\ ..lN '-\ C!l I 1- \
National Research Conseil national l+I Council Canada de recherches Canada
Reprinted from Reimpression de la · Genome Genome
Genetic diversity within the species Arachis duranensis Krapov. & W;C. Gregory, a possible progenitor of cultivated peanut
H.T. Stalker, J.S. Dhesi, and G. Kochert
Volume 38 • Number 6 • 1995
Pages 1201-1212
Canada 1201 Genetic diversity within the species Arachis duranensis Krapov. & W.C. Gregory, a possible progenitor of cultivated peanut
H.T. Stalker, J.S. Dhesi, and G. Kochert
Abstract: Eighteen accessions of a diploid wild peanut species (Arachis duranensis) were analyzed using morphological, intercrossing, cytological, and RFLP data. Abundant variation was found for morphological characters and for RFLP patterns both between and within accessions, and each
accession could be uniquely identified by RFLP pattern. Several plants were found to be F1 hybrids between different accessions, indicating that intercrossing had occurred when these were planted for seed increase. Patterns of RFLP diversity were found to correspond with geographic distribution. Analysis of the number of RFLP fragments observed per accession indicates that additional field collections of this complex of taxa will yield additional genetic variability. Key words: peanut, Arachis hypogaea, Arachis spp., RFLP, variation, genetic diversity.
Resume : Dix-huit accessions d'une espece diplo!de sauvage d'arachide (Arachis duranensis) ont ete examinees a l'aide de criteres morphologiques, de croisements, d'analyses cytogenetiques et RFLP. Une abondante variation a ete observee tant au niveau morphologique qu'au niveau des RFLP parmi et entre les accessions. Chaque accession pouvait etre identifiee a I' aide de motifs RFLP distinctifs. Plusieurs plantes se sont averees des hybrides F 1 entre differentes accessions, indiquant que des croisements s'etaient produits naturellement !ors de la multiplication des graines. La diversite genetique au niveau des RFLP etait correlee avec I' origine geographique des accessions. Une analyse du nombre de fragments RFLP observes pour chaque accession.indique que des recoltes additionnelles permettraient de faire ressortir encore davantage de variabilite genetique chez ce complexe taxonomique. Mots cles : arachide, Arachis hypogaea, Arachis spp., RFLP, variation, diversite genetique. [Traduit par la Redaction]
Introduction donors to the cultivated peanut, including A. cardenasii Krapov. & W.C. Gregory, A. villosa Benth., A. duranensis The cultivated peanut, Arachis hypogaea L., originated in Krapov. & W.C. Gregory, and A. spega:u.inii (=A. duranensis) the southern Bolivia - northern Argentina region of South (see Stalker and Moss 1987; Kochert et al. 1991). Arachis America. The species is an annual allotetraploid with A duranensis is the most widely accepted donor species of the and B genomes (2n = 4x = 40) and is widely grown in A genome to cultivated peanut (Kochert et al. 1991 ). An tropical and subtropical regions. Arachis monticola Krapov. analysis of intraspecific variation in A. duranensis would et Rigonii is the only closely related tetraploid species be valuable for determining its status as a progenitor species native to northern Argentina and interspecific hybrids of the cultivated peanut, for understanding evolution in between the two taxa are fully fertile (Stalker and Moss the genus Arachis, and for assessing issues of germplasm 1987). All other closely related species of Arachis are conservation. diploids (2n = 2x = 20) with A, B, or D genomes. At least Although A. batizocoi Krapov. & W.C. Gregory has four species have been proposed as possiblii A-genome been the assumed B-genome donor to A. hypogaea (Smartt et al. 1978), recently obtained molecular evidence indicates Corresponding Editor: J.P. Gustafson. that it was probably not involved in the ancestry of the Received January 20, 1995. Accepted July 28, 1995. cultivated peanut (Kochert et al. 1991; Halward et al. 1992; H.T. Stalker. Department of Crop Science, North Carolina Paik-Ro et al. 1992; Bianchi-Hall et al. 1993; Stalker et al. State University, Raleigh, NC 27695-7629, U.S.A. 1994). The donor of the second genome of cultivated J.S. Dhesi. Department of Biology, North Carolina Central peanut appears, based on DNA restriction fragment length University, Durham, NC 27707, U.S.A. polymorphism (RFLP) evidence, to be A. ipaensis Krapov. G. Kochert. Department of Botany, University of Georgia, & W.C. Gregory or a closely related taxon (Kochert et al. Athens, GA 30602, U.S.A. 1991; G. Kochert and H.T. Stalker, unpublished data).
Genome, 38: 1201-1212 (1995). Printed in Canada I Imprime au Canada l I \ 1202 Genome, Vol. 38, 1995
Fig. 1. Locations of the Arachis accessions analyzed for variation.
Cusco• 500 Km
PERU •Trinidad BOLIVIA Arequipa • 30092 •La Paz • • 30084 • Cochabamba
esanta Cruz
20·
-Tropic- ol Capricorn- PARAGUAY Antofagasta e36003 36036• • 30061 •30060 .•30064 10038 •Asuncion ARGENTINA
San •Miguel Resistencia • de Tucuman 50· I
Arachis duranensis is an annual species native to north to several major disease and insect pests, including Puccinia ern Argentina and southern Bolivia (Fig. 1). The species is arachidis (Subrahmanyam et al. 1985), peanut stunt virus placed in section Arachis because it has a tap root and (Hebert and Stalker 1981), and potato leafhopper (Stalker pegs that elongate in a vertical direction after they penetrate and Campbell 1983) have been reported. The species also the soil. The first known accession of A. duranensis (col has moderate levels of resistance to corn earworm (Stalker lection K 7988, PI 212893) was collected in 1953 in north and Campbell 1983), Arachis duranensis will hybridize ern Argentina by A. Krapovickas (Gregory et al. 1973). with the 20-25 other species in section Arachis (Stalker In 1977 and 1980, 16 additional accessions were obtained et al. 1991), including A. hypogaea and A. monticola (Simpson and Higgins 1984; Table 1) and have since been (Seetheram et al. 1973). Hybrids with A-genome taxa gen analyzed for plant morphology (Stalker 1990), seed storage erally have fertility levels ranging from 25 to 80% and proteins (Bianchi-Hall et al. 1993; Singh et al. 1991 ), and meiotic chromosomes generally pair as bivalents. When isozymes (Lu and Pickersgill 1993; Stalker et al. 1994) as the B-genome species A. batizocoi or the D-genome species part of larger investigations of section Arachis. Accessions A. glandulifera Stalker are crossed with A. duranensis, of A. duranensis clustered into one broadly defined group hybrids are sterile and meiosis is irregular (Stalker and using morphological data (Stalker 1990) and into several Moss 1987; Stalker et al. 1991). groups using isozymes (Stalker et al. 1994). Seed storage The objective of this investigation is to characterize proteins are not useful for defining the species (Bianchi-Hall variation within the species A. duranensis. A combination et al. 1993). of morphological, cytological, and molecular data will be One accession of A. duranensis (K 7988) was analyzed used to describe variability found among the accessions using RFLPs (Kochert et al. 1991) and the single primer and the implications of this variability will be presented. polymerase chain reaction (Halward et al. 1992). Both studies grouped A. spegazzinii with A. duranensis and Materials and methods Krapovickas and Gregory (1994) described the respective taxa as A. duranensis. Eighteen accessions belonging to A. duranensis and 1 acces In addition to evolutionary and botanical interests, acces sion each of the related species A. magna Krapov. & sions of A. duranensis represent valuable genetic resources W.C. Gregory and A. kempff-mercadoi Krapov. & for improving the cultivated peanut. Resistance or immunity W.C. Gregory were propagated at the Sandhills Agricultural Stalker et al. 1203
Table 1. Accessions of Arachis analyzed morphologically, cytologically, or with molecular markers.
Year Province and Accession Collectora PI Species collected country Latitude Longitude
7988 K 219823 duranensis 1954 Salta, Argentina 22.19 63.43 l0038b,c GKP 262133 duranensis 1959 Salta, Argentina 24.47 65.32 30060c GKBSPSc 468197 duranensis 1977 Jujuy, Argentina 24.22 65.07 30061 GKBSPSc 468198 duranensis 1977 Jujuy, Argentina 24.16 65.12 30064 GKBSPSc 468200 duranensis 1977 Jujuy, Argentina 24.23 65.07 30065 GKBSPSc 468201 duranensis 1977 Salta, Argentina 23.04 63.53 30067 GKBSPSc 468202 duranensis 1977 Salta, Argentina 23.03 63.56 30069 GKBSPSc 475844 duranensis 1977 Tarija, Bolivia 21.48 63.33 30070 GKBSPSc 475845 duranensis 1977 Tarija, Bolivia 21.53 63.38 30071 GKBSPSc 475846 duranensis 1977 Tarija, Bolivia 21.41 63.45 30072 GKBSPSc 475847 duranensis 1977 Tarija, Bolivia 21.41 63.44 30073 GKBSPSc 468319 duranensis 1977 Tarija, Bolivia 21.44 63.33 30074 GKBSPSc 468320 duranensis 1977 Tarija, Bolivia 21.26 63.27 30075 GKBSPSc 468321 duranensis 1977 Tarija, Bolivia 21.18 63.27 30077 GKBSPSc 468323 duranensis 1977 Chuguisaca, Bolivia 20.37 63.13 30078 GKBSPSc 468324 duranensis 1977 Chuguisaca, Bolivia 20.45 63.08 36003 KSBScC 475883 duranensis 1980 Salta, Argentina Unknown Unknown 36036 KS Sc 475887 duranensis 1980 Salta, Argentina Unknown Unknown 30084 GKBSPScZ 468330 kempff-mercadoi 1977 Santa Cruz, Bolivia 17.19 63.18 30092 GKSSc 468337 magna 1977 Santa Cruz, Bolivia 16.36 62.14
"K, A. Krapovickas; G, W.C. Gregory; P, J. Pietrarelli; B, D.J. Banks; S, C.E. Simpson; Sc, A. Schinini; C, L. Coradin; and Z, H. Zurita 0. bThis accession has large leaf (normal) and small leaf (mutant) forms. cAccessions previously named A. spegazzinii.
Research Station, North Carolina (Table l; Fig. 1) or in a Based on the morphological analysis, 11 accessions rep greenhouse at the University of Georgia. For the morpho resenting the most unique types were used in crossing pro logical analysis, three plants each of the 18 A. duranen grams during 1991 and 1992. Flowers were hand emas sis accessions were randomly selected and scored for culated during the afternoon between 3 and 5 p.m. and 18 reproductive and 30 vegetative traits 90-95 days after pollinated the following morning between 8 and 10 a.m. plantings during the summer of 1989. The large leaf (1.1.) Pegs with the emasculated attached flower were tagged, and small leaf (s.l.) types of accession 10038 were treated and seeds from crosses harvested, dried, and planted the as separate entries for this study. Data for the following following spring in the greenhouse. At least 300 pollen traits were collected: presence or absence of flowers on grains of two flowers collected on 2 separate days were the mainstem; mainstem and lateral branches erect or pros analyzed as an estimate of male fertility. Meiotic chromo trate; length of mainstem; length of longest lateral branch; some relationships of hybrids were analyzed according to relative amount of lateral branching; length and width of the methods described by Stalker et al. (1991). In addition, mainstem and lateral leaves and leaflets; shape of main somatic chromosomes of six accessions, including 7988, stem and lateral leaflets; shape of the leaflet apex on main 30061, 30064, 30067, 30070, and 10038, were karyotyped stem and lateral leaves; stipule length and petiole length of using procedures described in Stalker and Dalmacio (1981). mainstem and lateral leaves; internode length between Eighteen accessions were evaluated for variability using leaflets on mainstem and lateral leaves; numbers of tri conventional RFLP methods as described by Kochert et al. chomes on stems, upper and lower leaflet surface, and (1991). DNA was isolated from leaves derived from single leaflet margins and pegs; length and type of trichomes on plants. Three to five plants from each accession were sep leaf and peg; flower height, width, shape and color, and arately isolated to check for variation within accessions. hypanthium length; peg length, width, and shape; pod retic Total DNA from each plant was digested with one of four ulation and beak size and seed length, width, color, retic restriction enzymes (DraI, EcoRV, HaeIII, HindIII) and ulation, and shape; and flowering pattern on lateral stems. separated in 0.8% agarose gels. Following Southern blotting, In addition, the following six derived variables were eval a set of 34 probes that had previously been mapped on uated: mainstem and lateral leaf length/width ratios, main the peanut RFLP map (Halward et al. 1993) were used as stem and lateral leaflet length/width ratios, and pod and hybridization probes. Because 6 probes did not detect vari seed length/width ratios. The three values for the three ation within or between accessions, only the 28 variable plants of an accession were averaged for each trait, pro probes were analyzed. These probes provided good coverage viding one value for each of the 54 variables measured for of the genome and included markers from most of the link an accession. age groups (Fig. 2). Fig. 2. Highlighted RFLP probes were used in surveying genetic variation in A. duranensis. All probes are cDNAs previously mapped in a F2 population I\) derived from a A. cardenasii X A. stenosperma cross (Halward et al. 1993). 0 .j::.. 1a 28 3 4 6 8 Xuga.cro14 Xuga.cr2n 6.6 _-fl-i%:1!A~#i~?t?td Xuga.cr105 Xuga.pg108 Xuga.csQ.46 17.9- 11.7- t 23.5- Xuga.cr132 30.2- 5.2 - Xuga.cr287 Xuga.cr209 rn®~g:rnrnm 21.9 - 1b 14.9- Xuga.cr108 11.5- Xuga.cr.266 Xuga.c:s065 :::::: 2b 15.1- 12.2- :::::::::::::::: 30.4 - 4.7- Xuga.r118b 9.2- Xuga.cr137 16.6- ...... ,.,..·~··:•:•:•:•:•:•:•:•:•:: 3.7--= ~~~~~~~·::~:t:: Xugacr078 Xuga.cr.246 1-ii!i+'i!h~.)foftl 21.9 - 6.5- 20.3- 11.8- 19.1- '.··~~tttt\ 26.9 - ::::::::: Xuga.cs046 24.2- 9.3- 13.0- Xuga.cr.225 Xuga.c:s001 9.5 - f:~M#.? e.s -f Xuga.cr197 15.4- Xuga.cr186 16.6-ll Xuga.cs026 Xuga.cs063 Xuga.pg068 3.6--= Xuga.cr163 17.2- 9.7- Xuga.cr.229 Xuga.cr145 19.7- Xugacro79 18.9- 9 Xuga.cr244 Xuga.pg069 ~ Xuga.cr.211 Xuga.cross Xugacr.274 10.7- 17.9-
7 Xuga.cs023 i:~'i#.~(:•••••••••••••••••• Xuga.cr171 Xuga.cr.288 8.7- 5.9- Xuga.cc315 Xuga.cr.235 Xuga.cr129 6.7- Xuga.cr021 Xuga.r118a 13.1- 1.3 Xuga.cr.219 13.0- Xuga.cr.239 Xuga.cro10 ,,,_~-~· Xuga.cr13a 23.6 - 33.0 -
26.1- Xuga.cr256 8.0 - f.%Ql..'iiimfffff 5a Xuga.cr.237 i~ii~Mii\~M?t•::• 5.4- Xuga.cr.250 18.3- ,'~ii¥#¥EHF 1Q tXuga.cr272 21.0- 26.4- 5b G) tt-Xuga.c:s035 Xuga.c:s029 9.1- CD :::J 6.2- Xuga.c126a 0 Xuga.pg082 Xuga.cr189 3 9.8- 16.0-,,~~-" CD Xuga.cr194 1ll < 3.8 - Xuga.cr255 11 0 illlll 5.4 - Xuga.cr140 (;) 7.9- 6.8 - Xuga.cro32 [ Xuga.cr086 co Xuga.cr177 3.2 _; ·:::;::::::::::: 26.5 - (0 15.1- (0 .7 II (J1 Xuga.cr107 Xuga.CS028 Stalker et al. 1205
Fig. 3. Principal component groupings of the 18 accessions analyzed for plant morphological traits.
•30069
•30077
4
•30075 •36036 II 30078 30060 •10038 SL 36003 7988 •10038 LL -4 30067 •30074 4 30073• •30061
•30065 •30071 •30070 •30072 •30064 -4
I
RFLP banding patterns were analyzed using one restric for 30065 X 30077, 30064 X 30077, and 36036 X 30077, tion enzyme for each probe (the number of unique frag respectively, to 87% for 30071 X 30069. Several other ments was scored for each accession). Each RFLP band hybrid combinations had greater than 95% fertility but was scored as present or absent in each plant. The com these could not be confirmed morphologically and thus bined data were analyzed using parsimony by the PAUP the data are not included in this report. Most intraspecific computer program (Swofford 1991). Heuristic searches hybrids, however, had fertility levels ranging from 25 to were performed with the Mulpars option set. Genetic dis 75% (Table 2). Hybrids between accessions 7988 and tance data were obtained from PAUP and used to construct 10038 (=A. spegazzinii) ranged from 54.7 to 58.5% fertility neighbor-joining trees using the JOIN program of the PHYLIP and univalents or multivalents were not observed. Most package obtained from br. Joe Felsenstein, University of other hybrids also had 10 bivalents in pollen mother cells Washington. (PMCs). Three crosses averaged more than 0.3 univalents per cell, including 30064 X 30071, 36036 X 30067, and Results 36036 X 30069. Each of these cases represents a cross between a southern and a northern accession. Comparing the Analyses of plant morphological traits clustered the average fertility levels with corresponding univalents or 18 accessions analyzed into two small and one diverse multivalents in hybrids listed in Table 2, there is no apparent group (Fig. 3). Accessions of the previously named species trend to account for sterility based on chromosome A. spegazzinii (10038 1.1., 10038 s.l., and 30060) clustered associations. together and .had negative principal component I values; Multivalents were found at a low frequency (0.01- 30069 and 30077 formed a second group, whereas a con 0.26/PMC) in 12 of the 27 hybrids. Nine of these crosses tinuous and variable group included the remaining 14 acces had either accession 30069 or 30075 as one of the parents. sions. The most important characters separating accessions The presence of quadrivalents most likely represents chro on principal component I were the lateral leaf length/width mosome translocation differences among the accessions. ratio, prominence of the pod peak, mainstem leaf length, Additional evidence of translocations was found after karyo mainstem leaf width, and lateral leaf width. Accessions typing six A. duranensis accessions (Table 3). Five of the six were separated on principal component II by the amount of accessions had a symmetrical chromosome 1, whereas in venation on pods, lateral leaf length, and the shape of the 30067, chromosome 1 was submedian. Little differentiation apex of lateral leaves. There was no apparent relationship was observed for chromosomes 2, 3, 4, 5, and 7, whereas between accessions within a group and geographical origin. only 7988 had an asymmetrical chromosome 6. Several For example, the collections of A. duranensis used in this accessions had either an asymmetrical chromosome 9 or 10 study can be separated into northern and southern groups but these two chromosomes were median in other geno (Fig. l); however, northern and southern accessions do types. Additional karyotypic evidence for intraspecific dif not cluster separately in the principal components analysis. ferentiation is seen in the different relative lengths of chro Intraspecific hybrids were obtained for 31 combinations mosomes 9 and 10. For example, chromosome 10 was only (Table 2). Fertility of F 1s ranged from 4.7, 5.6, and 9.7% 67% of the length of chromosome 9 for accession 7988, 1206 Genome, Vol. 38, I 995
Table 2. Fertility and chromosome associations in F 1 intraspecific A. duranensis hybrids.
Meiosis Fertility Mean frequency/cell: No. of Identity plants Mean No. of plants No. of cells I II III IV Chiasma
7988 x 30070 1 47.5 23 0.21 9.83 0 0 18.82 7988 x 30075 4 30.9 3 72 0.06 9.97 0 0 19.42
36036 x 30065 6 54.4 2 100 0.03 9.92 0.01 0.01 19.32 36036 x 30067 5 53.5 2 54 0.19 9.91 0 0 19.28 36036 x 30069 7 77.3 5 137 0.12 9.74 0 0.07 18.86 36036 x 30071 2 60.8 15 0 9.46 0 0.26 18.86 36036 x 30075 6 30.8 3 72 0.49 9.93 0 0.01 18.86 36036 x 30077 4 9.7 2 57 0.04 9.94 0 0 19.33
30065 x 30069 62.1 50 0.16 9.92 0 0 19.36 30065 x 30070 4 29.9 50 0.08 9.98 0 0 19.44 30065 x 30071 1 75.0 30065 x 30075 2 26.3 2 100 0.20 9.84 0 0.03 18.94 30065 x 30077 4.7
30067 x 30064 1 61.3 30067 x 30069 6 51.9 30067 x 30071 6 78.8 5 146 0.25 9.94 0 0 19.29 30067 x 30075 9 23.0 9 374 0.15 9.95 0 0 19.40
30064 x 30069 6 73.5 6 300 0.11 9.87 0 0.03 19.34 30064 x 30070 2 27.7 1 52 0.08 9.96 0 0 19.16 30064 x 30071 2 65.2 2 100 0.18 9.85 0 0 18.78 30064 x 30075 7 17.9 5 244 0.11 9.84 0 0.04 19.02 30064 x 30077 2 5.6 4 161 0.11 9.89 0 0.02 19.08
30070 x 30069 7 38. l 3 100 0.08 9.68 0 0.09 18.71 30070 x 30071 1 27.3 1 50 0 10.00 0 0 19.34 30070 x 30075 9 69.3 4 193 0.11 9.87 0 0.03 19.33 30070 x 30077 3 28.9 2 59 0.14 9.93 0 0 19.20
30071 x 30069 10 87.4 2 55 0 9.74 0 0.05 18.80 30071 x 30075 6 28.1 3 150 0.20 9.78 0 0.03 18.90
30069 x 30075 13 15.8 8 427 0.17 9.56 0 0 18.42
10038 x 7988 3 54.7 2 0 10.00 0 0 20.00
7988 x 10038 3 58.5 50 0 10.00 0 0 19.50
7988 ® 6 98.0 6 171 0.05 9.98 0 0 19.63
whereas chromosome 10 was 80% of the length of chro clear RFLP patterns (for example, the three probes shown mosome 9 in accession 30070. In summary, most karyo in Fig. 4). RFLP banding patterns showed abundant vari typic differentiation in the species was observed on the two ation between accessions and some variation within acces smallest chromosomes but also on chromosomes 1 and 6. sions (Tables 4 and 5). Most RFLP fragments were found The RFLP probes used in this study were all derived in more than one accession but some accession-specific from cDNA libraries produced from the cultivated peanut, fragments were also observed. For example, a 9.4-kilobase A. hypogaea (Halward et al. 1993). All the probes (kb) EcoRV fragment detected with probe Xuga.crl99 from hybridized well with A. duranensis DNA and produced linkage group lb was unique to accession 30074 (Table 4). Stalker et al. 1207
Table 3. Karyotypic data of six A. duranensis accessions.
Chromosome
Accession 2 3 4 5 6 7 8 9 10
7988 Total length 3.58 3.12 2.88 2.85 2.84 2.63 2.58 2.44 2.07 1.47 Short/long arm ratio 0.84 0.81 0.85 0.87 0.74 0.64 0.81 0.80 0.75 0.83 30061 Total length 3.27 3.11 2.91 2.74 2.63 2.56 2.50 2.37 1.97 1.54 Short/long arm ratio 0.92 0.77 0.84 0.86 0.86 0.88 0.76 0.89 0.55 0.69 30064 Total length 3.20 3.19 2.98 2.86 2.73 2.62 2.50 2.44 2.08 1.67 Short/long arm ratio 0.86 0.77 0.88 0.88 0.84 0.85 0.74 0.91 0.57 0.70 30067 Total length 3.43 3.00 2.83 2.74 2.72 2.71 2.65 2.54 2.50 1.82 Short/long arm ratio 0.46 0.82 0.89 0.87 0.81 0.78 0.88 0.82 0.77 0.69 30070 Total length 3.43 3.43 3.38 3.08 3.06 2.89 2.73 2.51 2.30 2.05 Short/long arm ratio 0.91 0.77 0.88 0.85 0.88 0.83 0.76 0.78 0.56 0.84 10038 Total length 3.18 2.97 2.87 2.62 2.51 2.43 2.27 2.03 1.91 1.51 Short/long arm ratio 0.87 0.87 0.90 0.86 0.80 0.87 0.82 0.84 0.52 0.76
Note: Chromosome length is given in µm.
Most plants exhibited little evidence of heterozygosity genetically very similar even though they are morpholog for different alleles at a RFLP locus, which would indicate ically distinct. that A. duranensis reproduces largely by selfing. However, RFLP variation appears to correspond with geographic two plants were heterozygous for all probes and appeared distribution. In general, accessions collected from adjacent to be F 1 or F 2 hybrids (plant A of accession 30070 and localities (e.g., 30077 and 30078; 30070, 30071, and 30072; plant A of accession 30071 (Fig. 4)). When the banding and 30060 and 30061) clustered on the neighbor-joining pattern for these plants was compared with that of the tree (Fig. 5). The grouping pattern of the tree reflects the other accessions, it was evident that 30070A could have northern and southern geographic groupings, except for been produced by hybridization between accession 30070 the anomalous placing of accession 30069, which was col and either accession 36003 or 30069. Plant 30071A could lected in the north but groups with the southern accessions have been produced by hybridization between accessions (Fig. 5). Also, plants from accessions 30069 and 36003 30071 and 36036. showed no RFLP differences with any of the 28 probes; When the RFLP data were analyzed by neighbor joining therefore, an error in sampling or seed source is suspected. based on genetic distance or by parsimony, relationships To determine whether the collection of additional acces between the accessions could be determined. Accessions sions of A. duranensis would yield additional genetic diver 30084 (A. kempff-mercadoi) and 30092 (A. magna) were sity, the number of different RFLP fragments per accession used as outgroups for the RFLP study because an earlier was determined. When this total number of RFLP fragments study (Halward et al. 1992) had placed these two accessions is plotted against the number of accessions, a relationship close to A. duranensis. Neighbor-joining and parsimony that approaches the linear is obtained (Fig. 6), suggesting analysis gave generally similar results. The tree generated that if more accessions were collected more genetic diversity by neighbor joining (Fig. 5) indicated that accessions for would be obtained. merly classified into A. duranensis and A. spegazzinii do not form discrete groups, i.e., A. spegazzinii accessions are Discussion included within the spectrum of diversity shown by A. duranensis. The small and large leaf forms of accession Arachis duranensis can be differentiated from other mem 10038 could be separated by either of two RFLP probes bers of section Arachis by a combination of traits, including (Table 5). Further, variation was found among the different annual habit, small flowers, obicular to obcordate leaflets, plants of the small leaf mutant. These results would be and small oblong seeds. The two taxa previously called unexpected if the small leaf type had been derived from A. duranensis and A. spegazzinii are morphologically dis a single plant via a simple genetic change. Arachis magna tinct because A. spegazzinii has elongated petioles and separated from A. duranensis as was expected for a unique stems, has a more delicate appearance, and has fewer tri species. However, A. kempjf-mercadoi clustered with the chomes. The species are not, however, genetically unique northern accessions, which means that the two species are based on isozyme data (Stalker et al. 1994) or the RFLP 1208 Genome, Vol. 38, 1995 - I cu ccu - -c ·c;, N t-- m 3.5_,. kb A 6.4_.. kb 8 ABCDE ABCDEABCDE ABCDABCDE 6.4_.. kb c Stalker et al. 1209 Fig. 4. RFLP patterns in accessions of the A. duranensis complex. DNA samples from individual plants of various accessions are denoted by letters such as A-E (S. hamata = Stylosanthes hamata). A sample of cultivated peanut ('Florigiant') is included for comparison. (A) A simple RFLP pattern showing largely uniform patterns within accessions (except for 30064) and variation between accessions. The probe was Xuga.cr015 and the enzyme used to digest the DNA samples was Dral. (B) A more complex RFLP pattern showing variable patterns within and between accessions. Plants 30070A and 3007 lA are hybrids. The probe was Xuga.cr230 and the enzyme used to digest the DNA samples was HindIII. (C) RFLP pattern showing variable patterns within and between accessions. Plant 30070A is clearly a hybrid and 3007 lA has apparently fixed a RFLP allele different from the other plants in this accession. The probe was Xuga.cr030 and the enzyme used to digest the DNA samples was HaeIII. Table 4. RFLP variability between and within accessions and accession-specific RFLP fragments. Accession-specific fragments Fragment length Probe-enzyme Variability Accession (kb) Xuga.crOOB-EcoRV Within and between Xuga.crOl 5-DraI Within and between Xuga.cr016-Dral Within and between Xuga.cr027-HaeIII Within and between 30075 1.8 Xuga.cr029-HindIII Within and between 30064 3.8 Xuga. cr030-Hae III Within and between Xuga.cr032-HindIII None detected Xuga.cs033-EcoRV Within and between Xuga. cs043-Hae III Within and between Xuga.cs044-EcoRV Within and between Xuga. c s047-HindIII Within and between Xuga. c s066-Hae III None detected Xuga. cr073-HindIII None detected Xuga.cr074-HindIII Within and between Xuga.cr090-HindIII Within and between Xuga.crlll-Dral None detected Xuga.cr119-HaeIII Within and between Xuga.crl21-DraI Within and between 30067 1.4 30074 4.3 Xuga.cr 122-EcoRV Within and between 30074 16.0 Xuga.crl 38-HaeIII Within and between 30072 4.0 Xuga.cr 142-HindIII Within and between 30070 3.8 Xuga.crl43-DraI Within and between Xuga.cr 159-Dral Within and between 30074 6.5 analysis described in this paper. The molecular data support accessions of one diploid Arachis species. This study grouping the taxa into a single species, as recently per demonstrates that a large amount of genetic diversity, as formed by Krapovickas and Gregory (1994). measured by morphological, cytological, molecular, and Genetic variation for isozymes (Stalker et al. 1994 ), intercrossing data, exists among the various accessions of RFLP (Kochert et al. 1991; Paik-Ro et al. 1992), and ran A. duranensis. dom amplified polymorphic DNA (Hal ward et al. 1992) Most RFLP variation observed was between accessions, has been reported for several species of Arachis including but variation was also found within accessions. For example, accessions of A. duranensis. The first study on RFLP diver three plants were studied from 30064, and plant A was sity in accessions of wild peanut was that of Kochert et al. different from the other two plants for nearly every probe (1991), which indicated that A. duranensis (PI 262133, studied (for example see Fig. 4). This could have resulted accession 10038) was a likely diploid progenitor of the from mislabeling or a seed mix-up at some time during allotetraploid cultivated peanut. Paik-Ro et al. (1992) com the collection and maintenance process because a matching pared four diploid wild species accessions with A. monticola pollen parent could not be identified in plant nurseries. and concluded that A. duranensis (PI 468201, accession RFLP variation within the two accessions 30070 and 30071 30065) was very closely related to A. hypogaea. The present clearly resulted from hybridization. For 30070A, the data work, however, is the first to study a large number of are consistent with either accession 36003 or 30069; for / 0\ 1210 Genome, Vol. 38, 1995 Fig. 5. A neighbor-joining tree of A. duranensis RFLP data. Data of a single plant from each accession were used to construct the tree. Accession 30092 was assigned as an outgroup. Accessions formerly termed A. spegazzinii are labeled in bold type. Groups of accessions from the northern and southern groups of the distribution (Fig. 1) are noted. ~------30074 '------30084 r----30077 30078 .------30070 Northern Accessions ~-----30071 ~-----30072 ~------7988 '------30075 ~---30060 Southern Accessions 30061 '------30064 36003 '------~ 30069 .------30067 30065 2.5 30092 Fig. 6. Number of unique RFLP fragment "alleles" found Table 5. Variability within accessions. per accession sampled. The 18 accessions surveyed were randomly placed in groups of 6 accessions each and the No. of probes detecting cumulative number of separate RFLP fragments was Plant plant-specific fragments plotted. 10038 s.1. A 120 10038 s.1. E 2 30064 A 2 30064 B 1 30064 c 10 100 30070 A 6 30070 E 30071 A 7 30074 A 1 80 30077 A 30078 c 36036 E 6 12 purity of peanut accessions, it will clearly be necessary to carefully plan planting schedules and the spatial distri Number of accessions sampled bution of plots to prevent cross-pollination. However, nearly all other individual plants studied were homozygous at all loci analyzed, which indicates that these plants repro 3007 lA, the pollen parent was apparently accession 36036. duce largely by selfing in nature. The results demonstrate that cross-pollination does occur Analysis of RFLP data from nuclear sequences has been even between accessions that were separated by 60 m or used to evaluate germplasm of several genera for genetic more and separated by incompatible genotypes. To maintain relatedness (Friar and Kochert 1991; Kesseli et al. 1991; {() \ Stalker et al. 1211 King et al. 1993; Lubbers et al. 1991; Miller and Tanksley data indicate that the molecular characters produce a better 1990; Nienhuis et al. 1993; Velasquez and Gepts 1994). fit to the geographical location where the accessions were However, if multiple enzymes are used with the same collected. Also, the codominant nature of RFLP markers probes, then, in some cases, independence of characters made it possible to clearly detect hybridization between can be compromised. We attempted to avoid this problem accessions and to determine the parents of the resultant by using probes for loci that have been mapped to widely hybrids. We can confidently predict that additional col separated parts of the Arachis genome and by using only lections of A. duranensis accessions will yield more genetic one enzyme for each probe. When analyzing data, we have diversity. All these factors illustrate the utility of molecular used a conservative approach, genetic distance analyzed markers in germplasm evaluation and maintenance. by neighbor joining. Further, we did perform PAUP analysis and found that nearly identical relationships were obtained. Acknowledgments All A. duranensis accessions are closely related based on crossability and meiotic cytogenetic relationships, even This research was supported by a grant from the United though fertility was low in several crosses. It is not entirely States Department of Agriculture, National Research clear, however, how to properly analyze relationships Initiatives Program, Plant Genome Initiative (No. 92-37300- among accessions of a species for which intercrossing 7 516) and funds from the Georgia Agricultural Experiment could have occurred in nature. Given the lack of evidence Station and the Peanut Collaborative Research Support for long-distance dispersal in Arachis, we assumed that Program, United States Agency for International Devel accessions which are located in geographic proximity would opment (grant No. DAN-4048-G-SS-2065-00). Technical be most closely related and that accessions from extremes assistance was provided by Betty Schroeder. of the range should cluster into northern and southern groups. First, crossing data supported separation into north References ern and southern groups because the largest number of univalents and the lowest overall fertility were observed Bianchi-Hall, C.M., Keys, R.D., Stalker, H.T., and Murphy, J.P. in hybrids between geographically distant accessions. The 1993. Diversity of seed storage protein patterns in wild peanut (Arachis, Fabaceae) species. Plant Syst. Evol. 186: RFLP data also fit expected patterns. An exception was 1-15. accession 30069, which is a northern accession grouping Friar, E., and Kochert, G. 1991. Bamboo germplasm screening with the southern accessions. No RFLP differences with with nuclear restriction fragment length polymorphisms. any of the 28 probes differentiated it from 36003. This Theor. Appl. Genet. 82: 697-703. unexpected result is likely the result of a mistake in labeling Gregory, W.C., Gregory, M.P., Krapovickas, A., Smith, B.W., and at some time since these accessions were collected. Further, Yarbrough, J.A. 1973. Structures and genetic resources of 30069 must be the mislabeled accession, since 36003 peanuts. In Peanuts: culture and uses. American Peanut appears to be properly placed in the neighbor-joining tree Research and Education Association, Stillwater, Okla. (Fig. 5). pp. 47-133. A linear relationship was found between the number of Halward, T.M., Stalker, H.T., LaRue, E.A., and Kochert, G. 1991. Genetic variation detectable with molecular markers RFLP alleles detected and the number of observed acces among unadapted germ-plasm resources of cultivated peanut sions. Based on this relationship we expect additional and related wild species. Genome, 34: 1013-1020. RFLP diversity would be found if more collections are Halward, T.M., Stalker, T., LaRue, E., and Kochert, G. 1992. made. Because cultivated peanut accessions represent a Use of single-primer DNA amplifications in genetic studies of very narrow germplasm base (Halward et al. 1991; Kochert peanut (Arachis hypogaea L.). Plant Mo!. Biol. 18: 315-325. et al. 1991), collection and use of wild peanut germplasm Halward, T.M., Stalker, H.T., and Kochert, G. 1993. is desirable. Accessions of A. duranensis should be espe Development of an RFLP linkage map in diploid peanut cially valuable if one of the genomes present in A. hypogaea species. Theor. Appl. Genet. 87: 379-384. was derived from this closely related species (Kochert et al. Hebert, T.T., and Stalker, H.T. 1981. Resistance to peanut stunt 1991; G. Kochert and H.T. Stalker, unpublished data). virus in cultivated and wild Arachis species. Peanut Sci. 8: A longer term justification for additional collection of 45-47. Kesseli, R., Ochoa, 0., and Michelmore, R. 1991. Variation A. duranensis accessions is to preserve genetic diversity at RFLP loci in Lactuca spp. and origin of cultivated lettuce before it is lost because of habitat destruction. The value of (L. sativa). Genome, 34: 430-436. accessions of A. duranensis in breeding programs ultimately King, G., Nienhuis, J., and Hussey, C. 1993. Genetic similarity will depend on utilizing genes of agronomic value that among ecotypes of Arabidopsis thaliana estimated by analysis may not be directly related to RFLP diversity. The key of restriction fragment length polymorphisms. Theor. Appl. question in deciding whether to collect more accessions Genet. 86: 1028-1032. of A. duranensis then becomes whether RFLP diversity Kochert, G., Halward, T., Branch, W.D., and Simpson, C.E. translates into increased variation for desirable traits. 1991. RFLP variability in peanut (Arachis hypogaea) cultivars Although this has not been determined for wild peanut and wild species. Theor. Appl. Genet. 81: 565-570. accessions, other potentially relevant studies have shown that Krapovickas, A., and Gregory, W.C. 1994. Taxonomy of the genus Arachis (Leguminosae). Bonplandia (Corrientes), 8: RFLP diversity is useful in predicting breeding value 1-186. (Melchinger et al. 1990, 1992; Messmer et al. 1991; Smith Lu, J., and Pickersgill, B. 1993. Isozyme variation and species et al. 1991, 1992). relationships in peanut and its wild relatives (Arachis L. - In summary, comparisons of the groupings produced Leguminosae). Theor. Appl. Genet. 85: 550-560. by analysis of RFLP markers with those by morphological Lubbers, E.L., Gill, K.S., Cox, T.S., and Gill, B.S. 1991. 1212 Genome, Vol. 38, 1995 Variation of molecular markers among geographically diverse of Arachis hypogaea. 1. Cytogenetic studies of putative accessions of Triticum tauschii. Genome, 34: 354-361. genome donors. Euphytica, 27: 665-675. Melchinger, A.E., Lee, M., Lamkey, K.R., and Woodman, W.L. Smith, J.S.C., Smith, O.S., Bowen, S.L., Tenborg, R.A., and 1990. Genetic diversity for restriction fragment length poly Wall, S.J. 1991. The description and assessment of distances morphisms - relation to estimated genetic effects in maize between inbred lines of maize. III. A revised scheme for inbreds. Crop Sci. 30: 1033-1040. the testing of distinctiveness between inbred lines utilizing Melchinger, A.E., Boppenmaier, J., Dhillon, B.S., Pollmer, W.G., DNA RFLPs. Maydica, 36: 213-226. and Herrmann, R.G. 1992. Genetic diversity for RFLPs in Smith, J.S.C., Smith, O.S., Wright, S., Wall, S.J., and Walton, M. European maize inbreds: 2. Relation to performance of 1992. Diversity of U.S. hybrid maize germplasm as revealed hybrids within versus between heterotic groups for forage by restriction fragment length polymorphisms. Crop Sci. traits. Theor. Appl. Genet. 84: 672-681. 32: 598-604. Messmer, M.M., Melchinger, A.E., Lee, M., Woodman, W.L., Stalker, H.T. 1990. A morphological appraisal of wild species Lee, E.A., and Lamkey, K.R. 1991. Genetic diversity among in section Arachis of peanuts. Peanut Sci. 17: 117-122. progenitors and elite lines from the Iowa Stiff Stalk Synthetic Stalker, H.T., and Campbell, W.V. 1983. Resistance of wild (BSSS) maize population: comparison of allozyme and RFLP species of peanut to an insect complex. Peanut Sci. 10: data. Theor. Appl. Genet. 83: 97-107. 30-33. Miller, J.C., and Tanksley, S.D. 1990. RFLP analysis of phylo Stalker, H.T., and Dalmacio, R.D. 1981. Chromosomes of genetic relationships and genetic variation in the genus Arachis species, section Arachis. J. Hered. 72: 403-408. Lycopersicon. Theor. Appl. Genet. 80: 437-448. Stalker, H. T., and Moss, J.P. 1987. Speciation, cytogenetics, Nienhuis, J., Slocum, M.K., Devos, D.A., and Muren, R. 1993. and utilization of Arachis species. Adv. Agron. 41: 1-40. Genetic similarity among Brassica oleracea L. genotypes Stalker, H.T., Dhesi, J.S., Parry, D.C., and Hahn, J.H. 1991. as measured by restriction fragment length polymorphisms. Cytological and interfertility relationships of Arachis section J. Am. Soc. Hort. Sci. 118: 298-303. Arachis. Am. J. Bot. 78: 238-246. Paik-Ro, O.G., Smith, R.L., and Knauft, D.A. 1992. Restriction Stalker, H.T., Phillips, T.D., Murphy, J.P., and Jones, T.M. fragment length polymorphism evaluation of 6 peanut species 1994. Variation of isozyme patterns among Arachis species. within the Arachis section. Theor. Appl. Genet. 84: 201-208. Theor. Appl. Genet. 87: 746-755. Seetheram, A., Nayar, K.M.D., Sreekantaradhya, R., and Subrahmanyam, P., Ghanekar, A.M., Nolt, B.L., Reddy, D.V. R., Archar, D.K.T. 1973. Cytological studies on the interspecific and McDonald, D. 1985. Resistance to groundnut diseases in hybrid of Arachis hypogaea X Arachis duranensis. Cytologia, wild Arachis species. In Proceedings of an International 38: 277-280. Workshop on Cytogenetics of Arachis, held 31 October - Simpson, C.E., and Higgins, D.L. 1984. Catalog of Arachis 2 November 1983 at Hyderabad, India. International Crops germplasm collections in South America, 1976-1983. Texas Research Institute for the Semi-Arid Tropics, Patancheru, Agricultural Experiment Station, College Station, and the Andhra Pradesh, India. pp. 49-61. International Board for Plant Genetic Resources, Rome. Swofford, D.L. 1991. PAUP: phylogenetic analysis using par pp . .1-38. simony, vers. 3.1. Illinois Natural History Survey, Champaign, Singh, A.K., Sivaramakrishnan, S., Mengesha, M.H., and Ill. Ramaiah, C.D. 1991. Phylogenetic relations in section Velasquez, V.L.B., and Gepts, P. 1994. RFLP diversity of com Arachis based on seed protein profile. Theor. Appl. Genet. mon bean (Phaseolus vulgaris) in its centres of origin. 82: 593-597. Genome, 37: 256-263. Smartt, J., Gregory, W.C., and Gregory, M.P. 1978. The genomes \,rv ' I