Extracto
Table of Contents
Dedication
Acknowledgements
Zusammenfassung
Abstract
Abbreviations
Table of Contents
List of Figures
List of Tables
1 Introduction
1.1 Honeybees and Man
1.2 The genus Apis
1.3 The species A. mellifera
1.4 A. mellifera in Africa
1.5 A. mellifera in Western Africa
1.6 Conservation of Local Populations of Honeybee
1.6.1 Breeding
1.6.2 The Varroa Problem
1.6.3 The Africanised honeybees
1.6.4 The Capensis Calamity
1.6.5 The Small Hive Beetle
1.6.6 The Honeybees of Africa
1.7 Methods of Characterising Honeybee Populations
1.7.1 Morphometry
1.7.2 Variation of Mitochondrial DNA
1.7.3 Microsatellite Polymorphism of Nuclear DNA
1.8 Research questions
1.9 Aim and Objectives of the Study
1.10 Research Hypothesis
2 Materials and Methods
2.1 Description of the Area of Study
2.2 Sampling the Area of Study
2.3 Collection of Samples of Honeybee
2.3.1 Fieldtrips
2.3.2 Sources of Honeybees
2.3.3 Sampling of Honeybees
2.4 Morphometry
2.4.1 Dissection of Honeybees
2.4.2 Measurements
2.4.3 Multivariate Statistical Analysis of Morphometric Data
2.4.3.1 Analysis of samples under investigation
2.4.3.1.1 Descriptive Statistics and Analysis of Variance
2.4.3.1.2 Principal Component Analysis
2.4.3.1.3 Hierarchical Cluster Analysis
2.4.3.1.4 Discriminant Analysis
2.4.3.1.5 Correlation Analysis
2.4.3.2 Analysis of samples under investigation together with reference samples
2.4.3.2.1 Principal Component Analysis
2.4.3.2.2 Discriminant analysis
2.5 Mitochondrial DNA Analysis
2.5.1 Analysis of the tRNAleu-COII intergenic region
2.5.1.1 Polymerase Chain Reaction (PCR)
2.5.1.2 DRA I Restriction Digestion
2.5.2 Sequence Variation of the Cytochrome b gene
2.5.2.1 Polymerase Chain Reaction (PCR)
2.5.2.2 Sequence alignment and phylogenetic analysis
2.5.3 Statistical Analysis of Mitochondrial Haplotypes
2.6 Microsatellite Polymorphism
2.6.1 Extraction and Genotyping of DNA
2.6.2 Statistical Analyses
2.7 Bibliography
3 Results
3.1 Morphometry
3.1.1 Variation between the Colonies under investigation
3.1.1.1 Descriptive Statistics and Analysis of Variance
3.1.1.2 Similarity of Colonies
3.1.1.3 Classification of Colonies Based on Localities
3.1.1.4 Classification of Colonies Based on Vegetation
3.1.1.5 Relationship between Morphology and Environmental Factors
3.1.2 Similarity of Colonies under Investigation and Reference Samples
3.2 Mitochondrial DNA
3.2.1 COI-COII Haplotypes
3.2.2 Cytochrome b Sequence Analysis
3.3 Microsatellite Polymorphism
4 Discussion
4.1 Morphometry
4.1.1 Morphometric Variation of Honeybees in the Area of Study
4.1.2 Morphometric Variation of the honeybees of the Study Area in the Context of Published Subspecies
4.2 Mitochondrial DNA Polymorphism
4.3 Microsatellite polymorphism
4.4 Conclusions
References
Appendix I
Appendix II
Curriculum Vitae
Acknowledgements
I wish to express my profound gratitude to my supervisors, Professor Dr. Bernd Grünewald and Dr. Marina D Meixner for their guidance, suggestions, constructive criticism and other forms of support leading to the preparation of this thesis. I am immensely indebted to PD Dr. Stefan Fuchs. His guidance in the analysis of data, valuable suggestions and untiring response to my questions have done a lot in improving the quality of this work. I thank Dr. Per Kryger of Aarhus University, Denmark, for tutoring me in population genetics and for his assistance in microsatellite analysis. Similarly, I wish to express my sincere gratitude to Ms Beate Springer of Institut für Bienenkunde, Oberursel, for the training in morphometric techniques and assistance in dissections and measurements, to the Institute for providing the facilities and to the members of our research group for sharing their knowledge. In particular, I wish to thank Ms Valentina Resnik for her invaluable assistance, especially in formatting and printing the thesis. Dr. Marina D Meixner translated the summary from English to German. I am grateful to the LLH Bieneninstitut Kirchhain and its parent organisation, Landesbetrieb Landwirtschaft Hessen, for accepting me as a visiting scientist to work in the institute's laboratories and for providing all the facilities and support I needed. In this regard I wish to particularly thank the head of the institute, Dr. Ralph Büchler and his entire staff, most especially the technicians – Ms Sandra Backhaus, Ms Anja von Gall, Ms Gerda Waldschmidt and Ms Elke Loider - who were always willing to attend to my numerous requests. My colleagues in the institute – Ms Ina Heidinger, Ms Vera Poker, Dr. Reinhold Siede and Dr. Gefion Brunnen-Stube - deserve special thanks for the friendship and support I enjoyed from them. My special thanks go to the numerous beekeepers, honey hunters, public servants, traditional rulers, my former students, friends and other well-wishers for their support during my fieldtrips in Nigeria, Niger, Cameroon and Chad. Likewise, I am grateful to my extended family, friends, colleagues and neighbours in Nigeria for their support throughout the period of my study. Particularly, I would like to express my profound gratitude to my friend, Professor Saminu Ibrahim, for his immense support. Finally, I wish to thank my employer, Abubakar Tafawa Balewa University, Bauchi, Nigeria for awarding me a study fellowship, funded by the Government of Nigeria, through TETFUND, to pursue this study in Germany.
Zusammenfassung
Einleitung
Durch ihre Bestäubung von Wild- und Nutzpflanzen sowie durch die Produktion der Wirtschaftsgüter Honig und Wachs sind Honigbienen von hoher gesellschaftlicher und wirtschaftlicher Bedeutung. So wurde zum Beispiel der Beitrag der Bestäubungsleistung von Honigbienen zur Ökonomie des Vereinigten Königreichs mit 191.80 Millionen £ geschätzt (Anonymous, 2009). Im Jahr 2011 produzierten die 20 Top-Produktionsländer zusammengenommen 1.26 Millionen Tonnen Honig und Wachs, die einen Wert von 3.16 Milliarden US$ hatten (Anonymous, n. d.). Seit über 4000 Jahren halten Menschen daher Honigbienen aus wirtschaftlichen Gründen (Crane, 1999).
Das natürliche Verbreitungsgebiet der westlichen Honigbiene, Apis mellifera, umfasst Europa, Afrika und das westliche Asien. Geographische Isolation und ökologische Anpassung führten zur Entstehung lokaler Populationen mit erheblicher geographischer Variation und Anpassung an regionale Bedingungen von Klima und Vegetation, sowie an Schädlinge und Krankheitserreger (Marina D. Meixner et al., 2013; F. Ruttner, Tassencourt, & Louveaux, 1978).
Durch Eingriffe des Menschen, z. B. in der kommerziellen Bienenhaltung, können diese Anpassungen auf verschiedene Weise verlorengehen: z. B. durch Konkurrenz um Nahrungspflanzen, Einfuhr fremden Genmaterials, oder durch Einfuhr von exotischen Schädlingen, Parasiten und Krankheiten (Pilar De la Rúa, Jaffé, Dall'Olio, Muñoz, & Serrano, 2009; Robin F. A. Moritz, Härtel, & Neumann, 2005). Für eine nachhaltige Bienenhaltung ist es daher notwendig, die genetische Diversität von Honigbienen in ihrer natürlichen Verbreitung zu schützen. Um einen solchen Schutz erreichen zu können, ist es zunächst jedoch notwendig, einheimische Bienenpopulationen zu charakterisieren, damit eine Einfuhr fremder Genotypen erkannt werden kann (Franck et al., 2001; Marina D. Meixner et al., 2013).
Die gebräuchlichsten Verfahren zur wissenschaftlichen Charakterisierung von Populationen der Honigbiene sind die Morphometrie sowie der Analyse der Variation von mitochondrialer DNA und Mikrosatelliten. Die Morphometrie beruht auf statistischen Analysen exakter Messungen an Einzelbienen und ihren Mittelwerten für ein Bienenvolk (F. Ruttner, 1988). Für diese Messungen bildet ein Katalog von morphologischen Merkmalen wie Körpergröße, Färbung, Behaarung und Form die Grundlage.
Die nur von der Mutter vererbte mitochondriale DNA enthält bei Apis mellifera eine Zwischengenregion, die nicht-kodierend und daher hochvariabel ist (weil sich Mutationen hier ansammeln können). Diese Region eignet sich daher gut als Marker zur Untersuchung von geographischer Variabilität und Introgression von Unterarten und wird für derartige Studien intensiv genutzt (J. M. Cornuet, Garnery, & Solignac, 1991; R. H. Crozier & Crozier, 1993).
Von beiden Eltern vererbt werden sogenannte DNA Mikrosatelliten, die auch als variable Tandemrepeats (VNTR), kurze Tandemrepeats (STR) oder „simple sequence repeats (SSR) bezeichnet werden. Die Motive von Mikrosatelliten sind typischerweise zwischen einer und sechs Basenpaare lang und können vielfach, bis zu 100mal, wiederholt werden. Mikrosatelliten treten über das gesamte Genom verteilt auf, sind aber in nicht-kodierenden Abschnitten häufiger anzutreffen. Aufgrund ihrer hohen Mutationsrate ist die Variabilität von Mikrosatelliten sehr hoch. Diese Eigenschaft macht sie, zusammen mit ihrer Häufigkeit und Verteilung über das gesamte Genom, zu nützlichen Markern in Studien zur genetischen Diversität, Populationsstruktur, Introgression, Genfluss und ähnlichen Fragestellungen (J.-M. Cornuet, Piry, Luikart, Estoup, & Solignac, 1999; Marina D. Meixner et al., 2013; Nedić et al., 2014; Rowe, Rinderer, Stelzer, Oldroyd, & Crozier, 1997; Sainudiin, Durrett, Aquadro, & Nielsen, 2004).
Innerhalb des Verbreitungsgebiets von A. mellifera ist Westafrika eine der bisher am wenigsten gut untersuchten Regionen. Aus dem gesamten Gebiet von West- und Zentralafrika, von Mauretanien und Senegal im Westen, bis nach Chad im Osten, im Süden bis Namibia und Sambia, wurden bisher Proben von nur 190 Völkern gesammelt und morphometrisch untersucht (H. R. Hepburn & Radloff, 1998), sowie Daten zur mitochondrialen DNA von 114 Völkern und Mikrosatelliten von nur 30 Völkern (Franck et al., 2001).
Das Ziel dieser Arbeit war es, durch die Untersuchung von Proben mit morphometrischen Methoden, sowie der Analyse von mitochondrialer DNA und Mikrosatelliten den Erkenntnisstand zur Variabilität der Honigbienen von Westafrika zu verbessern. Im Einzelnen sollten folgende Ziele erreicht werden:
i. Die in Westafrika vorkommenden Unterarten von A. mellifera zu bestimmen
ii. Die in Westafrika vorkommenden Populationen von A. mellifera genetisch zu charakterisieren
iii. Die genetische Variabilität dieser Bienen zu bestimmen
iv. Gründe für die beobachtete Variabilität oder deren Fehlen zu diskutieren
v. Die phylogenetische Verwandtschaft dieser Bienen zu bestimmen
Sammeln von Proben aus Bienenvölkern
Über einen Zeitraum von fünf Jahren wurden insgesamt 204 Bienenvölker an 44 verschiedenen Orten in vier Ländern – Nigeria, Niger, Chad, Kamerun – beprobt (Abb. 2.5 und Tabelle 2.1). Von jedem Volk wurden etwa 20 Arbeitsbienen gesammelt und in 70% oder 90% Ethanol konserviert.
Morphometrische Analyse
Zehn Arbeitsbienen aus jeder Probe wurden präpariert, und morphometrische Messungen von 35 Merkmalen nach F. Ruttner (1988) und F. Ruttner et al. (1978) wurden vorgenommen (Tabelle 2.4). Diese Messungen wurden an Proben von 86 Völkern aus 23 Standorten durchgeführt (Tabelle 2.2). Aus zeitlichen Gründen wurden von 66 weiteren Proben aus 20 weiteren Sammelorten und vier der oben enthaltenen Orte (Sarh, Bauchi, Jos und Umuahia) nur die Merkmale des Vorderflügels gemessen (Tabelle 2.3).
Zur Analyse der Variabilität wurden die morphometrischen Daten dieser Proben einer Hauptkomponentenanalyse (Principal Component Analysis, PCA) unterzogen. Drei Hauptkomponenten, mit Eigenwerten von 8.2, 2.1 und 0.9, wurden extrahiert, die 58.4%, 14.7 und 6.7% der Gesamtvarianz erklärten. Größenmerkmale zeigten eine hohe Ladung auf allen drei Komponenten. Zusätzlich hatte die Pigmentierung des Scutellum eine hohe Ladung auf Komponente 3 (Tabelle A7, Appendix II). Die Positionen der Proben auf den Hauptkomponenten sind in den Streudiagrammen der Abbildung 3.1 dargestellt.
Da sich die Proben in der PCA nicht in klar erkennbaren Gruppen anordneten, wurden sie im Folgenden einer hierarchischen Clusteranalyse unterzogen, der sich eine stufenweise Diskriminanzanalyse (DA) anschloss. Darin wurden drei Cluster erkennbar (Abb. 3.2 und 3.3).
Die morphologische Ähnlichkeit der Völker aus den vier Vegetationszonen des Studiengebiets wurde ebenfalls mit einer hierarchischen Clusteranalyse untersucht. Hierzu wurden die Mittelwerte der morphometrischen Merkmale aus den vier Vegetationszonen herangezogen. Hieraus ergaben sich drei überlappende Cluster, die durch eine Diskriminanzanalyse (DA) bestätigt wurden (Abb. 3.4. und 3.5).
Ein Konturplot der Hauptkomponente 1 aus der PCA gegen die geographischen Koordinaten der Sammelorte ist in Abb. 3.6 dargestellt und ergibt ein ungleichmäßiges Bild. Im westlichen Teil des Untersuchungsgebiets nimmt Komponente 1 mit der geographischen Breite zu, im östlichen Teil jedoch ab, wobei sie offensichtlich dem Relief des Gebiets folgt. (Abb. 2.4). Infolgedessen wurden die Proben in einer weiteren Analyse in zwei Gruppen geteilt: eine Atlantische Gruppe, die Proben aus dem westlichen Teil des Gebiets umfasste, sowie eine Tschadische Gruppe mit Proben aus dem östliche Teil des Untersuchungsgebiets.
Für jede Gruppe wurde eine Korrelationsanalyse (Pearson-Produktmoment) durchgeführt, um den Zusammenhang zwischen den drei Hauptkomponenten der PCA und den Faktoren geographische Länge und Breite, Meereshöhe, Temperatur, Regenmenge aufzuklären.
In der Atlantischen Gruppe zeigte die erste Hauptkomponente eine starke positive Korrelation mit geographischer Breite (r (47) = .571, p < .001), eine schwache positive Korrelation mit der Temperatur (r (47) = .292, p < .05) und eine moderate negative Korrelation mit der Regenmenge (r (47) = -.411, p < .01). Diese Komponente korrelierte nicht mit der geographischen Länge oder der Meereshöhe. Die Hauptkomponente 3 zeigte eine moderate positive Korrelation mit der geographischen Breite (r (47) = .334, p < .05), und eine moderate negative Korrelation sowohl mit der Meereshöhe (r (47) = -.403 p < .01) und der Regenmenge (r (47) = -.421 p < .01). Im Gegensatz dazu korrelierte die zweite Hauptkomponente mit keiner der genannten Variablen.
Da, im Gegensatz zur beobachteten positiven Korrelation zwischen Breite und Hauptkomponente 1, hier eine negative Korrelation erwartet worden war, wurde dieser Zusammenhang durch eine partielle Korrelationsanalyse erster Ordnung weiter untersucht um die Effekte von Temperatur, Regenmenge und Länge aufzuklären (also Variablen, die ihrerseits entweder mit der Breite oder mit der Hauptkomponente 1 korrelierten). Die Korrelation zwischen der geographischen Breite und Hauptkomponente 1 war dabei über die Effekte von Temperatur, Regenmenge und Länge hinaus statistisch signifikant (r (41) = .793, p < .001).
Zur Überprüfung wurde umgekehrt der Zusammenhang zwischen Temperatur und Hauptkomponente 1 mit einer partiellen Korrelation erster Ordnung untersucht, um die Effekte von Breite, Länge, Meereshöhe und Regenmenge aufzuklären (also Variablen, die ihrerseits mit Temperatur und Hauptkomponente 1 korrelierten). Diese partielle Korrelation war jedoch statistisch nicht signifikant (r (41) = -.191, p > .05).
In der Tschadischen Gruppe ergab sich für die erste Hauptkomponente eine moderate negative Korrelation sowohl mit der geographischen Breite (r (38) = -.413, p < .01) als auch der Länge (r (38) = -.384, p < .05), eine starke positive Korrelation mit der Meereshöhe (r (38) = .563, p < .001), eine moderate positive Korrelation mit der Regenmenge (r (38) = .453, p < .01) und eine starke negative Korrelation mit der Temperatur (r (38) = -.586, p < .001). Es bestand eine moderate negative Korrelation zwischen den Hauptkomponenten 1 und 2 (r (38) = -.340, p < .05) sowie zwischen der Hauptkomponente 2 und der Länge. Die Hauptkomponente 3 korrelierte moderat negativ sowohl mit der Breite (r (38) = -.362, p < .05) und der Temperatur (r (38) = -.398, p < .05) und moderat positiv mit der Meereshöhe (r (38) = .476, p < .01) und der Regenmenge (r (38) = .409, p < .05).
Die morphologische Variabilität der für diese Studie gesammelten Proben im Kontext von bereits publiziertem Referenzmaterial wurde zunächst ebenfalls mit einer PCA untersucht, die die morphometrischen Daten von 85 neu gesammelten Proben und 69 Referenzproben (jeweils 21 von A. m. jemenitica und A. m. scutellata, und 27 von A. m. adansonii) einschloss. Vier Hauptkomponenten mit Eigenwerten größer als eins wurden extrahiert, die zusammen 79.7% der Gesamtvarianz erklärten. Auf der Basis des Screeplots wurden die ersten drei Hauptkomponenten mit Eigenwerten von 8.5, 2.0 und 1.2, die 53.2%, 12.5% und 7.6% der Gesamtvarianz erklärten, beibehalten. Die Ladungen der Merkmale auf den Komponenten und die Kommunalitäten der rotierten Darstellung sind in Tabelle A12, Appendix II dargestellt. Alle größeren Ladungen korrelierten positiv mit der jeweiligen Hauptkomponente. Die Positionen der Proben auf den verschiedenen Achsen sind in den Scatterplots der Abbildung 3.7 dargestellt.
Im Anschluss an die PCA wurde eine stufenweise DA durchgeführt um die Gruppenzugehörigkeit von 83 neuen Proben zu den durch die Referenzproben repräsentierten Unterarten zu berechnen. Hier wurden 94.2% der Proben korrekt klassifiziert, wobei die durch eine F-Statistik berechneten Abstände zwischen den Zentroiden der drei Gruppen hochsignifikant waren (p < 0.0005). Von den Proben der neuen Sammlung wurden 61 als A. m. jemenitica und 22 als A. m. adansonii klassifiziert (Abb. 3.8 und 3.9).
Analyse der mitochondrialen DNA
Die Analyse der mitochondrialen DNA wurde an 148 Proben von 39 Sammelorten durchgeführt (Tabelle 2.5). Die DNA einer Biene pro Probe wurde mit dem DNeasy® Blood & Tissue Kit (QIAGEN, 2006a) nach der vom Hersteller für Insekten vorgesehenen Arbeitsanweisung durchgeführt (QIAGEN, 2006b). Der Extrakt wurde bis zur weiteren Verwendung bei -20°C aufbewahrt.
Der die Zwischengenregion enthaltende mtDNA-Abschnitt zwischen den Untereinheiten I und II des CytochromC-Oxidase-Gens (COI-COII) wurde mittels PCR amplifiziert, wobei das Primerpaar E2-H2 eingesetzt wurde (Garnery et al., 1993). Die Reaktionsbedingungen richteten sich nach Garnery et al. (1998); Garnery et al. (1993) und Kandemir et al. (2006). Um den Erfolg der PCR zu überprüfen sowie die Größe des Amplifikats zu bestimmen wurde ein 5µl Aliquot des PCR-Produkts in einem 1.5% Agarosegel einer Elektrophorese unterzogen, mit Ethidiumbromid angefärbt und unter UV-Beleuchtung fotografiert. Die restlichen 25 µl des PCR-Produkts jeder positiven Reaktion wurden mit dem Restriktionsenzym Dra I bei 37°C über Nacht verdaut. Die Restriktionsfragmente wurden auf 10% Polyacrylamidgelen separiert, mit Ethidiumbromid angefärbt und unter UV-Illumination fotografiert. Die Größe der Restriktionsfragmente wurde anhand von mitgelaufenen Größenstandards bestimmt und zur Identifizierung der Haplotypen mit den Mustern der bereits publizierten Haplotypen verglichen (Franck et al., 2001).
Für einen Teil der Proben wurde ein Teil des mitochondrialen CytochromB Gens sequenziert. Mit Hilfe der PCR wurde ein etwa 800 bp langes Fragment amplifiziert, wobei die Primer CB2 und tSer eingesetzt wurden (abgeleitet von Y. C. Crozier, Koullianos, and Crozier (1991), und Garnery et al. unpublished data). Die Produkte wurden auf 1.5% Agarosegelen separiert, mit Ethidiumbromid angefärbt und unter UV-Illumination fotografiert. Die Amplifikate positiver Reaktionen wurden aufgereinigt und in einer kommerziellen Einrichtung sequenziert (Seqlab, Göttingen, Germany).
Insgesamt wurden vier verschiedene Haplotypen in den untersuchten Proben identifiziert, die alle zur afrikanischen Lineage A gehören und bereits publiziert sind (Mogbel A. A. El-Niweiri & Moritz, 2008; Franck et al., 2001; Shaibi et al., 2009): A1 (n = 62), A4 (n = 70), A4' (n = 15) und A14 (n = 1). Die insgesamt beobachtete Diversität der Haplotypen war niedrig (h = 0.478 ± S. E. 0.057). Die Verteilung der Haplotypen im Zusammenhang mit Umweltbedingungen wurde mittels eines Chi-Quadrat-Tests zwischen Haplotyp und den Variablen Vegetation, geographische Breite, geographische Länge. Meereshöhe, Temperatur und Regenmenge geprüft. Dabei wurde ein statistisch signifikanter Zusammenhang zwischen dem Haplotyp und jeder der sechs Variablen festgestellt. Der Zusammenhang mit der Breite war dabei stark ausgeprägt, mit der Vegetation und der Regenmenge moderat, und mit den übrigen Variablen schwach ausgeprägt (Tabelle A14, Appendix 2; Abb. 3.10 und 3.11).
Bei der Amplifikation des mitochondrialen CytochromB Gens entstand ein Fragment von etwa 800 Basenpaaren; davon konnten 696 zwischen den 11 neu sequenzierten Proben und bereits publizierten Referenzproben aligned werden. Innerhalb des Alignments waren 37 Basen variabel, 28 davon waren unter dem Parsimony-Kriterium informativ. Im phylogenetischen Neighbour-Joining -Tree ordneten sich alle Proben aus dieser Studie (gesammelt an verschiedenen Orten in Westafrika) zusammen mit den Referenzproben von A. m. scutellata aus Kenya (Referenzproben aus Westafrika stehen in Genbank nicht zur Verfügung) eindeutig in einen einzigen Ast ein (Abb. 3.12). Innerhalb dieses von Proben aus dem sub-Sahara-Afrika gebildeten Astes war jedoch keine Auflösung zu erkennen.
Analyse von Mikrosatelliten
Die genetische Variabilität der Honigbienen im Studiengebiet wurde anhand von polymorphen nuklearen Mikrosatelliten untersucht. Dazu wurden 133 Arbeitsbienen (eine pro Bienenvolk) von 38 Sammelorten (Tabelle 2.6) zum Zweck der DNA- Extraktion und Genotypisierung an eine kommerzielle Einrichtung geschickt. Die folgenden 15 Mikrosatellitenloci wurden verwendet: A008, A014, A029, A079, A088, Ac011, Ac088, Ap085, Ap090, Ap224, Ap249, Ap273, At005, At163 und At188 (Arnaud Estoup, Garnery, et al., 1995; Solignac et al., 2003).
In den 133 Proben (eine Arbeiterin pro Bienenvolk) wurden an den 15 Loci insgesamt 292 verschiedene Allele beobachtet. Alle Loci waren polymorph, wobei die Anzahl der Allele pro Locus von 10 (in Locus At163) bis 31 (in Locus A029) reichte. In allen Loci war die Heterozygotie hoch. Die erwartete Heterozygotie, als bessere Entsprechung von Gendiversität, lag bei 0.861 ± S.E. 0.017 (Tabelle 3.4). Der Gesamt-Fst (der Inzuchtkoeffizient innerhalb von Subpopulationen relativ zur Gesamtpopulation), als Maß für die genetische Differenzierung von Populationen, war dabei sehr niedrig (Table 3.5): 0.007 ± S.E. 0.001 (0.001 - 0.014).
Schlussfolgerungen
Auf der Basis der morphometrischen Ergebnisse dieser Studie sind die Honigbienen von Westafrika als eine Einheit mit erheblicher Variation zu betrachten, die allerdings nicht geographisch abgegrenzt ist. Infolgedessen unterstützen die Ergebnisse, im Widerspruch zur bisher publizierten Literatur, die Aufteilung der Honigbienen dieser Region in die Unterarten A. m. adansonii und A. m. jemenitica nicht. Eine endgültige Beurteilung des taxonomischen Status der Honigbienen Westafrikas erfordert daher weitere Studien auf der Basis von morphometrischen, molekularen und physiologischen Daten.
In dieser Studie wird zum ersten Mal eine Ökokline für die Honigbienen der Region beschrieben, in deren Zentrum der Tschadsee liegt. Innerhalb dieser Kline nehmen die Bienen, vom Tschadsee ausgehend, in die umgebenden, höher liegenden Gebiete an Größe zu, wie es nach Bergmanns Regel zu erwarten ist.
Die Analyse der mitochondrialen DNA ergab eine geringe genetische Diversität der Honigbienen von Westafrika. Dabei ist das Fehlen „exotischer“ Haplotypen ein bedeutender Hinweis darauf, dass die Bienen der Region noch unverfälscht sind. Das heißt es wurden keine Königinnen oder Völker aus anderen Regionen eingeführt, oder diese konnten sich nicht etablieren. In der Analyse der CytochromB Sequenz bestätigt die Eingruppierung der Bienenproben aus dieser Studie mit den Referenzbienen von A. m. scutellata ihre Zugehörigkeit zu den Bienen des Afrika südlich der Sahara.
Im Gegensatz dazu ergab die Analyse der Mikrosatelliten eine extrem hohe Gendiversität und eine sehr niedrige genetische Differenzierung innerhalb der Bienen dieser Region. Obwohl, wie weiter oben angemerkt, dieses Ergebnis mit den bisher publizierten Studien über die Bienen Afrikas übereinstimmt, erscheint es möglich, dass die geringe genetische Differenzierung auf den extremen Polymorphismus der Mikrosatellitenmarker in afrikanischen Bienen zurückgeht. Dieser ist hier um ein Vielfaches stärker ausgeprägt als bei europäischen Bienen, für die diese Marker ursprünglich entwickelt wurden. Auf ähnliche Weise mag die COI-COII Zwischengenregion ein ungeeigneter Marker zum Studium der Bienen Afrikas sein, aber aus dem entgegengesetzten Grund – Mangel an Variabilität. Es erscheint daher angebracht, nach besser geeigneten molekularen Markern für die Honigbienen Afrikas zu suchen.
Die große morphometrische und mitochondriale Ähnlichkeit zwischen den Honigbienen in der Nähe des Tschadsees mit denen des feuchten Regenwaldes im Süden Nigerias (wie aus der ähnlichen Körpergröße und dem Vorherrschen des Haplotyps A1, sowie des Zusammenhangs dieses Haplotyps mit humiden Bedingungen hervorgeht) könnte eine Erklärung in der Geschichte des Tschadsees finden. Obwohl die Gegend um den Tschadsee heute ein semiarides Klima hat, können in der Vergangenheit durchaus humide Bedingungen geherrscht haben; die Verbreitung der Bienen wäre demnach als Reflektion der Vergangenheit zu betrachten. In paläontologischen Studien wird der heutige Tschadsee als winziges Überbleibsel eines prähistorischen Riesensees betrachtet, der als Mega-Tschad bezeichnet wird (Abb. 4.1). Auf dem Höhepunkt seiner Ausdehnung im frühen Holozän (vor 8000 – 10000 Jahren) umfasste der Mega-Tschad eine Fläche von mindestens 361,000 km², das heißt er war größer als jeder heutige See (Anonymous, 2015; Ghienne et al, 2002; Schuster et al, 2005; Drake und Bristow 2006). Nach Drake et al (2011) war die Trockenheit damals durch diesen See und zwei weitere Mega-Seen (Chotts Megalake and Lake Megafezzan) nicht nur in ihrer direkten Umgebung, sondern in der gesamten Zentralsahara vollständig verschwunden.
Abstract
In order to investigate the diversity of the western honeybee, Apis mellifera L., in West and Central Africa, a total of 204 colonies were sampled from 44 localities in four countries – Nigeria, Niger, Cameroon and Chad. 86 of these colonies, from 23 localities, were subjected to full morphometric analysis. In a principal component analysis (PCA) of the morphometric data, the colonies formed a single cluster. It also revealed that overall size of the body was the most important source of variation between the colonies. A hierarchical structure analysis, followed by a stepwise discriminant analysis, classified the colonies into three distinct morphoclusters; however, these clusters were not geographically demarcated. In another PCA carried out with the samples under investigation and reference samples of A. m. adansonii, A. m. jemenitica and A. m. scutellata, the colonies under investigation again formed one cluster which lying over and extended beyond the clusters of the reference subspecies. This is suggestive of a wider variation in size in the bees under investigation. In a stepwise DA, 94.2% of cross-validated grouped cases were correctly classified and the distances between group centroids were highly significant (p < 0.0005) according to F-statistic. 61 and 22 of the 83 colonies under investigation were assigned to A. m. jemenitica and A. m. adansonii, respectively. Mitochondrial DNA analysis was carried out on 148 colonies from 39 localities. Four mitochondrial haplotypes, previously reported from Africa and belonging to the African mitochondrial lineage, A, were detected: A1 (n = 62), A4 (n = 70), A4' (n = 15) and A14 (n = 1). The overall haplotype diversity was low (h = 0.478 ± S. E. 0.057). A chi-square test for association was conducted between haplotypes and type of vegetation, latitude, longitude, altitude, temperature and rainfall, severally. There was a statistically significant association between haplotype and each of the six variables and the association was strong with latitude, moderate with vegetation and rainfall and weak with the remaining variables. The neighbour-joining, maximum likelihood and maximum parsimony trees, obtained from sequence variation of the cytochrome b gene of mitochondrial DNA, showed that the samples, from the current study, unambiguously clustered with the reference sequences of A. m. scutellata from Kenya, but without showing further subdivision within this sub-Saharan cluster. 133 workers (one per colony) collected from 38 localities were subjected to microsatellite analysis. A total of 292 different alleles were recorded for the 15 microsatellite loci used. All microsatellite loci were polymorphic and the number of different alleles per locus ranged between 10, in locus At163, and 31, in locus A029. Heterozygosity (or gene diversity) was high in all loci. The unbiased expected heterozygosity, which is a better expression of gene diversity, was 0.861 ± S.E. 0.017. The overall FST value, which is a good estimate of genetic differentiation of populations, was very low: 0.007 ± S.E. 0.001 (0.001 - 0.014). AMOVA and Bayesian assignment showed no differentiation of the investigated populations. Based on morphometric analysis, the results of this study present the honeybees of western Africa as a single entity with an internal variation which lacks a geographical demarcation. Consequently the results do not support the splitting of the honeybees of the region into the two subspecies, A. m. adansonii and A. m. jemenitica, as reported in the literature. More morphometric, molecular, physiological and behavioural studies are required to confirm the taxonomic status of the honeybees of the region. Meanwhile, the use of A. m. adansonii, as the sole sub-specific name for the honeybees of West and Central Africa, is recommended.
Abbreviations
Abbildung in dieser Leseprobe nicht enthalten
List of Figures
Figure 1.1 Distribution of the subspecies of A. mellifera in Africa
Figure 1.2 Map of the mtDNA genome of Apis mellifera
Figure 2.1 The area of study
Figure 2.2 Mean annual temperature (°C) of the area of study
Figure 2.3 Mean annual rainfall (mm) of the area of study
Figure 2.4 Altitude (m above sea level) of the area of study
Figure 2.5 Localities in western Africa from which samples of honeybee were collected
Figure 2.6 Sources of honeybees
Figure 2.7 Honeybee parts prepared for morphometric measurements
Figure 2.8 Measurement of hair length on tergite 5 (h) and tomentum on tergite 4 (a, b)
Figure 2.9 Length of femur (Fe), tibia (Ti) and metatarsus (ML); MT width of metatarsus
Figure 2.10 Longitudinal diameter of tergite 3 (T3) and 4 (T4)
Figure 2.11 Measurements of sternite 3: longitudinal (S3), wax plate, longitudinal (WL) and transversal (WT) and distance between wax plates (WD)
Figure 2.12 Sternite 6, longitudinal (L6) and transversal (T6)
Figure 2.13 Fore-wing, length (FL) and width (FB); distances a and b of cubital vein
Figure 2.14 Measurement of 11 angles (A4 – O26)
Figure 2.15 Classes of pigmentation of tergites 2 – 4
Figure 2.16 Pigmentation of scutellum (Sc) and plates (K, B)
Figure 3.1 A-C. PCA plots using the colony means of 14 morphological characters of workers
Figure 3.2: Classification of 23 populations of A. mellifera
Figure 3.3 Distribution of members of three morphoclusters of A. mellifera
Figure 3.4 Classification of 3 populations of A. mellifera
Figure 3.5 Distribution of members of three morphoclusters of A. mellifera
Figure 3.6 Distribution of A. mellifera in the area of study
Figure 3.7 Variation of size of A. mellifera in western Africa
Figure 3.8 Plots of principal components extracted from a PCA
Figure 3.9 Prediction of group membership of 83 colonies of A. mellifera
Figure 3.10 Distribution of subspecies of A. mellifera
Figure 3.11 Distribution of mitochondrial DNA haplotypes of A. mellifera
Figure 3.12 Association between mitochondrial DNA haplotypes of A. mellifera and environmental variables
Figure 3.13. Molecular Phylogenetic analysis of cytochrome b sequence data of Apis mellifera, by the Neighbor-Joining method
Figure 3.14. Molecular Phylogenetic analysis of cytochrome b sequence data of Apis mellifera, by Maximum Likelihood method.
Figure 3.15 . Molecular Phylogenetic analysis of cytochrome b sequence data of Apis mellifera, by the Maximum Parsimony method.
Figure 3.16 Allelic patterns, based on 15 microsatellite loci of populations of A. mellifera
Figure 3.17 Results of Bayesian assignment test
Figure 4.1 Shuttle Radar Topography Mission (1 km resolution) Digital Elevation Model showing Saharan palaeolakes over 500 km2.
List of Tables
Table 2.1 Localities in western Africa from which honeybees where collected for this study.
Table 2.2 Details of the honeybee samples from which a complete set of 35 morphometric characters were measured.
Table 2.3 Details of the honeybee samples from which only characters of the fore-wing were measured
Table 2.4 List of characters measured for morphometry
Table 2.5 Details of the honeybee samples on which analysis of mitochondrial DNA was carried out
Table 2.6 Details of the honeybee samples on which analysis of microsatellite polymorphism was carried out
Table 3.1 Means and standard deviations of morphometric characters of 10 workers from N colonies of A. mellifera in West and Central Africa.
Table 3.2 Counts and (frequencies) of haplotypes of mitochondrial DNA of A. mellifera in four types of vegetation in western Africa.
Table 3.3: Allelic patterns (mean/locus) of two populations of A. mellifera in western Africa based on 15 microsatellite loci.
Table 3.4: Heterozygosity and related statistics of 15 microsatellite loci in two populations of A. mellifera in western Africa.
Table 3.5 Wright’s F-Statistics for 15 microsatellite loci of A. mellifera in western Africa
Table 4.1: Comparison of values (mean ± s.d.) of some morphometric characters of subspecies of A. mellifera from sub-Saharan Africa from various sources.
1 Introduction
1.1 Honeybees and Man
Honeybees have fascinated man since time immemorial and, indeed, this fascination is not without a reason. They produce a number of useful products in addition to presenting a way of life that attracts man’s attention and arouses his curiosity. These characteristics make the honeybee one of the most studied animals. Rock paintings, dating back to 13,500 B.C., at Altamira, Spain, suggest honey hunting while earliest records of beekeeping date back to 2,400 B.C in Egypt, where rock paintings, on the walls of the sun temple of King Nyuserre Ini, depict beekeepers using smoke to harvest honey from hives (Crane, 1999). They are of considerable economic importance, producing products of commercial value, such as honey and wax, and pollinating crops and wild plants. For example, Calderone (2012) estimated the value of honeybees, through pollination of crops, to be US$11.68 billion in the United States of America, for the year 2009. Similarly, the contribution of honeybees to the economy of the United Kingdom, through pollination, was estimated at £191.80 million (Anonymous, 2009). The combined production of honey by the top 20 producer-countries for 2011 was estimated at 1.26 million metric tonnes valued at US$3.16 billion (Anonymous, n. d.). In addition, products of honeybees, notably honey, royal jelly, propolis and bee venom, contribute to our well-being through their nutritional and therapeutic properties. For completeness, it is necessary to mention that honeybees also affect us negatively. For example, honeybee stings can cause a wide range of problems, to humans and livestock, due to varying degrees of reaction which range from small local reactions to anaphylaxis and even death (Brown, 2013). I have personally witnessed four instances of mass-stinging by honeybees in my home country, Nigeria. In two of these cases, two people were stung to death and in the other two livestock were killed: A donkey and a few turkeys. Other negative impacts of honeybees on the environment (where they are introduced) include competition with native fauna for food and shelter and the reduction of the production of seed by native plants due to their out-competing native pollinators or their inefficient pollination (Badano & Vergara, 2011; Krend & Murphy, 2003; Paton, 1996).
1.2 The genus Apis
Honeybees (genus Apis) belong to the family Apidae (social bees) and super family Apoidea (all bees) in the insect order Hymenoptera (bees, wasps, ants, etc.). Although there is no agreement, among taxonomists, on the number of honeybee species, the following nine are widely recognised (Sarah E. Radloff, Hepburn, & Engel, 2011):
i. A. andreniformis F. Smith
ii. A. cerana Fabricius
iii. A. dorsata Fabricius
iv. A. florea Fabricius
v. A. koschevnikovi Enderlein
vi. A. laboriosa F. Smith
vii. A. mellifera L.
viii. A. nigrocincta F. Smith
ix. A. nuluensis Tingek et al
While A. mellifera is native to western Asia, Africa and Europe, all the other species are restricted to Asia. Before interference by humans, A. mellifera was allopatric with the other species of Apis in Asia (F. Ruttner, 1988). Now it is sympatric with A. florea in Oman, Jordan and Sudan, following introduction of A. mellifera in Oman (Dutton, Ruttner, Berlkeley, & Manley, 1981) and A. florea in the other countries (Haddad, Fuchs, Hepburn, & Radloff, 2009; Lord & Nagi, 1987; Mogga & Ruttner, 1988). However, A. mellifera has been introduced, by man, to all other continents, except Antarctica (Meixner et al., 2013). A fossilised honeybee (Apis nearctica Engel et al) has been discovered recently in North America (Engel, 1999).
1.3 The species A. mellifera
The geographical origin of A. mellifera is not certain. Whereas some authors (Garnery, Solignac, Celebrano, & Cornuet, 1993; Han, Wallberg, & Webster, 2012; F. Ruttner, 1988; F. Ruttner et al., 1978; Wallberg et al., 2014) favour an Asian origin, others (Anonymous, 2006; Kotthoff, Wappler, Engel, & Ali, 2013; Whitfield et al., 2006; Wilson, 1971) favour an African one. Based on morphometric analysis, F. Ruttner (1988) classified A. mellifera into 24 geographic races (subspecies), ten of which are found in Africa. West and Central Africa is occupied by A. m. adansonii Latreille alone, except in Chad where A. m. jemenitica Ruttner takes its place. In a recent review of the intra-specific classification of A. mellifera, Engel (1999) added four subspecies and upheld the synonym, A. m. remipes Gerstäcker as the valid name for Ruttner's A. m. armeniaca. Thereafter, two new subspecies have been described, namely, A. m. pomonella (Walter S. Sheppard & Meixner, 2003) and A. m. simensis (Marina D. Meixner, Leta, Koeniger, & Fuchs, 2011). Details of these 30 subspecies of A. mellifera are given in Appendix I.
F. Ruttner (1988) further grouped the races into four phylogenetic lineages:
- M: Western Mediterranean and northwestern European races (A. m. intermissa, A. m. sahariensis, A. m. mellifera and A. m. iberiensis)
- C: Eastern Mediterranean and southeastern European races (A. m. siciliana , A. m. ligustica, A. m. carnica, A. m. macedonica and A. m. cecropia)
- A: Tropical African races (A. m. lamarckii, A. m. jemenitica, A. m. adansonii, A. m. scutellata, A. m. monticola, A. m. litorea, A. m. capensis and A. m. unicolor)
- O: Oriental and Mediterranean races (A. m. caucasica, A. m. anatoliaca, A. m. syriaca, A. m. meda, A. m. adami, A. m. cypria and A. m. remipes)
Subspecies described later, A. m. ruttneri (W. S. Sheppard, Arias, Grech, & Meixner, 1997), A.m. pomonella (Walter S. Sheppard & Meixner, 2003) and A. m. simensis (Marina D. Meixner et al., 2011) were assigned to M, O and A lineages, respectively (Marina D. Meixner et al., 2013).
In addition to the morphological lineages mentioned above, three major lineages, based on mitochondrial DNA, have been identified. These are M, C and A with the latter containing five sub-lineages, AI, AII, AIII, Y and Z (Alburaki, Moulin, Legout, Alburaki, & Garnery, 2011; Franck et al., 2001; Marina D. Meixner et al., 2013). However, these two kinds of lineage do not always concord, as their names may suggest. For example, whereas all the subspecies belonging to the A morphological lineage also belong to the A mitochondrial lineage, some members of the M morphological lineage (A. m. intermissa, A. m. sahariensis,) belong to the A mitochondrial lineage. It should, however, be noted that these mitochondrial lineages also exhibit a geographical pattern: honeybees of Western Europe are characterised by the M lineage, those of northern Mediterranean by the C lineage and those of Africa by the A lineage. Sub-lineage Y has so far been reported from Ethiopia only and sub-lineage Z (formally referred to as O lineage) from Egypt and western Asia (Alburaki et al., 2011; Franck et al., 2001; Franck, Garnery, Solignac, & Cornuet, 2000; Garnery, Cornuet, & Solignac, 1992; Kandemir, Meixner, Ozkan, & Sheppard, 2006; Marina D. Meixner et al., 2013).
1.4 A. mellifera in Africa
With an estimated 310 million colonies of A. mellifera, Africa harbours the largest diversity of the species and is the only area in the world where a naturally-occurring population exists on a large scale (Vincent Dietemann, Pirk, & Crewe, 2009). 11 out of the 30 recognised subspecies of A. mellifera are found in Africa. These are: A. m. adansonii Latreille , A. m. capensis Eschscholtz, A. m. intermissa Maa, A. m. jemenitica Ruttner, A. m. lamarckii Cockerell, A. m. litorea Smith, A. m. monticola Smith, A. m. sahariensis Baldensperger, A. m. scutellata Lepeletier, A. m. simensis Meixner et al and A. m. unicolor Latreille. The distribution of these subspecies (excluding A. m. simensis) is shown in Figure 1.1.
Abbildung in dieser Leseprobe nicht enthalten
Figure 1.1 Distribution of the subspecies of A. mellifera in Africa according to H. R. Hepburn and Radloff (1998)
1.5 A. mellifera in Western Africa
The history of the classification of the honeybees of western Africa started in 1804 when P. A. Latreille named a bee, collected from Senegal by Michael Adanson, Apis adansonii. However, he based his description on a single, poorly preserved specimen of a worker bee. Though Latreille's A. adansonii was dropped in a revision by Lepeletier in 1836, it resurfaced later, as A. m. adansonii, to refer to any yellow bees collected from sub-Saharan Africa. However, following the progress made in morphometric studies, especially in eastern and southern Africa, F. Ruttner (1975, 1982, 1988) suggested that the name A. m. adansonii be restricted to the bees of western Africa while the savanna honeybees of eastern and southern Africa should be referred to as A. m. scutellata. This area (the area regarded as western Africa in the context of the distribution of A. m. adansonii) appears too large and diverse to contain an undifferentiated population of bees. It contains plains, plateaus, forests, savannas, semi-deserts and deserts and extends from Mauritania and Senegal in the west, to Chad in the east, then south to Namibia, through Zambia. In fact many researchers have reported variations in colouration and behaviour among honeybees of Senegal, Togo, Ivory Coast, Benin, Angola, Congo, Gabon and the Democratic Republic of the Congo (H. R. Hepburn & Radloff, 1998; F. Ruttner, 1988). Although these observations were not based on morphometric studies, H. R. Hepburn and Radloff (1998) believe that “the observed variation remains a reality that must be resolved.”
F. Ruttner (1988) identified two important problems of the A. m. adansonii region and called for further studies and refined analysis: (1) A lack of evident correlation, in the case of A. m. adansonii, between phenotype and environment, as found in other African races. That is, although there is a phenetic north-south cline along the African west coast, no morphometric differentiation had yet been found, in spite of the huge geographic distance and important differences in humidity and altitude; (2) The sharp border between A. m. adansonii and A. m. jemenitica around the Niger/Chad border in the same ecological zone (Sahel) and without any physical barrier.
In a comprehensive and recent analysis of the honeybees of Africa, including the A. m. adansonii region, H. R. Hepburn and Radloff (1998) suggested the existence of a correlation between phenotype and environment, in the case of A. m. adansonii as well and the non-existence of the supposed border between this subspecies and A. m. jemenitica around the Chad/Niger border. According to their hypothesis, A. m. adansonii occupies the forest belt, along the west coast, from Senegal, in the west, to the Congo, in the east; while A. m. jemenitica occupies the Sahel from Mali, in the west, to Sudan, in the east. The savannas serve as a zone of hybridisation between these two races (Figure 1.1). However, they cautioned that: "in view of the enormity of the area considered as West Africa, the actual number of sampled localities is small and requires attention in future."
Indeed, with only 190 colonies morphometrically analysed (H. R. Hepburn & Radloff, 1998) and 114 and 30 colonies analysed for mtDNA and microsatellites, respectively (Franck et al., 2001), western Africa (Countries in West and Central Africa, from Mauritania and Senegal in the west, to Chad in the east, then south to Namibia, through Zambia) is evidently under-studied. Although a few morphometric studies have been carried out, recently, in Nigeria (Ajao, Oladimeji, Idowu, Babatunde, & Obembe, 2014; Oyerinde, Dike, Banwo, Bamaiyi, & Adamu, 2012; Yu et al., 2012), they have done little in improving the situation, due to their inadequacy in coverage or methodology.
1.6 Conservation of Local Populations of Honeybee
The natural range of the western honeybee, A. mellifera, is western Asia, Africa and Europe: From southern Scandinavia in the north to the Cape of Good Hope in the south, from Dakar in the west to the Urals, Mashhad and to the coast of Oman in the East. Geographical isolation and ecological adaptations resulted in the evolution of local populations showing considerable geographical variation, resulting in adaptation to local factors of climate, vegetation, pests and pathogens (Marina D. Meixner et al., 2013; F. Ruttner et al., 1978). These adaptations may be lost due to human activities in beekeeping that affect wild honeybees in different ways: competition for floral sources, introduction of exotic genes, pests, parasites and diseases (Pilar De la Rúa et al., 2009; Robin F. A. Moritz et al., 2005). Therefore it is necessary to protect the biological diversity of local populations of honeybees in their natural habitats. A few issues are discussed below to underscore the importance of this.
1.6.1 Breeding
Intensive beekeeping, characterised by selection for desirable traits such as gentleness and high yield, are contributing to the reduction of biodiversity of honeybees in Europe: The original geographic distribution pattern is being dissolved by mass importation of queens and colony movements and deliberate replacement of native subspecies by non-native ones (for instance, the replacement of A. m. mellifera in northern and central Europe with A. m. carnica or A. m. ligustica). These actions are leading to a loss of both genetic diversity and specific adaptations to local conditions (Bouga et al., 2011; Pilar De la Rúa et al., 2009; Marina D. Meixner et al., 2010; Marina D. Meixner et al., 2013) and this may cause unpredictable negative impacts. For example, it has been shown that colonies with reduced genetic diversity are less capable of controlling hive temperature (Jones, Myerscough, Graham, & Oldroyd, 2004) and more prone to develop diseases when challenged by parasites (Tarpy, 2003). This reduction in genetic diversity may also affect the capacity of honeybee populations to adapt to new threats, such as newly introduced parasites, like Varroa.
1.6.2 The Varroa Problem
The parasitic mite, Varroa destructor, is a natural parasite of the eastern honeybee, A. cerana, in Asia and it does not constitute a pest problem there. However, following the introduction of A. mellifera into Asia, the pest switched hosts to become the destructive pest of A. mellifera it now is (Robin F. A. Moritz et al., 2005). The mite feeds on both adult bees and brood by sucking their haemolymph (blood), thereby depriving them of vital nutrients while transmitting pathogens, such as viruses. Mild infestations do not cause any noticeable damage but heavy ones weaken colonies and ultimately lead to their collapse. V. destructor is regarded as the most economically important pest of A. mellifera and a major contributor to colony collapse disorder in Europe and North America (Anonymous, 2013; Chauzat et al., 2010; Vincent Dietemann et al., 2013; Guzmán-Novoa et al., 2010; Peter Neumann & Carreck, 2010; D. vanEngelsdorp, Hayes Jr., Underwood, & Pettis, 2008; Dennis vanEngelsdorp, Hayes Jr., Underwood, & Pettis, 2010).
On the other hand, the decline of the populations of the native A. cerana in China has been attributed to several factors including competition with A. mellifera and the transfer of the latter's pathogens and parasites to it (Li et al., 2012). The problem may be compounded by the fact that drones and queens of both species readily mate but do not produce viable offspring (Friedrich Ruttner & Maul, 1983), thus placing A. cerana at a disadvantage.
1.6.3 The Africanised honeybees
In order to increase the production of honey from the earlier introduced European honeybees, queens of the African honeybee, A. m. scutellata, were introduced to Brazil in 1956. However, some swarms, produced by some of these queens, escaped into the wild, thus starting a process that came to be known as Africanisation of the European bees which had already established feral colonies throughout most of South and Central America. By 1990 the so called Africanised bees had reached the southern parts of the United States of America. In spite of hybridisation with European bees for over half a century, the Africanised bees remain essentially African in terms of their behaviour, morphology and even genetic composition. In other words, the Neotropical honeybees of today are African bees with some traits of European bees. It appears that the combination of high colony mobility, intraspecific parasitism, and higher investment in colony replication underlies the invasion success of African A. mellifera subspecies. The downside of this invasion was the initial negative impact on the beekeeping industry which suffered a decline due to the well-pronounced defensive behaviour of the Africanised bees which earned them the name "killer bees". However, with time, beekeepers have learned to handle the bees and the industry has recuperated, at least in Brazil (Coulson et al., 2005; Francoy et al., 2009; H. Glenn Hall, Zettel-Nalen, & Ellis, 2014; Robin F. A. Moritz et al., 2005; Schneider, DeGrandi-Hoffman, & Smith, 2004).
1.6.4 The Capensis Calamity
Some workers of the Cape honeybee, A. m. capensis, produce female offspring through thelytokous parthenogenesis. This is a phenomenon in which unfertilised workers lay viable diploid eggs which usually hatch into other workers or, sometimes, queens. These workers exhibit a strange behaviour of parasitising colonies of other subspecies of A. mellifera. When such social parasites invade the nests of their host, they compete with the queen for egg-laying. The colony so parasitised starts to produce more parasitic workers, which hardly forage, at the expense of its own. This, in turn, leads to the decline and eventual collapse of the colony. The parasitic workers then invade other colonies. In 1990, beekeepers from the southern part of South Africa transported 400 colonies of A. m. capensis to the northern part, an area solely occupied by A. m. scutellata. This led to the infestation of colonies of the latter by parasitic workers of the former, an issue that caused an estimated annual loss of 100,000 host colonies in the following years, leading to the so-called "capensis calamity" for South African beekeeping enterprises (M. H. Allsopp, 1992; M. H. Allsopp & Crewe, 1993; Vincent Dietemann, Lubbe, & Crewe, 2006; H. R. Hepburn & Radloff, 1998).
1.6.5 The Small Hive Beetle
A scavenger and a minor pest of honeybees in its native range in sub-Saharan Africa, the small hive beetle, Aethina tumida Murray, became a serious pest after its introduction into other places. In its native range it only inflicts damage on weak colonies since the bees have developed strategies to combat it. First identified in southeastern United States of America (US) in 1996 and Australia in 2002, this pest has found the susceptible European subspecies of A. mellifera a suitable host. The beetle is now well-established and is causing a considerable damage to beekeeping through death of colonies and increased rate of absconding (Cuthbertson et al., 2013; Vincent Dietemann et al., 2009; Mogbel A. A. El-Niweiri, El-Sarrag, & Neumann, 2008; H. R. Hepburn & Radloff, 1998; Hood, 2000; Peter Neumann & Ellis, 2008). The pest was first detected in Europe, in Portugal, in September 2004, on queens imported from the US, but timely introduction of control measures prevented it from spreading and establishing (Murilhas, 2005). Recently (September 2014), the pest has been found again in Europe; this time in Italy. It was reported in three nuclei in an apiary near an international port (Gioia Tauro) in the southern part of the country. The discovery of this pest in this area is of particular concern being a major source of queens, not only for Italy, but for the European Union and other parts of the world. Control measures have since been put in place. These include demarcation of protection and surveillance zones of 20 km and 100 km radius, respectively; destruction of infected apiaries; treatment of the soil in infected apiaries with insecticides; restriction of movement of bees, their products and beekeeping equipment; and awareness drive targeting all stakeholders in the beekeeping industry (Mutinelli et al., 2014; Palmeri, Scirtò, Malacrinò, Laudani, & Campolo, 2014).
1.6.6 The Honeybees of Africa
Africa is the only area in the world where a natural population of A. mellifera exists on a large scale. With an estimated 310 million colonies in the wild and another 11 to 18 million managed colonies, themselves captured from the wild, it may be assumed that African honeybees are not threatened and, therefore, conserving them is not necessary (Vincent Dietemann et al., 2009). However, it is arguable that African honeybees need a preventive conservation for at least three reasons. Firstly, to preserve their genetic diversity which is responsible for their fitness; secondly, wild populations serve as a reservoir of genes from which managed colonies can be improved; and thirdly, to avoid the errors done elsewhere, for example in Europe, where intensive beekeeping has already compromised the diversity of local honeybee populations to near extinction in many areas (Pilar De la Rúa et al., 2009; Marina D. Meixner et al., 2014). Current threats to the conservation of honeybees in Africa include loss of habitat through deforestation; hunting for honey involving killing of wild colonies; poor enforcement of legislations aimed at protecting honeybees, in particular, and the ecosystem in general; and introduction of exotic honeybees and the parasitic mite, V. destructor (Vincent Dietemann et al., 2009; Eardley, Gikungu, & Schwarz, 2009). Several attempts have been made at introducing European subspecies to "improve" local stocks in Africa, including the recent introduction of A. m. ligustica to Nigeria from China (Yu et al., 2012). So far these efforts have failed and European bees have not become established on the continent (Vincent Dietemann et al., 2009; H. R. Hepburn & Radloff, 1998). A. florea, the dwarf Asian honeybee, was accidentally introduced into Sudan in 1985, probably via an air cargo from Pakistan, and is spreading in the country, apparently without constituting a threat to local honeybees (El Shafie, Mogga, & Basedow, 2002; Lord & Nagi, 1987; Mogga & Ruttner, 1988). However, we do not know what will happen in the future. V. destructor, on the other hand, as destructive as it is to European subspecies of A. mellifera in Europe and North America, appears to be of little harm to African honeybees. Introduced in the 1990`s (Mike H Allsopp, Govan, & Davison, 1997; Anonymous, 1992; Ellis & Munn, 2005; Majeed, 2000; Matheson, 1993), this pest is now found in all regions of Africa but it does not constitute a pest problem: The bees have adapted to living with it (Muli et al., 2014; Mumbi, Mwakatobe, Mpinga, Richard, & Machumu, 2014; Strauss, Pirk, Crewe, Human, & Dietemann, 2014). Similarly most diseases of honeybees are present in Africa without causing a serious harm, and so far, no epidemics have been reported (Vincent Dietemann et al., 2009). The factors that contribute to this, probably, include absence of large-scale beekeeping and breeding; high genetic diversity in the large wild stock population on which beekeeping relies; less migratory beekeeping; and less harvest of hive products. In addition to a higher genetic diversity, relative to European bees, African bees exhibit the following characters that could further contribute to their survival (H. R. Hepburn & Radloff, 1998; Robin F. A. Moritz et al., 2005): more swarms per colony; more migratory swarms; more drones; smaller colonies; a shorter generation time; a shorter queen development time; and no excessive honey hoarding necessary for overwintering.
Thus, from the foregoing, it can be strongly argued that, for a sustainable beekeeping, it is necessary to protect the genetic diversity of honeybees in their natural range. In order to achieve this, the bees must first be characterised so that any intrusion of foreign genotypes can easily be detected (Franck et al., 2001; Marina D. Meixner et al., 2013).
1.7 Methods of Characterising Honeybee Populations
1.7.1 Morphometry
The earlier methods of classifying honeybees were based mainly on qualitative descriptions of morphology. Though fairly adequate in discriminating higher taxa, these methods proved insufficient at discriminating subspecies of honeybees. This necessitated the evolution of morphometrics in the last century. Instead of mere description of characters of individual honeybees, the new method uses numeric data resulting from exact measurements from which means of colony characters are obtained for statistical analyses (F. Ruttner, 1988). A set of morphological characters of body size, colour, pilosity and shape is measured. Whereas classical morphometrics emphasizes the first three, geometric morphometric addresses the last. Although there is not yet a universally accepted standard suit of characters for use in the two forms of morphometry, the 36 characters compiled and used by Ruttner in his monogram (F. Ruttner, 1988) or a subset of them, appear to be most favoured for classical morphometry (Table 2.2). In geometric morphometrics, on the other hand, 18 landmarks taken from the fore-wing are most commonly used (Tofilski, 2004, 2008). In either case multivariate statistical analyses are used in analysing the acquired data in order to study the variation within and between subspecies of A. mellifera (Marina D. Meixner et al., 2013). With only one morphometric study in Cameroon (H. R. Hepburn & Radloff, 1997), the honeybees of western Africa are poorly studied.
1.7.2 Variation of Mitochondrial DNA
The mitochondrial DNA of A. mellifera, which is maternally inherited, contains a non-coding intergenic region which is highly variable (because it is free to acquire mutations) and has been used extensively as a genetic marker for studies of geographic variation and introgression of subspecies (J. M. Cornuet et al., 1991; R. H. Crozier & Crozier, 1993). This region, referred to as COI-COII or tRNAleu-cox2, lies between the sub-units I and II of the cytochrome c oxidase gene (Figure 1.2). It consists of the gene tRNA for leucine (71 base pairs) and a non-coding sequence which is 220 - 860 base pairs long. This sequence, in turn, is made of a region named Q and an optional region called P. When P is present, Q may be repeated up to four times. Thus, five basic combinations are possible: Q, PQ, PQQ, PQQQ and PQQQQ. However, due to the presence of at least three additional variants of P (Po, P1 and P2), the number of combinations may be large. Q is 192-196 base pairs and Po (regarded as the ancestral sequence) is 67-69 base pairs long. P is 15 base pairs shorter than Po, due to a deletion. P1 differs from Po in having a deletion of 17 base pairs at the 3' end. P2, on the other hand, has its own deletion near the 5' end. To date, the following combinations have been described for the various mitochondrial lineages: Q for C; PQ to PQQQQ for M; PoQ to PoQQQQ and P1Q to P1QQQ for A; PoQ to PoQQQ for Z (Formally O); and P2QQ for Y. Y and Z are now regarded as sub-lineages of A (Alburaki et al., 2011; Marina D. Meixner et al., 2013). In addition, this region shows restriction polymorphisms with Dra I, making it possible to characterize different haplotypes based on the number and size of fragments so produced (T. Collet, Ferreira, Arias, Soares, & Del Lama, 2006; J. M. Cornuet et al., 1991; R. H. Crozier & Crozier, 1993; Pilar De la Rúa et al., 2009; Franck et al., 2001; Franck, Garnery, Solignac, & Cornuet, 1988; Garnery et al., 1992).
African bees are poorly studied, with regard to mitochondrial variation, compared to their European counterparts and the situation is more serious in western Africa with only three localities (Nimba (Guinea), Maradi (Niger) and Yaoundé (Cameroon)) sampled (Franck et al., 2001).
Abbildung in dieser Leseprobe nicht enthalten
Figure 1.2 Map of the mtDNA genome of Apis mellifera: (a) Complete genome after R. H. Crozier and Crozier (1993); (b) The subunits I (COI) and II (COII) of cytochrome c oxidase gene after R. H. Crozier, Crozier, and Mackinlay (1989). [W = tRNA for tryptophan; L = tRNA for leucine; D = tRNA for aspartate; K = tRNA for lysine; and ? = non-coding inter-genic region]
1.7.3 Microsatellite Polymorphism of Nuclear DNA
Inherited from both parents, DNA microsatellites are a type of variable number of tandem repeats (VNTR), also known as short tandem repeats (STR) and simple sequence repeats (SSR). They are one to six base pairs long and may be repeated many times (up to 100). They can be found in any part of the genome, though they are more frequent in the non-coding part. Due to their high rate of mutation, microsatellites present a lot of variation. This high variation, together with their abundance and genome-wide distribution, make microsatellites useful genetic markers in the study of genetic identity and diversity, population structure, introgression, gene flow, rates of admixture and related disciplines (J.-M. Cornuet et al., 1999; Marina D. Meixner et al., 2013; Nedić et al., 2014; Rowe et al., 1997; Sainudiin et al., 2004).
Developed for honeybees in the 1990s (Arnaud Estoup, Garnery, Solignac, & Cornuet, 1995; A. Estoup, Solignac, Harry, & Cornuet, 1993; Arnaud Estoup, Tailliez, Cornuet, & Solignac, 1995; Solignac et al., 2003), the use of microsatellite markers has become routine in the study of genetic diversity of honeybees (Franck, Garnery, Celebrano, Solignac, & Cornuet, 2000; Franck et al., 2001; Garnery et al., 1998) and they are currently considered the most useful genetic markers in population genetics (Solignac et al., 2003). However, their application to African bees has been very limited. So far some study has been done on the genetic diversity of A. m. intermissa and A. m. sahariensis from three localities in Morocco; A. m. scutellata from two localities in South Africa; A. m. capensis from Cape Town, South Africa; A. m. monticola from Chelinda, Malawi; A. m. adansonii from Nimba, Guinea, using six loci: A113, A43, A28, A24, A88 and B124 (Arnaud Estoup, Garnery, et al., 1995; Franck et al., 2001; Franck et al., 1988). Shi (2001) studied effective mating frequencies of Kenyan honeybees using 29 colonies sampled from five localities using two loci: A1 and A76.
1.8 Research questions
From the foregoing discussion, the following questions arise:
i. To what extent do the honeybees of western Africa differ in respect of their morphology?
ii. What is the genetic structure of the population of honeybees of western Africa?
iii. What is the level of genetic differentiation among the honeybees in this area?
iv. Is the differentiation of honeybees in the area related to ecological variations?
1.9 Aim and Objectives of the Study
The aim of this study was to provide answers to the questions, raised above, by analysing honeybee samples from a large and diverse, but manageable, portion of western Africa attributed to A. m. adansonii. It was the first single study to cover so large an area in the region; the most comprehensive in terms of the number and distribution of localities and the number of colonies sampled; and the first to investigate the genotype and phenotype, simultaneously, using molecular (analyses of mitochondrial DNA and microsatellite polymorphism) and morphological (classical morphometric analysis) tools. The specific objectives of the study were:
i. To determine the subspecies of A. mellifera in western Africa.
ii. To genetically characterise the populations of A. mellifera in western Africa.
iii. To determine the degree of genetic variability among the bees.
iv. To suggest reasons for the observed variability or the lack of it.
1.10 Research Hypothesis
The null hypothesis for this study was: "There is no variation among the honeybees of western Africa.”
2 Materials and Methods
2.1 Description of the Area of Study
The area of study (Figure 2.1), which is about two million square kilometres (that is about 5½ times the area of Germany), lies approximately from 1˚ to 22˚ E and 4˚ to 15˚ N and covers the whole of Nigeria, southern part of Niger (up to 15°N), northern part of Cameroon (from 9°N) and southern part of Chad (up to 15°N). A summary of the important physical features of the area, based on H. R. Hepburn and Radloff (1998) is given below:
The area consists of four climatic zones, namely, equatorial, wet tropical, dry tropical and sahelian. These correspond, approximately, to four zones of vegetation (Figure 2.1): Tropical rainforest, Guinea savanna, Sudan savanna and Sahel. Sahel is the transitional zone between savanna and desert.
The forest zone is characterized by a very dense vegetation, very rich in species diversity; a very heavy rainfall and a long rainy season (more than half of the year); and a mean annual temperature of about 25˚C which varies a little (Figure 2.2). Plants flower throughout the year.
The savannas consist of a mixture of grasslands and woody vegetation; lighter rainfall with a short rainy season (less than half of the year); and a high variation of mean annual temperature (10 to 15˚C). The density of vegetation, species richness, amount of rainfall and length of the rainy season decrease with the increase in latitude (Figure 2.3). Annuals flower at the end of the rainy season while trees flower during the dry season.
The Sahel is characterized by scanty rainfall; very short rainy season; very sparse vegetation of short trees and grass; and frequent droughts.
Altitude varies from sea level, through 1200 metres on the Jos Plateau, to about 1800 metres on the Mambila Plateau (part of the Adamoua Massif), both in Nigeria (Figure 2.4).
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.1 The area of study (enclosed by the black thick line) which comprises southern parts of Niger and Chad, northern part of Cameroon and the whole of Nigeria. Inset: A map of Africa showing the location of the area of study. Information on vegetation was obtained from Hoyle (1958)
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.2 Mean annual temperature (°C) of the area of study in western Africa (enclosed by the black thick line) plotted using data in Table 2.1. Inset: A map of Africa showing the location of the area of study. Temperature varies a little (26 to 30°C) in the area of study.
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.3 Mean annual rainfall (mm) of the area of study in western Africa (enclosed by the black thick line) plotted using data in Table 2.1. Inset: A map of Africa showing the location of the area of study. Amount of rainfall decreases with the increase in latitude: with over 2000 mm in the south and less than 500 mm in the north.
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.4 Altitude (m above sea level) of the area of study in western Africa (enclosed by the black thick line) plotted using data in Table 2.1. Inset: A map of Africa showing the location of the area of study. Most of the area is below 500 m.
2.2 Sampling the Area of Study
44 localities were selected for sampling. Most of these were located along three north - south transects to capture the ecological variations in the area of study, the most important of which is vegetation which follows the distribution of rainfall. These transects were: (1) Lagos to Tahoua, about 980 km long; (2) Bonny to Maiadua, about 980 km long; and (3) Mundemba to Afunori, about 1200 km long. Localities lying outside these transects were intended to address special issues such as altitude and the supposed border between adansonii and jemenitica around Lake Chad (Figure 2.5).
Additional samples from four localities (Bamenda, Ngaoundéré, Touboro and Mundemba) in Cameroon, outside, but close to the study area, provided by colleagues, were also included in the study. The details of the sampled localities are given in Table 2.1.
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.5 Localities in western Africa from which samples of honeybee were collected for this study. The black thick line marks the borders of the area of study while the green lines indicate the major transects. Inset: A map of Africa showing the location of the area of study.
Table 2.1 Localities in western Africa from which honeybees were collected for this study.
Abbildung in dieser Leseprobe nicht enthalten
Table 2.1 (Continued)
Abbildung in dieser Leseprobe nicht enthalten
1 Source: (Anonymous, 2014a)
2 Source: Hoyle (1958)
2.3 Collection of Samples of Honeybee
2.3.1 Fieldtrips
In order to collect honeybees from the selected localities, 18 fieldtrips were made between 2005 and 2013. The trips were made in a private car or by public transport. They lasted 111 days and covered a distance of 29,180 km. The roads were generally good though in a few places they were extremely bad.
2.3.2 Sources of Honeybees
Prospective sources of bees (beekeepers or honey-hunters) were often identified in the field with the assistance of leaders of communities or relevant government officials. Then, depending on his preference, the beekeeper would collect the bees for the researcher or allow him make the collection. For security reasons honey-hunters always did the collection, although the investigator did accompany some of them. Bees were collected from wild nests or unmanaged traditional or top-bar hives populated by wild swarms (Figure 2.6).
2.3.3 Sampling of Honeybees
Whenever available, up to five colonies were selected at random in each locality. Then about thirty workers were collected by hand from the interior of the hive or nest (after excavation) and immersed in ethanol, immediately. 20 bees were immersed in 70% ethanol for morphometric analysis while 10 were preserved in 95% ethanol for analysis of DNA. In this way a total of 204 colonies were sampled from 44 localities (Figure 2.5 and Table 2.1). The collection of samples and the morphometric data generated therefrom are stored at the Institut für Bienenkunde in Oberursel (Polytechnische Gesellschaft), University of Frankfurt, Germany.
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.6 Sources of honeybees: (a) A natural nest in a tree; (b) A traditional straw hive; (c) An apiary of Kenya Top-bar hives.
2.4 Morphometry
2.4.1 Dissection of Honeybees
A set of 10 workers was dissected from each colony for taking morphometric measurements according to F. Ruttner (1988) and F. Ruttner et al. (1978). For every set of bees, the left fore-wings were removed and sandwiched between two microscope slides secured with Sellotape (Figure 2.7a). The left hind legs were then removed and embedded in Gum Arabic between two microscope slides (Figure 2.7b). After the removal of wings and legs, the bees were pinned on their sides in 70% ethanol in a Petri dish for measurement of the pigmentation of the scutellum and pilosity (Figure 2.7c). Thereafter, the third and sixth sternites were removed, rinsed in water, blotted with tissue paper and embedded in Gum Arabic between two microscope slides (Figure 2.7d). Finally, the second, third and fourth tergites were removed and laid in Gum Arabic on a glass rod and secured with Sellotape (Figure 2.7e).
2.4.2 Measurements
Morphometric measurements of 35 characters were taken from 86 colonies in 23 localities (Table 2.2) according to F. Ruttner (1988) and F. Ruttner et al. (1978). Due to the exigency of time, only characters of the fore-wing were measured in 66 additional colonies from 20 additional and four common (Sarh, Bauchi, Jos and Umuahia) localities (Table 2.3).
Measurements of hair and pigmentation were taken under a dissecting microscope, fitted with an eyepiece graticle, at a magnification of 40 x. Measurements of wings, legs and sternites were taken with the help of a CCD camera connected to a personal computer, using the measuring program Bee Morphometric, Version 1.02 (M.D. Meixner, 1994). The details of the variables measured are shown in Table 2.4 and Figures 2.8 to 2.16.
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.7 Honeybee parts prepared for morphometric measurements: (a) wings, (b) legs, (c) bees pinned in wax, (d) sternites and (e) tergites
Table 2.2 Details of the honeybee samples from which a complete set of 35 morphometric characters were measured.
Abbildung in dieser Leseprobe nicht enthalten
Table 2.3 Details of the honeybee samples from which only characters of the fore-wing were measured
Abbildung in dieser Leseprobe nicht enthalten
Table 2.4 List of characters measured for morphometry[1]
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.8 Measurement of hair length on tergite 5 (h) and tomentum on tergite 4 (a, b)[2]
Abbildung in dieser Leseprobe nicht enthalten
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.9 Length of femur (Fe), tibia (Ti) and metatarsus (ML); MT width of metatarsus
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.10 Longitudinal diameter of tergite 3 (T3) and 4 (T4)
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.11 Measurements of sternite 3: longitudinal (S3), wax plate, longitudinal (WL) and transversal (WT) and distance between wax plates (WD)
Figure 2.12 Sternite 6, longitudinal (L6) and transversal (T6)
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.13 Fore-wing, length (FL) and width (FB); distances a and b of cubital vein
Figure 2.14 Measurement of 11 angles (A4 – O26)
Abbildung in dieser Leseprobe nicht enthalten
Figure 2.15 Classes of pigmentation of tergites 2 – 4
Figure 2.16 Pigmentation of scutellum (Sc) and plates (K, B)
2.4.3 Multivariate Statistical Analysis of Morphometric Data
Statistical analyses were carried out in two steps. The first step involved the analysis of morphometric data of the honeybees collected for this study alone. This was necessary for an exploration of the variation between them while avoiding any bias that might be introduced by analysing them with samples from preconceived groups. The second step consisted of the analyses of the honeybees under investigation together with reference data for the purpose of comparison. All statistical analyses were carried out with IBM® SPSS® Statistics Version 20 with additional material from Burns and Burns (2008) and Anonymous (2014b).
2.4.3.1 Analysis of samples under investigation
In order to investigate the variation between the samples collected for this study alone, the morphometric data were subjected to a number of analyses. First, descriptive statistics, in form of means and standard deviations, were calculated for every colony for the purpose of exploring the data for any anomalies of measurement that might be present so that they could be corrected. This was followed by principal component analysis (PCA) and hierarchical cluster analysis in order to detect any clustering of samples. That is, grouping of the samples based on overall morphological similarity. Discriminant function analysis (DA) was then used to test the probabilities of the assignment of the samples to the identified groups. Correlation analysis was used to test the relationships between morphology and environmental factors, such as temperature and rainfall. Details of the analyses are as follows:
2.4.3.1.1 Descriptive Statistics and Analysis of Variance
First, the mean and standard deviation of each of the 35 morphometric characters were calculated for every colony. Then the means for colonies were used to calculate the means and standard deviations for localities. The data were subjected to a one-way analysis of variance (ANOVA) to compare different localities. Tukey HSD tests, on the means of the 35 characters, were used to detect significant differences between the localities.
2.4.3.1.2 Principal Component Analysis
A PCA, using colony means of 35 morphometric characters was run in order to detect any possible clusters. The suitability of PCA was assessed, prior to the analysis, using correlation coefficients of the variables, Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett's test of sphericity (Anonymous, 2014b; Burns & Burns, 2008). Any variable that did not have a correlation with at least one other variable where r ≥ 0.3 should be removed from the analysis. The KMO measure is used as an index of whether there are linear relationships between the variables. Its value can range from 0 to 1, with values above 0.6 suggested as a minimum requirement for sampling adequacy. Bartlett's test of sphericity tests the null hypothesis that there are no correlations between any of the variables and, therefore, the variables cannot be reduced to a smaller number of principal components. For a PCA to be feasible the null hypothesis must be rejected: The result of Bartlett’s test is used to take this decision.
2.4.3.1.3 Hierarchical Cluster Analysis
A hierarchical cluster analysis was carried out with the assumption that samples from each locality formed a distinct group. Thus, 23 groups were assumed and the analysis was carried out using the means of the 35 morphometric characters for the 23 localities in the study.
Thereafter, the analysis was repeated, using the means of the 35 morphometric characters for the four zones of vegetation (Rainforest, Guinea savanna, Sudan savanna and Sahel) in the area of study, based on the assumption that samples from each of the zones formed a distinct group. In either case, the final number of groups to be entered into discriminant analysis, for confirmation, was determined from the dendrogram generated by the analysis.
2.4.3.1.4 Discriminant Analysis
A stepwise DA was run in order to confirm the groups predicted by cluster analysis and to determine the discriminant characters. The suitability of DA was determined through log determinants and Box’s M test. In DA the basic assumption is that the variance-co-variance matrices are equivalent. For this assumption to hold the log determinants should be equal and the Box’s M test should not be significant. The latter tests the null hypothesis that the covariance matrices do not differ between groups formed by the dependent variable.
Wilk’s lambda was used to test the discriminatory power of the discriminant functions while the significance of the distance between group centroids was tested by F-statistic.
2.4.3.1.5 Correlation Analysis
A Pearson's product-moment correlation was run to assess the relationship between latitude, longitude, altitude, temperature, rainfall and the three principal components extracted by PCA (representing morphological characters loaded on them). Prior to analysis, the data were assessed for linearity, normal distribution and presence of outliers (Anonymous, 2014b; Burns & Burns, 2008). Interpretation of the magnitude of Pearson’s correlation coefficient, r, was based on Cohen (1988), that is: 0.1 < | r | < .3: small/weak correlation; 0.3 < | r | < .5: medium/moderate correlation; and | r | > .5: large/strong correlation.
2.4.3.2 Analysis of samples under investigation together with reference samples
In order to determine the subspecies to which the samples under investigation belonged, they were analysed with reference samples of subspecies presumed to exist in the area of study, using a PCA, in order to find out whether they would cluster with any of the reference samples. Finally a discriminant analysis was carried out to confirm the assignment of the colonies under investigation to any of the reference subspecies. The details of the analyses are given below:
2.4.3.2.1 Principal Component Analysis
A PCA was run, using colony means of 35 morphological characters of workers from 83 colonies of the honeybees under investigation and 69 reference colonies (21 each of A. m. jemenitica and A. m. scutellata and 27 of A. m. adansonii) from the morphometric data bank of the Institut für Bienenkunde, Oberursel, Germany. The suitability of PCA was assessed, prior to the analysis, using correlation coefficients of the variables, Kaiser-Meyer-Olkin (KMO) measures and Bartlett's test of sphericity (Anonymous, 2014b; Burns & Burns, 2008). Any variable that did not have a correlation with at least one other variable where r ≥ 0.3 should be removed from the analysis. The KMO measure is used as an index of whether there are linear relationships between the variables. Its value can range from 0 to 1, with values above 0.6 suggested as a minimum requirement for sampling adequacy. Bartlett's test of sphericity tests the null hypothesis that there are no correlations between any of the variables and, therefore, the variables cannot be reduced to a smaller number of principal components. For a PCA to be feasible the null hypothesis must be rejected: The result of Bartlett’s test is used to take this decision.
The following six characters were excluded from the analysis due to the large variance of their values with those of the reference samples collected from the same localities: Length of cover hair on tergite 5, width of tomentum, longitudinal diameter of wax mirror of sternite 3, cubital vein (distances a and b) and angle of wing venation G18. The three subspecies were chosen because they were the only ones, among the samples classified by F. Ruttner (1988) from sub-Saharan Africa, with a sufficient number of samples.
2.4.3.2.2 Discriminant analysis
Finally, a stepwise DA was carried out in order to predict the membership of the 83 colonies of A. mellifera, under investigation, among the three reference subspecies (A. m. jemenitica, A. m. scutellata and A. m. adansonii) and to identify the discriminant characters. The suitability of DA was determined through log determinants and Box’s M test. In DA the basic assumption is that the variance-co-variance matrices are equivalent. For this assumption to hold the log determinants should be equal and the Box’s M test should not be significant. The latter tests the null hypothesis that the covariance matrices do not differ between groups formed by the dependent variable. Wilk’s lambda was used to test the discriminatory power of the discriminant functions while the significance of the distance between group centroids was tested by F-statistic.
2.5 Mitochondrial DNA Analysis
Mitochondrial DNA analysis was carried out on 148 colonies from 39 localities (Table 2.5) using Dra I RFLP of the tRNAleu-COII inter-genic region. This region, from which about 100 haplotypes have been described, has been extensively used and has become a standard in studying the intra-specific variation of A. mellifera. Due to their geographical pattern of distribution, these haplotypes are effective in identifying local populations and detecting introgression of exotic genes. In addition, because the mtDNA genome is maternally inherited, they are useful in monitoring the movement of the colonies (or queens) of honeybees due to swarming, migration or importation (Achou et al., 2015; T. Collet et al., 2006; J. M. Cornuet et al., 1991; R. H. Crozier & Crozier, 1993; Pilar De la Rúa et al., 2009; Franck et al., 2001; Pierre Franck et al., 2000; Franck et al., 1988; Garnery et al., 1992; Garnery et al., 1998; Garnery, Mosshine, Oldroyd, & Cornuet, 1995; Marina D. Meixner et al., 2013; M. Alice Pinto et al., 2014).
Extraction of total DNA was carried out using DNeasy® Blood & Tissue Kit (QIAGEN, 2006a) according to the supplementary protocol for insects (QIAGEN, 2006b). The crude DNA extract was stored at -20°C until needed.
Table 2.5 Details of the honeybee samples on which analysis of mitochondrial DNA was carried out
Abbildung in dieser Leseprobe nicht enthalten
2.5.1 Analysis of the tRNAleu-COII intergenic region
2.5.1.1 Polymerase Chain Reaction (PCR)
The mtDNA fragment containing the inter-genic region between the sub-units I and II of the cytochrome c oxidase gene (COI-COII) was amplified by PCR, using the primer pair E2 - H2, according to Garnery et al. (1998); Garnery et al. (1993) and Kandemir, Meixner, et al. (2006) with a little modification. The reaction was performed in 30 µl containing 24 pmol of each primer, 25 nmol of each dNTP, 1.5 U of Taq polymerase, 1.5 mmol of MgCl2 and 1.5 µl of DNA extract. A positive control containing DNA of known size was included. Similarly a negative control containing distilled water in place of DNA was used.
The amplification was carried out in Eppendorf Gradient Master Cycler. It consisted of an initial denaturation step of 2 minutes at 92°C, followed by 35 cycles of 3 s at 92°C, 30 s at 47°C and 2 minutes at 63°C, followed by a final extension step of 10 minutes at 63°C.
An aliquot of 5 µL of the PCR product was electrophoresed on a 1.5% agarose gel, stained with ethidium bromide and photographed under UV illumination in order to test the quality of the PCR and to determine the size of the amplified DNA. A 100 base pair ladder was used as a reference. Electrophoresis was run for 70 minutes at 150 V.
2.5.1.2 DRA I Restriction Digestion
The remaining quantity (25 µl) of PCR product of each positive reaction was digested with the restriction enzyme Dra I at 37°C overnight. Restriction fragments were separated on 10% poly-acrylamide gels, stained with ethidium bromide, and photographed under UV illumination. The restriction fragments were scored and compared with published haplotype patterns (Franck et al., 2001) in order to identify haplotypes. The concordance of the haplotypes observed in this study with published sequences was confirmed by sequencing about five representatives of each haplotype. To this end, the amplified fragments were purified and sent to a commercial sequencing facility. Using ClustalW, inbuilt in Mega6 (K. Tamura, Stecher, Peterson, Filipski, & Kumar, 2013), the sequences were then aligned with reference sequences downloaded from GenBank (FJ477987, KJ661741), and alignments were corrected manually.
2.5.2 Sequence Variation of the Cytochrome b gene
2.5.2.1 Polymerase Chain Reaction (PCR)
For a subset of samples (Afunori, Garoua, Bonny, Tahoua, Maga, Touboro, N’Djamena and Ganye), a part of the mitochondrial cytochrome b gene was sequenced. The polymerase chain reaction was used to amplify a fragment of about 700 base pairs, using the following primers (modified from Y. C. Crozier, Koullianos, and Crozier (1991) and Garnery et al. unpublished data):
CB2 : 5’ ATTAcacctcctaatttattaggaat 3’
tSer: 5’ acttattcaagttcattaact 3’.
Reactions were performed in a total volume of 50 μL with a final concentration of 1X reaction buffer, 1.5 mM MgCl2, 0.2 mM of each dNTP, 0.8 μM of each primer and 2 units of taq polymerase. The amplification cycle consisted of 94°C (60 s), 44°C (80 s), 68°C (120 s) and was repeated 35 times, followed by a final extension step of 5 minutes at 72 °C. Products were electrophoresed on a 1.5% agarose gel, stained with ethidium bromide and photographed under UV light. Amplification products were purified and sent to a commercial sequencing facility (Seqlab, Göttingen, Germany).
2.5.2.2 Sequence alignment and phylogenetic analysis
The sequences were aligned using ClustalW, inbuilt in MEGA6 (K. Tamura et al., 2013) and adjusted manually where necessary. Phylogenetic analyses were also performed using MEGA6, following the protocol of B. G. Hall (2013). In order to better evaluate the phylogenetic relationship, since the sequence was relatively short and not expected to carry much variation, three methods were used: (1) neighbour-joining (NJ), based on the maximum composite likelihood method (K. Tamura, Nei, & Kumar, 2004); (2) maximum parsimony (MP); and (3) maximum likelihood (ML), based on the Tamura-Nei model (K. Tamura & Nei, 1993). The most suitable substitution models were determined through the "Find the Best Model" feature of MEGA6. It should, however, be noted that a rigorous phylogenetic analysis was never attempted as the data currently available from Africa do not allow this. The objective was to place the samples under investigation into the context of reference data so far available from Africa. In other words, it was to see whether the samples under investigation would cluster with the North African or the sub-Saharan subspecies. Published reference sequences from African A. m scutellata and A. m. intermissa were included in the analysis, together with reference sequences of representatives of mitochondrial lineages M, C, and Z (M. A. Pinto et al., 2007). Published reference sequences of A. m adansonii and A. m. jemenitica were not available.
NJ is a distance-matrix method which assumes an explicit substitution model. Construction of a phylogenetic tree proceeds in two steps. First, for all pairs of sequences, the genetic distance is estimated from the observed sequence dissimilarity by applying a correction for multiple substitutions. The genetic distance thus reflects the expected mean number of changes per site that have occurred, since two sequences diverged from their common ancestor. Second, a phylogenetic tree is constructed by considering the relationship between these distance values (de Peer, 2009). NJ is generally more efficient than the other distance-matrix methods in obtaining the correct tree (Saitou & Nei, 1987; Sourdis & Nei, 1988). The limitation of this method, like all distance-based methods, is loss of information when the original data is converted into distances (Penny, 1982).
The MP method (a character-based method) performs a site-by-site analysis. For each tree topology, it calculates the minimum number of nucleotide changes (substitutions) that are required to explain the observed site pattern. The numbers of changes are summed over sites to give a parsimony score for each tree topology, and the topology having the smallest total number of changes is taken as the estimate of the phylogeny, which is known as the most parsimonious tree (Yang, 1996). The major drawbacks of this method is that it does not assume an explicit substitution model and it does not use all the information available in the sequences and, consequently, the most parsimonious tree is not necessarily the correct tree (de Peer, 2009; Sourdis & Nei, 1988).
Like MP, ML is a character-based method; but unlike MP, it uses all the information in the sequences and assumes an explicit substitution model. This model accounts for multiple changes at a single position, so ML simultaneously estimates both branch lengths and phylogenetic tree topologies. It analyses many trees and proposes the tree with the highest likelihood. Moreover, the method seems to be robust and relatively insensitive to violations of the evolutionary model used or to unequal rates of evolution or nucleotide bias. ML is generally the most reliable method. Its only disadvantage is that it uses a lot of computer power which makes it unsuitable for analysing large datasets (DeBry & Abele, 1995; Opperdoes, 1997).
2.5.3 Statistical Analysis of Mitochondrial Haplotypes
Analysis of frequency of haplotypes and their diversity and analysis of molecular variance were carried out with GenAlEx 6.5 (Peakall & Smouse, 2012). A chi-square test for association was conducted between haplotypes and type of vegetation, latitude, longitude, altitude, temperature and rainfall, severally, using IBM® SPSS® Statistics Version 20 with additional material from (Anonymous, 2014b). For the purpose of this test, the four belts of vegetation were combined into two: Humid (Rainforest + Guinea savanna) and dry (Sudan savanna + Sahel) vegetation. Latitude was partitioned into three: Low (4 - 8°N), medium (8 -12°N) and high (Above 12°N). Longitude was also partitioned into three: Low (1 - 8°E), medium (8 -14°E) and high (Above 14°E). However, due to its narrow range, altitude was partitioned into two: Low (Below 360 m) and high (Above 360 m). In the same vein, temperature was partitioned into two: Low (Below 27.5°C) and high (Above 27.5°C). Rainfall was partitioned into three: Low (Less than 500 mm), medium (500 - 1000 mm)) and high (Above 1000 mm).
2.6 Microsatellite Polymorphism
2.6.1 Extraction and Genotyping of DNA
Microsatellite polymorphism of nuclear DNA was employed in order to assess the genetic variation of the honeybee populations of the area of study. 133 workers (one per colony) collected from 38 localities (Table 2.6) were sent to a commercial facility for the extraction and genotyping of DNA. Alleles were initially scored automatically, using the software GeneMapper (Applied Biosystems) and the result verified by visual observation and corrected manually where necessary. 20 microsatellite loci were used: A008, A014, A029, A079, A088, A113, Ac011, Ac088, Ac306, Ap085, Ap090, AP223, Ap224, Ap226, Ap249, Ap273, Ap274, At005, At163 and At188 (Arnaud Estoup, Garnery, et al., 1995; Solignac et al., 2003). However, the following five loci were removed from the analysis due to missing data of more than 50% of alleles: A113, Ac306, Ap223, Ap226 and Ap274.
Table 2.6 Details of the honeybee samples on which analysis of microsatellite polymorphism was carried out
Abbildung in dieser Leseprobe nicht enthalten
2.6.2 Statistical Analyses
For the purpose of these analyses, two a priori populations were assumed: Humid (Rainforest + Guinea savanna) and dry (Sudan savanna + Sahel) belt populations. It should be noted that this assumption does not in any way suggest relatedness between microsatellites and climate; rather microsatellites were being used to investigate a differentiation that might have been caused by a lack of gene flow between the two populations due to isolation by distance or genetic drift.
Indices of genetic variation such as number of different alleles, allele frequencies, number and frequency of private alleles, observed heterozygosity (proportion of heterozygotes within a population) expected heterozygosity (the probability of two randomly chosen genes to be different in the sample) were obtained using GenAlEx 6.5 (Peakall & Smouse, 2012).
Tests for significant departure from Hardy-Weinberg equilibrium were carried out using Arlequin version 3 (Excoffier, Laval, & Schneider, 2005). The Hardy–Weinberg equilibrium states that both allele and genotype frequencies in a population remain constant, from generation to generation. This principle assumes that, within a given population, mating is random; no mutations are arising; no gene flow; no natural selection and population size is infinitely large (Edwards, 2008). Thus a significant departure from this principle suggests a changing population due to the violation of one or more of these assumptions.
In order to establish whether the humid and semiarid populations were parts of a single population or they were separate populations, an analysis of molecular variance (AMOVA) was used to examine the partitioning of genetic variation. Exact tests of population differentiation, based on genotype frequencies, were carried out using Arlequin version 3.5 (Excoffier et al., 2005) following the procedure described in Raymond and Rousset (1995). This analysis is analogous to Fisher's exact test on a 2 x 2 contingency table extended to a r x k contingency table where all potential states of the contingency table are examined using a Markov chain. Here, the null hypothesis of non-differentiation is tested by examining the probability of observing a table less or equally likely than the observed sample configuration. Additional indices of population differentiation such as FST, FIS, FIT and Nei genetic distance were calculated with GenAlEx 6.5 (Peakall & Smouse, 2012).
FIS is the inbreeding coefficient within individuals relative to the subpopulation. It measures the reduction in heterozygosity of an individual due to non-random mating within its subpopulation. FIT is the inbreeding coefficient within individuals relative to the total. It takes into account the effects of both non-random mating within subpopulations and genetic differentiation among the subpopulations (Peakall & Smouse, 2012).
According to Wright (1978) values of FST of 0 to 0.05 indicate a negligible differentiation between subpopulations while 0.15 to 0.25 suggest a moderate differentiation and values higher than 0.25 mean a great differentiation. However, Hedrick (2000) has shown that FST values can be considerably lower for genetic markers with many alleles, than for those with very few alleles. Therefore, what is pertinent is whether a significant genetic differentiation (FST > 0) can be detected or not, and whether this differentiation is biologically meaningful. Procedures such as AMOVA allow for such statistical tests (Peakall & Smouse, 2012).
In order to assign genotypes to specific populations, a Bayesian statistical method implemented in STRUCTURE ver. 2.3.4 (Pritchard, Stephens, & Donnelly, 2000) was used. The program implements a model-based clustering method for inferring population structure using genotype data. The model assumes that there are K populations, each of which is characterized by a set of allele frequencies at each locus. Individuals in the sample are probabilistically assigned to one of the populations, or jointly to two or more populations if their genotypes indicate they are admixed. This procedure accounts for the presence of Hardy–Weinberg and linkage disequilibrium by introducing population structure, and attempts to find population groupings that (as far as possible) are not in disequilibrium. The Markov Chain Monte Carlo (MCMC) method can allow the posterior probability distribution to be computed for estimated parameters.
The admixture model, which assumes that each individual (i) has inherited some fraction of its genome from ancestors in all K populations, and the correlated allele frequency model were used. Such assumptions seem to be the most reliable for detecting structure between closely related or admixed populations (Falush, Stephens, & Pritchard, 2003), like the populations under investigation.
Various combinations of burn-in (1,000 to 100,000 iterations) and MCMC run (10,000 to 1,000,000 iterations) lengths, at several values of K (K=1 to K=6), were tried until consistent results were obtained.
2.7 Bibliography
References were cited and listed according to the American Psychological Association, 6th edition (Anonymous, 2010), using EndNoteTM version X7.
3 Results
3.1 Morphometry
3.1.1 Variation between the Colonies under investigation
In order to investigate the variation between the colonies under investigation, the morphometric data were subjected to a number of analyses. These included descriptive statistics, principal component analysis (PCA), hierarchical cluster analysis and discriminant function analysis (DA). Correlation analysis was used to test the relationships between morphology and environmental factors, such as temperature and rainfall. The results of these analyses are presented below.
3.1.1.1 Descriptive Statistics and Analysis of Variance
Means of the 35 morphometric characters for the sampled localities are given in Tables A1 - A6 in Appendix II. Table A1 shows the means of characters of body hair and pigmentation. The length of cover hair ranged from 0.09 ± s.d. 0.01 mm in Moundou to 0.20 ± s.d. 0.01 mm in Bauchi; the width of tomentum from 0.62 ± s.d. 0.05 mm in Touboro to 0.79 ± s.d. 0.02 in Bauchi and the pigmentation of the scutellum from 6.02 ± s.d. 0.24 in Sokoto to 7.60 ± s.d. 0.17 in N'Djamena. The mean values of the characters of the hind leg are given in Table A2: The bees of Maga had the shortest leg (6.97 ± s.d. 0.13 mm) while those of Sokoto had the longest (7.52 ± s.d. 0.07 mm). As may be seen in Table A3, the combined longitudinal diameters of tergites 3 and 4 varied between 3.79 ± s.d. 0.06 mm in Gitata and 4.09 ± s.d. 0.10 mm in Maiadua; while the sternite 6 index (index of slenderness) varied from 80.15 ± s.d. 1.77 in Okuta to 84.16 ± s.d. 0.66 in Afunori. The smallest fore-wing (length = 7.98 ± s.d. 0.10 mm; width = 2.71 ± s.d. 0.04 mm) was recorded in the bees of Abeokuta while the largest fore-wing (length = 8.69 ± s.d. 0.05 mm; width = 2.96 ± s.d. 0.01mm) was found in the bees of Kano (Table A4).
A one-way ANOVA revealed that all of the morphometric characters, except pigmentation of the second and fourth tergites, distance between wax mirrors of sternite 3, the two cubital distances and cubital index, differed significantly (p < 0.05) between sampled localities.
3.1.1.2 Similarity of Colonies
In order to investigate the similarity of the honeybee colonies under study, a PCA, using colony means of 14 morphometric characters (21characters were excluded from the analysis (See §2.4.3.1.2)) of 10 worker honeybees from each of 83 colonies at 21 localities, was run to detect possible clusters. The suitability of PCA was assessed prior to analysis. Inspection of the correlation matrix showed that all variables had the minimum requirement of at least one correlation coefficient greater than 0.3. The overall Kaiser-Meyer-Olkin (KMO) measure was 0.90 with individual KMO measures from 0.61 to 0.95, thus meeting the minimum requirement for sampling adequacy. Bartlett's test of sphericity was statistically significant (p < .0005), suggesting that the data could be appropriately analysed using PCA (Anonymous, 2014b; Burns & Burns, 2008).
Three principal components, with eigenvalues 8.2, 2.1 and 0.9 each and accounting for 58.4%, 14.7% and 6.7% of the total variance, respectively, were extracted. A Varimax orthogonal rotation was employed to aid interpretability. There were strong loadings of characters of size on all three components and pigmentation of scutellum on component 3 (Table A7, Appendix II). Component 1 was loaded with (and positively correlated with ) length and width of fore-wing, longitudinal and transversal diameters of sternite 6, longitudinal and transversal diameters of wax mirror of sternite 3, longitudinal diameter of sternite 3, length of femur, length of tibia and length of metatarsus. Component 2, on the other hand, was loaded with (and positively correlated with) longitudinal diameters of tergites 3 and 4 while component 3 was loaded with pigmentation of scutellum (with a negative correlation) and width of metatarsus (with a positive correlation). Thus, principal components 1 and 2 can be defined as components of size, whose values increase with the increase in the values of size, and principal component 3 as the component of pigmentation whose values increase with the decrease in the values of pigmentation. All the four characters of hind leg, mentioned above, loaded on all the three principal components.
Although scatter plots of the three components do not show a clear clustering of the colonies, they do show a considerable variation. For example, as could be seen from Figure 3.1A, the Sahelian bees occupy the whole range of component 1, from the smallest to the largest values, but only the lower range of component 2, except two colonies found in the upper half of the graph. The colonies from the rainforest, on the other hand, are spread along the whole range of component 2 and the lower range of component 1, with only two colonies found in the right half of the graph. The colonies from the savannas are more or less evenly spread, along both components, and are therefore fairly distributed in all four quadrants. Figure 3.1B shows that most colonies from the savannas have high values of component three and are thus found in the upper half of the graph. However, the situation is reverse in the bees from the Sahel. The colonies from the rainforest, on the other hand, are more or less equally shared between the upper and lower halves of the graph. The distribution of the colonies in Figure 3.1C is similar to that in Figure 3.1B.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.1 A-C. PCA plots using the colony means of 14 morphological characters of workers of
A. mellifera from 21 localities in western Africa. All three components were loaded with characters of body size. In addition component 3 was loaded with pigmentation of scutellum. The colonies are coded according to the type of vegetation of their origin.
3.1.1.3 Classification of Colonies Based on Localities
Since PCA did not produce a clear clustering of the colonies, an alternative method, hierarchical cluster analysis, using means of 35 morphometric characters for the 23 localities under investigation, was used. Based on the output of this analysis (Figure 3.2A and Table A8, Appendix II), three tentative clusters were assumed: Cluster 1 made up of colonies from Abeokuta (rainforest) and Afunori (Sahel), cluster 2 colonies from Bauchi, Sokoto, Maiadua (Sudan savanna), Tahoua (Sahel), Bamenda, and Jos (Guinea savanna) and cluster 3 colonies from Am Timan, Bousso, Sarh, Koumra Mubi, Yola, Niamey (Sudan savanna), Maga, N’Djamena (Sahel), Okuta, Touboro, Moundou, Nguroje, Gitata (Guinea savanna) and Umuahia (rainforest).
A stepwise discriminant analysis was run, in order to predict the membership of the 83 colonies of A. mellifera, among the three a priori groups defined by the cluster analysis, as a means of confirming these groups. 35 morphometric characters of workers were used as predictor variables. Significant mean differences (ANOVA; p < 0.05) were observed for all, but 12, of the variables. While the log determinants were similar (40, 35 and 30, respectively, for the three groups), Box’s M indicated that the assumption of equality of covariance matrices was violated (p < 0.05). However, given the large sample size of the dataset, this problem was not regarded as serious and, therefore, DA was deemed appropriate (Anonymous, 2014b; Burns & Burns, 2008).
Two canonical discriminant functions (with eigenvalues 3.8 and 1.1, respectively) were used in the analysis. They explained 77.5% and 22.5% of the total variance, respectively, and their Wilk's Lambda values, assessed by chi-square, were highly significant (p < 0.0005), suggesting they had sufficient discriminatory power to group the cases. The discriminant variables used in the analysis were: Length of cover hair on tergite 5, width of tomentum, longitudinal diameter of tergite 4, width of metatarsus, length of fore-wing and angles of wing venation E9, J16 and N23. The correlations of these variables with the discriminant functions are shown in Table A9, Appendix II.
Overall, 96.4% of cross-validated grouped cases were correctly classified. All the nine colonies in group 1 were correctly classified, 8 of which were classified with a posterior probability of 100%. For group 2, 15 of the 17 colonies were correctly classified, out of which 11 were classified with a minimum posterior probability of 99%. The two misclassified colonies, one each from Tahoua and Sokoto, were assigned to group 3 with a posterior probability of 97% and 79%, respectively. For group 3, 57 of the 58 colonies were correctly classified, out of which 53 were classified with a posterior probability of 95 - 100%. The only misclassified colony in this group was from Umuahia and it was assigned to group 1 with a posterior probability of 72%. In a nutshell, 64% of the colonies were assigned to group 3, 13% to group 2 and 10% to group 1 with a posterior probability of at least 95%; while 6%, 5% and 2% of the colonies were, respectively, assigned to these groups with lower posterior probabilities. A pairwise comparison of the groups, using F-statistic, revealed a very highly significant difference between the groups' centroids (p < 0.0005). A scatter plot of the two discriminant functions is shown in Figure 3.2B, mean values of the 35 morphometric characters for the groups in Table 3.1 and the distribution of the members of the groups (morphoclusters) in the area of study in Figure 3.3.
Abbildung in dieser Leseprobe nicht enthalten
(A)
Abbildung in dieser Leseprobe nicht enthalten
(B)
Figure 3.2: (A) Classification of 23 populations of A. mellifera from western Africa by hierarchical cluster analysis (using average linkage (between groups)) using means of 35 morphometric characters for the 23 localities. The tree was cut at distance 15 to yield three tentative groups for entry into stepwise discriminant analysis. (B) Confirmation of the three groups by discriminant analysis. 96.40% of cross-validated grouped cases were correctly classified into their original groups and the distances between group centroids were highly significant (p < 0.0005) according to F-statistic. Each group is made up of colonies from distant localities in different types of vegetation (See Figure 3.3).
Table 3.1 Means and standard deviations of morphometric characters of 10 workers from N colonies of A. mellifera in West and Central Africa.
Abbildung in dieser Leseprobe nicht enthalten
†The three groups were defined by hierarchical cluster analysis, using means of 35 morphometric characters for 23 localities, and confirmed by discriminant analysis. A pairwise comparison of the groups, using F-statistic, revealed a very highly significant difference between the groups' centroids (p < 0.0005). The eight discriminant variables are highlighted. Measurements of distance are in mm and of angles in degrees.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.3 Distribution of members of three morphoclusters of A. mellifera from western Africa defined by hierarchical cluster analysis and confirmed by stepwise discriminant analysis. 87% of the colonies were assigned with posterior probabilities of 95% and above (clear classifications), while 13% were assigned with lower probabilities (unclear classifications). Each cluster consists of colonies from distant localities in different belts of vegetation. Each circle represents one colony. The black thick line demarcates the area of study while the green lines are approximate boundaries of belts of vegetation. Inset is a map of Africa showing the location of the area of study.
3.1.1.4 Classification of Colonies Based on Vegetation
In order to determine the similarity of colonies from the four zones of vegetation in the area of study, another hierarchical cluster analysis, using means of 35 morphometric characters for four zones of vegetation was run. Based on the output of the analysis (Figure 3.4A and Table A10, Appendix II), three a priori clusters were assumed: Cluster 1 made up of colonies from rainforest, cluster 2 colonies from Guinea and Sudan savanna (= savanna) and cluster 3 colonies from Sahel.
A stepwise discriminant analysis was run, in order to predict the membership of the 83 colonies of A. mellifera, among the three a priori groups defined by the cluster analysis, as a means of confirming these groups. 35 morphometric characters of workers were used as predictor variables. Significant mean differences (ANOVA; p < 0.05) were observed for all, but 13, of the variables. While the log determinants were similar (33, 25 and 28, respectively for the three groups), Box’s M indicated that the assumption of equality of covariance matrices was violated (p < 0.05). However, given the large size of the dataset, this problem was not regarded as serious, thus DA was considered appropriate (Anonymous, 2014b; Burns & Burns, 2008).
Two canonical discriminant functions (with eigenvalues 0.9 and 0.6, respectively) were used in the analysis. They explained 63.0% and 37.0% of the total variance, respectively, and their Wilk's Lambda values, assessed by chi-square, were highly significant (p < 0.0005), suggesting they had sufficient discriminatory power to group the cases. The discriminant variables used in the analysis were: Length of cover hair on tergite 5, width of tomentum, length of tibia, longitudinal diameter of wax mirror, angle of wing venation N23 and pigmentation of plates of scutellum. The correlations of these variables with the discriminant functions are shown in Table A11, Appendix II.
Overall, 83.1% of cross-validated grouped cases were correctly classified. Five of the nine colonies in the rainforest were correctly classified with a posterior probability of 74 - 94%. Three of the misclassified colonies were assigned to the savanna group with a posterior probability of 69 – 89% while the remaining one was assigned to the Sahel group with a posterior probability of 65%. For the savanna group, 54 of the 57 colonies were correctly classified out of which 27 were classified with a posterior probability of 95 – 100%. Two of the misclassified colonies were assigned to the rainforest with a posterior probability of 99% and 52%, respectively. The remaining one was assigned to the Sahel with a posterior probability of 56%. For the Sahel group, 10 of the 17 colonies were correctly classified out of which 9 were classified with a posterior probability of 95 - 100%. The seven misclassified colonies were assigned to the savanna, three with a posterior probability of 97 - 100%. In other words, the assignment of the colonies, with posterior probabilities of 95% and above, was 35%, 11% and 1%, while the assignment with lower probabilities was 42%, 4% and 7%, respectively, to the savanna, Sahel and rainforest. A pairwise comparison of the groups, using F-statistic, revealed a very highly significant difference between the groups' centroids (p < 0.0005). A scatter plot of the two discriminant functions is shown in Figure 3.4B and the distribution of the members of the groups (morphoclusters) in the area of study in Figure 3.5.
Abbildung in dieser Leseprobe nicht enthalten
A
Abbildung in dieser Leseprobe nicht enthalten
B
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.4(A) Classification of 3 populations of A. mellifera from western Africa by hierarchical cluster analysis, using average linkage (between groups) and using means of 35 morphometric characters for four zones of vegetation. The tree was cut at distance 15 to yield three tentative groups for entry into stepwise discriminant analysis. (B) Confirmation of the three groups by discriminant analysis. 83.1% of cross-validated grouped cases were correctly classified into their original groups and the distances between group centroids were highly significant (p < 0.0005) according to F-statistic.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.5 Distribution of members of three morphoclusters (rainforest, savanna and Sahel) of A. mellifera from western Africa defined by hierarchical cluster analysis and confirmed by stepwise discriminant analysis. 47% of the colonies were assigned with posterior probabilities of 95% and above (clear classifications), while 53% were assigned with lower probabilities (unclear classifications). Each cluster consists of colonies from distant localities in different belts of vegetation. Each circle represents one colony. The black thick line demarcates the area of study while the green lines are approximate boundaries of belts of vegetation. Inset is a map of Africa showing the location of the area of study.
3.1.1.5 Relationship between Morphology and Environmental Factors
A contour plot of principal component 1 against the geographical coordinates of the sampled localities presents an X-shaped pattern of distribution of the honeybees in the area of study: While the values of the principal component decrease from the centre towards northeast and southwest, they increase from the centre towards northwest and southeast (Figure 3.6). In other words, in the western part of the area of study, principal component 1 increase with latitude and in the eastern part it decreases, with the increase in latitude, apparently following the relief of the area (Figure 2.4). That is, the western part drains southwards into the Atlantic and the eastern part drains northwards into Lake Chad. The values of the principal component appear to decrease down the gradients of these systems of drainage.
To further investigate the relationship between the morphology of the honeybees and environmental factors, a correlation analysis was carried out in two steps: First, with all samples in one group; and second, with samples split into two groups, western (Atlantic) and eastern (Chadic) samples, based on the trends mentioned above. The Atlantic group consisted of samples from Tahoua, Niamey, Sokoto, Okuta, Abeokuta, Gitata, Umuahia, Jos, Bauchi, Yola, Nguroje and Bamenda. Samples from Jos, Bauchi, Maiadua, Mubi, Afunori, Touboro, Moundou, Koumra, Sarh, Am Timan, Bousso, Maga and N’Djamena formed the Chadic group. Jos and Bauchi are found in both groups because they are part of both systems of drainage.
A Pearson's product-moment correlation analysis was run (on each of the three groups of sample, separately) to assess the relationship between latitude, longitude, altitude, temperature, rainfall and the three principal components extracted by PCA. Preliminary analyses showed the relationship to be linear with all variables normally distributed (as assessed by visual inspection of normal Q-Q plots) and there were no outliers (Anonymous, 2014b).
For the combined group, principal component 1 (a measure of overall size of the body) had a moderate positive correlation with latitude (r (81) = .324, p < .01), a weak positive correlation with altitude (r (81) = .268, p < .05) and a weak negative correlation with longitude (r (81) = -.253, p < .05). Component 3 (a measure of pigmentation), on the other hand, had a moderate negative correlation (r (81) = -.347, p < .001) with longitude only. Component 2 (a measure of length of abdomen) did not significantly correlate with any of these environmental variables and neither did temperature and rainfall correlate significantly with any of the principal components. Similarly, the three principal components did not significantly correlate with each other. There was a strong positive correlation between latitude and temperature (r (81) = .787, p < .001) and a strong negative correlation between latitude and rainfall (r (81) = -.942, p < .001), and a strong negative correlation between altitude and temperature (r (81) = -.617, p < .001).
For the Atlantic group, principal component 1 had a strong positive correlation with latitude (r (47) = .571, p < .001), a weak positive correlation with temperature (r (47) = .292, p < .05) and a moderate negative correlation with rainfall (r (47) = -.411, p < .01). This principal component did not significantly correlate with either longitude or altitude. Principal component 3 had a moderate positive correlation with latitude (r (47) = .344, p < .05), a moderate negative correlation with each of altitude (r (47) = -.403 p < .01) and rainfall (r (47) = -.421 p < .01). There was no correlation between this component and longitude or temperature. Principal component 2, on the other hand, did not significantly correlate with any variable. Likewise, the three principal components did not significantly correlate with each other. There was a strong positive correlation between latitude and temperature (r (47) = .793, p < .001), a strong negative correlation between latitude and rainfall (r (47) = -.957, p < .001) and a strong negative correlation between altitude and temperature (r (47) = -.619, p < .001).
Because a negative correlation was expected, between latitude and principal component 1, instead of the positive correlation reported above, this relationship was subjected to a first-order partial correlation, in order to explore the relationship, controlling for the effects of temperature, rainfall and longitude (that is, variables that correlated with either latitude or principal component 1). The first-order correlation was found to be statistically significant (r (41) = .793, p < .001), indicating that a relationship between latitude and principal component 1 exists above and beyond the effects of temperature, rainfall and longitude.
To serve as a check, the relationship between temperature and principal component 1 was similarly subjected to a first-order partial correlation, in order to explore the relationship, controlling for the effects of latitude, longitude, altitude and rainfall (that is, variables that correlated with either temperature or principal component 1). The first-order correlation was found to be statistically insignificant (r (41) = -.191, p > .05, indicating that a relationship between temperature and principal component 1 does not exist above and beyond the effects of latitude, longitude, altitude and rainfall. This underscores the importance of latitude in the variation of principal component 1 in the Atlantic catchment (Figure 3.7A).
For the Chadic group, principal component 1 had a moderate negative correlation with each of latitude (r (38) = -.413, p < .01) and longitude (r (38) = -.384, p < .05), a strong positive correlation with altitude (r (38) = .563, p < .001), a moderate positive correlation with rainfall (r (38) = .453, p < .001) and a strong negative correlation with temperature (r (38) = -.586, p < .001). There was a moderate negative correlation between principal components 1 and 2 (r (38) = -.340, p < .05). There was a moderate positive correlation between principal 2 and longitude. This principal component, however, did not significantly correlate with latitude, altitude, temperature, or rainfall. Principal component 3 moderately negatively correlated with each of latitude (r (38) = -.362, p < .05) and temperature (r (38) = -.398, p < .05) and moderately positively correlated with each of altitude (r (38) = .476, p < .01) and rainfall (r (38) = .409, p < .05). This principal component did not correlate with longitude or any other principal component. There was a strong positive correlation between latitude and temperature (r (33) = .638, p < .001), a strong negative correlation between latitude and rainfall (r (33) = -.982, p < .001) and a strong negative correlation between altitude and temperature (r (33) = -.819, p < .001).
In order to isolate the effect of latitude on the relationship between altitude and principal component 1, and vice vasa, the relationship of either variable with principal component 1 was subjected to a first-order partial correlation, controlling for the effects of the other.
The correlation between altitude and principal component 1, controlling for latitude, was found to be statistically significant (r (35) = .438, p < .01) and the correlation between latitude and principal component 1, controlling altitude, was found to be statistically insignificant (r (35) = -.139, p > .05). Thus the correlation between principal component 1 and altitude is not dependent on latitude while the correlation between principal component 1 and latitude is dependent on altitude. This underscores the importance of altitude in the variation of principal component 1 in the Chadic catchment (Figure 3.7B).
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.6 Distribution of A. mellifera in the area of study in western Africa (enclosed by the black thick line) based on variation in the first principal component, loaded with morphometric characters of body size, extracted by PCA. The numbers labelling the contour lines are factor scores. Thus it can be seen that while the values of the principal component decrease (< 0.0) from the centre of the area of study (0.0) towards northeast and southwest, they increase from the centre towards northwest and southeast (> 0.0). In other words, in the western (Atlantic) part of the area of study, principal component 1 increase with latitude and in the eastern Chadic) part it decreases, with the increase in latitude, apparently following the relief of the area. Inset: A map of Africa showing the location of the area of study.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.7 Variation of size of A. mellifera in western Africa, illustrated by scores of the first principal component of PCA (loaded with characters of body size). In the Atlantic catchment (A) size increases with latitude and in the Chadic catchment (B) it increases with altitude.
3.1.2 Similarity of Colonies under Investigation and Reference Samples
To explore the morphological similarity of the samples under investigation and reference samples, a PCA was run on 16 morphometric characters (19 variables were excluded from the analysis; see §2.4.3.2.1) of workers from 85 colonies of the honeybees under investigation and 69 reference colonies (21 each of A. m. jemenitica and A. m. scutellata and 27 of A. m. adansonii). The suitability of PCA was assessed prior to analysis. Inspection of the correlation matrix showed that all variables had at least one correlation coefficient greater than 0.3. The overall KMO measure was 0.9 with individual KMO measures at least 0.7. Bartlett's test of sphericity was statistically significant (p < .0005), suggesting that the data could be analysed using PCA (Anonymous, 2014b; Burns & Burns, 2008).
Four principal components, each with eigenvalues greater than one and collectively accounting for 79.7% of the total variance, were extracted. Based on the scree plot, the first three principal components with eigenvalues 8.5, 2.0 and 1.2 and accounting for 53.2%, 12.5% and 7.6% of the total variance, respectively, were retained. A Varimax orthogonal rotation was employed to aid interpretability. There were strong loadings of characters of size on principal components 1 and 2 and those of pigmentation on principal component 3. Principal component 1 was loaded with length of femur, length of tibia, length and width of metatarsus, longitudinal diameter of sternite 3, and transversal diameter of wax mirror of sternite 3, longitudinal and transversal diameters of sternite 6 and length and width of fore-wing. Principal component 2, on the other hand, was loaded with longitudinal diameters of tergites 3 and 4 while principal component 3 was loaded with pigmentation of tergites 2 and 3 and scutellum. Scatter plots of different combinations of these components are shown in Figure 3.8. Component loadings and communalities of the rotated solution are presented in Table A12, Appendix II. All major loadings were positively correlated with their respective principal components.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.8 Plots of principal components extracted from a PCA using the colony means of 16 morphometric characters of workers from colonies under investigation (Unknown) and reference colonies of A. m. jemenitica, A. m. scutellata and A. m. adansonii. Principal components 1 and 2 were loaded with characters of size while principal component 3 was loaded with those of pigmentation.
As can be seen in Figure 3.8A, samples from the three reference subspecies form separable, but overlapping, clusters arranged diagonally, with A. m. scutellata and A. m. adansonii mostly occupying the higher range of both principal components (top right quadrant) and A. m. jemenitica mostly restricted to their lower range (lower left quadrant).The colonies under investigation, on the other hand, occupy the other diagonal overlapping all the three reference subspecies, especially, A. m. jemenitica. In Figure 3.8B and C while most of the colonies under investigation occupy a narrow band in the higher range of principal component 3 (upper half of the graph), those from the reference subspecies occupy a wide band in the lower range (lower half of the graph).
The PCA was then followed by a stepwise DA in order to predict the membership of the 83 colonies of A. mellifera, under investigation, among the three a priori groups assumed for the analysis: Reference subspecies of A. m. jemenitica (21 colonies), A. m. scutellata (21 colonies) and A. m. adansonii (26 colonies). 35 morphometric characters of workers were used as predictor variables. Significant mean differences (ANOVA; p < 0.05) were observed for all, but nine, of the variables. While the log determinants were quite similar (7.2, 6.5 and 6.7, respectively for A. m. jemenitica, A. m. scutellata and A. m. adansonii), Box’s M indicated that the assumption of equality of covariance matrices was violated (p < 0.05). However, given the large size of the dataset, this problem was not regarded as serious and DA was considered appropriate (Anonymous, 2014b; Burns & Burns, 2008).
Two canonical discriminant functions (with eigenvalues 6.5 and 0.6, respectively) were used in the analysis. They explained 91.9% and 8.1% of the total variance, respectively, and their Wilk's Lambda values, assessed by chi-square, were highly significant (p < 0.0005), suggesting they had sufficient discriminatory power to group the cases. The discriminant variables used in the analysis were: Length of metatarsus, longitudinal diameter of tergite 4, length of fore-wing and pigmentation of scutellum. Their correlations with the discriminant functions are shown in Table A13, Appendix II.
Overall, 94.2% of cross-validated grouped cases were correctly classified. For the A. m. jemenitica group 19 of the 21 colonies were correctly classified, 14 of which were classified with a minimum posterior probability of 95%. For the A. m. scutellata group, all 21 colonies were correctly classified 18 of which were classified with a minimum posterior probability of 95%. For the A. m. adansonii group, 25 of the 27 colonies were correctly classified 19 of which were classified with a minimum posterior probability of 95%. For the ungrouped cases (the 83 colonies under investigation) 61 colonies were assigned to A. m. jemenitica, 48 of which were classified with a posterior probability of at least 95% and 13 with lower probabilities. The remaining 22 colonies were assigned to A. m. adansonii, 14 of which were classified with a posterior probability of at least 95% and eight with lower probabilities. In a nutshell, 58% of the colonies were classified as A. m. jemenitica, 17% as A. m. adansonii, with a probability of at least 95%, and none as A. m. scutellata. The remaining 25% were classified as either A. m. jemenitica or A. m. adansonii but with a low probability (less than 95%). A pairwise comparison of the groups, using F-statistic, revealed a very highly significant difference between the groups' centroids (p < 0.0005). A scatter plot of the two discriminant functions, showing the reference colonies forming three clusters and the ungrouped colonies forming a fourth one over those of A. m. jemenitica and A. m. adansonii, is given in Figure 3.9 and the distribution of the classified ungrouped cases in the area of study is shown in Figure 3.10.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.9 Prediction of group membership of 83 colonies of A. mellifera (Ungrouped cases) by discriminant analysis among three a priori groups of reference colonies of A. m. jemenitica (n = 21), A. m. scutellata (n = 21) and A. m. adansonii (n = 27). Overall 94.2% of cross-validated grouped cases were correctly classified and the distances between group centroids were highly significant (p < 0.0005) according to F-statistic. 61 and 22 of the ungrouped cases were assigned to A. m. jemenitica and A. m. adansonii, respectively. Y<
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.10 Distribution of subspecies of A. mellifera from western Africa defined by stepwise discriminant analysis. 75% (A. m. jemenitica = 58%; A. m. adansonii = 17%) of the colonies were assigned with posterior probabilities of 95% and above (clear classifications), while 25% (A. m. jemenitica = 15%; A. m. adansonii = 10%) were assigned with lower probabilities (unclear classifications). Each subspecies consists of colonies from distant localities in different belts of vegetation. Each circle represents one colony. The black thick line demarcates the area of study while the green lines are approximate boundaries of belts of vegetation. Inset is a map of Africa showing the location of the area of study.
3.2 Mitochondrial DNA
3.2.1 COI-COII Haplotypes
Four mitochondrial haplotypes, previously reported from Africa (Mogbel A. A. El-Niweiri & Moritz, 2008; Franck et al., 2001; Shaibi et al., 2009) and belonging to the African mitochondrial lineage A, were detected: A1 (n = 62), A4 (n = 70), A4' (n = 15) and A14 (n = 1). The highest frequency of A1 was in the rainforest, followed by the Guinea savanna, and the lowest in the Sudan savanna. The frequency of A4 was highest in the Sudan savanna, followed by the Sahel, but lowest in the rainforest. Its variant, A4’, was also most prevalent in the Sudan savanna, followed by the rainforest, and rarest in the Sahel. The singleton haplotype, A14, was the only private haplotype, found only in the Sudan savanna population. The overall haplotype diversity was low (h = 0.478 ± S. E. 0.057). With regard to individual populations, haplotype diversity was lowest in the rainforest (h = 0.314) and highest in the Sudan savanna (h = 0.582) followed by the Sahel (h = 0.551) and then Guinea savanna (h = 0.503). An analysis of molecular variance revealed a statistically significant genetic variation among the four populations (Φ PT = 0.128, p = 0.001). Details are shown in Table 3.2 and Figure 3.11.
To explore the relationship between the distribution of mitochondrial haplotypes and environmental variables, a chi-square test for association was conducted between haplotypes and type of vegetation, latitude, longitude, altitude, temperature and rainfall, severally. Haplotype, A4 and its variant A4' were combined and treated as one (due to their similar pattern of distribution) while the singleton haplotype, A14, was excluded from the analysis so as to have sufficiently high expected frequencies necessary for a valid analysis (Anonymous, 2014b). In the same vain, vegetation was reduced to two logical zones: Humid (rainforest + Guinea savanna) and semiarid (Sudan savanna + Sahel). All expected cell frequencies were greater than five. There was a statistically significant association between haplotype and each of the six variables and the association was strong with latitude, moderate with vegetation and rainfall and weak with the remaining variables (Table A14, Appendix 2).
While there was a dominance of A1 haplotype in the humid vegetation (Figure 3.12A) which is in the southern part of the area of study (Figure 3.12B) and the area with the highest rainfall (Figure 3.12F) and a slightly lower temperature (Figure 3.12E), A4 and its variant, A4’, was dominant in the semiarid zone (Figure 3.12A) in the north (Figure 3.12B) which is the area with lowest rainfall (Figure 3.12F) and slightly higher temperature (Figure 3.12E) and altitude (Figure 3.12D). Figure 3.12F shows a progressive increase in the prevalence of A1 from the zones of low, through medium, to high rainfall. A progressive decrease of A4 and its variant, A4’, from west to east, was observed (Figure 3.12C).
To explore the relationship between the distribution of mitochondrial haplotypes and the three morphometric clusters obtained in the present study (§ 3.1), a chi-square test for association was also conducted. Although the test suggested a significant association (χ2 (2) = 11.888, p = 0.003), the result was disregarded since 4/6 (66.7%) of the cells had expected cell frequencies of less than five.
Table 3.2 Counts and (frequencies) of haplotypes of mitochondrial DNA of A. mellifera in four types of vegetation in western Africa.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.11 Distribution of mitochondrial DNA haplotypes of A. mellifera across sampling locations in western Africa (enclosed by the black thick line). Each circle represents one colony. Inset: A map of Africa showing the location of the area of study. The distribution shows a preponderance of A1 haplotype in the southwestern (humid vegetation) and northeastern (semiarid vegetation) and A4 and A4’ (a variant of A4) in the northwestern (semiarid vegetation) parts of the study area. There was a statistically significant moderate association between haplotypes and type of vegetation: χ2 (1) = 19.349, p < 0.0005, Cramer’s V = 0.363.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.12 Association between mitochondrial DNA haplotypes of A. mellifera and environmental variables in western Africa. All associations were statistically significant (p < 0.05) according to chi-square test for association.
Abbildung in dieser Leseprobe nicht enthalten
3.2.2 Cytochrome b Sequence Analysis
The amplification of the mitochondrial cytochrome b gene produced a fragment of approximately 800 base pairs. Among all the eight sequenced samples, a fragment of 696 base pairs was successfully aligned with the published reference sequences. The phylogenetic relationship of the samples was investigated, based on the variation of this sequence, using neighbour-joining (NJ), maximum likelihood (ML) and maximum parsimony (MP) methods. In the first two methods, samples from the current study, collected in different regions in western Africa, unambiguously clustered with the reference sequences belonging to the A lineage of mtDNA (Dra I RFLP of tRNAleu-COII inter-genic region). Within this cluster, the samples under investigation clustered with A. m. scutellata from Kenya (reference sequences from western Africa were not available in GenBank), but without showing further subdivision within this sub-Saharan cluster. Similarly, other reference sequences also clustered according to their mtDNA lineages (Figures 3.13 – 3.15). However, although the MP method produced a tree showing a similar trend (Figure 3.15), the clustering was not as clear as in NJ and ML (Figures 3.14 and 3.15).
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.13. Molecular Phylogenetic analysis of cytochrome b sequence data of Apis mellifera, together with GenBank reference data of M. A. Pinto et al. (2007), by the Neighbor-Joining method (Saitou & Nei, 1987). The optimal tree with the sum of branch length = 0.04308104 is shown. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (2000 replicates) are shown next to the branches (Felsenstein, 1985). The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Maximum Composite Likelihood method (K. Tamura et al., 2004) and are in the units of the number of base substitutions per site. The analysis involved 28 nucleotide sequences. Codon positions included were 1st+2nd+3rd+Noncoding. All positions containing gaps and missing data were eliminated. There were a total of 697 positions in the final dataset. Evolutionary analyses were conducted in MEGA6 (K. Tamura et al., 2013). Clusters are highlighted according to their lineage of mtDNA (Dra I RFLP of tRNAleu-COII inter-genic region). The samples of the actual collection from western Africa, represented by the names of their localities, are printed in red.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.14. Molecular Phylogenetic analysis of cytochrome b sequence data of Apis mellifera, together with GenBank reference data of M. A. Pinto et al. (2007), by Maximum Likelihood method. The evolutionary history was inferred based on the Tamura-Nei model (K. Tamura & Nei, 1993). The tree with the highest log likelihood (-1070.3603) is shown. The percentage of trees in which the associated taxa clustered together is shown next to the branches. Initial tree(s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach, and then selecting the topology with superior log likelihood value. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The analysis involved 28 nucleotide sequences. Codon positions included were 1st+2nd+3rd+Noncoding. All positions with less than 95% site coverage were eliminated. That is, fewer than 5% alignment gaps, missing data, and ambiguous bases were allowed at any position. There were a total of 705 positions in the final dataset. Evolutionary analyses were conducted in MEGA6 (K. Tamura et al., 2013). Clusters are highlighted according to their lineage of mtDNA (Dra I RFLP of tRNAleu-COII inter-genic region). The samples of the actual collection from western Africa, represented by the names of their localities, are printed in red.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.15 . Molecular Phylogenetic analysis of cytochrome b sequence data of Apis mellifera, together with GenBank reference data of M. A. Pinto et al. (2007), by the Maximum Parsimony method. The bootstrap consensus tree inferred from 2000 replicates is taken to represent the evolutionary history of the taxa analyzed (Felsenstein, 1985). Branches corresponding to partitions reproduced in less than 50% bootstrap replicates are collapsed. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (2000 replicates) are shown next to the branches (Felsenstein, 1985). The MP tree was obtained using the Subtree-Pruning-Regrafting (SPR) algorithm (Nei & Kumar, 2000) with search level 1 in which the initial trees were obtained by the random addition of sequences (10 replicates). The analysis involved 28 nucleotide sequences. Codon positions included were 1st+2nd+3rd+Noncoding. All positions with less than 95% site coverage were eliminated. That is, fewer than 5% alignment gaps, missing data, and ambiguous bases were allowed at any position. There were a total of 705 positions in the final dataset. Evolutionary analyses were conducted in MEGA6 (K. Tamura et al., 2013). Clusters are highlighted according to their lineage of mtDNA (Dra I RFLP of tRNAleu-COII inter-genic region). The samples of the actual collection from western Africa, represented by the names of their localities, are printed in red.
3.3 Microsatellite Polymorphism
For the purpose of this section, the samples of honeybee under investigation were grouped into two a priori populations: Populations of the humid (rainforest + Guinea savanna) and semiarid (Sudan savanna + Sahel) zones. It should be noted that this assumption does not in any way suggest relatedness between microsatellites and climate; rather microsatellites were being used to investigate a differentiation that might have been caused by a lack of gene flow between the two populations due to isolation by distance or genetic drift. A total of 292 different alleles were recorded for 15 microsatellite loci in 133 colonies (one worker per colony). All microsatellite loci were polymorphic and the number of different alleles per locus ranged between 10, in locus At163, and 31, in locus A029, while allelic frequencies ranged from 0.007 to 0.569 (Tables A15 and A16, Appendix II). The mean number of different alleles per locus is 19.5 ± S.E 1.1. The frequency of private alleles (alleles found in one population only for any particular locus) was low in both populations: From 0.009 to 0.039 (mean = 0.016) in the humid zone and from 0.007 to 0.039 (mean = 0.015) in the semiarid zone (Table A17, Appendix II). The average number of private alleles per locus in the humid population was 1.733 and in the semiarid one 4.000. Allelic patterns of the two populations are summarised in Table 3.3 and Figure 3.16.
Table 3.3: Allelic patterns (mean/locus) of two populations of A. mellifera in western Africa based on 15 microsatellite loci.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3.16 Allelic patterns, based on 15 microsatellite loci of populations of A. mellifera in western Africa. Na = Number of Different Alleles; Ne = Number of Effective Alleles; I = Shannon's Information Index; He = Expected Heterozygosity; F = Fixation Index.
Values of the number of different alleles, number of effective alleles, Shannon's Information Index, observed heterozygosity, expected heterozygosity, unbiased expected heterozygosity and Fixation Index for all loci are given in Table 3.4. Heterozygosity (or gene diversity) was high in all loci in both populations. The unbiased expected heterozygosity, which is a better expression of gene diversity, was 0.861 ± S.E. 0.017 for the two populations together and 0.859 ± S.E. 0.023 and 0.864 ± S.E. 0.026 for the humid and semiarid populations, respectively.
Table 3.4: Heterozygosity and related statistics of 15 microsatellite loci in two populations of A. mellifera in western Africa. Na = Number of Different Alleles; Ne = Number of Effective Alleles; I = Shannon's Information Index; Ho = Observed Heterozygosity; He = Expected Heterozygosity; uHe = Unbiased Expected Heterozygosity; F = Fixation Index.
Abbildung in dieser Leseprobe nicht enthalten
Wright's F-statistics over loci are given in Table 3.5. The overall FST (the inbreeding coefficient within subpopulations relative to the total) value, which is a good estimate of genetic differentiation of populations, was very low: 0.007 ± S.E. 0.001 (0.001 - 0.014). A pairwise G-statistics analysis for FST between the two populations revealed that the FST value was significant (P (rand >= data) based on 999 permutations = 0.025). Values for pairwise population Unbiased Nei Genetic Distance and Unbiased Nei Genetic Identity for the two populations were, respectively, 0.030 and 0.970.
Table 3.5 Wright’s F-Statistics for 15 microsatellite loci of A. mellifera in western Africa
Abbildung in dieser Leseprobe nicht enthalten
Key: HT = Total Expected Heterozygosity.
He = Expected Heterozygosity.
Mean He = Average He across the populations.
Ho = Observed Heterozygosity.
Mean Ho = Average Ho across the populations.
F = Fixation Index (or Inbreeding Coefficient).
FIS = the inbreeding coefficient within individuals relative to the subpopulation.
FIT = the inbreeding coefficient within individuals relative to the total.
FST = the inbreeding coefficient within subpopulations relative to the total.
According to an analysis of molecular variance (AMOVA), the two a priori populations (humid and semiarid) were not significantly different from each other (p > 0.05). Furthermore, 99.95% of the total variation was due to the variation within populations and only 0.05% was due to that between populations (Table A18, Appendix II). Similarly, Raymond and Rousset’s exact test (Raymond & Rousset, 1995) of population differentiation (a variant of Fisher’s exact test) revealed that the two populations were not significantly differentiated at p = 0.05.
The results of the Bayesian assignment test, implemented in STRUCTURE Version 2.3.4 (Pritchard et al., 2000), at different run conditions, when two populations were assumed (K = 2) are shown in Figure 3.17. No clear grouping was produced and most of the bees were assigned to both of the putative ancestral populations with a high probability: Again, this is suggestive of a lack of a structure.
Abbildung in dieser Leseprobe nicht enthalten
(A) Burn-in length = 1,000 iterations; Run length = 10,000 iterations
Abbildung in dieser Leseprobe nicht enthalten
(B) Burn-in length = 10,000 iterations; Run length = 100,000 iterations
Abbildung in dieser Leseprobe nicht enthalten
(C) Burn-in length = 100,000 iterations; Run length = 1,000,000 iterations
Figure 3.17 Results of Bayesian assignment test implemented in STRUCTURE Version 2.3.4 at different run conditions (A – C). Each vertical bar represents one worker bee, and its parts – proportions of genes from putative ancestral populations.
4 Discussion
To be able to live in the diverse ecosystems of its natural range in Western Asia, Africa and Europe, the western honeybee, A. mellifera, has evolved into many subspecies or geographical races. Presently, 30 subspecies are recognised (Engel, 1999; Marina D. Meixner et al., 2011; F. Ruttner, 1988; Walter S. Sheppard & Meixner, 2003). An adequate knowledge of the natural diversity of local subspecies and ecotypes is essential for their management and conservation. However, whereas the European subspecies have been thoroughly studied, the study of their Asian and African counterparts is still in its infancy, in many places (Marina D. Meixner et al., 2013). With only 190 colonies (from 91localities) morphometrically analysed (H. R. Hepburn & Radloff, 1998) and 114 (from 10 localities) and 30 (from one locality) colonies analysed for mitochondrial DNA and microsatellites, respectively (Franck et al., 2001), the western part of Africa (Countries in West and Central Africa, from Mauritania and Senegal in the west, to Chad in the east, then south to Namibia, through Zambia) is evidently under-studied. Therefore, the present study attempts to improve our knowledge of the diversity of the honeybees of this region by analysing 204 colonies from 44 localities in four countries - Nigeria, Niger, Cameroon and Chad - through classical morphometry and molecular methods (mitochondrial DNA and microsatellite polymorphism).
4.1 Morphometry
In morphometry 35 characters were measured in 10 workers from each of 86 colonies at 23 localities (except characters of the fore-wing, in which case measurements were taken from 152 colonies in 43 localities). The mean values for a set of these characters (Table 4.1) are in general agreement with what has been reported for the subspecies of A. mellifera in sub-Saharan Africa (F. Ruttner, 1988; Yu et al., 2012), thus confirming the validity of the measurements taken.
Table 4.1: Comparison of values (mean ± s.d.) of some morphometric characters of subspecies of A. mellifera from sub-Saharan Africa from various sources.
Abbildung in dieser Leseprobe nicht enthalten
†The validity of these values is doubtful because they fall outside the range reported for all subspecies of A. mellifera. For example, in respect of the length of the fore-wing, the smallest value, 8.13 ± 0.19 mm, was reported for A. m. jemenitica and the highest value, 9.33 ± 0.11 mm, for A. m. mellifera (F. Ruttner, 1988). Measurements of distance are in mm and of angles in degrees.
4.1.1 Morphometric Variation of Honeybees in the Area of Study
As revealed by a principal component analysis (PCA), the most important morphological variation of the honeybees of this area is size, followed by pigmentation. As may be seen from the plots of the principal components (Figure 3.1), the colonies form one large mixed cluster: That is, they do not segregate according to the type of vegetation of their origin. This suggests a continuous variation in size, which is not related to the ecological variation of the area. For example, whereas most of the smallest bees are from the rainforest (the wettest part of the area) and Sahel (the driest part of the area) the latter also harbours some the largest bees. A contour plot of the first principal component of PCA (Figure 3.6), which is loaded with characters of body size, sheds more light on the distribution of the bees in the area of study: Large bees in the northwest (Sudan savanna and Sahel) and southeast (Guinea savanna), small bees in the southwest (rainforest) and northeast (Guinea and Sudan savanna and Sahel) and medium sized bees in the centre (Guinea and Sudan savanna). In other words, the bees increase in size northwards, in the western part of the area of study, and southwards, in the eastern part. This situation, in which two opposing trends are observed in the same area, requires an explanation. Friedrich Ruttner (1985) reported a gradual increase in the size of A. m. mellifera with the increase in latitude, north and south of the equator, including the western coast of Africa were two samples collected at 5 and 15°N (the same range with the present study) showed a variation in size along this cline. However, since in western Africa temperature generally increases inland from the coast (that is with the increase in latitude), a reversal of this trend, as observed in the eastern part of the area of study, is expected. Thus the possible explanation of having larger bees in the north (in the western part of the area of study, Figure 3.7A) may lie in historical patterns of distribution or some unknown local forces of evolution, rather than current environmental factors such as temperature. This view is supported by the strong positive correlation between latitude (location) and size of bees, even after removing the effects of temperature and rainfall, and the lack of significant correlation between size and temperature after removing the effects of latitude, as shown by partial correlation analysis.
The trend in the eastern part of the area of study, which apparently contradicts Bergmann’s rule, may be due to the effect of an ecocline - a graduation in measurable characters observed in zones of different altitudes (Huxley, 1938, 1939). Thus the size of the bees increases from north to south, with altitude, based on the relief of the area (Figure 2.4; Figure 3.7B). In other words, this trend is according to Bergmann’s rule explained by altitude – larger individuals at higher altitudes. This observation is supported by the strong positive correlation between size and altitude. Similarly, a gradual increase in size of honeybees, with increasing altitude, was observed in East Africa (Amssalu, Nuru, Radloff, & Randall Hepburn, 2004; F. G. Smith, 1961). The strong negative correlation between the size of the body and temperature supports the observation made by Friedrich Ruttner (1985) that the smallest bees are found in hot places.
It is worthy of notice that the observed significant moderate positive correlation between size and latitude and the significant weak positive correlation between size and altitude, obtained after combining the two groups in the analysis, does not reflect the actual pattern of distribution of the bees. This is because the results suggest an increase of size northwards and upland, throughout the area of study, which, certainly, is not the case. This underscores the importance of taking local conditions into consideration when carrying out field-based studies.
Although principal component 3 has a very strong loading of pigmentation of scutellum ( which may represent overall pigmentation of the body) it also has a strong loading of a character of size, width of metatarsus (Appendix II, Table A7). Thus, in spite of its correlation with environmental factors, interpreting this principal component as a component of pigmentation may not be appropriate. Moreover, there is, in fact, little variation in the colour of the honeybees in this area of study (Figure 3.7C and D; Appendix II, Table A1).
The colonies were classified, according to the localities from which they were collected, using a hierarchical structure analysis, followed by a discriminant analysis (DA). Although three morphometric clusters were obtained, these groups are not geographically defined: Membership of each cluster consists of colonies from different ecological zones (Figure 3.3). Thus, again, suggesting that the observed morphological variation is not related to adaptation to current environmental conditions of any particular part of the area of study.
Secondly, the colonies were classified, according to vegetation, through a hierarchical cluster analysis, followed by a discriminant analysis (§3.1.1.4). However, due to the lower percentage of correctly classified cases, coupled with lower proportion (less than half) of clearly classified colonies (classified with a posterior probability of at least 95%) and pronounced overlapping of clusters it can be assumed, with good reason, that classifying the bees according to vegetation is a weak one (Figure 3.5). This further supports the earlier observation that the morphological variation of the bees is not tied to the major ecological variations of the area of study. This observation agrees with that of F. Ruttner (1988): “Although a phenetic north-south cline was established along the African west coast, no morphometric differentiation has yet been found, in spite of the huge geographic distance and important differences in humidity and altitude.”
Thus, from the foregoing, it is clear that although a variation exists in the population of honeybees under investigation, based on which the population could be divided into three morphometric forms, this variation is not related to the present ecology or climate of the area and, consequently, the three forms lack a geographical demarcation. Similarly, although the distribution of the bees in the area of study presented a pattern which fits into the relief of the area, it does not tally with the overall ecological pattern of the area.
4.1.2 Morphometric Variation of the honeybees of the Study Area in the Context of Published Subspecies
In a PCA, carried out to investigate the similarity between the honeybees under investigation and three reference subspecies of sub-Saharan Africa, A. m. adansonii, A. m. jemenitica and A.m. scutellata, samples from the reference subspecies form separable, but overlapping, clusters (Figure 3.8). This is suggestive of a continuous morphological variation, mostly overall size of the body, between them. A. m. jemenitica is the smallest and A. m. scutellata the largest while A. m. adansonii falls in between. The same pattern of variation was reported by F. Ruttner (1988). Thus, because the variation is a continuum, any attempt at separating these bees as subspecies, geographical races or ecotypes, solely on morphometry (without considering other biological characteristics, like physiology and behaviour (F. Ruttner, 1988), can, at best, be arbitrary. Daly (1985) warns that: "Morphometric methods are powerful research tools when used in the context of sound biological knowledge." Since the cluster formed by the bees under investigation extends over all the three reference clusters and beyond, it means some of these bees are larger than the largest scutellata and others are smaller than the smallest jemenitica. This suggests that improved sampling reveals more variation.
It is observed that, unlike the reference bees which exhibit a large variation in pigmentation (yellow-dark), the bees under investigation are virtually uniformly yellow. This could be due to the low elevation of the area of study as most of it is below 500 m above sea level. Amssalu et al. (2004), in a study of the honeybees of Ethiopia, observed a gradual increase in size and change in pigmentation from yellow to black with increasing altitude. H. Randall Hepburn, Radloff, and Oghiakhe (2000), on the other hand, reported that while the high altitude bees of Tanzania and Kenya are more darkly pigmented than the bees immediately surrounding them at lower altitudes, the trend is opposite in other places such as Cameroon. Similarly, S. Radloff and Hepburn (2000) reported these contradicting trends in A. m. scutellata. The proximity of the area to the equator and the generally high temperatures (25 – 30°C) may, in part, explain the low pigmentation observed in these bees (Friedrich Ruttner, 1985).
With more than half of the colonies under investigation classified as A. m. jemenitica, and less than one fifth classified as A. m. adansonii, by stepwise DA (Figure 3.9 ), this result apparently contradicts the popular belief that A. m. adansonii is the dominant, if not the sole, subspecies in western Africa. For example, F. Ruttner (1988) mentioned that the range of A. m. jemenitica starts from the Arabian Peninsula in the east, through Sudan and Chad and ends abruptly around Lake Chad. From here the distribution of A. m. adansonii starts and continues westwards to the Atlantic, in all ecological zones. In a study along a north-south transect in western Cameroon, S. E. Radloff and Hepburn (1997) reported a morphometric variation in the honeybees of western Africa with A. m. adansonii in the south and A. m. jemenitica in the extreme north, at Maroua, and their hybrids in between, around Garoua. H. R. Hepburn and Radloff (1998) reported that A. m. adansonii is found in the rainforest, along the west coast of Africa while A. m. jemenitica occupies the Sahel, the semiarid belt at the fringes of Sahara desert. The vast savanna lying between these two belts serves as a zone of hybridisation between the two subspecies. However, it is important to mention here that the present study renders the use of A. m. adansonii or A. m. jemenitica, to describe the honeybees of western Africa, problematic. Firstly, because previous studies are based on a few samples, as highlighted at the beginning of this chapter and in §1.5. Secondly, the use of A. m. adansonii is based on antecedence rather than a thorough taxonomical work (See §1.5). Thirdly, the use of A. m. jemenitica is based purely on morphometric resemblance of the African bees with those of the Arabian Peninsula, named A. m. jemenitica by Ruttner (1975), while subspecies should be characterized by a multitude of morphological, ethological and ecological features, originating with evolution (M. Meixner, Koeniger, & Koeniger, 1989). Consequently, they cannot be defined on the basis of morphometry alone. F. Ruttner (1988) cautions that: “In a biologically meaningful classification, the morphometric analysis has to be supplemented by essential attributes of a biological unit like, for example, specific ecological adaptations, behavioural characteristics and clear geographic demarcation.” Thus, as a matter of caution, henceforth, the terms adansonii -like and jemenitica -like, as the case may be, will be used to describe the honeybees of the present area of study.
As mentioned above, most of the colonies under investigation (75%) have been classified as either jemenitica -like or adansonii -like with a posterior probability of at least 95%. However, albeit statistically significant, this classification is not biologically meaningful, due to the following reasons:
1. The distribution of the supposed subspecies is not geographically defined (Figure 3.10). They are dispersed throughout the area of study, irrespective of ecological conditions, for example, jemenitica -like is found in N´Djamena (Sahel), Yola (Sudan savanna), Moundou (Guinea savanna) and Abeokuta (rainforest). Similarly, adansonii -like is found in Tahoua (Sahel), Bauchi (Sudan savanna) and Bamenda (Guinea savanna). One colony in Umuahia (rainforest) was classified as adansonii -like with a low posterior probability. In some localities both kinds are found, for example, Niamey (Sudan savanna) and Touboro (Guinea savanna).
2. The colonies under investigation formed one cluster in PCA and DA, thus suggesting a continuous variation, rather than separate groups.
3. The preponderance of jemenitica -like bees, even in the very humid rainforest (1000-3000 mm of rainfall; 25-27°C), though A. m. jemenitica has been described as a bee of very dry areas, with a very low and irregular rainfall (50-300 mm per annum) and temperatures of 27-31°C (F. Ruttner, 1982, 1988).
Therefore, in light of these observations, the classification of the bees into two subspecies is not considered valid and, therefore, they are regarded as members of a single morphometric entity, simply referred to as a "morphocluster", within which some variation exists, pending further investigation. Similarly, F. Ruttner (1988) observed that: “Although a phenetic north-south cline was established along the African west coast, no morphometric differentiation has yet been found, in spite of the huge geographic distance and important differences in humidity and altitude.”
Consequently, it is argued that splitting the honeybees of this area into two subspecies, A. m. adansonii and A. m. jemenitica (H. R. Hepburn & Radloff, 1998; F. Ruttner, 1988), may be due to inadequate sampling. H. R. Hepburn and Radloff (1998) used 190 samples collected from 91 localities in the A. m. adansonii zone (out of which 54 samples from 11 localities were from the area of study of this work) in their analysis with a word of caution that, given what is regarded as western Africa, the number of sampled localities was inadequate; thus echoing an observation earlier made by (F. Ruttner, 1988). Even less samples were available to (F. Ruttner, 1988) since the former carried out the analysis after combining the latter`s databank with their own. Thus sparse sampling might be responsible for masking some of the variation which has been captured by the present study, due to improved sampling: More localities (44 versus 11) and samples (83 versus 45). This is in line with an earlier observation by Sarah E. Radloff and Hepburn (1998) that the greater the distance between samples, the more distinct morphoclusters are.
4.2 Mitochondrial DNA Polymorphism
The earliest study on the variation of mitochondrial DNA of honeybees, based on digestion of the whole genome with restriction enzymes, suggested that this genome might not possess sufficient variation among populations of honeybees (R. F. A. Moritz, Hawkins, Crozier, & Mackinley, 1986). However, subsequent studies, based on the length of the non-coding intergenic region, located between the COI and COII subunits, and its further digestion with DRA I revealed a substantial variation among populations of this species in Africa (Achou et al., 2015; P. De la Rúa, J. Galián, J. Serrano, & R. F. A. Moritz, 2001; De la Rua, Serrano, & Galian, 1998; Mogbel A. A. El-Niweiri & Moritz, 2008; Franck et al., 2001; Garnery et al., 1992; Garnery et al., 1995; Garnery et al., 1993; Rasolofoarivao et al., 2015; Shaibi et al., 2009; Techer, Clémencet, Simiand, et al., 2015), Europe (F. Cánovas, De la Rúa, Serrano, & Galián, 2008; De la Rua, Galian, Serrano, & Moritz, 2003; Pilar De la Rúa, Jiménez, Galián, & Serrano, 2004; Franck et al., 2001; Garnery et al., 1992; Garnery et al., 1998; Garnery et al., 1995; Garnery et al., 1993; Jensen, Palmer, Boomsma, & Pedersen, 2005; Kandemir, Meixner, et al., 2006; Miguel, Iriondo, Garnery, Sheppard, & Estonba, 2007; Muñoz, Stevanovic, Stanimirovic, & De la Rúa, 2012; Nedic, Stanisavljevic, Mladenovic, & Stanisavljevic, 2009; M. Alice Pinto et al., 2014; Susnik, Kozmus, Poklukar, & Meglic, 2004) the Middle East (Alattal et al., 2014; Alburaki et al., 2011; Pierre Franck et al., 2000; Nizar Haddad et al., 2009; Jang, Yılız, Fakhri, & Nobakht, 2011; Kandemir, Kence, Sheppard, & Kence, 2006; Kence, Farhoud, & Tunca, 2009; Özdïl, Yildiz, & Hall, 2009; Palmer, Smith, & Kaftanoglu, 2009; D. R. Smith, Slaymaker, Palmer, & Kaftanoglu, 1997; Solorzano et al., 2009; Zaitoun, Hassawi, & Shahrour, 2008) and the Americas (Clarke, Oldroyd, Javier, Quezada-Euan, & Rinderer, 2001; T. Collet et al., 2006; Franck et al., 2001; Prada, Duran, Salamanca, & Del Lama, 2009; Segura, 2000; Deborah Roan Smith, 1991; Szalanski & Magnus, 2010).
Initially the mtDNA data indicated three evolutionary lineages, with distinct geographical distributions, within Apis mellifera: branch A for African subspecies (intermissa, monticola, scutellata, andansonii and capensis), branch C for North Mediterranean subspecies (caucasica, carnica and ligustica) and branch M for the West European populations (mellifera) (Garnery et al., 1992; Deborah Roan Smith, 1991). Later, two additional lineages were proposed: O for the subspecies of the Middle East (Pierre Franck et al., 2000) and Y for those of Ethiopia (Franck et al., 2001). However, further studies suggest that O (renamed Z) and Y are sub-lineages of the African lineage, A, rather than stand-alone lineages (Alburaki et al., 2011; Marina D. Meixner et al., 2013). Based on the frequency of its haplotypes, this marker is also suitable for studying variation within these lineages. For example, whereas haplotypes A8, A9 and A10 are more frequent in subspecies of northwest Africa, A1 and A4 are most common in sub-Saharan Africa. Similarly, haplotypes A2 and A3 are principally observed in Spain and Sicily, and haplotypes of the AIII group (e.g. A11, A14, A15, A16, A20, A21, A35 and A42) are the most frequent within the populations from Portugal and the Canary Islands (Franck et al., 2001; Maria Alice Pinto, Muñoz, Chávez-Galarza, & De la Rúa, 2012). However, no diagnostic haplotypes have so far been found for any subspecies (Marina D. Meixner et al., 2013).
In addition, among other uses, the variation of the COI - COII intergenic region have been found useful in monitoring the genetic integrity of local populations and movement of queens and establishing the historical and geographical origin of local populations (Clarke et al., 2001; Clarke, Rinderer, Franck, Quezada-Euán, & Oldroyd, 2002; T. Collet et al., 2006; Kraus, Franck, & Vandame, 2007; M. Alice Pinto et al., 2014; Prada et al., 2009; Schiff, Sheppard, Loper, & Shimanuki, 1994; Segura, 2000; Szalanski & Magnus, 2010).
In order to further explore the variation of mitochondrial DNA of the honeybees of Africa, 148 colonies, from 39 localities, were investigated. Four, previously described mitochondrial haplotypes, belonging to the African mitochondrial lineage, A (A1, A4, A4' and A14), were detected. These haplotypes have been reported from many subspecies of A. mellifera from Africa and its islands, the Iberian Peninsula, the Mediterranean islands and the Middle East. For example, haplotypes A1 and A4, the most common haplotypes found in the African subspecies of A. mellifera (Franck et al., 2001), have been found in A. m. adansonii, A. m. scutellata and A. m. capensis in sub-Saharan Africa (Mogbel A. A. El-Niweiri & Moritz, 2008; Franck et al., 2001); A. m. intermissa and A. m. sahariensis in northwest Africa ( Achou et al., 2015; Franck et al., 2001; Shaibi et al., 2009); A. m. iberiensis in the Iberian Peninsula (F. Cánovas et al., 2008; Pilar De la Rúa et al., 2004; Franck et al., 2001; Miguel et al., 2007) and Africanised honeybees in the Americas (Clarke et al., 2001; T. Collet et al., 2006; Franck et al., 2001; Kraus et al., 2007; Prada et al., 2009; Segura, 2000; Szalanski & Magnus, 2010). These haplotypes have also been reported in A. m. unicolor from Madagascar and Mauritius, A. m. adansonii from Cabo Verde and São Tomé and A. m. siciliana from Sicily (Franck et al., 2001; Garnery et al., 1993; Rasolofoarivao et al., 2015). Haplotype A1 was reported in A. m. syriaca and A. m. jemenitica in the Middle East (Alattal et al., 2014; Alburaki et al., 2011); A. m. iberiensis in Canary and Balearic islands (P. De la Rúa et al., 2001; Pilar De La Rúa, José Galián, José Serrano, & R.F.A. Moritz, 2001) and A. m. unicolor from Seychelles and Reunion (Franck et al., 2001; Techer, Clémencet, Simiand, et al., 2015). The less common haplotype, A4', on the other hand, was found in A. m. adansonii in sub-Saharan Africa and São Tomé (Franck et al., 2001) and A. m. iberiensis in the Iberian Peninsula (Pilar De la Rúa et al., 2004). The singleton haplotype found in this study, A14, was reported in A. m. adansonii in Namibia and A. m. iberiensis in the Iberian Peninsula and Canary islands (F. Cánovas et al., 2008; Franck et al., 2001).
The overall unbiased haplotype diversity was low (0.495 ± S. E. 0.078). Similarly, Franck et al. (1988) reported a low unbiased haplotype diversity for the A lineage (0.428 ± S. D. 0.070 – 0.681 ± S. D. 0.063), throughout its range, from West Africa through northwest Africa to the Iberian Peninsula, as compared to the M lineage in France and the Iberian Peninsula (0.324 ± S. D. 0.054 – 0.771 ± S. D. 0.041). The low mitochondrial diversity reported by this and previous studies (Mogbel A. A. El-Niweiri & Moritz, 2008; Franck et al., 2001), suggests this to be a characteristic of the honeybees of Africa. For example, whereas F. Cánovas et al. (2008) reported 22 haplotypes from Spain alone, in a study involving 1017 colonies, sampled from 109 localities, Franck et al. (2001) found only 21 haplotypes, in the whole of Africa, from the 738 colonies analysed from 64 localities. Similarly, Solorzano et al. (2009) identified 12 haplotypes from 135 colonies, collected from 22 localities, in Turkey. Alburaki et al. (2011), on the other hand, identified 21 haplotypes from an analysis of 1837 colonies from the Middle East (Syria, Lebanon and Iraq). Given the high variability of the COI – COII intergenic region, as demonstrated by the over 100 distinct haplotypes described for the region (Marina D. Meixner et al., 2013), its low diversity in Africa is unexpected. Moreover, given the high population densities of the honeybees of Africa, together with their high migratory behaviour, and their historically more stable environment (no bottlenecks), the honeybees of Africa are expected to show a higher mitochondrial diversity than the European ones, just as they have shown with microsatellites. This anomaly has been explained by a lower rate of mutation in the COI – COII region due to its shorter length (Franck et al., 1988) because smaller sequences offer less targets for site mutations and less possibilities for the duplication or deletion of Q (J. M. Cornuet et al., 1991).
The statistically significant association between A1 and humid vegetation and between A4 and semiarid vegetation suggests that A1 is a humid zone haplotype while A4 is a dry zone haplotype (Figure 3.12A). This is further supported by the progressive increase of the frequency of A1 from areas of low rainfall to those of high rainfall (Figure 3.12F). Thaís Collet, Arias, and Del Lama (2007) observed a similar trend in Brazil. However, since the COI-COII inter-genic region is non-coding, the observed distribution of haplotypes may be due to a historical event, such as ancient environment conditions. Alternatively, the region may be linked to a coding region (Dr. Per Kryger, personal communication) or, probably, this marker is not selectively neutral (Clarke et al., 2001; Franck et al., 1988). Similarly, the observed progressive decrease of A4 and A4’ haplotypes, from west to east, may be due to the same reasons (Figure 3.12C).
The clustering of the honeybees under investigation with A. m. scutellata, in the cytochrome b analysis confirms their status as sub-Saharan honeybees.
4.3 Microsatellite polymorphism
Microsatellites have been used in studying the variation and structure of populations of A. mellifera in Africa (Franck et al., 2001; Franck et al., 1988; Loucif-Ayad, Achou, Legout, Alburaki, & Garnery, 2015; Rasolofoarivao et al., 2015; Techer, Clémencet, Simiand, et al., 2015; Techer, Clémencet, Turpin, et al., 2015), Europe (Fernando Cánovas, De La Rúa, Serrano, & Galián, 2011; Dall'Olio, Marino, Lodesani, & Moritz, 2007; De la Rua et al., 2003; P. De la Rúa et al., 2001; P. De la Rúa, Galián, Serrano, & Moritz, 2002; Arnaud Estoup, Garnery, et al., 1995; Garnery et al., 1998; Kandemir, Meixner, et al., 2006; S. R. Nikolova & Ivanova, 2012), the Middle East (Alburaki et al., 2013; Pierre Franck et al., 2000; Kence et al., 2009; Rahimi et al., 2014) and in Africanised honeybees in the Americas (Clarke et al., 2002; Kraus et al., 2007; McMichael & Hall, 1996; Segura, 2000). Also, microsatellite analyses have confirmed that A. mellifera evolved in three distinct and deeply differentiated lineages, previously detected by morphological and mitochondrial DNA studies (Arnaud Estoup, Garnery, et al., 1995). Furthermore, microsatellite analyses have been found useful in distinguishing some subspecies and group of subspecies (McMichael & Hall, 1996; Miguel et al., 2007; Nedić et al., 2014; Oleksa & Tofilski, 2014; Uzunov et al., 2014) and monitoring of the genetic integrity of local populations such as in breeding and conservation programmes (Bourgeois & Rinderer, 2009; Kraus et al., 2007; Muñoz et al., 2014; Muñoz & De la Rúa, 2012; S. Nikolova, 2011; Oleksa, Chybicki, Tofilski, & Burczyk, 2011).
To further investigate the variation of the honeybees of Africa, using microsatellite polymorphism, 133 colonies from 38 localities were analysed, using 15 loci. The mean number of different alleles per locus is 19.5 ± S.E 1.1. This is not surprising since the honeybees of Africa are known to have a high level of polymorphism at nuclear markers (Arnaud Estoup, Garnery, et al., 1995; Franck et al., 2001; Franck et al., 1988; H. G. Hall, 1992, 1998; Loucif-Ayad et al., 2015; McMichael & Hall, 1996). For example, a mean number of different alleles per locus of 16.6 ± S.D. 1.7 and 7.29 ± S.D. 0.73 - 13.79 ± S.D. 1.17 were reported for populations of honeybees in South Africa (Robin F. A. Moritz, Kraus, Kryger, & Crewe, 2007) and Algeria (Loucif-Ayad et al., 2015), respectively. Arnaud Estoup, Garnery, et al. (1995), on the other hand, reported a mean number of different alleles per locus of 4.1± S.E. 0.6 – 6.1± S.E. 1.2 and 7.01 ± S.E. 1.6 – 11.0 ± S.E. 1.0.8 for European and African populations, respectively.
The high unbiased expected heterozygosity, or gene diversity, reported by this study (0.861 ± S.E. 0.017) is typical of the African races of honeybee, which always show higher values than other races. This has been interpreted as a consequence of larger effective population sizes in Africa (Arnaud Estoup, Tailliez, et al., 1995; Franck et al., 2001). For example, Robin F. A. Moritz et al. (2007) reported an unbiased gene diversity of 0.767 - 0.905 for South African honeybee populations against 0.666 - 0.713 for German populations. Similarly, Alburaki et al. (2013) reported the following gene diversity values for the four morphometric lineages (§1.3) as follows: Lineage M (Western Mediterranean and Northwestern European races, 0.425 - 0.496), lineage C (Eastern Mediterranean and Southeastern European races, 0.478 - 0.529), Lineage O (Oriental and Mediterranean races, 0.612 - 0.672) and lineage A (Tropical African races, 0.757 - 0.811).
Gene diversity is high in both the humid and semiarid populations and, a genetic distance tending towards zero (Unbiased Nei Genetic Distance = 0.030) and a genetic identity tending towards unity (Unbiased Nei Genetic Identity = 0.970) suggest that the two assumed populations are not separate. In addition, both AMOVA and Bayesian assignment suggest a lack of differentiation of the honeybees under investigation. There are three possible explanations to this situation: Firstly, the observed apparent lack of differentiation of the population may be real and may be due to a high gene flow among the bees or the lesser impact on Africa by the Quaternary ice episodes which are thought to be the major cause of the differentiation of A. mellifera in Europe (Hewitt, 2000; F. Ruttner, 1988). Gene flow is facilitated by the high mobility of African honeybees through mating flights, swarming, abscondment and migration (Fletcher, 1978; Gruber, Schoning, Otte, Kinuthia, & Hasselmann, 2013; H. R. Hepburn & Radloff, 1998; Robin F. A. Moritz et al., 2005; Nightingale, 1976; Rashad & El-Sarrag, 1980; F. Ruttner, 1988; Woyke, 1993). Secondly, the apparent lack of differentiation of the population may be consequent to the observed high gene diversity which results in low FST values. Thirdly, the population may be structured but the analysis failed to detect the structure due to an inadequate sampling (Per Kryger, personal communication). Fourthly, microsatellites, originally developed for studying European honeybees, may not be suitable for studying African bees.
The lack of congruence between microsatellites and mitochondrial DNA (which may be due to their different rates of evolution) found in this study was earlier reported for honeybees of the Iberian Peninsula (Franck et al., 1988).
4.4 Conclusions
The present study improved, substantially, the coverage of western Africa in respect of the study of the diversity of the western honeybee, A. mellifera. Previous studies analysed 304 colonies collected from 101 localities in 20 countries. The present study analysed 204 colonies collected from 44 (38 of which are new) localities in four of the countries.
Based on morphometric analysis, the results of this study present the honeybees of western Africa as a single entity with an internal variation which lacks a geographical demarcation. Consequently the results do not support the classification of the honeybees of the region into the two subspecies, A. m. adansonii and A. m. jemenitica, as reported in the literature. However, their clustering with reference samples of sub-Saharan subspecies confirms their belonging to the A morphometric lineage. To confirm their taxonomic status, more morphometric, molecular, physiological and behavioural studies are required. Meanwhile, the use of A. m. adansonii, as the sole sub-specific name for the honeybees of West and Central Africa, is recommended.
This study is reporting, for the first time, the existence of an ecocline for the honeybees of the area studied, centred on Lake Chad. In this cline the bees increase in size, from the shores of the lake upland, towards the periphery of the basin, according to Bergmann's rule.
Analysis of mitochondrial DNA polymorphism reveals a low genetic diversity of the honeybees of western Africa. All the four haplotypes found, belonging to the mitochondrial sub-lineages AI and AIII of lineage A, are endemic. The lack of reporting of exotic haplotypes is a strong indication of the mitochondrial purity of the honeybees of this region. That is to say, exotic queens or colonies that might have been introduced into this region have failed to establish; or their genomes have been diluted to the level of insignificance. The clustering of the honeybees of the region with A. m. scutellata, in the cytochrome b analysis confirms their status as sub-Saharan honeybees.
Analysis of microsatellite polymorphism reveals very high gene diversity and a very low genetic differentiation in the honeybees of this region. Although, as indicated above, this result agrees with what has been reported for the honeybees of Africa, still there is the possibility that the low genetic differentiation may have been exaggerated by the very high polymorphism of this marker in African honeybees as compared to European bees, for which it was originally developed. In the same vain, the COI-COII intergenic region of mitochondrial DNA may also be unsuitable for the honeybees of Africa but for an opposite reason - in sufficient polymorphism. The search for alternative molecular markers for the honeybees of Africa is therefore recommended.
The close morphometric and mitochondrial resemblance of the honeybees around Lake Chad with those of the humid rainforest of southern Nigeria (as suggested by body size and preponderance of A1 haplotype and its significant association with humidity) may be explained by the history of the lake. Though the area around Lake Chad is semiarid today, it has been hypothesised to have been humid in the past and, therefore, the distribution of the bees is a reflection of this. Palaeontological studies suggest that modern Lake Chad is a tiny remnant of an ancient large lake called Lake Megachad (Figure 4.1). At its peak, during early Holocene (10,000 – 8,000 years ago), Megachad occupied an area of at least 361,000 km²: That is, it was larger than any lake existing today (Anonymous, 2015; N. A. Drake & Bristow, 2006; Ghienne, Schuster, Bernard, Duringer, & Brunet, 2002; Schuster et al., 2005). N. A. Drake, Blench, Armitage, Bristow, and White (2011) suggest that aridity was completely ameliorated, not only in this area, but in the whole of the central Saharan region, by the combined effect of this and two other mega-lakes (Chotts Megalake and Lake Megafezzan).
This study has been able to establish, through morphometric, mitochondrial DNA and microsatellite analyses, the existence of a genetic variation among the honeybees of western Africa. Therefore, the null hypothesis of the study – “There is no variation among the honeybees of western Africa” – is hereby rejected.
Abbildung in dieser Leseprobe nicht enthalten
Figure 4.1 Shuttle Radar Topography Mission (1 km resolution) Digital Elevation Model showing Saharan palaeolakes over 500 km2. (Image by Nick Drake; Source: The Sahara Megalakes Project, King’s College, London. Retrieved 22nd January 2015, from http://www.kcl.ac.uk/sspp/departments/geography/people/academic/drake/Research/The-Sahara-Megalakes-Project/The-Sahara-Megalakes-Project.aspx
References
Achou, M., Loucif-Ayad, W., Legout, H., Hmidan, H., Alburaki, M., & Garnery, L. (2015). An insightful molecular analysis reveals foreign honeybees among Algerian honeybee populations (Apis mellifera L.). Journal of Data Mining in Genomics & Proteomics, 6 (1), 166. doi:10.4172/2153-0602.1000166
Ajao, A. M., Oladimeji, Y. U., Idowu, A. B., Babatunde, S. K., & Obembe, A. (2014). Morphological characteristics of Apis mellifera L . (Hymenoptera : Apidae) in Kwara State , Nigeria. International Journal of Agricultural Sciences, 4 (4), 171-175.
Alattal, Y., Alsharhi, M., Alghamdi, A., Alfaify, S., Migdadi, H., & Ansari, M. (2014). Characterization of the native honey bee subspecies in Saudi Arabia using the mtDNA COI–COII intergenic region and morphometric characteristics. Bulletin of Insectology, 67 (1), 31-37.
Alburaki, M., Bertrand, B., Legout, H., Moulin, S., Alburaki, A., Sheppard, W. S., & Garnery, L. (2013). A fifth major genetic group among honeybees revealed in Syria. Genetics, 14. Retrieved from http://www.biomedcentral.com/1471-2156/14/117
Alburaki, M., Moulin, S., Legout, H., Alburaki, A., & Garnery, L. (2011). Mitochondrial structure of Eastern honeybee populations from Syria, Lebanon and Iraq. Apidologie, 42 (5), 628-641. doi:10.1007/s13592-011-0062-4
Allsopp, M. H. (1992). The capensis calamity. S. Afr. Bee J . 64, 52–55.
Allsopp, M. H., & Crewe, R. (1993). The Cape honeybee as a Trojan horse rather than the hordes of Jenghiz Khan. Am. Bee J., 133, 121-123.
Allsopp, M. H., Govan, V., & Davison, S. (1997). Bee health report: Varroa in South Africa. Bee World, 74 (4), 171-174.
Amssalu, B., Nuru, A., Radloff, S. E., & Randall Hepburn, H. (2004). Multivariate morphometric analysis of honeybees (Apis mellifera) in the Ethiopian region. Apidologie, 35 (1), 71-81. doi:10.1051/apido:2003066
Anonymous. (1992). World animal health 1991. VII. Animal health status and disease control methods. Part 2. Tables.. Retrieved from Paris, France:
Anonymous. (2006). Insights into social insects from the genome of the honeybee Apis mellifera. Nature, 443, 931-949. doi:10.1038/nature05260
Anonymous. (2009). The health of livestock and honeybees in England (HC 288 Session 2008-2009). Retrieved from London:
Anonymous. (2010). Publication manual of the American Psychological Association. (G. R. VandenBos Ed. 6th ed.). Washington, DC: American Psychological Association.
Anonymous. (2013). Managing Varroa. York, UK: The Food and Environment Research Agency.
Anonymous. (2014a). The Climate Change Knowledge Portal. Retrieved from http://sdwebx.worldbank.org/climateportal/index.cfm
Anonymous. (2014b). Laerd Statistics: IBM SPSS Tutorials and Statistical Guides. Retrieved from https://statistics.laerd.com/
Anonymous. (2015). The Sahara Megalakes Project. Retrieved from http://www.kcl.ac.uk/sspp/departments/geography/people/academic/drake/Research/The-Sahara-Megalakes-Project/The-Sahara-Megalakes-Project.aspx
Anonymous. (n. d.). FAOSTAT. Food and agricultural commodities production. Retrieved from http://faostat.fao.org/site/339/default.aspx
Badano, E. I., & Vergara, C. H. (2011). Potential negative effects of exotic honey bees on the diversity of native pollinators and yield of highland coffee plantations. Agricultural and Forest Entomology, 13 (4), 365-372. doi:10.1111/j.1461-9563.2011.00527.x
Bouga, M., Alaux, C., Bienkowska, M., Büchler, R., Carreck, N. L., Cauia, E., . . . Wilde, J. (2011). A review of methods for discrimination of honey bee populations as applied to European beekeeping. Journal of Apicultural Research, 50 (1), 51-84. doi:10.3896/ibra.1.50.1.06
Bourgeois, A. L., & Rinderer, T. E. (2009). Genetic characterization of Russian honey bee stock selected for improved resistance to Varroa destructor. J. Econ. Entomol., 102 (3), 1233-1238
Brown, T. C. (2013). Reactions to Honeybee Stings. Curr Opin Allergy Clin Immunol, 13 (4), 365-371.
Burns, R., & Burns, R. (2008). Business research methods and statistics using spss. London, Thousand Oaks, New Delhi, Singapore: SAGE Publications Ltd.
Calderone, N. W. (2012). Insect pollinated crops, insect pollinators and US agriculture: trend analysis of aggregate data for the period 1992-2009. PLoS One, 7 (5), e37235. doi:10.1371/journal.pone.0037235
Cánovas, F., De la Rúa, P., Serrano, J., & Galián, J. (2008). Geographical patterns of mitochondrial DNA variation in Apis mellifera iberiensis (Hymenoptera: Apidae). J Zool Syst Evol Res, 46 (1), 24–30. doi:10.1111/j.1439-0469.2007.00435.x
Cánovas, F., De La Rúa, P., Serrano, J., & Galián, J. (2011). Microsatellite variability reveals beekeeping influences on Iberian honeybee populations. Apidologie, 42, 235–251. doi:10.1007/s13592-011-0020-1
Chauzat, M.-P., Carpentier, P., Madec, F., Bougeard, S., Cougoule, N., Drajnudel, P., . . . Faucon, J.-P. (2010). The role of infectious agents and parasites in the health of honey bee colonies in France. Journal of Apicultural Research, 49 (1), 31-39. doi:10.3896/IBRA.1.49.1.05
Clarke, K. E., Oldroyd, B. P., Javier, J., Quezada-Euan, G., & Rinderer, T. E. (2001). Origin of honeybees (Apis mellifera L.) from the Yucatan peninsula inferred from mitochondrial DNA analysis. Mol Ecol, 10 (6), 1347-1355. doi:10.1046/j.1365-294X.2001.01274.x
Clarke, K. E., Rinderer, T. E., Franck, P., Quezada-Euán, J. G., & Oldroyd, B. P. (2002). The Africanization of Honeybees (Apis Mellifera L.) of the Yucatan:A Study of a Massive Hybridization Event across Time. Evolution, 56 (7), 1462. doi:10.1554/0014-3820(2002)056[1462:taoham]2.0.co;2
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (Second ed.). Hillsdale, New Jessy, Hove and London: Lawrence Erlbaum Associates.
Collet, T., Arias, M. C., & Del Lama, M. A. (2007). 16S mtDNA variation in Apis mellifera detected by PCR-RFLP. Apidologie, 38 (1), 47-54. doi:10.1051/apido:2006056
Collet, T., Ferreira, K. M., Arias, M. C., Soares, A. E., & Del Lama, M. A. (2006). Genetic structure of Africanized honeybee populations (Apis mellifera L.) from Brazil and Uruguay viewed through mitochondrial DNA COI-COII patterns. Heredity (Edinb), 97 (5), 329-335. doi:10.1038/sj.hdy.6800875
Cornuet, J.-M., Piry, S., Luikart, G., Estoup, A., & Solignac, M. (1999). New Methods Employing Multilocus Genotypes to Select or Exclude Populations as Origins of Individuals. Genetics, 153, 1989-2000. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1460843/pdf/10581301.pdf
Cornuet, J. M., Garnery, L., & Solignac, M. (1991). Putative origin and function of the intergenic region between COI and COII of Apis mellifera L. mitochondrial DNA. Genetics, 1128, 393-403.
Coulson, R. N., Pinto, M. A., Tchakerian, M. D., Baum, K. A., Rubink, W. L., & Johnston, J. S. (2005). Feral honey bees in pine forest landscapes of east Texas. Forest Ecology and Management, 215 (1-3), 91-102. doi:10.1016/j.foreco.2005.05.005
Crane, E. (1999). The world history of beekeeping and honey hunting. London: Gerald Duckworth and Company Ltd.
Crozier, R. H., & Crozier, Y. C. (1993). The mitochondrial genome of the honeybee, Apis mellifera: Complete sequence and genome organization. Genetics, 133, 97-117.
Crozier, R. H., Crozier, Y. C., & Mackinlay, A. G. (1989). The CO-I and CO-II region of honeybee mitochondrial DNA: Evidence for variation in insect mitochondrial evolutionary rates. Mol. Biol. Evol., 6 (4), 399-411.
Crozier, Y. C., Koullianos, S., & Crozier, R. H. (1991). An improved test for Africanized honeybee mitochondrial DNA. Experientia, 47, 968-969.
Cuthbertson, A. G. S., Wakefield, M. E., Powell, M. E., Marris, G., Anderson, H., Budge, G. E., . . . Brown, M. A. (2013). The small hive beetle Aethina tumida: A review of its biology and control measures. Current Zoology, 59 (5), 644–653.
Dall'Olio, R., Marino, A., Lodesani, M., & Moritz, R. F. A. (2007). Genetic characterization of Italian honeybees, Apis mellifera ligustica, based on microsatellite DNA polymorphisms. Apidologie, 38 (2), 207-217. doi:10.1051/apido:2006073
Daly, H. V. (1985). Insect Morphometrics. Annu Rev Entomol, 30, 415-438. doi:10.1146/annurev.en.30.010185.002215
De la Rua, P., Galian, J., Serrano, J., & Moritz, R. F. (2003). Genetic structure of Balearic honeybee populations based on microsatellite polymorphism. Genet Sel Evol, 35 (3), 339-350. doi:10.1051/gse:2003012
De la Rúa, P., Galián, J., Serrano, J., & Moritz, R. F. A. (2001). Genetic structure and distinctness of Apis mellifera L. populations from the Canary Islands. Mol Ecol, 10, 1733-1742. Retrieved from http://onlinelibrary.wiley.com/doi/10.1046/j.1365-294X.2001.01303.x/abstract
De La Rúa, P., Galián, J., Serrano, J., & Moritz, R. F. A. (2001). Molecular characterization and population structure of the honeybees from the Balearic islands (Spain). Apidologie, 32 (5), 417-427. doi:10.1051/apido:2001141
De la Rúa, P., Galián, J., Serrano, J., & Moritz, R. F. A. (2002). Microsatellite analysis of non-migratory colonies of Apis mellifera iberica from south-eastern Spain. J. Zool. Syst. Evol. Research, 40, 164-168.
De la Rúa, P., Jaffé, R., Dall'Olio, R., Muñoz, I., & Serrano, J. (2009). Biodiversity, conservation and current threats to European honeybees. Apidologie, 40 (3), 263-284. doi:10.1051/apido/2009027
De la Rúa, P., Jiménez, Y., Galián, J., & Serrano, J. (2004). Evaluation of the biodiversity of honey bee (Apis mellifera) populations from eastern Spain. Journal of Apicultural Research, 43 (4), 162-166.
De la Rua, P., Serrano, J., & Galian, J. (1998). Mitochondrial DNA variability in the Canary Islands honeybees (Apis mellifera L.). Mol Ecol, 7, 1543-1547.
de Peer, Y. V. (2009). Phylogenetic inference based on distance methods. In P. Lemey, M. Salemi, & A.-M. Vandamme (Eds.), The phylogenetic handbook: A practical approach to phylogenetic analysis and hypothesis testing. Cambridge: Cambridge University Press.
DeBry, R. W., & Abele, L. G. (1995). The relationship between parsimony and maximum-likelihood analyses: Tree scores and confidence estimates for three real data Sets. Mol Biol. Evol., 12 (2), 291-297.
Dietemann, V., Lubbe, A., & Crewe, R. M. (2006). Human factors facilitating the spread of a parasitic honey bee in South Africa. Journal of Economic Entomology, 99 (1), 7-13 doi:http://dx.doi.org/10.1603/0022-0493(2006)099[0007:HFFTSO]2.0.CO;2
Dietemann, V., Nazzi, F., Martin, S. J., Anderson, D. L., Locke, B., Delaplane, K. S., . . . Ellis, J. D. (2013). Standard methods for varroa research. Journal of Apicultural Research, 52 (1). doi:10.3896/IBRA.1.52.1.09
Dietemann, V., Pirk, C. W. W., & Crewe, R. (2009). Is there a need for conservation of honeybees in Africa? Apidologie, 40 (3), 285-295. doi:10.1051/apido/2009013
Drake, N. A., Blench, R. M., Armitage, S. J., Bristow, C. S., & White, K. H. (2011). Ancient watercourses and biogeography of the Sahara explain the peopling of the desert. Proc Natl Acad Sci U S A, 108 (2), 458-462. doi:10.1073/pnas.1012231108
Drake, N. A., & Bristow, C. (2006). Shorelines in the Sahara: geomorphological evidence for an enhanced monsoon from palaeolake Megachad. The Holocene, 16, 901-911.
Dutton, R. W., Ruttner, F., Berlkeley, A., & Manley, M. J. D. (1981). Observations on the morphology, relationships and ecology of Apis mellifera of Oman. Journal of Apicultural Research, 20 (4), 201-214.
Eardley, C. D., Gikungu, M., & Schwarz, M. P. (2009). Bee conservation in Sub-Saharan Africa and Madagascar: diversity, status and threats. Apidologie, 40 (3), 355-366. doi:10.1051/apido/2009016
Edwards, A. W. F. (2008). G. H. Hardy (1908) and Hardy–Weinberg Equilibrium. Genetics, 179, 1143-1150.
El-Niweiri, M. A. A., El-Sarrag, M. S., & Neumann, P. (2008). Filling the Sudan gap:the northernmost natural distribution limit of small hive beetles. Journal of Apicultural Research and Bee world, 47 (3), 184–185. doi:10.3827/IBRA.1.47.3.02
El-Niweiri, M. A. A., & Moritz, R. F. A. (2008). Mitochondrial discrimination of honeybees (Apis mellifera) of Sudan. Apidologie, 39 (5), 566-573. doi:10.1051/apido:2008039
El Shafie, H. A. F., Mogga, J. B. B., & Basedow, T. (2002). Studies on the possible competition for pollen between the honey bee, Apis mellifera sudanensis, and the imported dwarf honey bee Apis florea (Hym., Apidae) in North-Khartoum (Sudan). Journal of Applied Entomology, 126, 557–562. doi: 10.1046/j.1439-0418.2002.00711.x
Ellis, J. D., & Munn, P. A. (2005). The worldwide health status of honey bees. Bee World, 86 (4), 88-101.
Engel, M. S. (1999). The taxonomy of recent and fossil honey bees (Hymenoptera: Apidae; Apis). ]. Hym. Res., 8 (2), 165-196.
Estoup, A., Garnery, L., Solignac, M., & Cornuet, J.-M. (1995). 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., Solignac, M., Harry, H., & Cornuet, J.-M. (1993). Characterization of (GT)n and (CT)n microsatellites in two insect species: Apis mellifera and Bombus terrestris. Nucleic Acids Research, 21, 1427-1431.
Estoup, A., Tailliez, C., Cornuet, J.-M., & Solignac, M. (1995). Size homoplasy and mutational processes of interupted microsatellites in two bee species, Apis mellifera and Bombus terrestris (Apidae). Mol. Biol. Evol., 12 (6), 1074-1084. Retrieved from http://mbe.oxfordjournals.org/content/12/6/1074.full.pdf
Excoffier, L., Laval, G., & Schneider, S. (2005). Arlequin (version 3.0): An integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online, 1, 47-50.
Falush, D., Stephens, M., & Pritchard, J. K. (2003). Inference of Population Structure Using Multilocus Genotype Data: Linked Loci and Correlated Allele Frequencies. Genetics, 164, 1567–1587. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1462648/pdf/12930761.pdf
Felsenstein, J. (1985). Confidence limits on phylogenies: An approach using the bootstrap. Evolution, 39, 783-791.
Fletcher, D. J. C. (1978). The African Bee, Apis mellifera adansonii, in Africa. Annu Rev Entomol, 23 (1), 151-171. doi:10.1146/annurev.en.23.010178.001055
Franck, P., Garnery, L., Celebrano, G., Solignac, M., & Cornuet, J. M. (2000). Hybrid origins of honeybees from Italy (Apis mellifera ligustica) and Sicily (A. m. sicula). Mol Ecol, 9, 907-921.
Franck, P., Garnery, L., Loiseau, A., Oldroyd, B. P., Hepburn, H. R., Solignac, M., & Cornuet, J. M. (2001). Genetic diversity of the honeybee in Africa: Microsatellite and mitochondrial data. Heredity (Edinb), 86, 420-430.
Franck, P., Garnery, L., Solignac, M., & Cornuet, J.-M. (2000). Molecular confirmation of a fourth lineage in honeybees from the Near East. Apidologie, 31, 167–180.
Franck, P., Garnery, L., Solignac, M., & Cornuet, J. M. (1988). The origin of western European subspecies of honeybees (Apis mellifera): New insights from microsatellite and mitochondrial data. Evolution, 52 (4), 1119-1134.
Francoy, T. M., Wittmann, D., Drauschke, M., Steinhage, V., Müller, S., Cunha, D. R., . . . Gonçalves, L. S. (2009). Morphometric and genetic changes in a population of Apis mellifera after 34 years of Africanization. Genet. Mol. Res., 8 (2), 709-717.
Garnery, L., Cornuet, J. M., & Solignac, M. (1992). Evolutionary history of the honey bee Apis mellifera inferred from mitochondrial DNA analysis. Mol Ecol, 1, 145-154.
Garnery, L., Franck, P., Baudry, E., Vautrin, D., Cornuet, J.-M., & Solignac, M. (1998). Genetic diversity of the west European honey bee (Apis mellifera mellifera and A. m. iberica): 1. Mitochondrial DNA. Genet. Sel. Evol., 30 (Suppl 1), S31-S47.
Garnery, L., Mosshine, E. H., Oldroyd, B. P., & Cornuet, J. M. (1995). Mitochondrial DNA variation in Moroccan and Spanish honey bee populations. Mol Ecol, 4, 465-471.
Garnery, L., Solignac, M., Celebrano, G., & Cornuet, J. M. (1993). A simple test using restricted PCR-amplifled mitochondrial DNA to study the genetic structure of Apis mellifera L. Experimentia, 49, 1016-1021.
Ghienne, J. F., Schuster, M., Bernard, A., Duringer, P., & Brunet, M. (2002). The Holocene giant Lake Chad revealed by digital elevation models. Quaternary International, 87, 81–85.
Gruber, K., Schoning, C., Otte, M., Kinuthia, W., & Hasselmann, M. (2013). Distinct subspecies or phenotypic plasticity? Genetic and morphological differentiation of mountain honey bees in East Africa. Ecol Evol, 3 (10), 3204-3218. doi:10.1002/ece3.711
Guzmán-Novoa, E., Eccles, L., Calvete, Y., Mcgowan, J., Kelly, P. G., & Correa-Benítez, A. (2010). Varroa destructor is the main culprit for the death and reduced populations of overwintered honey bee (Apis mellifera) colonies in Ontario, Canada. Apidologie, 41, 443–450. doi:10.1051/apido/2009076
Haddad, N., Fuchs, S., Hepburn, H. R., & Radloff, S. E. (2009). Apis florea in Jordan: Source of the founder population. Apidologie, 40 (4), 508-512. doi:10.1051/apido/2009011
Haddad, N., Meixner, M. D., Fuchs, S., Migdadi, H., Garnery, L., & Sheppard, W. S. (2009). Mitochondrial DNA support for genetic reserves of Apis mellifera syriaca in Jordan. Journal of Apicultural Research and Bee world, 48 (1), 19-22. doi:10.3896/ibra.1.48.1.05
Hall, B. G. (2013). Building phylogenetic trees from molecular data with mega. Mol Biol. Evol., 30 (5), 1229–1235. doi:10.1093/molbev/mst012
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 amplifcation of a locus with RFLP allele specifc to African honey bees. . Biochem. Genet., 36, 351-361.
Hall, H. G., Zettel-Nalen, C., & Ellis, J. D. (2014). African honey bee: What you need to know: Fact Sheet ENY-114 (MG113), Entomology and Nemotology Department, University of Florida/IFAS Extension.
Han, F., Wallberg, A., & Webster, M. T. (2012). From where did the Western honeybee (Apis mellifera) originate? Ecol Evol, 2 (8), 1949-1957. doi:10.1002/ece3.312
Hedrick, P. W. (2000). Genetics of Populations (2nd ed.). Boston: Jones and Bartlett.
Hepburn, H. R., & Radloff, S. E. (1997). Biogeographical correlates of population variance in the honeybees (Apis mellifera L.) of Africa. Apidologie, 28, 243-258.
Hepburn, H. R., & Radloff, S. E. (1998). Honeybees of Africa. Berlin, Heidelberg: Springer-Varlag.
Hepburn, H. R., Radloff, S. E., & Oghiakhe, S. (2000). Mountain honeybees of Africa. Apidologie, 31, 205-221.
Hewitt, G. (2000). The genetic legacy of the Quaternary ice ages. Nature, 405, 907-913.
Hood, W. M. (2000). Overview of the small hive beetle, Aethina tumida, in North America. Bee World, 81 (3), 129–137.
Hoyle, A. C., Keay, R. W. J., Mendonca, F. A., & Pichi-Sermolli, R. E. G. (Cartographer). (1958). Vegetation map of Africa, south of the tropic of cancer
Huxley, J. (1938). Clines: An auxiliary taxonomic principle. Nature, 142, 219-220.
Huxley, J. (1939). Clines: An auxiliary method in taxonomy. Bijdragen for de Dierkunde, 27, 491.
Jang, J. P., Yılız, M. A., Fakhri, B., & Nobakht, A. (2011). A study of the diversity in COI-COII intergenic region of mitochondrial DNA in different Persian honeybee (A. mellifera meda) populations. J. Basic. Appl. Sci. Res., 1 (11), 2150-2154.
Jensen, A. B., Palmer, K. A., Boomsma, J. J., & Pedersen, B. V. (2005). Varying degrees of Apis mellifera ligustica introgression in protected populations of the black honeybee, Apis mellifera mellifera, in northwest Europe. Mol Ecol, 14 (1), 93-106. doi:10.1111/j.1365-294X.2004.02399.x
Jones, J. C., Myerscough, M. R., Graham, S., & Oldroyd, B. P. (2004). Honey Bee Nest Thermoregulation: Diversity Promotes Stability. Science, 305 (5682), 402-404 doi:DOI: 10.1126/science.1096340
Kandemir, I., Kence, M., Sheppard, W. S., & Kence, A. (2006). Mitochondrial DNA variation in honey bee (Apis mellifera L.) populations from Turkey. Journal of Apicultural Research and Bee world, 45 (1), 33-38.
Kandemir, I., Meixner, M. D., Ozkan, A., & Sheppard, W. S. (2006). Genetic characterization of honey bee (Apis mellifera cypria) populations in northern Cyprus. Apidologie, 37 (5), 547-555. doi:10.1051/apido:2006029
Kence, M., Farhoud, H. J., & Tunca, R. I. (2009). Morphometric and genetic variability of honey bee (Apis mellifera L.) populations from northern Iran. Journal of Apicultural Research, 48 (4), 247-255. doi:10.3896/ibra.1.48.4.04
Kotthoff, U., Wappler, T., Engel, M. S., & Ali, J. (2013). Greater past disparity and diversity hints at ancient migrations of European honey bee lineages into Africa and Asia. Journal of Biogeography, 40, 1832–1838. doi:10.1111/jbi.12151
Kraus, F. B., Franck, P., & Vandame, R. (2007). Asymmetric introgression of African genes in honeybee populations (Apis mellifera L.) in Central Mexico. Heredity (Edinb), 99 (2), 233-240. doi:10.1038/sj.hdy.6800988
Krend, K., & Murphy, C. (2003). The Relationship of Apis mellifera with exotic and native plants in Boulder County, Colorado. American Journal of Undergraduate Research, 2 (2), 5-12.
Li, J., Qin, H., Wu, J., Sadd, B. M., Wang, X., Evans, J. D., . . . Chen, Y. (2012). The prevalence of parasites and pathogens in Asian honeybees Apis cerana in China. PLoS One, 7 (11), e47955. doi:10.1371/journal.pone.0047955
Lord, W. G., & Nagi, S. K. (1987). Apis florea discovered in Africa. Bee World, 68 (2), 39-40.
Loucif-Ayad, W., Achou, M., Legout, H., Alburaki, M., & Garnery, L. (2015). Genetic assessment of Algerian honeybee populations by microsatellite markers. Apidologie, 46 (3), 392-402. doi:10.1007/s13592-014-0331-0
Majeed, Q. (2000). Evaluation of some natural products for controlling mites of Varroa sp. parasitizing honeybees in Sokoto. Shashpa, 7 (1), 49-52
Matheson, A. (1993). World bee health report. Bee World, 74 (4), 176-212.
McMichael, M., & 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., 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. D. (1994). Analyse polymorpher Subspezies von Apis mellifera L.: Morphometrische und molekulare Untersuchungen an den europäischen Rassen Apis mellifera carnica und ligustica und den afrikanischen Rassen Apis mellifera monticola und scutellata. Johann-Wolfgang-Goethe- University, Frankfurt.
Meixner, M. D., Büchler, R., Costa, C., Francis, R. M., Hatjina, F., Kryger, P., . . . Carreck, N. L. (2014). Honey bee genotypes and the environment. Journal of Apicultural Research, 53 (2), 183-187. doi:10.3896/ibra.1.53.2.01
Meixner, M. D., Costa, C., Kryger, P., Hatjina, F., Bouga, M., Ivanova, E., & Büchler, R. (2010). Conserving diversity and vitality for honey bee breeding. Journal of Apicultural Research, 49 (1), 85-92. doi:10.3896/ibra.1.49.1.12
Meixner, M. D., Leta, M. A., Koeniger, N., & Fuchs, S. (2011). The honey bees of Ethiopia represent a new subspecies of Apis mellifera: Apis mellifera simensis n. ssp. Apidologie, 42 (3), 425-437. doi:10.1007/s13592-011-0007-y
Meixner, M. D., Pinto, M. A., Bouga, M., Kryger, P., Ivanova, E., & Fuchs, S. (2013). Standard methods for characterising subspecies and ecotypes of Apis mellifera. Journal of Apicultural Research, 52 (4), 1-28. doi:10.3896/ibra.1.52.4.05
Miguel, I., Iriondo, M., Garnery, L., Sheppard, W. S., & Estonba, A. (2007). Gene flow within the M evolutionary lineage ofApis mellifera: role of the Pyrenees, isolation by distance and post-glacial re-colonization routes in the western Europe. Apidologie, 38 (2), 141-155. doi:10.1051/apido:2007007
Mogga, J., & Ruttner, F. (1988). Apis florea in Africa: Source of the founder population. Bee World, 69 (2), 100-103.
Moritz, R. F. A., Härtel, S., & Neumann, P. (2005). Global invasions of the western honeybee (Apis mellifera) and the consequences for biodiversity. Ecoscience, 12 (3), 289-301. doi:10.2980/i1195-6860-12-3-289.1
Moritz, R. F. A., Hawkins, C. F., Crozier, R. H., & Mackinley, A. G. (1986). A mitochondrial DNA polymorphism in honeybees (Apis mellifera L.). Experimentia, 42 (322-324).
Moritz, R. F. A., Kraus, F. B., Kryger, P., & Crewe, R. M. (2007). The size of wild honeybee populations (Apis mellifera) and its implications for the conservation of honeybees. Journal of Insect Conservation, 11 (4), 391-397. doi:10.1007/s10841-006-9054-5
Muli, E., Patch, H., Frazier, M., Frazier, J., Torto, B., Baumgarten, T., . . . Grozinger, C. (2014). Evaluation of the distribution and impacts of parasites, pathogens, and pesticides on honey bee (Apis mellifera) populations in East Africa. PLoS One, 9 (4), e94459. doi:10.1371/journal.pone.0094459
Mumbi, C. T., Mwakatobe, A. R., Mpinga, I. H., Richard, A., & Machumu, R. (2014). Parasitic mite, Varroa species (Parasitiformes: Varroidae) infesting the colonies of African honeybees, Apis mellifera scutellata (Hymenoptera: Apididae) in Tanzania. Journal of Entomology and Zoology Studies, 2 (3), 188-196.
Muñoz, I., Dall′Olio, R., Lodesani, M., De la Rúa, P., Schonrogge, K., & Brady, S. (2014). Estimating introgression in Apis mellifera siciliana populations: are the conservation islands really effective? Insect Conservation and Diversity, 7 (6), 563-571. doi:10.1111/icad.12092
Muñoz, I., & De la Rúa, P. (2012). Temporal analysis of the genetic diversity in a honey bee mating area of an island population (La Palma, Canary Islands, Spain). Journal of Apicultural Science, 56 (1), 41-49. doi:10.2478/v10289-012-0005-y
Muñoz, I., Stevanovic, J., Stanimirovic, Z., & De la Rúa, P. (2012). Genetic variation of Apis mellifera from Serbia inferred from mitochondrial analysis. Journal of Apicultural Science, 56 (1), 59-69. doi:10.2478/v10289-012-0007-9
Murilhas, A. M. (2005). Aethina tumida arrives in Portugal. Will it be eradicated? EurBee Newsletter No. 2 pp. 7-9.
Mutinelli, F., Montarsi, F., Federico, G., Granato, A., Ponti, A. M., Grandinetti, G., . . . Chauzat, M.-P. (2014). Detection of Aethina tumida Murray (Coleoptera: Nitidulidae.) in Italy: outbreaks and early reaction measures. Journal of Apicultural Research, 53 (5), 569-575. doi:10.3896/ibra.1.53.5.13
Nedić, N., Francis, R. M., Stanisavljević, L., Pihler, I., Kezić, N., Bendixen, C., & Kryger, P. (2014). Detecting population admixture in honey bees of Serbia. Journal of Apicultural Research, 53 (2), 303-313. doi:10.3896/ibra.1.53.2.12
Nedic, N., Stanisavljevic, L., Mladenovic, M., & Stanisavljevic, J. (2009). Molecular characterization of the honeybee Apis mellifera carnica in Serbia. Archives of Biological Sciences, 61 (4), 587-598. doi:10.2298/abs0904587n
Nei, M., & Kumar, S. (2000). Molecular Evolution and Phylogenetics.. New York: Oxford University Press
Neumann, P., & Carreck, N. L. (2010). Honey bee colony losses. Journal of Apicultural Research, 49 (1), 1-6. doi:10.3896/IBRA.1.49.1.01
Neumann, P., & Ellis, J. (2008). Guest Editorial: The small hive beetle (Aethina tumida Murray, Coleoptera: Nitidulidae) distribution, biology and controlof an invasive species. Journal of Apicultural Research, 47 (3), 181-183. doi:10.3827/ibra.1.47.3.01
Nightingale, J. M. (1976). Traditional beekeeping amaong Kenyan tribes and methods proposed for improvement and modernisation. Paper presented at the First International Conference on Apiculture in Tropical Climates, London.
Nikolova, S. (2011). Genetic variability of local Bulgarian honey bees Apis mellifera macedonica (rodopica) based on microsatellite DNA analysis. Journal of Apicultural Science, 55 (2), 117-129.
Nikolova, S. R., & Ivanova, E. N. (2012). Genetic variability in a local Bulgarian honey bee population. Acta zool. bulg., 64 (1), 199-204.
Oleksa, A., Chybicki, I., Tofilski, A., & Burczyk, J. (2011). Nuclear and mitochondrial patterns of introgression into native dark bees (Apis mellifera mellifera) in Poland. Journal of Apicultural Research, 50 (2), 116-129. doi:10.3896/ibra.1.50.2.03
Oleksa, A., & Tofilski, A. (2014). Wing geometric morphometrics and microsatellite analysis provide similar discrimination of honey bee subspecies. Apidologie, 46 (1), 49-60. doi:10.1007/s13592-014-0300-7
Opperdoes, F. (1997). Phenetics versus Cladistics and the pro's and con's of the various phylogeny inference methods. Retrieved from http://www.icp.ucl.ac.be/~opperd/private/phenetics.html
Oyerinde, A. A., Dike, M. C., Banwo, O. O., Bamaiyi, L. J., & Adamu, R. S. (2012). Morphometric and landmark based variations of Apis mellifera L. wings in the savannah agro - ecological zone of Nigeria. Global Journal of Science Frontier Research(D), 22 (1), 33-41.
Özdïl, F., Yildiz, M. A., & Hall, H. G. (2009). Molecular characterization of Turkish honey bee populations (Apis mellifera) inferred from mitochondrial DNA RFLP and sequence results. Apidologie, 40 (5), 570-576. doi:10.1051/apido/2009032
Palmer, M. R., Smith, D. R., & Kaftanoglu, O. (2009). Turkish honeybees: Genetic variation and evidence for a fourth lineage of Apis mellifera mtDNA. J Hered, 91 (1), 42-66.
Palmeri, V., Scirtò, G., Malacrinò, A., Laudani, F., & Campolo, O. (2014). A scientific note on a new pest for European honeybees: first report of small hive beetle Aethina tumida, (Coleoptera: Nitidulidae) in Italy. Apidologie. Retrieved from http://link.springer.com/article/10.1007%2Fs13592-014-0343-9
Paton, D. C. (1996). Overview of feral and managed honeybees in Australia: Distribution, abundance, extent of interactions with native biota, evidence of impacts and future research. Retrieved from
Peakall, R., & Smouse, P. E. (2012). GenAlEx 6.5: genetic analysis in Excel. Population genetic software for teaching and research – an update. Bioinformatics, 28, 2537-2539.
Penny, D. (1982). Towards a basis for classification: the incompleteness of distance measures, incompatibility analysis, and phenetic classification. J. Theor. Biol., 96, 129- 142.
Pinto, M. A., Henriques, D., Chávez-Galarza, J., Kryger, P., Garnery, L., van der Zee, R., . . . Johnston, J. S. (2014). Genetic integrity of the Dark European honey bee (Apis mellifera mellifera) from protected populations: a genome-wide assessment using SNPs and mtDNA sequence data. Journal of Apicultural Research, 53 (2), 269-278. doi:10.3896/ibra.1.53.2.08
Pinto, M. A., Muñoz, I., Chávez-Galarza, J., & De la Rúa, P. (2012). The Atlantic side of the Iberian Peninsula: a hot-spot of novel African honey bee maternal diversity. Apidologie, 43 (6), 663-673. doi:10.1007/s13592-012-0141-1
Pinto, M. A., Sheppard, W. S., Schiff, N. M., Johnston, J. S., Rubink, W. L., Coulson, R. N., . . . Patton, J. C. (2007). Honey bees (Hymenoptera: Apidae) of African origin exist in non-Africanized areas of the southern United States: Evidence from mitochondrial DNA. Ann. Entomol. Soc. Am., 100 (2), 289-295.
Prada, C. F. Q., Duran, J. T., Salamanca, G. G., & Del Lama, M. A. (2009). Population genetics of Apis mellifera L. (Hymenoptera: Apidae) from Colombia. Journal of Apicultural Research, 48 (1), 3-10. doi:10.3896/ibra.1.48.1.02
Pritchard, J. K., Stephens, M., & Donnelly, P. (2000). Inference of population structure using multilocus genotype data. Genetics, 155, 945–959.
QIAGEN. (2006a). DNeasy Blood & Tissue Handbook: QIAGEG.
QIAGEN. (2006b). QIAGEN Supplementary Protocol: Purification of total DNA from insects using the DNeasy Blood & Tissue Kit QIAGEN.
Radloff, S., & Hepburn, R. (2000). Population structure and morphometric variance of the Apis mellifera scutellata group of honeybees in Africa. Genetics and Molecular Biology, 23 (2), 305-316.
Radloff, S. E., & Hepburn, H. R. (1997). Multivariate analysis of honey bees, Apis mellifera Linnaeus (Hymenoptera: Apidae) from western central Africa. . African Entomology, 5, 195–204.
Radloff, S. E., & Hepburn, H. R. (1998). The matter of sampling distance and confidence levels in the subspecific classification of honeybees, Apis mellifera L. Apidologie, 29 (6), 491-501.
Radloff, S. E., Hepburn, H. R., & Engel, M. S. (2011). The Asian Species of Apis. In H. R. Hepburn & S. E. Radloff (Eds.), Honeybees of Asia. Berlin, Heidelberg: Springer-Verlag.
Rahimi, A., Mirmoayedi, A., Kahrizi, D., Abdolshahi, R., Kazemi, E., & Yari, K. (2014). Microsatellite genetic diversity of Apis mellifera meda skorikov. Mol Biol Rep, 41 (12), 7755-7761. doi:10.1007/s11033-014-3667-7
Rashad, S. E., & El-Sarrag, M. S. (1980). Some characters of the Sudanese honeybee Apis mellifera L. Paper presented at the Second International Conference on Apiculture in Tropical Climates, New Delhi.
Rasolofoarivao, H., Clémencet, J., Techer, M. A., Ravaomanarivo, L. H. R., Reynaud, B., & Delatte, H. (2015). Genetic diversity of the endemic honeybee: Apis mellifera unicolor (Hymenoptera: Apidae) in Madagascar. Apidologie. doi:10.1007/s13592-015-0362-1
Raymond, M., & Rousset, F. (1995). An exact test for population differentiation. Evolution, 49, 1280-1283.
Rowe, D. J., Rinderer, T. E., Stelzer, J. A., Oldroyd, B. P., & Crozier, R. H. (1997). Seven polymorphic microsatellite loci in honeybees. Insectes Sociaux, 44, 85-93.
Ruttner, F. (1975). African races of honeybees. Paper presented at the The 25th International Beekeeping Congress, Grenoble, France.
Ruttner, F. (1982). On the taxonomy of honeybees of tropical Africa. Apiacta 1. Retrieved from http://www.apimondiafoundation.org/cgi-bin/index.cgi?sid=&zone=cms&action=search&categ_id=146&search_ordine=descriere
Ruttner, F. (1985). Graded geographic variability in honebees and environment. Pszczel. Zesz. Nauk., 29, 81-92.
Ruttner, F. (1988). Biogeography and taxonomy of honeybees. Berlin, Heidelberg, New York: Springer-Verlag.
Ruttner, F., & Maul, V. (1983). Experimental analysis of reproductive interspecies isolation of Apis mellifera L. and Apis cerana Fabr. Apidologie, 14 (4), 309-327.
Ruttner, F., Tassencourt, L., & Louveaux, J. (1978). Biometrical-statistical analysis of the geographic variability of Apis mellifera L.: I. Material and methods. Apidologie, 9 (4), 363-381.
Sainudiin, R., Durrett, R. T., Aquadro, C. F., & Nielsen, R. (2004). Microsatellite mutation models: insights from a comparison of humans and chimpanzees. Genetics, 168 (1), 383-395. doi:10.1534/genetics.103.022665
Saitou, N., & Nei, M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol., 4, 406-425.
Schiff, N. M., Sheppard, W. S., Loper, G. M., & Shimanuki, H. (1994). Genetic Diversity of Feral Honey Bee (Hymenoptera: Apidae) Populations in the Southern United States. Annals of the Entomological Society of America, 87 (6), 842-848. doi:10.1093/aesa/87.6.842
Schneider, S. S., DeGrandi-Hoffman, G., & Smith, D. R. (2004). The African honey bee: factors contributing to a successful biological invasion. Annu Rev Entomol, 49, 351-376. doi:10.1146/annurev.ento.49.061802.123359
Schuster, M., Roquin, C., Duringer, P., Brunet, M., Caugy, M., Fontugne, M., . . . Ghienne, J.-F. (2005). Holocene lake Mega-Chad palaeoshorelines from space. Quaternary Science Reviews, 24, 1821-1827.
Segura, J. A. L. (2000). Highly polymorphic DNA markers in an Africanized honey bee population in Costa Rica. Genetics and Molecular Biology, 23 (2), 317-322.
Shaibi, T., Muñoz, I., Dall′Olio, R., Lodesani, M., la Rúa, P., & Moritz, R. F. A. (2009). Apis mellifera evolutionary lineages in Northern Africa: Libya, where orient meets occident. Insectes Sociaux, 56 (3), 293-300. doi:10.1007/s00040-009-0023-3
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., & Meixner, M. D. (2003). Apis mellifera pomonella, a new honey bee subspecies from Central Asia. Apidologie, 34 (4), 367-375. doi:10.1051/apido:2003037
Shi, W. (2001). Microsatellite investigations of effective mating frequencies in honey bees (Apis mellifera) from different locations of Kenya. Paper presented at the 37th Int. Apic. Congr., Durban, South Africa.
Smith, D. R. (1991). African Bees in the Americas: Insights from Biogeography and Genetics. Tree, 6 (1), 17-21.
Smith, D. R., Slaymaker, A., Palmer, M., & Kaftanoglu, O. (1997). Turkish honey bees belong to the east Mediterranean mitochondrial lineage. Apidologie, 28, 269-274.
Smith, F. G. (1961). The races of honeybees in Africa. Bee World, 42, 255–260.
Solignac, M., Vautrin, D., Loiseau, A., Mougel, F., Baudry, E., Estoup, A., . . . Cornuet, J.-M. (2003). Five hundred and fifty microsatellite markers for the study of the honeybee (Apis mellifera L.) genome. Molecular Ecology Notes, 3 (2), 307-311. doi:10.1046/j.1471-8286.2003.00436.x
Solorzano, C. D., Szalanski, A. L., Kence, M., McKern, J. A., Austin, J. W., & Kence, A. (2009). Phylogeography and Population Genetics of Honey Bees (Apis mellifera) From Turkey Based on COI-COII Sequence Data. Sociobiology, 53 (1), 237-246.
Sourdis, J., & Nei, M. (1988). Relative efficiencies of the maximum parsimony and distance-matrix methods in obtaining the correct phylogenetic tree. Mol. Biol. Evol., 5 (3), 298-311.
Strauss, U., Pirk, C. W. W., Crewe, R. M., Human, H., & Dietemann, V. (2014). Impact of Varroa destructor on honeybee (Apis mellifera scutellata) colony development in South Africa. Experimental and Applied Acarology, 65 (1), 89-106 doi:10.1007/s10493-014-9842-7
Susnik, S., Kozmus, P., Poklukar, J., & Meglic, V. (2004). Molecular characterisation of indigenous Apis mellifera carnica in Slovenia. Apidologie, 35 (6), 623-636. doi:10.1051/apido:2004061
Szalanski, A. L., & Magnus, R. M. (2010). Mitochondrial DNA characterization of Africanized honey bee (Apis mellifera L.) populations from the USA. Journal of Apicultural Research, 49 (2), 177-185. doi:10.3896/ibra.1.49.2.06
Tamura, K., & 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.
Tamura, K., Nei, M., & Kumar, S. (2004). Prospects for inferring very large phylogenies by using the neighbor-joining method. Proceedings of the National Academy of Sciences (USA) 101:11030-11035
Tamura, K., Stecher, G., Peterson, D., Filipski, A., & Kumar, S. (2013). MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Mol Biol Evol., 30 (12), 2725-2729. doi:10.1093/molbev/mst197
Tarpy, D. R. (2003). Genetic diversity within honeybee colonies prevents severe infections and promotes colony growth. Proc Biol Sci, 270 (1510), 99-103. doi:10.1098/rspb.2002.2199
Techer, M. A., Clémencet, J., Simiand, C., Portlouis, G., Reynaud, B., & Delate, H. (2015). Genetic diversity of the honeybee (Apis mellifera L.) populations in the Seychelles archipelago. Insect Conservation and Diversity. doi:10.1111/icad.12138
Techer, M. A., Clémencet, J., Turpin, P., Volbert, N., Reynaud, B., & Delatte, H. (2015). Genetic characterization of the honeybee (Apis mellifera) population of Rodrigues Island, based on microsatellite and mitochondrial DNA. Apidologie, 46 (4), 445-454. doi:10.1007/s13592-014-0335-9
Tofilski, A. (2004). DrawWing, a program for numerical description of insect wings. Journal of Insect Science, 4, 17.
Tofilski, A. (2008). Using geometric morphometrics and standard morphometry to discriminate three honeybee subspecies. Apidologie, 39 (5), 558-563. doi:10.1051/apido:2008037
Uzunov, A., Meixner, M. D., Kiprijanovska, H., Andonov, S., Gregorc, A., Ivanova, E., . . . Kryger, P. (2014). Genetic structure of Apis mellifera macedonica in the Balkan Peninsula based on microsatellite DNA polymorphism. Journal of Apicultural Research, 53 (2), 288-295. doi:10.3896/ibra.1.53.2.10
vanEngelsdorp, D., Hayes Jr., J., Underwood, R. M., & Pettis, J. (2008). A survey of honey bee colony losses in the U.S., fall 2007 to spring 2008. PLoS One, 3 (12), e4071. doi:10.1371/journal.pone.0004071
vanEngelsdorp, D., Hayes Jr., J., Underwood, R. M., & Pettis, J. S. (2010). A survey of honey bee colony losses in the United States, fall 2008 to spring 2009. Journal of Apicultural Research, 49 (3), 7-14. doi:10.3896/IBRA.1.49.1.03
Wallberg, A., Han, F., Wellhagen, G., Dahle, B., Kawata, M., Haddad, N., . . . Webster, M. T. (2014). A worldwide survey of genome sequence variation provides insight into the evolutionary history of the honeybee Apis mellifera. Nature Genetics, 1-10. doi:10.138/ng.3077
Whitfield, C. W., Behura, S. K., Berlocher, S. H., Clark, A. G., Johnston, J. S., Sheppard, W. S., . . . Tsutsui, N. D. (2006). Thrice Out of Africa: Ancient and Recent Expansions of the Honey Bee, Apis mellifera. Science, 314, 642-645. Retrieved from www.sciencemag.org
Wilson, E. O. (1971). The insect societies. Harvard: Belknap Press.
Woyke, J. (1993). Some behavioural characteristics of the Sudanese honey bee. Bee World, 74 (3), 133-140.
Wright, S. (1978). Evolution and the Genetics of Populations. Variability within and among natural populations. (Vol. 4). Chicago: The University of Chicago Press.
Yang, Z. (1996). Phylogenetic analysis using parsimony and likelihood methods. J Mol Evol, 42, 294–307.
Yu, L., Xie, W., Wu, H., Zou, Y., Nan, Q., Zhu, L., . . . Wu, Q. (2012). Morphological characteristics and microsatellite DNA genetic diversity of Nigeria African honey bee, Anhui Apis mellifera and their hybrid generation II. Acta Ecologica Sinica, 32 (11), 3555-3564. doi:10. 5846 / stxb201104060442
Zaitoun, S., Hassawi, D. S., & Shahrour, W. (2008). Origin of Jordanian honeybees Apis mellifera (Hymenoptera: Apidae) using amplified mitochondrial DNA. Eur. J. Entomol., 105 (41-44).
Appendix I
Subspecies of Apis mellifera L, based on Ruttner (1988), Engels (1999), Sheppard and Meixner (2003) and Meixner et al (2011).
Abbildung in dieser Leseprobe nicht enthalten
Appendix II
Table A1: Means and standard deviations of morphological characters of body hair (mm) and pigmentation of 10 worker bees each from (N) colonies from 21 localities.
Abbildung in dieser Leseprobe nicht enthalten
*Significant (p < 0.05); ***highly significant (p < 0.001); †not significant (p > 0.05); according to one-way ANOVA. Means within a column followed by the same letter are not significantly different at p = 0.05 according to Tukey HSD test. Hair: Cover hair on tergite 5; Tom1: Width of tomentum; Tom2: Width of stripe behind tomentum; Toind: Tomentum index; Pt2,3 & 4: Pigmentation of tergites 2, 3 & 4; Scut1,2: pigmentation of scutellum and its plates. Numbers in brackets are Ruttner numbers.
Table A2: Means and standard deviations (mm) of characters of the hind leg of 10 worker bees each from (N) colonies from 21 localities in western Africa.
Abbildung in dieser Leseprobe nicht enthalten
***Highly significant (p < 0.001), according to one-way ANOVA. Means, within a column, followed by the same letter are not significantly different at p = 0.05 according to Tukey HSD test. Numbers in brackets are Ruttner numbers.
Table A3: Means and standard deviations (mm) of characters of the abdomen of 10 worker bees each from (N) colonies from 21 localities.
Abbildung in dieser Leseprobe nicht enthalten
*Significant (p < 0.05); ***highly significant (p < 0.001) and †not significant (p > 0.05), according to one-way ANOVA. Means, within a column, followed by the same letter are not significantly different at p = 0.05 according to Tukey HSD test. Lt3, 4: Tergites 3, 4, longitudinal; Lst3: Sternite 3, longitudinal; Lwm: Wax mirror, longitudinal; Wwm: Wax mirror, width; Dwm: Distance between wax mirrors; Lst6: Sternite 6, longitudinal; Wst6: Sternite 6, width. Numbers in brackets are Ruttner numbers.
Table A4: Means and standard deviations of characters of the fore-wing of 10 worker bees each from (N) colonies from 40 localities.
Abbildung in dieser Leseprobe nicht enthalten
Table 4 continued
Abbildung in dieser Leseprobe nicht enthalten
*Significant (p < 0.05); **highly significant (p < 0.01); ***Very highly significant (p < 0.001) and †not significant (p > 0.05), according to one-way ANOVA. Means, within a column, followed by the same letter are not significantly different at p = 0.05 according to Tukey HSD test. Measurements of length are in mm and of angles in degrees. Fwl: Fore-wing, length; Fww: Fore-wing, width; Cub-a, b: Cubital vein, distance a, b; A4-O26: 11 Angles of wing veins. Numbers in brackets are Ruttner numbers.
Table A5: Means and standard deviations of morphological characters of 10 worker bees each from (N) colonies collected from 21 localities (cf Table 3) in western Africa.
Abbildung in dieser Leseprobe nicht enthalten
† All measurements are in mm except those of pigmentation which are in counts.
Table A6: Means and standard deviations of characters of the fore-wing of 10 worker bees each from (N) colonies collected from 40 localities in western Africa.
Abbildung in dieser Leseprobe nicht enthalten
† Measurements of distance are in mm and of angles in degrees.
Table A7: Rotated component matrix for PCA, with Varimax rotation, of morphometric characters of A. mellifera colonies from western Africa.
Abbildung in dieser Leseprobe nicht enthalten
† Major loadings for each character are shown in bold.
Table A8: Proximity (dissimilarity) matrix of 23 populations of A. mellifera from western Africa in squared Euclidean distance.
Abbildung in dieser Leseprobe nicht enthalten
Table A9: Pooled within-groups correlations between discriminating variables and standardized canonical discriminant functions.
Abbildung in dieser Leseprobe nicht enthalten
† The largest absolute correlation between each variable and any discriminant function is shown in bold.
Table A10: Proximity (dissimilarity) matrix of four populations of A. mellifera from western Africa in squared Euclidean distance.
Abbildung in dieser Leseprobe nicht enthalten
Table A11: Pooled within-groups correlations between discriminating variables and standardized canonical discriminant functions.
Abbildung in dieser Leseprobe nicht enthalten
† The largest absolute correlation between each variable and any discriminant function is shown in bold.
Table A12: Rotated component matrix for PCA, with Varimax rotation, of morphometric characters of A. mellifera colonies.
Abbildung in dieser Leseprobe nicht enthalten
† Major loadings for each character are shown in bold.
Table A13: Pooled within-groups correlations between discriminating variables and standardized canonical discriminant functions.
Abbildung in dieser Leseprobe nicht enthalten
† The largest absolute correlation between each variable and any discriminant function is shown in bold.
Table A14: Association between mitochondrial DNA haplotypes of A. mellifera and environmental variables in western Africa.
Abbildung in dieser Leseprobe nicht enthalten
Table A 15: List of 164 alleles found in 15 microsatellite loci in A. mellifera in western Africa.
Abbildung in dieser Leseprobe nicht enthalten
Table A15 continued
Abbildung in dieser Leseprobe nicht enthalten
Table A15 contidued
Abbildung in dieser Leseprobe nicht enthalten
Table A16: Allele frequencies of 15 microsatellite loci in two populations of A. mellifera in western Africa.
Abbildung in dieser Leseprobe nicht enthalten
Table A16 continued
Abbildung in dieser Leseprobe nicht enthalten
†Humid = population of humid zone; Semiarid = population of semiarid zone; N = sample size.
Table A17: Frequency of private alleles of microsatellite loci in two populations of A. mellifera in western Africa.
Abbildung in dieser Leseprobe nicht enthalten
Table A17 continued
Abbildung in dieser Leseprobe nicht enthalten
Table A18 Analysis of molecular variance of microsatellite variation of two a priori populations (humid and semiarid) of A. mellifera
Source of Sum of Variance Percentage
variation d.f. squares components of variation
Among
populations 1 4.592 0.00214 Va 0.05
Within
populations 264 1139.058 4.31462 Vb 99.95
Total 265 1143.650 4.31675
Fixation Index FST : 0.00050
Significance tests (1023 permutations)
Va and FST : P(rand. value > obs. value) = 0.39296
P(rand. value = obs. value) = 0.00098
P-value = 0.39394+-0.01494
Curriculum Vitae
1. Education
i. Goethe Universität, Frankfurt am Main (2011 to date: Pursuing a PhD programme)
ii. Bayero University, Kano: M.Sc Zoology, 1989
iii. University of Sokoto: B.Sc Zoology, 1983
iv. Government Secondary School, Bauchi: West African School Certificate, 1978
v. Central Primary School Dukku: First School Leaving Certificate, 1973
2. Publications
i. Hamid, A., Abubakar, U., Dukku, U.H. and Hamid, B.L. (1987) A preliminary report on beekeeping in Sokoto State; Nigeria: Nest construction in the Dange Area. Nigerian Journal of Basic and Applied Sciences 1:1-9.
ii. Dukku, U.H., and Muhammadou, S. (1990) Performance of Rhizopertha dominaca (F) (Coleoptera: Bostrychidae) in three samples of Sorghum grains. Nigerian Journal of Basic and applied Science 4(1&2): 113-117.
iii. Dukku, U.H.(2003) Acacia ataxacantha: A nectar plant for honeybees between two dearth periods in the Sudan Savanna of Northern Nigeria. Bee World 84(1):32-34.
iv. Adamu H M , Abayeh,O J, Agho, M O, Abdullahi, A L, Uba, U, Dukku, U H and Wufem, B M.(2005) An ethnobotanical survey of Bauchi State herbal plants and their antimicrobial activity. Journal of Ethnophamacology 99:1-4
v. Adamu, S.U. and Dukku, U.H.(2009) Antischistosomal effect of seed oil of Nigella sativa on Schistosoma mansoni in mice. Biotropic Research International Journal 1 (1):50-53.
vi. Dukku, U.H.(2010) Parkia biglobosa: An important honeybee forage in the Savanna. Bee World 87(2):28-29.
vii. Dukku, U.H.(2010) Vitellaria paradoxa: An important nectar plant in the Savanna. Bee World 87(3):59-60.
viii. Dukku, U. H.(2010) Guiera senegalensis: An important bee forage in the Sudan Savanna and Sahel. Bee World 87(4):77.
ix. Dukku, U. H.(2011) Eucalyptus torelliana: An important naturalized bee plant in northern Nigeria. Bee World 88(1):16-17.
x. Dukku, U. H.(2011) Nymphaea lotus: A cosmopolitan beeplant. Bee World 88(2):38-39.
xi. Dukku, U.H.(2011) Delonix regia: An important naturalized beeplant in northern Nigeria. Bee World 88(3):71-72
xii. Dukku, U.H.(2011) Khaya senegalensis: An important naturalized bee plant in the Savanna. Bee World 88(4):92
xiii. Dukku, U. H.(2012) Trees bees use: Bombax costatum. Bees for Development Journa l 102:7
xiv. Dukku, U. H.(2013) Identification of plants visited by the honeybee, Apis mellifera L. in the Sudan Savanna zone of northeastern Nigeria. African Journal of Plant Science 7(7):273-284
xv. Dukku, U.H., Russom, Z. and Domo, A.G.(2013) Diurnal and seasonal flight activity of the honeybee, Apis mellifera L, and its relationship with temperature, light intensity and relative humidity in the Savanna of northern Nigeria. Global Journal of Science Frontier Research 13(4 Ver.1):1-6
3. Conferences
Dukku, U. H., Fuchs, S., Grünewald, B., Meixner, M. D. (2012) Genetic Diversity of Apis mellifera L. in Parts of West and Central Africa: A Preliminary Report. Proc. 59th annual meeting German bee research institutes, Bonn. 2012
Dukku, U. H., Fuchs, S., Grünewald, B., Meixner, M. D. (2014) The genetic diversity of Apis mellifera L. in western Africa based on mitochondrial DNA. Proc. 61st annual meeting German Bee Research Institutes, Marburg. 2014
Dukku, U. H., Fuchs, S., Kryger, P., Francis, R, M., Grünewald, B., Meixner, M. D.(2014) The genetic diversity of Apis mellifera L. in western Africa. Proc. 6th European Conference of Apidology, Murcia, Spain. 9 – 11 September 2014.
4. Unpublished Work
i. Evaluation of the Hadejia-Nguru Wetlands Conservation Project. Nigerian Environmental Study/Action Team, Ibadan, 1997.
ii. Prospects of Traditional Beekeeping in Bauchi State. Nigerian Environmental Study/Action Team, Ibadan, 1998.
iii. A Preliminary Environmental Impact Assessment of Specialty Metals Mining in the Udegi (Ogapa-Oto) Area of Nasarawa State. Global Minerals Limited, Abuja, 1998.
iv. A Preliminary Environmental Impact Assessment of Kaolin Mining at Major Porter in Plateau State. Global Minerals Limited, Abuja, 1998.
v. Ethnobotanical Survey of Medicinal Plants in Selected Local Government Areas of Bauchi, Gombe and Plateau States. ATBU/NIPRD, 1998.
vi. Environmental Baseline Studies of the Gongola River Basin. ATBU/FEPA, Phase I 1999.
vii. Environmental Baseline Studies of the Gongola River Basin. ATBU/FEPA, Phase II 2002.
5. Adhoc Scientific Activities
i. Investigator/Principal Investigator, Beekeeping Research Project, Usmanu Danfodiyo University, Sokoto, 1985-91.
ii. Resource person, Expert meeting on the Conservation of Biological Diversity in the Yobe Chad Basin, Bauchi, 18th & 19th May, 1992.
iii. Resource person, International Workshop on the Preparation of a National Water Resources Management Plan for Nigeria, Abuja, 6th – 9th December, 1993.
iv. Participant, International Workshop on Beekeeping and Extension, Israel, 20th April – 11th May, 1994.
v. Principal Investigator/Officer in Charge, Beekeeping Development Project, FEPA, Bauchi, 1994-1995.
vi. Resource person at a one day Workshop for the Development of Biodiversity Conservation Action Plan for Nigeria, Yankari National Park, May, 1997, FEPA/UNDP.
vii. Team leader, need-assessment survey on tradition beekeeping in Taraba State, 1994, FEPA.
viii. Resource person, Beekeeping Training Workshop, Yankari National Park, National Parks Board/Embassy of Israel. March, 1999.
ix. Participant, Mushroom Cultivation Technology Workshop, Africa University, Mutare, Zimbabwe, UNDP/ZERI Foundation/Africa University, May, 1999.
x. Participant, Training Workshop on the management of an Integrated Biosystems Farm, Songhai Centre, Porto Novo, Republic of Benin. July, 1999.
xi. Participant, Stakeholders’ Workshop on the management of natural resources in the Hadejia-Jama’are-Kamodugu-Yobe Basin. Dutse, Jigawa State. January, 2001. DFID.
xii. Resource person, Monthly Technology Review Meeting, GSADP, Gombe, March 2001 (Subject: Beekeeping Technology).
xiii. Resource person, Monthly Technology Review Meeting, BSADP, Bauchi, October, 2001 (Subject: Beekeeping Technology).
xiv. Resource person, Monthly Technology Review Meeting, BSADP, Bauchi, March, 2002 (Subject: Beekeeping Technology).
xv. Resource person, Beekeeping Sensitization Workshop, ECWA Church, Bauchi, June 2006
xvi. Resource person, Beekeeping Training Workshop for Rural Women, Federal Ministry of Agriculture/GSADP, Gombe, December 2007.
xvii. Member, Project Steering Committee, Building Nigeria’s Response to Climate Change (BNRCC) Project, Marbek/CUSO-VSO/NEST, 2008 to 2011.
6. Record of Employment
i. Senior Lecturer, Abubakar Tafawa Balewa University, Bauchi, 1999 to date.[ Teaching and research in entomology, beekeeping, taxonomy and ecology of the honeybee; medicinal properties of native plants; team leader, environmental baseline study of the Gongola basin: development of an IBS (integrated biosystems) farm; sensitizing industries and the general public on the operation and benefits of the ZERI(Zero Emission Research Initiative) concept. ]
ii. Lecturer I, Abubakar Tafawa Balewa University, Bauchi, 1995 to 1999.[Same as above]
iii. Assistant Chief Conservation Officer, Federal Environmental Protection Agency, The Presidency, Abuja, 1992 to 1995.[ Identification of sites and species of conservation interest in Bauchi and Gombe States; establishment of a pilot beekeeping project at the Maladumba Forest Reserve in Misau L.G.A. of Bauchi State. ]
iv. From Graduate Assistant to Lecturer I, Usmanu Danfodiyo University, Sokoto, 1984 to 1991. Teaching of zoology courses; research in the traditional beekeeping of Sokoto.
v. Biology Teacher (NYSC), Community Secondary School Iwok, Akwa-Ibom State, 1983-1984.
[...]
[1] Sources: F. Ruttner (1988) and F. Ruttner et al. (1978)
[2] Figures 2.8 – 2.16 were taken from F. Ruttner (1988); F. Ruttner et al. (1978).
- Citar trabajo
- Usman Dukku (Autor), 2015, Studies on the Diversity of Apis mellifera L. in Parts of West and Central Africa, Múnich, GRIN Verlag, https://www.grin.com/document/511313
Textos similares
Así es como funciona
Comentarios