GODLEY: VEGETATION OF AUCKLAND Is. 57

REFERENCES KENNY, R., and HAYMAN, N., 1962. Ecology of rocky shore organisms at Mflcquarie Island. Paci/. Sci. 16: BRODIE, J. VV., and DAWSON, E. W. (in press). Mor- 245-263. phology of North Macquarie Ridge, Nature. KNOX, G. A., 1960. Littoral ecology and biogeography BURLING, R. W., 1961. Hydrology of circumpolar waters of the southern oceans. Proc. Roy. Soc. Land. B, south of . N.Z. D.S.I.R. Bull. 143; t52, 577-624. 1-66 (N.Z. Oceanagr. Inst. Mem. 10). CLARK, A. M., 1962. Asteroidea. B.A.N.Z.A.R. Exped. KNOX, G. A., 1963. The biogeography and intertidal Rep. (B) 9, t-t04. ecology of the Australasian coasts. Oceanogr. !lIar. D WSON, E. W., 1963. Oceanography in the Sub- BioI. Ann. Rev. 1: 341--404. antarctic. Antarctic 3; 312-314. McKNIGHT, D. G. (in press). An outline distribution of n\'\VSON, E. W., 1964. Antarctic oceanography. Antarc- the New Zealand Shelf fauna; the Echinoidea. tic 3;430-432. N.z. Oceanogr. Imt. Mem. 37. DAWSON, E. W. (in press). The marine benthic fauna McKNIGHT, D. G. (in prep.). An outline distribution of Foveaux Strait, Southern New Zealand. N.Z. of the New Zealand Shelf fauna: the Asteroidea. Oceanogr. Inst. Mem. 34. N.Z. Oceanogr. Inst. Mem. 44. DELL, R. K., 1952. Marine biology. In The Antarctic MAWSON, D., 1943. . Its geography Today, ed. Simpson, pp. 129-150. Reed, Wellington. and geology. Austral. Antarct. Exped. 1911-14. Sci. DELL, R. K., 1956. The archibenthal of New Rep. (A) 5, t-t94. Zealand. Bull. Dam. Mus. 1S: 1-235. MORTENSEN,TH., 1921. Echinoderms of New Zealand DELL, R. K., 1962. New Zealand marine provinces~do and the Auckland and Campbell Islands. Part 1. they exist? Tuatara 10; 43-52. Echinoidea. Vidensk. Medd. dansk. naturh. Foren. DELL, R. K., 1963a. The littoral marine mollusca of 73, 139-t98. the Snares Islands. Rec. Dam. Mus. 4: 221-229. MORTENSEN, TH., 1925. Echinoderms of New Zeala~d DELL, R. K., 1963b. Archibenthal mollusca from nor- and the Auckland and Campbell Islands. Parts III thern New Zealand. Trans. Roy. Soc. N.Z. Zool. -V. Asteroidea, Holothuroidea and Crinoidea. 3, 205-216. Vidensk. Medd. dansk. naturh. Foren. 79: 261--420. DELL, R. K., 1964. Marine mollusca from Macquarie MORTENSEN, TH., 1926. Goniocidaris umbraculum, a and Heard Islands. Rec. Dam. Mus. 4: 267-301. brood-pl'otecting species. N.z.J. Sci. Tech. S: 192. FELL, H. B., 1953. Echinoderms from the OUVER, R. L., FINLAY, H. J., and FLEMING, C. A., 1950. islands of New Zealand: Asteroidea, Ophiuroidea The geology of Campbell Island. N.Z. D.S.l.R. Cape and Echinoidea. Rec. Dom. Mus. 2: 73-111. Exped. Ser. Bull. 3: 1-62. FELL, H. B., 1962a. A dangerou3 sea-urchin. Tuatara 9, 84. PARROTT,A. W., 1958. Fishes from the Auckland and FELL, H. B., 1962b. West-wind drift dispersal of Campbell Islands. Rec. Dam. Mus. 3: 109-119. echinoderms in the Southern Hemisphere. Nature PAWSON, D. i L., 1964. A new Cidaroid from New Zea- t93, 759-761. land waters. Trans. Roy. Soc. N.z. Zool. 5: 67-70. FELL, H. B., 1962c. Native Sea Stars. Reed, Wellington. PAWSON, D. L. (in press). Echinozoa from the New FINLAY, H. J., 1925. Some modern concepts applied Zealand Subantartic, Macquarie Island, and the to the study of Cainozoic Mollusca from New ZeQ- Chatham Rise. N.Z. Oceanogr. Inst. Mem. 41. land. Verbeek Mem. Birthday Vol.: 161-172. POWELL, A. W. B., 1951. Antarctic and Subantarctic FLEMING, C. A., REED, J. J., and HARRIS, W. F., 1953. Mollusca. "Discovery" Rep. 15: 151-222. The geology of the Snares Islands. N.z. D.S.I.R. POWELL, A. W. B., 1955. Mollusca of the Southern Cape Exped. Ser. Bull. 13;1-42. Islands of New Zealand. N.Z. D.S.I.R. Cape Exped. GARNER, D. M., 1959. The sub-tropical convergence in Sa. Bull. 15: 1-152. New Zealand surface waters. N.Z.J. Geol. Geophys. POWELL, A. W. B., 1957. Mollusca of Kerguelen and 2, 315-337. Macquarie Islands. BA.N.z.A.R. Exped. Rep. (B) HOUTMAN, TH. J. (in press). Water masses and fronts 6, 107-t50. in the , south of New Zealand. POWELL, A. W. B., 1961. New Zealand biotic provinces. N.Z. Oceanogr. Inst. Mem. 36. Tuatara 9: 1-S.

NOTES ON THE VEGETATION OF THE E. J. GODLEY Botany Division, D.S.l.R., Christchurch

The following observations on the vegetation ALTITUDINAL DISTRIBUTION of the Auckland Islands, made between 26 December 1962 and 20 January 1963, are Hooker (1847: 2) made the following impor- additional to or amplify those of Hooker tant note on the vegetation in the north of (1847), Cockayne (1904, 1909) and Moar Auckland 1. "It is especially towards the sum- (1958a, b). mits of these hills that the most striking 58 ECO'LOGY OF SUBANTARCTIC ISLANDS are found, vying in brightness of colour with speciosum, Polystichum cystostegia, Ranunculus the Arctic Flora, and unrivalled in beauty by pinguis, Rostkovia magellanica, and Schizeilema those of any other Antarctic country. Such reniforme. (A larger list could be made of are the species of Gentian, and a Veronica species confined to lower altitudes.) with flowers of the intensest blue, several magnificent Compositae, a Ranu"nculus, a As most visitors to the Auckland Is. do not Phyllachne and a Liliaceous whose dense go far from the coast these striking endemics spikes of golden flowers are often so abundant are seldom seen and it is natural that their as to attract the eye from a considerable exceptional occurrence at a low altitude in distance. " an easily accessible position should become noted. Such a situation occurs at "Fairchild's Cockayne (1904) accepted this statement Garden", on Adams I. at the western entrance but in 1909 wrote as follows: "Although there to Carnley Harbour, described by Chapman is a well-marked zonal distribution of the vege- (1891) as follows: tation~ the majority of the species may occur at any altitude, from sea-level to the highest "It extends from the strait at the north \vest summits." And he italicised the following: end of the island along the shore to the first "It is the formations-i.e. the distinct physio- piece of bush, and thence up to and over the gnomic combinations-which mark the zones, summit of the hill-in all perhaps 400 acres- and not the occurrence of special species for one of the most wonderful natural gardens the the first time." He mentions four exceptions extra-tropical world can show." This area to this statement~ namely Grammitis pumila, of open ground can be clearly seen in Falla's Plantago aucklandica and M arsippospermum figure 12 (1948). Presumably because of the gracilis at higher altitudes, and Ranunculus exposed situation inimical to tree growth, the aucklandicus near the shore. My observations upland species have here been able to descend confirm those of Hooker as the following field to sea-level. noteS show. These points are reinforced by a perusal of the list of plants collected by Moar (1958 b). "28.12.62. We went from Ranui Cove via Meg's Hill as far as the base of Mount Eden. During this Such striking endemics as Anisotome antipoda, time we did not see Bulbinella rossii, Celmisia Celmisia vernicosa, PleurophyZZum criniferum, vernicosa or any species." P. speciosum and Hebe benthamii were not "29.12.62. Suddenly, as we climbed above the recorded from the northern coastal areas visited, scrub, we came upon many of the species that we had been seeking. We entered a New Zealand sub- and only one plant of Bulbinella rossii was antarctic species zone. Below this most of the plants observed. All these species were recorded how- had been mainland species, but now we found Hebe ever from "Adams Island", and the actual benthamii, Celmisia vernicosa and Abrotanella locality was Fairc.hild's Garden, as shown on spathulata. However not until we reached the extensive rock outcrop which forms the summit of the map given by Leamy and Blakemore Mt. Eden did we at last see Bulbinella rossii, (1960). Pleurophyllum hookeri, P. criniferum and many other sub-antarctic endemics." The northern slopes of Adams 1. should be examined further. to see whether the situation In my experience, there exists in the north at Fairchild's Garden is exceptional. The other of the Auckland Is. (if not throughout the striking occurrence at sea~level- that of whole group) a distinctive group of species BulbineZZa rossij on Enderby- I.-is discussed charac.teristic of or confined to the higher later. altitudes, and these include many of the endemics of the New Zealand subantarctic LOWLAND SCRUB AND Oreobolus LANES islands. Lowland scrub quickly replaces coastal rata I place at least 18 species in this category: forest, which is usually no more than 50 yards Abrotanella spathulata, Anisotome anlipoda, in depth even on the valley floors at the heads Bulbinella rossii, Cardamine subcarnosa, Carpha of sheltered inlets. On the steep sides of tbe alpina, Celmisia vernicosa, Gaimardia pallida V-shaped valleys scrub is at its densest (Fig. (a new record for the Auckland Is.), Geum 1) but on the long tongue of gradually rising parviflorum var. albiflorum, Hebe benthamii, land stretching inland from Ranui Cove Helichrysum bellidioides, Marsippospermum towards Mt. Eden, the vegetation is a combina- gracilis, Plantago aucklandicus, P. hooken', P. tion of two entirely different communities. GODLEY: VEGETATION OF AUCKLAND Is. 59

Coastal Rata Forest Scrub Chionochloa Tussock Grassland Fell-field

FIGURE 1. Zonation of vegetation in the Chambres Inlet-Stony Peak sector, northern .

(1) The scrub occurs in patches or in scattered about, approximately 10 in. below the roughly parallel lines, often in an east-west present peat surface. Specimens are not yet direction. The predominant species are identified. Metrosideros umbellata, longi- folium and Myrsine divaric'ata, while UPLAND VEGETATION foetidissima, Cassinia vauvilliersii, Hebe odora Hitherto the only investigations at high and Neopanax simplex are less impcrtant. In levels have been those of Cockayne above many places the pattern of the scrub has been Camp Cove, Carnley Harbour in the south. determined by colonisation of sheltered hollows Here he described a Pleurophyllum hookeri but patches of scrub also occur on convexities association in "sopping wet through frequently in the ground. stony ground." In the north, between Bivouac (2) Within the scrub are extensive areas and Bleak Hill (4t miles) the peat and its of open ground in the form of lanes, covered attendant grassland reaches to about 1600 ft. with hard cushions of Oreobolus peetinatus in Above this is rock, or a brown mineral soil which grow Cyathodes empetrifolia, scattered interspersed with rock fragments, extending to gentians, orchids, and Sehizaea fistulosa. the western cliffs and the highest peaks. The Scattered tussocks of Chionoehloa antarctica following features were noted. and stunted plants of the eight scrub species (a) Fell-field supporting a diversity of mentioned earlier also occur. Most of these species is predominant, with Pleurophyllum species, whether in the scrub or in the hookeri one of the most prominent plants. Oreobolus lanes, are restricted to lower alti- (b) On the summit of Bivouac, and half a tudes. mile north of Bleak Hill are fiat poorly drained areas dominated by Marsippospermum gracilis. Viewed from the western high country a (c) On the north side of Mt. Easton is a similar pattern of vegetation was seen to the striking deep-green Gaimardia pallida bog. south on both Twin Peaks and Musgraves Knob (d) Carpha alpina predominates on some 15 peninsulas. Cockayne (1909) described a acres of the north slope of Mt. Easton. similar situation above Carnley Harbour and suggested that the rows of woody plants result THE Ross HARBOUR ISLANDS from the invasion of grassland by the tree The islands of Rose, Enderby and Ewing, species, and that the pattern results from mutual lying so close together, yet with such interesting protection under the influence of strong pre- differences in vegetation, provide valuable vailing winds. Du Rietz (in Oliver 1927) sug- opportunities for studying the effects of intro- gested however that the lanes were caused by duced animals and man on the original vege- local bog conditions. Future expeditions should tation. check whether or not the trend is towards an expansion of the area under scrub. Rose Island (Fig. 2) A considerable amount of fossil wood is In my opinion the vegetation on Rose is exposed in eroded Chionochloa tussock-grassland the least modified of these islands. On the at 1450 It. between Mt. Eden and Cloudy Peak. south side is scrub and rata, forest, an eastern Stems or branches about 2 in. in diameter lie portion of which has been almost destroyed. 60 ECOLOGY OF SUBANTARCTIC ISLANDS

The greater part of the northern half of the in which breviglumis and Agrostis island is tussock-grassland with a magellanica are prominent. sward grassland in the centre. Hitherto the I suggest that the west and much of the only extensive grassland of Poa litorosa north of Enderby 1. was once under Poa recorded for the Auckland Is. is on Disappoint- litorosa grassland (as can still be seen on Rose ment 1. (Cockayne 1909). Moar (1958 b: 1.) and that this has been modified to either 471), who did not visit Rose, was told that a grassland sward or Bulbinella. The Bulbinella Bulbinella was "abundant" there, but I could could have been derived from scattered plants find none. along the cliff edges, and have invaded tussock, depleted by rabbits, cattle and burning. The facts that Bulbinella tends to die back to the ground in winter, and has numerous fleshy roots, make it less susceptible to the effects of burning than other plants, which it may gradually replace.

FIGURE 3. : FIGURE 2. Rose Island: general disposition of of vegetation. vegetation. Ewing Island (Fig. 4) Falla (1948) referred to the "dense Olearia Enderby Island (Fig. 3) forest which surrounds a core of rata forest" Cockayne (1904) has left the following use- on Ewing 1. The situation in 1964 is shown ful records: in Fig. 4, based on an aerial photograph. (a) "On Enderby Island there is a consider- able quantity of rata forest and scrub which forms a somewhat narrow belt on the east and south sides, the whole of the west side of the island, according to Captain J. nollons, being a tussock formation." (P. 244). (b) "Proceeding through the forest to the north, the Metrosideros becomes less in quantity and more stunted, while Cassinia vauvilliersii makes its appearance, this shrub finally becom- ing dominant." (P.244). (c) ".. . on Enderby Island rata gives place on the windward side to Cassinia scrub, and this in its turn to tussock." (P. 254). The rata belt and the Cassinia scrub still exist, but the remaining area is mainly covered by almost pure stands of Bulbinella rossii on FIGURE 4. Ewing Island: central rata forest the north and north-west or by a close-cropped surrounded by Olearia lyallii; rock platforms sward on the very exposed western slopes, unshaded. Based on an aerial vhotogravh. 1964. GODLEY: VEGETATION OF AUCKLAND Is. 61

.olearia Lyallii-A RECENT IMMIGnANT? places on the island were occupied by Poa litorosa growing in large tussocks". Cockayne This striking tree-composite, with leaves (1909), on the other hand, still wrote of O. amongst the largest in the New Zealand flora, lyaUii thot "this formation occupies a small is usually recognised as occurring only on the portion of Ewing Island." Snares Is., and at the northern extremity of the Auckland Is., where it is mainly concentrated Oliver spent half an hour on Ewing 1. in at Erebus Cove and on Ewing 1. Falla (1948, 1927, b;.;t only noted that "the scrub here is and in Moar 1958 b) pointed out that the area mainly tupari." of O. lyallii on the headland at Erebus Cove Whether Olearia on Ewing 1. now grows had grown up in a man-made clearing; and he on a man-made clearing, as tentatively sug- suggested that the same might have occurred gested by Falla (1948), is by no means as on Ewing I. It is of interest to discuss whether certain as it is for Erebus Cove. The area now O. lyallii may in fact be a recent immigrant under Olearia on Ewing seems much larger and not, as Cockayne thought, a remnant of than needed for the beaching of whale boats a primeval forest which has been ousted by and for shore equipment, if this in fact changing conditions. occurred during the brief unsuccessful period in 1850-52. Ewing Island Recalling Chapman's and Dorrien-Smith's In 1840 O. lyallii was rare in Ross Harbour. references to tussock on Ewing, and the pre- One specimen was brought from Ewing I. sent tussock-grassland on Rose I., it is possible by Lyall, and Hooker described the plant thus: that Ewing 1. originally had a central stand "a short stout trunk rises a few inches above of rata surrounded by Poa litorosa tussock- the ground and then sends off horizontally grassland. The grassland could have been p3tent branches which radiate as from a com- invaded by Olearia which grew up in the mon centre for 10 or 12 feet on all sides a shelter of the large tussocks. little above the surface of the earth." This description suggests a stunted plant in an Erebus Cove (Auckland Island) exposed situation. Hooker continues: "It is a rare plant on the islands now under considera- O. lyallii was not seen here by Hooker in tion and will probably be found to be a native 1840 although his vessel was anchored at the of the southern extremity of New Zealand." Cove. Nor was the species seen by the French M'Cormick (1884), who also landed on Ewing and American expeditions in the same year. I. in 1840, does not mention the species. During the Enderby settlement (1850-52) the Bolton collected a specimen for Hooker in 1850 forest was cleared behind the beach, across the but Dr. R. Melville of Kew tells me that it is base of the headland, and on the headland only labelled "Auckland Islands". itself. It is on the headland that O. lyallii thrives today. In 1890 Chapman recorded, "We walked for some hours over this flat island, among fairly The scanty botanical records of the German grown rata trees, which occasionally bore Transit of Venus party (1874) give no relevant bunches of glorious crimson flowers, finding information. Nor did Chapman and Kirk, magnificent Olearias on the sea-shore. Every during their 1890 visit, record O. lyallii here. now aud then we stumbled upon huge sea-lions Chapman only notes (1891: 499): "All that among the tussocks. . ." (Chapman 1891: 500). remains of Governor Enderby's settlement is a piece of country which looks as if it had In 1903 Cockayne observed that O. lyallii been cleared, with stumps sticking up here occupied only a narrow zone on the sheltered and there. . ." Cocka yne in 1903 examined side of the island giving way to rata forest the regeneration and recorded: "The new (Cockayne 1904: 254). In 1907 Cockayne formation consists of Metrosideros lucida, again landed on Ewing I., with Dorrien-Smith Panax simplex, Coprosma foetidissima, Draco- (1908), but their records are contradictory. phyllum longifolium and Phormium tenax Dorrien-Smith wrote: "The chief interest the whole forming a most beautiful natural anlong the plants was the forest of Olearia shrubbery in various hues of green" (Cockayne lyallii which covered the low island from one 1904: 303). He is presumably describing the end to the other. . ." ~ and that "the clear area at the beach and across the base of the 62 ECOLOGY OF SUBANTARCTIC ISLANDS headland (where the patch of Phormium tenax of the depOl" (i.e. Erehus Cove); and may still be seen). All that Cockayne notes Dorrien-Smith (1908) refers to "a small patch in 1904 (other than his references to Ewing 1.) on the main island, at Port Ross." Photograph is that O. lyallii occurs "perhaps to a very No. 63 of the 1907 expedition in the Canter- limited extent in one or two places on Auck- bury Museum shows that O. lyalli was plenti- land Island itself." After the 1907 visit ful on the headland by that time, although the Cockayne (1909) wrote more definitely that area appears slightly smaller than at present, "There are a few trees in the neighbourhood and the trees appear thinner on the ground.

Enderby J

Rose I

,Ocean I I Ewing I Tucker Pt ,

, .-Erebus Cove ,1, numerous 'Davis I , - seedlings .Webling Bay

AUCKLAND J

FIGURE 5. Distribution of Olearia lyallii on the Auckland Islands: locations observed in 1962- 63, with single plants marked. (Falla (in Moar 1958 b) refers to "A small patch at the head of Laurie Harbour on a site described by Ross (1847) as a clearing for huts of French whalers". Neither plants or reference was found.) FALLA: BIRDS AND l\1AMMALS 63

General pollen analysis may not be fruitful in elucidat- ing the history of this noble composite in the The luxuriant growth of O. lyallii, and Auckland Islands. the large numbers of young plants, with germination even in cracks in the bark of ACKNOWLEDGEMENTS prostrate trunks, have often been noted-for instance by Cockayne (1904) and Dorrien- I am very grateful to Dr. R. A. Falla for a vegetation map of Ewing Island on which Fig. 4 is based and Smith (1908). Although Cockayne could see for his interest in these problems; and to Mr. K. R. no reason why it "should not be the dominant West for preparing the illustrations. forest of the Southern Islands", O. lyallii has always been considered as fighting a losing REFERENCES battle in the Auckland Islands. To Dorrien- Smith (1908) it was swamped out by rata, CHAPMAN, F. R., 1891. The outlying islands of New Zealand. Trans. N.Z. Inst. 23: 491-522. "hence its disappearance from the main COCKAYNE,L., 1904. A botanical excursion during mid- islands"; to Cockayne (1904) it was perhaps winter to the southern islands of New Zealand. a relic of a former primeval forest, "ousted by Trans. N.Z. Inst. 36: 225-333. a new formation as the conditions changed"; COCKAYNE, L., 1907. In southern seas. The Auckland Islands. N.Z. Times, 11 Dee. and he also considered "that the tree in question COCKAYNE,L., 1909. The ecological botany of the sub- may have been a member of the now vanished antarctic islands of New Zealand. In The Sub- and problematical forest of an ancient antarctic Islands of New Zealand. Phil. Inst. of Antarctic continent" (Cockayne 1907; d. also Canterbury. COCKAYNE, L., 1927. New Zealand plants and their 1928: 174). story. 3rd. edn. Govt. Printer, Wellington. I suggest that O. lyallii is in fact increasing COCKAYNE, L., 1928. Vegetation of New Zealand. 2nd its area on the Auckland Islands, and that the edn. Engelmann, Leipzig. DORRIEN-SMITH, A. A., 1908. The southern islands present populations are derived from a rela- expedition. Kew Bull. 6: 239-249. tively recent immigration, probably in the early FALLA, R. A., 1948. The outlying islands of New Zea- 19th century, with Ewing 1. as the point of land. N.Z. Geographer 4:127-154. entry. From here, it is suggested, the species HOOKER, J. D., 1847. . Vol. 1. Reeye Bros., London. moved east to points around Ross Harbour, and HOOKER, J. D., 1867. Handbook of the New Zealand soutbwards along the coast (Fig. 5). Flora. Reeve & Co., London. The source of O. lyallii would probably be LEAMY, M. L., and BLAKEMORE,L. c., 1960. The peat soils of the Auckland Islands. N.Z. J. Agr. Res. 3: the Snares, the only other locality where it is 626-546. usually recognised as occurring. But when M'CORMICK, R., 1884. Voyages of discovery in the speculating on the history of this species it Arctic and Antarctic seas. Pt. 1. Sampson, London. should be recalled that specimens from Milford MOAR, N. T., 1958a. Contributions to the quaternary history of the New Zealand Flora. 1. Auckland Sound (Hooker 1867) and from Stewart Island Island peat studies. N.Z. J. Sci. 1: 449-465. and the Bluff (Cockayne 1927, 1928) have been MOAR, N. T., 1958b. Notes on the botany of the Auck. considered similar to or identical with O. lyallii. land Islands. N.Z. J. Sci. 1: 466-479. OLIVER, W. R. B., 1927. New Zealand's subantarctic As Moar (1958 a) mentions the difficulty of isles. Typescript held at Museum, Well- distinguishing O. lyallii pollen from that of ington. (Printed with modifications in the Evening the Pleurophyllums, particularly P. hookeri, Post, Wellington, 16 April 1927.)

BIRDS AND MAMMALS OF THE SUBANTARCTIC ISLANDS

R. A. FALLA Dominion .Museum, Wellington

If comprehensive limits are allowed for a northern zone vvith some derivative sub- zone in the New Zealand area designated antarctic elements only and this is clearly 'Subantarctic', it would include the outlyers distinguisha ble from the middle zone which of Stewart 1., the Snares and the Chathams would include Auckland, Campbell, Antipodes in the north, and Macquarie I. in the south. and Bounty Is. The southern zone represented There is, however, a very mixed fauna in the by MacquCtrie is distinguishable again, its