HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Supplemental Data Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (46) Citing Articles via Google Scholar Google Scholar Articles by Beardsley, P. M. Articles by Olmstead, R. G. Search for Related Content PubMed Articles by Beardsley, P. M. Articles by Olmstead, R. G. Agricola Articles by Beardsley, P. M. Articles by Olmstead, R. G.

Redefining Phrymaceae: the placement of Mimulus, tribe Mimuleae, and Phryma¹

Department of Botany, Box 355325, University of Washington, Seattle, Washington 98195 USA

Received for publication August 3, 2001. Accepted for publication February 19, 2002.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Chloroplast trnL/F and nuclear ribosomal ITS and ETS sequence data were used to analyze phylogenetic relationships among members of tribe Mimuleae (Scrophulariaceae) and other closely related families in Lamiales. The results of these analyses led to the following conclusions. (1) The Australian genera Glossostigma and Peplidium and the taxonomically isolated Phryma join four genera of tribe Mimuleae to form a well-supported clade that is distinct from other families in the Lamiales. We refer to that clade as the subfamily Phrymoideae. (2) The genera Mazus and Lancea (tribe Mimuleae) together form a well-supported clade that we recognize as the subfamily Mazoideae. Mazoideae is weakly supported as sister to Phrymoideae. We assign Mazoideae and Phrymoideae to a redefined family Phrymaceae. (3) Mimulus is not monophyletic, because members of at least six other genera have been derived from within it. In light of the molecular evidence, it is clear that species of Phrymaceae (about 190 species) have undergone two geographically distinct radiations; one in western North America (about 130 species) and another in Australia (about 30 species). Phylogenetic interpretations of morphological evolution and biogeographical patterns are discussed.

Key Words: ETS • ITS • Mimuleae • Mimulus • Phryma • Phrymaceae • Scrophulariaceae • trnL/F

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Species in the genus Mimulus have become model systems for the study of evolutionary processes in nature. The most intensely studied species are part of the radiation of Mimulus in western North America. North American Mimulus was established for the study of evolution by the classic work of scientists at the Carnegie Institution. They investigated biosystematics, genetics, and physiological ecology in sect. Erythranthe (Hiesey, Nobs, and Bjorkman, 1971 ). Research on Mimulus has continued with studies of inbreeding depression (Carr and Dudash, 1996 ; Carr, Fenster, and Dudash, 1997 ; Karron et al., 1997 ; Dudash and Carr, 1998 ; Willis, 1999 ) the genetics of speciation (Bradshaw et al., 1995 , 1998 ), mating system evolution (Leclerc-Potvin and Ritland, 1994 ), and the ecological effects of hybridization (Beeks, 1962 ; Waayers, 1996 ).

As currently described, Mimulus contains approximately 120 species, of which some 75% occur only in western North America. However, Mimulus is worldwide in distribution. Four species exist in Australia (Grant, 1924 ; Barker, 1982 ) (one of them also occurs in South Africa and India), 10 in Chile (von Bohlen, 1995a ), approximately 19 in Mexico (Grant, 1924 ; Vickery, 1997 ), 4 in the Himalayas (Yamazaki, 1993 ), 1 in Madagascar, and 2 in eastern North America. The relationships among western North American Mimulus, Mimulus species distributed outside western North America, and several genera putatively closely related to Mimulus remain uncertain.

Dumortier (1829) first proposed tribe Mimuleae (Scrophulariaceae). Von Wettstein (1891) placed Mimulus with Mazus, Dodartia, Monttea, Melosperma, and Lancea in subtribe Mimulinae of tribe Gratioleae. Leucocarpus, Hemichaena, and Berendtiella were placed in tribe Scrophularieae. Pennell (1920) redefined tribe Mimuleae to contain Leucocarpus and Mimulus, though in subsequent works, Pennell placed Mimuleae in tribe Gratioleae (Pennell, 1935 , 1947 ). Pennell (1935) also transferred Hemichaena and Berendtiella to tribe Gratioleae, noting a similarity in floral structure to Mimulus. Burtt (1965) reestablished Mimuleae as a tribe, diagnosed on the presence of two characters: (1) tubular, toothed calyces and (2) bilamellate stigmas that are receptive only on the inner surface and that close together with contact. Thieret (1954 , 1967) suggested the removal of Melosperma and Monttea to Melospermeae. This left tribe Mimuleae sensu Argue (1984) composed of Mimulus (125 species), Berendtiella (5 species), Hemichaena (1 species), Leucocarpus (1 species), Dodartia (1 species), Lancea (2 species), and Mazus (25 species). The relationships among Mimulus and other members of Mimuleae are poorly understood. In his analysis of pollen, Argue (1980, 1984) found no character or combination of characters that separates Dodartia, Lancea, Leucocarpus, Mazus, or Mimulus.

Barker (1982) considered the Australian genera Glossostigma, Peplidium, and Elacholoma (subtribe Limoselleae of tribe Gratioleae, sensu Bentham [1876] ) to be members of Mimuleae. Similar to the traditional members of Mimuleae, these Australian genera possess tubular, toothed calyces, though in Glossostigma the calyx has three or four lobes instead of the typical five lobes. The stigmas of these Australian genera are structurally different from those of species traditionally assigned to Mimuleae. In the monotypic Elacholoma, the stigma lobes are relatively long and are receptive over most of their length (Barker, 1982 ). The stigmas of Glossostigma and Peplidium contain one large and one vestigial stigma lobe. The large lobe covers the mouth of the corolla and is receptive only on the outer surface (Barker, 1982 ). Upon being touched, this large stigma lobe moves from covering the mouth to pressing against the upper corolla lip, exposing the anthers. In order to assess the homology of these stigmatic movements to the touch-sensitive stigmatic movements in traditional Mimuleae, a sound phylogenetic hypothesis is needed.

Though much has been learned about processes of evolution in Mimulus, the systematic placement of the genus and the relationships among species within it remain unresolved. Molecular studies have demonstrated that the traditionally recognized Scrophulariaceae are not monophyletic (Olmstead and Reeves, 1995 ; Olmstead et al., 2001 ). In their analysis of Scrophulariaceae using three chloroplast genes, Olmstead et al. (2001) found that the single representative of Mimulus formed the sister group to Orobanchaceae and Paulownia in some trees and to Lamiaceae in others. Additional taxa were not sampled within Mimulus or tribe Mimuleae. Olmstead et al. (2001) dismembered the traditional Scrophulariaceae but did not reassign Mimulus to any clade of family rank.

An additional intriguing result of molecular analyses of Lamiales is the close relationship between Mimulus and the taxonomically isolated Phryma (Phrymaceae) (R. G. Olmstead, unpublished manuscript). Phryma is monotypic with a disjunct distribution in eastern Asia and eastern North America. The taxonomic position of Phryma has been unsettled for years: some authors place it in Verbenaceae (e.g., Cronquist, 1981 ) and others in its own family Phrymaceae (e.g., Schauer, 1847 ; Moldenke, 1971 ; Lu, 1990 ). Phryma has a distinctive pseudomonomerous gynoecium, which develops into a one-seeded fruit (Chadwell, Wagstaff, and Cantino, 1992 ). A close relationship between Phryma and Mimulus has never been suggested.

Relationships within Mimulus have also been a source of taxonomic controversy. Grant (1924) divided the approximately 120 species into two subgenera and ten sections. Subgenus Mimulus (Synplacus sensu Grant) is based on the character of the placentae being firmly united to form a central column or separating only at the apex. Taxa within subgenus Schizoplacus possess placentae that are divided to the base. The subgeneric assignments made by Grant have not been challenged, although differing views on relationships among and within sections have been proposed and some sections have been elevated to generic rank (McMinn, 1951 ; Pennell, 1951 ). Vickery (1969) recognized seven major sections in Mimulus (Mimulus, Erythranthe, Simiolus, Paradanthus, Eunanus, Oenoe and Diplacus) and five monotypic sections. Thompson's (1993) treatment of Mimulus in California follows Vickery, but reassigns M. mohavensis (monotypic sect. Mimulastrum) and M. pictus (monotypic sect. Pseudoenoe) to sect. Mimulastrum and places M. pygmaeus (monotypic sect. Microphyton) in sect. Oenoe. The monophyly of some of the sections is suspect, because other sections have been suggested to be derived from within them (Grant, 1924 ; D. Thompson, personal communication).

The goals of this study are to: (1) develop a rigorous phylogenetic hypothesis for the higher level relationships of Mimulus and other members of the tribe Mimuleae, (2) examine the relationship of Mimulus to putatively closely related genera, (3) test the hypothesis of monophyly of the genus Mimulus, (4) analyze subgeneric and sectional relationships within Mimulus, and (5) assess biogeographical relationships and morphological character changes in a phylogenetic context.

We report the results of analyses of DNA sequences from both the chloroplast and nuclear genome. Data from the chloroplast genome come from the leucine (trnL) intron and the intergenic spacer between trnL and trnF (trnL/F) (Gilley and Taberlet, 1994 ). This region has been shown in previous studies to provide good phylogenetic resolution within families and genera in the Lamiales (McDade and Moody, 1999 ). The nuclear genome is represented by sequences of both the internal transcribed spacer (ITS) and external transcribed spacer (ETS) regions (Baldwin et al., 1995 ; Baldwin and Markos, 1998 ) of nuclear rDNA. We collected nuclear genome data for two reasons: (1) to compare phylogenetic hypotheses from different genomes and (2) to help resolve more distal portions of the phylogeny. Our expectation in starting this project was that the substitution rate of trnL/F, ETS, and ITS would overlap, but overall the trnL/F region would have more slowly evolving sites. This information would be useful for resolving the deeper nodes in the phylogeny. The ETS and ITS would show higher overall substitution rates that would aid in resolving the distal portions of the phylogeny.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Taxon sampling
All genera within the traditionally described tribe Mimuleae were sampled except Dodartia, a monotypic genus from Russia, and the monotypic Australian genus Elacholoma (Table 1). The Australian genera Glossostigma and Peplidium were sampled, as was Phryma from eastern North America and China. Within Mimulus, we included at least two representatives of each of the seven major sections. An attempt was also made to sample broadly across the geographical distribution of Mimulus. Species from western North America, eastern North America, Chile, Australia, and China were included. Molecular systematic analyses of nearly every species of Mimulus currently underway (P. Beardsley, unpublished data) indicate that our sample of species within Mimulus results in the same pattern of relationships as denser sampling. Also represented are members of the clades identified by Olmstead et al. (2001) as most closely related to Mimulus, including Paulowniaceae, Lamiaceae, Orobanchaceae, Pedaliaceae, and Acanthaceae. Some of the trnL/F and ITS sequences for the preceding families were obtained from GenBank. Species used in this study, voucher specimens, and GenBank accession numbers are listed on the Botanical Society of America website (http://ajbsupp.botany.org/v89/). A representative of Veronicaceae, Mohavea breviflora, was used as an outgroup in all analyses, a choice that is supported by previous molecular analyses of Lamiales (Olmstead et al., 2001 ).

The trnL/F region was amplified using the trn-c and trn-f primers (Taberlet et al., 1991 ). We had difficulty amplifying both the intron and spacer as one fragment for three species (Mimulus uvedaliae, Lancea tibetica, and Berendtiella rugosa). For these taxa, the intron and spacer were amplified separately, using the trn-c and trn-d primers to amplify the trnL/F intron and trn-e and trn-f to amplify the spacer. The entire ITS region was amplified using the its4 and its5 primers (Baldwin, 1992 ). In Hemichaena fruiticosa and Mimulus gracilis, we had difficulty amplifying the entire fragment so the ITS1 and ITS2 were amplified separately, using the its5 and its2 primers for ITS1 and the its3 and its4 primers for ITS2. To amplify a portion of the 3' end of the ETS, we used the 3' 18S-IGS primer of Baldwin and Markos (1998) . The 5' primer, named ETS-B (5'-ATAGAGCGCGTGAGTGGTG-3') (A. Yen, University of Washington, unpublished manuscript) was designed using Mimulus sequences as a reference. The polymerase chain reaction (PCR) conditions for all three DNA regions were as follows: 35 cycles of 94°C for 1 min, 50°C for 1 min, and 72°C for 1 min. The PCR products were purified using Qiagen Qiaquick spin-columns according to the manufacturer's protocol.

Sequences of both strands of the PCR product were generated on an ABI 377 (Applied Biosystems, Foster City, California, USA). To improve sequence quality for the trnL/F region of all sampled taxa, the internal primers of Taberlet et al. (1991) , trn-d and trn-e, were used for sequencing as were two primers that we designed using Mimulus sequences as a reference, trnL-2C (5'-ATCGGTAGACGCTACGGACT-3') and trnL-2F (5'-CGGGATAGCTCAGCTGGTAG-3') that are just internal to trn-c and trn-f, respectively. The ITS was sequenced using the external PCR primers, its4 and its5, and the two internal primers, its2 and its3. The ETS was sequenced using the 18S-E primer of Baldwin and Markos (1998) , which is slightly internal to 18S-IGS primer, and the Mimulus specific ETS-B primer. Electropherograms for each region were compiled and compared using the program Sequencher version 3.0 (Gene Codes Corporation, Ann Arbor, Michigan, USA), from which a consensus sequence was generated.

Analyses
Consensus sequences for trnL/F, ITS, and ETS of all taxa were aligned manually using the program Se-Al version 1 (A. Rambaut, University of Oxford, Oxford, United Kingdom). Some regions were too different to be confidently aligned for ITS and ETS sequences, but would be valuable for analyses within subgroups. Regions that were difficult to align with confidence were excluded. Alignments are available at the Botanical Society of America website (http://ajbsupp.botany.org/v89/).

The three DNA regions were analyzed both individually and in combination using the program PAUP* 4.0b3a (Swofford, 1998 ). The default PAUP settings were used except as noted. Parsimony searches were conducted using heuristic searches with 1000 random sequence addition replicates, to find multiple islands of trees, if present (Maddison, 1991 ). Gaps were scored as missing data. Support for individual branches was estimated using bootstrap values (Felsenstein, 1985 ). Bootstrap values were calculated using 1000 full heuristic search replicates.

Maximum likelihood estimates of the phylogeny were also made using PAUP* for the nrDNA, trnL/F, and combined data sets. In each case, one of the most parsimonious trees was selected and PAUP* was used to estimate the transition : transversion ratio (ti : tv) (with two rate categories) and the shape parameter for the gamma distribution. Base frequencies were determined using the empirical values. The estimated values were entered as fixed data and were used in a heuristic search. Using these parameters corresponds to the HKY85 model of evolution.

The incongruence length difference (ILD) test (Farris et al., 1994 , as implemented in PAUP*) was used to assess potential conflicts between the phylogenetic signal from different DNA fragments. The following comparisons were made: ITS vs. ETS and trnL/F vs. combined ITS and ETS. For each test, 100 replicates were analyzed with an heuristic search, each with ten random sequence addition replicates.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
The alignment of the trnL/F sequences was straightforward, though numerous short gaps were introduced. We had difficulty amplifying trnL/F from Glossostigma and were only able to amplify the intron for Berendtiella. The total aligned length of the trnL/F intron and spacer is 1024 base pairs (bp). The trnL/F alignment had 339 variable and 181 parsimony-informative sites. Analyses using maximum parsimony (MP) as the optimality criterion resulted in 700 most-parsimonious trees of 521 steps found on six different islands (consistency index [CI] = 0.785, retention index [RI] = 0.823, rescaled consistency index [RC] = 0.646). For maximum likelihood (ML) analyses of trnL/F sequence data, the following parameters were estimated on one of the most parsimonious trees: ti : tv ratio = 1.035, gamma = 0.863. The most likely tree given our data set was one of the 700 MP trees (Fig. 1).

View larger version (37K): [in this window] [in a new window]   Fig. 1. (A) The maximum likelihood (ML) tree resulting from the analysis of trnL/F data; this tree is also one of the 700 most-parsimonious (MP) trees. An asterisk indicates nodes that collapse in the strict consensus of MP trees. (B) The ML tree resulting from the analysis of nrDNA ITS and ETS sequence data. An X on this tree indicates nodes that differ from the MP tree. In both trees, numbers above the branches indicate bootstrap values estimated using parsimony

Compared to ITS, the approximately 450 bp of the 3' end of the ETS that we analyzed was relatively easy to amplify and sequence. Complete sequences for both strands required only two sequencing reactions. The aligned ETS sequences were 491 bp in length. Pairwise distances among all sampled taxa ranged from 0.7% to 40.1% for ETS compared to a range of 0% to 27.9% for ITS. The GC content of ETS (56.2%) was similar to that of ITS (59.2%). Seven regions totaling 123 bp could not be aligned confidently and were excluded from the analysis (bases 93–104, 159–200, 263–284, 368–373, 390–398, 439–451, and 470–488 in the alignment).

The ETS and ITS region are closely linked in the rDNA, therefore, we will only present results for analyses of combined rDNA data. Results from the partition homogeneity test (see below) justify our decision. The combined nrDNA data set was 1187 bases in aligned length. After excluding ambiguously aligned sites, 874 bases were used in the analysis, of which 540 were variable and 365 were parsimony informative. Parsimony analyses resulted in one most-parsimonious tree of length 1586 (CI = 0.541, RI = 0.656, RC = 0.355). For the ML analysis of nrDNA sequence data (Fig. 1), the following parameters were estimated on one of the most-parsimonious trees: ti : tv ratio = 2.037, gamma = 0.634. The MP tree differed from the ML tree in the resolution of the deeper nodes and the ML tree resolves M. floribundus as sister to M. nepalensis while the MP tree resolves M. nepalensis as sister to a clade containing M. floribundus, M. guttatus, M. tilingii, and M. depressus.

Results of the partition homogeneity test for ITS vs. ETS and ITS/ETS vs. trnL/F showed that none of the data sets were significantly different from random pairwise partitions of the data (P = 0.14, P = 0.67, respectively). Therefore, we combined all three data sets in subsequent analyses.

The combined chloroplast and nrDNA data set was 2211 bases in aligned length. A total of 313 sites were difficult to align and were excluded, resulting in 1898 included characters, of which 879 were variable and 546 were parsimony informative. Parsimony analyses of the combined data resulted in three most-parsimonious trees of length 2117 (CI = 0.599, RI = 0.691, RC = 0.413). For the ML analysis of the combined data, the following parameters were estimated on one of the most parsimonious trees: ti : tv ratio = 1.164, gamma = 0.471. The ML tree is congruent with the strict consensus of the three MP trees and has a parsimony length of 2118.

Figure 1 compares the ML trees resulting from the analysis of chloroplast DNA and nrDNA. The phylogeny estimated using trnL/F shows good resolution of deep nodes among genera of Mimuleae but poor resolution among the families in the Lamiales. The phylogeny estimated using nrDNA resolves some of the deeper nodes in the phylogeny differently, though none are well supported. Analyses of the combined data (Fig. 2) resulted in a topology that is congruent with the phylogeny estimated using only chloroplast data, though more resolved, for the relationships of species of Mimulus, Phryma, Leucocarpus, Hemichaena, Berendtiella, and Peplidium with the exception of the placement of M. pulchellus. Relationships among and within these genera are identically resolved in the combined ML tree and one of the combined MP trees. The relationships of families in the Lamiales to each other are generally unresolved in all analyses. The major clades identified in our analyses are detailed below.

View larger version (35K): [in this window] [in a new window]   Fig. 2. Maximum likelihood tree inferred using combined data. Numbers above the branches indicate bootstrap values based on parsimony analyses. An asterisk indicates the nodes that collapse in the strict consensus of the three most parsimonious trees. Sampled genera and sections within Mimulus are listed to the right of the tree. Capital letters denote groups mentioned in the text

Within Phrymoideae, relationships among all sampled taxa are fully resolved in the combined analysis, with the exception of one node in the MP analyses, and most clades are well supported. Of the 20 resolved nodes in the MP analyses, only one has a bootstrap value <76%, and only three have bootstrap values <92%. Clade D (BS = 100% all analyses) contains members of Mimulus sect. Mimulus (sect. Eumimulus sensu Grant [1924] ), including the type species, M. ringens, and the Australian genera Glossostigma and Peplidium. Derived from within clade D is clade E (BS = 96%), containing Glossostigma, Peplidium, and M. uvedaliae, which also is Australian in distribution. A sister relationship between sampled members of Glossostigma (which was not sampled for trnL/F) and Peplidium was recovered in analysis of nrDNA (BS = 90%).

All Mimulus in western North American plus the genera Berendtiella, Hemichaena, and Leucocarpus form clade F (BS = 63%). Phryma from eastern North America and from eastern China form a clade (BS = 100%). Phryma is sister to clade F in the ML analysis of combined data and in two of the three MP trees. In one MP tree, Phryma is sister to the entire Phrymoideae. Clade G (BS = 93%) contains Leucocarpus and its sister clade H (BS = 100%), which contains Mimulus sects. Simiolus, Erythranthe, and Paradanthus. Within clade H are clades that correspond to sect. Erythranthe (M. cardinalis and M. lewisii [BS = 93%]) and sect. Simiolus (M. guttatus, M. tillingii, and M. depressus [BS = 100%]). Within Simiolus, the two species from western North America form a clade (BS = 92%) that is sister to the species from Chile (M. depressus). Section Paradanthus is paraphyletic.

Berendtiella and Hemichaena, together with all members of Mimulus subgenus Schizoplacus form clade I (BS = 92%). All sampled species in Mimulus subg. Schizoplacus are monophyletic (BS = 100%) and together comprise clade J that is sister to a clade comprised of Hemichaena and Berendtiella (BS = 99%). Within clade J are clades that correspond to sect. Eunanus (M. whitneyi and M. brevipes [BS = 100%]) and sect. Diplacus (M. aurantiacus and M. clevelandii [BS = 99%]). The latter plus two species of Mimulus sect. Oenoe form a clade (BS = 92%). Diplacus appears to be derived from within a paraphyletic Oenoe.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Estimates of phylogeny from chloroplast and nuclear data
Data from the chloroplast and the nucleus are largely congruent with respect to phylogenetic inference and, in combination, provide a powerful data set for resolving relationships within the group. The nrDNA and cpDNA data are complementary in that the nrDNA provides better resolution near the tips of the tree whereas the cpDNA resulted in more strongly supported nodes at deeper levels within the Phrymoideae. This result was expected based on the results of ITS analyses of other closely related families in Lamiales (McDade et al., 2000 ). Neither source of data provided resolution of the relationships of Phrymaceae to other families in the Lamiales. The chloroplast data were less homoplastic than the nuclear data (trnL/F CI = 0.776 vs. nrDNA CI = 0.518) for analyses of all taxa. The combined analysis reflects the strongly supported components of the individual data sets and results in a well-supported estimate of phylogeny in Mimulus and related genera in tribe Mimuleae from which significant systematic and evolutionary conclusions can be drawn.

Systematic conclusions
(1) Mimulus and the genera derived from within it form a strongly supported clade referred to as subfamily Phrymoideae (Fig. 2). This is not nested in any clade traditionally considered to have the rank of family in this study. Similarly, the recent study of Olmstead et al. (2001) found that Mimulus was not associated with any of the named clades of Lamiales. Phrymoideae has a wide geographic range and contains approximately 160 species that are divergent morphologically and ecologically.

(2) There is weak support for a more inclusive clade, Phrymaceae, in which Phrymoideae and Mazoideae are sister groups. This relationship is recovered in the ML analysis of the combined data, in two of the three MP trees, and in 39% of the parsimony bootstrap data sets. Given that Mazus and Lancea share a set of morphological characters with species in Phrymoideae, we tentatively assign these genera to Phrymaceae with the caveat that further research is needed to confirm their sister relationship to Phrymoideae. Therefore, we propose that the clade recovered in this analysis that contains all the representatives of Mimulus, as well as the genera Phryma, Glossostigma, Peplidium, Leucocarpus, Berendtiella, and Hemichaena, Mazus, and Lancea be recognized at the rank of family; Phrymaceae (Schauer, 1847 ) has priority. A distinct advantage in recognizing Lancea and Mazus within the Phrymaceae is that synapomorphies can be diagnosed for Phrymaceae, as will be discussed below in the section on morphological evolution. The scope of this newly defined Phrymaceae is very different from its historical antecedents. Previously, the family was monotypic and limited in geographic range to eastern North America and eastern China. In our circumscription, Phrymaceae encompasses approximately 190 species that are distributed worldwide.

(3) Relationships of Phrymaceae to other families in the Lamiales are generally unresolved. A similar lack of resolution has been encountered in other molecular analyses of the Lamiales (Olmstead et al., 2001 ; B. Oxelman, Uppsala University, unpublished manuscript). Sampling in this study was designed to include groups putatively closely related to Mimulus and tribe Mimuleae, based on other molecular studies of Lamiales. Given a limited sampling of these groups, Paulownia has the strongest support (albeit weak; BS = 37%) as sister to Phrymaceae. This relationship is recovered in the ML analysis and in two of the three MP trees. The third MP tree resolves Paulownia as sister to Phrymoideae and they, in turn, are sister to Mazoideae. If this latter relationship turns out to be correct, Paulownia would simply be included within Phrymaceae. At present, we assign Paulownia to the monogeneric family Paulowniaceae.

(4) The genus Mimulus is not monophyletic. Glossostigma, Peplidium, Phryma, Leucocarpus, Hemichaena, and Berendtiella are derived from within Mimulus. Though it was not sampled, morphological evidence strongly implies that the Australian genus Elacholoma is also derived from within Mimulus (Barker, 1982 ). The monophyly of Mimulus was not found in analyses of any of the individual data sets or in any combination of the data sets. In the combined data set, trees that are constrained to make Mimulus monophyletic are 94 steps longer. The nonmonophyly of Mimulus is not surprising, given that no clear synapomorphy for the group has been identified. In her monograph of Mimulus, Grant (1924) suggested that calyx characters best diagnosed the genus. Calyx characters within Mimulus, however, cannot be described in any way that would consistently differentiate Mimulus from other genera in tribe Mimuleae.

(5) It is clear that generic boundaries within Phrymaceae require extensive redefinition in order for genera to be monophyletic. Options for generic revisions include: (i) retain the traditional Mimulus sensu lato (s.l.) as one genus and reassign species of the six other genera to Mimulus or (ii) break up the traditional genus Mimulus. The advantages of option i are that fewer species in Phrymaceae would require a name change and all of the intensely studied species in western North America would retain the name with which they are currently identified. In this scenario, Phrymoideae would be synonymous with Mimulus. The primary disadvantages of option i are that at least six genera would require name changes and that all the diversity contained in Phrymoideae would be assigned to one genus. The latter situation is only a problem, however, if one assumes that different ranks within a hierarchical classification have implications with regards to amounts of diversity. Option ii would result in more genera in Phrymaceae than option i but would require 105 to 185 name changes, depending on where generic boundaries are drawn. Redrawing generic boundaries within Phrymaceae and in-depth discussions of the implications of using traditional and phylogenetic nomenclatural systems (deQueiroz and Gauthier, 1992 ; Cantino et al., 1999 ) are beyond the scope of this paper, but will be the subject of a future publication.

(6) Some subgeneric and sectional relationships are apparent. Subgenus Schizoplacus is shown to be monophyletic (BS = 100%) whereas subgenus Mimulus is not. Preliminary evidence exists for the monophyly of sects. Simiolus (BS = 100%), Erythranthe (BS = 93%), Diplacus (BS = 99%), and Eunanus (BS = 100%). Section Mimulus is not monophyletic, because Glossostigma and Peplidium are derived from within it, nor is section Paradanthus. In defining sect. Paradanthus, Grant (1924) admitted that the species within it were probably not closely related. The nonmonophyly of this section was also postulated from pollen data (Argue, 1980 ). Mimulus bicolor (sect. Paradanthus) is closely related to species in Erythranthe, whereas other species in Paradanthus are more closely related to sect. Simiolus. Section Oenoe may also not be monophyletic. More thorough analyses of relationships within Mimulus in western North America and Australia are in preparation.

Morphological evolution
Species within the redefined Phrymaceae are diverse in both life history and morphological traits. This clade includes species with habits ranging from annuals, attaining a final plant height of a few centimeters, to perennials, some of which are woody and attain a height of 4 m. Variation for reproductive mechanisms also exists, including species that have outcrossing, mixed-mating, selfing, and asexual breeding systems. At least five species in Phrymaceae, M. congdonii, M. douglasii (Thompson, 1993 ), M. nasutus (Diaz and Macnair, 1998 ), M. pictus (Thompson, 1997 ), and Glossostigma cleistanthum (Barker, 1982 ), have some populations that are cleistogamous. Within Phrymaceae, insects pollinate flowers of most of the species, but hummingbird pollination also exists (M. cardinalis [Bradshaw et al., 1998 ] and M. flemingii [Stebbins, 1989 ]). Many species have corollas with bilateral symmetry, but flowers of other species are radially symmetric. Barker (1982) notes that radially symmetrical corollas seem to be associated with a prostrate habit in Australian species of Mimulus and Peplidium. The most common fruit type in Phrymaceae is a readily dehiscent capsule containing numerous seeds, but exceptions exist. Phryma leptostachya has an achene (Whipple, 1972 ). Leucocarpus perfoliatus has a baccate, indehiscent fruit (Burtt, 1965 ) described as a white berry, with thin skin and with most of the substance of the fruit derived from the fleshy placenta. Many species of Peplidium possess capsules that are thick-walled and open only after the plant has senesced. Grant (1924) described a similar capsule in some species of Mimulus sect. Oenoe. The phylogenetic distance between Mimulus sect. Oenoe and Peplidium (Fig. 2) demonstrates that this character evolved convergently in taxa from two different continents, possibly in response to common environmental conditions. Species in Phrymaceae also show variation in habitats, inhabiting sites as varied as desert, riparian, and alpine environments. Some species extend over a large geographic range and have developed many different races (e.g., M. glabratus) whereas others are local endemics and several are rare (e.g., M. exiguus).

It is of great practical importance to provide morphological synapomorphies for described clades. Synapomorphies for some of the recovered clades are described below, along with evidence for or against previous systematic and morphological hypotheses. A summary of the major morphological shifts in the family is presented in Fig. 3.

View larger version (60K): [in this window] [in a new window]   Fig. 3. Inferred morphological changes in Phrymaceae using the ML estimate of the phylogeny from the combined data

Mimulus subgenus Schizoplacus
Grant (1924) diagnosed subgenus Schizoplacus primarily on the basis of having a placenta that is divided to the base. Other traits associated with Schizoplacus were relatively short pedicels (usually shorter than the calyx) and glandular pubescent styles. Sister taxa Hemichaena and Berendtiella have united placentae thus making divided placentae a synapomorphy for Schizoplacus. Berendtiella and Hemichaena have pedicels that are shorter than the calyx and glandular pubescent styles, making these characters synapomorphies for the more inclusive clade.

Mimulus section Mimulus
The characters that Grant (1924) used to diagnose sect. Mimulus (Eumimulus sensu Grant [1924] ) were tubular calyces with equal teeth and a distinctly bilabiate, blue corolla with glabrous palatine ridges. Presumably, the description of tubular calyces was meant to distinguish the calyces in sect. Mimulus from those in sect. Simiolus, which are strongly inflated and sagitally compressed, and sects. Paradanthus and Erythranthe, which have sharp, definite angles and flat sides. However, according to our results, these characters are not consistent within the clade containing sect. Mimulus. The Australian genera Glossostigma and Peplidium are clearly derived from within sect. Mimulus and do not have equal calyx teeth, many have radially symmetric flowers, many others have flowers that are not blue, and some have sharply angled calyx lobes. Overall, the distribution of morphological characters in this clade needs to be reexamined in light of the phylogenetic evidence. At present, no clear synapomorphies exist for the clade containing Mimulus sect. Mimulus.

Glossostigma/Peplidium
Barker (1982) described two derived morphological characters for these genera: one-celled anthers and the presence of one large and one vestigial stigmatic lobe. In Glossostigma and Peplidium, the large stigmatic flap covers the corolla mouth and, upon being touched, is triggered back against the upper corolla lip. This stigmatic movement is apparently homologous to and derived from the stigmatic condition in the rest of the Phrymaceae in which two equal or nearly equal stigma flaps, receptive on the inner surface only, close upon touch (Fetscher and Kohn, 1999 ).

Phyrma
The monotypic genus Phryma has been difficult to place taxonomically due to its having a pseudomonomerous gynoecium that develops into an achene. The results presented here indicate that Phryma is derived from within the traditionally recognized genus Mimulus. In comparison to other species in Phrymaceae, Phryma shares some floral characteristics, but many of its reproductive characters are derived. All other taxa in the Phrymaceae are bicarpellate and have numerous seeds, thus the enigmatic gynoecium and fruit of Phryma is derived. Whipple (1972) described the calyx of Phryma as persistent, zygomorphic, and as having five ridges with three hooked adaxial lobes and two short subulate abaxial lobes. The three elongated and hooked adaxial corolla lobes can be interpreted as being derived from the ancestral condition of relatively short, uncurved adaxial lobes. It is possible that this morphological switch is an adaptation to epizoochorous dispersal, in that these hooked upper lobes allow the fruit to become attached to animal fur (Holm, 1913 ). The flowers of Phryma have two stigmatic surfaces at the tip of the style, but, in general, flowers are much reduced, which makes determination of stigmatic characters difficult. Phryma is also characterized by the reflexed movement of the mature calyx and fruit (Whipple, 1972 ). While in flower, the calyx is perpendicular to the stem, while in fruit the calyx points abaxially and is parallel to the stem. Its common name "lopseed" reflects this fruit orientation.

Diplacus
The following characters have been used to diagnose Diplacus: (1) plants are shrubs, semi-shrubs, or perennial from a woody caudex; (2) glutinous exudation from the leaves; (3) prismatic calyx; and (4) investing, linear, oblong capsule (Grant, 1924 ). This combination of characters separates Diplacus from other groups in Phrymaceae.

Hemichaena-Berendtiella
In an analysis of morphological characteristics of Hemichaena, Berendtiella, and Leucocarpus, Thieret (1972) suggested that Hemichaena and Berendtiella could be distinguished from other closely related genera by their inflorescences of bracteolate cymes. Species in this clade are woody, as are members of the Diplacus clade, suggesting the convergent evolution of wood in these two groups (Fig. 3). Additionally, species in both Diplacus and the Hemichaena-Berendtiella clade have leaves that are revolute at their margins, a character that is possibly a convergent adaptation to the dry habitats in which plants of both groups live. Thieret (1972) postulated that Leucocarpus and Hemichaena were closely related due to their possession of reticulate seeds and distinctive calyces and stigmas. The estimated phylogeny (Fig. 2), however, shows that Leucocarpus and Hemichaena are not closely related. The calyx and stigma characters are sympleisiomorphies.

A final observation made by Thieret (1972) on the Hemichaena-Berendtiella clade is the curious presence of two flowers in each axil that he called superposed inflorescences. The presence of two supernumerary buds in the axil of each leaf has also been reported in M. guttatus and M. gemmiparus (Moody, Diggle, and Steingraeber, 1999 ). In each of these species, the distal bud ultimately becomes a flower or a lateral branch. In M. guttatus, the proximal bud remains inactive for the entire life of the plant. In M. gemmiparus, the proximal bud becomes a brood bulbil that functions as an asexual propagule, which is the primary means of propagation of this species (Beardsley, 1997 ). Therefore, among North American Phrymaceae, supernumerary buds can be seen to take on at least three different forms: a flower in Hemichaena-Berendtiella, a brood bulbil in M. gemmiparus, and dormant in M. guttatus.

Biogeography
Within Phrymoideae are two clades that represent distinct radiations and document two centers of diversity. The first center of diversity is in western North America and is well recognized and often studied. From within this radiation of species in western North America, several species have colonized other continents, possibly through long-distance dispersal. Mimulus depressus (Chile) is a member of sect. Simiolus, indicating a relatively recent establishment of Mimulus in South America. The presence of a suite of morphological traits makes it highly probable that most species of Mimulus in Chile are closely related and reflect one colonization event. Pollen (Argue, 1981 ) and other morphological data (von Bohlen, 1995b ) indicate that M. crinitus and M. bridgesii are not closely related to other Chilean Mimulus and reflect at least one additional colonization event. Mimulus nepalensis (Tibet and China) is derived from within a clade of taxa that are found primarily in California and the Pacific Northwest. In addition, morphological evidence (Grant, 1924 ) and preliminary sequence data (P. M. Beardsley, unpublished data) suggest that M. sessilifolius represents a second, distinct, relatively recent establishment of Mimulus in Asia.

The greatest number of species in North America occurs in California, with declining numbers of species to the north, extending to southern Alaska. Fewer species occur in the Great Basin, Rocky Mountains, and the desert Southwest. Some species not sampled in this study in sections Erythranthe, Simiolus, and Paradanthus are also found in Mexico. Our data suggest that these species were derived relatively recently from ancestors in California.

The Hemichaena-Berendtiella clade is found primarily in central and southern Mexico, and one species, Hemichaena fruiticosa, is native to southern Mexico, Guatemala, and Costa Rica (one of the two species in the family with a tropical distribution). Hemichaena fruiticosa is found in diverse habitats, both disturbed and undisturbed, in the mountains from 900 to 3500 m (Thieret, 1972 ). The second species in Phrymaceae found in Central America is Leucocarpus perfoliatus, which grows along stream banks in forests from 1350 to 2000 m (Pennell, 1920 ). Leucocarpus is sister to a large clade (approximately 60 species) that radiated extensively in western North America, indicating that some combination of higher rates of speciation or lower rates of extinction in temperate habitats than in tropical habitats exists in this clade.

Phrymaceae have undergone a second radiation in Australia. All Australian taxa within Mimulus, Glossostigma, and Peplidium, together comprising approximately 30 species, fall within one clade. The geographic division between the Australian species and the rest of the taxa in Phrymoideae, which are primarily North American, is certainly of ancient origin. The reconstructed ancestral distribution of the common ancestor of Phrymoideae is unresolved.

Populations of Phryma leptostachya exist in eastern North America (var. leptostachya) and eastern China and Japan (var. asiatica). Similar to other studies (Lee et al., 1996 ; Xiang et al., 2000 ), we recovered substantial divergence in the sampled DNA sequences between the two varieties. Using a molecular clock based on divergence in rbcL sequences, Xiang et al. (2000) estimated the time of divergence between the two varieties at 5.45 ± 2.47 million years. Hara (1962) demonstrated limited morphological divergence between the two varieties. Differences were limited to the shape, size, and pubescence of leaves. This lack of differentiation is perhaps explained by Phryma's greatly reduced floral display that offers limited opportunities for morphological change. Closely related groups of species in the newly defined Phrymaceae often differ morphologically in floral characters, while retaining pleisiomorphic vegetative characters.

The results of this study begin to clarify patterns of relationship in Mimulus, tribe Mimuleae, and other closely related genera. In light of this molecular evidence, we propose the recognition of a dramatically expanded Phrymaceae containing many species that have become model systems for the study of evolutionary processes. Future phylogenetic studies in Phrymaceae will continue to inform the systematics of the group and will serve as a useful starting point for further comparative analyses.

	FOOTNOTES

 
¹ The authors thank W. Barker, P. Cantino, D. Crawford, L. McDade, B. Oxelman, K. Ritland, S. Schoenig, and J. Willis for providing plant material or DNA and P. Cantino, L. McDade, P. Reeves, J. Reveal, and D. Schemske for helpful discussions. Funding was provided by grants to PMB from ASPT, Sigma Xi, and the University of Washington Department of Botany and NSF awards to RGO #DEB-9509804 and #DEB-9727025.

² Author for reprint requests (pbeard{at}u.washington.edu )

	LITERATURE CITED

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Argue C. L. 1980 Pollen morphology in the genus Mimulus (Scrophulariaceae) and its taxonomic significance. American Journal of Botany 67: 68-87[CrossRef][ISI]

Argue C. L. 1981 The taxonomic implications of pollen morphology in some South American species of Mimulus (Scrophulariaceae). American Journal of Botany 68: 200-205[CrossRef][ISI]

Argue C. L. 1984 Pollen morphology in Dodartia, Lancea, Leucocarpus, and Mazus and an analysis of pollen morphotypes in the Mimuleae (Scrophulariaceae). Canadian Journal of Botany 62: 1287-1297

Baldwin B. G. 1992 Phylogenetic utility of the internal transcribed spacers of nuclear ribosomal DNA in plants: an example from the Compositae. Molecular Phylogenetics and Evolution 1: 3-16[CrossRef][Medline]

Baldwin B. G. S. Markos 1998 Phylogenetic utility of the external transcribed spacer (ETS) of 18S-26SrDNA: congruence of ETS and ITS trees of Calycadenia (Compositae). Molecular Phylogenetics and Evolution 10: 449-463[CrossRef][ISI][Medline]

Baldwin B. G. M. J. Sanderson J. M. Porter M. F. Wojciechowski C. S. Campbell M. J. Donoghue 1995 The ITS region of nuclear ribosomal DNA: a valuable source of evidence on angiosperm phylogeny. Annals of the Missouri Botanical Garden 82: 247-277[CrossRef][ISI]

Barker W. R. 1982 Evolution and biogeography of arid Australian Scrophulariaceae. In W. R. Barker and P. J. M. Greenslade [eds.], Evolution of the flora and fauna of arid Australia, 341–350. Peacock Publications, Adelaide, South Australia, Australia

Beardsley M. R. 1997 Colorado's rare endemic plant, Mimulus gemmiparus, and its unique mode of reproduction. Master's thesis, Colorado State University, Fort Collins, Colorado, USA

Beeks R. M. 1962 Variation and hybridization in southern California populations of Diplacus (Scrophulariaceae). Aliso 5: 83-122

Bentham G. 1876 Scrophulariaceae. In G. Bentham and J. D. Hooker [eds.], Genera Plantarum, vol. 2, 946–958. Reeve, London, UK

Bradshaw H. D., Jr. K. G. Otto B. E. Frewen J. K. McKay D. W. Schemske 1998 Quantitative trait loci affecting differences in floral morphology between two plant species of monkeyflower (Mimulus). Genetics 149: 367-382[Abstract/Free Full Text]

Bradshaw H. D., Jr. S. M. Wilbert K. G. Otto D. W. Schemske 1995 Genetic mapping of floral traits associated with reproductive isolation in monkeyflowers (Mimulus). Nature 376: 762-765[CrossRef]

Burtt B. L. 1965 The transfer of Cyrtandromoea from Gesneriaceae to Scrophulariaceae, with notes on the classification of that family. Bulletin of Botanical Surveys of India 7: 73-88

Cantino P. D. H. N. Bryant K. De Queiroz M. J. Donoghue T. Eriksson D. M. Hillis M. S. Y. Lee 1999 Species names in phylogenetic nomenclature. Systematic Biology 48: 790-807[CrossRef][ISI][Medline]

Carr D. E. M. R. Dudash 1996 Inbreeding depression in two species of Mimulus (Scrophulariaceae) with contrasting mating systems. American Journal of Botany 83: 586-593[CrossRef][ISI]

Carr D. E. C. B. Fenster M. R. Dudash 1997 The relationship between mating-system characters and inbreeding depression in Mimulus guttatus. Evolution 51: 363-372[CrossRef][ISI]

Chadwell T. B. S. J. Wagstaff P. D. Cantino 1992 Pollen morphology of Phryma and some putative relatives. Systematic Botany 17: 210-219[CrossRef][ISI]

Cronquist A. 1981 The evolution and classification of higher plants. Houghton Mifflin, Boston, Massachusetts, USA

deQueiroz K. J. Gauthier 1992 Phylogenetic taxonomy. Annual Review of Ecology and Systematics 23: 449-480[ISI]

Diaz A. M. R. Macnair 1998 The effect of plant size on the expression of cleistogamy in Mimulus nasutus. Functional Ecology 12: 92-98[CrossRef][ISI]

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

Dudash M. R. D. E. Carr 1998 Genetics underlying inbreeding depression in Mimulus with contrasting mating systems. Nature 393: 682-684[CrossRef]

Dumortier B. C. 1829 Analyse des familles des plantes, avec l'indication des principaux genres qui s'y rattachent, vol. 23. Tournay, Hainaut, Belgium

Farris J. S. M. Kallersjo A. G. Kluge C. Bult 1994 Testing significance of incongruence. Cladistics 10: 315-319[CrossRef][ISI]

Felsenstein J. 1985 Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791[CrossRef][ISI]

Fetscher A. E. J. R. Kohn 1999 Stigma behavior in Mimulus aurantiacus. American Journal of Botany 86: 1130-1135[Abstract/Free Full Text]

Gilley L. P. Taberlet 1994 The use of chloroplast DNA to resolve plant phylogenies: non-coding versus rbcL sequences. Molecular Biology and Evolution 11: 769-777[Abstract]

Grant A. L. 1924 A monograph of the genus Mimulus. Annals of the Missouri Botanical Garden 11: 99-389[CrossRef]

Hara H. 1962 Racial differences in widespread species with special reference to those common to Japan and North America. American Journal of Botany 49: 647-652[CrossRef][ISI]

Hiesey W. M. M. A. Nobs O. Bjorkman 1971 Experimental studies in the nature of species. V. Biosystematics, genetics, and physiological ecology of the Erythranthe section of Mimulus. Carnegie Institution of Washington Publication 628, Washington, D.C., USA

Holm T. 1913 Phryma leptostachya L., a morphological study. Botanical Gazette 56: 306-318[CrossRef]

Karron J. D. R. T. Jackson N. N. Thumser S. L. Schlicht 1997 Outcrossing rates of individual Mimulus ringens genets are correlated with anther-stigma separation. Heredity 79: 365-370[CrossRef][ISI]

Leclerc-Potvin C. K. Ritland 1994 Modes of self-fertilization in Mimulus guttatus (Scrophulariaceae): a field experiment. American Journal of Botany 81: 199-205[CrossRef][ISI]

Lee N. S. T. Sang D. J. Crawford S. H. Yeau S. C. Kim 1996 Molecular divergence between disjunct taxa in eastern Asia and eastern North America. American Journal of Botany 83: 1373-1378[CrossRef][ISI]

Lu A. 1990 A preliminary cladistic study of the families of the superorder Lamiiflorae. Botanical Journal of the Linnean Society 103: 39-57

Maddison D. R. 1991 The discovery and importance of multiple islands of most-parsimonious trees. Systematic Zoology 40: 315-328[CrossRef]

McDade L. A. S. E. Masta M. L. Moody E. Waters 2000 Phylogenetic relationships among Acanthaceae: evidence from two genomes. Systematic Botany 25: 106-121[CrossRef][ISI]

McDade L. A. M. L. Moody 1999 Phylogenetic relationships among Acanthaceae: evidence from noncoding trnL-trnF chloroplast DNA sequences. American Journal of Botany 86: 70-80[Abstract/Free Full Text]

McMinn H. E. 1951 Studies in the genus Diplacus (Scrophulariaceae). Madrono 11: 1-32

Moldenke H. N. 1971 A fifth summary of the Verbenaceae, Avicenniaceae, Stilbaceae, Dicrastylidaceae, Symphoremaceae, Nyctanthaceae, and Eriocaulaceae of the world as to the valid taxa, geographic distribution and synonomy. The William Paterson College of New Jersey, Wayne, New Jersey, USA

Moody A. P. K. Diggle D. A. Steingraeber 1999 Developmental analysis of the evolutionary origin of vegetative propagules in Mimulus gemmiparus (Scrophulariaceae). American Journal of Botany 86: 1512-1522[Abstract/Free Full Text]

Olmstead R. G. C. W. DePamphilis A. D. Wolfe N. D. Young P. A. Reeves 2001 Disintegration of the Scrophulariaceae. American Journal of Botany 88: 348-361[Abstract/Free Full Text]

Olmstead R. G. P. A. Reeves 1995 Evidence for the polyphyly of the Scrophulariaceae based on chloroplast rbcL and ndhF sequences. Annals of the Missouri Botanical Garden 82: 176-193[CrossRef][ISI]

Pennell F. W. 1920 Scrophulariaceae of Colombia. Proceedings of the Academy of Natural Sciences, Philadelphia 72: 136-188

Pennell F. W. 1935 Scrophulariaceae of eastern temperate North America. Proceedings of the Academy of Natural Sciences, Philadelphia, Monograph No. 1. Wickersham, Lancaster, Pennsylvania, USA

Pennell F. W. 1947 Some hitherto undescribed Scrophulariaceae of the Pacific states. Proceedings of the Academy of Natural Sciences, Philadelphia 99: 151-171

Pennell F. W. 1951 Mimulus. In L. Abrams [ed.], Illustrated flora of the Pacific states, vol. 3, 688–731. Stanford University Press, Palo Alto, California, USA

Schauer J. C. 1847 Phrymaceae. In A. De Candolle [ed.], Prodromus 11: 520.

Stebbins G. L. 1989 Adaptive shifts toward hummingbird pollination. In J. H. Bock and Y. B. Linhart [eds.], The evolutionary ecology of plants, 39–60. Westview Press, Boulder, Colorado, USA

Swofford D. L. 1998 PAUP*: phylogenetic analysis using parsimony. Sinauer Associates, Sunderland, Massachusetts, USA

Taberlet P. L. Gielly G. Patou J. Bouvet 1991 Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105-1109[CrossRef][ISI][Medline]

Thieret J. W. 1954 The tribes and genera of Central American Scrophulariaceae. Ceiba 4: 164-184

Thieret J. W. 1967 Supraspecific classification in the Scrophulariaceae: a review. SIDA Contributions to Botany 3: 87-106

Thieret J. W. 1972 Synopsis of Hemichaena, including Berendtiella (Scrophulariaceae). Fieldiana: Botany 34: 89-99

Thompson D. M. 1993 Mimulus. In J. C. Hickman [ed.], The Jepson manual, 1037–1051. University of California Press, Berkeley, California, USA

Thompson D. M. 1997 The genus Mimulus. In B. C. O'Brien, L. C. Fuentes, and L. F. Newcombe [eds.], Rancho Santa Ana botanic garden occasional publications No. 1, symposium proceedings, 112–125. Rancho Santa Ana Botanic Garden, Claremont, California, USA

Vickery R. K. 1969 Crossing barriers in Mimulus. Journal of Japanese Genetics 44: supplement 1 325-336

Vickery R. K. 1997 A systematic study of the Mimulus wiensii complex (Scrophulariaceae: Mimulus section Simiolus), including M. yeocorensis and M. minutiflorus, new species from western Mexico. Madrono 44: 384-393

von Bohlen C. 1995a El genero Mimulus l. (Scrophulariaceae) en Chile. Gayana Botanica 52: 7-28

von Bohlen C. 1995b Mimulus crinitus A. L. Grant (Scrophulariaceae), transferido de la seccion Simiolus Green a la seccion Paradanthus A. L. Grant. Gayana Botanica 52: 1-5

von Wettstein R. 1891 Scrophulariaceae. In A. Engler and K. Prantl [eds.], Die naturlichen Pflanzenfamilien, vol. 4 (3b), 39–107. Wilhelm Engelmann, Leipzig, Germany

Waayers G. M. 1996 Hybridization, introgression, and selection in Mimulus aurantiacus ssp. australis and Mimulus puniceus. Master's thesis, San Diego State University, San Diego, California, USA

Whipple H. L. 1972 Structure and systematics of Phryma leptostachya L. Journal of the Elisha Mitchell Scientific Society 88: 1-17

Willis J. H. 1999 The role of genes of large effect on inbreeding depression in Mimulus guttatus. Evolution 53: 1678-1691[CrossRef][ISI]

Xiang Q. Y. D. E. Soltis P. S. Soltis S. R. Manchester D. J. Crawford 2000 Timing the eastern Asian-eastern North American floristic disjunction: molecular clock corroborates paleontological estimates. Molecular Phylogenetics and Evolution 15: 462-472[CrossRef][ISI][Medline]

Yamazaki T. 1993 Two new species of Mimulus from the Himalayas. Journal of Japanese Botany 68: 23-26

This article has been cited by other articles:

D. C. Tank and R. G. Olmstead From annuals to perennials: phylogeny of subtribe Castillejinae (Orobanchaceae) Am. J. Botany, May 1, 2008; 95(5): 608 - 625. [Abstract] [Full Text]