Am. J. Bot. Tips for Better Browsing
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bain, J. F.
Right arrow Articles by Golden, J. L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Bain, J. F.
Right arrow Articles by Golden, J. L.
Agricola
Right arrow Articles by Bain, J. F.
Right arrow Articles by Golden, J. L.
(American Journal of Botany. 2003;90:1087-1094.)
© 2003 Botanical Society of America, Inc.

Systematics

Phylogeographic relationships within Packera sanguisorboides (Asteraceae), a narrow endemic species that straddles a major biogeographic boundary1

John F. Bain2 and Joanne L. Golden

Department of Biological Sciences, University of Lethbridge, Lethbridge, Alberta, Canada T1K 3M4

Received for publication October 15, 2002. Accepted for publication February 13, 2003.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Packera sanguisorboides is endemic to the Sangre de Cristo and Sacramento Mountains of New Mexico, USA. As such, its distribution spans the boundary between two major floristic regions: the southern Rocky Mountain region and the Madrean region. Chloroplast DNA haplotype polymorphism patterns in populations from both regions show that most of the molecular variance exists among populations rather than between mountain ranges and that hybridization with at least one other Packera species containing distinct cpDNA haplotypes has contributed to the cpDNA diversity within P. sanguisorboides. Based on morphological and cpDNA data, the closest relatives to P. sanguisorboides are thought to be mesic species similar to P. glabella and P. sanguisorbae.

Key Words: Asteraceae • biogeography • cpDNA • haplotypes • Mexico • Packera • phylogeography • southwestern USA


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
The wide-ranging North American genus Packera (= aureoid Senecio complex) has often been cited as a taxon whose evolution has been significantly shaped by geohistorical events (e.g., Barkley, 1962 , 1988 ; Kowal, 1975 ; Bain, 1988 ; Bain and Jansen, 1996 ; Golden and Bain, 2000 ). Both morphological and molecular data have been used to support the notion that events such as Pleistocene glaciation and mountain building have played a major role in shaping species relationships within the genus by promoting opportunities for hybridization and introgression (Barkley, 1988 ). From an evolutionary biologist's standpoint, the most attractive aspect of the variation patterns in Packera is the wide taxonomic and geographical distribution of this hybridization and introgression, because it allows patterns of hybridization in different regions of North America to be compared within the same genus, allowing a better assessment of the effects of different climates and historical events on the species in question. Furthermore, because of its wide distribution, an understanding of overall gene flow patterns among species of Packera from different regions is likely to provide insights into the evolution of North American flora similar to those provided for European flora by studies of wide-ranging tree species in that region (e.g., Desmesure et al., 1996 ; Dumolin-Lepegue et al., 1997 ).

Bain and Jansen (1996) identified a number of areas in North America where high levels of intrapopulational chloroplast DNA (cpDNA) diversity indicated regions of possible hybridization and introgression. These areas included two regions at the boundary of continental Pleistocene glaciation (Alaska–Yukon, southwestern Alberta, Canada) and one from southwestern United States (Arizona–New Mexico).

In a follow-up study, Golden and Bain (2000) provided a detailed phylogeographic analysis of haplotype diversity in four Packera species from southwestern Alberta. They concluded that haplotypes from the two major cpDNA clades within Packera (coastal vs. widespread) were present in all species and in most populations from the region. Therefore, past hybridization and introgression had been important in shaping the diversity within the species and populations. They also showed that the pattern of haplotype diversity in alpine species differed from that found in the montane and prairie species of lower elevations, the former having greater levels of population differentiation. Tests for isolation-by-distance (IBD) effects on two species (one alpine, one montane) revealed a negative correlation between distance and number of migrants in both species as expected for IBD, but also revealed a number of cases of apparently high amounts of gene flow among geographically distant populations. In spite of the IBD effects, no strong geographical pattern to the haplotype diversity was evident in alpine P. contermina (Greenm.) Bain, suggesting that other factors were also playing a role in the evolution of the species. Because of the complexity of the haplotype variation patterns, these factors could not be identified and clearly described, although the authors did suggest that past hybridization events may provide a partial explanation for the unusual patterns.

In the Arizona–New Mexico region, the most southern of the three areas of haplotype diversity identified by Bain and Jansen (1996) , the biogeographic factors affecting the evolution of the flora, including Packera species, are different from those in southwestern Alberta in a number of important ways. The flora has existed in the region for a significantly longer period, and the diverse geological landscape has allowed diversification of the flora into distinct assemblages with distinct histories (Axelrod and Raven, 1985 ). Among these distinct assemblages are the floras of the high mountain regions of New Mexico, which have been segregated as constituents of two different floristic provinces (sensu Peet, 1988 ). The southern Rocky Mountain province includes the mountain floras between southern Montana and the Sangre de Cristo Mountains of northern New Mexico, while the Madrean province is centered in the Sierra Madre mountain ranges of Mexico and extends north into isolated ranges in the southern United States, such as the Chiricahua (Arizona), Davis (Texas), and Sacramento Mountains (New Mexico). Thus, the desert region of New Mexico that separates the Sangre de Cristo and Sacramento mountain ranges acts as the approximate boundary between the southern Rocky Mountain and Madrean floristic provinces. High elevation mountain floras on these two ranges have remained separated from each other even during the Pleistocene. (Van Devender et al., 1983 , 1987 ).

Packera sanguisorboides (Rydb.) W.A. Weber & Á. Löve is endemic to the Sangre de Cristo and Sacramento Mountains of New Mexico, so its very restricted distribution nevertheless spans two floristic provinces. It is a common element of mesic, subalpine, high-elevation forests in both mountain ranges but is especially abundant in the Sierra Blanca region of the Sacramento Mountains (Dye and Moir, 1977 ). Based on morphological and ecological similiarity, its closest taxonomic affinities are with Madrean Packera species (Greenman, 1915 ; Barkley, 1968 ), probably lying with mesophytic taxa like Packera sanguisorbae (DC.) C. Jeffrey and P. coahuilensis (Greenm.) C. Jeffrey of the Sierra Madre in Mexico and with P. glabella (Poir.) C. Jeffrey and P. tampicana (DC.) C. Jeffrey, two widespread annual species of lower elevations along the Gulf coastal plain of Mexico, adjacent Texas, and regions to the north (Barkley, 1978 ; Freeman and Barkley, 1995 ). However, molecular data from the internal transcribed spacer (ITS) region of nuclear rDNA suggest a relationship to more xeric taxa such as P. multilobata (Torr. & A. Gray) W.A. Weber & Á. Löve and P. neomexicana (A. Gray) W.A. Weber & Á. Löve (Bain and Jansen, 1995 ; Bain and Golden, 2000 ). Thus, the possibility exists that ancient hybridization has contributed to the evolution of P. sanguisorboides just as it has for Packera species in southwestern Alberta.

Further evidence that P. sanguisorboides may include greater genetic diversity than is suggested by its uniform morphology exists in the report of cpDNA polymorphisms within the one population examined by Bain and Jansen (1996) . Although these authors did not identify individual haplotypes within the species, the cpDNA population profile includes some unusual restriction site polymorphisms, indicating that at least some of the haplotypes present in P. sanguisorboides are absent from most Packera populations sampled across western North America. The haplotype profile most similar to P. sanguisorboides in the Bain and Jansen (1996) study is found in the southern California endemic, P. bernardina (Greene) W.A. Weber & Á. Löve (haplotype H18 as described in Golden and Bain, 2000 ).

In the present study, a detailed analysis of cpDNA variation in P. sanguisorboides was undertaken to identify the haplotypes present, to describe the pattern of haplotype variation that exists within and between the two distinct mountain ranges, and to clarify and elucidate the phylogenetic relationships of P. sanguisorboides by generating a more complete cpDNA phylogeny. An understanding of the dynamics of evolution of a single endemic species within these two mountain ranges will contribute further insight into the complex evolutionary dynamics uncovered within Packera throughout its range.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 LITERATURE CITED
 
Plant samples
Six populations of Packera sanguisorboides were collected as individuals so that chloroplast haplotypes could be unambiguously identified for each individual. Three of the populations were collected from the more northern Sangre de Cristo Mountains and three from the southern Sacramento Mountains. From each population, samples of 13–31 individuals were collected. Two or three leaves were taken from individual plants, packed in foil packages, frozen, and stored at –80°C. A total of 142 individuals were examined.

To obtain a broader sampling of haplotype variation in related taxa, total DNA was also extracted from bulk collections of three other Packera species from the southwestern United States and Mexico. Two species, P. sanguisorbae and P. glabella, are morphologically similar to P. sanguisorboides, while the third Mexican species, P. bellidifolia (Kunth) W.A. Weber & Á. Löve, is related based on results of a phylogenetic analysis of ITS sequence data (Bain and Jansen, 1995 ; Bain and Golden, 2000 ). Collection data for all populations are summarized by collection number and locations in Table 1.


View this table:
[in this window]
[in a new window]
  		Table 1. Collection number, collector, and regional and geographical locations of Packera bellidifolia, P. bernardina, P. glabella, P. sanguisorbae, and six populations of P. sanguisorboides. Vouchers were deposited in University of Lethbridge herbarium (LEA) unless otherwise indicated

 
DNA isolation and purification
Total genomic DNA was isolated from frozen tissue of individual plants by the method of Doyle and Doyle (1987) using 2x cetyltrimethyl ammonium bromide (CTAB) extraction buffer, chloroform : octonal partitioning solvent, and isopropanol precipitation. DNA purification was carried out by ultracentrifugation on a CsCl gradient or by treatment with a caylase cocktail according to the methods of Okada et al. (1997) as modified by Golden and Bain (2000) .

Chloroplast restriction site analysis
Samples of individual DNA suspensions for all populations were digested using restriction enzymes AseI, AvaI, AvaII, BanII, BstNI, DraI, EcoRI, HaeIII, and SspI.

Restriction enzyme fragments were separated by electrophoresis in 1.3–1.5% agarose gels in 1% TAE (Tris-acetic acid-EDTA) buffer and transferred to nitrocellulose membrane by southern blotting for 12–24 h. Filters were heat-fixed for 2 h at 80°C according to the methods of Jansen and Palmer (1987) . Probes were cloned from the lettuce chloroplast genome, radio-labeled with 32P-dATP by nick translation, and hybridized to Packera cpDNA fragments at 65°C for 16–36 h, according to the methods of Palmer (1986) and Jansen and Palmer (1987) . Hybridized filters were exposed to multipurpose phosphor-imaging screens for 1–3 h and read by the Optiquant imaging analysis system (Packard Instruments, Meriden, Connecticut, USA). Enzyme-probe combinations used were those shown to be variable in previous studies (Bain and Jansen, 1996 ; Yates et al., 1999 ; Golden and Bain, 2000 ) and are listed in Table 2. Restriction sites were scored as present or absent and collated to describe individual haplotypes. New haplotypes were named following the format adopted by Golden and Bain (2000) beginning with H19 (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 2. Restriction site mutations detected in Packera by Bain and Jansen (1996) and used in the analysis of P. bellidifolia, P. bernardina, P. sanguisorbae, and P. sanguisorboides. Probes are cloned Lactuca cpDNA fragments. Parentheses indicate inferred fragments

 

View this table:
[in this window]
[in a new window]
  		Table 3. Restriction site characters for nine haplotypes isolated from Packera bellidifolia, P. bernardina, P. glabella, P. sanguisorbae, and P. sanguisorboidesa from southwestern North America

 
Statistical analysis of haplotype variation
Haplotype frequencies in P. sanguisorboides populations were calculated and graphed for each region. A molecular analysis of variance (AMOVA) was performed to examine the distribution of genetic variation within and among populations and between regions. The overall measures of subdivision are expressed as a {Phi}ST statistic and calculated using the AMOVA program (Excoffier et al., 1992 ) in which the frequency of haplotypes as well as the number of mutations between them is taken into account. The effective number of migrants between all populations, Nm, was calculated as described by Golden and Bain (2000) .

Haplotype diversity values (gST) within each population were calculated, and a hierarchical analysis of haplotype diversity performed using the methods of Holsinger and Mason-Gamer (1996) .

To determine possible IBD effects, the correlation of haplotype variation and geographic distribution among populations was examined. A pairwise matrix of geographic distances was compared in a Mantel test with a corresponding matrix of log-transformed co-ancestry coefficients between pairs of populations. Geographic distances were calculated using latitudinal and longitudinal coordinates recorded from topographic maps or by on-site GPS readings. The population genetic structure was calculated using the formula (–ln[1 – FST] = t/2N) in the AMOVA program (Excoffier et al., 1992 ). The Mantel test result is reported as an r value and with a corresponding significance value based on permutation of 1000 multiple random matrices in the R Package computer program (Legendre and Vaudor, 1991 ).

A minimum spanning network was generated using the AMOVA program (Excoffier et al., 1992 ).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 LITERATURE CITED
 
Eight new haplotypes were found in the four species examined. Six of these (H19, H20, H21, H22, H23, H24) plus H18 were present in the six populations of Packera sanguisorboides. The other two, H25 and H26, were found in P. sanguisorbae and P. bellidifolia, respectively. Packera sanguisorboides shares H24 with P. glabella and H18 with P. bernardina (Bain and Jansen, 1996 ; Golden and Bain, 2000 ). Only populations of P. sanguisorboides were polymorphic.

The minimum spanning network combining the eight new haplotypes with the 18 described by Golden and Bain (2000) is shown in Fig. 1. It reveals the presence of two more distinct haplotype groups (C and D) in addition to the two (A and B) identified by Golden and Bain (2000) . Haplotypes H18 and H19 form the very distinct Group D, differing from the other haplotypes in Group C by a minimum of three mutations. Group C differs from Group B by a minimum of two mutations.



View larger version (71K):
[in this window]
[in a new window]
  		Fig. 1. Minimum spanning network of 26 haplotypes identified from Packera. H1–H18 and Groups A and B are described in Golden and Bain (2000) . H19–H26 and Groups C and D are described in the present study (Table 3 ). Each haplotype is separated by a single mutation. Blank boxes represent unidentified haplotypes

 
The nine haplotypes in Group C and D are all restricted to populations and species from Mexico and the southwestern United States. Seven of these haplotypes (five from Group C and two from Group D) are found in P. sanguisorboides. Morphologically, the Packera species in which these haplotypes occur also form two groups, but they do not align with the haplotype groups (C and D). One group includes P. bernardina and P. bellidifolia, two species with entire, tomentose leaves, while the other group approximates Greenman's (1915) sect. Sanguisorboidei and includes the species with large, lobate, mostly glabrous leaves (P. glabella, P. sanguisorbae, and P. sanguisorboides). Haplotypes 19–25 are characteristic of the Sanguisorboidei species, while haplotype 26 is found in the tomentose P. bellidifolia. Haplotype 18 is the only haplotype found in both morphological groups, present in the tomentose California endemic P. bernardina as well as in five of the six populations of P. sanguisorboides. The seven haplotypes found in P. sanguisorboides encompass a broad range of diversity. They are separated by as few as one step (e.g., H18 and H19) and as many as seven steps (H19 to either H20 or H23, Fig. 1).

The distribution and frequency of haplotypes among populations of P. sanguisorboides are shown in Table 4 and Fig. 2. Five of the seven haplotypes are restricted to particular regions. Haplotypes H19, H21, and H24 are northern haplotypes restricted to the Sangre de Cristo Mountains, while H20 and H23 are only found in the southern Sacramento Mountain populations. Haplotype 18 and H22 are constituents of both northern and southern populations. Haplotype 22 is found in low frequency in one northern (8%) and one southern population (18%), while H18 is found in five of the six populations, that is, all but the monomorphic collection (number 569). Population number 552 shows the highest percentage of H18 (52%), and the closest population to it (number 571) is the next highest (31%). Only one H18 individual is present in each of the southern populations.


View this table:
[in this window]
[in a new window]
  		Table 4. Sample size (N), haplotype frequencies, and haplotype diversity (gST) in six populations of Packera sanguisorboides from New Mexico, USA

 


View larger version (28K):
[in this window]
[in a new window]
  		Fig. 2. Haplotype frequencies in six populations of Packera sanguisorboides. Population numbers and locations are in Table 1 . Sample sizes and precise haplotype frequencies are in Table 4

 
The variance of chloroplast types among all populations as detected by the AMOVA was high ({Phi}ST = 0.785), confirming that the haplotypes are highly subdivided among the populations in the study area. The AMOVA further showed that 60.9% (P = 0.0509, 1000 permutations) of the variation occurs within the groups (among populations), 21.5% (P = 0.0010, 1000 permutations) within populations, and 17.6% (P = 0.0010, 1000 permutations) among the groups (between regions). Haplotype diversity values for all P. sanguisorboides populations are presented in Table 4. They range from 0.0 (number 569) to 0.013158 (number 571).

The Mantel test of the correlation between co-ancestry coefficients and geographical distance resulted in an r value of 0.74180, significant at the {alpha} = 5% level (P = 0.03097, 1000 permutations). The effective number of migrants (Nm) is 0.137. The high positive r value and correspondingly low Nm indicates a strong IBD effect among the populations. Combined with the high AMOVA value ({Phi}ST = 0.785), the positive Mantel result indicates that the intraspecific patterns of genetic variation may be the result of an overall lack of migration among populations. Figure 2 shows that the highest frequencies of shared haplotypes are in geographically proximal populations (H23 is shared between numbers 580 and number 581, H21 between numbers 569 and 571, H18 between numbers 552 and 571) but that otherwise the populations are very distinct.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
LITERATURE CITED
 
The identity and distribution of the cpDNA haplotypes found in P. sanguisorboides provide significant insights into the origin and evolution of the species. As well as providing further evidence that P. sanguisorboides is related to morphologically similar species like P. glabella and P. sanguisorbae, the haplotype population profiles identify hybridization and drift as important forces in the evolution of the species and further suggest that the separation of P. sanguisorboides across two mountain ranges and floristic provinces has been evolutionarily less important.

Packera glabella and P. sanguisorbae are the taxa from this study most morphologically and ecologically similar to P. sanguisorboides. The three species also collectively share life history characters. Packera sanguisorbae is a perennial from mesic, high-elevation sites in central Mexico, while P. glabella is a low-elevation, annual or biennial species found mostly in the Gulf Coast region of Texas. Barkley (1978) describes P. sanguisorboides as "biennial or perennial (or annual?)" so it combines the life history traits of both P. glabella (annual or biennial) and P. sanguisorbae (perennial). It is more similar to P. sanguisorbae in its habitat preference for high-elevation understory. The three species also share some Group C haplotype similarity, since P. glabella shares haplotype H24 with one population of P. sanguisorboides and H25, the P. sanguisorbae haplotype differs from H24 by only a single mutation. The higher haplotype diversity within P. sanguisorboides of Group C haplotypes (five vs. two for Group D) and their wider distribution among morphologically related taxa (P. sanguisorbae and P. glabella) provides evidence that they represent the haplotypes found originally within P. sanguisorboides. However, molecular data from the nucleus suggests neither P. glabella nor P. sanguisorbae is likely the sister group of P. sanguisorboides because neither shares the ITS characters that unite P. sanguisorboides with P. bellidifolia, P. hartiana (Heller) W.A. Weber & Á. Löve, P. quercetorum (Greene) C. Jeffrey, and P. multilobata (number 452) (Bain and Jansen, 1995 ; Bain and Golden, 2000 ). So far, the only species identified with the same combination of cpDNA and ITS characters as P. sanguisorboides is P. bellidifolia, which, like P. bernardina, is part of a distinct morphological group characterized by entire, tomentose leaves and not including P. sanguisorboides. Packera bernardina in turn shares Group D haplotypes (H18) with P. sanguisorboides but not P. bellidifolia. The lack of congruence among the three data sets (ITS, cpDNA, morphology) suggests the Packera lineages in the southwest are not well differentiated, in spite of the morphological distinctions that exist between species groups. The shared presence of H18 in the morphologically and geographically distinct P. bernardina and P. sanguisorboides is perhaps most difficult to explain, but some insight may be gleaned from data provided by Bain and Jansen (1996) on population (number 452) of P. multilobata collected from Cuba, New Mexico.

When compared with other P. multilobata populations, number 452 is unusual in two ways. It appears to be made up of annual, rather than perennial, individuals, and its cpDNA profile matches that of P. sanguisorboides more than that of other conspecific populations examined (Bain and Jansen, 1996 ). The actual haplotype profile for the population is not certain because only a bulk sample was examined and polymorphism exists within the population, but, based on the polymorphism pattern, a minimum of two and a maximum of four Group D haplotypes are present. Two of the haplotypes could be H18 and H19, while the other two, if they are present, have yet to be described. The presence of H18 and H19 alone does not explain the profile, so at least one new haplotype must be present. Numerous attempts to recollect the population as individuals have failed because no plants have been found, perhaps because the plants are moisture dependent and rainfall has never reached the levels that occurred during May 1992 (Western Regional Climate Center, 2002 ), when the population was first collected. According to Barkley (1978) , P. multilobata, like P. sanguisorboides, may behave as an annual, biennial, or perennial. Because Group D haplotypes have not been found elsewhere in P. multilobata (Bain and Jansen, 1996 ; Golden and Bain, 2000 ) and the annual habit is relatively uncommon within the species, the shared variation in life history and shared cpDNA haplotypes between P. multilobata (number 452) and P. sanguisorboides may best be explained as resulting from hybridization, either simply between P. multilobata and P. sanguisorboides or by two independent hybridization events involving P. sanguisorboides and P. multilobata each with the same, as yet unidentified, monocarpic species containing Group D haplotypes. The latter hypothesis is favored because it better explains how Group D haplotypes were introduced into both species as well as their wide and apparently disjunct geographic distribution (southern California and New Mexico). The most likely extant monocarpic candidate for the hybridizing species is P. tampicana because its distribution extends south into parts of Chihuahua and north into the central United States as far as Kansas (Freeman and Barkley, 1995 ), including New Mexico (Kartesz, 1999 ). Haplotype characterization for P. tampicana has yet to be done.

The strongest evidence for a role of hybridization in the evolution of P. sanguisorboides is that all populations except the monomorphic number 569 population contain haplotypes from the two distinct haplotype groups (C and D; Fig. 2), and some haplotypes in each of the five polymorphic populations differ by at least four mutations. Such large intrapopulational differences support the idea that all the haplotypes did not evolve in situ and coexist now as a result of hybridization. However, any hybridization that has taken place has not resulted in appreciable homogenization of the populations because the level of genetic subdivision among the populations ({Phi}ST = 0.785; Nm = 0.137) is greater than any other Packera species examined to date except the alpine P. subnuda (DC.) Trock and Barkley ({Phi}ST = 1.0; Nm = 0.0) (Golden and Bain, 2000 ). This could be either because the hybridization occurred a long time ago and drift has subsequently allowed the populations to diverge or the hybridization is contemporary and has had only minimal effect on the level of genetic subdivision among populations. However, since both Group C and D haplotypes are present on both mountain ranges, and gene flow between ranges is likely low, contemporary hybridization, involving a species that has yet to be clearly identified, would have had to have occurred separately on each mountain range. We therefore think it more likely that the hybridization is ancient, perhaps during the Pleistocene, when documented changes in vegetation patterns (Van Devender et al., 1987 ) may have resulted in a wider distribution within this region, of species related to P. sanguisorboides.

Of the two Group D haplotypes (H18 an H19) found in P. sanguisorboides populations, only H18 is thought to be of direct hybrid origin. The two differ by a single mutation and H19 is restricted to a single population (number 552) whose only other haplotype is H18. We therefore believe that H18 was introduced into P. sanguisorboides populations via hybridization and that H19 subsequently evolved from H18 in one population (number 552). Although H18 has so far only been identified from P. bernardina, we anticipate, as already discussed, that it is more widespread and is likely present in more populations of annuals like the P. multilobata (number 452) population described from New Mexico by Bain and Jansen (1996) and perhaps in P. tampicana. The ultimate origin of the Group D haplotypes is unclear. Although current knowledge indicates the separation of Group C and Group D haplotypes probably occurred in Mexico, not enough is known at this time about the spatial and taxonomic distribution of the haplotypes in Mexico to allow further speculation.

The other factor, besides hybridization, that we predicted as important in the evolution of P. sanguisorboides was its north–south distribution across two mountain ranges and two floristic provinces, but our results do not fit the expectations of a vicariance model of divergence. The results of the AMOVA clearly show that most of the variation present in P. sanguisorboides is apportioned between populations (60.9%) and not mountain ranges (17.6%) and that the hierarchical analysis of haplotype diversity (Fig. 3) does not separate the populations based on mountain range distribution. Furthermore, neither mountain range contains a subset of the haplotypes found on the other, so there is no clear evidence of a "stepping-stone" colonization pattern between ranges. All of these data indicate that gene flow between all these populations, not just those on separate mountain ranges, has been low to nonexistent, a result similar to that found by Strand et al. (1996) in their study of Aquilegia from the same region.



View larger version (10K):
[in this window]
[in a new window]
  		Fig. 3. Hierarchical analysis of haplotype diversity in Packera sanguisorboides. Numbers at the nodes are distances (gij) between the daughter nodes. P values are the probability of obtaining a distance as great or greater under the null hypothesis of no daughter-node differentiation

 
Overall, we believe the complex pattern observed in P. sanguisorboides is most compatible with the hypothesis that hybridization has injected high levels of haplotype diversity into populations of a relatively restricted species and that drift has subsequently allowed the populations in both regions to further diverge, obscuring any previous pattern left either by hybridization events or by any divergence that may have occurred between mountain ranges.

Although the notion that ancient hybridization has shaped variation patterns in Packera has been previously presented and is well accepted (Barkley, 1988 ; Bain and Jansen, 1996 ; Golden and Bain, 2000 ), the possibility that ephemeral species are adding another level of complexity to the patterns by periodically hybridizing with other Packera species is new and makes their interpretation even more challenging.

Packera sanguisorboides provides a unique opportunity to study historical evolutionary interactions among cpDNA clades at the northern edge of their distribution. It also provides a useful basis for comparison of patterns of hybridization and haplotype diversity between southern and northern haplotype groups. Overall, the pattern of combining two distinct haplotype groups within a single population is common to both the southern (New Mexico) and northern (southwestern Alberta) regions, suggesting that hybridization among distinct groups has been an important generator of intrapopulational and intraspecific haplotype diversity within Packera throughout its range.


View this table:
[in this window]
[in a new window]
  		Table 5. Geographic and genetic distances between pairs of populations of Packera sanguisorboides from New Mexico, USA. Pairwise geographic distances (in kilometers) are shown above diagonal; pairwise coancestry coefficients are shown below the diagonal (calculation shown in text)

 

   FOOTNOTES
 
1 The authors thank T. Lowry, J. Mygatt, R. Sivinski, and R. Spellenberg, for assistance with plant distribution information; Fernando Chiang and Bonnie Clark for providing plant material; and E. Schultz and D. Sillito for field assistance. The staff at Lincoln National Forest, New Mexico, USA, also provided valuable assistance with field work. Back

2 Author for reprint requests (Bain{at}uleth.ca ) Back


   LITERATURE CITED
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Axelrod D. I. P. H. Raven 1985 Origins of the cordilleran flora. Journal of Biogeography 12: 21-47

Bain J. F. 1988 Taxonomy of Senecio streptanthifolius Greene. Rhodora 90: 277-312[ISI]

Bain J. F. J. L. Golden 2000 A phylogeny of Packera (Senecioneae; Asteraceae) based on ITS sequence data and a broad sampling of outgroups. Molecular Phylogenetics and Evolution 16: 331-338[CrossRef][ISI][Medline]

Bain J. F. R. K. Jansen 1995 A phylogenetic analysis of the aureoid Senecio (Asteraceae) complex based on ITS sequence data. Plant Systematics and Evolution 195: 209-219[CrossRef][ISI]

Bain J. F. R. K. Jansen 1996 Numerous chloroplast DNA polymorphisms are shared among different populations and species in the aureoid Senecio (Packera) complex. Canadian Journal of Botany 74: 1719-1728[ISI]

Barkley T. M. 1962 A revision of Senecio aureus Linn. and allied species. Transactions of the Kansas Academy of Science 65: 318-408[CrossRef]

Barkley T. M. 1968 Taxonomy of Senecio multilobatus and its allies. Brittonia 20: 267-284[CrossRef][ISI]

Barkley T. M. 1978 Senecio. North American Flora. Series II 10, 50–139

Barkley T. M. 1988 Variation among the aureoid Senecios of North America: a geohistorical interpretation. Botanical Review 54: 82-106

Desmesure B. B. Comps R. J. Petit 1996 Chloroplast DNA phylogeography of the common beech (Fagus sylvatica L.) in Europe. Evolution 50: 2515-2520[CrossRef][ISI]

Dumolin-Lepegue S. B. Demesure S. Fineschi V. Le Corre R. J. Petit 1997 Phylogeographic structure of white oaks throughout the European continent. Genetics 146: 1475-1487[Abstract]

Doyle J. J. J. L. Doyle 1987 A rapid isolation procedure for small quantities of fresh tissue. Phytochemical Bulletin 19: 11-15

Dye A. J. W. H. Moir 1977 Spruce-fir forest at its southern distribution in the Rocky Mountains, New Mexico. American Midland Naturalist 97: 133-146[CrossRef][ISI]

Excoffier L. P. E. Smouse J. M. Quattro 1992 Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131: 479-491[Abstract]

Freeman C. C. T. M. Barkley 1995 A synopsis of the genus Packera (Asteraceae: Senecioneae) in Mexico. Sida 16: 699-709

Golden J. L. J. F. Bain 2000 Phylogeographic patterns and high levels of chloroplast DNA diversity in four Packera (Asteraceae) species in southwestern Alberta. Evolution 54: 1566-1579[ISI][Medline]

Greenman J. M. 1915 Monograph of the North and Central American species of the genus Senecio, Part II. Annals of the Missouri Botanical Garden 2: 573-626[CrossRef]

Holsinger K. E. R. J. Mason-Gamer 1996 Hierarchical analysis of nucleotide diversity in geographically structured populations. Genetics 142: 629-639[Abstract]

Jansen R. K. J. D. Palmer 1987 Chloroplast DNA from lettuce and Barnadesia (Asteraceae): structure, gene localization, and characterization of a large inversion. Current Genetics 11: 553-564

Kartesz J. T. 1999 A synonymized checklist and atlas with biological attributes for the vascular flora of the United States, Canada, and Greenland, 1st ed. In J. T. Kartesz and C. A. Meacham [eds.], Synthesis of the North American flora, Version 1.0. North Carolina Botanical Garden, Chapel Hill, North Carolina, USA

Kowal R. R. 1975 Systematics of Senecio aureus and allied species on the Gaspé Peninsula, Quebec. Memoirs of the Torrey Botanical Club 23: 1-113

Legendre P. A. Vaudor 1991 The R Package: multidimensional analysis, spatial analysis. Département de sciences biologiques, Université de Montréal, Montreal, Quebec, Canada

Okada M. R. Whitkus T. K. Lowrey 1997 Genetics of adaptive radiation in Hawaiian and Cook Islands species of Tetramolopium (Asteraceae: Astereae). I. Nuclear RFLP marker diversity. American Journal of Botany 84: 1236-1246[Abstract]

Palmer J. D. 1986 Isolation and structural analysis of chloroplast DNA. Methods in Enzymology 118: 167-186[ISI]

Peet R. K. 1988 Forests of the Rocky Mountains. In M. G. Barbour and W. D. Billings [eds.], North American terrestrial vegetation, 63–101. Cambridge University Press, New York, New York, USA

Strand A. E. B. G. Milligan C. M. Pruitt 1996 Are populations islands? Analysis of chloroplast DNA variation in Aquilegia. Evolution 50: 1822-1829[CrossRef][ISI]

Van Devender T. R. J. L. Betancourt M. Wimberly 1983 Biogeographic implications of a packrat midden sequence from the Sacramento Mountains, south-central New Mexico. Quaternary Research 22: 344-360[CrossRef][ISI]

Van Devender T. R. R. S. Thompson T. R. Betancourt 1987 Vegetation history of the deserts of southwestern North America; the nature and timing of the LateWisconsin-Holocene transition. In W. F. Ruddiman and H. E. Wright [eds.], The geology of North America, vol. K-3, North America and adjacent oceans during the last deglaciation, 323–352. Geological Society of America, Boulder, Colorado, USA

Western Regional Climate Center. 2002 Monthly total precipitation, Cuba, New Mexico. Website: http://www.wrcc.dri.edu/cgi-bin/cliMONtpre.pl?nmcuba

Yates J. S. J. L. Golden J. F. Bain 1999 A preliminary phytogeographical analysis of inter- and intrapopulational chloroplast DNA variation in Packera pseudaurea (Asteraceae: Senecioneae) from southwestern Alberta and adjacent Montana. Canadian Journal of Botany 77: 305-311[ISI]





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bain, J. F.
Right arrow Articles by Golden, J. L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Bain, J. F.
Right arrow Articles by Golden, J. L.
Agricola
Right arrow Articles by Bain, J. F.
Right arrow Articles by Golden, J. L.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS