Abstract

Background and Aims

The genus Erodium is a common feature of Mediterranean-type climates throughout the world, but the Mediterranean Basin has significantly higher diversity than other areas. The aim here is to reveal the biogeographical history of the genus and the causes behind the evolution of the uneven distribution.

Methods

Seventy-eight new nrITS sequences were incorporated with existing plastid data to explore the phylogenetic relationships and biogeography of Erodium using several reconstruction methods. Divergence times for major clades were calculated and contrasted with other previously published information. Furthermore, topological and temporal diversification rate shift analyses were employed using these data.

Key Results

Phylogenetic relationships among species are widely congruent with previous plastid reconstructions, which refute the classical taxonomical classification. Biogeographical reconstructions point to Asia as the ancestral area of Erodium, arising approx. 18 MYA. Four incidences of intercontinental dispersal from the Mediterranean Basin to similar climates are demonstrated. Increases in diversification were present in two independent Erodium lineages concurrently. Two bursts of diversification (3 MYA and 0·69 MYA) were detected only in the Mediterranean flora.

Conclusions

Two lineages diverged early in the evolution of the genus Erodium: (1) subgenus Erodium plus subgenus Barbata subsection Absinthioidea and (2) the remainder of subgenus Barbata. Dispersal across major water bodies, although uncommon, has had a major influence on the distribution of this genus and is likely to have played as significant role as in other, more easily dispersed, genera. Establishment of Mediterranean climates has facilitated the spread of the genus and been crucial in its diversification. Two, independent, rapid radiations in response to the onset of drought and glacial climate change indicate putative adaptive radiations in the genus.

INTRODUCTION

Traditionally, disjunct distributions within taxonomic groups have been attributed to plate tectonic shifts or the historical presence of land bridges, such as the Beringian land bridge or the North Atlantic land bridge. These continental connections provide a clear explanation for the distribution of many groups, such as the Tertiary relict floras (Milne, 2006). However, the use of molecular techniques for dating divergence times is continually revealing younger ages than previously accepted for many taxa, uncovering patterns that are incompatible with tectonic shifts (Milne, 2006). An alternative hypothesis to explain distribution patterns is long-distance dispersal of organisms, a concept that is increasingly accepted as the last resort. In fact it has been pointed out that one single seed dispersal ‘every few millions of years can have a large impact on biogeography’ (Milne, 2006). Investigation of the role of dispersal in the evolutionary process has shed light on many aspects of evolutionary biology, such as selection pressures, dormancy, altruism and senescence (Ronce, 2007). The existence of many plant groups with disjunct distributions which are present in the Mediterranean Basin was thought to happen via arid corridors (Coleman et al., 2003). However, this hypothesis has been refuted, and long-distance dispersal has been proposed as an alternative explanation for the colonization of these Mediterranean climate zones (Coleman et al., 2003). The Mediterranean Basin region is a hotspot of global diversity, and evolutionary studies of its flora contribute to our understanding of Mediterranean ecosystems. Given its worldwide distribution and high diversity, especially in Mediterranean regions, Erodium provides an excellent model system to study these phenomena.

Under the current circumscription, Geraniaceae sensu stricto comprises five genera: Geranium, Monsonia, Pelargonium, California and Erodium. The genus Erodium has 74 species and is distributed on all continents, excluding Antarctica (Fig. 1; Fiz et al., 2006). A major centre of diversity is observed in the Mediterranean Basin region (62 species), whereas, the other continents harbour only a few native species: one each in North and South America, five in Australia and four in Asia. The dispersal and colonization ability of Erodium is reflected in this broad distribution. In fact, four species (E. botrys, E. brachycarpum, E. cicutarium and E. moschatum) each occupy Mediterranean floristic regions on all continents where these habitats exist. The monophyletic clade containing Erodium, California and Geranium has its origin in Asia, having migrated from southern Africa (Fiz et al., 2008). The origin of Erodium is less clear as California, its monotypic sister genus, is endemic to California.

Fig. 1.

Map detailing the distribution and areas of endemism for Erodium species (shaded). The arrows represent the four possible intercontinental dispersals (see main text). a = North Africa, b = Italy and Balkans, c = Iberian Peninsula, d = The Middle East, e = Asia, f = North Europe, g = North America, h = South America, i = Australia.

Fig. 1.

Map detailing the distribution and areas of endemism for Erodium species (shaded). The arrows represent the four possible intercontinental dispersals (see main text). a = North Africa, b = Italy and Balkans, c = Iberian Peninsula, d = The Middle East, e = Asia, f = North Europe, g = North America, h = South America, i = Australia.

The Mediterranean Basin can be divided into two subregions of high species diversity for Erodium: subsection Absinthioidea (13 species) is centred in the eastern region, and the western region contains the core of section Barbata. In the western Mediterranean, 22 species are endemic to Spain and Morocco, eight of which are restricted to one or a few small populations. Rapid species radiations have been described for a number of genera with Mediterranean distribution (Valente et al., 2010). This, together with the high rate of nucleotide substitution in Erodium (Fiz, 2005; Guisinger et al., 2008), makes this genus an excellent candidate to explore putative rapid radiations.

It was not possible to resolve low-level relationships within Erodium by analysing the trnL-F gene region and morphological characters (Fiz et al., 2006, 2008). However, these studies demonstrated the monophyly of Erodium and its division into two main clades, placed California macrophylla as its sister species, and highlighted an ancient diversification pattern. Finer-scale analyses of molecular regions with higher variability are needed to investigate more recent species formation and colonization patterns in Erodium. Sequences of nuclear ITS have been frequently used to study low-level phylogenetic relationships and to infer reticulation effects on phylogeny (Campbell et al., 1997; Fuertes Aguilar and Nieto Feliner, 2003), to infer cases of allopolyploidy (Whittall et al., 2000) and historical biogeographical patterns (Sang et al., 1995; Campbell et al., 1997), and as such this region is an appropriate candidate for further investigation of evolution and dispersal within Erodium.

The main objective of the study presented here is to examine the historical biogeography of Erodium in an attempt to explain its disjunct distribution over five continents. To this end, variation in newly sequenced nrITS sequences is used to search for fine-scale phylogenetic relationships and this is compared with phylogenetic information gathered using previously published plastid sequences (Fiz et al., 2006, 2008). Greater phylogenetic resolution was necessary to reconstruct the evolutionary history of Erodium. A total evidence approach is adopted (combining nrITS and trnL-F matrices) and used to infer general biogeographical patterns in Erodium. Diversification rate shifts were also explored in an effort to elucidate putative, rapid radiations throughout the genus. These data along with divergence time estimates will lead to a better understanding of the evolutionary history of Erodium.

MATERIALS AND METHODS

Plant material and PCR

Seventy-eight nrITS sequences were generated, corresponding to 58 Erodium taxa and two individuals of California macrophylla (Appendix). DNA extracted from a previous study (Fiz et al., 2006) was used, and DNA of several additional individuals was extracted from herbarium and field collections (Appendix) using DNeasy Plant Mini Kit (QIAGEN Laboratories, Germany).

The ITS region was amplified using the polymerase chain reaction (PCR) with following primer pairs ITS-4 and ITS-5 (White et al., 1990) and 17SE and 26SE (Sun et al., 1994. In addition, 0·5 µL of dimethyl sulfoxide was added to each reaction. PCR conditions for amplification were as described in Fiz et al. (2002). The PCR-Beads kit (‘puRetaq Ready-To-Go’; Amersham Biosciences) was occasionally used for poorly preserved DNA from herbarium specimens. Amplified products were then purified using spin filter columns (PCR clean-up kit; MoBio Laboratories, CA, USA) following the protocols provided by the manufacturer. Forward and reverse sequences from cleaned products were then directly sequenced as in Fiz et al. (2002).

Alignment and phylogenetic analyses

Sequence data were edited using the program Seqed (Applied Biosystems). The limits of the ITS1-5·8S-ITS2 region were determined by comparison with secondary structure of the homologue in Asteraceae (Goertzen et al., 2003). CLUSTAL W (Thompson et al., 1994) was used as a first approach to the alignment of the sequences, followed by manual adjustment. IUPAC symbols were used to represent nucleotide ambiguities.

Three different matrices were constructed and analysed: the first included all 79 nrITS sequences (76 Erodium, two California and one Geranium); the second consisted of 60 nrITS sequences corresponding to one sequence per taxon (58 Erodium, one California and one Geranium); the third matrix combined nrITS and trnL-F sequences from 69 taxa (67 Erodium, one California and one Geranium) taken from a previous study (Fiz et al., 2006). Sequences of the two molecular markers from the same individual (i.e. the same DNA extraction) were used for the combined matrix with the exception of 11 taxa (Appendix).

Matrices were analysed using maximum parsimony (MP) and Bayesian inference (BI) methods. Geranium and California were included as outgroups. Models were chosen based on the Akaike information criterion (AIC) as implemented in MrModeltest 1·1b (Nylander, 2002), which is a simplified version of Modeltest 3·06 (Posada and Crandall, 1998). The MP phylogenetic analyses were conducted using Fitch parsimony as implemented in PAUP* (Swofford, 1999), with unordered and equal weighting of all characters. All phylogenetic analyses were performed for the whole nrITS and for ITS1, 5, 8S and ITS2 independently. Heuristic searches were conducted using random taxon-addition sequences (100 replicates), tree bisection–reconnection branch swapping and with the options MULPARS and STEEPEST DESCENT in effect. Relative support for clades identified by parsimony analysis was assessed by both ‘fast’ bootstrapping (10 000 re-samplings using the heuristic search strategy as indicated above (Mort et al., 2000) and ‘full’ bootstrapping (1000 re-samplings with simple taxon addition and SPR branch swapping but permitting only 10 trees per replicate to be held). Phylogenetic reconstructions using distance-based reconstructions were also explored in PAUP* using the neighbor-joining method. Genetic distances obtained from this analysis were used to study nrITS interspecific variability (Table 1).

Table 1.

Genetic distances (%) for nrITS at genus, species and intraspecific level

Genus level: between different genera 35·23 % to 25·61 %: Geranium biumcinatum/Erodium 
26·69 % to 25·81 %: Geranium biumcinatum/California macrophylla 
25·32 % to 16·68 %: California macrophylla/Erodium 
Species level: between Erodium species 20·96 %: E. laciniatum 1/E. crassifolium 
0 %: E. botrys/E. brachycarpun 
0 %: E. glandulosum1/E. rupestre 2/E. lucidum/E. antariense 2/E. foetidum/E. foetidum cheilanthifolium/E. foetidum celtibericum 
0 %: E. cossoni/E. laciniatum 
0 %: E. touchyanum 2/E. moschatum 
0 %: E. mouretti/E. moschatum and E. touchyanum 
Intraspecific level: between populations of the same species 6·18 %: California macrophylla 
0·674 %: E. gruinum; 0·497 %: E. touchyanum; 0·162 %: E. ciconium, E. cygnorum, E. laciniatum and E. cazorlanum 
0 %: E. arborescens, E. hoeftianum, E. alpinum, E. rupestre, E. antariense, E. reichardii, E. sanguis-christi, E. recoderii, E. rupicola, E. sebaceum, E. moschatum, E. tordylioides 
Genus level: between different genera 35·23 % to 25·61 %: Geranium biumcinatum/Erodium 
26·69 % to 25·81 %: Geranium biumcinatum/California macrophylla 
25·32 % to 16·68 %: California macrophylla/Erodium 
Species level: between Erodium species 20·96 %: E. laciniatum 1/E. crassifolium 
0 %: E. botrys/E. brachycarpun 
0 %: E. glandulosum1/E. rupestre 2/E. lucidum/E. antariense 2/E. foetidum/E. foetidum cheilanthifolium/E. foetidum celtibericum 
0 %: E. cossoni/E. laciniatum 
0 %: E. touchyanum 2/E. moschatum 
0 %: E. mouretti/E. moschatum and E. touchyanum 
Intraspecific level: between populations of the same species 6·18 %: California macrophylla 
0·674 %: E. gruinum; 0·497 %: E. touchyanum; 0·162 %: E. ciconium, E. cygnorum, E. laciniatum and E. cazorlanum 
0 %: E. arborescens, E. hoeftianum, E. alpinum, E. rupestre, E. antariense, E. reichardii, E. sanguis-christi, E. recoderii, E. rupicola, E. sebaceum, E. moschatum, E. tordylioides 

Bayesian approaches were conducted using MrBayes 3·0b4 (Ronquist and Huelsenbeck, 2003), sampling for two million generations with four Markov Chain Monte Carlo (MCMC) chains (chain temperature 0·2; sample frequency 100). Combined analysis (nrITS and trnL-F) was partitioned, and different substitution models were applied and each partition had its own parameters. As such nrITS was coded with the GTR + I + Γ model and trnL-F with the GTR + Γ model. For all analyses burn-in was discarded. Posterior probabilities were examined to avoid sampling prior to convergence and mixing, and finally a majority-rule tree was reconstructed after discarding 1 × 105 generations.

For the biogeographical study, nine areas of endemism were defined, each having at least one endemic taxon present (Fig. 1). Lagrange v. 2·0·1 (Ree et al., 2005; Ree and Smith, 2008) and DIVA (Ronquist, 1996) were used to estimate ancestral areas and dispersals along the evolution of Erodium. Both analyses were performed in one of the posterior probability trees from the Bayesian analysis of the combined matrix. The cosmopolitan outgroup taxon Geranium was removed from the analyses. The human-mediated recent colonizations of the New World and Australia by E. botrys, E. brachycarpum, E. cicutarium and E. moschatum were not taken into account in Lagrange analysis. Including them in the analysis resulted invariably in a ‘convergence error’. All possible area combinations with a maximum of six simultaneous areas were permitted except for those containing non-adjacent areas and for intercontinental ranges of more than two areas. Four ‘continents’ or main regions weakly connected between them are considered here: North America, South America, Australia and Africa–Eurasia. The restrictions applied are justified given the distributions of extant taxa. The biogeographical model used was constant through time. Dispersals between all neighbouring areas were permitted bidirectionally, but they were given three different levels of probability (see Table S1 in Supplementary Data, available online).

Divergence times were reconstructed for nrITS (58 Erodium species plus California) using BEAST v. 1·4·7 (Drummond and Rambaut, 2007) under the uncorrelated lognormal dating method and the GTR + I + Γ substitution model. Assignment of fossils to nodes is a significant problem in phylogenetic tree calibration, but the use of multiple fossils may reduce the effect of incorrect placements (Rutschmann et al., 2007). The tree was constrained at the divergence between Erodium and California using a normal prior distribution [20·34 million years ago (MYA); after Fiz et al., 2008] and a uniform prior distribution to constrain the dating to the only fossil available (8 MYA on Clade II; Fiz et al., 2008). Ten million generations, with sampling every 1000th generation, were simulated and a tree was reconstructed after examination of all parameters, discarding the generations before MCMC-chain convergence. This analysis was repeated using extra outgroups [four from Geraniaceae and two from other families of Geraniales sensu APG III (2009)] to check for any distortion of the reconstructed divergence times. An additional calibration point was included with a normal prior distribution (42 MYA mean, four standard deviations) for the Geraniaceae crown node (Fiz et al., 2008). Inclusion of more distantly related species was not possible as the high rate of mutation in Geraniales made homology assessment for the alignment unclear. All temporal reconstructions were compared with dates retrieved from the rbcL marker for all main nodes in Geraniaceae (Fiz et al., 2008).

Three approaches were implemented for the diversification analysis. For the first, diversification rate shifts were examined, within the same tree as used in the analysis of biogeography, using a topological method. The likelihood-based diversification rate shift statistics, Δ1 (Moore et al. 2004) and the Slowinski–Guyer statistic (Slowinski and Guyer, 1993) as implemented in SymmeTREE (Chan and Moore, 2005) were calculated. One hundred thousand trees equal in length to the study tree were generated by Monte Carlo simulation in order to estimate the null distribution of the test statistic, Δ1. For the second, a temporal method (LASER; Rabosky, 2006) was used to test constancy of the overall rate of diversification. Since this model was rejected, to find out which rate-variable model fitted best to the data, between two (one rate change) and five rate parameters (four rate changes) were incorporated, enabling the detection of temporal increases in diversification rates. For the third, LASER was used to identify the position on the tree at which the diversification rate shift had the highest likelihood. All 74 species of Erodium were incorporated into the tree by assigning the number of species to each clade on the nrITS chronogram (with only California as an outgroup) (following Fiz et al., 2006). When rate constancy was rejected, two values for the extinction rate (0 and 0·9) were taken into account.

RESULTS

Phylogenetic analysis

The length of the nrITS1 is 228 bp in California macrophylla and 208–231 bp in Erodium. The length of nrITS2 is 232–233 bp in C. macrophylla and 232–235 in Erodium. The numbers of variable and potentially parsimony informative sites in the nrITS matrix are 251/193 [687 steps; consistency index (CI) 0·51; retention index (RI) 0·86] and 209/276 (903 steps; CI 0·62; RI 0·86) in the combined matrix.

All phylogenetic analyses confirm the monophyly of Erodium. California macrophylla has lower genetic distance to Erodium (16·68 %) than to Geranium (25·81 %) (Table 1). BI and MP reconstructions using the nrITS region revealed two main clades (I and II; Figs 2 and 3). Clade I [73 % BS (bootstrap), 0·94 pp (posterior probability)] includes subgenus Erodium and subgenus Barbata subsection Absinthioidea. Clade II (59 % BS, 1·0 pp) comprises the remainder of subgenus Barbata and can be divided into a further two subclades: one comprising only Australian species (100 % BS, 1·00 pp) and the second containing the rest of the predominantly Mediterranean species of subgenus Barbata (50 % BS, 0·99 pp). Within the core subgenus Barbata, two cosmopolitan species (E. brachycarpum and E. botrys) are sister to the remainder; the rest of the core subgenus Barbata (81 % BS, 1·00 pp) is represented by several nested clades: section Malacoidea and subsection Petraea are successively sister to section Cicutaria. Topological resolution and support are congruent between MP and BI analyses of the nrITS sequences (Fig. 2). The incongruence length difference test shows significant incongruence between ITS1 vs. trnL-F (P = 0·01) and ITS2 vs. trnL-F (P = 0·019). In contrast, congruence between ITS1 and ITS2 is high (P = 0·92).

Fig. 2.

Majority rule consensus tree obtained using Bayesian analysis of all ITS sequences available in this study. Posterior probabilities (pp) are given as values above branches. Bootstrap support (BS) values >50 % are below the branches. An asterisk denotes 1·0 pp or 100 BS. Diversification rate shifts (obtained from SymmeTREE and LASER) are indicated by circles. Clade names (I and II) according to Fiz et al. (2006) are highlighted in squares and taxonomic information according to Guittonneau (1990) is indicated on the right.

Fig. 2.

Majority rule consensus tree obtained using Bayesian analysis of all ITS sequences available in this study. Posterior probabilities (pp) are given as values above branches. Bootstrap support (BS) values >50 % are below the branches. An asterisk denotes 1·0 pp or 100 BS. Diversification rate shifts (obtained from SymmeTREE and LASER) are indicated by circles. Clade names (I and II) according to Fiz et al. (2006) are highlighted in squares and taxonomic information according to Guittonneau (1990) is indicated on the right.

Fig. 3.

Combined majority rule consensus tree (nrITS and trnL-trnF sequences) obtained from Bayesian analysis using one sequence per taxa. Values above branches are posterior probabilities (pp) and those below correspond to bootstrap (BS) supports >50 %. An asterisk denotes 1·0 pp or 100 % BS. Ancestral ranges recovered with the highest probability with Lagrange are shown at internal nodes (letters following those in Fig. 1). No letter indicates invariable estimated range for the remaining nodes of that lineage. Arrows highlight cases of intercontinental dispersals (see Fig. 1). Current ranges for each taxon are indicated on the right. For these reconstructions, it was assumed that cosmopolitan taxa have been recently introduced to the New World and Australia. Clade names (squares I and II, and 1–4) are also indicated.

Fig. 3.

Combined majority rule consensus tree (nrITS and trnL-trnF sequences) obtained from Bayesian analysis using one sequence per taxa. Values above branches are posterior probabilities (pp) and those below correspond to bootstrap (BS) supports >50 %. An asterisk denotes 1·0 pp or 100 % BS. Ancestral ranges recovered with the highest probability with Lagrange are shown at internal nodes (letters following those in Fig. 1). No letter indicates invariable estimated range for the remaining nodes of that lineage. Arrows highlight cases of intercontinental dispersals (see Fig. 1). Current ranges for each taxon are indicated on the right. For these reconstructions, it was assumed that cosmopolitan taxa have been recently introduced to the New World and Australia. Clade names (squares I and II, and 1–4) are also indicated.

In the combined tree (nrITS and trnL-F; Fig. 3) the species of subgenus Erodium are monophyletic, whereas in the nrITS reconstruction they are split between two clades (Fig. 2). Furthermore, in the nrITS reconstruction both clades form a polytomy with the clade containing E. stephanianum and subsection Absinthioidea (Fig. 2), whereas in the combined tree it appears as monophyletic with the E. stephanianum clade as sister group. There is a conflicting signal given by the two molecular datasets in clade 2 (Fig. 3). In the combined tree E. hoeftianum is highlighted as the first branching lineage, whereas in the nrITS reconstruction E. ciconium is placed as sister to other taxa in subsection Absinthioidea with poor statistical support. On the other hand, within subgenus Barbata (clade II) relationships among the E. cygnorum group, E. botrys group, subsection Petraea, section Malacoidea and section Cicutaria are completely congruent between nrITS and combined trees. Incongruences inside section Malacoidea include: (a) in the combined tree, E. reichardii and E. boissieri are sister to the core group, whereas in the nrITS tree they form a polytomy and (b) in the combined tree the E. moschatum clade is sister to the rest of section Cicutaria whereas in nrITS it is not. However, as both alternatives within section Cicutaria have low statistical support the phylogenetic relationship of these clades remains unclear.

The greatest nrITS sequence divergence between species was in the comparison of E. laciniatum 1 and E. crassifolium (20·96 %, GTR + Γ model). On the other hand, a total of 21 species pairs showed no divergence whatsoever. Intraspecific genetic distances range from 0·674 % in E. gruinum to 0 % within 12 other species (Table 1).

Biogeography and diversification analysis in Erodium

Both Lagrange and DIVA reconstruction methods broadly agree in assigning ancestral areas. As reconstructed by Lagrange, the estimated origin of Erodium is in Asia or North America; subsequent diversification extended the genus into the rest of Asia, the Mediterranean Basin and Africa. The DIVA analysis resulted in one optimal reconstruction requiring 79 dispersals and a combination of multiple areas is retrieved from DIVA (results not shown) for the origin of Erodium. Subgenus Erodium is assigned to North Africa while its sister subsection Absinthioidea is assigned to the eastern Mediterranean (plus Asia and Asia Minor with DIVA). The ancestral area of both clades is Asia (plus Asia Minor with DIVA). The ancestor of clade II is assigned to Asia or a combination of areas (Fig. 3) or to the Iberian Peninsula, Australia and South America when using DIVA. Within this clade, the ancestral area for sections Malacoidea and Cicutaria and subsection Petraea is the Iberian Peninsula. The ancestor of each individual clade stems from the Iberian Peninsula, but section Cicutaria is also assigned to six combined areas (Lagrange; Fig. 3) or to North Africa (DIVA). Biogeographical analysis suggests four long-distance dispersal events: two groups are endemic to Australia and two species are endemic to America (Fig. 3). The Australian group (clade 3) and a further Australian species nested within the body of sect Malacoidea (E. aureum) provide the ambiguity in the biogeographical origin of Clade II in the DIVA analysis. Moreover, the distant evolutionary relationship of E. geoides (South America) and E. texanum (North America) indicates two independent colonizations of the American continents.

The reconstructed divergence times of the major clades when using the nrITS marker are broadly congruent with those reported using the plastid rbcL (Fig. S1 in Supplementary Data, available online). Confidence intervals are broader when six outgroups are included (see Fig. S1 in Supplementary data). The use of California as a sole outgroup retrieved ages more congruent with those derived from rbcL for the crown age of the genus Erodium (18·34 MYA), the crown age of Clade II (14·81 MYA) and stem age of the core of Clade II (11·73 MYA) (chronogram in Supplementary Fig. S1). On the other hand, the age retrieved for clade I (13·68 MYA) was more similar when six outgroups were used, as was the crown age of section Absinthioidea (4·4 vs. 3 MYA). However, if E. hoeftianum is not taken into account, the age of this clade is similar among the three reconstructions (approx. 3 MYA). The use of six outgroups caused alignment ambiguities (an increase of 37 nucleotide positions), a result of the large genetic distance of Erodium from the majority of the outgroups included. This factor, coupled with the distortion of the crown and stem ages of Erodium from those expected from analyses of rbcL (Fig. S1 in Supplementary data), justify the rejection of reconstructed ages of the six-outgroup analysis.

Although no significant diversification rate shifts were detected with SymmeTREE, two marginally significant cases were identified (Fig. 2). The first shift (P = 0·11) occurs in subsection Absinthioidea after the divergence of E. ciconium and is significant when using and Slowinski–Guyer statistic (P = 0·04). The second (P = 0·08) occurs after the divergence of the group containing E. botrys, along the branch subtending the core of subgenus Barbata (Fig. 2). All other branches show balance in the diversification rate (P = 1).

The likelihood ratio test (LASER) for the 74 species tree, based on the nrITS chronogram (with California as the sole outgroup), also supported diversification rate heterogeneity when extinction was zero (P = 0·005) but not when extinction was 0·9 (P = 0·16). This analysis highlighted a diversification rate shift at the crown node of core subgenus Barbata in agreement with the second diversification rate shift detected using SymmeTREE (Fig. 2). Overall rate constancy was rejected (delta AICrc = 2·78); a three-rate variable model (Yule-3-rate) fitted the data most effectively. This analysis detected two inflection points at 0·069 and 3·09 MYA, indicating two increases in diversification rates (Fig. 4). An inflection point at 3·18 MYA (Yule-2-rate) was also detected from the six-outgroup chronogram (Fig. 4).

Fig. 4.

Lineage-through-time plot for nrITS chronogram using California as the outgroup and using six outgroups, as indicated. The diversification rate shifts located by LASER are indicated as triangles.

Fig. 4.

Lineage-through-time plot for nrITS chronogram using California as the outgroup and using six outgroups, as indicated. The diversification rate shifts located by LASER are indicated as triangles.

DISCUSSION

This study confirms that Erodium is a natural group as previously demonstrated using plastid DNA regions and morphology (Fiz et al., 2006, 2008). A consistent structure is obtained across all phylogenetic analyses of Erodium, with two main clades (I and II) present in all analyses (trnL-F and rbcL; Fiz et al., 2006, 2008). In a similar fashion, subclades 2, 3 and 4 are always recovered in phylogenetic analyses of nuclear and plastid markers. Subgenus Erodium (clade 1) is paraphyletic in the analyses of the nuclear marker (nrITS), but is a monophyletic group in the trnL-F (Fiz et al., 2006) and combined analyses. This is probably an indication of reticulate evolution, potentially due to hybridization between members of section Erodium and subsection Absinthioidea. Section Absinthioidea (sensu Guittonneau) is split into three separate groups, and the paraphyly of section Malacoidea is also confirmed. Although there are different phylogenetic incongruences between markers at the species level (e.g. E. cicutarium), none of them is strongly supported. In these cases incomplete lineage sorting and plastid capture may also explain these differences. Hence, the current and previous studies (trnL-F and rbcL; Fiz et al., 2006, 2008) agree that a revision of the subgeneric and sectional circumscription of Erodium species is required.

Although special care must be taken with the nrITS marker when dealing with branch length in angiosperms (especially at low scales, Soria-Hernanz et al., 2008; see also Bromham et al., 2000), a signal was recovered that is broadly congruent with that gained from plastid regions in Erodium (Fiz et al., 2008). The reconstruction of divergence times places the origin and diversification of Erodium in the early Miocene. The extension of steppes at the end of the Tertiary (Kers, 1968; Venter, 1983) has been linked to the evolution of California, Erodium and Geranium (Geraniaceae; Fiz et al., 2008). As the American genus California is sister to Erodium, an origin in Asia spreading through to North Africa seems more probable [as for instance in Zygophyllaceae (Beier et al., 2004) and Plantaginaceae (Meyers and Liston, 2008)]. One explanation for the disjunction between North America (California) and Asia (Erodium) is the breaking of the Beringian land bridge approx 20 MYA.

Multiple intercontinental colonizations

The phylogenetic evidence presented here highlights four incidences in which sister species/clades possess allopatric distributions on separate continents. Furthermore, biogeographical analyses using two reconstruction methods support the hypothesis that these four colonizations constitute long-range, intercontinental dispersal events. Erodium texanum stems from a predominantly Mediterranean lineage originating in Africa. Its presence in North America constitutes a convincing example of long-distance dispersal. The only endemic South American species, E. geoides, is sister to the clade formed by the cosmopolitan species E. botrys and E. brachycarpum. This clade also has its origins in Africa, supporting at least one long-distance dispersal between the continents. Two separate colonizations of Australia have occurred. The ancestor of E. aureum dispersed relatively recently from the Mediterranean Basin and Middle East (ACD), whereas the E. cygnorum clade represents a more ancient dispersal event, most likely from Asia.

Although E. geoides arrived in South America as long ago as 12 MYA (Fiz et al., 2008 and the present study) when the Beringian land bridge and North Atlantic land bridge were in existence, a dispersal event across the Atlantic should not be ruled out as its origin appears to be in Africa. However, the alternative explanation that the ancestor of the clade was a more widespread species that evolved in Africa, spread to the New World via land bridges and diversified into the three modern species has some merit, as previously suggested for subfamily Betoideae [Chenopodiaceae (= Amaranthaceae sensu APG III, 2009); Hohmann et al., 2006]. In contrast, the derived position of E. texanum, nested within a Mediterranean lineage, implies a recent (0·6 and 0·9 MYA from nrITS reconstructions) dispersal of Erodium making the North Atlantic land bridge migratory route impossible and strongly implicating a translocation across the Atlantic from the Mediterranean Basin. Therefore, this is confirmed as another case of long-distance dispersal between Mediterranean–North Africa and western North America for taxa adapted to semi-arid habitats, as reported in other genera, e.g. Senecio (Asteraceae; Coleman et al., 2003), Oligomeris (Resedaceae; Martin-Bravo et al., 2009) and Plantago (Plantaginaceae; Meyers and Liston, 2008).

The most striking cases presented are the colonization of Australia by E. aureum and the E. cygnorum group. Human introduction of these species potentially explains these distributions, but the divergence of E. aureum towards the end of the Pliocene (from nrITS reconstructions) and that of E. cygnorum towards the end of the relatively warm Miocene (Fiz et al., 2008, and the present study) pre-date human evolution, ruling out an anthropogenic introduction to Australia. This leaves long-distance dispersal as the most plausible explanation, especially as this type of colonization between the northern Hemisphere and Australia has been described previously [Ceratocephala (Garnock-Jones, 1984) and Scleranthus (Smissen 2003; Besnard et al., 2009)]. Erodium aureum is closely related to a number of species with broad Mediterranean distributions indicating a propensity for dispersal, also seen in the relatives of E. geoides. Additionally, the occurrence of seeds of E. cygnorum in the stomach of an Australian babbler (Ridley, 1930) also suggests a potential for bird-mediated long-distance dispersal in this clade.

Although the Mediterranean Basin harbours the largest diversity of Erodium (Fiz et al., 2006), many species are widespread, and four species have cosmopolitan distributions. Introductions due to human migration, such as that reported for E. cicutarium to California by Spanish missionaries in 1769 (Mensing and Byrne, 2003), may explain some of these broad distributions. Different features may also explain the vast distribution of the genus Erodium. In Israel, Zeide (1976) observed long-distance wind dispersal of mericarps of a desert species of subgenus Erodium (E. crassifolium). Most species in subgenus Erodium have plumose fruits (Fiz et al., 2006), which may have facilitated the dispersal of this early lineage through anemochory. According to Stamp (1989) and Van Rheede and Van Rooyen (1999), Erodium species exhibit a combination of ballistic primary dispersal followed by hygroscopic secondary dispersal called trypanospermy. These authors suggested that the distances covered by this type of dispersal are short, and plant-colonization success depends on hygroscopic burial of mericarps. Also, Stamp (1989) stated that Erodium diaspores never dispersed >3 m in experimental conditions. It is difficult to reconcile this dispersal pace with a colonization of America or Australia from the Mediterranean Basin. Alternatively, Shmida and Ellner (1983) found a large number of E. malacoides diaspores attached to sheep wool in the Mediterranean chaparral of Israel. Thus phenomena such as epizoochory and endozoochory (E. cygnorum, see above) may operate in Erodium and potentially explain its present worldwide distribution.

Two periods of Mediterranean radiations in Erodium

The present study revealed two periods of significant increases in diversification rate within Erodium. As diversification in lineages occupying semi-arid regions outside the Mediterranean is minimal, the causes of these shifts in diversification are certainly linked to factors only affecting the lineages of the Mediterranean Basin. The most recent radiation was detected at approx. 0·69 MYA. This is coincident with the onset of the largest Mediterranean glaciation periods (0·65 MYA; Médail and Diadema, 2009), possibly indicating a role of isolation in glacial refugia as a major promoter of speciation and diversification in Erodium. Médail and Diadema (2009) identified 52 glacial refugia scattered throughout the Mediterranean Basin, 33 of which are located in montane or submontane regions. These data might explain the high level of species diversity and endemism for Erodium in the Mediterranean Basin. This is particularly evident in montane areas; different species of subgenus Barbata (e.g. E. boissieri and E. trifolium) are endemic to mountains from Iberian Peninsula and Morocco. The core of subgenus Barbata harbours over 70 % of the species in Erodium and it is mainly distributed in the western Mediterranean, the Iberian Peninsula and western North Africa, which were connected several times during the Quaternary (e.g. the Messinian Salinity Crisis, 5 MYA; Bocquet et al., 1978).

The temporal analyses indicate that the second, more ancient shift occurs specifically in two lineages, the core of subgenus Barbata (section Cicutaria, section Malacoidea and subsection Petraea) and subsection Absinthioidea, in the last approx. 3 MYA. These groups diversified over the last 1–6 MYA (nrITS and rbcL), a period when seasonal drought was already present. Definitive stabilization of the summer drought in the Mediterranean Basin was at approx. 2·8 MYA (Suc, 1984). The data presented here could be interpreted as evidence of ancient divergence associated with adaptation to the onset of seasonal drought (see also Fiz et al., 2006). The Mediterranean Basin contains the highest percentage of annual species (31–54 %) among the five Mediterranean climates (Fiz et al., 2002), and this strategy is common in Erodium. Over 32 % of Erodium species are autogamic annuals (Fiz et al., 2008). Long-lasting disturbance regimes are associated with annuality in Mediterranean floras (Shmida, 1981; Pons and Quezel, 1985; Herrera, 1991). Erodium has the greatest number of species growing in disturbed habitats in Geraniaceae (Fiz et al., 2006). Selfing dominates in the 22 annual Erodium species, which are widely distributed, suggesting a possible correlation with the ability to colonize new areas (see Fiz et al., 2006). The colonization and radiation of Erodium in the Mediterranean could also have been facilitated by acquisition of a new set of pollinators (Fiz et al., 2008). These authors identified an association between a shift to generalist pollinators and higher reproductive flexibility with the colonization of disturbed habitats in the Mediterranean Basin (Fiz et al., 2008). This evidence provides some insight into how this radiation may have been an adaptive response to regular drought in the Mediterranean Basin.

Subsection Absinthioidea arose from an Asian ancestor, and the earliest diverging species (E. hoeftianum and E. ciconium) inhabit disturbed sites from the east Mediterranean Basin to central Asia. In general, species of subsection Absinthioidea inhabit mountain crevices, shrublands and meadows from Turkey, Caucasus and Greece, with distributions matching the SW Asia Tertiary refugia (Tiffney and Manchester, 2001). This link to Tertiary flora refugia is also supported by the divergence from its sister group at the end of the Miocene warm period. However, this radiation may have happened after diversification of the core group approx. 3 MYA (nrITS and rbcL) in agreement with diversification rate shift, and so it is unclear whether it is a tertiary lineage or it colonized these Tertiary refugia as a response to the start of Mediterranean climate. The two main features of this subsection are that dioecy is common (>70 % of the group) and that many species are endangered. It is common for clades of angiosperms to show low levels of dioecy compared with their sister taxa (Heilbuth, 2000; Vamosi and Vamosi, 2005). However, high proportions of dioecy have been found in rapid radiations of the flora of Hawaii (Sakai et al., 1995) and of the dioecious group of Momordica (Cucurbitaceae; Schaefer and Renner, 2010). Furthermore, the dioecious group within Gaertnera (Rubiaceae) has been reported to have the highest rate of diversification and of nucleotide substitution (Malcomber, 2002). A striking feature of subsection Absinthioidea is that it has the highest rate of nucleotide substitution, not only within Erodium (Fiz, 2005; Guisinger et al., 2008), but also within Geraniaceae (Guisinger et al., 2008). This is particularly remarkable as Geraniaceae has one of the fastest substitution rates for organellar genomes within the angiosperms (Parkinson et al., 2005; Bakker et al., 2006; Guisinger et al., 2008). A close relationship between rates of nucleotide substitution, rates of diversification and dioecy can be inferred for Erodium subsection Absinthioidea. Dioecy may provide an escape from the negative effects of inbreeding (Baker and Cox, 1984; Freeman et al., 1997), a useful strategy in isolated refugia.

Conclusions

The evidence presented here clearly demonstrates four cases for which the current distribution of sister species or clades can only be explained by intercontinental dispersal events between regions with Mediterranean-type climates.

Although Erodium has fewer species than most other Geraniaceae (e.g. Geranium or Pelargonium), it can be seen that it has undergone significant shifts in diversification that are probably adaptive responses to radical changes in climate. Acquisition of some traits, such as selfing, annuality and fruit-dispersal structures, may have been crucial key innovations for the successful colonization of arid steppes and dry habitats, whereas dioecy may have allowed colonization of mountane crevices in subsection Absinthioidea. These adaptive responses may have allowed the survival and diversification of the genus during the onset of Mediterranean drought regimes and the Quaternary and Tertiary glaciation periods. Confirmation of correlations between the particular traits and the environments suggested here, and reconstructions of trait variation on the phylogenetic trees could potentially reveal that the rapidly radiated sections within Erodium have, in fact, undergone adaptive radiations.

SUPPLEMENTARY DATA

Supplementary data are available online at www.aob.oxfordjournals.org and consist of the following files. Table S1: Matrix of weights of dispersal events between areas used for Lagrange analyses. Fig. S1: Divergence times and 95 % HPD retrieved from BEAST on the ITS tree.

ACKNOWLEDGEMENTS

We thank I. Gillespie and N. López, for help with specimens and literature and two anonymous reviewers for their helpful comments. We are also grateful to the herbarium curators and collectors for the material used for DNA analyses. This work was partly financed by the Spanish Dirección General de Investigación Científica y Técnica (DGICYT) through the research project REN2000-0818/GLO and REN2003-04397/GLO.

LITERATURE CITED

APG III
An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG III
Botanical Journal of the Linnean Society
 , 
2009
, vol. 
161
 (pg. 
105
-
121
)
Baker
HG
Cox
PA
Further thoughts on dioecism and islands
Annals of the Missouri Botanical Garden
 , 
1984
, vol. 
71
 (pg. 
244
-
253
)
Bakker
FT
Breman
F
Merckx
V
DNA sequence evolution in fast evolving mitochondrial DNA nad1 exons in Geraniaceae and Plantaginaceae
Taxon
 , 
2006
, vol. 
55
 (pg. 
887
-
896
)
Beier
BA
Nylander
JA
Chase
MW
Thulin
M
Phylogenetic relationships and biogeography of the desert plant genus Fagonia (Zygophyllaceae), inferred by parsimony and Bayesian model averaging
Molecular Phylogenetics and Evolution
 , 
2004
, vol. 
33
 (pg. 
91
-
108
)
Besnard
G
Rubio de Casas
R
Christin
PA
Vargas
P
Phylogenetics of Olea (Oleaceae) based on plastid and nuclear ribosomal DNA sequences: Tertiary climatic shifts and lineage differentiation times
Annals of Botany
 , 
2009
, vol. 
104
 (pg. 
143
-
160
)
Bocquet
G
Widler
B
Kiefer
H
The Messinian model: a new outlook for the floristics and systematics of the Mediterranean area
Candollea
 , 
1978
, vol. 
33
 (pg. 
269
-
287
)
Bromham
L
Penny
D
Rambaut
A
Hendy
MD
The power of relative rates tests depends on the data
Journal of Molecular Evolution
 , 
2000
, vol. 
50
 (pg. 
296
-
301
)
Campbell
CS
Wojciechowski
MF
Baldwin
BG
Alice
LA
Donoghue
MJ
Persistent nuclear ribosomal DNA sequence polymorphism in the Amelanchier agamic complex (Rosaceae)
Molecular Biology and Evolution
 , 
1997
, vol. 
14
 (pg. 
81
-
90
)
Chan
KMA
Moore
BR
SYMMETREE: whole-tree analysis of differential diversification rates
Bioinformatics
 , 
2005
, vol. 
21
 (pg. 
1709
-
1710
)
Coleman
M
Liston
A
Kadereit
JW
Abbott
RJ
Repeat intercontinental dispersal and Pleistocene speciation in disjunct Mediterranean and desert Senecio (Asteraceae)
American Journal of Botany
 , 
2003
, vol. 
90
 (pg. 
1446
-
1454
)
Drummond
AJ
Rambaut
A
BEAST: Bayesian evolutionary analysis by sampling trees
BMC Evolutionary Biology
 , 
2007
, vol. 
7
 pg. 
214
 
Fiz
O
Systematics and evolution of the family Geraniaceae: phylogenetic relationships of the genus
 , 
2005
Spain
Universidad Autonoma de Madrid
 
Erodium. PhD Thesis
Fiz
O
Valcárcel
V
Vargas
P
Phylogenetic position of Mediterranean Astereae and character evolution of daisies (Bellis, Asteraceae) inferred from nrDNA ITS sequences
Molecular Phylogenetics and Evolution
 , 
2002
, vol. 
25
 (pg. 
157
-
171
)
Fiz
O
Vargas
P
Alarcón
ML
Aldasoro
JJ
Phylogenetic relationships and evolution in Erodium (Geraniaceae) based on trnL-trnF sequences
Systematic Botany
 , 
2006
, vol. 
31
 (pg. 
739
-
763
)
Fiz
O
Vargas
P
Alarcón
ML
Aedo
C
García
JL
Aldasoro
JJ
Phylogeny and historical biogeography of Geraniaceae in relation to climate changes and pollination ecology
Systematic Botany
 , 
2008
, vol. 
33
 (pg. 
326
-
342
)
Freeman
DC
Dust
JL
EL-Keblawya
A
Miglia
KJ
Durant
E
Sexual specialization and inbreeding avoidance in the evolution of dioecy
Botanical Review
 , 
1997
, vol. 
63
 (pg. 
65
-
92
)
Fuertes Aguilar
J
Nieto Feliner
G
Additive polymorphisms and reticulation in an ITS phylogeny of thrifts (Armeria, Plumbaginaceae)
Molecular Phylogenetics and Evolution
 , 
2003
, vol. 
28
 (pg. 
430
-
447
)
Garnock-Jones
PJ
Ceratocephalus. New Zealand Journal of Botany
 , 
1984
, vol. 
22
 (pg. 
135
-
137
)
Goertzen
LR
Cannone
JJ
Gutell
RR
Jansen
RK
ITS secondary structure derived from comparative analysis: implications for sequence alignment and phylogeny of the Asteraceae
Molecular Phylogenetics and Evolution
 , 
2003
, vol. 
29
 (pg. 
216
-
234
)
Guisinger
MM
Kuehl
JV
Boore
JL
Jansen
RK
Genome-wide analyses of Geraniaceae plastid DNA reveal unprecedented patterns of increased nucleotide substitutions
Proccedings of the National Academy of Sciences, USA
 , 
2008
, vol. 
105
 (pg. 
18424
-
18429
)
Guittonneau
GG
Voster
P
Taxonomy, ecology and phylogeny of genus Erodium LHer. in the Mediterranean region
Proceedings of the International Geraniaceae Symposium
 , 
1990
South Africa
Stellenbosch University
(pg. 
71
-
91
)
Heilbuth
JC
Lower species richness in dioecious clades
American Naturalist
 , 
2000
, vol. 
156
 (pg. 
221
-
241
)
Herrera
J
Reproductive characteristics of a stable-dune therophyte community in Doñana (southern Spain)
Israel Journal of Botany
 , 
1991
, vol. 
40
 (pg. 
295
-
304
)
Hohmann
S
Kadereit
JW
Kadereit
G
Understanding Mediterranean-Californian disjunctions: molecular evidence from Chenopodiaceae-Betoideae
Taxon
 , 
2006
, vol. 
55
 (pg. 
67
-
78
)
Kers
LE
Contributions to a revision of Monsonia (Geraniaceae)
Botaniska Notiser
 , 
1968
, vol. 
121
 (pg. 
44
-
50
)
Malcomber
ST
Phylogeny of Gaertnera Lam. (Rubiaceae) based on multiple DNA markers: evidence of a rapid radiation in a widespread, morphologically diverse genus
Evolution
 , 
2002
, vol. 
56
 (pg. 
42
-
57
)
Martín-Bravo
S
Vargas
P
Luceño
M
Is Oligomeris (Resedaceae) indigenous to North America? Molecular evidence for a natural colonization from the Old World
American Journal of Botany
 , 
2009
, vol. 
96
 (pg. 
507
-
518
)
Médail
F
Diadema
K
Glacial refugia influence plant diversity patterns in the Mediterranean Basin
Journal of Biogeography
 , 
2009
, vol. 
36
 (pg. 
1333
-
1345
)
Mensing
S
Byrne
R
Pre-mission invasion of Erodium cicutarium in California
Journal of Biogeography
 , 
2003
, vol. 
25
 (pg. 
757
-
762
)
Meyers
SM
Liston
A
The biogeography of Plantago ovata Forssk. (Plantaginaceae)
International Journal of Plant Sciences
 , 
2008
, vol. 
169
 (pg. 
954
-
962
)
Milne
RI
Northern Hemisphere plant disjunctions: a window on tertiary land bridges and climate change?
Annals of Botany
 , 
2006
, vol. 
98
 (pg. 
465
-
72
)
Moore
BR
Chan
KMA
Donoghue
MJ
Bininda-Emonds
ORP
Detecting diversification rate variation in supertrees
Phylogenetic supertrees: combining information to reveal the tree of life
 , 
2004
Dordrecht
Kluwer
(pg. 
487
-
533
)
Mort
ME
Soltis
PS
Soltis
DE
Mabry
M
Comparison of three methods for estimating internal support on phylogenetic trees
Systematics Biology
 , 
2000
, vol. 
49
 (pg. 
160
-
171
)
Nylander
JAA
Ronquist
F
Huelsenbeck
JP
Nieves-Aldrey
J-L
Bayesian phylogenetic analysis of combined data
Systematic Biology
 , 
2004
, vol. 
53
 (pg. 
1
-
21
)
Parkinson
CL
Mower
JP
Qiu
Y-L
, et al.  . 
Multiple major increases and decreases in mitochondrial substitution rates in the plant family Geraniaceae
BMC Evolutionary Biology
 , 
2005
, vol. 
5
 pg. 
73
 
Pons
A
Quézel
P
Gomez-Campo
C
The history of the flora and vegetation and past and present human disturbance in the Mediterranean region
Plant conservation in the Mediterranean area. Geobotany 7
 , 
1985
Dordrecht
Kluwer
(pg. 
25
-
43
)
Posada
D
Crandall
KA
Modeltest: testing the model of DNA substitution
Bioinformatics
 , 
1998
, vol. 
14
 (pg. 
817
-
818
)
Rabosky
DL
LASER: a maximum likelihood toolkit for detecting temporal shifts in diversification rates from molecular phylogenies
Evolutionary Bioinformatics Online
 , 
2006
, vol. 
2
 (pg. 
257
-
260
)
Ree
RH
Smith
SA
Maximum likelihood inference of geographic range evolution by dispersal, local extinction, and cladogenesis
Systematic Biology
 , 
2008
, vol. 
57
 (pg. 
4
-
14
)
Ree
RH
Moore
BR
Webb
CO
Donoghue
MJ
A likelihood framework for inferring the evolution of geographic range on phylogenetic trees
Evolution
 , 
2005
, vol. 
59
 (pg. 
2299
-
2311
)
Ridley
HN
The dispersal of plants throughout the world
 , 
1930
 
Ashford, Kent: L. Reeve & Co
Ronce
O
How does it feel to be like a rolling stone? Ten questions about dispersal evolution
Annual Review of Ecology, Evolution, and Systematics
 , 
2007
, vol. 
38
 (pg. 
231
-
253
)
Ronquist
F
DIVA version 1·1. Computer program and manual available by anonymous FTP from Uppsala University
1996
Ronquist
F
Huelsenbeck
JP
MrBayes 3: Bayesian phylogenetic inference under mixed models
Bioinformatics
 , 
2003
, vol. 
19
 (pg. 
1572
-
1574
)
Rutschmann
F
Torsten
E
Kamariah
A
Conti
E
Assessing calibration uncertainty in molecular dating: the assignment of fossils to alternative calibration points
Systematic Biology
 , 
2007
, vol. 
56
 (pg. 
591
-
608
)
Sakai
AK
Wagner
WL
Ferguson
DM
Herbst
DR
Origins of dioecy in the Hawaiian flora
Ecology
 , 
1995
, vol. 
76
 (pg. 
2517
-
2529
)
Sang
T
Crawford
D
Stuessy
T
Documentation of reticulate evolution in peonies (Paeonia) using internal transcribed spacer sequences of nuclear ribosomal: implications for biogeography and concerted evolution
Proceedings of the National Academy of Sciences, USA
 , 
1995
, vol. 
92
 (pg. 
6813
-
6817
)
Schaefer
H
Renner
SS
A three-genome phylogeny of Momordica (Cucurbitaceae) suggests seven returns from dioecy to monoecy and recent long-distance dispersal to Asia
Molecular Phylogenetics and Evolution
 , 
2010
, vol. 
54
 (pg. 
553
-
560
)
Shmida
A
Mediterranean vegetation in California and Israel: similarities and differences
Israel Journal of Botany
 , 
1981
, vol. 
30
 (pg. 
105
-
123
)
Shmida
A
Ellner
S
Seed dispersal on pastoral grazers in open Mediterranean chaparral, Israel
Israel Journal of Botany
 , 
1983
, vol. 
32
 (pg. 
147
-
159
)
Slowinski
JB
Guyer
C
Testing whether certain traits have caused amplified diversification: an improved method based on a model of random speciation and extinction
American Naturalist
 , 
1993
, vol. 
142
 (pg. 
1019
-
1024
)
Smissen
RD
Phylogenetic analysis of ITS sequences suggests a Pliocene origin for the bipolar distribution of Scleranthus (Caryophyllaceae)
Australian Systematic Botany
 , 
2003
, vol. 
16
 (pg. 
301
-
315
)
Soria-Hernanz
D
Fiz-Palacios
O
Braverman
JM
Hamilton
MB
Reconsidering the generation time hypothesis based on nuclear ribosomal ITS sequence comparisons in annual and perennial angiosperms
BMC Evolutionary Biology
 , 
2008
, vol. 
8
 pg. 
344
 
Stamp
NE
Seed dispersal of four sympatric grassland annual species of Erodium
Journal of Ecology
 , 
1989
, vol. 
77
 (pg. 
1005
-
1020
)
Suc
JP
Origin and evolution of the Mediterranean vegetation and climate in Europe
Nature
 , 
1984
, vol. 
307
 (pg. 
429
-
432
)
Sun
Y
Skinner
DZ
Liang
GH
Hulbert
SH
Phylogeneticanalysis of Sorghum and related taxa using internal transcribed spacer of nuclear ribosomal DNA
Theoretical and Applied Genetics
 , 
1994
, vol. 
89
 (pg. 
26
-
32
)
Swofford
DL
PAUP*: phylogenetic analysis using parsimony
 , 
1999
Champaign, IL
Illinois Natural History Survey
 
Version 4·0
Tiffney
BH
Manchester
SR
The use of geological and paleontological evidence in evaluating plant phylogeographic hypotheses in the Northern Hemisphere Tertiary
International Journal of Plant Sciences
 , 
2001
, vol. 
162
 (pg. 
S3
-
S17
)
Thompson
JD
Higgins
DG
Gibson
TJ
CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalities and weight matrix choice
Nucleic Acids Research
 , 
1994
, vol. 
22
 (pg. 
4673
-
4680
)
Valente
LM
Savolainen
V
Vargas
P
Unparalleled rates of species diversification in Europe
Proceedings of the Royal Society. B. Biological Sciences
 , 
2010
, vol. 
277
 (pg. 
1489
-
1497
)
Vamosi
JC
Vamosi
SM
Present day risk of extinction may exacerbate the lower species richness of dioecious clades
Diversity and Distributions
 , 
2005
, vol. 
11
 (pg. 
25
-
32
)
Van Rheede
K
Van Rooyen
MW
Dispersal biology of desert plants
 , 
1999
Berlin
Springer-Verlag
Venter
HJT
Phytogeography and interspecies relationships in Monsonia L. (Geraniaceae)
Bothalia
 , 
1983
, vol. 
14
 (pg. 
865
-
869
)
White
TJ
Bruns
T
Lee
S
Taylor
J
Innis
M
Geldfand
D
Sninsky
J
White
T
Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics
PCR protocols: a guide to methods and applications
 , 
1990
San Diego, CA
Academic Press
(pg. 
315
-
322
)
Whittall
J
Liston
A
Gisler
S
Meinke
RJ
Detecting nucleotide additivity from direct sequences is a SNAP: an example from Sidalcea (Malvaceae)
Plant Biology
 , 
2000
, vol. 
2
 (pg. 
211
-
217
)
Zeide
B
Dispersal patterns in Erodium hirtum Willd
Israel Journal of Botany
 , 
1976
, vol. 
25
 (pg. 
221
-
224
)

APPENDIX

nrITS accessions provided for individuals used in the present work, including their locality, voucher and herbarium and GenBank accession numbers. trnL-F sequences are taken from a previous work (Fiz et al., 2006).

Taxa Geographical origin Voucher sample ITS Genbank accession no(s) trnL-F Genbank accession no(s) 
California 
macrophyllumUSA, California, Riverside Co., Murrieta Region: Skink Hollow, Santa Gertrudis Creek Drainage J. Easton s.n. (MA) EF185337 DQ072015 
macrophyllumUSA, California, Riverside Co., Temescal Valley, 0·9 miles SE of Indian Truck Trail and 30 m south of De Palma Ra. I. Gillespie 10 (MA) EF185338 – 
Erodium 
absinthioides Turkey, Bursa, Uludag G. Nieto Feliner 1580 (MA-393124) EF185348 DQ072034 
acauleItaly, Sicily, Palermo, La Pizzuta, Portella della Paglia C. Aedo & al. 5677 (MA-646287) EF185392 DQ072089 
aguilellae Cultivated in MA, from seeds collected in Castellón, Onda, Sitjar J. Aldasoro 2826 (MA) EF185401 DQ072090 
alnifolium Tunisia, Nefta, 5 km to Segename J. Aldasoro 2865 (MA) EF185391 DQ072064 
alpinumItaly, Abruzzo, pendici del Mt Rosa Pinnola, Bisegna, L'Aquila F. Conti 1656 (MA) EF185352 DQ072029 
alpinumItaly C. aedo CA8129 (MA) EF185351 – 
antarienseMorocco, Alto Atlas, Tizi-n-Aït-Hamed J. Güemes 1549 (MA) EF185373 DQ072078 
antarienseMorocco Staudinger 4800 EF185374 – 
arborescensTunisia, Skhira J. Aldasoro 3053 (MA) EF185340 DQ072018 
arborescensCultivated in MA, from seeds collected in Israel, Nahal Yarqon J. Aldasoro 3488 (MA) EF185341 – 
asplenioides Tunisia A2935(1) EF185390 DQ072065 
astragaloides Spain, Granada, Dilar, Trevenque, Los Alayos C. Navarro & al. 2246 (MA-625117) EF185400 DQ072091 
aureum Australia, Coolgardie, Eyre Higway, 59 km W of Madura B. Archer 15 (MEL-2039223) EF185385 DQ072066 
beketowi Ukraine, Biespars, Stavropol Smababanova s.n. (LE) EF185358 DQ072030 
boissieri Spain, Granada, La Zubia, Cortijo de la Cortichuela, Trevenque M. Velayos & Navarro 9676 (MA-644606) EF185378 DQ072054 
botrys USA, California, San Francisco, Mt Tamalpais S. Castroviejo & al. 14575 (MA-590950) EF185365 DQ072049 
brachycarpum Spain, Madrid, Rozas de Puerto Real N. López 499 (MA) EF185366 DQ072050 
carolinianum Australia, Olympic Dam Mine, Gairdner-Torrens F.J. Badman 3597 (MA-592447) EF185364 DQ072046 
carvifolium 2 Spain, La Rioja, Montenegro de Cameros, N Puerto de Santa Inés P. Vargas 230PV99 (MA) EF185399 DQ072094 
cazorlanumSpain, Jaen, Sierra de Cazorla, Cortijo de la Cabrilla C. Navarro & Benavente 3025 (MA-628379) EF185404 DQ072097 
cazorlanumSpain, Jaen MA580102 EF185403 – 
cedrorum Cultivated in MA from seeds collected in Bolkar Daglari, Nigde, Turkey J. Aldasoro 3489 (MA) – DQ072031 
chium Spain, Cádiz, Monte Tavirana, Ronda C. Navarro 3450 MA EF185384 DQ072067 
chrysanthum Greece, Peloponeso, Killíni, N a NE-Siete des Gipfelmassiv E. Hörandl & F. Hadacek 7612 (W) EF185361 DQ072032 
ciconiumItaly, Abruzzo, L'Aquila, pr. Santo Stéfano de Sessanio C. Aedo & al. 8108 (MA) EF185354 DQ072039 
ciconiumGreece, Fthiotis, Othrys, Dhivri, 1991 Willing14594 EF185355 – 
cicutariumCalifornia CA4439 EF185393 – 
corsicumItaly, Sardinia, Santa Teresa de Gallura, Capo Testa C. Aedo & al. 9120 (MA-702081) EF185380 DQ072061 
cossoniiMorocco, Haut Atlas, Tiz-n- Test J. Fernández Casas & al. 3277 (MA-252363) – DQ072073 
cossoniiMorocco Staudinger2852  EF185388 
crassifoliumTunisia, Coutinedes, cerca de Gabes Aldasoro 3069 (MA) EF185339 DQ072020 
cygnorumCultivated in MA from seeds collected in Great Victoria Desert, Camp Aldasoro 2842 (MA) EF185362 DQ072044 
cygnorumCultivated in MA MEL1580207 EF185363 – 
daucoidesSpain, Palencia, Velilla del río Carrión, Peña Cueto C. Navarro & al. 1602 (MA-559982) EF185402 DQ072096 
foetidum subsp. celtibericum Spain, Tarragona, ports de Beseit, ĹEngrillo LL. Sáez s.n. (MA) EF185375 DQ072081 
foetidum subsp. cheilanthifolium Spain, Granada, Sierra de Arana, Cueva del Agua P. Vargas 100PV00 (MA) EF185372 DQ072080 
foetidum subsp. foetidum Spain, Gerona, Cabo Norfeu, Rosas, Gerona C. Aedo et al. 4920 (MA) – DQ072079 
gaillardotti Turkey, Malatya, Darende, 27 km de Gürün to Darende F. Muñoz-Garmendia & al. 4567 (MA) EF185353 DQ072035 
geoides Chile, Coquimbo, Choapa province, 1 km N of the border of Petarca province Taylor 10620 (MO) – DQ072048 
glandulosumSpain, Leon, Puente de la Palanca C. Aedo & Patallo 4451 (MA-621226) EF185367 DQ072082 
glaucophyllumSpain, Barcelona, Montcau, St Llorenç de Munt LL. Saéz 5001 (MA) – DQ072083 
gruinumJordania, Gerassa (Jerash) P. Vargas (MA) EF185359 DQ072037 
gruinumCultivated from Iran Davis56574 EF185360 – 
guicciardii Cultivated in MA from seeds collected in Ohrid, Macedonia, Aldasoro 2842 EF185356 DQ072036 
guttatumTunisia, Feriana J. Aldasoro 2973 (MA) EF185344 DQ072024 
hoefftianumTurkey, Göreme, Ask Vadisi, dept. Nevsehir F. Muñoz-Garmendia & al. 4626 (MA) EF185349 DQ072033 
hoefftianumIran J. Aldasoro 10000 (MA) EF185350 – 
jahandiezianum Morocco, Anti-Atlas, Igherm F. Gómiz s.n. (BC). EF185342 DQ072022 
janszii Australia, Far Western Plains, near Mt Robe, 35 km of Broken Hill M.G. Corrick 7271 (MEL-592017) – DQ072047 
laciniatumLetur, Albacete I. Alvarez 1239 (MA-591697) EF185387 DQ072071 
laciniatumPalestine, Petra, Jordania P.Vargas (MA) EF185386 – 
lucidum Cultivated in MA, from seeds taken in Huesca, Aneto J. Aldasoro 2821(MA) EF185370 DQ072084 
macrocalyx Spain, Cuenca, Tragacete C. Navarro 2469 (MA) EF185405 DQ072087 
malacoidesAustralia, Volcanic plain, SE Organ Pipes, S side of Jacksons Creek V. Stajsic 852 (MEL-2020988) EF185383 DQ072070 
manescavi Cultivated in MA, from seeds taken in Valle de Ossau, France J. Aldasoro 2829 (MA) EF185398 DQ072098 
maritimumCultivated in MA, from seeds taken in Devon, Great Britain J. Aldasoro 905 (MA-614528) – DQ072057 
maritimumSpain, A Coruña, San Andres de Teixido MAL129(2) EF185379 – 
moschatumAustralia, Southern Lofty, Angas River, Strathalbyn N.M. Smith 2393 (MEL-1621233) EF185409 DQ072086 
moschatumMorocco, Chefchaouen, Campsite O. Fiz 152of00 EF185408 – 
mouretii Spain, Alange, Castillo M.A. Moreno 9 (MA-643352) EF185414 DQ072099 
nervulosum Morocco, Ifrane to Inmouzer Mateos and Montserrat 6038 (BC-826634) – DQ072072 
oxyrhynchum Cultivated in MA from seeds collected in Egypt, Cairo-Suez Desert Road J. Aldasoro 3487 (MA) EF185343 DQ072023 
paularense Spain, Guadalajara, Cañamares, Atienza C. Aedo 4097 (MA-588866) EF185371 DQ072077 
pelargoniflorum Cultivated in MA from seeds collected in SE Turkey. J. Aldasoro 2838 (MA) EF185347 DQ072041 
recoderiiSpain, Málaga, Monte Tavirana, Ronda C. Navarro 3449 (MA-685241) EF185394 DQ072104 
recoderiiSpain, Cádiz, Pto Palomas JR4, 4-5-2000 EF185395 – 
reichardiiSpain, Baleares Islands, Palma de Mallorca, Lluc, collado de Massanella R. Morales et al.1831 (MA-618180) EF185377 – 
reichardiiCultivated in MA from seeds collected in Menorca, Cabo Favaritx J. Martinez 173JM03 EF185376 DQ072063 
rupestreSpain, Lérida, Pallars Jussa, Trem, Serra de Gurp, Roques de Codó C. Aedo & Pedrol 4782 (MA) EF185368 DQ072085 
rupestreBarcelona, Montserrat Ll. Saenz 17-III-2002 EF185369 – 
rupicolaSpain, Granada, Guejar Sierra, Barranco del Guarón, 30SVG4106469 M. Ruiz & S. Vidal (GDA-41392) EF185397 DQ072105 
rupicolaSpain, Almeria GDAC40231 EF185396 – 
ruthenicum Ukraine, Dniepopetrovskaia, Sabrilovka, SE Kiev Deryiova s.n. (LE) – DQ072042 
sanguis-christiSpain, Murcia, La Azohia, castillo C. Navarro & al. 1922 (MA-612356) EF185381 DQ072055 
sanguis-christiSpain, Castellón, Peñíscola, Barranco de la Torre Nova C. Fabregat & al. 51 (MA-580737) EF185382 – 
sebaceumMorocco, middle Atlas, Ben Smine, Azrou, 67 S of Meknes F. Damblon 82/36 (MA-596076) EF185407 DQ072103 
sebaceumMorocco, Boumia, 8 km NW of Er-Rachidia Podlech 43213 (MA-464889) EF185406 – 
stephanianum China, Qinghai, Nangqên Xian, NW of Jangkar, E side of Za Qu (upper Mekong), on road between Jangkar and Yushu Ho & al. 2892 (MO). EF185346 DQ072027 
tataricum Russia, Jakasia, Payon, Ust-Bior M. Voroniena s.n. (LE) – DQ072028 
texanum Cultivate in MA, from seeds taken in Yavapai Co., Arizona, USA J. Aldasoro 3492 (MA) EF185345 DQ072026 
tordylioidesSpain, Huesca, Agüero, Los Mallos C. Navarro 3485 MA EF185412 DQ072101 
tordylioidesMorocco Saudinger 2856 EF185413 – 
touchyanumMorocco, Sk-el-Had-de-Reggada J. Arrington & al (MA-654483) EF185411 DQ072088 
touchyanumIran E91240 EF185410 – 
trichomanifolium Turkey, Palandoken Dag, Erzürüm A. Herrero 1705 (MA) EF185357 DQ072040 
trifoliumTunisia, Rohnia a Maktar, 30 km of Rohnia J. Aldasoro 2936 (MA) EF185389 DQ072075 
Geranium biuncinatum Jebel Burá, between Hilla and Attuba J.R.I. Wood 3126 (MA648734) Yeo3 DQ525076 – 
Monsonia speciosa   AF505648 – 
Pelargonium zonale   DQ345326  
Geraniales: Melianthaceae 
Bersama lucens   DQ435401 – 
Melianthus elongatus   DQ435418 – 
Taxa Geographical origin Voucher sample ITS Genbank accession no(s) trnL-F Genbank accession no(s) 
California 
macrophyllumUSA, California, Riverside Co., Murrieta Region: Skink Hollow, Santa Gertrudis Creek Drainage J. Easton s.n. (MA) EF185337 DQ072015 
macrophyllumUSA, California, Riverside Co., Temescal Valley, 0·9 miles SE of Indian Truck Trail and 30 m south of De Palma Ra. I. Gillespie 10 (MA) EF185338 – 
Erodium 
absinthioides Turkey, Bursa, Uludag G. Nieto Feliner 1580 (MA-393124) EF185348 DQ072034 
acauleItaly, Sicily, Palermo, La Pizzuta, Portella della Paglia C. Aedo & al. 5677 (MA-646287) EF185392 DQ072089 
aguilellae Cultivated in MA, from seeds collected in Castellón, Onda, Sitjar J. Aldasoro 2826 (MA) EF185401 DQ072090 
alnifolium Tunisia, Nefta, 5 km to Segename J. Aldasoro 2865 (MA) EF185391 DQ072064 
alpinumItaly, Abruzzo, pendici del Mt Rosa Pinnola, Bisegna, L'Aquila F. Conti 1656 (MA) EF185352 DQ072029 
alpinumItaly C. aedo CA8129 (MA) EF185351 – 
antarienseMorocco, Alto Atlas, Tizi-n-Aït-Hamed J. Güemes 1549 (MA) EF185373 DQ072078 
antarienseMorocco Staudinger 4800 EF185374 – 
arborescensTunisia, Skhira J. Aldasoro 3053 (MA) EF185340 DQ072018 
arborescensCultivated in MA, from seeds collected in Israel, Nahal Yarqon J. Aldasoro 3488 (MA) EF185341 – 
asplenioides Tunisia A2935(1) EF185390 DQ072065 
astragaloides Spain, Granada, Dilar, Trevenque, Los Alayos C. Navarro & al. 2246 (MA-625117) EF185400 DQ072091 
aureum Australia, Coolgardie, Eyre Higway, 59 km W of Madura B. Archer 15 (MEL-2039223) EF185385 DQ072066 
beketowi Ukraine, Biespars, Stavropol Smababanova s.n. (LE) EF185358 DQ072030 
boissieri Spain, Granada, La Zubia, Cortijo de la Cortichuela, Trevenque M. Velayos & Navarro 9676 (MA-644606) EF185378 DQ072054 
botrys USA, California, San Francisco, Mt Tamalpais S. Castroviejo & al. 14575 (MA-590950) EF185365 DQ072049 
brachycarpum Spain, Madrid, Rozas de Puerto Real N. López 499 (MA) EF185366 DQ072050 
carolinianum Australia, Olympic Dam Mine, Gairdner-Torrens F.J. Badman 3597 (MA-592447) EF185364 DQ072046 
carvifolium 2 Spain, La Rioja, Montenegro de Cameros, N Puerto de Santa Inés P. Vargas 230PV99 (MA) EF185399 DQ072094 
cazorlanumSpain, Jaen, Sierra de Cazorla, Cortijo de la Cabrilla C. Navarro & Benavente 3025 (MA-628379) EF185404 DQ072097 
cazorlanumSpain, Jaen MA580102 EF185403 – 
cedrorum Cultivated in MA from seeds collected in Bolkar Daglari, Nigde, Turkey J. Aldasoro 3489 (MA) – DQ072031 
chium Spain, Cádiz, Monte Tavirana, Ronda C. Navarro 3450 MA EF185384 DQ072067 
chrysanthum Greece, Peloponeso, Killíni, N a NE-Siete des Gipfelmassiv E. Hörandl & F. Hadacek 7612 (W) EF185361 DQ072032 
ciconiumItaly, Abruzzo, L'Aquila, pr. Santo Stéfano de Sessanio C. Aedo & al. 8108 (MA) EF185354 DQ072039 
ciconiumGreece, Fthiotis, Othrys, Dhivri, 1991 Willing14594 EF185355 – 
cicutariumCalifornia CA4439 EF185393 – 
corsicumItaly, Sardinia, Santa Teresa de Gallura, Capo Testa C. Aedo & al. 9120 (MA-702081) EF185380 DQ072061 
cossoniiMorocco, Haut Atlas, Tiz-n- Test J. Fernández Casas & al. 3277 (MA-252363) – DQ072073 
cossoniiMorocco Staudinger2852  EF185388 
crassifoliumTunisia, Coutinedes, cerca de Gabes Aldasoro 3069 (MA) EF185339 DQ072020 
cygnorumCultivated in MA from seeds collected in Great Victoria Desert, Camp Aldasoro 2842 (MA) EF185362 DQ072044 
cygnorumCultivated in MA MEL1580207 EF185363 – 
daucoidesSpain, Palencia, Velilla del río Carrión, Peña Cueto C. Navarro & al. 1602 (MA-559982) EF185402 DQ072096 
foetidum subsp. celtibericum Spain, Tarragona, ports de Beseit, ĹEngrillo LL. Sáez s.n. (MA) EF185375 DQ072081 
foetidum subsp. cheilanthifolium Spain, Granada, Sierra de Arana, Cueva del Agua P. Vargas 100PV00 (MA) EF185372 DQ072080 
foetidum subsp. foetidum Spain, Gerona, Cabo Norfeu, Rosas, Gerona C. Aedo et al. 4920 (MA) – DQ072079 
gaillardotti Turkey, Malatya, Darende, 27 km de Gürün to Darende F. Muñoz-Garmendia & al. 4567 (MA) EF185353 DQ072035 
geoides Chile, Coquimbo, Choapa province, 1 km N of the border of Petarca province Taylor 10620 (MO) – DQ072048 
glandulosumSpain, Leon, Puente de la Palanca C. Aedo & Patallo 4451 (MA-621226) EF185367 DQ072082 
glaucophyllumSpain, Barcelona, Montcau, St Llorenç de Munt LL. Saéz 5001 (MA) – DQ072083 
gruinumJordania, Gerassa (Jerash) P. Vargas (MA) EF185359 DQ072037 
gruinumCultivated from Iran Davis56574 EF185360 – 
guicciardii Cultivated in MA from seeds collected in Ohrid, Macedonia, Aldasoro 2842 EF185356 DQ072036 
guttatumTunisia, Feriana J. Aldasoro 2973 (MA) EF185344 DQ072024 
hoefftianumTurkey, Göreme, Ask Vadisi, dept. Nevsehir F. Muñoz-Garmendia & al. 4626 (MA) EF185349 DQ072033 
hoefftianumIran J. Aldasoro 10000 (MA) EF185350 – 
jahandiezianum Morocco, Anti-Atlas, Igherm F. Gómiz s.n. (BC). EF185342 DQ072022 
janszii Australia, Far Western Plains, near Mt Robe, 35 km of Broken Hill M.G. Corrick 7271 (MEL-592017) – DQ072047 
laciniatumLetur, Albacete I. Alvarez 1239 (MA-591697) EF185387 DQ072071 
laciniatumPalestine, Petra, Jordania P.Vargas (MA) EF185386 – 
lucidum Cultivated in MA, from seeds taken in Huesca, Aneto J. Aldasoro 2821(MA) EF185370 DQ072084 
macrocalyx Spain, Cuenca, Tragacete C. Navarro 2469 (MA) EF185405 DQ072087 
malacoidesAustralia, Volcanic plain, SE Organ Pipes, S side of Jacksons Creek V. Stajsic 852 (MEL-2020988) EF185383 DQ072070 
manescavi Cultivated in MA, from seeds taken in Valle de Ossau, France J. Aldasoro 2829 (MA) EF185398 DQ072098 
maritimumCultivated in MA, from seeds taken in Devon, Great Britain J. Aldasoro 905 (MA-614528) – DQ072057 
maritimumSpain, A Coruña, San Andres de Teixido MAL129(2) EF185379 – 
moschatumAustralia, Southern Lofty, Angas River, Strathalbyn N.M. Smith 2393 (MEL-1621233) EF185409 DQ072086 
moschatumMorocco, Chefchaouen, Campsite O. Fiz 152of00 EF185408 – 
mouretii Spain, Alange, Castillo M.A. Moreno 9 (MA-643352) EF185414 DQ072099 
nervulosum Morocco, Ifrane to Inmouzer Mateos and Montserrat 6038 (BC-826634) – DQ072072 
oxyrhynchum Cultivated in MA from seeds collected in Egypt, Cairo-Suez Desert Road J. Aldasoro 3487 (MA) EF185343 DQ072023 
paularense Spain, Guadalajara, Cañamares, Atienza C. Aedo 4097 (MA-588866) EF185371 DQ072077 
pelargoniflorum Cultivated in MA from seeds collected in SE Turkey. J. Aldasoro 2838 (MA) EF185347 DQ072041 
recoderiiSpain, Málaga, Monte Tavirana, Ronda C. Navarro 3449 (MA-685241) EF185394 DQ072104 
recoderiiSpain, Cádiz, Pto Palomas JR4, 4-5-2000 EF185395 – 
reichardiiSpain, Baleares Islands, Palma de Mallorca, Lluc, collado de Massanella R. Morales et al.1831 (MA-618180) EF185377 – 
reichardiiCultivated in MA from seeds collected in Menorca, Cabo Favaritx J. Martinez 173JM03 EF185376 DQ072063 
rupestreSpain, Lérida, Pallars Jussa, Trem, Serra de Gurp, Roques de Codó C. Aedo & Pedrol 4782 (MA) EF185368 DQ072085 
rupestreBarcelona, Montserrat Ll. Saenz 17-III-2002 EF185369 – 
rupicolaSpain, Granada, Guejar Sierra, Barranco del Guarón, 30SVG4106469 M. Ruiz & S. Vidal (GDA-41392) EF185397 DQ072105 
rupicolaSpain, Almeria GDAC40231 EF185396 – 
ruthenicum Ukraine, Dniepopetrovskaia, Sabrilovka, SE Kiev Deryiova s.n. (LE) – DQ072042 
sanguis-christiSpain, Murcia, La Azohia, castillo C. Navarro & al. 1922 (MA-612356) EF185381 DQ072055 
sanguis-christiSpain, Castellón, Peñíscola, Barranco de la Torre Nova C. Fabregat & al. 51 (MA-580737) EF185382 – 
sebaceumMorocco, middle Atlas, Ben Smine, Azrou, 67 S of Meknes F. Damblon 82/36 (MA-596076) EF185407 DQ072103 
sebaceumMorocco, Boumia, 8 km NW of Er-Rachidia Podlech 43213 (MA-464889) EF185406 – 
stephanianum China, Qinghai, Nangqên Xian, NW of Jangkar, E side of Za Qu (upper Mekong), on road between Jangkar and Yushu Ho & al. 2892 (MO). EF185346 DQ072027 
tataricum Russia, Jakasia, Payon, Ust-Bior M. Voroniena s.n. (LE) – DQ072028 
texanum Cultivate in MA, from seeds taken in Yavapai Co., Arizona, USA J. Aldasoro 3492 (MA) EF185345 DQ072026 
tordylioidesSpain, Huesca, Agüero, Los Mallos C. Navarro 3485 MA EF185412 DQ072101 
tordylioidesMorocco Saudinger 2856 EF185413 – 
touchyanumMorocco, Sk-el-Had-de-Reggada J. Arrington & al (MA-654483) EF185411 DQ072088 
touchyanumIran E91240 EF185410 – 
trichomanifolium Turkey, Palandoken Dag, Erzürüm A. Herrero 1705 (MA) EF185357 DQ072040 
trifoliumTunisia, Rohnia a Maktar, 30 km of Rohnia J. Aldasoro 2936 (MA) EF185389 DQ072075 
Geranium biuncinatum Jebel Burá, between Hilla and Attuba J.R.I. Wood 3126 (MA648734) Yeo3 DQ525076 – 
Monsonia speciosa   AF505648 – 
Pelargonium zonale   DQ345326  
Geraniales: Melianthaceae 
Bersama lucens   DQ435401 – 
Melianthus elongatus   DQ435418 – 

Comments

0 Comments