Study of the length of the mouthparts of Africanized, Caucasian and Africanized/Caucasian honey bee crosses, and relationships between glossa size and food gathering behavior

Study of the length of the mouthparts of Africanized, Caucasian and Africanized/Caucasian honey bee crosses, and relationships between glossa size and food gathering behavior

Maria Izabel Barnez Pignata ¹, Antonio Carlos Stort ² and Osmar Malaspina ²

¹ Centro de Ensino e Pesquisa Aplicada à Educação (CEPAE), Universidade Federal de Goiás, Caixa Postal 131, 74001-970 Goiânia, GO, Brasil. Send correspondence to M.I.B.P. E-mail: [e-mail]

² Departamento de Biologia, Instituto de Biociências, UNESP, Caixa Postal 199, 13506-900 Rio Claro, SP, Brasil.

ABSTRACT

The lengths of the mouthparts of bees, the glossa, paraglossa, stipes, galea, labial palpus, maxillary palpus, cardo, lorum, mentum and prementum, were studied in Caucasian and Africanized bees and in their F₁ descendants. Only the lengths of the paraglossa, stipe, galea, mentum, prementum and maxillary palpus differed significantly between these two bee types. These six variables were studied in the F₁ descendants of two types of crosses, i.e., Caucasian queens x Africanized males (cross 1) and Africanized queens x Caucasian males (cross 2). Multidimensional analyses were also performed and the generalized Mahalanobis distances (D²) between the F₁ descendants and the parental lines were determined. There was an apparent dominance of Africanized bees in both unidimensional and multidimensional analyses. Correlation analysis showed that bees with longer glossae collected more food (sugar syrup) and flew more slowly from the colony to the food source.

INTRODUCTION

The mouthparts of bees are highly specialized for nectar collection. The mandibles, maxillae and lips are the three major components of the mouthparts of adult bees. The latter two have been extremely modified to form a composite structure, the proboscis. The proboscis, frequently called the tongue of the bee, is used to suck nectar from flowers.

Several morphometric studies of the mouthparts of bees, mainly European bees, have been conducted, especially with respect to the length of the tongue as a factor of selection and discrimination among the various subspecies of Apis mellifera (Hachinohe and Onisshi, 1954; Weiss, 1955; Goetze, 1956; Akahira and Sakagami, 1959; Hawkins, 1969; Abdellatif et al., 1977; Michener and Brooks, 1984; Mattu and Verma, 1983). However, little is known about the morphometry of the various mouthparts of Africanized bees. Even the extensive studies by Gonçalves (1970) and Stort (1979) morphometrically comparing more than 150 traits of Africanized and Italian bees made no reference to the mouthparts.

Among the few reports available are those of Cosenza and Batista (1974), who showed that the length of the glossa of Africanized bees is shorter than that of Caucasian bees, and of Rinaldi et al. (1970), who established indices for different traits, including the glossae of Africanized, Italian and Caucasian bees, in order to differentiate these bee types. Funari (1983) also showed that the length of the glossa is important for the separation of bee groups, and Woyke (1977), in a study of Africanized and Italian bees, showed that diploid drones are closer to workers and haploid drones are closer to queens in measurements of lips and labial palpi.

Hybrids of crosses between Caucasian and Africanized bees have an intermediate glossal length (Cosenza and Batista, 1974). Foraging behavior studies by Beig et al. (1972), Nuñez (1974), Neves-Fermiano and Stort (1985) and Malaspina and Stort (1987) have described some traits of Africanized bees, but information about a possible influence of mouthpart size on foraging is lacking.

The objective of the present investigation was to analyze morphometrically and comparatively the mouthparts of Africanized and Caucasian bees. The mouthparts of the descendants of the crosses of these two groups of bees were also analyzed in order to obtain information about the type of inheritance. The data for glossal length were compared to food gathering data obtained by Malaspina and Stort (1987) for these same bees.

MATERIAL AND METHODS

Controlled crosses using instrumental insemination were made between Caucasian queens and Africanized drones (cross 1) and Africanized queens and Caucasian drones (cross 2) for the comparative study of the mouthparts of Africanized and Caucasian bees and of their F₁ descendants.

We used an Africanized parental colony, a Caucasian parental colony, 10 F₁ colonies originating from cross 1, and 10 F₁ colonies originating from cross 2. Samples of 20 bees were taken from each parental colony, and samples of 10 bees from each F₁ colony.

The mouthparts were removed from workers under a stereomicroscope with the aid of fine-pointed pliers, mounted on slides with clear enamel, without a coverslip, and measured with an ocular micrometer with 10X magnification. The length of the following structures was measured: glossa, paraglossa, stipes, galea, labial palpus, maxillary palpus, cardo, lorum, prementum and mentum. In analyses at the unidimensional level, the value of each variable of Africanized parental bees was compared with the value of each correspondent variable of the Caucasian parental bees. The t-test was used for comparison. The homogeneity of variances was checked by the F-test (Li, 1966) and the normality of the distribution by the test of Levy (1974).

Only the variables that differed in the parental lines were analyzed in the F₁ descendants. Each variable was submitted to analysis of variance and the differences in mean values between the parental lines and each F₁ generation were compared by the Tukey test. A comparison of the data for the F₁ descendants with the parental data at the multidimensional level, i.e., with all variables analyzed as a whole, was made by calculating the generalized Mahalanobis distances (D²) (Rao, 1952).

Bees from the Caucasian and Africanized parental colonies and from the F₁ colonies had been trained to collect sugar syrup on a scale and data about the following variables were obtained for each bee: bee weight, weight of honey stomach load, time spent on food source, time between foraging flights, time for one unloaded bee to fly from hive to feeder, time for a loaded bee to fly from feeder to hive and time taken to alight on food source (Malaspina and Stort, 1987). After the data for 20 foraging trips were obtained, each bee was captured, killed in an ether chamber, fixed in Dietrich for 24 h and stored in 70% alcohol. The length of the glossae of these bees was measured by Pignata (1990). Spearman correlation coefficients (r_s) (Siegel, 1975) were calculated to test for significant correlations between the glossal length of workers that had been trained and the seven foraging behavior variables.

RESULTS AND DISCUSSION

Of the 10 traits considered, the following six differed between parental bee colonies: length of the paraglossa, galea, stipes, maxillary palpus, prementum and mentum, all of these values being higher in the Caucasian than in the Africanized bees (Table I).

Comparison of European bees with African bees (Apis mellifera scutellata) has shown that African bees are smaller for most traits (Kerr et al., 1967; Kerr, 1969). Gonçalves (1970) found that only five of the 63 head and thorax traits analyzed (middle ocellum diameter, width of the mandible base, length of the propodeal spiracle, number of hamuli, and basitarsal width) were larger in Africanized than in Italian bees (Apis mellifera ligustica). Stort (1979) showed that only five of 111 traits analyzed in African (Apis mellifera scutellata), Italian (Apis mellifera ligustica) and German bees (Apis mellifera mellifera) (right eye width, middle ocellum diameter, distance between antennal alveoli, width of the area with no sensory structures on segment 10 of the antennae, and number of middle sensory hairs on segment 8 of the antennae) were larger in African bees. Stort (1979) showed that of 12 traits analyzed in the abdomen, only the length of the 6th tergite was greater in Africanized than in Italian bees.

Data obtained for bee samples collected from six different regions in Brazil showed that most of the bees were Africanized (closer to African than to European bees) and the mean values of the 10 traits studied were higher in Africanized than in African bees (Stort and Bueno, 1985). Thus, for most traits and in most cases Africanized bees have been found to be smaller than European bees (Italian, German and Caucasian) and larger than African bees, but closer to African bees (Stort, 1979; Daly, 1991).

The analyses of variance for the six traits that differed between the parental lines gave significant F values for the colonies of cross 1 (stipes = 4.63, galea = 4.42, mentum = 17.06, paraglossa = 584.03, maxillary palpus = 2.97, prementum = 7.03) and cross 2 (stipes = 5.82, galea = 2.82, mentum = 10.46, paraglossa = 574.90, maxillary palpus = 3.78, prementum = 4.40). In cross 1 (Caucasian queens x Africanized males), most F₁ colonies had mean values identical to those of the Africanized parental colony for almost all structures analyzed, i.e., length of stipes, galea and maxillary palpus (in 100% of cases), and length of mentum and prementum (in 90% of cases), with paraglossa length being different in 100% of cases (Tukey test). In cross 2 (Africanized queens x Caucasian males), the F₁ colonies had mean values identical to those of the Africanized parental line for length of stipes, galea, mentum, prementum and maxillary palpus (in 100% of cases), with paraglossa length being different in 100% of cases.

Thus, five of six traits showed apparent dominant inheritance from the Africanized bees. The number of genes involved could not be determined because we only had the F₁ generation and there was no control for heterozygosity of the two parental colonies (via inbreeding, for example). On the other hand, since there were several significant correlations between the various traits under study both in the parental lines and the descendants, some traits may be controlled by the same genes or by linked genes.

The data obtained for bees from the Caucasian and Africanized parental colonies and for their F₁ descendants (Table II and ^III) were also compared at the multidimensional level by calculating the Mahalanobis D². These analyses permitted joint comparison of all variables for each of the 20 F₁ colonies in relation to the parental colonies (Figures 1 and 2). Three D² values were higher in relation to the Africanized parent (those for colonies 98, 110 and 123) (Figure 1), whereas all F₁ colonies presented higher D² values in relation to the Caucasian parent (Figure 2), i.e., all F₁ descendants from cross 2 were closer to the Africanized parent. These differences between these two types of reciprocal crosses may be a consequence of a maternal effect, but the data obtained in the present study cannot explain the origin of these differences. It was also found at the multidimensional level that most F₁ colonies (85%) were closer to the Africanized parent.

Figure 1
- Relations between F₁ colonies from cross 1 (Caucasian queens x Africanized drones) and the Caucasian and Africanized parents in terms of generalized Mahalanobis distances.

Figure 2
- Relations between F₁ colonies from cross 2 (Africanized queens x Caucasian drones) and the Caucasian and Africanized parents in terms of generalized Mahalanobis distances.

The data obtained for glossa length in Caucasian and Africanized parental bees and in their F₁ descendants were compared with the data concerning seven variables of food gathering behavior obtained by Malaspina and Stort (1987). Since there was no difference in mean glossa length between the bees of the Caucasian and Africanized parental colonies (t = 0.822, nonsignificant), no comparisons were made between the mean values obtained for these bees and the foraging behavior variables. However, the method used to train workers individually and the later morphometric analysis permitted us to compare the glossa length of each bee with its own foraging behavior, for each colony.

Only one significant r_s value was obtained for the Africanized parental bees (Table IV), indicating that bees with a longer glossa collect a larger amount of syrup. No significant value was observed in the Caucasian parental bees.

In the comparisons with hybrid F₁ bees from cross 1, two significant correlation coefficients were obtained, indicating that bees with a longer glossa collect larger amounts of syrup and take a longer time to reach the food source. In contrast, comparison of hybrid F₁ bees from cross 2 did not show any significant r_s value, suggesting that the type of cross performed affects these characters.

According to Neves-Fermiano (1981), Africanized hives containing heavier bees have proven to be better honey producers. It could be that heavier bees have longer glossae and therefore collect more nectar, as demonstrated in the present study. Bees with longer glossae could potentially collect a larger amount of nectar because they possibly have better access to flowers with longer corollas and therefore could be better equipped to probe and collect more nectar. According to Roberts (1961), the heritability of tongue length is quite high among European bees (0.85). Almeida (1995) also showed high heritability of glossa length in Africanized bees.

ACKNOWLEDGMENTS

The authors wish to thank Dr. Odair Correa Bueno for help with the statistical analysis and the referees for corrections and suggestions. This research was supported by CAPES, CNPq and FAPESP. Publication supported by FAPESP.

RESUMO

Os comprimentos das peças que constituem o aparelho bucal, glossa, paraglossa, estipite, gálea, palpo labial, palpo maxilar, cardo, lorum, mento e pré-mento foram estudados a nível unidimensional em abelhas caucasianas, africanizadas e nos descendentes F₁. Somente a paraglossa, estipite, gálea, palpo maxilar, mento e pré-mento mostraram ser diferentes entre esses 2 tipos de abelhas. Essas 6 variáveis foram estudadas nos descendentes F₁, tendo sido utilizados 2 tipos de cruzamentos: rainhas caucasianas x machos africanizados (cruzamento 1) e rainhas africanizadas x machos caucasianos (cruzamento 2). Análises multidimensionais também foram realizadas, tendo sido obtidas as distâncias generalizadas de Mahalanobis (D²) entre os parentais e os descendentes F₁. Tanto nas análises unidimensionais como nas multidimensionais houve aparente dominância das abelhas africanizadas, mas o número de genes não pôde ser conhecido porque só havia a geração F₁ e não houve controle da heterozigosidade das 2 colônias parentais. Os cálculos dos coeficientes de correlações de Spearman mostraram que as abelhas com glossae mais longas coletaram mais xarope de açúcar e voaram mais lentamente da colônia para a fonte de alimento.

REFERENCES

Abdellatif, M.A., Abou-Elnaga, A.M., Ali, M.H., Shakir, P.M. and Al-Jalili, M.K. (1977). Biometrical studies on Iraqi honey bees. J. Apic. Res. 16: 143-144.

Akahira, Y. and Sakagami, S.F. (1959). A biometrical study on the Japanese honey bee; observations upon some populations on Kyushu (studies on the Japanese honey bee Apis cerana cerana). J. Hokkaido Gattugei Univ. 2: 353-362.

Almeida, M.J.O. de F. (1995). Avaliação de fatores genéticos e ambientais interferentes na produtividade de Apis mellifera. Master's thesis, UFG, Goiânia.

Beig, D., Pisani, J.F. and Kerr, W.E. (1972). Capacidade estomacal das abelhas operárias de duas subespécies de Apis mellifera L. (Hymenoptera, Apoidea). Cienc. Cult. 24: 464-468.

Cosenza, G.W. and Batista, J.S. (1974). Morfometria da Apis mellifera adansonii (abelha africanizada), da Apis mellifera caucasica (abelha caucasiana) e suas híbridas. Cienc. Cult. 26: 864-866.

Daly, H.V. (1991). Sistematics and identification of Africanized honey bees. In: The African Honey Bee (Spivak, M., Fletcher, D.J. and Breed, M.D., eds). Westview Press, Inc., Boulder, CO, pp. 13-44.

Funari, S.R.C. (1983). Estudo morfométrico de Apis mellifera Linée, 1758. Análise de agrupamento e caracteres discriminantes. Master's thesis, UNESP, Botucatu.

Goetze, G. (1956). Method of selection of the honey bees for length of tongue. Insectes Soc. 3: 335-346.

Gonçalves, L.S. (1970). Análise genética do cruzamento entre Apis mellifera ligustica e Apis mellifera adansonii. Escolha e análise genética de caracteres morfológicos da cabeça e do tórax. Doctoral thesis, USP, Ribeirão Preto.

Hachinohe, Y. and Onisshi, N. (1954). Studies on the length of proboscis of the worker honey bee (Apis mellifera). Bull. Nat. Inst. Agric. Sci. Ser. G 8: 9-16.

Hawkins, R.P. (1969). Length of tongue in a honey bee in relation to the pollination of red clover. J. Agric. Sci. 73: 489-493.

Kerr, W.E. (1969). Some aspects of the evolution of social bees (Apidae). Evol. Biol. 3: 119-175.

Kerr, W.E., Gonçalves, L.S., Stort, A.C. and Bueno, D. (1967). Biological and genetical information on Apis mellifera adansonii. Summary of the XXI International Congress of Apiculture, Apimondia, pp. 76.

Levy, K.J. (1974). Testing that K independent random samples were drawn from K normal populations. Psychometrika 39: 363.

Li, J.C.R. (1966). Statistical Inference I. Edward Brothers Inc., Ann Arbor, MI.

Malaspina, O. and Stort, A.C. (1987). Sucrose syrup collecting behavior in Africanized and Caucasian bees and in the descendants of their crossings. Rev. Bras. Genet. 10: 459-469.

Mattu, V.K. and Verma, L.R. (1983). Comparative morphometric studies on the Indian honey bee of the north-west Himalayas. 1. Tongue and antenna. J. Apic. Res. 22: 79-85.

Michener, C.D. and Brooks, R.W. (1984). Comparative study of the glossae of bees (Apoidea). Contrib. Am. Entomol. Inst. 22: 1-73.

Neves-Fermiano, L.H.M. (1981). Estudo do comportamento de coleta de alimento em duas subespécies de Apis mellifera (Hymenoptera, Apidae) e nos descendentes F₁ resultantes de seu cruzamento. Master's thesis, USP, Ribeirão Preto.

Neves-Fermiano, L.H.M. and Stort, A.C. (1985). Study of food gathering behavior in Italian bees (Apis mellifera ligustica), in Africanized bees and in the descendants of their crossings. Rev. Bras. Genet. 8: 29-36.

Nuñez, J.A. (1974). Estudio cuantitativo del comportamiento de Apis mellifera ligustica Spinola y Apis mellifera adansonii Latreille: Factores energeticos y informacionales condicionantes y estrategia del trabajo recolector. Cienc. Cult. 26: 786-790.

Pignata, M.I.B. (1990). Análise genética e morfológica do aparelho bucal de algumas espécies de abelhas da família Apidae. Master's thesis, UNESP, Rio Claro.

Rao, R.C. (1952). Advanced statistical methods in biometric research. John Wiley & Sons Inc., London.

Rinaldi, A.J.M., Pailhé, L.A. and Popolizo, E.R. (1970). Indices alares, tarsales y glossales. Anais do 1º Congresso Brasileiro de Apicultura, Florianópolis, pp. 228-233.

Roberts, W.C. (1961). Heterosis in the honey bee as shown by morphological characters in inbred and hybrid bees. Ann. Entomol. Soc. Am. 54: 878-882.

Siegel, S. (1975). Estatística não Paramétrica (Para as Ciências do Comportamento). McGraw-Hill do Brasil Ltda., São Paulo, pp. 350.

Stort, A.C. (1979). Estudo genético de caracteres morfológicos e suas relações com o comportamento de defesa de abelhas do gênero Apis. Livre Docência thesis, UNESP, Rio Claro.

Stort, A.C. and Bueno, O.C. (1985). Are Apis mellifera bees morphologically Africanized in Brazil? Rev. Bras. Biol. 45: 393-397.

Weiss, K. (1955). Measurements of tongue length and tongue reach in nigra and Carniolan bees. Z. Bienenforsch 3: 53-56.

Woyke, J. (1977). Comparative biometrical investigation on diploid drones of the honey bees. 1. The head. J. Apic. Res. 16: 131-142.

(Received March 18, 1997)

Publication Dates

• Publication in this collection
  01 Mar 1999
• Date of issue
  Dec 1998

History

• Received
  18 Mar 1997