KlOGlNKIRAPHICAT, HELATIONS O F T H E NEW ZEALAND REGION.
99
Biogeographical Itelations of the New Zealand Region.
By W. lt. U. OLIVEII, F.l,,S., F.Z.S.
(With 7 Text-figures.)
[Read 2nd April, 19245.1
CONTENTS.
1. Ways of Dispersal ...................................
I . I m d Bridges ....................................
4. Ocerin Curreiits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3. Atmospheric Conditious ..........................
4. Animals ........................................
IT. Means of Ijispersal ....................................
111. The 3Iovonient of Bpecien . . .
.... . . . . . . . . . . . .
1V. The Illovenlent of I ~ and
w Farinas . , , . .
1'. Ilisotrg of t h e New Zealand Flora a d I?"tiunn
.......
I . RIesozoic Floras .................
2. Malayin Land Connecrt.ioii ........................
3. The Intliieiice of ;Iiitarr.ticit. .......................
4. Species dispersing Overseas. .......................
VI. Summery . . . . . . . . . . . . . . . ..........................
YII. I,iteratui*e ............................................
.
Page
100
100
104
105
106
100
107
I18
119
1%
138
1:15
138
TKTROINJCTION.
THEgeographical fe:iturcts which
111i11te
New Zealand of interest to thc? plant,
geog::1.iil)hur:ire the extensive land areas lying i i i the Pacific
C)ce:tit f a r distalit from the nearest continent, the diversity of their physical
clisracters, :tnd the great, depth o f the surrountiing ocean. From a. biologicla1
virw-point, the outstanding ch:rractc.ristics of the fnuiia nnd flor:~of the New
Ze:iI;lnd region are t,hc ahseiice of ni:itrirn:tl~,contrasting it with the reiiiaillder
of I lit! wo:.ltl except Ant:trulJica, the marked dissiruilarity of it,s p1:int altd
aiiiiiial productions to those of Austr:ilia, and tlie presence of an element
c~~innioii
to two or more of Bhe southero land masses. No woacier is it, thea,
that ttict history of its farina r t r d flora lias of'ten been il iiiuttt!r for tli.u:cussion
:L111ongI)ioIogisks. T 1 ~ tNew Zealand has been n long time isolatpd frollr
a n y ot.hc:r large 1:tnd mass 110 one doubts, a.nd that most of its plants alld
ariiiiials are tlwwided froin those which rctlched i t over dry land is generally
agreed ~ i p o i iby both botanists a d zoologists, but in which direction the IrJn(j
hritlge o r land bridgc!s I:ty and a t what periods they existed :Ire 1)y iio'x1ieaIis
undispnted I)oint,s.
:Lii{I : t i i i i i i i i I
LINN. J0UKN.-BOTANY,
VOL. XLVII.
I
100
MR. W . K. D. OLTVEIl O N THE 11IOG&OQliAPHICAL
In the present paper :in attempt is r n : i d c s hy :in ana.lysis of certain classes
of p1ant.s and :i.nintitls of New Ze:tliind to d ( h - m i n c t h e c.oniitricts between
wliicli there has heen ail intercliaiige of species. I lie miin principle nritlerlying the investigation is one consist,rrit with ihe theory o f descent with
modificntion-namc!l~, that e:icli group Itas had a single place of origin from
which its ii)enil)srshave tlis1)crsetl. The place of origin is not, necess;rrily a
sinall locality, hiit is coter~iiiiiouswith tlie nrca occupied hv the parent species
of' the group i n qucst,ion. l'he prosent area of greatest development OE a
genus i s :issunietl t,o be the 1tlac.r. of origin unless other evidence is available
to show that, it has shifted. 'l'hcentre of dislwrsal of :t genus is, therefore,
trot, necessarily siil)~Bose(lto rcmsin :ilwa.ys :it the place of origin. I n the
case of a genus entlcniic in tlitr New %e:ilarid regioii the relations of the
fmriily which includes it :ire niattle tile hasis for tletertnining t,he point, of origin
of it,s antecetlcnts, it. 1)eing :rssiiniecl t h t t,he :iucc+stors of the genus if traced
far enough b:tcli would be species hiving closer relatioils to those of some
d h e r country. For instance, tlie New Zealand entleniic genus Xporodanthus
may he considered as ii dctsccvid:rnt of it species wliic*hwould be included in
the same genus a s the :Incastors of I,e;rpotIia wit11 Australia w s the centre of
dispersal.
Witliout being in the least concerned with the place of origiri arid dispersal
of the great pliyla of p1:tiits ii,tid :tnimals, it must be conceded that if New
Zealand contains or contaiiitd fa,tinas or floras which require continuous land
lor their dispers:tl l i e n 1:md connectmionwith some other country is proved.
But the deniaiidi~igof laiitl hritlges is not to he lieltl to imply that all the
present xnimals and plants ol' New Zealand or their ~ L I I C ~ S ~have
O ~ S arrived
from sonic other country. On the contrary, the :Lutlior believes with
Dr. Cocliayne ('Vegetation of New Zeala.ii(1,' p. 1.92, 1921) that many
groups have arisen within tlie Kcm Zealilntl territ,ory, itntl furt.her tliitt inany
species have migrated from it to the surrounding countries.
Before attentpting a history of tlie New Zealand fauna ant1 flora., sonie
general rcniarks relative to t,lie clispersal of orgnnisins in tlie Southern
Hemisphere will be offered.
I ,
I. WAYSOF DISPERSAL.
1. Land Bridges.--l\losk tliscussioiis concerning tlie origin of tbe fauns and
flora of New Zealand, and esl)ec:i:tlly tliat sectioii of it which has been given
tile name of Antarctic., have been around tlie question of t'onner land
connections. At one time or another such land bridges have bean proposed
towards nearly all ijoints of the conilws. The easy inetliod by which these
hypothetical continents can be brought tip froin the depths OF tlie ocean has
\'robably been one cause why the study of geographical distribution has
inade little advance in receiit years ; for once a land bridge is assunled,
there is no incentive to further investigation. Now, land bridges required on
101
RELATIONS O W THE NEW XEALANJ) RKOION.
bio1ogic:il consider;itions slionld not conflict with geological cvidence-tlitLt is,
the e'videiice of the struoture of the eurtli's crust. Gleologists, I find, generully rely on palaon tological evidence wlien prolmsing former land connections.
Certain geological evidence, liowcver, will lie rsviewed l~ecanseit indictitos
the direct.ion and time of possihla l i i n i l tiritlgeg.
Chttowr yf' Ocmn;jh)ov.-On the principle of thc periiiiineiicw of coil tinental
: r i d occariiG iirt.as, p d land conncctioiis odd tie itidicritetl 1 ) suhniarine
~
ridges.
The Lroiid features of' the contour o f tlie ocean-floor in soutliem rrgioiis
consist of ~ubititirinr 1 ) I i t t l ~ ) r l )less
~ ~ t.liaii 2000 fritliosis HIO OW the siirt'i~:e
ratliiiting froin t*lieAiit:rrctic cwiitiiient to New Ze:iland. tlit! tyistern I'acific,
Soiitli America, the central Atl:intic, i i r d the soiitli I r i t l i m Ocean. A siiiiilur
extmsion, hiit of gra:iter depth, is intlic;it.c~cltowards Tasiniinia., wliile Houtli
Africa is joiiicti to tlw centml .\tl:intie ridge. The ttioiiafitid-fatltotii lint!
oomplrtely snrround?;the Antarctic contiiieiitz,so tlittt iuiy direct Irrntl bridge
with tlre north itinst Iiiive lwcn :it ;i rc?iiiot,epriotl.
New %eal:indis flitlllied e:ist,sntl west ti? tleep oce;iii-t~rouglis. Soutllwartls
is the broad snbiii:irint! ridge just i n e i i t i o i i t d , wliile iiortht\.il'r(l is :I shallon er
aud more broken eutcmsioii. Sw-tmttoni iint1e.r :I thouetlnii f:itIiotns is
continuous hetween New Z t d a n i ~ ,I,orcl Howa 1 sliincl, and New Oaledoilia.
More brokrii contoiirs voiinect, New Xc:ilsiid with the weshrn I'acific ty w:iy
of Norfolk Island m t 1 tlitJ Kc!rrii:ieiecs rc.spee:tivcly. (!oil tinental t!onirectiorls,
aocording to .the evidence of the oc(::iii-Aooi*, itre thus inrlicated from New
Zeulaiui to tIiU r i o r t . ~atit1
~ soiitJi, but tiot, to tlie east or west.
A1 tliougli i n 1.11~ niajority of the series of ~)alsoyrapliicma1)s Bensoii
(Trans. N.Z.Iiist. vol. liv. .I923) shows ti s h e - l i n e soinewhern near the
e;istern coast of Aii.4xali;iy lie c!videni.ly t l o e ~not consitler tlie Tasniirii Sea
tt*ongh :t fo:itiirct of tliu Pal:rwoic and Mesozoic periods. Bnt he bases his
results largely on tlie relictions of iuarine faimis, wliich shoeld be used with
ciiiition, ;is mtny littord :itlintaIs :ire distributed by means of pelirgic larve or
floatiiig ohjecta rind Cityeiid inore OII temper:tture and ocean cnrrenh than
coast-lines. ICegiin h i s enililiitsiml this point with regird to fishes Terra
Nov:r ' Report, Xool. vol. i. 1). 149, 1'31.6).
Tho opinion o l geologists tiiffor widely regarding the date of origin of the
TtiRnian Sea. Mwslial I :ind Morgan t l h k that during the early and irtidMesozoic era New XetIlit1ld wiis tlic shore-line of o continent stretching to
the weptward or north-westwircl, while 13eiisoii draws a map of Australasia
iii Jurassic titires illuatruting such :I disposition of t.he land. On t,he other
Iiand, Arldt, Scliuchert, ant1 Stq)lieiia (:is quoted by Ehnson) believed the
Tasinan 8e;i t.o have I)ecii early lor~iivd,existing in thc Jurassic period. I t
should be pointed out t h t the existence of t,he 'l'asiiiuii Sea since tht: Jurassic
1)eriod is necessary iF the former iidepentle~rt julictiou of Australia and
New Zea1:intl with Aiitarct,ic H during lute Mesozoic times be iioceyted.
12
102
M l t . ’CV. R. B . OLIVER ON THE BIOGEOGRAPHICAL
Periods of Idlevation i t , Neio %~alattd.--C*cologic;ll evidence points to the
late Palzoxoic as a period when the New Zraland area was elevated, and the
land may have e s t e n d d so as to join other lands. During Paleozoic times
folding, prob:it)ly pre- Permian, took plrice along north-\\ est to south-eaet
lines, and at times New Zedand proba!)ly formed part of a continerit (Morgtin,
N.Z. Journal Sci. & Tech. vol. v. p. 49, 1922). Recently Park discovered
striated boulders in a breccia near Taieri Mouth. He favours a glacial origin,
and st;Ltes that the rocl\s M o n g to the ‘I’e Anau series, or Upper Carboniferous (Park, Trans. N.Z. I n d . vol. lii. p. 107, 1920).
Glaciation during the I’ernio-( hrboniferous period is known to have been
widespread i i i the Southern Heiuisphere. Striated boulders, tillite, or other
cvidence h i s lwen detected in Australia, South Al’rica, South America, 1 ndia,
:Lnd New Zealand In Auetra1i:i there are two distinct, horizons of glacial
origin-one, the liuttung Series, Cnrboniferons, and the otlier, the Bolwarra
ConglSloiner,ite,Permian (Diivid, ‘Guide to Hunter River District,’ p. 35,1923).
The glacial plienoiriana would semi to point to the elevation of the land in
various ~)ortionsof the Southern Hemisphere, and thus connections might
t)c n i d e between Antarctica and ont! or more of the southern continental
lands. Sucl~:I radid tlispoaition of tlte Iantl, instead of the east to west
direction that the conception of Gond wanaland implies, might explain tho
distributioti oE the glacial phenomena in late Palzeozoic times.
Stephens believed th:it during 1’el.mo-Carboniferous times eastern Australia
and New Zealand were independeiit groups of islands each united with
Antarctica (l’roc. Linn. Soc. N.S.W. vol. siv. p. 349, 1889).
The evidence for an extension of‘ land i n Permian or early Triassic times
is not clear. It, rests 011 the difference between the fossils of the Maitai
(Yul.rt~o-(lar~)oniferons)
and Triassic rocks and on tlre supposed intrusion of
plutonic r o c l r ~ In Now Caledonia middle Triassic rocks tire frequently
missing.
A contiiiuous series of sediments, mostly nnfossilifrrous, covers the period
frorn iniddle Triassic to lower Cretaceous in New Zoal.land. The direction of
the land horn which those are derived has not been nscertained, but whether
it be east or west, it might fa11 w i t h i the h i i t s of the submarine plateau 011
which New Zealand stands-that is, the Tasrnan Sea might have intervened
between the land and tho Australian terrain.
The Lower Cretnceons period was rnarked by intenso crustal movement i n
the New Zealand area, resiiltilfg in the folding of all the Mesozoic rocl~s
deposited up to that time. According to Morgan, in the south aiid in the
extreme north of New Zealaud the folding followed already existing Palsozoic folding, but elsewhere it coininonly took a new direction, north-eabt to
south-west, almost a t right angles to the older folding. This was a period oE
elevation aird extension of the land. I n New Caledonia no middle Cretaceous
roclss are known, so that a long emergence must have been there the feature
of that period (Benson, Trans. N.Z. Inst. vol. liv. p. 49, 1923).
RELATlONH OF T H E NEW ZEALAND REGION.
103
Following another period of elevation iu Eocene times was a lollg period
of subsidence in the New Ze:iland area, iq'prently reaching its lowest limit
in tlie Miocene. EIev;it,ion gg.:iin took @ice in late Tertiary times, resulting
in inucli Mock-faulting and tilting.
Sur1ini:irizing the geological evidence for land conliections between New
Zealand and other lands, it iiiay be said that elevation is indicated for
Perniiitn kind perhaps I'riassic times, : ~ n dm c i w certainly in the Lower
Cretaceous period. Elevation also txiok place a t tlie begiiining and near tlie
c1Im of the Tertiary epoch. Whettier or iiot land connections with other
countries were a c t i d l y iiiade a.t any of these periods can best .be determined
hy the biological evidence.
Narine Fuzcrias.--I pass over niarine faunas as evidunce of land connections
for the I:ollowing reasons :(1) Nariiie ;iniiiials ill most cases have t'rcx!-swirnining larvw, many of
them being pc1:igic. III sonie groups, l a r v s remain in thn swimming stage
for :I. coiisit1cr:tble tinic. Mathews records tltiit tlic y o u ~ i gof Alytilus sdulis
liatclied oii May 21st,, 1912, wore swiiiiniiiig on August 15th (Journ. Marine
Biol. Assn. vol. ix. p. 557, 1913). The pelagic larva of littoral nlolluscs
have on several occasions been given distinct generic iiarnes, as Sii~usiyeru,
~~u~~:Llillicr~~yu,~u,
C'/ieZotmpis, anti so 011. &ny coastal fislies have in their
life-history a pelagic pIi:ise, which 111:iy he the larval statu or the young fish.
r,
l h e marine stages of certain froshwater fishes such :IS C;aluAus and Cteotria
come under tliis head.
(2) Even it' niaririe faunas :ire held to indicate bile prcsence of' a coastline, cotitiiiuous laid cotiiieetiori does riot iiecessarily follow. Tliere may be
one or niore straits easily crossed by inarine ~inini:ils~
especially if the water
be shallow but iinpass;thle to land plants a n d n i i i i i i & .
(3) The community of species in marine I'ossil fauiias iiidicutev like
coiiditions of temperature from whicIi the ;listrit)utioli ot' ocean currents may
be iiiferreci. It would scarcely be safe, however, to map laiid-lines from t,his
evideiice, as currents of different temperaturtas are sonietintes tound side by
side. The tnarinc f:tuna in the south of New Zealalitl is clifferent from that in
the itorth. Here are two ocean currents affecting tile coast, but they are not
separated by :I IanJ Ixirricr. A Le,tter-l;nown iiistance of tliff'erent faunas oii
the sanie coast is that of the eastern United States, where tire northern aiid
southern faunas meet a t Cape Cod. The marine faunas oi' the present day
are limited usitally by tcmper:itiire i i i a north a i d south direction and by
l a i d harriers i t i an enst and west direction. W i t h a knowledge only of the
liiuits of the Indo-Pacific region from its f a m n no one cot~ldpossibly map
the shore-line. An island aren like the Pacific would coinpletely baffle any
attempt to do so.
Yalieozoic l~'h~as.-No plant-reiiiains have so far beon detected in rocks of
PalBozoic age in New Zeiiland. This, of coiirsc, does not mean that land did
not exist in or n c t r the arcst. On tlie coiitrary, it is evident that there was
I04
N K . W. R .
H. OI,l'i'EK
ON T H E IllOGROGRAI'HICAI~
RELA'I'IOXS O P THE N E W ZEALAND REGION.
105
would be throughout tlic Tertiary period westerly winds or easterly-moving
storms in thr soiit,h t,enipcmte region. The surface currents in accordnnce
with t.he attitospheric circuliitioli wort10 carry drift in an easterly direction,
and an int(!rchaitge might. eiisily take placc l)et.wc?enthe Antarctic continent
and the New Zealaticl and Au*:tralian continetits.
The directioii of tlie drift of the surfwc water of the Southern Ocean a t
the present day is eviclcAnced by tbe general northward and eastward inoveinent of icebergs from Antarctic regiotts. In the South Pacific they reach
the latitude of New %t.nl:intl. 1 c c ~ t ) e r phave been stranded a t the Chat1i:im
Ishinds. I n the Solitti Atlatitic ; i n d 1ntli:tn Oceiins, icebergs pass thc 40th
parallel of South latilridr. Antarctic aninials are sometimes stranded on the
shores of New Zealand :ind Ai&t*:tli:i.
r J h cmb-c?:ttiiig seal, for instiince,
has been recorded twicc: i i i New Zeiilund :md twice in Australia. I'ossibIy,
however, t,he nortliwartl drift frorii t,lie Antarctic continent in early Tortiary
titnes wheti the const at least, supportetl vegetation may not have h e n SO
p ro~ ~ ounccd
:is it. is now.
The investiptions of Gn11l)y(' €'I:ttits, S c d s , and Current,s in the West>
Indies :md Q%(ires,'1). 310, I ! ) ] 7) show t h t , from the present direction of currents in t.he Soutlierii Hetuispliere, Aiwtr:iIi:i would receive drift troiii k'uegia,
the isi:intls of tlie Bo~t~licrri
O C ( U I I:mcl
~ South Africa, and t l i s t i i h t c it to tlie
north 01' Now %enl:miI ; w i i i l t ? X(aw Zc:il:intl would receive drift from Yuegi:i,
the Antarctic cwnt,inc!iit., l h e islands of i l : h~~ t l i e r nOcean, Tiisniiinia, and
Southern Australiii, a i d its ~oiitheriiend woulcl distribute it t,o South Oliile.
At pix~sotit:i w:~rni c:uri.rtit rutis soiith dong the cast Australian coast',
tiiriiiitg :ihout, t'lie 1:ititndc of 'l'xstii:ini:i towvarcls New Zealand.
This cumtiit
would not COIIIC' into (?xi n(!o while tlic T:is~iianSea. was ulosetl to the nortli.
Instead, the east.rrly currt:nt iiiiglit tie deflected nwthwards 011 reaclring the
N(:w Zo:iIand continent, atid co:ist roiiii(1 the Taeiiiaii Sen, carrying clritt to
Anstrali:t.
1bfPing Punzice.-'l'lirrct is ail agont of dis11era:iI snflicient t.0 accoiiiit l t ~ r
the transference o f c * o : d . i i l iiiariiie forir:s to all parts of tlie Pacific, ~i:in~ely
driftitig piinice. It is*cast 111) on tlie h o t w 01' Awtrali:i, N e w %ral;tnd,a i i d
all the ishnds of I'olyiiwin, :ill({s i i l i p o r t s corals (Kent,, ' Great Barricr lieef,'
p. 122, 1 Y!)3), harniicles, and no tloit td, tn:iriy other foritis of 1n:irine life.
3. Atmospheric Conditions.--The (Ioiiiiii;iitt feature of the clirnate of tlic
south teiiipi.nto regions is ~ h t ?~~iiss:ige
I I ; I S ~any given 1loiiit of :t wries of
easterly-ino~ingcyc*loliiostorms. 'I'liej t,ake i t ItNJrC! southern rorttr i n wiiitei.
than in s i t i i r ~ ~ i o r . A sccond t ~ p of
e cyc*lottic: storms conies to Bew Z d a n d
t'roirr the north-west a t i d affects tlic Iiort,lic'rn portion of tile Doininion.
These are iisti:illy s r ~ i i ~ t t i ( visitors.
~r
Tlic rate 01' niovenient of cyclotlic:
storms varies considernhly, hut :iverng:.t.sa.bout 400 miles pet. day (1'mik)erton,
3.2.Jonrn. Rci. & Tech. vol. ii. I). 165, 1!,19>.
Tlic path of c:yclotric stortiis in t1.e Inte Tci6iii.j period woiilJ dtapei~lon
the extent, of tlie gl;ici:ited itrea of Antarctica, for these storms coast round
the anticyclone area of the po1:~r ice-c:tp. During tlie Pleistocene glaciation,
therefore, they would he considerably f d h e r nortli thau at present.
4. Animals.-Birds as an agent, of dispersd need only lie referred l o fiere
tiy noting thitt, the Soutliern Ocean swarms with long-distance flying petrel&,
which breed in countless ~iuiritiersanlong sariil) ant1 tussock vegetation on
the Subantarct.ic islands. Probably occasional opportnnitics are givon lor
the tronsfereiice of seeds oE 1)lants aiid eggs of aiiinials froin isl:riitl to islniitl
1 ) these
~
birds.
I r. MEANS O F 2)181JEk8.4L..
All lant3 plants pass t,lirongli ;i stage i i i thei I* life-history specially fitted to
endure unfavour;ihlt: conditions. Ti1 spermophyt,es it is the seed, in 1,t)eridophytes anti lower pluiit,s tlie sl)ore. I t is diiriilg this stag(: t h t dispersal
most effkctively takes pl:ic.e, aiid the opportu~iit~y
is give11 for t i ~ ; ~ ~ ~ s ~ ) o r ~ : t t i o i i
over loiig distaiioas. A classiticatioii of pl:int>sa,ccording to whether sved or
spore carriage is hy wind, water, or aniiiials would 110 I1;tscti on infweiice
rather than on observation or axperinient ; nioreovcr, die seeds or sliores of
a species might be carried by inore than onc of thesc :tgent,s. In the followiiig
arialysis 1 have given in pcrcrritages ct:rtain 1):irticulars of tlie floras ot
Australia :tiid S e w Ze:tlanti imd of the non-endemic vascu1;ir pla~itsi n Ec!w
Zealand. Figures are first given for those systciriat,ic ~ ~ O I I I which
I S
appr
specially fitted for wide tli,.peiYal-ii;inirl~, the ptcritlopltytes, orcliitls, wtiiposites, grasses, and sedges. The reiiiainder of tlie New Zealaiicl p1:iiits arc
then divided accordiog to whether the fruit is fleshy or dry. Tlic re3iilts
are apparently contradictory, but when otlier factors, such as tlie geiier;tl
direction of the iiiovenient of plants in the Southern Heiiiisphere arid the
age of the groups, are taken into coiisideration, explanations may Lc g i w i
for these :ipprent aiiomalies.
Australia.
l'teridopllytes . . . . . .
Orchids . . . . . . . . . .
Composites ........
(iraswb
..........
Sedges . . . . . . . . . .
Plants with fleshy fruits.
Plants with dry liuits not
elsewhere included.
t
Zeitl~iid
New %e:iland. New
non-endeinic.
10,670 Epecies. 1570 species.
10
3
3
4
370 species.
''6
'4
6
4
11
5
7
4
8
8
la
3
41
14
41
79
The first tliree groups are specially atlapted for dispersal by wind, yet
each gives a different result when the New Zealaiitl non-eiideniic syecies are
compared with thc floras of Australia :~ndNew %eal:ind. The high proportion of pteridophytes is what might be expected in plants mith minute
RELATIONS OF T H E NEW ZEALAND REGION.
107
spores capahle of heing carried long distances by wind. Orchids show no
similar high proportion, perhaps because they are a group recently evolved
and much speci:ilizod. Composites in thc New Ze:tl:ind non-endemic plants
show a proportion siniilnr to that of orchids when hotth are coinpared with the
Australian flor:i, brit when coinpared with the New Zealand flora they are
found to be only about, or~e-t~hird
as numerous. They are probably ail old
groyp, hit; thcy show a greater development in New Zealitnd t.ltan in
Australia, i d , as will he pointod out later, migration in the south temperate
region is niuinly from w e s t to east. Hence the proportion of composites
found in the New Zealand non-endemic plants, which aro mainly also
Australian, is low.
r 7
1he sceds of grasses :ind sedges inay be carried by all agencies, hut niainly
perhaps by wind itnd aniniaIs. I3oth, :is might be expected, are well represented in the New Zea1:tnd non-endemic species. Plants po~sessingflesliy
fritit,s are usually corisitlereci :is speci:illy :i,dapted for dispersal by attimals.
I t is significant,, tlierefore, that the proportioil o f these in the New Zealand
flora., which I consider a continental type, is high, while the -percentage in
the New Zealand iion-cntletnic plmtls is quite low.
As most plants are capable occ*asionnllyor accidentally of crossiiig stretches
oE watctr, I do not rely grcat,Iy on iiieans OE dispcrsal to juclge whether’ they
require continiious liind conncctions to explain their present distri hition.
1l:ither do I contend that where tlierc has been connection hy land the flora.
which occupietl it will, when afterwards divided, show by comparison of the
stq):ir:itd p r t s that they were originally one. The conimon eletiient will be
large a i d Fiintlaniental. There will not be two distinct floras each liaving
bnt, frapntents of tlie othcr, as appears when the floras of Aiistralia and New
Zeillailtl o r New Ze:ilatid :ind South America are conlpared. I n one case
original continuity of the Isntl is indicated, in the other it is not.
111. THE MOVEMENT
OF SPECIES.
li$ oj’ CI Spcies.--A point to he considered in connection with the niovement of species is the length of life of a species. A species changes in the
course of tinic ; so that whether it gives rise to more tlian one or not, it eveiit ~ i ~ lchange.
ly
into whut would be considered a distinct species, provided of
course it does not beconit. extinct. This statement requires niodif! ing o d y
by saying that sonic species change more quickly than others. I n a change
siich :is this it is evident that isohtion is :in important conditioii in originating
now species, for t11o.c individuals M ltich are free to cross 11 ill deterntiiie the
limits of the clianging spvies. ‘1‘11~ p;il~ontoIogicalrecords hhOW that vcry
few species exist as long its h e dulation of the Tertiary era. Proiri this it
follow, tliat if a species is found in lands presiinitd to Iiave been separated
duriiig the M hole of the ’l”c.rti:iry period, t h e proha\,ility is that dispersal of
that species botween the countries in which it is now fotirid is still goiiig on.
108
MR. W. R. B. OLIVER O N THE RIOGEOGRAPHICAL
The phenomena of swamped genera- th:it; is, those non-cncleniic but represented in Now Zealiintl hy enclttniic slwcies onlj-ii;iturally follows from the
fact of siiecies (:hanging in course of tinie, i n diis CXSH tile New Zealand
section of the original specitls r u n n i i i g its own course through isolation.
Many grlieril nnil inore species have cotitc! into esistence as sucli in the New
Zealatid region, anti the process is +till going on.
l h e age of very few recent species of New Zealand plants CiliI I)e known,
as the pa1w)ntological records are scanty. I n sotiie cases the n>lative ages
of two groups niay be judged hy inorpho1ogic:il charxcters, I J Ua1~w i i v s with
a degree of uncertaiiity. Willis's 'Age and Area' hypothesis L I I I L ~he useful
in suggesting t,he relxtive ages of species or the lcngth of tilie they Iliive
been in New Zealand, but intiependent contirrnation is tteetletf.
It sliould be 1)ortie in mind that the tliseiit,ari~:.letnent
of the present flora
and fauna is coiiiplioatetl hy ttic fact that dispersal II:IR tmcn continuous
thronghout the ages. Species have arrived and species have tlepartetl. Tliere
appear, indeed, to be regular migration routes.
Simihr .r!ai*ietirs arising ia two ( J Y ~ more loculities.-Al tltough it is VOIIceivtiblt? that, :t spec ios u n tl e r si Iti ila I' con cl i t i o n F: in two w i t l r l -sepira
~
ted
countt.it.s niight in eaclt give rise to varietic:a w l t i c - l i on coinpariaon would
appear ident,ical, it is iinprolidh tliat t,he new varieties would cottlinue to
reinain alike for long. The tendency would illwitys bt: towards ditferentiation into distinct species. Thus C+uppybelieves that the v:iriety Caturuc-toe
of Care;<(Xderi Found i n New Be:il:itid, South Africa, and Soritjlt America is
a corresponding varietal niodificat~ionwhich Iias t':ilien ljlitoo irr e:ioli regioa.
Likewise I h:rve reoortletl niy opinion that. the sinlililrity o f tlw ntoutitaiii
species of Co~iaria( C . t l ~ p m ) W ain tlir A n d e s and C. lrwitlu i i t N o w Zo:ilund)
map be due to tlie fact that each is a derivative of t h wiclespred (,'. rusei/ d i u (Trans. N.Z. Iiist. vol. liii. 1). 363, 1921).
&Single point qj' o~igin.-IE the plants nnd aniniiils of Sew Zealand be
exuniinetl from tbIwpoint of view of their p r o b a b l ~place ot origin ~ i n dsubseyiient dispersal, n basis will be estnblislted for tleteriiiining 1 1 i c origin ;iud
inovenirnts of tlie florii and faun:i. Iflor, assunling the nicnilxrs of it faniily
or generic group to he derived f i w i n a coiiinion ancestral slwies, a single
point of origin and dispersal follows. Tlte regioii where the grc?at.esttlevelopinent of a group occurs will iisualip bc where t.he group origiiiated, hut the
centre of dibpersal may shift. In this case indications 01' tslte ~)l:iceof origiti
inay be traced t)y considering tlie relationships of' the group in question, as
in t,he fainilies StJjlidiacex! antl Itc,st,i:icca t o he nisutioncd pr~?sri~tly.
I n very old gt.oups tltere i1i.e prolr)a.hly suvwal u Iiosc? gre:rtret c~eveloyinent
is IIOW in the s o i i t . h e t x &ntispllei%, though they originaleti in the Northern
Hemisphere, where some of tlteni :LI*A Found foshil i t i Cretaceous and Tertiary
rocks. Sonie of' tlie Coniferales: as dycir/~i.s,rlmiteur*ia,antl I'li,yZloc/utdr.is, also
Pugus (including Notl~~fayus),appear to conte under this category. The
principle of the spreading out of successive waves ot' migrat,ion, eac!i :i stage
I ,
RYLATIONS OF THE NEW XEALANU REGION.
109
higher than the preceding one, so that old tylJeS are fouiid farthest from the
centre of dispersal, him been recently pointed out by Matthew (Ann. N.Y.
Acad. Sci. vol. xxiv. 1915).
Willis ( b A g e and Area,’ p. 60, 1922) has discovered im itnportilnt [act in
plant-distIibution-ii:uriely,
h a t if the endemic species of u coutitry be
arranged according to t,he size of their areas of dismibution, then inore ouc:upy
sui:tll than large areas. This gives strong support to the princi1)Ie of a single
point of origin for cach species. Willis concludes that the older :t species
the wider its distribution. Tliere caii be little doubt, however, that agents
for dispersal and oplmrtutdy for estabIishnient :ire the deciding f:ictors in
deteriiiiiiiiig the area a species occupies.
In a faitlily, as it extends its area, new points of dispersal arise. Thus the
faiiii ly Stylidiaceze (br Caiid 01 1ence;E) I)resuinably origiii:it eti in Australia,
wliere nine-tentlis of tIic species are now found, and S ~ J ~ C X Ithence eastward.
Retichiiig Ncw Zoalancl, i i new p i n t of tlisliersal wkis f(JI’nlCL!,resulting i n the
multiplication of t,lie spciee of ForYte,ru, and the origiii ot‘ (hvo~tylidiimtand
Pliylluchie. lCvun if it he argued t h i t these two genera 1ii;iy Iiiive ariseti in
Australia, travellefl to New Zealand, and subseqaently bc+canie extinct in
Australia, it makes oiily :i difference of degree-namely, whether the ancestril
species wlieii they crossed the Tasnian Sea area were ditferentialed a s gellera.
or not. l’/iyZlach,ris Iias rearlied still farther east,ward :is fhr as Soulh
Aiiierica. Equ:~lly interesting resulk are obtained !.rain a st.udy of the
family ltestiacez, which ~irobablyoriginated ill &mth Alrica, whrre liiost of
the gtrnera and sliecies are now found. Some generit like Kestio and Hypo/wnu have species in both countries bnt fewer in Auutrali:i ; Lrptoeurpus has
niorc species in Australia than in South Africa, and tilere!.:ire eleven genera
of srnall and nicdiorri size confined to Australia. The Further exreiisiun of
the family eastward to New Zealand is precisely of tlie character of its
extension froin South Africa to Australia. Of thc tliree species in New
Zetllittid one belongs to an enticiiiic genus, Sporodantkus, rehted to a genus,
Lepyrodia, endemic: in Australia ; another, Leptocuq)us siniplr.c, is endeiiiic,
but belongs to a genus tnainly Australian ; while tlic tliird, liypolwnu Luterij o r a , is identicnl with an Australian species, and the genus is niainly South
African. The distribution of the whole fiimily 1i;irmonises with tlie theory
of origin in Soul11Africa, dispersal to Australia by chance crossings during
:I loiig perioti of time, and thence extension to Sew Zealaiid iii a similar way
with long intervals hetween the arrival of tlie speciw.
TV.
‘L(Hlr: 31tlVERIF:K.T O F F1,OItAS AKL) YAliNAS.
1)ewetit of Fauaitrs urto! k%o~u.s.--Afaunit or flora, like an individual, is the
liiiwl dcscrndant OF a previous fauna or flora that in course ot’ titile lias been
niodifiecl by inhcwiit changes in the species, : i i d by iinniigrations and
eiuigrations due to changes in c1iui;tte or other causes. The study of past floras
110
MR. W. H. B. OLIVER ON THE BIOGEOGRAPBICAL
and faunas, especially of the Tertiary period, aniiily justifies this i)rinc*iple.
The Eocene gymnoupernis ot‘ New Zealant1 are inoi’e closely related to existing
species than are tho ,t* from the Cretaceous. The theory of Ettingsbausen
that the Ter.ti;ivg floras of Aiistridi:t and New Zealand resembled t h t of
Europe inore thaii they rcseinhlrd the jjresont tloras of Aastr:t!ia ant3 New
Zealand, hits p i n e d wide acceptance aotwithstaiidiiig the fact that it hiis
either been rejected or ignored by most botanists who have since discussed
the relationships of the flot as. Ettiiigshaitsen’s theory rests on the identification of fossil leaves, whicli sl~owshow the use of characters of sinall
taxonomic value may lead to rasrrlts inconsistent with the principles of distribution and descent. I n point of fact, if Ettingshausen’s identifications be
accepted, his conclusions would scarcely be justified, as the Tertiary floras
of both Australia and New Zealand contain large cletnents related to the
existing floras and not to that of Europe. The reference of European
Tertiary species to southern genera like Eucalypttcs 11:~sbeen discredited
and needs revision, as does also the reference of Australian and New Z e a l a d
Tertiary plants to northern genera.
Miyrutlon of Ft!ni,as.--ln order to make coiiiparisons of diflerent floras
with it view to deterniining thwe whicli migrated as a whole and those which
have been transferred by occasional means, I have made a comparative analysis
of thre vascular floras of New Zwilantl, Lord Howe Island, the Kermadec
Islands, and the plants coininon to New Zealand and Australia and New
Zealand and South Atnerica(see I). 110). First the plrrnts were divided iiiI.0 five
iliain groups according to lrabititt-namely, (1) coastal, including all coastal
formations subject to the influence of stilt air or water; ( 2 ) freshwater,
comprising swmnp, lake, and bog associations ; (3) scrub and grasdand below
the upper Iiinia of forest; (4)forest ; and ( 5 ) mountain formations, including
scrub and grassland above the upper limit of forest. The South Island of
New Zealand was taken as the basis for the definition of mountain plants.
I t was necessary to have some criterion such as this because such inountain
plants as reach the islands of the Soutliern Ocean there descend to sea-level.
Each of these five main groups was next divided according to their method
and opportunities for dispersal thus : Plants belonging t o four groups based
on systeinatic affinity were first counted ; these groups are pteridophytes,
orchids, grasses, sedges and rushes, and composites. The balance were
coiinted according to whether they were herbs or woody platits. ThiJ
grouping is adinittedlj*mixed, being based partly on aysteitiacic and pnrtly
011 ecological characters.
But from t!ie point of Jiew ot dispersal they are
comparable, ant1 the features brought out by this c1asuific:ition are, I believe,
of soiite \ altie. Ertch flors w m therefore divided into thirty p : i h ; these were
then reduced to percentages a i d plotted in the accompanying diagrams.
r i
l h e total area is the saiiie i r i each diagram. For exp1anotio:i of the
diagrams see fig. 6, p. 134.
RELATIONS OF THE NEW ZEALAND REQION.
113.
The New Zealand Flora (fig. l).-New Zealand is an extremely diversified
country with a long and varied coast-line, abundance of freshwater streams
and lakes ; scrub, grassland, and forest extensively developed in both
islands ; tind large areas of mountain country in the South Island, with a
smaller area in the North Island. Its flora, therefore, might well serve as
a standard €or coin parison with those of otlier countries, but iny investigations have not led me so far. The percentages of the five niain groups
utilised in the claqsiticaiion adopted comes very near to multiples of tens,
there being two groups e:tch about 10 per cent. of the flora., coastal and
firshwater plants ; one about 20 per cent., lowland scrub iind gra*sland ;
and two, each 30 per cent., forest and mountain.
The composition of
these groups rnay now be considered. The lending features are these :There is a small proportion of wood.~plants in the coastal group, the bulk
beitig herbs, grasses, and sedges. The freshwater plants are about half
Iierhs and ahout half grasses and sedges. I n lowland scruh and grassland
FIG.1 .
I
I
..,"I* ...
* I
basta1 FteshwaterScrub 8 Forest Mountain
Grassland
R'ew Zedand Flora (see p. 134).
$111 g r o i i p ~are fairly well reljresented. Tile coinposition of the forest is
imt'ortant froiii R distributional standpoint,, as the characteristics of forests
in continental masses and in oceanic islaiida tliffer essentially in features
wIiicIi I hsve endeavoured t o portray. 1 infer, therefore, that the forest on
:In islantl having tlie same characters as that o f a land mass would indicate
that, it, InipriLted thither by a continuous land connection. The New Zealand
forest contains 43 per cent. of woody plants, 23 per cent. of pteridophytes,
and the halance about equally divided between the other groups of orchids,
pissen, coinpouites, and herhs. The New Zealand mountain plants are strong
in woody plants, herhs, and composites.
Lorti IIoolos lsland.Flora (fig. 2).--The
flora of Lord Howe Tslmd is
introduced for coniparison wit8hthat of New Zealtint1 because the island Ilas
at O l l e time been in direct land coniiection with New Zealand and New
CaIetloni:r, hut has heen tt long t h e , perhaps for the greater portion of the
Tertiary period, isolated and hence subject to oceanic conditions as regards
112
MR. W. K. B. OLIVER ON THE BIOGEOGRAI'HI(!AL
dispersal. The flor:~then slionld est1il)it the characters of n I a r p Inild area
with oceanic elements :idded, ant1 sucli, I hrliew, tho an:11ysi,s1 1 s (Jagrain~
niatically given shows. The island is almost entirely covered witli forest,, wllictl
reaches t o t h e highest point ; hence tnonntnin pl:int,.: iire itllsent. Freshwater
condit,ions are lilrem-ise scarcely represotitetl. Tho perceiit:rge of plilntci i n
the coastal group is twice :IS high i t s in New %wl:tntl Tllis wollid foIIow
from t h e greater length of coast-line i n proportioit to area in the t w o ~ t ] ; l c e ~ .
The scrub arid grasslitnd group is sinnller th:in in New Ze:ilunti, hut, like it,
cont:iins representait,iives of all the classes of plants, Iierhs .I)cing espcially
:rbundant. The scrub 011 I m d How(? Island is inainly fo1111d011 t,llc! high
cliffs. Porest is nat,urslly the chief fea.tu.ro of the veget,;itivr cwvrt-ing of
FIQ.2.
.....
.,/'..
!--I
r
*
.
I
.
.
.
I
.*
i..
.
.
.
.
.
I
.
.
I
x.....
,
r
.
*
.
.
.I.*.,
...
I...i.
x
.I.
.
I ,
1."
Coastal Freshwater Scrub & b e s t
Grassland
Lord Howe Island Flora.
RELATIONS OF THE NEW ZEALAND REGION.
113
Fnrtlier evidence of the fornier connection of Lord Howe Island with New
Ze:iland :tilt1 New Ch1edoni:i is furnished by tlie lJTeS011Ct: of a flightless rail,
friclioli~ntmssylaestwk, :I l i t 1 scvoral species of large land shells, iiicludiiig
Ylacostyhs /,k~rr~ic.osics
(see 01iver, Tr:~ns.inst. N.Z: Iiist. vol. xlix. p. 111,
l917j.
?'uw f,'loras in. Nm %ealwi.tE.-The analysis so far given does not disclose
the presence of differcwt Hor:is in tlie saiiie area. In New Zealmd, for
institnee, t,tiere ap~~tt:tr
to J J ~ ? interiningletl IWO floras, one in wliich podocarps, IJities, and trces of M:iI:i?.;i.n :iffinities are dominant, and another in
whicli ~Yot//~ojiLyus
is tlio prcxvalont tree. Brit the plant,s associated witjh
A'ot/ic$i~ycis art? likewise tiiainly plaiits of Malayan alliances, so that the distinctness of the: two types of forest is probably dirc: to age. One is the result
of :in e:trlier periotl of prevalence in New Ze:ilniid than the other, the 8otltofagus forcst t~eingtlio t>:it.lier. It has for the tilost part been displaced by
the mixed forest. I%ot,lik'orests are siiiiilar in t.he clliiraoters brought out in
tlic (liagrmi, so t h t both :ire of ooiitinent;tl type. Were t.he iVotlioftcgrcs
forest to he the portion of :I I'orest iriwlitig S e w Zealttncl froiti the south,
it shonld 1)e accoinI)ariietl by a flora of southern facies, South Aiiieric:m for
instmce, : i d one would expect it to be hest represented in t,lie islands to the
soiitli of New %eal:inil. h c h , however, is not the case. Tile forest of the
AucBland lslancls does not even include i1'othof'ugv.s :is one of it,s inembers.
Its doininant t m e is the sont,lierii rata, ;1/etr(~sitle~~*o.s
Zucida, a tree I)elongilrg
to a genns of M:iI:tyan origin.
Dislrwnio& Fboius.--If a flora o f tlie constittition of th:i.t of New %e:ila.nd
bc? trrrned harrnonic, tlrcri one (Iepart.ing Eroni it in the oliaracter :ind
proportion of :ill it.s nitiiti gron~)s,Ibut especially of its forest, ni:iy he oirlletl
dishnrmonic. This term T have \)orrowed froill tlte writings of zoogeogralJiers, For tlit! floras wliich CI'OSS si;ret,olir?sof ocean t)p occ::tsiona.I meaiis
of t,ra.nsport obey diffrrctut laws to tlioue whicli Iiiigrate overland. They are
the result of ail :icctimnl:tt,ioti of species which liave accidentally come
together ; Iience they would I Jexpected
~
to lack sonie of the essential
chiiracters of' Iinriiionic floras a n d have others enhanced. Such a disharirionic
flora is that of ail o c w i i c islaatl-that is, one that has received its entire
flora overseas. Chastai antl frc:uhw:iter floras exhibit no such differknces as
those foiiiicl betwceri coiitirient:tl anti oceanic forests, but scrub tinti grassland
differ to ii sindl extthtit in contincant,al and island areas. Before nnalysiiig a
typical island of the oceauic cl:iss, soiiir gener:il remarks on the migration of'
plants in tlie solitherti toniperate region will be made.
T'latit;/brmatioizu t i t i t 1 with tlispevsal.- Wi(Iely-distributed species of plant,s
are i i i 1ie:irly all C:ISBS those Leloiiging to 1)l;Liit-fortriatioiIs wliicli are tilore or
less open :ind exposed. Forest, plants usually are iiot widely distributed.
The reasons for this pro1):ihly are that the opportnnities for having their
seeds removed by wind or birds, or, in tlie case of coastd formations, by oceaii
r .
114
Mlt.
W. 1L
B. OLIVER ON T H E BTOGEOGRAPHICAL
currents, are best aff'ordetl in low, exposed formations. The opportunities for
establishment are likewise more frequent in open formations than in closed
ones. Hence coastal, swamp, scrub, itnd mountain plants intike u p the bulk
of the widely-riingiiig species corninon to two or more of the southern land
masses.
The wide distribution of freshwater plants hiis often been eoniuiented on.
They form 10 per cent. of the flora of New Zcalantl, 25 per twit. of the 320
species coinmon to Australia and New Zeiiland, and 35 per cent. of the 80
cosnio~~olita~i
species i n New Zealand. Thus the wider ranging the group
thr higher the percentago of' freshwater plants.
Certain portioiiv of White Island off the north-east cnast of New Zealnnd
:ire path suininer occupied by ganncts, which destroy the vegetati I I where
tliey breed. When they leave the islend a rank growth of herbs springs up
in thrir place. Tlie seeds of these have without doubt been carried, pi-otmhly
hy wind and birds, froin the mainland, h i t the hirds Iiwe macle the opportunity for their establishment.
BistriGtition in 7'fJ)nperuture%ones.-On
comparing the tlistrihution in
their respective countries of plitiits :ind :miinnls coininon to Austrtllia and
New &dand, son18 are fonnd to be restricted t o belts bouiidetl by isothermal
lines. For institncr, a few ni;irine niolluscs of southern distribution in
Australia are found in the saiite latitndes i n New %e:tland. iVcytilus p / a w latus and Argohtcccinum t tnnidum iire esatripl es. Si I oiliirly T o i l r i a cererisiita,
Ostreu cuctilIcIIu, a i d Alz'tru curhona~iuoccnr in the north of New Zealand
and in corresponding latitutles in Austrdiii. Plants foitncl in T:istii:iiiia but
not in Australiii and with a southern tlistrihution in New %cnl:iiid :ire
Oa ult lieria depressu, Do ticit iu f l n w-Zealaiidiw, ['t ricda~*ia
ncoirurit Iios, Liyuvoylyllum Guiuiii, aiid Frlieroclhloe f+ciseri. A nioiig tliosc forintl in Australia h i t
not in T:wnani:t :tilt1 with a northern tli-tril)rition in New Zcn1:iiirl :ire
Calysteggia trturginafa, Broniiis areitarirts, ~ ~ p c o * g c i r ~ istib!yloboairrti,
ri~n
(bss!ltha
p 7 l i C f l l a t U , and botiie orchid-.
Tliese t.pecies two iiieittionetl to show tll;tt
distrilmtion Ii;is prob:ibly taken p1:tce direct :icrodq the Tasm:in SI:I, :t- tltosr
plants of soutlmn distribution could not tolei att. migration hy tc:iy of it 1:ultl
bridge i n a w:irtner region. Most specie+ are itiore generdly tlistrikmtecl 013
both bides of thc. Tikstiiall Sea than those just tiieiitionetl, aiid this would
naturally be the case with plants aiid ~niiiiiilsthat litwe the itieans of cro+iiig
an expanw of ocean.
Tlie West to &st i2lovet,zent.--erliai)s
the ittost important inovt~nientof
organisnis inigratitig tiy ine:tns ot' wind, cnrrents, or anintalw in the suuih
temperate region is that in an eaqterly tlirection. T t is slwcially eiicleiit in
plants where large genera it1 the continental regions have onc or two ontliers
to the eastw;ird. For instance, aiatiy Australian gonera containing lip to 60
species (Persoonin) have ii few representatives i n New Zealand, PheCuZiunl,
s,
and others may be mentioned. Large
/leptospernuvvi, H d o r u ~ ~ iEpacris,
RELATIONS OF THE NEW XEALAND REGION.
115
New Zealand genera like Hehe * and Coprosma have one or two species in
South America. There are Sout,h Africa11 genera like Hypoxis, Tetrayoilia,
Mesembryanthemunz, Restio, Hypoletia, and Wahlenbergia, with few species
in Australia j while there is a regular trail of Fuegian plants eastwards to
the Fdklands, South Georgia, and other subantarctic islands.
The East to West Moopment.-l!his is very small when compared with that
from west to east. Possibly it may be accounted for by occasional bird
carriage or upper air currents. Outlying species of the New Zealand genera
Celniisia, Ourisia, Hebe, and PsychoplLyytou are found in Tasmania. The
Australian genera €filAertia, Keraurbenia, and Rulingia each have one or
two species in Madagascar.
I'h? Flora and Fauna of Oceanic Islands.-'Flie study of the plants and
animals of islands whicli on geological evidence appear never. to have been
united with a larger land area should throw light on the nature of those
which are able to cross wide expanses of ocean. The Kermadec Islands may
be taken as an instance. According to my own observations (Trans. N.Z.
Inst. vol. xliii. p. 524, 1911), these islands have been built up of volcanic
niaterials on a subinarine bank. Their plants and animals must therefore all
have crossed at least 600 miles of ocean. And it is significant that the bulk
of them are related to species found in New Zealand, froni which direction
come the prevailing winds. The surface currents, as evidenced by logs of
New Zetlluud origin cast up on Sunday Island, flow in the same direction.
The vasciilar plants of the Kermadecs consist of 38 pteridophytes, 20 grasses
and sedges (including Juncus and !l:yyha), 2 orchids, 9 composites, 27 other
herbs of which 1 7 are coastal and none forest, and 15: woody plants of which
14 are forest. Included in the foregoing total are 9 species with succulent
fruits. The flora is fragmentary in the sense that there are only one or very
few to each genus, and largc Npw Zealand genera, such as Hebe and Coprosnia,
are represented in this way. Thus, although iiiost ecological groups occur in
the Kermadecs, including trees with succulent fruits which almost certainly
depen(j for their transference on birds, most of the species are terns, grasses,
sedges, composites, and plants of open formations. Of the 114 species, 86
found in two or more of the adjacent regione of Australia, Polynesia,
atid New Ze:iland (see Oliver, Trans. N.Z. Inst. vol. xlii. 1). 149, 1910).
The main features of the flora are shown in the diagr:im (fig. 3). The
proportion of coastal plants is much higher than in Lord Howe Island.
'JJlis, of course, is due to the fact that ocean .currents are one of the means of
trrinsport, and the land pl:tntq depending on occasional means of dispersal
number less than half' as inany as in TJord Howe Islaiid. The coastal plants
* The following groups are here usedin
R,
generic sense, though appearing only as ~ectioi;v
of genera in Cheesernnn's ' Mnnual of the New Zealad Flora,' 1906:-Hebr, fiJgwm(J
[both = Veropaiea], Schizekima [= Azorella], Edwwdsin [ =rSophorn], Gymnelrea [ = Olm],
Leucoyenes Reauv.
LINN. JOURN+-BOTANY, VOL. XLVIIK
116
YR. W. R. B. OLIVER ON TEE BIOGEOC?RAPEICAL
of the Kermadecs include relatively more composites, grasses, and sedges
than do those of Lord Howe Island. Freshwater plants are few and
mountain plants altogether wani,ing. The scrub plants of the Kermadecs
compare with those of Lord HOW Island, though exhibiting a higlier
proportion of pteridophytes and composites.
Even here the insular
charsct>er of the flora is evident. But it is in forest plants that the
Kennndecs differ esseritinllly from continental forests such as Lord Howe
Island and New Zealand. The proportion to the whole flora is high because
practically the whole island is under forest. The proportion of species of
trees in the forest, rvliich is 43 per cent. iri N e w Zealand and 40 per cent. in
Lord Howe Island, falls to 30 per cent. in the Kermadecs, whereas the
percentage of pteridophytes in the forest, from 23 in New Zealand and 33 in
Lord Howe Island, rises to GO per cent. in the Kermadecs.
FIG.3.
Coashl
Freshwater Scrub 8, forest
Grassland.
Kermadec Islands Flora.
Plants common to Australia and iVew Zealand (fig. 4).-In
the light, of
results obtained by comparing the floras of oceanic islands like the
Kerlnndecs with continental floras on a large land mass as in New Zealand
or isolated a 3 on Lord Howe Island, it will be profitable to analyse in a
similar way the plants couimon to New Zealand arid other countries,
Beginning first with those found in Australia or Tasmania and New Zealand,
320 in numher, they may be expressed i n diagrammatic form as with the
floras already discussed. CoasLal plants are in the percentage to the flora
here under examination midway between those of New Zealand and Lord
How(? Island, with a composition similar to the latter. Freshwater plants
Occur in large proportion, 25 per cent. of the flora, and consist of nearly half
herbs and two-fifths grasses, sedges, and rushes. As has already been
pointed out, freshwater plants are apparently easily transported over oceanic
areas, so that their fewness on Lord Howe Island aud the Kermadecs must
be due solely to the fact that the conditions reqnired for their establishnlent
RELATIONS OF THE NEW ZEALAND REGION.
117
are there of small extent. Scrub and gr,zsslnnd plants are well represented,
and in their composition resemble very closely those of the Kermadecs.
Forest plants consist of 75 per cent. pteridophytes and the balance herbs,
grasses, and sedges. The mountain plants forin the sni:illest group, but this
may be esp1:lined by the fact t8hatthese plaiits are hut a sinall proportion of
the vegetation of south-eastern Australia a r i d Tasniania.
Coinparing the plants common to Australia and New Zealand with those of
the Kermadecs, it will bs seen that in those characteristics by which the flora
of tlie Kermadecs differs froin that of New Zealand, the Australian element
in the Kew Zealand flora differs in a still greater degree. It is iiideed more
“oceanic ” in character than the flora of au oceanic island. This element is
of course complex, consisting of species which have reached New Zealand
and Australia indopendently from the Malayan region, species which bavc
reached New Zealand overseas from Australia, and species which have
migrated in the reverse direction.
FIG.4.
Coastal. Freshwater.Scrub & Forest Mountain
Grassland
Austrslia-New Zealand Fauna.
P l a n t s rontmon to New Zealand and South A m e i i c a (fig. 5).-Now compare the diagrammatic representation of the 70 species of plants common to
New Zealand and South America with that of tlie 320 cominon to Australia
and New Zealand. I n the relative proportions of the five main ecological
groups there is an apparent difference owing to the different, proportions of
the scrub and grassland and niountain groups. But this is due to the latitude
in which dispersal takes place, for the scrub and grassland plants of the
islands of the southern oceans are mountain plants in New Zealand and are
counted :is such for the puryoae of the diagrams. Now, these form a large
proportion of the plants corninon to New Zealand and South America.
Hence, if the scrub and mountain groups be counted as one, the agreement of
the diagrams is remarkably close. The percentages for the Australian
element in the New Zealand flora are coastal 14, freshwater 25, scrub and
grassland 42, forest 19. I n the South American element the corres~ionding
figures are 17, 28, 36, and 19. The conipoiients of these groups are also in
K2
118
MR. W. R. B. OLIVER ON TEE BIOGEOGRAPHICAL
the two elements quite similar. The most n.oticeable difference is in the
forest group, there being in the South American element two trees. These
are Edwavdsia micropliylla and Coriaria ruscifoolia, both species which a t the
present day are, there can be little doubt, :tctually being dispersed, both being
found on intermediatt, islands in the Pacific Ocean. Guppy found that the
seeds of Edwardsia microphylla germinatod after floating for seven months
in sea water. Goriaria ruscifolia is a plant of open scrubland niore than of
forest, and bears nuuwrous sinall sncculeiit fruits which are greedily eaten by
birds, though the chances of these being carried by sea-birds would possibly
be through sticking to the plumage.
FIG.6.
Coastal Freshwater Scrub 8 Forest Mountain
Grassland
New Zealand-South America Flora
Taken as a whole thc plants coinmon to New Zedand and South America
belong to groups coniparable in their characteristics with those common t o
New Zealand and Australia, and they exhibit in a high degree those features
which character& the floras of oceanic islands. In reality, as will appear
in the final part of this paper, the South Arrierican element in the New
Zealund flora is a complex one, consisting of (1) species which have reached
both countries by migration from the Northern Hemisphere, (2) species
which have migrated from New Zealand to South America overseas, and (3)
species which have migrated from South America to New Zealand overseas.
The possibility of m y of the specias of plallts at present common to New
Zealand and South America being due to their having crossed by a direct
land connection is not here admitted.
V. HISTORY
OF THE NEWZEALANDFLORA
AND FAUNA.
1. Mesozoic Floras.-It is insisted by all who study them that the Triassic
an(] Jurassic floras are similar in type throughout the world. They are
known from all the continental masses, incliiding Antarctica and New
Zealand. As in the earlier floras, therefore, land connections in any definite
(jirection are scarcely indicated. Of the Jurassic plants known in New
Zealnnd six extend to the British Isles, six to India, nine to Australia, and
five to @&an1 Land. The species common to Australia and New Zealand are
RELATION8 OF T H E NEW ZEALAND REGION.
119
given by Arber (N.Z. Geol. Surv. Pal. 13ull. No. 6, p. 24, 1917) as follows :
Cladoplilebis australis, Thinpt$eldia lancijXiu, T . otlontopteroides, T. Peistmanteii, Tmniopteris Ilaititreti, 7’. c.rassineruis, Coniopteris Iiy~neno;c7ti~lloiaPs,
Sphenopteris Currani, IGtocladus corfiyta. Of these, Ectocladus belongs to
the Coniferales ; the remairider are fern-like plants and may be seed-bearing ;
further, of the 45 Mesozoic plants known froiii New Zealaud, six :we Uycadofilices, eight Coniferules, and of tho 27 fern-like plants included in the
remainder many may prove to be seed-bearing. A land connection is thcrcfore demanded with some other portion of the world. Nine species of the
New Zealand Mesozoic plants occur in the Upper Triassic (Hh~etic)beds, which
would place the land bridge before this time. I t might well have occurred
during the early Triassic period when 011 geological ovidence, according to
Marshall, Park, and others, a break in the faunal succession ;ind a period of
orogeny took place. As to the direction in which this land bridge lay, it is
not necessary to assunio that there was continuous land in temperate regions
joining New Zealand and Austriilia. This is suggested by Arber and mapped
by Benson, though Arber states that the comp:irison of the New Zealnnd
with the Australian and l’asmanian Jurassic floras is more remote tlinn iiiight
be anticipated. An extension of land to the north along the routc afterwards
taken by the Malayan flora would be the probable connection in early Mesozoic times.
2. Malayan Land Connection.-A large liroportion of the plants and
animals at present living in New Zealand, perhaps the bulk of them, are such
as require continuous land connection for their dispersal. Tlieir presence
demands th:it at some period ill the past, New Zealand was joined to the other
land mass of the globe. Most of theBe animals and plants are related to
species now found in lands to the north, and an explanatiou of the origin of
these must be consistent with the fact of the fundaniental difTerences between
the faunas and floras of thc south temperate land masses.
Distribution of Coniferales.-The distribution of the Australian and New
Zealand Coniferales must be considered here. The presence of all may be
explained on the assumption of a northern origin. Aruucaria, Ayuthis
( =Bamnzara), Libocedrus, Potbocarpus, and Yliyllocladiiu or allied forms all
occur in the Cretaceous and Tertiary of Europe and North America. Thus
, their presence in the southern land masses is explained by iiiigration along
land lines from the north. Knoche (Etude Phytog6ogr. Iles Ualdares, p. 155,
1923) holds this view regarding Liboeedrus and other genera. The only
other New Zealand genus, Ducrytlium, is represented by scveral species in
the Malayan region and New Caledonia. There are seven species in New
Zealand, one in Tasmania, ilnd oiie in Chile. Fossil species have been
described from New South Wales and New Zealand.
The Australian genera iiiclude all those in New Zealand except Lilrocedrus,
120
MIL W. X. B. OLIVER ON THE BIOGEOGRAPHICAL
and, besides these, six genera confined to Australia and Tasmania. I ) i s e l w ,
Microcachys, and Athrotuxis are found in Tasmania only, the first two, and
Yherosphccru with one specie8 in Tasiiittnia arid one in New South Wales, arc
closely related to l h r y d i u m . 4throtaxis had allies in the Tertiary of
Europe. CdZitris and its ally Bctinostrohus are related to African genera.
The Coniferales, being an old order, show a good deal of diversity both in
Australia and New Zealanti. The congregation of‘ genera in Tasmania
perhaps shows, as in the case of New Zealand, some former land extension
followed by contraction.
Tlie joining up to Australia and subsequent
separatiou would possibly account for this,
Ilistribution qf J’AGUN.-L’hc history of Pqus (including N o t l ~ o f u p is
~)
apparently precisely similar to tlint of Armccuriu and Agnthis. Nothofagus
diflers from Fugus onlg in the m d e r size of the flowers and leaves and
i n tIie fewer flowers in the tiide catkins. Some species in Australia
(3’.
~Mooi*ei)and South Aiiierica have large leaves. It is found in South
America (8), Tasmania and Eastern Australia (3), and New Zealand (4 species
and several hybrids), whereas Pugus is confined to the north temperate
region, iiicluding Japan. But Pugus hns heen described from the Upper
Cretaceous of Kansas and various Tertiary localities in the United States,
British Columbia, Alaska, and Europe. Fossil plants assigned to both Pugus
and I\.’othojbyus have heen described from the Oligocene of Graham Land,
whilst possibly certain Tertiary p1:uits fronl Austra1i:L and New Zealarid may,
a s Ettingshausen believed, be referred to Fa921s. It is probaLle that Fayus
and iVothojayus originated in Nortli Aiiierica and spread thence east, south,
and west. The western moiety l)nssed, via Japan, round the Pacific, reaching
Australia and New Zeal:ind. A similar place of origin and routes of dispersd
would explain the past and present distribution of Arar~cariaaiid Agathis.
But these two genera have become extinct in North America, whereas Fugus
still persists. If the characters by which h7)tl~ofugusis separated from Pugus
be considered primitive, then these t w o genera exemplify the principle
enunciated by Mathcw, which states t h a t a group should be most advanced
at its point of origiiiul dispersal, the most consurvative stages being farthest
from it. I find that Guppy believes in the northern origin of the New
Zea1:ind and South Ainerican spcciea of P’uyus (including NotAojhgus) (Plants,
Seeds, and Currents in the W e s t Tndies and Azores, 1). 326, 1917).
Upper Ci.rtuceous Rlora.--The late Cretaceous and early Tertiary floras of
New Zealand are known only from the determinations of Ettingslrausen,
whose iclentifications have not been generally accepted. Many of the plants
were referred to northern genera, one was compared with :L Greenl:Lnd
species, and tlie conclnsion arrived at that the Tertiary flora of New Zealand
was a part of that 11niverw1original flora from which all living floras of the
globe descend. Ettingshausen supposed that from one part of the Tertiary
RELATIONS OF THE N E W ZEALAND REGION.
121
flora of New Zealand the present flora was descended, whilc the other
portion became extinct. Probably he has erred on the side of referring too
many of the plants to northern genera, tliough there is nothing inherently
improbable in supposing that the first dicotyledonous flora soon became
widely distributed, and that the modern floras have differentiated from and
displaced it. But such a universal flora would be Mesozoic, not Tertiary.
Ettingshausen referred the Shag Point and related plant beds to the
Tertiary, whiie the Nelson and Westland series (Paltawau, Wangapeka,
Reefton, and Grey River) was classed as Cretaceous. This order is now
generally reversed. I am indebted to Mr. P. G. Morgan, Director of the
New Zealand Geographical Survey, for kindly supplying me wiCh information as to the relative ages of the principal New Zealand plant beds.
Oinittiiig those identifications not f'nunded 011 leaves, i 6 iriay be profitable
to analyse in a general way the floras described by Ettingshausen. Those
which may be considered as of late Cretaceous age consist of 35 species from
Shag Point, besides a few frorn Malvern Hills, l'aparoa, liedclifl'e Gully, and
Murderer's Creek, Of the plants from Shag Point there are two ferns, one
of which, said by Ettingshausen to occur also a t Dunstan, a rnid-Tertiary
locality, may be compared with the recent Dryopteris pennigera. The
gymnosporms comprise two species of A g u t l h , two of Araucaria, three of
four podocarps and Scqzcoia Novce-%eulantEiB. Judging by both the present
and Tertiary distribution of these genera, northern relationsllips are indicated.
The dictoyledons include eight species with simple ell tire leaves, a type
characteristicof the existing flora. Three spccies refrrred to Picus, lledycurp,
and Cinnamoinum respectively likewise iudicate an alliance with the north.
I n addition to these there are eleven species having sprrated pinnately-veined
leaves and two with palmate leaves. These are niainly referred to the
Cupuliferse, Myricacez, and Ulmacez, families which, except for Notlwfugus,
are scarcely characteristic of the present flora. Nothing can be said with
any degree of certaintj regarding the relationships of' these plants, but
if Ettingshausen's determinations have any value, they would SUpiJOrt the
d
towards the nortli.
evidence of the gymnosperm for a l i ~ ~ iconnection
Such a connection would have been during Cretaceous times.
-Eocene FZora.-Thc plants referred to the Cretaceous period by Nttingshausen are in reality of later date than the Shag Point fossils, and p i obably
should be classed R S Eocene. They consist of leaf and other impressions from
the Nelson and Westland districts (Piikawau, Wangapelta, Reefton, Grey
River). Four species of ferns are described, of which Blciclu?nia obscura and
BEechiiuin priscum show relationsliips to recent tropical species. A fan palm
nained Flabellaria subl~n~qiracliis
was present. The gymnosperms consist of
Ginkgoelailus NovwZealundice, a relatioil of Pl~yllocladus; D u m m w u Muntelli,
scarcely distinguishable from Ayathis australis ; and six of the appearance of
122
NK. W. R. B. OLIVER ON THE BlOGEOGRAPHlUAL
podocarps, a group well represented in the living flora of New Zealand.
Tlie ferns and gymnosperms, therefore, show unmistakable evidence of
relationships with the Malayan elenient of the New Zealand flora. The
dicotyledons include f i v ~species wit11 simple entire leaves and five with
serrated leaves. These simple leavos are quite characteristic of tlie present
flora of New Zealand. Two leaves named Ficus similis and CinnamoniunL
E-laastii belong to types iiot now found in Ncw Zealand, though, as Ficus and
Cryptocarya, reaching Lord H owe Island. Besides these there are eight
species referred to the Cupuliferz and Ulmiipes.
As in the late Cretaceous flora, the eviderwe of the dicotyledons for determining rulationsliips is inconclusive. Tlie gymnosperms, liowever, hy the
absence of Araucaria and tlie presence of Ilummura Mantelli, Ginkqocladue
Novce-Zealandice, Potlocarpium prteducrydioidcs, arid Uacrydium cupressirrum
show closer relationship with the existing flora of New Zealand, and indicate
the same alliances. Whether Tertiary or present gymnospermous floras be
compared, therefore, the result is the same-namely, a forriier northern land
extension is proclaimed.
The Tertiary flora of Seymour Tsliind, held by Dusen to be Oligoccne,
contains the genera Luureliu, Jbrirnys, Kniglitia, n’otliofagus, a d AiBuucaria.
In all these, cxcept Kiiiglitia, the relationship of the species is with South
America. If correctly determined, the leaf referred to Knightia is of considerable interest.
Pa1ceo:eaEondic C;enera.-Clhsracteristic of New Zealand are many genera
and family groups so distinct from any known elsewhere, yet taken together
obviously descended from animals slid plant,s that inust have existed at a
period when there was land coiiiiection hetwcen the New Zealaiid area and
some other land mass, that tliey iiidicate i t loiig period since the connection
was severed. The last date that New Zealand formed part of this continent
may be taken as so111e time in the Cretaceous period. It cannot have been
later, otherwise land maninials wonld have entered the New Zealand portion.
Cockayne’s term Ynlaeozealandic (Veg. N.Z. 1’. 315, 1921) niight be applied
to the genera Irere listcd. His group is of mixed origin according to niy
views, some of the geiier:t having aristm from the original continental flora,
others from species that have afterwards come overseas.
The continental genera I include under the general heading Malayan
elenlent, because t!iey represent ilie earliest of the higher aninials and plants
to people the New Zealand area, which I believe would be washed by the
ocean except towiirds the north-west, and in their broader affinities may in
most cases be compared with groups of nortliern origin.
The dieerence btstween northern and souther11 distribution is not so well
marked with birds 11s with plants.
KELATlONS OP TEE N E W ZEALANI) BEQlON.
NOBTHBRN
DISTRIBUTION.
(a) Genera w i t h
l'lants.
Entelea, illelieytus, A lretnJon, A s t e l k ,
Lrerba, Dactyknthuu, dlseaosmia.
Birds.
Hetwolucha, Calhaa, Crerulion, Turncigra.
123
SOUTHERN~IIBTRIBUTION.
120
tiear relations.
Plants.
Coralbpnrtium, iYotospartiuna, Chordoq a r t i u m , Carmichelici, Ansotome,
Aciphylla, Coxelia, Padbyeladon, NothothInspi, StSt41bocu,~apa,,~yosot~diu9n,
Celnaisiu,
Ilaastia, Leucoyeues, Yhoriniiim,Hoheriu,
Raoule'a,
Pseudopanax,
Sisimnpliciu,
Psychrophyton, Pleurophyllum.
Birds.
Bowdleria, Neoolimnas, Cubabs, Notornis,
Diaphoropteryr, Apter.y.r, Di9&ornithzd@,
Xenicus,
Traversia,
Acanthobittn,
Nesonetta.
Mammals-ik'ystarops.
Reptiles-Sphenodon,
Uatraohia,-&opeha.
(b) Genera whose relationships are w i t h Alalaya and New Culedoniii.
l'lt111ts.
Rhabdothnntnus, Gymnelcca, &q?odetws,
Knightita.
Birds.
Henip hnga.
Plants.
S@honid%m, Hebe, Coprosma, I'ygnma,
Chrysobactron.
Birds.
Nestor, Gallirallus.
( c ) Genera whose relotionships are w i t h Australia.
Birds.
Birds.
Sceloglatu, Myionwwa, Anthornis, 8trigops.
Ziro, Notomystis, Prosthemadera.
(d)
Genera whose relationships are with Amrrica.
Plants.
Corokia, Lonoma.
Plants.
Hectorella.
Tile New Zealand Continent.--Perhaps the outstanding feature of these
lists is the large number of genera, includiiig many with a large number of
species, with a southern distribution. I take this as indicating a foriiier
considerable extension of land about and to the sout2h of South Idand, but
not necessarily connected with the Antarctic Continent. The date can only
be conjectured, but it may be put down as one of the periods when, judging
by the geological history of N e w Zealand, there was a general uplift,
perhaps in the Cretaceous and Eocene anti again In the late Pliocene periods.
T w o biological considerations indicate the continental charaoter of the land.
First, tilere is great diversity of species in the soutlierii genera, species with
their areas of distribution overlapping. UinornitliidE, Apteryx, Gallirallus,
Hebe, Coprosmu, Curmiehueliu, Aciphylla, Icuoulia, Celmisia, and Anisotome
may be mentioned. The crowding together of these specics suggests a tornier
wider area where they differentiated. As the land area diminished they have
124
M R . W.
K. H. O L I V E R ON THE BlOGEOGRAPHICAL
come together, so that many allied species are found in the locality. Willis
interprets theso facts as a southern invasion taking place later than a northern
one (Ann. Bot. vol. xxxiii. p. 40, 1919). Where this so-called invasion came
from we are not told.
Secondly, there are xerophytic characters in inany New Zealand plants
either in their whole life-history or during a portion of it, in a cliinate in
which at prasent one would expect only iiiesophytes. Diels appears to have
heon the first to suggest that a continental (5stension was necessary to explain
the presetice of xerophyte plants in New %eal:ind, while Cockajrne explains
the developrncnt.al stages of many plants on tile same assurqltion.
As tliese modificittions affect entire genera, it must be prcsuiiicd that ttie
coritinental exten4on wllicll induced t'tic4r development was at the period
when these genera were differentiated. Th:rt is, it inust be p1:rcecl early in the
Tertiary period, and inny theref'oro be ment.ionod in connection with the
multiplication of species ref erred to the same cause.
On account of the 1n:tiiy lines of evidence, both geological and biological,
pointing to IL former New Zoa1:tnd coi~tinent,it has beer1 ncceptetl by geologists and biologists alike. Ttia controversial points concern tthe area it
occupied, the titne it existed, and the lands it joined. The extension of' land
above indicated with a northern connection would eviclently be suficient to
expliiin the Malayan hasis of the New Zealaiid f:tuna and flora and the
diversity of life now crowded in a coniparatively smiill area. The early New
Zealand continent would be a centre for the tlevelopnient and dispersal of
many of the forms of life so characteristic of southern regions, including
much of the so-called '' Antarctic " flora. and fauna. Thc penguins, shags,
and petrels among birch, t,he G:tl:ixiad~among fish, the &otri(h> and perhaps
sotne marine niolluscs, ui her invertebrates and algae characteristic of southern
regions and whose hmdquarters are in New Zcaland, owe their development
and distribution to the New Zealand continent, which in former times
stretched towards Antarctica, the shore of' which would act as a route for
dispersal. The only other co:ltinent in the same latitude was South Anierica.
Australia and Tasmania since the Jurassic period do not appear to have
extended inuch farther south than at present.
Endemic Species qf Llluluyan urul Australian Geiaeru.--Beside
the genera
already mentioned tliere are in New Zealand representatives of itutny genera
which Show their greatest area of developinent in the Malaym region. The
speciei in New Zealand helonging t o these genera are ali eiideniic, and their
ancestors would enter by ttie northern land bridge described above. They
would, in fact, be the last to enter by such a way. Hence t,he distribution of
these species in New Zealand is for the most part northern.
The principal genera of Angiosperms in the New Zealand flora cbniing
under the present heading are the tbllowing :-ilristoteliu, I)raptes, C'oriaria,
Edwardsia, Uagnisia, Nothopanax, Hedycarya, Melieope, Litma, Beilschmiedaa,
'
RELATIONS OF TEE NEW ZEALAND REGION.
125
Weinmannia, Corynocarpus, Meryta, Schflera, Gaultheria, Pratia, Paratrophis, and ivetrosideros. I cannot name any birds coming under the
present heading; perhaps they change at a quicker rate than (lo plants.
Certain land molluscs, liowevcr, sach its Plucostyku, Rliytida, and ./'ayplianta
may be mentioned here.
Among genera characteristically Australian are soiiie which possibly
indicate migration over a coiitinuous land surface. These may have entered
by wily of Northern Austr;diii, New Caledoni:i, and Lord Howe Island.
The followiiig genera have species iii one or more of these places :-Olearia,
UracoplLyllum, l l y menantheru, Yennantiw, E,:cocarpus, and Rhipoyon~nz.
lhree genera, E'usanus, Quintiniu, and Ackamu, are confined to the north of
New Zealand, while Pluyicrritlius is tound in tlic south a5 well. Thus the
Maliiyiiii land connection may explain the presence in New Zealand of
Auhalian genera with all endemic species in New Zealand. That such
genera as Drucophyllum and Oleuria have been in the New Zealand area
il long time is indicated by tho large numlrer of species belonging to each in
New Ze:iland and their mainly southern distribution.
3. The Influence of htsrctica.--Pcrliaps n o point concerilirig the origin
and distribution of the New Zealand fauna and flora has given rise to more
controversy tlian the so-called " Antarctic " element. This appeclrs to me to
be a mixture of several elen~entsM hicli are considered in different places iu
this paper. From thc genera and species of plants coiiinion to New Zealand
and South America I have first eliminatcd those which may be explained by
migration froin the north overland and from the west overseas. Hut there
remains a residue which seem to deniand a iiiore direct land route between
Ncw Zealand and South America. By most authorities a land bridge is
considered necessary. Thus Hutton, Benharn, Utiilton, and Cockayiie in
New Zealand and Hodlcy in Australia favour a continental connection.
(~heescinaii, Schucher, and Schenck, however, on the evidence of the tlora
think a closer approach of the land areas sufficient.
On account of the relatively sniall proportion of the New Zealand flora
with '' Antarctic " affinities and the larger Malayan element, also the contour
of the ocean bottoni and physical conditions of the Atitarctic continent, it
seems safe to assume that the most active period of transfer between the
South America 1 and New Zedand floras anti faunas must liave been a t the
time of New Zcalaiid's greatest exteitsion in late Mesozoic or etrly Tertiary
times.
Thr genera of plants which, judging from their present distribution, have
their greatest development in South Anictrica, arid therefore i(r0 prtasuined to
have supplied thence the New Zealand representatives, are Griseliniw,
Ourisia, l)isca~ia,and Gaya with R predominantly southern distribution in
New Zealand, and Eiuci&u, Joaellaria, Laurelia, Plwy$antAus, and Mueiilenbeckia with a mor0 northern or general distribution.
r y
126
MR. W. R. B. OLIVER ON THE BIOGEOGRAPHICAL
Unfortunately there is nothing known of the early Tertiary Antarctic flora
beyond a few Oligocene plants froin Graham Iland. These are in the main
South American types. However, if they are to be taken as an indication
of t.he flora of the Antarctic coast a t that time, then i t is evident that New
Zealand received no inore of it than fragments that might have crossed, with
the assistance of birds or wind, a siiiall expanbe of ocean.
Griselznia h:is four s p ~ c i c sin Chile arid two in New Zealand, the latter
species being different in appearance and perhaps belongiug to a distinct
section of the genus.
Ourisia h:ir 15, species in South America, eight in New Zedand, and one in
Tasinania. They are inainly plants of ~nountninousdistricts, where opportimities for dispersal and establislinient are frequent.
Discaria has about 18 species in extra-t ropical and Andine South America,
one in Australia, and one in New Bedand. The two last are closely
allied, and 11. discolor of South America is related to 1). tounaatozr of New
Zealand.
The New Zealand species of G a p has ~nnchlarger flowors than any of
the LO Sontli American species.
Fuchsiu has :houtj 60 species i n America from Mexico and Fuegia. Of
the three New Zealand species t w o :we closely allied, and the third is local
in the northern portion of thc Dominion.
Neither of the two species of Jowellaria in New Zealand is generally distributed. There arc two or inore species in Uliile and Peru.
Laurelin has two species in Soiitli Cliile and one in New Zealand. An
extinct species has been described froin the Oligocene of Graham Land. The
genus is nearly allied to the Australian AtkProspzrma.
PArygilanthus has about 20 species in South America, four in Australia,
and two in New Zealand (both rare).
Mueklenheckia has 10 species in South America, seven in Australia (one
extending to New %e:ilaud), four others in New Zealand, and one in the
Solomon Islands. Of' the New Zealand species three are iiiainly coastal and
another occurs in mountain localities.
The species above mentioned show ii certain amount of distinctness from
the related South Anit~icaiiforins, thus indicating the lapse of a long period
of time since disporsal took place. They iiiiglit well be the descendants of
stray immigrants that crossed the seu that soparated ttie late Mesozoic or
early Tertiary New &alaiid continent from Antarctica.
Besides the genera listed ahove urhich indicate the derivation of New
Zealand species from a South American source, there are a few others which
point to migration in the opposite direction.
U a c y d i u m , a genus probably of Malayan origin with its present greatest
devclopiiient in New Ze.iland, has in Chile a single species, B. Ebnckii,"
related to D . 1a;zifolizm of New Zealand.
*
Hutchinson (Kew Bulletin, 1924, p. 64) omits Chile in giving the range of Uucrydium.
RELATIONS OF THE NEW ZEALAND REQION.
137
Pseudopanax has five specie5 in New Zealand and two in South Chile.
Perhapc, bicentric genera like br&da, and Gunnera with the +ubgenus
Milliganiu confined to New Zealand and Tasinania, and the closely-related
subgenus Jlinonrlm t o Chile, Fuegia, and the Falklarid Islands, owe their
distribution to the former pi esence of a habitable Antarctic continent.
Discassing the distribution of Cinci~iia,Guppy (‘Plants, Seeds, and Currents
in the West Indies and Azores,’ p. 50L, 1917) comes to tlie conclusion that
whilst South America was the original differentiating ground of the genus,
New Zealand with a single section h:is been i n later times Inore vigorous
and productit(. of species. The sanie author, however (p. 328), thinks that
Antarctica lins not shared in the history of the plant world since the
appearance of Dicotyledons.
The genera enuiiiernted in the proceding pwagraphs niust, be taken as
indicating at least an approximation of the New Zealand area to that of
Antarctica at some time in the past. It is known that Graliain Land supported a land flora as late as tlie Oligoccne period-that
is, long after
Dicotyledons appeared. But a direct land connection does not appear to be
necessary, hecause of their fragmentary nature and, as has already been
pointed out, the species now common to South Aiiierica and New Zealand
forin a disharinonic community.
There does not exist in New Zealnnd a plant association related to any in
South America, all doininant plant speciss in Now Zealand, including Notliofagus, being of northern derivation. I t may l)e pointed out too that both
the Tertiary, so far as is known, :lid the recent floras of New Zealand and
South America a r e funtiamentally different. The views hereiir expressed
coincide almost uxactly with those of Cheeseman (Rep. Aust A n t . Exp.,
Rot. vol. vii. pt,. 3, 1). 53).
South A merican- Tasmcinian Biological Relations.-The relationsliip of the
Hora of Tasmania to thal of South America is far less than tliat of New
Zealand to South America. TIie New Zealand continent, indeed, seems to
tynve been the source of such plants in the Twmanian flora as Acipli?$la,
Psychrophayton,and Schi:eleimu, and, if so, probably that portion of the South
Aineriran element in the Tasintlnian flow represented by Guni~ernand Ourisia
was received hy way of the New Zealand continent.
The South Ainerican element in the ‘hsmanian flora is apparently quite
small, and that portion which it has not also in conimon with New Zealand
might be explained by tlrit’t from tho Antarctic continent when its shore supported vegetation and the land extended froin AuJtralia to the south of
Tasim nia.
Lomntia has three species in Chile, four are described froin the Oligocene
of Graliam Lalid, and thcre arc six existing species in Tasmania and Eastern
Australia.
Ewtbothrium has four species in South America and one in Eastern
Australia. Ewrypkia has t w o species in Chile, one in Tasmania, and one
128
MIL 7V. R. R. OLIVER ON THE BIOGEOURAPHICAL
in New South Wales. Prioiiites has one species in Fuegia and one in
Tasmania. On thc other hand, two Chilean species belonging to the genera
Fitzroya, (allied to Iliselma) and O r i t t s may indicate drift from the
T:ismanian land extension to the Antarctic shore-he.
Much has been made in the past of the relationship of the mammals of
South America and Australia, hit it has been shown that the South American
ProthyZucinus is a creodont, and that the so-called Diprotodonts of South
America nre not to he included with the true Australian Diprotodonts, but
are in reality a distinct group descended from Ainericari Polyprotodonts.
Likewise the reptile from the South American Tertiary supposed to be near
the Austrdian Miolunia has been shown to he quite distinct (Regail, Terra
Nova Exp., Zool. vol. i. pt. 1, p. 41, 1914). Finally, Dunrl (Amer. Naturalist, vol. lvii. p. 135, 1923) h a y pointed out that all Aniphibinn distribution
can br explained without recourse to land bridges save connections in the
north between the northern land masses.
4. Species Dispersing Overseas. Species dispersiny from Australia to New
zealand. (a) Zndetnie New Zeciland genera.-The constant arrival of species
of Australian plants in New ZenlLtnd overseas during the Tertiary epoch
would result in different degrees of endemisin according to the time since
the species established themselves in New Zealand. Species arriving early
and not subject to later additions might become so different as to be classed
as (]istinct genera, others might differentiate into distinct species, but those
species that were constantly dispersing individuals to New Zealand or which
arrived recently woiild be iden tical with Australian ones.
These conditions would result if during past ages there had been a steady
easterly movement, including chance arrivals and regular niigrants, ot' plants
across the Tasmaii Sea. The first degree of endemisin by which there result
in New Zealand distinct generil derived from Australian species may be
represented hy OreostylitEiiim, Pli9lyllachtie, Colensoa, IlgcEutella, Sporoduntlius,
and OreoboZus. Of these, PhylEacline and Or~obolushave supl)lied species
farther eastit ard to South America.
(b) Australian genera with e7ulew& species ira New %ecclund.--(l) PLANTS.
Included here are :L nuinher of fairly large Australian genera with one
or few species in New Zealand all or some of which are endemic. The
following genera have each one endeiriic species in New Zealand, the
reIl1:tinder of the species being Australian : PliebuZium (27 species in
Australia), Epawis (23), Persooriia (60), Myoporun~(25), Swuinwnia (32),
and Logunia (18).
J'g,elia has 80 species in Australia and 12 in New Ze:ilaatl. They :Ire
mostly plants of the coast and mountain scrub. f1aZomyi.s hns 41 specieh in
Australia and Tasmania, of which four extend to New Zealand, and there is
an endemic species in New Zealand as well. L+ospernaum has 25 species in
Australia and three in New Zealand (one of which is Audraliun). (7e?ttroZepis
RELATIONS OF THE NEW ZEALAND REGION.
129
has 18 ‘species in Australia, one in Now Guinea, and two in New Zealand
(one of which is Australinn). Other Australian genera with a few species in
New Zealand are Pornaderris, L>’rackyeome, Cyutiboiles, and Leucopogon.
Smaller genera t11a.t may be mentioned liere are Artimpodium, Yoraiztlhera,
Archeria, and Forstera.
In all the nl)ove genera, the New Zealand species, judging from the’ sinall
nnmber reliresented, may be assriined to be descendaiits of c1i;ince arrivals.
I n some cases the distribution suggefits thitt Australian genera suppIy
specics to South Anieric~adirect, or by way of New Zeahnd. Thus Yratia,
a genus with its 11e:dquarters in Australiii, but extending to New Guinea
and the Hinialayas, has three specics in New ;(lt!aland and one in South
Americ:i, Lagsriopitora lias four species in Australia and Taslnania, six in
New Zealand, two in Polynesia, and foul. in Soiitli America. Abrotanellu is
very similar in distribution, having t’liree species in Tasmania and Victoria,
one in New Guinea, seven in New Zealand, three in South America, and one
in Rodriguez. 11aloragis erscta extends to Juan Fernandez. This type of
distribution is also shared by the South African genus Leptocarpus, which
has 11 species in Australia, one in New Zealand, and one in South America.
(2) BIRDS. There are three species of New Zealand birds which seem
referable to the category of endemic species differentiated from Australian
arrivals-namcly,
Coturnix noc~-:eulandie, Casarca variegata, and Anthus
nova?-zealandie. All are birds of the open country.
( C j Species identical in Australia and New Zealand : Plants.--The
species
of plants coinrnon t o Australia and New Zealand may be considered in
groups :iceording t,o their distribution beyond these two countries.
It is not cont,ended here that the explanation of the presence in New
Zealand of all the species mentioned in the following paragraphs is due to
their having migrated from Australia across the Tasnian Sea. But it is suggested, &at on account. of their existing as identJcal species in the t.wo regions
which have been separated during the greater portion of the Tertiary period,
and the general eastward movement of I)lants in the South Temperate ltegion,
the probability is that the bulk of them have made the passage overseas and
in the direction wcst to cast.
(1) Closmopolitali species are here defined as those which extend to the
continents of beth the Eastern and Western Hemisphcres. Besides 1 9 species
of pteridophytes tliere :ire in New %.dalld 6 1 flowering plants of‘ this
nature. They consist oE 24 grasses, d g e s and rushes, five composites, 1 2
coastal plants, 12 freshwater plants, : i d eight others (all herbs). It is
evident that all these specics owe their wide distribution to means of or
opportunities for dispersal. TO say that tlley are 111 or mainly olti species
one lllight infer from Willis’s ‘ Age and Area ’ hypothesis gives no satisfactory explanation. When indcpendent cvideiice of the age of these species
is forthcoming it should prove or disprove Willis’s theory.
130
MR. W. R . B. OLIVER ON TEE BTOQEOGRAPHICAL
(2) There are a number of species (T have listed 25 of flowering' plants)
which are distributed over Australia, Tasmania, portions of' Malaya, Polynesia,
and New Zealand, but do not extend to South America. Half of them are
grasses and sedges, the remainder herbaceous plants of the coast, fresh water,
scrub, or inoiintain. Forty species of pteridopliytes belong to this group,
including 7bdea barbaru, found only in South Africa, Australia, Tasmania,
and New Zealand.
(3) About 12 species of flowering plants and nine pteridophytes have
a southern distribution that would suggest Australia as a starting point,
distrihuting thence to New Zealand and South America. Thcy are of the
same general nature as those already mentioned, as far as ineans and opportnnities for dispersal are concerned.
They include Ualoragis erecta,
Myrioph$um elatinoides, Geraniicm sessil$foruni,Sellirra radicans, Gratiola
peruaiana, Carez p u n d a , and Jicncus planlifolius. Two species, Mesembrqyavrthemum cequilaterale and Tetrugonia e,Tpansa, have their distribution, in
addition to the countries mentioned, recorded as Chlifornia and Japnn
respectively, while three have :tpparentlp continued their easterly route
beyond South America, A b t e r a depressa reaching Tristnn d'Acunha, and
Scirpus nodosus and Ayium prostratuin South AErica.
(4) There are 135 species of flowering plants and 19 pteridophytes confined to Australia, Tasmania, and New Zealand. The flowering plants inay
he classed under the following systeina tic and ecological groups :--Grasses,
18 species ; sedges and rushes, 19; orchidp, 15 j composites, 9 ; coastal
plants, 11; swamp and other freshw:tter plants, 23 ; scrub plants, 20 (of
which five are woody) ; mountain plants, 14 (of which four :ire woody) ;
forest plants, six (five herbs and Pornaderris apetala). This list may be
compared with that of the Kerinadecs previously mentioned. It has, in fact,
the characteristics of the flora of an ocBeanic island. The inference is that
these plants might have crossed the Tasman Sea between Australia and New
Zealand by occasional means of trailsport,. A few of thein are qnite rare in
New Zealand.
The New Zealand orchids are mainly of Australian :iffinity. Of the genera,
14 (including 39 species) may be described as Australiiin, whilst seven
(including 17 species) are mainly Malayan. This distribution would suggest
that the family reached its highest development after the connection hetween
New Zealand and the North was severed, and the species found their way by
wind carriage ; hence the prepondorance of Australian forms.
The characteristics of the 320 plaiits coininon to New Zealand and
Australia have already been indicated. They are plants of the shore, lake,
swamp, scrub, grassland, and mountain. Taken as a whole they correspond
with the type of flora found on oceanic islands, and therefore do not require
the hypothesis of a land conricctioii with Australia to explain their presence
in New Zealund. I n each coiintry those large genera which stamp the flora
as distinct from that of other lands and form the bulk of the forest vegetation
181
RELATIONS 3F THE NEW ZEALAND REGION.
have few or no representatives co~ninonto the t8wo countries. The plants
whicli are coinmoit give no indication whatever of any niigration as a whole
flora from one country to another.
,BBirds.--There are three spwirs of land birds coininon to Australia and
New Zealand : Ninox now-zealatitliw, Rli@dura j?abelli/era, imd %o.rtevops
lateralis. Of these the last appears to have found its way froni Tasmania
since New Zealand wits bettled by Europeiins, it having bcen first noted in
the south-west of Otago in 1832. Of birds listed as stragglers-that is, those
which have been recorded in New Zedand from ii f e w specimens presunially
carried accidentally by storms-there are about 10 species, not counting petrels.
They include two land birds (Graucalzis robustiis and Coluia caruncxlata) and
seven species of rails and herons, which a r e iri1iul)itants of fresh water. To
these may be added three species of ducks. These last nine species are of
interest as indicating how seeds of water plants, grasses, and sedges niiglit
be transported. The Grey Duck (Anus superciliosu) wanders far ; it and
the Harrier (Circus approainians) are regular visitors to Sunday Island,
600 miles to the north-east of New Zeuland.
Species dispersing from South America to Tasinatiia and Australia.-That
there is B continuous stream of migrants originating in South America. and
moving eastwartls is a conclusion that seeins evident from the facts of distribution of' the plants of' the southern portions of' Chile and Patagonia. Moat
of them, one must assume, perish at sea, tnany reach the Falklands, fewer
South Georgia, still fewer the islirnds of the S0ut.h Indian Ocean, some
Tasmania, and more New Zealand and the islands to the south. It appears
to be a question of latitude and distance. Thus the islauds closest to South
America are most favoured by these Frregian plants, whilst of those farthest
anay the inore southern, as the South Island of New Zealand, receive niore
than those such as Tasinania, which lie sliglitly farther to the north. Certain
South American genera have their farthest, eastward riinge in Taurnuiiia--Eucryphia, Prionites, and B c m a section Acroliyssinoides. Others occur in
New Zealand as well. Such are Pernetlya (20 in South America, one in
Tasmania, one in New Zealand), Gentiana section Aizdicolu (50 in South
America, one in Tasmiinia, one in New Zealand), and Ureomnywliis.
Species dispevszny frona !.South America to ATew %ealawd.-Three E'uegian
species of flowering plants are, in the New Zealatid region, found only on
the islarids of the route and east. Rostkociu ,m!p&nica and Azorella Seluyo
have s h m II their route and origin by their occurrence oil intermediate islands,
but Cares Darwinii outside South America has been found only in the
Chatham Islands. Besides these a fern (I'olysticlmm uioliriodes) is found ;it
the Auckland Islands.
A class of plants not far removed from these so far us distribution is
concerned is that consisting of those species occurring in South America,
the Subantarctic Islands and mainland of New Zealand, and sometimes
LINN. J0URN.-BOTANY,
VOL. XLVlI.
L
132
YR. W. K. B. OLIVER ON THE BlOQEOGRdPHICAL
intermer1i:tte isliinds. Snch are Cardatnine glacialis, Tillca moschata,
Carex tr@de, a i d Iiununeitlus acaulis.
Just :is tliere :ire New Zealand genera of plants having representatives in
South America, so thore :ire South American genera or sections of genera
with species in New Zealiind. They are fewer, howerer, and more southern
in their New Zealanti distribution. Besides those already mentioned as
occurring also in Tasiiiania, there are Bnargea, Illarsippospermam, Cawx
seotions Bractioscc and Aciculares, Geranium sections Chilensia and Andha,
AcEiia section J‘:i~ancistrum,
and Cultha section Psyc*lirophylla. Colobanthus
is a genus whose species appear still t,o be dispersing both froin South
America and New Zealand. There are 10 species in Andine South Anierica,
tIie Palklnnds, South Georgia, and Gfiiham Land (C. crassifohs). Two of
these are found in New Zealand. There are nine species in New Zealand,
one in Kerguelen, another in New Amsterdam Island, and a third in Victoria
and Tasmania.
Species dispersing from South A#frica.--It is not here contended that any
species have arrived in New Zealand direct from South Africa, but there are
several genera which have their centre of dispersal in South Africa and
which extend castward through Australia to New Zealand. Such are
Leptocarpus, Hypolam, Hyp&s, Wuhlenberyin, Tetragonia, and Mesembryant henzum.
Species dispersing from Polynesia to A’ew Zealand.-A few species of New
Zealand plants may bo r e p r d e d as having been derived from Polynesia by
tra1ls-oceaiiic migration. DianeLla intermedia, 1’q>eromia Urvilliana, Macropiper excelsum, Yisottia Jhwioniana, and the ferns IIipZazium japonieum,
Hymenophyhin deniisstcm, 11. dilatatunc, and jhyopturis Il’helypteris are
suggested :is coining under the prusent heading, while the orchids of the
genera Earina, I)eniirobium, Bulbophyllum, Gastrodia, and Corysanthes inny
be descendants of f’olynesian inimigrants. Ascu&za is a genus of forest trees
tiiat seems capable of crossing wide expanses of ocean. There are three
species in New Cakedoniii and one each in the Philippines, Fiji, Samoa,
‘I’ahiti, Ilaratonga, Kermadecs, and New Zenlancl. The species from the
Rerinadecs, Fiji, sit tiloit, and liaratonga. are wry closely allied, and related
to A . lucida of New Zealand.
S1)ecies d i s p r s i n ~,from N&u Zealand to Australia und ITasnzania.-’l!he
number of genera having many representatives in New Zealand and with
few species in Australia atid Tasmania is small compared \\ ith those large
Austra1i;in genera which Ililve one or few species in New Zealand. Furthermore, in the case of tlie Net\ Zealanti genera lierein mentioned, the species
in Tasmania and Australia are all endemic, showing that transport is quite
occasional, if indeed it was not confined to the period when the New
Zealand continent extended towards Tasmania. The following New Zealand
RELATIOKS OF THE NEW ZEALAND KEGION.
133
genera are considered to have distributed species overseas to westward :
Gunnera subgenus J4illigania (nine species in New Zealand, one in Tasmania) ;
Axorella, sect. Scliizeleinza (nine in New Zealand, one in Australia, two in
South America); Ourisia (29 in South America, eight in New Zealand, one
in Tasmanitt): Acipliylla (14 in New Zealand, four in Tasmania and Australia) ; CeZntisia (43 in New Zealand, one in S.E. Australia) ; Psychrop1uhyton
(seven in New Zealand, one in Tasmania); IIebt (90 in New Zealand, two
in Tasmania and Eastern Australia). In nearly all the above-mentioned
cases the species are mountain plants, :ind the range on the western side of
the Tasnian Sea is Tasmania and South- East Auqtralia.
Species dispersiriy ,from N e w Zealand to Po[ipesica.--There are a few genera
of Angiosperms whose centre of dispersal is apparently New Zealand arid
whicli have one or two species in the Pacific Islands. A ste2ia has one species
in Piji and Samoa, two in the Hawaiian Islands, and one i n Tahiti. There
R
i a specie3 of Oreohlus recorded from the Hawaiian Islands. Coprosma
occurs in several islands of the Pacific, as far as Hawaii and Tahiti. Nelicytzls raiiiifEoriis is found in Norfolk Island, the Kermadecs, Eua (Tonga
Group), and Fiji. Coriariu ruscyolia has reached the Kermadecs, Banks
Islands, Fiji, Sanioa, Tahiti, and South America ; whilst Edwardsia microplylla occurs in Easter Island, J u a n Fernandez, South America, and Gough
Island. There are also three species of ferns widely distributed in New
Zealand but only known elsewhere from one island in the Pacific ; they
are Lomaria jilifoi-mis and Polysticfmn Ricliardi in Fiji and Polypoikium
dictyopteris in New Hebrides.
Xpeccies dispersing f r o m hTew Zealarid to Soutli America.--In considering
the origin of the New Zedand element in the South American flora we may
omit the ferns, and about 13 otliors which are cosinopolitan in their
distribution, but take into accountt those species in South A inerica whicli
are closely allied to New Zealand species. The feature of these plants is
that most of them belong to genera whose centre of dispersal is apparently
Zealand or Australia. And this fact, together with the oceanic chtmwter
of the species, taken a s a whole ruggests trans-oceanic migration as the
explanation of their present distribution.
Among Australian genera with spcios in South America identical or
allied to species in New Ze:iland the following rnay he nientioried :Ahrotanella, the three South American species are rclated to New Zeuland
Ones ; Gaimardia, belonging to :m Australian f:imily, has one species in
Tasmania, one in New Guinea-G. setucea in New %ealaiitl and G. uustrulis
(allied to G. setucea) in Fuegia :md the Fdklands. Lugertophora, iiiainly
Australian and New Zealand, has four species in Andine South Anierica,
one of which is closely related to Ti. p o d u of New Zealand ; I'ratia repens
in Chile, Fuegi:,, and the Fallilands is related to the New Zealand Y.angu&.jc ;
L2
134
M R . W. R. B. OLIVER ON THE BIOGEOGRAPHICAL
Haloru& erecta of Australia a i d New Zealmd extends to Juan Fernandez;
Leptocurpus has one species in South America ; Urosera section Ysychro.
phyllu has oiie species in Tasm:mia, Australia, and New Zealand, another in
New Zealand, and a third in Chile, Fuegia, and the Falklands.
The genera or sections of genera which may bu considered of New Zealand
origin and having oiie or a few species in s o u t h America are as follows :Hebe, with about 90 species in New Zealand, has H.elliptica in the south of
South America, Subantiiretic Islands, Fuegia, Chile, and the Pallilands ;
whilst H. saZic[foZia of New Zealand has a n ally (H. Fonckii) in South
America. Coprosma has one species (11.trifEora) in Juan Pernandez;
Myosotis is a North Tairiperate genus with 24 species in New Zealand, of
which one extends to Patagonia. and there is another, related to a New
Zealand species, in Magellan ; Bstelia, whose centre of dispersal is New
Zealand, has one species in Fuegin and the Falklands ( A . punzila) related
to A . linearis of New Zealand; Scliizeleima (section), with nine species in New
FIG.6.
a
Herbs
Grasses 8
Orchids
Trees 8
Composites
Shrubs
ijiiiiil
......
Pteridophykj
Sedges
Explanation of signs used in Figs. 1-6.
Zealand and one in Australia, has two in South America ; some srnall genera
with one or few species in New Zealand and one in South America should
Donatia.
probably be classed here- Tetrachondru, Yl~~ylLachne,
Therv are a few genera which I class as of Malayan origin, but New Zealand
is probably the centre from which the South Ainericari species have been
derived. They are Aristotelia, Nertera, Coriaria, and E d w a r d s k .
Enough has now been given to show that the New Zealand region ailpears
to have been x centre of dispersal for lnahy species that reach as far eastward as South America, and that in their characteristics as regards 111eansof
dispersal and occupying h:\bitzlts giving opportunities for dispersal and
ektablishinent, the plants common to New Zealand and Ponth America
compare with those found on an oceanic island. The coiiclusion seems to
be inevitable that plants have been carried from New Zealand to South
America hy agencies comparable to those which populate the reniote islands
RELATIONS OF THE NEW- ZEALAND REGION.
135
of Polynesia. Exactly what these agents are may require long and close
observations in inhospitable climates.
Circurn-austral Species.-As if showing that distribution is now actually
taking place, there are several circum-austral species which, beginning in one
of the southern land masses, have completed the circuit of the glohe. Such
are iVertera depressa, Raniinculus biternatus, W l @ a moscliatu, Callitriche
antarctica, Festuca erecta, Edwardsia micropl~!jlla,Scirpus iiodosus, Apirim
prostratum, Agrostis mafyellasica, Crantzia lineatu; Oreornyrrhis antlicola,
and Acsna adscendens.
VI. SUMMAHY.
1. Thrre has been a continuous land surface in the New Zealand area
since the beginning of the Mesozoic epoch. ‘I’lie early Triassic period
appears to h w e been a time when there was a direct land connection with
north. I n the late Triassic a flora consisting of Equisctale~,FilicnIee,
iind Ginkgoales was conirno~i to New Zriiland and other southern lands.
This florii, together with S’henoilon, Liopellma, and possibly Peripatus,
would date from the time of the former northern lard connection.
2. I n the later Triassic and in the Jurassic periods there were tiines when
the land was lower and a long series of niariiie sedinients was laid down.
Throughout this time there flourished on the land surface a fiora including
FilicaIes, Cycadofilices, Podozamites, and the early forms of Coniferdes.
T]le Jurassic flora was nearly uniform in character tlirougllout the world,
extending beyond the Arctic and Antarctic circles.
3. In Cretaceous times the land in tlie New Zealand area was of ~011tinental dimensions. It extended to the north SO as to connect with New
Guinea and North-Eastern Australia, but Western Australia was sepdfijted
frorll tliis continent by an arm of the sea. A sea also intervened between
the New Zealand area and Southern Australia (with Tasniania), but the lalid
extelided to the south and east SO as to include the area of the submarine
plateau on which now stands the Auckland, Campbell, Antipodes, anti
Chatham Islands. The Antarctic continent during this period of elevation
no doubt extended farther to the north, approaching perhaps within a few
Ilundred miles of the New Zealand continent. Possibly Macquarie Island
w;ls much larger than a t present. Mr. H. Hamilton informs me th:lt it
contains altered sedinientary rocks of unknown age. According to Thoinson
(Hept. Austr. Ant. Exp., ZOO^. V O ~ .iv. p. 60, 1‘318), the late 3ur:tssic or early
Cretaceous was a period of emergent lands all round the Pacific.
The map given by Hedley (Proc. Linn. SOC.
N.S.W. vo). xxiv. p. 404,18$!9)
represents precisely what, judging from t h present New 2eahnd flora and
the evidence gradually accumulating of the flora of New Guinea and tile
islalids of the western Pacific, I think necessary for a land connection in
the cretaceous period. The route taken by plants and animals 1nigr:tting
136
MR. W. R. B. OLIVER ON THE BIOGEOGRAPBIOAL
botwcen New Zeltland and the north is as clearly marked by the present
distribution of Apthia ant1 Arauearia as by that of Placostylus. The
accoinpanying m p , therefore, shows a former land bridge by way of
Lord Howe Island, New Caledonia, and the New Hebrides, but I cannot
follow Mr. Hedlcy in hi,.. Aiitarctic conncctions to the southward.
ThH Crctaceous period was important in the history of New Zealnnd, which
afterwards was not again united with any other land. The period of land
FIG.7.
I
>lap of South-wwt Pacific showing greatest extension of land required
for the dispersal of Ypermophytes in late Mesozoic times.
connection with the north must have lasted sonic t h e , for two continental
fioras succeeded one another in the New Zealand area. The first comprised
the modern types of (:oniferalrs aiitl Filicalea, and the niorc priiilitive
Angiospemns such as l~Totli+pis. Such geiiera as Avaucaria, Libocedms,
Phylloclatlics, and Nothofagus appear to 1i:ivtA arisen in North America and
migrated along the western diore of the Pacific; hence their presence in
Australia anti New Zealand hnt :ibsencc from Africa. The second flora
includttd the bulk OF the ancestors ot' the Malayan eleinent in the present
flora. 1 t included Angiosperms, and with it were associated birds, lizards,
insdcts, and other animals.
RELATIONS OF THE N E W ZEALAND REGION.
137
Overseas came many animals and plants, some from Australia and
Tasmania across the Tasinan Sea, and a fcw from the shores of the Antaretic
continent, which supported vegetation.
The New Zealand continent not oniy received but, gave to neigkbouring
lands some of its productions. It W R R a centre for the developiuent of many
peculiar groups of plants and animals. Shut off froin iiiainin:tl~which spread
over the world in late Mesozoic and early Tertiary times, its birds tillcd their
place, and a great variety of flightless forins--Dinortiithitlae, dpteryz, rails
-originated.
I n the same diversified and extensive lend area the plallts
likewise increased and diff erentiatcd along linrs adapted to different stations.
Hence arose the many species o f ' Hebe, Coprosma, Celnaisia, Olearia.,
Camtichaelia, and ot.liers. I t was i n the southcrn portion of the continental
area that this new world of life (:mie iot,o existence. Some of these forins
wandered back along the land bridge to the north, as Curmiclwliu and
Phormiuin ; a few found their way to the southern part o f Australia and
Tasmania, as Aciphylylla, Celnzisiu, and PsyclLropliyton ; while some even
reachod the shores of Antarctica, as P s d o p u n u x aud perhaps I hcrydiwm.
4. On the breaking down of the land connection to the north the exchange
of species between New Zealiind and other countries was coufined to such BY
could by chance cross a corifiidorable stretch of ocean. Nevertholess, a great
many species of +its both arrived aiid departed from New Ze:tland, the'lands
both receiving and giving being mainly those in the same latitudes -Australin,
Tasmania, and South Anlericii. A small north :tnd south inovcrneiit between
POlyIIBSiit and New Zealand also took place. But in accordance with the
means of and opportunities for dispersal this moving population has the
charyteristics of the inhabitants of truly ocwnic islands. The priiicipal
sections of the flora received since New Zealand severed its last, direct land
connection are the orchids and the Aiistralian species.
The flora as it stands today I have endeavoured to represent by means of
a diagram (fig. 1 ). I t s derivation.for the most part by direct land connection
in the north, gives its forests which have nearly half o € their species woody
plants (sonie trees and shrubs are included in the Composites), and also
n considerable proportion o f woody species in the scrub and grassland
formations.
Continental conditions including diversified mountainous
country are shown by the mountain pliints equalling the forest plants
(30 per cent. each). From Ur. Cockape's work on the vegetation of New
Zealand, I gather that he considers that the mountain plants wcre mainly
differentiated in late Terti:iry tiiries. This may I Jso,
~ but. large distinct
genera evidently require a longer period for their differentiation.
The paucity of orchids, so abundantly developed in New Guinea ant1 New
Caledonia, leads one to conclude that this t'amilyireached its highest development after New Zealand's connection with the north had been severed.
Composites, which figure so largely in the New Zealand flora (14 per cent.),
138
MIX. W. R. B. OLIVER ON THE BIOGEOGRAPHLCAL
are mainly plants of the scrub and grassland areas. Their great development
is perhaps a result of continental conditions in both early and late Tertiary
times.
Tn the preparation of this pqwr
I hnvo endo;ivonrcd to group the plants arid
some ofthe animals of’ New Zealatid according to their place of‘ origin. For
the facts of plant distribution I atit especially iudeltted to the works of
Cheeseman, Cockayne, and Skotdmrg. Tile main groups of nnirnals not
tie:ilt with-earthworms, insects, sl)idt)rs, a n d crustacea-have all been used
to support the theory of :in Aiititrctic connection in late Mesozoic or early
Tertiary times. But opinion is not unanimous on t,his point, and I ventnre
to predict that, as methods o f tlisperssl itlnong the invertfibrates ure hetter
known, the arguments for trans-oceanic migration will he drenghhened,
VII.
IATERATURE.
(Titles of the principal recent publications bearing on the geographical distribution of
the flora and fauna of New Zealand.)
ilfesozoir: Floras.
HALLE
T.~G.-Nesozoic Plorn OP rirahamlnnd. Bchwed. Sudp. Exp. nd. iii, If. 14, 1913.
ARRRR,E. A. N.--Earlier Mesozoic Flora6 of New Zealaud. N.Z. Geol. Surv. Pal. Bell.
No. 6, 1917.
WALKOM,
.4. 13.-Geology of Lower Mesozoic Rocks of Queensland. Proc. Linn. 8 0 ~
N.S.W. vol. xliii. pp. 37-116, 1S18.
WALKOM,
A. B.-Queei~sland l(:umil Plorr~s. I’ruc. Itoy. SOC.Q.vol. xxxi. pp. 1-20, 1920.
Il’eriimy Floras.
I)TJRRN,P --Tertiare Flora du Seymour Insul. Schwed. Sudp. Exp. Bd. iii. If. 3, 1908.
ETTINQYHAUSEN,
C.-Fossil Flora of New Zealand. Trans. N.Z.Inst. vol. xxiii. pp. 237310,1891.
EyrINc,wrAiism, C.-Tertiary Flora of Australia. Pal. IIeni. Deol. Surv. N.S.W. NO.2,
1888.
QeoTogicnl Histoi y of New Zealcind.
I’ARK,J.-The Urology of New Zenland. 1910.
I~ARBHALL,
P.-Kew Zealaiid and Adjacent Islands. 1912.
NOROAN,
1’. G.-Geology of Kew Zealaud. N.Z. Journ. Sci. Tech. vol. v. pp. 4667, 1922.
RICNBON,W . N. -1tecent Advances ill New Zealand Geology. lbpt. A. A. A. S. vol. xv
pp. 46-133,1922.
Bp;NBON, w. N.-~’alm~zoic rtud -1leaozuic Seas in Australasia. Trans. N.Z. Iust. VOI. iiv.
pp. 1-62,1923.
I’ARK,J.-Birth and Devrlop~nentof New Zealand. ’I%ans. N.2. Inst. vol. fiii. pp. 73-76,
1921.
Biqqtographitd 1ieZatiom- ffmiaal.
WALLACE,
A. R.--Island I.ife. 1880.
MATTHEW,W. D.-Climate and Evolution. Ann. X . Y . Accsd. Sci.TOI.
xxiv. pp. 171-318,
1915.
RELATIONS OF THE NEW ZEALAND REGION.
130
GUPPY, H. B.-Plants, Seeds, and Currents in the West Indies and Azores. 1917.
WILLIS,J. C.-hge and Area. 1933.
DIJNN,E. R.-Geographical Distribution of Amphibians. hmer. Nat. vol. lvii. pp. 129
136,1323.
Malaya a i d Polyneaiu.
J.-Xorphology,
Chsdication, and Zoogeography of Cndian Oligochseta
Proc. Zool. Soc. 1921, pp. 103-142.
XIERRILL, E. D.-Distribution of Uipterucaryacea. Phil. Journ. Sci. vol. xxiii. pp. 133,
1923.
GIRRS, I,. S.-Pliyto~eogr~~pliyarid Flora of the hrfak Mountains. 1917.
HEULEY,
C.-A Zoogeographic Scheme for the Mid I’acitic. Proc. Linn. Soc. N.S.W.
vol. xxiv. pp, 391-417, 1899.
GUPPY,11. ~.-obsarvations of H. Nrtturaliat in the Pacific. 11. Plant Disperaal. 1m.
STEPHENSON,
Australirc a d TusPnnlbia.
Z~BDLEY, C.--The Xarine Fauiia of Queensland. Ilept. ,%.A.A. Y. vol. xii. p. 329-373,
1910.
HEDLRY,C.--’allieogeo;raphical
Relations of hutarctica. I’roc. Linn. SOC.
(Lond.) pp. 8090, 1912.
SPENCER,W. B.-Faunti and Zoologieal Eelationships of Tasmania. Hept. A. A. A. S.
vol. iv. pp. 82-124, 1895.
SMITH,GI.-A Naturalist in Tnsmania. 1909.
GIBBS,
L. S.-Phytogeogrctphy and Flora of Meuntaiu Summit Plateaux of Tasmania.
Journ. Ecol. vol. viii. pp. 1-17, 89-117, 1920.
New Zealand.
IIUTTON,
F. W.-lndex Fauna Nova Zealandie, Introduction. l’p. 1-23, 1904.
CHILTON,C.--I+iological lteliltions of the Subantarctic lslands of New Zedand. Subant.
Is., N.Z. vol. ii. pp. 733407, 1909.
C‘HEEBBMAN, T. I~.-Spternatic Hatany of the I’laats to the South of New Zealand.
Subwt. Is., N.Z. vol. ii. pp. 389-471, 1909.
CHEESEMAN,
T. F.--The Vascular Flora of Macquarie Island. Itept. h u s k Ant. Exp., Bot.
vol. vii. pt. 3, 1919.
SKOTTSBRRO,
C.-Relations between the Floras of Subantarctic America and New Zealand.
Plant World, vol. xviii. pp. 130-142, 1915.
COCKAYNR,
L.--The Veptation of New Zealend. 1921.
Reyt. Austr. Ant. Exp., Zool. vol. iv. pt. 3, 1918.
THOMEON,
J. A.-Brachiopoda.
BENIIAM,W. B.-Oligochseta of Macquarie Islmd. Rept. Anstr. Ant. Exp., Zool. vol. Ti.
pt. 4, 1932.
140
BIOGEOGRAPHICAL RELATIONS OF THE NEW ZEALAND REGION.
1
No. of Species.
____
ji
I
Percentages.
I
.II
..
Trees and shrubs ....... 25
Herbs. ................. 56
4
6s
..
Orchids ............... .I
l'teridophytes ..........
I
70
.I
Cornposites ........... 18
Grasses, sedges, and rus1iea.j 22
5
8
!;I24
-3150
-.
I--
New Zeulaiid and Stewurt Isluqd.
Fa 183
.. 4 13 6 a5
92 A9
5
4 12 31
6
4
2
7 14
48 I a4
4 I6
4 2
4
* . 1 2 .. I 3
2
1 I 11
7
1
- 21 30 30 'loo
,
..
Lord Howe Isiund.
I
Trees and shrubs ....... ' I 8
Herbs., . . . . . . . . . . . . . . . . 24
Compo~iites . . . . . . . . . . . . 3
Grasses, hedges, and rushea. 3
Orchids . . . . . . . . . . . . . . . . .
Pteridophptes . . . . . . . . . . I
39
- -.
Trees :md shrubs . . . . . . . . lj
Ilrrbs.. . . . . . . . . . . . . . . . 17
Composites . . . . .
4
t;riLsws, sedges, mid
Orchids . . . . . . . . . . . . . . , I . .
l'teridophytes . . . . . . . . . ,I 2
1
14
r
.,
1
~~es.1
1 1
2 ;
1 I ,.
24 . .
a
1
Auctrdici-New
Zeulund.
..I
Trees and shrubs . . . . . . . . 2
Herbs.. ..............
29
Trees and shrubs . . . . . . . 1 I
Herbs . . . . . . . . . . . . . . . . . 8
Composites . . . . . . . . . . . . . .
Grasses, sedges, and rushes. 2 j
3rchids . . . . . . . . . . . . . . . . . . ,
Pteridophytes . . . . . . . . . .
I
...
10 .
I
8
3
4
1
3
..
1
'
~
....
13
~
. .
..
1
1.7 I70
17 28
16
84
9
IR
4
33
02