Academia.eduAcademia.edu
South African Journal of Botany 89 (2013) 244–256 Contents lists available at ScienceDirect South African Journal of Botany journal homepage: www.elsevier.com/locate/sajb Wood anatomy of the tribe Podalyrieae (Fabaceae, Papilionoideae): Diversity and evolutionary trends A.V. Stepanova a,b, A.A. Oskolski a,b, P.M. Tilney a, B.-E. Van Wyk a,⁎ a b Department of Botany and Plant Biotechnology, University of Johannesburg, P.O. Box 524, Auckland Park 2006, Johannesburg, South Africa Komarov Botanical Institute of the Russian Academy of Science, Prof. Popov Str. 2, 197376 St. Petersburg, Russia a r t i c l e i n f o Available online 12 September 2013 Edited by JS Boatwright Keywords: Fabaceae Fynbos Grouped vessels Helical thickenings Podalyrieae Shrubs Trees Wood anatomy a b s t r a c t Detailed wood anatomical data for 32 species from all nine genera of the tribe Podalyrieae are presented, together with numerical analyses and the mapping of character states onto the latest available molecular phylogeny. It was found that trees (Cadia, Calpurnia and Virgilia) have vessels in small isolated groups, whilst fynbos shrubs (the remaining genera: Amphithalea, Cyclopia, Liparia, Podalyria, Stirtonanthus and Xiphotheca) commonly show highly grouped narrow vessels (frequently in a dendritic pattern), and helical thickening on the vessel walls. Comparisons of the main character state changes with the molecular phylogeny of the tribe show that the wood structure of trees probably represents the basic condition in the tribe; character states present in shrubs appear to have arisen a few times and very likely represent adaptations to seasonal water stress. In general, the wood anatomy is congruent with current subtribal and generic delimitations. Fire-survival strategy is reflected in the rays, with seeders having mostly procumbent cells whilst sprouters have square and upright cells. The close similarity in wood anatomy between Cadia and Calpurnia is in agreement with the transfer of Cadia to the Podalyrieae. A remarkable diversity of crystals was found, including prismatic, acicular and navicular crystals, the last two of which may occur singly or in sheaf-like aggregates. © 2013 SAAB. Published by Elsevier B.V. All rights reserved. 1. Introduction The tribe Podalyrieae is subendemic to the Cape Floristic Region of South Africa, with six of the nine genera confined to the Fynbos biome with its cold wet winters and dry summers, namely Amphithalea Eckl. & Zeyh. (42 spp.), Cyclopia Vent. (23 spp.) (Schutte et al., 1995), Liparia L. (20 spp.), Stirtonanthus B.-E. Van Wyk & A.L. Schutte (3 spp.), Xiphotheca Eckl. & Zeyh. (10 spp.) and Virgilia Poir. (2 spp.). Podalyria Willd. has 16 of the 17 species endemic to fynbos (one species extends into the grassland region of South Africa) (Schutte-Vlok and Van Wyk, 2011); Calpurnia E. Mey. has only one species in the Cape Floristic Region, whilst the other seven species occur in the summer rainfall Grassland biome of South Africa, with one [C. aurea (Lam.) Benth.] extending to tropical east Africa and India (Beaumont et al., 1999); Cadia Forssk. has one species in tropical Africa and six in Madagascar. The current tribal and generic classification system (Van Wyk and Schutte, 1995; Schutte and Van Wyk, 1998; Van Wyk, 2005) was based on morphological, chemical and cytological evidence, which closely agreed with later molecular phylogenetic analyses. The molecular studies showed that Cyclopia forms an isolated group sister to the ⁎ Corresponding author. Tel.: +27 115592412; fax: +27 115592411. E-mail address: bevanwyk@uj.ac.za (B.-E. Van Wyk). 0254-6299/$ – see front matter © 2013 SAAB. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.sajb.2013.07.023 remaining genera (Van der Bank et al., 2002) and that the genus Cadia, despite its radially symmetrical flowers, should be included in the tribe (Boatwright et al., 2008). In the current classification system, Amphithalea and Xiphotheca are placed in the monophyletic subtribe Xiphothecinae whilst the remaining genera are placed in the paraphyletic subtribe Podalyriinae, the latter shown to comprise three main clades: Liparia/Podalyria/Stirtonanthus, Calpurnia/Virgilia, and Cyclopia (Boatwright et al., 2008). Although it has been shown that wood anatomy is useful in exploring the systematics and evolution of Fabaceae (Baretta-Kuipers, 1981; Gasson, 2000), such studies have thus far focussed mainly on groups which consist predominantly of trees. All species of Podalyrieae are woody plants – mostly shrubs or small trees but ranging from tall trees to subshrubs and creeping shrublets – but their wood anatomy has remained largely unknown. Some tree species of Cadia, Virgilia and Calpurnia have been studied in detail (Gasson, 1994; Fujii et al., 1994) but data for the remainder of the tribe are limited to the observations by Van Wyk and Schuttte (1995) and Schutte and Van Wyk (1998), who noted and illustrated the diversity of vessel grouping. This paper describes, for the first time, the wood anatomy of all genera of the tribe Podalyrieae, and attempts to relate the diversity and main discontinuities of wood anatomical characters to the current state of knowledge regarding the systematics and phylogenetics of the group. Table 1 Quantitative wood characters of Podalyrieae. I II III IV V VI VII VIII IX X XI XII XIII XIV Cadia purpurea Ait. KK 33-10 Calpurnia aurea Baker KK 39-11 Calpurnia intrusa E. Mey. BEVW 3249 C Calpurnia sericea Harv. KK 35-11 Cyclopia aurescens Kies. ALS + BEVW 771 Cyclopia burtonii Hofmeyr & E. Phillips ALS 641 Cyclopia buxifolia (Burm.f.) Kies ALS 544 Cyclopia genistoides (L.) Sieber ex C. Presl ALS 615 Cyclopia genistoides ALS 624 Cyclopia intermedia E. Mey. ALS 646 Cyclopia intermedia ALS 647 Cyclopia maculata (Andrews) KiesALS 636 Cyclopia maculata ALS 528 Cyclopia plicata Kies ALS 670a Cyclopia subternata Vogel KK 43–11 Liparia hirsuta Moench JSB 595 Liparia myrtifolia Thunb. ALS 727a Liparia racemosa A.L. Schutte ALS 642a Liparia splendens (Burm.f.) Bos & de Wit BEVW 3147 Podalyria calyptrata (Retz.) Willd. KK 03-10 Podalyria lanceolata Benth. ALS s.n. Podalyria myrtifolia D. Dietr. ALS + BEVW 166 24 ± 1.3 12–38 35 ± 1.5 18–48 41 ± 2.7 11–61 45 ± 2.6 18–74 8 ± 0.3 3–18 13 ± 0.5 4–31 19 ± 0.8 5–44 12 ± 0.7 6–22 16 ± 1.7 7–30 18 ± 1.4 7–36 11 ± 0.5 3–39 18 ± 1.2 6–34a 31 ± 2.8 8–60 7 ± 0.4 3–22 23 ± 2.1 10–48 19 ± 1.0 11–33 22 ± 1.5 10–44 25 ± 1.0 16–42 16 ± 1.0 9–28 47 ± 2.5 22–78 26 ± 2.3 8–61 21 ± 1.2 12–36 145 119–165 73 60–91 44 23–63 75 49–102 792 706–877 658 486–788 228 110–411 411 199–575 418 308–610 390 240–637 503 226–706 247 192–329 143 121–170 1615 1322–2144 470 384–555 280 158–356 545 412–688 358 265–574 433 229–593 7156–84 2.0 7 22.0 11 12.5 1.8 10 29.8 1.8 5 31.0 160.5 461 0 476 0 81.2 407 0.1 14.2 40 0 29.6 84 0.1 44.8 122 0.2 118.8 496 0.1 61.6 287 0.5 4.6 18 5.0 a a a 504 ± 19.7 290–740 601 ± 21.5 279–807 577 ± 17.3 396–828 453 ± 18.5 269–760 63 ± 15.3 448–817 53216.4 329–673 601 ± 19.8 262–814 709 ± 24.1 422–1019 537 ± 22.9 226–801 751 ± 34.2 424–1064 654 ± 32.7 300–1181 602 ± 15.8 442–810 719 ± 22.8 557–866 542 ± 17.8 346–714 744 ± 21.1 512–983 495 ± 14.3 367–611 411 ± 37.3 135–767 605 ± 17.5 456–821 757 ± 17.4 625–952 602 ± 17.4 381–783 841 ± 23.3 654–1085 511 ± 16.0343–678 18 ± 0.8 12–31 27 ± 1.4 14–43 24 ± 1.3 14–41 22 ± 0.9 11–32 56 ± 2.2 138.2 128 ± 4.3 79–178 191 ± 8.0 102–287 151 ± 4.5 95–214 150 ± 4.5 106–204 228 ± 6.1 164–323 172 ± 7.0 87–245 227 ± 8.5 132–347 322 ± 13.2 165–483 245 ± 11.6 148–393 278 ± 12.7 158–405 174 ± 8.5 97–298 207 ± 9.0 12–5324 167 ± 6.6 109–232 161 ± 5.5 102–232 249 ± 11.3 154–425 217 ± 7.1 159–313 191 ± 6.5 115–266 224 ± 7.5 131–323 198 ± 6.8 125–269 216 ± 9.3 94–306 271 ± 11.1 177–404 161 ± 7.6 80–257 3.9 2.5 3.9 2.8–5.8 3.9 2.7–5.3 5.7 4.4–6.7 4.9 3.8–5.9 5.4 4.0–7.8 4.5 3.6–5.7 5.1 4.0–7.3 5.4 4.0–7.8 4.9 3.6–6.5 5.0 3.7–6.6 4.5 3.4–5.8 5.0 4.0–7.2 5.1 4.1–8.2 4.2 3.2–5.6 5.1 3.1–6.8 6.0 4.6–11.6 4.0 2.7–4.9 5.6 4.1–7.1 6.7 4.8–8.9 5.2 4.2–6.3 5.4 4.2–7.2 5.1 3.8–6.8 135 ± 6.7 83–244 217 ± 12.9 105–396 216 ± 14.7 107–476 225 ± 30.8 88–875 798 ± 67.3 238–v 377 ± 27.9 141–879 282 ± 25.0 121–635 605 ± 77.7 137–1421 331 ± 18.7 145–577 729 ± 76.2 267–1887 453 ± 37.2 178–923 418 ± 22.8 239–765 311 ± 27.7 116–759 227 ± 18.2 73–429 375 ± 27.6 173–771 233 ± 15.7 128–456 456 ± 51.602 160–13 304 ± 26.9 146–656 475 ± 38.6 106–1182 259 ± 19.6 84–687 257 ± 15.5 123–415 314 ± 29.4 158–515 3.7 2.1–6.3 8.7 6.3–11.6 7.8 5.3–11.6 5.8 3.2–8.4 2.9 1.1–5.3 4.6 3.2–7.4 3.7 0–6.3 1.6 0.0–4.2 2.6 1.1–4.2 6.5 5.3–7.4 7.3 5.3–12.6 4.7 3.2–7.4 4.5 4.2–5.3 4.8 3.2–6.3 4.5 3.2–6.3 4.9 3.2–7.4 4.5 3.2–6.3 5.4 ± 0.4 3.2–7.4 4.0 ± 0.2 1.9–5.6 4.9 3.2–6.3 5.1 3.2–6.3 4.83.2–6.3 5.2 3.2–7.4 2.4 1.1–4.2 1.9 0.0–3.2 3.4 0.0–8.4 4.2 2.1–6.3 2.7 1.1–5.3 4.4 2.1–6.3 7.2 4.2–9.5 4.2 3.2–6.3 1.2 0–3.2 3.3 2.1–5.3 3.3 1.1–5.3 3.2 2.1–6.3 6.5 4.2–9.5 3.8 1.1–6.3 3.1 1.1–5.3 2.6 1.1–5.3 1.5 ± 0.3 0.0–3.2 7.2 ± 0.3 4.6–10.2 3.4 1.1–5.3 1.7 0.0–4.2 2.51.1–6.3 8.9 5.3–12.6 11.2 9.5–12.6 9.7 7.4–12.6 9.2 5.3–12.6 7.1 5.3–9.5 7.4 6.3–8.4 8.1 5.3–10.5 8.8 5.3–11.6 6.8 5.3–10.5 7.7 6.3–8.4 10.5 8.4–14.7 8.0 6.3–9.5 7.7 6.3–10.5 11.4 9.5–13.7 8.3 6.3–10.5 8.0 6.3–9.5 7.1 5.3–8.4 6.9 5.3–9.5 11.2 9.3–13.9 8.3 5.3–10.5 6.7 5.3–8.4 7.45.3–9.5 199 164–235 2250 200–26 – – – 102.4 277 0.04 5.4 29 7.4 15.7 84 1.0 18.9 78 1.0 8.0 44 4.0 2.1 7 20.0 11.7 54 16.7 3.0 26 13.0 3.2 3.8 3.0 2.8 39–80 3.1 2.6 2.2 2.2 2.7 3.8 2.9 4.3 3.4 3.0 2.2 2.7 3.8 2.8 3.1 3.2 51 ± 2.2 29–73 24 ± 0.7 16–31 36 ± 1.2 29–52 88 ± 10.6 38–320 42 ± 1.6 27–59 44 ± 1.3 33–59 40 ± 3.0 19–90 34 ± 1.5 2–48 46 ± 2.2 27–86 34 ± 1.2 24–49 44 ± 1.7 18–62 34 ± 1.7 19–58 52 ± 2.6 29–87 36 ± 1.6 22–71 44 ± 1.6 26–60 32 ± 1.1 22–45 25 ± 0.8 15–31 A.V. Stepanova et al. / South African Journal of Botany 89 (2013) 244–256 Species and voucher specimens (all in JRAU) (continued on next page) 245 246 Table 1 (continued) I II III IV V VI VII VIII IX X XI XII XIII XIV Podalyria rotundifolia (P.J. Bergius) A.L. Schutte ALS s.n. Stirtonanthus insignis (Compton) B.-E.van Wyk & A.L. Schutte BEVW 3331 Stirtonanthus taylorianus (L. Bolus) B.–E.van Wyk & A.L. Schutte BEVW 3169 Virgilia divaricata Adamson AO 16-09 Virgilia divaricata KK 02-10 Virgilia oroboides subsp. oroboides (P.J. Bergius) T.M. Salter BEVW 5722 Virgilia oroboides subsp. ferruginea B.-E.van Wyk KK 44-11 Amphithalea ericifolia (L.) Eckl. & Zeyh. ALS 617 Amphithalea rostrata A.L. Schutte & B.-E.van Wyk ALS 629 Amphitalea vlokii (A.L. Schutte & B.–E.van Wyk) A.L. Schutte ALS 743 Xiphotheca canescens (Thunb.) A.L. Schutte & B.-E.van Wyk ALS 595a Xiphotheca elliptica (DC.) A.L. Schutte & B.-E.van Wyk ALS 600 Xiphotheca phylicoides A.L. Schutte & B.–E.van Wyk ALS 648 Xiphotheca tecta A.L. Schutte & B.–E.van Wyk ALS 597 25 ± 2.1 10–48 18 ± 1.3 40 13–59 159 3.9 10 16.0 44 6.2 189 ± 7.3 108–273 169 ± 8.1 660 ± 21.0 414–846 438 ± 17.7 3.5 4.5 5.4 3.5–8.5 4.2 2.6 30 ± 1.2 16–42 24 ± 1.0 286 ± 20.6 107–607 342 ± 34.9 6.7 3.2–9.5 3.6 3.1 1.1–5.3 3.4 9.8 6.3–13.7 6.9 6–36 16 ± 1.7 97–207 317 17.7 89 1.0 3.6–5 5.8 85–267 350 ± 13.5 254–663 701 ± 27.1 2.0 14–33 25 ± 1.2 102–658 181 ± 9.2 2.1–6.3 4.6 2.1–5.3 3.4 5.3–8.4 8.0 4–47 74 ± 3.1 39–108 69 ± 3.2 35–98 85 ± 3.9 286–367 18 15–20 17 4–19 18 1.5 5 40.2 1.5 5 44.7 1.5 5 39.5 4.4–7.8 5.8 4.4–8.2 6.4 5.0–8.1 7.3 228–512 291 ± 10.3 194–410 321 ± 9.8 229–484 288 ± 11.5 457–1025 636 ± 35.2 306–1016 615 ± 39.4 375–1088 995 ± 40.8 16–39 28 ± 0.7 20–35 27 ± 0.9 16–36 31 ± 0.9 87–325 244 ± 13.6 130–404 275 ± 22.4 136–572 386 ± 36.6 1.1–6.3 4.8 3.2–7.4 2.3 1.1–4.2 4.7 2.1–5.3 2.2 0–4.2 2.4 0.0–4.2 2.3 5.3–10.5 7.1 5.3–10.5 4.7 3.2–6.3 7.1 49–130 80 ± 3.6 37–121 14 ± 0.9 6–26 10 ± 0.7 5–18 13 ± 1.3 13–21 18 15–21 294 240–333 295 263–374 104 1.5 6 38.4 2.3 18 23.0 3 19 16.0 2.8 17 23.0 5.2–10.4 6.2 4.6–8.7 3.8 2.9–5.1 5.0 3.2–7.7 2.5 153–406 274 ± 13.3 120–424 176 ± 7.6 118–265 175 ± 7.5 84–258 135 ± 5.3 586–1314 813 ± 30.9 367–1114 430 ± 13.8 258–586 546 ± 20.9 256–815 572 ± 13.7 4.2 22–44 30 ± 0.9 20–38 29 ± 1.0 16–40 20 ± 1.0 12–32 22 ± 0.8 198–1060 310 ± 19.8 121–537 237 ± 20.8 78–602 242 ± 24.5 98–616 276 ± 20.5 2.1–6.3 2.8 1.1–5.3 4.9 3.7–6.5 4.4 2.1–8.4 6.2 1.1–3.2 2.3 0 – 4.2 5.6 3.7–8.4 8.6 4.2–13.7 5.2 5.3–9.5 5.2 4.2–6.3 10.6 8.4–13.0 13.1 11.6–15.8 11.4 4–34 16 ± 1.1 67–163 281 15.9 174 13.8 1.4–4.1 4.8 74–187 193 ± 8.6 458–704 517 ± 15.4 2.7 16–37 27 ± 0.9 117–511 33 ± 21.5 3.2–8.4 3.6 2.1–9.5 3.4 9.5–12.6 7.1 6–27 17 ± 1.7 254.5–334.6 134 14.6 38 2.3 3.9–6.1 5.3 121–300 205 ± 9.0 348–678 576 ± 20.6 2.8 16–40 32 ± 1.5 162–564 336 ± 19.2 1.1–6.3 4.7 0.0–10.5 3.8 5.3–1.6 8.6 8–40 13 ± 1.0 124–154 275 16.1 51 20.0 4.0–7.0 5.0 111–345 160 ± 10.5 335–813 421 ± 11.0 2.6 21–52 31 ± 1.1 157–650 230 ± 15.9 2.8–6.5 2.6 1.9–6.5 8.6 7.4–10.2 11.2 4–16 18 ± 1.6 6–35 259–303 415 314–581 16.4 120 2.0 3.6–6.3 6.7 5.3–8.1 112–445 158 ± 6.1 93–229 344–571 590 ± 16.1 394–764 20–44 29 ± 0.7 22–38 103–513 336 ± 26.2 39–670 0.9–3.7 3.8 2.1–6.3 6.5–10.2 5.1 3.2–7.4 8.4–13.0 8.9 6.3–10.5 Legend: I — Tangential diameter of vessels, μm. II — Number of vessel lumina per mm2. III — Mean number of vessel lumina per group. IV — Max number of vessel lumina per group. V — Percent of solitary vessels. VI —Intervessel pit diameter, μm. VII — Length of vessel elements, μm. VIII — Length of fibres, μm. IX — F/V ratio (Fibre length/Vessel element length ratio). X — Width of multiseriate rays, μm. XI — Height of multiseriate rays, μm. XII — Number of multiseriate rays per mm. XIII — Number of uniseriate rays per mm. XIV — Total number of rays per mm. a In Сyclopia plicata all vessels are united in one group. 2.2 1.9 3.5 3 2.4 3.1 3.7 A.V. Stepanova et al. / South African Journal of Botany 89 (2013) 244–256 Species and voucher specimens (all in JRAU) A.V. Stepanova et al. / South African Journal of Botany 89 (2013) 244–256 2. Material and methods We studied the wood anatomy of all nine genera of Podalyrieae, with a representative sample between one and nine species of each genus. The source of material was the collection of FAA-fixed samples, housed in the herbarium of the Department of Botany and Plant Biotechnology at the University of Johannesburg. As shown in Table 1, each sample has a voucher specimen (in the UJ Herbarium, acronym JRAU), collected by A.-L. Schutte (ALS), A.A. Oskolski (AO), A.V. Stepanova (AS), B.-E. van Wyk (BEVW), JSB (J.S. Boatwright) and E.L. Kotina (KK). The taxa studied, collection numbers, and stem diameters are shown in the Results section (in the generic descriptions of the wood anatomy and also in the summary of the results in Table 1). The age of the woody stems or shoots was no less than three years, with diameters ranging from 4 to 55 mm. Transverse, radial, and tangential sections were made on freezing (shoots with a small diameter) or rotary microtomes (Ernst Leitz 247 GMBH, Wetzlar, Germany and Jung AG Heidelberg, Germany), and then stained with a 1:1 alcian blue/safranin mixture (Jansen et al., 2004). Macerations were made using Jeffrey's solution (Johansen, 1940). Where necessary, the wood was softened in ethylenediamine (Carlquist, 1982) followed by soaking in 15% glycerol in ethanol or boiling in 10% glycerol solution (Jansen et al., 1998). The ethylenediamine method for softening very hard woods was first used by Kukachka (1977) and also by Maclachlan and Gasson (2010) but we followed the protocol of Carlquist (1982). The descriptive terminology used is in accordance with that of the IAWA Committee (1989). A special method was used for counting the number of vessels in wood with a dendritic pattern of vessel arrangement. Since the groups are often very large whilst the vessels are very small, it is difficult to count the number of vessels per group. For every sample, we chose five groups of vessels, measured their size and then counted the number of vessels in each group. The mean ratio of vessel number per unit area Fig. 1. Wood porosity and vessel arrangement, LM, TS. (A) diffuse-porous wood, vessels solitary and in radial multiples, Calpurnia sericea; (B) diffuse-porous wood, vessels in clusters, Cyclopia maculata; (C) diffuse-porous wood, tangential multiples at the beginning of growth ring, Podalyria myrtifolia; (D) semi-ring-porous wood (weakly so), vessels in a dendritic pattern, Cyclopia plicata; (E) diffuse-porous wood, vessels in a dendritic pattern, Liparia racemosa. Scale bars: 200 μm. 248 A.V. Stepanova et al. / South African Journal of Botany 89 (2013) 244–256 of the group was then calculated. Using this ratio, we determined the vessel number of 30 groups for each of the samples. Standard methods were used (IAWA Committee, 1989) for all other parameters. Measurements and photographs were made, respectively, with an Olympus ColorView Soft Imaging System (Olympus Soft Imaging Solutions, Stream Essentials version 1.8). Crystals were studied under polarised light. Scanning electron microscope (SEM) observations and photographs were made on a Tescan Vega TC SEM. Samples for SEM study were first preserved in 96% ethanol and then dried in an Edward Tissue Dryer (ETD 4). Evolutionary pathways for wood anatomical features were reconstructed by mapping their character states on a subsample of one of the 140 equally most parsimonious trees from the combined molecular analysis of ITS and rbcL data for Podalyrieae taken from Boatwright et al. (2008). Character optimisation along tree branches was visualised using the parsimony reconstruction method with the “Character History Tracing” option in the computer package Mesquite 2.0 (Maddison and Maddison, 2007). 3. Results 3.1. General wood description Wood mainly diffuse-porous (Fig. 1A–C, F). Semi-ring porosity was observed in some species of Cyclopia (Fig. 1D). Growth rings commonly distinct to faint or absent in two species of Podalyria, some species of Cyclopia, and some specimens of Virgilia, Liparia splendens and Amphithalea ericifolia. Width of growth rings commonly 0.4–1 mm in Fig. 2. Vessels characters. (A) Vessels with simple perforation plates and spiral thickening, Cyclopia plicata, LM, RS; (B) spiral thickening, Amphithalea rostrata, SEM; (C) vestured intervessel pits with scanty vestures located at edge of inner aperture viewed from the outside, appearing as simple rounded and weakly branched warts, Cyclopia buxifolia, SEM; (D) vestured intervessel pits with numerous vestures varying from simple and unforked warts to strongly branched fine protuberances viewing from the outside, Calpurnia sericea, SEM; (E) vestured intervessel pits with numerous seemingly unbranched vestures near inner pit apertures viewed from the outside, Virgilia oroboides subsp. oroboides, SEM; (F) deposits inside vessels, Xiphotheca canescens, LM, TS; (G) deposit inside vessels, Virgilia oroboides subsp. oroboides, LM, RS. Scale bars: 100 μm (A, F), 20 μm (B), 5 μm (C, D), 10 μm (E), 200 μm (G). A.V. Stepanova et al. / South African Journal of Botany 89 (2013) 244–256 249 Fig. 3. Fibre wall thickness and axial parenchyma distribution, LM, TS. (A) Thin- to thick-walled fibres, Stirtonanthus taylorianus; (B) very thick-walled fibres, Stirtonanthus insignis; (C) gelatinous fibres, Podalyria rotundifolia; (D) axial parenchyma scanty paratracheal, Cadia purpurea; (E), axial parenchyma scanty paratracheal and marginal, (also vasicentric and in narrow bands), Amphithalea rostrata; (F) axial parenchyma confluent, Calpurnia intrusa. Scale bars: 50 μm (A–C), 100 μm (D), 200 μm (E, F). shrubs and 1.4–2.5 mm in trees. Boundaries of growth rings may be marked by rows of flattened fibres or flattened vascular tracheids and/ or narrow vessels (in species with very thick-walled fibres) and/or by marginal parenchyma. Vessels narrow, round to angular in outline, relatively rare (Fig. 1A) to very numerous (Fig. 1D, F), commonly in dendritic pattern (Fig. 1D, F), less often in radial and oblique multiples and clusters (Fig. 1A, B), or in tangential multiples in Podalyria, Virgilia and Xiphotheca (Fig. 1C). Perforation plates simple (Fig. 2A), intervessel pits alternate, vestured, 3–8 μm in diameter (up to 10 in Virgilia spp.), round and oval to polygonal with narrow oval to slit-like apertures (Fig. 2C–E). Vessel-ray pits similar to intervessel ones in size and shape. Helical thickenings in vessels absent in large trees (Cadia, Calpurnia and Virgilia) and in two shrubby species (Amphithalea rostrata and Liparia splendens), but present in almost all species of shrubby genera (Fig. 2A, B), occurring throughout the bodies of the vessel elements of all diameters (and not only in the tails). Brown to yellowish-brown deposit (Figs. 1C, 2F, J) in vessels, vascular tracheids and adjacent parenchyma cells in all genera examined except Stirtonanthus. Vascular tracheids commonly present, more or less numerous in woods with dendritic pattern of vessel arrangement and rare in woods with isolated vessel groups. Fibres nonseptate, thin- to thick-walled (Fig. 3A) or very thick-walled (Fig. 3B, E), with small simple pits, more common on the radial walls; thin- to thick-walled fibres often gelatinous (Fig. 3C). Axial parenchyma scanty paratracheal to vasicentric, rarely confluent (Fig. 3D–F), often also in narrow bands of 1–6 cells wide (Fig. 3E), sometimes marginal, fusiform (Fig. 4C), and in 2, sometimes up to 4 cells per strand (Fig. 4A). Rays commonly 1 or 2- or 1–3-seriate (Fig. 4B–E), rarely broader (up to 6), 1–6 times higher than axial parenchyma strands, composed of procumbent, square and upright cells, perforated ray cells often present (Fig. 4F). Prismatic crystals occur in ray cells in some species of Calpurnia and Amphithalea (Fig. 5A, B); navicular crystals, often aggregated (Fig. 5C) and crystal sand common for two species 250 A.V. Stepanova et al. / South African Journal of Botany 89 (2013) 244–256 Fig. 4. Axial parenchyma strands and rays. (A) Axial parenchyma strand of 4 cells, Podalyria lanceolata, LM, TLS; (B) axial parenchyma in 2 or 3 cells per strand and 2 or 3-seriate rays, Calpurnia intrusa, LM, TLS; (C) fusiform axial parenchyma cells, Xiphotheca canescens, LM, TLS; (D) 1 or 2-seriate rays, Virgilia oroboides subsp. ferruginea, LM, TLS; (E) 1 or 2-seriate rays, Cyclopia buxifolia, LM, TLS; (F) perforated ray cell, Stirtonanthus taylorianus, LM, RLS. Scale bars: 50 μm (A, C, F), 100 μm (B), 200 μm (D, E). of Amphithalea; acicular crystals in sheaf-like aggregates present in almost all species of Cyclopia and Virgilia (Fig. 5D, E), and also in one species of Calpurnia. Silica bodies commonly occur in rays and axial parenchyma cells (Fig. 5F) in almost all genera, but their presence often varies within genera. They are small (up to 3 μm in diameter), spheroidal and yellowish, and usually solitary. Storied structures were not found. thickenings not found. Vascular tracheids rarely occur. Fibres very thick-walled. Axial parenchyma fusiform and in strands of 2 cells, scanty paratracheal, in solitary strands near vessels, and sometimes diffuse. Rays 1- or 2-seriate, formed mostly by procumbent and square cells; sometimes with 1 or 2 marginal rows of upright cells. Crystals not observed. Silica bodies in ray and axial parenchyma cells. Brown deposits in a few vessels. 3.2. Generic wood descriptions 3.2.1. Cadia Material studied: C. purpurea Ait. (KK 33–10, 15 mm). Wood diffuse-porous. Growth rings indistinct, marked by 1–3 layers of radially flattened fibres. Vessels narrow, very thick-walled, rounded in outline, solitary and in short radial multiples of 2–7. Intervessel pits alternate, vestured, with abundant seemingly unbranched vestures filling the pit chambers. Vessel-ray pits with distinct borders. Helical 3.2.2. Calpurnia Material studied: C. aurea Baker (KK 39–11, 22 mm), C. intrusa E. Mey. (BEVW 3249 C, 22 mm), C. sericea Harv. (KK 35–11, 16 mm). Wood diffuse-porous. Growth rings absent in C. intrusa, or distinct, marked by flattened latewood fibres and commonly also by marginal axial parenchyma. Vessels narrow, rounded and angular in outline, solitary and in radial multiples of 2–6. Intervessel pits alternate, vestured, in C. sericea with rather numerous vestures, adjacent to the A.V. Stepanova et al. / South African Journal of Botany 89 (2013) 244–256 251 Fig. 5. Inclusions, SEM. (A) Prismatic crystals in ray cells, Calpurnia aurea; (B), starch grains and very small prismatic crystals in ray cell, Amphithalea rostrata; (C) navicular crystals in aggregates in ray cell, Amphithalea rostrata; (D) acicular crystals in sheaf-like aggregates in ray cell, Cyclopia aurescens; (E) acicular crystals in sheaf-like aggregates in vessel, Virgilia oroboides subsp. oroboides; (F) silica body in ray cell, Cyclopia aurescens. Scale bars: 10 μm (A, B, D, F), 2 μm (C), 50 μm (E). inner pit apertures, varying from simple and unforked warts to strongly branched fine protuberances (vestures in C. aurea and C. intrusa were not examined by SEM). Vessel-ray pits with distinct borders. Helical thickenings not found. Vascular tracheids rare. Fibres very thickwalled (C. intrusa) or thin- to thick-walled. Axial parenchyma fusiform or in strands of 2 or 3 cells (C. sericea), or of (2–)3–4(−5) cells (C. aurea and C. intrusa). Axial parenchyma scanty paratracheal, in incomplete uniseriate sheaths near vessels (C. aurea and C. sericea), or vasicentric, sometimes aliform to confluent (C. intrusa), and also banded in 2–8-seriate tangential bands (in C. aurea and C. introsa also in marginal bands). Rays are 1–2(3)-seriate (C. sericea) or 1–3(4)–seriate. Multiseriate rays mostly of square and upright cells with few procumbent cells scattered throughout (C. aurea and C. sericea), or mostly of procumbent cells with 1 or 2 marginal rows and incomplete sheaths of square and upright cells (C. intrusa). Prismatic crystals in ray cells of C. aurea and C. sericea. Аcicular crystals in sheaf-like aggregates in vessels and parenchyma cells of C. intrusa. Silica bodies in ray and axial parenchyma cells of C. intrusa. Brown deposits in vessels rarely in C. intrusa and C. sericea. 3.2.3. Cyclopia Material studied: C. aurescens Kies. (ALS + BEVW 771, 9 mm), C. burtonii Hofmeyr & E. Phillips (ALS 641, 10 mm), C. buxifolia (Burm.f.) Kies (ALS 544, 10 mm), C. genistoides (L.) Sieber ex C. Presl (ALS 615, 7 mm; ALS 624, 10 mm), C. intermedia E. Mey. (ALS 646, 6 mm; ALS 647, 10 mm), C. maculata (Andrews) Kies (ALS 636, 9 mm; ALS 528, 14 mm), C. plicata Kies (ALS 670, 14 mm), C. subternata Vogel (KK 43–11, 22 mm). Wood semi-ring-porous (at least weakly so) in C. plicata, C. aurescens and C. buxifolia, or diffuse-porous. Growth rings absent (C. maculata and C. subternata), distinct (C. aurescens, C. burtonii, C. buxifolia and C. plicata) or inistinct, marked by 1 or 2 rows of flattened latewood fibres, narrow vessels and vascular tracheids, tangentially extended groups of vessels and vascular tracheids in some species (C. aurescens, 252 A.V. Stepanova et al. / South African Journal of Botany 89 (2013) 244–256 C. buxifolia and C. plicata) form more or less distinct rings of wider earlywood vessels; commonly growth ring boundaries also marked by marginal axial parenchyma. Vessels narrow (5–50 μm in diameter) and numerous (N125 per mm2), in clusters and radial multiples arranged in diagonal rows (C. maculata [AL 528]), or mostly in large groups (N10 vessels) arranged in diagonal to dendritic pattern, or in continuous dendritic aggregations (C. plicata) or diagonal bands (C. subternata) with co-occurrence of a few solitary vessels (no solitary vessels found in C. aurescens, C. genistoides [AL 615], and C. plicata) and groups of 2–8 vessels. Intervessel pits alternate, vestured, with scanty vestures located at edge of inner aperture, appearing as simple rounded (C. burtoni, C. intermedia and C. subternata) and also weakly branched (C. buxifolia and C. genistoides) warts (vestures in C. aurescens, C. maculata and C. plicata not examined by SEM). Vessel-ray pits with distinct (C. burtonii) or reduced borders. Helical thickenings present. Vascular tracheids common. Fibres thin- to thick-walled in C. maculata, very thick-walled in other species. Axial parenchyma fusiform (C. plicata) or also in strands of 2 cells (up to 3 cells in C. aurescens and C. maculata [AL 636]), vasicentric (C. intermedia, C. plicata and C. subternata) or scanty paratracheal (mostly as incomplete sheaths near vessel groups), also in 1–3-seriate tangential (sometimes marginal) bands (up to 5seriate in C. burtonii, up to 7-seriate in C. aurescens and C. gentionoides, up to 12-seriate in C. plicata). Rays 1–2(3)-seriate (C. buxifolia and C. genistoides) or 1–3 (4)-seriate (up to 5-seriate in C. burtonii and C. maculata [AL 636]), composed mostly of procumbent cells (C. burtonii, C. maculata, C. plicata and C. subternata) with square and upright cells in few (up to 5 in C. maculata) marginal rows and in incomplete sheaths (C. maculata and C. plicata), or mostly of square cells with procumbent and upright cells mixed throughout. Acicular crystals in sheaf-like aggregates in vessels and ray cells in all species examined except C. plicata. Small bands of acicular crystals in intervessel pits in C. intermedia. Silica bodies in ray and axial parenchyma cells. Yellowish-brown deposits common in vessels of C. intermedia, rare in C. maculata and C. subternata. 3.2.4. Liparia Material studied: L. hirsuta Moench (JSB 595, 13 mm), L. myrtifolia Thunb. (ALS 727, 12 mm), L. racemosa A.L. Schutte (ALS 642, 19 mm), L. splendens (Burm.f.) Bos & de Wit (BEVW 3147, 11 mm). Wood diffuse-porous. Growth rings absent (L. splendens), indistinct (L. myrtifolia) to distinct (L. hirsuta and L. racemosa), marked by flattened latewood fibres, and also by tangentially extended groups of vessels and vascular tracheids in earlywood, and by difference in vessel diameter between late- and earlywood (L. hirsuta and L. racemosa). Vessels narrow, rounded and angular in outline, mostly in radial multiples and clusters (L. splendens), or in nearly continuous diagonal bands and dendritic patterns, rarely solitary and in small clusters. Intervessel pits alternate, vestured, in L. splendens with scanty vestures at edge of inner aperture, appearing as simple warts (vestures in L. hirsuta, L. myrtifolia and L. racemosa not examined by SEM). Vessel-ray pits with reduced borders. Helical thickenings absent (L. splendens) or present. Vascular tracheids common. Fibres thin- to thick-walled. Axial parenchyma commonly fusiform, rarely in strands of 2 cells, scanty paratracheal, in incomplete sheaths near vessel groups, and also in 2–6-seriate tangential bands (L. splendens). Rays 1 or 2-seriate, of square and upright cells with few procumbent cells (L. splendens), or 1–3(4)seriate (up to 5-seriate in L. racemosa), with predominance of procumbent cells in both multi- and uniseriate rays; square and upright cells in 1 or 2 marginal rows and occasionally as solitary sheath cells. Crystals not observed. Silica bodies in ray cells. Brown deposits in a few vessels of L. myrtifolia. 3.2.5. Podalyria Material studied: P. calyptrata (Retz.) Willd. (KK 03–10, 22 mm), P. lanceolata Benth. (ALS s.n., 23 mm), P. myrtifolia D. Dietr. (AS + BEVW 166, 7 mm), P. rotundifolia (P.J. Bergius) A.L. Schutte (ALS s.n., 24 mm). Wood diffuse-porous. Growth rings absent (P. calyptrata and P. rotundifolia), or distinct, marked by flattened latewood fibres, 2–4seriate bands of marginal axial parenchyma, and by difference in vessel diameter between late- and earlywood. Vessels narrow to moderately wide, rounded in outline, solitary and in clusters and radial multiples of 2–10, tending to form diagonal to dendritic arrangement most distinctive in P. lanceolata and P. myrtifolia. Intervessel pits alternate, vestured, in P. rotundifolia with scanty vestures located at edge of inner aperture, appearing as flattened weakly branched protuberances (vestures in P. calyptrata, P. lanceolata and P. myrtifolia not examined by SEM). Vessel-ray pits with distinct borders. Helical thickening present. Vascular tracheids common. Fibres thin- to thick-walled. Axial parenchyma in strands of 2–4 cells, scanty paratracheal to vasicentric, in incomplete to complete 1–3-seriate sheaths near vessel groups, sometimes aliform and confluent, and also marginal in 1–4-seriate bands (P. lanceolata and P. myrtifolia). Rays are 1 or 2(3)-seriate (up to 4-seriate in P. rotundifolia), composed mostly of square and upright cells (P. myrtifolia), or mostly of procumbent cells, with 1–4 marginal rows of square and upright cells. Crystals not observed. Silica bodies in ray cells of P. lanceolata and P. myrtifolia. Yellowish-brown deposits in a few vessels of P. myrtifolia. 3.2.6. Stirtonanthus Material studied: S. insignis (Compton) B.-E.van Wyk & A.L. Schutte (BEVW 3331, 14 mm), S. taylorianus (L. Bolus) B.-E.van Wyk & A.L. Schutte (BEVW 3169, 28 mm). Wood diffuse-porous. Growth rings distinct, marked by flattened latewood fibres, and also by marginal axial parenchyma (S. insignis) or by difference in vessel diameter between late- and earlywood (S. taylorianus). Vessels narrow, rounded to angular in outline, in dendritic pattern. Intervessel pits alternate, vestured, with scanty vestures located at edge of inner aperture, appearing as simple rounded warts. Vessel-ray pits with reduced borders. Helical thickenings present. Vascular tracheids common. Fibres thin- to thick-walled in S. taylorianus and very thick-walled in S. insignis. Axial parenchyma in strands of 2(3) cells, sometimes fusiform, scanty paratracheal to vasicentric and confluent, in incomplete to complete uniseriate (sometimes up to 3seriate) sheaths near vessel groups, and also banded, in interrupted tangential (sometimes marginal) 1–3-seriate bands (S. insignis). Rays 1–3seriate, mostly of procumbent cells, sometimes with 1–3 marginal rows of square and upright cells (S. taylorianus), or mostly of square cells mixed with procumbent and upright cells throughout ray (S. insignis). Crystals not observed. Silica bodies in ray cells. Deposits in vessels not observed. 3.2.7. Virgilia Material studied: V. divaricata Adamson (AO 16-09, 43 mm; KK 0210, 28 mm), V. oroboides subsp. oroboides (P.J. Bergius) T.M. Salter (BEVW 5722, 55 mm), V. oroboides subsp. ferruginea B.-E. van Wyk (KK 44-11, 23 mm). Wood diffuse-porous. Growth rings absent or indistinct, marked by tangential rows of larger vessel groups (V. divaricata [AO 16-09]). Vessels narrow to relatively wide, angular to rounded in outline, solitary, in clusters and radial multiples of 2–4(6). Intervessel pits alternate, vestured, in V. oroboides subsp. oroboides with rather numerous seemingly unbranched vestures near inner apertures (vestures not examined by SEM in V. divaricata and V. oroboides subsp. ferruginea). Vessel-ray pits with distinct borders. Helical thickenings absent. Vascular tracheids rare. Fibres thin- to thick-walled. Axial parenchyma in strands of 2–5 cells, scanty paratracheal to vasicentric, sometimes aliform, in complete or incomplete 1–3-seriate sheaths near vessels and vessel groups. Rays 1–3-seriate. Bi- and triseriate rays of procumbent cells, sometimes with 1–2(4) marginal rows of upright and square cells. Uniseriate rays consist mostly of square and upright cells, with few procumbent cells. Acicular crystals in sheaf-like aggregates in ray and axial parenchyma cells. A.V. Stepanova et al. / South African Journal of Botany 89 (2013) 244–256 253 Silica bodies in ray cells of V. divaricata (AO 16-09). Yellowish-brown deposits in a few vessels. a predominance of procumbent cells in the rays, whereas the sprouters share rays that consist mostly of square and upright cells. 3.2.8. Amphithalea Material studied: A. ericifolia (L.) Eckl. & Zeyh. (ALS 617, 4 mm), A. rostrata A.L. Schutte & B.-E. van Wyk (ALS 629, 5 mm), A. vlokii (A.L. Schutte & B.-E.van Wyk) A.L. Schutte (ALS 743, 5 mm). Wood diffuse-porous. Growth rings absent (A. ericifolia) or distinct to faint marked by marginal parenchyma. Vessels very narrow (5–18 μm in tangential diameter), mostly rounded in outline, in radial multiples and clusters of 2–10 vessels, rarely solitary, with tendency to dendritic arrangement. Intervessel pits alternate, vestured (vestures not examined by SEM). Helical thickenings on vessel walls common (A. vlokii), rare (A. ericifolia) or absent (A. rostrata). Vessel-ray pits with distinct borders. Vascular tracheids common. Fibres very thickwalled. Axial parenchyma fusiform and in strands of 2 cells (A. vlokii) or 2–4 cells (A. rostrata), scanty paratracheal (A. vlokii) to vasicentric (A. rostrata), and marginal, in 1 or 2-seriate tangential rows (A. vlokii) or 2–4-seriate bands (A. rostrata). Rays 1 or 2(3)-seriate, mostly of square cells mixed with procumbent cells throughout ray. Small prismatic crystals, solitary or aggregated small navicular crystals, and crystal sand in ray cells of A. rostrata and A. vlokii. Silica bodies common in ray and axial parenchyma cells of A. rostrata. Brown deposits in vessels in all studied species. 3.4. Character evolution within the Podalyrieae 3.2.9. Xiphotheca Material studied: X. canescens (Thunb.) A.L. Schutte & B.-E.van Wyk (ALS 595a, 9 mm), X. elliptica (DC.) A.L. Schutte & B.-E.van Wyk. (ALS 600, 8 mm), X. phylicoides A.L. Schutte & B.-E.van Wyk (ALS 648, 5 mm), X. tecta (Thunb.) A.L. Schutte & B.-E.van Wyk (ALS 597, 11 mm). Wood diffuse-porous. Growth rings absent (X. tecta) or distinct, marked by marginal parenchyma, flattened latewood fibres, and sometimes also by tangential band of vessels and vascular tracheids in earlywood (X. canescens). Vessels narrow, rounded to angular in outline, mostly in clusters of 3–6 and large radial multiples (X. tecta), or in dendritic pattern (X. canescens). Intervessel pits alternate, vestured in X. tecta with scanty vestures located at edge of inner aperture, appearing as weakly branched coarse protuberances (vestures not examined by SEM in X. canescens, X. elliptica and X. phylicoides). Vessel-ray pits with reduced to distinct borders. Helical thickenings present. Vascular tracheids common. Fibres very thick-walled. Axial parenchyma fusiform and in 2-celled strands, scanty paratracheal (X. tecta) to mostly vasicentric (X. canescens), and also in uniseriate marginal rows (X. canescens) or in 1–4 seriate tangential bands (X. tecta). Rays 1 or 2(3)-seriate, up to 4-seriate in X. elliptica. Multiseriate rays mostly of square and upright cells mixed with few (X. tecta) or quite numerous (X. canescens) procumbent cells throughout ray. Crystals not observed. Silica bodies in ray and axial parenchyma cells. Yellowish-brown deposits in a few vessels of X. canescens and X. tecta. 3.3. Numerical analysis The effects of fire-survival strategies on wood structure were assessed by one-way analysis of variance (ANOVA) using the F-test for quantitative wood characters (Table 1), and the Chi-square test for two types of rays (with a predominance of procumbent cells vs square and upright cells). The variability of wood characters was compared amongst and within two groups of samples representing seeders (reseeders) and sprouters (resprouters). As these groups differ greatly in the average size of the wood samples (mean sample radius 10.8 mm for seeders and 5.0 mm for sprouters), the 12 largest samples of both groups (with a radius N9 mm) were excluded from the data matrix to avoid the potential influence of sample size on the results. A comparison of the wood anatomical data between the two groups revealed a statistically significant effect (p b 0.0001) of fire-survival strategy on the ray composition. All the seeder samples included show Character states of the quantitative wood features listed in Table 1 as well as some qualitative ones were plotted on the tree topology based on the combined simultaneous parsimony analysis of ITS and rbcL data for Podalyrieae (Boatwright et al., 2008). Patterns of variation for a few wood features [the occurrence of a dendritic vessel arrangement (Fig. 6A), helical thickenings (Fig. 6B), fibre wall thickness (Fig. 6C) and acicular crystals in sheaf-like aggregates (Fig. 6D)], are apparently consistent with the topology of the phylogenetic tree and support the four main clades in Fig. 6. 4. Discussion The wood diversity within the tribe Podalyrieae is due mainly to variations in the grouping and arrangement of vessels, as well as in the occurrence of helical thickenings on the vessel walls. Most members of the genera Amphithalea, Cyclopia, Liparia, Podalyria, Stirtonanthus, and Xiphotheca share large vessel grouping, commonly arranged in dendritic or diagonal patterns, and the presence of helical thickenings. In Cadia, Calpurnia and Virgilia as well as in the single species of Liparia (L. splendens), however, smaller vessel groups (up to 7 vessels), without a distinct arrangement, co-occur with the absence of helical thickenings. An apparent loss of helical thickenings in combination with large vessel grouping is found only in Amphithalea rostrata, whereas Cyclopia maculata shows relatively small vessel grouping and very prominent helical thickenings. An increasing degree of vessel grouping — a transition from solitary vessels and small isolated groups to large groups and a dendritic pattern, has been noted in genistoid tribes closely related to Podalyrieae, i.e. in Genisteae, Thermopsideae and Crotalarieae (Metcalfe and Chalk, 1950; Yatsenko-Khmelevsky, 1954; Grosser, 1977; Baretta-Kuipers, 1981; Fahn et al., 1986; Schweingruber, 1990; Van Wyk and Schutte, 1995; InsideWood, 2004–onwards; Schweingruber et al., 2011; and our unpublished observations) as well as in some members of the tribes Sophoreae (Gasson, 1994), Hedysareae (Benkova and Schweingruber, 2004); Galegeae (Fahn et al., 1986), Loteae (Schweingruber et al., 2011) and the subfamily Caesalpinioideae (Cozzo, 1951; Schweingruber et al., 2011). This trend can be postulated to be an adaptation to water stress, because the vessel groups provide the bypass water conduits in the case of an embolism of some vessels (Zimmermann, 1983; Carlquist, 1987; Sperry, 2003). This hypothesis on the adaptive value of vessel grouping was recently demonstrated in experiments (Lens et al., 2011). Helical thickenings are mostly present in shrubs growing in Mediterranean climates. High percentages of species with spiral thickening were noted in Californian chaparral as well as in maquis and batha biomes in Israel (Baas and Carlquist, 1985; Carlquist, 1987). Batha is a secondary vegetation type comprising Mediterranean semi-shrubs that colonise abandoned cultivated lands in Israel. The shifts in fire-survival strategy between seeders and sprouters are considered by Schnitzler et al. (2011) to be one of the most important factors for the evolutionary diversification of the Podalyrieae. Recently, Pratt et al. (2012) found that sprouters show lower water stress resistance but higher efficiency of xylem transport than seeders. Within Podalyrieae, no significant differences were found in the wood characters related to the safety and efficiency of water transport (vessel diameter, frequency and grouping) between the species with these two strategies. Nevertheless, we found that the rays in seeders consist of numerous procumbent cells whilst those of sprouters have mainly square and upright cells. This difference in ray composition is unlikely to be related to the safety or efficiency of xylem transport. Rather, this character reflects a difference in the rate of radial growth (especially during the 254 A.V. Stepanova et al. / South African Journal of Botany 89 (2013) 244–256 Fig. 6. Distribution of the wood character states reconstructed onto a subsample of the tree based on the combined simultaneous parsimony analysis of ITS and rbcL data for Podalyrieae (modified from Boatwright et al., 2008). (A) Occurrence of dendritic vessel arrangement: white — no arrangement, black — diagonal or dendritic arrangement; (B) occurrence of helical thickenings: white — absent, black — present; (C) fibre wall thickness: white — thin or thin to thick, black — very thick; (D) occurrence of acicular crystals in sheaf-like aggregates: white — absent, black — present. early stages of stem development) that seems to be faster in seeders than in sprouters. This explanation agrees with the difference in habit between these fire-survival strategies. Commonly, seeders form a single, tall and relatively thick stem, whereas sprouters produce numerous thin stems from an underground lignotuber (Schutte et al., 1995). The wood anatomy of all tree genera is very similar, and also strongly supports the inclusion of Cadia in the Podalyrieae (Boatwright et al., 2008). It was shown that Cadia should be excluded from the Cadia group of Sophoreae s.l. because it has very narrow rays and no crystals (Gasson, 1994). Our study shows that most of the Podalyrieae have 1 or 2- or 1–3-seriate rays. In four genera (Liparia, Podalyria, Stirtonanthus and Xiphotheca), crystals were not observed. The similarity in wood anatomy is especially close between Cadia and Calpurnia, the main difference being the presence of crystals in the latter. As molecular phylogenetic analysis (Boatwright et al., 2008) suggested, Cadia belongs to the basal-most lineage of Podalyrieae. Thereafter the Cyclopia clade diverges. Then Calpurnia and Virgilia form a weakly supported clade that is sister to the rest of Podalyrieae. The members of these basally diverged clades show some common wood features. Cadia, Calpurnia and Virgilia have pits with numerous vestures, vessel-ray pits with distinct borders as well as the absence of helical thickenings and relatively small vessel groups. On the other hand, Calpurnia Cyclopia and Virgilia, share the very distinctive feature of the occurrence of acicular crystals in sheaf-like aggregates. Although the A.V. Stepanova et al. / South African Journal of Botany 89 (2013) 244–256 patterns of variation in these wood characters (Fig. 6A, B and D) are consistent with the topology of the phylogenetic tree (including the monophyly of the Virgilia/Calpurnia clade) proposed by Boatwright et al. (2008), the wood anatomical data do not contribute to the clarification of relationships between the early diverged lineages of Podalyrieae which remain poorly resolved. As for the rest of the Podalyrieae, the members of the Liparia/ Podalyria/Stirtonanthus subclade have mostly thin- to thick-walled fibres whereas all species of its sister subclade, comprising Amphithalea and Xiphotheca, show the presence of very thick-walled fibres. As the character mapping suggests, modifications from very thick-walled fibres in Podalyrieae to thinner fibre walls occurred at least three times in the course of its evolution (i.e. in the Cyclopia, Virgilia/Calpurnea, and Liparia/Podalyria/Stirtonanthus clades). The Podalyrieae show a remarkable diversity of crystals in the axial and radial parenchyma. Amongst its related genistoid tribes Thermopsideae, Genisteae and Crotalarieae, the presence of crystals has thus far been reported only in a single species of Calobota belonging to the Crotalarieae (Metcalfe and Chalk, 1950; YatsenkoKhmelevsky, 1954; Grosser, 1977; Fahn et al., 1986; Schweingruber, 1990; Schweingruber et al., 2011; InsideWood, 2004–onwards; Oskolski et al., in prep.). Prismatic crystals are commonly found in Fabaceae wood (Gasson, 1994), whilst in Podalyrieae we also observed acicular crystals in sheaf-like aggregates, solitary and aggregated navicular crystals, as well as crystal sand. Different types of crystals may be present in the same genus as, for example, in Calpurnia. Acicular crystals in sheaf-like aggregates have been described in the bark of Virgilia (Kotina et al., 2013-in this issue). It is also noteworthy that in Fabaceae, crystals occur mainly in axial parenchyma cells, commonly in chambered parenchyma and less often in ray cells. In a few well-studied tribes of Papilionoideae, prismatic crystals in ray cells were not found in Dipterygeae, whilst they do occur in Millettieae, Sophoreae, and Swartzieae (Fujii et al., 1994; Gasson, 1994, 2000; Gasson et al., 2004). However, in all these tribes, species with crystals in the rays are much less numerous than those with crystals in chambered axial parenchyma cells. In the species of Podalyrieae, crystals are localised in ray cells but acicular crystals in sheaf-like aggregates are also often found inside vessels. The occurrence of acicular crystals in the intervessel pits of Cyclopia intermedia is especially noteworthy. Such crystals are very rare within angiosperms and have been described in detail only in the fern genus Botrichium Sw. (Morrow and Dute, 2002). The common occurrence of silica bodies in the ray and axial parenchyma cells in most species of Podalyrieae (this trait is absent or lost only in particular species of Amphithalea, Calpurnia, Podalyria and Virgilia) is rather surprising. Silica bodies occur in several genera of Caesalpinioideae, namely Apuleia Gaertn., Dialium L., Dicorynia Benth., Distemonanthus Benth., Loesenera Harms., Sclerolobium Vogel and Tachigali Aubl. but are not found in any Mimosoideae (InsideWood, 2004). They are also absent in all but one species of Papilionoideae investigated (Gasson, 1994; Fujii et al., 1994; Gasson et al., 2004; et al.). The occurrence of silica bodies in ray cells was reported in Calobota saharae (Coss. and Dur.) Boatwr. & B.-E.van Wyk (InsideWood, 2004, as Genista saharae Coss. & Dur.). This species belongs to the tribe Crotalarieae that is closely related to the Podalyrieae. Thus the occurrence of silica bodies seems to be a shared character state of these two tribes. However, further wood anatomical studies of other genistoid groups are necessary to clarify the distribution and evolutionary trends for this character. 5. Conclusions Our wood anatomical study of several members of all genera belonging to the tribe Podalyrieae has led to the following new insights: (1) wood anatomical characters allow for the distinction between some genera, especially if unique combinations of characters are taken into account; (2) variation in some wood characters (vessel grouping 255 and arrangement, and occurrence of helical thickenings) is consistent with the idea that their evolution occurred independently in different lineages of Podalyrieae in the course of adaptation to seasonally arid environments; (3) unlike the related genistoid tribes, the Podalyrieae show a remarkable diversity of crystals in the axial and radial parenchyma; (4) the placement of Cadia in the Podalyrieae is confirmed by wood anatomical data; (5) the occurrence of sheaf-like crystals and the thickness of fibres are apomorphic features for some subclades of Podalyrieae; and (6) within the Podalyrieae, seeders can be distinguished from sprouters by some ray features related to their habits, but they show no differences in the wood characters related to the safety and efficiency of water transport. Therefore, the observed range of wood diversity related both to habit and habitat, and to the phylogenetic relations within Podalyrieae, provides us with a better understanding of the trends and conditions of plant diversification that have occurred in the Cape region. Acknowledgements The authors thank the National Research Foundation (South Africa), the University of Johannesburg (UJ) and the Russian Foundation of Basic Research (grant # 12-04-01684) for funding. We also thank the personnel of the SEM unit at UJ for their help. References Baas, P., Carlquist, S., 1985. A comparision of the ecological wood anatomy of the floras of Southern California and Israel. IAWA Bulletin 6, 349–353 (n.s.). Baretta-Kuipers, T., 1981. Wood anatomy of Leguminosae: its relevance to taxonomy. In: Polhill, R.M., Raven, P.H. (Eds.), Advances in Legume Systematics 1. Royal Botanical Gardens, Kew, pp. 677–705. Beaumont, A.J., Beckett, R.P., Edwards, T.J., Stirton, C.H., 1999. Revision of the genus Calpurnia (Sophoreae: Leguminosae). Bothalia 29, 5–23. Benkova, V.R., Schweingruber, F.H., 2004. Anatomy of Russian woods. An atlas for the identification of trees, shrubs, dwarf shrubs and woody lianas from Russia. Swiss Federal Institutes for Forest, Snow and Landscape Research.Bern, Stuttgart, Birmensdorf. Boatwright, J.S., Savolainen, V., Van Wyk, B.-E., Schutte-Vlok, A.L., Forest, F., Van der Bank, M., 2008. Systematic placement of the anomalous genus Cadia and the phylogeny of the tribe Podalyrieae. Systematic Botany 33, 133–147. Carlquist, S., 1982. The use of ethylenediamine in softening hard plant structures for paraffin sectioning. Stain Technology 57, 311–317. Carlquist, S., 1987. Diagonal and tangential vessel aggregations in wood: function and relationship to vasicentric tracheids. Aliso 11, 451–462. Cozzo, D., 1951. Anatomia del leno secundario de las Leguminosas Mimosoideas y Caesalpinoides Argentinas silvestres y cultivadas. Revista del Museo Argentino de cienicas Naturales “Bernardino Rivadavia” e Instituto Nacional de Investigacion de las Ciencias Naturales (Argentina). Botanica 2, 63–146. Fahn, A., Werker, E., Baas, P., 1986. Wood anatomy and identification of trees and shrubs from Israel and adjacent regions. Israel Academy of Sciences and Humanities, Jerusalem. Fujii, T., Baas, P., Gasson, P., Ridder Numan, J.W.A., 1994. Wood anatomy of the Sophora group (Leguminosae). In: Ferguson, I.K., Tucker, S. (Eds.), Advances in Legume Systematics 6. Royal Botanic Gardens, Kew, pp. 205–249. Gasson, P., 1994. Wood anatomy of the tribe Sophoreae and related Caesalpinioideae and Papilionoideae. In: Ferguston, I.K., Tucker, S. (Eds.), Advances in Legume systematics 6. Royal Botanical Gardens, Kew, pp. 165–203. Gasson, P., 2000. Does wood anatomy support tribal and generic classification in papilionoid Leguminosae? In: Herendeen, P.S., Bruneau, A. (Eds.), Advances in Legumes Systematics 9. Royal Botanical Gardens, Kew, pp. 201–215. Gasson, P., Wray, E., Schrire, B.D., 2004. Wood anatomy of the tribe Millettieae with comments on related papilionoid Leguminosae. IAWA Journal 25 (4), 485–545. Grosser, D., 1977. Die Hölzer Mitteleuropas. Ein mikrophotographischer Lehratlas, Springer, Berlin. IAWA Committee, 1989. IAWA list of microscopic features for hardwood identification. IAWA Bulletin 10, 221–331. InsideWood, 2004. Onwards, Published on the Internet http://insidewood.lib.ncsu.edu/ search (date of accession: 29.03.2013). Jansen, S., Kitin, P., Pauw, H. De, Idris, M., Beeckman, H., Smets, E., 1998. Preparation of wood samples for transmission light microscopy and scanning electron microscopy. Belgian Journal of Botany 131, 41–49. Jansen, S., Choat, B., Vinckier, S., Lens, F., Schols, P., Smets, E., 2004. Intervascular pit membranes with a torus in the wood of Ulmus (Ulmaceae) and related genera. New Phytologist 163, 51–59. Johansen, D.A., 1940. Plant Microtechnique. McGraw-Hill, New York. Kotina, E.L., Oskolski, A.A., Stepanova, A.V., Van Wyk, B.-E., 2013. Notes on the taxonomic and ecological significance of bark structure in the genus Virgilia (Fabaceae, Podalyrieae). South African Journal of Botany 89, 240–243 (in this issue). Kukachka, B.F., 1977. Sectioning Refractory Woods for Anatomical Studies. USDA Forest Research Note FPL-0236. USDA Forest Service Forest Products Laboratory, Madison, Wisconsin. 256 A.V. Stepanova et al. / South African Journal of Botany 89 (2013) 244–256 Lens, F., Sperry, J.S., Christman, M.A., Choat, B., Rabaey, D., Jansen, S., 2011. Testing hypotheses that link wood anatomy to cavitation resistance and hydraulic conductivity in the genus Acer. New Phytologist 190, 709–723. MacLachlan, I., Gasson, P., 2010. Quantitative wood anatomy using multivariate principal components analysis for identification of the CITES listed Pterocarpus santalinus (Dalbergioideae, Papilionoideae, Leguminosae). IAWA Journal 31, 121–138. Maddison, W.P., Maddison, D.R., 2007. Mesquite: a modular system for evolutionary analysis. Version 2.00. Published on the Internet http://mesquiteproject.org (date of accession: 29.03.2013). Metcalfe, C.R., Chalk, L., 1950. Anatomy of the dicotyledons, 1. Oxford University Press, Oxford. Morrow, A.C., Dute, R.R., 2002. Crystals associated with the intertracheid pit membrane of the woody fern Botrychium multifidum. American Fern Journal 92, 10–19. Pratt, R.B., Jacobsen, A.L., Jacobs, S.M., Esler, K.J., 2012. Xylem transport safety and efficiency differ among fynbos shrub life history types and between two sites differing in mean rainfall. International Journal of Plant Sciences 173, 474–483. Schnitzler, J., Barraclough, T.G., Boatwright, J.S., Goldblatt, P., Manning, J.C., Powell, M.P., Rebelo, T., Savolainen, V., 2011. Causes of plant diversification in the Cape biodiversity hotspot of South Africa. Systematic Biology 60, 343–357. Schutte, A.L., Van Wyk, B.-E., 1998. Evolutionary relationships in the Podalyrieae and Liparieae (Fabaceae) based on morphological, cytological, and chemical evidence. Plant Systematics and Evolution 209, 1–31. Schutte, A.L., Vlok, J.H.J., Van Wyk, B.-E., 1995. Fire-survival strategy — a character of taxonomic, ecological and evolutionary importance in fynbos legumes. Plant Systematics and Evolution 195, 243–259. Schutte-Vlok, A.L., Van Wyk, B.-E., 2011. A taxonomic revision of Podalyria (Fabaceae). Systematic Botany 36, 631–660. Schweingruber, F.H., 1990. Anatomy of European Woods. Verlag Paul Haupt, Bern and Stuttgart. Schweingruber, F.H., Börner, A., Schulze, E.-D., 2011. Atlas of stem anatomy in herbs, shrubs and trees, vol. 1. Springer, Heidelberg. Sperry, J.S., 2003. Evolution of water transport and xylem structure. International Journal of Plant Sciences 164, S115–S127. Van der Bank, M., Chase, M.W., Van Wyk, B.-E., Fay, M.F., Van der Bank, F.H., Reeves, G., Hulme, A., 2002. Systematics of the tribe Podalyrieae (Fabaceae) based on DNA, morphological and chemical data. Botanical Journal of the Linnean Society 139, 159–170. Van Wyk, B.-E., 2005. Tribe Podalyrieae. In: Lewis, L., Schrire, B., Mackinder, B., Lock, M. (Eds.), Legumes of the World. Royal Botanic Gardens, Kew, pp. 267–271. Van Wyk, B.-E., Schutte, A.L., 1995. Phylogenetic relationships in the tribes Podalyrieae, Liparieae and Crotalarieae. In: Crisp, M., Doyle, J.J. (Eds.), Advances in Legume Systematics 7: Phylogeny. Royal Botanic Gardens, Kew, pp. 283–308. Yatsenko-Khmelevsky, A.A., 1954. Caucasian woods, 1. Academy of Sciences of the Armenian Soviet Socialist Republic, Yerevan (in Russian). Zimmermann, M.H., 1983. Xylem Structure and the Ascent of Sap. Springer, Berlin.