Published online 30 July 2007
Published in Crop Sci 47:1401-1406 (2007)
© 2007 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
CROP BREEDING & GENETICS
Breeding System of the Aerial Flowers in an Amphicarpic Clover Species: Trifolium polymorphum
Daniel Reala,b,c,*,
Marco Dalla Rizzad,
Rafael Reynoa and
Kenneth H. Quesenberrye
a Forage Legume Dep., National Institute of Agricultural Research, INIA Tacuarembó, Ruta 5 Km 386, Tacuarembó, Uruguay
b Cooperative Research Centre for Plant-Based Management of Dryland Salinity, Univ. of Western Australia, University Field Station, 1 Underwood Ave., Shenton Park, WA 6009, Australia
c School of Plant Biology, Faculty of Natural and Agricultural Sciences, Univ. of Western Australia, 35 Stirling Hwy., Crawley, WA 6009, Australia
d Biotechnology Unit, National Institute of Agricultural Research, INIA Las Brujas, Ruta 48 Km 10, Canelones, Uruguay
e Dep. of Agronomy, Univ. of Florida, Gainesville, FL 32611-0500
* Corresponding author (dreal{at}cyllene.uwa.edu.au).
 |
ABSTRACT
|
|---|
Two perennial Trifolium, T. polymorphum Poir. and T. argentinense Speg., are American clovers unique within the genus for being amphicarpic. There is no consensus in the literature regarding the breeding system of the aerial flowers of T. polymorphum, therefore, the breeding system was studied. In 1997 T. polymorphum was collected in Uruguay and evaluated at INIA Tacuarembó. In 2001, 10 field patches were marked and in 2004, 20 plants per patch were characterized with simple sequence repeat markers. Patch J10 showed a particular molecular profile, therefore, 198 open-pollinated progenies freely visited by honeybees were studied. In 2005, at the University of Florida, Gainesville, different hand-pollination treatments were conducted within an accession from Paraguay. Trifolium polymorphum was able to cross-pollinate with all the known pollen donors molecularly marked that surrounded plants from patch J10 (30%), also with some nonmarked native ones from the vicinity (10%) as well as with itself (60%), when allowed to be visited by honeybees. However, when there are no pollinators, the selfing rate is minimal. The proposed classification for the breeding system is an allogamous, self-compatible species that benefits from pollinators to set seed.
Abbreviations: PIC, polymorphism information content • SSR, simple sequence repeat
|
INTRODUCTION
|
|---|
THE GENUS Trifolium comprises 245 species with a wide distribution throughout the temperate and subtropical regions of the world. North and South America have 26% of these species and 13 species occur in South America (Zohary and Heller, 1984). The perennial forage legume Trifolium polymorphum Poir. is one of the most widespread South American clovers. It is endemic to Uruguay, Argentina, southern Brazil, Chile, Peru, and Paraguay (Burkart, 1952; Zohary and Heller, 1984; Izaguirre and Beyhaut, 1998). Trifolium polymorphum is also distributed in North America, in the USA (east Texas and west Lousiana) previously classified as T. amphianthum Torr. and Gray (Zohary and Heller, 1984), but this might be a secondary dispersion area. There are only two other Trifolium species; T. argentinense Speg. and T. riograndense Burkart in the Campos region of South America (Uruguay, southern Brazil, and eastern Argentina), both are perennials with narrow distributions. Trifolium polymorphum has a stoloniferous prostrate habit and roots at the nodes and is reported to be found in the wild as diploid (2n = 2x = 16) and tetraploid (2n = 4x = 32) (Zohary and Heller, 1984; Vizintin et al., 2006). It grows in coexistence with a mixture of C3 and C4 grasses, the main components of the native grasslands of Campos region. Trifolium polymorphum is the most broadly adapted perennial forage legume in all the soil types of Uruguay. It has potential to be bred and sold as improved seeds to farmers in the Campos region.
Trifolium polymorphum and T. argentinense are unique species in their genus for being amphicarpic. These two species have typical aerial flowers in clusters and also small cleistogamous buried basal flowers. Trifolium polymorphum usually grows in uniform patches of approximately 1-m diameter and when flowering, the flower color is a distinct feature among patches. Genomic simple sequence repeat (SSR) transferable markers have proven that plants within patches are mainly uniform, but patches are different among themselves (Dalla Rizza et al., 2005).
There is no consensus in the literature regarding the breeding system of the aerial flowers of T. polymorphum (Schifino-Wittmann, 1985; Coll and Zarza, 1992; Speroni, 2000; Speroni and Izaguirre, 2001, 2003). Moreover, the necessity to determine the reproduction mode of the aerial flowers and which form of reproduction is given priority by the species was reported by Speroni and Izaguirre (2003).The objective of this study was to determine the mode of reproduction of the aerial flowers of T. polymorphum.
|
MATERIALS AND METHODS
|
|---|
Plant Materials
In November 1997, 64 accessions of T. polymorphum were collected in a large array of soils, from shallow stony soils to deep clay soils and deep acid sands, in Uruguay. The pH of these soils varied from 4.5 to 6.5 and rainfall is all year round with an average of 100 mm mo–1. At each collection site, a minimum of 100 flowers were collected. In 2001, at INIA Tacuarembó Research Station (31°44'20'' S, 55°58'46'' W), 10 distant patches of approximately 1 m2 of T. polymorphum were marked and labeled J1 to J10. In 2004, 20 individual plants were physically mapped per patch (J1–J10), and subsequently transplanted to pots in a naturally lit glasshouse at INIA Tacuarembó. In 2005, 198 open-pollinated progenies of plants from patch J10 were planted. A Paraguayan genotype (USDA GRIN, PI 233554, collected on the grounds of the Agricultural School, Pilar, Paraguay) of T. polymorphum was utilized in 2005 for a pollination study in an insect-proof glasshouse at the University of Florida, Gainesville.
Plant Characterization
Morphological Characters
In June 1998, 20 seeds per accession were surface disinfected by 4 min soaking in 70% alcohol, then 8 min in 10% hypochlorite, followed by four rinses with sterile water. Seeds were scarified with sterile sandpaper and sown in petri dishes. Petri dishes were placed in a growth cabinet, and once the seeds germinated they were transplanted to 10 pots per accession inside a screenhouse and subsequently characterized for variability of several traits. The measurements taken were leaf mark (no leaf mark, white central leaf mark, or red leaf mark), leaf margin (smooth or serrated), length of central leaflet, hairiness (0–3, 0 being glabrous plants and 3, hairy plants), number of stolons and yield scores (1–9, 1 being the smallest plants and 9, the largest plants) on 14 Dec. 1998 and 25 Jan. 1999.
The 200 plants transplanted in 2004 from the 10 patches were also measured for size and assigned to one of three leaf mark categories: no leaf mark, white central leaf mark, or red leaf mark. The next generation of 198 open-pollinated progenies from patch J10 had leaf marks assigned to the three categories mentioned above.
Molecular Markers
Plant material of the 200 plants (20 plants per 10 patches) was sampled for DNA extraction (Dalla Rizza et al., 2004). For cross–T. polymorphum amplification, 50 ng of genomic DNA was used for polymerase chain reaction, following the amplification procedure described by Kölliker et al. (2001). Four perfect (sequences containing at least five uninterrupted dinucleotide repeats), three imperfect (three or more uninterrupted tetranucleotide repeats), and two compound (four uninterrupted trinucleotide repeats) SSR classified markers (Table 1) were evaluated, using silver stain to reveal the amplicons in 8% nondenaturing polyacrylamide gel electrophoresis. The primers were screened for their ability to yield an amplification product and to detect repeatable polymorphisms. For primers that detected polymorphism, the number of alleles and the polymorphism information content (PIC) of the SSR was calculated as described by Saal and Wricke (1999):
 | [1] |
where pi is the frequency of the ith allele out of the total number of alleles and k is the number of different alleles in the sample.
View this table:
[in this window]
[in a new window]
|
Table 1. Primer characteristics (primer sequences, polymorphism, allele number, polymorphism information content [PIC]) found in T. polymorphum.
|
|
Pollination Studies
The collection of 640 plants sown in 1998 was evaluated for seed production in a screenhouse at INIA Tacuarembó. In December 1998, 10 plants corresponding to 10 different collection sites were taken to an insect-proof glasshouse, where all exiting flowers were removed and new flowers tagged on a weekly basis and hand-tripped. Another 10 plants from the same collection sites were taken outside the screenhouse, where they were frequently visited by honeybees. There were also many native T. polymorphum plants growing in the nearby area.
In 2004, nine individual plants corresponding to patch J10 were surrounded by four plants of the other nine patches (nine sets of five plants) and allowed to open-pollinate. In 2005, seeds of these nine J10 plants were harvested and 198 progenies were grown in a naturally lit glasshouse at INIA Tacuarembó. Similar methodology was applied by Real et al. (2004) to study the breeding system of the forage legume Lotononis bainesii Baker from South Africa.
In 2005, with the Paraguayan genotype in an insect-proof glasshouse at the University of Florida, 21 flowers were marked. The three treatments imposed to them were (i) trip individual florets as they opened with a small emery paper covered toothpick, (ii) roll gently entire head between finger and thumb to effect tripping of florets, and (iii) do nothing except tag flowers after the florets were all opened.
|
RESULTS
|
|---|
Plant Characterization
Trifolium polymorphum was very variable in all characteristics measured in 1998 and 1999 from the germplasm collected in 1997. No leaf marks were observed for 80% of the plants, 11% had a white central leaf mark, and 9% a red leaf mark. Leaf border was 98% smooth and 2% serrated. Length of central leaflet varied from 0.3 to 2.6 cm. There were 8% glabrous plants, 47% plants of hairiness 1, 34% of hairiness 2, and 11% of hairiness 3. Yield scores (1–9) are presented in Fig. 1
. Number of stolons was also very variable with an average of 1.3 stolons and a maximum of 23 stolons per plant. These results were further analyzed by Real and Ferreira (1999) and Real et al. (2005) to determine best plant collection strategy. The best 50 plants from evaluations in 1998 and 1999 were selected and transplanted into 10-L pots, and in spring 1999 they were harvested and the seed was stored in the INIA genebank as breeding germplasm.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 1. Qualitative yield score (1–9) of 640 T. polymorphum plants measured on 14 Dec. 1998 and 25 Jan. 1999 at INIA Tacuarembó.
|
|
From the nine microsatellites primer pairs screened in 2004, five showed cross-amplification success in T. polymorphum (TRSSRA01H11, TRSSRA02B08, TRSSRA02C02, TRSSRA02C03, and TRSSRAXX31) comparable in size to the product reported by Kölliker et al. (2001) for T. repens L. (Table 1). Perfect primers had the highest PIC values and detected more alleles than compound or imperfect SSRs. As observed for T. repens (Kölliker et al., 2001), imperfect SSRs did not detect polymorphisms. PIC values obtained with primers TRSSRA01H11, TRSSRA02C03, and TRSSRAXX31 were 0.80, 0.86, and 0.85, respectively. These PIC values are highly unexpected considering the differences in geographical origin of the two species. They are higher than the reported values by Kölliker et al. (2001) for T. pratense L. but less than values reported for T. ambiguum M. Bieb and T. nigrescens Viv. At least 12 alleles were observed with the perfect microsatellite TRSSRA01H11 in 10 patches (Fig. 2
and 3B
), more than reported in T. repens.
View larger version (80K):
[in this window]
[in a new window]
|
Figure 2. Microsatellite marker TRSSRA01H11 amplification on T. polymorphum genotypes. Three plants belonging to nine patches. Molecular ladder on the right.
|
|
View larger version (137K):
[in this window]
[in a new window]
|
Figure 3. (A) Microsatellite marker TRSSRA01H11 amplification on 40 plants of T. polymorphum corresponding to patches J9 and J10 (1–20 = 20 plants from J9; 21–40 = 20 plants from J10; 41 = DNA ladder, 100 bp [Fermentas, Hanover, MD]). (B) Microsatellite marker TRSSRA01H11 amplification on 16 plants of T. polymorphum corresponding to patch J6 (molecular weight marker on the right, DNA ladder, 100 bp [Fermentas]).
|
|
In 2004, 10 patches were evaluated by SSR–cross-amplification and showed a high level of polymorphism (Fig. 2 and 3B) among patches. The 20 plants in six patches were monomorphic for the analyzed loci (examples presented in Fig. 3A) and four patches were dimorphic (example presented in Fig. 3B).
Pollination Studies
During evaluation of 640 plants in the screenhouse at INIA Tacuarembó in 1998, the aerial flowers did not set seeds, indicating that pollinators were required. The 10 plants evaluated inside the insect-proof glasshouse and their tripped flowers also did not set seed, reinforcing the observations in the 640 plants. Conversely, the 10 plants taken outside the screenhouse and allowed to be visited freely by honeybees did set seeds. At the University of Florida, with the Paraguayan genotype (USDA GRIN, PI 233554), the aerial flowers were able to set 1.57, 1.95, and 2.27 seeds per floret when tripped, left untouched, and rolled, respectively (Table 2).
View this table:
[in this window]
[in a new window]
|
Table 2. Effects of different flower manipulations on seed production in glasshouse conditions by a Paraguayan genotype of T. polymorphum (USDA, GRIN PI 233554)
|
|
For determination of the reproductive system of the aerial flowers, plants of patch J10 were used as mother plants because they did not have any high molecular weight band as did all the other nine patches (Fig. 2 and Fig. 3B). Patch J10 had no leaf mark, however, from the 198 progenies; 11 had white leaf marks and 11 had red leaf marks. When the 198 progenies of patch J10 were analyzed with the molecular marker TRSSRA01H11, nonmaternal alleles were observed in 78 individuals (39.4%). Within the 78 individuals with bands not explained by selfing of the mother plants, 58 had bands that were present in at least one of the other nine surrounding molecularly characterized pollen donors and 20 had bands that were not explained by the studied pollen donors and may have come from native T. polymorphum in the vicinity of the trial.
The 22 plants with either central white leaf mark or red leaf mark were analyzed separately with molecular marker TRSSRA01H11 to determine if all of them had high molecular weight bands and were already considered products of cross-pollination or if new information was gained by adding the morphological data to the molecular marker analysis (Fig. 4
). All the plants with leaf marks were already among the 78 plants that had high molecular weight bands.
View larger version (54K):
[in this window]
[in a new window]
|
Figure 4. Microsatellite marker TRSSRA01H11 amplification on 22 progenies of patch J10 that had leaf marks. Molecular ladder on the left (DNA ladder, 100 bp [Fermentas, Hanover, MD]).
|
|
|
DISCUSSION
|
|---|
The preliminary results obtained in 1998 with plants not setting seed inside the screenhouse and failing to set seed when located in an insect-proof glasshouse were evidence that the species is not cleistogamous. These results are in agreement with Speroni and Izaguirre (2003) that reported the species was chasmogamous. However, when individual florets were tripped, these too did not set seed, indicating that either there is an incompatibility system in the species or there is different maturity among sexes in the same floret. When the plants were allowed to be visited by honeybees and open-pollinate, seeds were produced readily. Due to the fact that this species has a double flowering system with aerial and subterranean flowers, it is difficult to explain a genetic incompatibility system that operates in the aerial flowers but not in the subterranean flowers. Therefore, a differential maturity of sexes in the same floret is more likely to be the cause of failure to self, instead of an incompatibility system.
In the hand-pollination experiment, with only one studied accession that selfs quite readily, there was no significant difference in seed production between the rolling treatment (2.27 seeds per floret) than the tripping treatment (1.57 seeds per floret). The rolling treatment may be more similar to insect visitation and manipulation of florets where different maturity stages are all cross-pollinated within the same flower. Although some selfing did occur in this accession we believe the above data does not negate our overall determination that the reproductive system of the aerial flowers is primarily cross-pollination in the field.
There is not a consensus in the literature regarding the breeding system of the aerial flowers of T. polymorphum. Schifino-Wittmann (1985) was not able to come to a conclusion about the breeding system of the aerial flowers. Coll and Zarza (1992) reported them as allogamous but did not support their observations with detailed studies. Speroni (2000) and Speroni and Izaguirre (2001), concluded that aerial flowers were autogamous. The evidence for autogamy was not very strong due to a very poor seed production in all pollination treatments in the experiments reported by Speroni (2000) and Speroni and Izaguirre (2001), consequently, Speroni and Izaguirre (2003) concluded that even though they had found a few cases of autogamy, further studies were needed to determine if the aerial flowers are predominantly autogamous or allogamous.
When we studied the variability of patches in nature, we observed that they were variable among themselves and mainly uniform within a patch (Fig. 2, 3A and 3B), indicating that new patches were most likely formed by cross-pollinated seed from aerial flowers. If the plants of a single patch predominantly originated by vegetative reproduction as reported above, each patch might consist of clonal progeny from a single maternal genotype; the percentage of polymorphic SSR loci observed between patches was very high, as would be expected for a highly heterogeneous cross-pollinated species. The PIC was comparable to values found in other cross-pollinated species such as T. repens (Kölliker et al., 2001) and Secale cereale L. (Saal and Wricke, 1999).
The effective use of transferable markers among species from diverse geographical origins suggests that the South American species T. polymorphum and T. argentinense could be useful as sources of genetic variability for future improvements in the genus Trifolium. Additionally, in the context of comparative genomics, the exploitation of cross-genomic synteny and orthology of genes involved in complex pathways is a powerful tool to survey allelic diversity not available or expressed in any one species.
In summary, all the information regarding the pollination biology of aerial flowers of T. polymorphum suggests that the aerial flowers are able to cross with all the pollen donors surrounding the target plants, with pollen coming from native sources in the vicinity, as well as with its own pollen. However, in absence of pollinators, the evidence for producing self-pollinated seed was limited to a few cases. Therefore, the proposed classification for the breeding system of T. polymorphum is an allogamous, self-compatible species that benefits from pollinators to set seed.
|
ACKNOWLEDGMENTS
|
|---|
The authors wish to thank Prof. G. Spangenberg (DPI, Victoria, Australia) for kindly sharing the T. repens microsatellites used in this study and also to thank D. Torres, P. Díaz, M. Zarza, A. Viana, and R. Mérola from INIA for their practical contributions. We would also like thank Dinacyt-MEC Project No. 8151, INIA, and the program TSTAR from USDA for the partial funding provided for this project.
|
NOTES
|
|---|
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.
Received for publication November 24, 2006.
|
REFERENCES
|
|---|
- Burkart, A. 1952. Las leguminosas argentinas silvestres y cultivadas. Editorial ACME, Buenos Aires, Argentina.
- Coll, J., and A. Zarza. 1992. Leguminosas natives promisorias: Trébol polimorfo y babosita. (In Spanish.) Boletín de Divulgación 22:19.
- Dalla Rizza, M., D. Real, K.H. Quesenberry, and E. Albertini. 2004. Plant reproductive system determination under field conditions based on codominant markers. J. Genet. Breed. 58:47–54.
- Dalla Rizza, M., D. Real, R. Reyno, and K.H. Quesenberry. 2005. Use of cross-species amplification markers for pollen-mediated gene flow determination in Trifolium polymorphum Poiret. p. 194. In M.O. Humphreys (ed.) Molecular breeding for the genetic improvement of forage crops and turf. Proc. of the 20th Int. Grassland Congr., Aberystwyth, Wales. July 2005. Wageningen Academic Publishers, Wageningen, the Netherlands.
- Izaguirre, P., and R. Beyhaut. 1998. Las leguminosas en Uruguay y regiones vecinas. Part 1. Editorial Hemisferio Sur, Montevideo, Uruguay.
- Kölliker, R., E.S. Jones, M.C. Drayton, M.P. Dupal, and J.W. Forster. 2001. Development and characterisation of simple sequence repeat (SSR) markers for white clover (Trifolium repens L.). Theor. Appl. Genet. 102:416–424.[CrossRef][Web of Science]
- Real, D., M. Dalla Rizza, K.H. Quesenberry, and M. Echenique. 2004. Reproductive and molecular evidence for allogamy in Lotononis bainesii Baker. Crop Sci. 44:394–400.[Abstract/Free Full Text]
- Real, D., and G.A. Ferreira. 1999. Native germplasm collection for genetic improvement in rangeland fields under various production systems. p. 665–666. In Proc. of the Int. Rangeland Congr., Vol. 2, Townsville, QLD. 19–23 July 1999. CSIRO, Elect Printing, Fyshwick, ACT.
- Real, D., R. Reyno, M. Dalla Rizza, M. Zarza, R. Mérola, A. Viana, N.A. Altier, D.F. Risso, C.A. Labandera, M. Jaurena, K.H. Quesenberry, and M.J. Williams. 2005. Leguminosas nativas evaluadas en INIA Tacuarembó desde 1997 al presente. (In Spanish.) IX Congreso Nacional de Ingenieros Agrónomos, II Encuentro regional de Ingeniería agronómica, Montevideo, Uruguay. 24–25 Oct. 2005. AIAU, Montevideo, Uruguay.
- Saal, B., and G. Wricke. 1999. Development of simple sequence repeat markers in rye (Secale cereale L.). Genome 42:964–972.[Medline]
- Schifino-Wittmann, M.T. 1985. Estudios citogenéticos em Trifolium riograndense Burkart, T. polymorphum Poir. e T. repens L.: Inducao de poliploidia, numero cromossomico, cariotipo, comportamento meiotico. Ph.D. diss. Universidad Federal de Rio Grande do Sul, Rio Grande Do Sul, Brazil.
- Speroni, G. 2000. Aspectos de la biología reproductiva de Trifolium polymorphum Poir. (Fabaceae, papilionoideae). M.S. thesis. Universidad de la República, Montevideo, Uruguay.
- Speroni, G., and P. Izaguirre. 2001. Morfología y Esporogénesis en flores aéreas y subterráneas de la especie anficárpica Trifolium polymorphum (Fabaceae, Papilonoideae). (In Spanish, with English abstract.) Boletín Sociedad Argentina de Botánica 36(3–4):253–265.
- Speroni, G., and P. Izaguirre. 2003. Características biológicas de la leguminosa nativa promisoria forrajera Trifolium polymorphum Poir (Fabaceae, Papilonoideae). (In Spanish, with English abstract.) Rev. Agrociencia 3:68–76 (UDELAR).
- Vizintin, L., B. Jarornik, and B. Bohanec. 2006. Genetic characterization of selected Trifolium species as revealed by nuclear DNA content and ITS rDNA region analysis. Plant Sci. 170:859–866.
- Zohary, M., and D. Heller. 1984. The genus Trifolium. Ahva Printing Press, Israel Academy of Science and Humanities, Jerusalem.
TOP
ABSTRACT
INTRODUCTION
