European Journal of Experimental Biology Open Access

Keywords

Domestic bee; Biometric study; Statistical analysis; Algerian north-east

Introduction

Pollinating insects play a key role in cultivated ecosystems [1,2]. Among them, the Bees who belong to the order of Hymenoptera, the superfamily of Apoidea which forms the widest and specialized group [3,4]. This specialization is the result of a long co-evolution between bees and angiosperms group which have begun after synchronous evolutionary radiation in the mid-Cretaceous [5,6]. Today, there are 11 bee families with about 700 genera and 20,000 living species [7-10]. Its abundance and density are higher in the temperate zone due to genera: Apis which is characterized by a highly social behavior and comprises four species: Apis dorsata, Apis florea, Apis cerana, and Apis mellifera (honey bee) (Linné, 1758) (Apidae), that is exploited by the Man in beekeeping for approximately 7000 years [3,11,12].

In Algeria, 02 breeds have been identified: The first one, Apis mellifera intermissa (Tellian bee) described by Buttel-Reepen (in Ruttner), it is a breed of north Africa found in northern Algerian Sahara and Libya to Morocco [13-15]. The second breed, was successively described by Baldensperger and also by Haccour [16,17]. Apis mellifera sahariensis is found in the south of Morocco and Algeria [14]. The studies aiming to know and characterize morphologically and biologically the breed which populates the Algerian territory are of a great interest in so far as they constitute the basis of any work of selection. It can be obtained using many morphological characters used in systematic. We therefore consider biometrics as a working method to highlight the purity of race or on the contrary, the hybrid character of the selected colonies. Several works were realized in certain regions of Algeria [18-23]. Our present study at 03 sites in the province of Jijel aims to complement and deepen the knowledge of Algerian bee populations.

Materials and Methods

Equipment for collection of samples and the measurement of studied morphological characters

The bees come from healthy apiaries. Shaking the frame covered with bees on top of a small plexiglass hive (6 × 8 × 10 cm) and transported to the laboratory where they were transferred in a freezer at -20°C waiting to be measured. Morphometric measurements of 15 characters on a sample of 69 bees are performed using a binocular magnifying glass (Carl Zeiss, GERMANY) and that of coloration is carried out under a microscope (OLYMPUS OPTICAL, JAPAN) equipped with an ocular micrometer calibrated with objective (X10) [21,23].

Description of the area of study

The sampling of worker bees of unknown ages has been carried out at the level of 03 stations (Tahir, Al-Ancer and Ziama) in the Province of Jijel; located at the extreme north-east of the Algerian coast during the beginning of the winter period (December 2014) (Figure 1). This region is located in the bioclimatic floor wet temperate and characterized by a type of diverse vegetation [24].

Figure 1: Geographical situation of the study sites in the Province of Jijel.

In each apiary, it was randomly sampled 23 bees, a total of 69 individuals (Table 1). All resident bees in hives of type "Langstroth". It has been verified that they were not object of transhumance and diseases such as it has been advocated by Cornuet et al. and Leporati et al. [25,26].

Table 1: Sites, geographic coordinates, number of worker bees measured per apiary of sampling.

Morphological characters studied and methods of measurement

The measures carried out on the workers relate to 15 characters which have been selected on the basis of their biological significance; these characters contribute much in the production of honey and wax.

The settings selected are: length of the tongue, length and width of the anterior wing, length and width of the posterior wing, length of the nervure A and B, the cubital index (ratio A/B), length and width of the metatarsus, length of the femur, length of the tibia, coloration, and length of the sternite 1 and 3.

The bees as well chosen are dissected one by one. The articles are collected and assembled between blade and cover glass in a few drops of a middle of mounting (glycerine gel) [27,28].

Statistical Analysis

For each studied site we have calculated the means, the standard deviations and the extreme values of each variable.

A comparative study of the means between the 3 sites for each variable (character) was carried out by using the analysis of variance to one classification criterion fixed template (ANOVA) [29,30].

The comparison of 3 sites, between them, for all 15 variables was performed by the analysis of variance multivariate (MANOVA) while using three statistical tests namely: Wilk's Lambda-Lawly Hotteling and Pillai's Trace. However none of these tests is considered as the most powerful and none of them can’t be recommended a preferential manner [29].

According to Huberty, the test Wilk’s is the most commonly used. This analysis method is an extension of the univariate analysis, to several variables [31].

We used also the analysis in main component (ACP), in order to look for eventually the groups of apiaries homogeneous and the groups of variable. This method was applied to the matrix of the correlation of the averages of the dimension data (nxp) with n=3 sites and p=15 variables in order to study the structure of bee colonies.

All statistical analyses were performed using MINITAB Software (Version 16, Penn State College, PA, USA).

Results

Description of biometric data

The mean ± standard deviations (SD) and the two extreme values calculated for 03 sites are shown in Table 2.

Table 2: Statistical parameters of 15 morphological characters for the whole of 03 sites.

Analysis of the variance (ANOVA)

The test results of analysis of the variance univariate applied to each of the 15 variables measured, for all 3 sites. Sampling show that there are very highly significant differences concerning the length of the tongue, the length and width of the anterior wing and the posterior wing, the length of the nervure B of the cubital index, the cubital index (Ratio A/B), the length of the metatarsus, femur, tibia, the sternite 1 and 3, and coloration (P = 0.001), is significant concerning the width of the metatarsus (P = 0.05), and non-significant concerning the length of the nervure A of the cubital index (Table 3 and Figure 2).

Table 3: Inter-site comparison of medium sized relating to 15 biometric variables of 03 sites sampling (ANOVA).

Figure 2: Morphometric study of the bees at the level of the three sites during the winter (December 2014) (m ± s, n=23), (L: Length, w: width).

Analysis of the variance multivariate or MANOVA

The results of the analysis of variance multivariate are shown in Table 4. We distinguish the values of Fobs with associated probabilities, and this for each of the test for equality of vectors of the averages of all the characteristics of 03 sites sampled. For this comparison Inter-apiaries, the three tests (Wilk’s Lambda, Hotelling-Lawley and Pilla’s Trace) of MANOVA lead to the rejection of legality of the vectors averages for all 03 sites with a very highly significant difference.

Table 4: Comparison between sites, vectors of the means of 15 different morphometric variables using MANOVA

Analysis in main components (ACP)

The eigenvalues of the correlation matrix, so the percentages and the cumulative percentages of variation explained by each of the components main for all of the data of the matrix (nxp) (Table 5 and Figure 3).

Table 5: Eigen values and percentages of total variation explained by the four main components and the cumulative percentages.

Figure 3: Cone diagrams of main components and eigenvalues.

We note that the first component CP1 explains himself 38.6% of the total variation of the initial variables, the first two component explain the all 50.6%; the first three component explain together 59.5%; and finally the fourth first component explain together 66.8 %. So, these four components or synthetic indices summarize the better the information given by the 15 initials biometric variables. From the 5 the components are less useful and correspond to eigenvalues less than unity (<1).

Table 6 presents the correlation matrix of the initial variables with the first four principal components. The correlation analysis of the variables with the first component CP1 shows that all the variables : length of the tongue (X1), length and width of the anterior wing (X2, X3), length and width of the posterior wing (X4, X5), length of the nervure A (X6) and B (X7), length and width of the metatarsus (X9, X10), length of the femur (X11), length of the tibia (X12), coloration (X13), and length of the sternite 1 and 3 (X14, X15) are positively correlated with the first main component, With the exception of variable (8) which represents the cubital index. The second components CP2 are correlated positively with length of the anterior wing (X2), Width of the posterior wing (X5), Length of the nervure A (X6), Cubital index (8), Length of the metatarsus (X9), Length of the femur (X11), Coloration (X13), Length of the sternite1(X14), Length of the sternite 3 (X15). The third components CP3 are correlated positively with 10 variables: length of the tongue (X1), length and width of the anterior wing (X2, X3), length and width of the posterior wing (X4, X5), length of the nervure A (X6) and B (X7), cubital index (8), width of the metatarsus (X10) length of the tibia (X12).

Table 6: Correlation matrix of the initial variables with the first four principal components.

The study of the correlations of the 15 initial variables relative to axes 1 and 2 of ACP watch that the variable (X7) represents the length of the nervure B is clearly positively correlated according to axis 1, while the variables (X1, X2, X3, X4, X6, X9, X10, X12) corresponding to length of the tongue, length and width of the anterior wing, length of the posterior wing, the length of the nervure A, length and width of the metatarsus length of the tibia get closer to 0.

Finally, the fourth components CP4 correlated positively with length of the tongue (X1), length of the posterior wing (X4), Length of the nervure A (X6) and B (X7), length of the tibia (X12), coloration (X13), and length of the sternite 1 and 3 (X14, X15).

The variable (X8) represent cubital index is negatively correlated according to axis 1. On the other hand, the group formed by variables (X5, X11, X13, X14) are positively correlated according to axis 1 and 2, are respectively width of the posterior wing, length of the femur, coloration, length of the sternite 1 and 3 (Figure 4).

Figure 4: Graphical representation of the correlations of the 15 initial variables on the plan composed of axis 1 and 2 of ACP.

Discussion

The biometric analysis and statistical analysis that we performed on samples of bees North-east Algeria used to define the position of this breed compared to others within the Apis mellifica from the point of view morphological

The description of the data by the univariate statistical method revealed that the bee of North-east Algeria is large compared to that of Marocco, Tunisia and the south Algerian [23] for most morphological characters [23,25,32-34].

Our morphometric results obtained compared with those obtained by several study especially, Ruttner, and Cornuet et al. on the Moroccan bees, or Grissa et al. on the Tunisian bees, finally Achou and Bendjedid and Achou on the Algerian bees [21,23,25,32,33].

The mean length of the anterior wing obtained is 13.56 mm (S1), 12.72 mm (S2) and 13.28 mm (S3). It is clearly very high compared to those found respectively by Ruttner, Achou and Bendjedid and Achou which are of the order of 9.19 mm, 8.58 mm and 7.24 mm [21,23,32].

The mean width of the anterior wing is 4.58 mm (S1), 4.29 mm (S2) and 4.45 (S3). It is greater than those found by Ruttner which is of the order of 3.08 mm, Achou 3.05 mm and by Bendjedid and Achou which is 2.39 mm [21,23,32].

The mean value of the width of the yellow band (length of the nervure A is of 8.10 mm (S1), 8.15 mm (S2) and 8.25 mm (S3). It is much higher respectively to those made by Cornuet et al. and Grissa et al. which are of the order of 0.544 mm and 0.563 mm, Achou which remain of a same order 0.50 mm and by Bendjedid and Achou who is from 0.41 mm [21,23,25,34].

The mean value of the width of the yellow band length of the nervure B obtained is 4.34 mm (S1) and 3.66 mm (S2, S3). This value is also higher compared to those found by Cornuet et al. or by Grissa et al. which are of the order of 0.222 mm and of 0.249 mm respectively the observed values by Achou is 0.210 mm, but by Bendjedid and Achou is from 0.176 mm [21,23,25,34].

The mean value of the width of the yellow band (Coloration) of the three stations studied equals at 1.37 mm and 1.47 mm. It is higher than that given by Achou and Bendjedid and Achou which are of the order of 0.40 mm and 0.45 mm, Grissa et al. who is from 0.19 mm [21,23,34].

This differentiation of the size is due to the high richness of the vegetation and the moderate climate of North-east Algeria. Thus, this bee has a body more large to browse the long distances in search of its food.

We can assign this differentiation to the existence of a northsouth gradient for some morphological characters. In effect, one of the first examples of the North-South gradient has been provided for the bees by a number of russian authors, as Chochlov, Michailov, Aplatov and Skorikov in Ruttner, who have found that the length of the tongue decreased gradually from north to south [35-38]. In our country, to the north-east, the length of the tongue is of 6.146 mm, it is therefore noted, based on our result and the value of 8.92 mm (S1), 8.16 (S2) and 8.78 (S3), a similarity of the length of the north to the south [19].

This large size in the Length of the tongue of bees, at the levels of sites presented in this study, is an adaptation to several types of flowers on which the bees feed. Indeed, this would lead to admit the existence of a parallel gradient of the average depth of the corollas of species melliferous. The hypothesis, advanced by Ruttner, we seems more likely [39]. It consists to admit that the environmental rules of Bergmann and Allen set for vertebrates to warm-blood apply here on bees [40]. These rules stipulate that the races cold climates tend to be bigger that the races of the same species living in warm climates (rules of Bergmann).

The comparison between the 3 apiaries for each of the 15 characteristics perform with the analysis test of univariate variance reveals that there exist between the three apiaries the differences very highly significant for the average of all the morphometric variables except for the width of the metatarsus shows a significant differentiation and non-significant for the length of the nervure A of the cubital index. These results are confirmed with the analysis of the variance multivariate (MANOVA) which shows that differences was very highly significant between the three apiaries and this for the majority of the 15 characteristics taken in consideration simultaneously.

This variation inter-apiary cannot be attributed to eco-climatic factors only, given that the bees live in the same biotope. We can be attributed to the size of the cells of the brood, because it is at this level that takes place the embryonic development of bees, consequently, the size of the bees may be affected [19,41-43]. In addition, the parasitism, in particular ectoparasites (for example: Varroa destructor) have an effect on the phenotype of bees [39,44,45].

The principal component analysis applied to our bee colonies made it possible to retain the first four components which together account for 66.8% of the total variation. The first component to it only takes into account 38.6% of variability. These four component or synthetic indices summarize the better the information brought by the 15 variables biometrics.

According to our results, we find that the values obtained are closer to the giant bee (Apis dorsata) and this is declared by Ruttner and that goes back to the importation of the species by the beekeeper, but we have not yet confirmed, because we are trained to do molecular biology [39].

Conclusion

In conclusion, our results indicate that morphometry study of the Algerian bees in three different sites can be used as a modern program for selection and determination of the different breeds. As results, the most studied characteristics revealed similar in the three tested sites.

Competing Interests

The authors declare that they have no competing interests.

Acknowledgements

This work is part of a doctoral thesis by Houda Bouzeraa. We would like to thank the « the Ministry of High Education and Scientific Research of Algeria » for their financial support.

References

1. Breeze TD, Bailey AP, Balcombe KG, Potts SG (2011) Pollination services in the UK: How important are honeybees? Agric Ecosyst Environ 142:137-143.
2. Ollerton J, Winfree R,Tarrant S (2011) How many flowering plants are pollinated by animals? Oikos 120: 321-326.
3. Michener CD (2000) The bees of the world. John Hopkins Univ. Press, Baltimore, Maryland, USA p: 913.
4. Le Conte Y, Navajas M (2008) Climate change: impact on honey bee populations and disease. In: Cimate change impact on the epidemiology and control of animal diseases. Rev sci tech off int Epiz 27: 499-510.
5. Crane E (1999) The world history of beekeeping and honey hunting. Taylor & Francis.
6. Danforth BN, Sipes S, Fang J, Brady SG (2006) The history of early bee diversification based on five genes plus morphology. PNAS 103: 15118-15123.
7. Michener CD (1974) The social behavior of the bees: a comparative study. Cambridge Mass: Harvard University Press.
8. Michener CD, Greenberg L (1980) Ctenoplectridae and the origin of longtonged bees. Zool J Linn Soc 69: 183-203.
9. Malyshev SI (1968) Genesis of the Hymenoptera. London Methuen.
10. Michener CD (1969) Comparative social behavior of bees. Ann Rev Entomol 14: 299-342.
11. Crane E (1976) The world’s beekeeping - past and present. Dadant and Sons (ed.), The Hive and the Honey Bee. Dadant and Sons, Inc, Hamilton, Illinois, USA pp: 1-38.
12. Crane E (1990) Bees and Beekeeping: Science, Practice and World Resources. Comstock Publishing Associates (Cornell University Press), Ithaca, New York.
13. Buttel-Reepen H von (1906) Apistica Contributions to the systematics, biology, as well as to the historical and geographic distribution of the honey bee (Apis mellijica L.), its varieties and the ilibigen Apis species. Zoological Museum in Berlin 3: 121-196.
14. Ruttner F (1968) Les races d'abeilles. In: Traité de Biologie de l'Abeille. Chauvin R (Ed.), Masson et Cie, Paris, France pp: 27-44.
15. Frère A (1955) A search of the best bee lines (Second Voyage). Belgium Beekeeping 19: 72-80.
16. Baldensperger PJ (1932) Bee varieties in North Africa, Proceedings of the 5thInternational Congress on Entomology, Paris, France pp: 829-839.
17. Haccour P (1960) Research on the Saharan bee in Morocco. Comptes Rendus, Society of Natural and Physical Sciences of Morocco 6: 96-98.
18. Loucif W (1993) Biometric study of populations Of bees from east Algeria Apis Mellifica intermissa. Thesis of magisterium; Ecology and animal physiology. ISN. Annaba University.
19. Achou M, Soltani N (1997) Impact of Varroa jacobsoni Oud. on the morphometry and biochemical composition of hermolymph in honeybees Apis mellifera intermissa L. Parasitica 53:127-134.
20. Doumandji H (2006) Biometric study of populations Of bees from Northern Algeria Apis Mellifica intermissa. Thèse de magistère ; Science animale. El-Harrach-Alger National Agricultural Institute.
21. Achou M (2007) Morphometric, biochemical and molecularPopulations of honeybees in eastern Algeria - Physiopathological effectsof its major parasite Varroa destructor. Doctoral thesis in animal biology. University of Annaba, Algeria p: 136
22. Barour C (2012) Analyse De La Biodiversité Des Populations D’abeilles Mellifères Apis Mellifera Intermissa (Buttel-Reepen, 1906) (Hymenoptera: Apoidea) Dans Le Nord Algérien: Morphométrie Moderne Basée Sur la Configuration des Points-Repères (Landmark). Thèse de Doctorat en biologie animale. Université de Annaba, Algérie.
23. Bendjedid H, Achou M (2014) Etude de la diversité morphométrique de deux populations d’abeilles domestiques (Apis mellifera intermissa et Apis mellifera sahariensis) du sud Algérien. Rev Sci Technol Synthèse 28: 84-95.
24. Greininger O (1982) L'apiculture en Algérie, Gazette Apicola 888: 131-141.
25. Cornuet JM, Daoudi A, Mohssine EH, Fresnaye J (1988) Etude biométrique de populations d'abeilles Marocaines. Apidologie 19: 355-366.
26. Leporati M, Valli M, Cavcchi S (1984) Biometric study of the geographical variability of Apis mellifera populations in northern Italy. Apidologie 15: 285-302.
27. El Banby MA, Aboukorah M (1976) The inheritance of certain quantitative characters. In: ligusticaxfaciata Honeybee Hybrid, The Libyan Journal of Agriculture 5: 131-137.
28. Hassanein MH, El Banby MA (1956) Studies on the biometrics of Egyptian honey bee, Apis mellifera fasciata Latr. (Hymenoptera: Apoidae). Bulletin de la Société Entomologique d'Egypte 42: 127-130.
29. Dagnelie P (1986) Multivariate statistical analysis. Gembloux Presses agronomiques p: 362.
30. Dagnelie P (1999) Theoretical and applied statistics. Tome 2 Bruxcelles, De Boeck et Larcier p: 659.
31. Huberty CJ (1994) Applied discriminated analysis. New York Wiley p: 466.
32. Ruttner F (1975) African races of honey bees. Proceedings XXV International Apiculture Congress, Grenoble pp: 325-344.
33. Cornuet JM, Fresnaye J, Tassencourt (1975) Discrimination and classification of bees from biometric characters. Apidologie 9: 145-187.
34. Grissa K, Cornuet JM, Msadda K, Fresnaye J (1990) Biometric study of populations of Tunisian bees. Apidologie 21: 303-310.
35. Chochlov BA (1916) Investigations on the length of the bee tongue (In Russian). Ministry of Agriculture, Petrograd pp: 17-41.
36. Michailov AS (1926) On the increasing variability of the honey bee from north to south (In Russian). Opytnaja Paseca 12: 13-14.
37. Alpatov WW (1925) On the reduction of the length of the honey bees from the south to the north. Zoologische Anzeiger 65: 103-111.
38. Ruttner F, Pourasghar D, Kauhausen D (1985) Die honigbienen des Iran. 2. Apis mellifera meda Skorikow, die Persische Biene. Apidologie 16: 241-264.
39. Ruttner F (1988) Biogeography and Taxonomy of honeybees. Springer-Verlag, Berlin, Germany p: 284.
40. Mayr E (1963) Animal Species and evolution. Belkhap Press of Harvard University Press, Cambridge, Mass, USA.
41. Abdellatif MA, Abou-E-Naga AM, Ali MH, Shakir PM, Al-Jaltli MK (1977) Biometrical studies on Iraq honeybees. J Apicult Res 16: 143-144.
42. Alpatov WW (1929) Biometrical studies on variation and races of the honey bee Apis melllifera L. Q Rev Biol 4: 1-58.
43. Goetz B, Koeniger N (1992) Structural features trigger capping of brood cells in honey bees. Apidologie 23: 211-216.
44. Daly HV (1985) Insect morphometric. Annu Rev Entomol 30: 415-438
45. Radloff SE, Hepburn HR (2000) Population structure and morphometric variance of the Apis mellifera scutellata group of honeybees in Africa. Genet Mol Biol 23: 305-316.