Genetic Characterization of Three Varieties of Astragalus Lentiginosus (Fabaceae)

Genetic characterization of three varieties of Astragalus lentiginosus (Fabaceae)

BRIANJ. KNAus,' RICHC. CRONN,AND AARONLISTON

Knaus, B. J. (Oregon State University, Department of Botany and Plant Pa- thology, 2082 Cordley Hall, Corvallis, OR, 97331-2902, U.S.A.; e-mail: [email protected]), R. C. Cronn (USDA Forest Service Pacific Northwest Research Station, 3200 SW Jefferson Way, Corvallis, OR, 97331, U.S .A.; e-mail: [email protected]) & A. Liston (Oregon State University, Depart- ment of Botany and Plant Pathology, 2082 Cordley Hall, Corvallis, OR, 97331- 2902, U.S.A.; e-mail: [email protected]).Genetic characterization of three varieties of Astragalus lentiginosus (Fabaceae). Brittonia 57: 334-344. 2005.-Astragalus lentiginosus is a polymorphic species that occurs in geologically young habitats and whose varietal circumscription implies active morphological and genetic differentiation. In this preliminary study, we evaluate the po- tential of amplified fragment length polymorphism (AFLP) markers to resolve infraspecific taxa in three varieties of Astragalus lentiginosus. Distance-based principle coordinate and neighbor-joining analyses result in clustering of individuals that is congruent with population origin and varietal circumscription. Anal- ysis of molecular variance of two Oregon varieties demonstrates that varietal categories account for 11% of the total variance; in contrast, geographic proximity does not contribute to the total variance. AFLPs demonstrate an ability to discriminate varieties of A. lentiginosus despite a potentially confounding geographic pattern, and may prove effective at inferring relationships throughout the group. Key words: AFLP, amplified fragment length polymorphism, Astragalus lentiginosus, genetic differentiation, infraspecific taxa.

Astragalus lentiginosus Dougl. ex Hook. populations and varieties may have oc- is a polymorphic species consisting of 40 curred as recently as the late Pleistocene1 varieties (Isely, 1998) distributed through- early Holocene period. This study of A. len- out the Intermountain and Desert Southwest tiginosus represents an investigation of a di- regions of North America. The current va- verse group whose circumscription reflects rietal circumscription reflects Barneby's dramatic morphological variation and im- (1945) view that A. lentiginosus comprises plies genetic differentiation. multiple lines of evolutionary radiation, as Jones (1923) provided the first revision well as occasional reticulation, both of of Astragalus during the twentieth century. which may explain the morphological in- Among his innovations was the reduction termediates between varieties. Many of the of the section Diphysi to a single species, habitats where extant populations occur Astragalus lentiginosus. He characterized (e.g., inland dune systems, desert seeps, this species as the "most variable of all As- mountain ridges) were profoundly different tragali" (Jones, 1923, p. 123). As defined during the Pleistocene (Grayson, 1993). Oc- by Jones (1923), this species consisted of currence on these geologically young hab- 18 varieties, many of which were originally itats suggests that the separation of current described as species. Within his varieties Jones included "forms" (never validly published) that had originally been recognized To whom correspondence should be addressed. as species or varieties.

Brittonia, 57(4), 2005, pp. 334-344. ISSUED: 28 December 2005 O 2005, by The New York Botanical Garden Press, Bronx, NY 10458-5126 U.S.A. 20051 KNAUS ET AL.: ASTRAGALUS LENTIGINOSUS 335

Applying a very different taxonomic ap- 1976), but phylogenetic inference was lim- proach, Rydberg (1929) split Astragalus ited by the large size of the group and ap- into 28 genera (summarized by Barneby, parent homoplasy for this character. Molec- 1964). In doing so, he divided A. lentigi- ular phylogenetic studies of nuclear ribo- nosus between the genera Cystium (33 spe- somal DNA internal transcribed spacers cies, all currently known as A. lentiginosus) (Wojciechowski et al., 1993), chloroplast and Tium (39 species, only three of which DNA (Liston, 1992; Sanderson & Doyle, are currently known as A. lentiginosus). 1993), and combined nuclear and chloro- This revision resulted in most of Jones's plast DNA datasets (Wojciechowski et al., (1923) varieties (and forms) being raised to 1999) have demonstrated the monophyly of species (Barneby, 1945). the New World aneuploid species of As- Barneby's treatments (1945, 1964, 1989) tragalus. However, none of these studies reduced the group back to a single species, has resolved interspecific relationships Astragalus lentiginosus, with ca. 40 varie- within this clade. ties. This classification came from the re- Amplified fragment length polymor- duction of many of Rydberg's (1929) spe- phism (AFLP) analysis (Vos et al., 1995) is cies to varieties, or similarly elevating an anonymous genetic fingerprinting tech- many of Jones's (1923) "forms" to varie- nique developed for plant breeding that has ties. Since Barneby 's 1945 treatment there been adapted to studies of natural popula- has been a reduction in the original number tions (Wolfe & Liston, 1998; Mueller & of names accepted as varieties in part due Wolfenbarger, 1999). The AFLP method to additional collecting that has blurred has been used to infer interspecific (Abdel- some intervarietal distinctions (Barneby, fattah et al., 2002; Beardsley et al., 2003) 1964, 1989). This trend has been partially and intraspecific (Brouat et al., 2004; Juan offset by the description of new varieties et al., 2004; Travis et al., 1996) relation- since 1964 (e-g., Barneby, 1977; Welsh, ships as well as population level dynamics 1981). While the currently recognized num- (He et al., 2004). This method was applied ber of varieties is 40 (Isely, 1998) there is by Travis et al. (1996) to evaluate popula- still an active debate as to what constitutes tion-level differentiation of Astragalus A. lentiginosus (Alexander, 2005) as well as cremnophylax var. cremnophylax occurring to the validity of the varieties. at the Grand Canyon, U.S.A. Results from The varieties of Astragalus lentiginosus that study showed strong differentiation of are morphologically distinct when observed north and south rim populations, suggesting at distant stations, yet when geographically that AFLPs are sufficiently sensitive to dis- proximal their distinctiveness may become cern genetic differences in recently di- obscured to the point where they are indis- verged lineages, such as A. lentiginosus. tinguishable (Barneby, 1964, p. 922). Most In this preliminary study we utilize of the group is characterized by an inflated AFLP analysis to test Barneby's (1945, bilocular pod with a false septum that in- 1964, 1989) taxonomic treatment of three trudes from the abaxial surface and is in- Astragalus lentiginosus varieties. We focus complete in the beak. Barneby (1945, 1964, on the question of whether the sampled 1989) created major divisions within the populations of A. lentiginosus varieties dis- group based on flower size (keel greater or play genetic relationships that can be attri- less than 8.5 mm), raceme length (greater buted to geographic proximity, or whether or less than 4 cm), and flower color (purple the varieties exhibit genetic cohesiveness. or white). Attempts to evaluate taxonomic interpre- Methods tations within Astragalus utilizing molecular methods have been complicated by low levels of divergence among the New World Plants included in this study were select- species. Chromosome number variation was ed to determine whether genetic differenti- identified as potentially useful at the infra- ation could be discerned across the latitu- generic level (Barneby, 1964; Spellenberg, dinal extent of Astragalus lentiginosus. 336 BRITTONIA [VOL. 57

TABLEI COLLECTIONSOF Astragalus lentiginosus

Altitude Population N Habitat Latitude Longitude (m) State County lentiginosus-1 8 Juniper woodland 42.76 - 118.74 1915 OR Harney lentiginosus-2 8 Ponderosa pine forest 42.27 - 121.30 1490 OR Klarnath salinus-1 8 Sage shrubland 42.43 -1 18.08 1325 OR Harney salinus-2 7 Sage shrubland 43.32 -121.06 1318 OR Lake variabilis-1 8 Disturbed desert 34.57 -1 17.41 873 CA San Bemardino variabilis-2 7 Dune system 36.65 -116.57 744 NV Nye

Leaves were collected during the summer the Columbia Basin to northern California of 2004 (Table I), tissue was desiccated in and the northwestern Great Basin. Barneby silica gel and stored at 4OC until DNA ex- (1945, 1964, 1989) placed this taxon among traction. Collections were made from six lo- the A. lentiginosus possessing short flowers cations (Fig. I), with two locations per va- (keel length < 8.5 mm), short racemes (axis riety. Geographic distances between sample < 4 cm long in fruit), and white flowers. locations were calculated with ArcView Astragalus lentiginosus var. lentiginosus is GIs 3.2 (Environmental Systems Research distinctive among the varieties of A. lenti- Institute, Inc.). ginosus in having moderately inflated pods Three morphologically distinct varieties with a coarse texture (thick walls), de- were included in this study. Astragalus Zen- scribed as stiffly papery, leathery, or woody tiginosus var. lentiginosus is a slender, de- (as opposed to thinly papery or membra- cumbent to prostrate perennial occurring in naceous). Ponderosa pine or juniper woodland from Astragalus lentiginosus var. salinus (Howell) Barneby is a short-lived perennial consisting of diffuse and ascending stems occurring in sagebrush and alkaline flats from eastern Oregon to southern Idaho and throughout the northwestern Great Basin. This variety possesses short flowers (keel length < 8.5 mm long), short racemes (axis < 4 cm long in fruit), and white flowers. Astragalus lentiginosus var. salinus is geographically proximal to A. 1. var. lentiginosus and can appear identical to the latter except for the thinly papery texture and in- flation of the pods. Astragalus lentiginosus var. variabilis Barneby is a monocarpic to short-lived perennial plant consisting of erect to ascending diffuse stems occurring on sandy flats, washes, desert playas, and sometimes on inland dunes throughout the Mojave Desert of southern California and southern Neva- da. This variety possesses large flowers (keel length > 8.5 mm), long racemes (axis FIG. 1. Map of western North America showing > 4 cm long in fruit) and purple flowers. ranges for three varieties of Astragalus lentiginosus This variety was chosen because it is geo- and collections used in this study. Horizontal lines indicate the range of A. 1. var. salinus, vertical lines in- graphically disjunct from the above varie- dicate the range of A. 1. var. lentiginosus, and dots ties. In addition to these traits, A. 1. var. var- indicate the range of A. 1. var. variabilis. iabilis differs from the previous two varie- 20051 KNAUS ET AL.: ASTRAGALUS LENTIGINOSUS 337 ties in its robust and sometimes erect habit EcoRI+ACAIMseI+GAT, in a 10 p1 final and its frequently cinereous to canescent volume [lX PCR buffer (Fisher Scientific), vestiture. 0.2 mM dNTPs, 1.5 mM MgCl,, 3.75 pmol fluorescent1y labeled EcoRI primer, 5.0 pmol DNA EXTRACTION unlabeled MseI primer, and 0.625 U Taq DNA extractions were based on a CTAB polymerase]. selective primer pairs were ch& procedure modified from Chen and Ronald sen by screening ten primer pairs and select- (1999). Exceptions were that leaflet tissue ing for maximum number of peaks (results (ca. 50 mg) was homogenized in a bead not shown). Cycling conditions were 2 min- mill (Fastprep 120; Qbiogene) using 0.5 utes at 94°C followed by 9 cycles of 30 sec- cm3 of 2.5 mm zirconia/silica beads (No. onds at 94OC, 30 seconds at 65°C (-1°C per 11079125~-B; Biospec Products) and 800 cycle), and 2 minutes at 72°C. An additional pl2X CTAB buffer [2% wlv CTAB, 1.4 M 25 cycles were performed at 30 seconds at NaCl, 0.1 M Tris-HC1 (pH 8.0), 20 mM 94"C, 30 seconds at 56"C, and 2 minutes at EDTA, 2% wlv PVPP]. Samples were ho- 72"C, followed by a final extension at 60°C mogenized for 2 cycles of 20 seconds at a for 45 minutes. setting of 5. DNA was resuspended in 30 Fragments were resolved on an Applied p1 of buffer [lo mM Tris-HC1, 0.1 mM Biosystems Inc. 3 100 capillary fragment EDTA (pH KO)] and then evaluated for analyzer (36 cm capillary; POP4 polymer) quality by gel electrophoresis and quanti- using a 1:2 dilution of " + 3" product. Trace fied by fluorometry (VersaFluor; BioRad files were scored using ABI GeneMapper Laboratories). v3.0 following Rinehart (2004). To de- crease the probability of including homo- AFLP SCREENING plastic AFLP bands we included only bands between 200-490 bp with peak heights Amplified fragment length polymor- above 200 relative fluorescent units (Koop- phism (AFLP) analysis was performed uti- man & Gort, 2004). lizing a modification of the Vos et al. (1995) method. Template DNA (-200 ng) was digested using 12 U of EcoRI and 8 U of MseI (New England Biolabs) for 2 hours at 37°C in a 20 p1 volume. Adapters were AFLP products were scored as the pres- ligated to restricted DNAs by adding 20 p1 ence (1) or absence (0) of bands. Population ligation mix [I X final concentration ligase and variety level band frequencies were cal- buffer, 8 U T4 DNA ligase (New England culated using GenAlEx v5.1 (Peakall & Biolabs), and 5 pmol EcoRI and 50 pmol Smouse, 2001). Two distances were calcu- MseI adapters] and incubating for 3 hours lated from these data. Pairwise squared Eu- at 16°C. Enzymes were heat denatured at clidian distances (E2, = xk[Xki-Xkj]2,where 65°C for 20 minutes after each treatment. i and j represent a pair of objects, and k Pre-selective (" + 1") amplification was denotes cases where y, and ykj are both pre- performed in 10 p1 [1X PCR buffer (Fisher sent) were calculated using GenAlEx and Scientific), 0.2 mM dNTPs, 1.5 mM MgCl,, used as input for analysis of molecular var- 8 pmol each primer, 400 nglpl BSA, 0.5 U iance (AMOVA; Excoffier et al., 1992). Taq polymerase] utilizing 1 p1 of diluted (1 : The Lynch shared band similarity index (S, 5) ligation product. PCR cycling conditions = 2NXy/(NX+ N,) where N, is the number consisted of 2 minutes at 75°C followed by of bands in sample x, N, is the number of 19 cycles of 30 seconds at 94"C, 30 seconds bands in sample y, and N,, is the number at 56"C, and 2 minutes at 72°C; a final ex- of bands shared between both samples tension of 30 minutes was performed at (Lynch, 1990; Lamboy, 1994), as calculated 60°C. using NTSYSpc version 2.1 (Rohlf, 2000) Selective (" + 3" ) amplifications were per- was used for neighbor-joining and ordina- formed using two primer combinations, tion analyses. Because this index is equiv- [FAM]EcoRI+ACA/MseI+ GAC and [HEX] alent to one minus the Dice (1945) coinci- 338 BRITTONIA ~VOL.57

TABLE11 AFLP BAND FREQUENCIES

Population n #loci #bands PLP #private

- - (n), number of samples; (#loci), total number of loci; (#bands), number of bands scored as present in each population; (PLP), percent loci in each population; (#private), number of bands occurring in a population that are not present in other populations. dence index, we refer to this measure as and the third accounted for 10.3% of the "Dice distance. " variation (49.6% of the total variation). Principle coordinates analysis (PCoA) of Populations of Astragalus lentiginosus var. Dice distances was used to evaluate inter- variabilis formed a single discrete cluster individual distances in low dimension space while most individuals of A. 1. var. lentigi- to determine whether a priori population nosus also formed a cluster. In contrast, the and varietal designations matched patterns two populations of A. 1. var. salinus clus- of genetic variation. Dice distances were tered separately from each other. Individu- evaluated using the neighbor-joining algo- als lentiginosus-1-5, lentiginosus-1-6, len- rithm (NJ; Saitou & Nei, 1987). Both PCoA tiginosus-1-9, lentiginosus-2-1, and sali- and NJ were implemented in NTSYS-pc nus-1-1 1 (A, B, C, D, E, respectively; Fig. (Rohlf, 2000). Partitioning of genetic vari- 2) fall in a region of intermediacy between ation between populations nested within va- populations salinus-1 and lentiginosus-1. rieties was evaluated by AMOVA using the However, individual salinus-1-1 1 (E) is approach of Excoffier et al. (1992), as im- separated from the others on the third axis plemented with GenAlEx v5.1 (Peakall & by its strongly negative eigenvector Smouse, 2001). For variance partitioning, (-0.0446) as opposed to the positive values we looked at two groupings: at populations on the third axis of individuals lentigino- within varieties or regions (Q,,), and sus-1-5, lentiginosus-1-6, and lentigino- among varieties or regions (Q,,). Signifi- sus-1-9 (0.0166, 0.0176 and 0.0212 respec- cance was measured with 999 permutations tively). Individual lentiginosus-2-1 (D) is of individuals among hierarchical levels notable in its intermediate third axis eigen- and recalculating null distributions of the vector (-0.0230) and its position as closer test statistic (Excoffier et al., 1992). to the origin than the other members of its population. Close inspection of AFLP trace Results files show that these were high quality amplifications, so intermediacy is not the result Two primer pairs resulted in 184 loci of aberrant amplification. scored between 200 and 490 bp with an av- Cluster analysis utilizing the neighbor- erage of 36.5 fragments per individual. Ten joining method (Fig. 3) resulted in a den- loci were monomorphic and 174 were poly- drogram that reflects ordination results with morphic. Seventy three of the bands (40%) each variety forming a distinct cluster. In- had a frequency of less than 0.05. Number dividual lentiginosus-2-1 (D; Fig. 2) was of loci and private alleles per population are the sole exception in that it neither clustered summarized in Table 11. with its source population or variety. Indi- Ordination by principle coordinates anal- viduals showing intermediacy on PCoA ysis (PCoA; Fig. 2) shows varieties to oc- (Fig. 2; lentiginosus-1-5, lentiginosus-1-6, cupy distinct regions of coordinate space. lentiginosus-1-9 and salinus-1-1 1; A, B, C The first axis accounted for 22.8% of the and D respectively) cluster with their re- variation, the second accounted for 16.5%, spective populations (Fig. 3). 20051 KNAUS ET AL.: ASTRAGALUS LENTIGINOSUS 339

FIG.2. Principle coordinates analysis (PCoA) based on a matrix of Dice genetic distances. A = lentiginosus-l- 5. B = lentiginosus-1-6. C = lentiginosus-1-9. D = lentiginosus-2-1. E = salinu~l-11.F = salinus-1-12.

Dice genetic distances show a general sampled sites. Analyses were conducted pattern where within population genetic dis- three ways to evaluate diversity at different tance was lowest (Table 111; mean = .0823 scales. Analysis #1 included all varieties + 0.0202, n = 154), within variety genetic and populations to address the question of distance was slightly higher (mean = whether there was structure to the sampled 0.1055 2 0.0201, n = 176), and between populations and varieties of Astragalus len- variety genetic distances were greatest tiginosus throughout its range. Variance de- (mean = 0.1216 L 0.0176, n = 705). composition showed that most of the vari- Among-population genetic distances for As- ation in A. lentiginosus was contained with- tragalus lentiginosus var. salinus (0.1 19) is in populations (67.3%, p < 0.001; Table comparable to the average among popula- IV), due in large part to the abundance of tion divergence. bands restricted to one or a few individuals Analysis of molecular variance revealed within a population. The remaining 32.7% significant structure within the sampled va- of the total variance was apportioned rieties and populations across the range of among populations, with 13.4% of the var- 340 BRITTONIA [VOL. 57

I lent igjno~k-2-3 I lent iginoas-2-6

FIG. 3. Neighbor-Joining dendrogram built from a matrix of Dice genetic distances. Individuals labeled A, B, C, D, E, and F are the same as in Figure 2. 20051 KNAUS ET AL.: ASTRAGALUS LENTIGINOSUS 341

lentiginosusl lentiginosus2 salinus-1 salinus2 variabilis-1 variabilis2

Above the diagonal are geographic distances (km). Below the diagonal are Dice genetic distances. Along the diagonal are intra-population Dice genetic distances. iation distributed among varieties (p < lations lentiginosus-1 and salinus-1 were 0.001) and 19.3% distributed among pop- grouped into one region while populations ulations within varieties @ < 0.001). lentiginosus-2 and salinus-2 were grouped Because Astragalus lentiginosus var. var- into the second region. This grouping ac- iabilis is several hundred kilometers dis- counted for none of the variance at the re- junct (Table 111; Fig. I), it may have con- gional level of hierarchy (@,, = 0, p > tributed disproportionately to the varietal 0.999; Table IV). These results indicate that component of the variance. Analysis #2 was the variation revealed by AFLP shows a performed to determine whether geograph- strong within-variety signal, and that geo- ically proximal populations retained the va- graphic proximity does not contribute to the rietal signal exhibited in the full dataset. measured variance. Removing A. 1. var. variabilis populations from the analysis resulted in little change in Discussion the among-variety variance component (a,, = 10.9%, p < 0.001; Table IV). Barneby7s (1 945, 1964, 1989) circum- The third analysis was conducted to test scription of Astragalus lentiginosus repre- whether geographic proximity accounted sents an exceptionally detailed account of a for more of the variance than varietal cir- group of organisms that appear too different cumscription. The Oregon populations were to be considered the same taxon but also grouped based on geographic proximity (re- exhibit intergradation that precludes one gion) instead of taxonomic variety. Popu- from delimiting species; hence he chose the

TABLEIV ANALYSISOF MOLECULAR VARIANCE (AMOVA)

% total Source df SS MS Est. var. var. Stat Value p-value* Analysis 1: all populations Among Varieties 2 Among Pops.1Varieties 3 Indiv./Within Pops. 40 Analysis 2: OR populations Among Varieties 1 Among Pops.1Varieties 2 Indiv./Within Pops. 27 Analysis 3: OR populations1 geographic grouping Among Regions 1 Among Pops./Regions 2 Indiv./Within Pops. 27 * p-values computed based on a simulation of 999 permutations. BRITTONIA taxonomic rank of variety to describe these INTER-VARIETALINTERMEDIACY taxonomic units. This results in a species WITHIN A. LENTIGINOSUS that includes an unusually high number of varieties and has been described as un- Four individuals of Astragalus lentigi- wieldy and vastly multiracial (Isely, 1998). nosus var. lentiginosus and one of A. 1. var. To date there has been no test of this cir- salinus were positioned near the origin in cumscription utilizing molecular methods, the PCoA analysis (Fig. 2). The difference perhaps due to the complex nature of this between varieties is resolved on the third diverse taxon. axis where they are clustered according to variety (except individual lentiginosus-2-1) and in the neighbor-joining dendrogram, suggesting differentiation among the varieties. The position of these individuals, par- ticularly lentiginosus2-1, near the origin, Taxonomic circumscription within this may indicate limited intergradation among group was explored with ordination by the varieties. Due to the inclusion of only principle coordinates analysis (PCoA), three varieties in this analysis, inferences which resulted in each variety occupying about the cause of this finding cannot yet distinct coordinate space. Cluster analysis be made. using the neighbor-joining method largely concurred with ordination results. A series COMPARISONWITH A. CREMNOPHYLAX of AMOVAs were then employed to pro- VAR. CREMNOPHYLAX vide statistical significance to varietal circumscription. These data suggest that inter- Travis et al. (1996) used AFLPs to dis- breeding occurs preferentially within the tinguish populations of Astragalus crern- sampled varieties as opposed to between in- nophylax var. crernnophylax on the north tervarietal populations in close geographic and south rim of the Grand Canyon. AMO- proximity. VA results partitioned 63% of the variance The comparison of Dice genetic distance amongst the north and south rim popula- with geographic distance (Table 111) shows tions, as contrasted to 32% variance among interpopulation genetic distance within As- populations of A. lentiginosus reported tragalus lentiginosus var. variabilis and A. here. Principle coordinate analysis showed 1. var. lentiginosus to be relatively small. In discrete clusters in the A. crernnophylax contrast, the two populations of A. 1. var data as compared to the clusters that in- salinus, supported in both PCoA and clus- cluded intermediate individuals reported in tering, show considerably greater diver- the current study. This suggests that popu- gence than the other two varieties. lations of A. crernnophylax var. crernnophy- One of these populations, salinus-1, was lax are more structured and divergent than collected near Whitehorse Ranch, a site the varieties included in this current study mentioned by Barneby (1964, p. 922) as despite our sampling of A. lentiginosus over possessing individuals that defy classifica- a much larger geographic range (1000 km). Recent allozyme data (Allphin et al., in tion (due to intermediate morphologies with press) suggest the north and south rim pop- Astragalus lentiginosus var. Jloribundus). ulations warrant specific rank. This inter- Genetic divergence between the sampled pretation seems to be in accordance with populations of A. 1. var. salinus may be due the amount of AFLP divergence found be- to factors such as introgression or hybrid- tween varieties in this study. ization with other varieties (e.g., A. 1. var. fioribundus; not included in this study), or it may simply represent a greater molecular CoNCLUoING REMARKS diversity in this variety. The inclusion of In this pilot study we have demonstrated more varieties (such as A. 1. var. Jloribun- genetic differentiation between varieties of dus) in future studies may provide addition- Astragalus lentiginosus. If populations of A. a1 evidence for this observation. lentiginosus represented a simple pattern of 20051 KNAUS ET AL. : ASTRAGALUS LENTIGINOSUS 343 isolation by distance one would expect a Barneby, R. C. 1945. Pugillus Astragalorum IV: the cline of molecular diversity across the tax- section Diplocystium. Leaflets of Western Botany 4: 65-147. on's range. Populations collected at distant . 1964. Atlas of North American Astragalus. sites would appear distinct, yet intermediate Memoirs of The New York Botanical Garden 13: samples would fall along a continuum, both 91 1-958. morphologically and in terms of the molec- . 1977. Dragma hippomanicum 111: novitates ular data. However, in the present study va- Californicae. Brittonia 29: 376381. . 1989. Astragalus lentiginosus. Pages 157- rietal distinction has been demonstrated in 164. In: A. Cronquist, A. H. Holmgren, N. H. both disjunct and geographically proximal Holmgren, J. L. Reveal & I? K. Holmgren, editors. varieties suggesting that simple isolation by Intermountain Flora 3B. The New York Botanical distance may not be responsible for all of Garden, New York. the differentiation among varieties. While Beardsley, P. M., A. Yen & R. G. Olmstead. 2003. AFLP phylogeny of Mimulus section Erythranthe inferences about mechanisms are premature and evolution of hummingbird pollination. Evolu- at this time, these data do not contradict tion 57: 1397-1410. Barneby's circumscription of these three Brouat, C., D. McKey & E. J. P. Douzery. 2004. varieties. Differentiation in a geographical mosaic of plants coevolving with ants: phylogeny of the Leonardoxa This study accompanied with that of africana complex (Fabaceae: Caesalpinioideae) us- Travis et al. (1996) show AFLPs to be a ing amplified fragment length polymorphism mark- promising tool in the study of Astragalus ers. Molecular Ecology 13: 1 157-1 171. species. Here we have demonstrated the Chen, D. H. & P. C. Ronald. 1999. A rapid DNA ability of AFLPs to discriminate infraspe- minipreparation method suitable for AFLP and other PCR applications. Plant Molecular Biology Re- cific relationships within A. lentiginosus de- porter 17: 53-57. spite a potentially confounding geographic Dice, L. R. 1945. Measures of the amount of ecolog- pattern. This suggests their use may prove ical association between species. Ecology 26: 297- effective at inferring relationships through- 302. out the group. Excoffier, L., P. E. Smouse & J. M. Quattro. 1992. Analysis of molecular variance inferred from met- ric distances among DNA haplotypes: applications Acknowledgments to human mitochondria1 DNA restriction data. Ge- netics 13l : 479-491. B. J. K. would like to thank the Hardman Grayson, D. K. 1993. The desert's past. Smithsonian Foundation, the Native Plant Society of Institution Press, Washington, D.C. Oregon, and the Nevada Native Plant So- He, T., S. L. Krauss, B. B. Lamont, B. P. Miller & ciety for grants that helped fund travel and N. L. Enright. 2004. Long-distance seed dispersal in a metapopulation of Banksia hookeriana inferred lab costs. Additional support for this project from a population allocation analysis of amplified was provided by the USDA Forest Service fragment length polymorphism data. Molecular Pacific Northwest Research Station. We Ecology 13: 1099-1109. would also like to thank Timothy J. Motley Isely, D. 1998. Native and naturalized Leguminosae and an anonymous reviewer for comments (Fabaceae) of the United States. Monte L. Bean Life Science Museum, Provo, Utah. that improved the text. Jones, M. E. 1923. Revision of North American As- tragalus. Published by the author. Salt Lake City, Literature Cited Utah. Juan, A., M. B. Crespo, R. S. Cowan, C. Lexer & Abdelfattah, B., H. E. Shazly, H. E. Rabey & L. E. M. F. Fay. 2004. Patterns of variability and gene Watson. 2002. Systematic relationships in Lathy- flow in Medicago citrina, an endangered endemic rus sect. Lathyrus (Fabaceae) based on amplified of islands in the western Mediterranean, as revealed fragment length polymorphism (AFLP) data. Ca- by amplified fragment length polymorphism nadian Journal of Botany 80: 962-969. (AFLP). Molecular Ecology 13: 2679-2690. Alexander, J. A. 2005. The taxonomic status of As- Koopman, W. J. M. & G. Gort. 2004. Significance tragalus mokiacensis (Fabaceae), a species endem- tests and weighted values for AFLP similarities, ic to southwestern United States. Brittonia 57: 320- based on Arabidopsis in silico AFLP fragment 333. length distributions. Genetics 167: 1915-1928. Allphin, L., N. Brian & T. Matheson. In press. Re- Lamboy, W. F. 1994. Computing genetic similarity productive success and genetic divergence among coefficients from RAPD data: the effects of PCR varieties of the rare and endangered Astragalus artifacts. PCR Methods and Applications 4: 3 1-37. cremnophylax (Fabaceae) from Arizona, USA. Liston, A. 1992. Variation in the chloroplast genes Conservation Genetics. rpoCl and rpoC2 of the genus Astragalus (Faba- 344 BRITTONIA [VOL. 57

ceae): evidence from restriction site mapping of a cytotaxonomic significance for North American As- PCR-amplified fragment. American Journal of Bot- tragalus (Fabaceae). Taxon 25: 463-476. any 79: 953-961. Travis, S. E., J. Maschinski & P. Keim. 1996. An Lynch, M. 1990. The similarity index and DNA fin- analysis of genetic variation in Astragalus crem- gerprinting. Molecular Biology and Evolution 7: nophylax var. cremnophylax, a critically-endan- 478-484. gered plant, using AFLP markers. Molecular Ecol- Mueller, U. G. & L. L. Wolfenbarger. 1999. AFLP ogy 5: 735-745. genotyping and fingerprinting. Trends in Ecology Vos, P., R. Hogers, M. Bleeker, M. Reijans, T. van and Evolution 14: 389-394. de Lee, M. Hornes, A. Frijters, J. Pot, J. Pele- Peakall, R. & P. E. Smouse. 2001. GenAlEx V5: Ge- man, M. Kuiper & M. Zabeau. 1995. AFLP: a netic analysis in Excel. Australian National Uni- new technique for DNA fingerprinting. Nucleic Ac- versity, Canberra, Australia. http://www.anu.edu. ids Research 23: 4407-441 4. au/BoZo/GenAlEx/. Welsh, S. L. 198 1. New taxa of western plants-in trib- Rinehart, T. A. 2004. AFLP analysis using Gene- ute. Brittonia 33: 294-303. Mapper@software and an Excel@' macro that aligns Wojciechowski, M. F., M. J. Sanderson, B. G. Bald- and converts output to binary. Biotechniques 37: win & M. J. Donoghue. 1993. Monophyly of an- 186-188. euploid Astragalus (Fabaceae): evidence from nu- Rohlf, F. J. 2000. NTSYSpc v2.1. Exeter Software. clear ribosomal DNA internal transcribed spacer sequences. American Journal of Botany 80: 71 1- Setauket, New York. -nn Rydberg, P. A. 1929. Astragalanae. North American ILL. --& J.-M. Hu. 1999. Evidence on the Flora 24: 251-462. , monophyly of Astragalus (Fabaceae) and its major Saito, N. & M. Nei. 1987. The neighbor-joining meth- subgroups based on nuclear ribosomal DNA ITS od: a new method for reconstructing phylogenetic and chloroplast DNA trnL intron data. Systematic trees. Molecular Biology and Evolution. 4: 406- A- - Botany 24: 409-437. 423. Wolfe, A. D. & A. Liston. 1998. Contributions of Sanderson, J. & J* J. Doyle. 1993. Ph~logenetic PCR-based methods to plant systematics and evo- relationships in North American Astragalus (Fa- lutionary biology. Pages 43-86. In: D. E. Soltis, F? baceae) based on chloroplast DNA restriction site S. Soltis and J. J. Doyle, editors. Molecular system- variation. Systematic Botany 18: 395-408. atics of plants 11: DNA sequencing. Kluwer Aca- Spellenberg, R. 1976. Chromosome numbers and their demic, Boston, Massachusetts.