Flora of New Zealand Series

PREFACE

1. Historical

"Is there a fatherless book, an orphan volume in this world? A book that is not the descendant of other books? . . . Is there creation without tradition?" In How I wrote one of my Books Carlos Fuentes (1988) posed these questions. This grass Flora is the direct descendant of John Buchanan's Indigenous Grasses of New Zealand, the first New Zealand book devoted solely to grasses, originally published in three parts - Plates 1-20 in 1878, Plates 21-40 June 1879, and Plates 41-58, 17(2), 26(2), and 36(2) in June 1880, and printed in Imperial Quarto (14½ inches × 10 inches) because the plates were to be nature-printed. The half-bound edition weighs 2.66 kg. As James Hector remarked in his Preface to the 1880 Royal Octavo (9½ inches × 6 inches) complete edition, the large size of the original was inconvenient, but the grasses would now be illustrated at half their natural size. The arrangement of the text differs between the 1880 Imperial Quarto (price 3 guineas) and the Royal Octavo (price 7s.6d.) editions, particularly in the placing of the nomenclaturally important Addenda et Corrigenda. The differing pagination of the indexes to Popular Names and to Genera and Species is not critical, but the omission of Buchanan's "Remarks on the Distribution of Grasses in New Zealand" of pages 8-10 of the Introduction denied general access to one aspect of its original purpose.

Thus Buchanan's illustrated Manual; but Buchanan was guided by J. D. Hooker's Handbook of the New Zealand Flora (1864, 1867) with its 26 genera of grasses, because Buchanan dealt with the same 26 genera although he increased them to 28 by plates for Stipa petriei and Deyeuxia scabra as Addenda. Buchanan is a Hooker descendant; his lineage is secure. Hooker's Handbook was "Published under the Authority of the Government of New Zealand." Buchanan's was "Published by Command"; there was discussion about it, in the House of Representatives (Parliamentary Debates 20: 217-219, 340 [1876]) notably by Sir George Grey K.C.B., and its preparation was resolved in Parliament on 29 June 1876.

Our Flora may also be interpreted as the linear successor of H. H. Allan's Introduction to the Grasses of New Zealand (1936), price 4s. Allan's purpose was "to provide a means of recognising the different species" and to encourage ". . . others to join the present small but enthusiastic band of collectors." That band provided the basis for this Flora. A more correct interpretation might be that Flora Volume V is the final step in the sequence from Allan's Introduction. The compass is the same. Allan included 76 genera and both native and naturalised species; he provided keys often as workable now as then. In 160 pages there is ample introduction, as one of us remembers well, and there are 102 figures. Figures 1-4 are explanatory of grass features, 5-102 are species habit or detail; some include several taxa, e.g., in fig.12 there are six taxa, in fig. 13 there are 10 taxa. What has never been discussed is the number of plates drawn by V. D. Zotov whom Allan acknowledges as providing ". . . the detail drawings of indigenous grasses." By our count Zotov drew 27 plates incorporating 130 taxa. There were 72 habit drawings and 387 details of spikelets, florets, and flowers; often the detail is too small, and the captions all lack any explanation. Those drawings of native grasses were never used again even though Allan assured us that ". . . they were prepared for another purpose."

H. H. Allan in his Handbook of the Naturalised Flora of New Zealand (1940) listed the grasses in two parts, 87 naturalised species, and a further 60 species in the Appendix, some of which were ". . . non-persistent escapes from cultivation." Five were thought to have fairly good claim to naturalisation: Avena strigosa, Entolasia marginata, Glyceria striata, Nassella trichotoma, Oryzopsis miliacea; all are now.

Flora of New Zealand Vol. V Gramineae will meet the modern requirements of a flora - descriptions, distributions, keys, uncertainties, modern classification, notes, reference to allied studies, and to the relevant revisions. It will inform New Zealand about the grasses resident here, regardless of their status as native or naturalised entities. It will tell the world outside New Zealand of the kinds of grasses here, especially those with which they are unacquainted. It will tell other parts of the world about the variation, behaviour, ecology, and distribution of their grasses in the New Zealand environment.

We are unable to meet some of the requirements of a flora that Funk (1993) elaborated, such as that of solely monophyletic taxa, and a phylogenetic arrangement of them; we do not know either, but opinions are widely available in up-to-date summaries by Hsiao et al. (1999) and by Soreng and Davis (1998). We do know that we tired of referees' comment on our revisions "Where is the cladogram?" We do endeavour, however, to provide comparable and general descriptions, and synopses where the groups are large. Endemism is clearly designated; but not always can we indicate the closest relative of any taxon because that is not the stage of evolution of our flora writing. No-one should have difficulty with the detection of the circumpolar, the subantarctic, or the Pacific elements, and thus the subregions within the New Zealand Botanical Region.

Nowhere in this Flora will there be a philosophical discussion of the Gramineae - the absence of epiphytism, parasitism, succulence, or heterostyly for example. The evolution of the endemic grass florula is not attempted here, and although well aware of the Gondwanaland influences we find discussion of them unsuited to a flora. Nor will we debate the outcomes of studies deriving from molecular biology; those conclusions are readily available and recent as in Soreng and Davis (1998) and Hsiao et al. (1999). We choose only to note two novel grass universals, the small trnT inversion in cpDNA (Doyle et al. 1992), and the crimp mark on the leaf-blades indelibly present through the action of the ligule (Espie, Connor and McCracken 1992), and note also the suggestion of the pathway to one morphological character that defines the subfamily Panicoideae (Le Roux and Kellogg 1999).

The role of experiment as an adjunct to taxonomy came to the fore after World War II. At Botany Division, D.S.I.R., experimental projects included growing plants from different localities and habitats in a common garden, and the production of interspecific and intraspecific hybrids in several grass genera. One example of large scale experiment was in Cortaderia, a danthonioid genus of five endemic species, the toetoes, and two naturalised species, the pampas grasses of South American origin.

Gynodioecism where the populations consist of plants bearing hermaphrodite flowers and plants with female (male sterile) flowers is present in all endemic species of Cortaderia, except in C. turbaria from the Chatham Is. Extensive experiments involving just more than 13,000 plants in cultivation mostly of South Island C. richardii, but including the other endemics, attempted to interpret the genetics of male sterility. Because each generation of plants from controlled pollination first flowered after four years these species were not as ideal subjects for experiment as the naturalised pampas grass, C. selloana, which comes to flower in its first growing season. Families of C. selloana, totalling 17,000 plants, were raised, and analysis of the frequencies of male steriles showed that sterility was under recessive genic control, the double recessive at any of up to three complementary loci causing male sterility. No adequate solution accounted for male sterility in endemic species other than to note that control was genic, probably recessive, and common to all four species.

Trials with grasses in cultivation may not always be simple or rapid. From seeds Festuca novae-zelandiae and F. matthewsii come to flower in their second growth year, and Chionochloa spp. raised from seeds come into flower in their third or fourth year. Some endemic species of Poa flower in their first year, but not all, and endemic species of Elymus also come to flower in their first year from seeds. Large scale experiments with native grasses have not always been easily or quickly managed.

2. Taxonomic Treatment

In this volume of the Flora of New Zealand we have followed the taxonomic system that W. D. Clayton and S. A. Renvoize set out in their Genera Graminum: Grasses of the World (1986). Amendments to their scheme and some nomenclatural variation especially of tribal names will be evident, but our departures are so insignificant as to cause little disquiet and every agrostologist will easily find the related group. At the generic level the defence of our choices, should defence ever be needed, is usually in one of our revisions that have preceded this volume.

Volume V in the Flora of New Zealand series departs from Vols I-IV, as they from each other, by presenting endemic, indigenous, and naturalised taxa equally and without any discriminating sign. Volume V follows the style of Vols I and II by presenting the full synonymy for endemic species, and for indigenous species as it affects New Zealand; for naturalised taxa it offers the best binomial or polynomial, just as in Volumes III and IV.

The only generic name of uncertain application is Kampmannia Steud. , Syn. Pl. Glum. 1: 34 (1854) with the single species K. zeylandica Steud. The type, said to be of New Zealand origin, has been lost and the protologue does not allow of any positive interpretation.

Two issues of concern to taxonomists, though not necessarily so to all users, are nomenclature and typification.

Typification

The typification of all New Zealand endemic and indigenous taxa is indicated throughout; holotypes are signified and for lectotypes the designators and references are given in full. Typification of a further 11 taxa is listed on p. xxxiv; we do it in this Flora to avoid a separate, and probably later, publication. Some indigenous taxa are based on New Zealand plants; in those, should the necessity for lectotypification arise, it is seen as a local responsibility; by contrast, for indigenous taxa based, for example, on Australian collections, we refrain from typification on the grounds that the responsibility is best assumed by botanists with local knowledge. Thus for dioecious Spinifex sericeus, shared with Australia, the R. Brown sheet 6147, East Coast, Broad Sound, BM! comprises both sex forms of which the female is a solitary inflorescence, the lectotypification by Craig (1984) is incomplete.

The rare neotype, or replacement of a lectotype known to have been destroyed, is clearly defined. It is our belief that this emphasis on and clear identification of holotypes and lectotypes will benefit future New Zealand agrostographers.

Nomenclature

For endemic or indigenous taxa the appropriate synonymy is given. Endemic taxa are no problem, but for indigenous species, those shared perhaps with Australia, the holotype or lectotype may be of Australian provenance, and the synonymy may not correctly apply to New Zealand plants, e.g., in Hierochloe redolens the binomial Disarrenum antarcticum Labill. , Nov. Holl. Pl. Spec. 2: 83 t. 232 (1807), with its subsequent combinations, is based on an Australian provenance. Melica magellanica Desr. in Lamarck Encycl. Méth. Bot. 4: 72 (1797) with its subsequent combinations, is based on Commerson's specimen from the Straits of Magellan. Both names are treated by all authorities as synonymous with H. redolens. We have not seen the types for either, and having limited ourselves to material originating in New Zealand, have not included these names in our synonymy.

We have endeavoured to verify, unify, correct, and bring up-to-date all nomenclatural references; in doing so there will be some differences - dates in particular - from those commonly found in earlier works. Poa, Agrostis, and Rytidosperma are examples of our approach to nomenclature; the status of the plant in New Zealand determines our treatment. Synonymy of naturalised species is never given, nor is typification indicated except by way of offering older or historical names.

Hawksworth (1992) asked how much time taxonomists spent on nomenclatural matters, time that could well have been better devoted to taxonomic studies. Our answer for our studies is simple - no longer than necessary. Homotypic synonyms are relatively easily determined and, apart from customary verification of date and place, often pose no major problem other than to ensure that the epithet in question is not a later homonym and that transfers were not precluded by earlier names. Heterotypic synonymy involves taxonomic judgement, and it is the taxonomy not the nomenclature that is demanding. We have devoted our time to taxonomy and attended to the nomenclature as its consequence. Never for a moment would we deny that some nomenclatural issues were not a source of concern - worry, more often. In Lachnagrostis where many combinations were illegitimate but the varietal epithets were valid, confusion may seem to rule even though it never rises above initial uncertainty.

Autonyms, names generated automatically as the consequence of subdivisions of a taxon - species in this Flora - apply in several genera; we clearly indicate their use.

The names of New Zealand grasses reflect nothing of indigenous origin at generic level, and at specific level "toetoe" and "unarede" are the only uses of Maori words. Chionochloa flavicans forma temata invokes a Maori place name - even if its author was moved to its use by the name of a local wine estate. Rakiura, the Maori name for Stewart Island is reflected in Poa aucklandica subsp. rakiura. No political overtones, no conservation ploy, no historic person, event, or place is called into commemoration even though there were many opportunities. Our grass nomenclature is singularly free of overtones.

3. Structure of the Grass Flora

An analysis of the structure of the grass flora is in Table 1; there arranged by tribes, the number of species in each, and their classification into endemic, indigenous, naturalised, or transient (zeta entries, ξ) is shown. Of these four classes only the last may need definition - these are the species found once, or perhaps twice, and not collected again. Their share is 10% of the entries in this Flora, and of course increases the incidence of the alien elements here.

The endemic florula is 157 species, c. 38% of the total, and has its highest concentration of taxa in three tribes where the frequencies are almost equal - Poeae 49 spp.; Agrostideae 45 spp.; Danthonieae 43 spp. Addition of the indigenous element is only conspicuous in Agrostideae, the remaining taxa are spread very thinly among the tribes.

On biogeographic grounds the expectation for indigenous Bambuseae must be low, and in oligotypic tribes the chances for an indigene or endeme are greatly reduced. The signal lack is of chloridoid and panicoid endemics; the significant entry is Agrostideae in all classes - endemics, indigenous, and naturalised taxa. World-wide that tribe has an estimated 1000 spp. mostly of temperate areas.

The naturalised grass florula exceeds the total of the endemic and indigenous components, and the evidence of the number of species which had escaped from cultivation or lived a short time after diaspores were first dispersed is indicative of a further probable increase in alien grasses.

It may be of no great significance but only in the tribes Poeae, Agrostideae and Danthonieae does the number of endemic and indigenous species exceed the alien element (Table 1).

Endemic Grasses

Genera There are five endemic genera of New Zealand grasses, two of them are monotypic and the others are oligotypic: Anemanthele 1 sp. (Stipeae), Pyrrhanthera 1 sp. (Danthonieae), Simplicia 2 spp. (Agrostideae), Stenostachys 3 spp. (Hordeeae), Zotovia 3 spp. (Ehrharteae).

Linder and Verboom (1996) reduced Pyrrhanthera to synonymy in Rytidosperma, and Willemse (1982) concluded that Zotovia spp. (as Petriella) should be included in Ehrharta. The generic status of Anemanthele, Simplicia, and Stenostachys has not been assailed. None is diploid, and all have discernible relatives except Simplicia.

Species The frequency of endemic species, arranged by tribes, is inTable 1, but that simple analysis hides the detail. High concentrations of endemic species (n ≥ 10) are: Agrostis 10 spp. (Agrostideae), Chionochloa 22 spp. (Danthonieae, c. 90% of the genus), Lachnagrostis 10 spp. (Agrostideae, c. 50% of the genus), Poa 37 spp. (Poeae), Rytidosperma 15 spp. (Danthonieae). This is a small assemblage; the high frequencies are in cosmopolitan genera such as Poa and Agrostis, and in austral danthonioids.

At lower levels of endemism, 4-9 species per genus, there are: Cortaderia 5 spp. (Danthonieae), Deschampsia 4 spp. (Agrostideae), Deyeuxia 4 spp. (Agrostideae), Elymus 7 spp. (Hordeeae), Festuca 9 spp. (Poeae), Hierochloe 6 spp. (Agrostideae), Trisetum 9 spp. (Agrostideae). Apart from the preponderance of agrostidoids, 23 of 45 taxa, this assemblage is undistinguished; all are cosmopolitan genera except austral Cortaderia and Deyeuxia.

Single endemic species are few: Achnatherum (Stipeae), Australopyrum (Hordeeae), Dichelachne (Agrostideae), Imperata (Andropogoneae). Of these Australopyrum is the only known endemic pooid diploid in New Zealand (2n = 14), and Achnatherum petriei is one species in a genus of c. 500 spp.

Calcicoles Molloy (1994) identified the endemic grasses that are specialists of limestone or detrital marble habitats. This list now includes: Australopyrum calcis, Chionochloa flavescens subsp. lupeola, C. flavicans forma temata, C. spiralis, Dichelachne lautumia, Elymus sacandros, Festuca deflexa, F. luciarum, F. multinodis, Poa acicularifolia, P. spania, P. sudicola, P. xenica, Simplicia laxa. By Molloy's criteria Poa schistacea and Rytidosperma tenue of southern South Island schists could also be included here. Four of these are in the Threatened Plants list below.

Three species of ultramafic soils may be mentioned here as there is no other particular place to list them; they are Chionochloa defracta, Festuca ultramafica and Trisetum serpentinum.

Conservation of Endemic Species

Among the many plants at conservation risk in New Zealand there are some grasses (de Lange et al. 1999). No grass is presumed now extinct but the class "Threatened" which includes 108 taxa whose survival is a matter of concern and priority contains several grass species. We use the classification of de Lange et al. (1999) for the details below.

Critically endangered (3 grasses among 24 taxa): Amphibromus fluitans (Agrostideae), Australopyrum calcis subsp. calcis (Hordeeae), Poa spania (Poeae).

Endangered (4 grasses among 31 taxa): Chionochloa flavicans forma temata (Danthonieae), Cortaderia turbaria (Danthonieae), Puccinellia raroflorens (Poeae), Simplicia laxa (Agrostideae).

Vulnerable (2 grasses among 52 taxa): Australopyrum calcis subsp. optatum (Hordeeae), Deschampsia cespitosa (Agrostideae).

Should the group "Declining" be added to the preceding, the only entry is Austrofestuca littoralis (Poeae).

Naturalised Grasses

The naturalised element exceeds the total of endemic and indigenous taxa (Table 1). Tribes with significantly high representation are Paniceae 42 spp., Agrostideae 36 spp., Poeae 31 spp.; in total they comprise 48% of the naturalised taxa. At a second level of high frequency are Hordeeae 18 spp., Bambuseae 16 spp., and Stipeae 14 spp. - tribes in which endemism is very low or nil except for Hordeeae. The most striking element here is the number of panicoids - seven species each of Panicum, Pennisetum, and Setaria and six in Digitaria and Paspalum; most of these have North Island distribution but many extend well into South Island. Indigenous panicoids, by contrast, number three including the coastal psammophilous Spinifex sericeus.

Most naturalised species in Stipeae are of Australian provenance especially as Austrostipa spp., but South American Nassella has contributed three species. Poa with 10 naturalised species has the largest number of naturalised Poeae, all European except two from Australia. Other taxa in Poeae are principally European. Agrostidoid genera with naturalised species total 14, but only in Agrostis 4 spp., Avena 5 spp., and Phalaris 5 spp. are the numbers relatively high. Bromus (Bromeae) represented by 15 naturalised species and five transients, is the genus with most naturalised taxa; most are Eurasian (7 spp.) and South American (5 spp.).

Weediness is not our chief concern, but some genera were destined to be weeds of arable land, e.g., Elytrigia repens, or weeds of pastoral land, e.g., species of Austrostipa, and Nassella, N. trichotoma in particular. Zizania and Spartina are notorious in aquatic systems. Species of Critesion (Hordeeae) are prone to damage livestock, meat and pelts, through their penetrating awns.

About 40% of the naturalised species are Eurasian taxa; some of them probably arrived here via Australia, our nearest neighbour, but just how many cannot be determined. From Australia itself 35 species are naturalised here, half of them as species of Austrostipa and Rytidosperma. The other southern continents, Africa and South America, make equal but smaller contributions though the genera are rarely species rich here except Bromus as already remarked, and five African species of Pennisetum.

Two naturalised species of Cortaderia (Danthonieae) are especially undesirable in plantation forests notably so in North Island and Nelson; one of them C. selloana, pampas grass, had earlier enjoyed a reputation for both shelter and fodder for cattle. In the popular days of pampas grass culture one selection, "MacLeans", was of a tall, vigorous, large white-plumed, female line; seeds were never set because no pollen bearing plants were planted with them. Later commercial releases of pampas grass comprised both female and pollen-producing plants like the natural populations at some North Island sites; seed set was inevitable and the consequences are evident.

Surveillance Grasses

Some grasses are considered potentially so ecologically dangerous that they have been designated Surveillance Grasses. Under the Biosecurity Act 1993 and its amendments, the sale, propagation, and distribution of particular grasses is prohibited. Species such as Phragmites australis, common reed, Zizania latifolia, Manchurian wild rice, Pennisetum macrourum, African feather grass, and the two naturalised species of Cortaderia, the pampas grasses C. selloana and C. jubata, are among 10 or more subject to Plant Pest Management Strategies under that Act. Currently the Surveillance List contains reference to species of Stipa; in this Flora there are no species included in Stipa sens. strict. and species formerly included there are now referred to Achnatherum, Austrostipa, and Nassella.

Johnson grass, Sorghum halepense (Andropogoneae), is a weed of national concern, classified under the Biosecurity Act as a Notifiable Organism, one that must be reported to the Ministry of Agriculture and Forests. This species has the highest ranking for biological success and weediness among the grasses discussed in Esler et al. (1993).

4. Chromosome Numbers

In the list below (p. 1-9) prepared by M. I. Dawson, Landcare Research, Lincoln, there are chromosome counts for 88 species among 21 genera; most of these (c. 50%) are in polytypic Poa and Chionochloa. Other counts are in genera where intensive work had been done, e.g., Elymus, Festuca, Cortaderia. By our estimate the chromosome numbers are known for c. 50% of endemic and indigenous species. The area of greatest deficiency lies in Agrostideae where no chromosome numbers are known for Deyeuxia or Lachnagrostis. Few numbers are recorded for New Zealand Rytidosperma (Danthonieae), by comparison with the detail and the wide range of levels of infraspecific polyploidy described in Australia (Abele 1959; Brock and Brown 1961).

Two significant features emerge from the data in Dawson's list - endemic Australopyrum calcis (Hordeeae) is diploid (2n = 14) as are its Australian congeners, and dioecious Spinifex sericeus (Paniceae), shared with eastern Australia is also diploid (2n = 18). All other taxa are polyploids. In subantarctic Poa litorosa 2n = c. 263-265 (Hair and Beuzenberg 1961) or 2n = c. 266 (Hair 1968); in Festuca contracta 2n = c. 170 (Moore 1960); no other chromosome number of New Zealand grasses approaches these levels of ploidy; in both Poa chathamica and P. cockayneana 2n = 112 (16x) and this level is not reached elsewhere in New Zealand grasses either. The Poa count of 2n = c. 265 is the highest chromosome number recorded in any grass in the world.

Where there is little cytological information, e.g., Dichelachne, a count in a second species is confirmatory, and in Microlaena, cytologically unknown until recently, the four counts are coincident (2n = 48). For the endemic genus, Zotovia (Ehrharteae), the counts, even though one is approximate, are concordant with Microlaena and South African species of Ehrharta. The count in endemic Anemanthele lessoniana is inexact, but approximates to an expectation in Stipeae where x = 11. Chromosome numbers 2n = 28 are of no particular assistance in assessing the true relationships of Simplicia, a ditypic endemic genus.

When Hair (1966) discussed the overall frequency of polyploidy in New Zealand grasses, he used much of his data which were to be published later; he could not have known about diploid Australopyrum calcis, nor that Chionochloa would not be a dibasic genus with x = 6 and x = 7, but exclusively based on x = 7. For the rest his assessment holds true - 98% of the grasses are polyploid - but there is no way of knowing which among them are palaeopolyploids.

5. Grasses and Animal Health

Pastures in New Zealand are of diverse structure and their products are a substantial part of New Zealand's GDP. Some of the grasses used in these pastures may cause seasonal and acute disorders in livestock; these were discussed in Connor (1977). But since that time advances in animal husbandry have replaced many of the data there. Ryegrass staggers, a neuromuscular disease, is now known to be caused by toxins produced by the fungal endophyte Neotyphodium lolii (Fletcher and Harvey 1981). The main toxin is the tremorgen lolitrem B (Gallagher et al. 1984). Ryegrass, Lolium perenne, is thus the host of the disease-producing organism. A full set of papers on ryegrass staggers and on the associated fescue toxicosis is in Neotyphodium/Grass Interactions (1997 Bacon and Hill, Eds). Fescue toxicosis, one syndrome known locally as fescue foot, has been associated in New Zealand with grazing of Festuca arundinacea since the 19th century. The alkaloid ergovaline, synthesised by the endophyte N. coenophialum, is a vasorestrictor and the chief associate of the syndrome in animals; ergovaline is also synthesised by ryegrass.

Neotyphodium was isolated from indigenous Echinopogon ovatus and other Australian species of Echinopogon; it is a fungus which is serologically related to Neotyphodium lolii and other species of Epichloe and Neotyphodium which cause staggers in livestock. This is the first time an endophytic species of Neotyphodium has been identified in grasses native to New Zealand and Australia (Miles et al. 1998).

Although the name "phalaris staggers" is thought to be an inaccurate terminology for the suite of disorders of livestock caused by alkaloids in Phalaris aquatica, phalaris staggering is one syndrome which also results from grazing P. arundinacea, reed canary grass.

Facial eczema, an important photosensitivity disease of livestock, is caused by toxins produced by the fungus Pithomyces chartarum, but photosensitisation has been caused here by the grasses Japanese millet, Echinochloa esculenta and Panicum miliaceum, broomcorn millet.

The first grass shown to produce hydrocyanic acid was Cortaderia selloana, pampas grass. All species of Cortaderia in New Zealand, endemic and naturalised, are cyanophoric (Tjon Sie Fat 1979), but none has any deleterious effect on livestock. The glycoside there is triglochinin. In Glyceria maxima, reed sweet grass, hydrocyanic acid concentrations were high enough to poison cattle, just as in Sudax, a commercial hybrid involving Sorghum sudanense, Sudan grass.

In The Poisonous Plants in New Zealand (Connor 1977) oat, Avena sativa, is included because it is an accumulator of nitrates even to very high levels. These nitrates are reduced to toxic nitrites in the rumen of sheep and cattle, leading consequently to methaemoglobinaemia. Oats, like ryegrass and tall fescue, are vehicles for the accumulation and transport of toxins but not their syntheses.

6. Flora of Australia, Poaceae

Flora of Australia Vol. 43 and Vol. 44, Poaceae, are well advanced in the editing process and expected to start appearing in 2000 (Orchard 1999, and pers. comm.). The coincidence of publication of grass floras on both sides of the Tasman Sea should evoke some excitement. Their grass flora, much larger than ours, may contain some conflicts in treatment but there will be many correspondences, and these may allow comparisons to be made that were formerly impossible, thus pleasing Funk (1993).

We take this opportunity to congratulate Australian agrostographers on the completion of a task three times larger than ours.

7. De Auctoribus

This, the last volume of the Flora of New Zealand series, unlike Buchanan's Indigenous Grasses is unheralded by Parliamentary discussion or entitled "By command", but should bear the annotation "Published in the Fulfilment of Promises." Those promises were never originally made by either author, but became accepted by them when external circumstances altered. Both of us have worked on this Flora in our retirement from official positions at the then Botany Division, Department of Scientific and Industrial Research, Connor since late 1982 and Edgar since late 1988. We have, between us, published fifteen revisions involving 18 genera (p. xxxii), all of which were the necessary precursors to this modern Flora, and none of them gives the appearance of ease of completion. V. D. Zotov had earlier treated Simplicia, Zoysia, and Hierochloe; these were re-examined and resulted in some adjustments reported here. Deschampsia needed attention, too, especially the question of typification of seven taxa.

Is there a millennial messianic message in this liber peroptatus? It must be that the next grass to become naturalised in New Zealand is not yet established. The latest on that list is Achnatherum caudatum whose arrival from South America has been awaited by one of us (H.E.C.) for more than 40 years; we record it here on p. 65.

Although the authorship of this Flora is not that planned in the 1950s, Elizabeth Edgar, the senior author has spent her whole professional life writing floras, or preparing revisions of monocotyledonous Glumiflorae in preparation for floras both in New Zealand and Australia. Her associate author, Henry Connor, has no such experience at flora writing but is more practised at grasses. Despite this broad experience, there is still what Hilaire Belloc in his pilgrimage book The Path to Rome (1902) called the "Difficulty of Ending". We eschew gratuitous exhortations on the benefits of cladistic analyses or the singular results of molecular biology; unsolicited advice of this sort is all too common, and is invaluable. Belloc's preface - "Praise of this Book" - is silent on the point but in "Introductory Rambling" he gives a clear method for an ending. One, he wrote, ". . . is to rummage about one's manuscripts till one has found a bit of Fine Writing (no matter upon what subject), . . . to introduce a row of asterisks, and then paste on to the paper below these the piece of Fine Writing one has found."

* * * * * * * * * * * * * * * * * * * *

What is the meaning of the differences that separate the Gramineae so delicately, yet so definitely, from any other order, and that so prevail that a grass remains a grass, however freely the type may vary? To attribute these differences to genic constitution is an 'explanation' of a merely descriptive kind; it enables us, indeed, to assign a place to them in the mental framework which we impose upon reality, but in so doing we have shelved, not solved, the problem. The mystery abides. (Agnes Arber The Gramineae, 1934).

Elizabeth Edgar

Henry Eamonn Connor

Christchurch

SS Thomas More and John Fisher

June 1999

REFERENCES

Abele, K. 1959: Cytological studies in the genus Danthonia. Trans. Roy. Soc. S. Aust. 82: 163-173.

Allan, H. H. 1936: An introduction to the grasses of New Zealand. N.Z. DSIR Bull. 49.

Allan, H. H. 1940: A handbook of the naturalized flora of New Zealand. N.Z. DSIR Bull. 83.

Bacon, C. W. and N. S. Hill (Eds). 1997: Neotyphodium/Grass Interactions. New York, Plenum Press.

Belloc, H. 1902: The Path to Rome. London, Allen and Unwin.

Brock, R. D. and J. A. M. Brown. 1961: Cytotaxonomy of Australian Danthonia. Aust. J. Bot. 9: 69-91.

Buchanan, J. 1878: Indigenous Grasses of New Zealand. t. 1-20. Wellington, James Hughes.

Buchanan, J. 1879: idem, t. 21-40.

Buchanan, J. 1880: idem, t. 41-58, 17(2), 26(2), 36(2).

Buchanan, J. 1880: idem, consolidated.

Clayton, W. D. and S. A. Renvoize. 1986: Genera Graminum; Grasses of the World. London, H.M.S.O.

Connor, H. E. 1977: The Poisonous Plants in New Zealand. Wellington, Government Printer.

Craig, G. F. 1984: Reinstatement of Spinifex sericeus R.Br. and hybrid status of S. alterniflorus Nees (Poaceae). Nuytsia 5: 67-74.

de Lange, P. J. et al. 1999: Threatened and uncommon plants of New Zealand. N.Z. J. Bot. 37: 603-628.

Doyle, J. D. et al. 1992: Chloroplast DNA inversions and the origin of the grass family (Poaceae). Proc. Nat. Acad. Sci. 89: 7722-7726.

Esler, A. E., Liefting, L. W. and P. D. Champion. 1993: Biological Success and Weediness of the Noxious Plants of New Zealand. Ministry of Agriculture and Fisheries.

Espie, P. R., Connor, H. E. and I. J. McCracken. 1992: Leaf-blade crimping in grasses: a new measure of growth. Experientia 48: 91-94.

Fletcher, L. R. and I. C. Harvey. 1981: An association of a Lolium endophyte with ryegrass staggers. N.Z. Vet. J. 29: 185-186.

Fuentes, C. 1988: How I wrote one of my Books. London, Andre Deutsch.

Funk, V. A. 1993: Uses and misuses of floras. Taxon 42: 761-772.

Gallagher, R. T. et al. 1984: Tremogenic neurotoxins from perennial ryegrass causing ryegrass staggers disorder of livestock: structure elucidation of lolitrem B. J. Chem. Soc. Chem. Comm. 1984: 614-616.

Hair, J. B. 1966: Biosystematics of the New Zealand flora, 1945-1964. N.Z. J. Bot. 4: 559-595.

Hair, J. B. 1968: Contributions to a chromosome atlas of the New Zealand flora - 12. Poa (Gramineae). N.Z. J. Bot. 6: 267-276.

Hair, J. B. and E. J. Beuzenberg. 1961: High polyploidy in New Zealand Poa. Nature 189: 160.

Hawksworth, D. L. 1992: The need for a more effective biological nomenclature for the 21st century. Bot. J. Linn. Soc. 109: 543-567.

Hooker, J. D. 1864: Handbook of the New Zealand Flora. Part 1. London, Reeves.

Hooker, J. D. 1867: idem. Part II.

Hsiao, C. et al. 1999: A molecular phylogeny of the grass family (Poaceae) based on the sequence of nuclear ribosomal DNA (ITS). Aust. Syst. Bot. 11: 667-688.

Le Roux, L. G. and E. A. Kellogg. 1999: Floral developments and the formation of unisexual spikelets in the Andropogoneae (Poaceae). Amer. J. Bot. 86: 354-366.

Linder, H. P. and G. A. Verboom. 1996. Generic limits in the Rytidosperma (Danthonieae, Poaceae) complex. Telopea 6: 597-626.

Miles, C. O. et al. 1998: Endophytic fungi in indigenous Australasian grasses associated with toxicity to livestock. App. Envir. Microbiol. 64: 601-606.

Molloy, B. P. J. 1994: Observations on the ecology and conservation of Australopyrum calcis (Triticeae: Gramineae) in New Zealand. N.Z. J. Bot. 32: 37-51.

Moore, D. M. 1960: Chromosome numbers of flowering plants from Macquarie Island. Bot. Not. 113: 185-191.

Orchard, A. E. 1999: ABRS Report. Aust. Syst. Bot. Soc. Newsletter 98: 4-5.

Soreng, R. J. and J. I. Davis. 1998: Phytogenetics and character evaluation in the grass family (Poaceae): simultaneous analyses of morphological and chloroplast DNA restriction site character sets. Bot. Rev. 64: 1-88.

Tjon Sie Fat, L. 1979: Cyanogenesis in some grasses. IV. The genus Cortaderia. Proc. Kon. Nederl. Akad. Wetensch. C. 82: 165-170.

Willemse, L. P. M. 1982: A discussion of the Ehrharteae (Gramineae) with special reference to the Malesian taxa formerly included in Microlaena. Blumea 28: 181-194.