Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

Tuatara: Volume 11, Issue 3, September 1963

New Caledonia and New Zealand — A Botanical Comparison

page 178

New Caledonia and New Zealand — A Botanical Comparison

In View of the Frequently Suggested Subtropical Affinities of a part of the New Zealand flora a close look at the flora of our nearest neighbour in a tropical direction, New Caledonia, is clearly called for. New Caledonia lies 1.000 miles north-west of New Zealand at a point approximately half-way between New Zealand and New Guinea. The island's dimensions, about 250 x 30 miles, are very similar to those of the North Auckland peninsula, although the former is much more mountainous with altitudes up to 5.500 feet. Geologically New Zealand and New Caledonia have had related histories and the two countries are at the present time connected by a series of submarine ridges. The number of seed plants native to New Caledonia is estimated at about 3,000 far exceeding the approximately 1,750 species native to New Zealand. Furthermore, if the comparison were restricted to comparable types of vegetation the discrepancy would be even greater, as the New Zealand alpine vegetation has no counterpart in New Caledonia.

Fig. 1: New Caledonia showing routes followed.

Fig. 1: New Caledonia showing routes followed.

page 179
Fig. 2: Melaleuca woodland, East coast. Note pale bark blackened at the base by fire.

Fig. 2: Melaleuca woodland, East coast. Note pale bark blackened at the base by fire.

Fig. 3: Serpentine scrub, Mt. Kohgi. Dracophyllum sp. centre, Meryfa sp. right.

Fig. 3: Serpentine scrub, Mt. Kohgi. Dracophyllum sp. centre, Meryfa sp. right.

page 180

Despite the lack of alpine habitats the island is hy no means entirely forested. The prevailing winds are easterly, so rainfall is highest on the east coast as well as on the upper slopes of the mountains and rainforest is largely restricted to these situations. On the drier western side of the island the prevalent vegetation cover is a dry open woodland dominated by Melaleuca leucadendron (Niaouli). This overall pattern is complicated by the occurrence of large areas of serpentine rock. The largest such area occupies all of the southern third of the island apart from a western strip and there are also smaller ‘islands’ of serpentine spaced along the north-western coast. These latter cause a striking change in the vegetation cover as Melaleuca leucadendron is rarely found on serpentine, being replaced there by a lower, denser cover of Acacia spirorbus. The pattern of vegetation on serpentine in the wetter areas, notably the south-east, is much more complex. Rainfall in the south-eastern serpentine area is mostly over 100 inches which would be more than adequate for forest under normal circumstances. However at lower elevations, below about 1,500 feet on the average, the vegetation is ‘scrub’ formed by a surprising variety of shrubs and some herbs of a distinctly xerophytic appearance. The terrain clothed by serpentine scrub is of relatively low relief with a deep, red soil of sandy texture. According to Sarlin (1954) this soil may be metres to dozens of metres thick and he attributes the xerophytic nature of the vegetation to the extreme permeability of the soil, which he feels more than counter balances the high rainfall. On the other hand Birrell and Wright (1945) suggest that the absence of forest on these soils may be due to toxicity of the chromium and nickel compounds present.

Above 1.500 feet the slopes become much steeper and support extensive, species-rich forests. Rainfall is higher and according to Sarlin this factor, combined with the steep slopes, allows heavy haching of the serpentine minerals and their rapid transport to lower levels. The soil is skeletal, consisting of a thin layer of clay without any red colouration.

The presence of forest at higher levels on serpentine is puzzling as it is generally thought that with steep slopes and skeletal soils the toxic effect of serpentine increases, while here the reverse appears to be the case.

My main aim in visiting New Caledonia was to examine the forests there in order to compare them with those of New Zealand. Two main types of forest are recognised in New Caledonia. These are termed by Sarlin (1954) ‘middle altitude forest’ and ‘conifer forest’. The former ranges from approximately 1.300 feet to 3.250 feet in altitude and appears to fit the concept of tropical rain-forest as described in Richards (1952), the latter ranges from 3.250-5.400 feet and has much in common with the New Zealand page 181 podocarp-dicotylous forest or ‘bush’. In this paper I shall refer to the two types of forest as ‘lowland’ and ‘montane’ respectively.

I was able to examine lowland forest on the Mount Kohgi Range near Noumea and also patches of regenerating forest of this type on the east coast near Poindimié. A good example of montane forest was studied on Mt. Ignambi in the far north.

Lowland Rain Forest

Although sharing the same range of plant forms with the New Zealand lowland forest — trees, shrubs, lianes and vascular epiphytes — the lowland rain forest in New Caledonia differs from ours in several respects. On entering the forest two differences are immediately apparent, firstly general leaf size is much greater, a fact which is first observed in the leaf litter, and secondly lianes and epiphytes are neither so abundant nor so luxuriant. Shrub epiphytes appear to be absent. In structure the forest is similar to ours, being multi-storied with an upper level of emergent trees and below that upper and lower canopy layers and a shrub layer. The emergents however are flowering plants while ours, with the exception of Metrosideros robusta, are conifers. The Metrosideros becomes an emergent by virtue of establishing itself as a so-called ‘strangling’ epiphyte on an emergent conifer.

Fig. 4: Lowland forest, Mt. Kohgi. Uneven canopy with scattered emergents.

Fig. 4: Lowland forest, Mt. Kohgi. Uneven canopy with scattered emergents.

page 182

The number of species in this New Caledonian forest is bewilderingly great. In one small valley on Mount Kohgi I collected 60 species of trees, shrubs and lianes and the collection was by no means complete.

Ferns, bryophytes and lichens appeared to be less common than in our forest and tree ferns in particular were quite infrequent. In gullies were one might expect to find tree ferns there were instead impressively large plants of Marattia. In some cases the root stocks were several feet high and wide with fronds up to 20 feet long.

In my notes on the Mt. Kohgi forest I describe the emergent trees as being 20 or more feet apart and up to 100 feet high with relatively slender, gradually tapering trunks 2-3 feet in diameter. A few of the emergent species have elaborate plank buttresses at the base, while others are only slightly buttressed. The trunks for the most part are free of epiphytes and climbers. Occasional plants of Asplenium nidus (Bird's nest fern) occur at branch forks and also the more diffuse fern Drynaria rigidula, but there does not seem to be any tendency for these to be aggregated together into ‘epiphyte gardens’ as is the case with Astelia and Collospernum in New Zealand.

The main canopy trees are closer together, lack buttresses, are mostly no more than a foot in diameter and range up to 70 feet high.

The sub-canopy trees are up to 40 feet high with trunks six inches or less in diameter.

An occasional large tree supports a strangling fig (Ficus spp.). These differ from the New Zealand Metrosideros robusta in that the descending roots form a complete network about the trunk of the host. In Metrosideros robusta the roots are usually disposed to one side of the host trunk.

Montane Rain Forest

When we reached the montane forest on Mt. Ignambi the similarities with New Zealand lowland forest were immediately apparent. Leaf size was greatly reduced, species were fewer, and most of the canopy trees belonged to familiar genera — Weinmannia, Metrosideros, Elaeocarpus. Tree ferns were quite common and ferns generally, bryophytes and lichens were more evident than in the lowland forest. Replacing the ‘bird's nest’ ferns of lower altitudes was an epiphytic species of Astelia (A. neocaledonica), occurring quite abundantly on the trunks and branches of many of the trees. It was interesting to discover that at least one of the species of Metrosideros can act as a ‘strangler’ epiphyte and that a species of Weinmannia sometimes begins life as a low epiphyte, as is the case with Waracemosa in New Zealand. At about 4.000 page 183
Fig. 5: Lowland forest, Mt. Kohgi. Plank buttresses on an emergent tree.

Fig. 5: Lowland forest, Mt. Kohgi. Plank buttresses on an emergent tree.

Fig. 6: Lowland forest, Mt. Kohgi. Descending roots of a strangling fig (Ficus sp.).

Fig. 6: Lowland forest, Mt. Kohgi. Descending roots of a strangling fig (Ficus sp.).

feet on Mt. Ignambi I noted a case where a host tree supported an epiphytic Metrosideros and an epiphytic Weinmannia. The host trees was dead with a steeply inclined trunk about one foot in diameter, broken off about 15 feet from the ground. The roots of the Metrosideros were wrapped around the base of the host and could be traced from there along the upperside of the trunk from whence several branch roots descended to the ground. The root-stem junction was judged to be about 20 feet from the page 184
Fig. 7: Young fronds, Marattia sp. Lowland forest, Mt. Kohgi.

Fig. 7: Young fronds, Marattia sp. Lowland forest, Mt. Kohgi.

ground and the total height of the tree 40-50 feet. The Weinmannia was attached to the underside of the host about 10 feet from the ground and its roots ran from there to the base of the host trunk. The trunk of the Weinmannia was about six inches in diameter and the height of the crown about 30 feet. Astelia was quite common along the upperside of the combination.
page 185

As in New Zealand the tree fern trunks supported a number of epiphytes, including Tmesipteris and seedlings of Weinmannia and Metrosideros.

The chief difference between this forest and the lowland forest in New Zealand was the absence of emergent trees. The reason for this may be the scarcity of conifers, which in New Zealand provide most of the emergents. In the Ignambi forest there were only two conifer species — Dacrydium taxoides, a scattered undershrub reminiscent of sapling Podocarpus ferrugineus in New Zealand, and Austrotaxus spicata which according to Sarlin, can become a fairly large tree 50-80 feet high and up to three feet in diameter. However on Mt. Ignambi the occasional trees of this species were quite small, possibly because they were above the range of 1,600-2.600 feet given by Sarlin.

The conifers in New Caledonia are particularly interesting. There are 34 species, 14 more than in New Zealand. Of these 16 belong to the family Podocarpaceae (Podocarpus 191, Dacrydium (5), Acmopyle (2)); 11 to the Araucariaceae (Araucaria (8), Agathis (3); six to the Cupressaceae Libocedrus (3), Callitris (2). Callitropsis (1)); and one to the Taxaceae (Austrotaxus (1)). The surprising fact is that 25 of these species are restricted to serpentine in the southern half of the island. Of the nine other species three occur only on non-serpentine in the north
Fig. 8: Montane forest, Mt. Ignambi. Fairly even canopy without emergents.

Fig. 8: Montane forest, Mt. Ignambi. Fairly even canopy without emergents.

page 186 (Austrotaxus spicata, Araucaria montana and Agathis moorei) and six occur on serpentine in the south as well as non-serpentine in the north. Three of the nine species in the north do not usually occur in forest, including the tall pencil-like Araucaria cookii along the coasts.

Despite the very minor role played by conifers in the Ignambi forest its points of agreement with the New Zealand lowland podocarp forest are so striking that it can only be regarded as belonging to the same vegetation type. This type of vegetation has its closest affinities with tropical rain forest and for this reason has often been termed subtropical rain forest. Certainly tropical and subtropical rain forests are much more closely related to each other than either is to other world types of forest vegetation, e.g. to temperate deciduous forest, and it would seem appropriate that the names applied to them should reflect their affinity. Whatever names are used it is essential that disjunct occurrences of what are judged to be the same world vegetation type should not be given different names, e.g. rain forest of the type here termed subtropical should not be called ‘subtropical’ where it occurs near the tropics and ‘warm temperate’ where it occurs at higher latitudes.


It is only recently that the genus Nothofagus has been known to occur in New Caledonia. At the present time five species are recognised, all belonging to the section of the genus (N. brassii group) otherwise known living only in New Guinea. As extensive Nothofagus forests occur in New Zealand I was interested to discover whether such forests occurred in New Caledonia and what relation they have to the rain forest.

Unfortunately I did not see Nothofagus for myself, but I gathered from discussions with local botanists that there was nothing that could really be described as Nothofagus forest. Trees of this genus mostly occur in rain forest, although tending there to form distinct groves. Evidently in the Montagne des Sources area near Noumea Nothofagus becomes dominant on the higher ridges recalling the similar rain forest/Nothofagus forest pattern in New Zealand. This may be the nearest approach to Nothofagus forest in New Caledonia at the present time.

The Role of Fire

According to Sarlin rain forest does not occur below 1,300 feet as a general rule, being replaced at the lower altitudes by Melaleuca woodland, or serpentine scrub. This applies even on the wet east coast where the annual rainfall at sea level is mostly 80in. page 187
Fig. 9: Podocarpus sylvestris, Mt. Kohgi. The climber left and right is a small species of Freycinetia.

Fig. 9: Podocarpus sylvestris, Mt. Kohgi. The climber left and right is a small species of Freycinetia.

Fig. 10. Descending root of Metrosideros sp., Mt. Kohgi.

Fig. 10. Descending root of Metrosideros sp., Mt. Kohgi.

page 188 or more and in this case Sarlin attributes the lack of forest to the drying effect of the easterly winds. He notes that the vegetation pattern here tends to be Melaleuca on the exposed easterly slopes, rain forest on the sheltered westerly slopes and in the valley bottoms. Above 1,300 feet increased rainfall counterbalances the drying effect of the wind and rain forest becomes continuous. My own impression is that, although the drying effect of the wind is a factor, fire may be the basic cause of the pattern Sarlin describes. Fires on the seaward slopes of the hills are obviously very frequent as the many blackened ares testify and, according to local people, most of them are started by the natives for no particular reason. Fires on seaward facing slopes would be fanned by the wind and would mostly burn to the crest of a ridge and stop, resulting in a very sharp boundary at the ridge crest between forest and the scrubby, pioneer vegetation following fire.

We examined several patches of rain forest on sheltered western slopes and they all appeared to be second growth. This would suggest that fires also occur on the sheltered westerly slopes, but sufficiently infrequently to allow re-establishment of forest. On the eastern slopes fires are so frequent that rain forest would have little chance of re-establishing. This state of affairs may have obtained for many centuries as Cook in 1772 describes Melaleuca as clothing the hills near the sea.

A very similar pattern caused by recurrent fires can be found in New Zealand, although here prevailing winds are westerly, so west-facing slopes are burnt most frequently and carry scrubby early regeneration stages, while forest or regenerating forest occupies the sheltered east-facing slopes. It is interesting to note that Virot (1956) regards all Melaleuca woodland as induced vegetation, presumably resulting from fire.

On the east coast in New Caledonia the pioneer plants in most situations are abundant Gleichenia linearis and Lycopodium cernuum with Pteridium aquilinum (looking very similar to our bracken), Metrosideros sp. (a small shrub), Baeckea sp. (related to Leptospermum) and Gahnia spp. more scattered. Melaleuca grows through and eventually above this cover. Once established the Melaleuca savannah, as it is often called, is very resistant to fire. Fires burn through the ground cover blackening the trunks without killing the trees. If the fire is very intense the foliage may be destroyed, but even then the trees are often capable of producing new leaves.

Sarlin mentions that Melaleuca trees sometimes invade rain forest, but that shading and sometimes strangling figs usually cause them to die out. Another possibility is that the Melaleuca trees were on such sites first as a sub-climax and that rain forest species have established beneath them, eventually overtopping them and causing them to die out by shading.

page 189
Fig. 11 : Regeneration pattern following fire, East coast. Foreground, early regeneration stage with abundant Gleichenia and scattered Gahnia and Melaleuca. Middle distance, second growth forest.

Fig. 11 : Regeneration pattern following fire, East coast. Foreground, early regeneration stage with abundant Gleichenia and scattered Gahnia and Melaleuca. Middle distance, second growth forest.

Such, in brief, is the present vegetation pattern of New Caledonia. To understand this pattern fully it would be necessary to know something about its history, particularly during the last glacial period and during the warmest part of the Tertiary. Fleming (1963) suggests that during the last glaciation vegetation zones were approximately 3,000 feet lower than now. In New Caledonia page 190 this would mean montane forest down to present sea level and lowland rain forest, if it survived at all, near the sea level of that time, which is estimated to have been about 400 feet below that of the present. Above the montane forest there were possibly Nothofagus forests and it seems likely that there would have been small areas of alpine vegetation on the higher mountains.

During the period of maximum warmth during the Tertiary it is suggested that New Caledonia was much less mountainous than now, so probably only forest of the present lowland type would be present. Serpentine rocks were present in New Caledonia throughout the Tertiary, so serpentine scrub has probably always been a feature of the vegetation.

As a means of elucidating vegetation history in New Caledonia, a study of plant fossils, particularly fossil pollens, is urgently needed.

Floristic Comparison

This comparison is based on Guillaumin (1948) for New Caledonia. Allan (1961) for New Zealand gymnosperms and dicotyledons and Cheeseman (1925) for New Zealand monocotyledons. As the New Caledonian flora is still incompletely known and as Cheeseman's account of the monocotyledons is more than 30 years old, the following statistics must be regarded as approximate.

There are more than twice as many genera of seed plants recorded for New Caledonia as for New Zealand, 793 against 344. The number of genera shared is 124 representing 37% of the New Zealand total, but only 16% of the New Caledonian.

The discrepancy between total numbers of species is not quite so great however, 3,000 versus about 1,750, owing to a smaller average number of species per genus in New Caledonia (3.8) than in New Zealand (4.9). A check revealed that the proportions of monotypic genera in the two floras is not greatly different — in New Caledonia 48% and in New Zealand 44%. Possibly the larger land area in New Zealand has allowed greater scope for speciation. although it should be pointed out that even the average number of species per genus in New Zealand is very low by comparison with floras of less isolated regions.

A comparison of the number of species in New Caledonia with those of areas of comparable size and diversity in New Zealand reveals that, on the average, the concentration of species in New Caledonia is about three times that of New Zealand.

The common genera can be placed in two categories according as they are judged to belong to forest or non-forest vegetation. The numbers in the following lists refer to the number of species in New Caledonia and in New Zealand respectively.

page 191
Common Forest Genera (44)
Agathis 5,1
Dacrydium 5,6
Libocodrus 3,2
Podocarpus 9,6
Alectryon 1,1
Ascarina 4,1
Avicennia 1,1
Corynocarpus 1,1
Dodonaea 1,1
Dysoxylum 26,1
Elaeocarpus 29,1
Eugenia 30,1
Geniostoma 17,1
Hedycarya 17,1
Knightia 2,1
Korthalsella 3,2
Litsea 12,1
Melicope 11,2
Meryta 10,1
Metrosideros 10,11
Muehlenbeckia 2,5
Myoporum 8,2
Myrsine 28,9
Neopanax 1,6
Nothofagus 5,4
Parsonsia 36,2
Peperomia 12,2
Pittosporum 30,26
Planchonella 31,1
Quintinia 6,3
Schefflera 21,1
Sophora 2,3
Vitex 5,1
Weinmannia 5,2
Acianthus 14,1
Astelia 1,10
Dendrobium 36,1
Bulbophyllum 10,2
Earina 5,3
Freycinetia 15,1
Gastrodia 1,1
Lyperanthus 11,1
Sarcochilus 5,1

Species totals: New Caledonia 501 (11.4 per genus); New Zealand 133 (3.0 per genus).

Common Open Habitat Genera (82)
Actinotus 1,1
Alternanthera 2,1
Apium 2,2
Atriplex 1,2
Bidens 2,1
Brachycome 2,3
Callitriche 2,5
Calystegia 2,4
Cassytha 1,1
Centella 1,1
Centipeda 1,1
Chenopodium 3,4
Clematis 1,10
Cotula 1,24
Dracophyllum 7,35
Drosera 1,6
Epacris 1,2
Erechtites 1,8
Euphorbia 11,1
Exocarpus 4,1
Gnaphalium 3,14
Haloragis 1,7
Helichrysum 1,9
Hibiscus 9,2
Hypericum 1,2
Ipomea 17,2
Lagenophora 2,5
Lepidium 3,9
Leucopogon 13,8
Sonchus 1,2
Suaeda 1,1
Urtica 1,6
Utricularia 2,9
Viola 1,3
Vittadinnia 1,1
Wahlenbergia 1,10
Arthropodium 2,2
Caladenia 1,4
Calochilus 1,2
Cladium 5,10
Carex 7,55
Cordyline 1,4
Corysanthes 1,8
Cyperus 7,2
Dianella 8,1
Eleusine 1,1
Fimbristylis 7,1
Gahnia 4,8
Imperata 2,2
Juncus 2,17
Killingia 2,1
Lemna 2,2
Lepidosperma 2,2
Mariscus 3,1
Microtis 5,1
Oplismenus 3,1
Orthoceros 1,1page 192
Mitrasacme 2,2
Oxalis 4,3
Polygonum 5,1
Ranunculus 1,43
Rhagodia 2,1
Rubus 3,5
Rumux 2,2
Salicornia 1,1
Samolus 1,1
Senecio 1,16
Siegesbeckia 1,1
Solanum 21,3
Paspalum 6,3
Potamogeton 3,5
Prasophyllum 1,4
Pterostylis 6,13
Ruppia 1,1
Sehoenus 5,7
Sclipus 5,13
Spinifex 1,1
Sporobolus 2,1
Thelymitra 5,14
Typha 1,1
Uncinia 1,14
Xeronema 1,1

Species totals: New Caledonia 253 (3.0 per genus); New Zealand 478 (5.8 per genus).

In New Caledonia the largest genera belong to the forest and in New Zealand to vegetation of open habitats, particularly alpine habitats. This difference is clearly reflected in the common genera.

Probably a number of the common forest genera have reached New Zealand via New Caledonia and some of the shared open habitat genera in New Caledonia may be immigrants from New Zealand.

No complete account of the New Caledonian pteridophytes is available, but resemblance to New Zealand species was quite striking in some cases. The bracken fern (Pteridium aquilinum) looked very similar to the New Zealand form and there were species very like our Gleichenia microphylla and G. linearis, Lycopodium cernuum and L. volubile. The form of Tmesipteris appeared somewhat different with the leaves very regularly flattened into two rows. A species of the fern genus Davallia was very common on rocky outcrops and looked very similar to Davallia tasmanii restricted to the Three Kings Islands in New Zealand.

Among seed plants the absence of the genus Beilschmiedia, woody Compositae and the genus Coprosma provides a marked contrast with New Zealand.

In conclusion it seems hardly necessary to emphasise the value of a more detailed comparative study of the New Zealand and New Caledonian floras. In particular, from the New Zealand point of view, further ecological and historical information about the montane forest, serpentine vegetation and Nothofagus in New Caledonia would be especially interesting and might aid us in the interpretation of similar types of vegetation in New Zealand.


Allan, H. H., 1961. Flora of New Zealand, Vol. 1, Government Printer, Wellington.

Birrel, K. S., and Wright, A. C. S., 1945. A Serpentine Soil in New Caledonia. N.Z. Journ. of Science and Tech., 27: 72-76.

page 193

Cheeseman, T. F., 1925. Manual of the New Zealand Flora. 2nd. Ed., Government Printer, Wellington.

Fleming, C. A., 1962. New Zealand Biogeography. A Paleontologist's Approach. Tuatara 10 (2): 53-108.

1963. Age of the New Zealand Biota. N.Z. Ecological Society Proceedings 10, Wellington.

Guillaumin, A., 1948. Flore de la Nouvelle-Caledonie. Paris.

Richards, P. W., 1952. The Tropical Rain Forest. Cambridge University Press.

Sarlin, P., 1954. Bois et Forets de la Nouvelle-Caledonie. Paris. Virot, 1956. La Vegetation Canaque. Mémoires du Muséum National d'Histoire Naturelle, Série B, Botanique, Tome 7, Paris.


I should like to acknowledge the University Grants Committee and the Victoria University of Wellington for grants enabling me to carry out this study.