates of nucleotide substitution in Drosophila mitochondrial DNA and nuclear DNA are similar. Proc. Natl. Acad. Sci. USA 83:9090- 9093. Powers, T. O., E. G. Platzer and B. C. Hyman, (1986). Large mitochondrial genome and mitochondrial DNA size polymorphism in the mosquito parasite, Romanomermis culicivorax. Curr. Genet. 11: 71-77. Punchihewa, R. W. K., Koeniger N., Nevan, P. G. and Gadawski, R. M. (1985). Observations on the dance communication and natural foraging ranges of Apis cerana, Apis dorsata and Apis florea in Srilanka. J. Apic. Res. 24 (3): 168-175. Radloff, S. E., Hepburn, H. R., (1997a). Multivariate analysis of

256 honeybees, Apis mellifera Linnaeus. (Hymenoptera: Apidae), of the Hom of Africa. African Entomology 5: 57- 64. Radloff, S. E., Hepburn, H. R., (1997b). Multivariate analysis of honeybee populations, Apis mellifera Linnaeus (Hymenoptera: Apidae) from western central Africa: morphometrics and phermones. African Entomology 5: 195- 204. Radloff, S. E., Hepburn, H. R., Villet, M. H. (1997). The honeybees Apis mellifera Linnaeus (Hymenoptera: Apidae), of woodland savannah of southeastern Africa. African Entomology 5: 19- 27. Radloff, S. E., Hepburn, H. R., Fuchs, S. (1998). Ecological and morphological differentation of the honeybees, Apis mellifera Linnaeus (Hymenoptera: Apidae), of west Africa. (African Entomology in press). Rand, D. M. and R. G. Harrison, (1989). Molecular population genetics of mtDNA size variation in crickets. Genetics 121: 551-569. Rainey, R. C. (1963). Meterology and the migration of Desert Locust. Tech. Notes Wld. met. Org., 54: 1-115. Rashad, S. E. and El Sarrag, M. S. A. (1977). Beekeeping in the Sudan. Bee World 59 (3): 105-111. Rashad, S. E. and El Sarrag, M. S. A. (1980). Some characters of the Sudanese honeybee Apis mellifera L. 2nd . Int. Conf. Apic. In trop. Climates 301-309. Rashad, S. E. and El Sarrag, M. S. A. Saeed, A. A. and Hussein, M. A. (1984). Morphometrical studies on the Sudanese honeybees. Un published paper (cited from Mogga, 1988). Raven, P. H. and Axelord, D. I. (1974). Angiosperm biogeography and past continental movements. Ann. Mo Bot. Gard. 61:539-673. Ride, W. D., Sabrosky, C. W., Bernardi, G., Melville, R. V. (1985). International Code of Zoological Nomenclature. London: Int.

257 Trust Zool. Nomencl. 3rd ed. Rincon, M., Roubik, D. W., Findegan, B., Delgado, D. and Zamora, N. (1999). Understory bees and floral resources in logged and silviculturally treated Costa Rica rainforest plots. J. Kans. Entomol. Soc., 72: 379-393. Rinderer, T. E., N. Koeniger, S. Tingek, M. Mardan, and G. Koeniger. (1989). Amorphological comparison of the cavity dwelling honeybees of Borneo Apis koschevikovi (Buttel-Reepen, 1906) and A. cerana (Fabricius, 1793). Apidologie 20:405-411. Rinderer, T. E., Daly, H. V., Sylvester, H. A., Lancaster, V. A., Collins, A. M., Buco, S. M., Hellmich, R. L., II. Danka, R. G. (1990). Morphometric differences among Africanized and European honeybees and their FI hybrids (Hymenoptera: Apidae). Ann. Entomol. Soc. Am. 83: 346- 351. Rinderer, T.E., Stelzer, J.A., Oldroyd, B.P., Buco, S.M. and Rubink, W.L. (1991). Hybridization between European and Africanized honey bees in the Neotropical Yucatan Peninsula. Science 253: 309- 311. Rinderer T.E., Oldroyd B.P., Wongsiri S., Sylvester H.A., De Guzman L.I., Potichot S., Sheppard W.S., Buchmann S.L. (1993) Time of drone flight in four honey bee species in south-eastern Thailand, J. Apic. Res. 32, 27–33. Rinderer T.E., Oldroyd B.P., Wongsiri S., Sylvester H.A., De Guzman L.I., Stelzer J.A., Riggio R.M. (1995) A morphological comparison of the dwarf honey bees of south-eastern Thailand and Palawan, Philippines. Apidologie 26, 387–394. Rinderer T.E., Wongsiri S., Kuang B.Y., Liu J.S., Oldroyd B.P., Sylvester H.A., De Guzman L.I. (1996) Comparative nest architecture of the dwarf honey bees.

258 J. Apic. Res. 35, 19–26. Robert, W. C. (1961). Heterosis in the honeybee as shown by morphological characters in inbred and hybrid bees. Ann. Entomol. Soc. Am. 54: 878- 882. Rotter, E. (1920). The Egyptian bee (Apis mellifera var. fasciata, Latr). Bee World 2: 76- 81. Rotter, E. (1921). Di ägyptische Biene. Archiv fur Bienenkunde 3: 1-8. Rueher, J. B., (1926). Les Abeilles de l‘Afrique- Ēquatoriale francaise. Marcel Roux, Paris, France. Ruttner, F. and Mackensen, O. (1952). The genetics of the honeybee. Bee World 33 (4): 53-62. Ruttner, F., (1956a). Zur Frage der spermaübertragung bei der Bienenkönigin. Insectes Sociaux 3: 351- 359. Ruttner, F., (1956b). The mating of the honeybee. Bee World 3: 2- 15, 23- 24. Ruttner, F., (1975). Biological and statistical methods as means of an ingenous intraspecific classification of honeybee. XXVth. Int. Apic. Cong: 360-362. Ruttner, F., (1975a). Races of bees. In The hive and the honeybee, ed. C. P. Dadant, pp. 19- 38. Hamilton, III., Dadant. Ruttner, F., (1975b). African races of honeybees. Proc, XXVth. Int. Apic. Cong. Of Apim: 325-344. Ruttner, F., (1976). The present knowledge on the taxonomy of African races of bees. Proc. Apim. Int. Symposium-Apimondia: 50-54. Ruttner , F. (1976). Les races d’abeille de l’Afrique. In: XXV Congres International d’Apiculture, pp. 347-367, Grenoble. Ruttner, F., N. Koeniger, and H. J. Veith. (1976). Queen substance bei eierlegenden Arbeiterinnen der Honigbiene (Apis mellifera L.). Naturwissenschaften 63: 434- 435.

259 Ruttner, F., Tassenocourt, L. Lauveaux, J. (1978). Biometrical-statistical analysis of the geographic variability of Apis mellifera L. 1. Material and Methods. Apidologie 9 (4): 363-381. Ruttner, F., and Maul, V. (1983). Experimental analysis of the reproductive interspecific isolation of Apis mellifera L. and Apis cerana Fabr. Apidologie 14: 309-327. Ruttner, F. (1983). Queen rearing. Apimondia, Bucharest, Romania. Ruttner, F., and D. Kauhausen. (1985). Honeybee of tropical Africa: Ecological diversification and isolation. 3rd. Int. conf. Apic. Trop. Climates, Nairobi, 1984: 45-57. Ruttner, F., D. Pourasghar, and D. Kauhausen. (1985). Die honigbienen des Iran, II. Apis mellifera meda Skorikow, die persische biene. Apidologie 16: 241- 264. Ruttner, F., (1986). Geographical variability and classification. Bee genetics and breading: 23-56 Academic Press Orlando FA. Ruttner, F., (1987). Biography and Taxonomy of Honeybees. Springer- Verlag Berlin Heidelberg, W. Germany. Ruttner, F., (1988a). “Biogeography and Taxonomy of Honeybees,” Springer-Verlarg, Berlin, Germany. Ruttner, F. (1988b). Principles of geographic variations in honeybees. In Africanized Honey Bees and Bee Mites, ed. G. R. Needham, R. E. Page, Jr., M. Delfinado- Baker, C. E. Bowman, pp. 250- 259. Chichester, England: Ellis Horwood Limited. Ruttner, F. (1988). Biogeography and taxonomy of honeybees. pp: 163- 257. Springer-Verlag, Berlin. Ruttner, F. (1988). Biogeography and taxonomy of honeybees. 284p., Springer-Verlag, Berlin. Ruttner F. (1992) Naturgeschichte der Honigbienen, Ehrenwirth, Műnich. Sakagami, Sh. F., Matsumura, T., and Ito, K. (1980). Apis

260 laboriosa in Himalaya: The little known worlds largest honeybee (Hymenoptera: Apidae). Insecta Matsumurana NS 19: 47-77. Saeed, A. A. (1981). Morphometrical studies on honeybees (Apis mellifera L. Apidae, Hymenoptera) in Eastern, Western and Central Sudan. M. Sc. Thesis. Khartoum University. Sambrook , J. , Fritsch , E. and Maniatis , T. (1989). Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Satta, Y., H. Ishiwa, and S. I. Chigusa. (1987). Analysis of nucleotide substitutions of mito- chondrial DNAs in Drosophila melanogaster and its sibling species. Mol. Biol. Evol. 4:638- 650. Sawadogo, M. (1993). Contribution à I‘étude du cycle des miellées et du cycle biologique annuel des colonies d‘abeilles Apis mellifera adansonii Lat. à l‘ouest du Burkina Fso. Thése Universite de Ouagadougou, Ouagadougou, Burkina Faso. Schiff , N. M. and Sheppard , W. S. (1993). Mitochondrial DNA evidence for 19th century introduction of African honey bees into the United States. Experientia, 49, 350-352. Schmutterer, H. (1965). Pests of Crops in North-east and Central Africa. Stuttgart, Gustav Fischer Verlag. W. Germany. Schneider S., Blyther, R. (1988). The habitat and nesting biology of the African honeybee Apis mellifera scutellata in the Okavango River delta, Botswana, Africa. Insects Sociaux 35: 167- 181. Seeley, T. D. R. H. Seeley, and P. Akratanakul. (1982). Colony defense strategies of the honeybees in Thailand. Ecol. Monog. 52: 43- 63. Severson, D. W., Aiken, J. M., Marsh, R. F. (1988). Molecular analysis of North American and Africanized honeybees. . In Africanized Honey Bees and Bee Mites, ed. G. R. Needham, R. E. Page, M. Delfinado- Baker, C. E. Bowman. pp. 2984- 302. Chichester,

261 England: Ellis Horwood Limited. Sihanuntavong, D., Sittipraneed, S. and Klinbunga, S. (1999). Mitochondrial DNA diversity and population structure of the honeybee, Apis cerana, in Thailand. J. Apic. Res., 38: 211-219. Simon, C., Frank, A. and Martin, A. (1991). The polymerase chain reaction: DNA extraction and amplification. Molecular Techniques in Taxonomy (Hewit, G. M., Johnston, A. W. B. and Young, J. P. W., eds), NATO Advanced Studies Institute, H57: 329-355. Springer, Berlin. Simon , C. , Frati , F. , Beckenbach , A. , Crespi , B. , Liu, H. and Flook , P. (1994). Evolution, weighting and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers. Ann. Entomol. Soc. Am., 87, 651-701. Sheppard, W. S., Berlocher, S. H. (1984). Enzyme polymorphism in Apis mellifera from Norway. J. Apic. Res. 23:4-69. Sheppard, W. S., Berlocher, S. H. (1985). New allozyme variability in Italian honeybees J. Hered. 76:45-48. Sheppard, W. S., Mcpheron, B. A. (1986). Genetic variation in honeybees from an area of racial hybridization in West Czechoslovakia. Apidologie 17:21-32. Sheppard, W. S., Huettel MD. (1988). Biochemical genetic markers, intraspecific variation, and population genetics of honeybee, Apis mellifera. In: Africanized Honey Bees and Bee Mites (Eds GR Needham, RE Page, M Delfinado-Barker, CE Bowman), 281- 286, Ellis Horwood, Chichester, UK. Sheppard, W. S., and Berlocher, S. H. (1989). Allozyme variation and differentation among four Apis species. Apidologie 20: 41. Sheppard , W. S. , Soares , A. E. E. and DeE jong , D. (1991). Hybrid

262 status of honey bee populations near the historic origin of the Africanization in Brazil. Apidologie, 22, 643-652. Sheppard W.S., Rinderer T.E., Mazzoli J., Stelzer J.A., Shimanuki H. (1991) Gene flow between African- and European-derived honey bee populations in Argentina. Nature 349, 782–784. Sheppard, W.S., T.E. Rinderer, M.D. Meixner, H.R. Yoo, J.A. Stelzer, N.M. Schiff, S.M. Kamel and R. Krell. (1996). HinfI variation in mitochondrial DNA of old world honey bee subspecies. J. Hered., 87: 35-40. Sheppard W.S., Arias M.C., Grech A., Meixner M.D. (1997). Apis mellifera ruttneri, a new honey bee subspecies from Malta. Apidologie 28, 287–293. Sheppard W.S. (1997). Subspecies of Apis mellifera, in: Morse R.A., Flottum K. (Eds.), Honey Bee Pests, Predators and Diseases, A.I. Root Co., Medina, OH, USA, pp. 519–533. Sheppard , W. S. , Rinderer , T. E. , Garnery , L. and Shimanuki, H. (1999). Analysis of Africanized honey bee mitochondrial DNA reveals further diversity of origin. Genet. Mol. Biol., 22, 73-75. Smith F. (1858) Catalogue of the hymenopterous insects collected at Sarawak, Borneo, Proc. Linn. Soc. London 2, 42–130. Smith, F. G. (1865). On the species and varieties of the honeybees belonging to the genus Apis. Ann. Mag. Nat. Hist. 3: 372- 380. Smith (1960). Beekeeping in the tropics. London: Longmans. Smith, F. G. (1961). The races of honeybees in Africa. Bee World 43 (10): 255- 260. Smith, F. G. (1972). African raceas of Apis mellifera. Apimondia Sc. Bull: 363 –366. Smith, D. R. (1988). Mitochondrial DNA polymorphisms in five Old World subspecies of honey bees in New World hybrids. pp. 303-

263 312 in M. Delfinado-Baker and C. E. Bowman, eds. Africanized honey bees and bee mites. Halsted, Needham, & Page, New York. Smith , D. R. and Brown , W. M. (1988). Polymorphisms in mitochondrial DNA of European and Africanized honeybees (Apis mellifera). Experientia, 44, 257-260. Smith, R. K. (1988). Identification of Africanization in honeybees based on extracted hydrocarbons assay. In Africanized Honey Bees and Bee Mites, ed. G. R. Needham, R. E. Page, Jr., M. Delfinado- Baker, C. E. Bowman, pp. 275- 280. Chichester, England: Ellis Horwood Limited. Smith, D. R. (1990). Mitochondrial DNA and biogeography of Apis species. In “proceedings, XI Congress, IUSSI, India,” pp. 113- 114. Smith, D. R., and W. M. Brown. (1990). Restriction endonuclease cleavage site and length polymorphisms in mitochondrial DNA of Apis mellifera mellifera and A. m. carnica (Hymenoptera: Apidae). Ann. Entomol. Soc. Am. 83: 81-88. Smith, D. R. (1991). “Diversity in the Genus Apis,” Westview press, Boulder, CO. Smith, D. R., Villafuerte, L., Otis, G. and Palmer, M. R. (2000). Biogegraphy of Apis cerana F. and A. nigrocincta Smith: insights from mtDNA studies. Apidologie, 31:265-279. Smith, D. R., M.F. Palopoli, B.R. Taylor, L. Garnery, J.M. Cornuet, M. Solignac and W.M. Brown. (1991). Geographical overlap of two mitochondrial genomes in Spanish honeybees (Apis mellifera iberica). J. Hered., 82: 96-100. Smith , D. R. (1991a). African bees in the Americas: insights from biogeography and genetics. Trends Ecol. Evol., 6, 17-21. Smith, D. R. (1991b). Mitochondrial DNA and honeybee biogeography. In

264 D. R. Smith (ed.), Diversity in the Genus Apis, pp. 131-176. Westview Press, Boulder, Colorado. USA. Smith , D. R., Glenn, T. C. (1994). Allozyme polymorphisms in Spanish honeybee (Apis mellifera iberica). Journal of Heredity 86: 12- 16. Smith , D. R. and Hagen , R. H. (1996). The biogeography of Apis cerana as revealed by mitochondrial DNA sequence data. J. Kansas Entomol. Soc., 69, 294-310. Snyder, M., A. R. Fraser, J. Laroche, K. E. Gartner -Kepkay, and E. Zouros. (1987). Atypical mitochondrial DNA from the deep-sea scallop Placopecten magellanicus. Proc. Natl. Acad. Sci. USA 84:7595-7599. Solignac, M., and M. monnerot, (1986). Race formation, speciation, and introgression within Drosophila simulans, D. mauritiana, and D. sechellia inferred from mitochondrial DNA analysis. Evolution 40: 531- 539. Solignac, M., M. monnerot., and J. C. Mounolou, (1986). Concerted evolution of sequence repeats in Drosophila mitochondrial DNA. J. Mol. Evol. 24: 53-60. Spivak, M., T. Ranker, O. R. Taylor, W. Taylor & L. Davis (1988). Discrimination of Africanized honeybees using behavior, cell size, morphometrics, and a newly discovered isozyme polymorphism, pp. 313-324. In G. R. Needham et al., (eds.), Africanized honeybees and mites. Horwood, Chichester, England. Spivak, M., Zeltzer, A., DeGrandi- Hoffman, G., Martin J. H. (1990). The influnce of population temperature on the colour patterns and development time of queen honeybees. American Bee Journal 130: 418.

265 Starr. C. k., Schmidt, P. J., and Schmidt, J. O. (1987). Pan-pac. Entomol. 63: 37 - 42. Stauffer, P. H. (1979). A fossil honeybee comb from late Cenozoic cave deposits at Batu Caves, Malay Peninsula. J. Paleontol. 53: 1416- 21. Sylvester, H. A. (1976). Allozyme variation in honeybees (Apis mellifera). Ph.D. disertation, University of California, Davis. Sylvester, H. A. (1982). Electrophoretic identification of Africanized honeybees. Journal of Apicultural Research 21:93-97. Szabo, T. L., Lefkovitch, P. L. )1992). Heritability of colour patterns in a closed population of honeybees. Apidologie 23: 151- 159. Vawter, L., and W. M. Brown. (1986). Nuclear and mitochondrial DNA comparisons reveal extreme rate variation in the molecular clock. Science 234: 194- 196. VlasakK, I., S. Burgschwaiger, and G. Kreil. (1987). Nucleotide sequence of the large ribosomal RNA of honeybee mitochondria. Nucleic Acids Res. 15:2388. Tahmasebi G., Ebadi R., Tajabadi N., Akhoundi M., Faraji S. (2002) The effects of geographical and climatic conditions on the morphological variation and seperation of Iranian small honeybee (Apis florea F.) populations, J. Sci. Technol. Agric. Nat. Res., Iran 6, 169–176. Tamura , K. and Nei , M. (1993). Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Mol. Biol. Evol., 10, 512-526. Tanaka, H., Roubik, D. W., Kato, M., Liew, F. and Gunsalam, G. (2001). Phylogenetic position of Apis nuluensis of northern Bormeo and phylogeography of Apis cerana as inferred from mitochondrial DNA sequences. Insects Soc, 48: 44-51.

266 Tingek, S., M. Mardan, T. E. Rinderer, N. Koeniger, and G. Koniger. (1988). Rediscovery of Apis vechti (Maa, 1953): the Saban honeybee. Apidologie 19: 97-102. Tingek, S., Koeniger, G. and Koniger, N. (1996). Description of a new cavity nesting species of Apis (Apis nuluensis n. sp.) from Sabah, Borneo, with notes of its occurrence and reproductive biology (Insecta: Hymenoptera: Apoidea: Apini). Senckenbergiana Biol., 76: 115-119. Thompson , J. D. , Higgins , D. G. and Gibson , T. J. (1994). Clustal W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position specific gap penalties and weight matrix choice. Nucl. Acids Res., 22, 4673-4680. Thompson , J. N. (1994). The coevolutionary process. 376., University of Chicago, lllinois, USA. Thorpe, R. S. McGregor, D. P., Cumming, A. M. (1993). Molecular phylogeny of the Canary Island Lacertids (Gallotia): mitochondrial DNA restriction fragment divergence in relation to sequence divergence and geological time. Journal of Evolutionary Biology, 6: 725-735. Thorpe, R. S. McGregor, D. P., Cumming, A. M., Jordan, W. C. (1994). DNA evolution and colonization sequence of island lizards in relation to geological history: mtDNA RFLP, Cytochrome B, Cytochrome oxidase, 125 Rrna AND nuclear RAPD analysis. Evolution, 48: 230-240. Tomassone, R., Fresnaye, J. (1971). Etude d‚une methode biometrique permettant la discrimination et la classification de populations d‚abeilles (Apis mellifera L.). Apidologie 2: 49- 65. Tsurata, M., Matsuka, M (1989). Temperature as a causative factor in the seasonal colour dimorphism of Apis cerana japonica workers.

267 Apidologie 20: 149- 155. Uhlenbusch, I., A. Mccracken, and G. Gellissen. (1987). The gene for the large ( 16s) ribosomal RNA from the Locusta migratoria mitochondrial genome. Cm-r. Genet. 11:63 l- 638. Wafa, A. K.; Rashad, S. E. and Mazeed, M. M. (1965). Biometrical studies on the Egyptian honeybee. J. Apic. Res. 4(3): 161 – 166. Wallis, G. P., (1987). Mitochondrial DNA insertion polymorphism and germ line heteroplasmy in the Triturus cristatus complex. Heredity 58: 229-238. Wayne, M. W. and Hendrickson (1973). Numerical taxonomy. An. Rev. Ent. 18: 227 – 258. Winston, M. L., and Michener, C. D. (1977). Dual origin of highly social behavior among bees. Proc. Natl. Acad. Sci. (U.S.) 74: 1134- 1137. Winston, M. L. (1978). “The Biology of the Honeybee.” Harvard Univ. Press, Cambridge. Winston, M. L.; Taylor, O. R. and Otis, G. W. (1979). Absconding behaviour of the Africanized honeybees in South America. J. Apic. Res. 18 (2): 85 – 94. Winston, M. L.; Taylor, O. R. and Otis, G. W. (1983). Some differences between temperate European and Tropical African and South American honeybees. Bee World 64 (1): 12 – 21. Winston, M. L (1988). The impact of a tropical-evolved honeybee in temperate climates of North America. In: Africanized Honey Bees and Bee Mites (Eds GR Needham, RE Page, M Delfinado- Baker,CE Bowman), 135- 140, John Wiley, New York, USA. Whitecombe, R. P. (1985). Aspects of the biology and management of Apis florea in Oman 3rd. Int. Conf. Apic. Trop. Climates, Nairobi, 1984: 96 – 103.

268 Wille, H. (1979). Promoting bee culture in Democratic Republic of Sudan. Report to Director General, Agriculture Research Crop. Willis , L. G. , Winston , M. L. and Honda , B. M. (1992). Phylogenetic relationships in the honeybee (genus Apis) as determined by the sequence of the cytochrome oxidase II region of mitochondrial DNA. Mol. Phylogenet. Evol., 1, 169-178. Wilson, A. C., Cann, R. L., Carr, S. M., George, M., Gyllensten., U. B., Helme- Bychowski, K. M., Higuchi, R. G., Palumbi, S. R., Parager, E. M., Sage, R. D., Stoneking, M. (1985). Mitochondrial DNA and two prespectives on evolutionary genetics. Biol. J. Linnean Soc. 26: 375- 400. Wongsiri, S., K. Limbipichai, P. Tangkanasing, M. Mardan, T. Rinderer, H. A. Sylvester, G. Koeniger, and G. Otis. (1990). Evidence of reproductive isolation confirms that Apis andreniformis (Smith, 1958) is a separate species from sympatric Apis florea (Fabricius, 1787). Apidologie 21: 47-52. Wolstenholme, D. R., J. L. Macfarlane, R. OKIMOTO, D. 0. Clary, and J. A. Wahleith- Ner. (1987). Bizarre tRNAs inferred from DNA sequences of mitochondrial genomes of nem- atode worms. Proc. Natl. Acad. Sci. USA 84: 1324- 1328. Woyke, J. (1977). Comparative biometrical investigation on diploid drones of the honeybee. 1. The head. J. Apic. Res. 16 (3) 131 – 142. Woyke, J. (1978). Comparative biometrical investigation on diploid drones of the honeybee. 11. The Thorax. J. Apic. Res. 17 (4) 195– 205. Woyke, J. (1989). Biology and management of African honeybees Apis mellifera adansonii in Africa. Thirty- second International Congress on Apiculture, Apimondia, Bucharest, Romania, pp. 44- 47. Woyke, J. (1993). Some behavioural characteristics of the Sudanese

269 honeybee. Bee World 74: 133- 140. Woyke, J. (1977). The density of britles covering the wings as discrimination value between African and other races of honeybees. Proc. Apim. Int. Symp. Pretoria 1976: 15 – 24. Wu, C. I., and W. H. LI. (1985). Evidence for higher rates of nucleotide substitution in rodents than in man. Proc. Natl. Acad. Sci. USA 82: 174 1- 1745. Wu, Y., and Kaung, B. (1986). A study of genus Micrapis (Apidae) Zool. Res. 7: 99- 102. Zeuner, F. E. Manning, F. J., and Morris, S. F. (1976). A monograph on fossil bees (Hymenoptera: Apiodea). Bull. Brit. Mus. Nat. Hist. 27: 142-268.

270

CHAPTER SEVEN: APPENDIXES Appendix (A): Abbreviations of the morphometric characters used in the study: Number Measured character Abbreviation 1- Hair length (mm.) Hair 2- Tomentum 1. """" Tom1 3- """""""""""" 2. """" Tom2 4- Proboscis. """"" prob 5- Femur. """"" fem 6- Tibia. """"" tib 7- Tarsus length.""""" ltar 8- """“ Width. """"" wtar 9- Tergite2 pigment. pt2 10- """""3 """"""""" pt3 11- """""4 """"""""" pt4 12- Tergite3 length (mm.). lt3 13- """"""''''4 '''''''''''''''' '''''''''''''' lt4 14- Sternite3 """"" """""" lst3 15- Wax mirror length.(mm.). lwm 16- """"" """""" width. wwm 17- Distance between wax dwm mirrors.(mm). 18- Sternite 6 length. (mm.). lst6 19- """"""""""" width. wst6 20- Forewing length. """""" lfw 21- """""""""" width. """""" wfw 22- Scutellum 1 pigment. scut1 23- """"""""""" 2 """"""""" scut2

271

Appendix (A) continued: 24- Labrum 1 pigment plab1 25- """""""" 2 """""""" plab2 26- Cubital vein1length. (mm.). cub1 27- Cubital vein2. """". """""" cub2 28- Angle A4. ( Degree.). a4 29- """""" B4. """""""""" b4 30- """""" D7. """""""""" d7 31- """""" E9. """""""""" e9 32- """""" G18. """"""""" g18 33- """"""J10. """"""""" j10 34- """"" J18. """"""""" j16 35- """"" K19. """"""""" k19 36- """"" L13. """"""""" l13 37- """" N23. """"""""" n23 38- """" O26. """"""""" o26 39- Hind leg total length. (mm.). Leg 40- Tergite3 + Tergite 4. """"""" lt3lt4 41- Metatarsal index L/W. """"" lw_mtar 42- Wax mirror """"" """". """"" lw_wxm 43- Sternite 6 """" """". """"" lw_st6 44- Cubital vein index a/b. """"" Cind 45- Tomentum index (mm.). Toind 46- Body size/Leg. """""" bs_leg 47- Forewing Length/Width. (mm). Lwfw

272

Appendix (B):

Climatologically & Rainfall averages for at least 30 years from climatological stations in or / near the samples collection areas.

KHARTOUM:

Weather station KHARTOUM is at about 15.60°N 32.50°E. Height about 382m / 1253 feet above sea level.

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year mm 0 0 0.2 0.4 3.7 7.2 48.6 69.1 20.9 4.5 0.2 0 155.5 Inches 0 0 0 0 0.1 0.3 1.9 2.7 0.8 0.2 0 0 6.1

Source: KHARTOUM data derived from GHCN 1. 1088 months between 1899 and 1989.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 31 33.2 36.9 40 41.8 41.2 38 36.7 38.6 39.2 35.231.8 37 °F 87.8 91.8 98.4 104 107.2 106.2 100.4 98.1 101.5 102.6 95.4 89.2 98.6

Source: KHARTOUM data derived from GHCN 2 Beta. 444 months between 1950 and 1987.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 15.6 17 20.3 23.5 26.7 27.2 25.8 25 25.7 25.3 20.9 17 22.5 °F 60.1 62.6 68.5 74.3 80.1 81 78.4 77 78.3 77.5 69.6 62.6 72.5

Source: KHARTOUM data derived from GHCN 2 Beta. 443 months between 1950 and 1987.

SHENDI:

Weather station SHENDI is at about 16.70°N 33.40°E. Height about 360m / 1181 feet above sea level.

273

MADANI:

Weather station WAD MEDANI is at about 14.40°N 33.40°E. Height about 408m / 1338 feet above sea level.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 33.1 35 38.3 40.8 41.5 39.6 35.9 34 35.5 37.9 36.333.5 36.8 °F 91.6 95 100.9 105.4 106.7 103.3 96.6 93.2 95.9 100.2 97.3 92.3 98.2

Source: WAD MEDANI data derived from GHCN 2 Beta. 455 months between 1950 and 1987.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov DecYear °C 14.1 15.8 18.9 21.4 24.3 24.7 23 22.3 21.9 21.7 18.2 15 20.1 °F 57.4 60.4 66 70.5 75.7 76.5 73.4 72.1 71.4 71.1 64.8 59 68.2

Source: WAD MEDANI data derived from GHCN 2 Beta. 455 months between 1950 and 1987.

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year mm 0 0 0 0.2 9.2 26.7 103.2 119 50.8 10 0 0 319.5 inches 0 0 0 0 0.4 1.1 4.1 4.7 2 0.4 0 0 12.6

Source: BARAKAT data derived from GHCN 1. 444 months between 1929 and 1983.

EL-HISSAHISA:

It is about 14.75°N. and 33. 35°E.

274

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year mm 0.0 0.0 0.2 0.1 5.4 23.1 88.8 115.6 46.1 6.1 0.0 0.0 282.7 inches 0.0 0.0 0.0 0.0 0.2 0.9 3.5 4.6 1.8 0.2 0.0 0.0 11.1

Source: RUFA'A data derived from GHCN 1. 437 months between 1913 and 1988.

PORT SUDAN: Weather station PORT SUDAN is at about 19.57°N 37.20°E. Height about 3m / 9 feet above sea level. Average Maximum Temperature

Source: PORT SUDAN data derived from GHCN 2 Beta. 483 months between 1906 and 1947. Average Minimum Temperature

Source: PORT SUDAN data derived from GHCN 2 Beta. 488 months between 1906 and 1947. Average Rainfall

Source: SUAKIN data derived from GHCN 1. 1092 months between 1891 and 1987. 275

KOSTI:

Weather station KOSTI is at about 13.17°N 32.60°E. Height about 381m / 1250 feet above sea level.

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year mm 0 0 0.3 2.8 16.1 45.8 106 137.4 61.1 18.4 1.3 0 391.2 inches 0 0 0 0.1 0.6 1.8 4.2 5.4 2.4 0.7 0.1 0 15.4

Source: KOSTI data derived from GHCN 1. 967 months between 1909 and 1989.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 31.5 33.5 37.5 40 41.2 39.5 36.1 34.5 36.2 38 35.4 32.2 36.3 °F 88.7 92.3 99.5 104 106.2 103.1 97 94.1 97.2 100.4 95.7 90 97.3

Source: ED DUEIM data derived from GHCN 2 Beta. 440 months between 1950 and 1987.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 16.4 17.7 20.5 22.7 24.9 25.2 23.9 23.3 23.4 23.9 20.9 17.6 21.7 °F 61.5 63.9 68.9 72.9 76.8 77.4 75 73.9 74.1 75 69.6 63.7 71.1

Source: ED DUEIM data derived from GHCN 2 Beta. 436 months between 1950 and 1987.

276 DOKKA:

Weather station is at about 12.75°N 35.90°E. Height about 764m / 2506 feet above sea level.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 34.7 36.3 39 40.6 40.3 37.2 33 31.8 33.7 36.7 37 35 36.3 °F 94.5 97.3 102.2 105.1 104.5 99 91.4 89.2 92.7 98.1 98.6 95 97.3

Source: GEDAREF data derived from GHCN 2 Beta. 453 months between 1950 and 1987.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 17.1 18.4 21.5 23.8 24.9 23.1 21.4 21.0 21.2 21.9 20.9 18.1 21.1 °F 62.8 65.1 70.7 74.8 76.8 73.6 70.5 69.8 70.2 71.4 69.6 64.6 70.0

Source GEDAREF data derived from GHCN 2 Beta. 453 month between 1950 – 1987.

GALLABAT:

Weather station GALLABAT is at about 12.80°N 36.17°E. Height about 764m / 2506 feet above sea level.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 35.9 37 38.6 39.3 37.6 33.5 29.6 29.1 30.8 34 35.9 35.7 34.7 °F 96.6 98.6 101.5 102.7 99.7 92.3 85.3 84.4 87.4 93.2 96.6 96.3 94.5 Source: GALLABAT data derived from GHCN 2 Beta. 337 months between 1906 and 1940.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 16.3 18.2 20.3 22.5 22.5 20.2 19.2 18.9 18.8 18.3 16.8 15.7 18.9 °F 61.3 64.8 68.5 72.5 72.5 68.4 66.6 66.0 65.8 64.9 62.2 60.3 66.0

Source: GALLABAT data derived from GHCN 2 Beta. 301 months between 1906 and 1940.

277 El HAWATA:

Weather station HAWATA is at about 13.40°N 34.60°E. Height about 440m / 1443 feet above sea level.

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year Mm 0.0 0.0 0.0 1.2 15.1 99.5 154.1 201.7 79.5 15.8 0.6 0.0 567.8 Inches 0.0 0.0 0.0 0.0 0.6 3.9 6.1 7.9 3.1 0.6 0.0 0.0 22.4

Source: HAWATA data derived from GHCN 1. 420 months between 1950 and 1988.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 34.7 36.3 39.0 40.6 40.3 37.2 33.0 31.8 33.7 36.7 37.0 35.0 36.3 °F 94.5 97.3 102.2 105.1 104.5 99.0 91.4 89.2 92.7 98.1 98.6 95.0 97.3

Source: GEDAREF data derived from GHCN 2 Beta. 453 months between 1950 and 1987.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 17.1 18.4 21.5 23.8 24.9 23.1 21.4 21.0 21.2 21.9 20.9 18.1 21.1 °F 62.8 65.1 70.7 74.8 76.8 73.6 70.5 69.8 70.2 71.4 69.6 64.6 70.0

Source: GEDAREF data derived from GHCN 2 Beta. 455 months between 1950 and 1987.

GALA El- NAHAL:

Weather station is at about 13.60°N 34.75°E. Height about 440m / 1443 feet above sea level.

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year Mm 0.0 0.1 1.0 4.8 25.8 94.5 177.3 192.1 97.6 26.9 4.5 0.0 626.4 Inches 0.0 0.0 0.0 0.2 1.0 3.7 7.0 7.6 3.8 1.1 0.2 0.0 24.7

Source: GEDAREF data derived from GHCN 1. 1047 months between 1903 and 1990.

278 Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 34.7 36.3 39.0 40.6 40.3 37.2 33.0 31.8 33.7 36.7 37.0 35.0 36.3 °F 94.5 97.3 102.2 105.1 104.5 99.0 91.4 89.2 92.7 98.1 98.6 95.0 97.3

Source: GEDAREF data derived from GHCN 2 Beta. 453 months between 1950 and 1987.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 17.1 18.4 21.5 23.8 24.9 23.1 21.4 21.0 21.2 21.9 20.9 18.1 21.1 °F 62.8 65.1 70.7 74.8 76.8 73.6 70.5 69.8 70.2 71.4 69.6 64.6 70.0 Source: GEDAREF data derived from GHCN 2 Beta. 455 months between 1950 and 1987. SINGA: Weather station SINGA is at about 13.15°N 33.95°E. Height about 436m / 1430 feet above sea level.

Average Maximum Temperature

Source: SINGA data derived from GHCN 2 Beta. 321 months

between 1914 and 1943.

Average Minimum Temperature

Source: SINGA data derived from GHCN 2 Beta. 338 months between 1914 and 1943.

279

Average Rainfall

Source: SENNAR data derived from GHCN 1. 945 months between 1907 and 1989.

ROSEIRES: Weather station ROSEIRES is at about 11.85°N 34.38°E. Height about 467m / 1532 feet above sea level. Average Rainfall

Source: ROSEIRES data derived from GHCN 1. 796 months between 1903 and 1987. Average Maximum Temperature

Source: ROSEIRES data derived from GHCN 2 Beta. 483 months between 1905 and 1947.

280

Average Minimum Temperature

Source: ROSEIRES data derived from GHCN 2 Beta. 494 months between 1905 and 1947.

KADUGLI:

Weather station KADUGLI is at about 11.00°N 29.70°E. Height about 499m / 1637 feet above sea level.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 34.9 36.5 39.1 39.8 38.2 34.9 31.8 31.2 32.5 34.9 36.5 35.0 35.4 °F 94.8 97.7 102.4 103.6 100.8 94.8 89.2 88.2 90.5 94.8 97.7 95.0 95.7

Source: KADUGLI data derived from GHCN 2 Beta. 453 months between 1950 and 1987.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year

°C 17.2 19.2 21.8 23.2 23.6 22.4 21.5 21.1 20.6 20.0 18.4 17.7 20.5

°F 63.0 66.6 71.2 73.8 74.5 72.3 70.7 70.0 69.1 68.0 65.1 63.9 68.9

Source: KADUGLI data derived from GHCN 2 Beta. 450 months between 1950 and 1987.

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year mm 0.0 0.8 1.8 14.4 74.9 112.7 150.4 152.8 140.0 73.3 2.9 0.0 725.5 281 inches 0.0 0.0 0.1 0.6 2.9 4.4 5.9 6.0 5.5 2.9 0.1 0.0 28.6

Source: KADUGLI data derived from GHCN 1. 950 months between 1910 and 1989.

El DALANG:

Weather station DILLING is at about 12.00°N 29.60°E. Height about 670m / 2198 feet above sea level.

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year Mm 0.0 0.0 1.8 5.6 33.8 80.5 143.5 159.4 119.4 37.9 0.4 0.0 565.1 inches 0.0 0.0 0.1 0.2 1.3 3.2 5.6 6.3 4.7 1.5 0.0 0.0 22.2

Source: DILLING data derived from GHCN 1. 439 months between 1950 and 1988.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 34.9 36.5 39.1 39.8 38.2 34.9 31.8 31.2 32.5 34.9 36.5 35.0 35.4 °F 94.8 97.7 102.4 103.6 100.8 94.8 89.2 88.2 90.5 94.8 97.7 95.0 95.7

Source: KADUGLI data derived from GHCN 2 Beta. 453 months between 1950 and 1987.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 17.2 19.2 21.8 23.2 23.6 22.4 21.5 21.1 20.6 20.0 18.4 17.7 20.5 °F 63.0 66.6 71.2 73.8 74.5 72.3 70.7 70.0 69.1 68.0 65.1 63.9 68.9

Source: KADUGLI data derived from GHCN 2 Beta. 450 months between 1950 and 1987.

UM RAWABA: Weather station UMM RUWABA is at about 12.80°N 31.20°E. Height about 450m / 1476 feet above sea level. Average Rainfall

282

Source: UMM RUWABA data derived from GHCN 1. 916 months between 1912 and 1989. Average Maximum Temperature

Source: EL OBEID data derived from GHCN 2 Beta. 496 months between 1905 and 1947.

Average Minimum Temperature

Source: EL OBEID data derived from GHCN 2 Beta. 484 months between 1905 and 1947. KUBBUM: Weather station KUBBUM is at about 11.80°N 23.80°E. Average Rainfall

283

Source: KUBBUM data derived from GHCN 1 484 months between 1943 and 1985. ZALINGEI: Weather station ZALINGEI is at about 12.90°N 23.30°E. Height about 900m / 2952 feet above sea level.

Average Rainfall

Source: ZALINGEI data derived from GHCN 1. 620 months between 1929 and 1986. Average Maximum Temperature

Source: Source: NYALA data derived from GHCN 2 Beta. 452 months (1950 - 1987.) Average Minimum Temperature

284

Source: NYALA data derived from GHCN 2 Beta. 440 months between 1950 and 1987.

WAU-SHOLOK:

Weather station is at about 9.75°N 31.80°E. Height about 390m / 1279 feet above sea level.

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year Mm 0.0 0.4 3.2 19.5 65.6 115.3 159.7 174.2 135.8 79.7 4.6 0.0 758.5 Inches 0.0 0.0 0.1 0.8 2.6 4.5 6.3 6.9 5.3 3.1 0.2 0.0 29.9

Source: KODOK data derived from GHCN 1. 816 months between 1903 and 1978.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 35.7 37.2 38.9 38.6 35.8 33.0 31.1 30.9 32.2 34.0 35.6 35.7 34.9 °F 96.3 99.0 102.0 101.5 96.4 91.4 88.0 87.6 90.0 93.2 96.1 96.3 94.8

Source: MALAKAL data derived from GHCN 2 Beta. 391 months between 1915 and 1947.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 18.3 19.9 21.8 23.6 23.0 21.9 21.5 21.4 21.7 21.7 19.4 18.2 21.0 °F 64.9 67.8 71.2 74.5 73.4 71.4 70.7 70.5 71.1 71.1 66.9 64.8 69.8

285 Source: MALAKAL data derived from GHCN 2 Beta. 372 months between 1915 and 1947.

MALAKAL:

Weather station MALAKAL is at about 9.60°N 31.70°E. Height about 388m / 1273 feet above sea level.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 35.1 36.9 38.8 38.5 35.8 32.8 30.9 30.8 31.8 33.4 35.1 34.9 34.6 °F 95.2 98.4 101.8 101.3 96.4 91.0 87.6 87.4 89.2 92.1 95.2 94.8 94.3

Source: MALAKAL data derived from GHCN 2 Beta. 443 months between 1950 and 1987.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 18.4 20.0 22.8 23.9 23.4 22.2 21.7 21.7 21.7 21.7 19.7 18.2 21.3 °F 65.1 68.0 73.0 75.0 74.1 72.0 71.1 71.1 71.1 71.1 67.5 64.8 70.3

Source: MALAKAL data derived from GHCN 2 Beta. 441 months between 1950 and 1987.

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year Mm 0.0 0.3 2.1 14.6 40.3 98.3 139.3 154.4 106.0 71.9 2.8 0.0 647.6 Inches 0.0 0.00.1 0.6 1.6 3.9 5.5 6.1 4.2 2.8 0.1 0.0 25.5

Source: MELUT data derived from GHCN 1. 362 months between 1951 and 1981.

DOLIEB:

Weather station DOLEIB is at about 9.30°N 31.63°E. Height about 391m / 1282 feet above sea level.

Average Maximum Temperature

286 Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year

°C 36.2 37.9 39.7 39.7 37.0 34.3 32.5 32.2 33.7 35.0 36.4 36.0 35.9

°F 97.2 100.2 103.5 103.5 98.6 93.7 90.5 90.0 92.7 95.0 97.5 96.8 96.6

Source: DOLEIB HILL data derived from GHCN 2 Beta. 410 months between 1906 and 1943.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year

°C 18.2 19.8 21.8 23.2 22.7 21.6 21.2 21.2 21.4 21.3 19.2 17.7 20.8

°F 64.8 67.6 71.2 73.8 72.9 70.9 70.2 70.2 70.5 70.3 66.6 63.9 69.4

Source: DOLEIB HILL data derived from GHCN 2 Beta. 416 months between 1907 and 194

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year Mm 0.0 0.3 5.9 28.0 98.5 134.9 204.9 211.7 155.6 92.4 6.9 0.0 944.1 inches 0.0 0.00.2 1.1 3.9 5.3 8.1 8.3 6.1 3.6 0.3 0.0 37.2

Source: FANGAK data derived from GHCN 1. 691 months between 1922 and 1981.

GANAL:

Weather station is at about 9.25.00°N 31.20°E. Height about 390m / 1279 feet above sea level.

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year

Mm 0.0 0.3 2.1 14.6 40.3 98.3 139.3 154.4 106.0 71.9 2.8 0.0 647.6

Inches 0.0 0.0 0.1 0.6 1.6 3.9 5.5 6.1 4.2 2.8 0.1 0.0 25.5

Source: MELUT data derived from GHCN 1. 362 months between 1951 and 1954. 287 Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 18.3 19.9 21.8 23.6 23.0 21.9 21.5 21.4 21.7 21.7 19.4 18.2 21.0 °F 64.9 67.8 71.2 74.5 73.4 71.4 70.7 70.5 71.1 71.1 66.9 64.8 69.8

Source: MALAKAL data derived from GHCN 2 Beta. 372 months between 1915 and 1947.

JUBA:

Weather station JUBA is at about 4.87°N 31.60°E. Height about 460m / 1509 feet above sea level.

Average Maximum Temperature

Jan Feb MarApr MayJun Jul Aug Sep Oct NovDec Year °C 36.9 37.5 37.1 34.9 33.2 32.0 30.8 30.9 32.4 33.5 34.7 35.8 34.1 °F 98.4 99.5 98.8 94.8 91.8 89.6 87.4 87.6 90.3 92.3 94.5 96.4 93.4

Source: JUBA data derived from GHCN 2 Beta. 453 months between 1950 and 1987.

Average Minimum Temperature

Jan Feb MarApr MayJun Jul Aug Sep Oct NovDec Year °C 19.6 21.3 23.1 23.0 22.3 21.4 20.7 20.6 20.7 20.9 20.5 19.5 21.1 °F 67.3 70.3 73.6 73.4 72.1 70.5 69.3 69.1 69.3 69.6 68.9 67.1 70.0

Source: JUBA data derived from GHCN 2 Beta. 452 months between 1950 and 1987.

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year 288 Mm 3.5 11.9 42.1 104.7 155.0 114.4 128.3 137.9 113.7 107.5 41.8 9.7 971.4 inches 0.1 0.5 1.7 4.1 6.1 4.5 5.1 5.4 4.5 4.2 1.6 0.4 38.2

Source: JUBA data derived from GHCN 1. 1045 months between 1901 and 1988.

BANGO:

Weather station is at about 4.90°N 31.50°E. Height about 457m / 1499 feet above sea level.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 36.9 37.5 37.1 34.9 33.2 32.0 30.8 30.9 32.4 33.5 34.7 35.8 34.1 °F 98.4 99.5 98.8 94.8 91.8 89.6 87.4 87.6 90.3 92.3 94.5 96.4 93.4

Source: JUBA data derived from GHCN 2 Beta. 453 months between 1950 and 1987.

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 19.6 21.3 23.1 23.0 22.3 21.4 20.7 20.6 20.7 20.9 20.5 19.5 21.1 °F 67.3 70.3 73.6 73.4 72.1 70.5 69.3 69.1 69.3 69.6 68.9 67.1 70.0

Source: JUBA data derived from GHCN 2 Beta. 452 months between 1950 and 1987.

LIRRIA:

Weather station is at about 4.25°N 32.25°E. Height about 460m / 1509 feet above sea level.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 37.5 37.1 36.3 35.1 33.3 32.1 30.4 30.8 32.7 34 35.4 36.1 34.1

289 °F 99.5 98.8 97.3 95.2 91.9 89.8 86.7 87.4 90.9 93.2 95.7 97 93.4

Source: TORIT data derived from GHCN 2 Beta. 138 months (1922 - 1940.)

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 19 20.2 21 20.8 20.3 19.5 19.1 18.9 18.7 18.9 18.3 18.1 19.3 °F 66.2 68.4 69.8 69.4 68.5 67.1 66.4 66 65.7 66 64.9 64.6 66.7

Source: TORIT data derived from GHCN 2 Beta. 192 months (1922 - 1940.)

Average Rainfall

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year mm 5.8 16.3 52.8 62 85.8 84.7 117.8 120 82.2 65.7 55.5 24.8 785 inches 0.2 0.6 2.1 2.4 3.4 3.3 4.6 4.7 3.2 2.6 2.2 1 30.9

Source: KAPOETA data derived from GHCN 1. 362 months (1951 -1981.)

KHOUR-MAQUIRE:

Weather station is at about 5.25°N 31.75°E. Height about 422m / 1384 feet above sea level.

Average Maximum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 37.5 37.1 36.3 35.1 33.3 32.1 30.4 30.8 32.7 34 35.4 36.1 34.1 °F 99.5 98.8 97.3 95.2 91.9 89.8 86.7 87.4 90.9 93.2 95.7 97 93.4

Source: TORIT data derived from GHCN 2 Beta. 138 months (1922 -1940).

Average Minimum Temperature

Months Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year °C 19.0 20.2 21.0 20.8 20.3 19.5 19.1 18.9 18.7 18.9 18.3 18.1 19.3 °F 66.2 68.4 69.8 69.4 68.5 67.1 66.4 66.0 65.7 66.0 64.9 64.6 66.7

Source: TORIT data derived from GHCN 2 Beta. 192 290 months between 1922 and 1940.

YEI:

Weather station YEI is at about 4.00°N 30.60°E.

Average Rainfall

Source: YEI data derived from GHCN 1. 362 months between 1951

and 1981.

RAJA: Weather station WAU is at about 7.70°N 28.00°E. Height about 438m / 1437 feet above sea level.

Average Rainfall

Source: WAU data derived from GHCN 1. 1008 months between 1904 and 1987.

291

Appendix (C): Taxonomic relationships between bees in the family Apidae. 292

Appendix (D): Different species of the Genus Apini (Institute Für Bienenkunde, Oberursel, Germany).

293

Appendix (E): Natural distribution of Honeybee Species (Institute Für Bienenkunde, Oberursel, Germany).

294

Appendix (F): Geographical distribution of Genus Apis (Institute Für Bienenkunde, Oberursel, Germany).

295

Appendix (G): Distribution of geographical honeybee races and mean of annual temperature (F. Ruttner, Institute Für Bienenkunde, Oberursel, Germany).

296

Appendix (H): Multivariate ANOVA Table: Sum of squares, dF, mean square, F values and significances for each phenotypic character from the measured individuals. (Sudanese honeybee Apis mellifera L.) No. Character Sum of df Mean F Signi- squares square ficances 1- Hair 6.097 2 3.048 0.471 0.633 2- Tom1 776.250 2388.125 6.178 0.010 3- Tom2 100.282 250.141 0.558 0.583 4- Prob 596.188 2298.094 11.523 0.001 5- Fem 120.115 260.057 2.574 0.107 6- Tib 172.043 286.022 1.700 0.214 7- Ltar 350.731 2175.365 9.608 0.002 8- Wtar 102.291 251.146 8.179 0.004 9- Pt2 0.514 20.257 0.328 0.725 10- Pt3 3.448 21.724 2.590 0.106 11- Pt4 1.486 20.743 2.857 0.087 12- Lt3 58.133 229.066 1.435 0.267 13- Lt4 114.567 257.283 3.566 0.052 14- Lst3 512.573 2256.287 12.300 0.001 15- Lwm 111.078 255.539 10.207 0.001 16- Wwm 229.209 2114.605 15.324 0.000 17- Dwm 9.656 24.828 1.251 0.313 18- Lst6 449.644 2224.822 7.809 0.004 19- Wst6 450.963 2225.481 11.075 0.001 20- Lfw 2276.432 21138.216 10.940 0.001 21- Wfw 571.699 2285.849 15.010 0.000 22- Scut1 0.105 20.053 0.119 0.888 23- Scut2 1.894 20.947 0.245 0.786 24- Plab1 25.219 212.610 6.378 0.009

297

Appendix (H) continued:

No. Character Sum of df Mean F Signi- squares square ficances 25- Plab2 2.511 2 1.255 4.404 0.030 26- Cub1 127.136 2 63.568 12.643 0.001 27- Cub2 4.329 2 2.164 1.125 0.349 28- A4 3.769 2 1.885 1.511 0.250 29- B4 32.003 2 16.001 4.991 0.021 30- D7 6.317 2 3.159 0.764 0.482 31- E9 2.545 2 1.272 2.557 0.109 32- G18 25.617 2 12.809 7.477 0.005 33- J10 24.535 2 12.268 3.941 0.041 34- J16 33.840 2 16.920 5.743 0.013 35- K19 28.469 2 14.235 2.430 0.120 36- L13 10.506 2 5.253 1.108 0.354 37- N23 39.370 2 19.685 4.966 0.021 38- O26 71.139 2 35.570 12.461 0.001 39- Leg 1570.136 2 785.068 4.073 0.037 40- Lt3lt4 315.184 2 157.592 2.497 0.114 41- Lw_mtar 1.717 2 0.859 0.591 0.566 42- Lw_wxm 2.868 2 1.434 1.092 0.359 43- Lw_st6 3.408 2 1.704 0.875 0.436 44- Cind 0.601 2 0.301 6.124 0.011 45- Toind 1.525 2 0.762 0.735 0.495 46- Bs_leg 0.536 2 0.268 0.137 0.873 47- Lwfw 1.051 2 0.526 2.522 0.112 p ≤ 0.005

298

299

300 301 302

304

305

306

307

308

309

310

312

313

314

315 316

317

318

319

320

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

352