Journal of Botany

ISSN: 0028-825X (Print) 1175-8643 (Online) Journal homepage: http://www.tandfonline.com/loi/tnzb20

Biosystematics of higher in New Zealand 1965–1984

H. E. Connor

To cite this article: H. E. Connor (1985) Biosystematics of higher plants in New Zealand 1965–1984, New Zealand Journal of Botany, 23:4, 613-643, DOI: 10.1080/0028825X.1985.10434233 To link to this article: http://dx.doi.org/10.1080/0028825X.1985.10434233

Published online: 04 May 2012.

Submit your article to this journal

Article views: 141

View related articles

Citing articles: 12 View citing articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=tnzb20

Download by: [125.239.168.206] Date: 22 August 2017, At: 08:23 Nell' Zealand journal of' BOlan)'. 1985. 1'01. 23 : 613-644 613 0028-825X/85/2304-0613$2.50/0 © Crown copyright 1985

Biosystematics of higher plants In New Zealand 1965-1984

CONTENTS H. E. CONNORt Page Centre for Resource Management INTRODUCTION ...... 614 University of Canterbury C·ytology ...... 614 Christchurch, New Zealand Experimental hybrids ...... 614 HIGHER PLANTS ...... 614 A. GYMNOSPERMAE ...... 614 P()[)OC-\RPACEAE ...... 614 a Natural Hybridism ...... 614 Abstract During 1965-1984 progress in the h Ecotyp) ...... 615 study of biosystematics of the New Zealand flora B. DICOTYLEDONES ...... 615 has advanced an understanding of the monocoty­ I. COMPOSIT·\E ...... 615 ledonous clement more than the dicotyledonous. Hrachl'f;/'>III.I ...... 617 ( .c/ IIIlsia ...... 617 In Lleula. Fi.l'IIlU.\', and CO/'Iadc/'ia experimentally (','Ili/U ...... 61 ~ produced hybrids between New Zealand species and SCI1('ClO ...... 619 overseas taxa have measured gene-exchange, 2. FAGACEAE - .\'oliIo/Uglls ...... b 19 allowed generic realignment or confirmed taxon­ .1 CiOODENIACEAE -'ie/liera...... 619 omic judgements. ChiolloCh/OQ, Gramineae. lies to 4. HALORACiACEAE - l/u/oragll ...... 619 5. MYRTACEAE...... 620 the forefront in experimental work at specific and l.£'f>IO.IPl'l'IlI II 111 ...... 62() infraspecific levels. ecologically. physiologically, 11.'1111::1'0 ...... 620 biochemically and morphologically. and will con­ 6. ONAGRACEAE - F"i/o/>llIm ...... 620 tinue to yield information needed to comprehend 7. RllBIACEAE - (·opro.l//IlJ ...... 622 its evolution in New Zealand. Among dicotyledons 8. SCROPHliLARIACFAL...... 622 }'u/'iIhclw ...... 622 Epilohi II III. Onagraceae, is the best documented }/c/>c...... 622 genus because of the extent of the study of experi­ 9. 11MBELLIFERAE ...... 622 mental hybrids. natural hybridism. gene-exchange. a Natural Hybridism ...... 622 and cytology. b Cultivation Experiments ...... (2) Experimental studies on economically important 10. WINTERACEAE - 1'.I·l'lIlilm·illlcra ...... 623 plants with a long lifespan e.g., NOlho(aglls arc few. II. HYBRIDISA nON WITH ALIEN FLORA ...... 6~3 and even in the study of hybrid asteliads. at the C. \10NOCOTYLEDONES ...... 6~3 end of 12-13 years there had been no flowering to I. ASPARACiALES ...... 6~3 .·jsll'lili and (·oI/O\[il'l'lIlIllII ...... 623 allow a first assessment of pollen or seed fertilitv. ('ordI'IiIlC ...... 624 Experimental studies needed for the recogniti~n }'hOflllllllll ...... 6~ 5 of ecotypes are also few but data and responses to ~. LlLlALES - 1.1/>l'I'lIu ...... 625

Downloaded by [125.239.168.206] at 08:23 22 August 2017 ultramafic soils and geothermally determined soils 3 . .IUNCALES ...... 6~5 are available. I.II::II/a ...... 6~5 Wild hybrids, their frequency. their fertility or a Hybrids between N.Z. taxa ...... 6~5 b Cienetic relationships between N.Z. and sterility, have been studied in a sporadic way. l:'pi­ overseas taxa ...... 627 lo/Jiufll excepted, and have not yielded the data ./IIIICIIS ...... 627 needed for writing the flora of New Zealand. The 4. (·YPERALES ...... 627 Cockayne-Allan pedestal of 50 years ago awaits its ('arc\, ...... 627 column and capital. L cl/CIIIlU ...... 6~7 5. GRAMINALES ...... 628 a Danthonieae ...... 628 Keywords Biosystematics: New Zealand: natu­ (·hi(lllnch/nu ...... 6~8 ral hybrids: experimental hybrids; fertility: gene­ (i) Hybridism ...... 628 exchange; infra-specific differentiation; ecotypy (ii) lnfraspccific differentiation ...... 6~8 (iii) Cultivation and transplant cxperimcnts ...... 630 ~·ogft~~;·~~··=·j,:i;:;;;;;:;::::::::::::::::::::::::: ::::::::::::::::::::::::::~~~ c Poeac - Fl's/IICO ...... 634 tMiss E. L. Hellaby Indigenous Grasslands Research CONCLUSION ...... 635 Fellow ACKNOWLEDGMENTS ...... 637 RCi'cil'cd 2 .-IIIgllSI 19115 REFERENCES ...... 637 614 New Zealand Journal of Botany, 1985, Vol. 23

INTRODUCTION synthesise the hybrid from the supposed parents'. The experimental hybrid Sophora tetraptera (n = This descriptive account of progress in biosyste­ 9) X S. prostrata (n = 9) has been made (Godley matics of the New Zealand flora is one part of a 1979) in part ' .... to clinch the argument .. .' that S. symposium to honour Dr E. J. Godley who began microphylla arose in New Zealand, and derived work in "Biosystematics' or the 'New Systemat­ some of its present variation, especially its juvenile ics", at Cambridge University in 1945 on Agro­ forms, from S. prost rata and from the older species pyron (Godley 1951). This review continues that of S. tetraptera. The presence of juvenile forms is of my late colleague John Bruce Hair for the period considerable significance. Extensive studies on S. 1945-1964 (Hair 1966), a period which encom­ microphylla (Godley & Smith 1977, Godley 1979) passed the post-war boom in botany to which he showed that the possession of a juvenile form is was a substantial contributor. the rule in South Island, except for S. microphylla Included here are those studies which incorpo­ var. /ongicarinata from northern Nelson. Plants rate the classical biosystematic catalogue of eco­ from the Chatham Is, some North Auckland popu­ typy, transplant experiments, natural hybridism, lations, and some central North Island populations measurement of gene-exchange in experimental and lack the divaricating juvenile phase found else­ natural hybrids, fertility-sterility assessments, and where in North Island. infraspecific differentiation. Godley (1979) indicated that S. prost rata could Of the main sustaining elements in biosyste­ not be, as proposed by Cockayne (1912), a fixed or matics - cytology, genetics, experimental hybrids, permanently juvenile (neotenous) form of S. micro­ morphology, and ecology - substantial contribu­ phylla, and suggested the reverse, namely, that the tions in New Zealand have been made in cytology juvenile form of S. microphylla is derived from S. and experimental hybrids. prost rata. The F, S. tetraptera X S. prostrata grown by Cytology Godley (op. cit.), and probably subsequent gener­ In his review ofbiosystematics of the New Zealand ations, is critical to his hypothesis; flavonoid pro­ flora Hair (1966) indicated that the chromosome files should help distinguish stages in the proposed numbers of just less than 40% of the 1800 indige­ evolution of S. microphylla because the suggested nous vascular plants were known. Chromosome parents have distinctive complements (Markham numbers are now published in the series 'Contri­ & Godley 1972). The outcome will be important butions to a chromosome atlas of the New Zealand to other parts of the Pacific basin. flora' for all the taxa that Hair, Beuzenberg, and The many experimental hybrids made during this Groves have examined, and these approximate to twenty year period are central to this review. Any half New Zealand's indigenous flowering plants tendency to depend on barriers to crossability for (Beuzenberg 1970, 1975, 1983; Beuzenberg & species discrimination seems to have diminished Groves 1974; Beuzenberg & Hair 1983, 1984; in this same period. Groves 1977· Groves & Hair 1971; Hair 1967a, b, 1968, 1970, i 977, 1980a, b, 1983; Hair & Beuzen­ berg 1966, 1968; Hair, Beuzenberg & Pearson 1967; HIGHER PLANTS

Downloaded by [125.239.168.206] at 08:23 22 August 2017 Hair & Kalin Arroyo 1984; Hair, Webb & Beu­ Individual taxa studied are grouped and presented zen berg 1980; Post 1983). under the three major headings of Gymnospermae, Because Hair commented at length on the level Dicotyledones and Monocotyledones. The level of of polyploidy, both primary and secondary, it is association of taxa into other taxonomic groups is appropriate only to reinforce the generalisations of a matter of convenience. Beuzenberg (1982) that (i) in many polytypic gen­ era there is a uniform level of polyploidy, (ii) in A. GYMNOSPERMAE some genera several levels of ploidy occur, (iii) high The greatest activity in the study of New Zealand secondary basic numbers are found in many gen­ gymnosperms in the past two decades has been in era, and (iv) the highest chromosome numbers yet taxonomic rearrangements. Of the 20 species in recorded in the angiosperms are known from New Allan (1961), distributed among five genera, these Zealand plants viz. Poa litorosa 2n = c. 266 (Hair same species are now allocated to 11 genera (see 1968), and Olearia albida (Compositae) 2n = c. Edgar & Connor 1983 for summary). 432 (Beuzenberg & Hair 1984). Podocarpaceae Experimental hybrids a Natural hybridism Godley (1949) in a discussion of hybridism wrote: Two natural hybrids have been discussed in the 'The final experiment to clinch the argument is to last decade. In Podocarpus Wardle's report (1972) Connor-Biosystematics: higher plants 615

poses the question of the time involved between possibility and although pollen fertility falls ofr initial hybridisation and the stabilisation of intro­ quickly with time. gymnosperm pollen is reputed gressants; in Lepidothamnlls Quinn & Rattenbury to have a lag period of up to 16 h and a relatively (1972) reported a hybrid that may be the conse­ slow rate of pollen tube growth (Hoekstra 1983). quence of longish distance pollen or seed dispersal. The least satisfactory solution may involve the Neither problem is simple in resolution. elimination of L intermedius from the district some In south Westland on alluvial vallev floors totara time after hybrid formation. is a uniform. stable entity distinct from Poc/ocarpus b Ecotypy totara var. totara in northern and eastern South [n dioecious Ph.l'//oc/adus a/pinus. Wardle (1969) Island. Wardle (1972) interpreted these south refers to two ecotypes - one "timberline" the other Westland plants. which he named P. lolara var. "lowland". Plants on well-drained. rocky slopes lI'aihocnsis. as the result of introgression between above timberline cast of the Main Divide in South P. aculijiJ/ius (2n = 34) and P. 100ara var. tolara Island form broad. layering thickets with phyllo­ (2n = 34). clades 4-20 mm long; and these contrast with plants A very recent origin is postulated for this taxon bearing phylloclades 10-50 mm long which may as Wardle suggests that it was selected and stabi­ become erect trees up to 9 m tall and grow on lised from among hybrids of P. acut(jIJ!ius X P. poorly drained. leached soils at low altitudes in totara since the return of warmer conditions after western South Island. The morphological features the Otiran. persisted in plants in cultivation. The expansion of P. totara var. lVai}wcllsis Freezing resistance of phylloclades in P. a/pinus. following the return to warmer conditions after the -IO°C for lowland plants. and .- 18°e and -22°C Otira glacial maxi1TIum would appropriately for those from higher altitudes. coincided with the account for the current distribution. But for initial ecologically and morphologically differentiated hybridisation and many generations of recurrent lowland and subalpine ecotypes (Sakai et al. 1978. backcrossing to the long-lived. slow-maturing P. 1981 ). to/ara. and for the elimination of P. totara var. {otara itself the time involved is unlikely to be less than 250 000 years which would place the original hybridisation in the Pleistocene as seems quite B. DlCOTYLEDONES logical. This. so far as I can judge. is one among few A wide assortment of studies is included here rang­ recent reports in New Zealand of a hybrid giving ing from what Turrill (1940) called the " ... simple rise to a stable and distinct taxon; Sophora micro­ cultivation under approximately uniform condi­ phyl/a has a similar evolutionary background tions ... " experiment. to the artificial production of (Godley 1979). In Epi/obium Raven & Raven (1976) interspecific hybrids and reports on natural outlined pathways of Pleistocene interspecific hybridism. hybridisation in species formation. Amphiploidy is The arrangement in this section is alphabetical not involved in these genera. by families. Time of formation may also be important in Downloaded by [125.239.168.206] at 08:23 22 August 2017 hybrids in Lcpidoth aIII III us. In that genus. segre­ 1. COMPOSIT AE gated from Dacrydium by Quinn (1982). six iso­ Two genera in the Compositae have been exam­ lated. uniform populations of infertile F, L. ined for the extent and frequency of natural hybrid­ /axi{o/ius X L. intermedius were found in central ism. Scnecio-Brachyg/ottis (Senecioneae) and North Island by Quinn and Rattenbury (1972). In Ce/misia (Astereae). Cockayne & Allan (1934) and F, IS" is frequent, but 14" + 2, or 12" + 4, were Allan (1961) listed c. 30 natural hybrids in Cc/­ also found; the hybrid is heterozygous for two par­ misia and c. 6 in the shrubby Senecioneae. acentric inversions. Cytologically there is no alter­ These two genera have contrasting taxonomic native to the parentage suggested because these two histories. Ce/misia has a constant generic treat­ species arc the only indigenous species where ment. only one or two taxa having been placed in 2n = 30. Hybrid seed was easily produced Aster by Forster f. The history of generic transfers experimentally. between Senecio and Brachyg/ottis and the debate, No plants of L. intermedius currently occur in is given in Edgar & Connor (1983) and repeated the district. but may have been present in the past, only in its essentials. or F, seeds on crimson, succulent receptacles could An emergent theme here is that despite uniform­ have been transported by birds from some other ity of chromosome numbers in the members of both part of the country. This latter is the least demand­ genera. gene-exchange is often limited or ing solution. Long distance pollen transport is a constrained. 616 New Zealand Journal of Botany, 1985, Vol. 23

revo/uta

buchananii haastii

cassinioides

bellidioides

monroi

rotundifolia G) compacta ~G G greyi" G

bidwillii repanda Downloaded by [125.239.168.206] at 08:23 22 August 2017

perdicioides J south/andica G~L...... -----I ------...rhectori ~

_____ High frequency natural hybrids

• - - - • Low frequency natural hybrids High frequency spontaneous garden hybrids _ G - Fig. I Crossing diagram in G:.:::. Low frequency spontaneous garden hybrids Brachyglottis; data from Drury -=-= (1973); nomenclature after N Of­ • includes B. /axifolia denstam (1978); n = 30 in all taxa (Beuzenberg 1975). Connor-Biosystematics: higher plants 617

Brachyglollis of Ceillll.IW subgenus Pelliculatae section Petiola­ The recent taxonomy of the tribe Senecioneae in tae an up-to-date list of (i) recognisable, (ii) very New Zealand was set in train by Drury (1973) and probable. and (iii) putative natural hybrids between completed by Nordenstam (1978). The major result species of that group and taxa from other subge­ is the placing of all shrubby species of .I.,'eflecio in nera. Unlike Drury he used the same techniques Brachl'glollis except S. kirkii which constitutes the for hybrid identification which Cockayne & Allan monotypic genus U'roslclI/on; the herbaceous S. (op. cit.) had used except that there is one chro­ Ival/ii and S. scor::oncroidcs arc segregated into mosome number and one estimate of pollen fer­ Do/ichog/ollis. Scnccio s.s. is reserved for species tility reported. Given's list contained two significant treated by Allan (1961) as Erecll/iles and some her­ features (i) there is little substantive change from baceous species. the list of natural interspecific hybrids in Allan Drury (1973) had discussed the shrubby Sene­ (1961). (ii) free gene-exchange is probable in three cioneae and their hybrids. describing the latter in hybrid combinations among 25. Chromosome anatomic detail as well as in morphological fea­ number differentiation alone does not account ti)f tures. Beuzenberg (1975) reported 2n = 60 for the extent of limited gene-exchange. the same members of the complex and for four hybrids. gametic number (n -- 54) occurring in the species Half of the hybrids Drury accepts arose spon­ studied except for some populations of C. graci­ taneously in cultivation. Specimens of wild hybrids ICllIa and C. gralllilllfil/ia, which cannot at a lirst arc usually few (Drury op. cit.). but there arc excep­ glance be implicated as particular parents in the tions. e.g .. B. cassinioides X IJ. rc\'Olwa. and B. hid­ hybrids. Il"il/ii x B. c/aeagn(f{)lia. Most natural and Species of ('('/lIIisia subgenus Pelliculatae section spontaneous hybrids are low frequency hybrids (Fig. Petiolatae form high and low frequency interspe­ I). cific hybrids within the subgenus (Fig. 2): high fre­ Four species arc central in the crossing diagram quency hybrids are C. s/Il'clahi/is X C. fral'cl'sii, and in Fig. I: C 11'I.1I'enii x C. I'crha.lci/illia; C. corcialifil/ia X C (i) IJ. cassillioidl's occurs frequently in natural hybrid spcclahili.1 was formed perhaps once only; ('. lIIack­ combinations; alii and C. hookeri form no natural hybrids with (ii) B. monroi and 13. grcyi link the group composed any species of sect. Pctiolatae 1(.)[ reasons of geo­ primarily of spontaneous hybrids from cultivation graphic distribution. with two groups of natural hybrids. and The hybrid ('. lrarer.li, X ('. l'erh(JsCl/i}lia is the (iii) n. rcpancia-grcl'i links a second group. mostly of natural hyhrids. into the lIlonroi-greyi complex. only intrasectional one where. because of the var­ There is no infrageneric treatment of the group. iability in populations. gene-exchange is indicated; hut as rosette herbs hybridise with trees or shrubs. the high frequency hybrids C. sJlcClllhilis X C. lral'­ boundaries separating taxa on crossability grounds crsii arc interpreted as F, plants. Figure 2 illustrates may not be easily detected. Jeffrey (1979) indicated the crossing behaviour for high frequency hybrids within the section. son;e suhgroups in Bracllrglotlis (which he treats conservatively as Scnccio); intergroup hybrids This array of interspecific hybrids within section occur. e.g .. 11. heclori x 11. e/aeagn(/()/ia and 11. !lcr­ Petiolatae contrasts with the high frequency of

Downloaded by [125.239.168.206] at 08:23 22 August 2017 dicioides x B. claeagni/il/ia. It may be more impor­ hybrids and gene-exchange between it and other tant to note that Brachyg/ollis was never listed as subgenera (Fig. 2). forming hybrids with l.'roslcmon or f)o/ichoglollis. Nine species form low frequency hybrids with C. though in this latter natural hybrids arc said to spcc/ahih.l. C lral'cr.lli or C. rcrhasc(/iJha. and five occur (Cockayne & Allan 1934). species form high frequency hybrids. Among the There may be some gene-exchange between spe­ common hybrids there is evidence of gene-exchange cies of Brachyglollis, but pollen fertility is often low in C. spcclahilis X C. Iral/ii (C X pseudo/ral/ii), in although in the B. compacla-monroi-greyi hybrids C. l'cl'hascifi)lia x C. Iml/ii. and in C. frarel'sii X pollen fertility exceeds 50%. There arc no data on (' I'cl'hascjfi)/ia. . seed setting in hybrids but viable seeds are pro­ Chromosome numbers for species of ('c/lllisia duced in some back crosses. were reported in Hair (1980a). Parent of both fer­ tile and sterile hybrids share the common comple­ Ceimisia ment 2n = 108; hybrids between 2n = 108 species Natural hybrids in Celmisia listed in Allan (1961) and those where 2n = 216 arc at a low frequency. are based on Cockayne & Allan (1934) but with the Lee and Given (1984) demonstrated experi­ rider ..... some with good to fair field evidence, occ. mentally that C. spedcmi is physiologically adapted supported by herbarium specimens, others more to growth on ultramafic soils in Southland; it is the speculative." Given (1984) appended to a revision sale ultramafic endemic in southern South Island. 618 New Zealand Journal of Botany, 1985, Vol. 23

r-____ Subgenus Celmlskl ___--. ~ ___ Subgenus Peiliculalae ___--,

sect Pulvlnatae sect Celmlsla sect. Pelilculatae

series Llneanfoilatae senes Pelilculatae

Subgenus Pelilculatae

~ ______sect Peholatae ______---'

High frequency natural hybrids

Downloaded by [125.239.168.206] at 08:23 22 August 2017 gene·exchange probable

• M • M. Low frequency natural hybnds

Fill. 2 Crossing diagram in Celmisia subgenus Pelliculatae sect. Petiolatae and of high frequency hybrids between it and other subgenera. Nine low frequency intersectional hybrids are omitted. Chromosome numbers in Hair (. 980a); n = 54 in most taxa.

Cotula islands 40 km distant and (iii) yet another particu­ In cultivation experiments to determine the growth lar to Pyramid Rock, south of Pitt L Lloyd inter­ habits of indigenous species of CotuJa, Lloyd (1981) preted the former as the one which was common grew plants from Chatham Is. and the mainland. to the much larger land mass of the Chathams at He showed (Lloyd 1982) that Chatham Is. C. leath­ the time of the last glacial maximum (c. 20000 erstonii consists of three distinctive forms: (i) one years BP); the others were isolated from this com­ common to much of the Chathams and its lesser mon stock and from each other by a deeper ocean islands, (ii) another, once known as C. renwickii, before that glacial maximum. No taxonomic rank particular to the Forty Fours, a group of small is accorded the three forms. Connor-Biosystematics: higher plants 619

Although primarily concerned with the breeding low altitudes: system in dioecious species of Callila, Lloyd (1975) (iii) "Undifferentiated /I:. solandri" from sites more reported that 563 plants of 163 hybrid combina­ mesic than those typical of mountain beech and tions (some of them intraspecific) were raised. The black beech and from low altitude: so far known hybrids were easily made - even intersectional only in South Island. ones - and F, 's were fertile, but the particular The possibility that "Undifferentiated N. so/an­ details were not presented. Natural hybrids occur ciri" are ancient and stable N. solallcirt x /\'. Ii/SUI in relatively few combinations, C. delldri x C. hybrids is raised by Wilcox & Ledgard (op: cit.) atrata and C. albida x C. pcc/inata being ihe most who suggest that allozyme studies may test this common (Lloyd 1972). hypothesis. The work of Hillis & Orman (1962) COlilla pyrethrd"olia var. Iincari/iJlia was shown suggests that flavonoid analyses may be more easily by Lee et at. (1983) in experiments with ultramafic applied. but the results of those analyses while spe­ and contrasting soils to be an edaphic ecotype of cies specific may not contain as many data as could ultramafites confirming Lloyd's earlier suggestion come from the allozyme studies. The F, hybrid (1972). made artificially by Poole (1951). N. sofandri var. SCl1ccio solandri (n = 13) x N. lim'a (n = 13), may still Sykes (1971) described the behaviour of the Ker­ be available and would be an important datum. madec Is. endemic SCI1ccio /aUllis var. cspcr('flsis Differences in N. sofandri var. clil/imioidcs for in a cultivation experiment. On the southernmost freeLing resistance of leaves relative It) altitude were point of the Kermadec Is, L'Esperance, S. laUllls reported by Sakai & Wardle (1978) and Sakai et al. grows on sites rich in guano. Sykes grew S. fallflls (1981) and although treated as ecotypic by Sakai & var. ('spcrcnsis in a glasshouse for two generations Wardle (1978) the results arc probablv better inter­ and showed it was a short-lived annual which in preted as ecoelinal because of the ~Ievation gra­ cultivation did not retain the height, leaf size, or dient. But in experiments with plants grown in succulence found in the wild. Leaf shape was fairly ultramafic soils and in contrasting soils, and with constant and also .... the degree of marginal inden­ plants from ultramafic sites and different habitats, tation". He believes that var. csperensis behaves as Lee et al. (1983) found no evidence of selection for a permanent juvenile because the leaves never ultramafism in I\'. so/alldri var. cl(//iJrt/()idcs. become deeply dissected as do mature plants of var. faufus. 3. GOODENIACEAE - SeI!icra Se/licra mdicans is a small herb with creeping stems 2. FAGACEAE - Notholagus and branches found on the coast in sandhills, estu­ The genus NOlholagus has not been subject to the aries and salt marshes, and inland on the margins genetic or cultivation experiments that are war­ of streams, lakes and tarns in Hushed areas. In New ranted by its ecological and physiognomic impor­ Zealand 2n = 16 (Hair & Beuzenberg 1960): in tance in New Zealand forests. The work of Wilcox Tasmania 2n = 16 and 48 (Jackson 1958). & Ledgard (1983) corrects in part this deficiency Ogden (1974) conducted a randomised 'rans­ experiment with plants from an estuarine salt and reveals some population differentiation. Seed marsh population and plants from a sand plain. of the species and varieties of indigenous beeches Downloaded by [125.239.168.206] at 08:23 22 August 2017 from 50 localities were sown, and seedlings raised The two were genetically distinct especially in leaf size and shape, and in Hower width and in pedicel to 2 1/2 years old at Rangiora and Rotorua. In N. menziesii (silver beech) leaf size and shape length. Plants from the dunes were rhizomatous with appressed leaves but those from the estuary differed, but there was no association between these bore erect lea ves on stolons. characters and their localities. Red beech (N fifSca) was described as comparatively uniform, prove­ Ogden (lac. cit.) treats the two forms as ecotypes: nances not being distinctive in the nursery although there may be others among inland plants. There is some seed lots showed evidence of hybridism with no suggestion in Australia that more than one spe­ N solandri. cies is involved despite differences in chromosome In N so/andri three entities were recognised: number. (i) var. clWarlioidcs: mountain beech with an erect habit, green triangular, inrolled, distichous leaves 4. HALORAGACEAE - /-Ia/oragis in populations from high altitude sites on moun­ In /la/magis erecla (2n = 14) Forde (1964) made tain sides, or at lower altitudes on poorly drained over 100 intraspecific crosses, 91 of which flowered sites; in their first year. Pollen fertility was reduced to c. (ii) var: solandri: black beech with an interlacing 50% in two families: no estimates of seed-set were branchmg habit, bronze oval to orbicular flat dis­ possible because the fruit is indehiscent. Forde tant leaves in populations from ridges o~ sp~rs at concluded that there was little evidence of sterility 620 New Zealand Journal of Botany, 1985, Vol. 23

between regional forms but did not estimate the was expressed in terms of the number of layers of extent to which gene-exchange could occur. She aerenchyma, - coast > inland. A difference in could not sustain H. co/ensoi as a separate species. cell number, length and width in the innermost The experimental F, hybrid H. erect a X H. car­ aerenchyma layer was not statistically significant. ti/aginca had reduced pollen fertility (52%); F, seed Manuka seedlings from 17 localities lying did not germinate; both qualities are indicative of between Kaitaia and Stewart Island were measured some barriers to gene-exchange. Orchard (1975) when 16 months old for seven morphological and treated H. erecta as two subspecies H. erecta ssp. habit characters (Yin et al. 1984). The populations erecta and /I. erecta ssp. cartilaginea which appears could be segregated into four sets: somewhat contrary to Forde's biosystematic (i) a high altitude Stewart Island population from evidence. 570 m, (ii) northern North Island populations ranging from sea-level to 330 m and occurring south to latitude 5. MYRT ACEAE 38°S, Lcpt osperm u m (iii) populations from relatively moist sites at low­ Two major experiments with L. scoparium to mid-altitudes, (manuka) have been reported recently, Lee et al. (iv) less precisely, populations from relatively dry (1983) on response to ultramafic soils and Yin et sites of central North Island and in North al. (1984) on the seedling behaviour of New Canterbury. Zealand-wide provenances. Yin et al. (1984) claim evidence for ecotypic dif­ Leplospermum scoparium from ultramafic sites ferentiation associated with environmental varia­ was grown in experiments on two ultramafic soils bles, stating that altitude and distance from the - Red Hill and Black Ridge, Southland - and in coast are probably the most important in gener­ a potting mixture; these were contrasted with a ating the responses. population from a hill soil (Lee et al. 1983). No Manuka is versatile in its responses, and it is evidence for ecotypic differentiation was detected; improbable that all the variants have been detected and only the Black Ridge soil depressed growth in the widespread, habitat-varied species. significantly. Kun::ea Because on Dun Mountain (Nelson) ultramafic Kanuka, until recently always treated as Leptos­ soils manuka demonstrated a capacity for accu­ pennum ericoides , has been transferred to Kunzea mulating Ni. or Cr and Ni. by contrast with plants as K ericoides by Thompson (1983). from an andesetic site on Mt Egmont, it was In an area of geothermally determined soils at regarded as an ecotype by Lyon et al. (1968. 1969, Karapiti, near Wairakei, Given (1980) described 1970. 1971). As Lee et al. (1983) pointed out, and heated ground covered, or almost so, with pros­ Yin et al. (1984) also noted, there was no experi­ trate K ericoides that did not exceed c. 30 em in mental dem )nstration of ecotypy in L. scoparium height - " ... a habit which is retained in cultiva­ by Lyon et ,,1. (op. CiL), but all results - Lyon's tion." In seedlings of K ericoides grown in soil from and Lee's - are concordant and show that toler­ their original site and in a potting mix, the branch­ ance to ultramafites is within the genetic compe­ ing angles indicated that the prostrate habit " ... will tence of that species.

Downloaded by [125.239.168.206] at 08:23 22 August 2017 be retained." The varietal name micro/lora attaches The evidence that L. scoparium is a Cr-accu­ to these plants. Known variability w'ould indicate mulator (Lyon et al. 1968, 1969, 1970) on ultra­ that it is quite unlikely that there are only two eco­ mafic sites and without morphological types in K ericoides, one the typical variety and differentiation (Lee et al. op. ciL), is in contrast with the other var. micro/lora. genetically determined prostrate forms from gley podzol soils in south western Nelson (see Wardle et al. 1973 p. 624). There is a further contrast between that Nelson genoecodeme and the response 6.0NAGRACEAE to shade and infertile soils in Fiordland which Two genera of the family Onagraceae occur in New depresses manuka into a rhizomatous shrub whose Zealand - Epilobium and Fuchsia. The latter has habit is not genetically determined (Wardle et al. attracted little recent attention since Godley's study 1973). of sex forms (Godley 1955, 1963). That the experi­ Another physiological differentiation was mental F, Fuchsia excorticata (n = II) X F. pro­ revealed in a study of the relationship between cumbens (n = II) was made once again by Diane manuka on waterlogged and freely drained soils H. Percy, may easily escape detection, included, as (Cook et al. 1979). There is some evidence of eco­ it is, among a list of ornamental species of Fuchsia typic differentiation between populations from being studied for ftavonoids (Williams et al. 1983); Central (dry) and coastal (waterlogged) Otago which no genetic data are yet available. Connor-Biosystematics: higher plants 621

Epilobium v High frequency and low frequency hybrids The monumental programme of hybridisation occur; ecological instability is associated with the experiments carried out by the late Walter Boa former more often than not; habitat stability is often Brockie with 37 native species of Epilobium associated with the latter. (Brockie 1966, 1970, and W. B. Brockie in Appen­ vi Sympatry is relatively high. Not all sympa­ dix to Raven & Raven 1976) allowed Raven & tric taxa form natural hybrids, and ecological spe­ Raven (1976) to say" ... the New Zealand species cialisation precludes, to a considerable extent, of Epilobium are probably better known in terms opportunities for sympatry. of their crossing relationships than any other group of plants of comparable size anywhere in the C Experimental hybrids world". vii Experimental hybrids are easily produced, Hair (1977) reported 2n = 36 in all species. New and Brockie successfully made c. 900 of them. As Zealand species have the BB chromosomal summarised by Raven & Raven (op. cit.) fertile F, arrangement characteristic of Eurasian species, but which produced F2's occurred in 75% of the hybrid some plants of E. microphyllum differing in one combinations; 25% of the combinations produced reciprocal translocation are B,B,. abnormal weak plants which rarely flowered; 2 A summary of the critical data presented by hybrids were sterile but could be backcrossed. Raven & Raven (1976) on natural and artificial Hair et al. (1977) reported 2n = 36 mostly as 18 11 hybrids can be expressed in 10 points: in the hybrids, although 16 11 and a ring or chain of 4 was detected. A Reproductive Biology Vlll (a) 27 species of the 37 in New Zealand are i Self-pollination and self-fertilisation (auto­ fully intercompatible. gamy) prevail in 33 protogynous species, with c1eis­ (b) Some species form fertile artificial togamy obtaining in eight of them; this proportion hybrids with almost all others e.g., E. rostratum. of autogamy is higher than anywhere else in the (c) Four peregrinating species - E. brun­ world. Three species are aJlogamous in an appre­ nescens. E. pernitens. E. nerterioides. E. num­ ciable degree together with many populations of a mularifolium - are intercompatible but do not fourth species. Autogamy is the most important form hybrids with many other species. constraint to natural hybridism; it is one of the most (d) E. pallidiflorum. E. billardiereanum ssp. important attributes in New Zealand in that it cinereum. E. pictum. E. alsinoides ssp. atripli­ ensures the perpetuation of adapted genotypes. Four ci/olium form hybrids with several species but species strictly autogamous in New Zealand have the frequency of abnormal weak plants is very strongly outcrossing populations in Australia - E. high. pallidiflorum. E. billardiereanum, E. gunnianum ix (a) There are no barriers among species at and E. hirtigerum. the level of decreased hybrid fertility; even ii Gene-exchange within natural populations is hybrids which are difficult to obtain are fertile. small; exchange between populations is likely to be (b) Failure to set seed is not an active barrier even less. The exceptions are E. chionanthum. E. in interspecific hybridisation, nor is failure of wilson ii, some large flowered populations of E. gla­ seed germination. (c) Abnormal hybrids apart, and even though Downloaded by [125.239.168.206] at 08:23 22 August 2017 bellum and E. angustum which are organised for out-crossing. reported in 25% of hybrid combinations these may not occur in all circumstances, the New B Natural hybridism Zealand species form a complex of completely iii Natural hybridism in New Zealand occurs interfertile taxa in contrast with other parts of in at least 38 interspecific combinations, and there the world. may be another 10 that could be very difficult, if x Hybrids between Northern Hemisphere taxa not impossible, to detect in the wild because of and New Zealand species are often, but not invar­ similarity to one parent. This number is vastly in iably, of reduced pollen fertility. Seeds may arise excess of the estimates of natural hybridism in Epi­ from backcrossing to either or both parents. lobium made by Cockayne & Allan (1934). Perhaps Autogamy, poorly developed internal barriers to E. alsinoides and its three subspecies are more hybridisation, and rather frequent natural hybrid­ commonly involved than any other individual ism characterise Epilobium sect. Epilobium in all taxon. parts of the world. New Zealand is no different, iv Natural hybridism occurs here between four except that adaptive radiation during the Pleisto­ (and possibly five) indigenous species and the natu­ cene allowed the development of a large number ralised E. ciliatum of North America, and one of species - all 2n = 36 - compared with other ~ybrid with E. obscurum, naturalised from Europe, countries. 1S recorded. The contemporary taxa here are competent to 622 New Zealand Journal of Botany, 1985, Vol. 23

exchange genes at a relatively high level, and it is martinii in Marlborough. The geographic distri­ unlikely that they were not so in the past. The con­ bution can be adequately explained by changes in tribution made by this historic gene-exchange is the land mass of New Zealand since the Otiran. assessed by Raven & Raven (1976) as important Grayer-Barkmeijer (1978) identified two flavon­ in allowing recombinants to form, to become pro­ oid chemodemes which corresponded to ssp. calar­ tected by prevalent autogamy and habitat prefer­ ractae and ssp. martin ii, and a third for ssp. diffusa ences, and to have played an important role in the and ssp. lanceolata jointly. Ravonoid analysis of formation of the taxa that now occupy New Fl hybrids P. hookeriana X P. decora and P. calar­ Zealand, a country which has " ... the greatest con­ ractae X P. hookeriana showed most of the char­ centration of species of Epilobium found in any acteristic compounds of the parents, but in neither region of comparable size in the world". was the flavonoid profile completely additive (Grayer-Barkmeijer 1978). Experimentally made P. catarractae (n = 21) X 7. RUBIACEAE - Coprosma P. hookeriana (n = 21) and P. catarractae X P. Coprosma was among the first genera discussed by /yallii (n = 21) were moderately fertile (Garnock­ Cockayne (1923) when natural hybridism became Jones & Langer 1980); no data were appended. recognised in New Zealand. Allan (1924, 1926a, Hybrids P. catarractae X P. /ya/lii have been I 929b, 1931, 1938) described in morphological detected in Fiordland; the other combination has detail the artificial interspecific hybrid C. propin­ been collected on the Volcanic Plateau. Garnock­ qua~ X C. robustao, and its filial progenies, Jones made intraspecific P. catarractae hybrids the reporting as well sex-form segregation in Fl (100, parentages of which are not disclosed; they are 9~ ) and F2 (180, 26~ ). All generations up to F) reported to be " ... nearly as fertile as their parents", were fertile (Allan 1940). but there are no supporting data. Ravonoid analyses by Wilson (1979) showed that Hebe plants predicted by A. P. Druce as C. robusta X C. In Hebe, Garnock-Jones & Molloy (1982) showed repens, and C. repens X C. crassifolia possessed that dense leaf and stem pubescence in H. allan;i additively the flavonoid spectra of both parents. is dominant over the variable glabrousness seen in Later work by Wilson (1984) reports the flavonoid H. amplexicaulis. Garnock-Jones & Molloy treat profiles of the species already recognised, and for the two as one species - H. amplexicaulis. the undescribed entities that A. P. Druce accepts. The way is clear in Coprosma to use these profiles in the analysis of interspecific gene-exchange. 9. UMBELLIFERAE In addition to reporting the chromosome num­ a Natural hybridism bers of the recognisable taxa in Coprosma, Beu­ Although natural hybridism in the Umbelliferae is zenberg (1983) listed the chromosome numbers of more frequent in New Zealand than elsewhere in seven natural and artificial interspecific hybrids. All the world (Webb & Druce 1984), New Zealand's are regular with 2n = 44 except C. depressa X C. umbellifers have received most attention in the last ciliata where 2n = 6x = 132; 2n = 2x = 44 is two decades in taxonomic revisions and in studies attributed to C. depressa, and 2n = 4x = 88 for of reproductive biology (see, for example, Dawson Downloaded by [125.239.168.206] at 08:23 22 August 2017 C. ciliata . The hybrid may be an amphiploid I 967a, b, 1974; Dawson & Webb 1978; Lloyd 1973; involving doubling of the Fl complement; its con­ Webb 1979, 1981). Chromosome numbers are in tinued resemblance to an Fl interspecific hybrid Hair (1980b); in Aciphylla, Anisotome and Gingidia would thus be assured. Plants of C. depressa X C. 2n = 22. ciliata were not examined by Wilson (1979). The natural intergeneric hybrid Aciphylla squar­ rosa X Gingidia montana from Marlborough is described by Webb & Druce (1984) with details of' 8. SCROPHULARIACEAE about 25 individual characters; most are interme­ Parahebe diate between the parents. The hybrid was seed Garnock-Jones & Langer (1980) used multivariate sterile and produced 3% stainable pollen. No other analytical methods and cultivation of 35 popula­ putative hybrid involves these two genera. But Aci­ tions under uniform glasshouse conditions to study phyl/a X Anisotome is proposed several times (see the infraspecific taxonomy of Parahebe catarrac­ table 3 in Webb & Druce 1984) even though some­ tae. Their analyses revealed discontinuities between times qualified by doubt. There is, however, no four allopatric groups which were ultimately treated doubt about the frequency of the interspecific as (i) P. catarractae ssp. diffusa in central New hybrids Gingidia decipiens X G. montana, Aniso­ Zealand, (ii) ssp. lanceolata in Taranaki and Coro­ tome haastii X A. pilifera, or A. aromatica X A. mandel, (iii) ssp. catarraclae in Fiordland, (iv) ssp. haas!ii; all are frequent (Webb & Druce op. cit.). Connor-Biosystematics: higher plants 623

In dioecious Aciphylla, a genus of about 40 spe­ Ogden (1978) in a study of Calystegia (Con vol­ cies, Webb & Druce (op. cit.) lists 27 interspecific vulaceae) in New Zealand endeavoured inter alia hybrids, one third of them involving A. colensoi. to discriminate native from introduced species and For once, the list of Cockayne & Allan (J 934) is to detect natural hybrids. No experiments, cultural exceeded. Judged solely from the Webb-Druce or genetic, were carried out. remarks, gene-exchange between A. poppelwellii and His results are interpreted as revealing (i) indige­ A. scott-thomsonii is not restricted. The same may nous C. sepium X naturalised C. sepium, with the obtain in other hybrids. New Zealand element dominant in the gene pool; b Cultivation experiments (ii) suggestive evidence of introgression between C. The recognition that Aciphylla townsonii Chee­ sepium (indigenous) and C. silvatica (naturalised); sem. is a juvenile form of A. hookeri Kirk was (iii) the possibility of hybrids C. tuguriorum obtained from a simple cultivation experiment (Le (indigenous) X C. sepium. Stace (1961) had syn­ Comte & Webb 1981). thesised the fertile hybrid C. sepium X C. silvatica The two taxa were known to coexist in Nelson in Great Britain. and Westland. Plants of both, and of a putative hybrid, were transplanted at Lauriston on the Can­ terbury Plains. During the next year the townsonii C. MONOCOTYLEDONES plants and the suspect hybrids " ... began to resem­ ble or exactly match plants of A. hookeri ... " The Genera among five orders are discussed. As with change involved a switch from simple I-pinnate the Dicotyledones there are varied experiments to grasslike leaves to 2-pinnate leaves with short broad report, though experimental hybridisation has been more frequently practised than elsewhere among segments. Plants raised from seed of A. hookeri begin life the native plants. Despite this clear preponderance, with townsonii leaves and during the next four years the range of studies is on much the same scale. begin, and continue, to produce hookeri leaves. Two genera deserve special mention as they were Reversion leaves in A. hookeri are of townsonii involved in experiments with overseas taxa, viz. Luzula and Elymus. Chionochloa, another grass form. genus, lias an ever-increasing literature. 10. WINTERACEAE - Pseudowintera Sampson (1980) described in detail F, hybrids 1. Pseudowintera colorata (n = 23) X P. traversii (n = 23); pollen fertility of F, and the parents is Among the indigenous spathaceous and petalous > 90%; some fruit was set in F,. asparagoid Monocotyledonae three genera of tall It is generally asserted that P. colorata and P. tufted herbs, (Phormiaceae), Astelia and axillaris hybridise freely (Allan 1961, Cockayne & its segregate genus Col/ospermum (Asphodeli­ Allan 1934, Vink 1970) but Sampson (op. cit.) is aceae), and one, mostly trees, Cordyline (Aspho­ unconvinced. Fruits were set after experimental deliaceae) have been studied experimentally, pollination of P. colorata by P. axillaris. Although particularly as interspecific hybrids. In Cordyline P. c%rata is self-incompatible (Godley & Smith there are no barriers to interspecific hybridisation.

Downloaded by [125.239.168.206] at 08:23 22 August 2017 1981), Sampson is cautious enough to indicate that Phormium seems equally freely open to hybridisa­ plants must be raised before he will be sure that tion. Very large differences in chromosome num­ he has in fact made a verifiable hybrid. ber between Astelia and Col/ospermum do not deter hybridisation. 11. HYBRIDISATION WITH ALIEN FLORA Astelia and Collospermum Two genera in which hybridism was studied involve Astelia is a dioecious genus of 13 species where the species and genera now naturalised in New Zealand. flowers are fragrant and nectar is present. It is not Cockayne & Allan (1934) had cited seven cases of regarded as a genus in which hybridism is wide­ hybridism between exotic and indigenous species. spread, and Moore (1966) was unable to confirm Of the world Aizoaceae only Disphyma australe the putative hybrids listed by Cockayne & Allan is native to coastal New Zealand; on many parts (1934). The chromosome numbers in Astelia and of the coast Carpobrotus edulis and C. aequilaterus Collospermum are: are naturalised. Chinnock (1972) found 27 chro­ A. banks;; n = 35; A. solandr; n = 35; A. fragrans mosome, sterile, intergeneric hybrids F, D. australe 2n = 70; A. trinervia n = 105; A. nervosa n = 105; (n = 18) X C. edulis (n = 9) on both coasts in A. grandis 2n = 140; C. hastatum n = 35; C. North Island, and on the eastern coast of South microspermum n = 35, 2n = 70 (Wheeler 1966; Island. The hybrid (2n = 27) D. australe X C. see also Moore 1980). aequilaterus is known from Patea only. Wheeler (1966) judged that hybridism took place 624 New Zealand Journal of Botany, 1985, Vol. 23

between A. solandri and A. banksii on Rangitoto Is., Auckland Harbour. Pollen fertility of the hybrids was nil; some seed was present in fruits on female putative F, plants and may have arisen from backcrossing. Kaspar x australis Moore (1980) reported on experimental hybrids among As/elia plants, some hundreds of which were in cultivation at Botany Division, Lincoln. There was one successful hybrid, A. fragrans~ X A. ner­ vosa S d , among II attempts. In the F, family there was an evident segregation, 6 plants resem­ bling A. fragrans and 3 resembling A. nervosa. Moore is very positive that the A. fragrans female plant is pure A. fragrans, but does not comment on the lineage of A. nervosa. The family was lost natural hybrids by misadventure when it was 5 years old and plants artificial hybrids had not flowered. Co/lospermum, also dioecious, is a generic seg­ regate from Astelia. Moore (1980) reported on Fig. 3 Crossing polygon in Cordyline; n = 19 (Hair & experimental intergeneric hybrids thus: Beuzenberg 1968). i a C. microspermum~ X A. nervosa S d : non­ flowering; at 12 years one plant resembled C. microspermum. in hybrids where chromosome imbalance would b C. microspermum ~ X A. nervosa S d : the occur. Wheeler's (1966) logical conclusion that progeny died, morphology unrecorded, by about hybridism between species with n = 35 and n = the time it was 8 years old. 105 was unlikely, is shown not to be so by the c C. microspermum ~ X A. nervosa S d : at 8 experimental hybrids Moore (1980) described. years plants resembled A. nervosa; none flow­ These data Moore said " ... should help to give cre­ ered nor lived to 12 years. dence to field reports of hybrid asteliads", but ii C. has/alum ~ X A. nervosa B d : non-flowering should also be noted for studies on the evolution and smaller than parents. of dioecism. iii C. has/alum ~ X A. banksii d' : the hybrids Field hybrids include A. solandri X A. banksii though showing characters of both parents super­ (Wheeler 1966); A. nivicola X A. petriei (Wilson ficially resembled the largest plants of A. nivicola 1976); and possible A. nervosa X A. graminea of Fiordland! On one plant at 3'/2 years a female (Moore \980). The hybrid C. hastalum X A. solan­ inflorescence appeared, but at 13 years no others dri which interested Skottsberg (1934), though not had flowered in the then depleted family. Because synthesised by Moore, may not exceed the bounds of flowering time differences Moore thinks this of possibility. hybrid is unlikely to occur in nature even though Cordyline the two species coexist in the Auckland area. Of the cabbage trees, Cordyline, Moore (in Moore Downloaded by [125.239.168.206] at 08:23 22 August 2017 In describing interspecific and intergeneric & Edgar 1970) recognised five endemic species; all hybrids Moore commented very particularly on age flowers are fragrant and nectar is plentiful. to flowering; she had already described asteliad Natural hybrids are reported (Fig. 3). The only inflorescences and flowering (Moore 1970). Collos­ hybrids unrecorded in nature are C. australis X C. permum may flower precociously producing indivisa and C. indivisa X C. pumilio (Moore 1975); "reduced" inflorescences; adult plants bear well­ C. kaspar is the sole species on the Three Kings, developed spathes each subtending a simple raceme. and is locally endemic there. In all species n = 19 Species of Astelia of similar size and habit to Col­ or 2n = 38 (Hair & Beuzenberg 1968). lospermum do not flower precociously, but pro­ Cordyline kaspar and C. pumilio are self-incom­ duce inflorescences related to their stature and patible (Beever 1981, 1983). Experimental crosses habitat. In intergeneric hybrids the Astelia char­ by Beever showed that seed was formed in C. kas­ acter dominates the precocity known in Collosper­ par X C. australis; a putative F, C. kaspar X C. mum except perhaps for one plant in C. hastalum australis was successfully backcrossed to both spe­ X A. banksii. cies. The same presumed F, was interfertile with The ASlelia-Co/lospermum account contains C. pumilia. Abundant seed was set from all several unresolved problems some of which were pollinations. insoluble even after growing hybrids for as long as Gene-exchange in Cordyline, as distinct from 16 years. Matrocliny and patrocliny are indicated hybrid formation, was shown by Moore (1975) Connor-Biosystematics: higher plants 625

when she grew families from seeds of two hybrid Chromosome numbers from Hair et al. (1967) are: (or putative hybrid) plants. Plants were raised at L. pulchella 2n = 2x = 38; L. peregrinans 2n = Botany Division, Lincoln from seeds from a soli­ 6x = 114; L. grandiflora 2n = 6x = 114; L. ixioides tary open-pollinated C. banksii X C. pumilia in the 2n = 12x = 228. There is some conflict over chro­ Christchurch Botanic Gardens, and from C. aus­ mosome counts, or identification, between Hair et tralis X C. pumilia? seeds from one plant originally al. (1967) and Kenton & Heywood (1984) who from North Cape. Moore assumed that the plant reported 2n = c. 230 in L. grandiflora from Nel­

in the Botanic Gardens was either an F J hybrid or son, and 2n = c. 72 in L. ixiaides from Chile. a segregate from C. banksii X C. pumilia because grandiflora is self-fertile; plants attrib­ it possessed characters specific to each, and because uted to this hexaploid species are described by in the families raised from seed at least one bank­ Moore as "... heterogeneous in appearance ... " sii-type plant occurred. The family from the plant Flowering times are not synchronous for plants of North Cape origin was not pure C. pumilia but from different districts indicating some physio­ contained some C. australis influence. It did not logical differentiation. No infraspecific rank was flower. accorded to any variant because she found it was Beever (1981, 1983) who demonstrated self­ impossible to aggregate the variation. The other incompatibility in Cordyline hybrids doubted hexaploid L. peregrinans, by contrast, is very uni­ Moore's (1975) basic assumption that the proge­ form apart from colour and size differences asso­ nies she grew from single plants could take their ciated with habitat. In dodecaploid L. ixiaides, genesis in selfing. He favoured the view that the plants are strictly tufted and liable to environmen­ hybrid seed parents would be pollinated by some tal modification of size. other plants, even by other hybrids. Moore's third In diploid L. pulchella only could Moore (op. generation hybrids (her "second progeny") could cit.) be reasonably sure of fairly clear-cut regional have arisen from F2 interpollination, or from pol­ differentiation, but then only for minor characters lination by other species or hybrids in the Lincoln and again not awarded infraspecific rank. Three collection. phenodemes seem to be indicated: a Westland Whatever their origin, pollen fertility was esti­ phenodeme with the range Fiordland to Buller, a mated at 90% in all plants studied by Moore and northwest Nelson - Buller phenodeme, and a third by Beever except for one plant (C) in C. australis for North Island plants which extend also south X C. pumilio? Seeds were set in Moore's plants into Nelson. Perhaps the diploid alone is old enough just as in Beever's backcrosses. Interspecific bar­ for the evolution of such a diversification to have riers in Cordyline are genetically insignificant and begun; the higher levels of ploidy may represent the self-incompatibility system which operates may progressively less differentiation. effectively increase the extent of gene-exchange by allowing compatible pollination between hybrids and other taxa. 3. JUNCALES Phormium Two genera of small herbs of world-wide distri­ The two species of Phormium (Phormiaceae) in bution Luzula and Juncus have been involved in New Zealand were experimentally crossed by Allan interspecific hybridisation studies. Luzuia, through

Downloaded by [125.239.168.206] at 08:23 22 August 2017 & Zotov (1937). Allan (1940) referred to a large FJ the efforts of Nordenskiold (1966, 1969, 1971, progeny raised by 1. Swindlehurst at Moutoa; that 1977), backed by the taxonomic work of Edgar hybrid was P. tenax

uniform conditions at Botany Division, Lincoln. in F J but reduced fertiIty and lowered vitality 626 New Zealand Journal of Botany, 1985, Vol. 23

~o ~'lf ~ '\,e.. .~~ (~.~ ~~ ~e..~ ~<:' ·V\.~ " V '0 ~~ New Zealand ruta 6 3.0 - -

banksii 6 2.8 3.0 -

pieta 6 2.7c 3.0c 2.0c

pumila 6 1.3 - -

leptophylla 6 d d d

Australia meridionalis 6 1.8 O.Sc 2.0c

flaeeida' 6 2.3c 1.7 2.0c

australasica2 6 3.0 3.0 2.5

novae-eambriae3 6 2.3c 2.3 2.0

modesta 4 12 0.8 1.0 -

aeutifolia 6 - 1.3 0

Europe pallescens 6 1.0 0.5 2.0

eampestris 6 0 0 0

U.S.A. maerantha S 6 0.4 0 0

Japan capitata 6 1.0 0 O.B

scale

Downloaded by [125.239.168.206] at 08:23 22 August 2017 , reported as L.meridonalis var. f/aceida 0= 0-25 percent fruit set 2 reported as L. oldfieldii var. oldfieldii 1 = 26-50 percent fruit set 3 reported as L. oldfieldii var. angustifolia 2 = 51-75 percent fruit set • reported as L. australasica 3 = 76-100 percent fruit set 5 L.campestris var. maerantha c_ = cushion-form in progeny d = died before flowering

Fig. 4 F J fertility of interspecific hybrids with New Zealand species of Luzula. All data from Nor­ denskiold (1966, 1971); F J L. pumila x L. capitata, and F J L. pumila x L. campestris var. macrantha were sterile. Connor-Biosystematics: higher plants 627

occurred among F2 L. ruja var. ruja and F2 L. pieta sphere plants are, on the other hand, almost exclu­ var. piela (Nordenskiold 1966). sively self-pollinating (Nordenskiold 1966), and In intervarietal hybrids FI was usually fertile, but produce invariable families. in F2 var. ruja X var. albieomans there were chlo­ Natural hybridism, as reported in Edgar (1966) rophyll-deficient plants, and in var. piela X var. for New Zealand, is supported by the ease of pro­ limosa there was reduced fertilty in FI (Edgar 1966, duction of fertile hybrids. Their recorded fre­ Nordenskiold 1966). quency in nature is not high except for L. banksiana At the interspecific level (Fig. 4) plants of L. lep­ var. banksiana X L. piela var. pieta, and L. bank­ tophylla X L. ruja, X L. banksiana, and X L. pieta siana var. migrata X L. eolensoi; none involves died; FI fertility was high except in L. ruja X L. species naturalised here. pumila; chlorophyll-deficient plants occurred else­ Juneus where in all F2 families (Nordenskiold 1966). In New Zealand Juneus is a genus with 16 native species and 31 introduced taxa (Healy & Edgar Genetic relationships between New Zealand and 1980). Edgar (l964a) attempted crosses among overseas taxa native species and achieved about half of the pos­ There are 11 species of Luzula in Australia (Nor­ sible combinations. Fertile FI progeny was raised denskiold 1969, Edgar 1975, and see the nomen­ in J. australis X J. distegus, and J. distegus X J. clatural corrections by Jansen 1978); none is in gregiflorus which probably also occurs in nature. common with New Zealand. Natural hybrids occur FI J. australis X J. pallidus was infertile - no seed there (Edgar 1975). Chromosome numbers are in was set on selfing and no pollen shed from the Fig. 4. anthers. Interrelationships in Luzula as a whole, includ­ Natural hybridisation in Juneus as assessed from ing Northern and Southern hemispheres, are evidence from field studies, knowledge of the described in the addition,al crossing experiments of reproductive systems, experimental hybrids, and a Nordenskiold (1971). New Zealand species crossed more accurate identification of specimens in her­ with Australian taxa yield relatively fertile FI; baria, is unlikely to be a high frequency event. hybrids between New Zealand taxa and Northern Cockayne & Allan (1934) had listed hybrids among Hemisphere taxa are less fertile except for L. piela most of the then recognised species. XL. palleseens (Europe) and L. ruja X Japanese L. eapitala. . Nordenskiold (1971) concluded that m the L. 4. CYPERALES eampestris - L. multiflora complex the relation­ Two significant cyperaceous genera in New Zealand ships between regions, as measured by hybrid fer­ are Carex, 73 indigenous species, and Uncinia with tility, are roughly proportional to their geographic 32 indigenous protogynous or protandrous species. separation. Although FI interspecific hybrids between meso­ Carex phytic New Zealand taxa are the most fertile among In Carex where Cockayne & Allan (1934) saw about all the hybrids Nordenskiold had produced, their 12 hybrid combinations, Edgar (in Moore & Edgar F/s displayed the highest frequency of abnormal­ 1970) failed to confirm rife hybridism. Edgar ities she had encountered. The presence of cushion (1964b) in a detailed morphometric analysis showed Downloaded by [125.239.168.206] at 08:23 22 August 2017 plants among F2 progeny of L. ruja var. albieomans that FI C. diandra X C. seeta occurred at several X L. piela var. piela, and in L. banksiana var. localities in New Zealand. Sterile pollen is present banksiana X L. piela var. pa/lida as well as in in indehiscent anthers, and seed-bearing utrides are hybrids involving graminiform Australian species not found. C. diandra occurs in the Northern and L. pieta, L. ruja, and L. banksiana, is the more Hemisphere as well, and Kiikenthal (1909) men­ remarkable because the cushion habit is not a fea­ tioned some natural hybrids involving it. ture of these taxa even though in New Zealand the Uneinia cushion-habit occurs in both the raeemosa and Uncinia was subject to controlled cross-pollination campestris - multiflora groups. by Edgar (in Moore & Edgar 1970). Seeds were set As well, a high level of Chlorophyll-deficient in just over half of those attempted - of which seedlings occurred in New Zealand hybrids quite most were intra-series crosses. Inter-series crossing unlike Australian experimental hybrids, although produced endosperm less seeds except series Lep­ somewhat alike to North American hybrids where tostachyae X series Ripariae. No details have since about 25% of F2 families contained albino seed­ been published about the FI's. lings. In New Zealand, species though self-fertile, Cytological examination of 31 species showed 2n are not exclusively self-fertilising, and Norden­ = 88 in all except the two species of series Macro­ skit)ld detected a lack of morphological uniformity lepidae C. sinclairii (2n = 94, 2n = 96), and C. in progenies grown from seed. Northern Hemi- elegans (2n = 94) (Beuzenberg 1970). 628 New Zealand Journal of Botany, 1985, Vol. 23

5. GRAMINALES leaf wax; C. rigida is not known to synthesise this Among indigenous grasses the genus Chionochloa compound. On the Paparoa Range with rigida-type ~as been most extensively studied, as befits the plants are (i) C. jlavescens which in that area does Importance of its ecological role in New Zealand not synthesise TMEs, and (ii) C. rubra which syn­ alpine and subalpine grassland. Such studies had thesises cycloartenol ME (Connor & Purdie 1976b. been urged in 1954 (Sewell et al. 1954). Lower 1981). We mterpreted the rigida-type plants as downslope, species of Festuca are very abundant hybrids between C. rubra and C. /lavescens. and oft~n dominate the grassland landscape; two Chionochloa flavescens x C. ruhra is the hybrid hexaplOId species were involved in a biosystematic most frequently found: it is at least on the Tararua evaluation. Though never as common as Festuca Range (Greenwood 1953, Druce 1957, Williams et the New Zealand members of the tribe Triticea~ al. 1978c), at Jacks Pass, Marlborough, (Connor are important components of several plant com­ 1967), in Central Otago (Williams et al. 1978a), and munities. Here experimental hybridisation and the on the Paparoa Range (Connor & Purdie 1981). measurement of chromosome pairing in those Williams (1975) reported an increase in C. f/a­ hybrids were used to determine the genomic con­ vescens x C. pal/ens hybrid frequency with increas­ stitution of the New Zealand taxa and to indicate ing altitude on the southern Tararuas; in South their generic placement. Island at various sites Williams et al. (I 978b) invoke hybridism between C. pal/ens and C. rubra, and C. a Danthonieae pal/ens and C. macra, and suggest " ... obvious Chionochloa introgression between several taxa at ... " Temple Chionochloa is a danthonioid genus segregated from Basin where C. crassiuscula, C. ./lavescens, c. pal­ Danthonia by Zotov (1963); all 23 species in the lens and C. rubra co-occur. There are no support­ genus are found in the New Zealand region except ing data. On Mt Stokes C. pal/ens x C. flavescens C. Jng/da and C. pal/ida of Australia. Chionochloa is not infrequent (Park 1968). is one of the most important grass genera in New Purdie and I found little or no evidence for con­ Zealand; it is the physiognomic dominant over temporary interspecific gene-exchange as measured millions of hectares of South Island uplands. by TME presence, although we sampled at many sites where several species were sympatric. At the (i) Hybridism level of gene-exchange and introgression, the TME Experimentally produced and spontaneously profiles of some chemodemes of C. pal/ens, and of occurring hybrids between hexaploid (2n = 42) C. flavescens, are suggestive of an ancient hybrid­ species of Chionochloa were reported earlier (Con­ isation and persistent backcrossing in one direction nor 1967); no hybrid was sterile; seed-set was rela­ (Connor & Purdie 1976b, 1981). In particular, C. tively high in most crosses but was lower in two pal/ens from the vicinity of its locus classicus clearly hybrids involving C. cunninghamii, and usually indicates the possibility of a parkeol ME - arun­ higher in hybrids involving C. rigida. Pollen stain­ doin penetration of the cycloartenol ME chemo­ ability was invariably high. deme at some time in the past. A possible donor No barriers to gene-exchange were detected is C. rubra from its north. From the same locality, except that C. flavicans, the rupestral species found Mt Holdsworth, C. flavescens was undoubtedly fr.om Coromandel. and East Cape to Hawkes Bay,

Downloaded by [125.239.168.206] at 08:23 22 August 2017 influenced by a species which synthesised miliacin dId not form hybnds with other species when used and arundoin; no lupeol ME, characteristic of C. as a female parent although it was successful as a flavescens on the Rimutaka Range and in Nelson­ pollen parent. Those hybrids reputed to occur Marlborough, was isolated from Mt Holdsworth. naturally were made artificially, or occurred spon­ The miliacin-arundoin contributor to C. flavescens taneously in the experimental gardens. is most likely to be C. rubra, but a chemotype of No attempts have been made to produce hybrids that composition has, at present, a north Canter­ between 2n = 6x = 36 species and those where bury-Nelson distribution. The evident contradic­ 2n = 6x = 42. tion is that two distinct chemotypes of C. rubra, Triterpene methyl ether (TME) composition of one northern and the other southern, are invoked the leaf-surface wax provided a methodology to help for two events at one place. The two events need resolve problems of natural hybridism in Chion­ not have occurred at the same time, but there is ochloa. Triterpene methyl ethers found in Chion­ no method of demonstrating such a possibility, no ochloa, including some novel ones, are listed in more than suggesting that C. rubra alone is the Russell et al. (1976), and Pauptit et al. (1984). donor species. On the Paparoa Range, Westland, there are plants that resemble C. rigida but these lie outside the (ii) Infraspecific differentiation expected geographic distribution of this species. Some species have been studied in detail at the Cycloartenol ME was the sole TME present in their infraspecific level, in particular the three wide- Connor-Biosystematics: higher plants 629

spread species C. flavescens, c. pal/ens and C. rubra, In North Island Williams et al. (1977) recognised and, to a lesser extent because of its more restricted that C. pallens north of the type locality in the Tar­ distribution, C. rigida. arua Range produced more robust shoots and leaves Chionochloa flavescens because it is so clearly less V-shaped in section. In South Island Williams site-selective in eastern South Island, and because et al. (1976) identified a phenotype which extends those plants are so ecologically, morphologically, from Fiordland to a line connecting the Godley and physiologically different from those on the Valley in the east and the Cook River in the west Tararua Range, (type locality, Zotov 1963), has (Fig. 5); these plants are robust with broad flatter incited discussion, experiment and debate led by leaves, and their blades disarticulate at the ligule. P. A. Williams. In C. pallens north of that line blades are persistent Williams (1975) and Williams et al. (1976, 1977, (Williams et al. 1977). These characteristics 1978b) argued very persuasively that the ecological remained in plants grown in pots. Genes control­ similarity between C. flavescens Tararua Range and ling leaf blade abscission have no ecological or geo­ C. flavescens Murchison Mountains, Fiordland, graphic coincidence with any other genes recognised should be matched by a taxonomic similarity, and to date. that eastern South Island phenodemes attributed to Triterpene methyl ether analyses revealed infras­ C. flavescens should be distinguished in some tax­ pecific chemodemes in C pallens (Connor & Pur­ onomic way. Connor & Purdie (I 976b, 1981) die 1976b, 1981); chemodemes synthesising offered some taxonomic realignment by treating cycloartenol ME are separated by a miliacin "Fiordlandflavescens" as the ~-amyrin methyl ether chemodeme of central and southern South Islands chemodeme of C. rigida which extends on leached (Fig. 5). Interspersed among this pattern is a spo­ soils in western South Island from Lake Kaniere radic distribution of arundoin, either on its own or to at least Doubtful Sound (Fig. 7). Veritable C. in combination with miliacin, but never with flavescens extends in south Westland to the Clarke cyc!oartenol ME except at Field Peak in the Tar­ River at least. For eastern South Island C. flaves­ arua Range where parkeol ME intrudes as well. cens they did not proffer any solution; those plants In C. crassiuscula of higher South Island alti­ do not synthesise triterpene methyl ethers and are tudes, there are two chemodemes: a northern mil­ biochemically distinct from Tararua Range plants iacin chemodeme reaching to Fox Glacier, and an which synthesise arundoin and miliacin. . arundoin chemodeme south of there. Bell (1973), working in the Owen Range ill Nel­ Red tussock, C. rubra, does not display triter­ son, found what he referred to as C. sp. aff. flaves­ pene methyl ether diversification to the same extent cens was calcicolous, dominating on marble and on as other equally widely distributed taxa (Fig. 6). calcium-rich soils at lower altitudes. Connor & The only distinctive series of populations are (i) a Purdie (l976b) found that it synthesises lupeol ME cycloartenol ME chemodeme from mid-altitudes as do all Nelson Province plants of C. flavescens. (c. 1000 m) in Westland on terraces with modestly There is a good element of agreement for South young soils, (ii) a parkeol ME-arundoin chemo­ Island with what Burrows (1967) recognised in C. de me of the Volcanic Plateau in North Island and, flavescens: thus the western Nelson form is the less definitely, (iii) a miliacin-arundoin chemo­ lupeol ME chemodeme; plants from the Haast Pass deme from northern South Island. and throughout Fiordland are the ~-amyrin ME Chionochloa rigida is distinguished in western Downloaded by [125.239.168.206] at 08:23 22 August 2017 chemodeme of C. rigida; the central South Island South Island by the presence of a ~-amyrin ME form with leaves c. 14 mm wide is the phenodeme chemodeme; for the rest, non-synthesis ofTMEs is of young sites; and the Stewart Island popUlation characteristic (Fig. 7). In a species of even more is differentiated by the synthesis of a mixture of restricted distribution - C. cheesemanii - two miliacin and arundoin (Connor & Purdie 1981). No chemodemes are found. North Island populations TME analyses were made of C. flavescens from Mt synthesise two additional TMEs (which have not Stokes, Marlborough, which A. P. Druce in Park been identified to date) and these discriminate it (1968) ascribed to the Tararua Range phenodeme; from the South Island chemodeme (Fig. 7); lupeol Purdie and I are therefore unable to comment. ME and arundoin are common to both. Chionochloa pallens is a common enough moun­ Lipid composition of leaf wax (Cowlishaw et al. tain species in South Island but is restricted in 1983) showed infraspecific differentiation of much North Island to the Tararua and Ruahine Ranges. the same order as with TMEs. Burrows (1967) distinguished North Island popu­ These results emphasise that the geographic dis­ lations from South Island ones, but did not describe tribution of a number and a variety of genes, or the differences. In South Island he recognised four gene complexes, in species of Chionochloa has kinds. A. P. Druce in Park (1968) considered the become well documented, but the relationships forms of C. pal/ens on the Tararua Range and Mt between those genes and ecological factors remain Stokes, Marlborough, to be the same. unanswered. 630 New Zealand Journal of Botany, 1985, Vol. 23

~~+-arundoin·parkeol ME­ cycloartenol ME

~---- cycloartenol ME

v--...... ----·miiacin

~+_____1-___r_~~----cycloartenol ME

o 100 200 km I I I

170 0 180 0

Downloaded by [125.239.168.206] at 08:23 22 August 2017 Fig. 5 Triterpene methyl ether distribution in Chionochloa pal/ens.

(iii) Cultivation and transplant experiments Mark (1965) conducted reciprocal transplant I can find no evidence that I recorded of C. macra, experiments mostly with C. rigida in Central Otago. that when grown under uniform conditions with Although Mark (op. cit.) identified two ecotypes other species of Chionochloa, it retained its char­ in C. rigida, one of them from high altitudes was acteristic slim stature, short leaves, early flowering, later recognised by Molloy & Connor (1970) as c. and non-chaffing sheath. This was, however, macra Zotov and thus its status as an ecotype is important in urging the formal recognition of the to be disregarded. taxon because it could be shown that plants in cul­ Growth and development was recorded in Otago ture for 5-10 years could be distinguished from a populations of C. rigida from (i) Maungatua 870 range of species. While I never acknowledged these m, (ii) Old Man Range (OM) 910 m (iii) OM 1220 data, neither did Zotov (I970) when he described m cross-transplanted, including 1590 m on OM and C. macra, but it was a lesser concern to him; Mol­ in the University garden, Dunedin, (Mark, 1965). loy & Connor (1970) are also silent. Mark (1965) Of total variance in tussock growth, about 40% was showed that in cultivation at Dunedin it retained attributed to the genetic variation contributed by its characters just as it did at two other sites. the populations. Connor-Biosystematics: higher plants 631

miliacin---~-:r-l

cycloartenol ME --~

~ _____ arundoin or TME absent

100 200 km I !

180 0

Downloaded by [125.239.168.206] at 08:23 22 August 2017 Fig. 6 Triterpene methyl ether distribution in Chionochloa rubra.

Mark (1965) showed that the transplants tended The results from these experiments were inter~ to become the same size as resident tussocks, but preted by Greer as revealing a high level of inher­ that the relative order of height was preserved, ent infraspecific plasticity for leaf growth and indicating genetic determination. Inflorescence flowering. This plasticity he attributed to post­ height was influenced both by the environment and Pleistocene selection pressures during a period when the genetic background. there was a very variable climate, especially during Greer (1979) re-reciprocally transplanted clones the last 5000 years. That inherent flexibility has over the same sites to determine the effect oflong been reinforced by, or even overtaken by, recently term (14 years) preconditioning on subsequent evolved local variants adapted to positions on the growth. He indicated that his results, like those of altitudinal range of the species in Otago. But no Mark, revealed a strong genetic component in evidence for differences in photosynthetic activity differences in growth pattern associated with alti­ was discovered by Greer (1984) which would sup­ tude of origin. He further recognised ecotypic dif­ port that particular hypothesis. ferentiation in a lowland population from Shag Recency in the evolution of localised biotypes is Point, a population which he introduced into the visualised by Mark (1965) and by Greer (1979) after trials but which Mark had not grown. C. rigida expanded its territory by occupying in 632 New Zealand Journal of Botany, 1985, Vol. 23

C. cheesemanii arundoin-Iupeol ME---+'-4 ~~L-.f--arundoin-Iupeol ME and 2 unknown

C.rigida ~---­ TME absent. or rarely arundoin

o 100 200 km I I I

170 0 1800

Downloaded by [125.239.168.206] at 08:23 22 August 2017 Fig. 7 Tritcrpene methyl ether distribution in Chionochloa rigida and C. cheesemanii.

quite recent times (c. 1000 years) land formerly because intraspecific gene-exchange remains under forest. (see Molloy et al. 1963; McGlone unexamined. 1983). The only inter-regional comparison of growth in What Mark and Greer described in C rigida C rigida or C. macra is that of Williams (1977) could adequately be identified as a combination of who compared leaf growth and leaf death in plants the alpine genotypes from upslope and the contri­ in the Ashburton Catchment, Canterbury, with bution made by smaller populations from subal­ those of Mark (1965) and Rowley (1970) in Otago. pine and montane sites. Relative to Otago the peak of leaf elongation in I had proposed earlier (Connor 1964, 1967) that both species occurred a month earlier in Canter­ gene-exchange between populations of C rigida, bury; leaves did not die-back as extensively in which had become free to exchange genes after the autumn; there was no marked increase in leaf lon­ removal of barriers separating them for long gevity with altitude. These comparisons might periods. would contribute to its plasticity. There indicate physiological differences between C. macra was no evidence for it then. and there is none now and C rigida in the two provinces but. because of Connor-Biosystematics: higher plants 633

difficulties in the interpetation of growth per se, In the experimental sterile F, hybrids C. arau­ Williams cautioned care in their interpretation. cana (Chile, n = 36) x C. toetoe all plants were of Nevertheless, these Canterbury results and those female sex-form (Connor 1983). This hybrid family, from Otago indicate incipient ecotypy, the popu­ and the sterile C. toetoe x C. selloana family, con­ lation from sea-level excepted. firm the taxonomic relationship accorded them only This discussion on ecotypy has concentrated recently by Zotov (1963). around eastern South Island C. rigida; there are data The other naturalised species of Cortaderia found which allow the conclusion that western South in New Zealand, C. jubata (2n = 12x = 108), is Island and Fiordland ~-amyrin ME C. rigida is eco­ autonomously aposporous (Philipson 1978); it will typically differentiated from that in the east, and not be involved in genetic intrusion into the other that in C. flavescens ecotypy is evident. But in many five species found here. widespread species, the nature of the relationship between ecological and morphological differentia­ b Triticeae tion is unexplored. Chionochloa remains a treasure Elymus trove. The New Zealand wheatgrasses were studied exten­ sively in the late 1940s and 1950s (Connor 19S4a, b, 1956, 1957b, 1962 a, b). In Agropyron scabrum Cortaderia (2n = 6x = 42) where there was (i) a series of Cortaderia, the first grass in which gynodioecism plants of different appearance; (ii) these differences was recognised (Connor & Penny 1960), has been were preserved in cultivation at five sites; (iii) each extensively studied (see Connor 1974). The four phenodeme occupied a geographical territory of its indigenous (2n = lOx = 90) species preserve their own; and (iv) many of these phenodemes were distinctive characteristics in cultivation at Botany interfertile with each other but in some intraspe­ Division, Lincoln. cific hybrids reduction in fertility occurred at F, or Interspecific hybrids are easily made and were F2• Hair (1956) identified apomixis in some phen­ used primarily for studies on the inheritance of male odemes which have since been recognised as Aust­ sterility (Connor 1965b), and for the genetics of the ralian migrants (L6ve & Connor 1982). synthesis and inheritance oftriterpene methyl ethers Five other indigenous species were involved: A. (Martin-Smith et al. 1967, 1971; Connor & Purdie enysii (2n = 4x = 28), A. kirk;; (2n = 6x = 42) 1976a). F, fertility is very high (Connor 1983); per­ and A. tenue (2n = 8x = 56). These were crossed centage seed-set in selfed hermaphrodites of F. C. experimentally; most hybrids were sterile (Connor richardii x C. fulvida was 71.0, 71.2, 82.2, 83.8 1956; 1962, a, b). Two species of Cockaynea. C. (Connor 1983) and in F, C. toetoe X C. richardii gracilis and C. laevis were also involved; in both was 74.0, 74.3, 87.6. 2n = 4x = 28. No intrinsic barriers to gene-exchange have been Classification of the Triticeae rarely satisfied tax­ detected among the four indigenous species, but the onomists at the generic level; currently there is opportunities for natural hybridisation are few much redistribution of species among several gen­ because of differences in flowering times and geo­ era primarily based on genome analysis (Dewey graphic distribution (Connor 1965a). I know of no 1982, 1983a, b, 1984; Barkworth et al. 1983; LOve natural hybrids, though C. splendens X C. fulvida 1982, 1984). The level of satisfaction may not be Downloaded by [125.239.168.206] at 08:23 22 August 2017 is a possible combination in North Island and increased. would leave viable descendants.! indicated (Con­ New Zealand taxa were incorporated into a series nor 1965a) that hybrids between the naturalised of experimental crosses which involved a sector of sexually reproducing Argentinian-Uruguayan C. the world spectrum of the Triticeae of known gen­ selloana (2n = 8x = 72) and decaploid indigenous omic constitution: Critesion marinum (2n = 2x = species were improbable because of (i) chromo­ 14) genome H; Elymus hystrix (2n = 4x = 28) some imbalance, (ii) time of flowering difference, genome HS; Elymus longearistatus (2n = 6x = 42) and (iii) geographic distribution differences that genome HSS; Elymus transhyrcanus (2n = 6x = would limit many combinations. The sterile (2n = 48) genome HSS - as E. stewartii; Elymus tra­ 9x = 81) F. C. toeloe X C. selloana was made chycaulis ssp. novae-angliae (2n = 4x = 28) gen­ experimentally by using an out of season flowering ome HS; Elytrigia repens (2n = 6x = 42) genome C. selloana as male parent. In the F. family there EJS; Pseudoroegneria spicata (2n = 2x = 14) gen­ are vigorous plants as well as some that are genet­ ome S. Of these there is a difference of interpre­ ically inharmonious. There will be no genes tation with Dewey (1984) over the genome for exchanged between the two; there may be no genes Elytrigia, Dewey favouring SX where X is of exchanged in a parallel hybrid Nothofagus menzie­ unknown or unspecified origin. Critesion marinum sii (New Zealand) x N. obliqua (Chile) which has is included in Hordeum sect. Stenostachys by von occurred spontaneously in Britain (Wigston 1979). Bothmer et al. (1983). 634 New Zealand Journal of Botany, 1985, Vol. 23

~, ~~ ~v<:> ~(Q ",,0 . ",,10 ,s.,t:, v 0~ ~v"" e f;:-O ,0 ,10 \'b (' 010 0~~

Critesion marinum 7 H 7rr +7r 7 -9rr + 10-14r

Pseudoroegneria spicta 7 S 511 + 111 3m+4n + 1h

Elymus enysii 14 3m+ 11rr+ 41 Up to 10n 12n + 117

E longearistatus 14 HS 14rr

E trachycaulis 14 HS 13Il + 21

Ehystrix 14 HS 13n + 21

E multiflorus 21 3m+ 11rr+ 41 21rr

E transhyrcanus 21 HSS 21rr

Elytrigia repens 21 EJS 411 Up to 7rr

Fig. 8 Chromosome pairing in interspecific hybrids between New Zealand species of Elymus and overseas taxa. Data from Love & Connor (1982); nomenclature, after Love & Connor. See also Dewey (1984) for comment on the extent of pairing.

The results of these crosses are reported in Love In the Elymus hybrids seeds were set in FI E. & Connor (1982). Other crosses repeated those per­ rectisetus x E. transhyrcanus among extra-terri­ formed earlier in New Zealand and showed that torial crosses, 2n = 42 in both species; all the others some early crosses failed through external rather failed, or nearly so, to set seed giving constant evi­ than internal influences. From analyses of the dence of the failure of gene-exchange. chromosome pairing in hybrids, and their relation­ The most significant feature of this work was the ship to taxa of known genomic constitution, the demonstration of the relationship in inter­ overall genomic constitution of the New Zealand nationalIy accepted genomic terms of the endemic Downloaded by [125.239.168.206] at 08:23 22 August 2017 species was determined, and the individual hap­ genus Cockaynea to endemic species of Agropyron lome in each genome identified (Fig. 8). New and to other members of the Triticeae. Zealand tetraploid taxa were allocated the genomic formula HS, the hexaploids HSS, and the octoploid c Poeae HHSS; HS is the basic Elymus genome. Dewey Festuca (1984) expressed doubt over the interpretation of Two tussock grassland species of much promi­ the extent of pairing in some hybrids, concluding nence are the hexaploids (2n = 42) Festuca novae­ that they are impossible to reconcile with the gen­ zelandiae and F. matthewsii. Usually they are alIo­ omic constitution of E. repens in particular. The patrie, the former at low altitudes and the latter genomic structure and composition of some New upslope, and although they do meet at a good many Zealand species are, he concluded, uncertain. sites in South Island I have no evidence of natural The conclusions expressed in taxonomic terms hybridisation between them (Connor 1968). The are that the New Zealand species currently placed two species preserve their characteristic growth in Agropyron and Cockaynea should be transferred patterns and morphology in cultivation. to the genus Elymus, but in two different sections. Experimental FI interspecific hybrids were fertile In addition, the Central Otago. strict, wide-angled­ but did not set as much seed as either parent, spikeleted. blue, long-anthered plants were finally although by FJ seed-set was equal to parental values. given specific rank as Elymus apricus. Pollen fertility never felI below 90% (Connor 1968). Connor-Biosystematics: higher plants 635

A high-altitude form of F. novae-zelandiae which from a "List of Supposed Hybrids" and the well­ occurs in north and south Canterbury was recog­ established record of the 15 hybrids mentioned in nised some time ago (Connor 1960). It is distinc­ Cockayne (1923), and from the substantial list tive in stature and in panicle characters both in the included in J.P. Lotsy's "Evolution considered in wild and in cultivation. Scott (1970) grew F. novae­ the light of Hybridization" (1925), itself published zelandiae collected from 1800 m and from 700 m within four years of the discovery of widespread in the Mackenzie basin, south Canterbury, under hybridism in Nothofagus (Cockayne 1921; see also controlled temperatures. The lower altitude collec­ Godley 1979, and Thomson 1979, 1980). Allan's tion recorded its highest relative growth rate at 18°C final reflections on natural hybridism and the 1934 and the higher altitude material at 12°C; a signifi­ list are embodied in his "Flora of New Zealand" cant difference in temperature response. The Vol. 1 (Allan 1961); they reveal a subdued attitude experimental results and the behaviour of the higher to hybridism relative to that of Cockayne. A dec­ altitude material in cultivation are a clear indica­ ade earlier he foreshadowed that attitude (Allan tion of ecotypic differentiation. Experimentally 1951 ). produced F, hybrids between the high altitude form The study of wild hybridism has not expanded and lowland plants were fertile, but there was an on the bases built by Cockayne and by Allan, and inexplicable lowering of seed-set in F2 (Connor apart from Raven & Raven (1976) on Epilobium, 1968). and Fisher (1965) on alpine epirote ranunculi (see the discussion in Hair 1966) natural hybrids have either been reduced to brief sentences in taxonomic CONCLUSION treatments, or been treated as introgressants as Orchard (197 5) interpreted plants he was unable to What I have discussed shows that New Zealand classify in Gonocarpus. Gardner (1978) brought no plant biosystematic studies have made important, evidence, compelling or otherwise, in favour of his if strictly conventional, advances in the last twenty conclusion that Alseuosmia banksii X A. macro­ years, but the investigations are heavily weighted phylla is a frequent hybrid which should be recog­ in favour of - grasses, sedges, nised as A. X querci/olia. asteliads, wood-rushes. These are not unrelated to the two monocotyledonous floras which have been Wardle (1968) advanced reasons for the recog­ published in the "Flora of New Zealand" series nition of Pseudopanax macintyrei as a tetraploid (2n 48) calcicolous species of the northern South (Moore & Edgar 1970; Healy & Edgar 1980). = Epilobium, which had the three-fold advantage Island; it had often been awarded hybrid status (Cockayne & Allan 1934, Allan 1961). Such eli­ of the late W. B. Brockie's abundant experimental hybrids, the international taxonomic confidence of sions from the list of hybrids ate rare. P. H. Raven, and the cytological resolve of the late Melville (1976) invoked novel interactions to J. B. Hair, should need attention to details of resolve the old question of hybridism in Parsonsia reproductive biology only. Luzula and Elymus, (see Cockayne 1912, Allan 1926c). No experiments where international co-operation was also involved, were conducted; no studies of pollen fertility made; should be let rest. We know much of the interre­ no approximation of actual gene-exchange lationships among species in those genera. Cort­ attempted; the solution Melville offered follows Downloaded by [125.239.168.206] at 08:23 22 August 2017 aderia, but for an adequate genetic solution on my from a study of leaf polymorphism and persistent own part to the question of the inheritance of male juvenility, and from his concept of the glossopterid sterility, will yield diminishing returns. leaf syndrome. Chionochloa (Gramineae) is the only genus of If ever the time was appropriate to eschew reports economic importance to have received significant of natural hybridism which are based on such evi­ attention. But Nothofagus is possibly even more dence "as apparently intermediate" or some such important economically than the snow-tussocks. euphemism, that time is now. With the common Until M. D. Wilcox and N. J. Ledgard of the New availability in the 1980s of such modem techni­ Zealand Forest Service started work on it a few ques as thin layer chromatography, gel electropho­ years ago, it had been left where A. L. Poole laid resis, SEM photography, computing programs, and it (Poole 1950a, b, 1952, 1958). Weinmannia, Dac­ such traditional disciplines as genetics, cytology and rycarpus, Podocarpus, Beilschmiedia, Kunzea, and morphology, statements such as "putative" or Leptospermum are tree genera which warrant "apparent" hybridism unsupported by suitable data national attention. should be unacceptable. It is, as I write, 50 years since the publication of Allan established contemporary solutions with the list of wild hybrids in New Zealand by Cock­ (i) his series "Illustrations of wild hybrids in the ayne & Allan (1934). That list, which has caused New Zealand flora" (Allan 1925, 1926b, c, 1927a, unease and uncertainty, developed very rapidly b, 1929a; Allan, Simpson & Thomson 1926); (ii) 636 New Zealand Journal of Botany, 1985, Vol. 23

detailed measurements of many character-states; Biosystematics may not meet the needs or expec­ (iii) artificial hybrids in Coprosma (Allan I 926a, tations of early 21 st century New Zealand botany 1929b, 1938). Fuchsia (Allan I 927b). Rubus (Allan in much the way that Hagen (1983) found that there 1927c, 1928, 1949), Phormium (Allan & Zotov was no revolution in pre-war taxonomy. But bio­ 1937); (iv) use of the experimental garden; (v) the systematists must find resolution to the question assessment of pollen sterility. of interspecific gene-exchange and the taxonomy of The onus for contemporary solutions 50 years higher plants; they could be expected to clarify later lies clearly with modern botanists. And if the whether or not natural hybridism is more impor­ work in the intervening 50 years has been tant in the evolution of the New Zealand flora than inadequate or even antiquated, the responsibility it is elsewhere as Raven (1973) asked, or as Rat­ - or fault if it is a fault - is also clear; if the work ten bury (1962) proposed. in the next half century is to be adequate, and even At the level of ecological investigation of the modern, the responsibility is equally clear. Data underlying causation of interpopulation differentia­ from plants in cultivation which at the very least tion Wardle (1969) and Sakai et al. (1978, 1981) in reflect the different responses plants make, e.g .. Phyllocladus, Wardle et al. (1973) in Leptosper­ juvenile forms, plasticity, ecotypic differentiation, mum scoparium , Given (1980) in Kunzea eri­ are abundant and needed, but unpublished. Fla­ coides, and Lee et al. (1983) in a number of genera, vonoid profiles could be used in assessing gene­ have demonstrated recognisable genetic compo­ exchange among genera such as Coprosma (Wilson nents. The causes underlying the ecological selec­ 1984), Carmichaelia (Purdie 1984), and Sophora tion of these gene complexes is the proper object (Markham & Godley 1972). Divarication is com­ of genecology (Turesson 1923). mon enough in all three genera. Godley (1979) went further than positing the Biosystematic studies are more easily carried out origin of Sophora microphylla; he advanced the than genecological ones; hybrids can be made hypothesis that the nine species in New Zealand experimentally and gene-exchange measured. with divaricating juvenile forms resulted from Genecology demands different kinds of experi­ hybridisations involving species of divaricating ments before adequate ecological explanations shrubs and arborescent species. The two qualities underpin interpopulation differentiation. Many of divarication and juvenility should be examina­ experiments in New Zealand reveal morphological ble in terms of the contemporary evidence of their and physiological differences in populations from genetic behaviour, but to be able to do so one must varied localities and from varied ecological con­ depend on accurate descriptions ofF, hybrids. Allan ditions, e.g. Connor (I 954a,b) in E/ymus rectisetus (1925, 1926b, I 927a, 1929a) contributes most of (as Agropyron scabrum), Ornduff (1960, 1964) in the evidence, although there is a list of 54 species the Senecio lautus alliance; Garnock-Jones & Lan­ of divaricating plants in the New Zealand flora in ger (1980) in Parahebe catarractae, and Yin et al. table I of Greenwood & Atkinson (1977), together (1984) in Leptospermum scoparium. Coincidence with an indication of interspecific hybridisation apart, none reveals genetic-ecological causation. with open shrub or tree forming species. Much conservation effort of the 1970s and 1980s Data on the inheritance of genes controlling has in New Zealand been centred on forest com­ juvenility are insufficient to be of much use to

Downloaded by [125.239.168.206] at 08:23 22 August 2017 munities especially, but not restrictively, on low­ Godley's hypothesis (op. cit.); but the general wealth land podocarp forest; parallel ancillary action has of information on juvenile forms should allow, after taken place over wetlands but a basic cause here is a complete morphological examination of hybrids, game birds and not wetlands per se. With the an assessment of the inheritance of this spectacular development of the concept of Ecological Districts characteristic. (see Simpson 1982) - and their misleadingly Divarication, an equally spectacular character and named Ecological Regions (Simpson op. cit.) - one involved in current debate (Greenwood & conservation must expand correctly to embrace all Atkinson 1977; Atkinson & Greenwood 1980; major members of the great plant communities­ McGlone & Webb 1981), should be more easily grasslands, forests, alpine fell fields, shrublands, assessed in terms of its inheritance. From the evi­ sand dune communities, and aquatic systems. Other dence that is available, most of it again from Allan than by exclosure, by noxious weed and animal (see above and 1924, 1937), the divaricating habit control, by preventive action against fire and loot­ does not appear to be under the control of domi­ ing, and by legislation, the conservation effort nant genes by comparison with genes controlling should be directed towards the understanding of open tree shape and form; but the same remarks the genetic differentiation in the dominant plants about the possibility of improving our knowledge of those communities. Genecological - ecogenetic of the inheritance of juvenility apply to the inher­ - investigation must become, and quite easily itance of the divarication syndrome. could become, part of conservation practice in New Connor-Biosystematics: higher plants 637

Zealand. It is inconsistent to assert that the con­ the century is to see its adequate replacement _ servation of species is a safeguard against the pos­ as it must - we have seen its half-life already. sible loss of genes which may become valuable to Biosystematics, with its obligatory tradition of man in the future without the logical investment experiment, must therefore make progress apace in of genetic research in those same plants which the New Zealand Dicotyledones. dominate forests and grasslands and shrublands and herbfields and sand dunes and lakeshore commu­ nities. The ardent conservation clamour will ACKNOWLEDGMENTS become credible through willingness to complete I am grateful to my successor Dr Warwick Harris. Direc­ research work, or even to fund it, rather than arro­ tor, Botany Division, DSIR, for the opportunity to speak gantly to demand it. on Biosystematics at the symposium to honour Dr E. J. Godley, a former Director of Botany Division. For more At a level of a conservation much below the pro­ than 20 years he supported biosystematics at the Divi­ tection concept, should the number of individual sion, sustaining, in particular, cytology and experimental plants of a species locally fall, or be reduced by programmes in the many genera which are discussed man or predators, below critical reproductive levels, above. As one who benefitted greatly from that support enrichment by the addition of individual plants of I acknowledge his guidance, liberalism, and friendship. the same species is a practice to be commended. That the individuals being added must be mem­ bers of the same ecotype as the depleted popula­ REFERENCES tion is a sine qua non. Allan, H. H. 1924: On the hybridity of Coprosma cun­ A third level in this series from (i) widespread ninghamii Hook. f. New Zealand journal ofscience and abundant, to (ii) locally diminished is (iii) that and technology 6: 310-318. group of plants described in detail by Given (1981) ---1925: Illustrations of wild hybrids in the New as rare, and/or endangered. These plants, above all, Zealand flora I. Genetica 7: 287-292. should be objects of careful and detailed genecol­ ---- 1926a: The F, progeny resulting from the crossing of Coprosma propinqua with C. robusta. ogical research, and not solely matters for legisla­ Genetica 8: 155-160. tion, or for sequestration. ---- 1926b: I\lustrations of wild hybrids in the New In the evolution of the kind of botanical knowl­ Zealand flora II. Genetica 8: 369-374. edge needed in New Zealand with its short history --- 1926c: I\lustrations of wild hybrids in the New of European settlement, we have passed from the Zealand flora III. Genetica 8: 525-536. frontier days of the requirement of tree identifi­ --- 1927a: Illustrations of wild hybrids in the New cation manuals, through the manual of useful plants Zealand flora IV. Genetica 9: 145-156. phase, faster through to the weed manual phase --- 1927b: Illustrations of wild hybrids in the New because of the economic importance of weeds to Zealand flora V. Genetica 9: 499-515. agriculture and livestock health, to this present time ---- 1927c: An artifical Rubus hybrid. Transactions where conservation competes with maximum uti­ of the New Zealand Institute 58: 51-54. lisation of natural resources. The needs of the ---- 1928: Further notes on an artificial Rubus colonial phase were the more easily met than those hybrid. Transactions of the New Zealand Institute of the late 20th century; modern methods in botan­ 59: 643-644. Downloaded by [125.239.168.206] at 08:23 22 August 2017 ical research should allow the current needs to be ---- 1929a: I\lustrations of wild hybrids in the New satisfied rapidly. They must satisfy rapidly if no Zealand flora VI. Genetica 11 : 491-508. more genes are to be lost from such important ---1929b: The F, progeny resulting from the plants as the beeches, kamahi, tawa, and rimu than crossing of Coprosma propinqua with C. robusta. have already been lost to attrition by axe and fire, Genetica J J : 335-346. and by domestic and wild animals. ---1931: The significance of hybridism in the New Biosystematics in this vigesimal period was very Zealand flora. Report of 20th meeting of the Ausht­ strongly developed in monocotyledonous plants ralian and New Zealand Association for t e Advancement of Science. Vol. 20: 429-477. primarily because that element of the flora of New Zealand was receiving concerted attention. That ----1937: Wild species - hybrids in the phanero­ gams. Botanical review 3 : 593-615. most of them were herbaceous, easily propagated, sma of modest size, and flowered promptly in cultiva­ ----1938: Fruit characters in the F, ofa Copro cross. Transactions of the Royal 'Society of New tion, Astelia excepted, and were easily manipulated Zealand 68 : 34-36. for crossing experiments, were qualities to their ---- 1940: Natural hybridization in relation to tadx)­ advantage for biosystematic studies. onomy. In: The new systematics. Huxley, J. (e . It is now almost twenty-five years since H. H. Pp. 515-528 Oxford. Clarendon Press. Allan's "Flora of New Zealand Volume I, Indige­ --- 1949: Wild species - hybrids in the phanero- nous Tracheophyta" was published. If the turn of gams II. Botanical review 15: 77-105. 638 New Zealand Journal of Botany, 1985, Vol. 23

---- 1951: Botany in the service of the State. New ---- 1970: Artificial hybridisation in Epi/obium Zealand science review 9: 124-130. involving New Zealand, European, and North ---- 1961: Flora of New Zealand. Vol. I. Welling­ American species. New Zealand journal (if botany ton. Government Printer. 8: 94-97. Allan, H. H.; Simpson, G.: Thomson, J. S. 1926: A wild Burrows, C. J. 1967: Progress in the study of South Island Hebe community in New Zealand. Genetica 8: alpine vegetation. Proceedings ollhe New Zealand 375-388. Ecological SOCiety 14 : 8-13. Allan, H. H.; Zotov, V. D. 1937: An artificial cross between Chinnock, R. J. 1972: [1973] Natural hybrids between Phormium colensoi and P. ten ax. New Zealand Disphyma and Carpohrotlls (Aizoaceae) in New journal of science and technology 18: 799-804. Zealand. New Zealand journal of bolanv /0 : 615- 626. Atkinson, I. A. E.; Greenwood. R. M. 1980: Divaricating plants and moa browsing: a reply. New Zealand Cockayne, L. 1912: Observations concerning evolution journal of ecology 3: 165-167. derived from ecological studies in New Zealand. Transactions of the Nell" Zealand Institute 44: I- Barkworth. M. E.: Dewey, D. R.; Atkins, R. J. 1983: New SO. . generic concepts in the Triticeae of the intermon­ tane region: keys and comments. Great Basin ---- 1921: The southern·beech (Notho/agus) forests naturalist 43: 561-572. of New Zealand. New Zealand journal (if agricul­ Beever. R. E. 1981: Self-incompatibility in Cordyline kas­ ture 23: 353-360. par (Agavaceae). New Zealand journal of botany ---- 1923: Hybridism in the New Zealand tlora, New 19: 13-16. phylologist 22: 105-127. ---- 1983: Self-incompatibility in Cordyline pumi· Cockayne, L.: Allan, H. H. 1934: An annotated list of lio (Agavaceae). New Zealand journal lilbotany 21 : groups of wild hybrids in the New Zcaland flora. 93-95. Annals (il botany 48: I-55. Bcll. C. J. E. 1973: Mountain soils and vegetation in the Connor. H. E. 1954a: Studies in New Zealand Agropyron Owen Range, Nelson. 2. The vegetation. New I. Descriptions and distribution of species and Zealand journal of botany II : 73-102. infraspecific units. New Zealand journal (if science Bcuzenberg. E. J. 1970: Contributions to a chromosome and technology 35B: 315-333. atlas of the New Zealand flora - 14 Uncinia ---- 1954b: Studies in New Zealand Agropyron II. (Cyperaceae). New Zealalld journal of hotal/)' 8 : 260-261 . Breeding systems, New Zealand journal of science and technology 35B: 333-343. ---- 1975: Contributions to a chromosome atlas of the New Zealand flora - 17 Senecio (Composi­ ---- 1956: Interspecific hybrids in New Zealand Agropyron, Evolution 10: 415-420. tae). Nell" Zealand journal of botanv 13: 345-353. ----1982: Cytology. In: Triennial Report Botany ----1957: Agropyron x wallii a natural hybrid in Division 1979-1981. Connor. H. E. ed. Pp. 37-38. New Zealand. Transactions (if the Royal Society of Christchurch. Caxton Press. New Zealand 84: 757-760. ---- 1983: Contributions to a chromosome atlas of ---- 1960: Variation in leaf anatomy in Festuca the New Zealand tlora - 24 Coprosma (Rubi­ novae-::clandiae (Hack.) Cockayne and F. mal­ aceae). Nell" Zealand journal of hot any 21: 9-12. thell'sii (Hack.) Cheeseman. New Zealand journal (if science 3: 468-509. Bcuzenberg. E. J.: Groves, B. E. 1974: Contributions to a chromosome atlas of the New Zealand tlora - ---- I 962a: Studies in New Zealand Agropyron Part 16 Miscellaneous families. Nelt' Zealand journal of 3. Intraspecific hybrids in A. scahrulll (R. Br,)

Downloaded by [125.239.168.206] at 08:23 22 August 2017 botany 12: 211-217. . Beauv. New Zealand journal (if science 5 : 99-115. Beuzenberg, E. J.; Hair, J. B. 1983: Contributions to a ---- 1962b: Studies in New Zealand Agropyron Part chromosome atlas of the New Zealand tlora - 25 4. Interspecific hybrids A. scabrum (R. Br.) Beauy, Miscellaneous species, Nell' Zealand journal ol X A. kirkii Zotov. New Zealand journal of science botany 21: 13-20. . 5: 116-119, ---- 1984: Contributions to a chromosome atlas of ---- 1964: Tussock grasslands communities in the the New Zealand tlora - 27 Compositae. Nelt' Mackenzie Country, South Canterbury, New Zealand journal (il botany 22: 353-356. Zealand. New Zealand journal (~f botany 2: 325- Bothmer. R. von; Flink, J.: Jacobsen, N.; Kotimaki. M.: 351. Landstrom, T. 1983: Interspecific hybridization ---- 1965a: Breeding systems in New Zealand with cultivated barley (Hordeum vulgare L.). Her· grasses V. Naturalised species of Cortaderia. New editas 99: 219-244. Zealand journal of botany 3: 17-23. Boyce, W. R.: Newhook. F. J. 1953: Investigations into ---- 1965b: Breeding systems in New Zealand yellow-leaf disease of Phormium I. History and grasses VI. Control of gynodioecism in Cortaderia Symptomatology. Nell' Zealand journal of science richardii (End!.) Zotov. New Zealand journal Qf and technology 34A supp!. I-II. botany 3 : 233-242. Brockie. W. B. 1966: Artificial hybridisation of New ---- 1967: Interspecific hybrids in Chionochloa Zealand species and varieties of Epilobium. Nell' (Gramineae), New Zealand journal of botany 5: 3- Zealand journal of botany 4: 366-391. 16. Connor-Biosystematics: higher plants 639

---- 1968: Interspecific hybrids in hexaploid New Drury, D. G. 1973: Annotated key to the New Zealand Zealand Festuca. New Zealand journal of botany shrubby Senecioneae-Compositae and their wild 6: 295-308. and garden hybrids. New Zealand journal of botany ---- 1974: Breeding systems in Cortaderia (Gra­ 11: 731-784. mineae). Evolution 27: 663-678. Edgar, E. 1964a: The leafless species of Juncus in New ---- 1983: Cortaderia (Gramineae): Interspecific Zealand. New Zealand journal of botany 2: 177- hybrids and the breeding system. Heredity 51 : 395- 204. 403. ---- 1964b: The natural sterile hybrid Carex dian­ Connor, H. E.; Penny, E. D. 1960: Breeding systems in dra x secta. New Zealandjournal q{botany 2: 279- New Zealand grasses II. Gynodioecy in Arundo 285. richardii. New Zealand journal of agricultural ---- 1966: Luzula in New Zealand. New Zealand research 3: 725-727. journal of botany 4: 159-184. Connor, H. E.; Purdie, A. W. 1976a: Inheritance of tri­ ---- 1975: Australasian Luzula. New Zealand jour­ terpene methyl ethers in Cortaderia (Gramineae). nal of botany 13: 781-802. Phytochemistry 15: 1937-1939. Edgar, E; Connor, H. E. 1983: Nomina Nova III, 1977- ---- 1976b: Triterpene methyl ether differentiation 1982. New Zealand journal of botany 21 : 421-441. in Chionochloa (Gramineae). New Zealand journal Fisher, F. J. F. 1965: The alpine Ranunculi of New of botany 14: 315-326. Zealand. New Zealand Department of Scientific and ---- 1981: Triterpene methyl ethers in Chionochloa Industrial Research bulletin 165. (Gramineae): Distribution in western South Island, Forde, M. B. 1964: Ifaloragis erecta: a species complex New Zealand. New Zealand journal of botany 19: in evolution. New Zealand journal of botany 2 : 161-170. 425-453. Cook, I. M.; Mark, A. F.; Shore, B. F. 1980: Responses Gardner, R. O. 1978: The species of Alseuosmia (Alseuos­ of Leptospermum scoparium and L. ericaides miaceae). New Zealand journal afbotany 16 : 271- (Myrtaceae) to waterlogging. New Zealand journal 277. of botany 18: 233-246. Garnock-Jones, P. J.; Langer, H. J. 1980: Parahebe catar­ Cowlishaw, M. G.: Bickerstaffe, R.; Connor, H. E. 1983: ractae (Scrophulariaceae): infraspecific taxonomy. Intraspecific variation in the epicuticular wax New Zealand journal of botany 18: 285-298. composition of four species of Chionochloa. Bio­ chemical systematics and ecology 11 : 247-259. Garnock-Jones, P. J.; Molloy, B. P. J. 1982: Polymorph­ ism and the taxonomic status of the Hebe amplex­ Dawson, J. W. 1967a: The New Zealand species of Gin­ icaulis complex (Scrophulariaceae). New Zealand gidium (Umbelliferae). New Zealand journal of journal of botany 20: 391-399. botany 5 : 84-116. Given, D. R. 1980: Vegetation on heated soils at Kar­ ----1967b: New Zealand Umbelliferae - Ligno­ apiti, central North Island, and its relation to carpa gen. nov. and Scandia gen. nov. New Zealand ground temperature. New Zealand journal q{botany journal of botany 5: 400-417. 18: 1-13. ----1974: Validation of Gingidia (Umbelliferae). ---- 1981: Rare and endangered plants of New Kew bulletin 29: 476. Zealand. Wellington, A. H. & A. W. Reed Ltd. Dawson, J. W.; Webb, C. J. 1978: Generic problems in ---- 1984: A taxonomic revision of Celmisia Australasian Apioideae (Umbelliferae). Pp. 21-32. subgenus Pelliculatae section Petiolatae (Compos­ In: Actes du 2eme symposium international sur les itae-Astereae). New Zealand journal of botany 22: ombelliferes (Perpignan, 1977). Contributions 139-158. Downloaded by [125.239.168.206] at 08:23 22 August 2017 pluridisciplinaires a la systematique. Godley, E. J. '1949: Cytology and genetics and their appli­ Dewey, D. R. 1982: Genomic and phylogenetic relation­ cation to New Zealand plants. Tuatara 2: 109- ships among North American perennial Triticeae. 115. In: Grasses and grasslands: systematics and ecol­ ----1951: Two natural Agropyron hybrids occur­ ogy. Estes, J. R.; Tyrl, R. J.; Brunken, J. N. (ed.). ring in the British Isles. Annals of botany 15: 535- Pp. 51-88. Norman, University of Oklahoma Press. 544. ---- 1983a: Historical arid current taxonomic per­ ---- 1955: Breeding systems in New Zealand plants spectives of Agropyron, Elymus, and related gen­ 1. Fuchsia. Annals of botany 19: 549-559 era. Crop science 23: 637-642. ---- 1963: Breeding systems in New Zealand plants. ----1983b: New nomenclatural combinations in 2. Genetics of the sex forms in Fuchsia procum­ North American perennial Triticeae (Gramineae). bens. New Zealand journal of botany 1 : 48-52. Brittonia 35 : 30-33. ---- 1979: Leonard Cockayne and evolution. New ---- 1984: The genomic system of classification as Zealand journal of botany 17: 197-215. a guide to intergeneric hybridisation with the per­ Godley, E. J.; Smith, D. H. 1977: Kowhais and their flow­ ennial Triticeae. In: Gene manipulation and plant ering. Royal New Zealand Institute of Horticulture improvement. Gustafson, J. P. (ed.) Pp. 209-279. annual journal 5: 24-31. Plenum Publishing. ----1981: Breeding systems in New Zealand plants. Druce, A. P. 1957: The vegetation of Mt Kaiparoro. Bul­ 5. Pseudowintera colorata (Winteraceae). New letin afthe Wellington Botanical Society 29: 7-13. Zealand journal of botany 19: 151-156. 640 New Zealand Journal of Botany, \985, Vol. 23

Grayer-Barkmeijer. R. 1. 1978: Ravonoids in Parahebe ---- 1983: Contributions to a chromosome atlas of and Veronica: a chemosystematic study. Biochemi­ the New Zealand flora - 23 Ranunculus (Ran­ cal systematics and ecology 6: 131-137. unculaceae). New Zealand journal of botany 21 : 3-7. Greenwood. R. M. 1953: Red tussock in the Tararuas. Bulletin of the I-Vellington Botanical Society 26 : Hair. J. B.; Beuzenberg, E. J. 1960: Contribution to a 29-30. chromosome atlas of the New Zealand Flora - 4 Miscellaneous families. New Zealand joul'IJal oj Greenwood. R. M.; Atkinson. I. A. E. 1977: Evolution of science 3 : 432-440. divaricating plants in New Zealand in relation to ----; 1966: Contributions to a chro- moa browsing. Proceedings of the New Zealand mosome atlas of the New Zealand flora - 7 Mis­ Ecological Society 24 : 21-33. cellaneous families. New Zealand journal ()( botany Greer. D. M. 1979: Effects of long-term preconditioning 4: 255-266. on growth and flowering of some snow tussock ----; 1968: Contributions to a chro- (Chionochloa spp.) populations in Otago. New mosome atlas of the New Zealand flora - II Mis­ Zealand. Australian journal of hotany 27: 617-630. cellaneous families. Nell' Zealand journal of botany ---- 1984: Seasonal changes in photosynthetic 6: 19-24. activity of snow tussocks (Chionochloa spp.) along Hair. 1. B.; Beuzenberg, E. 1.; Pearson. B. 1967: Contri­ an altitudinal gradient in Otago, New Zealand. butions to a chromosome atlas of the New Zealand Oecologia 63: 271-274. flora - 9 Miscellaneous families. New Zealand journal of hotany 5: 185-196. Groves. B. E. 1977. Contribution to a chromosome atlas of the New Zealand flora - 19. Gnaphalium Hair, J. B.; Kalin Arroyo, M. T. 1984: Contributions to (Compositae). Nell' Zealand jOlll'lJal o( hotany 15 : a chromosome atlas of the New Zealand flora - 17-18. 28 Ourisia (Scrophulariaceae). New Zealand jour­ ilal of botany 22: 357-359. Groves. B. E.; Hair. 1. B. 1971: Contribution to a chro­ mosome atlas of the New Zealand flora - 15. Hair, 1. 8.; Raven, P. H.; Seavey, S. R. 1977: Cytological Miscellaneous families. New Zealand joul'IJal oj affinities of the Australasian species of Epilobium botany 9: 569-575. (Onagraceae). New Zealand journal ()( botany 15 : 1-4. Hagen. J. B. 1983: The development of experimental methods in plant taxonomy. 1920-1950. Taxon 32 : Hair. J. 8.; Webb, C. J.; Beuzenberg, E. J. 1980: Contri­ 406-416. butions to a chromosome atlas of the New Zealand flora - 22 Genliana (Gentianaceae). New Zealand Hair. J. B. 1956: Subsexual reproduction in Agropyron. journal of botany 18: 563-564. Heredity 10: 129-160. Healy, A. J.; Edgar, E. 1980: Flora of New Zealand. Vol. ---- 1966: Biosystematics of the New Zealand flora, III. Adventive cyperaceous, petalous and spatha­ 1945-1964. New Zealand journal of botany 4 : 559- ceous Monocotyledons. Wellington, Government 595. Printer. ---- 1967a: Contributions to a chromosome atlas Hillis. W. F.; Orman, H. R. 1962: The extractives of New of the New Zealand flora - 8 Lu::ula (Juncaceae). Zealand Nolhofagus species. Journal o( the Lin- Nell' Zealand journal of botany 5: 17-21. nean Society. hotany 58: 175-184. . ---- 1967b: Contributions to a chromosome atlas Hoekstra, F. A. 1983: Physiological evolution in angios­ of the New Zealand flora - 10 Hebe (Scrophular­ perm pollen: possible role of pollen vigour. In: Pol­ iaceae). New Zealand joul'IJal of hotany 5: 322- len: Biology and implications for plant breeding. (ed.). Downloaded by [125.239.168.206] at 08:23 22 August 2017 352. Mulcahy, D. L.; Ottaviano, E. Pp. 35-41. New York, Elsevier Science Publishing Co. Inc. ---- 1968: Contribution to a chromosome atlas of the New Zealand flora - 12 Poa (Gramineae). Nell' Jackson, W. D. 1958: Chromosome numbers in Tasman­ Zealand journal of botany 6 : 267-276. ian Goodeniaceae and Brunoniaceae. Papers of Royal Society of Tasmania 92: 161-163. ---- 1970: Contributions to a chromosome atlas of the New Zealand flora - 13 Parahebe and Pyg­ Jansen, M. E. 1978: Notes on Lu::ula (Juncaceae) in maea (Scrophulariaceae). New Zealand joul'IJal ()( Malesia, Tasmania, and Australia. Blumea 24: botany 8: 255-259. 527-532. Jeffrey, C. 1979: Generic and sectional limits in Senecio ---- 1977: Contribution to a chromosome atlas of (Compositae) II. Evaluations of some recent stud­ the New Zealand flora - 18 Epilobium (Onagra­ ies. Kew bulletin 34: 49-58. ceae). Nell' Zealand joul'IJal (!( botany 15 : 5-15. ---- I 980a: Contributions to a chromosome atlas Kenton, A.; Heywood C. A. 1984: Cytological studies in South American Iridaceae. Plant systematics and of the New Zealand flora - 20 Ce!misia and Dam­ namenia (Compositae). New Zealand joul'IJal ()[ evolution 146: 87-104. botany 18: 553-558. Kiikenthal. G. 1909: Cyperaceae - Caricoideae. Pilan- ---- 1980b: Contributions to a chromosome atlas zen reich 38: 1-824. . of the New Zealand flora - 21 Umbelliferae (Mis­ Le Comte. J. R.; Webb C. J. 1981: Aciphylla townsonit' cellaneous genera). New Zealand joul'IJal ()( botany - a juvenile form of A. hookeri (Umbelliferae). 18: 559-562. New Zealand journal of botany 19: 187-191. Connor-Biosystematics: higher plants 641

Lee, W. G.; Given, D. R. 1984: Celmisia spedenii G. Martin-Smith, M.; Subramanian, G.; Connor, H. E. 1967: Simpson, an ultramafic endemic, and C. markii sp. Surface wax components of five species of Cortad­ nov., from southern New Zealand. New Zealand eria (Gramineae): a chemo-taxonomic compari­ journal of botany 22: 585-592. son. Phytochemistry 6: 559-572. Lee, W. G.; Mark, A. F.; Wilson, J. B. 1983: Ecotypic Martin-Smith, M.; Ahmed, S.; Connor H. E. 1971: Sur­ differentiation in the ultramafic flora of the South face wax components in species and hybrids of Island, New Zealand. New Zealand journal of Cortaderia. Phytochemistry 10: 2167-2173. botany 21: 141-156. Melville, R. 1976: Neoteny, evolution and the New Lloyd, D. G. 1972: A revision of the New Zealand, Zealand Parsonsia hybrids (Apocynaceae). Botan­ Subantarctic, and South American species of ical journal Qfthe Linnean Society 72: 171-189. Cotula, section Leptinella. New Zealand journal of Molloy, B. P. J.; Connor, H. E. 1970: Seedling growth in botany 10 : 277-372. four species of Chionochloa (Gramineae). New ---- 1973: Sex ratios in sexually dimorphic Umbel­ Zealand journal of bolany 8: 132-152. liferae. Heredity 31 : 239-249. Molloy, B. P. J.; Burrows, C. J.; Cox, C. J.; Johnston, J. ---- 1975: Breeding systems in Cotula. III. Dioe­ A.; Wardle, P. 1963: Distribution of subfossil forest cious populations. New phylologisl 74: 109-123. remains, eastern South Island, New Zealand. New ---- 1981: Evolution of prostrate and erect habits Zealand journal of botany 1 : 68-77. in Cotula section Leptinella and other New Zealand Moore, L. B. 1966: Australasian Asteliads (Liliaceae). New plant groups. New Zealand journal of botany 19: Zealand journal of botany 4: 201-240. 247-253. ---- 1967: The New Zealand species of Libertia ---- 1982: Variation and evolution of plant species (Iridaceae). New Zealand journal of botany 5: 255- on the outlying islands of New Zealand, with par­ 275. ticular reference to Cotulafeatherstonii. Taxon 31: ---- 1970: Some implications of precocious flow­ 478-487. ering in Collospermum. Pacific science 24: 409- Lotsy, J. P. 1925: Evolution considered in the light of 413. hybridization. Christchurch, Andrews, Baty & Co ----1975: Hybridism in Cordyline(Agavaceae). New Ltd. Zealand journal of botany 13: 305-316. LOve, A. 1982: Generic evolution of the wheat grasses. ---1980: Hybrid Astelias (Liliaceae). New Zealand Biologisches Zentralblatt 101: 191-212. journal of botany 18: 37-42. ---- 1984: Conspectus on the Triticeae. Feddes Moore, L. B.; Edgar, E. 1970: Flora of New Zealand Vol. Repertorium 95: 425-521. II. Wellington Government Printer. LOve, A.; Connor, H. E. 1982: Relationships and taxon­ Nordenskiold, H. 1966: Cytogenetic studies in the genus omy of New Zealand wheatgrasses. New Zealand Luzula in New Zealand. New Zealand journal of journal of botany 20: 169-186. botany 4: 185-195. Lyon, G. L.; Brooks, R. R.; Peterson, P. J.; Butler, G. W. ---- 1969: The genus Luzula in Australia. Botan­ 1968: Trace elements in a New Zealand serpentine iska Notiser 122: 69-89. flora. Plant and soil 29: 225-240. ---- 1971: Hybridization experiments in the gt;nus Lyon, G. L.; Peterson, P. J.; Brooks, R. R. 1969: Chro­ Luzula IV. Studies with taxa of the campes ris - mium-51 distribution in tissues and extracts of multiflora complex from the northern and ',outh­ Leptospermum scoparium. Planta 88: 282-287. em hemispheres. Hereditas 68: 47-60. Lyon, G. L.; Brooks, R. R.; Peters~n, P. J.; Butler, G. W. Nordenstam, B. 1978: Taxonomic studies in the tribe 1970: Some trace elements III plants from serpen­ Senecioneae (Compositae). Opera Botanica 44: 1- Downloaded by [125.239.168.206] at 08:23 22 August 2017 tine soils. New Zealand journal ofscience 13: 133- 84. 139. Ogden, J. 1974: Observations on two coastal ecotypes of ---- 1971: Calcium, magnesium and trace elements Selliera radicans Cav. (Goodeniaceae) growing in in a New Zealand serpentine flora. Journal of ecol­ the Manawatu District of New Zealand. New ogy 59: 421-429. Zealand journal of botany 12: 541-550. McGlone, M. S. 1983: polynesian deforestation in New ----1978: Variation in Calystegia R. Br. (Con vol­ Zealand: a preliminary synthesis. Archaeology in vulaceae) in New Zealand. New Zealand journal Oceania 18: 11-25. of botany 16: 123-140. McGlone, M. S.; Webb, C. J. 1981: Selective forces influ­ Orchard, A. E. 1975: Taxonomic revisions in the family encing the evolution of divaricating plants. New Haloragaceae. I. The genera Haloragis. Halora­ Zealand journal of ecology 4: 20-28. godendron, Glischrocaryon. Meziella and Gono­ Mark, A. F. 1965: Ecotypic differentiation in Otago popu­ carpus. Bulletin of the Auckland Institute and lations of narrow-leaved snow tussock Chionoch­ Museum No. 10: 1-299. loa rigida. New Zealand journal of botany 3 : 277- Ornduff, R. 1960: An interpretation of the Senecio lautus 299. complex in New Zealand. Transactions ofthe Royal Markham, K. R.; Godley, E. J. 1972: Chemotaxonomic Society of New Zealand 88 : 63-77. . studies in Sophora I. An evaluation of Sophora ---- 1964: Evolutionary pathways of the Senecio microphylla Ait. New Zealand Journal of botany lautus alliance in New Zealand and Australia. 10: 627-640. Evolution 18: 349-360. 642 New Zealand Journal of Botany, 1985, Vol. 23

Pauptit, R. A.; Waters, J. M.; Rowan, D. D.; Russell, G. Sakai, A.; Wardle, P. 1978: Freezing resistance of New B.; Connor, H. E.; Purdie, A. W. 1984: The struc­ Zealand trees and shrubs. New Zealand journal of ture of 19cxH-lupeol methyl ether from Chionoch­ ecology 1 : 51-61. loa bromoides. Australian journal of chemistrv 37: 1341-1347. .. Sampson, F. B. 1980: Natural hybridism in Pseudo win­ tera (Winteraceae). New Zealand journal of botany Park, G. N. 1968: The vegetation and flora of Mt Stokes. 18: 43-51. Bulletin of the Wellington Botanical Societv 35 : 43-53. . Scott, D. 1970: Relative growth rates under controlled temperatures of some New Zealand indigenous and Philipson, M. N. 1978: Apomixis in Cortaderia jubata introduced grasses. New Zealand journal of botany (Gramineae). New Zealand journal of hotany 16 : 8: 76-81. 45-59. Poole, A. L. 1950a: Studies of New Zealand Nothofagus Sewell, T. G.; Holloway, J. T.; Moore, L. B.; Raeside, J. 1. Taxonomy and floral morphology. Transactions D.; Saxby, H. S.; Connor, H. E. 1954: The high of the Royal Society of New Zealand 78: 363-380. altitude snow-tussock grassland in South Island, New Zealand. New Zealand journal of science and ---- 1950b: Studies of the New Zealand Nothofagus techology 36A : 335-364. species. 2. Nut and cupule development. Trans­ actions ()( the Royal Society (}f New Zealand 78 : Skottsbcrg, C. 1934: Studies in the genus Astelia Banks 502-508. et Solander. Kungliga Svenska I'etenskapsakade­ miens handlingar 14 (2): 1-106. ---- 1951: Hybrid southern beeches. New Zealand journal (~f forestry 6 : 250-253. Simpson, P. [Compiler] 1982: Ecological regions and dis­ ---- 1952: The dcvclopment of Nothofagus seed tricts of New Zealand. A natural subdivision. Bio­ (including a preliminary account of the embry­ logical Resources Centre Publication I. New ogeny etc). Transactions (}fthe Royal Society ofNew Zealand, Wellington. Zealand 80 : 207-212. Stace. C. A. 1961: Some studies in Calystegia: compati­ ---- 1958: Studies of the New Zealand Noth(}fagus bility and hybridisation in C. sepium and C. sil­ species III - The entire leaved species. Transac­ mtica. Walsonia 5: 88-105. tions of the Royal Society (}f New Zealand 85 : 551- Sykes, W. R. 1971: Senecio lautus in the Kermadec 564. Islands. New Zealand journal of botany 9: 533- Post, D. M. 1983: Contributions to a chromosome atlas 538. . of the New Zealand flora - 26 Gentiana (Gen­ Thompson, J. 1983: Redefinitions and nomenclatural tianaceae). New Zealand journal (}f botany 21 : 229. changes within the Leptospermum suballiance of Purdie, A. W. 1984: Some flavonoid components of Car­ Myrtaceae. Telopea 2: 379-383. michaelia (Papilionaceae) - a chemotaxonomic Thomson, A. D. 1979: Annotated summaries of letters to survey. New Zealand journal (}f botany 22: 7-14. colleagues by the New Zealand botanist Leonard Quinn, C. J. 1982: Taxonomy of Dacrydium Sol. ex Lamb. Cockayne - I. New Zealand journal of botany 17 : emend. de Laub. (Podocarpaceae). Australian jour­ 389-416. nal of ;'otany 30: 311-320. ---- 1980: Idem - 2. New Zealand journal of Quinn, C. J.; Rlttenbury, J. A. 1972: Structural hybridity botany 18: 405-432. in New Lcaland Dacrydium. New Zealand journal of botany 10: 427-436. Turesson, G. 1923: The scope and impact of genecology. Hereditas 4: 171-176. Rattenbury, J. A. 1962: Cyclic hybridization as a survival mechanism in the New Zealand forest flora. EI'olu­ Turrill, W. B. 1940: Experimental and synthetic plant tax­

Downloaded by [125.239.168.206] at 08:23 22 August 2017 tion 16: 348-363. onomy. In: The New Systematics. Huxley, J. (ed.) Pp. 47-71. Oxford, Clarendon Press. Raven, P. H. 1973: Evolution of subalpine and alpine plant groups in New Zealand. New Zealand jour­ Vink, W. 1970: The Winteraceae of the Old World. I. nal of botany 11: 177-200. Pseudowintera and Drimys - morphology and taxonomy. Blumea 18: 225-354. Raven, P. H.; Raven, T. E. 1976: The genus Epilobium (Onagraceae) in Australasia: a systematic and evo­ Wardle, P. 1968: The taxonomy and distribution of the lutionary study. New Zealand Department (}f Sci­ stipitate species of Pseudopanax in New Zealand. entific and Industrial Research bulletin 216. New Zealand journal of botany 6: 226-236. Rowley, J. 1970: Effect of burning and clipping on tem­ ---- 1969: Biological Flora of New Zealand 4. Phyl­ perature, growth, and flowering of narrow-leaved loc/adus alpinus Hook. f. (Podocarpaceae) moun­ snow tussock. New Zealand journal of botany 8: tain toatoa, celery pine. New Zealand journal of 264-282. botany 7: 76-95. Russell, G. B.; Connor, H. E.; Purdie, A.W. 1976: Tri­ ---- 1972: Podocarpus totara var. waihoensis var. terpene methyl ethers of Chionochloa (Grami­ nov.: the result of introgressive hybridization neae). Phytochemistry 15: 1933-1935. between P. totara and P. acutifolius. New Zealand journal of botany 10: 195-201. Sakai, A.; Paton, D. M.; Wardle, P. 1981: Freezing resist­ ance of trees of the South Temperate Zone, espe­ ---- 1979: Variation in Phorrnium cookianum cially subalpine species of Australasia. Ecology 62: (Agavaceae). New Zealand journal of botany 17: 563-570. 189-196. Connor-Biosystematics: higher plants 643

Wardle, P.; Baylis, G. T. S.; Mark, A. F. 1973: Vegetation Williams, P. A.; Mugambi, S.; O'Connor, K. F. 1977: and landscape of the West Cape District, Fiord­ Shoot properties of nine species of Chionochloa land, New Zealand. New Zealand journal oj botany Zotov (Gramineae). New Zealand journal oJbotany 11 : 599-626. 15: 761-765. Webb, C. J. 1979: Breeding systems and the evolution of Williams, P. A.; Mugambi, S.; Nes, P.; O'Connor, K. F. dioecy in New Zealand apioid Umbelliferae. 1978a: Macro-element composition of tall-tussock Evolution 33: 662-672. (Chionochloa) in the South Island, New Zealand ---- 1981: Test of a model predicting equilibrium and their relationship with soil chemical proper­ frequencies of females in populations of gynodioe­ ties. New Zealand journal oj botany 16: 479-498. cious Angiosperms. Heredity 46 : 397-405. Williams, P. A.; Grigg, J. L.; Mugambi, S.; O'Connor, K. Webb, C. J.; Druce, A. P. 1984: A natural intergeneric F. 1978b: Soil chemical properties beneath tall-tus­ hybrid, ACiphylla squarrosa x Gingidia montana socks (Chionochloa) in South Island, New Zealand. and the frequency of hybrids among other New New Zealand journal oj science 21: 149-156. Zealand apioid Umbelliferae. New Zealand jour­ nal oj botany 22: 403-411. Williams, P. A.; Grigg, J. L.; Nes, P.; O'Connor, K. F. 1978c: Macro-elements within shoots of tali-tus­ Wheeler, J. M. 1966: Cytotaxonomy of the large asteliads socks (Chionochloa), and soil properties on Mt (Liliaceae) of the North Island of New Zealand. Kaiparoro, Wairarapa, New Zealand. New Zealand New Zealand journal oj botany 4: 95-113. journal oj botany 16: 255-260. Wigston, D. L. 1979: NothoJagus Blume in Britain. Wat­ sonia 12: 344-345. Wilson, H. D. 1978: Wild plants of Mount Cook National Park. Christchurch, Field Guide Publications. Wilcox, M. D.; Ledgard, N. J. 1983: Provenance varia­ tion in the New Zealand species of NothoJagus. New Wilson, R. D. 1979: Chemotaxonomic studies in the Zealand journal oj ecology 6: 19-31. Rubiaceae. I. Methods for the identification of WiJliams, C. A.; Fronczyk, J. H.; Harborne, J. B. 1983: hybridization in the genus Coprosma J. R. et G. Forst. using flavonoids. New Zealand journal oj Leaf flavonoids and other phenolic glycosides as indicators of parentage in six ornamental Fuchsia botany 17: 113-116. species and their hybrids. Phytochemistry 22: 1953- ---- 1984: Chemotaxonomic studies in the Rubi­ 1957. aceae 2. Leaf flavonoids in New Zealand copros­ WiJliams, P. A. 1975: Studies of the tall-tussock (Chion­ mas. New Zealand journal oj botany 22: 195-200. ochloa) vegetation/soil systems of the southern Yin Ronghua; Mark, A. F.; Wilson, J. B. T. 1984: Aspects Tararua Range, New Zealand. 2. The vegetation of the ecology of the indigenous shrub Leptosper­ /soil relationships. New Zealand journal oj botany mum scoparium (Myrtaceae) in New Zealand. New 13: 269-303. Zealand journal oj botany 22: 483-507. ---- 1977: Growth, biomass, and net productivity of tall-tussock (Chionochloa) grasslands, Canter­ Zotov, V. D. 1943: Certain changes in the nomenclature bury, New Zealand. New Zealand journal oj botany of New Zealand species of Gramineae. Transac­ 15: 399-442. tions oj the Royal Society oj New Zealand 73 : 233- 238. Williams, P. A.; Grigg, J. L.; Nes, P.; O'Connor, K. F. 1976: Vegetation/soil relationships and distribu­ ---- 1963: Synopsis of the grass subfamily Arun­ tion of selected macro-elements within the shoots dinoideae in New Zealand. New Zealand journal of tall tussocks on the Murchison Mountains, oj botany 1 : 78-136. Fiordland, New Zealand. New Zealand journal oj ---- 1970: Chionochloa macra (Gramineae): a new botany 14 : 29-53. species. New Zealand journal oj botany 8 : 91-93. Downloaded by [125.239.168.206] at 08:23 22 August 2017