Nothofagus, key genus of plant geography, in time

and space, living and fossil, ecology and phylogeny

C.G.G.J. van Steenis

Rijksherbarium, Leyden, Holland

Contents

Summary 65

1. Introduction 66

New 2. Caledonian species 67

of and Caledonia 3. Altitudinal range Nothofagus in New Guinea New 67 Notes of 4. on distribution Nothofagus species in New Guinea 70

5. Dominance of Nothofagus 71

6. Symbionts of Nothofagus 72

7. Regeneration and germination of Nothofagus in New Guinea 73 8. Dispersal in Nothofagus and its implications for the genesis of its distribution 74

9. The South Pacific and subantarctic climate, present and past 76

10. The fossil record 78

of in time and 11. Phylogeny Nothofagus space 83

12. Bi-hemispheric ranges homologous with that of Fagoideae 89

13. Concluding theses 93

Acknowledgements 95

Bibliography 95

Postscript 97

Summary

Data are given on the taxonomy and ecology of the genus. Some New Caledonian in descend the lowland. Details the distri- species grow or to are provided on bution within New Guinea. For dominance of Nothofagus, and Fagaceae in general, it is suggested that

this. Some in New possibly symbionts may contribute to notes are made onregeneration and germination

Guinea.

A is devoted a discussion of which to be with the special chapter to dispersal appears extremely slow, implication that Nothofagus indubitably needs land for its spread, and has needed such for attaining its

colossal range, encircling onwards of New Guinea the South Pacific (fossil pollen in Antarctica) to as far

as southern South America. Map 1.

An is other chapter devoted to response ofNothofagus to the present climate. The possibility is envisaged inthe that it could have grown along the border of the Antarctic Continent during a milder climate Creta-

ceous and Tertiary.

The fossil record is ample, both by macrofossils and fossil pollen. Ofthe three pollen types, the brassii and

in Cretaceous in New in fusca types are already found the Upper Australia, Zealand, and Antarctica, and the in the found hitherto earlier than the Lower Table Eocene Fuegia menziesii type being not Tertiary. 1. No There it is reliable Nothofagus fossils have ever been found on the northern hemisphere. represented

its with which it forms the of the Macrofossils by counterpart, Fagus, subfamily Fagoideae family Fagaceae. also from the the northern hemi- ofFagus are known from the Tertiary and possibly Upper Cretaceous (on

sphere to a fairly high latitude. Map 1.

called for because it the three criteria for safe Nothofagus is a key genus plant-geography meets biogeo-

viz. it has sound an fossil and for which graphical reasoning, a taxonomy, ample record, diaspores long distance dispersal is excluded.

a remarkable with the contraction in the Fagoideae occupy hour-glass-shaped bi-hemispheric range, Malesian Whereas the of of tropics. Map 1. evenat present largest amount morphological diversity Fagaceae

65 BLUMEA No. 66 VOL. XIX, I, 1971

in the world is found in Malesia (Lithocarpus: various sections, Quercus, Castanopsis, Trigonobalanus, and

Nothofagus), with Castanea and Fagus in the Sino-Himalaya, it is likely that the cradle of Fagoideae was

somewhere in the region stretching from Yunnan to Queensland. It is reasonable to assume that from this

initial Cretaceous matrix in the northern and in pre-Upper Fagaceous Fagus emerged part Nothofagus for which looks the the southern part, Queensland most likely.

It is shown that the evolution and of be in unforced under the spread Nothofagus can explained an way

steady stateprinciple, which implicates that certain oceanic areas in the South Pacific have been land in the

Cretaceous or Early Tertiary.

north-south of distribution is found in the rank This bi-hemisphere pattern many groups, sometimes in sometimes in rank in of family, the lower of genus or even rare cases of species. Ranges may be almost

intact the be in which almost in tropical-montane zone, as happensto Euphrasia possesses a range replicating In the transition absent. The that of Fagoideae. Map 2. others tropical relicts are very rare or even rare ones isolated and unusual are sometimes taxonomically highly interesting by being systematically showing

characters.

advocated that the of this of distribution have their It is tropical montanerepresentatives pattern gained series is range synchronously or almost so, as the of opportunities enabling such ranges to develop supposed

to have happened only ‘once’ in geological time. This implies that taxa belonging to ‘morphologically

advanced’ groups are of greater antiquity than formerly often supposed.

The main conclusions have been framed in a number of theses (Chapter 13).

1. Introduction

A few words of introduction for those who have no ready access to my monograph

and notes (van Steenis, 1953) subsequent (van Steenis, 1953a, 1954, 1955).

& southern halfof the Southern beeches range (van Steenis van Balgooy, 1966) in the

0 0 Pacific in South America from c. 33 S southwards to c. 55 S, and occur further in New

Zealand, Tasmania, East Australia, New Caledonia, and New Guinea (inch the d'Entre-

casteaux Islands Goodenough and Normanby, and New Britain). confined There is the leaf-shedding section Nothofagus (7 or 8 spp.) which is to South for the America except Tasmanian N. gunnii (Hk. f.) Oerst.

28 South American The evergreen section Calusparassus (c. spp.) is, 3 species excepted, confined the this the to Old World. Within section 21 Papuan—New Caledonian species

1 a that are by occupy special taxonomieposition in they characterized 2-valved cupules ) and therefore accomodated in subsection series are a special Bipartitae. This comprises a

New which the Triflorae, with 4 species in New Guinea and 4 in Caledonia, in cupule

others in which the contains 3 nuts, the belonging to series Uniflorae cupule has I female

flower, hence I nut. who the Baumann-Bodenheim (1953), referred New Caledonian species to a separate

stated that all New Caledonian have female flowers genus Trisyngyne , species 3 per

cupule, but this has turned out to be an erroneous generalisation: N. aequilateralis and N. codonandra indeed have but least discoidea has female flower 3, at N. only 1 per cupule. affinities The of the species within sect. Bipartitae run, however, across the distinction of these two series showing that this systematic division is artificial and shouldbe aban-

doned. the of three female flowers a of course Morphologically, presence in cupule is than of more primitive one, but this reduction has obviously taken place independently in several specific lineages.

1 related the ofthis I mention in remarkable ) Thoughnot immediately to purpose essay, want to passing a

morphologicalreplica of the Fagaceous cupule in a totally unrelated plant, Blepharocarya involucrigera F. v.

M., belonging to the Anacardiaceae or closely related to these. The flowers in this dioecious tree are borne

in a bracteate woody involucre which according to Shaw (1966) is formed 'by the flattened, externally

grooved and partly concrescentultimate branches of the inflorescence', that is, exactly as I have formerly (1953) interpreted the Fagaceous cupule. This remarkable plant is found (by coincidence?) in N. and E.

Australia. C. G. G. J. van Steenis: Nothofagus 67

2. NEW CALEDONIAN SPECIES

In New Caledonia at least 5 well distinct species occur; all belong to the subsect. in New Guinea. Two of these Bipartitae which otherwise only occurs show a distinct

resemblance to New Guinean species, viz. N. discoidea resembles N. perryi from New

Guinea and N. aequilateralis resembles N. starkenborghii from New Guinea, emphasizing

the ancient plant-geographical relation between New Caledonia and New Guinea. In

passing it may be said that in spite ofintentional intensive research by Professor Corner c.s. turned Solomons. on Guadalcanal at suitable height, no Nothofagus has up in the This

absence from the Solomons is shared by Araucaria and Agathis. The absence of these three

if of these and genera is not unexpected the very poor dispersal capacities genera the of the Solomon taken consideration. relatively young age present Is. are into See chapter 8. difference from the The New Caledonian species exhibit a remarkable vegetative

Papuan species, not realized before, namely that whereas the Papuan ones have a strictly

distichous phyllotaxis, the New Caledonian ones share a spiral phyllotaxis. This dualism and in phyllotaxy is, however, also found in some other Fagaceous genera though

remarkable, because clearly geographical in Nothofagus, need not alarm us about the

In addition I the that in generic consistency. may point to fact Vink noted that young

seedlings of N. rubra the first leaves are always spirally arranged, and that distichy starts

the first to occur on lateral twigs (Vink 17473). The New Caledonian species have all

thick large leaves with entire or crenate margin.

In New Caledonia Nothofagus shows also an ecological behaviour slightly different

from that in New Guinea, viz. in altitude.

ALTITUDINAL OF NOTHOFAGUS IN NEW GUINEA 3. RANGE C.A. AND NEW CALEDONIA

Nothofagus was hitherto assumed to be bound to rain-forest under cold to cool climatic conditions. Accordingly it is found towards the tropics at steadily increasing altitude.

— of New Guinea proper. In New Guinea, at 7° S, 98 % the collections were made between and m. 1000 3100

All species ascend to various heights above 2000 m, two excepted, viz. N. nuda and known from N. cornuta which are both only their type specimens, collected at 1200 and Six found than five 1750m respectively. species are not higher 2400 m, others occur as the sofar recorded for high as 2800—3000 m, highest locality is 3100 m N. pseudoresinosa. viz. N. Besides N. nuda, there are four species descending below 1500 m, womersleyi,

N. carrii,N. starkenborghii, and N. flaviramea. The first two of these are too rarely collected idea oftheir total The latter and to gain agood range. two, however, are common are also foundbelow 1000 m. The occurrence of Nothofagus below this critical altitudefor moun- Rand tain plants in Malesia was already reported by Brass (in Archbold, & Brass, 1942) in of. dominated of an 'outlier body Nothofagus forest' at 850—900 m southwest Bernhard voucher Camp, West New Guinea, but I have not seen specimens sustaining this record. made of dominance However, recently a similar report was even by Nothofagus in Fago- forest Schodde Hitchcock Lake Kutubu 0 Lauraceous by & (1968) near at c. 143 15' E and

800 The 6° 20' S, found on karst limestone topography at —900 m. voucher specimens belong to N. starkenborghii. has also been collected and but N. flaviramea repeatedly at 750, 850, 900 m, not espe- Genofa. cially stressed as dominant, though common on Mt. 68 BLUMEA No. VOL. XIX, I, 1971

D'Entrecasteaux Islands. — Hitherto Nothofagus has been found in two islands of All this group which is adjacent to the north of the eastern tip of New Guinea. four

collections are sterile, save one which carries male flowers. I have tentatively referred and if these identifications them to two species are correct the data are as follows:

elevation found crest In Goodenough I. (highest c. 2500 m) N. carrii is on a narrow at in New Guinea 1750 m. This species ranges from 2850 m downwards to 1400 m.

In Normanby I. (highest elevation c. 1000 m) Brass collected N. carrii once on the

summit of Mt. Pabirama at 900 m as a major constituent in shrubby growths 2 —5 m

high. This occurrence is distinctly below its lowest altitude in the mainland of New

Guinea.

I. Brass collected N. rubra at m as a tree in tall In Normanby once 850 common canopy

forest and m stunted forest. This also mossy (male fl.) once at 750 in open Dacrydium is

lower than the mainlandof where rubra from — distinctly in New Guinea N. ranges 1750

2850 m.

In both islands the lowest localities of the genus are the same as in the mainland ofNew

viz. Guinea, c. 750 m.

New Britain. — In New Britain (highest elevation of the non-volcanic mountains c.

has been twice 1900 m) N. starkenborghii collected (both specimens sterile), once at 1350 m

0 and once at 600 m (area between 5° 30'—6° S, 151 —151° 30' E). In the New Guinea

mainland N. from m down to this starkenborghii ranges 2400 750 m. In New Britain shows hence the lowest of in lat. species occurrence the genus Papuasia sens.

I have formerly tentatively concluded that specimens below 1000 m remain sterile, and

that the lowest records would concern occasional specimens. The new records show

that this both male 1 is not true, and female are collected at these lowest altitudes ). Though it remains to be proved that low occurrences are found throughout New Guinea, there

several localities where is abundant are Nothofagus at 750—900 m. Furthermore, they are found fertile at these altitudes. Obviously lowest occurrences concern largely two species,

N. starkenborghii and N. flaviramea. But it should be added that there are no signs that descends the lowland that would the Nothofagus in Papuasia to or species solely occur in lowland.

There that descend lower altitudes is, however, a tendency Nothofagus species to on smaller and lower, more oceanic lands as shown by N. carrii, N. rubra, and N. starken- borghii.

the New Guinea leaves N. N. Among species some possess large (N. perryi, grandis, and flaviramea), some very thick-coriaceous leaves (N. rubra, N. brassii), some always small leaves (N. pullei) and it might be supposed that leaf-size and leaf texture go parallel with the first lower andthe other altitude altitude, categoryindicating two categories high there thin species. However, is hardly any correlation in this aspect. Two species with and fairly large leaves, N. resinosa and N. pseudoresinosa, belong to the highest ascending

both exudate of resin the leaves and species, admittedly sharing a conspicuous waxy on flush.

Caledonia. — New This oceanic island is situated at c. 20—22° S, that is still within the tropics. Its highest elevation is considerably lower than that of New Guinea, the two

x The Melchior that is ) statementby (1964) Nothofagus is dioecious an error; alle are monoecious. C. G. G. J. van Steenis: Nothofagus 69

Five in highest peaks reaching only c. 1640 m altitude. species occur; proportion to the all and New Guinea species they are large-leaved very coriaceous. of in Caledonia Well-documented data on altitudinal range Nothofagus New are as

mentioned Virot. Baumann-Bodenheim yet scanty and are not by According to (1953)

all are confined to serpentine bedrock.

Combining the published data by Baillon (1873) with the collections at hand and

unpublished data kindly provided by Dr. H. S. McKee, the following picture emerges

of the altitudinal range of the New Caledonian species: N. discoidea c. 150—230 m, N.

—600 N. codonandra N. balansae aequilateralis c. 200 m, c. 150—1350 m, c. 300—1350 m, and data N. baumanniae c. 1000—1200 m. These altitudinal agree with the altitudes given of by Baumann-Bodenheim (1956); he even admits descent three species to the lowland

at 100 m.

in Guinea in The situation is obviously very different from that New that New Cale-

donian Nothofagus clearly does not shun the tropical lowland.

the in Judging from macroclimate, one would expect New Caledonia an altitudinal

600 behaviour similar to that in New Guinea (750—3100 m, m in New Britain), as New Caledonia situated from that within the of is 20—22° S, is Tropic Capricorn, anyway much its in New not so different as altitudinal range Caledonia really is, namely from Caledonia further from the than 150—1350 m. New is admittedly away equator New be The with Guinea, but I doubt whether this can the reason. experience other mountain

plants is, that we have for example never found a consistent difference in altitudinal

of behaviour the same mountainplant species in Java at 7° S or Central Celebes at o° as

compared with northern Luzon which is at 20° N. One would have expected that in New CaledoniaNothofagus wouldbe crowded in the

summit zone of the higher peaks and, giving allowance to the elevation effect (van

Steenis, 1961; Backhuys, 1968), estimate a descent to possibly 500 m, but then of the the same species, and no species confined to lowland. As of climate in that to response altitude with we have, Malesia, the impression some-

times lower altitudes are reached in a seasonally dry climate, namely in the Lesser

in Sunda Islands. But several New Caledonian species do not behave this way, though it

should be admitted that New Caledonia is averagely less wet in total annual rainfall as

with rainfall hence compared New Guinea. Lower may contribute to some degree to

lower occurrence of the genus in New Caledonia.

that But it should be kept in view that it is not so much the descent of species, but two

species are confined to the lowland and low hills. And I have shown formerly (van

that in to correlate altitude with climate be Steenis, 1933) attempting one must very careful in dataof ofdifferent All comparing the same species, not species. New Caledonian

species are certainly specifically different from those of New Guinea.

well correlate the In conclusion, it appears not possible to altitudinal behaviour of

in New For I Nothofagus Caledonia with environmental factors. the present believe we

have just to accept that the New Caledonian species behave as they do and show a diffe- rent ecological response from the New Guinea species. Obviously the total ecological than tolerance of the genus is somewhat wider hitherto assumed, running from tropical

to cold-temperate.

There are other that isolated oceanic islands in the West examples on Pacific genera lower occur at altitude than on larger land masses, but a true comparison can only be made if I in this relates to exactly the same species. refer this respect to the general

remarks on altitudinal behaviourof Nothofagus made above underNew Britain, in which island reaches lowest altitude it its in Papuasia at 600 m. No. 70 BLUMEA VOL. XLX, i, 1971

It would be worthwhile to make some experiments with Nothofagus, viz. to transplant

Guinea New and vice in order how behave. New species to Caledonia versa, to see they

4. NOTES ON DISTRIBUTION OF NOTHOFAGUS SPECIES IN NEW GUINEA

of of At the time of my monograph the genus (1953) the total number of collections

the 16 species distinguished was only 47, including the sterile ones. Largely through the ofthe of and the LandResearch of greatactivity Division Botany, Lae, section C.S.I.R.O., the collections. Canberra, material amounts at present to some 380

There is gradually coming more insight in common and rare species: of N. grandis,

and have and of some N. perryi, N. pullei I now over 40 collections, N. flaviramea 30.

collections were known either from the eastern the western In 1953 many only or part,

6 of the but this restricted occurrence has a tendency to fade away, commoner species being

known now from both East and West New Guinea. But it appears that 2 very common

species remain to be restricted to East New Guinea, viz. N. perryi and N. grandis. collections abundant the be Though are now complete inventory cannot yet made, I because in and am afraid, still new, clearly good species turn up, N. womersleyi i960 N. nuda each of collection. in 1966, consisting a single have the I been compelled to reduce a few species and among species with reduced

is cupules there is still some uncertainty about specific status for a few. This not so much about the due to an assumed indistinctness of specific delimitation or doubt importance

of the cupule characters of Papuan species of Nothofagus, but to the inadequacy of most collections. material but sterile Complete fruiting gives never difficulty in identification, often taken from and those with male flowers with collections, too young trees, only or

very young female flowers are sometimes difficult to place, because in contrast to the

Caledonian which there half New species, can all easily be recognized in sterile state, are dozen which much alike in characters. a Papuan species are very vegetative characters belief Though they are often closely resembling in vegetative my strong is that there least dozen derive this from the fact that of are at over a good species. I a number of have of which goodly species I now 10—40 exactly matching specimens the first the cupule characters are completely constant. In addition, since writing attempt towards revision in I have found the detail characters of the male a 1953 that, though their be flowers provide no clue, number per catkin, one or three, has proved to constant

found for the species, most having three, one being only in N. pullei, N. crenata, N.

resinosa, and N. pseudoresinosa.

Hybridisation is known from Nothofagus, both in South America and New Zealand, would and could occur in New Guinea as well. Though this would infer that such hybrids bear whether these fruits fruit which should not occur with good species, we do not know

are viable.

In dissecting nuts in the herbarium of indubitably good species, one finds many nuts climatic conditions barren, as also often happens in Fagus, often depending on the year's

favourable or not for pollination. As male anthesis is short in New Guinean Nothofagus,

rains in that heavy particular period may be unfavourable for fruit-setting.

In I observe of Zealand passing may that in comparison with specimens New and South beeches American theNew Guinean(and NewCaledonian) species carry very few cupules

per collection. It is not only that in a number of species the cupules are small and incon- but also the real of spicuous or even absent, number nuts per sheet is extremely small, has often the o—1( —3). One to scan the specimen carefully to find them. Of course trees in with columnar bole of are extremely large good situations, 30—45 metres and over a C. G. G. J. van Steenis: Nothofagus 71

record thickness of the bole c. I—1| metre through (the in being 2\ m dbh, by a of noted and the number of height 45 m, in N. grandis, Cruttwell 899) nuts produced by hence be considerable. of the be one tree must But a large proportion nuts appears to and become earth-covered not viable, possibly only those which will germinate success- fully.

5. DOMINANCE OF NOTHOFAGUS

the of the of In New Guinea highlands Nothofagus is one most important components the frame of the whether and forest, or not mossy, at altitudes between 1500 3000 m, lower. rarely Frequency on labels reads often 'abundant', 'common, 'very common'.

the This tropical-montane mesophytic forest, generally known as Fago-Lauraceous forest for of type, consists a large part Fagaceae (Castanopsis and Lithocarpus); other

genera often frequent are Araucaria, Libocedrus, Weinmannia, Engelhardia, Dacrydium,

and of is often Podocarpus, Eugenia, Cryptocarya. Frequency Nothofagus very high and is

especially easily observed when the trees are in reddish flush.

field labels of almost all of Dominance is reported on species, in particular N. pullei

(8 times), N. grandis (5 x), N. perryi (5 x), N. starkenborghii, N. rubra, and N. flaviramea

(each 3 X). the of Giluwe Dominance is easily observed from air and maps (e.g. Mt. area) can be

composed in which the stands of dominant Nothofagus are indicated in detail. Of course

through human interference there are also degraded areas where Nothofagus occurs scat-

restricted of tered, in groves or stands, etc. And there is plenty forest where Nothofagus

is not dominant.

Kalkman and Vink confirmed older data of Brass and of Robbins that the co-domi-

nance with other Fagaceae, notably Castanopsis and Lithocarpus, is in their experience not

manifest collectors have observed and Kubor that so as suggest. They on Arfak, Doma, the than the proper Nothofagus zone is always situated higher Lithocarpus or Castanopsis dominated lower forest. also studied They have in more detail (Kalkman & Vink, 1970)

the regeneration on Doma Peaks and found Nothofagus in restricted, sharply demarcated

stands with plenty regeneration inside, but not beyond the stands, and also no Nothofagus

in forest. and this is of beech trees surrounding Obviously, no news, dispersal nuts must slow indeed. go very

This other of viz. certain of brings me to an general aspect interest, why groups plants

such a dominant role numbers the over surfaces and play in in vegetation immense get

dominance or co-dominance by their social behaviour, and maintain this, as proved in least onwards of the similar Nothofagus, for geological epochs, at Cretaceous. A thing can be said of Conifers, of the Dipterocarps in western Malesia, and ofEucalyptus and Acacia the latter the in Australia, though in two cases mass development is certainly less ancient.

bold their Is it too to put forward the hypothesis that virility as exhibited by abundance due their either exudation is to capacity to produce toxic substances, by or exhalation,

or substances in their litter which in some way or other are toxic to or inhibit germination

in the of other or growth soil plants? This concept, called allelopathy by Molisch, is

difficult to because can the extremely prove, in experiments one not exactly imitate

processes going on in nature, especially the decomposition of litter by micro-organisms, the of and besides calculate or eliminate root-competition, reaction mycorrhiza, the

whole process being an extremely complicated complex of interacting factors. Though in it rather clear that feel some simple cases seems allelopathy exists, I that the concept has which the not yet been sufficiently examined is probably reason that it is hardly 72 BLUMEA VOL. XIX, No. I, 1971

introduced into ecology. Mostly the dominanceof the said groups is claimed to be due to

the excessive shade they cast, prohibiting growth of shade-intolerantother plants. But to rain-forest without dominance co-dominants this can be opposed that mixed or provides

also abundant shade, but nevertheless remains mixed. The same can be opposed to declare has been it a matter ofroot-competition. For Conifers still an other explanation advanced, viz. oflitter humous the fact that many plants cannot germinate on a layer or too topsoil, and that for successful germination their seeds have to be in contact with the mineral soil. This in countries and in mountain may be true temperate rain-forest where such humus accumulation is often found. And it is also true that in mixed lowland tropical rain-

forest abundant litter is mostly hardly existent because of its rapid decomposition. But

such reasoning could also be explained by the allelopathy hypothesis (Grtimmer, 1959; accumulation Rovira, 1969), whereby of humus produces a high percentage of toxic substance whereas rapid decomposition combined with constantly high rainfall dissolves

substances such low be more effective. In the toxic to percentages to no dry climates,

as for Eucalyptus, litter is so slowly decomposed not to have sufficient effect.

I am myself not certain about the significance of allelopathy in ecology, in particular be in the wet tropics, but the idea that it may important should not be suppressed. What- have that the solution ever the merits given to it, I come to believe for the dominance in of Fagaceae and others lies probably rather their living as symbionts with ectotrophic

mycorrhiza. I am much impressed by the data provided by mycologists.

6. SYMBIONTS OF NOTHOFAGUS

the A striking fungus occurring on the branches of Nothofagus is Discomycete genus

Cyttaria; it is peculiar to Nothofagus, causing big galls covered by gelatinous stromata. has of the of This not been found on any species the New Guinean-New Caledonian It is has it. Heim subsect. Bipartitae. a parasite, not a true symbiont: Nothofagus no profit of

(1951) reviewed the fungus flora of the New Zealand beech forests and remarked that

there are which both many genera, especially among Agaricales, are profusely represented

in the Eurasian Fagaceous forests and in the Nothofagus forest.

The of offer most feature. Morello mycorrhiza Fagaceae a interesting Singer & (i960)

generalized the occurrence of ectotrophic mycorrhiza which occur abundantly on the northern hemisphere with Pinaceae, Salicaceae, Betulaceae, Alnus, and Fagaceae (also and these that Fraxinus, Allophylus, some Caesalpinoideae). They are so closely bound to the symbiosis is by them consideredto be of a similar nature as in lichens. In the Americas Columbia and connected southwards the ectotrophs extend southwards on Quercus to are

montane line the of with the southern Notho- in a narrow along Andes, consisting Alnus, The fagus belt. fungus components largely belong to the Agaricales of which genera as well and Russula, Cortinarius, Lactarius, Inocybe, Boletus, etc. are extremely represented, some Gastromycetes and Aphyllophorales. Quoting these authors: 'These southern ecto- trophs find their natural geographical continuation in the Nothofagus areas of New Zea- The land, E. Australia, New Caledonia, and New Guinea. ectotroph as a biological unit

than forest element without generally possesses higher plasticity any comparable mycor- the the element with the and rhiza. In an association, ectotroph is usually highest fidelity regularity of distribution, and is commonly the first to establish itself among its conso- ciates. The distribution of mycorrhizal dominants and subdominants of this type shows an even distribution throughout the stand which causes the ectotroph mycelium to the soil almost and be occupy homogeneously. Fidelity regularity may explained by trophic reasons favoring the ectotroph over the non-mycorrhizal elements and making Steenis: C. G. G. J. van Nothofagus 73

climatic conditions and it more agressive and often more resistant to unfavorable to and diseases. The development of ectotrophs is limited by two factors, (i) dryness, (ii)

humidity if accompanied by generally high temperatures throughout the year; in very

humid climates with relatively low temperatures ectotrophs are common, frequency

and abundance increasing with a humid heterothermic climate.' which Singer & Moser (1965) explored forests south of Valdivia at 39° 30' S, in they stands the of is corroborated that in Nothofagus-free percentage ectotroph-forming fungi of zero. In the Nothofagus forest they found a clavaroid ectotroph fungus Ramaria zippelii, in tropical Asiatic distribution. They found it reasonable to assume 'that its origin must be and Guinea'. a region where the areas of Quercus Nothofagus 'touch', viz. New

Moser (1967) stressed again that ectotrophs have wider capacities than endotrophs or

disturbed after and near trees without mycorrhiza, for example as pioneers on soils, fire,

the tree limit. In his list of ectotrophs are also included, besides almost allPinaceae, Fagaceae, African and Betulaceae, Salicaceae; a few Rosaceae, some tropical Caesalpiniaceae, Tilia, Eucalyptus. essential role Notho- Summarizing, it seems that the ectotrophic mycorrhiza play an in had role also could well be that fagus forests, and must have that in past geological ages. It and dominance. through them they are capable to maintain their abundance, virility,

REGENERATION AND NEW GUINEA 7. GERMINATION OF NOTHOFAGUS IN

forest About fire I can be short: Nothofagus is not encouraged by fire. In New Guinea

officers have never observed actual basal suckering after a lightning strike or wind damage.

in N. do side shoots which Fallen trees mossy forest, e.g. of pullei, produce grow vertically

and also produce roots. This leads to thickets of saplings, but Mr. John Womersley, to

these field has how if reach whom I owe observations, no idea many, any, eventually

mature tree size.

reservoir of seed No regeneration has been observed in New Guinea which suggests a in is short the soil. The impression that the seed has a very viability. from of it that New seedlings may rise dispersal trees in adjacent forests. But as seems

is devastated dispersal goes slowly and only of short-distance reach, large areas offer,

even in upspringing secondary growth only an opportunity to regenerate Nothofagus

along the margins bordering old forest with Nothofagus. of the Germination is said to be epigeal, but in the old meaning word this means that

the seed is elevated by the hypocotyl above the ground. In the modernimproved meaning between the of this concept it means that the stem is produced centrally spreading coty-

ledons, irrelevant whether these enlarge or not and whether the seed is on the soil or be the elevated on the hypocotyl. Also in this meaning epigeal germination must surely

have a the case in Nothofagus, as it is in Fagus, though I never seen young seedling in herbarium.

would like know addition another of However, I to in ecological aspect germination, the viz. whether successful germination takes place if the seed is exposed on soil, or and whether the seed must be buried in the soil. The experience is namely that in Castanea the soil Quercus all nuts lying on are attacked by fungi and doomedto decay; only those into which happen to come the top soil (worm holes, washing over and covered by some soil, moved by ground animals, etc.) produce seedlings. Kalkman and Vink found in

Lei (Doma Peaks) the largest number of seedlings often on fallen, deadtree trunks which covered were thinly with moss.

In East New Guinea Papuans plant Nothofagus in their homesteads, along tracks and I3LUMEA VOL. No. 74 XIX, i, 1971

it is is of N. around their garden lands, said for ornamental purpose. This reported grandis,

N. perryi, and N. pullei. But they do not raise them from seed, but invariably transplant seedlings from the forest. and desired. More information on germination regeneration by seedlings is

8. DISPERSAL IN NOTHOFAGUS AND ITS IMPLICATIONS

FOR THE GENESIS OF ITS DISTRIBUTION

Distribution always brings along the question of dispersal. It has been observed that

in both New this goes Nothofagus very slowly, in Guinea and in New Zealand, and over very small distances, that is, effective dispersal including regeneration. found In addition there are experiments recorded by Holloway (1954) who that neither

Conifers nor stand besides nuts sink. the Nothofagus can transport by seawater;

Streams and rivers can effect dispersal over larger distances, but of course only down- stream.

Transport of the wingless nuts by wind is also only a matter of hundreds of metres;

the absolute maximum and that Preest (1963) put at 2 —3 kilometers, concluded long- be distance transoceanic dispersal by wind cannot considered.

also the the birds He reviewed situation in dispersal by birds, but nut-eating are all endemic and sedentary. Besides nuts would not survive passage through the gut. Epizoic dispersal of the nuts can readily be discarded.

would not as nuts cannot survive Transport by icebergs carry any living seed, freezing. The first for these who point to make is, clearly, reasons, that to any plant-geographer himself and critical branch of respects wants to develop plant-geography as a science, the of the land. dispersal Nothofagus is bound to presence of ofland and This leads to the question how under formerly different distribution sea, the broken has present range of Nothofagus has developed in time. Because Nothofagus an

and South Pacific fossils the enormous history behind it its range at least from Upper

will have land where there is The Cretaceous, we to accept none today. cornerstone of the explanation of the relict distribution of Nothofagus is the 'filling' of the three and oceanic gaps in the continuous range round the South Pacific, that between Fuegia Seymour Island, between Antarctica (McMurdo Sound) and New Zealand, and between these Antarctica and Tasmania/Australia. Accepting more land in gaps does not mean that this should all have been continuous land, and moreover need not to have been strictly synchronous. And if the land-connections were 'insular landbridges', these must have

had the geography of 'very nearby-continuous' — even less spaced than in the Malesian

landbridge — to permit short-distance dispersal.

This is the view held by a number of former plant-geographers (e.g. Diels) and advo- cated by myself (van Steenis, 1962) and by the late palaeobotanist Dr. Florin (1963).

Corner in be Also (1969) for his classical research the genus Ficus shares this view as may illustrated by a few citations from his latest summary on Melanesian Ficus: 'Theevolution of Melanesia Ficus in reflects its geological history. Ficus is not distributable except over land or through archipelagoes. There is no reason to invoke the risky means of long- distance the least and freak dispersal by chance, except to explain most casual, effective, occurrences. There is no evidence of random drift in the distribution of Ficus.’ In short: distribution land means (in litt.).

the land the To me the strongest evidence for theory is distribution of Nothofagus, living and fossil, while Florin's data are independently derived from his classic study of Conifers and Taxads. C. G. G. J. van Steenis: Nothofagus 75

I and that not to us was, am still, aware geomorphologists are yet so generous grant ancient land it this (cq. land-connections), even though appears irrelevant from which have have fore- source it might come and by what processes it might gone: foundered

continental continents, subsided continents, expanding or shrinking earth, drift, or plate tectonics.

be the To contemporary geomorphologists continental drift seems at present to more attractive theory. The land theory under the steady state postulate seems to some just heresy.

remind tries But I want to that to anyone who to follow the development of geomor- phological knowledge it is clear that geomorphologists are continuously struggling with alternative solutions or models on the basic principles about the evolution of the crust of the Earth, onetheory or reconstruction following an other in rapid succession. Geophys- ical flux thinking is at present in such a that it is impossible, at least for a biologist, to select reconstructions interpretations, datings, and which appear to be of lasting value.

It appears therefore premature to venture on a highly hypothetical correlation.

Reversely, botany has some unshakable factual material for phylogenetical and plant- geographical synthesis, viz. (i) the taxonomic relations within the living plant world,

(ii) the fossil record restricted to trustworthy evidence, (iii) the plant-geographical record of both living and fossil plants. of whether One can course dispute taxonomy and fossil evidence is 'factual' material

is in the the level of and it agreed that too many cases evidence has not reached factual material. For Fagaceae, however, I feel there is no doubt whatsoever, if we discard some doubtfully identified macrofossils which are besides irrelevant. of factual if The plant-geographical record I regard the same level as material, it is carefully sorted into intelligent distribution patterns.

Form-making, speciation, is admittedly erratic. For the present flora it is incompre- hensible: tell from the or the of it is impossible to past predict why array living beings

the observe it, went extinct, some is composed in way we why groups gradually, rich and other some rather abruptly, why certain groups exploded into development diverse groups on levels did not develop into profuseness, why some groups were of short duration and others maintained themselves for colossal periods of time though not necessarily attaining large size.

Biogeography is, however, not erratic, because once a species (hence necessarily a will And this genus) is successfully created, it start to expand. expansion process is govern- ed, restricted, or directed, or whatever term one may prefer, by its environment, geo- graphical and ecological. It will take advantage ofthis environmentwithinthe capacity of and this differ from another. all its dispersal rate, may one plant to But plants associated least the with it will take (or at have) same advantages within their ecological capacity, that means develop homologous patterns of distribution. And the closer their ecological

the the that will Plants amplitude more homologous patterns develop. spread together. be clear that the of be- From these considerations it must study homologous ranges

material. I the Malesian comes trustworthy factual Long ago have analyzed mountain flora by this methodology (van Steenis, 1933 —34) with remarkably successful results. There who this random are botanists doubt reasoning in accepting a hypothetical long- distance dispersal which to them would have played an enormous role, because this supposed dispersal would have gone on, admittedly rarely, through geological time,

it which would compensate for rarity. If this would have been really a general rule, be- comes clear that the orderliness of the homologous ranges would have been destroyed

least teaches or at completely obscured. As experience to increasing degree, however, BLUMEA VOL. No. 7 6 XIX, I, 1971

there exists distribution in with the an undeniable, clear, orderly array of patterns fitting

taxonomical data, so that we can safely discard hypothetical random long-distance dis-

in contributed the of distribution persal as a major factor having to genesis present

patterns. It is clear that the vicissitudes of the physiography of the earth, distribution ofland and

extension mountain been shared the sea, and height of ranges, and climate, have by plant

world. The botanical net results of this history, that is, present plant distributions, must

consequently ultimately be reconciled with a unanimously agreed history of the phys- latter branch of has been stabilized iography. Obviously the science not yet sufficiently

for such final correlation. Provisionally, the botanist has to work out his own data.

and the three criteria The distribution of Nothofagus Fagus agrees completely with

critical correlation on a factual basis: required for a biogeographical (i) taxonomists agree

unanimously on the taxonomicrelations, (ii) there is an ample trustworthy fossil evidence, and (iii) random long-distance dispersal is for these genera completely out of question.

this sound basis the frame the of the a Working on purpose is to history Fagoideae into what the time reasonable picture of must have happened in course of to cause the present will also land and pattern with its disjunctions. This show what physiographic changes (in

climate) are required as implications. The simplest way towards this goal is to relate this evolved to the steady state at least onwards of the time that the Angiosperms or were

expanding; it can be added in passing that most of the assumed major drifting of conti-

nents is usually supposed to have taken place before the Upper Cretaceous. be As will be shown, all data, including the ecological behaviour of Nothofagus, can correlated and unforced normal nicely in an way by geological processes. David's Geology of the Commonwealth of Australia (1950) (Burbidge, i960) shows

the complicated geology of the Australasian region, with immense changes in the distri-

of land and and subsidences. illustrated this for the New bution sea, large Fleming (1962)

Zealand region showing that New Zealand and the remote, wide-flung, speck-like it the of immense drowned is islands 'surrounding' are emergent remnants an (that

subsided) Tasmantis bank stretching almost as far as New Caledonia. A glance at modern atlasses shows immense submerged banks between Australia and New Zealand and And this for evidence that the land all without Antarctica. is me sufficient theory is not at vital background. How and when precisely these land connections have been situated the is not my task to unravel and entirely outside my capacity to reconstruct. But rough be clear and sufficient for picture seems to my purpose.

9. THE SOUTH PACIFIC AND SUBANTARCTIC CLIMATE,

PRESENT AND PAST

the fact that the broken In conjunction with the conclusion in the preceding chapter on be reconciled of the range of Nothofagus today cannot with the present physiography

South Pacific but must be correlated with palaeographic conditions, we will have to

consider the climatic conditions, present and past, in relation to Nothofagus ecology.

Except in New Caledonia where some species are growing under almost tropical of the cool lowland conditions, the majority species grow under cool to rather everwet

(rain-forest) conditions, the southernmost localities of N. antarctica and N. pumilio being 0 where extend far the limit. at c. 55 S in Fuegia, they as as tree

We have to realize that we still five in the tail of the Pleistocene Glacial Epoch, maybe

in We even an inter-glacial period. are so used to contemporary climatic conditions, historic that that observing almost no change in time, it is insufficiently realized our C. G. G. J. van Stbenis: Nothofagus 77

It is present climatic conditions at higher latitudes are still distinctly Ice-Age-featured. difficult butnevertheless reasonable that the to visualize, beyond any doubt, palaeoclimate Cretaceous and the of the the of the large part Tertiary up to Pliocene was very much milder. This have caused much diminished least must (or even absent) polar ice caps. At

the borderlands of Antarctica could easily have carried a temperate to cold-temperate

forest, suitable to the growth of Nothofagus. A climate shift of 15 latitudinal degrees from

Fuegia where the deciduous N. antarctica is found today, to about 70° S would allow Antarctica. Nothofagus to grow on This shift seems to be quite acceptable in view of

what we know of the magnitude of Glacial Epoch shifts on the northern hemisphere.

Even today, under the still prevailing Ice Age-featured climatic conditions as compared

with those before the onset of the Ice Age in the Pliocene, growth oftrees in the Arctic for of goes as far as 72° N for Larix and to 70° Juniperus; for smaller woody plants and Salix, Betula, Vaccinium even to 77° N. Though their growth must be slow, they get sufficient obviously light in quantity to maintain themselves.

How much better should they be situated under the ameliorated Cretaceous and

Tertiary palaeoclimates! northern With these hemisphere data in mind it seems to me clear that this milder

have palaeochmate must been sufficient, both as to light and temperature, to allow the

growth of Nothofagus on the borderlands of the antarctic continent. Cranwell believed 'could have Mrs. (1962, 1964) that Nothofagus never maintained itself far southwards outside the belt and for a the under temperate cries out place in sun

conditions comparable with those holding between 40° and 50°. How could massive forests and light-demanding have survived long in an Antarctica characterized by a long and dark, inevitably cold polar winter?Antarctica as it is placed today could never have been its cradle.'

My comment on this is firstly, that there is of course no question of a 'cradle', as this

would imply an Austral origin in the phylogenetic sence. But surely, Mrs. Cranwell cannot mean to divorce Nothofagus from the Fagaceous family and certainly not from the southern Fagus? From what ancestors on hemisphere would Nothofagus then have of emerged, quite independent the rest of Fagaceae? ¹) other she In an place said (Cranwell, 1963: 396) 'that the highly speculative tropical origins would please some phytogeographers', but it is surely not to please phytogeog- but for of the raphers an utter impossibility taxonomists to accept a biphyletic origin subfamily Fagoideae.

ancient is Bi-hemispheric counterparts are not restricted to Fagoideae, where one genus northern, the other southern hemisphere. Florin (1963) stressed the early bi-hemispheric which he held the evolution conifer and taxad geography as most significant in of the have stocks. I called attention(van Steenis, 1962: 285) to a similar tendency still discernible

various at different levels Such for among Phanerogams (family, tribe, genus). 'pairs' are example: Staphyleaceae-Cunoniaceae, Magnoliaceae-Winteraceae, Ericaceae-Epacridaceae, Cy- peraceae-Restionaceae, Poterium-Acaena, Carex-Uncinia, Dillenia-Hibbertia, Veronica-Hebe

complex, and to a fairly large extent Umbelliferae-Araliaceae and Cruciferae-Capparaceae ofwhich the two members (or at least one of them) are largely developed on one hemi-

be but sphere. Many more examples can unearthed, this requires a separate study. In my

x A is ) similarly erroneous premise the basis of the phytogeographical analysis of a certain number of which would all have extended their from the into the taxa ranges temperate zones tropics including

Nothofagus, by Bader (i960: 79, 81, 463). Many of them must stem initiallyfrom the tropics and have but later extended to higher latitudes. BLUMEA VOL. No. 78 XLX, I, 1971

opinion these pairs still reflect signs of early phylogenetical differentiation from common

tropical, tropical-montane, or subtropical matrices. second that and My comment is, even today N. antarctica N. pumilio grow profusely far 0 that as as 55 S and further nobody suggested that in Antarctica 'big massive forests' that for which the occurred, nor Nothofagus is a distinctly light-demanding plant summer

at 70° S would not provide sufficient light to grow. The latter assumption is certainly

in correct with the data known from the northern hemisphere even of today. And it is

most peculiar thatMrs. Cranwell completely neglects the significant fact that no less than

4 species are dominants as the most advanced woody plants at the tree limit, viz. N. antarctica and N. pumilio in Fuegia and N. cliffortioides and N. menziesii in New Zealand. both And they belong to sections, the first two are deciduous, the latter two evergreen. all And They are perfectly adapted to grow at a tree limit in Antarctica at 70° S. as

there well been other cold- many fossils are described may have species adapted to endurance which are now extinct.

Cranwell's additional to the fossil finds of Mrs. hypothesis, necessary explain Nothofagus

on Antarctica under her assumption, namely a close proximity of Antarctica and New Zealand the and able in Upper Cretaceous early Tertiary by which Nothofagus was to and then which grow in Antarctica, a southward drift of Antarctica by Nothofagus got be there extinct, appears to a completely unnecessary complication. That towards the end of the Tertiary, with the onset of the Glacial Epoch, the borderlandantarctic Nothofagus is in the Pacific the forests were wiped out, of course clear. Everywhere South Ice Age

must have taken a heavy toll of the flora of southern latitudes, probably even more than the because of the then lack of severely in northern, already existing escape possibil- ities. Mrs. Cranwell (1964: 90) rejects the possibility of long-distance dispersal in Notho- fagus, unfit to cross salt water.

this could Concluding subject, it seems beyond doubt that Nothofagus have grown on

antarctic conditions under Creta- the borderland under present physiographic but a mild and climate. ceous Tertiary

I will not reject Mrs. Cranwell's reference to Irving's palaeomagnetic data, indicating

that Antarctica and Australia were close together in the early Tertiary, swinging down by the later. Miocene, and that Antarctica and New Zealand were separated

What I wanted to show is that also without continentaldrift hypothesis the borderland of

Antarctica it is situated would allow for under milder as today Nothofagus to grow a climate.

10. THE FOSSIL RECORD. Table 1

There evidence This is, fortunately, a profuse trustworthy fossil of Nothofagus. consists

of of In be remarked that the partly macrofossils, partly pollen. passing it may curiously pollen of Nothofagus is very distinct from that of other Fagaceae and had it been known only from the fossil state its correlation would not have been possible, I am told by palynologists.

the claimed Among macrofossils Von Ettingshausen (1888) to have found various of and also Australian fossils. species Fagus Betula in Tertiary There seems, however, no doubt that these are all Nothofagus.

the should be advance that there are three the As to pollen it explained in pollen types, and fusca menziesii types and the brassii type. These coincide only partly with the taxon- of the which omy genus, is first subdivided into two sections, evergreen Calusparassus The and deciduous Nothofagus. fusca and menziesii types occur in both sections. However, C. G. G. J. van Steenis: Nothofagus 79

SOUND

1. stratigraphical

record of GUINEA CALEDONIA ZEALAND

NOTHOFAGUS

pollen FUEGIA NEW NEW AUSTRALIA NEW Mc.MURDO SEYMOUR

m b f m b f m b f m b f m b f m b f b f m

RECENT 1 1 i

PLIOCENE

UPPER MIOCENE

LOWER MIOCENE

OLIGOCENE I I

EOCENE I - I ■ - - j 1 PALAEOCENE 1 1

UPPER CRETACEOUS 1

Table and fossil known from the record 1. Nothofagus, recent as yet pollen (after Cookson, 1959; Couper, Dettmann those of McMurdo Sound = brassii 1960; & Playford, 1968); unspecified as to type (b type;

f = m = menziesii For emendation see fusca type; type). Postscript p. 97.

the brassii type is amongst living species strictly confined to subsect. Bipartitae of Calus- the and Caledonian parassus, comprising New Guinean New species. Precisely as in the of case Metasequoia, Nothofagus subsect. Bipartitae was described as pollen before botanists the Pike described living plant. Cookson & (1955), who studied the pollen of the species this stated that different be of subsection, species can not always discriminated by their the fossils there have been far 'botanical pollen. Among may consequently more species' than 'pollen species'!

For that will later will the fossils the total reasons appear I now review along range, the Pacific Table country by country, going from New Guinea along South arc. 1.

— About all with brassii Fossil of New Guinea. 15 living species, pollen type. pollen

this type has been found in SE. New Guinea from the Upper Miocene (A. M. Khan, in litt.). No older records known; it should be realized that the fossil record of New Guinea

is extremely poor.

New Caledonia. — At least 5 living species. No fossil records known to me.

Australia Tasmania. — There deciduous in & are 3 living species, one (N. gunnii Fossil the and Tasmania). pollen of both fusca menziesii types occur in abundance in

Australia. In 1946 Cookson described fossil pollen from SE. Australia from strata dated from the that as Oligocene to probably Miocene. Later she compared pollen types with of the living New Guinea species (1952) and concluded that one of the Australian pollen

much similar to the brassii types was very type (Cookson & Pike, 1955).

We have now a recent record of an Upper Cretaceous (Santonian to Uppermost of brassii from Cretaceous) pollen by Dettmann & Playford (1968) the type the Otway

Basin in Victoria. This was expected by me in the light ofthe Upper Cretaceous findings BLUMEA VOL. XIX, No. 80 i, 1971

in New Zealand. It is most important to note that during the Tertiary Period the brassii

was the and distributed found and also type commonest widely pollen in East, South,

West Australia (Cookson, 1954). From this follows that in the Tertiary Nothofagus must been of have one the most important genera of Australian forests, obviously wiped out

by the onset of the continental desiccation following the disappearance of the epeiric earlier As seas which caused the pluvial climates. all three pollen types are represented,

and 'pollen species' may have represented more than one or even several 'botanical

rich species', Australia must in the past have been a country in Nothofagus, possibly some

10—20 species.

New Zealand. — There are 4 living species, all evergreen. As early as 1887 von Et- had tingshausen described fossil leaves from New Zealand which he attributed to

and if Fagus (4 species), Quercus, Alnus, Ulmus. In my opinion the larger part of these, not

all, must belong to Nothofagus. Oliver (1936, 1950) described macrofossils from the

Upper Cretaceous to Miocene and assigned them partly to Nothofagus, partly to Fagus,

and to was well with the partly a new genus Parafagus. He only acquainted living species Zealandand fromNew was not aware of the existence of large-leaved species belonging from and to subsect. Bipartitae New Guinea and New Caledonia, obviously inadequately

informed about the South American species. The characters he used to distinguish the three he derived fromthe and leaf genera mainly venationpatterns margin. In my mono- have shown that all the and graph (van Steenis, 1953: 322, fig. 1) I venation leaf-margin described him in of I structures by are also present living species Nothofagus and con-

cluded that all, or almost all, his fossil leaf material must be referred to Nothofagus.

record Fortunately there is a profuse fossil pollen of Nothofagus in New Zealand. This

Mrs. knowledge started with a paper by Cranwell (1939) who recorded fossil pollen from with of the all Tertiary lignites, in conjunction a study pollen of living species of then This Nothofagus known. fossil pollen was later named N. cranwellae Couper the (!953). representing cranwellae pollen type.

The oldest pollen record was described by Te Punga (1947) as N. kaitangata from the

Lower Coal Measures of the Lower Cretaceous. Couper (1952a) reinvestigated these coal and confirmed this measures record. Later Couper (1960a) corrected the dating to Upper I.e. that other coal Cretaceous (Senonian, p. 6), adding in Upper Cretaceous measures

the same species occurs abundantly. In the Upper Eocene to Lower Oligocene deposits

of four other discovered and in this makes 80 pollen species was some places up % or

more of the forest pollen. Their pollen was first considered as belonging to the cranwellae

the type, but this was later found to belong to brassii type. The of these data that the brassii far the oldest. fact great importance is type is by In the earliest identifiable fossil Nothofagus ranges among generically Angiosperm pollen of the world. types

The brassii type is perpetuated during the Tertiary till the Pliocene, but already in the the the the Upper Cretaceous it is joined by fusca type; menziesii type appears later, in recorded Oligocene, as by Couper (1960a). derive data It is tempting to on the New Zealand palaeoclimate from the occurrence of fossil Nothofagus and other fossils found with it. In the Miocene (or ± younger) Cocos and ?Avicennia in are among the latter. Couper (1952) reviewing Nothofagus as a climate

indicator concluded that the occurrence of Cocos together with Nothofagus is 'anomalous'.

But that time he could the brassii at not know that some Nothofagus species of pollen type

still grow in New Caledonia under tropical conditions almost to sealevel. Besides, it

should be observed that one cannot simply infer that a fossil Cocos must have possessed the C. G. G. J. van Steenis: Nothofagus 81

the Cocos same restricted thermo-ecological range as present nucifera. Furthermore, northern of Zealand. the Avicennia is even at present growing in the tip New Finally, climate of Zealand been than it Tertiary New must have distinctly warmer is at present.

So I feel that with our increased knowledge the term 'anomalous' is no longer justified.

existed the According to the pollen record many species of Nothofagus in New Zealand all found Tertiary, but not are throughout this period. and all According to Couper (1953a, 1953b) Couper & Harris (i960), species of the

cranwellae group—ibrassii type became extinct before the Pliocene, or latest in the Upper found the Pleistocene. Pliocene; none is ever in The fossil record till the Recent. fusca and menziesii types were perpetuated in the up

A noteworthy remark by Couper (1953b) is that in comparing fossil New Zealand

with Australian he found some similar. This does pollen types ones 'pollen species' very because have not prove that New Zealand shared 'botanical species' with Australia, we

than botanical but it seen that a pollen species may cover more one species, probably indicates close affinity.

I in the lists of fossils found with in New may add that scanning together Nothofagus in this of the Zealand and Fuegia, there appears to be a striking homogeneity part Sub-

antarctic, and this affinity is still expressed generically in the present flora, pointing to a

hence at least onwards of the great common floristic affinity, common history, Upper

Cretaceous.

of Concluding, New Zealandmust have harboured a large number Nothofagus species, the brassii with late maybe some 20—30; in the Cretaceous mainly of pollen type one of all three Several record of the menziesii type, in the Tertiary pollen types. species for the There with Australian lived only a certainperiod in Tertiary. was a good similarity

'pollen species'.

from where Antarctica. — The first fossil records were from Seymour Island, Dusen of (1908) attributed macrofossils to species Fagus, from Oligocene deposits, mentioned also from the correlated by Cranwell (1963: 395) Upper Cretaceous. In part they were

doubt that all these to with some other fossil species from Fuegia. There is no belong Nothofagus.

describe microfossils from I. situated Mrs. Cranwell (1959) was the first to Seymour of 0 and them of on the tip Antarctica (64° W, 57 S) opposite Fuegia, among are pollen the which is Miocene. Nothofagus. She could correct Dusen about the age of strata Early

she corrected the of the material In a very recent paper (1969) age Younger Seymour which is probably older, viz. Paleocene (Maastrichtien); the Snow Hill samples and

Older Seymour material is Upper Cretaceous (Campanian) and contains the fusca type.

The dominants among the pollen were Conifers and Nothofagus, a striking parallel with and the New Zealand! The pollens belong to the brassii fusca types; surprisingly, menziesii

abundant South was not found. type which is so in living American species

A second record, remote from the above, was published by Cranwell, Harrington &

Speden (i960), viz. from the McMurdo Sound District, Grahamland, Ross Sea (c. 165° E,

0 to be late to Oligocene. These are 77 S), opposite New Zealand, supposed Cretaceous believed be Eocene. the now by Mrs. Cranwell (1969) to Upper Among pollens, none identical with those of of which was found Seymour I., Nothofagus predominates, pollen with and of conifers being in smaller quantity palmoid proteaceous types.

Antarctica: data there to be an Summarizing on though are yet few, appears astounding parallelism between the fossil evidence with that of New Zealand. It is ofeminent value

that the brassii found and close South America. type was so early so to

81 82 BLUMEA VOL. XLX, No. I, 1971

Southern South America. — Ten living species, of which 3 evergreen and 7 decid-

uous. Dusen (1899), Gothen, and others described various macrofossils, attributed to

Fagus etc. They are certainly Nothofagus species. fossils According to Archangelski (1970) no Nothofagus are recorded from the South

American (Argentine) Cretaceous. In the Eocene flora of Rio Turbia (Santa Cruz Prov-

0 is It ince), at 52 S Nothofagus dominant (leaf impressions, wood). was said that Fagus

wood occurs here, but no pollen records sustain this.

0 In the Oligocene floras of Rio Pichileufi (Bio Negro Province) at 41 S in Argentina,

and the similar one at Concepcion-Arauco, at 37° S in Chile no Nothofagus was found.

Recently Cookson & Cranwell (1967) have published the first Nothofagus pollen records

from the Fuegian mainland in the valley of the Rio Lena Dura at c. 53° S. The age of

these strata is Eo-Oligocene (probably Eocene) and they contain pollen of the brassii

and fusca types.

Summarizing this chapter, the main results are:

( i) Nothofagus species of the brassii pollen type existed in Australia, New Zealand, and

Antarctica (Seymour I.) already at least from the Upper Cretaceous. the (ii) This pollen type was abundantly perpetuated and is known in Tertiary fromNew

Guinea (onwards of the Upper Miocene), Australia, New Zealand, Seymour Island,

and Fuegia. the the known from (iii) Nothofagus species of fusca pollen type: oldest occurrence is the

Upper Cretaceous in Seymour I. (Antarctica); it occurs further widely distributed all

over Australia, New Zealand, ?McMurdo Sound, and southern South America.

The earliest know wide record of the menziesii pollen type is as far as we of less dis- tribution, dating in New Zealand from the Eo-Oligocene; older records from New

Zealand are as yet uncertain.

is there been (iv) The impression that must have many dozens of botanical species. The further that if searched (v) impression is more places were as intensely as New similar abundances will be found and filled the record. Zealand, gaps in pollen from (vi) Including the fossil records there is thus a tolerably complete, continuous range

New Guinea, through Australia, New Caledonia, Tasmania, New Zealand, northern

Antarctica to southern South America.

The this should summary on impressive pollen record include a statement on the rec-

ords of Nothofagus pollen from the northern hemisphere. It is curious that claims of from the fossil Nothofagus pollen the northern hemisphere keep appearing, even in past

Khin & decade, e.g. by Potonie (i960), Ma Sein (1961), Ames Riesel (1962), Kedves

also the for (1964), Ramanujam (1966), and Kuprianova (1967). This is case Podocarpus and identifications debat- pollen. Significantly, they represent rare occurrences are

able or definitely erroneous. If some are genuine I feel inclined to attribute their occur- forests rence to long-distance dispersal from the south at a time when the Nothofagus

far than at Such were more extensive, especially in Australia, they are present. long- of the distance dispersal has, admittedly in the southern zone roaring forties, beenfound

Hafsten. Such accidental to occur, notably in recent peats on Gough I. by occurrences

can, however, certainly not be taken as an indication for the existence of Nothofagus

the northern or Podocarpus trees growing on hemisphere. This conclusion is agreed on and Cranwell by the great Nothofagus pollen specialist Couper (in litt.) by Mrs. (1963) and reflects also the opinion of Muller (personal communication). C. G. G. J. van Steenis: Nothofagus 83

11. PHYLOGENY OF NOTHOFAGUS, IN TIME AND SPACE

ofthe southern An interpretation origin and spread of the hemisphere genus Nothofagus be cannot divorced from considering that of the northern hemisphere genus Fagus, the

pair forming together the subfamily Fagoideae. At present the smallest distributional

between these two and southern China and gap genera is Fagus in Formosa Nothofagus

in New Guinea.

In this connectionit is noteworthy to observe that within this same area, ranging from the Guinea all other of Sino-Himalayan area to New and Queensland 5 genera Fagaceae¹) and are represented. Four ofthemoccur inprofusion in Malesia, viz. Lithocarpus Castanop-

sis ofthe Castanoideae and Quercus and Trigonobalanus of the Quercoideae, only Castanea (a

close ally of Castanopsis) being Sino-Himalayan-Japanese. Thus the the at present entire major morphological diversity is represented in area between Yunnan and Queensland.

This leads to the question whether this centre is of secondary origin or whether it is a

Fagaceous cradle-area persisting to the present day.

Besides the fact that in this region almost the entire gene pool of the family is concen- there additional the trated, are two important arguments for accepting it as cradle-area of the family.

the first fossils not recorded the In place it is argued that, though pre-Tertiary are yet, fossils of in the southern Pacific sites indubitable Upper Cretaceous Nothofagus prove that before that the the tie of Nothofagus withFagus must date from even time; ancient ances-

tral Fagaceous matrix from which they differentiated must have existed in the present transtropical disjunction of Fagoideae between southern China and Queensland. the of Secondly, there is remarkable genus Trigonobalanus, which one species, T.

verticillatus, is endemic on the Sunda shelf (Malaya, NW. Borneo) and the other, T.

Thailand. This remarkable several in the first doichangensis, in N. genus is in aspects, the the rather and place as vegetative morphology of two species is different, in the essential characters wood second place because its more generic (floral structure, anatomy,

pollen) show partly Fagoid and partly Quercoid affinity or features. Even so much that,

whereas Forman (1964) reckons Trigonobalanus to Quercoideae, Soepadmo is inclined to find with its affinity somewhat more Fagoideae.

An other recently found significant aspect of Trigonobalanus is, that T. verticillatus has

2 42 ± 2 chromosomes ) in the diploid and c. 21 in the haploid fase. This strongly contrasts

with chromosome counts in other Fagaceae where the numbers 20 and 24 in diploid fase

prevail. Obviously, Trigonobalanus is an ancient (? dysploid) palaeo-allotetraploid, vir- relict tually a which could well stem from an ancient Fagaceous matrix.

that the consists of Finally the fact genus Trigonobalanus two rather different local the character This species point also to relict of both. conclusion has received recent

support by Mai (1970) who accepts that several species of Trigonobalanus occurred in the from the European Lauro-Fagaceous mesophytic forest climax mid-Eocene to the mid-

Miocene as important dominants or co-dominants, associating the cupules and nuts with described leaves of a fossil genus of Fagaceae, as Dryophyllum.

The above-mentioned three arguments represent, I believe, convincing evidence that

Even *) if an 8th genus, Chrysolepis, is admitted, as is done by Forman (1966), ofwhich one species

be the Dr. communicated occurs in California, this seems to represented in cradle-area, as Soepadmo to me

that if this is second in Indo-China. genus recognized a species occurs 2 I this vital information who is this ) owe to Dr. Ding Hou preparing a paper on subject. BLUMEA VOL. No. XIX, I, 1971 84

the region where the present concentration of Fagaceous morphology is found, broadly

ranging from Yunnan to Queensland, must be the cradle-area of Fagaceae where the the located. ancestral matrix to present genera was

in I for the Only passing remark that possibility of a neotropical Fagaceous lineage

the there are no such valid this be elimi- across tropics arguments; neotropical origin can nated with confidence.

is obvious that for also search for its real It then Nothofagus ancestry we must cradle the needs this somewhere in Indo-Malesian region. One not to pin down in New Guinea

or any other single island, but consider the entire region from Yunnan and Indo-China

through Malesia to Queensland as the matrix area of the ancestral stock of which we material. And ifthis found have no fossil was it would not even be certain that we should

be able to recognize it. In this connection I might point in passing to the fact that if would been described Nothofagus were entirely extinct, the pollens have as some 'Stephano-

but also the porites', not referred to Fagaceae¹). We can not hypothesize about question off from of whether Nothofagus split a sort primitive Fagus or the reverse. It would be

reasonable to ancestral distributed from more assume a common generic stock, Yunnan the Chinese of to Queensland, in more northern, southern part which a genus emerged northern which led to present hemisphere Fagus and independently in the more southern

somewhere the - ofwhich what part in New Guinea Queensland area emerged we now

know as the southern hemisphere Nothofagus.

The assumed cradle-area of coincides in marked with the Fagaceae a way gradually

developed idea that this Yunnan-Queensland area could well have been the critical

theater in which Angiosperms evolved. This has been stressed recently by Takhtajan and A. C. Smith the latter also (1969) by (1970), defending my steady state concept as and far as evolution of Phanerogams is concerned, saying that: 'After establishment

initial diversification in this' (Indo-Australian cradle-) 'area, they spread over the rest different the of the world's surfaces, which in configuration were not vastly in Middle

and Period I wish make Late Cretaceous than they are today'. Though do not to any about the absolute of their record suggestion age Amentiferous orders, fossil is among

the earliest their is old that of Ranales. known, though ancestry possibly not so as I believe it is futile to speculate beyond the general importance of this cradle-area.

be to derive from table from which that far known It might tempting 1, appears as as the fossil and New Cale- the brassii pollen type is only type among living Nothofagus in

donia and New Guinea, that Nothofagus arrived there later, after having evolved in

northern Australia. This could then also lead to the view that the lowland occurrence of

brassii in New Caledonia would be extension of the the type a secondary megatherm generic ecology. in few About Fagaceae on the northern hemisphere, and Fagus particular (see map 1) a

words first. There are abundant macrofossils on the northern hemisphere, and it seems

certain according to Kirchheimer (1957) that they partly belong to several genera of

*) As regards pollen morphology, Soepadmo (1968: 362) stated that the pollen of Castanopsis and Litho-

to that of and to second whileNothofagus carpus belong one pollentype, Fagus, Quercus, Trigonobalanus a type, stands with third distinct The close of and does quite apart a type. affinity Lithocarpus Castanopsis, therefore,

show in the On the other the difference between the of and up pollen morphology. hand, pollen Fagus

Nothofagus does not reflect their taxonomic affinity and their forming together a distinct subfamily of the Fagaceae. For establishing systematic affinity, pollen morphology is often, necessarily, less valuable than complete

macrofossil material, because the latter possesses far more criteria for the interpretation of morphological

structure of the complete plant. C. G. G. J. VAN Steenis: Nothofagus 85

Fagaceae, but some to other groups (Alnus). Unfortunately the pollen of the northern and features hemisphere Fagaceae gives no clue possesses no peculiar as Nothofagus. Any- indubitable amber way Fagus is found in of Oligo-Miocene age. Weyland (1964) admits

Fagus also for the Tertiary and possibly also for the Cretaceous, saying that such forms

cannot immediately be tied up with recent species (which is not surprising). There are these uncertain of Cretaceous for fossils the Upper and Tertiary form genera (Dryo-

phyllum¹) and Quercophyllum) which could well belong to an ancient matrix which be- which came largely extinct but from modern Fagus emerged. Summarizing: Fagus the but had then obviously became clear in Oligocene, already a lineage in the warm- northern behind There chance temperate to temperate hemisphere it. remains a good that it dates back from the Upper Cretaceous.

occurs cradle Although Nothofagus at present in the supposed area only in New Guinea

and a few adjacent islands, its ancestors might have occurred in other Malesian islands or

is sustain North Queensland as well, but there as yet no fossil pollen to this. No Nothofagus pollen has been found during the extensive research done in Sarawak

and Brunei in strata ranging from the Middle Cretaceous to the Pliocene according to Muller and (1968, pers. comm.).

Also in NW. New Guinea the fossil Nothofagus pollen record is meagre and only located

as far down as the Upper Miocene. But it would be too rash to conclude from this

hitherto negative evidence that Nothofagus migrated into the Arfak Mountains only in needed geologically recent time. Far more data are from other strata in New Guinea to

give evidence pro or con of its early history in this huge island.

remark that that the Notho- In passing I may one could hypothesize upland growth of could been for of fagus in New Guinea (c. 750—3100 m) have responsible the absence

pollen in marine sediments and that the upland deposits havebeen destroyed. Such reason-

ing has sometimes even been made to explain the sudden origin of Angiosperms thought

be an which later the to originally upland development only replaced gymnospermous

lowland forests, an idea repeatedly advanced by Axelrod (1970). From research done with recent pollen and experience with fossil pollen this is, however, untenable: most pollen direct catchment soil gathered by montane river systems, by or by erosion is transported

downstream and incorporated in deltaic and offshore sediments. Also a considerable the like part of pollen production, especially of anemophilous trees Nothofagus, is directly

transported to the sea and settles in the marine sediment. In Venezuelan marinesediments and mangrove pollen pollen from montane plants is consequently found together.

Anyway New Guinea occupies together with New Caledonia certainly a key position

in that they harbour species belonging to subsect. Bipartitae of which all species are char- acterized of the and of the by one two most ancient most widespread types pollen, brassii type. therefore reasonable that formed least of the It is to assume New Guinea at part region of the ancient Fagaceous matrix from which Nothofagus emerged.

From the fact of the Upper Cretaceous occurrence of Nothofagus pollen in Australia,

New Zealand, and Seymour I., the conclusion is that the cradle of Nothofagus must date at

least from pre-Upper Cretaceous time. And this was probably Early Cretaceous in view of the needed obtain the wide distribution the time to already present in Upper Cretaceous and the inevitable conclusion that it must have been tied and from up sprung an early

This leaf has been tied Mai with in inter- *) type quite recently up by (1970) Trigonobalanus an extremely

which into hands while the MS was in esting publication came my already press. 86 BLUMEA VOL. No. XLX, I, 1971

Fagaceous matrix which — if not on the northern hemisphere — must have been at the

closest distance at least in the Malesian (montane) tropics. To the ofthe ancient is regard present species array as representative stock very unlikely, but there that with lamellar is reason to assume species a well-developed cupule containing

three nuts and male flowers in triads are morphologically undoubtedly more primitive

than species with an elamellate or even ecupular nut and solitary male flowers.

of this are found and also Species primitive category in New Guinea, in New Cale-

donia, the land crowded with ancient primitive plants. could well be that the of the Caledonian It supposed genepool New species approaches

a slightly more ancient condition, because of the more varied ecology (both lowlandand

mountains) as compared with the New Guinea assemblage which lacks the capacity the and the less to grow in proper lowland, strict phyllotaxis.

It be that the may original stock of ancient Nothofagus brassii type in New Guinea

the and menziesii it be in ancient developed fusca types, may that this occurred Queensland.

The great fact is that Nothofagus brassii ancestors succeeded in spreading (almost)

throughout Australia and to New Zealand, and onwards to Seymour Island and Fuegia. Whether this of Tasmania New Caledonia is was by way or by way of questionable. harbours Tasmania N. gunnii, a characteristic deciduous species of sect. Nothofagus, which is from New southern absent Zealand, but reappears in South America, where is found

its nearest congener N. antarctica. N. gunnii is certainly a primitive species in sect. Notho-

fagus, by its 4-valved cupule with entire lamellae and containing three nuts.

The pollen of N. antarctica and N. gunnii belong to the fusca type, which is, according the fossil the second-oldest recorded for the to record, precisely pollen type as by Couper

Upper Cretaceous.

In America Australasia and South the fusca pollen type occurs among the living repre- of habit sentatives both evergreen and deciduous species. It is likely that the deciduous in Australia. It well be it in ancient already developed may that already existed the brassii the matrix; unfortunately, pollen can give no clue about this character for the fossils.

to that and As New Caledonia, it is curious no fusca menziesii types occur. It might the in point to an early distribution of brassii type the Cretaceous, but later isolation the (with New Guinea) from Australian speciation centre where the fusca and menziesii

this trend of would that Caledonia did not types developed. In thought it appear New of serve as a post-Cretaceous pathway Nothofagus towards New Zealand. The New

Hebrides mountains must be still explored.

In Australia recently the fossil recordwas extended to the Upper Cretaceous in Victoria. The further in is moist abounded story Australia rather simple: the warm, Tertiary with

of all three the onset of the desiccation and the species pollen types; Ice Age were respon- and sible for a wholesale extinction in the Pliocene, from which the most temperature

moistness-sensitive of the brassii fell victim and suffered extinction. species type a entire survived restricted of the A meagre triplet in extremely places, one fusca type (N. gunnii in the Tasmania), two of menziesii type (N. moorei and N. cunninghamii), vestiges of a

glorious past.

As to New Zealand, at least a Cretaceous land connection with Australia or Tasmania be because must suggested, in the first place the oldest brassii type reached it in the Upper but besides the and menziesii in in Cretaceous, fusca types were there profusion the Upper Cretaceous And ofthe and Early Tertiary respectively. as Couper remarkedsome Tertiary New Zealand similar Australian which the pollens are very to ones, includes possibility

that they had species in common on this Australasian continent. There is of course no

necessity to assume this 'continental land' to have been intact all through this epoch, but C. G. G. J. van Steenis: Nothofagus 87

there must have been periods of intense contact to allow dispersal of these southern beeches.

which have with In the New Zealandian Cretaceous region must had land connections Antarctica least in and the as well, at both the Early Cretaceous in Early Tertiary as testified the finds of the brassii Island and the of the and by type in Seymour occurrence menziesii shown fusca types in South America, there must have been a profuse speciation as by the

many fossil species described and occurring throughout the Tertiary. Not all species survived this whole period, some became extinct in the Miocene, or even Eocene. And

there was an abrupt end to all of the brassii type and most of the other two types in the

Lower Pliocene, an extinction doubtless due to the onset of the Ice Age. in Whether there have been the Australasian region very primitive species with more than the derived with 3 nuts per cupule (as in South American N. alessandrii) or types adorned with of such still South cupules an intricate pattern appendages, as are living in America, is of unknown. the carried the which course But gene pool by representatives containedall three and the borderland of have been types spread along Antarctica must richer than the distinctly is reflected by two extremely meagre but most important

fossil finds at McMurdo Sound and Seymour Island. In both places, however, the accom- panying fossil assemblage is similar to that found in New Zealand at comparable geolog- ical this age, giving assumption a not unreasonable background.

The Cretaceous-Tertiary borderland antarctic Nothofagetum will have allowed

for N. only hardy species, evergreen and deciduous, similar in ecology to N. antarctica, and pumilio, N. cliffortioides which also now grow to the tree limit at 55° S. The antarctic

have been similar those the northern tree limit conditions at c. 70° S will to in hemisphere, with denser, higher forests along the coastal regions, and more stunted, gradually dwarf- ing and spaced to the tree limit.

As we have seen there are from southern South America Eocene records of the occur- of the brassii and and there with the and rence fusca types are living species menziesii fusca the is pollen types. Though record very incomplete, through the meagre research as yet reasonable that will have differed performed, it is to expect the situation in Fuegia not from that in Australia and New Zealand. This means that it can confidently be expected that there willbe continuous the also a record for thefusca type till present, and probably the while be that the brassii till of menziesii type, it can expected type will be found up the Pliocene. There is also good reason to assume, judging from the abundanceof macro- fossils and the of and with the finds similarity accompanying trees, comparison on that the been in the Seymour I., Nothofagus assemblage must have very much richer than The deterioration of the climate Tertiary it is at present. through the Tertiary, but the the with initially slow, gainingstrength towards end of Tertiary, came to a climax

1 thePleistocene Ice Age which played havoc with much of the austral mesophytic forest ).

*) The destructive influence of the deteriorating climate on the 'Arcto-Tertiary' Fago-Lauraceous

in the mesophytic forest on the northern hemisphere resulted a reflected image of same process of gradual extinction, finally accelerated towards the end ofthe Tertiary with the onset ofthe Ice Age in the Pleistocene. China and local forest because Except in areas in West Asia most of this mesophytic became extinct, the east-west directed mountain massifs did allow for for few not escape possibility southward, except a genera, e.g. Liquidambar, Platanus, etc., which managed to survive as local relicts in the subtropics. In China the

forest could southward with the southward latitudinal of the mesophytic zone move move warm-temper- ate and subtropical climate zones, and furthermore maintain its footholds in Japan and Formosa thanks to the ameliorate oceanic climate of these West Pacific lands.

This broad sufficient for the whole of extinction of explanation is, however, in my opinion not story

these much — this mesophytic forest, as several of relicts are now so very restricted as for example Sequoia,

the Platanus and — that additional Sequoiadendron, primitive kerrii, Metasequoia, Trigonobalanus, many more BLUMEA VOL. XIX, No. 88 I, 1971

crosses.

by indicated

fossils Nothofagus;

and

Fagus

of Distribution

1.

Map Steenis: C. G. G. J. van Nothofagus 89

found It could be questioned why Nothofagus, which is now living to 33° S in S.

far the record is in the S. America, as as meagre fossil goes absent Oligocene at 41—37°

Especially if one takes into consideration that it must have had time to spread during

almost the entire Tertiary.

A reasonable explanation for this could be that by the passage eastwards along the

could cold borderland of Antarctica, only species succeeded which stand a temperate to and in this ancient of the ancestral stock climate, way the capacity to produce more the selection. megatherm species was underway eliminated from gene pool by

This microthermic ecology must also have been shared by species of the brassii type.

This is still of this several in New present among living species subsection; species are

Guinea only found in the subalpine zone, for example N. resinosa and N. pseudoresinosa

of hardiness. which with the (2300—3100 m), token They are not microphyllous, agrees

Fuegian macrofossils. In living Nothofagus leaf-size itself is not a measure for hardiness.

12. BI-HEMISPHERIC RANGES HOMOLOGOUS WITH THAT OF FAGOIDEAE

& (Maps 1 2)

Ranges matching or approximating that of Nothofagus are in the southern Pacific not that of the northern rare. Ranges similar to Fagus are common on hemisphere.

there are few which show a similar of But very pattern an undoubtedly genealogical that of such the of and lineage, is a higher order, as Fagoideae, ranges Fagus Nothofagus

combined into a characteristic bi-hemispherical range, 'hour-glass-shaped' narrowed in

the Malesian middle. tropical See map 1.

have that In chapter 9 I already referred to the fact there are, at several levels (genus, and which northern subfamily, family) bi-hemispheric relationships, in one counterpart is affinities and the other southern. Obviously such are lineages sprung from a common

ancestral matrix which is likely to have been situated halfway, that is, in the tropical

and subtropical zone.

It is not surprising that not all taxa of higher order which occur on both hemispheres still this exhibit picture in a very clear and distinct way. Starting from these ancient

divided into matricesit will usually not have occurred that the gene pool was sharply two distinct but would the the rule. parts; this might occur seem exception rather than Thus find that for we example Ericaceae and Epacridaceae are not sharply counterparts, as for

example Gaultheria is found both northern and southern. It can be expected, therefore,

that sharply separated complements are rather rare.

Such distributions are found in all sorts and to various degree. They have often been obscured later evolution the and become then by deviating in counterparts no longer have different recognizable. Evolutionary trends and speed, and rate of spread, been very

for various groups, and extinction as well. The term 'random' is here not well in place,

one could better call the situation 'erratic', by which I mean that there is no single rule

its and about this for all groups. Each group behaved, according to capacities potentialities, let alone in an opportune way. One can simply not imagine, explain, why certain taxa

has be called the deterioration of climate insufficient explanation to for, (cold or drought) being as a single

I is factor. This touches the general problem of extinction, am afraid, which still very much obscure. biotic factors role in this than is understood. think this Possibly, may play a more important presently I in in elm matter of viruses, bacteria, and fungi plants, as suggested by the and chestnut diseases, and the

large-scale extinction of Zostera marina. Also in the fauna colossal shifts seem to occur by biotic factors

as shown by the recent destruction of corals in the Pacific. BLUMEA VOL. XIX, No. i, 1971 90

differentiated to a large degree and why form-making was poor in others. It goes beyond

certain of our comprehension why taxa developed a large number species or genera, and others others did not, and especially why some diverged and kept their precise identity similar Whether for very long geological epochs, as Gingko, Araucaria, etc. we will ever be able to dig such secrets out of the genome structures where they must be seated, remains questionable.

However, Fagoideae is a remarkably good illustrative example of 'keeping identity', as be and the its genealogical lineage cannot doubted, ranges of Fagus and Nothofagus are

exemplary complementary. See map I.

Both Fagus and Nothofagus are not especially microtherm, but they are certainly largely

mesotherm to temperate. Nothofagus has the largest eco-climatical amplitude of the pair

by growing at present in the tropics besides montane and subalpine montane also with

some New Caledonian species in the tropical lowland.

reasonable to that mesotherm and microtherm are the Therefore, it is assume groups

most likely to show a similar bi-hemisphere distribution through the Malesian montane the pathway connecting two hemispheres. have erratic Here, again, we will to face an array, 'virility', keeping identity, extinction, of these themselves etc. Some genera or species will show to have been able to maintain the on single tropical montane stations, others became extinct in tropical zone. They are various then expected to represent equiformal or homologous ranges to degree.

within this of mesotherm in Besides, group to temperate taxa, they vary thermophilous and be response capacity, and none can be expected to exactly identical in this respect other. have behaved to an They as an assemblage varying from tropical montane to alpine.

treatise flora of attention In my of the Mount Kinabalu (1964) I have called to this

of and of which almost in- most interesting group plants given examples ranges appear that with tact, as ofEuphrasia, but addedothers which have only still a few stations: Poa 3 and (Kinabalu, Central Celebes, New Guinea), Oreomyrrhis with 2 (New Guinea Kina-

balu), Trisetum with 1 (Kinabalu), Eryngium with 1 (Ceram), Centunculus with I (Timor). there But are numerous others which have no present tropical-montane station in the

Malesian tropics although their range extends over it, for example: Aphanes, Asperula, Barbarea, Caltha, Capsella, Centaurium, Eritrichum, Erodium, Gypsophila, Lepidium, Limosella,

Linum, Lotus, Lythrum, Mentha, Myosurus, Papaver, Prunella, Samolus, Scleranthus, Thlaspi,

Tillaea, Trigonella, etc. etc. The become range of such genera can no longer be called bi-hemispheric, but has

bipolar when their stations in the tropical and sometimes adjacent subtropical zones have and their northernand southern considerable disappeared partial ranges are separated by a transtropical disjunction.

also be sometimes aberrant found the It is to suspected that ancient or species are in

isolated tropical montane stations, and in this respect I remind of the rather aberrant

features ofEryngium moluccanum and of Ranunculus fasciculiflorus, both alpines of Mount Binaya in Ceram I. (Moluccas).

Imagining the vicissitudes of the geographical physiography of Malesian tropics

during the huge period from the Cretaceous to the present and the unknown biogeo-

graphical role played by the Tethys Sea, it is clear that only by exception represent- of such could themselves. lived atives taxa, mostly genera, maintain They a precarious life. With mountain ranges dwindling down by tropical erosion footholds were lost, and others for all by migration were gained; it was: to spread or die. C. G. G. J. van Steenis: Nothofagus 91

Spreading in a geological period means inevitably jumping from one decaying moun- another tain to coming up. Hence, I have emphasized (van Steenis, 1967) that many will be older than the mountains be plants they grow on at present; they may even

endemic and nonetheless not have had their cradle on the mountain on which they are endemic. This will be clear who now to any chorologist wants to get some insight by historical least faint idea of the thinking in terms necessary to get at some implications of the time dynamics of plant wandering.

It is therefore not likely that species, for example of Mount Kinabalu, originated

there; the occurrence of Euphrasia certainly must be ascribed to a later accession. It is

that mountain of lowland real neo-endemics. Mountains more likely species genera are with rising in lowland countries or peneplains no neighbouring mountains with a moun- flora available for will and mountain flora recruted tain exchange, develop carry only a

from the lowland genera, say ofLauraceae, Symplocos, Ericaceae, Theaceae, Myrtaceae, etc.

Such situation the Malesian with their a is of course seldom found, especially in tropics

and would account for for the widespread active orogeny, present past. But it instance

almost absence of true mountain plants in Guadalcanal, which developed obviously

without access of such alpines within effective dispersal distance.

have forward these considerations for a better of the I put understanding generic of which is in remarkable of range Euphrasia (Scrophulariaceae) my opinion a example a the also range which is homologous with that of Fagoideae. It is ecologically comparable

somewhat cold with the understanding that it is eco-climatically more temperate to fashioned. Therefore the somewhat elevation than it occurs in tropics at higher Nothofagus, from the northern viz. 2000—4000 m. Consequently its range on hemisphere goes farther and that of suffered also less from north is more extensive than Fagus. Probably it the than than Pleistocene Ice Age Fagus, and spreads furthermore more rapidly Fagus.

See map 2.

A of further between the of and significant point agreement ranges Fagoideae Euphrasia

is that, similarto the situationin Fagoideae, at present the greatest morphological diversity the southern Zealand of Euphrasia is found on hemisphere. In New even two endemic entities, obviously representing specialisations, are sometimes regarded as closely allied, but viz. and separate genera, Anagosperma Siphonidium. other with the be the An parallel Fagoid range may represented by Veronica-Hebe- Parahebe-Detzneria complex. This also shows a most remarkable form-making in the austral region. One would even be inclined to parallelize the position of Hebe in this

complex with that of the Bipartitae in Nothofagus, that is, the most ancient type.

It is understood that of the easily many 'hour-glass-shaped, bi-hemispheric, trans- have Malesian-tropics ranges' (see p. 77,89) or sometimes bipolar ranges (see p. 90) come ofland the side of Pacific from the into being by means areas along western the ranging the Siberian shield to Antarctica, notnecessarily continuous landall time, as some unimag-

inative opponents of the landbridge, better 'land theory' inplant-geography, sometimes unreasonably have deduced.

It also involves to accept the steady state principle in geomorphology, at least as far as and that least from the Conifers, Taxads, Angiosperms are concerned, is at Cretaceous. There be such can no question to explain ranges by a southern Gondwana land origin, classic the continental as correctly pointed out by Diels (1928) in his critique of drift theory

inits original concept for the Australian-Papuan flora. Anyway, it requires a close contact between northernand southernland masses (ofAsia and Australia) in theEarly Cretaceous. BLUMEA VOL. XLX, No. 92 I, 1971

Euphrasia.

of Distribution

2.

Map C. G. G. J. van Steenis: Nothofagus 93

in the 1 because be well that all fossil And may be even Upper Jurassic ), we must aware finds absolute And has give a minimum age, not an age. experience taught that with ofresearch in hitherto almost all be older than progress palaeontology groups appear to considered half a century ago.

It stands to reason that the 'phylogenetic chorology', that is phylogenetic development

extension in time and of and and other and range space, Fagoideae Euphrasia many taxa their have If the of distributional type (see pp. 77, 90) must gone synchronously. Fagoideae

from the that of and the other range dates Upper Cretaceous, so must Euphrasia, taxa. be of It is also clear that bi-hemispheric ranges of this kind must high antiquity. that such find still in From this follows ranges as we largely essentially unimpaired

and be rare the because such Fagoideae Euphrasia must in present plant world, to maintain characters which still with full confi- range and besides maintain taxonomic by we can

denceconsider the taxa as representing a true genealogical lineage (a genus as Euphrasia, that a complex as of Hebe-Veronica, a subfamily as Fagoideae) it requires a high taxonomic and towards ecological stability against evolutionary pressure or tendency change, towards and will have been diversity consequent frequent extinction. Most plant groups and evolution will have caused less stable in these aspects a so much chequered phylogeny will that taxonomists only very hesitatingly suggest lineages, or are even not capable to with distinguish lineages (‘Sippen’) any confidence. The uncertainty of discerning such in and is the lineages more rapidly evolving diverging groups enlarged by greater prob- that such has been and has led obliteration ability in groups extinction more severe to of too affinities and face us with too links. many many missing

THESES 13. CONCLUDING

The detailed study of Nothofagus, recent and fossil, leads to several conclusions of more for and for general bearing, both taxonomy chorology.

Because of the restricted means of effective dispersal of Nothofagus by which long-

distance dispersal is indubitably excluded, and because of its wide and well-known distribution and and fossil evidence of ecology, an ample trustworthy great age, it is a pre-eminent example for geobotanical conclusions. have alluded of the rules which In the preceding chapters I already to some general

we can derive from this detailed study, and the corollaries involved. Clearly a 'New needed from Chorology' is in plant-geography. I am aware that it will drastically differ

some currently accepted views.

I find it desirable to summarize into a number of theses.

(i) The present centre of morphological diversity of Fagaceae where all genera still

x ) Axelrod (1970: 285 —290) has convincingly summarized evidence and reasoning for a pre-Cretaceous origin of Phanerogams.

On pollen evidence Muller (1970: 428, 444) concluded that 'except for the possibility that the Magnoliales

considerably preceded the development of other groups, the palynological evidence does not support the view that the Angiosperms had behind them a long evolutionary history, starting as early as the Permian',

as postulatedby Axelrod and Eames. He tabulated the plants possibly ancestral to Polycarpiceae (‘Clavatipolle-

nites’), from the Upper Jurassic,but the 'modern' branches of Phanerogamsfrom the Cretaceous. However,

Muller (I.e. 429) leaves the door open for presence of their pre-Cretaceous ancestors which could possibly have had primitive pollen types — evolution of pollen types lagging behind morphological

evolution — and admitting that palynology has some (regrettable) limitations. No. 94 BLUMEA VOL. XIX, I, 1971

occur, viz. from Yunnan to Queensland, appears to be the 'cradle-area' of the family Fagaceae which harboured its ancestral matrix (chapter u).

In the southern part of this matrix Nothofagus differentiated and spread in austral

other in with the the regions, genera evolved this matrix spread mesophytic forests over

northern hemisphere, westwards towards Europe and eastwards via Beringia to North

and Central America.

Land (ii) A Gondwana origin ('cradle') of Fagaceae is taxonomically an impossible

assumption, a biphyletic origin of Fagoideae is still more so.

(iii) The fact that Nothofagus occurred already in the Upper Cretaceous in Australia,

New Zealand, and Antarctica, and that its birth-place from the ancestral Fagaceous stock have the can at most been closest to the fossil sites in tropics, involves that it must the of the stock have taken some time to reach New Zealand, which dates age ancestral

back to the Upper Jurassic, at least Lower Cretaceous.

themselves derived which (iv) As Fagaceae are morphologically certainly plants cannot

in the least be tied with other up directly Pteridosperms or potential Gymnosperm ances- than the tors, the origin of Angiosperms must be considerably older Cretaceous.

ofthe and (v) The shrinking genus Nothofagus in space size is exemplary for extinction South Pacific brought along by the two major environmental changes affecting the at

the end of the Tertiary, viz. desiccation of the Australian continent and the Ice Age.

The ancient in vi) range of Nothofagus canbe explained an unrestrained way under the

steady state theory, provided the filling at some period in the Upper Cretaceous of the between the borderland of and South and present oceanic gaps Antarctica America Australasia before the the brassii and reach Upper Cretaceous (allowing fusca types to Seymour Island in the Upper Cretaceous and Fuegia in the Early Tertiary).

occurred (vii) If continental drift ever in the austral regions it must have taken place the if long before Cretaceous. The Fagaceous matrix can only have developed already in

the (at least) Late Jurassic southern and northern hemisphere continental lands have been

in contact in the Malesian tropics.

of that (viii) The occurrence ranges which are equiformal (homologous) to ofFagoideae and which show a similar bi-hemispheric either level affinity on family (see p. 77) or

level assumed an on generic (see p. 90) is to point to ancient synchronous development of unrelated the both and herbaceous many groups in plant kingdom, involving ligneous

plant groups.

(ix) For the cradle of these lineages the Sino-Malesian subtropics and montane tropics crucial represent a region.

Among these lineages Euphrasia is exemplary in exhibiting a remarkably unimpaired,

ecological similar distribution and speciation pattern to such an extent that it must be of concluded that this distributionis as ancient as that Fagoideae, Magnoliaceae/ Winteraceae, and other mentioned many groups (pp. 77, 90, 91,).

(x) Although Euphrasia and many other lineages are morphologically far more 'ad- C. G. G. J. van Steenis: Nothofagus 95

with vanced' in the taxonomic system as compared Fagaceae or Magnoliaceae/ Winteraceae, they must be of great antiquity and date back to the Cretaceous. the We must obviously abandon current view that morphologically advanced groups

are necessarily of young age.

have the view size (xi) We to abandon common, light-heartedly accepted that and of structure of diaspores is necessarily a measure for degree of success dispersal. Range extension in primary vegetation goes slowly, by plant associations and ecological groups and rather independent of the structure of diaspores.

(xii) The degree of evolutionary development, form making, and extinction is un- predictable and erratic (see p. 75).

Biogeography is, however, not erratic as plants which evolved more or less simultane-

ously will take (or at least have) the same opportunity to spread according theirecological and the ahead of the all: capacity geomorphological future them, which is same for plants spread togedier.

ACKNOWLEDGEMENTS

I wish sincere thanks Mr. Head of the Division of for to express my to J. Womersley, Botany, Lae, this Port information and for his reading an abstract of study at the A.N.Z.A.A.S. meetings at Moresby,

which is in Transactions August 1970, to be published the Proceedings and of the Papua & New Guinea Scientific Society, Port Moresby.

Warm thanks are due to Mr. J. Muller,palynologist at the Rijksherbarium, Leyden, for his assistance in

the his constructive Mr. M. M. completing published data on fossil pollen and criticism, and to J. van

of the for assistance in Balgooy, Rijksherbarium, drawing accurate maps. I Dr. E. and information he various appreciate the discussion on Nothofagus with Soepadmo provided on

I subjects ofFagaceous systematics. am finally thankful for obtaining the unpublisheddata on the chromo-

from somes of Trigonobalanus Dr. Ding Hou.

BIBLIOGRAPHY

W. AMES, H., & RIESEL, 1962. Pollen et Spores 4: no. 2.

S. Univ. La Fac. Cienc. Nat. & ARCHANGELSKI, 1970. Fundamentos de paleobotanica. Plata, Museo, ser. Tecnica Didactica y no. 10 (Revista): 347 pp., 213 fig., 22 tab., especially pp. 294—296.

ARCHBOLD, R., A. L. RAND & L.J. BRASS. 1942. Summary of the 1938 —1939 New Guinea Expedition.

Am. Mus. Nat. 282. Bull. Hist. 97: 201—288, especially p.

AXELROD, D. I. 1970. Mesozoic palaeogeography and early Angiosperm history. Bot. Rev. 36: 277—319.

W. Der Schweiz. BACKHUYS, 1968. Elevations-Effekt bei einigen Alpenpflanzen der Blumea 16: 273 —320.

BADER, F. J. i960. Die Verbreitung borealer und subantarktischer Holzgewachse in den Gebirgen des

Nova Acta N. F. Tropengiirtels. Leopold. 23, no. 148: 544 pp., maps.

BAILLON, H. 1873. Nouvelles observations sur les Adansonia 11: 136—137 (Trisyngyne).

Mus. Paris BAUMANN-BODENHEIM, G. 1953. Fagac6es de la Nouvelle Caledonie. Bull. Hist. Nat. II, 25:

419—421.

1956. Uber die Beziehungen der neu-caledonischen Flora zu den tropischen und den siidhemi-

spharisch-subtropischen bis -extratropischen Floren und die giirtelmassige Gliederung der Vegetation

Ber. Inst. von Neu-Caledonien. Geobot. Forsch. Riibel, Zurich fUr 1955: 64—74. N. T. Bot. BURBIDGE, i960. The phytogeography of the Australian region. Austr. J. 8: 75—212, especially

three the maps, fig. 9—11. COOKSON, I. C. 1946. Pollens of Nothofagus Bl. from Tertiary deposits in Australia. Proc. Linn. Soc.

N.S.W. 71: 49—63.

New New 1952. Identification of Tertiary pollen grains with those of Guinea and Caledonian beeches.

Nature 170: 127.

in Western Austr. Sc. 1954. The occurrence of an older Tertiary microflora Australia. J. 17: 37—38.

1959- Fossil pollen grains of Nothofagus from Australia. Proc. R. Soc. Vict. n.s. 71: 25—30. BLUMEA VOL. No. XIX, i, 1971 96

CRANWEI and from southern- & L. M. L. 1967. Lower Tertiary microplankton, spores pollen grains —216. most Chile. Micropalaeontology 13: 204

& K. M. PIKE. 1955. The pollen morphology of Nothofagus Bl. subsect. Bipartitae Steen. Austr. J.

Bot. —206. 3: 197

CORNER, E. J. H. 1969. Ficus. Phil. Trans. R. Soc. Lond. B 255: 567—570.

R. A. The and flora of the North Auckland. COUPER, 1952. spore pollen Cocos-bearing beds, Mangonui,

Trans. R. Soc. New Zeal. 79: 340—348.

New B 1952a. Plant microfossils of Zealand coal measures. New Zeal. J. Sc. Techn. 34: 110—117. from Zealand. 1953. Upper Mesozoic and Cainozoic pollen grains New Palaeontol. Bull. New Zeal.

22: 4—77-

1953a. Plant microfossil dating of some New Zealand Upper Tertiary rocks. New Zeal. J. Sc. Techn.

B 34: 373—377-

1953b. Distribution of Proteaceae, Fagaceae and Podocarpaceae in some southern hemisphere Creta-

New Sc. B ceous and Tertiary beds. Zeal. J. Techn. 35: 247—250.

i960. Southern hemisphere Mesozoic and Tertiary Podocarpaceae and Fagaceae and their palaeogeo-

Proc. R. Soc. B graphic significance. Lond. 152: 491 —500.

1960a. New Zealand Mesozoic and Cainozoic Plant Microfossils. Palaeontol. Bull. New Zeal. 33:

1—87.

& W. F. HARRIS, i960. Pliocene and Pleistocene plant microfossils from drillholes near Frankton, New

Zealand. New Zeal. J. Geol. Geophys. 3: 15—21.

CRANWELL, L. M. 1939. Southern-beech pollens. Rec. Auckl. Inst. Mus. 2: 175—196.

1959. Pollen from Seymour Island, Antarctica. Nature 184: 1782—1785. Cradle for Pollen 1962. Antarctica: or grave Nothofagus? et Spores: 190—192 (Abstract). In 1963. Nothofagus: Living and Fossil. Gressitt (ed.), Pacific Basin Biogeography: 387—400.

Cradle for its Ancient Pacific Floras: 1964. Antarctica: or grave Nothofagus? 87—93. Antarctic 1969. Antarctic and Circum-Antarctic palynologicalcontributions. J.U.S.A. 4: 197—198.

H.J. HARRINGTON & I. G. SPEDEN. i960. Lower Tertiary microfossils from McMurdo Sound, Antarc- ,

tica. Nature 186: 700—702. of and from DETTMANN, M. E. & G. PLAYFORD. 1968. Taxonomy some Cretaceous spores pollen grains

Eastern Australia. Proc. R. Soc. Vict. n.s. 81: 69—93.

DIELS, L. 1928. Kontinentalverschiebung und Pflanzengeographie. Ber. Deut. Bot. Ges. 46, Gen. Vers. Heft: (49)—(58).

DUSEN, P. 1899. Uber die Tertiare Flora der Magellanslander. Wiss. Ergebn. Schwed. Exped. Magcllans-

landern, 1: no. 4.

1908. fjber die Tertiare Flora der Seymour-Insel. Wiss. Ergebn. Schwed. Siidpolar Expedition,

1901—1903, 3: no. 3.

C. Kenntnis fossilen Neuseelands. Denkschr. Ak Wiss. ETTINGSHAUSEN, VON. 1887. Beitrage zur der Flora

Wien 53.

1888. Contributions to the Tertiary Flora of Australia. Mem. Geol. Surv. N.S.W. Palaeontol. no. 2.

—108. FLEMING, C. A. 1962. New Zealand biogeography. A palaeontologist's approach. Tuatara 10: 53

R. and in Time Horti FLORIN, 1963. The distribution ofConifer Taxad genera and Space. Acta Bergiani 20:

121—312, 68 maps. L. FORMAN, L. 1964. Trigonobalanus,a new genus of Fagaceae, with notes on the classification ofthe family.

Kew Bull. 17: 381 —396, 3 fig.

18: 1966. Generic delimitation in the Castaneoideae. Kew Bull. 421 —426. hoherer GRUMMER, G. 1959. Die gegenseitige Beeinfliissung Pflanzen: 162 pp. (Jena). R. Un HEIM, 1951. mycologue dans la hetraie australe. Rev. Int. Bot. Appl. & Agric. Trop. 31: 54—69.

HOLLOWAY, J. T. 1954. Forest and climates in the South Island of New Zealand. Trans. R. Soc. New Zeal.

82: 329—410.

KALKMAN, C., & W. VINK. 1970. Botanical Exploration in the Doma Peaks Region, New Guinea, Blumea

18: 88—135.

KEDVES, M. 1964. Pollen et Spores 6: 195—207.

F. Die der Braunkohlenzeit: KIRCHHEIMER, 1957. Laubgewachse 783 pp., 55 Taf., especially pp. 382—384.

KUPRIANOVA, L. A. 1967. Pollen et Spores 9: 95—100.

Die Tertiaren Forman in MAI, D. H. 1970. Arten von Trigonobalanus (Fagaceae) Europe. Jb. Geologie 3

fur 1967: 381—409, t. I—4. MA KHIN SEIN. 1961. Nature 190: 1030—1031.

MELCHIOR, H. 1964. Fagaceae. In Engler's Syllabus der Pflanzenfamilien ed. 12, 2: 49—51. MOSER, M. 1967. Die ektotrophe Ernahrungsweise an der Waldgrenze. Mitt. Forstl. Bundesversuchsanstalt

Wien Heft 75: 359—380. C. G. G. J. van Steenis: Nothofagus 97

MUIXER, J. 1968. Palynology ofthe Pedawan and Plateau Sandstone Formations (Cretaceous-Eocene) in

Sarawak, Malaysia. Micropalaeontology 14: 1—37.

1970. Palynological evidence on early differentiation of Angiosperms. Biol. Rev. 45: 417—450. OLIVES, W. R. B. 1936. The Tertiary Flora of the Kaikorai Valley, Otago, New Zealand. Trans. R. Soc.

New Zeal. 66: 289—293.

1950. The fossil Flora of New Zealand. Tuatara 3: 6.

R. in Burma. PoTONif, i960. Sporologie der eozanen Kohle von Kalewa Senck. Leth. 40: 451 —481.

PREEST, D. S. 1963. A note on the dispersal characteristics of the New Zealand Podocarps and beeches and their biogeographical significance. In Gressit (ed.), Pacific Basin Biogeography: 41s—423. New PUNGA, M. T. TE. 1947. Nothofaguspollen from the Cretaceous coal measures at Kaitangata, Otago,

Zealand. New Zeal. J. Sc. Techn. B 29: 32—35.

K. RAMANUJAM, C. G. 1966. Pollen et Spores 8: 149—203.

ROVIRA, A. D. 1969. Plant root exudates. Bot. Rev. 35: 35—58.

W. B. HITCHCOCK. SCHODDE, R., & 1968. Report on the birds of the Lake Kutubu Area, Territory of

Wildlife Papua and New Guinea. CSIRO, Div. Res. Techn. Pap. 13: 7 —73, 5 fig., especially p. 7.

SHAW, H. K. AIRY. 1966. A dictionary of the Flowering Plants and Ferns (ed. Willis), ed. 7: 140.

SINGER, R., & J. H. MORELLO. i960. Ectotrophic forest tree mycorrhiza and forest communities. Ecology

4i: 549—551-

& M. MOSER. 1965. Forest mycology and forest communities in South America. Mycopathologia

26: et Mycol. Appl. 129—191.

SMITH, A. C. 1970. The Pacific as a Key to Flowering Plant FEstory. Univ. Hawaii. Harold L. Lyon Arbo-

Lectures 28, — retum 1: 1— 4 maps, especially pp. 12 23. revision of the L. Schneider in SOEPADMO, E. 1968. A genus Quercus subgen. Cyclobalanopsis (Oersted)

Malesia. Gard. Bull. Sing. 22: 355 —427. mountain Bot. STEENIS, C. G. G.J. VAN. 1933. Occurrence of plants at low altitudes. Bull. Jard. Buitenzorg

HI, 13: 37—56-

On Bot. III. — 1 933 —34- the Origin of the Malaysian Mountain Flora. Bull. Jard. Buitenzorg 13: 135

262, 289—417, 10 fig.

Arn. 1953. Papuan Nothofagus. J. Arb. 34: 301 —373, 22 fig.

1953a. Proposal to conserve the generic name Nothofagus Blume. Taxon 2: 177.

Arn. 1954- Additional note on Nothofagus. J. Arb. 35: 266—267.

1955. Nothofagus flaviramea n. sp. Nova Guinea n.s. 6: 281—282, fig. I.

1961. An attempt towards an explanation of the effect of mountain mass elevation. Proc. Kon. Ned.

Akad. Wet. Amsterdam C64: 435 —442.

1962. The Landbridge Theory in Botany. Blumea 11: 235 —372.

1964. Plant Geography ofthe MountainFlora ofMt Kinabalu. Trans. R. Soc. Lond. B 161: 7—38,

especially p. 29.

1967. The Age of the Kinabalu Flora. Mai. Nat. J. 20: 39—43.

& M. M. BALGOOY. 1966. BL. Pacific Plant Areas 163. J. VAN Nothofagus 2: 290—292, map L. Plants: and TAKHTAJAN, A. 1969. Flowering Origin Dispersal: 310 pp. H. WEYLAND, 1964. Lehrbuch der Palaobotanik ed. 2: 594 pp., especially pp. 409—414.

POSTSCRIPT

When this in communicated paper was galley proof most important news was by Dr. S. Archangelsky, Divisio Paleobotanica, Museo de Ciencias Naturales, La Plata,

which he Argentina, generously permitted me, through Mr. J. Muller, to quote, for I sincere thanks. which express my 1) The oldest pollen record from Patagonia is from Lower Maestrichtian Jaguel This of brassii Formation, Upper Cretaceous, Prov. Rio Negro. is the type, which small the Paleocene. the of all three perpetuates in quantities in In Eocene pollen types and both in and Chile. (brassii, menziesii, fusca) are abundant, Argentina Also the brassii still has that 2) at present type is represented in Argentina, as it appeared of alessandrii pollen the very rare, South American living species N. Espinosa belongs to

this type. Pollen ofN. alessandrii had been described by Miss Cranwell (Rec. Auckl. New

her the Zeal. Inst. 2: 190. 1939) and was attributed by to fusca type, but this record cannot be checked number convinced that based as no collecting is given. I am it is on erro-

97 8 BLUMEA VOL. XIX. No. 9 2, 1971

neously identified material supplied to her by Skottsberg from the Arnold Arboretum Herbarium.

The is bound the West Pacific 3) brassii type through this no longer to purely tropical alessandrii the sect. Calusparassus subsect. Bilamellatae, as N. belongs to leaf-shedding This sect. Nothofagus. makes it more comprehensible that the brassii type is among the

in Two in oldest the genus. small gaps table 1 can now be reduced. 4) This better understanding is strengthened by the morphologically logical deduction

made that which I formerly (van Steenis, 1953, p. 310, 335; 1954, p. 267, fig. re) of N. alessandrii is the most primitive among the living species sect. Nothofagus (7 9 flowers,

a lamellate, 4-partite cupule).

Concluding, the new facts confirm the theory and strengthen it. I hope that the

will be filled in future: records of the three the remaining gaps pollen types through further the and Tertiary ofPatagonia, details on three types in Antarctica, Upper Creta- and ceous records from Queensland possibly New Guinea.