Phylogeographical patterns in chloroplast DNA variation within the Acacia acuminata (Leguminosae: Mimosoideae) complex in Western Australia

Phylogeographical patterns in chloroplast DNA variation within the Acacia acuminata (Leguminosae: Mimosoideae) complex in Western Australia M. BYRNE, B. MACDONALD & D. COATES Science Division, Department of Conservation and Land Management, Locked Bag 104, Bentley Delivery Centre, WA, Australia

Keywords:

Abstract

Acacia; cpDNA; evolution; nested clade analysis; phylogeny; phylogeography.

The Acacia acuminata complex includes three taxa, A. acuminata ssp. acuminata, A. acuminata ssp. burkittii and A. oldfieldii, along with several informal variants of A. acuminata. It is widespread throughout southern Australia with the centre of diversity in south-west Western Australia. Phylogeographical patterns in the complex were investigated using a nested clade analysis of cpDNA RFLPs from 25 populations in Western Australia. Except for A. oldfieldii that was clearly identified as a distinct entity, haplotypes were not restricted to sub-specific taxa or variants within A. acuminata. There was significant association between phylogenetic position of many haplotypes and their geographical distribution. The fine-scale phylogeographical patterns were complex but at deeper levels in the phylogeny there was evidence of divergence between two lineages. The pattern of shared haplotypes between lineages suggests retention of ancestral polymorphism as a result of incomplete lineage sorting. The divergence of these lineages is consistent with fragmentation caused by climatic instability during the Pleistocene.

Introduction Phylogeographical analyses have, over the last decade, played a pivotal role in linking population genetics, phylogenetics and biogeography by using the spatial distribution of genealogical lineages to deduce the influence of historical processes on the evolution of populations and species (Avise, 2000). In plants, cpDNA has proven valuable in assessing historical processes influencing population structure on a broad scale (Schaal et al., 1998; Ennos et al., 1999; Avise, 2000; Schaal & Olsen, 2000). For example, recent cpDNA phylogeographical studies have indicated two glacial refugia in north-west America (Soltis et al., 1997), and have lead to the reconstruction of postglacial recolonization routes for oaks, beech and alder in Europe (Ferris et al., 1993; Demesure et al., 1996; Dumolin-Lape`gue et al., 1997; Correspondence: M. Byrne, CALM Science, W.A. Herbarium, Department of Conservation and Land Management, Locked Bag 104, Bentley Delivery Centre, WA 6983, Australia. Tel.: +618 93340503; fax: +618 93340515; e-mail: [email protected]

576

Ferris et al., 1998; King & Ferris, 1998). Phylogeographical studies of Australian plant species are also beginning to appear (Byrne & Moran, 1994; Butcher et al., 1995; Steane et al., 1998; Byrne et al., 1999; Jackson et al., 1999; McKinnon et al., 1999; Byrne & Macdonald, 2000), and are now leading to a much better understanding of the influence of historical processes on the geographical distribution of genetic variation and phylogenetic patterns in components of the Australian flora. The flora of the south-west corner of Australia is recognized for its remarkable diversity and endemism, and the diverse array of evolutionary patterns that combine refugial species in higher rainfall areas with fragmented relictual species, and suites of newly derived taxa, in the more arid areas (Hopper et al., 1996). Compared with many other floras, particularly in North America and Europe, this flora has persisted for an extremely long period, probably well back into the Cretaceous, without any large scale extinction episodes associated with glaciation. However, significant climatic and habitat instability has been experienced in this

J. EVOL. BIOL. 15 (2002) 576–587 ª 2002 BLACKWELL SCIENCE LTD

Phylogeography of A. acuminata

region since the late Tertiary leading to cyclic expansion and contraction of the mesic and arid zones (Hopper, 1979; Hopper et al., 1996). It is postulated that during this period many relictual taxa have become locally extinct, but have survived as fragmented remnants particularly through the semiarid transitional rainfall zone between the mesic and arid zones (see Hopper et al., 1996) and now persist in geographically restricted and fragmented distributions. Evidence for this comes primarily from isozyme studies in fragmented, geographically restricted species (see Coates, 2000) and has been supported by cpDNA studies in some species (Byrne et al., 1999, 2001b). These studies show significant levels of genetic divergence between populations and indicate that genetic processes associated with historical ecogeographical barriers to gene flow have led to the formation of distinct evolutionary lineages within many species. The presence of distinct intraspecific lineages has been observed in rare and geographically restricted species in south-west Australia (Byrne et al., 1999; Coates 2000; Byrne et al., 2001b), however, less is known about the patterns of genetic variation in more widespread species. There are several patterns that could be observed in widespread species. If widespread species have also been affected by historical range fragmentation and localized extinction caused by climatic fluctuations over extended periods of time, they would be expected to show similar patterns of intraspecific lineage divergence. This pattern has been observed in the widespread species Santalum spicatum which showed the presence of two divergent cpDNA lineages that were geographically separated (M. Byrne, unpublished data). In contrast widespread species which represent recent speciation and range expansion would be expected to show less genetic structuring in cpDNA patterns, such as that observed in the Eucalyptus kochii complex (Byrne, 1999; Byrne & Macdonald, 2000). However, genetic structuring which does not necessarily reflect current taxon boundaries is possible in an ancient landscape such as south-west Western Australia, where cpDNA variation may reflect historical processes prior to speciation. Alternatively a complex phylogeographical pattern that does not reflect taxon boundaries may be because of widespread hybridization. There are few discrete widely distributed species in south-west Australia and many species have regional ranges frequently occurring in allopatric replacement series with close relatives (Hopper et al., 1996). Acacia acuminata Benth. (Acacia section Juliflorae) is one such widespread species complex that occurs throughout southern Australia with a centre of diversity in southwest Western Australia. Acacia acuminata (common name Jam) has two described taxa which have received taxonomic treatment either as separate species, A. acuminata and A. burkittii (Maslin, 2001), or as subspecies, A. acuminata ssp. acuminata and A. acuminata

J. EVOL. BIOL. 15 (2002) 576–587 ª 2002 BLACKWELL SCIENCE LTD

577

ssp. burkittii (Kodela & Tindale, 1998). In this study, we will follow the Flora of Australia (Orchard & Wilson, 2001) treatment which treats them as subspecies. Within A. acuminata there are also two informal variants, narrow phyllode and small seeded forms, and some undefined variants with various combinations of morphological features. Acacia acuminata ssp. acuminata (a tree with wide flat phyllodes and long flower spikes) is largely confined to the south-west region of Western Australia and A. acuminata subsp burkittii (a shrub with terete phyllodes and short flower spikes) is widespread throughout the arid region of Western Australia and extending into eastern Australia. The narrow phyllode variant is found between the two subspecies and the small seeded variant is found in the north-west of the distribution near Geraldton. The undefined variants also occur in the Geraldton region. A study of the genetic diversity in the complex using isozymes showed low differentiation between subsp acuminata, the narrow phyllode variant and subsp burkittii but greater distinction between these taxa and the small seeded variant (Broadhurst et al., 2002). A second species, Acacia oldfieldii Muell. has been recognized in the complex and has a very restricted distribution north of Geraldton (Maslin & Pedley, 1982). Genetic diversity using isozymes showed A. oldfieldii to be genetically quite distinct (Broadhurst et al., 2002) but it will be included in this study as it has traditionally been included in the A. acuminata complex and will be useful to clarify the species relationships. This study investigates the phylogenetic and evolutionary relationships between populations and taxa in the A. acuminata complex using RFLP analysis of the chloroplast genome to determine the phylogeographical patterns within the species complex. Given the current taxonomic confusion within the complex this study also aimed to investigate whether genetic patterns reflected taxon ⁄ variant boundaries.

Materials and methods Plant collections and RFLP procedures Populations were sampled across the range of morphological variation in the A. acuminata complex, including A. acuminata ssp. acuminata, A. acuminata ssp. burkittii small seeded variant, narrow phyllode variant, the undefined variants and A. oldfieldii. Phyllode samples were collected from 25 populations and the distribution of taxa and location of sampled populations is shown in Fig. 1. Details of populations sampled are given in Table 1. Collections were made from five individuals from each of 30 populations, although only 25 populations were analysed in this study. Phyllode samples were also collected from two individuals of both A. ephedroides and A. anfractuosa for use as outgroups in the analysis. DNA was extracted from the phyllodes of the 129 individuals as described in Byrne et al. (1993)

578

M. BYRNE ET AL.

Fig. 1 Natural distribution of taxa within the Acacia acuminata complex and location of sampled populations (r) A. acuminata ssp. acuminata; (j) A. acuminata (narrow phyllode); (v) A. acuminata (small seed); (m) A. acuminata (undefined variants); («) A. acuminata ssp. burkittii; (J) A. oldfieldii.

except that 0.1 M sodium sulphite was added to the extraction buffer to prevent degradation of the DNA (Byrne et al. 2001a). For each sample 2 lg DNA was digested with six restriction enzymes (BclI, BglII, DraI, EcoRV, HindIII, XbaI) and hybridized with heterologous probes covering the majority of the chloroplast genome.

Eight petunia cpDNA probes were used, P1, P3, P4, P6, P8, P10, P12, P14 (details given in Sytsma & Gottlieb, 1986), plus three tobacco cpDNA probes, pTBa1, pTB22, pTB29 (Shinozaki et al., 1986; Suguira et al., 1986). Restriction digestion and hybridization were as described in Byrne & Moran (1994), and probe inserts

Table 1 Location and identity of populations of the Acacia acuminata complex sampled for chloroplast diversity analysis. Voucher specimens are deposited in the Western Australian Herbarium. No.

Provenance

Taxon

Lat. (S)

Long. (E)

Voucher no.

1 2 3 4 5 6 7 9 10 11 12 13 14 15 16 17 20 21 22 23 24 26 27 28 29

Moora Coorow Bolgart York Corrigin Borden Williams Mukinbudin Kalannie Latham Binnu East Yalgoo Murchison Binnu Mullewa Nerren Nerren Mingenew Wiluna Yalgoo Ninghan Goongarrie Lake Moore Paynes Find Murchison Hawkes Head

ssp. acuminata ssp. acuminata ssp. acuminata ssp. acuminata ssp. acuminata ssp. acuminata ssp. acuminata Narrow phyllode Narrow phyllode Narrow phyllode Small seed Small seed Undefined variant Undefined variant Undefined variant Undefined variant Undefined variant ssp. burkittii ssp. burkittii ssp. burkittii ssp. burkittii ssp. burkittii ssp. burkittii A. oldfieldii A. oldfieldii

3025¢00¢¢ 2934¢17¢¢ 3113¢22¢¢ 3200¢45¢¢ 3221¢37¢¢ 3404¢56¢¢ 3300¢59¢¢ 3043¢39¢¢ 3026¢07¢¢ 2942¢44¢¢ 2802¢28¢¢ 2822¢39¢¢ 2749¢24¢¢ 2803¢53¢¢ 2834¢36¢¢ 2711¢45¢¢ 2913¢16¢¢ 2640¢05¢¢ 2822¢39¢¢ 2924¢25¢¢ 3003¢¢ 2956¢45¢¢ 2910¢09¢¢ 2743¢05¢¢ 2447¢21¢¢

11602¢29¢¢ 11549¢07¢¢ 11629¢09¢¢ 11647¢33¢¢ 11745¢03¢¢ 11812¢46¢¢ 11654¢57¢¢ 11820¢39¢¢ 11722¢51¢¢ 11625¢20¢¢ 11459¢08¢¢ 11619¢53¢¢ 11441¢51¢¢ 11440¢05¢¢ 11535¢27¢¢ 11436¢58¢¢ 11528¢43¢¢ 12001¢59¢¢ 11619¢53¢¢ 11713¢38¢¢ 12109¢¢ 11740¢57¢¢ 11739¢08¢¢ 11440¢35¢¢ 11428¢01¢¢

BRM 7782 BRM 7783 BRM 7845 BRM 7847 BRM 7848 BRM 7849 BRM 7850 BRM 7820 BRM 7835 BRM 7843 BRM 7796 MM 2590 BRM 7788 BRM 7795 BRM 7801 MM 2580 BRM 7784 MM 2498 MM 2600 MM 2605 PJ1 BRM 7815 BRM 7809 BRM 7790 BRM 7791

J. EVOL. BIOL. 15 (2002) 576–587 ª 2002 BLACKWELL SCIENCE LTD

Phylogeography of A. acuminata

were amplified then labelled with priming method.

32

P using the random

Data analysis Banding patterns obtained were interpreted in terms of restriction site or length mutations, and assessed as presence or absence of mutations (not presence or absence of bands). Fragment patterns for consecutive cp probes were compared to ensure that each mutation was correctly interpreted and counted only once. Where a length mutation was detected by more than one restriction enzyme it was counted as only one mutation. All individuals were scored except for three individuals from one population of A. oldfieldii where the DNA was degraded. Nucleotide diversity was calculated for restriction site mutations using HAPLO (Lynch & Crease, 1990), and partitioned within and between populations. Haplotype diversity was calculated using Nei’s gene diversity measures (Nei, 1977) for haplotypes in the total sample and in each population. A parsimony analysis of haplotype relationships characterized by the presence or absence of each mutation was carried out using P A U P (Swofford, 1991). Bootstrap analysis used 100 replications and heuristic search with TBR branch swapping and MULPARS on. To test for association between phylogenetic position of haplotypes and their geographical distribution a nested clade analysis (Templeton et al., 1995) was carried out. The nested cladogram was drawn from the P A U P cladogram and parsimony of all connections determined using the program ParsProb (http://bioag.byu.edu/zoology/crandall_lab/programs.htm) based on algorithms described in Templeton et al. (1992). Geographical associations of haplotypes was determined using the program GeoDis (Posada et al., 2000). Interpretation of the nested clade analysis followed the inference key given in Templeton et al. (1995). An addendum to the nested clade analysis (Templeton, 2001) was used to test for gene flow between the lineages identified in the original nested clade analysis. The average pairwise distances between geographical centres of the haplotypes and clades found at each site were calculated and plotted in order of increasing latitude of the sites. This was done for all nesting levels in the cladogram.

Results Polymorphism in cpDNA This study analysed restriction sites in the cp genome and revealed polymorphism with all enzymes used. Within the A. acuminata complex (including A. oldfieldii) 76 mutations were detected; 17 of these were restriction site mutations and 58 were length mutations. One inversion was present in both populations of A. oldfieldii. Five mutations occurred in the inverted repeat and the rest of

J. EVOL. BIOL. 15 (2002) 576–587 ª 2002 BLACKWELL SCIENCE LTD

579

the mutations occurred in the large and small single copy regions of the chloroplast genome. In addition to the 76 mutations in the A. acuminata complex 11 mutations were detected in either or both of the two outgroup species, A. ephedroides and A. anfractuosa. The 76 mutations in the A. acuminata complex were distributed over 38 haplotypes. There were two haplotypes (1 and 27) that were more common than all other haplotypes with frequencies of 18 and 17%, respectively. The remaining haplotypes were present at much lower frequencies with most haplotypes present in less than 2% of individuals and unique to single populations. A large amount of the variation occurred within populations with 20 of 25 populations showing polymorphism, and one population where all five individuals sampled had different haplotypes. Within A. oldfieldii there was substantial polymorphism with six haplotypes present in seven individuals. Both populations of A. oldfieldii were polymorphic, although only two individuals were scored in one population. Haplotype relationships A phylogenetic parsimony analysis of haplotypes gave six trees all of the same length, with a consistency index of 0.930. A consensus tree with A. ephedroides and A. anfractuosa as outgroups is presented in Fig. 2. The phylogenetic tree shows the presence of three main clades. Clades B and C are sister clades and are distinguished from Clade A by 29 mutations. Clade A encompasses all samples from the two A. oldfieldii populations. Clades B and C were not specific to taxa, but rather showed a geographical pattern of distribution that is shown in Fig. 3. Clade B is dominated by populations of the narrow phyllode variant and the most south-westerly populations of ssp. burkittii (populations 23, 26 and 27). Haplotypes from Clade B were also present in some individuals of ssp. acuminata (populations 2 and 5) and the undefined variants (population 14). Clade C is comprised of the remainder of the populations of ssp. A. acuminata, small seeded variant, undefined variants and the eastern populations of ssp. burkittii (populations 21, 22 and 24). Within Clade C there is a structured subclade D that was present in all taxa except the narrow phyllode variant. The rest of Clade C is made up of a polytomy of two groups of haplotypes and six unstructured haplotypes. Clade C is widespread across the complex distribution. Clade B is restricted to the middle of the distribution in south-west Australia but with some representatives in the more peripheral areas. The subclade D occurs to the north of the distribution of Clade B but is also represented in some more distant populations. Nucleotide diversity Nucleotide diversity, the average number of nucleotide differences per site between two sequences (Nei, 1978),

580

M. BYRNE ET AL.

Fig. 2 Phylogenetic parsimony tree of haplotype relationships in the Acacia acuminata complex. Consensus tree based on 100 boot strap replications. Numbers above lines refer to individual mutations. Numbers in italics below lines at nodes represent bootstrap confidence values. Letters at nodes identify clades. Symbols after Haplotype number refer to taxa (see Fig. 1). Numbers to left of symbols are population numbers (see Table 1) with number of individuals with that haplotype in each population in parentheses.

can be determined for restriction site but not length mutations. Nucleotide diversity, averaged over all pairs of individuals, in A. acuminata (excluding A. oldfieldii) was 0.079%. Nucleotide diversity between A. oldfieldii (Clade A), and A. acuminata (Clades B and C combined) was 0.287%. Nucleotide diversity between Clade B and Clade C was 0.083%, and the diversity between the sub-clade D and the rest of Clade C was 0.027%. The diversity within the clades varied with Clade C having the greatest diversity (0.050, or 0.048% when sub-clade D was excluded). The nucleotide diversity within Clade B was 0.023%, and within sub-clade D was 0.018%. The

proportion of nucleotide diversity between populations of A. acuminata, NST, was 0.63%. Haplotype diversity was calculated using haplotype frequencies based on all mutations. Total haplotype diversity, HT, in A. acuminata was 0.9196, and mean haplotype diversity within populations, HS, was 0.4417. The proportion of haplotype diversity between populations, GST, was 52%. Nucleotide diversity and haplotype diversity were analysed at the species level, excluding A. oldfieldii, rather than the infraspecific taxa level since the taxa grade into each other and the patterns of chloroplast diversity detected were not associated with taxa.

J. EVOL. BIOL. 15 (2002) 576–587 ª 2002 BLACKWELL SCIENCE LTD

Phylogeography of A. acuminata

581

Fig. 3 Geographical location of clades influenced by pastfragmentation within Acacia acuminata. Haplotypes from the widespread clade shown in white (clade 3–2 ⁄ C). Haplotypes from fragmented clades shown in black (clade 3–1 ⁄ B) or grey (clade 2–3 ⁄ D).

and displacement values using the inference key of Templeton et al. (1995) (Table 3). At lower nesting levels in the cladogram association between haplotypes or clades and their geographical distribution were generally influenced by restricted gene flow but some cases of long distance colonization were inferred because of the significant displacement of haplotypes or clades from the centre of their nested clades. Inferences of past fragmentation were made at higher nesting levels for the widespread Clade 3–2 and at the total cladogram level. This was due to the mostly nonoverlapping distribution of the restricted clades (Clade 2–3 and 3–1), their significantly small dispersion values, and large interior – tip distances. The tip-interior status of the three-step clades is not clear from the topology of the cladogram but coalescent theory predicts that ancestral haplotypes will be the most frequent, occur

Nested clade analysis The nested clade analysis was also carried out for A. acuminata (not including A. oldfieldii) without reference to intraspecific taxa. The nested clade structure is shown in Fig. 4. All the most parsimonious connections had a probability greater than 99% of being true. Permutation analyses based on 1000 resamples estimated significant association with geographical structure for three one-step clades, two two-step clades, both threestep clades and the total cladogram. The dispersion (Dc) and displacement (Dn) values, and their probability of being significantly large or small, for the clades showing significant geographical structure are given in Table 2. Interpretation of historical process influencing the geographical structure of haplotypes at each clade level was made on the basis of the significance of the dispersion

Total cladogram A 3–1

B

2–2

C

3–2

10

2–4

1–5

26 1–10

0

25

Fig. 4 Haplotype network showing nested clades for Acacia acuminata. Haplotype numbers are those from Fig. 2. Interior haplotypes not detected in the sample are represented by 0. Each line connecting haplotypes represents a single mutational change. One-step clades are indicated by thin-lined boxes, two- and three-step clades by heavier-lined boxes. The equivalent clade name from the parsimony analysis is given in parentheses after the nested clade number.

11 23

1–6

2–1

1–2 2

0

0

0

24 1–12

1–13 32

22 1–11 21 1–1 3

4

5

1

1–4 9

0

0

0

0

0

13 0

20

27

2–3

1–9 14

0 12

17

15

16 1–7

7

8 0

28 29 1–15

30

31

19 1–14

2–5 6 1–3

J. EVOL. BIOL. 15 (2002) 576–587 ª 2002 BLACKWELL SCIENCE LTD

0 18 1–8

582

M. BYRNE ET AL.

Table 2 Levels of clade, nested clade and interior-tip distance in clades with significant geographical association in Acacia acuminata. Nested clade

Clade ⁄ haplotype

Dc

Probability

Dn

Probability

Clade 1–4

Hap 1 (interior) Hap 7 (tip) Hap 8 (tip) Hap 5 (interior) Hap 9 (interior) Interior vs. tip Hap 22 (tip) Hap 21 (interior) Interior vs. tip Hap 31 (tip) Hap 30 (tip) Hap 29 (tip) Hap 28 (tip) Hap 27 (interior) Interior vs. tip Clade 1–2 (tip) Clade 1–3 (tip) Clade 1–4 (interior) Clade 1–1 (tip) Interior vs. tip Clade 1–15 (interior) Clade 1–11 (tip) Clade 1–13 (tip) Clade 1–14 (interior) Interior vs. tip Clade 2–1 (tip) Clade 2–2 (tip) Clade 2–5 (interior) Clade 2–3 (tip) Clade 2–4 (tip) Interior vs. tip Clade 3–1 (tip) Clade 3–2 (interior) Interior vs. tip

67.0000 0.0000 0.0000 0.0000 0.0000 58.9600 0.0000 306.4613 306.4613 0.0000 0.0000 0.0000 141.6074 231.6004 137.1955 0.0000 0.0000 92.7519 0.0000 92.7519 244.8049 268.2873 0.0000 75.1533 )27.2254 100.8090 120.0090 304.9672 87.5531 187.5381 182.8039 115.0747 259.9867 144.9120

S0.0080 1.0000 0.1210 0.1230 1.0000 0.3640 S0.0020 0.1810 L0.0020 1.0000 1.0000 0.0730 S0.0190 0.2760 0.0510 1.0000 S0.0060 0.1410 0.1090 L0.0030 S0.0020 0.2490 1.0000 S0.0010 0.1010 S0.0140 0.3450 L0.0000 S0.0000 0.1440 L0.0000 S0.0000 L0.0000 L0.0000

67.3978 417.5037 31.8007 149.9991 333.1978 )75.7305 166.4814 353.1255 186.6441 289.5709 289.5709 300.4403 260.5806 229.2331 )42.8225 79.6011 147.1047 93.2232 148.1719 )42.9867 274.8068 429.5431 265.9838 248.9580 145.0830 101.8275 227.6761 301.9020 166.1339 199.7443 124.1337 125.4811 266.6911 141.2100

S0.0080 L0.0370 S0.0260 0.1420 L0.0670 0.1550 S0.0020 L0.0020 L0.0020 0.5400 0.5400 0.2800 0.4160 0.2260 0.2260 0.5650 0. 1720 0.1620 0.1380 0.1620 S0.0090 L0.0000 0.4330 0.1160 S0.0010 S0.0210 L0.0210 L0.0000 S0.0010 0.0820 L0.0000 S0.0000 L0.0000 L0.0000

Clade 1–11

Clade 1–15

Clade 2–1

Clade 2–5

Clade 3–1 Clade 3–2

Clade 4–1

Dc, clade distance (dispersion), Dn, nested clade distance (displacement). L, probability of larger than expected value, S, probability of smaller than expected value. Significant probabilities in bold.

Table 3 Phylogeographical inferences from nested clade analysis of Acacia acuminata. Numbers in Key refer to options in inference key in Templeton et al. (1995). Nested Clade

Key

Clade Clade Clade Clade Clade Clade

1, 1, 1, 1, 1, 1,

1–4 1–11 1–15 2–1 2–5 3–1

Clade 3–2 Clade 4

2, 2, 2, 2, 2, 2,

Inference 3, 3, 3, 3, 3, 3,

5, 6, 13 yes 4 no 4 no 4 no 5, 6, 13 yes 5, 6, 7 yes

1, 2, 3, 4, 9 no 1, 2, 3, 4, 9 no

Long distance colonization Restricted gene flow Restricted gene flow Restricted gene flow Long distance colonization Restricted gene flow with some long distance dispersal Past fragmentation Past fragmentation

in the greatest number of populations and have the greatest number of mutational connections (Crandall & Templeton, 1993). Using these criteria, Haplotype 27

nested in Clade 3–2 is the most ancestral haplotype suggesting that Clade 3–2 represents an interior clade in the total cladogram. Clade 3–2 is also more geographically widespread than Clade 3–1 which is characteristic of interior clades (Templeton et al., 1995). Clade 3–2 was therefore designated as an interior clade in the analysis. The inference of past fragmentation at the total cladogram level is also supported by the larger than average number of mutations that separate the two clades and the pattern of increasing average clade distance with increasing nested clade level until the total clade level where this pattern is reversed. The geographical distribution of the clades influenced by past fragmentation is shown in Fig. 3. Clade 3–1 in the nested clade analysis is equivalent to Clade B in the parsimony analysis and Clade 3–2 is equivalent to Clade C. There are four populations that share haplotypes from the two major clades. These populations occur throughout

J. EVOL. BIOL. 15 (2002) 576–587 ª 2002 BLACKWELL SCIENCE LTD

Phylogeography of A. acuminata

583

location distances were detected for populations throughout the range of the complex at all levels from haplotypes to 3-step clades. The level of pairwise location distances remained high and did not tend to zero at higher clade levels. At the 3-step clade level, where the major fragmentation event was detected, the four sites with mixed haplotypes show lower pairwise location distances than generally detected previously and are not geographically clustered.

Discussion

Fig. 5 Average pairwise location distances for haplotypes and clades for the Acacia acuminata cladogram. The 23 populations are ordered by latitude from north (left) to south (right). Average pairwise location distances for each population between the geographical centres of (a) haplotyes, (b) 1-step clades, (c) 2-step clades and (d) 3-step clades, are plotted for each population.

the range of the complex and are not clustered together. To test whether the populations share haplotypes due to gene flow between the clades, the average pairwise distance between the geographical centres of the haplotypes or nested clades was plotted for each population ordered from north to south (Fig. 5). Non-zero pairwise

J. EVOL. BIOL. 15 (2002) 576–587 ª 2002 BLACKWELL SCIENCE LTD

Phylogeographical patterns are evident in this cpDNA phylogeny in the A. acuminata complex. However, apart from A. oldfieldii which is clearly identified as a distinct entity, haplotypes were not restricted to sub-specific taxa or variants but did show clear coalescence to a common ancestor with strong bootstrap support. Although the unusually high level of cpDNA diversity within populations makes interpretation of fine-scale patterns difficult there is clear evidence for two distinct lineages within A. acuminata, a widespread clade that contains representatives from all taxa except narrow phyllode variant, and a restricted clade that contains all the sampled narrow phyllode variant populations but also contains representatives from all the other taxa. The sharing of haplotypes between taxa may occur through hybridization or through incomplete lineage sorting, and it is often difficult to distinguish between these two alternatives as the pattern of incongruence is similar (Wendel & Doyle, 1998). Incomplete lineage sorting is a factor that complicates the use of cpDNA in phylogenies as it can be easily misinterpreted as evidence of interspecific gene flow (Schaal & Olsen, 2000). The distribution pattern of shared haplotypes can give some clues to the cause of the incongruence between morphological and molecular phylogenies (Wendel & Doyle, 1998), and the distribution of cp variation detected here in A. acuminata suggests that incomplete lineage sorting is the major influence on the pattern of shared haplotypes between taxa. All of the taxa contain both shared and unique haplotypes, and the shared haplotypes are often ancient, interior haplotypes rather than recent derived ones, and are not restricted to neighbouring populations. The unique haplotypes are both divergent haplotypes and those derived from shared ancestral haplotypes. This pattern is consistent with incomplete lineage sorting of variation in the ancestral taxon prior to divergence of the nuclear genome in the two recognized taxa, ssp. acuminata and ssp. burkittii. There are several other examples where incomplete lineage sorting of ancestral polymorphisms has been identified as the most likely cause of the incongruence between cpDNA diversity and morphology (Lavin et al., 1991; Mayer & Soltis, 1994; Mason-Gamer et al., 1995). At the species level geographical associations were present within the phylogeny of A. acuminata. At deeper

584

M. BYRNE ET AL.

levels of the phylogeny nested clade analysis identified past fragmentation as having significant effects with two lineages at the base of the phylogeny showing ancient separation, and a second separation into isolated groups present within one of these lineages. In both cases, the clades influenced by fragmentation do not have allopatric distributions, rather one clade is localized within the distribution of the other more widespread clade. There are no clear geological features that would explain a fragmentation event in the sampled distribution of the complex although historical climatic change with absence of major barriers to dispersal can lead to species surviving in isolated habitat islands (Templeton et al., 1990). The scale and intensity of such microvariance will effect lineage differentiation but if populations are isolated for a significant period of time the effect of fragmentation will be detectable even if secondary contact occurs (Templeton et al., 1995). Climatic instability during the Pleistocene caused by cyclic contraction and expansion of the mesic and arid zones in south-west Western Australia, could lead to fragmentation and isolation of populations in this area. Later periods of suitable climatic conditions would allow range expansion from relictual areas and bring isolated lineages back into contact. These historical influences would leave a pattern of nonallopatric fragmentation as detected here in A. acuminata. The localization of the restricted lineages to the middle of the species distribution is also consistent with the prediction that the major effects of climatic instability will be in the semiarid area between the mesic and arid zones (Hopper et al., 1996). The presence of the fragmentation effects at different levels in the phylogeny and the different estimated time of divergence between the fragmented lineages indicates that vicariance events may have occurred several times in the evolutionary history of A. acuminata. For the chloroplast genome it is estimated that 0.1% divergence represents one million years separation (Zurawski et al., 1984). Using this estimate, the time of divergence between the two main lineages within A. acuminata is in the order of 800 000 years ago, in the middle of the Pleistocene. Although this value should be treated as a broad indicator only, it is consistent with the hypothesis of mid-Pleistocene climatic instability leading to divergence between lineages in the semiarid zone. The lower divergence of the sub-group from the widespread lineage suggests that climatic pressures in the northern coastal area of the region have been more recent than in the inland centre of the zone. An alternative explanation for the origin of two lineages within a species, particularly where one lineage has a restricted geographical distribution, is introgression of one chloroplast genome from another species. Ancient hybridization and introgression of chloroplast genomes can have significant impacts on intraspecific chloroplast variation (Soltis et al., 1992; Ennos et al., 1999). Introgression of chloroplast lineages where hybridization was not suspected has been observed in a number of species,

e.g. Tellima grandiflora (Soltis et al., 1991), Populus nigra (Smith & Sytsma, 1990), Salix melanopsis (see Reiseberg & Brunsfeld, 1992). Introgression from another species is unlikely in A. acuminata as the two lineages coalesce to a common ancestor. In this case, if one of the lineages had been derived from introgression the donor species would be expected to come from the same evolutionary lineage as A. acuminata. Based on morphological criteria there are no other extant species closer to A. acuminata than A. oldfieldii and the evidence presented here shows A. oldfieldii to be clearly phylogenetically distinct from A. acuminata. Further, studies of cpDNA diversity in other woody angioserm species have identified the presence of divergent lineages that were not due to interspecific hybridization (Lavin et al., 1991; Sewell et al., 1996). Interpretation of the patterns of variation within A. acuminata at a finer scale is more difficult because of the large amount of variation detected within and between populations. The nested clade analysis inferred restricted gene flow with some long distance colonization events as the main influence on the geographical patterns observed at shallower levels of the phylogeny. This inference results from the sharing of some identical interior haplotypes across large geographical distances throughout the species range. Although gene flow through long distance colonization events have been shown to lead to a patchy distribution of genotypes (Nichols & Hewitt, 1994), the large distances, and occurrence in several different geographical areas, suggests it is not likely in this species. The pattern of pairwise location distances for populations, identified through the addendum to the nested clade analysis, was not what would be expected if sharing of haplotypes between clades was because of gene flow. In contrast incomplete lineage sorting of ancestral polymorphism, which was identified at higher levels of the phylogeny, would also influence patterns at lower levels. This would result in a patchy retention of ancestral haplotypes similar to that detected here, and is a more likely explanation for shared haplotypes than isolated cases of gene flow over such large distances. The inadequacy in the inferences from the nested clade analysis at lower levels in the phylogeny is most probably because of the complex patterns of haplotype distributions that can arise in an ancient tectonically stable landscape where evolutionary influences have operated over long time frames. Another study has noted some inadequacies in the inferences derived from nested clade analysis when compared with extensive biogeographical data relating to expansion from glacial refugia (Seddon et al., 2001). However, in this study at the higher level of the phylogeny the predictions from the nested clade analysis are consistent with hypothesized climatic and biogeographical history in the region, and the nested clade analysis has been informative in establishing the broad phylogeographical patterns in this species complex although the fragmentation events in this case are not associated with

J. EVOL. BIOL. 15 (2002) 576–587 ª 2002 BLACKWELL SCIENCE LTD

Phylogeography of A. acuminata

significant eco-geographical barriers. The usefulness of cpDNA markers for phylogeography can be limited by the often low level of variation in the chloroplast genome and the slow rate of evolution (Schaal et al., 1998; Ennos et al., 1999; Schaal & Olsen, 2000). A large amount of cpDNA variation was detected in this study and it has been successful in identifying broad phylogeographical patterns in this species. The slow mutation rate of cpDNA may be less problematic in determining phylogeographical patterns in this landscape where the processes affecting haplotype distribution have been active over long time scales compared with those where haplotype distribution is most influenced by post-Pleistocene events. The cpDNA analysis clearly identified A. oldfieldii as distinct from the rest of the A. acuminata complex. Using the estimate of chloroplast sequence divergence the time of divergence of A. oldfieldii from A. acuminata is in the order of 2.5–3 million years ago towards the end of the Tertiary period. This level of divergence between A. oldfieldii and A. acuminata was unexpected given the relatively small morphological differences between the species (A. oldfieldii has glabrous phyllodes and shortly pedunculate spikes compared with short hairs on fimbriolate phyllodes and sessile spikes in A. acuminata). Lack of concordance between morphological and genetic divergence is a pattern that is evident within a significant proportion of species and species complexes investigated in south-west Western Australia (Coates, 2000). It indicates that many apparently morphologically uniform taxa, are likely to consist of multiple evolutionary lineages and highlights the value of molecular systematic and phylogenetic studies in resolving and understanding the complex evolutionary patterns evident in this flora. Investigations of diversity in the cp genome of Australian plants have often revealed a high degree of intraspecific variation (Byrne & Moran, 1994; Butcher et al., 1995; Steane et al., 1998; Byrne et al., 1999; Jackson et al., 1999; McKinnon et al., 1999; Byrne & Macdonald, 2000). The level of diversity detected here in the A. acuminata complex is also high, both within the complex and populations. The value of nucleotide diversity in A. acuminata (0.079%) was similar to other widespread species, e.g. E. kochii complex (0.082%) and E. nitens (0.084%), but the level of haplotype diversity (0.9196) was higher than in those species, e.g. E. kochii (0.8161), E. nitens (0.8706) (Byrne & Moran, 1994; Byrne & Macdonald, 2000). The proportion of diversity maintained between populations was lower in A. acuminata than in the eucalypt species for both nucleotide diversity and haplotype diversity. This reflects the unusually high level of polymorphism within populations in A. acuminata. The high levels of variation in cpDNA in the species suggests that population sample sizes should have been larger to adequately sample the unexpected level of variation in the populations. The phylogeographical pattern and genetic relationships within the A. acuminata group are complex, but are

J. EVOL. BIOL. 15 (2002) 576–587 ª 2002 BLACKWELL SCIENCE LTD

585

consistent with previous proposals of biogeographical patterns within south-western Australia (Hopper, 1979; Hopper et al., 1996). The pattern of divergence that has often been identified in fragmented species with restricted distributions has also been identified here in a widespread species. This divergence also appears to have arisen through historical fragmentation although in widespread species subsequent expansion of lineages may bring them back into contact as appears to be the case in the A. acuminata complex.

Acknowledgments We thank Bruce Maslin, Maurice MacDonald and Peter Jones for phyllode collections, and John Maslin for help in producing the maps. We also thank Loren Rieseberg and two anonymous reviewers for helpful comments on the manuscript. Financial support was provided by the Forest Products Commission of Western Australia.

References Avise, J.C. 2000. Phylogeography. The History and Formation of Species. Harvard University Press, Cambridge, MA. Broadhurst, L., Coates, D. & Maslin, B. 2002. Genetic diversity within and divergence between rare and geographically widespread taxa of the Acacia acuminata (Mimosaceae) complex. Heredity 88: 250–257. Butcher, P.A., Byrne, M. & Moran, G.F. 1995. Variation within and among the chloroplast genomes of Melaleuca alternifolia and M. linariifolia (Myrtaceae). Pl. Syst. Evol. 194: 69–81. Byrne, M. 1999. High genetic identities between three oil mallee taxa, Eucalyptus kochii ssp. kochii, ssp. plenissima and E. horistes, based on nuclear RFLP analysis. Heredity 82: 205–211. Byrne, M. & Moran, G.F. 1994. Population divergence in the chloroplast genome of Eucalyptus nitens. Heredity 73: 18–28. Byrne, M. & Macdonald, B. 2000. Phylogeography and conservation of three oil mallee taxa, Eucalyptus kochii ssp. kochii, ssp. plenissima and E. horistes. Aust. J. Bot. 48: 305–312. Byrne, M., Moran, G.F. & Tibbits, W.N. 1993. Restriction Map and maternal inheritance of chloroplast DNA in Eucalyptus nitens. J. Heredity 84: 218–220. Byrne, M., Macdonald, B. & Coates, D. 1999. Divergence in the chloroplast genome and nuclear rDNA of the rare Western Australian plant Lambertia orbifolia Gardner (Proteaceae). Mol. Ecol. 8: 1789–1796. Byrne, M., Macdonald, B. & Francki, M. 2001a. Incorporation of sodium sulphite into extraction protocol minimizes degradation of Acacia DNA. Biotechniques 30: 742–744. Byrne, M., Tischler, G., Macdonald, B., Coates, D.J. & McComb, J. 2001b. Phylogenetic relationships between two rare acacias and their common, widespread relatives in south-western Australia. Conservation Genet 2: 157–166. Coates, D.J. 2000. Defining conservation units in a rich and fragmented flora: implications for management of genetic resources and evolutionary processes in south-west Australian plants. Aust. J. Bot. 48: 329–339. Crandall, K.A. & Templeton, A.R. 1993. Empirical tests of some predictions from coalescent theory with applications to intraspecific phylogeny reconstruction. Genetics 134: 959–969.

586

M. BYRNE ET AL.

Demesure, B., Comp, B. & Petit, R.J. 1996. Chloroplast DNA phylogeography of the common beech (Fagus Sylvatica L.) in Europe. Evolution 50: 2515–2520. Dumolin-Lape`gue, S., Demesure, B., Fineschi, S., Le Corre, V. & Petit, R.J. 1997. Phylogeographic structure of white oaks throughout the European Continent. Genetics 146: 1475–1487. Ennos, R.A., Sinclair, W.T., Hu, X.-S. & Langdon, A. 1999. Using organelle markers to elucidate the history, ecology and evolution of plant populations. In: Molecular Systematics and Plant Evolution (P. M. Hollingsworth, R. M. Bateman & R. J. Gornall, eds), pp. 1–19. Taylor & Francis, London. Ferris, C., King, R.A., Va¨ino¨la¨, R. & Hewitt, G.M. 1998. Chloroplast DNA recognizes three refugia sources of European oaks and suggests independent eastern and western immigrations to Finland. Heredity 80: 584–593. Ferris, C., Oliver, R.P., Davy, A.J. & Hewitt, G.M. 1993. Native oak chloroplasts reveal an ancient divide across Europe. Mol. Ecol. 2: 337–344. Hopper, S.D. 1979. Biogeographical aspects of speciation in the southwest Australian flora. Ann. Rev. Ecol. Syst. 19: 399–422. Hopper, S.D., Harvey, M.S., Chappill, J.A., Main, A.R. & Main, B.Y. 1996. The Western Australian biota as Gondwanan heritage – a review. In: Gondwanan Heritage. Past Present and Future of the Western Australian Biota (S. D. Hopper, J.A. Chappill, M. S. Harvey & A. S. George, eds), pp. 1–46. Surrey Beatty & Sons, Chipping Norton. Jackson, H.D., Steane, D.A., Potts, B.M. & Vaillancourt, R.E. 1999. Chloroplast DNA evidence for reticulate evolution in Eucalyptus (Myrtaceae). Mol. Ecol. 8: 739–751. King, R.A. & Ferris, C. 1998. Chloroplast DNA phylogeography of Alnus glutinosa (L.) Gaertn. Mol. Ecol. 7: 1151–1161. Kodela, P.G. & Tindale, M.D. 1998. The reduction of Acacia burkittii to Acacia acuminata subsp. burkittii (Acacia sect Juliflorae: Fabaceae, Mimosoideae). Telopea 7: 415–417. Lavin, M., Mathews, S. & Hughes, C. 1991. Chloroplast DNA variation in Gliricidia sepium (Leguminosae): interspecific phylogeny and tokogeny. Am. J. Bot. 78: 1576–1585. Lynch, M. & Crease, T.J. 1990. The analysis of population survey data on DNA sequence variation. Mol. Biol. Evol. 7: 377–394. Maslin, B.R. 2001. WATTLE: Acacias of Australia. CD-Rom, Version 1.0. Australian Biological Resources Study, Canberra and Department of Conservation and Land Management, Perth. Maslin, B.R. & Pedley, L. 1982. The distribution of Acacia (Leguminosae: Mimosoideae) in Australia. Part 1. Species distribution maps. W.A. Herbarium Res. Notes 6: 1–128. Mason-Gamer, R.J., Holsinger, K.E. & Jansen, R.K. 1995. Chloroplast DNA haplotype variation within and among populations of Coreopsis grandiflora (Asteraceae). Mol. Biol. Evol. 12: 371–381. Mayer, M. & Soltis, P.S. 1994. The evolution of serpentine endemics: a chloroplast DNA phylogeny of the Strepanthus glandulosus complex (Cruciferae). Syst. Bot. 19: 557–574. McKinnon, G.E., Steane, D.A., Potts, B.M. & Vaillancourt, R.E. 1999. Incongruence between chloroplast and species phylogenies in Eucalyptus subgenus Monocalyptus (Myrtaceae). Am. J. Bot. 86: 1038–1046. Nei, M. 1977. F-statistics and analysis of gene diversity in subdivided populations. Annal. Hum. Genet. 41: 225–233. Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89: 583– 590.

Nichols, R.A. & Hewitt, G.M. 1994. The genetic consequences of long distance dispersal during colonization. Heredity 72: 312–317. Orchard, A.E. & Wilson, A.J.G. (eds) .2001. Flora of Australia Vol. IIB Mimosaceae, Acacia Part II. ARBS ⁄ CSIRO Publishing, Melbourne. Posada, D., Crandall, K.A. & Templeton, A.R. 2000. GeoDis: a program for the cladistic nested analysis of the geographical distribution of genetic haplotypes. Mol. Ecol. 9: 487–488. Reiseberg, L.H. & Brunsfeld, S.J. 1992. Molecular evidence and plant introgression. In: Molecular Systematics of Plants (P.S. Soltis, D. E. Soltis & J. J. Doyle, eds), pp. 151–176. Chapman & Hall, New York. Schaal, B.A., Haywood, D.A., Olsen, K.M., Rauscher, J.T. & Smith, W.A. 1998. Phylogeographic studies in plants: problems and prospects. Mol. Ecol. 7: 465–474. Schaal, B.A. & Olsen, K.M. 2000. Gene genealogies and population variation in plants. Proc. Natl. Acad. Sci. USA 97: 7024–7029. Seddon, J.M., Santucci, F., Reeve, N.J. & Hewitt, G.M. 2001. DNA footprints of European hedgehogs, Erinaceus europaeus and E. concolor: Pleistocene refugia, postglacial expansion and colonization routes. Mol. Ecol. 10: 2187–2198. Sewell, M.M., Parks, C.R. & Chase, M. 1996. Intraspecific chloroplast DNA variation and biogeography of North American Liriodendron L. (Magnoliaceae). Evolution 50: 1147–1154. Shinozaki, K., Ohme, M., Tanaka, M., Wakasugi, T., Hayashiuda, N., Matsubayashi, T., Zaita, N., Chunwongse, J., Obokata, J., Yamaguchi-Sinosaki, K., Ohto, C., Torazawa, K., Meng, B.Y., Sugita, M., Deno, H., Kamagashira, T., Yamada, K., Kusuda, J., Takaiwa, F., Kato, A., Tohdoh, N., Shimada, H. & Sugiura, M. 1986. The complete nucleotide sequence of the tobacco chloroplast genome: its gene organization and expression. EMBOJ. 5: 2043–2049. Smith, R.L. & Sytsma, K.J. 1990. Evolution of Populus nigra (Sect. Aigeiros): Introgressive hybridization and the chloroplast contribution of Populus alba (Sect. Populus). Am. J. Bot. 77: 1176–1187. Soltis, D.E., Gitzendanner, M.A., Strenge, D.D. & Soltis, P.E. 1997. Chloroplast DNA intraspecific phylogeography of plants from the Pacific Northwest of North America. Pl. Syst. Evol. 206: 353–373. Soltis, D.E., Mayer, M.S., Soltis, P.S. & Edgerton, M. 1991. Chloroplast DNA variation in Tellima grandiflora (Saxifragaceae). Am. J. Bot. 78: 1379–1390. Soltis, D.E., Soltis, P.S. & Milligan, B.G. 1992. Intraspecific chloroplast DNA variation: systematic and phylogenetic implications. In: Molecular Systematics of Plants (P.S. Soltis, D. E. Soltis & J. J. Doyle, eds), pp. 151–176. Chapman & Hall, New York. Steane, D.A., Byrne, M., Vaillancourt, R.E. & Potts, B.M. 1998. Chloroplast DNA polymorphisms signal complex inter-specific interactions in Eucalyptus (Myrtaceae). Aust. Syst. Bot. 11: 25– 40. Suguira, M., Shinozaki, K., Zaita, N., Kusuda, M. & Kumano, M. 1986. Clone bank of the tobacco (Nicotiana tabacum) chloroplast genome as a set of overlapping restriction endonuclease fragments: mapping of eleven ribosomal protein genes. Pl. Sc. 44: 211–216. Swofford, D.L. 1991. Paup: Phylogenetic Analysis Using Parsimony, Version 3.1.1. Computer program distributed by the Illinois Natural History Survey, Champaign, IL.

J. EVOL. BIOL. 15 (2002) 576–587 ª 2002 BLACKWELL SCIENCE LTD

Phylogeography of A. acuminata

Sytsma, K.J. & Gottlieb, L.D. 1986. Chloroplast DNA evolution and phylogenetic relationships in Clarkia sect. Peripetasma (Onagraceae). Evolution 40: 1248–1261. Templeton, A.R. 2001. Using phylogeographic analyses of gene trees to test species status and processes. Mol. Ecol. 10: 779– 791. Templeton, A.R., Crandall, K.A. & Sing, C.F. 1992. A cladistic analysis of phenotypic associations with haplotypes inferred from restriction mapping and DNA sequence data. III. Cladogram estimation. Genetics 132: 619–633. Templeton, A.R., Routman, E. & Phillips, C.A. 1995. Separating population structure from population history: a cladistic analysis of the geographical distribution of mitochondrial haplotypes in the tiger salamander, Ambystoma tigrinum. Genetics 140: 767–782.

J. EVOL. BIOL. 15 (2002) 576–587 ª 2002 BLACKWELL SCIENCE LTD

587

Templeton, A.R., Shaw, K., Routman, E. & Davis, S.K. 1990. The genetic consequences of habitat fragmentation. Ann. Miss. Bot. Gard. 77: 13–27. Wendel, J.F. & Doyle, J.J. 1998. Phylogenetic incongruence: window into genome history and molecular evolution. In: Molecular Systematics of Plants II: DNA Sequencing (D. E. Soltis, P. S. Soltis & J. J. Doyle, eds), pp. 265–296. Kluwer Academic Publishers, Norwell. Zurawski, G., Clegg, M.T. & Brown, A.H.D. 1984. The nature of nucleotide sequence divergence between barley and maize chloroplast DNA. Genetics 106: 735–749. Received 27 February 2002; accepted 20 March 2002