Other formats

    TEI XML file   ePub eBook file  


    mail icontwitter iconBlogspot iconrss icon

The Vegetation of New Zealand

Chapter II. — The New Vegetation

Chapter II.
The New Vegetation.

1. General.

Although the communities dealt with in this chapter fall conveniently under certain heads, there are not only intermediates between them, but it is sometimes difficult to decide into what class any particular combination of plants should go. Even in a series apparently so clear-cut as artificial page 356farmlands, nature, simultaneously, or very soon, may alter the intention of the farmer by bringing in more or less weeds; indeed, the latter element not infrequently dominates the community which is then transformed from one purely artificial to one of an exotic-induced or even an indigenous-induced character, As for the communities themselves, they are generally successions, the progress of which cannot usually be predicted for certain. Even when an apparently well-established association is present over wide areas (e. g. Ulex thicket, manuka shrubland, water-cress tangle), there is no telling what may eventually come abut through the establishment of some species new to the community, or some disease attacking the dominant member. Here, then, the term community is generally used rather than one speaking more definitely of fixity.

In considering this new vegetation as a whole, one fact — surprising enough to those taught to believe in the extreme aggressiveness of what Hooker termed the "Scandinavian flora" — is the far-greater aggressiveness—in their native land—of the New Zealand indigenous species themselves. Where rain-forest or swamp-forest has been felled and artificially replaced by European pasture-plants, as detailed further on, were it not for the constant presence of the farmers grazing and browsing animals, there would be rapid reversion to forest. In fact, it is hardly going too far to declare, that were such animals entirely removed from North Island, the whole of the present "permanent pastures" would in one hundred years, or less, be well on the road once more towards dense rain-forest!

Certainly the effect of the grazing and browsing mammal cannot be overestimated. Even, a plant-covering, so persistent as Pteridium heath (as detailed further on), is constantly being turned into grassland by aid of overstocking and burning, but alone the latter does nothing. If all stock, including rabbits, were removed from the induced steppe of semi-arid Central Otago, experiment has clearly demonstrated (Cockayne, L., 1922 a: 142) that, in certain habitats, there would be a gradual return to something not unlike the primitive plant-covering, while, if suitable seeds were sown, the process of pasture-establishment would be rapid, except in the most arid and stony places (lo. cit.: 142–43).

2. Modified communities.


A plant-community, without destroying its primitive stamp and general ecological and floristic structure, may readily be altered by a change in the relative abundance of its constituents, which may be so marked that a species of but little moment in the primitive community, or one introduced, may become common, or even dominant, and in this way a new community come into being.

The modifications may consist merely of increase or decrease in abundance of certain species already present but, on the other hand, species either page 357indigenous, exotic, or both together, frequently join the community and become definite members.

The causes of modifications are principally grazing or burning, or these combined; also seeds may be purposely sown or plants planted, or land may be drained, forest felled; indeed, there are many causes. In associations containing edible species, the relative palatability1 of their various members plays the foremost part and it can be seen how grazing, or browsing, may rapidly alter the composition of certain classes of pasture or forest without affecting their general physiognomy.

Fire, of course, functions in proportion to the damage done and in regard to the species destroyed or damaged. In forest, regeneration after fire will depend, to some extent, on the ability of certain species to put forth suckers or the contrary. It also destroys the humus of the forest-floor, so bringing about new conditions for the germination of seeds.

Both grazing &c. and burning lead to increase in light-demanding species, both in number and variety, but such additions as arrive come, almost invariably, from plants very near the scene of damage. On the other hand, stock brought from some distant locality frequently bring in seeds of plants not present in the association (e. g. species of Acaena, Danthonia pilosa, Carex resectans, Poa maniototo).

Some examples of modified communities.

The student of New Zealand vegetation has usually not far to go in order to see examples of modified associations; indeed, much of the apparently primitive vegetation he will encounter in the lowland and montane belts will be more or less modified. This has already been stressed in Part II, where various modified communities are dealt with. As the changes in these partly-modified communities become, by degrees, intensified, according as exotic or indigenous species rule, so will the communities become exotic-induced or indigenous-induced. All the same, so long as the ecological conditions governing a modified community remain constant, it will bear the stamp of fixity. Thus, many modified types of vegetation may well be expected to remain for a long time much as they now are provided no new disturbing factor appears. Virtually all lowland and montane tussock-grassland, swamp and salt-meadow, and a good deal of forest, dune and even rock-vegetation come into this class.

In some of the modified communities certain conspicuous species play page 358such an important part as to affect their physiognomy. In what follows various examples are given.

Where Ammophila arenaria has been purposely planted to arrest dune-movement, self-sown plants may invade the adjacent primitive sand-grass association and grow side by side with Desmoschoenus spiralis or Spinifex hirsutus; so, too, Oenothera odorata may become abundant and make many parts of more stable dunes, and also sandy river-beds, gay with its yellow blossoms. Sambucus nigra has become firmly established in certain lowland forests (its seeds brought by birds) where the light has been greatly increased through destruction of undergrowth; also, in the different environment of shady gullies in semi-arid Central Otago, it is now a common member of certain patches of scrub. Hypericum Androsaemum is frequent both in damaged lowland forest and in montane Pteridium heath (SO.), it having come in after fire. Senecio mikanioides climbs over small trees and shrubs on the outskirts of semi-coastal forest in some parts of North Island and north of South Island. Verbascum Thapsus is now a most important member of the plant-covering of river-terraces and other stony or dry spots in the montane belt of South Island; by the settlers it is well known by the name "tobaco-plant". Plantago Coronopus forms extensive colonies in certain salt-meadows looking exactly as if indigenous; so, too, Lepturus incurvatus is equally at home in the same formation; in places, some of the indigenous species of salt-meadow are the hosts of the parasitic Cuscuta Epithymum. Pinus radiata has extended into the Leptospermum shrubland (VP.) and, in course of time, should make a distinct association. Hypochoeris radicata is a most important constituent of low tussock-grassland.

1 Thus, in the great tussock-grassland formation, nearly all the exotic edible plants possess palatability to a far higher degree than do the indigenous grasses. For example, Festuca rubra var. fallax, a species of low palatability, ranks in this respect with Agropyron scabrum and Poa intermedia — species of great repute among high-country sheep-farmers — and the dominant Festuca novae-zelandiae, also regarded as of high palatability, is not eaten at all by sheep except the young leaves produced by burning. I have also observed sheep grazing in semi-coastal forest and how various shrubs &c. were eaten, some with avidity (e. g. Brachyglottis repanda) and others rejected.

3. Exotic-induced communities.

a. Tree associations.

At Waitati, near Dunedin (SO.) the burning of dense Leptospermum shrubland — really an indigenous-induced community — allowed the seeds from a mature tree of Eucalyptus radiata to germinate and there is now Eucalyptus forest, thanks to the destruction of the Leptospermum having provided the essential light-requirement for the development of the seedling eucalypts, and their far more rapid growth than those of L. scoparium. In Kawarau Gorge (NO.), Eucalyptus globulus has formed dense groups on rocky ground, the seeds having come from neighbouring trees, originally planted by gold-diggers in the "sixties" of last century.

Salix fragilis and S. babylonica, to a lesser extent, planted in the first instance on banks of rivers, thanks to the rooting-power of their broken twigs, line the margins of many streams.

Albizzia lophantha is self-established in many parts of the North Auckland district. Acacia dealbata, both on dunes and the northern gumlands, page 359in places, spread from parent trees by means of creeping, underground stems and forms dense groves. So, too, Robinia Pseud-acacia makes close thickets in the neigbourhood of old mission stations in several localities in North Island.

Near Lake Wanaka (NO.) windborne seed of Cupressus macrocarpa and Pinus radiata, trapped by bushes of Discaria toumatou, germinates freely and many trees have been so established in the presence of stock and rabbits thanks to their natural spiny protector.

b. Shrub communities.
Leguminous shrubs.

Ulex europaeus1, Cytisus scoparius, Lupinus arboreus and, to a much lesser degree, Cytisus candicans, invade open communities replacing and displacing the primitive vegetation. The seeds germinate in the neighbourhood of the parent shrub, and, in the first instance, the invasion was from shrubs purposely planted. At the present time, there are vast impenetrable thickets of pure (usually) Ulex and Cytisus on stony river-bed and hillsides where forest has been burnt, which when in blossom are a glorious spectacle; both species, too, are abundant on lowland and lower-montane tussock-grassland and fixed dune. At an altitude of 700 m. in South Island, the above are no longer aggressive.

Lupinus arboreus2 is confined to dry sandy or stony stations forming close thickets, 1.8 m. or more high, on fixed or semi-stable dunes; it also occurs to some extent on river-bed.

Rosaceous shrubs.

Rosa Eglanteria, R. canina3Rubus fruticosus in a wide sense and R. laciniatus form individually, in many localities, extensive thickets which differ from those of the Leguminosae inasmuch as they owe their distribution to birds4 while climate restricts Rubus to wet and Rosa to dry areas.

Rubus thicket is especially aggressive in forest-clearings. Though occurring abundantly in many places, it attains the greatest luxuriance in the East Cape, Western, North-western, Egmont-Wanganui and Chatham districts.

Rosa thicket becomes readily established on low tussock-grassland, making eventually a pure association. It is specially abundant in the North-

1 1) This was early on introduced as a hedge-plant and is still extensively used for that purpose. Where left uncut, the hedge exceeds 3.6 m. in height and seed is shed in profusion. Burning has no effect in eradicating the plants; seedlings also are produced in millions and grow with great rapidity.

2 This shrub was purposely planted, or sown, on dunes, in the first instance, in order to check drifting sand. This it is unable to do, but it forbids all sand-movement on the ground it occupies.

3 Of local occurrence.

4 Some forms of Rubus fruticosus spread vegetatively by means of natural "layering".

page 360eastern district, ascending to about 900 m. In Central Otago it is common on old mining tailings and in gullies in company with Discaria and the introduced Sambucus nigra.
Australian shrubs

1. Hakea acicularis2, Epacris purpurascens, E. microphylla and E. pulchetta form extensive colonies in Leptospermum shrubland (NA., SA.) the first-named being wide-spread, but the three epacrids being confined to one locality. The Hakea spreads after Leptospermum scoparium is burnt and probably the other shrubs first made their appearance after fire.

c. Communities of herbaceous and semi-woody plants.
Thymus vulgaris (garden thyme) association.

This well-known culinary plant occurs sporadically in several parts of semi-arid Central Otago, but on old mining tailings at Ophir several hectares are occupied by a close growth of this species, the flowers of which vary from crimson to white in many intermediate shades.

Some herbaceous plant communities.

Centranthus ruber occasionally forms a distinct and beautiful association with its abundant red, white and pink blossoms. The substratum preferred appears to be rocky slopes. One area under my close observation for some years is gradually extending its limits.

Where forest has been burned in areas subject to a considerable rainfall Digitalis purpurea forms colonies which extend for several kilometres at a time over hillsides. It also occurs in profusion on old mining tailings (SO.), rocky ground (NA.), and in the Taieri Gorge (SO.) along with Leptospermum scoparium until the average limit of the south-westerly downpour is reached, when the community halts, all on a sudden, and the induced steppe country begins.

The case of Eschscholzia californica is of particular interest. On the old flood-plain of the R. Clutha near Clyde and Cromwell, (NO.) on sandy ground, there is virtually a pure association of the above extending for several hectares at a time which, when in full bloom, is a dazzling spectacle with the yellows, brilliant oranges and creams of the fully opened

1 1) It is obvious that Australian trees and shrubs, if sufficiently hardy, are much better suited to New Zealand conditions than was supposed. The belief that Australian plants would not thrive was partly based on the statement that Bidwill was in the habit of habitually scattering Australian seeds during his travels through the country. But we have no evidence as to the nature of the seeds, their age, the soil or locality where scattered or the time of year; indeed the experience of Bidwill, even if a fact, proves nothing.

2 2) Hakea saligna forms close thickets on sour, boggy soil near Collingwood (NW.), but F. G. Gibbs has informed me that they have originated from seed purposely sown, but probably seedlings are being established from seed shed from this artificial shrubland. Where H. acicularis equals the Leptospermum in quantity, or is not so abundant, the community will be merely a modified one.

page 361flowers under the cloudless sky. The species also extends on to the depleted land. It is common, too, on the Marlborough Plain (NE.), but there the plants are more scattered. Also, it is greatly in evidence on certain stony river-beds (E.). The following species, growing under similar conditions to Eschscholzia — which also may be present — on more or less depleted grassland, and forming a distinct association are: — Echium vulgare, Acaena ovina and Gilia squarrosa and with is frequently Marrubium vulgare which, also, along with Urtica urens and Hordeum murinum makes a community on ground where sheep camp.

4. Indigenous-induced communities.

a. General.

There is nothing out of the way in indigenous and exotic species coming together and forming communities, or in the exotics making pure associations or colonies of themselves alone; on the contrary, these are what all phytogeographers must expect. But, that the indigenous plants, in the presence of hundreds of exotics dispersing far and wide highly viable seeds, can construct pure association, must cause some surprise, which cannot fail to be intensified on learning that communities, purely New Zealand in facies, but unknown in the primitive vegetation, have come into being when in competition with well established exotic-induced associations. In other words, contrary to what has been authoritatively reiterated ad nauseam, the indigenous species, in their own domain, are far more aggressive than the exotics. That is, that after all, the life-forms and "adaptations" of New Zealand plants are better suited for New Zealand conditions than those of any class of plants as yet introduced into the Region. In fact, wherever any part of New Zealand is in its primitive condition and uninterfered with by man or the animals he has introduced, none of the exotics have gained a foothold, their great powers of dissemination notwithstanding, although the virgin area may be pierced in all directions by ground occupied by man where there are introduced species in plenty. On the other hand, where man has separately, or collectively, brought into play fire, draining, cultivation and introduced domestic or feral animals, he has created a new environment where indigenous and introduced species, if the latter be present, alike go to the wall and new associations arise, or are purposely produced, made up, it may be, purely of exotics, though frequently of such and native plants. Thus, there are two distinct areas, the one dominated by primitive New Zealand conditions and the other by such as approximate to those of Europe, while between these extremes is a gradual series of intermediates.

Between the modified, exotic-induced and indigenous-induced communities there are frequently many transitions or successions leading from one to the other, so that it sometimes becomes a matter of personal opinion into which class this or that community should go.

page 362

At a low estimate. there are probably 50 more or less well-marked indigenous-induced associations or successions, some of trifling extend and others occupying wide areas; here only a limited number can be dealt with.

b. The grassland series.
Danthonia pilosa grassland.

As far as can be judged from an examination of the present modified tussock-grassland, and from and estimate of its primitive structure, it seems certain that Danthonia pilosa1 was rare, or, in places, absent in the unmodified formation. It would, however, be fairly common on rocks, open stony ground, between low shrubs, and like places. Certainly there was no primitive grassland where it was dominant.

With the introduction in 1850 of sheep-farming in South Island a powerful factor — the grazing mammal—, unknown in primitive New Zealand, came into play. Nor was this all, for the dominant tussock-grasses (Festuca novae-zelandiae and Poa caespitosa were quite neglected by the sheep. So, in order to add to the scanty food supply, the tussocks were set ablaze, many hectares at a time being burnt in order that there would be plenty of young green leaves for food. The vigorous tussocks of the early days would tolerate a good deal of burning, but as this was carried out to excess, and at all seasons, the tussocks became smaller, and some after repeated burning, followed by grazing, died outright and so, as the years sped, more and more bare ground became available for recolonization. Such was quickly occupied, light-demanding species, previously kept in check by the tussocks, seized on the new ground, so that the modified tussock-grassland of the present day came into being, with its mat-plants, low cushion-plants, small grasses and dwarf shrubs. Perhaps, first of all, it was in the dry North-eastern district that Danthonia pilosa, so well-equipped for occupying the open spaces, gradually increased in amount. Nor was this because it was inedible, but quite the contrary, since it is liked by sheep especially in spring, while the grazing is really beneficial since it encourages the turf-making habit. Nor is this all, for "danthonia" — as it is now called — not only tolerates burning, but the young leaves are specially palatable. It is no wonder, then, that the North-eastern pastures, over wide areas are now devoid of tussocks and present a more or less continuous turf of Danthonia pilosa3, not merely the association but the formation being changed.2

1 1) The species is compound but, so far only one jordanon has been recognized (var. racemosa). Probably it crosses with D. semiannularis, another compound species. D. pilosa is low-growing, of a more or less turf-forming habit (intensified by grazing), with leaves narrow, flat or involute (according to circumstances). D. semiannularis is of tufled habit. Both are light-demanding, hence their absence in dense tussock-grassland.

3 On stony or wet ground, exotics and indigenous enter in, but occupy little space compared with the dominant danthonia. Exotic grasses, of highly aggressive power elsewhere, are powerless, e. g. Anthoxanthum odoratum, Agrostis tenuis, A. alba (in a wide sense) and Poa pratensis,

2 Many areas were not grazed until 1860, or later.

page 363
Induced steppe.

In particularly dry parts of South Island the primitive tussock-grassland has been entirely replaced by a desert-like formation, here designated steppe. How this striking transformation has come about, and of what the new vegetation consists, are matters of far more than local interest.1

The maximum steppe development occurs in the most arid part of New Zealand, the upper basin of the R. Clutha2. Originally the valleys and hillsides, now virtually desert up to about an average altitude of 900 m., at the advent of sheep-farming were clothed with low tussock-grassland which must have existed for a very long period. Its subsequent disappearance, then, is due to those new conditions imposed by the sheep-farmers. But such were identical with those experienced for the same period by the formation elsewhere, yet, with the exceptions noted, steppe is unknown. Plainly, then, the sole difference for the North Otago grassland lay in its special rainless climate3 and its temperature, except in winter, easily the highest in New Zealand.

In order to render their harsh feed palatable for sheep, as already explained, the tussocks were burnt at all seasons. Now, in an arid climate or situation, a tussock can badly tolerate burning even during a moist period and the subsequent eating of the young leaves is highly detrimental to its well-being, while burning during hot dry weather may cause death. Also, consider the effect of a second burning on tussock not recovered from the previous attack, and then think of the result of indiscriminate burning coupled with heavy grazing year in and year out!

The lower mountain slopes and valleys, now depleted to the full (Figs. 82, 104), are where the flocks have been wintered yearly, nor can they be driven on to the high pastures until late in spring, hence this lower country has been heavily overstocked. So, what with burning and overstocking, more and more bare ground would gradually appear, the palatable plants would be eaten out, and the tussocks become weaker and weaker until they too vanished. In some parts of the valleys blown sand and gravel would be an additional agent of destruction (Fig. 59).

1 1) An account of such sowing is given when dealing with artificial communities near the end of this chapter

2 2) Steppe in the making can be observed in other parts of the great lowland-montane tussock-grassland formation, notably on the Mackenzie basin-plain (north of NO.) and near the source of the R. Awatere (NE.), so that the factors concerned stand out clearly enough.

3 3) That the special climate, above all else, is the deciding factor with regard to depletion, is further supported by the facts that in the arid area itself, as the rainfall gradually increases, so does depletion gradually decrease and that the sunny slope of even a shallow gully may be fully depleted while on the sunless slope there will be more or less tussock.

page 364

Finally, in the early "eighties" of last century, or a little earlier, the rabbit arrived from the lowlands and, as the supply of food1 became insufficient, crept higher and higher up the mountains. With an eminently favourable climate, abundant food, soil suitable for burrowing or rocks in plenty for their homes, these rodents increased almost beyond belief, so that with them and the many thousands of sheep the runs became greatly overstocked. Every plant at all palatable was eaten to the ground, the depleted area ascended higher and higher, those plants alone surviving which increased rapidly from seed, spread by means of underground stems, or were not eaten at all or only to a negligible extent.

At the present time, the scene is changed beyond recognition, instead of a close array of tussocks there is a man-made desert. With the tussocks are likewise gone nearly all the members of the ancient grassland. Viewed from a distance the mountains resemble giant sand-dunes. A close view shows, however, on sunny slopes a surface of bare, hard soil in places destitute of all visible plant-life except on rocks (Fig. 82), but usually having growing upon it at from about 60 cm. to several metres apart, that hall-mark of the rabbit, the hard, flat, silvery cushions or mats of Raoulia lutescens which are frequently 1 m. or more diam. (Figs. 59, 104). There are also more or less of the thinner, open mats of R. apice-nigra and occasionally those of R. australis, while, in the extreme west of the depleted area, the soft, silvery mats of R. Parkii abound and with them are many dull whitish-green, soft, hairy cushions of Pimelea sericeo-villosa and glaucous-green mats of Acaena Buchanani. Various plants, usually inconspicuous, are dotted about the otherwise bare ground, or epiphytic on the Raoulia cushions, especially the little tufted grass, Poa maniototo. the tiny hard brown cushions of Colobanthus brevisepalus and small brown mats of Stellaria gracilenta. Where conditions are slightly more favourable perhaps a dozen more species are present2, most of which along with

1 1) Besides the indigenous edible species, various exotic plants of higher palatability had become wide-spread, easily the most important of which are Rumex Acetosella and Erodium cicutarium. In spring, there are various annual grasses, especially Hordeum murinum and species of Bromus, several of which including Hordeum are eaten when young. Carduus pycnocephalus is extremely abundant but, as certain palatability experiments conducted by me proved, it is only eaten when there is little else; so, too with the abundant Cirsium arvense. Where a few tussocks of Poa caespitosa still persist on badly-depleted ground, notwithstanding the host of rabbits they remain untouched.

2 2) The slender Poa Lindsayi, a small tufted jordanon or epharmone of Danthonia semiannularis (eaten to the ground by rabbits but never killed), green mats of Carex resectans, occasional small rosettes of Geranium sessiliflorum var. glabrum, tufts of stunted Oxalis corniculata (in a wide sense), the small erect Hypericum gramineum, rigid open cushions of Hymenanthera alpina (Fig. 83) and the erect slender reddish Epilobium Hectori. On sunless slopes are many colonies of the exotic Cnicus arvensis. Rumex Acetosella is at home almost anywhere. On sunny slopes are Reseda luteola and Verbascum Blattaria both of which increase greatly from seed when protected from sheep and rabbits. Where there are loose stones the endemic far-creeping Urtica aspera is frequently abundant.

page 365those already cited would be either absent or rare in the original grassland. Where there is rather more rain various species of Acaena, particularly A. Sanguisorbae var. pilosa, are in extreme abundance and, on the less sunny slopes, Chrysobactron Hookeri in great quantity — an unexpected species which usually grows in very wet ground.

Near the upper limit of depletion (800 to 1100 m.), dead and moribund tussocks of Agropyron scabrum and Poa intermedia appear in quantity and, crossing this belt, the almost primitive subalpine tussock-grassland is entered with P. intermedia dominant.

In addition to the perennial steppe-vegetation there is a remarkable vernal florula1 made up largely of annuals, many of the members growing close together, so that for a brief period many more or less shady hillsides and the valley floors, desert-like for most of the year, become green and look quite fertile. If the flora of the whole Central Otago induced-steppe be considered, it consists of about 91 species (54 indigenous, 37 exotic) about 40 of which are common, but not more than 30, indigenous and exotic, occur on the worst depleted ground.

The evolution of induced steppe can be seen near the source of the R. Awatere (NE.). There, on the rather loose soil of gentle slopes (altitude about 800 m.), where there are thickets of Discaria toumatou, are many rabbit warrens, in the vicinity of which every stage2 of grassland depletion is present.

Celmisia spectabilis and Chrysobactron colonies.
These two communities, though of different life-forms, are taken together because both

1 1) The following are the principal species: — (exotic) Bromus hordeaceus, B. sterilis, Hordeum murinum, Aira caryophyllea, Festuca Myurus, Urtica urens, Gypsophila tubulosa, Cerastium vulgatum, Ranunculus falcatus, Draba verna, Erodium cicutarium (but of far wider distribution), Anagallis arvensis, Gilia squarrosa, Myosotis arvensis, Carduus pycnocephalus (but of far wider distribution); (indigenous) Chrysobactron Hookeri, Myosurus novae-zelandiae, Claytonia australasica (summergreen, grows on shady slopes in ground wet in winter), Myosotis pygmaea (a linneon, with 2 or more distinct jordanons amongst the Central Otago forms, and apparantly many hybrids between them).

2 2) Where the tussocks are close, bare earth dotted with rabbit-holes stands out conspicuously. By degrees, the tussocks are killed, not merely because they are eaten more or less, but through the intolerable manuring they undergo, and this area with holes and moribund tussocks is encircled by a wall of healthy tussocks. A little later, a temporary association occupies the new ground consisting of: abundant Rumex Acetosella, Geranium sessilifiorum var. glabrum (inedible), Epilobium novae-zelandiae (inedible), more or less circular mats of Acaena Sanguisorbae var. pilosa and A. inermis, Carex breciculmis, Cnicus lanceolatus, Stellaria media and Cerastium vulgatum. In course of time, as the depleted ground increases, there are established on its oldest part, their "seeds" wind-borne, low circular cushions of Raoulia lutescens between which, where fully exposed to the wind, no other plants are present, the dry soil having been blown away.

page 366arise from repeated burning and the reason for their remarkable increase is similar.

Celmisia spectabilis is common in dry mountain fell-field (SN., NE., E.) and forms large, green, circular semi-cushions made up of erect leaf rosettes crowded together. It also descends into the montane low tussock-grassland and was probably a rare member of the virgin community in its upper part. At the present time, in many places (E.) the tussocks are much stunted or eradicated and Celmisia cushions are everywhere at a distance apart of 2 m. or more.

On examining an area where burning has taken place it looks as if the Celmisia — its leaves burnt to their bases — was irretrievably damaged but not so. The leaf-buds are closely surrounded by the woolly sheaths of the burnt leaves and they also contain much moisture. Thus, the young leaves altogether escape damage from the fire and, in due course develope, so that, in quite a short time, the plant is a vigorous as ever. Further, as burning the tussocks leads gradually to extension of the bare ground, in which the Celmisia seed can germinate, in a comparatively short time C. spectabilis becomes the physiognomic plant, the open cushion replacing the tussock-form. Other species behave similarly, e. g. C. coriacea var. stricta in the Takitimu Mountains (SO.), Fig. 55, and C. dubia on the coastal mountains of the North-western district; indeed, wherever great fields of Celmisiae occur burning may be suspected and the community be indigenous-induced.

In the case of Chrysobactron it is its summergreen habit and tuberous roots which save it from grassland fires. Though naturally a plant of wet ground, as already noted, it can also grow under dry conditions, so it, too, can readily occupy bare ground resulting from burning tussock-grassland, while its rapid spread is furthered by its absolute unpalatability for any kind of grazing animal.

The incoming and spreading of Lycopodium fastigiatum.

At the head of Lake Wanaka (east of W.) there is a good deal of Lolium perenne pasture which has been laid down by aid of the plough. On what I feel pretty certain was an area of this kind — though unfortunately my notes say nothing on this head — Lycopodium fastigiatum has come into the community by means of its spores and, by its unexpected centrifugal growth, has formed huge "fairy rings" many metres diam. (Fig. 105). The plant extends by means of underground stems which branching put forth orange-coloured erect shoots 10 to 12.5 cm. high. Within the rings there is much dead Lycopodium; evidently as it exhausts the supply of nutritive salts in the soil, the stolons extend outwards into the new ground, the ring gradually increasing in diameter at the expense of the artificial pasture.

Between the rings there is a close turf of small indigenous plants common in tussock-grassland, together with others of more restricted page 367distribution, e. g. Herpolirion novae-zelandiae, Triodia australis, Poa pusilla, Acaena Buchanani, Stackhousia minima, Gentiana Grisebachii, Plantago triandra and Cotula squalida.

Invading shruh associations.

Montane low tussock-grassland, montane ploughed land (neglected), lowland artificial meadow and some other classes of vegetation are frequently invaded by different species of Cassinia. These are heath-like, erect composite shrubs ± 2 m. high, but much lower when growing closely. The seed is carried by the wind and by the sheep. As the species are of rapid growth, and seedlings abound, a closed association rapidly comes into being.

c. The swamp series.
The Phormium association.

In various parts of New Zealand proper, especially where large rivers overflow their banks, there are extensive areas of ground, perhaps extremely wet in winter, but quite dry in summer, occupied by bushes (tussocks) growing closely of Phormium tenax, making apparently a pure association. To one unversed in this class of vegetation, it would unhesitatingly be considered primitive, yet it has arisen solely through draining the Typha-Phormium swamp described in Part II. Nor would it remain stable for long were it not that various invading species are removed by the owners of the "flax-swamps", so called.

A has been explained, the primitive swamp is wet at all seasons and contains abundance of Typha angustifolia (in a wide sense) but, as for Phormium, it only occurs — more or less stunted — here and there in the shallowest water or fringing the margin of the swamp. Into the latter deep drains are cut, its water-content is greatly lessened, and in a year or two the Phormium present attains its full dimensions; then the subtratum becoming drier the Typha will languish and young Phormium plants come up thickly all over the drained ground. In a few years such attain their full dimensions and the former swamp will be almost pure Phormium 3.6 m. high, under favourable conditions.

In North Island "flax-swamps" the bushes about every 4 years are cut to within some 65 cm. of the base and as much as 100 quintals of leaves per acre are secured.

Generally, there are spaces between the bushes or groups of bushes; also paths are cut through the jungle of Phormium to enable the crop to be conveyed to the mill. Such spaces and paths are soon occupied by invaders — some exotic, some indigenous — but the latter. which form colonies of purely indigenous species, are alone considered here, though, it must be pointed out they have to compete with the eminently aggressive exotics, Festuca gigantea and Rubus fruticosus which form dense colonies. The indigenous invaders consist chiefly of the following: — (ferns) Blechnum procerum, Hypolepis rugosula; (grass-form) Arundo conspicua, Carex page 368secta; (herbaceous perennials) Epilobium junceum (in a wide sense), E. erectum, Eydrocotyle pterocarpa; (shrubs) Hebe salicifolia var. palustris (R. C.), Coprosma propinqua, C. robusta, ( C. prorobusta, C. tenuicaulis, Olearia virgata; (trees) Cordyline australis, Hoheria angustifolia', (lianes) Muehlenbeckia australis, M. complexa var. trilobata, Rubus schmidelioides var. coloratus, Parsonsia capsularis, Convolvulus Septum. The Cordyline is frequently dotted about the "swamp" and the woody lianes soon form dense masses; where the ground remains specially wet Typha abounds. Cattle pastured in the "swamp" help to dry and consolidate the soil, destroy the undergowth and make space for invading species.

Successions following felling of swamp-forest.

Cabbage-tree (Cordyline australis) swamp is now a common feature of lowland plains and valleys of North Island and has a truly pimitive aspect. But it has come into being after the destruction of Podocarpus dacrydioides forest and it is distinguished by the close groves of the Cordyline which are frequently accompanied by more or less Typha.

Divaricating-shrub swamp is a succession after the felling of Podocarpus dacrydioides forest when more or less of the undergrowth is left standing. Its composition depends, in addition, upon the incoming of species not present in the original forest, especially the following: — Arundo conspicua (wetter parts), Cordyline australis, Edwardsia microphylla, Hoheria angustifolia, Hebe salicifolia var. palustris (RC). The lianes — Muehlenbeckia australis, M. complexa var. trilobata, Rubus schmidelioides and Parsonsia heterophylla — become most luxuriant and bind the shrubs together, so that the thicket becomes impenetrable. Great tussocks, 3 m. high, of Gahnia xanthocarpa are abundant. Various divaricating-shrubs represent the original undergrowth especially: Paratrophis microphylla, Melicope simplex, Nothopanax anomalum, Coprosma rigida, C. tenuicaulis, C. propinqua, C. robusta (bushy shrub) and ( C. prorobusta.

An induced Phormium association has arisen at the base of Ruapehu (VP.) after the burning of the Dacrydium Colensoi association, and, without good evidence, such an origin could never be suspected. On the drier ground is Leptospermum scoparium thicket.

d. Some high-mountain indigenous-induced communities.
Hebe shrubland.

Probably a considerable percentage of Hebe shrubland has originated from the burning of various kinds of vegetation, notably, Nothofagus forest, tall tussock-grassland and subalpine-scrub. On Mount Dick (SO.) at 830 m. altitude Hebe buxifolia (in a wide sense) shrubland has originated most likely from burning forest and, at 1200 m., that of H. propinqua var. major from burning tall tussock-grassland. On Arthur's Pass (W.), at 900 m. altitude in 1908, several species of Hebe, with H. subalpina dominant, had taken possession of ground where scrub containing no Hebe page 369had been burned some 18 years previously. On Mount Egmont there is now pure H. salicifolia var. egmontiana (may = var. paludosa) where at about 1100 m. altitude the vegetation had been removed.

Induced fell-field or herb-field.

On Mount Miromiro near Hanmer (NE.), at an altitude of 840 m. on stony, wind-swept ground the Nothofagus forest on which had been burnt perhaps 20 years before, there is, as the dominant Helichrysum bellidioides, dwarfed to the ground and of raoulia form and other common members of the induced fell-field are Poa Colensoi, Holcus lanatus (exotic), Uncinia divaricata, dwarf Leptospermum scoparium, Leucopogon Fraseri, Cyathodes acerosa, Pentachondra pumila, Helichrysum microphyllum and Raoulia glabra.

By degrees as the altitude increases more typically high-mountain species come in, e. g. Geum parviflorum, Celmisia Traversii (a little), and a plant or two of the alpine C. sessiliflora. At a little higher altitude still, the following characteristic high-mountain species appear: Pratia macrodon, Raoulia bryoides, Celmisia Du Rietzii, C. viscosa, Helichrysum Mackayi and Leucogenes grandiceps.

At only 980 m. altitude, the species grow so closely that, in a sheltered gully, true herb-field appears. Here Celmisia Traversii — at at least 300 m. below its usually lowest range — is so plentiful as alone to catch the eye at first glance; there is also some C. coriacea and naturally some of the hybrid swarm between the two species of Celmisia — ( C. Morrisonii. Also the following are more or less common members of this remarkable community: — Lycopodium fastigiatum, Hypolepis Millefolium, Danthonia setifolia, Luzula picta, Astelia Cockaynei, Geum parviflorum, Acaena Sanguisorbae var. pilosa, Oxalis lactea, Epilobium chloraefolium var. verum, Drapetes Dieffenbachii, Anisotome aromatica, Hebe buxifolia (prostrate), Ourisia macrophylla, Coprosma ramulosa, C. parviflora, Brachycome sp. of the B. Thomsonii group, Hypochoeris radicata (exotic) and Traversia baccharoides. Here and there stand the burnt tree-stumps of the former forest. The gully lies right in the track of the north-west wind.

Induced cushion-plant association.

On the flat saddle connecting Mount Judah with Stone Peak (Richardson Mountains, SO.), at 1300 m. altitude, the primitive tall tussock-grassland has been burnt and is now replaced by a most unexpected type of vegetation. Dotted about everywhere are yellowish-green cushions of Phyllachne Colensoi, warm-brown mats of Dracophyllum uniflorum and straw-coloured tussocks of Festuca novae" zelandiae. In the adjacent unburnt grassland there is no trace of the Phyllachne but, in places, the Dracophyllum grows through it. The other principal species of the new association are: Lycopodium fastigiatum, Drapetes Dieffenbachii, Gaultheria depressa, Hebe dasyphylla, Ourisia caespitosa and Celmisia Lyallii.

page 370

5. Artificial communities

Under this head come all those communities — some designed to last for a considerable time, and others of brief duration — which are produced directly by man in his ordinary agricultural, forestry (so far as direct planting, sowing and thinning goes) and horticultural operations. Here only two classes of communities are dealt with, but they are of particular interest representing, as they do, agricultural proceedings unknown in Europe.

Displacement of rain-forest and swamp-forest by burning and replacement by artificial pasture without ploughing the ground.

Some 4050000 hectares of forest, the greater part in North Island, have been converted into grassland. For many years from 400 to 800 sq. km. were dealt with in each year but, at the present time, only a comparatively small area is available and the bush-buru days of New Zealand's pastoral development have almost passed away. The process of conversion of forest to grassland is of extreme phytogeographical interest since in one year's time a formation;, apparently attuned to a special habitat, is replaced by another ordinarily supposed to depend upon altogether different conditions, which more-over remains permanent so long, at any rate, as it is kept fully grazed. The processes involved consist of the following phases: — 1. Felling the forest. 2. Burning the fallen timber. 3. Sowing the seed. 4. Stocking the ground. 5. Burning the old and fallen logs. 6. Stumping the ground.

The forest is felled in winter and early spring. First, the undergrowth is cut and allowed to lie where it falls in order to provide the actual kindling-wood for the succeeding fire. Next, all the trees having a smaller diam. than 90 cm. are cut down, the others being left untouched, but the procedure differs in different localities and larger trees are felled on level ground than on steep slopes. Burning usually takes place after Christmas but the date depends entirely upon the state of the weather, since on a successful "burn" depends the future success of the subsequent operations. The following conditions are essential: — 1. the fallen trees &c. must be dry enough; 2. the weather must keep fine during the burn; 3. the wind must be favourable both in direction and intensity. The lighting of the fires takes place along as long a line as possible at right angles to the wind. Rapidity of burning is essential, so that on a large "block" as many as forty men may be required. Almost before the ashes have stopped smoking, and certainly before they are cold, the seed is sown.

The sowers sow by hand, carrying the seed1 in bags, which are ordinary

1 1) The seed is packed on horses to the ground from the nearest road in bags containing 35 kg., all the different seeds having been previously mixed together, and they are placed in position over the burnt area.

page 371sacks cut half-way down, a flap being thus formed in front, while a hole is cut in the back through which the man thrusts his head. The seed is scattered right and left, both hands being used, and, at each step, a handful of seed is thrown. The sowers form a diagonal line, so that one man slightly overlaps the work of another. The sight of the line of sowers crossing a log-strewn area where walking unburdened is no easy matter for a novice, each carrying a heavy bag and scattering the seed without cessation is not easily forgotten.

The amount sown per hectare varies from 20 to 30 kg. The following are the chief species sown: — Dactylis glomerata, Lolium perenne, L. italicum, Phleum pratense, Alopecurus pratensis, Poa pratensis, Cynosurus cristatus, Agrostis alba, Trifolium pratense, T. repens, T. hybridum. Other grasses (Festuca spp.) and clovers are occasionally used but the bulk of the seed consists of the first two species in the list1. A certain amount of rape, mustard and soft turnip is included in the mixture, so as to provide food for such stock as are turned on to the land within a few months of sowing.

Within 12 months from felling the forest the land is fully stocked and the trampling of the animals consolidates the ground and greatly assists in forming a sward. Where hilly, sheep are generally pastured, but the richer bottomlands are used for cattle.

By slow degrees, in process of time, the unburned logs decay or are burnt, the standing trees fall, and, if the ground is to be cropped, the stumps are extracted. At present, every stage of the conversion of forest into meadow is to be seen, but there are many areas where no vestige remains of the original plant-covering.

If the grass does not entirely cover the ground, certain indigenous shrubs may become abundant, especially Aristotelia serrata. Also, plants not present in the original vegetation may appear, especially Leptospermum scoparium and, near the coast, Cassinia leptophylla. Pteridium heath may also enter in, and were it not for abundant "stocking" would become permanent.

Replacement of Pteridium heath by artificial grassland ("Fern crushing").
As seen in Part II, there are large areas of Pteridium heath, much of which is certainly not primitive. Many areas of such heath have been converted into artificial grassland by a somewhat similar process as that just described. The heath ("fern" of the settlers) was burnt in the early autumn and from 15 to 30 kg. per hectare of grass and clover seeds, similar to those used on bush-burns, sown by hand. Between autumn and early spring little, if any, growth of Pteridium occurs, so time is afforded

1 1) In many localities Lolium perenne does not persist, so that eventually Dactylis and Trifolium repens dominate, and such form the basis of many meadows on soils which are "fertile", but over wide areas where lower "fertility" conditions were present the dominant species of grasses consist of such as were not originally sown but species of Danthonia (indigenous) or Agrostis (exotic).

page 372for the germination and establishment of sown grasses and clovers. But in the spring new leaves of Pteridium are rapidly developed and, if left to grow, the whole area during the one growing-season would be dominated by the fern and Pteridium heath hold sway, the young grass and clover being killed. However, as soon as the young leaves commence to develop and before their circinate tips have unrolled, large numbers of either dry sheep or dry cattle, depending upon which was the more economic, are grazed on the area and, to use the expressive term of the farmer, the fern is "crushed out". The stock are kept on the ground until lack of feed leads to considerable loss in body weight at which time they are removed. The area is then spelled until the fern begins to put forth fresh leaves when the crushing is repeated. According to the topography of the area, and the number of the stock available for crushing, in a variable time the fern becomes so weakened that the artificial grassland obtains the mastery and can be continuously grazed with that number of animals that, without falling away in condition, can make use of the feed produced. Not in every case, by any means, is success attained after a single burning of the fern, but, in certain instances, several resowing of grass seeds become necessary and in many cases this leads to the incoming of Leptospermum shrubland.

The underlying principle of fern-crushing is one that also is regularly adopted on bush-burn areas which are developing into induced Pteridium heath. In most cases, the application of crushing is successful, but in certain localities heavy crushing of such heath is followed by the establishment of circular sheets or deep mats of Paesia scaberula, and if the crushing is continued, the Paesia patches will coalesce and unpalatable species become dominant.