Gene dispersal by bumblebees between two lines of faba bean.
The cross-fertilization rate in faba bean is 50% on average, but estimates range from 4 to 84% (Bond and Poulsen, 1983; Link, 1990; Metz et al., 1993). According to Bond and Poulsen (1983) and Link et al. (1994), genetic and environmental factors, and also the method used for assessing the hybridization rate, may account for this high variability. Among environmental factors, pollinator activity may play a prominent role in this variability. In fact, the apoids guild contributing to faba bean pollination varies among fields and years. Each species can be characterized by behavioral traits that indicate whether the bee is an efficient pollen carrier, for example, the ability to obtain nectar at the mouth of flower without piercing the corolla base, the speed of flower visitation, and alternate foraging between lines (Poulsen, 1973: Tasei 1976: Currie et al., 1990). Therefore, we can assume that different proportions of bee species result in variations of pollen carryover and allogamy rates.
Most of the data on gene dispersal in faba bean come from the experiments of Bond and Pope (1974). These authors improved the conditions of line isolation but did not report on pollinator behavior, pollination intensity, or the relationship between pollen dispersal and gene flow. More recently, pollen transfer effectiveness was studied between male-fertile and male-sterile lines of faba bean Carre, et al., 1994). As gene flow is potentially different from pollen flow pattern because of postpollination events (Handel, 1983), an adequate assessment of gene dispersal is essential, especially in breeding programs.
Our objective was to estimate the consequences of visits to faba bean flowers by B. terrestris on cross-fertilization rate as influenced by (i) the position of recipient flowers in the visitation sequence, (ii) the position of flower nodes on the stem, and (iii) the ovule rank in the ovary.
MATERIAL AND METHODS
Pollen donor line D-27 is self-fertile and the recipient line D-23. with low self-fertility, requires tripping for fertilization (Mesquida et al., 1990). Under selfing conditions, 38 and 14% of their flowers produce pods, respectively (Le Guen et al., 1992). These two spring lines obtained by selection at the INRA Plant Breeding Station in Dijon (France) belong to the equina group. Le Guen et al. (1992) demonstrated that when D-23 and D-27 were cross pollinated by bumblebees, the outcrossing rates were 16 and 7%, respectively. Preliminary observations revealed that pollen production of both lines was similar, 24 200 [+ or -] 1300 pollen grains per flower (mean [+ or -] SE) for line D-27 (unpublished data) and 21000 [+ or -] 1500 pollen grains per flower for line D-23 (Carre et al.. 1994). Their flowering period was also similar. Seeds of D-27 and D-23 were sown on 15 March 1990 and plants were grown individually in 4-L pots under a greenhouse covered by 2-mm, nylon mesh. Flowering began during the last week of May, and pollination experiments with B. terrestris were conducted during the second week of June. Before presenting the plants to the foragers, we suppressed a part of the bloom to leave only five flowers per node at the open-flower stage, as defined by Marcellos and Perryman (1990). After pollination, the plants were kept under insect-proof conditions until harvest.
The B. terrestris colonies were reared in a glasshouse with plants of Lupinus, Medicago, and Melilotus spp. Pollen-collecting workers were caught on flowers, isolated in plastic screened tubes and kept in the dark for 30 min until introduction in a cage with donor plants of D-27. These short-tongued bumblebees constantly exhibited pollen collecting behavior on faba bean in our experimental conditions. Pollen carried on the head and thorax of B. terrestris workers after a 10-flower visit ranged between 1620 and 6300 grains (mean [+ or -] SE = 3150 [+ or -] 430) Carre, et al., 1994).
The experimental device used to study cross pollination by B. terrestris consisted of 11 nylon mesh cages (0.6 by 0.5 by 1.5 m each) joined side by side and fitted with sliding plexiglass partitions. The cages were kept in an insect-proof greenhouse. A pollen donor plant D-27 was introduced in the first cage while each of the other cages contained one recipient plant of D-23. Pollination trials were performed between 1100 and 1600 h. In front of each cage, a nylon screen attached with "Velcro" strips facilitated plant and bumblebee handling. Seventeen B. terrestris workers were used to perform one foraging run each on D-23 plants following controlled visitation of D-27. After each trial, the worker loaded with D-23 pollen was considered as not appropriate to a second experimental run. Therefore, it was possible to assess intra-population but not intra-individual variability. For each run, the tube containing one worker was opened in front of the lowest flower node of the D-27 plant in the first cage. The B. terrestris workers usually move upwards without revisiting flowers. When a worker had visited 10 D-27 flowers, the sliding plate was removed and the worker was moved to the second cage to continue the foraging sequence on the D-23 recipient flowers. The number of open flowers varied from 12 to 18 [plant.sup.-1]. Unless the worker stopped foraging, it visited at least 10 flowers before it was moved to the next cage with another D-23 plant. After each positive flower visit, a mark of acrylic paint was applied on the standard petal according to a sequential code. After each trial, this code was replicated on the calyx to permit the identification of the visit order until pod maturation. The 17 foraging sequences induced the tripping of 1261 D-23 flowers which produced 1040 pods containing 2812 seeds harvested for paternity assessment. At harvest, dry pods were isolated in small paper bags. The flowering node, flower position within the visiting sequence and flower and pod abortions were noted for each plant. Ovule position within the ovary was also recorded for each seed.
Isozyme Analysis of Seeds
Alcohol dehydrogenase (ADH) and Shikimate dehydrogenase (SKD) enzyme systems were used to characterize the genotypes of the two inbred parental lines. Fertilization of D-23 plants by the pollen of the donor line D-27 was easily recognizable through the heterozygote pattern of seeds compared with the homozygote pattern of self-fertilized seeds. Isozyme analysis was performed following the method developed by Carre et al., (1993): samples were prepared for starchgel electrophoresis by drilling cotyledons to obtain approximately 40 mg of flour per seed. Extracts were mixed with 100 mL of buffer (0.1 M Tris HCI. pH 7.5) and centrifuged at 14 840 X g for 10 min. The supernatant was used directly for electrophoresis (starch-gel concentration: 12% w/v) for 3 h at 250 V. Concentrate solution for electrode and gel buffer was L-histidine-citric acid (pH 6.5); staining solutions were described by Pasteur et al. (1987) for ADH and by Wendel and Weeden (1989) for SKD.
The relationship between flower and pod abortion rates and visitation order was tested by linear least squares regression. Chi-squared was used to compare differences among classes of cross-fertilized seeds according to flowering node and position of the seed in the pod. Due to the low number of hybrid seeds produced, the variability of allogamy rates among bumblebee runs was estimated after grouping the flowers; by five or 10 according to visitation order.
The mean number of visited flowers per run was 74 and ranged from four to 147. The mean number of visited plants per run was 5.3 and ranged from one to 10. In only three runs, were there less than 35 flowers visited. The total number of cross-fertilized seeds was 78. Individual worker bees induced from one to 9 hybrid seeds during their runs. The fertilization success of D-27 pollen transferred to D-23 was influenced by variability among workers. Cross fertilization rarely occurred beyond the tenth flower visited in the recipient line. These first 10 flowers accounted for 86% of all hybrid seeds. The average cross-fertilization rate for a 1/1 flower ratio of donor/recipient was 17.5% of the seeds for these first 10 flowers. Except for four isolated hybrid seeds between flower numbers 30 and 57, no gene transfer was observed beyond the twentieth flower. Considering the first five flowers visited in all runs, the percentage of hybrid seeds was 21.3 [+ or -] 9.4% of the seeds (Table 1). Flower and young pod (2 cm long) abortions averaged 17.5% of the tripped flowers. Linear regression indicated that abortion was not related to visitation order: y = -0.0005x + 0.202 (P = 0.065, [R.sup.2] = 0.023, d.f = 145), where y = abortion rate and x = visitation order. The percentage of pods containing one, two or three hybrid seeds was 78.0. 13.5, and 8.5%, respectively, beyond the first 10 pollinated flowers, no more than one cross-fertilized seed per pod was observed. The rate of hybridization was not influenced by the positions of the floral node or the ovule within the ovary ([X.sup.2] = 0.52 and 0.37, respectively).
Table 1. Variation of allogamy rate among worker runs according to visitation order. Flower visitation Number of Hybrid seeds order runs % 1 to 5 17 21.3 [+ or -] 9.4([dagger]) 6 to 10 16 9.4 [+ or -] 4.9 11 to 20 16 1.4 [+ or -] 1.1 21 to 60 14 0.4 [+ or -] 0.0
([dagger]) Standard error.
In an earlier cage experiment, either honeybees or B. lapidarius queens induced a cross-fertilization rate of 16.3% on the D-23 line grown in two rows adjacent to two rows of the D-27 line Carre, et al., 1991). Hybridization rate of D-23 was 10% of the seeds when visited by B. lapidarius queens carrying pollen from the donor line (T.201). This trial (unpublished data) was conducted in enclosures containing one plant of each line. This low allogamy tendency is confirmed by the 17.5% allogamy rate resulting from the visit of 10 flowers of the D-23 line by B. terrestris workers loaded with pollen of the D-27 line.
Therefore, three experimental protocols, using three different pollinators and either random or controlled cross pollination of D-23 by two different lines, resulted in a restricted range of hybridization rates. In recent unpublished experiments, cross pollination of 21 recipient lines of faba bean with the same donor (T.201) by B. lapidarius queens resulted in a hybridization range of 0 to 23% of the seeds. The allogamy rates ranged from 13 to 83% of the seeds in T.201 considered as the recipient line. We could conclude that a particular line may show an average low or high capacity to produce hybrid seeds, but the allogamy level will depend on the male parent. Pollen production, germination, and fertilizing capacity are more likely to cause such variations than behavioral traits of pollinators.
In our experiment where pollination was under control, the frequency of hybrids was similar regardless of the flowering nodes as reported in the field experiment of Duc (1990). As already stated by Hanna and Lawes (1967), Poulsen (1975), and Duc (1990), cross fertilization is unrelated to the position of the ovule in the ovary suggesting that allopollen and autopollen tube growth are identical.
The fast decrease of the hybridization rate after a few visits indicates that auto-pollen deposits rapidly exceed those of the donor line. Pollen collecting on the recipient line and grooming behavior could explain the dilution of foreign pollen and the reduced dispersion area previously reported on Erythronium (Thomson, 1986) and Diervilla (Thomson and Plowright, 1980). Moreover, variations in cross fertilization could result from postpollination processes. As no enhancement of germination and tube growth of foreign pollen has been demonstrated (Duc and Rowland, 1990; Hanna and Lawes, 1967), the respective amount of pollen from the recipient and donor line becomes the deciding factor. Besides, Carre et al. (1994) working on pollen deposits on the stigma of a male-sterile line with a similar visitation protocol, demonstrated that only 22% of the visited flowers had deposits greater than 10 grains and 60% of the visited flowers received no grain at all. In this experiment, pollen deposit on flowers successively visited did not always decrease monotonically and displayed a pattern in aggregates which is consistent with the distribution of self and cross-fertilized pods observed in this study.
In breeding procedures, spatial arrangement of lines in rows may reduce bumblebee movements between lines and create conditions unfavorable to a panmictic gene dispersal (Mesquida et al., 1990). Our study gives new insight into the feasibility of producing faba bean hybrids. It is obvious that a high frequency of alternate foraging between parental lines is needed to ensure a high percentage of allogamy. Our experimental procedure, implying controlled movements of a single insect on parental lines is useful to compare the pollen carryover by different pollinators and to study the variability of the allogamy tendencies of plants. In addition. cross fertilization may also be affected by the self-fertility of partners (Link, 1990) or by gametic competition phenomena (Le Guen, 1983). As a consequence, the possibility of creating commercially successful synthetic hybrid faba bean seems to be limited (Ederer et al., 1992; Metz et al., 1993). However, for better exploitation of heterosis than in conventional synthetics Link et al. (1994) investigated the prospects of "quasi-hybrids". Their method employs a large-seeded line as a pollen parent and one to three small-seeded lines as seed parents of the cultivar. In that system, it would be of interest to find out which plant arrangements facilitate pollen dispersal between lines and ensure the highest cross-fertilization rate.
The authors are grateful to A. Boe (South Dakota State University) for critical reading of the manuscript and help in language improvement.
Abbreviations: ADH, alcohol dehydrogenase; SKD, shikimate dehydrogenase
Bond, D.A., and M. Pope. 1974. Factors affecting the proportion of cross-bred and selfed seed obtained from field bean (Vicia faba L.) crops. J. Agric. Sci. (Cambridge) 83:343-351.
Bond, D.A., and M.H. Poulsen. 1983. Pollination. p. 77-101. In P.D. Hebblethwaith (ed.) The faba bean (Vicia faba L.); a basis for improvement. Butterworths, London.
Carre, S., I. Badenhausser, J.N. Tasei, and J. Le Guen. 1994. Pollen deposition by Bombus terrestris L. between male-fertile and malesterile plants in Vicia faba L. Apidologie 25:338-349.
Carre, S., J.N. Tasei, J Le Guen. J. Mesquida. and G. Morin. 1993. The genetic control of seven isozymic loci in Vicia faba L. Identification of lines and estimates of outcrossing rates between plants pollinated by bumblebees. Ann. Appl. Biol. 122:555-568.
Carre, S., J.N. Tasei, J. Mesquida. and J. Le Guen. 1991. Estimate of outcrossing rate between lines of field beans (Vicia faba L.) in various conditions with isozymic markers. Acta Horticulturae 288:354-358.
Currie, R.W., S.C. Jay, and D. Wright. 1990. The effects of honeybees (Apis Mellifera L.) and leafcutter bees (Megachile rotundata F.) on outcrossing between different cultivars of beans (Vicia faba L.) in caged plots. J. Apic, Res. 29:68-74.
Duc, G., and G.G. Rowland. 1990. The effect of ovule position on the frequency of hybrid and inbred seed in faba bean. Can. J. Plant Sci. 70: 79-82.
Ederer, W., and W. Link. 1992. How shall we select the components for open pollinated faba bean (Vicia faba L.) varieties? p. 75-76. In Proc. I. Europ. Conf. on Grain Legumes, Angers. France. 1-3 June 1992. Edition Soft Publicite, Reims. France.
Handel, S.N. 1983. Pollination ecology, plant population structure. and gene flow. In L. Real (ed.) Pollination biology. Academic Press. Orlando. FL.
Hanna, A.S., and D.A. Lawes, 1967. Studies on pollination and fertilization in the field bean (Vicia faba L.). Ann. Appl. Biol. 59: 289-295.
Le Guen, J. 1983. Incompatibilite unilaterale chez Vicia faba L. I. Analyse globale de croisements intraspecifiques entre quatre sous-especes. Agronomie 3:443-449.
Le Guen, J., J. Mesquida. G. Morin. F. Brunnet, J.N. Tasei, and S. Carre, 1992. Effect of insect pollination on abortion rate in faba bean. Fabis Newsl. 31:25-28.
Link, W. 1990. Autofertility and rate of cross-fertilization: Crucial characters for breeding synthetic varieties in faba bean (Vicia faba L.). Theor. Appl. Genet, 79:713-717
Link, W., Ederer. P. Metz. H. Buiel. and A.E. Melchinger, 1994. Genotypic and environmental variation for degree of cross-fertilization in faba bean. Crop Sci. 34:960-964.
Marcellos, H., and T. Perryman. 1990. Controls over pollination and ovule fertilization in Vicia faba . Euphytica 49:5-13.
Mesquida, J., J. Le Guen. J.N. Tasei. S. Carre, and G. Morin. 1990. Modalites de la pollinisation chez deux lignees de feverole de printemps (Vicia faba L. var. equina Stendel). Effets sur les coulures, la productivite et les taux de croisements. Apidologie 21: 511-525.
Metz, P.L.J., A.A.M. Buiel. A. van Norel, and J.P.F.G. Helsper, 1993. Rate and inheritance of cross-fertilization in faba bean (Vicia faba L.). Euphytica 66:127-133.
Pasteur, N., G. Pasteur, F. Bonhomme, J. Catalan, and J. Britton-Davidian. 1987. Manuel technique de genetique par electrophorese des proteines. Lavoisier. Paris.
Paul, C., P. Gates, N. Harris, and D. Boulter, 1978. Asynchronous sexual development determines the breeding system in field beans. Nature (London) 275:54-55.
Picard, J. 1960. Donnees sur l'amelioration de la feverole de printemps Vicia faba L. Annales de l'Amelioration des Plantes II:121-123.
Poulsen, M.H. 1973. The frequency and foraging behaviour of honeybees and bumble bees on field beans in Denmark. J. Apic. Res. 12:75-80.
Paulsen, M.H. 1975. Pollination, seed setting, cross-fertilization and inbreeding in Vicia faba L. Z. Pflanzenzucht 74:97-118.
Stoddard, F.L., and D.A. Bond. 1987. The pollination requirements of the faba bean. Bee World 68:144-152.
Tasei, J.N. 1976. Les insectes pollinisateurs de la feverole d'hiver (Vicia faba equina L.) et la pollinisation des plantes males steriles en production de semence hybride. Apidologie 7:1-38.
Thomson, J.D. 1986. Pollen transport and deposition by bumble bees, and its transport in Erythronium: Influence of floral secretions and bee grooming. J. Ecol. 74:329-341.
Thomson, J.D., and R.C. Plowright. 1980. Pollen carryover, nectar rewards, and pollinator behaviour with special reference to Diervilla lonicera. Oecologia, 46:68-74.
Wendel, J.F., and N.F. Weeden. 1989. Visualization and interpretation of plants isozymes, p. 5-45. In D.E. Soltis and P.E. Soltis (ed.) Isozymes in plant biology. Chapman and Hall. London.
S. Carre,(*) J. N. Tasei, I. Badenhauser, J. Le Guen, G. Morin, and J. Pierre
S. Carre, J.N. Tasei and I. Badenhauser, INRA, Laboratoire de Zoologic. 86600 Lusignan, France; J. Le Guen, G. Morin. J. Pierre. INRA. Station d'Amelioration des plantes, 35650 Le Rheu. France. Received 10 Nov. 1995. (*) Corresponding author (Serge.Carre@lusignan.inra.fr).
Published in Crop Sci. 38:322-325 (1998).
|Printer friendly Cite/link Email Feedback|
|Author:||Carre, S.; Tasei, J.N.; Badenhauser, I.; Le Guen, J.; Morin, G.; Pierre, J.|
|Date:||Mar 1, 1998|
|Previous Article:||Responses to seven methods of recurrent selection in the BS11 maize population.|
|Next Article:||Yield improvement in a multistage breeding program for cassava.|