Cytogenetics of hybrid introgression in Icelandic birch

Transcription

1 Cytogenetics of hybrid introgression in Icelandic birch KESARA ANAMTHAWAT-J6NSSON and THORSTEINN TOMASSON Agricultural Research Institute of Iceland, Keldnaholt, Reykjavik 112, Iceland ANAMTHAWAT-JONSSON, K. and TOMASSON, T Cytogenetics of hybrid introgression in lcelandic birch. - Heredirus /12: Lund, Sweden. ISSN Received October 23, Accepted January 4, 1990 Cytogenetic analysis has made it possible to describe the mechanism of introgression between two Icelandic birch species, dwarf arctic birch (Betula nana L.) and pubescent birch (Betulapubescem Ehrh.). It is suggested that the gene flow is predominantly from B. nana (2n=28) to B. pubescens (2n=56) via hybridization and subsequent backcrossing of the hybrids with B. pubescens. accompanied by selection of the introgressant type. The mechanism of such transfer is from diploid to tetraploid levels via intermediate triploids, not through the formation of aneuploid complexes. Kesara Anamthawat-Jonsson, Agricultural Research Institute of Iceland, Keldnaholt, Reykjavik 112. Iceland Icelandic birches are regarded as comprising two species, Betula nana L., dwarf arctic birch, and Betula pubescens Ehrh., pubescent birch. The two species differ in chromosome number; B. nana is diploid with 2n=28 whereas B. pubescens is tetraploid with 2n=56 (DARLINGTON and WYLIE 1955). They are distinctive in many characteristics, including growth habit, twig, bark, leaf and inflorescence morphology. B. nana shows little variation in morphology within the species. B. pubescens, on the other hand, is highly variable and therefore the name B. pubescens s.lat. was given as one collective species (LOVE and LOVE 1956). Birch populations in north-west Iceland were studied using histograms of hybrid indices based on morphological characteristics, and it was concluded that the variation in pubescent birch was due to introgressive hybridization between B. nana and B. pubescens (ELKINGTON 1968). The occurrence of interspecific hybridization among Betula species is common in the sub-arctic regions: Fennoscandia, Alaska, the Highlands of Scotland and Iceland (KALLIO et al. 1983). The Icelandic birches, B. nana and B. pubescens, are sympatric species. The short growing season, a characteristic feature of the sub-arctic regions, brings them closer in terms of growth, maturation and flowering. Considerable overlap in time of microsporogenesis of the two taxa has been noted in popu- lations Of Icelandic birch (ANAMTHAWAT-J6NSSON, unpublished). In Finland, anthesis of B. pubescens, B. pendula and also B. nana was found to be syn- chronous in the north, whereas there were differences of several days in the south (KALLIO et al. 1983). Factors such as temperature have been found to influence the genetic compatibility in birch, and it has been suggested that low temperature at the time of pollination might reduce the incompatibi- lity (HAGMAN 1971). Thus the environmental conditions of the sub-arctic regions appear to be highly favourable for hybridization of birches. Despite the wide recognition of introgression in the genus Betula, very little is known of its mechanism. Understanding the process of introgression requires cytogenetic information, which so far has been limited due to the lack of a reliable chromosome preparation technique. Birch chromosomes are very small, with an average length of less than one pm, and they tend to aggregate into clusters. We have previously described a method which enables the successful study of birch chromosomes (ANAMTHAWAT-JONSSON et a[. 1986). In the present paper, we examine chromosome complements of Icelandic birches, both from natural populations and from crossing experiments, and discuss the mechanism of gene flow. Materials and methods The birch material used in this study came from two sources; from crossing experiments at Korpa Agricultural Research Station in Reykjavik, and from

2 seeds collected in Brekkuskogur and Kirkjubaejnrskogur birch forests in the south and east of Iceland. These two birch populations were selected for the study because they are geographically distant (about 300 km apart) and both populations show extremely high morphological variability among the pubescent birches. Crosses between B. nunu and R. pitbescens were obtained, where the former was always the seed parent. One F,-hybrid plant. referred to as la. was produced in 1984, and in 1987 about forty F,-hybrid plants were produced. Twenty were randomly selected for chromosome counts and numbed from to Backcrosses between the F,-hybrid la as seed parent and R. pubescens as pollen parent gave rise to nine plants. numbercd from to Seed\ were collected from open-pollinated trees of B. pi~b~.w~n,\ in Brekkuskogur and Kirkjubaejarskogur forests and chromosome numbers were determined in greenhouse-grown seedlings. One plant from each of four single-tree progenies from Brekkuskogur (Korpa and 17) and one from each of twenty single-tree progenies from Kirkjubacjarskogur (F. 1.S4 to F.32.84) were USC~ for chromosome investigations. Root tips were collected during the period of active growth. particularly in June and August. Excised root tips were treated in 0.13 '50 colchicine tor 2-4 hours prior to fixation in 1:3 acetic acid: ethanol. Chromosome preparations were produced ugng the Feulpen-Giemsa double staining method o\ et al. 1986). with minor modifications, including the use of Leishman stain in place of Giemsa. Two to seven roots were used per plant. Chromosome counts were made using a microwope at a magnification of x and the counts were confirmed froni photographs. Meiosis of birch chromosomcs was observed from material collected in Brekkuskogur. Male catkins were slit lengthwise and immediately fixed in Carnoy's solution. The method of chromosome preparation was the same as for the root tip material. Results As shown in Table out of 15 F,-hybrid plants had 42 chromosomes (Fig. la), the expected number for triploid hybrids between the diploid B. nanu (2n=28) and the tetraploid B. pubescens (2n=56). These triploid plants represented a wide range of phenotypic intermediates between the two species for characteristics including leaf shape, hairiness of stems and leaves, branching patterns, and plant height. The remaining four F,-hybrid plants had 28 chromosomes (Fig. lb). Morphologically, the four hybrids resembled B. nnnu morc closely than B. pubescens and were easily distinguished from the triploid hybrids. Out of nine plants obtained in the backcrosses. four plants had 56 chromosomes (Fig. lc), like the male parent R. pube.scens. and two plants had 42 chromosomes (Fig. Id), the triploid number (Table I). Chromosome number could not be determined in three plants. The backcross plants were highly variable in morphology and resembled B. pubescens from natural populations. The plants were transferred to a natural habitat, where they are being grown for future morphological and adaptability assessmen t 5. Table 2 shows that the chromosome number in all scorablc cells from plants originating from natural populations was always 2n=56 (Fig. le), characteristic of B. pubcwens. However. a large number of mctaphases were uncountable because the chromosome5 tended to aggregate in clusters. Microsporogenesis was examined in samples collected from 29 plants of R. pubescens at Brekkuhkogur. The whole range of meiotic stages was observed, with the majority of cells being in early 7Wc I Chromiisorne numbers <if F,-hybrid plmts (A. mi~iu. scctl parent x H piibcwri\, pollen parent) and hackcross progenies (F,-hybrid 1.1. wcd piirrnt r 19 prrhrrwnr. pollen pirent) from crosaing experiment\ h'iirnher of cell\ 'indised per plant (range) with - 2n = 78 2n=42 k=56 I I 4

3 H~redito 112 (19901 HYBRID INTROGRESSION IN ICELANDIC BIRCH 67 Fig. 1 a-f. Photomicrographs of chromosomes of Icelandic birches, bar = 5 Frn. a and b Mitotic metaphases of F,-hybrids (B. nana, seed parent X B. pubescens, pollen parent), 42 and 28 chromosomes respectively. c and d Mitotic rnetaphases of backcross hybrids (F,-hybrid la, seed parent X B. pubescens, pollen parent), 56 and 42 chromosomes respectively. e Mitotic rnetaphase of a seed progeny of B. pubescens from Kirkjubaejarskogur, 56 chromosomes. f Meiotic metaphase of a pollen mother cell of B. pubescens from Brekkuskogur, probably with 28 bivalents.

4 ~ Table 2. Chromosome numbers of seed-grown plants derived from natural populationsof B. pubescens at Brekkuskogur and Kirkjuhaejarskogur Population Number of plants - Number of cells analysed per plant (range) with k.28 2n=42 2n = 56 Brekkuskogur 4 - >lo(mean= 7) Kirkjuhaejarskogur (mean= 5) prophase. Chromosome pairing was apparently normal in most of the microsporocytes (Fig. If). Some cells appeared to show multivalent-like structures together with normal bivalents. but this could not be confirmed because of the tendency of the chromosomes to aggregate into clusters. Other signs of irregular meiosis were observed, such as laggards at anaphase I and there was evidence of some chromosomes left behind at telophase 11. In one plant, the majority of pollen grains werz aborted and shrunken while pollen of the other plants appeared normal. Discussion Interspecific hybridization allows genetic recombination between species, and the stability of the transfer is achieved by the process of introgression through backcrossing (or introgressive hybridization). The mechanism of gene flow between B. pirhescens (2n=S6) and B. iium (2n=28) was suggested to be either direct from diploid to tetraploid levels or via intermediate triploids rather than through the formation of aneuploid complexes (Ei.tm~rou 1968). Triploid hybrids (2n=42) between B. pubescens and B. nana were reported to occur in Iceland (LOVE- and LOVE 1956). North Scotland (KENWOKTHY et al. 1972). and Northern Finland (SULKINOJA and VAI.AUNE 1980). LO~F antl Loi~ (1956). however. concluded that because the hybrids were sterile and only a very few hybrids were found. such introgressive hybridization did not account for the variation in Icelandic populations of 8. pirhescms, The results of the present study strongly indicate that triploid hybrids are not completely sterile. The F,-hybrid la produced some seeds and seedlings which were normal. although overall seed set antl germinability wcre very poor. Viable pollen grains were found among aborted pollen in male flowers of natural triploid hybrids occurring in Finland (SLYKIW)J,.\ and V,\i.,wNE- 1980). In a process of introgression, triploid hybrids may function as a bridge permitting the transfer of genes from one species to another. Only a very few such hybrids would be required for this process to occur, and there is no necessity for successful establishment as a population. There is a question of how the triploid hybrid serves as a bridge for gene transfer. Triploid hybrids may produce mainly two kinds of viable gametes (gametophytes), with n=28 and n=14. Two possible mechanisms can be suggested: firstly, both types of viable gametes may have contained chromosomes originating predominantly from one parent or the other. For example. in the egg cells with n=28, B. pubescens chromosomes would be retained after recombination but chromosomes with B. nana centromeres would be lost. Exchange of genetic materials or gene recombination may have occurred, presumably at homoeologous regions between the two genomes, before the chromosome elimination. Secondly, the two types of functional gametes, with n=28 and n= 14, may consist of various combinations of B. pubescens and B. nana chromosomes. Homology is likely to exist between the two birch species, which are cvolutionarily related. In the meiosis of the triploids, the chromosomes from both taxa can probably form multivalents, bivalents and univalents, depending on the number and size of homoeologous regions among the chromosomes. This would allow a reassortment of genes by chromosome recombination between the two species. Birch trees with 2n=42 from natural populations in Scotland, a rare cytotype which may or may not have hybrid origin involving B. pubesceris (2n=S6) and B. pendula (2n=28), showed a high frequency of multivalents and other abnormalities in the meiosis of pollen mother cells (BROWN and AL-DAWOODY 1979). It was also suggested that the basic chromosome number of the genus is x=7 rather than the usually accepted x=14. In this case, homology between genomes could be even greater than expected. If n=28 female gametes of the triploid hybrid

5 HYBRID JNTROGRESSION IN ICEL.4NDlC BIR('H 69 B. nana x 3. pubescens 2n =2x=28 2n =4x=56 B. nana 2n = 2x = 28 \ n= 14 \ X F1 hybrids, triploid x B. pubescens 2n =3x=42 2n=4x = 56 " 2n=2x = 28 2n=3x =42 2n= 4x=56 Backcross progenies Fig. 2. A diagram showing crosses and chromosome numbers of parents and progenies studied, x=14. were fertilized by normal n=28 pollen of B. pubescms, progenies with 56 chromosomes would be produced (Fig. 2). Chromosome stabilization at the tetraploid level would be achieved in one single step, which is the most efficient mechanism of introgression. Some plants from backcrosses to B. pubescens (Table 1) and all seed progenies derived from natural populations (Table 2) had 56 chromosomes, the chromosome number of the tetraploid level. The chromosome stabilization is important as a means to achieve relatively regular meiosis and, hence, high fertility. Female gametophytes with n=14 may be less viable. From the crossing experiment, two out of six backcross plants had 42 chromosomes (Table 2), probably as a result of fertilization of n=14 female gametes of the F,-hybrid with n=28 pollen of B. pubescens (Fig. 2). These individuals with 42 chromosomes and the triploid F,-hybrids are cytologically indistinguishable and probably share the same function in the introgression process. In subsequent generations, the chromosomal constitution of the progenies of these plants would be shifted quickly towards 2n=56. The introgression under study is from B. nana to B. pubescens, a common characteristic of diploid- tetraploid introgression. It is possible, however, that gene transfer may also occur in the opposite direction, although much less extensively. Fertilization of n=14 gametes of the triploids by normal n=14 gametes of B. nana would result in hybrid derivatives with chromosomal stabilization at the diploid level 2n=28 (Fig. 2). These backcross products should be able to interbreed with B. nana without hindrance. As a consequence, variation in B. nana is broadened by having received gene transfer from B. pubescens. The comparison of Icelandic and Scottish birches by the use of hybrid indices (ELKINGTON 1968) showed that the index score of B. nana from north-west Iceland was wider than that of B. nana from Scotland (10-24 and 10-19, respectively) and that the score of Icelandic B. nana overlapped the score of B. pubescens from the same localities. The hybrids between B. nana (seed parent) and B. pubescens (pollen parent) with 2n=28 instead of the expected triploid chromosome number (Table 1) may be an example of introgression from B. pubescens to B. nana. Male and female gametes of triploid hybrids, as proposed earlier, could be mainly of two types; n=14 and n=28. It is thus suggested that there was some n=14 pollen, prod-

6 uced by triploid plants. in the pollen pool collected from R. puhescens in the natural populations, and that this pollen was included in crossing experiments under study. When then= 14female gametes of B. num were fertilized with n=l4 pollen from the triploids, hybrid plants with 2n=28 were produced. Morphologically, these hybrids resembled B. nana more closely than B. pitbeseem. Later observation showed that after the first winter dormancy. most of these 2n=28 hybrids died. This may be the reason why such introgression is uncommon in nature. The present study has provided us with a genetic background useful for birch breeding programmes. Being the most dominant tree species in Iceland, its value in forestation and landscape is significant. 4[ktior\,lrdypnit,iir -- The authors than!, the Science Researc!i Council 0 1 Iceland for financial support. References A\AVTHAH 4 I -Jo\sso\. K.. ATIPAYL MI'AI, L., TIGERSTIIDT, P. M A. andtohl4~~oti.t TheFeiilgen-Giemsamethod for chromosomer of Berula species. - Hereditus BROWU. I. R. and AL-DANOODY, D Observations on melosib in three cytotypes of Berula albn L.. -New Phytol XI1 DARLIYGTON. C. D. and WYI.IE. A. P. IYSi. Chromosome atlas of flowering plants. --Allen and Unwin Ltd., London ELKINGTON. T. T Introgressive hybridization between Rrrzrlu nonu L. and B. puhescens Ehrh. in north-west Iceland. - New Phyrol H.4GMAN. M On self- and crosc-incompatibility shown by Herulo wrrucow Ehrh. and Berula puhrscens Ehrh - Comrnun. Insr Foresr. Fenrr. 7.7: K.&I.L.IO. P.. NIEMI. S.. SULKINOJA. M. and VALANNE, T The Fennoscandian hirch and its evolution in the marginal forest zone. -,Wordcann KENWORTHY. J. B.. ASTON. D. and BUCKNALL, S. A A study of hybrids between Rerukupuhescens Ehrh. and B. nana L. from Sutherland ~ and integrated approach. - 7 ~1,s Rot. Soc Edinh LOVE. A. and LOVE. D Cytotaxonomical conspectus of the Icelandic Flora. - Acra Horri Gotoburgensis 20: SL'I.KlhOJA. M. and VAIANNE, T Polyembryony and ahnormal germination in Berula puhescen? subsp. forfuo~a. - Rep. Kevo Suhurcric Re.?. Stn. 16: 31-37