| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics and Phytogeography |
2Laboratorio Ecotono, Universidad Nacional del Comahue, CRUB, Quintral 1250 Bariloche, RN 8400 Argentina; 3Department of Plant Biology, Southern Illinois University, Carbondale, Illinois 62901-6509 USA
Received for publication April 7, 2006. Accepted for publication February 21, 2007.
ABSTRACT
Phylogenies can provide valuable information on biotic and abiotic factors associated with speciation. We examined species relationships in Tristerix (Loranthaceae), a genus of 11 species with an Andean distribution from Colombia to Chile. A previous classification divided Tristerix into subgenera Tristerix (two species) and Metastachys (nine species). We tested this classification by generating a molecular phylogeny of the genus using nuclear ribosomal DNA ITS and chloroplast atpB-rbcL intergenic spacer and trnL-F regions. All partitions generally gave congruent trees, thus a combined analysis was conducted. Tristerix was composed of a northern clade (six species) and a southern clade (four species). Tristerix verticillatus and T. penduliflorus (Metastachys) were strongly supported as members of the (southern) subgenus Tristerix clade. Speciation appears to be correlated with the emergence of matorral and cloud forest biomes and is driven by interactions with pollinators and seed dispersers. Tristerix aphyllus is sister to T. corymbosus of the matorral, not to neighboring temperate forest populations, thus rendering the latter species paraphyletic. This ecological speciation event may have occurred in sympatry. Tristerix provides excellent examples of how, during the orography of the Andes, many dynamic and interacting ecological factors have influenced their speciation.
Key Words: atpB-rbcL • chloroplast DNA • Loranthaceae • paraphyletic species • ribosomal DNA • South America • sympatric speciation • trnL-F region
Advances in phylogenetic methods have led to rapid progress in addressing plant speciation such that students of this topic are no longer "evolutionary biologists' poor cousins, doomed to eternal speculation about untestable theories" (Rieseberg and Wendel, 2004
, p.3). Even if there is general acceptance that species are biologically real entities (Rieseberg et al., 2006
), the mode in which speciation takes place remains controversial and poorly understood (Kirkpatrick and Ravigné, 2002
). Mechanistic classifications of speciation modes include polyploidy, genetic drift, divergence under uniform selection, and ecological speciation (Schluter, 2005
). The latter process, in which divergent selection acts on traits that represent adaptations to contrasting environments, can occur in allopatry or sympatry. Convincing evidence of sympatric speciation in plants was recently reported for oceanic palms (Savolainen et al., 2006
), but questions remain as to whether or not this is an isolated example with little relevance to diversification of continental floras (Jiggins, 2006
). Even in phytophagous insects, which represent the best cases of sympatric speciation in animals, discounting allopatric speciation is challenging (Berlocher and Feder, 2002
).
The botanical equivalent of phytophagous insects are parasitic plants (Pennings and Callaway, 2002
), particularly mistletoes that occur on aerial portions of woody hosts. Thus, mistletoes represent good candidates to further study modes of speciation. Although all mistletoes are found in the order Santalales (families Loranthaceae, Misodendraceae, Santalaceae, and Viscaceae), the term refers to a habit, not to a specific taxonomic group. Mistletoes experience particularly constrained life histories because of associations with other organisms in their environment (Watson, 2001
; Aukema, 2003
). Specifically, all mistletoes are primarily dependent upon their host plant, thus their distributions obligatorily track those of their hosts (Norton and Carpenter, 1998
). Moreover, many loranthaceous mistletoes are highly adapted to either insect or bird pollination, and most are dispersed by animals, primarily birds (Amico and Aizen, 2000
; Watson, 2001
; Restrepo et al., 2002
). These close coevolutionary relationships with animals have resulted in numerous modifications of mistletoe flower and fruit features (Kuijt, 1969
; Reid, 1991
).
The largest mistletoe family is Loranthaceae (73 genera, ca. 900 species), which includes root and stem parasitic plants. Root parasitism is found in three genera: Nuytsia of western Australia, Atkinsonia of eastern Australia, and Gaiadendron of the neotropics. Loranthaceae have reached their greatest degree of differentiation in the southern hemisphere and have long been considered a family of Gondwanan origin. Of the 16 loranth genera found in South America, Gaiadendron, Desmaria, Ligaria, Notanthera, and Tristerix are considered relictual based on their base chromosome numbers, low numbers of species, and restricted distributions (Barlow, 1983
). According to the most recent taxonomic monograph of the genus (Kuijt, 1988
), Tristerix consists of 11 species endemic to the Andes of South America. Distributions range from subparamo elevations in north-central Colombia to low elevation and adjacent areas of the Andes of Argentina and south-central Chile (Fig. 1, Table 1). Tristerix is most diverse in Peru where four of its seven species are endemic. Tristerix aphyllus is found only in central Chile, and T. secundus is restricted to the Cordillera Occidental of Colombia. Only three species have wide distributions: T. longebracteatus in northern South America (from central Colombia to central Peru, Table 1) and T. verticillatus and T. corymbosus in southern South America. Tristerix verticillatus is found from southern Bolivia to the eastern slopes of the Andes in Argentina and central Chile. Tristerix corymbosus is found along the western slopes of the Andes in Chile and in Argentina northwest of Patagonia. All but three species of Tristerix are sympatric with at least one other species (Table 1).
|
|
; Kuijt, 1988
). The other host-specific species is T. chodatianus of Peru, that occurs only on Polylepis (Rosaceae). Although the relationship is not obligate, T. verticillatus and T. penduliflorus are frequently found on Schinus (Anacardiaceae).
Tristerix can be distinguished from other South American Loranthaceae by the presence of simple, terminal, racemose inflorescences with large, brilliant red or red and yellow flowers. These are tetramerous or pentamerous and the stamens are isomorphic with versatile anthers. Among the five relictual South American genera, only Tristerix and Ligaria lack epicortical roots. Only Tristerix has undergone a radiation into multiple species; the other four genera are monospecific. Tristerix also differs from these four genera, which are restricted to temperate or high altitude biomes, in that it also occurs in tropical areas (Table 1). In his revision, which included descriptions of three new species, Kuijt (1988)
divided the genus into two subgenera, Tristerix (T. aphyllus and T. corymbosus) and Metastachys (the remaining nine species). This classification was based mainly in the number of petals and the presence or absence of bracteoles. The objective of the present study is to test this subgeneric classification by reconstructing the phylogeny of the genus using molecular markers—the first phylogeny for any genus in Loranthaceae. Moreover, we will use the molecular phylogeny to address the distribution of various morphological characters and life history features, particularly with respect to how abiotic and biotic factors have influenced speciation in the genus.
MATERIALS AND METHODS
Taxon sampling and molecular methods
Ten of the 11 species of Tristerix were sampled for this study (Fig. 1, Table 2); only T. secundus of Colombia was not obtained. Individuals from different populations were used to test for monophyly in seven species. For two accessions of T. grandiflorus, the atpB-rbcL spacer was not obtained. For one accession of T. aphyllus and one of T. penduliflorus, ITS sequences are lacking. Ligaria cuneifolia (Ruiz & Pav.) Tiegh. and Notanthera heterophylla (Ruiz & Pav.) G. Don were also sampled and used as outgroups.
|
). One-tenth dilutions of the genomic DNA were used for PCR amplifications for the three gene partitions. Typical PCR amplification reactions included: 1x buffer (Promega, Madison, Wisconsin, USA; 10 mmol/L Tris HCl, 50 mmol/L KCl, pH 8.3), 1.5 mmol/L MgCl2, 50 µmol/L dNTPs, 1 unit Taq polymerase, 0.4 µmol/L of each primer, and ca. 30 ng genomic DNA. For ITS (nuclear internal transcribed spacer and 5.8S ribosomal DNA) amplifications, 5% dimethylsulfoxide (DMSO, final concentration) was added. Ready-to-Go PCR Beads (Amersham Biosciences, Piscataway, New Jersey, USA) were utilized for some problematic PCR reactions.
The ITS region was amplified and sequenced using the primer pair 18S 1830 forward (5'-AAC AAG GTT TCC GTA GGT GA-3') and 26S 40 reverse (5'-TCC TCC GCT TAT TGA TAT GC-3'). In addition, two regions from the chloroplast genome were amplified and sequenced: the spacer between the genes atpB and rbcL using the primer pair atpB 1 reverse (5'-GAA GTA GTA GGA TTG ATT CTC AT-3') and rbcL 60 reverse (5'-CAG GAG TAT AAT AAG TCA TTG-3') and the trnL-F region (including the trnL intron) using the two primers described in Taberlet et al. (1991)
.
The PCR thermal cycle profile for ITS was 5 min at 95°C, 35 cycles of 1 min at 94°C, 1 min at 52°C, and 1 min at 72°C, with a final extension of 10 min at 72°C. For the chloroplast regions, a touch down profile was used consisting of 5 min at 95°C, 5 cycles of 30 s at 94°C, 30 s at 52°C, and 1 min at 72°C, followed by 33 cycles of 30 s at 94°C, 30 s at 48°C, and 1 min at 72°C, with a final extension of 10 min at 72°C. In all reactions negative controls, lacking genomic DNA, were run to check for DNA contamination. Cycle sequencing reactions were performed directly on the purified PCR products following standard protocols using BigDye terminator Cycle Sequencing Ready Reaction Kit with AmpliTaq DNA Polymerase (Applied Biosystems, Foster City, California, USA) with Better Buffer (The Gel Co., San Francisco, California, USA). Sequences were determined with an ABI 377 automated sequencer (Applied Biosystems). All sequences generated in this study have been deposited with NCBI GenBank under the following accession numbers: DQ442916 to DQ442986.
Phylogenetic analyses
Because PCR amplifications were problematic for several DNA samples obtained from herbarium specimens, taxon sampling between the ITS and chloroplast data sets was not identical. Individual sequences were aligned manually using BioEdit (Hall, 1999
). PAUP* (Swofford, 2002
) was used to conduct maximum parsimony (MP) analyses; the branch-and-bound search option was used for all separate and combined partitions. Nodal support was assessed using the nonparametric bootstrap (BS) (Felsenstein 1985
) with 1000 pseudoreplicates using a heuristic search with the tree-bisection-and-reconnection option for branch swapping. Indels in the ITS data set did not provide any phylogenetic information and were treated as missing data. In contrast, indels were informative for the chloroplast partitions and thus were manually coded as present or absent. Gaps were considered homologous only when they shared identical boundaries and length. Insertions generally consisted of tandem duplications and only those of the same length were used in gap coding. Moreover, only those gaps shared with the ingroup were coded for the outgroup. A total of 27 indels were coded for the atpB-rbcL, and 11 for the trnL-F partitions. All alignments used in this study can be found at http://www.parasiticplants.siu.edu/Alignments/Alignments.html.
Models of sequence evolution for each of the partitions and for all combined data sets were determined using MrModeltest (Nylander, 2004
). The selected model was used to perform maximum likelihood (ML) and Bayesian inference (BI) analyses. For the nuclear partition, the HKY85+
model (Hasegawa, Kishino and Yano, 1985
) was selected, whereas for the chloroplast partition the GTR+ (General Time Reversible model, Tavaré, 1986
) was used. When all three partitions were combined the selected model was GTR+. Maximum likelihood analysis was performed using PAUP*, and nodal support was obtained using the nonparametric bootstrap with 100 pseudoreplicates. Bayesian inference was performed using MrBayes (Huelsenbeck and Ronquist, 2001
) wherein two independent analyses were run with four chains each, for 5 x 106 generations. Trees and parameters were saved every 100 generations, producing 50,000 trees. Starting model parameters were assigned uniform prior probabilities and estimated as part of the analysis, but in cases in which more than one partition was analyzed, the estimates were unlinked between data partitions, thus allowing each to vary independently. The burn-in was determined by stationary in the –ln likelihood score. The split frequency (variance between the two independent runs) in all cases was below 0.001, thus confirming that sampling was from the posterior probability distribution. In the case of the chloroplast partition in which gaps were coded, a mixed matrix was used as an input file in MrBayes. Here the nucleotide sequences were analyzed with the selected model, whereas the gaps were treated as restriction site data.
Eleven discrete morphological characters were scored for each species using data from Kuijt (1988)
and from direct observation of living plants and herbarium specimens. Morphological character state transformations were optimized using MacClade (Maddison and Maddison, 2000
) (see Appendix S1). Photographs of eight of the 11 species of Tristerix can be found on the Parasitic Plant Connection web site (Nickrent, 1997 onward
).
RESULTS
The ITS partition had 695 aligned positions and contained essentially no ambiguous positions. Including the outgroup, this partition had 178 variable sites, of which 74 were parsimony informative, whereas 44 of the 122 variable sites were parsimony informative for the ingroup alone. Maximum parsimony analysis resulted in eight equally parsimonious trees (L = 269). Double peaks at several sites in the sequence electropherograms were found in different accessions of Tristerix aphyllus and T. corymbosus. A total of 16 polymorphic sites were found, and these were coded as ambiguous. Five of these sites were polymorphic in all individuals of T. corymbosus and T. aphyllus.
With the outgroup included, the atpB-rbcL intergenic spacer had 899 aligned positions with 155 variable sites, of which 70 were parsimony informative. Considering just the ingroup, the number of variable sites dropped to 93, of which 46 were parsimony informative. The trnL-F region had 720 aligned positions with 133 variable sites, of which 65 were parsimony informative. The ingroup had 83 variable sites for this region, of which 45 were parsimony informative.
All analytical methods (MP, ML, and BI) of both chloroplast regions (atpB-rbcL, trnL-F) resulted in trees with congruent topologies; thus they were combined into one chloroplast partition and analyzed together. Eighteen equally parsimonious trees (L = 383) were obtained from an analysis of this combined chloroplast partition. The nuclear ITS tree (with smaller taxon sampling) generally had a topology similar to that chloroplast tree (but was less resolved; see Fig. 2); thus these data were combined and a "total evidence tree" was generated (Fig. 3). In this analysis, three most parsimonious trees were found of 678 steps.
|
|
DISCUSSION
Results of this molecular phylogenetic study of Tristerix provide, for the first time, compelling evidence that can be used to generate a new classification of the genus as well as data useful in addressing the evolution of morphological features and life history traits. As will be highlighted, this study also provides an example in which the phylogenetic structure of populations brings to the fore issues such as paraphyletic species and sympatric speciation.
Classification
Our results clearly differentiate two major clades within Tristerix, one composed of six species with northern distributions (Colombia, Peru, and Ecuador) and another with four species from more southern regions in South America (southern Peru and Bolivia, Chile, and Argentina). The differentiation of these two clades is well supported by both nuclear and chloroplast genes using all analytical methods. These clades do not conform to the previously proposed subgeneric classification by Kuijt (1988)
but appear to correlate more with geographic distribution (Fig. 1, Table 1). Two features of the Kuijt (1988)
classification appear valid, i.e., the recognition of a close relationship between T. corymbosus and T. aphyllus (subg. Tristerix) and the affinity of seven of the nine remaining subg. Metastachys species. Molecular phylogenetic analyses strongly support the transfer of two species, T. verticillatus and T. penduliflorus, from subg. Metastachys to subg. Tristerix. Given its morphology and geographical distribution, we can assume that T. secundus (not sampled in this study) is also a member of Metastachys.
Paraphyletic species
We do not advocate changing the current circumscriptions for any Tristerix species, although adherence to one of several competing species concepts might compell us to do so. Two cases of nonmonophyly for the chloroplast genome were seen: T. longebracteatus and T. corymbosus. For the former, the two accessions sampled were unresolved relative to the T. grandiflorus clade; however, this likely stems more from lack of data (missing atpB-rbcL spacer sequences) than from an actual genetic basis. The tree shown in Fig. 3 indicates that the three populations of T. aphyllus represent an "apospecies" and that the four populations of T. corymbosus could be considered a "metaspecies" (sensu Donahue, 1985
; De Queiroz and Donoghue, 1988
). But the data presented herein show that T. corymbosus is "positively paraphyletic." Specifically, the chloroplast gene tree strongly supports a topology in which T. aphyllus is more closely related to the TA and CH populations of T. corymbosus than to the other two populations of that species (LI and LL). Paraphyly is not shown by the ITS tree (Fig. 2); however, given that coalescence occurs four times faster in maternally inherited chloroplast genes than in biparentally inherited nuclear genes (Hudson, 1990
), this result is not unexpected. Moreover, we outline next a scenario that describes how this speciation event may have occurred. For these reasons, we consider T. corymbosus a "plesiospecies" (following Olmstead, 1995
), as opposed to splitting it into several smaller monophyletic units as required by the geneological species concept (Baum and Shaw, 1995
). Our rationale for this is that T. aphyllus is a genetically and morphologically distinct, clearly defined taxon that derives from populations representing one portion of the distribution of a wider-ranging taxon, T. corymbosus. This type of speciation in plants likely occurs very commonly (Rieseberg and Brouillet, 1994
).
The polymorphisms found in ITS between these two species might suggest a scenario for hybridization; however, some of the individuals that shared polymorphisms were widely separated geographically (Fig. 1). If the ITS ribosomal cistrons were distributed on separate chromosomes, these polymorphisms could have arisen independently via incomplete concerted evolution. Although Tristerix aphyllus and T. corymbosus are closely related, showing genetic differentiation only in the chloroplast genes, several morphological autapomorphies are found in the former species: the absence of leaves, red fused cotyledons, spherical white fruits, red inflorescences, extensive endophytic growth, and erect flowers. We feel these are compelling reasons to retain T. aphyllus as a species.
Morphological character evolution
Using the combined data, the topology of the most likely Bayesian tree was simplified by reducing most species to a single terminal; the exceptions were T. corymbosus and T. verticillatus, that were polymorphic for several features among populations. This simplified tree (Fig. 4) was then used to explore the evolution of 11 morphological characters as well as associated biomes and pollinators.
|
Character 1 (haustorial endophyte) appears to be a synapomorphy for Tristerix. The two outgroup genera have localized infections, whereas members of the southern clade all spread within the host cortex, thus forming new shoots at regions distal to the original infection. Unfortunately, data on the endophyte behavior is lacking for the northern clade, except T. longebracteatus, that has the spreading cortical strands. The occurrence of this state in this one taxon results in a most parsimonious reconstruction wherein it is present in all members of the northern clade. Further observations are needed to confirm this hypothesis.
We have not identified any morphological features that support the clade composed of all southern taxa, despite the strong support this clade obtained from molecular data (Fig. 3). Within this clade, the close relationship between Tristerix aphyllus and T. corymbosus is supported by two morphological characters that distinguish them from the rest of the species: the presence of bracteoles subtending each flower (character 5) and four petals (character 8). Indeed it was these two features that lead Kuijt (1988)
to segregate them in subgenus Tristerix. The inclusion of T. verticillatus and T. penduliflorus with the southern clade taxa is strongly supported. Given its inflorescence structure (flowers in whorls), pollinators, and distribution, T. penduliflorus would most certainly be placed with the northern clade taxa, but this conflicts with the molecular data (nuclear and combined analyses) that strongly support its position as sister to T. verticillatus of the southern clade.
As mentioned, the northern clade is supported by the presence of red-yellow-red banded flowers. Tristerix peruvianus was described by Kuijt (1988)
as having fenestrated flower buds, and he postulated that this is unique among Tristerix species. This feature, apparent during floral bud maturation, is recognized by the separation of the petals from one another approximately midway along the length of the corolla tube, thereby forming elongated vents. We found that T. pubescens, T. longebracteatus, and possibly T. chodatianus also have fenestrated floral buds. More field observations are necessary, but it is possible that this character is an additional synapomorphy for the northern clade.
Molecular data strongly support a clade composed of Tristerix longebracteatus, T. grandiflorus, T. peytonii, and T. pubescens, and this clade is marked by one synapomorphy, a foliaceous bract below the flower (character 6). These four closely related species are distinguished mainly by fruit color and the size and shape of the flower. In the combined data tree, no support was obtained for within-clade relationships (Fig. 3), thus the homologous condition for the two characters states shared by T. grandiflorus and T. peytonii (character 7, floral bud >8 cm and character 9, flower curved) cannot be determined. If one looks at the three possible alternative resolutions of the polytomy, then two favor parallelism and the other is uncertain, i.e., it is equally parsimonious to invoke parallelism or homology. On the other hand, when optimized for the most likely tree topology from the combined analysis, parallel evolution for these characters is suggested (Fig. 4). The four species in this clade show continuous interspecific variation in pubescence and intraspecific polymorphism in leaf shape, thus highlighting their close relationship (see Kuijt, 1988
, p. 35). Taken together, these data suggest that this species complex has diverged only relatively recently, in agreement with geological data that indicate that high Andean habitats have been available for colonization only since the end of the Tertiary (Simpson, 1975
). This also parallels other neotropical taxa such as Inga, in which diversification has been recent and is associated with Andean orogeny (Richardson et al., 2001
).
Well supported by molecular data is the sister relationship between the T. chodatianus and T. peruvianus clade and the remaining members of the northern clade. These two species both have red fruits and may share an additional synapomorphy: a prominent calyculus that spreads when dried. Kuijt (1988)
states that T. peruvianus is similar to T. secundus (not sampled in this study) but differs by having smaller, thinner, and more glaucous leaves. Tristerix chodatianus shares with these two species the characteristic of flowers occurring in whorls of three or four, but as T. peytonii and T. penduliflorus also share this feature, it does not appear to be diagnostic for this clade.
Tristerix chodatianus has a distinctive leaf that bears a sclerotic tip (character 4). Strangely, a similar leaf tip is seen in Ligaria cuneifolia (used here as an outgroup), but not in any other taxa. Ligaria is polymorphic for this feature; in some individuals it is prominent, whereas in others it is inconspicuous. If the character is homologous, then at least three losses are required to explain its distribution on the phylogenetic tree (Fig. 4); thus parallelism is a more parsimonious scenario for this character.
Pollination syndromes and speciation in Tristerix
Among the South American extant relictual mistletoe genera, only Tristerix and Ligaria, with their large tubular red flowers, are bird pollinated (Kuijt, 1988
; Rivera et al., 1996
; Medel, 2002; Aizen, 2003
). The other genera (Notanthera, Desmaria, and Gaiadendron) have small, light-colored (pink, orange, and yellow) flowers that are pollinated mainly by insects (Kuijt, 1963
, 1985
; G. C. Amico, unpublished data). Although all these genera have similar geographic distributions, Tristerix and Ligaria flower during winter, whereas the other genera flower in summer. The evolution of cold-season flowering may be related to the ability of homothermous birds to function as pollinators during both winter and summer, as opposed to poikilothermous insects that can only function during the summer.
Hummingbirds are most diverse in the Andes, and several species occur along the northern range of Tristerix at elevations as high as 5000 m a.s.l. (Fjeldsa and Krabbe, 1990
; Altshuler et al., 2004
). To date there is no published information on hummingbird pollination of Tristerix from the high Andes. We observed this in the southern part of their distribution, where the monochromatic red flowers of Tristerix species are pollinated solely by hummingbirds (mainly two species: Sephanoides sephaniodes and Patagonas gigas). In the northern part of their range, Tristerix species are pollinated by both humingbirds and flower-piercers. At two localities in Peru (Huascarán National Park and Cuzco), we observed Colibri coruscans pollinating T. pubescens and T. penduliflorus (the latter species was also visited by Lesbia nuna). These mistletoes were also visited by the flower-piercer Diglossa brunneiventris, a species documented as the pollinator of T. longebracteatus in northern Peru (Graves, 1982
). The genus Diglossa, with 16 primarily high-elevation Andean species, has a distribution that coincides with that of Tristerix (Fjeldsa and Krabbe, 1990
). These birds are generally considered nectar robbers, but they also function as authentic pollinators of several plants, including Tristerix (Graves, 1982
; Maloof and Inouye, 2000
). Eight of the 11 species of Tristerix overlap with the distribution of Diglossa spp., and all these species have red-yellow-red bicolored flowers (Fig. 4). As discussed, some of the northern clade Tristerix have fenestrated flowers, and all but T. peruvianus also have the bicolored corollas. We suggest that these function as floral guides and are recognized by Diglossa spp. that open the flower at the fenestration, thereby affecting pollination. Although pollinated by Diglossa, the floral banding pattern in T. penduliflorus is different (red-yellow), and this species does not show fenestrae, thus suggesting an independent evolution of the yellow floral guide.
Two species of Tristerix have colonized cloud forests: Tristerix peytonii of southern Peru and T. grandiflorus in northern Peru and Ecuador (Fig. 4, Table 1). It is noteworthy that these species have the longest corolla tubes among all species in the genus. For T. grandiflorus, flowers may be up to 16 cm long, thus suggesting directional selection being driven by the pollinating bird. Kuijt (1988)
hypothesized that the distinctive flowers of T. grandiflorus are associated with the cloud forest hummingbird Ensifera ensifera, and this correlates well with its beak length (10 cm).
Seed dispersal in Tristerix
Relatively few studies have been carried out on seed dispersal in Tristerix, but both mammals (Dromiciops, discussed later) and birds are involved (Parker, 1980
; Amico and Aizen, 2000
; Medel et al., 2004
). The Gondwanan families Contingidae and Tyrannidae represent the oldest bird lineages in South America and were present there prior to 40 million years before present (mybp; Ericson et al., 2003
; Barker et al., 2004
). Tristerix chodatianus and T. peruvianus of central Peru are dispersed by the cotinga, Zartornis stresemanni (Parker, 1980
), whereas T. aphyllus, T. corymbosus, and T. verticillatus of Chile are dispersed primarily by the mockingbird Mimus thenca (Mimidae) and secondarily by the flycatcher Elaenia albicepts and a thrush Turdus falcklandii. There is no information on the seed dispersal of the other six Tristerix species, but based on their distributions, the bird families Contingidae, Tyrannidae, and Turdidae are the most likely candidates. Turdidae and Mimidae have a Laurasian origin and appeared on the South American continent much later (2–5 mybp) than Contingidae and Tyrannidae (Ericson et al., 2003
and references therein). We therefore suggest that the early ancestors of Tristerix were dispersed by Cotingidae, Tyrannidae, and Dromiciops and more recently by Turdidae and Mimidae.
In the temperate forests of Chile and Argentina, seeds of T. corymbosus are dispersed by Dromiciops gliroides, an old marsupial lineage with a Gondwanan origin (Amico and Aizen, 2000
), which feeds on the ripe green fruits. Fruits are yellow in the matorral populations of T. corymbosus, and seeds are dispersed by birds (Hoffmann et al., 1986
; G. C. Amico, unpublished data). Seeds from the sympatric species T. aphyllus with white fruits are dispersed by the same bird (Medel et al., 2002
). The topology of the molecular tree (Fig. 3) indicates that the temperate forest populations (LL and LI) diverged first and that the matorral populations (CH and TA) are more derived. Optimization of fruit color on the molecular tree does not help to identify an ancestral state because reconstruction is equivocal; however, we hypothesize that the green color found in the forest populations results from heterochrony in fruit development and not to selection pressure from the different dispersers (G. C. Amico, unpublished manuscript).
Did Tristerix aphyllus arise via sympatric speciation?
It has been proposed that host races of phytophagous insects represent an intermediate stage between polymorphic populations and reproductively isolated species (Berlocher and Feder, 2002
; Drés and Mallet, 2002
). According to Drés and Mallet (2002, p. 471), "Host races are genetically differentiated, sympatric populations of parasites that use different hosts, and between which there is appreciable gene flow." Host races have been described numerous times for mistletoes based on seed inoculation experiments (Clay et al., 1985
), allozyme studies (Glazner et al., 1988
; Nickrent and Stell, 1990
), and AFLPs (Jerome and Ford, 2002
); however, despite implications that this process leads to speciation (Norton and Carpenter, 1998
), robust documentation is lacking. Although objections to sympatric speciation have been voiced (Futuyma and Mayer, 1980
; Felsenstein, 1981
), theoretical studies show that this process is plausible under real biological conditions such as when host-specific deleterious and beneficial mutations exist (Kawecki, 1996
, 1997
) and when genes control assortative mating (Johnson et al., 1996
; Dieckmann and Doebeli, 1999
). In addition, pleiotropy between host choice and assortative mating is a likely route toward the evolution of linkage disequilibrium that leads to distinct genetic entities (Drés and Mallet, 2002
).
The morphological and genetic differences provide strong evidence that T. aphyllus is a species distinct from T. corymbosus. Among Loranthaceae, Tristerix aphyllus is unique because it exists as an endophyte within the cactus cortex and its only emergent parts are inflorescences (Mauseth et al., 1984
, 1985
). This extreme reduction in vegetative morphology and dependence upon the host for most carbon (Kraus et al., 1995
) parallels other mistletoes that approach holoparasitism, such as Arceuthobium (Viscaceae). Given the topology of the molecular tree (Fig. 3), reciprocal monophyly for the two species is not met. Aside from the classification issues this topology presents (Kizirian and Donnelly, 2004
), it also shows that T. aphyllus is derived from matorral populations (CH and TA) of T. corymbosus. The Chilean matorral formed during the Pliocene-Pleistocene (Hinojosa and Villagran, 1997
), i.e., relatively recently compared with the high-elevation Andean forests that began in the Miocene (Taylor, 1991
). We propose that the emergence of this new biome brought the ancestral Tristerix into contact with new host niches, specifically the cactus genera Echinopsis and Eulychnia. The model of sympatric speciation developed by Dieckmann and Doebeli (1999)
performs best in recently colonized habitats where no competition exists with other sympatric species, as is the case with the Chilean matorral. With phytophagous insects, mating on the host is the most straightforward path by which host shifts can reduce gene flow (Bush, 1969
). In Tristerix mistletoes, such host fidelity is attained via the behavior of the seed-dispersing bird (Mimus thenca) that visits both species and prefers high perches (Martínez del Rio et al., 1995
; Medel et al., 2002
, 2004
). This behavior results in more seeds being deposited on cactus hosts (the highest vegetation in the area) than on other hosts. We suggest that, over time, pleiotropy between host choice and assortative mating genes lead to linkage disequilibrium and genetic isolation of the two mistletoe species. At present, peak flowering times overlap only slightly between T. aphyllus and T. corymbosus, thus limiting interspecific pollen transfer. The two species appear reproductively isolated and artificial inoculations of T. corymbosus seeds onto cactus hosts are not successful (W. Gonzalez, personal communication). Peak fruiting periods are also different, thus reinforcing host specialization. Although we cannot completely discount an allopatric origin of Tristerix aphyllus, we feel these initial data are compelling and set the stage for further work.
Conclusions
To reconstruct the complex historical patterns and relationships among organisms, information from both ecology and phylogenetics is mutually informative. Tristerix provides an example of a genus that expanded both north and south while diversifying in response to ecological and geographical isolation coinciding with Andean orography. Although we have not established an absolute time scale for these divergences (no fossil record of Tristerix exists), many appear to be recent events as gauged by low levels of molecular divergence between various species pairs. Overall, we have shown that Tristerix is composed of northern and a southern clades. The molecular data strongly support the transfer of Tristerix verticillatus and T. penduliflorus from subgenus Metastachys to subg. Tristerix. This is despite the presence in the latter species of floral and fruit characteristics that would suggest an alliance with the northern clade. Selectional forces acting upon morphological features of the flowers and fruits stem mainly from pollinating and fruit-dispersing birds, and these forces changed as new biomes (such as matorral, cloud forest, etc.) became available since the Tertiary. Autochthonous genera, represented in our study by the outgroup Notanthera, have small, insect-pollinated flowers. This ancestral syndrome evolved into that found in bird-pollinated mistletoes such as Ligaria and Tristerix with large, tubular red flowers that are presently pollinated by hummingbirds. Members of the northern clade of Tristerix that inhabit high Andean or cloud forest biomes are also pollinated by Diglossa, that appears to forage based on corolla red-yellow banding and possibly fenestration. With the appearance of the dry matorral association in Chile, temperate-forest populations of T. corymbosus adapted to hosts in this new environment whose climate favored the development of yellow fruits at maturity. Some populations of this plesiospecies gave rise to mistletoes that could parasitize the cactus genera Echinopsis and Eulychnia. Subsequently, these populations became reproductively isolated from T. corymbosus, an isolation driven by behavior of the main seed-dispersing bird, Mimus thenca. This new apospecies, T. aphyllus, is one of the most remarkable species in Loranthaceae, given its reduced vegetative morphology and physiology that approaches holoparasitism.
FOOTNOTES
1 The authors thank M. Aizen, W. Gonzáles, M. Rodríguez Cabal, G. Servat, L. Suarez, and N. Tercero for help in the field obtaining specimens for this project; the two anonymous reviewers for making valuable comments on an earlier version of the manuscript; the Corporación Nacional Forestal (Chile) and Llao Llao Municipal Park (Argentina) for granting permits to collect these mistletoes; and S. Sipes for generously allowing use of her automated DNA sequencer. Funds for field work in Peru were provided by the Organization for Tropical Studies (OTS). The following herbaria permitted access to specimens for observation: CONC, CUZ, SI, USL, USM. Herbarium specimens used for molecular work were obtained from MO. Financial support was provided by a Ph.D. fellowship from Consejo Nacional de Investigacion Científicas y Técnicas (CONICET) and grants from the National Geographic Society (7365–02) and the National Science Foundation (to D.L.N.). 
4 Author for correspondence (e-mail: gamico{at}crub.uncoma.edu.ar
)
LITERATURE CITED
Aizen M. A.. 2003. Influences of animal pollination and seed dispersal on winter flowering in a temperate mistletoe. Ecology 84: 2613-2627..[CrossRef][ISI]
Altshuler D. L. Dudley R. McGuire J. A.. 2004. Resolution of a paradox: hummingbird flight at high elevation does not come without a cost. Proceedings of the National Academy of Sciences, USA 101: 17731-17736..
Amico G. C. Aizen M. A.. 2000. Mistletoe seed dispersal by a marsupial. Nature 408: 929-930..[Medline]
Aukema J. E.. 2003. Vectors, viscin, and Viscaceae: mistletoes as parasites, mutualists, and resources. Frontiers in Ecology and the Environment 1: 212-219..[ISI]
Barker F. K. Cibois A. Schikler P. Feinstein J. Cracraft J.. 2004. Phylogeny and diversification of the largest avian radiation. Proceedings of the National Academy of Sciences, USA 101: 11040-11045..
Barlow B. A.. 1983. Biogeography of Loranthaceae and Viscaceae. In D. M. Calder, P. Bernhardt, [eds.], The biology of mistletoes, 19-45. Academic Press, New York, New York, USA..
Baum D. A. Shaw K. L.. 1995. Geneological perspectives on the species problem. In P. C. Hoch, G. D. Stephenson, [eds.], Experimental and molecular approaches to plant biosystematics, 289-303. Missouri Botanical Garden, St. Louis, Missouri, USA..
Berlocher S. H. Feder J. L.. 2002. Sympatric speciation in phytophagous insects: moving beyond controversy. Annual Review of Entomology 47: 773-815..[CrossRef][ISI][Medline]
Bush G. L.. 1969. Sympatric host race formation and speciation in frugivorous flies of the genus Rhagoletis (Diptera, Tephritidae). Evolution 23: 237-251..[CrossRef][ISI]
Clay K. Dement D. Rejmanek M.. 1985. Experimental evidence for host races in mistletoe (Phoradendron tomentosum). American Journal of Botany 72: 1225-1231..[CrossRef][ISI]
De Queiroz K. Donahue M. J.. 1988. Phylogenetic systematics and the species problem. Cladistics 4: 317-338..[ISI]
Dieckmann U. Doebeli M.. 1999. On the origin of species by sympatric speciation. Nature 400: 354-357..[CrossRef]
Donahue M. J.. 1985. A critique of the biological species concept and recommendations for a phylogenetic alternative. Bryologist 88: 172-181..[CrossRef]
Drés M. Mallet J.. 2002. Host races in plant-feeding insects and their importance in sympatric speciation. Philosophical Transactions of the Royal Society of London 357: 471-492..[CrossRef][ISI][Medline]
Ericson P. G. P. Irestedt M. Johansson U. S.. 2003. Evolution, biogeography, and patterns of diversification in passerine birds. Journal of Avian Biology 34: 3-15..[CrossRef][ISI]
Felsenstein J.. 1981. Scepticism towards Santa Rosalia, or why are there so few kinds of animals. Evolution 35: 124-138..[CrossRef][ISI]
Felsenstein J.. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791..[CrossRef][ISI]
Fjeldsa J. Krabbe N.. 1990. Birds of the high Andes. Apollo Books, Svendborg, Denmark..
Fry J.. 2003. Multilocus models of sympatric speciation: Bush versus Rice versus Felsenstein. Evolution 57: 1735-1746..[ISI][Medline]
Futuyma D. Mayer G.. 1980. Non-allopatric speciation in animals. Systematic Zoology 29: 254-271..
Glazner J. T. Devilin B. Ellstrand N. C.. 1988. Biochemical and morphological evidence for host race evolution in desert mistletoe, Phoradendron californicum (Viscaceae). Plant Systematics and Evolution 161: 13-21..[CrossRef][ISI]
Graves G. R.. 1982. Pollination of a Tristerix mistletoe (Loranthaceae) by Diglossa (Aves, Thraupidae). Biotropica 14: 316-317..[CrossRef][ISI]
Hall T.. 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series 41: 95-98..
Hasegawa M. Kishino H. Yano T.. 1985. Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution 22: 160-174..[CrossRef][ISI][Medline]
Hinojosa L. F. Villagran C.. 1997. Historia de los bosques del sur de Sudamérica, I: antecedentes paleobotánicos, geológicos y climáticos del Terciario del cono sur de América. Revista Chilena de Historia Natural 70: 225-239..[ISI]
Hoffmann A. J. Fuentes E. R. Cortes I. Liberona F. Costa V.. 1986. Tristerix tetrandrus (Loranthaceae) and its host-plants in the Chilean matorral: patterns and mechanisms. Oecologia 69: 202-206..[CrossRef][ISI]
Hudson R. R.. 1990. Gene geneologies and the coalescent process. Oxford Surveys in Evolutionary Biology 7: 1-44..[Medline]
Huelsenbeck J. Ronquist F.. 2001. MrBayes: Bayesian inference of phylogeny, version 3.1.2. Bioinformatics 17: 754-755..
Jerome C. A. Ford B. A.. 2002. The discovery of three genetic races of the dwarf mistletoe Arceuthobium americanum (Viscaceae) provides insight into the evolution of parasitic angiosperms. Molecular Ecology 11: 387-405..[CrossRef][Medline]
Jiggins C. D.. 2006. Sympatric speciation: why the controversy?. Current Biology 16: 333-334..[CrossRef][ISI][Medline]
Johnson P. A. Hoppensteadt F. C. Smith J. J. Bush G. L.. 1996. Conditions for sympatric speciation: a diploid model incorporating habitat fidelity and non-habitat assortative mating. Evolutionary Ecology 10: 187-205..[CrossRef][ISI]
Kawecki T. J.. 1996. Sympatric speciation driven by beneficial mutations. Proceedings of the Royal Society London, B, Biological Sciences 263: 1515-1520..[CrossRef]
Kawecki T. J.. 1997. Sympatric speciation via habitat specialization driven by deleterious mutations. Evolution 51: 1751-1763..[CrossRef][ISI]
Kirkpatrick M. Ravigné V.. 2002. Speciation by natural and sexual selection: models and experiments. American Naturalist 159: S22-S35..[CrossRef][ISI]
Kizirian D. Donnelly M.. 2004. The criterion of reciprocal monophyly and classification of nested diversity at the species level. Molecular Phylogenetics and Evolution 32: 1072-1076..[CrossRef][ISI][Medline]
Kraus R. Trimborn P. Ziegler H.. 1995. Tristerix aphyllus, a holoparasitic Loranthaceae. Naturwissenschaften 82: 150-151..[CrossRef][ISI]
Kuijt J.. 1963. On the ecology and parasitism of the Costa Rican tree mistletoe, Gaiadendron punctatum (Ruiz and Pavón) G. Don. Canadian Journal of Botany 41: 927-938..
Kuijt J.. 1969. The biology of parasitic flowering plants. University of California Press, Berkeley, California, USA..
Kuijt J.. 1985. Morphology, biology and systematic relationships of Desmaria (Loranthaceae). Plant Systematics and Evolution 151: 121-130..[CrossRef][ISI]
Kuijt J.. 1988. Revision of Tristerix (Loranthaceae). Systematic Botany Monograph No. 19, American Society of Plant Taxonomists..
Maddison D. R. Maddison W. P.. 2000. MacClade: analysis of phylogeny and character evolution, version 4.0. Sinauer, Sunderland, Massachusetts, USA..
Maloof J. E. Inouye D. W.. 2000. Are nectar robbers cheaters or mutualists?. Ecology 81: 2651-2661..[CrossRef][ISI]
Martínez del Rio C. M. Hourdequin M. Silva A. Medel R.. 1995. The influence of cactus size and previous infection on bird deposition of mistletoe seeds. Australian Journal of Ecology 20: 571-576..[CrossRef][ISI]
Mauseth J. D. Montenegro G. Walckowiak A. M.. 1984. Studies on the holoparasite Tristerix aphyllus (Loranthaceae) infecting Trichocereus chilensis (Cactaceae). Canadian Journal of Botany 62: 847-857..
Mauseth J. D. Montenegro G. Walckowiak A. M.. 1985. Host infection and flower formation by the parasite Tristerix aphyllus (Loranthaceae). Canadian Journal of Botany 63: 567-581..
Medel R. Botto-Mahan C. Smith-Ramirez C. Mendez M. A. Ossa C. G. Caputo L. N. Gonzales W. L.. 2002. Historia natural cuantitativa de una relación parasito-hospedero: el sistema Tristerix-cactaceas en Chile semiarido. Revista Chilena de Historia Natural 75: 127-140..[ISI]
Medel R. Vergara E. Silva A. Kalin-Arroyo M.. 2004. Effects of vector behavior and host resistance on mistletoe aggregation. Ecology 85: 120-126..[CrossRef][ISI]
Nickrent D. L.. 1997. onward. The parasitic plant connection. Southern Illinois University. http://www.parasiticplants.siu.edu/ [February 2005]..
Nickrent D. L Stell A. L.. 1990. Electrophoretic evidence for genetic differentiation in two host races of hemlock dwarf mistletoe (Arceuthobium tsugense). Biochemical Systematics and Ecology 18: 267-280..[CrossRef][ISI]
Norton D. A. Carpenter M. A.. 1998. Mistletoes as parasites: host specificity and speciation. Trends in Ecology and Evolution 13: 101-105..[CrossRef]
Nylander J. A.. 2004. MrModelTest, version 3.1.2. Distributed by the author. Evolutionary Biology Centre, Uppsala University, Uppsala, Sweden..
Olmstead R.. 1995. Species concepts and plesiomophic species. Systematic Botany 20: 623-630..[CrossRef][ISI]
Parker T. A.. 1980. Distribution and biology of the White-cheeked Cotinga Zaratornis stresemanni, a high Andean frugivore. Bulletin of the British Ornithologists' Club 101: 156-265..
Pennings S. C. Callaway R. M.. 2002. Parasitic plants: parallels and contrasts with herbivores. Oecologia 131: 479-489..[CrossRef][ISI]
Reid N.. 1991. Coevolution of mistletoes and frugivorous birds?. Australian Journal of Ecology 16: 457-469..[CrossRef][ISI]
Restrepo C. Sargent S. Levey D. J. Watson D. M.. 2002. The role of vertebrates in the diversification of New World mistletoes. In D. J. Levey, W. R. Silva, M. Galetti, [eds.], Seed dispersal and frugivory: ecology, evolution and conservation, 83-98. CAB International, Wallingford, UK..
Richardson J. E. Pennigton R. T. Pennigton T. D. Hollingsworth D. P. M.. 2001. Rapid diversification of a species-rich genus of neotropical rain forest trees. Science 293: 2242-2245..[CrossRef][ISI][Medline]
Rieseberg L. H. Brouillet L.. 1994. Are many plant species paraphyletic?. Taxon 43: 21-32..[CrossRef][ISI]
Rieseberg L. H. Wendel J. F.. 2004. Plant speciation—rise of the poor cousins. New Phytologist 161: 1-21..[CrossRef][ISI]
Rieseberg L. H. Wood T. E. Baack E. J.. 2006. The nature of plant species. Nature (London) 440: 524-527..[CrossRef][Medline]
Rivera G. L. Galetto L. Bernardello L.. 1996. Nectar secretion pattern, removal effects, and breeding system of Ligaria cuneifolia (Loranthaceae). Canadian Journal of Botany 74: 1996-2001..
Savolainen V. Anstett M.-C. Lexer C. Hutton I. Clarkson J. J. Norup M. V. Powell M. P. Springate D. Salamin N. Baker W. J.. 2006. Sympatric speciation in palms on an oceanic island. Nature 441: 210-213..[CrossRef][Medline]
Schluter D.. 2005. Ecology and the origin of species. Trends in Ecology and Evolution 16: 372-380..[CrossRef]
Simpson B. B.. 1975. Pleistocene changes in the flora of the high tropical Andes. Paleobiology 1: 273-294..[Abstract]
Swofford D. L.. 2002. PAUP*: phylogenetic analysis using parsimony (* and other methods), version 4.0b10. Sinauer, Sunderland, Massachusetts, USA..
Taberlet P. Gielly L. Pautou G. Bouvet J.. 1991. Universal primers for amplification of three noncoding regions of chloroplast DNA. Plant Molecular Biology 17: 1105-1109..[CrossRef][ISI][Medline]
Tavaré S.. 1986. Some probabilistic and statistical problems on the analysis of DNA sequences. Lectures on Mathematics in the Life Sciences 17: 57-86..
Taylor D. W.. 1991. Paleobiogeographic relationships of Andean angiosperms of Cretaceous to Pliocene age. Palaeogeography Palaeoclimatology Palaeoecology 88: 69-84..[CrossRef]
Tel-Zur N. Abbo S. Myslabodski D. Mizrahi Y.. 1999. Modified CTAB procedure for DNA isolation from epiphytic cacti of the genera Hylocereus and Selenicereus (Cactaceae). Plant Molecular Biology Reporter 17: 249-254..[CrossRef][ISI]
Watson D. M.. 2001. Mistletoe—a keystone resource in forests and woodlands worldwide. Annual Review of Ecology and Systematics 32: 219-249..[CrossRef][ISI]
This article has been cited by other articles:
|
|
 |
|
  R. Vidal-Russell and D. L. Nickrent Evolutionary relationships in the showy mistletoe family (Loranthaceae) Am. J. Botany, August 1, 2008; 95(8): 1015 - 1029. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||
