Flora of New Zealand Series

INTRODUCTION

1. History of exploration and documentation of the flora

First European visits to New Zealand

James Cook served as commander of three expeditions to the South Seas, with the first two most relevant to New Zealand botany. The first expedition (1768–1771) was commissioned to observe the 1769 transit of the planet Venus across the disc of the sun. Cook was commander of the Endeavour, and biological observations were the responsibility of Mr. Joseph Banks and Dr. Daniel Solander, a student of Linnaeus. The botanists collected plants at Poverty Bay (near the present city of Gisborne) and other localities on North Island, as well as Queen Charlotte Sound in the northern portion of the South Island (see comments in Cheeseman [1906, 1925] and Cockayne [1919]). Apparently no hepatics were published, but Hymenophyton flabellatum was collected and misidentified as a Trichomanes species (duplicate at WELT!). The second and greatest voyage of Cook (1772–1775) was on the vessels Resolution and Adventure, with John Reinhold Forster and his son, John George Adam Forster, serving as naturalists. The noted botanist Anders Sparrman, also a student of Linnaeus, joined the expedition in Cape Town. The observations made during this expedition "revolutionised knowledge of the southern hemisphere" (Godley, 1965, p. 146). New Zealand localities visited were Dusky Sound and Queen Charlotte Sound (see remarks in Cheeseman, 1906, 1925). Forster and Forster (1776) in their Characteres Generum Plantarum listed only one genus of cryptogams, Aytonia, based on A. rupestris (≡Plagiochasma rupestre), but this was collected in the Madeira Islands (Schuster, 1992b). Of interest is the description of the new genus and species, Monoclea forsteri Hook. in Hooker (1818–1820). The protologue includes "Forster’s MSS., Hab. Insulae australes. D. G. Forster (in Herb. Lambert.)." Given this information, the distribution of Monoclea forsteri and the route of the expedition, one of the Forsters probably collected the first New Zealand hepatic. Note here, incidentally, that Grolle and So (2003) concluded that the type specimen of Jungermannia violacea Ach. (≡Metzgeria violacea), previously regarded as having been collected at Dusky Bay, New Zealand, on this expedition, actually was gathered in Tierra del Fuego on the same expedition (see also Hodgson, 1961). Cook’s third and last voyage was with the Resolution and Discovery (1776–1780); William Anderson served as surgeon on the Resolution and made natural history collections. The expedition visited Queen Charlotte Sound, and although botanical collections were made at this site, liverworts apparently were not collected.

The Scottish surgeon-botanist Archibald Menzies (1754–1842) sailed on board H.M.S. Discovery as part of Vancouver’s round-the-world voyage (1791–1795) and collected a number of liverworts as well as mosses and ferns at Dusky Sound during the period 2–22 November 1791 (for an account of this voyage see the volumes of Lamb, 1984). The liverwort collections, which are the first in New Zealand that are documented, were published in William Hooker’s Musci Exotici (1818–1820). The Musci Exotici is noteworthy not only because of the introduction to science of such striking and conspicuous New Zealand hepatics as Schistochila nobilis, S. appendiculata, S. glaucescens, Lepicolea scolopendra, Temnoma pulchellum and several others, but also because of the handsome, accurate colored plates that remain useful nearly two centuries later. He is commemorated by Jungermannia menziesii Hook. (≡Lepidogyna menziesii) of southern South America. Menzies published little himself, but the impact of his hepatic collections is highly significant. For information on his botanical activities in New Zealand see Godley (1960), Galloway (1985, 1991, 1995) and Galloway and Groves (1987).

French, American and British natural history expeditions of the early nineteenth century

During his first voyage to the Pacific (1822–1825), Jules Sébastien César Dumont d’Urville (1790–1842) visited New Zealand on board the Coquille under the command of L. I. Duperrey. Lieutenant d’Urville, an ardent botanical collector, was assisted by René Primivère Lesson, who served as naval doctor and naturalist. The expedition remained at Bay of Islands for nearly two weeks. The second voyage to the Pacific by Dumont d’Urville was aboard the same vessel, which was renamed the Astrolabe (1827–1829), under his command, and Pierre Adolphe Lesson, brother of René, accompanied the voyage as botanist. In New Zealand botanical collections were made at Tasman Bay and the "French Pass," which separates d’Urville Island and the mainland, and Tolaga Bay and Auckland Harbour on North Island. Achille Richard (1832) published a botanical account of this expedition and included the species collected on the earlier d’Urville expedition as well as most of the collections made by Forster on Cook’s second voyage (Cheeseman, 1906, 1925). However, the account of the Hepaticae in Richard (1832) included enumeration of 28 species, and of these, all except four were cited as having been collected by Menzies. No collector is cited for the remaining four species; whether they were collected on either the first or second expedition of d’Urville, or by the Forsters, remains unknown. d’Urville’s third voyage to the Pacific is the most important for our purposes and involved two corvettes, the Astrolabe and the Zélée, during the years 1837–1840. d’Urville served as commander of the Astrolabe, and botanical collecting was the charge of Jacques Hombron, the senior surgeon on the Astrolabe. The Zélée was commanded by Charles Jacquinot; Honoré Jacquinot served as junior surgeon on this ship and was responsible for the collection of plants. The expedition visited New Zealand and (12–20 March 1840) Auckland Island, the latter botanized for the first time. The hepatics were studied by Camille Montagne, who published a preliminary account (Montagne, 1843) that included the introductions of Jungermannia schismoides Mont. (≡Anastrophyllum schismoides) and Madotheca elegantula Mont. (≡Porella elegantula), both described from the Auckland Islands. The full account appeared in Montagne (1845a) and, together with Montagne (1843), included a number of types based primarily on the collections of Hombron and Jacquinot.

The second United States Exploring Expedition (1838–1842) was prompted by a desire to obtain information on the South Pacific, an area that was rapidly becoming of interest to American traders and whalers. Commanded by Lieutenant Charles Wilkes of the United States Navy, it included a contingent of scientists among whom was botanist William Rich (1800–1864); for further comments see Godley (1965). The expedition briefly visited New Zealand and also Auckland Island in 1840. Any hepatics collected in New Zealand were apparently unpublished; Sullivant (1850) published on the hepatics collected at Orange Harbour, Tierra del Fuego.

Of the expeditions to Austral areas during this time period, the most important was that of the British voyages of Erebus and Terror (1839–1843). The period of this exploration is also notable, since it coincides with the beginnings of systematic colonization of New Zealand by Europeans in 1840. Botanists on the expedition were Joseph Dalton Hooker (1817–1911, son of William Jackson Hooker), who served as assistant surgeon on the Erebus (James Clark Ross was captain), and David Lyall, assistant surgeon on the Terror (under F. R. M. Crozier, commander). Hooker made the first extensive collection of plants from Auckland Island on 20 November–12 December 1840 (Hombron having collected there earlier that year) and was the first to make botanical collections on Campbell Island, which he visited during the period of 13–17 December 1840 (Godley, 1965). The vessels spent the period of 18 August–17 November 1841 at the Bay of Islands, and during the period Hooker actively collected plants, often in the company of William Colenso and Dr. Andrew Sinclair (see below). Many of the liverwort collections from Auckland and Campbell Islands and New Zealand are types.

In the autumn of 1843 Hooker invited Thomas Taylor (1786–1848; cf. Sayre, 1983, 1987) to examine some or all of the cryptogams that were collected as part of the expedition. Taylor replied (Sayre, 1987, p. 423), "‘I should be very proud to be associated once again in any scientific exertion with the name of Hooker; I will therefore most cheerfully undertake to examine, and if you please it, give specific characters and short descriptions of any Cryptogamic plants you may commit to me’ (TT to JH, 20 November 1843, K)." Taylor received the first specimens from Hooker in November 1843. Sayre (1987, p. 423) provided valuable insights in the way of Taylor’s involvement with the hepatic publications. "The first question was whether Taylor should use the new generic names proposed by Nees von Esenbeck. ‘While it would be desirable to keep pace with modern improvements on the one hand, on the other, the improvements are questionable, the generic characters being scarcely sufficient or strictly drawn from the parts necessary to fructification. In other words, the difficulty of subdividing Jungermannia, with few exceptions, remains as great as it was 30 years ago’ (TT to JH, 20 November 1843, K). Taylor recommended that all leafy liverworts should be called Jungermannia, but arranged in sections headed with Nees’s generic names. This system was followed in the series of four papers published in 1844 and (the Supplement) 1845 to secure the priority of the new names, and in Hooker’s Botany of the Antarctic Voyage in 1845 and 1847.

"He also had to take account of Montagne’s Plantes cellulaires in Dumont d’Urville’s Voyage au Pôle sud (published in parts in 1843 and 1844). Hooker had copied out the appropriate parts of Montagne for Taylor. Taylor had sent many Antarctic hepatics to Montagne and received in return seven of the thirteen species that had already been published. Taylor wrote to Hooker, ‘I am delighted that we shall, so far, be spared the shame of publishing as new what Montagne had already given.... This is a noble Frenchman!’ (TT to JH, 7 August 1844, K)."

The hepatic collections from Auckland and Campbell Islands and New Zealand were worked up by Thomas Taylor, and details first appeared in the London Journal of Botany (Hooker and Taylor, 1844a, 1844b, 1845). It is notable to call attention again here that Taylor received the first specimens from Hooker in November 1843, and, by 1 February 1845 the above-mentioned three papers were published in the Journal (see Bibliography). Taylor alone, or, more often, with J. D. Hooker, published as new 183 of the 260 taxa included in the two articles and the supplement (comprising Hooker and Taylor, 1844a, 1844b, 1845 [Sayre, 1983]). A few additions to the New Zealand flora appeared in Taylor (1846); these are attributed to "Hook. f. & Tayl." In Hooker and Taylor (1845) specimens of Cunningham and some from the herbarium of Robert Greville were added to those of Hooker. Hooker’s involvement and perspective may be found in the preface to his account of hepatics in the Handbook (Hooker, 1864–1867, p. 498), where he remarked that the hepatics "were placed, at the late Dr. Taylor’s request, in his [Taylor’s] hands for immediate publication, and the descriptions of many appeared, in 1844, in Hooker’s ‘London Journal of Botany,’ and again in the ‘Botany of the Antarctic Voyage.’ At that time the valuable ‘Synopsis Hepaticarum’ of Gottsche, Lindenberg and Nees von Esenbeck was not published (it was not completed till 1847), and the difficulty of ascertaining and defining species was very great indeed. Owing to this cause and to Dr. Taylor’s precipitancy in publication, much confusion crept into his work; the same plant appearing under several generic names, and some descriptions answering to subsequently cancelled species, having been also published as good species." Hooker continued but, qualifying his comments somewhat, remarked (p. 499) that "whilst feeling it is my duty to make these facts clearly known, I must add, that those who study this extensive Order of plants by the aid of modern works, can have no conception of the difficulties which Dr. Taylor had to overcome, twenty-two years ago, in classifying and naming 300 or 400 Hepaticae with which I supplied him, and of which not fifty had been published; imperfect and hasty as his work was, it showed great skill, no little sagacity, and indomitable perseverance as a microscopic investigator."

The papers published in the Journal appeared prior to publication of the liverwort sections in the Flora Antarctica— the portion covering Auckland and Campbell Islands appeared in Taylor and Hooker (1845), which appeared on 1 May 1845, a few months after publication of Hooker and Taylor (1845), which appeared on 1 February. Sayre (1983, p. 467) provided insight into the role of Hooker in joint publication with Taylor, with Hooker stating that "‘after a thorough examination by Dr Taylor, the specimens have twice passed under my own eyes, preparatory to the completion of the descriptions and the drawings’ (JH’s footnote, p. 114). Actually, the descriptions were completely rewritten from their first form in the Journal of Botany. Taylor commented: ‘I hardly recognized the brats in their Sunday clothes’ (to JH, 20 April 1845, K). A number of species had been added, included 16 new, but almost all the new taxa of the Journal paper have been retained." Taylor died before the publication of the Flora Novae-Zelandiae volume, and William Mitten (1819–1906) prepared the hepatics for the volume covering New Zealand (Mitten, 1854–1855) and Tasmania (Mitten, 1854–1855). Some of the species enumerated in the Flora Antarctica are illustrated with hand-colored illustrations.

Little information relevant to the specific localities of liverwort collections made on the Erebus and Terror expedition may be obtained from the publications of the hepatics in the London Journal of Botany or in the Flora Antarctica (see below). For example, the only specific locality cited for the 37 Hooker species enumerated in the Hepaticae section published in Hooker’s Flora Novae-Zelandiae (Mitten, 1854–1855) is Bay of Islands.

Collectors in the early New Zealand British colony

The Flora Antarctica included not only the specimens of Joseph Hooker, but also a number of collections that were sent to Kew by various New Zealand correspondents of Joseph Hooker and his father, William Hooker (see Galloway, 1998, 2004). These citations provide insight into the identity of the collectors and the localities visited by those individuals who were engaged in botanical pursuits that included the collection of liverworts in the early New Zealand colony. The collectors are:

Lieutenant-Colonel Daniel Bolton—6 species cited, all from Auckland. Bolton collected on the Auckland Islands in 1850 (Godley, 1970, p. 49; 2003); all "Auckland" records in Mitten (1874–1855) are from the vicinity of Auckland city.

William Colenso—collections of 129 species cited, all from North Island (see below).

Allan Cunningham—5 species cited, all from Bay of Islands where a specific locality is mentioned. It should be noted that Allan Cunningham (1791–1839), Government Botanist and Superintendent of the Sydney Botanical Gardens, traveled to New Zealand in 1826 and spent over five months collecting. His brother, Richard Cunningham (1793–1835), traveled to New Zealand in 1833. Both collected in the Bay of Islands, Whangaroa and Hokianga districts.

John Edgerley—1 species from the North Island.

John Jolliffe—5 species cited, 4 from Hokianga, 1 cited only as North Island.

Dr. David Lyall—60 species, with the following localities mentioned: Auckland and Port Nicholson in the North Island; Port Cooper, Canterbury Plains, Chalky Bay [Chalky Inlet], Port Preservation, Thomson’s Sound, Bligh’s Sound and Milford Sound in the South Island; and Port William on Stewart Island. David Lyall (1817–1895) served as surgeon-naturalist on the Acheron, under the command of Captain Stokes, which surveyed the coast of New Zealand, particularly the western and southwestern sectors, in 1847–1851; Lyall gave particular attention to lower plants. Localities visited during that expedition included Milford Sound, Chalky Inlet, Dusky Bay, Preservation Inlet and both shores of Foveaux Strait. He is commemorated by Isotachis lyallii Mitt.

Archibald Menzies—31 species from Dusky Bay made during the Vancouver expedition (see above).

Etienne Raoul—3 species cited, all from Akaroa, Banks Peninsula, South Island (see below).

William Rich—1 collection from a "small island off Auckland" (Mitten, 1854–1855, p. 127); see above under United States Exploring Expedition.

Dr. Andrew Sinclair—34 species cited, all from North Island, including Auckland, Bay of Islands, East Cape, Waikehi and Mt. Egmont (Mt. Taranaki). Sinclair (1794–1861) was colonial secretary of the British administration of New Zealand in 1844–1856 and actively pursued plant collecting, particularly in retirement; he drowned in an attempt to ford the Rangitata River on an expedition. Sinclair initially collected in the North Island, and later in Nelson as well as inland Canterbury, where he participated in botanical and geological surveys. Joseph Hooker considered Sinclair second only to Colenso as a botanical explorer (Cockayne, 1919, p. 10); Sinclair sent his materials to Kew.

Sowerby—1 species with no specific locality provided.

W. Stephenson (?1810–ca. 1863)—23 species cited; most have a specific locality and all of these are from the Wellington area.

Hooker played an important role in the development of New Zealand botany during these years "for he provided a direct and personal link between the few isolated plant collectors in New Zealand, and the wider botanical and scientific world of which the Royal Botanic Gardens at Kew was such a centre and focus" (Galloway, 1998, p. 35).

Volume one of Hooker’s Flora Novae-Zelandiae, containing the flowering plants, appeared in 1853, and the second volume, covering the cryptogams, appeared in 1854 and 1855. The decade that followed saw considerable botanical activity in the New Zealand colony. The period also had rapid settlement of the South Island, which was followed by exploration and botanical investigations of that island. A need for a concise account of the flora of the colony emerged, and Charles Knight (see below), a long-time correspondent with Hooker, obtained a government commission for Joseph Hooker to prepare such a publication. Hooker published the Handbook of the New Zealand Flora in two parts; the flowering plants and ferns appeared in 1864 and the concluding part, containing Hepaticae, mosses, lichens, fungi and algae, appeared in 1867. Hooker initially intended the Handbook to cover only the flowering plants, and in preparation of the second part, he relied on several cryptogamists for preparation of specific groups. Hooker (1864–1867), in the preface of the hepatic account, remarked (p. 499) that "In the present work I have followed in Mr. Mitten’s footsteps, he having drawn up all the descriptions for the ‘Flora Novae-Zelandiae,’ with great care; and to him, too, I am indebted for naming the species of this Order received since the publication of that work; he has also detected many of Dr. Taylor’s errors, and rediscovered, in later collections, many of his lost species." The treatment of liverworts was the first and one of only two accounts that provided a more comprehensive treatment of the hepatics of New Zealand. The publication of the Handbook is an important milestone in New Zealand botany, and in this connection Galloway (1998, p. 31) remarked that "for a small colonial government to commission one of the scientific world’s most illustrious botanists to devote at least three years of his life to such a task was no mean achievement." Moreover, the Handbook "put New Zealand botany on the map, both nationally and internationally, in a manner that was accessible and interesting to a wide audience who were clamorous for information on the natural history of the colonies" (Galloway, 1998, p. 32).

During a similar time frame that Hooker visited New Zealand, Etienne Fiacre Louis Raoul (1815–1852) served as surgeon on the French corvette L’Aube (later replaced by L’Allier) and was stationed at Akaroa on the Banks Peninsula (1840–1843). Raoul was the first botanist to collect on the eastern side of the South Island and sent a complete set of specimens to Kew where they were made available to Thomas Taylor. As a result, Raoul’s new finds from Akaroa were published by Hooker and Taylor or Hooker and Levier, and listed in the Choix de Plantes de la Nouvelle Zélande (Raoul, 1846). This publication included a list of New Zealand plant species known to that date, including 10 liverwort and one hornwort species (Macmillan, 1998).

Mid-nineteenth-century synoptic work

During the mid-nineteenth century several major synoptic works appeared. Johann Georg Christian Lehmann (1792–1860; Jungermannia lehmanniana Lindenb. ≡Schistochila lehmanniana) published his Novarum et Minus Cognitarum Stirpium Pugillus series (1828–1857), and five species from New Zealand were described as new, four based on Menzies collections from Dusky Bay (Anthoceros denticulatus, A. giganteus, Jungermannia anisostoma, J. novae-zeelandiae and J. praenitens). Gottsche, Lindenberg and Nees (1844–1847) published a compilation of the hepatics known to that date in their Synopsis Hepaticarum and included numerous references to New Zealand hepatics, many in newly introduced genera that segregated a number of species from the large portmanteau genus Jungermannia (see above comment on Taylor, who worked during the same period).

A German expedition to observe the transit of Venus visited the Auckland Islands, arriving on the Alexandrine on 15 October 1874 and remaining until 6 March 1875 (Godley, 1965). Hepatics were collected by Hermann Krone, who served as photographer on the expedition; at least some of Krone’s hepatic collections were studied by Stephani.

New Zealand–based collectors and botanists of the late nineteenth century

During the late nineteenth century several advancements were made relevant to New Zealand liverworts. Foremost among contributors is the Reverend William Colenso, who "was the last, and the greatest, of the missionary explorers" (Temple, 1985, p. 11). Colenso traveled nearly the whole of the North Island, all on foot, and was the first European to penetrate inland to any appreciable extent. For example, he was the first European to reach Lake Waikaremoana and the first to cross the Ruahine Range. Colenso published 59 botanical papers in the Transactions of the New Zealand Institute, of which 13 included hepatics. His papers that included liverworts spanned the years 1881 to 1893 and contained 217 new species. All but about 25 of these names are synonyms, but Cheeseman (1906, p. xxv; 1925, p. xxvi) provided an interesting perspective, "that [this] is a circumstance which must not detract from the recognition of his undoubted services to the botany of New Zealand." A complete list of Colenso’s papers that included hepatics are included in the Bibliography section of this volume; for his collections held at WELT see Hamlin (1971). Stephani (1892) described a number of new species collected by Colenso and provided a disposition of the species described by Colenso. Colenso is commemorated by several liverwort names, among them Schistochila colensoana Steph. (≡Pachyschistochila colensoana), Mastigobryum colensoanum Mitt. (≡Acromastigum colensoanum) and Lopholejeunea colensoi Steph. A few other individuals were cited in Colenso’s works as having gathered liverworts, among them Augustus Hamilton (1854–1913), director of the Colonial Museum in Wellington between 1903 and 1913 who collected Hepaticae at Okarito and the Hawke’s Bay District in the foothills of the Ruahine Range and places on the eastern side of North Island. Mr. Henry Hill (1849–1933), school inspector and mayor of Napier, collected in Hawke’s Bay, Gisborne, Urewera, Ruahine Range and Tongariro National Park between 1885 and 1898 and sent his specimens to Colenso.

Charles Knight (1808–1891), the New Zealand Government Auditor-General, published a series of papers on lichens, his main botanical interest, and a few on mosses during the period of 1860–1884 (Galloway, 1990, 1998). Knight also collected liverworts and is commemorated by several names, among them Marsupidium knightii Mitt. and Chiloscyphus knightii Steph. (≡Heteroscyphus knightii).

Dr. Lauder Lindsay (1829–1880), a British physician, collected in eastern Otago in 1861–1862. Data on liverworts were published in Lindsay (1867), and a list of 10 species of hepatics and one anthocerote appeared in Lindsay (1868), which presently is a notably rare publication.

Thomas Kirk (1828–1898; cf. MacLeod, 1993; Moore, 1973) arrived in New Zealand in 1863, residing for 10 years in Auckland before moving to Wellington, and for many years "held the position of leader of botanical thought in New Zealand" (Cockayne, 1919, p. 11). Kirk published many papers (nearly 150, mostly in the Transactions of the New Zealand Institute), chiefly on flowering plants, during the period of 1869 to 1899 (for a list see Allan, 1982). These included "The Forest Flora of New Zealand" (Kirk, 1889), a major contribution. His botanical excursions included both North Island and South Island. In 1890 Kirk visited several islands on board the Hinemoa : Auckland Islands (10–12 January 1890), Campbell Island (13–14 January 1890), Snares Islands and Antipodes Islands. During his botanical excursions Kirk collected a number of liverworts, many of which were sent to Stephani for study. Kirk’s collections from Great Barrier Island are particularly noteworthy. Kirk is commemorated by the following Stephani species: Frullania kirkii (≡Frullania patula), Jamesoniella kirkii, Lepidozia kirkii, Cephalozia kirkii (=Paracromastigum macrostipum) and Schistochila kirkiana Steph.

Other New Zealand botanists with a primary interest in flowering plants also collected liverworts. Among them was Thomas Frederick Cheeseman (1845–1923; cf. Cockayne, 1923), botanist, author of Manual of the New Zealand Flora (Cheeseman, 1906, 1925), director of the Auckland Institute and Museum for 50 years, and author of over 240 angiosperm names (Ewen Cameron, in litt.); he was commemorated by Lophocolea cheesemanii Steph. (=Chiloscyphus lentus). Others are Donald Petrie (1846–1925), who resided in Otago for over 20 years and was Chief Inspector of Schools for Otago and later held the same position in Auckland. While in Otago he traveled to the eastern, central and southern portions of Otago Province and was the first to provide a description of the flora and vegetation of Stewart Island. Leonard Cockayne (1855–1934; cf. Laing, 1936) published works during the period 1899–1934 and made important contributions to higher plants, including The Vegetation of New Zealand (Cockayne, 1921) and the popular work New Zealand Plants and Their Story (Cockayne, 1919). He collected liverworts on Stewart Island in 1908 for the Department of Lands and Survey (Cockayne, 1909a), and his collections were studied by Stephani. Cockayne is commemorated by Calycularia cockaynei Steph. (≡Allisonia cockaynei). Henry H. Travers (1844–1928) of Nelson visited the Chatham Islands twice, in 1864 to collect for Ferdinand von Mueller of Melbourne, and in 1870. Travers made a living as a collector of natural history material, selling live and pressed plants to horticulturists and museums in New Zealand and overseas. Some of his hepatic collections were cited by Stephani in his Species Hepaticarum (Lyon et al., 1971, p. 704).

John Buchanan (1819–1898; cf. Adams, 2002), naturalist and draughtsman, published a number of papers, mainly on flowering plants, during the period of 1870 to 1885 (for a list see Allan, 1982). Buchanan collected on mainland New Zealand and also sailed aboard the Stella, a New Zealand Government steamer, to Campbell and Auckland Islands in 1883. There remains some uncertainty as to the whereabouts of Buchanan’s collections. Adams (2002, p. 109) remarked that "until a search is undertaken, especially in the collections of Glasgow and Edinburgh, our knowledge must remain as fragmentary as the known Buchanan specimens that have survived in New Zealand, mainly in the herbarium of the Colonial Museum (Te Papa) and the Otago Museum. The latter collection has been deposited by the Trustees of the Otago Museum in the herbarium of Te Papa (WELT)." Some Buchanan collections came to Te Papa from the Otago Museum, as mentioned by Adams (2002). Lyon et al. (1971) noted that some bryophyte collections are at the University of Otago (OTA) on indefinite loan from the Otago Museum, that these are in good order, that some were named by J. D. Hooker and that the unnamed specimens were revised by K. W. Allison. John Steel (in litt.) is of the opinion that the specimens were transferred to the University of Otago in the early 1960s. Barry Sneddon of the herbarium at Te Papa commented (in litt.) that "there is a Buchanan guard book bearing the label ‘N.Z. Hepaticae’ on the front cover. This book contains approximately 85 pages of which ca. 63 have 1–4 liverwort species per page and the rest of the pages have no specimens. In spite of its title, the book does contain a few foreign liverworts (all[?] from Britain), but the great majority are from New Zealand. There is a wide range of genera represented, but the accompanying information is restricted to genus and species and (not always) locality (Otago & Wellington). There are no dates. Below one specimen Buchanan has written ‘ Plagiochila subsimilis n.s. Colenso’ as the only information. Te Papa also holds 3 boxes of his specimens labelled ‘Buchanan – Miscellaneous.’ ... One box has a folder on which someone (Nancy Adams, perhaps) has pencilled: ‘J. Buchanan ex O.M. / Bryophytes mounted/named no locality, in "Otago Witness" [a Newspaper] of Sep. 2 1887.’ This folder contains a mixture of liverworts and (preponderantly) mosses."

Franz Stephani (1842–1927; cf. Beauverd, 1927; Geissler, 1982) played a significant role in the advance and development of New Zealand hepaticology, since for a number of years many individuals in varied circumstances sent him New Zealand material. This resulted in many names described by him. Many of the new species appeared in Hedwigia, e.g., Blepharostoma corrugatum Steph. (≡Trichotemnoma corrugatum), and a great number were published in "Species Hepaticarum" (1898–1924).

Richard Helms (1842–1914; cf. Godley, 2001) was a self-taught naturalist of Greymouth who sent moss and liverwort specimens from the Paparoa Range and the vicinity of Greymouth to Cosmo Melville in England. These were forwarded to Henry Boswell and were later seen by Müller, who published a number of new mosses from this material, and by Stephani, who described Anthoceros helmsii, Lejeunea helmsii and Lophocolea helmsiana (≡Chiloscyphus helmsianus) from Helms’s collections.

T. W. N. Beckett (1839–1906) was an orchardist who immigrated to Christchurch in 1883 and who collected liverworts mainly on the Christchurch to Greymouth rail route, but also on Banks Peninsula, North and South Canterbury foothills, Hawke’s Bay and Kaikoura during the period of 1883–1904 (Godley, 1967; Lyon et al., 1971). Beckett was an important collector of mosses; his private herbarium, which is now housed at CHR, contains 15,000 moss specimens, and he exchanged material with leading overseas muscologists such as Brotherus (Fife, 1985). He sent liverwort specimens to Stephani and Levier. Five hundred hepatic specimens are in Beckett’s herbarium at CHR, and while most specimens are Beckett’s, there are also duplicates of collections made by Buchanan, Cheeseman, Cox, Hill, Kirk, Petrie and Rutland. Stephani named two species after Beckett, Mastigophora beckettiana Steph. (=Triandrophyllum subtrifidum) and Lepidozia beckettiana Steph. (=Telaranea praenitens).

European collectors in the late nineteenth century

Max Fleischer (1861–1930) collected liverworts in New Zealand in 1903 during his return to Germany from Java, where he gathered mosses for his Die Musci der Flora von Buitenzorg (Fleischer, 1904–1923). In New Zealand he visited Dunedin, Queenstown, Mt. Egmont, Rotorua and Auckland; his collections are widely distributed (see Schultze-Motel, 1982).

During this period several European collectors visited New Zealand. Dr. Henri Filhol (1843–1902) was surgeon, biologist and geologist on the vessel Vire as part of the 1874 French expedition to Campbell Island to observe the transit of Venus. The hepatics were reported by Kirk (1882), who lists Jungermannia monodon and Marchantia tabularis, and possibly by Stephani.

Dr. Sven Berggren (1837–1917; cf. Anon., 1917), a Swedish botanist from the University of Lund and an excellent draughtsman, visited New Zealand during the period of 1 January 1874 – April 1875. The expedition was encouraged and possibly suggested by Lauder Lindsay (Bagnall, 1970a). Bagnall (1970a, 1970b) enumerated portions of Berggren’s diary that are relevant to his travels in New Zealand, including notes on localities and dates. Berggren (1898) published a single account of his trip to New Zealand, the excellent "On New Zealand Hepaticae," which included a number of new species (e.g., Cesia cuspidata and C. stricta, both later transferred to Gymnomitrion), as well as excellent descriptions and illustrations. He is commemorated by the names Chloranthelia berggrenii (Herzog) R.M.Schust., Lepidolaena berggrenii E.A.Hodgs. and Pachyschistochila berggrenii J.J.Engel & R.M.Schust.

The German morphologist Karl I. E. Goebel (later Ritter von Goebel, 1855–1932; cf. Bower, 1933; Lloyd, 1935) traveled to New Zealand and Australia in 1898–1899. Goebel accompanied Leonard Cockayne to Westland for two weeks of field work. Cockayne (1933) recalled, "In a moist gully at the base of Mount Alexander in Westland crowned with liverworts of every size and shape (many being giants of the family) von Goebel’s delight rose to its highest pitch. ‘Here,’ he cried, ‘were it possible for me, I would build a cottage and spend the rest of my life studying the bryophytes of this gully.’" Goebel was chiefly interested in the form of plant organs in relation to their function and how these concepts may be explained in physiological, morphological and anatomical terms. His investigations included the formation of leaves in liverworts, the form and function of elaters, the dehiscence of antheridia and the anatomy of the sporophyte capsule (see also comments on fungal components; p. 53). Goebel published major works that were based in part on New Zealand materials that he collected; these publications include Beiträge zur Kenntnis australischer und neuseelandischer Bryophyten (Goebel, 1906) and Organographie der Pflanzen (Goebel, 1915, 1930). Goebel is commemorated by the names Goebeliella Steph. and Goebelobryum Grolle.

Early twentieth-century activity

Apart from the continued work of Stephani, little significant activity occurred during the early twentieth century in the area of liverwort collecting and publication. Lilian Gibbs (1870–1925; cf. Godley, 2004; Gibbs, 1911), a British botanist who collected in Africa, New Guinea and Fiji, visited New Zealand in 1907–1908 and collected liverworts at Nihotupu Hills in the Waitakere Range near Auckland and Mt. Te Aroha. Her collections were determined by Stephani and included four new species that Stephani later published; these were Marchantia laceriloba Steph., Aneura papulolimbata Steph., Calobryum gibbsiae Steph. (≡Haplomitrium gibbsiae) and Lepidozia gibbsiana Steph. (≡Telaranea gibbsiana). William Henry Pearson (1923a, 1923b) published several new species and provided observations based on the collections from New Zealand of William A. Setchell (1864–1943), an algologist and professor of botany at the University of California, Berkeley); examples are Frullania setchellii Pearson and Riccardia riccioides Pearson. Theodor Herzog (1880–1961; cf. Grolle, 1961e) of Munich and later Jena, Germany, published several papers that dealt with novelties of New Zealand hepatics, mostly based on specimens sent to him by E. A. Hodgson (e.g., Herzog, 1935, 1938); he is commemorated by, e.g., Herzogianthus R.M.Schust., Herzogobryum Grolle and Lepidozia herzogii E.A.Hodgs. (≡Telaranea herzogii). Alexander W. Evans (1868–1959, Yale University, New Haven) authored several papers that included observations on New Zealand specimens, e.g., on species of Marchantia (Evans, 1917), Metzgeria (Evans, 1923) and Acromastigum (Evans, 1934).

G. Einar Du Rietz (1895–1967), a Swedish botanist, visited New Zealand in 1926 and 1927 and collected extensively in a number of plant groups, e.g., the genus Euphrasia, and lichens (see Galloway, 1985, p. xxi). He and his wife Greta Du Rietz collected liverworts on Auckland, Campbell and Antipodes islands in 1927, and Hodgson (1962a) and Hodgson and Sainsbury (1948) published information on these collections.

Mid-twentieth-century activity

Collectors and publishers during the mid-twentieth century included most notably E. A. Hodgson (1888–1983), a resident of Wairoa, Hawke’s Bay. Mrs. Hodgson published a number of papers during the period of 1930 to 1972; for a necrology see Godley (1984). Mrs. Hodgson did relatively little field work, and "otherwise she was dependent on other collectors of whom the earliest were K. W. Allison, G. O. K. Sainsbury, J. H. McMahon (Marlborough), V. D. Zotov, Lucy B. Moore, G. Simpson and J. S. Thomson, V. W. Lindauer, Mrs. J. D. Smith (Stewart Island), B. Teague, Ruth Mason, Ella Campbell, and her daughters Lucy and Helen. Later there were her son-in-law, A. P. Druce, Margaret Simpson, William Martin, Jane Taylor, C. J. Burrows, L. Visch and G. A. M. Scott" (Godley, 1984, p. 23). Several collaborators of Mrs. Hodgson are particularly notable. G. O. K. Sainsbury (1880–1957; cf. Martin, 1957; Radula sainsburiana E.A.Hodgs. & Allison), who authored A Handbook of the New Zealand Mosses (Sainsbury, 1955), collaborated with Hodgson in many ways, particularly in the very early days of her career as an hepaticologist. Godley (1984, p. 21) remarked that Sainsbury in 1925 wrote to Mrs. Hodgson asking her to collect plant material for him, and "thus began one of the most important friendships in Mrs. Hodgson’s life."

The hepatic collections of Victor D. Zotov (1908–1977; cf. Thomson, 1977) of the Plant Research Bureau, Wellington, are mainly from North Island and in particular were gathered from the Tararua and Ruahine Ranges in the 1930s and 1940s; he visited Campbell Island in 1961. His collections were published in various papers by Hodgson (e.g., Hodgson, 1962a on the Campbell Island collections) and some by Herzog (1935). Zotov is commemorated by Lophocolea zotovii Herzog (=Hepatostolonophora rotata).

A. P. Druce’s (1920–1999; cf. Atkinson, 1999) collections of liverworts also were included in several studies of Hodgson’s, and he is credited, for example, with making the first collection of the remarkable Phyllothallia nivicola, a genus and species described by Hodgson (cf. Hodgson, 1964a). Druce is commemorated by Pseudocephalozia drucei R.M. Schust. (≡Paracromastigum drucei).

William Martin (1886–1975; cf. Galloway, 1976) published a number of papers on higher plants, particularly vegetation studies. The Flora of New Zealand (Martin, 1961) included several pages devoted to hepatics, and his two papers on "The Bryophytes of Stewart Island" (Martin, 1949, 1950) are noteworthy. He is commemorated by several names, including Lepidozia martinii E.A.Hodgs. (≡Telaranea martinii).

A number of individuals spent 12 months at a time on Auckland and Campbell islands during World War II, having been sent there by the New Zealand government as coastwatchers for enemy ships (see Turbott, 2002) after a German ship (the Erlangen) refueled at Carnley Harbour (Auckland Island) by cutting rata (Metrosideros umbellata) when escaping from New Zealand in 1939. Zoologists, surveyors and meteorologists made up the highly secret Cape Expedition. They were selected for their interests and skills to help occupy the many idle hours available between shifts when on duty at the lookout hut. Cape Expedition staff who collected bryophytes were:

W. H. Dawbin—Auckland Island, 1943. A zoologist who collected many flowering and non-flowering plants.

George Easton ("M. E. Easton" in Hodgson [1962a])—Auckland and Campbell islands, 1944–1945. A surveyor.

Robin L. Oliver—Auckland and Campbell islands, 1944. A zoologist.

Jack H. Sorensen—Campbell Island, 1942, 1943 and 1945. A zoologist and botanist.

E. G. (Graham) Turbott—Auckland Island, 1944–1945. Ornithologist and director of the Auckland Museum (1964–1979). (Turbott visited the Auckland Islands again in 1950.)

The specimens collected by these individuals were made available to Hodgson, who published her "Hepatics from the subantarctic islands of New Zealand including ‘Cape Expedition’ collections from the Auckland and Campbell Islands" (Hodgson, 1962a).

K. W. Allison (1894–1976; cf. Macmillan, 1978) made important contributions during this period. Allison, a forester who worked for the New Zealand Forest Service at Rotorua, Waipoua and Dunedin, was an avid collector of mosses and liverworts and exchanged materials with Hodgson, with whom he coauthored a number of species. His collection of 7600 hepatics is at CHR. Despite gaining a knowledge of the New Zealand Hepaticae more or less equal to that of Hodgson through exchange of material with her, he published little independently, but coauthored with John Child The Liverworts of New Zealand (Allison and Child, 1975), the second and only reasonably comprehensive account of New Zealand hepatics since Hooker’s Flora of New Zealand. Allison collected the distinctive hepatic flora of the pumicelands south of Rotorua before they largely disappeared under extensive Pinus radiata forests planted from the 1930s onwards.

Raymond E. Hatcher (1930–1967; cf. Fulford, 1968b) traveled to New Zealand as a Fulbright scholar and did field work during the period January 1954 – February 1955. Hatcher authored papers on Trichocolea in New Zealand (Hatcher, 1958), a world monograph of Isotachis (1960–1961) and several other papers (see Bibliography). Hemstead Castle (1894–1974, Yale University, New Haven) published a number of papers dealing with Radula, several involving New Zealand species (see Bibliography).

Jane Taylor (1924–), a student of Margaret Fulford (University of Cincinnati), was research fellow at the Botany Division of the Department of Scientific Research (DSIR) in New Zealand between 1961 and 1963 (Botany Division, 1963) and was interested in Lepidoziaceae (especially Lepidozia). While in New Zealand she mainly attempted chromosome studies on New Zealand liverworts. Hodgson named Metahygrobiella tayloriae E.A.Hodgs. (=Metahygrobiella drucei) for her. Her papers are cited in the Bibliography, including the unusual study of spores and elaters of New Zealand taxa that included stereo pairs of scanning electron micrographs (Taylor et al., 1974).

Mid-to-late twentieth-century activity

During the second half of the twentieth century, Rudolf M. Schuster visited New Zealand during the following periods: 1962, 1967, 1969–1970 (Campbell Island), 1983–1984 and 1995. Schuster published many papers based on specimens that he collected. His collections and publications were fundamental in establishing a modern taxonomy for liverworts (see the Bibliography for a list of his contributions to the New Zealand hepatic flora).

Riclef Grolle (1934–2004; cf. Zündorf et al., 2005; Gradstein, 2005) of Jena, Germany, and a student of Herzog, published a number of papers that considerably improved our knowledge of New Zealand hepatics, e.g., on Kurzia (1964b), Herzogobryum (1966b), Lepidolaenaceae (1967a), Jamesoniella and Cryptochila (1971a), and Lejeuneaceae (1973, 1975a, 1980b, 1980c, 1984).

Ella O. Campbell (1910–2003; cf. Rapson, 2004), botany lecturer at Massey University, Palmerston North, published many papers dealing with New Zealand liverworts and hornworts, including detailed morphological studies (see Bibliography for a list of her papers that deal with liverworts).

W. R. Sykes of Botany Division, DSIR, visited the Kermadec Islands in 1966, 1968 and 1970 and collected bryophytes there; these were identified by Ella Campbell (cf. Sykes, 1977).

Bruce Hamlin (1929–1976; cf. Yaldwin, 1976) initially trained under Victor Zotov at the Botany Division, DSIR, Wellington. As botanist at the National Museum, Wellington (1954–1976), he worked on the genera Uncinia and Carex as well as producing popular books on native trees and ferns, but later he became interested in the liverwort flora and compiled a comprehensive checklist of New Zealand hepatics (Hamlin, 1972, 1973a), as well as several other papers (see Bibliography). Peter Johnson of Botany Division, DSIR, Dunedin, made ca. 2800 collections of hepatics on the Auckland Islands in 1973 on behalf of Hamlin; these are lodged at WELT.

George A. M. Scott (1933–1998; cf. Dalton, 1998) originally from Glasgow, Scotland, took up a lecturing position in 1961 at the Botany Department of the University of Otago. In 1970 he moved to Melbourne to teach at Monash University and then Melbourne University, where he continued his interest in the Australasian hepatic flora (cf. Scott, 1985). He was particularly interested in bryophyte ecology and published a number of papers on New Zealand mosses and hepatics. He is commemorated by Protomarsupella scottii R.M.Schust. (≡Allisoniella scottii).

John Child (1922–1984; cf. Olssen, 1998), of Dunedin, was an economist by vocation, teaching at the University of New South Wales and Otago University, and wrote a number of popular books on a wide variety of natural history topics in both Australia and New Zealand, culminating in the noteworthy collaboration with Allison mentioned above. His collections, numbering ca. 5000, are at CHR and F. He is commemorated by Pachyschistochila childii R.M. Schust. & J.J. Engel and Marsupella sparsifolia subsp. childii R.M. Schust. and honored by the annual John Child Bryological Workshop.

John Bartlett (1945–1986; cf. Galloway, 1987), an Auckland school teacher with an exceptional eye for the unusual, collected bryophytes and lichens widely in New Zealand in the period 1981–1986. He is commemorated by Kymatolejeunea bartlettii Grolle.

John Engel conducted field work in New Zealand during the periods of 1971, 1982–1983, 1995, 1997, 2003, 2006 and 2007. The 1971 period in New Zealand was prior to participation in an expedition to Kerguelen Island under the auspices of the United States Antarctic Research Program (USARP). Field work involving collection of hepatics as part of the USARP program also was conducted by Rudolf Schuster and Dale Vitt (Campbell Island, cf. Vitt, 1974) and Ralph Common (Auckland Island, December 1972 – January 1973).

Late twentieth and early twenty-first century collectors and publishers include David Glenny and John Braggins, the latter of whom trained several hepatic students at the University of Auckland, e.g., Elizabeth Brown, Matt von Konrat and Matt Renner, with theses on Riccardia, Frullania and Radula, respectively. Rodney Lewington of Wellington is involved in collecting projects, most recently centering on Fiordland.

2. Biogeography of New Zealand’s liverwort flora in the world context

The Southern Hemisphere flora

New Zealand (Map 1), with, to date, a flora of 595 liverwort and 13 hornwort species, has a strong representation of a cool-temperate Southern Hemisphere flora that is distinct from the floras of the tropics and the Northern Hemisphere. Examples of genera with a mainly Southern Hemisphere distribution are Adelanthus, Balantiopsis, Chiloscyphus, Clasmatocolea, Gackstroemia, Herbertus, Heteroscyphus, Lepidolaena, Marsupidium, Pachyschistochila, Schistochila, Temnoma, Triandrophyllum and Tylimanthus. Most of these genera are shared mainly by Australia, New Zealand and South America, but some have a few species in the Northern Hemisphere and/or the tropics. Characteristically, whereas the Southern Hemisphere species in these genera have quite narrow ranges, the species in genera that are also present in the Northern Hemisphere have wide ranges, e.g., Adelanthus lindenbergianus, with a distribution that includes the Mascarene Islands, Réunion, Madagascar, South Africa, Uganda, Congo, the Tristan group, South Georgia Island, the Falkland Islands, southern South America, the Andes north to Mexico, western Ireland and northwestern Scotland.

In contrast to its rich leafy liverwort flora, the Southern Hemisphere is comparatively poor in the complex-thallose Marchantiales. Of the 27 genera in this order, Australasia has only 10. Neohodgsonia, a monotypic genus of New Zealand, Tristan da Cunha and Gough Island, is the only Marchantialian genus confined to the Southern Hemisphere.

New Zealand is the distributional center of a number of families and genera of Jungermanniales and Metzgeriales as measured by its relative richness at all taxonomic levels in these groups and by the presence of many isolated species that may be paleoendemics. Examples of this richness and paleoendemism are the families Lophocoleaceae, Schistochilaceae, Treubiaceae, Trichocoleaceae, and the genera Telaranea and Zoopsis. Paleoendemism is exemplified not only by many endemic species without close affinities to other taxa (e.g., Archeophylla schusteri, Dendromastigophora flagellifera, Temnoma quadrifidum, various Lepidolaena spp. and various Lepidozia spp.), but by isolated endemic genera, especially Allisonia, Echinolejeunea, Herzogianthus, Jubulopsis, Lembidium, Megalembidium and Verdoornia, all of which lack relatives with close affinities at the genus level.

Endemism

About 50% of the 595 New Zealand liverwort species are endemic. This percentage is about the same as for the ferns (Brownsey, 2001), lower than for the seed plant flora (83% according to Wilton and Breitwieser, 2000), but over twice the number of mosses (23% according to Glenny and Fife, in prep.). This degree of endemism is much higher than for any Northern Hemisphere liverwort flora. For comparison, ca. 20% of Australia’s liverwort flora is endemic.

New Zealand has two endemic families, both monotypic: Jubulopsidaceae, comprising Jubulopsis novae-zelandiae, a twig and bark epiphyte of the Auckland Islands, Stewart Island, Fiordland and Westland (see Engel, 2005b); and Verdoorniaceae, comprising Verdoornia succulenta, a simple thallose liverwort of the North and South Island penalpine and alpine zones.

New Zealand has 18 endemic genera: Amphilophocolea, Allisonia, Bragginsella, Castanoclobos, Cryptostipula, Dinckleria, Echinolejeunea, Eoisotachis, Herzogianthus, Jubulopsis, Kymatolejeunea, Lamellocolea, Lembidium, Megalembidium, Pseudolophocolea, Stolonivector, Verdoornia and Xenothallus. Fourteen of these genera are monotypic.

Flora shared with Australia

New Zealand has 49 families, and apart from the two that are endemic, only seven families (Allisoniaceae, Antheliaceae, Cephaloziaceae, Goebeliellaceae, Monocleaceae, Phyllothalliaceae and Ptilidiaceae) are not shared with Tasmania or mainland Australia. Of New Zealand’s 159 genera and excluding our 18 endemic genera, only 13 genera (Anthelia, Archeophylla, Austrolophozia, Austrometzgeria, Cephalozia, Chloranthelia, Goebeliella, Leiomitra, Lepidogyna, Monoclea, Pachyglossa, Phyllothallia and Ptilidium) are not shared with Tasmania or mainland Australia.

Approximately 284 species are in common to New Zealand and Australia, almost all of New Zealand’s non-endemic species. Of these, ca. 60 species are exclusively in common to New Zealand and Australia. Examples are the monotypic Chaetophyllopsis, which occurs in New Zealand and Tasmania and also Australia (Victoria, Australian Capital Territory, New South Wales, South Australia and Western Australia; see Engel 2005b). Cephalomitrion is present in New Zealand, Tasmania and mainland Australia.

By far the largest element exclusively shared with the New Zealand flora is Tasmania. Four genera are shared exclusively: Brevianthus, Eotrichocolea, Neogrollea and Trichotemnoma; all are monotypic. About 55 species are shared exclusively by New Zealand and Tasmania. Examples are Acromastigum cavifolium, A. verticale, Adelanthus gemmiparus, A. occlusus, Austrolejeunea olgae, Clandarium xiphophyllum, Dendromastigophora flagellifera, Drepanolejeunea aucklandica, Haplomitrium gibbsiae, Herbertus oldfieldianus, Heteroscyphus cymbaliferus, H. decipiens, Hymenophyton leptopodum, Isolembidium anomalum, Lepidolaena clavigera, Metzgeria submarginata, Pachyschistochila succulenta, P. trispiralis, Paraschistochila pinnatifolia, Plagiochilion conjugatum, Radula multiamentula and Temnoma pulchellum. These species are taxonomically spread very evenly across the families and genera. No genera and only a few species are exclusively shared with mainland Australia. These are mostly shared with the southeastern states Victoria and New South Wales (e.g., Drucella integristipula, Lepicolea attenuata, Lepidolaena palpebrifolia and Pedinophyllum monoicum) and may extend into Queensland (e.g., Telaranea elegans).

Flora shared with southern South America

Several genera are shared exclusively with southern South America (i.e., Acrolophozia, Archeophylla, Austrolophozia, Isophyllaria, Lepidogyna, Pachyglossa and Phyllothallia). Monoclea has two species, M. forsteri of New Zealand and M. gottschei in tropical and subtropical America and warm-temperate regions of Chile and Argentina. Some genera are shared by New Zealand, Tasmania (and in some instances mainland Australia) and southern South America, e.g., Allisoniella, Austrolejeunea, Blepharidophyllum, Clandarium, Gackstroemia, Hepatostolonophora, Herzogobryum, Lepicolea, Leptophyllopsis, Nothogymnomitrion, Wettsteinia, and of these, one (Wettsteinia) extends into Asia. In the case of the seven genera shared exclusively with southern South America, no species within these genera are shared. Species shared exclusively by New Zealand and southern South America are rather few: Acromastigum cunninghamii, Cephaloziella hispidissima, Herzogobryum molle, Nephelo-lejeunea taylinayi, Riccardia pallidevirens, Saccogynidium australe, Symphyogyna hymenophyllum and Temnoma quadripartitum. A number of species are shared by New Zealand, Tasmania (and at times mainland Australia) and southern South America, e.g., Anastrophyllum schismoides, Chiloscyphus austrigenus, C. bispinosus, Clasmatocolea notophylla, Cololejeunea pulcherrima, Nothogymnomitrion erosum, Riccardia crassa and Triandrophyllum subtrifidum. Metzgeria decipiens occurs in southern South America, New Zealand, Tasmania, mainland Australia and north to Japan (see map in Engel, 1990a, fig. 85). Zoopsidella has an interesting distribution in that most species are present in tropical sectors of South America while the genus is absent in southern South America; one species is shared by New Zealand and New Caledonia.

These various common elements may reflect ancient land connections, or they may suggest very infrequent transoceanic dispersal events across the ca. 6000 km of the Pacific Ocean. Southern South America also has a significant endemic element. The endemic elements at the family level are Grolleaceae and Vetaformaceae, both monotypic (see Engel, 2005b), and at the generic level are Evansianthus, Herzogiaria, Nothostrepta, Pedinophyllopsis, Perdusenia, Pigafettoa, Pisanoa, Pleurocladopsis, Pseudolepicolea and Roivainenia.

There is a circumsubantarctic element in the flora in New Zealand, which is present on most subantarctic islands such as Kerguelen Island: Clasmatocolea humilis, Herzogobryum teres, H. vermiculare, Lepidozia laevifolia and Temnoma quadripartitum. Within New Zealand such species are primarily alpine or are distributed in southern New Zealand as they are cool-temperate species.

Flora shared with the South Pacific islands

Of the Pacific Islands to the north of New Zealand, the strongest affinities are with New Caledonia. Three genera are shared exclusively with New Caledonia (Austrometzgeria, Chloranthelia and Goebeliella). There are 16 species in common but only three of these (Zoopsis caledonica, Frullania chevalieri and Goebeliella cornigera) are exclusive to the two areas; the remainder are also present in Australia (e.g., Goebelobryum unguiculatum, Lethocolea pansa, Geocalyx caledonicus and Acrobolbus cinerascens). One species, Pseudomarsupidium piliferum, occurs in New Zealand, New Caledonia and in southern South America. Otherwise, the flora in common with the Pacific Islands is composed of widespread tropical species, e.g., Lejeunea flava and L. tumida.

Bipolar element

A significant bipolar element is present in the New Zealand alpine liverwort flora. These are mostly species that are widespread in northern sectors of the Northern Hemisphere but represented in the Southern Hemisphere in New Zealand (and sometimes also in southern South America): Anthelia juratzkana, Calypogeia sphagnicola, Cephaloziella varians, Chiloscyphus cuspidatus, Lophozia excisa, L. bicrenata (also present in southeastern Brazil), Marsupella sparsifolia, M. sprucei, Ptilidium ciliare, Scapania nemorosa, S. undulata and Tritomaria exsecta.

Minor biogeographical elements

New Zealand has several representatives of a pan–southern-temperate flora, e.g., Acrobolbus ochrophyllus, Bazzania nitida, Chiloscyphus semiteres, Clasmatocolea vermicularis, Cryptochila grandiflora, Jamesoniella colorata and Marchantia berteroana are present in New Zealand, Australia, temperate regions of South America and in South Africa (see Engel, 1978, 1990a). To this group also belongs Hyalolepidozia, with the generic type (H. bicuspidata) in southern South America, the Falkland Islands, Tristan da Cunha and South Africa and an endemic species (H. microphylla) in New Zealand. A few species present in New Zealand have distributions that include Australia but also extend into Southeast Asia, e.g., Marsupidium knightii and Metzgeria scobina.

One species, Neohodgsonia mirabilis, is shared by New Zealand and the South Atlantic Ocean islands Tristan da Cunha and Gough.

A small cosmopolitan or subcosmopolitan element is present in the New Zealand flora, almost without exception belonging to the Marchantiales and Metzgeriales : Dumortiera hirsuta, Fossombronia pusilla, F. wondraczekii, Metzgeria furcata, M. leptoneura, Plagiochasma rupestre, Reboulia hemisphaerica, Riccia bifurca, R. crozalsii, R. crystallina, R. fluitans and Ricciocarpos natans. New Zealand is poor in Riccia species (eight species, including three adventive species) and none are common, in contrast to Australia where Riccia is a diverse genus of 43 species.

Adventives

New Zealand has only five liverwort species that are believed to be adventive: Lunularia cruciata (Campbell, 1965a), Riccia bullosa, R. ciliata and R. glauca (Campbell, 1975, 1977a) and Marchantia polymorpha. None of these appear to be abundant or increasing, and only Lunularia is widespread in unmodified habitats. This situation is in marked contrast to that for the flowering plants in New Zealand where the number of introduced species exceeds the number of native species. The situation is similar in Australia where only three liverwort species are believed to be adventive (Marchantia polymorpha, Sphaerocarpos texanus and Chonocolea doellingeri).

3. Theories on origins of the present-day New Zealand liverwort flora

Fulford (1951, 1963c) appears to have been the first to propose continental drift as an explanation for the distribution patterns in the temperate Southern Hemisphere liverwort floras. She viewed continental drift as an explanation for the distribution of genera such as Acromastigum, Clasmatocolea, Lembidium, Lepicolea, Neolepidozia (now placed in Telaranea), Triandrophyllum and Bazzania sect. Vittatae that have a mostly New Zealand–Australian–South American distribution but sometimes extend to southern Africa. This view was presented in much greater detail and with many examples by Schuster (1969a, 1972b, 1979b, 1981b, 1982b), who viewed New Zealand’s liverwort flora as predominantly a flora of southern and ancient origin. In addition, Schuster (1981b, 1982b) argued that the Jungermanniales and Metzgeriales had a Gondwanan origin in a consistently cool wet climate, while the Marchantiales had a Laurasian origin in a continental climate of wet and dry extremes. Many genera in the Jungermanniales, Metzgeriales and Treubiales show a pattern of distribution with an Australasian–South American–Melanesian concentration of species with a thinning of species in Asia (Schuster, 1982b). For instance, Schuster (1982b) pointed out that 21 of 50–52 families of Jungermanniidae are either confined to the Southern Hemisphere or are represented in the Northern Hemisphere by one or two species. Wallace’s line is the point at which many of these Southern Hemisphere genera have their limit or become sparse (see Schuster, 1972b). An example of this is provided by the genus Zoopsis with eight New Zealand species (six endemic), four Australian species (none endemic) and two Southeast Asian species (one endemic to that region).

This Gondwanan-origin view has been given a strongly vicariance interpretation by some subsequent authors, who interpreted the presence of the same species in New Zealand and Tasmania or New Zealand and southern South America to mean that these species were present in Gondwana 80 Mya and that speciation and mutation rates in the DNA of moss and liverwort genera studied has been especially slow (especially Frey et al., 1998 on the moss Lopidium; but also Meissner et al., 1998 on Monoclea; Stech et al., 2000 on Treubia; Schaumann et al., 2003 on Symphyogyna). The concept of "old," "senescent" and stenotypic taxa in the liverworts was put forward by Schuster (1979b) and adopted by Frey et al. (1998), who introduced the term "stenoevolution" for taxa appearing to have very low rates of DNA mutation. McDaniel and Shaw (2003) considered these rates, which are much lower than those reported for any group of land plants, to have a more likely explanation in recent spore dispersal. Heinrichs et al. (2006) argued that their DNA sequence data for Plagiochila, the largest liverwort genus, was consistent with a Gondwanan origin for the genus as well as very infrequent long-distance dispersal events.

Van Zanten (1978) found a strong correlation between desiccation-resistance of spores and breadth of distribution in Southern Hemisphere mosses. Van Zanten and Gradstein (1988) experimentally measured the resistance of Neotropical moss and liverwort spores to drying, freezing and ultraviolet light and concluded that the spores of most Neotropical liverworts (including some species present in New Zealand) were insufficiently drought- or ultraviolet-resistant to survive in high-altitude (10–12 km) jet stream winds for the duration of transoceanic dispersals, but did not rule out dispersal on more moist lower-altitude winds. Van Zanten (1978) took the view that both long-distance dispersal and vicariance are responsible for present-day distribution patterns. Where genera but not species were shared (as is commonly the case between New Zealand and southern South America), vicariance was the explanation, but where species were shared (as is commonly the case between New Zealand and Tasmania), long-distance dispersal was responsible. Schuster (1979b) argued that van Zanten’s (1978) experimental results implied that transoceanic dispersal of bryophytes should be so common that endemism would be almost nonexistent. He argued that other factors needed to be taken into account to explain observed levels of endemism: (1) whether the spore is unisexual or bisexual, (2) whether asexual reproduction is possible, (3) the habitat of the taxon (e.g., whether the habitat is open to wind) and (4) competition between new arrivals and existing species assemblages. Schaumann et al. (2003) used van Zanten’s (1978) data on the lack of freeze and desiccation resistance of Lopidium spores as support for their vicariance explanation of the distribution of the genus. Muñoz et al. (2005) presented evidence that bryophyte, fern and lichen distributions in the Southern Hemisphere correlated with "wind connectivity" between islands better than with geographical proximity of islands, but did not consider that wind connectivity may link islands of similar climate.

Schuster (1983, p. 600) concluded that all taxa of the liverwort floras of the Atlantic volcanic islands of 5–20 Mya (Tristan da Cunha, Gough and Nightingale islands) "had to arrive within a geologically limited time span." The same argument applies to New Zealand’s subantarctic islands, which are of recent (Miocene to Pleistocene) and of volcanic origin. Their floras, with a total of ca. 175 species, are almost entirely composed of species shared with the main islands of New Zealand and include species that are endemic to the New Zealand botanical region (e.g., Archeophylla schusteri, Lembidium nutans and Monoclea forsteri). The exceptions are two endemic species (Plagiochila fuscella and Riccardia umida) and one shared exclusively with southern South America (Riccardia pallidevirens).

The affinities of the New Zealand flora lie most strongly with its near neighbor, Tasmania. The 2000 km between New Zealand and Tasmania is not incomparable with the 820 km between New Zealand and Campbell Island (although in the case of Campbell Island, the Auckland Islands may have acted as a stepping stone for some species). The two landmasses share an oceanic cool-temperate climate. The climate and landform histories of New Zealand and Tasmania need to be compared in order to explain the persistence and evolution of a Gondwanan liverwort flora throughout the Tertiary in Australasia.

New Zealand’s climate and land area were rather different from the present during parts of the Tertiary: the Miocene climate is estimated to have been ca. 10°C warmer, and during the Oligocene New Zealand’s land mass was ca. 15% of its present area (Cooper and Cooper, 1995). This accords with the view, now widely held by New Zealand biogeographers, that only a small percentage of New Zealand’s vascular species had ancestors in New Zealand that were present at the 80-Mya separation of New Zealand from Gondwana. The implications of these factors for both the continuous presence of a liverwort flora in New Zealand and the composition of that flora throughout the Tertiary have not yet been discussed in the literature.

In the past 10 years, transoceanic dispersal, at least with regard to vascular plants, has gained acceptance as the major source of New Zealand’s flora (Pole, 1994; Brownsey, 2001; McGlone et al., 2001; Winkworth et al., 2002; Knapp et al., 2005) even in Nothofagus, a much-cited example of vicariance. DNA-based phylogenies of vascular plants have made it apparent that New Zealand has received most of its vascular flora since the Miocene from Australia, South America and the Pacific Islands, and has contributed to those other floras over the same period. The rate of successful dispersal and establishment seems to have been low in many vascular plant genera, as it is common to discover from a DNA-based phylogeny that a genus in New Zealand has radiated from a single arrival.

However, evaluation of the extent of long-distance dispersal as it relates to Austral hepatics must take into account factors inherent to this plant group. Schuster (1984b, p. 602) remarked that "much has been written on long-distance dispersal; I can only reemphasize what seems obvious: in ‘old,’ usually stenotypic taxa, such as the subantarctic monotypes (e.g., Herzogiaria and Isophyllaria), and in such ‘relict’ taxa as Dendrobazzania, lack of both sexual and asexual dispersal devices excludes these taxa from consideration in any discussion of long-distance transport. In such instances, invoking long-distance transport is manifestly absurd and the disjunct ranges are at least in part the result of contraction of formerly more widespread ranges." Moreover, it must be remembered that the liverwort gametophyte is haploid, and that "spores of Jungermanniae are produced on low sporophytes attached to low gametophytes: they are produced by plants usually growing in moist and sheltered sites screened from drying winds. Hence opportunity for wind dispersal of disseminules is slightly less than is the case with the taller seed plants" (Schuster, 1969a, p. 49). Also, these concepts must be coupled with the ecological behavior of liverwort taxa. Many occur in sheltered, shaded, permanently moist sites both under alpine and notably also in protected niches within forests (see discussions in Schuster, 1969a, p. 86 and Schuster, 1979b, p. 2182); these niches all or for the most part are not subjected to wind currents relevant to long-distance dispersal. Givnish and Renner (2004, p. 5) aptly addressed these arguments regarding disjunct plant distributions and concluded that such studies "require the judicious use of complementary data on phylogeny, biogeography, ecology, and palaeodistributions in time and space."

4. Ecological provinces of New Zealand

Cockayne (1928) formalized a system of botanical provinces for New Zealand. Wardle (1991) adopted this system but modified the names and boundaries to make them compatible with the system of 85 ecological regions and 268 ecological districts (McEwen, 1987) that is in widespread use in New Zealand.

This flora uses Wardle’s (1991) botanical provinces and their boundaries (Maps 2,3), but refers to them as ecological provinces (EPs) to convey that they can be considered part of the hierarchical system of provinces, regions and districts. The system makes more biogeographical sense than the land districts adopted by the Flora of New Zealand series (Allan, 1982; Edgar and Connor, 2000; Galloway, 1985; Healy and Moore, 1980; Moore and Edgar, 1970; Webb et al., 1988) and printed on the rear endpapers of each of those volumes, but in some cases (e.g., Canterbury), the names and boundaries are the same.

The ecological regions (ERs) are nested within ecological provinces with the exception of North Westland ER, which is split between Westland and Western Nelson EPs. Wardle’s boundary between Westland and Western Nelson EPs is more biogeographically meaningful than the ecological region boundary. An outline of these provinces is presented here, but for more detail see Wardle (1991, chapter 6).

Campbell Ecological Province

The province includes Macquarie, Campbell, Auckland, Antipodes and Bounty islands (Wardle, 1991, fig. 6.1). Macquarie Island is excluded from this flora and the following discussion, as the island lies equidistant between Tasmania and the rest of New Zealand; it is politically part of Australia and included in the Flora of Australia series (but not in McCarthy, 2003).

The islands are eroded volcanoes of Miocene to Pleistocene age on a granite basement. They are 460–820 km from New Zealand’s South Island at latitudes 47–52° S. The Bounty Islands have only one vascular species (Lepidium olearaceum; cf. Amey et al., 2007) and probably no bryophyte flora, and are excluded from the following discussion. The islands have a blanket of peat except on coastal cliffs and below the island summits (644 m on the Auckland Islands). The climate is cool but without extreme temperatures (at sea level a mean of 8°C, ranging between 4 and 16°C on the Auckland Islands), sunshine hours are low, and rain falls on most days but is not heavy (annual rainfall is 1.0–1.5 m on the Auckland Islands). The vegetation of the Auckland Islands at sea level is Metrosideros umbellata forest, and above 200 m becomes stunted scrub mixed with Dracophyllum longifolium and Myrsine divaricata. Chionochloa antarctica tussocklands occupy the zone from 200 to 450 m. Near the island summits there are low rush-herbfields of Carpha alpina, Marsippospermum gracile and Rostkovia magellanica with the large herbs Pleurophyllum hookeri and Bulbinella rossii and many small herbaceous species. Oreobolus pectinatus–Centrolepis ciliata–Astelia subulata–A. linearis cushionfields occupy raised bogs at all altitudes. Campbell Island lacks tall forest, and most of the lower slopes of the island are scrub and dwarf forest up to 5 m high of Dracophyllum longifolium and D. scoparium, with Myrsine divaricata and Coprosma species. Between ca. 150 and 300 m, Chionochloa antarctica tussockland was the original vegetation, reduced by pastoral farming and now occupied largely by tall herbfields. Tall macrophyllous herbs (e.g., Anisotome latifolia, Pleurophyllum speciosum) and short herbfields occupy the zone between 300 m and the summit at 570 m (McGlone, 2002).

Endemism in the vascular flora of the whole group is 14%. The largest island group, the Auckland Islands, is 630 km2 in area and has a vascular flora of 233 species, of which 2% are endemic.

There are approximately 175 liverwort species known from the group, with the Auckland Islands having ca. 100 species, Campbell Island ca. 160 species (Schuster, 1979b, p. 2183) and the Antipodes Islands ca. 45 species. Only two liverwort species are endemic, Plagiochila fuscella and Riccardia umida.

Rakiura Ecological Province

The province is comprised of Stewart Island and offshore islands including the Snares Islands and Solander Island. Stewart Island is at 47° S; Mt. Anglem in the north reaches 980 m high, and Mt. Allen in the south 750 m. The southern two-thirds of the island is granite, and the rest is composed of schist or recent sediments, particularly in the large central Freshwater River swamps. Annual rainfall ranges from 1 m at the coast to 3 m in the mountains. Forests of Dacrydium cupressinum, Metrosideros umbellata and Weinmannia racemosa cover large parts of the island, but extensive areas were burned in association with logging of the Dacrydium forests and these areas are now covered mostly in seral Leptospermum scoparium scrub and Empodisma minus sedgelands. Low forests and scrub of Dracophyllum longifolium, Halocarpus biformis, Lepidothamnus intermedius and Olearia colensoi are found in the south. Penalpine cushionfields and tussocklands composed of Chionochloa lanea, C. pungens, Donatia novae-zelandiae, Dracophyllum politum, Hectorella caespitosa and other species are found on Mt. Anglem, Mt. Rakeahua and the Tin Range. Stewart Island has a vascular flora of 585 species and has 28 endemic species (Wilson, 1987). The northern and southern parts of the island are separated by the Freshwater River valley and are floristically distinct. Endemic liverworts are Lepidozia serrulata, Plagiochila hatcheri and Radula parviretis. Several species are shared with Fiordland, Westland and Western Nelson EPs and have their greatest abundance in southern Stewart Island; these are Acromastigum verticale, Allisoniella scottii, Castanoclobos julaceus and Eoisotachis stephanii.

Fiordland Ecological Province

The province is formed by gneiss and granite mountains up to 2000 m high that are deeply incised by glaciers forming wide valleys unchanged in shape since the last glaciation and filled in places by large lakes and deep fiords. Annual rainfall is 6 m on the west coast, decreasing to 2 m on the province’s eastern boundary. Forests, mainly of Nothofagus menziesii, N. solandri and N. fusca, cover most of the province. An extensive tussock zone of several Chionochloa species occurs above the treeline. The province has a small number of endemic alpine herbaceous species. A notable 368 liverwort species have been recorded (R. Lewington, pers. comm., 2005), and endemics are Cephalozia pachygyna and Chiloscyphus tuberculatus; all other species are shared at least with Westland or Rakiura EPs.

Southland Ecological Province

The Livingstone (2000 m), Eyre (2020 m) and Garvie (1600 m) mountains in the north of the province (Mavora ER) are composed of greywacke and similar sandstones with bands of ultramafics. Forests in these ranges are Nothofagus solandri and N. menziesii, with Dacrydium cupressinum forests once common in the lowlands. Low hills and the extensive Southland Plain, now in pasture, form the rest of the province. The Catlins forests in the southeast have the largest remaining areas of forest on low hills (Ajax Hill, 700 m). The province has cool temperatures and low sunshine hours. The Eyre Mountains are a center of endemism for alpine genera such as Celmisia, Aciphylla and Ranunculus, and Seppeltia succuba of the Eyre Mountains is shared with Macquarie Island.

Otago Ecological Province

This is a very diverse province extending from the Main Divide mountains, which reach up to 3000 m high in the west with high rainfall and permanent snowfields (Lakes ER), to flat-topped ranges 1700 m high in Central Otago ER, to coastal hills and an eroded Tertiary volcano forming Otago Peninsula, which receives moderate rainfall (Otago Coast ER). Annual rainfall is less than 0.5 m in the central basins but ca. 2.5 m on the range summits and in the west of the province. The province lies almost entirely on a schist basement. Near the Main Divide Nothofagus forests predominate, but elsewhere almost none of the original fire-prone Podocarpus totara–Phyllocladus alpinus forests remain, and the extensive Chionochloa rubra tussocklands that replaced the forests are now occupied by pasture. Endemic vascular species are found in salt-pans in the basins, and a number of cushion-forming species are endemic to the cold, wind-swept central ranges. The liverwort flora is not large, and there are no endemics known, although Haplomitrium minutum may have originated there.

Canterbury Ecological Province

Greywacke mountains, which reach 4000 m (Aoraki/Mt. Cook), form the western boundary of the province; they have a high annual rainfall and extensive areas of permanent snow. There is a sharp rainfall transition to drier eastern mountain ranges (annual rainfall 0.9–3.0 m), which are occupied by Chionochloa macra and C. rigida tussocklands and large areas of bare scree and fellfield in the alpine zone. The extensive Canterbury Plain was formed by river alluvium and was once covered with lowland forests but currently retains virtually no native vegetation. Small areas of Tertiary limestone form low escarpments throughout the province. On the eastern coast, Banks Peninsula is a Miocene volcano that was until recently covered in lowland forest of Podocarpus totara, Prumnopitys taxifolia and Dacrycarpus dacrydioides emergent over a canopy of, among other species, Hoheria angustifolia, Elaeocarpus hookerianus, E. dentatus and Griselinia littoralis. The province has few endemic vascular species, but the most distinctive vegetational element is the mat and cushion plants of the wide gravel riverbeds, mountain screes and fellfields. Most of the province’s liverwort species are recorded from the wetter western edge of the province. Andrewsianthus hodgsoniae, Bragginsella anomala and Cephaloziella pellucida are endemic; all are found in the alpine zone in the west of the province.

Chatham Ecological Province

Chatham Island (950 km2) and several smaller islands 860 km from the South Island, all low in elevation (to 290 m), are composed of mid-Tertiary to Pleistocene volcanics interbedded with marine sediments that include limestone. Rainfall is low (ca. 0.8 m per year) and droughts are common in the summer, but peat soils nevertheless cover the southern tablelands of Chatham Island, which have the largest remaining area of forest of Coprosma chathamica, Dracophyllum arboreum, Melicytus chathamicus, Myrsine chathamica and Dicksonia squarrosa. The province has a vascular flora of 320 species, of which 10% are endemic, including many of the tree species, whose relatives are mostly New Zealand shrub species. There are ca. 155 liverwort species known from the province, none of which is endemic. The most conspicuous element of the flora is a lowland, warm-temperate element, e.g., Bazzania tayloriana, Cryptochila nigrescens, Harpalejeunea filicuspis, Lamellocolea grandiflora, Nephelolejeuna conchophylla and Trichocolea hatcheri. Most habitats of Chatham Island, including the southern tableland forests, are strongly disturbed by browsing animals. Collections have not been made from the smaller islands of the group.

Westland Ecological Province

The Hollyford Valley and Martins Bay form the southern boundary, while in the north the province includes Nelson Lakes National Park. Most of the province consists of the high mountains forming the Southern Alps, and annual rainfall in these mountains is very high (up to 14 m in the central Whataroa ER) and falls torrentially during storms from the northwest. Flat coastal land up to 10 km wide borders the province and is composed of alluvium and glacial outwash. The same alluvial and glacial landforms are found in the large Grey River valley. Both areas were formerly occupied by tall podocarp forest dominated by Dacrydium cupressinum, which was almost completely milled during the twentieth century. In the central region of the province, Whataroa ER, Nothofagus is completely absent and the montane forests are composed of Weinmannia racemosa, Quintinia serrata and Metrosideros umbellata. To the north and south of Whataroa ER, Nothofagus fusca, N. menziesii and N. solandri form the forest canopy or are mixed with Weinmannia racemosa, and the treeline is formed by Nothofagus menziesii or N. solandri. There is an extensive penalpine zone above the treeline with a scrub composed of Dracophyllum and Olearia and Chionochloa tussocklands, which gives way to alpine herbfields, open rock and snowfields. The province has no vascular endemics. Endemic liverworts are Cephaloziella grandiretis, Lophozia autoica var. alpina, L. nivicola and Telaranea pennata. Other species are shared with Fiordland and/or Western Nelson EPs, e.g., Hygrolembidum triquetrum and Lejeunea cyanophora.

Western Nelson Ecological Province

The province comprises North-west Nelson ER and the part of North Westland ER that includes Denniston and Stockton plateaus and the Paparoa Range. The province has moderately high rainfall (1.7–2.5 m annually) and consists mostly of mountains with deep river gorges and rounded summits to 1500 m. Granites predominate, but the province contains Ordovician-age marbles on the eastern side, Tertiary limestone in the Paparoa syncline, and Tertiary quartzose sandstones on the Denniston and Stockton plateaus north of Westport. Forests of Nothofagus fusca, N. menziesii, N. solandri and Weinmannia racemosa predominate, but in the Paparoa syncline some dense lowland podocarp forests of Dacrydium cupressinum remain. Above treeline, Chionochloa pallens and C. australis tussocklands predominate. The Denniston and Stockton plateaus are dominated by seral Leptospermum scoparium scrub. The province is floristically the richest in New Zealand for vascular plants as a result of its variety of substrate types and because the area has served as a glacial refugium. Riccardia alcicornis, R. intercellula, R. pseudodendroceros, Herzogianthus sanguineus, Austroscyphus nitidissimus and Cephaloziella aenigmatica are endemic to the province. The province has species shared with Rakiura EP, e.g., Acromastigum brachyphyllum and A. verticale, and the bryophyte floras of the Denniston and Stockton plateaus have affinities with the Tasmanian flora (e.g., Neogrollea notabilis).

Marlborough Ecological Province

The northern boundary of the province is the Wairau River, and the southern boundary lies just south of the Seaward Kaikoura Range and the Hamner Range. The province includes the Seaward and Inland Kaikoura mountains, the Raglan Range and other ranges up to 3000 m in elevation. The annual rainfall is low, ca. 0.6 m in the central part of the province; there is no permanent snow as a result of warm summer temperatures. With the exception of low escarpments of limestone and other marine sediments near Kaikoura, the geology is greywacke, which under such a dry climate forms extensive screes and fellfields. The most distinctive vascular flora of the province belongs to these screes and fellfields, and to the presence of limestone. The liverwort flora is small and has no known endemics.

Sounds–Nelson Ecological Province

The province is composed of two regions: Nelson ER, an extensive sheet of glacial outwash, the Moutere gravels, dissected into low hills (500–200 m) and a coastal alluvial plain; and Richmond ER, comprised of mountains up to 1750 m cut mostly from schist, and including the drowned river valleys of Marlborough Sounds (Mt. Stokes, 1200 m), as well as a number of small islands in Cook Strait that preserve a fauna and flora once present throughout New Zealand. The annual rainfall is low, 0.9 m at Nelson city, and sunshine hours are high. The province is the southern limit for some otherwise North Island species, e.g., Beilschmiedia tawa and the epiphytic moss Braithwaitea sulcata. Tall podocarp forests once occupied alluvial plains and fans but are rare now, while the hills and mountains support Nothofagus forests. There is a small endemic flora of alpine herbs on the Richmond Range. No endemic liverworts are known.

Southern North Island Ecological Province

The province includes the Manawatu Plain, a sandy coastal plain, the narrow Tararua and Ruahine ranges, and the Wairarapa and Hawke’s Bay plains and low hills. Extensive lowland forests in the Wairarapa and Hawke’s Bay were milled and burnt, and the lowland areas on both sides of the axial ranges are largely covered in pasture. The axial ranges themselves are covered in Nothofagus forest at higher altitudes and mixed Dacrydium cupressinum, Knightia excelsa and Beilschmiedia tawa forest at lower altitudes. Above treeline, Olearia colensoi scrub and Chionochloa flavescens tussockland dominate, but there is no alpine zone. Endemic vascular species are few, as the middle of the province has a floristic gap caused by submergence of most of the area until the late Quaternary uplift of the Ruahine and Tararua ranges. Endemic liverworts are Pachyschistochila papillifera, Cephaloziella nothogena and Triandrophyllum symmetricum.

Taranaki Ecological Province

The province is composed of two distinct areas: Egmont ER that includes the recently active and very young volcano, Mt. Taranaki (2500 m), its extensive ring plain, and two older eroded volcanoes forming the Pouakai and Kaitake ranges; and Taranaki ER, a large area of dissected low hills of Tertiary marine mudstone. Annual rainfall is ca. 1.5 m near the north Taranaki coast, but rises to 6.5 m at 1000 m on Mt. Taranaki, with torrential rain at times. Annual mean temperature is ca. 13°C at New Plymouth, declining to 9°C at 850 m (Clarkson, 1986). Mt. Taranaki has forests of Dacrydium cupressinum, Metrosideros umbellata and Weinmannia racemosa; Nothofagus is absent. The remaining inland forests of Taranaki ER are mainly Beilschmiedia tawa and Knightia excelsa, with patches of Nothofagus solandri and N. truncata forest on ridges.

Mt. Taranaki is too recent and too recently disturbed to support a diverse liverwort flora despite its high rainfall. Taranaki ER is bryologically poorly known. Allisoniella recurva is endemic to the province.

Gisborne Ecological Province

This is a province of mountain ranges, for the most part forested to their summit ridges, with a high point at Mt. Hikurangi (1750 m). The province is fringed by small coastal plains and low coastal hills, with annual rainfall only 1 m. The mountains are composed of greywacke, and the hills in the east are composed of Tertiary sandstones and mudstones, but rhyolite ash up to 3 m deep forms the soils in the Urewera forests. Forests at low altitudes are mostly Dacrydium cupressinum and Metrosideros robusta emergent over Dysoxylum spectabile, Beilschmiedia tawa, Laurelia novae-zelandiae and Litsea calicaris under an annual rainfall of 1.5–2.0 m. Dense tall podocarp forests of Dacrydium cupressinum, Prumnopitys taxifolia, Podocarpus totara and Dacrycarpus dacrydioides once filled valley floors on thick ash deposits but have survived logging only at Whirinaki. The Urewera Ranges have Nothofagus menziesii on their ridges, under an annual rainfall of 3 m. The forests experience strong disturbance from rain and wind accompanying tropical cyclones. Petalophyllum hodgsoniae is endemic to the province.

Volcanic Plateau Ecological Province

The province includes the central volcanoes of Tongariro National Park, including Mt. Ruapehu (2800 m), and the area, mainly to the northwest and extending to the Bay of Plenty coast, that was blanketed by ash showers from these central volcanoes and the huge Lake Taupo caldera in the recent past. A second volcanic center exists near Rotorua, which also distributed ash as recently as 1886 and has thermally heated ground and fumeroles (Clarkson et al., 1991).

Mt. Ruapehu has Nothofagus solandri forest on the western side under an annual rainfall of 2.0–3.0 m. Chionochloa rubra tussockland is extensive, a result of burning of forest in prehistoric and historical times. The alpine flora is sparse due to the recency of the volcanoes and the area’s remoteness from sources in the South Island. The remainder of the province has forests mainly of Beilschmiedia tawa, Laurelia novae-zelandiae and Knightia excelsa, with Weinmannia racemosa at higher altitudes; Dacrydium cupressinum was common in these forests before logging.

The province has a rich liverwort flora, which is expected due to its catastrophic volcanic activity. This suggests that destruction of forest habitats was never complete even on the slopes of Mt. Ruapehu itself. Species occur on Mt. Ruapehu that otherwise would have their northern limits in the South Island, e.g., Archeophylla schusteri, Diplophyllum gemmiparum and Xenothallus vulcanicolus. The tephra plains around Rotorua had extensive seral Leptospermum scoparium and Kunzea ericoides shrublands that resulted from Maori burning of forests. These and fire-induced Dracophyllum subulatum shrublands on valley floors were probably rich in liverworts including Chaetophyllopsis whiteleggei and Lophozia herzogiana, but are now converted to either pine plantation or pasture. Other endemics are Schistochila pellucida (in the ignimbrite-walled river gorges), Anastrophyllum papillosum, Cephaloziella invisa, C. muelleriana, C. nothogena and C. crassigyna.

Auckland Ecological Province

This is a province of diverse elements: Waikato peat domes and swamps, now very reduced in extent by draining and burning; Waitakere and Hunua ranges (500 m), on greywacke; Coromandel Peninsula with high points Mt. Moehau (900 m) and Mt. Te Aroha (960 m) on the Kaimai Range, a range of Tertiary volcanoes; and Tainui ER with a karst landscape. A recently active field of volcanoes is present in Auckland city (including Rangitoto Island, 260 m, still covered in forest). Both north and south of Auckland, the low hills once had a mosaic of kauri forest (Agathis australis), forest of emergent tall podocarps and Metrosideros robusta, and a canopy composed of a variety of dicotyledonous trees (Beilschmiedia tawa, B. tarairi, Weinmannia, Knightia and locally Nothofagus on the ridges). Plagiochila bazzanioides is endemic. Mt. Moehau is special in the province for having a liverwort flora that has many species in common with Western Nelson, Westland and Rakiura EPs, e.g., Mnioloma novae-zelandiae; it is also the northern geographical limit for ca. 20 flowering plant species.

Northland Ecological Province

Comprising the area north of Kaipara Harbour, the province notably includes Waipoua Forest in Western Northland ER and the distinctive subtropical area at the northern tip of the North Island, Te Paki ER. The province has a warm-temperate climate, and kauri (Agathis australis) forests dominate the uplands (to 780 m) on infertile podzolized soils. However, where kauri is absent, forests have Dacrydium cupressinum emergent over a canopy formed by, e.g., Metrosideros robusta, Elaeocarpus dentatus, Weinmannia sylvicola, Beilschmiedia tarairi and B. tawa. Typically valleys are dominated by broadleaf species of Beilschmiedia tarairi, Dysoxylum spectabile, Corynocarpus laevigatus and Vitex lucens, with tree ferns and Rhopalostylis common in the understory. Coastal forest is dominated by Metrosideros excelsa. Waipoua, on the western, wetter side of the province, has a rich liverwort flora. Schistochila nitidissima, Frullania wairua, Telaranea granulata, Cololejeunea falcidentata and Rectolejeunea denudata are endemic to the province. A number of subtropical species extend no further south than Northland, e.g., Dumortiera hirsuta and Stenolejeunea acuminata. The Three Kings Islands are included in the province and 19 liverwort species have been recorded there, none of which are endemic to the islands (Braggins, 1987).

Kermadec Ecological Province

This province is a group of Pleistocene-age volcanic islands situated 980 km north of the North Island at 29° S in the subtropical zone. Raoul is the main island, 33 km2 in area and 520 m high, and is volcanically active (as recent as March 2006). It has a vascular flora of 113 species, 10% of which are endemic. Metrosideros kermadecensis forms the forest canopy on Raoul Island and formerly did the same on Curtis and Macauley islands. Twenty liverwort species are present (Sykes, 1977), with Plagiochila pacifica and Radula cordiloba subsp. erigens the only endemics. Plagiochila kermadecensis is shared with Auckland EP. The province has a subtropical element, e.g., Archilejeunea planiuscula.

5. The climate of New Zealand

New Zealand is positioned in the largely land-free South Pacific Ocean mainly in the latitude zone termed the "roaring forties" for its strong westerly airstreams that are not much interrupted by land masses. Westerlies reaching New Zealand are little modified by the land mass of Australia because of the 2000-km-wide Tasman Sea that separates the two countries and because of New Zealand’s more southern latitude. Rainfall on the western side of the three main islands is high due to the blocking of the prevailing westerlies by the main mountain axis, the effect being strongest in the central South Island where the mountain ranges are highest. East of the axial mountain ranges annual rainfall decreases rapidly. The highest recorded rainfall for a 12-month period in the central Southern Alps, in the Cropp River valley, was 18 m, but the site with the highest mean annual rainfall is Milford Sound, Fiordland, with 6.7 m per year; the site with the lowest rainfall is Alexandra, Central Otago, with 0.36 m (NIWA science, 2006).

The country occupies the warm-temperate and cool-temperate zones. The Kermadec Islands are in the subtropical zone, but their vegetation is New Zealand warm-temperate in character. At sea level, the entire North Island and northern part of the South Island occupy the warm-temperate zone (mean January air temperature at sea level 17.5–20°C), while most of the South Island and the subantarctic islands occupy the cool-temperate zone (mean January air temperature at sea level 10–17.5°C).

An altitudinal zonation system of lowland, lower montane, upper montane, penalpine and alpine corresponds to these temperature-based zones. The scheme of Wardle (1991, fig. 5.2) is followed here (Fig. 1), but we avoid the confusing term "subalpine" to refer to forests below treeline. Instead we use the term "upper montane" for forests in the zone below the treeline in accordance with international usage, particularly in the tropics (e.g., Whitmore, 1975, p. 199). These are mostly stunted Nothofagus solandri and N. menziesii forests, but where beech is absent, e.g., in Central Westland, forest comprised of Dracophyllum traversii, Olearia ilicifolia, O. lacunosa, Libocedrus bidwillii, Archeria traversii and Halocarpus biformis occupies this zone, and on Mt. Taranaki, of Libocedrus plumosa, Podocarpus hallii, Griselinia littoralis and Olearia colensoi (Clarkson, 1986). In New Zealand the use of Wardle’s term "penalpine" for the zone of tall shrubs and tussocks above the treeline is increasing and seems useful to refer to what in New Zealand is a distinct vegetation zone while avoiding the confusion that exists with the term "subalpine." The tall shrubs of this zone are mostly Dracophyllum (especially D. longifolium and D. rosmarinifolium) and Olearia (O. colensoi, O. lacunosa and O. ilicifolia) and the tall tussocks mainly Chionochloa pallens, C. flavescens and C. rubra.

In this scheme, at 43° S, the latitude of Arthur’s Pass, lowland forests extend from the coast to an upper limit at 400 m, lower montane forests extend to ca. 800 m and upper montane forests extend to treeline at 1300 to 1350 m (depending on aspect). The penalpine zone at this latitude is between 1300 and 1500 m, and the alpine zone between 1500 and 2000 m (Fig. 1). Wardle (1991) defines a nival zone for areas of permanent snow above this, but we do not refer to this zone since the highest liverwort records are only slightly above the 2000 m level (Diplophyllum gemmiparum, 2025 m, leg. Engel and Braggins [Engel and Merrill, 1998]).

High and consistent rainfall and cool temperatures on the western side of the islands (particularly in Western Nelson, Westland and Fiordland EPs) result in impressively luxuriant bryophyte growth, with liverworts dominating the bryophyte component of the vegetation on the western side of the islands and on Stewart Island. In the western provinces, bryophytes cover a large percentage of the floor of lower and upper montane forests and can be a high percentage of their plant diversity. Some genera that can dominate the ground cover in these areas are Acromastigum, Bazzania, Chiloscyphus, Heteroscyphus, Lepicolea, Lepidozia, Lepidogyna, Plagiochila, Riccardia, Schistochila, Telaranea, Trichocolea and Tylimanthus. In eastern provinces where annual rainfall is much lower and summer temperatures reach 30–35°C, mosses dominate the bryophyte component of the vegetation, the bryophyte ground cover is usually less, and there are a limited number of dry-tolerant liverwort species (e.g., Chiloscyphus bispinosus, C. echinellus, C. lentus, C. semiteres, C. subporosus, Lepidozia laevifolia and epiphytes Frullania aterrima, F. deplanata, F. monocera, F. pentapleura, F. patula, F. rostrata, Lejeunea flava, Lepidolaena taylorii, Metzgeria flavovirens, M. furcata and Porella elegantula). However, constantly moist habitats near streams and in shaded gullies support a wider range of species such as Asterella tenera, Hymenophyton flabellatum, H. leptopodum, Monoclea forsteri, Plagiochila banksiana, Radula buccinifera and Symphyogyna hymenophyllum.

6. Geology and landforms of New Zealand

The most notable feature of New Zealand’s geology is its position at the junction of the Australian and Pacific tectonic plates. A transform fault, the Alpine Fault, runs through both main islands and has longitudinally offset the two halves of the South Island by 480 km over the last 25 My in which it has been active. However, it is doubtful whether this transform fault has produced disjunctions in plant distributions; none have been detected in the liverwort flora.

The Southern Alps of the South Island and axial mountains of North Island (the Tararua and Ruahine ranges) result from upward movement on the western edge of the Pacific Plate, which is overriding the Australian Plate. They are composed mainly of marine siltstones and mudstones of Mesozoic age, known in New Zealand as "greywacke." Their main physical feature is a tendency to fracture through faulting and freeze-thaw cracking. At high altitude in Canterbury and Marlborough, the shattering of the rock has resulted in extensive bare screes and fellfields. In Westland and Central Otago the same sediments were pressure-metamorphosed and occur as more durable schists. These greywackes and schists are unremarkable in their chemistry. The main effect of the Southern Alps and the axial ranges of the North Island is through the impact these mountains have on the rainfall as discussed above, and the cooler temperatures at higher altitude.

All Tertiary marine limestones occur at lowland to montane elevations and are found at Waitomo, Canterbury and Western Nelson (Paparoa syncline), but also are scattered through other areas, e.g., Southland and Wairarapa. A number of bryophyte species are confined to this rock type and include Chiloscyphus calcareus and Temnoma angustifolium. Characteristic communities of liverworts occur on limestone walls, particularly in river gorges (e.g., Asterella australis, Heteroscyphus triacanthus and Marchantia foliacea).

Weakly indurated calcareous siltstones called "papa" are common in the lower North Island and are exposed there in large road cuttings. Where there is shade and water seepage these support a distinctive flora and are the exclusive habitat of the hornwort Phaeoceros coriaceus.

Marble of Ordovician age occurs mainly in Western Nelson, forming the Arthur Range and Marino Mountains. Pure calcareous soils in these ranges are rare so that only epilithic bryophytes experience high calcium and magnesium concentrations. No liverwort species are known to be special to this substrate. However, sinkholes in these two ranges provide constantly cool and moist conditions, an ideal habitat for species such as Allisonia cockaynei, Haplomitrium ovalifolium and Phyllothallia nivicola (although these species are not restricted to this area or type of habitat).

Ultramafic bands of rock occur in Sounds–Nelson (Richmond Ra.), Otago (Livingstone Mountains) and Westland (Cascade moraine) EPs and consist mainly of weathered serpentine. The rock and derived soils are rich in calcium and magnesium but low in nitrogen and phosphorus. Nothofagus forests originally occupied these areas, but they are now largely bare or have Leptospermum scoparium scrub. The liverwort flora of these zones is sparse. Dry-tolerant Chaetophyllopsis whiteleggei is recorded from the Nelson ultramafic area, and Castanoclobos julaceus from the Cascade ultramafic moraine (Westland).

Rhyolite ash (tephra) and ignimbrite (hot aerial ash fused on deposition into a rhyolitic rock) blanket all of Volcanic Plateau EP, most of Taranaki EP and parts of Gisborne and Southern North Island EPs. Most of the material originates from the Taupo, Rotorua and Okataina calderas that were active over the last 0.4 My. The volcanoes of Mounts Tongariro, Ngauruhoe, Ruapehu and Taranaki have produced cones 2000–3000 m high during the last 0.5 My and have added to the ash cover of these provinces. The largest ash shower, ca. 800 km3 in volume, originated from the first eruption of the Taupo caldera 27,000 years ago. The tephra soils are deep, friable and free draining but not particularly fertile. Ignimbrite is exposed on cliffs and in river gorges. Heated ground and thermal vents near Rotorua have a distinctive bryophyte assemblage; the moss Campylopus capillaceus is an endemic of such sites, and Lepidozia and Telaranea species are common, but there are no liverwort species special to these areas.

Volcanoes are present in other parts of New Zealand, notably the Auckland volcanic field (recently active), Banks Peninsula (Pleistocene), the subantarctic islands (Miocene to Pleistocene), Otago Peninsula (late Miocene to Pliocene) and in Northland and the Coromandel Peninsula (Pliocene). The aa lava present on Rangitoto Island has created the most distinctive environment present on any New Zealand volcano and provides a habitat for a diverse bryophyte flora.

Granites make up almost all of Fiordland, most of Stewart Island and isolated batholiths in Westland (e.g., the Hohonu Range) and are present in Western Nelson EP (e.g., the Paparoa Range). All experience high rainfall and have well-developed peat soils because of the poor drainage and lack of mineral soil development. All have rich bryophyte floras that are associated with these organic soils rather than with the granite substrate itself. The exception is on Stewart Island where there are liverworts that grow mainly on the bedrock itself, kept moist by constant seepage over the rock surfaces (e.g., Eoisotachis stephanii).

Extensive lowland alluvial floodplains are present in all provinces of the main islands, particularly the large Canterbury Plains, and have the most fertile soils and mesic climates found in the country. These floodplains were, until human arrival 500 years ago, covered by tall podocarp forests that now exist only in small areas.

7. The vegetation of New Zealand

Forests

Tall conifer forests of the lowlands were once widespread in large areas on alluvium, tephra plains, glacial outwash terraces and estuarine flats, and were often a mixture of Dacrydium cupressinum, Dacrycarpus dacrydioides, Podocarpus totara, Prumnopitys ferruginea and P. taxifolia. Composition varied with soil fertility, drainage conditions and altitude. Dacrydium cupressinum is the most common of the tall podocarps in New Zealand and is common as an emergent in forests dominated by Nothofagus, Weinmannia and Beilschmiedia. In Northland and Auckland EPs, Agathis australis forest was a major lowland forest type, particularly on old infertile (gleyed and podzolized) soils, often with Gahnia xanthocarpa and Astelia trinervia forming an understory. In some places Agathis australis is emergent over a canopy of Beilschmeidia tawa or B. tarairi. In fertile and often swampy sites, Dacrycarpus dacrydioides forms pure stands (in wet sites associated with Phormium tenax), but also occurs on well-drained soils with a Coprosma understory. Forests of Prumnopitys taxifolia, Podocarpus totara and P. hallii were common in prehuman times on dry, fertile, well-drained sites such as the Canterbury Plains. On the deep tephra soils of the Volcanic Plateau EP, dense mixed forests of Dacrydium cupressinum, Dacrycarpus dacrydioides, Podocarpus totara, Prumnopitys ferruginea and P. taxifolia were common; this type of forest is now confined to small areas at Whirinaki, Tihoi and Pureora. Tall podocarp forests have a rather poor terrestrial bryophyte flora due to low light conditions on the floor, but on forest margins or where the canopy is less dense, the epiphytic flora is well developed.

Lowland and montane forests of the warm-temperate zone are a complex and varied assemblage of tree species, mainly Dacrydium cupressinum and Metrosideros robusta as emergents over a canopy of Beilschmiedia tawa (Auckland to Nelson), B. tarairi (Northland and Auckland EPs), Phyllocladus trichomanoides and Weinmannia sylvicola (Northland), W. racemosa (south of Auckland), and throughout the North Island, Knightia excelsa, Elaeocarpus dentatus and Weinmannia racemosa. Less widespread and common species that form the canopy are Ackama rosifolia, Corynocarpus laevigatus, Dysoxylum spectabile, Ixerba brexioides, Laurelia novae-zelandiae, Litsea calicaris, Metrosideros excelsa, Syzygium maire, Alectryon excelsus and Vitex lucens. These warm-temperate forests tend to have a low diversity of terrestrial bryophyte cover; relatively dry-tolerant mosses such as Leucobryum candidum dominate, and terrestrial liverworts are usually restricted to damp sites particularly near streams, or on leaf-litter-free sites such as tree fern bases. Rotting logs are less of a feature of the forest floor than in the cool-temperate zone and often are dominated by mosses. The epiphytic liverwort flora is abundant in high-rainfall areas.

Nothofagus occupies extensive mountain areas in the lower and upper montane forest zones of the South Island, but is more limited in its distribution in the North Island due to competition from warm-temperate species and from the smaller extent of higher-elevation land. There are four species: N. fusca occupies warmer and more fertile sites, especially on north-facing hillslopes and in frost-free valleys and between ridges; N. truncata is found mainly on ridges; N. menziesii occurs in cool and sometimes rather infertile sites especially in South Westland and Fiordland; and N. solandri is found in similar cool and somewhat infertile but drier sites. Nothofagus tends to exclude other canopy-forming species, but two tree species frequently share the canopy with Nothofagus : Griselinia littoralis and Weinmannia racemosa, the latter of which is sometimes a canopy co-dominant, particularly with N. fusca. Where Nothofagus is absent from montane forests, particularly in Central Westland, the midslopes of Mt. Taranaki and Stewart Island, Weinmannia racemosa forms forests with Metrosideros umbellata and Quintinia serrata with Libocedrus emergent. Nothofagus forests offer excellent bryophyte habitats because they offer a fairly low canopy, their small leaves do not smother terrestrial habitats with leaf litter, and understory shrubs and ground herbs are often quite sparse. Weinmannia–Metrosideros forests, on the other hand, are less favorable for bryophytes, but the large root systems of Metrosideros offer protected habitats to a distinctive assemblage of bryophyte species that includes, most notably, several Zoopsis species.

Seral forests of Aristotelia serrata, Fuchsia excorticata, Hoheria, Plagianthus (all deciduous), and Carpodetus serratus, Melicytus ramiflorus, Pseudopanax spp. and Schefflera digitata are common around slips and riverbanks in the lowland and montane zones. Due to their low canopy (ca. 10 m), proximity to streams and deciduousness, these forests provide a good liverwort habitat.

The most common scrub type in New Zealand is formed by Leptospermum scoparium, which occurs throughout the country except on the Kermadec Islands. It is a very common and abundant seral vegetation after burning, particularly on wet or organic soils. The taller Kunzea ericoides is often seral in similar habitats in lower rainfall areas.

A distinctive type of heathland grading to fern-rushland, termed "pakihi" in the South Island and Stewart Island and "gumlands" in Northland and Coromandel, forms on flat terraces or low-angled hills after removal of forest. Gumlands and pakihi differ in floristic composition, e.g., the species composition involving Dracophyllum and Baumea, but their soils, vegetation structure and history are broadly the same (Wardle, 1991, p. 201). The soils of these heathlands are gleyed and podzolized, with resulting perched water tables and low fertility. These communities are composed of varying and changing proportions of Leptospermum scoparium, Gahnia rigida, Baumea spp., Empodisma minus, Gleichenia dicarpa and G. microphylla.

Gumlands and pakihi occur in Northland, Coromandel, Western Nelson, Westland and Stewart Island, replacing mainly Agathis australis and Dacrydium cupressinum forests where these were logged and burned. Such communities can persist for decades, depending on drainage, soil fertility and frequency of fire, until the Leptospermum eventually forms a closed scrub canopy that forms shelter for seedlings of tree species. This vegetation type appears to have been created mostly by logging and burning of conifer forests, but pollen analysis of Northland gumlands indicates the vegetation type was present 16,000 years ago as a result of natural fires (Wardle, 1991). The extensive areas of these induced heathlands created in the twentieth century are now diminishing due to fire restrictions. Goebelobryum unguiculatum, Lethocolea pansa and Neogrollea notabilis are confined to this habitat in New Zealand, while Kurzia compacta, K. helophila, Riccardia multicorpora, R. pusilla, Telaranea herzogii and T. quinquespina occur in these habitats along with the distinctive mosses Pleurophascum ovalifolium and Pulchrinodus inflatus.

A distinctive liverwort community exists in what has been termed the "Murihiku alliance" (C. Meurk, pers. comm., 2005) of conifer scrub, particularly Lepidothamnus intermedius and Halocarpus biformis, mixed with low Nothofagus solandri and Metrosideros umbellata. Scrub of this type may have naturally formed a mosaic of scrub and pakihi, and most of it has been burned to some degree. Where burned, it has formed open fern-rushlands (pakihi). This vegetation occurs in southern Stewart Island, the highest station occurs at Waitutu Forest, Cascade Plateau, and small pockets remain throughout Westland, Stockton and Denniston plateaus, Gunnar Downs and in small areas in the North Island at treeline on Mt. Ruapehu and Mt. Moehau. This scrub is particularly rich in liverworts because the canopy is low, the soils are infertile (and therefore competition from vascular plants is low) and also the usual occurrence of peat over granite holds moisture well; the scrub is found only under high rainfall. Liverwort species occurring in this scrub are Acromastigum brachyphyllum, Blepharidophyllum vertebrale, Clandarium xiphophyllum, Eotrichocolea polyacantha, Isolembidium cucullatum, Megalembidium insulanum, Pseudolophocolea denticulata, Saccogynidium decurvum, Telaranea inaequalis, Zoopsis bicruris and epiphytes Goebeliella cornigera, Heteroscyphus menziesii, Jubulopsis novae-zelandiae and Schistochila pseudociliata. The combination of Acromastigum cavifolium, Archeophylla schusteri, Brevianthus flavus, Herzogianthus sanguineus and Leptoscyphus australis in this scrub type is distinctive.

Nothofagus solandri and N. menziesii commonly form the upper limit of forest and may form scrub 2–3 m high near treeline on exposed ridges. However, penalpine scrub composed mainly of Olearia colensoi, O. ilicifolia, Brachyglottis rotundifolia, Dracophyllum longifolium and D. rosmarinifolium commonly occurs above treeline.

Grasslands

Grasslands occur naturally in the penalpine and alpine zone, and in river valleys where forests are excluded by severe winter frosts. In such river valleys, and presently on extensive valleys and hillslopes formerly occupied by forest, Festuca novae-zelandiae, F. matthewsii, and in damper and more fertile soils, Poa cita form what are often termed "short tussocklands" to distinguish them from the tall Chionochloa tussocklands of higher elevations. These grasslands now are densely populated by adventive grasses such as Anthoxanthum odoratum and Agrostis capillaris and offer a bryophyte habitat only where soils are too thin for grasses to grow.

Chionochloa pallens and C. flavescens are the main species in the penalpine tussock zone in western provinces of the South Island and form large tussocks typically 0.5–1.0 m high. These are a good habitat for bryophytes where they grade into Celmisia herbfield, particularly on shaded south-facing slopes and around rock outcrops. Blepharidophyllum vertebrale, Clandarium xiphophyllum, Gackstroemia alpina, Heteroscyphus billardierei, Lepidogyna hodgsoniae, Paraschistochila conchophylla, Riccardia pennata, Schistochila monticola and the hornwort Megaceros giganteus are characteristic of this habitat, and Eotrichocolea polyacantha, Phyllothallia nivicola and Riccardia pseudodendroceros have been found in such sites. Chionochloa macra and C. rigida replace C. pallens in equivalent habitats east of the Main Divide, but the habitats are too dry for liverworts except around streams and seepages. Chionochloa crassiuscula and C. oreophila occur at higher altitudes in the alpine zone and are often mixed with Poa colensoi and Celmisia -dominated herbfields. Chionochloa rubra tussocklands occur above treeline in the North Island and northern South Island, particularly on low-angled and infertile organic soils, but are more often associated with wetlands on peaty soils below treeline. In the northern half of the South Island C. australis forms dense mats that tend to exclude bryophytes.

Wetlands

The many wetland types in New Zealand for the most part lack any significant diversity of bryophytes. Estuarine wetlands lack bryophytes because these plants are not salt-tolerant. In the large gravel riverbeds species of the moss Racomitrium may dominate, but the habitat is too dry and exposed to support a liverwort flora. Two broad classes of freshwater wetlands may be distinguished by their fertility and water flow: swamp and bogs.

Swamps, usually in the lowland zone, are dominated by Typha orientalis, Phormium tenax and the large sedge Carex secta; other Carex species such as C. sinclairii and C. gaudichaudiana can form extensive sedgelands. Floating plants are Azolla, Callitriche, Lemna, Potamageton and the liverwort Ricciocarpos natans. Swamps usually are not rich in bryophytes, and the liverworts to be found in swamps are usually rather common unspecialized species, e.g., Marchantia foliacea and Chiloscyphus species.

Bogs, by contrast, are a habitat for a diverse set of liverworts that require constant moisture but tolerate high light levels. Most bogs occur in the penalpine and lower alpine zone in the North and South islands but reach sea level on Stewart Island and the subantarctic islands. Two cushion-forming species, Oreobolus pectinatus and Donatia novae-zelandiae, commonly form much of the cover, and the mosses Sphagnum cristatum and S. falcatulum are present in well-inundated areas. Carpha alpina, Centrolepis ciliata, Chionochloa rubra, Halocarpus bidwillii and L. laxifolius are also common, and bogs often grade into low podocarp scrub. Liverwort genera that have their main occurrence in bogs are Austrofossombronia, Calypogeia, Cephalozia, Cephaloziella, Goebelobryum, Metahygrobiella and Pseudocephalozia. Certain species of Acromastigum, Allisoniella, Kurzia, Telaranea and Riccardia also have their main occurence in bogs.

Liverworts in flushes in the penalpine and alpine zones experience similar conditions to bogs but nutrient levels may be higher. Austrofossombronia, Cryptochila, Marchantia, Marsupella and Scapania are found in flushes, in addition to many of the genera and species of bogs listed above.

Aquatic liverworts occur mostly on rock, usually in very small streams, beside waterfalls and cascades, or are found on rock surfaces that are covered by a thin film of water seepage, as abrasion of leaves by sediment limits bryophyte growth in most large streams and rivers. Species and genera that have their main occurrence in this habitat are Allisoniella recurva, Cephalomitrion aterrimum, Cryptochila grandiflora, Eoisotachis stephanii, Hepatostolonophora rotata, H. paucistipula, Isophyllaria attenuata, Monoclea forsteri, Pachyglossa tenacifolia, Schistochila kirkiana, S. nitidissima, S. pellucida and species of Aneura, Lejeunea, Riccardia and Plagiochila.

Ricciocarpos natans and Riccia fluitans, both of cosmopolitan distribution, are the only free-floating aquatic species, but also can be found on mud in lowland swamps after water levels drop. Both appear to be much less common now than formerly.

Sparsely vegetated alpine habitats

In the alpine zone, sparsely vegetated surfaces provide a habitat for bryophytes: rock surfaces because the soil is too thin for vascular plants, and soil pockets and overhangs because they are protected from rain but receive cloud moisture from fog. A very distinctive bryophyte flora found on rock and thin mineral soils in exposed sites has Acrolophozia pectinata, Andreaea acuminata, A. acutifolia, Andrewsianthus cuspidatus, Diplophyllum domesticum var. icari, Gymnomitrion cuspidatum, Herzogobryum teres, Nothogymnomitrion erosum, Solenostoma rufiflorum and S. totipapillosum. Dracophyllum pronum is often associated with such sites. Conditions are less extreme on rock sites with water seepage (including snow melt) but otherwise free of vascular plants because of a lack of soil; Allisoniella, Cephalomitrion, Cephalozia, Hepatostolonophora, Solenostoma inundatum, Temnoma and Triandrophyllum subtrifidum are characteristic.

8. Principal substrates of New Zealand liverworts

The forest floor

The litter of some of the most abundant New Zealand trees, especially Nothofagus solandri and N. menziesii, but also all of the native conifers, is small enough that bryophytes growing on slightly uneven sites on the forest floor are able to survive inundation by the steady but light rain of leaf litter, undoubtedly assisted in this by the almost complete lack of deciduous tree species in New Zealand. Lepidoziaceae genera (Lepidozia, Bazzania, Megalembidium, Telaranea) are very successful in this situation, but other genera such as Trichocolea, Leiomitra, Lepidogyna, Chiloscyphus, Heteroscyphus, Plagiochila, Schistochila and Hymenophyton do equally well in these sites and are competitive against mosses such as Dicranoloma which specialize in this habitat. However, it is noticeable that in the northern half of the North Island, where larger-leaved tree species such as Dysoxylum, Elaeocarpus and Beilschmiedia dominate, terrestrial species are confined to sloping sites where leaf litter is unable to accumulate, or near streams where flooding regularly removes leaf litter. Where the ground is more uneven, particularly in the cooler southern forests, other less competitive genera, such as Acromastigum, Anastrophyllum, Chandonanthus, Gackstroemia and Lepidolaena, find a place on the forest floor.

Rotting logs and stumps are prime habitat for a diverse range of terrestrial liverworts because they shed leaf litter from their sloping surfaces, water is held by the decaying wood, and in the case of large logs they are comparatively well lit by being raised from the forest floor. Later in log or stump succession, bryophyte mats and the tight, interlaced root bases of hymenophylls are conducive to water retention. Rates of decay of logs and stumps are slow in the cool-temperate zone, and consequently they cover significant parts of the forest floor. Lophocoleaceae genera such as Chiloscyphus and Heteroscyphus excel in this habitat, as do Lepidoziaceae genera. In addition to all the forest floor genera listed above, Jamesoniella, Marsupidium, Plagiochilion, Riccardia, Schistochila, Tylimanthus and Wettsteinia have their main habitat on rotting logs.

Soil and humus under tree roots and in tree-root caves provide a habitat protected from the litter rain, but such sites are difficult because of their lack of moisture and dim light conditions. Zoopsis is extremely well adapted to these conditions by virtue of the large lens-like stem cortex cells that focus light on chloroplasts on the back walls of the cells and the reduced leaves that limit water loss. Other species adapted to these conditions are Bazzania nitida, Psiloclada clandestina and Zoopsidella caledonica.

Cliff faces, boulder faces, overhangs, stream banks and road and track cuttings within the forested environment (and at its margins) provide favorable environments for liverworts. Such sloping or overhanging sites are free of litter, are often well-lit, and often have water seepage from above or receive direct rainfall. The majority of terrestrial liverworts in New Zealand are found in such sites, with genera such as Aneura, Balantiopsis, Chiloscyphus, Cryptochila, Drucella, Geocalyx, Heteroscyphus, Isotachis, Jackiella, Lembidium, Metzgeria, Pallavicinia, Plagiochila, Radula, Riccardia, Saccogynidium, Solenostoma, Temnoma and Triandrophyllum prominent. Roadside banks in Westland and Fiordland are notable for their complete cover of single species such as Isotachis lyallii.

Epiphytes

At the base of trees there is no sharp distinction between epiphytes and terrestrial liverworts as soil accumulates and the surfaces are sloping and consequently leaf-litter free. Genera of Lepidoziaceae and Lophocoleaceae predominate. In Westland, because the humidity above ground level is consistently high, the distinction between terrestrial and epiphytic species is blurred and terrestrial species grow on sloping tree trunk bases up to about 2 m, e.g., Bazzania adnexa and Schistochila nobilis.

The epiphytes of tree trunks and tree branches can be rather arbitrarily divided into two groups, firstly those of upper montane forests where the canopy is low, light levels are high and strong drying winds at times occur. Species and genera typical of this habitat are Acrobolbus concinnus, Chandonanthus squarrosus, Goebeliella cornigera, Herbertus oldfieldianus, Heteroscyphus menziesii, Jamesoniella kirkii, Lepicolea attenuata, L. scolopendra, Paraschistochila tuloides, Pseudomarsupidium piliferum, and species of Lepidolaena and Frullania. These species for the most part exhibit some adaptations to a more rigorous habitat, e.g., strongly collenchymatous leaf cells, thick-walled stem cells and frequent presence of secondary pigments.

Secondly, epiphytes of lowland and montane forests with a high canopy experience lower light levels (except on forest margins), but the habitat is more constantly moist. Here, typically epiphytic Plagiochilaceae species are present, such as Acrochila biserialis, Plagiochila annotina, P. circinalis, P. deltoidea, P. fasciculata, P. incurvicolla, P. obscura, P. radiculosa, P. stephensoniana, P. strombifolia and Plagiochilion conjugatum. Also typical are Archilejeunea olivacea, Dendromastigophora flagellifera, Harpalejeunea latitans, Lejeunea flava, Mastigolejeunea anguiformis, Metzgeria furcata, Porella elegantula, Radula spp., and more occasionally, Bazzania nova, B. hochstetteri, Chiloscyphus muricatus and Heteroscyphus cymbaliferus.

The bryophyte flora of high tree canopies has not been studied; it is dominated by dry-tolerant epiphytic mosses such as Papillaria flavolimbata, Macromitrium spp., and the three Dicnemon species (D. calycinum, D. dixonianum and D. semicryptum). However, some liverwort species found in this habitat in Westland lowland forest canopies and at times elsewhere are Acrobolbus concinnus, Bazzania nova, Chandonanthus squarrosus, Cuspidatula monodon, Drepanolejeunea aucklandica, Diplasiolejeunea plicatiloba, Lepidolaena palpebrifolia, L. taylorii, Paraschistochila tuloides, Plagiochila deltoidea, Plagiochilion conjugatum and Pseudomarsupidium piliferum.

A distinctive epiphytic twig bryophyte flora is present in New Zealand, particularly at the margins of lowland forests. Coprosma species have the richest twig bryophyte flora, particularly C. propinqua, C. rhamnoides, C. rotundifolia and C. tayloriae. Austrolejeunea olgae, Austrometzgeria saccata, Cololejeunea spp., Colura pulcherrima var. bartlettii, Dinckleria pleurata, Diplasiolejeunea plicatiloba, Drepanolejeunea aucklandica, Harpalejeunea latitans, Lejeunea species, Lopholejeunea colensoi, Metzgeria furcata, M. scobina, M. flavovirens and Siphonolejeunea nudipes are common twig epiphytes of damp forest margins, and often are found with mosses such as Crosbya straminea, Ephemeropsis trentepohlioides and Tetraphidopsis pusillus. In drier, exposed habitats Frullania, and to a lesser extent Radula, are the only liverwort genera that compete with the more dry-tolerant mosses, while in exposed but high-rainfall situations Jubulopsis novae-zelandiae and Nephelolejeunea conchophylla are rare endemics, found mainly on twigs of Leptospermum and Metrosideros (both Myrtaceae).

Variation in the epiphytic liverwort flora between host species is quite noticeable, but insufficient collections recording the host have been made to quantify the differences. For example, Melicytus ramiflorus and Pseudowintera colorata and some Coprosma species are favored hosts, particularly for twig epiphytes, as well as Metrosideros, both vine and tree species, and other Myrtaceae genera such as Kunzea, Leptospermum and Neomyrtus, in addition to all of the native conifer species despite their peeling or flaking bark. However, a wide variety of tree species are also recorded as hosts.

Tree ferns (Cyathea and Dicksonia) are a conspicuous element of the understory in New Zealand forests of all types, particularly in canopy light gaps. Symphyogyna subsimplex is the only liverwort species confined to tree fern caudices; Bazzania tayloriana and Chiloscyphus helmsianus are mainly found on tree fern caudices, especially on their wide bases, and Bazzania nova, Podomitrium phyllanthus, Psiloclada clandestina, Schistochila pluriciliata and Zoopsis argentea are sometimes found in this habitat.

Epiphylls

Epiphyllous taxa occur only in high-humidity environments, especially on leaves or fern fronds close to the ground near streams. A number of biotic and abiotic factors impact the epiphyllous condition, such as high atmospheric humidity, leaf longevity and light (see Corley et al., 1993; Gradstein and Pócs, 1989; Richards, 1984; von Konrat and Braggins, 1999a). For a discussion of life history strategies and epiphyllous liverworts see Longton and Schuster (1983) and Schuster (1991). Among the ferns and fern allies, the main hosts of liverwort epiphylls are of the genus Blechnum, such as B. nigrum and B. colensoi (both ferns of very damp shady environments), B. chambersii, B. filiforme, B. discolor, B. minus, B. novae-zelandiae, B. procerum, Hymenophyllum and Trichomanes species, Polystichum vestitum, and rarely Microsorum, Lycopodium and Tmesipteris. Pseudowintera colorata is the most notable of the flowering plant hosts for epiphylls and may have complete liverwort cover on some leaves, and among the strap-leaved monocots, the most notable is Freycinetia baueriana. Epiphylls have been recorded on the following tree species: Alectryon excelsus, Ascarina lucida, Beilschmiedia tawa, Knightia excelsa, Laurelia novae-zelandiae, Nothofagus menziesii and the palm Rhopalostylis sapida. These fern and flowering plant hosts, with the exception of Nothofagus, are also the main hosts of foliicolous lichens. This suggests that a lack of tannins and a less waxy leaf surface may better enable spores and plants to adhere and become established, and may be partial explanations for their frequency as hosts.

Most genera of Lejeuneaceae can be found as epiphylls, but some species of Frullania (see von Konrat and Braggins, 1999a), Lepidolaena, Metzgeria, Riccardia and Telaranea can be epiphyllous. Kymatolejeunea bartlettii and Nephelolejeunea papillosa have only been found among the leaves of the large moss Dendroligotrichum dendroides, which is also commonly a host for Colura saccophylla and C. pulcherrima (but these two species also occur as twig epiphytes). Echinolejeunea papillata is found almost exclusively on fern fronds (Blechnum, Hymenophyllum, Trichomanes, Grammitis and Polystichum).

Cololejeunea hodgsoniae is often found on mosses and larger liverworts such as Plagiochila stephensoniana on lowland forest stream margins, and Cheilolejeunea campbelliensis is most commonly found on mosses and liverworts such as Breutelia elongata and Clandarium xiphophyllum in penalpine scrub.

9. Distribution and range of liverwort species within New Zealand

Liverworts appear to be somewhat less climatically limited in their range than the vascular plants of New Zealand. Only one liverwort species spans the range between the Kermadec Islands and Campbell Island: Marchantia berteroana. Marchantia foliacea ranges from the Kermadec to Auckland islands. However, a large number of liverwort species are present from the subantarctic islands to Northland; a greater proportion of the liverwort flora has this wide range than does the vascular flora. A number of species of Jungermanniideae have a broad range in New Zealand, notably Chandonanthus squarrosus, Cuspidatula monodon, Jamesoniella colorata, J. kirkii, J. tasmanica (all span Auckland Islands to Northland), but other very wide-ranging species are Bazzania adnexa, B. involuta, Chiloscyphus bispinosus, Drepanolejeunea aucklandica, Harpalejeunea latitans, Heteroscyphus coalitus, Hymenophyton flabellatum, H. leptopodum, Lamellocolea granditexta, Lejeunea primordialis, Lembidium nutans, Lepidolaena taylorii, L. clavigera, Lepidozia concinna, L. laevifolia, L. microphylla, Lopholejeunea colensoi, Metzgeria furcata, Monoclea forsteri, Pallavicinia xiphoides, Paraschistochila pinnatifolia, P. tuloides, Plagiochila gigantea, P. intertexta, P. radiculosa, Podomitrium phyllanthus, Porella elegantula, Psiloclada clandestina, Radula grandis, R. marginata, R. physoloba, R. silvosa, R. uvifera, Riccardia aequicellularis, R. cochleata, R. crassa, Saccogynidium australe, Tylimanthus tenellus and Zoopsis argentea. Fewer than average of these species (29%) are endemic to New Zealand, as might be expected, but Monoclea forsteri and Lembidium nutans (Campbell Island to Northland) are notable exceptions; both are very common and widespread New Zealand endemics. Many liverwort species have more restricted distributions, and in general two distribution types are the most common. First are southern species that have their northern limit in the South Island or, more commonly, in the North Island mountains. The second, less common distribution type consists of northern species that have their southern limit in the North Island or in northern or western lowlands of the South Island.

Cool-temperate species that are found almost exclusively in the South Island and further south and have a northern limit that may be limited by climate are Acrobolbus cinerascens, Acromastigum mooreanum, A. verticale, Archeophylla schusteri, Castanoclobos julaceus, Clandarium xiphophyllum, Cryptochila acinacifolia, Eoisotachis stephanii, Eotrichocolea polyacantha, Isolembidium cucullatum, Isophyllaria attenuata, Jubulopsis novae-zelandiae, Kurzia tenax, Lepidozia ulothrix, Megalembidium insulanum, Metzgeria rigida, Pachyglossa tenacifolia, Radula ratkowskiana, R. tasmanica, Schistochila lehmanniana, Telaranea tetradactyla, T. pulcherrima and T. tuberifera. Most of these species can be found near sea level in the south of New Zealand but are confined to the penalpine zone at their northern limit, which is in the Ruahine and Tararua ranges for some (e.g., Diplophyllum domesticum var. icari var. icari, Lepidozia ulothrix, Metahygrobiella drucei, Neohodgsonia mirabilis, Pachyschistochila parvistipula, Paraschistochila conchophylla and Phyllothallia nivicola), at Mt. Taranaki for others (e.g., Cheilolejeunea campbelliensis, Clasmatocolea strongylophylla, Hygrolembidium australe, Isotachis intortifolia, Lepidolaena reticulata, Plagiochila ramosissima, P. strombifolia, Telaranea quadriseta) or at Mt. Ruapehu (e.g., Archeophylla schusteri, Cephalomitrion aterrimum, Cryptochila pseudocclusa, Eotrichocolea polyacantha, Herzogobryum teres, Isophyllaria attenuata, Lepidozia ornata, Nothogymnomitrion erosum, Pseudolophocolea denticulata and Telaranea trilobata). Some southern species have their northern limit in the middle of the South Island, particularly in the mountains of Western Nelson EP. Examples are Telaranea meridiana, Wettsteinia schusteriana (southern end of Paparoa Range) and Cryptochila acinacifolia (Mt. Euclid in the central Paparoa Range).

Warm-temperate species that are mainly found in the North Island but can be found in Nelson and lowland parts of Westland are Acrolejeunea allisonii, A. mollis, Asterella australis, Bazzania tayloriana, Chaetophyllopsis whiteleggei, Chiloscyphus echinellus, Cololejeunea hodgsoniae, C. pulchella, Drucella integristipula, Geocalyx caledonicus, Jackiella curvata, Marchantia macropora, Mastigolejeunea anguiformis, Nephelolejeunea hamata, Paracromastigum furcifolium, Plagiochilion prolifer, Radula papulosa, Ricciocarpos natans, Telaranea gibbsiana and Trichocolea hatcheri.

Many tree species of northern New Zealand have their southern limit between Kawhia Harbour and East Cape, e.g., Agathis australis. Some liverworts seem similarly limited and include a subtropical element composed of Cololejeunea cardiocarpa, Dumortiera hirsuta, Harpalejeunea filicuspis and Leptolejeunea elliptica but also the endemics Frullania wairua, Telaranea granulata, T. perfragilis and Stenolejeunea acuminata, and the epiphytic hornwort Dendroceros granulatus. Von Konrat et al. (2006a) pointed out that the New Zealand distribution of Frullania chevalieri, a species otherwise known from New Caledonia, appears to be somewhat correlated with that of Agathis australis forest, and that it would be of interest to determine whether the southern range of F. chevalieri is the same as that forest type.

The following liverwort species show a disjunction between Northland EP (especially Waipoua) or Auckland EP (especially Coromandel Peninsula, Great Barrier or Little Barrier islands) and Westland–Fiordland–Rakiura EPs: Acrochila biserialis, Acromastigum cavifolium, A. marginatum, Brevianthus flavus, Goebelobryum unguiculatum, Lembidium longifolium, Lepidozia setigera, Mnioloma novae-zelandiae, Zoopsis ceratophylla and Z. nitida. The moss Pulchrinodus inflatus shows the same disjunction, as do a few vascular species such as Metrosideros parkinsonii. The reasons for this disjunction seem to be in part climatic and in part a disjunction in habitat such as pakihi/gumlands.

A number of alpine species are present in the South Island mountains but absent from Stewart Island or the subantarctic islands: Acrolophozia pectinata, Anthelia juratzkana, Austrolophozia paradoxa, Bragginsella anomala, Diplophyllum gemmiparum, D. verrucosum, Gymnomitrion cuspidatum, G. strictum, Herzogobryum filiforme, Heteroscyphus mononuculus, Isotachis intortifolia, I. olivacea, I. plicata, Marsupella sparsifolia, M. sprucei, Pachyschistochila altissima, P. berggrenii, P. childii, P. nivicola, P. parvistipula, P. succulenta, P. trispiralis, P. virescens, Ptilidium ciliare, Radula sainsburiana, Scapania nemorosa and S. undulata.

Some of these species (Anthelia juratzkana, Ptilidium ciliare, both species of Scapania and both species of Marsupella) are bipolar in range; they are found in the Northern Hemisphere in temperate to arctic conditions, and are pre-adapted to the alpine zone. There is a lack of obvious radiation in the alpine zone liverwort flora, due no doubt to the recency of the Southern Alps, and speciation seems to have been notably slower than in vascular plants, where major radiations have occurred (e.g., Aciphylla, Celmisia, Gentianella and Ranunculus). Only in Pachyschistochila can such a radiation be seen.

There are no known liverwort disjunctions corresponding to the beech gap in Westland between the Taramakau and Karangarua rivers.

Some species are known from a single region; examples are Cololejeunea falcidentata, Frullania wairua and Schistochila nitidissima of Northland, Schistochila pellucida of the Volcanic Plateau, and Pachyschistochila papillifera of the Ruahine Range. The complete range of such local endemics is not known and is likely to have been reduced by habitat loss. A much more comprehensive knowledge of the liverwort flora is needed to understand patterns of local endemism.

10. Rarity and conservation

New Zealand has its own threatened species classification (Molloy et al., 2002) developed by the Department of Conservation, which has the role of advocacy and management of threatened species. The classification is a two-tiered system:

Extinct : no reasonable doubt exists after repeated survey that the last individual has died. No New Zealand liverworts are in this category.

Acutely threatened (taxa that face a high risk of extinction in the wild):

Nationally critical : very small population or a very high predicted decline.

Nationally endangered : small population and moderate to high, recent or predicted decline.

Nationally vulnerable : small to moderate population and moderate decline.

Chronically threatened (taxa that face extinction but are buffered by a large total population or slow rate of decline):

Serious decline : moderate to large population and moderate to large decline.

Gradual decline : moderate to large population and small to moderate decline.

At risk (not presently facing a risk of extinction, but could be if new threats depleted their populations):

Sparse : naturally or unnaturally rare but not restricted to a small geographical area.

Range restricted : naturally or unnaturally rare, restricted to a small geographic area, a specific habitat or substrate.

Data deficient : information is too poor to assign them to one of the categories above.

The Department of Conservation maintains threatened species lists of most groups of organisms, including bryophytes. Listed species have no statutory protection. As of 2005 it contained 157 liverwort species, 41 of which were in the nationally critical category (Glenny and Fife, 2005). The composition of the list is not yet stable because of insufficient knowledge of distributions of the rarer species, the active state of New Zealand liverwort taxonomy and the recency of the list’s compilation.

11. Guidance to users of the flora

Geographical scope

The geographical coverage of the flora includes the three main islands of New Zealand, i.e., Stewart Island, South Island and North Island. It includes what are often termed the "outlying islands": Campbell, Auckland, Antipodes, Bounty, Snares, Chatham and Kermadec Islands, as well as all the "offshore islands" such as Three Kings Islands, Great Barrier Island, Little Barrier Island, Solander Island, Muttonbird Islands and Whenua Hou (Codfish Island) (Map 1). The flora excludes Macquarie Island, which although considered part of the New Zealand region, is politically part of Australia and is included in the Flora of Australia.

Intent of the flora

The intent of the flora is to treat all New Zealand liverworts and hornworts in a manner that is useful to a broad audience, e.g., to professionals, students, amateurs and conservation organizations. The treatment of each species includes a section on synonymy, a description of the plant, statements of world distribution, with more detailed comments of range within New Zealand, as well as a discussion of ecology. For ease in identification, full or nearly full descriptions are provided. Many groups have been subjected to study for the first time, and the flora therefore contains much information that is presented here for the first time. Likewise, many taxa have gametangia, sporophyte and oil-body detail described and often illustrated for the first time. Comments on extra-territorial taxa are typically included in the relevant taxonomic category, i.e., as part of discussions of families or genera.

Classification

The sequence of orders and families presented in this flora is based for the most part on the classifications of Schuster (1979a, 1984e, 2000a, 2002a). However, moderate to considerable variants are made in this treatment, e.g., the recognition of Megalembidium in its own subfamily juxtaposed to subfam. Lembidioideae, the placement of Neogrollea in its own suborder and the position of Trichotemnomaceae.

Within genera, the order in which species are treated follows a sequence from what we consider to be "most generalized" to what is considered to be "most derived."

A number of laboratories have been working on molecular phylogenetic studies of liverworts and hornworts in the past five years. The first molecular surveys of large numbers of families and genera reported results quite different from traditional classification schemes, but more careful and complete surveys at the same level are now reporting phylogenies that are very similar to traditional classifications. This gives confidence that traditional classifications are probably basically correct. It is a little too soon to modify traditional classifications with the results of molecular studies, and we have opted to maintain a more traditional approach.

Author citations follow the International Plant Names Index (IPNI, 2006, http://www.ipni.org/index.html) but G. L. Merrill is cited as "Merrill" instead of G.L.Sm. Journal citations follow the abbreviations in Botanico-Periodicum-Huntianum (B-P-H). New synonymy is noted by the designation "syn. nov."

Taxon names at all ranks have been italicized except where they appear as headings and accepted names in the Index. This is in accordance with the style adopted in the International Code of Botanical Nomenclature and recommended in the preface (McNeill et al., 2006, p. viii).

Types

Nomenclature and literature citations are included for each species, including all homotypic synonyms; this is followed by citation of the type. We have not used the term "holotype" for species described in the earlier literature, e.g., of the many species originally described by Hooker, Taylor, Colenso or Stephani. These individuals worked in a period prior to the establishment of the current type method, and, in our view, the terms holotype and isotype should not be applied to the original material on which their taxa were based. It should be emphasized that the terms holotype, lectotype, syntype and neotype, etc., were used in the International Code for the first time in 1952 (see Lanjouw, 1952). We have opted to utilize the more general term "type" for taxa in the earlier literature (see also Engel, 1978). If an author has merely said "type" when describing a species after 1952, we have used the term "holotype" if the location and identity of the type is clearly given. Herbarium citation is included only in those cases where we have examined a type; in cases where we have not studied the type, no herbarium designation is provided. Lectotypification is not attempted in the flora, a few instances excepted.

Materials used

The senior author has consulted specimens from numerous herbaria, the most significant being BM, CHR, FH, G, JE, NY, S and W. The junior author has examined all specimens at CHR and WELT except for species where large numbers precluded study of all of them.

Species descriptions are nearly all based solely on New Zealand material. Measurements of leaf cells are based on multiple measurements each comprised of a series of cells (ca. 2–5) per measurement rather than a measurement of length and width of an individual cell.

Dimensions are all given as width × length.

Oil-bodies

Oil-bodies are illustrated and data included for at least one species of each genus. Oil-bodies were measured and digitally photographed in New Zealand from fresh material collected by Glenny or Engel. Specimens of Glenny collections were then sent in the fresh state to Chicago for preparation of the figures, using the digital photographs as a reference for the fresh state.

We have attempted to categorize oil-body types based upon a number of criteria. In some cases, however, oil-bodies do not clearly and distinctly fit into a category and placement is somewhat subjective. This is particularly the case with boundaries between the "coarsely granular" and "finely papillose" types and between the "coarsely papillose" and "finely botryoidal" types. We have attempted to be consistent in adhering to the category definitions and characteristics outlined below. Examples of oil-body types are included within each category.

granular : appearing like fine or coarse grains of sand within a membrane; the spherules do not protrude from the relatively firm surrounding membrane; the oil-bodies totally lack segmentation (see ExamplesA, Acrobolbus concinnus andB, Archeophylla schusteri). Breakup occurs by swelling of the entire oil-body followed by a burst of the "grains" through the membrane (or a collapse of the membrane); the "grains" are then free within the cytoplasm (i.e., the "grains" or spherules do not coalesce).

papillose : oil-bodies with discrete segmentation between small (finely papillose) to large (coarsely papillose) spherules, the spherules protruding from the membrane, which is uneven in profile (ExamplesC, Adelanthus falcatus;D, Acrochila biserialis; andE, Pseudomarsupidium piliferum). Breakup is by way of coalescing of globules, which, with age, become increasingly large, and the internal segmentation gradually fewer.

botryoidal : with the characteristics of the "papillose" condition, but the protrusions greater, the globules larger and the segmentation distinct (ExamplesF, Mnioloma novae-zealandiae;G, Acromastigum mooreanum; andH, Leptoscyphus sp.); the coarsely botryoidal oil-bodies each appear like a cluster of grapes. Breakup is similar to the papillose condition.

homogeneous : either with no internal segmentation or with isolated incomplete or complete segments (e.g., Bazzania spp.), and the outer surface is smooth (ExamplesI, Bazzania tayloriana andJ, Herbertus oldfieldianus). Oil-bodies may have some internal differences in "cloudiness," probably due to differences in oil viscosity.

minute and numerous : the "Schrot-Korn Typus" of Müller (1939); oil-bodies are homogeneous and commonly only a few microns in diam. (ExampleK, Metahygrobiella drucei).

oil-bodies dimorphic : presence of two different oil-body types within the same cell, e.g., Trichotemnoma corrugatum has oil-bodies in all leaf cells, but dimorphic within each cell: one type occurs 2–4 per median leaf cell, burnt orange-brown, large, coarsely granular to finely papillose and irregular in shape; the other type is hyaline, numerous per cell, ± spherical, homogeneous, smooth, much smaller (only 1–1.9 µm), and occurs in the outer part of the cytoplasm (ExampleL, Trichotemnoma corrugatum). The size of the "grains" or "spherules" is much different, i.e., the "spherules" in the outer part of the cytoplasm do not originate from breakdown of the much larger other type.

cells dimorphic : presence of oil-bodies in some cells, while other cells completely lack them.

oil-bodies absent : We have placed species in this category if oil-bodies are lacking altogether (e.g., Megalembidium) or if only scattered minute lipid droplets are present (e.g., ExampleM, Drucella).

In some cases oil-body color is difficult to establish with certainty. Gradations occur from a hyaline, glistening condition to a smoky-grey, dull and opaque condition. Extremes of each condition are very distinct with regard to color and brightness. Intermediates, however, are more difficult to adequately and definitively describe. The color of brownish oil-bodies is often especially distinct at lower magnifications (e.g., at 200× or 400×).

The type of fungal component present in a New Zealand liverwort is included and placed at the end of the gametophyte description. The data have been kindly provided by Jeffrey Duckett, and details of some of the liverwort–fungi associations may be found in the account in Section 12 below and in papers cited there.

Bibliography

The bibliography section in this volume attempts to be comprehensive and includes references that we consider of interest to the student of New Zealand hepatics. It includes references cited in this volume as well as those that are not. The Literature Cited sections for volumes 2 and 3 will include only references cited in those volumes.

12. Endophytic fungi in New Zealand liverworts Jeffrey G. Duckett and Roberto Ligrone

Introduction

Liverworts contain a wide range of endophytic fungal associations matching mycorrhizas in higher plants. Although the existence of these associations had been known since the nineteenth century, it is only over that past 25 years or so that the groups of fungi have been identified by electron microscope studies (Duckett et al., 1991, 2006a) and most recently the fungal genera by molecular sequencing (Bidartondo et al., 2003; Kottke et al., 2003; Russell and Bulman, 2005; Upson et al., 2007). While axenic culturing of both partners in these associations and reinfection experiments are providing data on the host ranges of the fungi (Duckett and Read, 1995; Duckett et al., 2006a; Pressel et al., 2008) that in some cases extend to adjacent vascular plants, their functional significance has yet to be investigated (Read et al., 2000; Selosse, 2005).

The present account is based on detailed sampling of all the different major groups of liverworts collected during two visits to New Zealand in 1999 and 2001. All the specimens were examined in the living condition by light microscopy immediately after collection, whereas key taxa were subsequently studied by electron microscopy. The overall distribution of fungi in New Zealand liverworts closely mirrors that from earlier surveys of Northern Hemisphere taxa. However, this survey also provides new information on numerous taxa whose fungal status was previously unknown, as this information is completely lacking in Allison and Child (1975). Although our sampling included over 90 genera, it was far from exhaustive. Nevertheless, the general overview presented here is unlikely to change radically, although observations on additional taxa will doubtless reveal some surprises, particularly among the Jungermanniales.

The distribution of fungi across New Zealand liverworts presents a remarkably complete snapshot of the distribution of these associations worldwide; the only notable absences from New Zealand being the Blasiales and Sphaerocarpales (both fungus-free), the Pelliineae (typical glomeromycotes), the Arnelliaceae (distinctive basidiomycetes) and the Pleuroziaceae (fungus-free).

Marchantiales

All the New Zealand members of the Marchantiales, including Neohodgsonia and Monoclea but with the exception of the Ricciaceae, which lack endophytes (as is the case in the rest of the world), contain associations with glomeromycote fungi cytologically very similar to Paris-type vesicular arbuscular mycorrhizas formed by higher plants (Ligrone et al., 2007). The fungi occur in specific ventral cell layers in the midrib region throughout the thalli and in the smooth but not the pegged rhizoids. The endophytes do not, however, extend into the growing apices of the thalli and are also absent from the oil-body–containing cells, from around the sex organs and placental regions, and from the sporophytes. Hyphae have multiple entry sites into the rhizoids, from which they spread into the thallus cells proper. The fungi are entirely intracellular, with large infecting hyphae (Fig. 2a; trunk hyphae) spreading directly through the walls from cell to cell where they form numerous arbuscules comprising much-branched fine hyphae (Fig. 2a) and occasional multinucleate smooth-walled vesicles (Fig. 2b). Ubiquitous in the glomeromycote infections in liverworts (Duckett et al., 2006b) and also in pteridophytes, including both the gametophytes and sporophytes of Psilotum and Tmesipteris in New Zealand (Duckett and Ligrone, 2005), is the presence of gram-positive-like bacteria within the fungi (Fig. 2d).

Fungus colonization causes major changes in the host cells, including proliferation of the cytoplasm mitochondria and endoplasmic reticulum, replacement of a large central vacuole by numerous small ones, migration of the nucleus from a peripheral to a central location and a dramatic reduction in the starch content of the plastids. In later infection stages the arbuscules degenerate (Fig. 2c), and frequently a second colonization cycle can be seen.

This scenario for the fungi found in New Zealand Marchantiales is very much in line with expectation from previous information on Northern Hemisphere taxa (Read et al., 2000). The recent demonstration of a common endophyte (Glomus species) in New Zealand Marchantia foliacea and Podocarpaceae (Russell and Bulman, 2005) now invites experiments on possible nutrient sharing via the fungus as has been demonstrated previously between Cryptothallus and Betula in the Northern Hemisphere (Bidartondo et al., 2003).

Metzgeriales

The present survey produced some novel and highly distinctive results for the Metzgeriales. Strikingly similar to the marchantialean glomeromycote infections are the associations found throughout the families of the Metzgeriales now grouped together as Metzgeriidae clade I, viz. Fossombroniaceae, Allisoniaceae, Pallaviciniaceae and Hymenophytaceae (Crandall-Stotler et al., 2005; Forrest and Crandall-Stotler, 2004, 2005; Ligrone et al., 2007). Whereas in the first two families, as in the Marchantiales, they occur right up to the thallus apices within specific ventral cell layers, in the last two they are restricted to the subterranean rhizomatous axes lacking lateral laminae (Ligrone et al., 2007). The widespread occurrence of fungi in these axes has not been noted previously, indeed Hymenophyton was previously considered fungus-free (Nebel et al., 2004). In Metzgeriidae clade II fungi are absent from the basal genus Pleurozia (not present in New Zealand) and from Phyllothallia and Metzgeriaceae, while the Aneuraceae contain basidiomycetes (Fig. 3c) whose cytology, like that in Northern Hemisphere Aneura pinguis and Cryptothallus, is almost identical to orchidaceous mycorrhizas (Ligrone and Duckett, 1993). The fungi form hyphal coils within the host cells (Fig. 3b) that are then digested by the host (Fig. 3a). Multiple infection cycles are frequent. In each taxon the fungi occupy highly specific layers in the ventral region of the thallus. In Verdoornia all the cells in the ventral half of the thalli are infected (Fig. 3a, b), as is also found in Aneura pinguis in the Northern Hemisphere. In New Zealand Aneura species, however, the endophytes are restricted to 2–3 ventral cell layers (Duckett and Ligrone, 2008). In contrast to Aneura the vast majority of New Zealand Riccardia species are fungus-free, as is the case for all the European members of the genus. Exceptions are R. pennata, with prominent hyphae restricted to the ventral epidermal cells, and R. intercellula, where uniquely the hyphae lie within the inner walls of the ventral epidermis (Brown and Braggins, 1989). These highly distinctive fungal distributions are invaluable aids in identification.

Jungermanniales

Glomeromycote infections are unknown in leafy liverworts, and in New Zealand most families lack fungi, as is the case in the rest of the world. A few families, however, contain basidiomycetes, with ascomycetes being distributed more widely. The former associations would appear to be more restricted in the New Zealand flora compared with the Northern Hemisphere (Duckett et al., 2006a), and the only genuine examples to date are Diplophyllum (Fig. 3d, e) and Lophozia. A distinctive feature of these associations is their occurrence as a mosaic of infected and uninfected cells within the central cells of the stems and the absence of hyphal digestion by the hosts. Further studies are now needed to identify putative basidiomycetes in other genera in the Scapaniaceae and in Lethocolea pansa, where the purple-walled cells occupying the ventral half of the stems are packed with septate hyphae. Why many other leafy liverworts with swollen stems that grow on apparently highly suitable substrata, often in very close proximity to glomeromycote-containing thalloid genera, are fungus-free is a mystery.

Ascomycetes occur in the swollen rhizoids in most New Zealand genera in the Lepidoziaceae, Calypogeiaceae, Cephaloziaceae and Cephaloziellaceae. They are particularly prominent on flagelliform axes (also see Duckett et al., 1991) that may extend for tens of centimeters into peaty substrata. Indeed the prominence of these axes is perhaps one of the most distinctive features of New Zealand Lepidoziaceae. Whether the branched rhizoids in Acromastigum and Bazzania, which are also most prominent on flagelliform branches but rarely extend deep into the substratum, also form a functional relationship with fungi or are colonized only after death urgently requires investigation. The same is also true for the rhizoids on the rhizome systems in the Adelanthaceae. Swollen, fungus-packed rhizoids of unconfirmed identity are also a characteristic feature of the Balantiopsaceae, in contrast to the fungus-free rhizoids in members of the Jungermanniaceae that often grow with them.

The most surprising discovery of the present survey was that the highly distinctive branched and septate rhizoids in the Schistochilaceae are packed with fungi (Pressel et al., 2008). Electron microscopy identified the fungus as an ascomycete confirmed from sequencing as Rhizoscyphus ericae, the mycobiont in the Ericales. Culturing experiments have demonstrated that rhizoid branching and septation is in fact induced by this fungus, whereas apical branching without divisions in the rhizoids of other liverworts occurs as a consequence of contact with solid substrata.

Haplomitrium and Treubia

The glomeromycote associations found in Haplomitrium and Treubia are unique among liverworts and vascular plants (Carafa et al., 2003; Duckett et al., 2006b). Indeed Treubia contains one of the most highly differentiated symbiotic fungal associations in land plants. Unlike other liverworts in which thallus differentiation is dependent on cells of different sizes, wall thickness and the conducting strands, the unique anatomy of Treubia axes depends almost entirely on different host-fungus zones. Though differing in a few details, a suite of commonalities closely links the fungal associations of the two groups. Haplomitrium and Treubia are unique among liverworts in secreting huge quantities of mucilage. This is exuded directly from the epidermal cells lining the leafless subterranean axes in Haplomitrium, and from clefts between these cells from mucilage-filled intercellular spaces within the thalli in Treubia. Unlike the Golgi body–derived secretions from apical mucilage papillae in other liverworts, the mucilage in Haplomitrium and Treubia is a direct product of the endoplasmic reticulum. Within the mucilage spaces in the center of the Treubia thallus, the fungus proliferates to form pseudo-parenchymatous hyphal masses (Fig. 4c) and, in older thalli, highly distinctive spores with thick multilayered walls (Fig. 4d). The latter also occur within the mucilage surrounding the subterranean axes in Haplomitrium. Hyphae are obviously not associated with rhizoids in Haplomitrium, which lacks this feature, and Treubia is the only fungus-containing liverwort in which hyphae are absent from the rhizoids. In both genera the endophytes enter the epidermal cells directly, and in both the fungus has a very similar intracellular phase. In Haplomitrium the epidermal and subepidermal cells are packed with arbuscule-forming hyphae and short-lived fungal lumps (Fig. 4a) clearly distinct from the vesicles in other liverworts and vascular plant glomeromycotes. Exactly the same fungal lumps occur in the ventral cell layers in the Treubia thallus below the extracellular fungal zone (Fig. 4b). It is interesting to recall that in his original meticulous description of the endophyte in Treubia, Goebel (1891) speculated that the inter- and extracellular phases might be different fungi. Electron microscopy over a century later was needed to demonstrate hyphal continuity between the two, thus proving these to be one and the same fungus.

Ecological considerations

Correlations between ecology and the presence or absence of fungi in New Zealand liverworts also mirrors the situation in rest of the world (Duckett et al., 2006a). Leafy taxa with rhizoidal ascomycetes are most often found on waterlogged, humus-rich substrata, particularly peaty soils and rotting wood, whereas thalloid taxa are most frequently found growing on mineral soils. Few or no epiphytes contain endophytes (e.g., they are unknown in the Lepidolaenaceae, Porellaceae, Frullaniaceae and Lejeuneaceae), and the same is true of species growing directly on bare rocks, e.g., Gymnomitriaceae. Turning to anatomy, fungi are generally lacking in taxa with highly thickened stem cell walls and with erect stems bearing few or no rhizoids, e.g., Herbertus, Triandrophyllum, Trichocolea, Ptilidiinae and Lepidolaenaceae. Perhaps the most surprising fungus-free taxa, given their apparently suitable ecology and anatomy, are Goebelobryum, Jamesoniella, Marsupidium and Solenostoma.

Evolutionary considerations

A particularly interesting aspect of the liverwort and hornwort fungus associations is the new insights they are providing in liverwort classification and phylogeny and in the evolutionary origins of plant symbioses. Such considerations have become all the more significant since the almost unanimous recognition that liverworts are the earliest divergent clade of land plants (Renzaglia et al., 2007) and, more surprisingly, new molecular (Groth-Malonek et al., 2004) and immunocytochemical evidence (Carafa et al., 2005) positions the hornworts as sister to vascular plants. When considered against traditional liverwort classifications, the distribution and evolution of the fungal associations seem to make little or no sense with an almost random presence or absence of glomeromycetes, basidiomycetes or ascomycetes.

The twenty-first century has witnessed the first detailed molecular and total evidence liverwort phylogenies, and with these have come major advances in our understanding of the liverwort phylogenetic tree. The topology of the tree shows Treubia and Haplomitrium closely nested in a clade sister to all other land plants. It is perhaps not surprising therefore that their fungal associations share so many common features. The absence of fungi in Blasiales and Sphaerocarpales, the two clades immediately above Treubia and Haplomitrium and sister to all other liverworts, suggests the loss of the primeval associations and the secondary acquisition of the typical glomeromycote associations found throughout the Marchantiopsida and clade I in the Metzgeriidae. The topology of recent liverwort phylogenies (Forrest and Crandall-Stotler, 2005) strongly suggests that these typical associations with glomeromycotes long predated arbuscular mycorrhizas and that those in Haplomitrium and Treubia are the most ancient of all extant land plant–fungal symbioses.

The ascomycetes in leafy liverworts may then be considered to be a secondary acquisition of more recent origin following the loss of the primitive glomeromycote associates. Recent dating of the Schistochilaceae, the sister family to all other fungus-containing leafy liverworts (Heinrichs et al., 2007), in the Triassic 250 Mya suggests these long predated ericalean mycorrhizas, their vascular plant counterpart, dated at 106–114 Mya (Wikström et al., 2001), and whose endophytes they share (Read et al., 2000).

In leafy liverworts basidiomycete associations are most likely another example of secondary acquisition following the loss of ascomycetes. But, here again, the identification of their fungi as members of the Sebacinaceae (Kottke et al., 2003) suggests ancient origins. In contrast, the identification of the basidiomycetes with very different cytology in the Aneuraceae as Tulasnella (Bidartondo et al., 2003; Kottke et al., 2003; Nebel et al., 2004) indicates these to be much more recent. This finding is very much in line with this family forming the crown group in Metzgeriidae II (Crandall-Stotler et al., 2005). The final revelation from marrying fungi and phylogeny is the presence of a typical aneuracean basidiomycete in Verdoornia. This monotypic New Zealand genus, originally regarded as isolated and primitive (Schuster, 1964d, 1999d), is now nested within the Aneuraceae (Crandall-Stotler et al., 2005).

Mosses and hornworts

In concluding this account it must be emphasized that one of the most striking yet not widely cited differences between liverworts and mosses is that, in the latter, mutualistic fungal associations of any kind are unknown (Renzaglia et al., 2007). In hornworts, however, glomeromycote associations that are cytologically closely similar to those in thalloid liverworts have been noted in a few species, although their overall distribution requires further study (Read et al., 2000). The picture from New Zealand may well turn out to reflect the situation worldwide. We found the invariable presence of glomeromycote infections to be identical to those described from the Northern Hemisphere Phaeoceros carolinianus (Ligrone, 1988): in Anthoceros laminiferus, Phaeoceros coriaceus and P. hirticalyx. These are absent in Dendroceros validus and D. granulatus, mirroring the absence of fungi in epiphytic liverworts. Endophytes are also absent from the three New Zealand Megaceros species examined here (M. denticulatus, M. pellucidus and M. giganteus), again paralleling fungus-free liverworts growing in very wet habitats. It is noteworthy that the basal genus Leiosporoceros apparently lacks an endophyte (Villarreal and Renzaglia, 2006). Although at present highly provisional, when superimposed on the most recent hornwort phylogeny (Renzaglia et al., 2007), the endophyte distribution strongly suggests secondary acquisition of glomeromycetes independently of those in thalloid liverworts.

Genera examined

The order of genera follows Glenny (1998). A, ascomycete; B, basidiomycete; G, glomeromycote; ?, fungus of unknown identity; NP, fungus not present.

Anthoceros G, Dendroceros NP, Megaceros NP, Phaeoceros G, Haplomitrium G, Lepicolea NP, Herbertus NP, Triandrophyllum NP, Trichotemnoma NP, Archeophylla NP, Temnoma NP, Trichocolea NP, Acromastigum?, Bazzania?, Chloranthelia A, Hygrolembidium A, Isolembidium A, Kurzia A, Lembidium A, Lepidozia A, Megalembidium A, Neogrollea A, Pseudocephalozia A, Psiloclada A, Telaranea A, Zoopsidella A, Zoopsis A, Calypogeia A, Adelanthus?, Wettsteinia?, Cephalozia A, Cephaloziella A, Anthelia NP, Anastrophyllum NP, Chandonanthus NP, Cryptochila NP, Jamesoniella NP, Jungermannia NP, Lophozia B, Gymnomitrion NP, Herzogobryum NP, Marsupella NP, Diplophyllum B, Chiloscyphus NP, Clasmatocolea NP, Heteroscyphus NP, Leptoscyphus NP, Saccogynidium NP, Plagiochila NP, Plagiochilion NP, Acrobolbus NP, Goebelobryum NP, Lethocolea?, Marsupidium NP, Tylimanthus NP, Pachyschistochila A, Paraschistochila A, Schistochila A, Balantiopsis?, Eoisotachis?, Isotachis?, Radula NP, Ptilidium NP, Dendromastigophora NP, Gackstroemia NP, Lepidogyna NP, Lepidolaena NP, Porella NP, Frullania NP, all genera of Lejeuneaceae NP, G, Austrofossombronia G, Fossombronia G, Phyllothallia NP, Allisonia G, Verdoornia B, Pallavicinia G, Podomitrium G, Symphyogyna G, Xenothallus G, Hymenophyton G, Aneura B, Riccardia B and NP, Metzgeria NP, Monoclea G, Targionia G, Lunularia G, Dumortiera G, Asterella G, Reboulia G, Marchantia G, Neohodgsonia G, Riccia NP.

13. Acknowledgments

The senior author wishes to gratefully acknowledge support of the National Science Foundation (grants BMS76-03616 and DEB-8109680), which funded field work in New Zealand, Tasmania and Australia. The senior author also wishes to express deep appreciation for support of the National Geographic Society, which supported field studies in New Zealand (grants 5375-94, 5795-96, 7379-02, 7927-05 and 8132-06). The senior author wishes to thank the New Zealand Department of Conservation for the necessary collecting permits, and to especially acknowledge Paul Cashmore (Bay of Plenty Conservancy) for his kindness and valuable advice. The senior author acknowledges the Ngati Kuri runanga who, through their trust board Te Manawa O Ngati Kuri, granted permission to collect from their rohe. He also wishes to thank the Marutuahu and Te Arawa Trustees for permission to collect from their wahi tapu maunga Te Moehau, and Ngati Tuwharetoa for permission to collect from Tongariro National Park. Numerous individuals provided valuable field assistance to the senior author in New Zealand, and it is impossible to thank all of them here; in particular, thanks are due to John Braggins (AK) and to the late John Child for facilitating much of the field work in New Zealand, and for their warm field companionship. Thanks also to Peter de Lange for his interest and support as well as for valuable advice. Laura Engel provided valuable assistance in the field while in New Zealand (1997) and with literature searches; her help is acknowledged with thanks. The senior author wishes to extend special thanks to Zorica Dabich for her care in preparation of the habitus figures of the illustrations (which are otherwise those of the senior author), and to David Sollenberger for excellent preparation of many of the oil-body illustrations. The senior author also thanks Stefanie Stephens (Library, Field Museum) for her continued diligent tracking of literature.

The senior author wishes to thank the following individuals and institutions listed below for the loan of specimens, all made to Field Museum: John Braggins (AK), and also for his continued interest in this project; S. Blackmore and L. T. Ellis (BM); P. J. Brownsey and Barry Sneddon (WELT); Ewen Cameron and Mei Nee Lee (AK), with special thanks to Mei Nee Lee for specimen searches; Francoise Dreger-Jauffret (STR); Lesley van Essen (MPN), with special thanks for specimen searches; Allan J. Fife (CHR); the late Patricia Geissler and Michelle Price (G), with special thanks to Dr. Price for her invaluable help and kindness during herbarium visits; Gintaras Kantvilas and Kim Hill (HO); Nils Lundqvist (S); David A. Meagher (Surrey Hills, Victoria, Australia), with gratitude for his interest in the project and for forwarding specimens; F. K. Meyer (JE); Nicole Middleton (MELU); D. H. Pfister (FH); Alan Prather and Alan Fryday (MSC); C. Rausch de Traubenberg (PC); Harald Riedl (W); and William Buck and Barbara Thiers (NY).

A work of this magnitude could not have been possible without the constant help and encouragement of Karen R. Engel. Her assistance both in the field and in the herbarium, as well as at many and various stages in the development of this manuscript, is gratefully acknowledged with warmest thanks. The senior author also is deeply indebted to several other individuals that have been instrumental in the development of this flora project. Elizabeth Engel provided valuable assistance in the field (1982–1983) and in many and various stages of the manuscript. The senior author especially acknowledges Matt von Konrat, who has diligently assisted in numerous and varied ways, both during several expeditions to New Zealand and in the laboratory at the Field Museum; his interest in and support of this project are deeply appreciated.

Work by the junior author was funded by the Foundation for Research, Science and Technology. Thanks are due to Ilse Breitwieser for her support of the project. Sue Gibb, Michelle Breach and Jane Cruikshank helped with specimen preparation and databasing, and Sue Gibb also curated the collection at CHR. The junior author wishes to thank Allan Fife for discussion and time spent together in the field, and Matt Renner for his interest and enthusiasm for the subject. Thanks are also due to Christine Bezar for editorial assistance, and to Ewen Cameron and John Braggins for assistance with parts of the introduction. Thanks to Bill Malcolm and John Braggins for providing color photographs, to Cissy Pan for preparing the color plates, and to Bill Malcolm for discussion of the glossary. Eric Godley provided sound advice on the history section of the Introduction, Janet Wilmshurst, Peter Bellingham and Matt McGlone gave advice on bioclimatic zones, and Bruce Clarkson suggested use of ecological provinces rather than the traditional land districts. The junior author thanks the curators of other herbaria: Lesley van Essen (MPN) for help with Hodgson types, Simon Whittaker and Barry Sneddon for help with collections and databases at WELT, and John Braggins, Mei Nee Lee and Ewen Cameron at AK for help with databases. Barbara Polly and Alison Glenny gave hospitality during visits to WELT. Thanks also to friends for company and help in the field, especially Geoff Spearpoint, and to organizers of various John Child bryophyte workshops who created opportunities to collect. The junior author also wishes to thank Solid Energy and Ruth Bartlett of Kingett Mitchell Partnerships who made it possible to collect on a number of occasions at Stockton Plateau and Waimangaroa Valley.

Both authors sincerely extend thanks to the staff at the Missouri Botanical Garden Press. Gratitude is extended to Victoria Hollowell for valuable advice and suggestions, particularly during the formative stages of this project, and to Lisa Pepper for carefully proofreading most of the text before the typesetting. We want to especially call attention to the care, diligence and effort of Beth Parada, who also gave so freely of her time and patience. The quality of the volume is to a significant extent reflected by her considerable talents, and special gratitude is extended to her. We also thank Ingrid Willis and Aaron Wilton at Landcare Research, Lincoln, for formatting the text, and in particular the keys, and for creating the index.

J. G. Duckett and R. Ligrone thank the Department of Plant and Microbial Sciences, University of Canterbury, Christchurch, New Zealand, for providing laboratory facilities and the New Zealand Department of Conservation for granting collecting permits. Collection of the specimens used in the study of endophytic fungi in liverworts was facilitated by local knowledge of the sites by Brian Butterfield (University of Canterbury) and David Glenny (Landcare Research, Lincoln, New Zealand). J. G. Duckett was supported in New Zealand by an overseas travel grant from the Royal Society (UK) and by the New Phytologist Trust. R. Ligrone’s visit to New Zealand was supported by a short-term stay grant from CNR (Italy).

© All rights reserved. [Image: 4Y10]

© All rights reserved. [Image: 4Y11]

© All rights reserved. [Image: 4Y12]

© All rights reserved. [Image: 4Y13]

© All rights reserved. [Image: 4Y14]

© All rights reserved. [Image: 4Y15]

© All rights reserved. [Image: 4Y16]