Genetic structure of wild rice Oryza glumaepatula ... - CiteSeerX

Genetica (2005) 125:115–123 DOI 10.1007/s10709-005-4916-4

Springer 2005

Genetic structure of wild rice Oryza glumaepatula populations in three Brazilian biomes using microsatellite markers Rosana Pereira Vianello Brondani1, Maria Imaculada Zucchi2, Claudio Brondani1, Paulo Hideo Nakano Rangel1, Tereza Cristina De Oliveira Borba1, Priscila Nascimento Rangel1, Mara Rubia Magalha˜es1 & Roland Vencovsky2 1

Laborato´rio de Biotecnologia Vegetal, Embrapa-CNPAF, C.P. 179, 74001-970, Goiaˆnia-GO, Brazil (Phone: +55-62-533-2154; Fax: +55-62-533-2100; E-mail: rosanavb@cnpaf.embrapa.br); 2Departamento de Gene´tica – ESALQ/USP, Av. Pa´dua Dias, 11 CP 83, Piracicaba, 13400-970, SP, Brazil

Received 13 July 2004 Accepted 17 March 2005

Key words: genetic resources, genetic variability, molecular markers, population genetics, Oryza sp.

Abstract The existence of Oryza glumaepatula is threatened by devastation and, thus, the implementation of conservation strategies is extremely relevant. This study aimed to characterize the genetic variability and estimate population parameters of 30 O. glumaepatula populations from three Brazilian biomes using 10 microsatellite markers. The levels of allelic variability for the SSR loci presented a mean of 10.3 alleles per locus and a value of 0.10 for the average allelic frequency value. The expected total heterozygosity (He) ranged from 0.63 to 0.86. For the 30 populations tested, the mean observed (Ho) and expected heterozygosities (He) were 0.03 and 0.11within population, respectively, indicating an excess of homozygotes resulting from the preferentially self-pollinating reproduction habit. The estimated fixation index (FÎS ) was 0.79 that differed significantly from zero, indicating high inbreeding within each O. glumaepatula population. The total inbreeding of the species (FÎT ) was 0.98 and the genetic diversity indexes among populations, F^ST and R^ST , were 0.85 and 0.90, respectively, indicating high genetic variability among them. Thus, especially for populations located in regions threatened with devastation, it is urgent that in situ preservation conditions should be created or that collections be made for ex situ preservation to prevent loss of the species genetic variability.

Introduction Many wild species have been used as an additional source of genetic variability in breeding programs for cultivated species. In the case of rice, there are 21 wild species, and 6 of them constitute the primary gene pool that are diploid and have the AA genome that facilitates the obtainment of fertile hybrids (Khush, 1997). Four wild Oryza species have been identified in Brazil, but only the diploid Oryza glumaepatula (AA) belongs to the primary gene pool. Recent studies show that O. glumaepatula is predominantly an inbreeding species, and a prolific seed producer with an

essentially annual life cycle (Vaughan, Morishima & Kadowaki, 2003). Wild Brazilian rice, O. glumaepatula, completely adapted to the tropical soils and weather, is the most promising germplasm for rice breeding in Brazil, and has been successfully used as a donor of genes to cultivated rice, Oryza sativa (Brondani et al., 2001). Information on the genetic diversity of native O. glumaepatula populations is of fundamental importance for decisions about effective conservation management. In addition, the identification of genetically divergent populations is facilitating the selection of O. glumaepatula accessions as genitors in interspecific crosses with the

116 cultivated rice. Successful examples of introgression from wild rice species to cultivated rice include resistance to the tungro virus from O. nivara, resistance to the bacterial blight from O. longistaminata, new sources of cytoplasmic male sterility from O. rufipogon (Brar & Khush, 1997), and genes related to the increase of yield potential from O. glumaepatula (Brondani et al., 2002). The advances obtained with molecular genetics have broadened our knowledge on the wild species populations through studies of the polymorphism found at the DNA level using microsatellite markers. Also known as SSR (simple sequence repeats), microsatellites are outstanding markers because they are frequent and randomly distributed in the eukaryote genome, present co-dominant expression, and have high informative content due to multiallelism that results from variability in the number of repeated sequences. Besides, results can be easily obtained by the automated PCR technique. Microsatellite markers have been successfully used in characterizing plant populations (Collevatti, Brondani & Grattapaglia, 1999; Lemes, Brondani & Grattapaglia, 2002), gene flow and paternity (Wright & Bentzen, 1994). This study aimed to characterize O. glumaepatula populations from three Brazilian biomes (Amazon, Pantanal Wetlands and Cerrado), that posses predominantly distinct fauna and flora, by estimating population parameters based on microsatellite markers to generate useful information for conservation strategies and the use of this wild species as a donor of favorable genes in rice genetic breeding programs.

germination rate, a minimum of 3–24 individuals per population were analyzed, for a total of 414 samples (Table 1). Genome DNA extraction and analysis with microsatellite markers Twenty days after transplanting, leaf tissue samples were collected from each individual for genomic DNA extraction. The DNA was extracted following the protocol described by Doyle and Doyle (1987). The DNA concentration was estimated by electrophoresis in 0.8% agarose gel by comparison with a lambda DNA standard, adjusted to 3 ng/ll. Genetic analysis was made with previously published microsatellite markers (Akagi et al., 1996; Chen et al., 1997; Temnykh et al., 2000; Brondani et al., 2001). The amplification reactions were performed with a final volume of 13 ll containing 0.3 lM of each primer, 1 U of Taq DNA polymerase, 0.2 mM of each dNTP, 1 mM Tris– HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 13 ll of DMSO (50%) and 7.5 ng of template DNA. The PCR amplification reactions were performed in a PT-100 thermocycler (MJ Research) with the following program: one pre-cycle at 96 C for 2 min; followed by 30 cycles at 94 C for 1 min, 56 C for 1 min and 72 C for 1 min. Finally, na extension step at 72 C for 7 min. was performed. The amplification was checked by horizontal electrophoresis in 3.5% agarose gel containing TBE 1· buffer (0.09 M Tris–Borate and 2 mM EDTA, pH 8.3) and 0.2 lg/ml ethidium bromide.

Material and methods

Microsatellite screening

Populations assessed

A total of 133 microsatellite marker loci, 28 chosen from the literature based on their greater than 70% expected heterozygosity values and 105 derived from sequences of the rice genome sequencing program, formed the initial selection and were tested on 6 O. glumaepatula genotypes from different populations. After this analysis, 10 microsatellite marker loci that presented greater allelic variability, clear, specific and reproducible amplified products were used to genotype the 414 individuals. The allelic polymorphism was detected in 4% denaturing polyacrylamide gels containing 7 M urea and 1· TBE buffer, and visualized by

Thirty natural populations of O. glumaepatula were collected, 23 from the marshy valleys of the Cerrados in the Goia´s State, 2 from the Amazon region (Solimo˜es River) in the Amazon State, and 5 from the Pantanal wetlands in the Mato Grosso State. A total of 200 panicles were collected from each population, with an average of 40 seeds per panicle. The seeds were bulked, and 40 seeds of each population were germinated on paper rolls. After one week, the seedlings were transplanted to pots in a greenhouse. Depending on the

GEN 1234

GEN 1235

GEN 1241

GEN 1242

GEN 1244 GEN 1247

GEN 1237

GEN 1238

GEN 1239

GEN 1240

GEN 314

GEN 1245

3

4

5

6 7

8

9

10

11

12

13

GEN 1249

GEN 1250

GEN 1252

GEN 1253

GEN 1254

GEN 1255

GEN 1256 GEN 1257

GEN 256

GEN 291

GEN 311

GEN 315

GEN 313

GEN 319

GEN 1236

16

17

18

19

20

21

22 23

24

25

26

27

28

29

30 Mean

6

12

11

21

5

8

3

13

13 15

17

19

9

10

11

12

12 17

19

22

12

24

16

17

17 16

8

14

18

17

Dammed marshy valleys/Cerrado biome

Taguarı´ River Santa Aˆngela Farm

Negro River

Taguarı´ River

Solimo˜es River Solimo˜es River Paraguai River

Santo Antoˆnio Farm Santo Antoˆnio Farm

Dois Irmaõs Dairy Santo Antoˆnio Farm

Saõ Carlos Farm Saõ Carlos Farm

Rio Vermelho Farm

Rio Vermelho Farm

Rio Vermelho Farm Rio Vermelho Farm

Rio Vermelho Farm

Taguarı´ River

Santa Aˆngela Farm Santa Aˆngela Farm Santa Aˆngela Farm

Rio Vermelho Farm Rio Vermelho Farm Santa Aˆngela Farm

Santa Aˆngela Farm


Taquarı´ River/Pantanal biome

Negro River/Pantanal biome


Paraguai River/Pantanal biome

Solimo˜es River/Amazonia biome Solimo˜es River/Amazonia biome

Dammed marshy valleys/Cerrado biome Dammed marshy valleys/Cerrado biome



Intact marshy valleys/Cerrado biome




Intact marshy valleys/Cerrado biome Intact marshy valleys/Cerrado biome







Intact marshy valleys/Cerrado biome Intact marshy valleys/Cerrado biome


Dammed marshy valleys/Cerrado biome Dammed marshy valleys/Cerrado biome


Observation about the collection sites

Co´rrego Fundo Farm Co´rrego Fundo Farm Co´rrego Fundo Farm Santa Aˆngela Farm

Place

1.300 1.434

1.300

3.200

1.900

2.500

1.100

1.300

2.800 3.000

2.000

2.000

1.000

1.100

1.200

1.100

1.200 1.100

1.100

1.900

1.100

1.000

1.400

1.200

1.100 1.000

1.000

1.100

1.200

1.000

1.000

A

0.033 0.027

0.036

0.074

0.160

0.104

0.000

0.007

0.121 0.133

0.005

0.087

0.000

0.010

0.000

0.008

0.000 0.000

0.005

0.027

0.008

0.000

0.000

0.000

0.000 0.000

0.000

0.000

0.000

0.000

0.000

Ho

0.031 0.113

0.034

0.371

0.324

0.356

0.053

0.059

0.449 0.559

0.253

0.269

0.000

0.010

0.062

0.034

0.030 0.011

0.030

0.197

0.026

0.000

0.120

0.023

0.041 0.000

0.023

0.013

0.069

0.000

0.000

He

0.109 1.110 1.099 0.135

0.802 )0.047 0.761

0.294

0.161

0.000

0.068

0.149 0.130

0.011

0.190

–

1.000

0.000

0.133

0.000 0.000

0.091

0.073

0.188

–

0.000

0.000

0.000 –

0.000

0.000

0.000

–

–

ta

)0.052

0.546

0.722

1.000

0.873

0.740 0.770

0.977

0.680

–

0.000

1.000

0.765

1.000 1.000

0.833

0.864

0.684

–

1.000

1.000

1.000 –

1.000

1.000

1.000

–

–

f

N, number of individuals sampled; A, mean number of alleles; Ho, observed heterozygosity; He, expected heterozygosity; f, fixation index and ta, apparent outcrossing rate.

GEN 1246 GEN 1248

14 15

.

GEN 1233

individuals

bank number

2

Number of

Germplasm

1

Population

Table 1. Number of individuals sampled, geographical location of the collection areas of the populations and the genetic diversity parameter estimates in 30 O. glumaepatula populations using 10 SSR markers

117

118 silver staining (Bassam, Gresshoff, 1991).

Caetano-Anolles

&

Statistical analysis The loci that did not amplify consistently or had doubtful interpretation were eliminated from the analysis. The allele frequencies were submitted to Fisher’s exact test for Hardy–Weinberg equilibrium as shown by Weir (1996) using the TFPGA program (Miller, 1997). This test was applied through the conventional Monte Carlo method using 10 batches with 1000 permutations per batch. Genetic diversity and the F statistics were estimated considering a random model according to Weir (1996). The allelic frequencies, number of alleles per locus (A), observed (Ho) and expected (He) heterozygosities and Wright F statistics (FIS, FST and FIT) were estimated using the GDA program (Lewis & Zaykin, 2000). The apparent outcrossing rate (^ta ) was calculated from the inbreeding coefficient f^ on the bases of the average expected and observed heterozygosities. The mean number of migrants (Nm) per generation (gene flow) was estimated by the formula: Nm= (¼(1/ FST ) 1), where FST is the measurement of diversity among genetic populations (Wright, 1931). The RST statistic (Slatkin, 1995) and the gene flow (Nm) were also estimated by the RST Calc program (Goodman, 1997). Dendrograms were constructed from the genetic distance matrix generated from the J & C genetic distance coefficient (Jukes & Cantor, 1969) and grouped by the UPGMA clustering method, using the NTSYS program (Rohlf, 1989).

Results Microsatellite marker screening and characterization Of the 133 microsatellite markers screened, 10 were selected because of their greater capacity to detect allelic variability and quality of the amplified products. Seven of these markers were based on repeated dinucleotide sequences (OG7, OG17, OG44, OG89, RM224, RM248 and RM259). A total of 103 alleles were detected and the allelic variability levels in the 30 populations tested ranged from a maximum of 17 alleles for the RM259 locus (Figure 1) to a minimum of 7 for the RM224 and RG4961 loci, with a mean of 10.3 alleles per locus. The allelic frequency distribution ranged from 0.57 for alleles 133 and 137 of locus 4961 to 0.001 for alleles 133 of the OG7 and OG89 loci and allele 117 of locus 4879, with a mean value of 0.097. The number of alleles and their relative frequencies resulted in expected heterozygosity values that ranged from 0.63 for locus 4961 to 0.86 for locus RM259. Statistical analysis For the 10 SSR loci used, the average number of alleles/locus/population ranged from 1.0 to 3.2, with the higher values found in populations 28 (3.2) and population 23 (3 alleles), which were collected in the Pantanal and Cerrado biomes, respectively. The average genetic diversity values ^ e ) (Table 2) were great for the Pantanal (H

Figure 1. Microsatellite polymorphism in three wild O. glumaepatula populations. Allelic variation of locus RM259 visualized in silver-stained denaturing 4% polyacrylamide gels. The left and rightmost lanes contain the 10-bp Ladder size standard (BRL), with the sizes of fragments indicated in bp.

119 Table 2. Genetic diversity parameter estimates in 30 O. glumaepatula populations separated by biomes Biome

N

n

Lp

Ho

He

F

ta

Amazoˆnia

16

2

14

0.004

0.056

0.929

0.0363

Pantanal

67

5

45

0.080

0.257

0.688

0.184

Cerrado

331

23

19

0.018

0.089

0.797

0.113

Mean

414

30

26

0.027

0.115

0.761

0.135

N, number of individuals sampled; n, number of populations; Lp, number of private alleles in the populations; Ho, observed heterozygosity; He, expected heterozygosity; f, fixation index and ta, apparent out crossing rate.

populations 0.257 and low for the Amazon populations (0.056). Populations 22 (0.45) and 23 (0.56) presented the greatest values (Figure 2). A total of 29 exclusive or private alleles were observed, distributed over 12 of the 30 populations assessed. The greatest number of private alleles was observed in populations 12, 28 and 29 from the Pantanal biome. In populations from Cerrado

biome, which were collected in areas of dammed marshy valleys that had undergone fragmentation, ^ e ¼ 0:518) and private alleles a higher diversity (H (29) than those collected in intact marshy valley ê areas, which presented 14 private alleles and a H of 48.9%. The fixation index (f^ ) and apparent crossing rate (^ta ) estimated for the 30 populations were

Figure 2. Brazilian geographical distribution of the natural populations of O. glumaepatula.

120 0.761 and 0.135, respectively (Table 2). Considering each biome, the (f^ ) values detected were 0.929, 0.688 and 0.797, and the ^ta values were 0.036, 0.184 and 0.113 for the Amazon, Pantanal and Cerrado populations, respectively (Table 2). Total inbreeding rate for the whole set of populations (FÎT ) was 0.968 and the fixation index (FÎS ) average within populations was estimated as 0.794. The FST statistic and RST that measure genetic diversity among populations assuming different evolutionary models for the SSR loci, were 0.847 and 0.901, respectively. These estimates were congruent, and indicated that most of the molecular diversity was found among populations (Table 2). For each biome, the among population diversity was low for the Pantanal (F^ST ¼ 0:713) and high for Cerrado (F^ST ¼ 0:868) (Table 3). Gene flow (Nm) among populations, based on the ^ ST estimate, was 0.0268, that is, one migrant in R every 38 generations, indicating a low pollen, seed and floating plant migration rate. This value differed significantly from zero by the bootstrap resampling method, indicating a high rate of selffertilization within each O. glumaepatula population. From the J & C genetic distance data among the 30 populations, the dendrogram (Figure 3) shows clustering according to the O. glumaepatula population location. The Cerrado biome populations were subdivided into two main groups, one being formed by the populations from the Amazon and Cerrado (intact marshy valleys), and the other from the Pantanal and Cerrado (dammed marshy valleys) populations. The genetic distance analysis indicated that populations 1 and 2 are closest genetically (0.00) and populations 23 and 28 are more distant genetically (0.6905). Only four populations belonging to the Cerrado biome (20, 21, 22 and 23) and were located in neighboring geographic areas, clustered outside of the large group of the Cerrado biome.

Discussion Knowledge of the distribution of the genetic variability among and within O. glumaepatula populations is essential for adoption of strategies to conserve this germplasm in ex situ and in situ conditions. Depending on the germination rate, a minimum of 3–24 individuals per population was analyzed. The small size of some populations did not interfere in the genetic analysis, since, for genetic population structure studies in preferentially autogamous species, it is more coherent and informative to have a greater number of populations with a fewer number of individuals, than fewer populations with a greater number of individuals (Vencovsky & Crossa, 2003). The 10 SSR loci analyzed presented high multiallelism, with a mean of 10.3 alleles per locus. The mean value of the informative contents of these markers (PIC – Polymorphism Information Content), was 0.76, close to the PIC of 0.65 reported by Ishii, Xu and McCouch (2001), using 24 SSR loci to genotype five O. glumaepatula accessions. A high level of genetic variability was detected in 30 O. glumaepatula populations using SSR markers, in comparison to the molecular characterization obtained from allozyme and RAPD markers (Akimoto, Shimamoto & Morishima, 1998; Buso, Rangel & Ferreira, 1998) using O. glumaepatula populations collected in similar areas at the Amazon and Pantanal biomes. This indicates that SSR markers provide a useful tool in estimating genetic parameters, in accordance with many recent studies (Davierwala et al., 2000, Song et al., 2003, Zhou, Xie & Ge, 2003), better than other marker systems, due to the risk to underestimate the number and frequency of alleles in wild populations. Populations 22 and 23 showed the greatest genetic diversities (He) and crossing rates than the

Table 3. Wright’s F statistics estimates and the number of migrants per generation (Nm) considering the 30 O. glumaepatula populations, separated by biomes Estimative

FÎS

FÎT

F^ST

Nm

Amazonia biome

0.897

0.983

0.831

0.051

Pantanal biome

0.762

0.932

0.713

0.100

Cerrado biome

0.808

0.974

0.868

0.038

121

Figure 3. Genetic divergence pattern among 30 wild Oryza glumaepatula populations, obtained by the UPGMA grouping and J & C genetic distance coefficient using SSR data. Group CB is from Cerrado biome, group AB is from Amazonia biome and group PB is from Pantanal biome. Dotted squares indicate populations that grouped outside the Cerrado biome cluster.

other 28 populations that are completely isolated from rice fields (Table 2). The proximity of the commercial cropping areas to the O. glumaepatula populations could have been responsible for the introgression from cultivated rice, and consequently, increasing He, since some level of apparent outcrossing rate (11%) was detected in Cerrado populations. As the rice cropping area in Cerrado has being increased in recent years, its impact on O. glumaepatula populations should be determined to evaluate the risks of loss of the genetic identity of wild populations. Human interference also resulted in spatial isolation of O. glumaepatula populations in Cerrado biome. These facts could explain the restricted gene flow that resulted in a higher number of exclusive alleles (Slatkin, 1985) identified in the dammed marshy valley (29 private alleles) when compared with population from intact marshy valleys (14 private alleles). Analysis of O. glumaepatula populations in the three biomes (Table 2) showed the highest outcrossing rate in Pantanal populations (18.4%). There is consequently a great possibility of gene flow occurring among populations located in the hydrographic basins in the Amazon and Pantanal regions because plants become detached and float along the various rivers. However, it should be pointed out that only two populations were

analyzed in the Amazon biome, and the number of samples was limited when compared with five populations from the Pantanal biome and 23 from the Cerrado biome. This probably contributed to different cross-fertilization rates of 0.036, 0.184 and 0.113, for Amazon, Pantanal and Cerrado biomes, respectively. ^ ST indexes indicated that The high F^ST and R most of the existing genetic variability is found among populations, corroborating with all estimates reported for self pollinating species (Frankel, Brown & Burdon, 1998). For each biome, diversity among population was smaller for the Pantanal (F^ST ¼ 0:713) and greater for the Cerrado (F^ST ¼ 0:868) areas. The among population diversity (FST) is a measure of a remote inbreeding due to subdivision. As populations become more isolated and gene flow is reduced, genetic drift tends to manifest within each one by fixing different alleles, resulting, in the longer term, in the loss of intrapopulation genetic variability and in an increase of the genetic divergence among them. Therefore these results, when the three different biomes are compared, are in conformity with the geographic location of these populations. In the populations collected in the Cerrado marshy valleys, where geographic isolation predominates, pollen and seed dispersion is hindered, limiting or

122 even preventing gene flow among populations. The impact caused by human interference in the Cerrado region, resulting from expansion of agricultural and livestock in rearing areas is probably the main cause of the higher level of diversity among the wild O. glumaepatula populations, since such populations are subdivided, becoming genetically different from each other due to fixation of different alleles. The F^ST estimate was smaller for the Pantanal biome than for the Cerrado biome populations, probably due to the proximity among the populations in the Pantanal areas, allowing gene flow or individual plant migration among populations. High F^ST values have also been observed in related studies with O. glumaepatula (Buso et al., 1998) and O. granulate (Qian, Ge & Hong, 2001). Lower F^ST index, however, were observed in O. rufipogon (Song et al., 2003), that is perennial– annual, outbreeding Asian wide rice (Akimoto, Shimamoto & Morishima, 1999). The total inbreeding for the whole set of populations (FÎT ¼ 0:968) explains the high deviations from Hardy–Weinberg equilibrium frequencies, with great excess of homozygotes, in accordance with Buso et al. (1998), where a significant higher number of individuals per populations of O. glumaepatula collected in similar areas in Brazil was sampled. The FIT index, when separated by biome, was smaller for the Pantanal biome, which agrees with the smaller diversity index found among its populations. The present study reveals that high inbreeding values and relatively low diversity were observed within O. glumaepatula populations using SSR markers. However, the among population genetic variability in a certain sense compensates the within population homogeneity. This pattern of genetic diversity partitioned among population is correlated with the mating system and short-lived growth (Hamrick & Godt, 1989), suggesting that the O. glumaepatula species are essentially self-pollinating. Such a reduced intrapopulation diversity, associated with other factors, clearly increases the probability of extinction of some populations. There is a smaller risk of degradation of the O. glumaepatula populations in the Amazon region because areas of its occurrence are of difficult access and not suitable for agriculture. However, this risk is imminent in the Cerrado and Pantanal regions due to the fragmentation of the frontier of areas resulting from the expansion of agriculture

and livestock rearing. The Brazilian rice-breeding program also has great interest to preserve O. glumaepatula populations, since this species has being used as sources of genes with agronomic interest for cultivated rice. From the 30 populations of O. glumaepatula analyzed at the present study, 6 were collected from locations with high aluminum concentrations (2.5 ppm) and high iron concentrations (594 ppm), in the Cerrado areas, which are the most endangered regarding human interference regions. Studies are in progress to transfer genes related to aluminum and iron stresses tolerance from these populations to commercial rice cultivars. Extinction of O. glumaepatula populations, especially from Cerrado, which showed the lowest gene flow estimates, will cause the loss of important genes that could be used in the rice-breeding program. These results are extremely useful for the precise implementation of conservation management, and for designing strategies for collection, such as: (1) urgent establishment of protected areas for in situ conservation, with a safe isolation distance from cultivated rice to avoid gene flow and human disturbance. The preference for in situ conservation is based on the observation that O. glumaepatula seeds decline in vigor very fast in germplasm bank, in comparison to the cultivated rice, demanding seed multiplication every 5 years; (2) implementation of additional collection expeditions, particularly in most threatened areas, to identify, collect and preserve ex situ the existing genetic variability; (3) Sample more populations rather than sampling a large number of individuals from a single population. In addition, as O. glumaepatula has been employed as an important genetic resource for Brazilian rice breeding, a detailed knowledge of the amount and apportionment of genetic variation within and between natural populations from Brazil will allow for the identification of genetically divergent plants for use in inter-specific crosses with cultivated rice.

Acknowledgements We acknowledge financial support from the Ministry of Education – CAPES doctoral fellowship to M.I.Z.; Brazilian National Research Council – CNPq to research fellowship to R.P.V.B. and

123 graduate fellowship to T.C.O.B.; Brazilian Agricultural Research Corporation – EMBRAPA to graduate fellowship to M.R.M.

References Akagi, H., Y. Yokozeki, A. Inagaki & T. Fujimura, 1996. Microsatellite DNA markers for rice chromosomes. Theor. Appl. Genet. 93: 1071–1077. Akimoto, M., Y. Shimamoto & H. Morishima, 1998. Population genetic structure of wild rice Oryza glumaepatula distributed in the Amazon flood area influenced by its lifehistory traits. Mol. Ecol. 7: 1371–1381. Akimoto, M., Y. Shimamoto & H. Morishima, 1999. The extinction of genetic resource of Asian wild rice Oryza rufipogon Griff: a case study in Thailand. Gen. Res. Crop Evol. 46: 419–425. Brar, D.S. & G.S. Khush, 1997. Alien introgression in rice. Plant Mol. 35: 35–47. Bassam, B.J., G. Caetano-Anolles & P.M. Gresshoff, 1991. Fast and sensitive silver staining of DNA in polyacrylamide gels. Anal. Biochem. 196: 80–83. Brondani, C., R.P.V. Brondani, P.H.N. Rangel & M.E. Ferreira, 2001. Development and mapping of Oryza glumaepatula-derived microsatellite markers in the interspecific cross O. glumaepatula · O. sativa. Hereditas 134: 59–71. Brondani, C., P.H.N. Rangel, R.P.V. Brondani & M.E. Ferreira, 2002. QTL mapping and introgression of yieldrelated traits from Oryza glumaepatula to cultivated rice (Oryza sativa) using microsatellite markers. Theor. Appl. Genet. 104: 1192–1203. Buso, G.S.C., P.H.N. Rangel & M.E. Ferreira, 1998. Analysis of genetic variability of South American wild rice populations (Oryza glumaepatula) with isozymes and RAPD markers. Mol. Ecol. 7: 107–117. Chen, X., S. Temnykh, Y. Xu, Y.G. Cho & S.R. McCouch, 1997. Development of a microsatellite framework map providing genome-wide coverage in rice (Oryza sativa L). Theor. Appl. Genet. 95: 553–567. Collevatti, R.G, R.P.V. Brondani & D. Grattapaglia, 1999. Development and characterization of microsatellite markers for genetic analysis of a Brazilian endangered tree species Caryocar brasiliense. Heredity 83: 748–756. Davierwala, A.P., K.V. Chowdari, S. Kumar, A.P.K. Peddy, P.K. Ranjekar & V.S. Gupta, 2000. Use of three different marker systems to estimate genetic diversity of Indian elite rice varieties. Genetica 108: 269–284. Doyle, J.J. & J.L. Doyle, 1987. Isolation of plant DNA from fresh tissue. Focus 12: 13–15. Frankel, O.H., A.H.D. Brown & J.J. Burdon, 1998. The Conservation of Plant Biodiversity. Cambridge University Press, Cambridge. Goodman, S.J., 1997. RSTCalc: a collection of computer program for calculating estimates of genetic differentiation from microsatellite and determining their significance. Mol. Ecol. 6: 881–885.

Hamrick, J.L. & M.J.W. Godt, 1989. Allozyme diversity in plants, pp. 43–63 in Plant Population Genetics, Breeding and Genetic Resources, edited by A.H.D. Brown, M.T. Clegg, A.L. Kahler & B.S. Weir. Sinauer, Sunderland, Massachusetts. Jukes, T.H. & C.R. Cantor, 1969. Evolution in protein molecules, pp. 121–23 in Mammalian Protein Metabolism, edited by H.N. Munro. Academic Press, New York, NY. Ishii, T., Y. Xu & S.R. McCouch, 2001. Nuclear and chloroplast microsatellite variation in A-genome species of rice. Genome 44: 658–666. Khush, G. S., 1997. Origin, dispersal, cultivation and variation of rice. Plant Mol. Biol. 35: 25–34. Lemes, M.R., R.P.V. Brondani & D. Grattapaglia, 2002. Multiplexed systems of microsatellite markers for genetic analysis of Mahogany, Swietenia macrophylla King (Meliaceae), a Threatened Neotropical Timber Species. J. Hered. 93: 287–291. Lewis, P.O. & D. Zaykin, 2000. Genetic Data Analysis: Computer Program for the Analysis of Allelic Data. Version 1.0 (d15) Available from http://alleyn.eeb.uconn.edu/gda/2000. Miller, M., 1997. Tools For Population Genetic Analyses (TFPGA) 1.3: A Windows Program for Analyses of Allozyme and Molecular Population Genetic Data. Qian, W., S. Ge & D. -Y. Hong, 2001. Genetic variation within and among populations of a wild rice Oryza granulata from China detected by RAPD and ISSR markers. Theor. Appl. Genet. 102: 440–449. Rohlf, F.J., 1989. NTSYS-Pc: Numerical Taxonomy and Multivariate Analysis System. Exeter Publisher, New York, NY. Slatkin, M., 1985. Rare alleles as indicators of gene flow. Evolution 39: 53–65. Slatkin, M., 1995. A measure of population subdivision based on microsatellite allele frequencies. Genetics 130: 457–462. Song, Z.P., X. Xu, B. Wang, J.K. Chen & B.-R. Lu, 2003. Genetic diversity in the northernmost Oryza rufipogon populations estimated by SSR markers. Theor. Appl. Genet. 107: 1492–1499. Temnykh, S., W.D. Park, N. Ayres, S. Cartinhour, N. Hauck, L. Lipovich, Y.G.T. Cho & S.R. Ishii McCouch, 2000. Mapping and genome organization of microsatellite sequences in rice (Oryza sativa L.). Theor. Appl. Genet. 100: 697–712. Vaughan, D.A., H. Morishima & K. Kadowaki, 2003. Diversity in the Oryza genus. Curr. Opin. Plant Biol. 6: 139–146. Vencovsky, R. & J. Crossa, 2003. Measurements of representativeness used in resources conservation and plant breeding. Crop Sci. 43: 1912–1921. Weir, B.S., 1996. Genetics Data Analysis II – Methods for Discrete Population Genetic Data. Sinauer Associates, Inc. Publishers, Suderland, MA. Wright, S., 1931. The genetical structure of populations. Ann. Eugen. 15: 395–420. Wright, J.M. & P. Bentzen, 1994. Microsatellites: genetic markers of the future. Rev. Fish Biol. Fish. 4: 384–388. Zhou, H.-F. & Z.-W.S. Xie Ge, 2003. Microsatellite analysis of genetic diversity and population genetic structure of a wild rice (Oryza rufipogon Griff) in China. Theor. Appl. Genet. 107: 332–339.