Aspects of the Natural Regeneration of the Kauri (Agathis australis Salisb.)^*

R. V. Mirams

Climate

For the five years 1944–48, the average annual rainfall at three of the Waitakere stations was.

Waitakere 70.22 ins (178.5 cms)
Nihotupu 90.53 ins (230 cms)
Huia 84.66 ins (215 cms)

For comparison, the average rainfall in Auckland City for the same period was 51.44 ins (130.5 cms).

The rainfall everywhere in the ranges is high with quite an appreciable variation from station to station. The fall is fairly uniformly distributed throughout the year, although heavier during the winter months.

Measurements of temperature and relative humidity were made in the Leptospermum scrubland and in the kauri forest for a period of about fifteen months during 1949 and 1950.

During January and February, maximum temperatures as high as 80° F. were quite frequently recorded, but from May to August the maximum was often below 60° F. The minimum temperatures do not show these seasonal variations to such an extent. In the summer period the daily range of temperature is as much as 25° F. falling to 10°–15° F. in the winter.

The main feature of relative humidity readings was the long duration of high humidities in the scrub and forest. The relative humidity is usually at a level not far from saturation for many hours each week, it also does not often fall below a value of 50% (a level which is significant for the seeds to remain viable).

The daily temperature and humidity curves show an almost exactly opposite variation. The lowest temperature is usually reached just before sunrise, and from that time onward there is a rise in temperature over a period of four to five hours. The maximum value for the day is attained at noon or soon afterwards, and may last, with minor fluctuations, for periods of up to four hours, or it may commence to fall again almost immediately. The rapidity of this rise and fall of the temperature and the range is dependent on the season. During overcast and wet weather, there is a tendency for the temperature to remain at a fairly constant level for several days. The basic form of the diurnal variation curve is recorded, although it is very much flattened. The relative humidity is at its highest levels during the night hours, and there is a decrease as the temperature rises, often as much as 30% in one hour. The lowest reading for the day is during the noon hour. As with the maximum temperature, this minimum relative humidity can be either of short duration (most usual) or it may persist for some hours, with minor fluctuations. From the minimum level there is a steady rise until the maximum value near saturation point is attained In damp weather and when rain is falling, the humidity tends to remain static, in sympathy with the temperature at a uniformly high level. This phenomenon is more pronounced in winter and spring with their frequent rainy days than at other periods.

The regular daily patterns of temperature and relative humidity are essentially similar in the Leptospermum scrubland and in the kauri forest.

Soils

The underlying rock in the ranges is andesitic (basic), and this has given rise in most areas to soils of the brown granular clay sub-group of the brown loam group. Except for two very small areas of recent soils developed from alluvium, the remainder of the region is covered with soils belonging to the skeletal group (Grange-Soil Map of N.Z. Sheet 2. 194?). Wright, (1951) has pointed out that long cycles of kauri on an area will eventually lead to the impoverishment of the soil and the formation of swamp and barren “gumland” communities. This stage of development does not appear to have been reached in the Waitakere Ranges.

Part I.—Quadrat Investigations of Natural Regeneration

In the Waitakere Ranges, all degrees of inter-mingling of the forest are to be found from a pure kauri stand to a type of forest dominated mainly by species of Podocarpus and Dacrydium, the broad leaf, Beilschmiedia tawa not being common. These are some of the members of the ultimate climax, but the forests in this region do not appear to have reached a great degree of stability.

It is obviously impossible to follow a single succession from its initiation to the ultimate climax, and resource has therefore to be made to the study of a number of examples of the same succession at different developmental stages contemporaneously. Similarly it is not possible to see kauri forest initiation to-day. The commencement of secondary forest can, however, be observed in the natural regeneration which usually occurs after the forest has been destroyed or damaged. It is realised that the sequence of events which we see to-day giving rise to a new kauri forest may be, and probably are, vastly different from the succession which gave rise to the kauri forest which we now see and which is dominated by trees that are many hundreds of years old.

community and Agathis is certainly not doing this. It might be argued that there are sufficient individuals in class IV allowing for a considerable mortality, to maintain the stocking of class VI at its present level. However, the forest is surely old enough for individuals of Agathis intermediate between class IV and class VI to have developed and the fact that they have not already done so indicates that they may not under the present environmental conditions. At the same time, it does not seem possible to regard the members of class IV as potential replacements of the present mature individuals. The greatly reduced occurrence of seedlings of Agathis on quadrats within the mature forest is harder to explain for this community than for any of the others. Collections of seed by means of traps have shown that there is more than sufficient seed falling to the ground in one season to produce the number of seedlings of all ages recorded, let alone those which are only in the cotyledonary stage (size-age class I). The activities of at least one insect are responsible for the destruction of a large proportion of the seed; this does not explain, however, the complete absence of one of the size-age classes and the small number of individuals in the remainder.

Previous Views on Natural Regeneration

Although there is quite a volume of literature bearing upon the kauri, one finds on examination that most of it is in the broadest of terms. As far as the occurrence and distribution of natural regeneration is concerned, we have such statements as the following. Cockayne (1908 and 1928) says that “he notices usually nothing between seedlings and old trees. Whereas outside the forest in the manuka scrub kauri can be seen coming up in abundance and here trees of all ages are seen.” Cheeseman (1914a) also mentions “the comparative absence of young kauris in a mature kauri grove.”

Hutchins (1918) states that it is evident that the tea-tree can act as a nurse for the regeneration of kauri and the forest generally.

The phenomenon of succession in the New Zealand forests has been recognised from the earliest days. Cockayne (1928), while recognising the fact, does not regard the kauri forest as forming part of the climax community but considers that it will be replaced by a community dominated by broad-leaved trees, perhaps Beilschmiedia taraire, B. tawa or Weinmannia sylvicola.

One of the tests for a climax community is that it must be capable of reproducing itself, since it represents the last stage of succession so long as the climate remains the same. The climax must therefore be a stable community, in which the oldest individuals, dying, are replaced by their own kind, leaving the character of the community unchanged. As will be shown later, the kauri forest does not appear to be reproducing itself and it cannot therefore be considered to be the climax. It might be a subclimax community of long persistence apparently having a dominant similar in life form to that of the ultimate climax. It is also possible that the kauri forest may be a relict or post climax and that the few areas now remaining are survivals of once more extensive kauri forests. Wood indistinguishable from that of the present day Agathis australis has been found in Miocene coals in the south of the South Island (Evans, 1934).

It is known from the work of Cranwell and Von Post (1936) that climatic conditions in the Post-glacial were, at times, more favourable than they are now, and it is possible that during one of these Post-glacial warm periods in addition to one in earlier times, Agathis extended over more than its present range. It is unfortunate that kauri pollen does not preserve well (Cranwell, 1940). The kauri will today grow in gardens and parks as far south as Invercargill, possibly because of the lack of competition such as would occur in an integrated community.

Finally, there is the possibility of a cyclic complex such as is mentioned by Watt (1947), in which one community having developed to the maximum is replaced by

another which has grown up within it, it, in its turn, being replaced. It is suggested that the kauri forest alternates with and is replaced by a community dominated by some or other of the podocarps (Podocarpus and Dacrydium spp.) and broad-leaved species (Beilschmiedia and Weinmannia). Forest dominated by these species occurs intermingled with the kauri forest, there being a range from a pure podocarp-broad-leaved-community through a mixed podocarp-broad-leaved-kauri community to a pure stand of kauri.

The fact that the kauri forest may not be the climax community may possibly explain the apparent non-replacement of the species and the small number of individuals to be found in the mature forest as compared with the other successional phases. The early communities are on the upgrade while the kauri forest is apparently degenerating. The fact that Agathis does regenerate in the Leptospermum communities under the present climate indicates that the cause of its degeneration may be in its actual constitution and its effect on its own internal environment. All of the regenerating communities examined are on the sites of former kauri forests, and this is further evidence in support of the above conclusions.

Discussion

It has been shown that natural regeneration occurs more abundantly in the tea-tree communities than in the kauri forest. Not only are more individuals of Agathis found in the seral communities with the exception of the young Leptospermum, but in the mature forest there is a vast gap between the largest sapling, which is never more than two to three centimetres in diameter, and the adult trees one to two and more metres in diameter.

There must be a number of reasons for these differences in regeneration in the various communities. There is no evidence to show that seed supply is a factor limiting regeneration, the seed which germinates forming only a small fraction of the total which falls, even in the Leptospermum, where it must be transported often over considerable distances. As far as the kauri forest is concerned, although there are few seedlings compared with the number of seeds which were collected, it appears that the failure of regeneration in this particular community is in the subsequent lack of growth of small individuals of Agathis into saplings and small trees.

If the lack of seed is not the cause of the failure of natural regeneration in the kauri forest, we must look for differences in some of the factors operating in the various communities, or else in some variation brought about by the plants themselves. Variations in the site, aspect, topography, climate, soil, etc., cannot be ignored in attempting to explain the differences in the occurrence and distribution of natural regeneration. As far as their direct effect is concerned the differences, from the available data, are not sufficient to account for these wide variations in the numbers of individuals in the various communities. Light (a factor influencing the growth of Agathis but not directly analysed), while appearing to have no effect on germination, may be insufficient in the mature forest for the optimum growth of small individuals of kauri. A supposedly light-demanding tree such as the kauri cannot be expected to grow very successfully beneath a heavy canopy, formed in this case of individuals of the same species.

Root competition is another factor which may be responsible for the poor growth of kauri saplings in the kauri forest. McKinnon (1945) mentions it as a factor and says “kauri seedlings on a quadrat whose boundaries were trenched to below the general root level have shown a significant improvement in size and vigour as compared with an adjoining control quadrat”. The time between trenching and the publication of the above report was approximately three and a-half years.

It is interesting to note that remnants of former kauri forests, stumps, etc., have been found in many of the tea-tree communities investigated. This would suggest that the differences in regeneration are to be found in the dynamics of the various

communities and are not so much a result of physiographical and physical differences, except in so far as these may be modified by the present plant cover. It is possible that there are actions and reactions between the various plant species and combinations of species that may have an effect which is responsible for some of these differences. There is a great difference in the structure and physiognomy of the communities studied, but many, if not all of the characteristic species are to be found throughout the succession. There is some evidence from individual quadrats that the effect of one plant on another may be responsible for some of the differences of regeneration; in some cases the small number of individuals of Agathis and the reduced number of other species is due to the presence of one species tending to produce conditions not entirely favourable. The effect of these species—e.g., Gleichenia circinata, Schoenus tendo, Cyathea dealbata (leaf debris) and Freycinetia banksii is to produce a blanket-like layer of living and dead plant material over the surface of the soil which forms an unsuitable seed bed.

Sando (1936) lists the following as possible causes of the failure of kauri to regenerate:

• (1) Excessive leaf litter and dense floor covering of “kauri grass”.

• (2) The dense shade produced by Beilschmiedia taraire and Dysoxylum spectabile and also the dense undergrowth of Cyathea medullaris and C. dealbata on shady faces in old workings.

Cockayne (1928) and McKinnon (1945) also mention the aggressiveness of certain floor plants, among which are some of those listed above.

A further point which must be considered is whether Agathis australis is the dominant of the climax community. It is suggested that it is not and that it may either be a subclimax, a relict or post climax, or one of the communities of a cyclic complex. In each of these cases, there must be a replacement by or alternation with another community, and it seems probable that once the kauri forest has reached its full and characteristic development it commences to decline slowly within the succession prior to its replacement.

The fact that Agathis does regenerate in the Leptospermum communities under the present climate indicates that the cause of the degeneration of the present kauri forest is rather in the actual constitution of the forest community and its effect on its own internal community environment.

Part Two

In order to make use of this value as applied to the dispersal of seeds in the forest, a measurement of wind speed in the forest canopy was necessary. This was carried out by means of a Casella-Robinson three-cup anemometer. A pulley was attached at a height of 23 metres (80 feet) to one of the upper limbs of a large kauri. By means of a cord passing over this pulley it was possible to raise and lower a small wooden platform to which the anemometer was attached. The anemometer is an integrating instrument and only average values of the wind speed could be obtained for a given period. Measurements on any one day were taken over a total period of up to about seven hours, during which time the instrument might be read and reset to zero six or more times. The average wind speeds for these short daily periods were computed; the range of these speeds is shown in Table III together with the average wind speed for the total daily period. Observations were made for a total period of more than sixty-three hours spread over twelve days during the period of seed fall.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Table III.—Wind Speed Readings.

Date.	Period of Observation Minutes.	Range of Wind Speeds Feet per second.	Averag Wind-speed for Period Feet per second.
14/2/50	255	5.4 –5.45	5.4
15/2/50	236	3.6 –5.30	4.6
29/3/50	323	2.55 –4.25	3.3
31/3/50	432	3.1 –5.6	4.5
1/4/50	411	0.4 –2.3	1.8
2/4/50	346	5.4 –7 7	6.5
3/4/50	390	3.6 –4.3	3.8
4/4/50	420	1.7 –2.75	1.9
5/4/50	129	2.3	2.3
18/4/50	180	1.6 –4.1	2.8
19/4/50	390	0.3 –3.0	2.3
20/4/50	390	1.7 –3.0	2.3

The average wind-speed for the total period of the observations was 3.5 feet per second; for 54.3% of the time of the observations the average recorded wind-speed was below 3.5 feet per second. The wind conditions within and without the canopy layer must be vastly different, but it is considered that on every day there would be winds of sufficient velocity to transport the seed over some distance. The average wind-speed recorded in the forest canopy is almost identical with that found from the wind-tunnel as being of sufficient velocity to keep the seed afloat. Baldwin (1942, p. 24) says “that the distance that wind may carry the seed is conditioned very largely by the rate of fall”. The lack of high enough wind speeds in the kauri forest does not appear as if it would be a limiting factor in the carriage of kauri seed, although how far the seed will be carried is difficult to estimate.

Germination

Several writers in the past have expressed conflicting opinions on the longevity of kauri seed. Some, Sando (1936), McKinnon (1937b) and MacMorran (1946) saying the seed does not remain viable for much longer than one year under normal storage conditions. Hutchins (1918) on the other hand says that “the predominant opinion is that kauri seed may remain dormant for as long as thirty years”.

An attempt was therefore made to discover what time the seed will remain viable under various storage conditions and also to what extent various factors affect germination.

The “cutting” test as a means of ascertaining the proportion of sound seed in a sample has already been mentioned. It should be emphasised that the cutting test is not a measure so much of the germinability of seed but that the seed has been fully formed.

When this series of kauri seed experiments was commenced in 1948, germinations were carried out using nursery boxes with soil as the medium. A number of the boxes were in an unheated glasshouse, the remainder being in the open and shaded to various degrees.

Watt (1919) working on the English oakwoods, found that there were considerable variations in the rate of germination of the acorns when they were placed in different positions. It was thought that the position in which the kauri seed was placed in relation to the soil surface might have some bearing on germination.

Significantly better germinations were obtained in two cases. In 1949 and 1950, germination experiments were transferred to the laboratory, the seeds being sown in sand, in petrie dishes placed in incubators at 24° C%. The seeds were placed in the sand with the wing in the air, the lid of the petrie dish being raised by means of a square cover slip in order to ensure better air circulation. Good results were obtained using this method, and all of those which follow were obtained by using it.

Age of Seed and Retention of Viability

Rapid viability tests are of considerable importance where time does not permit of lengthy germination tests. In this series of experiments, use has been made of the salts, sodium biselenite (NaHSeO_a) and 2.3.5 triphenyl tetrazolium chloride. Viable tissue placed in solutions of these substances will, under suitable conditions, assume a red colour, its intensity and extent depending on the degree of viability of the material under test.

The technique for both salts was as follows: the seeds to be tested (one hundred sound ones in each case) were placed in a beaker containing distilled water and then put overnight into an incubator maintained at 24° C%. This preliminary soaking facilitates removal of the embryo, and in no way affects the results. The embryos were then excised and placed in the test solution for a period of up to eight hours, again at 24° C%. After this time the embryos were washed in water and sorted into the following categories:

• (1) at least half of the embryo intensely coloured.

• (2) the embryo showing a spot of colouration.

• (3) no colouration.

• (4) embryos obviously dead (decayed, etc.) on excision, and therefore not soaked in the test solution.

In the case of sodium biselenite, a 2% aqueous solution was used. The method is essentially a modification of Eidmann's as given in Baldwin (1942, p. 183).

The tetrazolium method was originally described by Lakon (1942) for cereals but the paper was unobtainable and a re-description of the method in English (Lakon, 1949) was seen only after most of the work described had been completed and use had therefore to be made of other methods of embryo classification. A 1% aqueous solution of 2.3.5. triphenyl tetrazolium chloride was used.

Within the limits of the apparatus available, as many replicates as possible were used in each series of experiments; in a few cases it was not possible to have any. In some of the later experiments, germinations indicated that the single sample result was probably correct, but in one or two cases further replicates were obviously necessary.

The course of germination is shown in Figure 1. There is a steady fall in the germination of kauri seed from 100% at the time of collection to only 5% in mid-July, four and a-half months later; and by mid-November, it is down to less than 1%. The results obtained by the rapid viability test methods, although somewhat higher, show the same trend as the actual germinations. It is evident from these results that the viability of kauri seed drops rapidly when the seed is stored under ordinary atmospheric conditions.

received all of its water at once dried out steadily during the course of the experiment, and at the conclusion was in fact drier than any of the other dishes. Seeds which were slow to start germinating under these conditions would thus encounter less favourable conditions towards the end of the experiment than those in dishes which had a more or less continuous though perhaps limiting supply of moisture during the whole period. The results of the second experiment are somewhat anomalous in that those seeds which received all of their water supply at once had as low a germination as those which received it throughout the course of the experiment. There has also been a drop in the overall germinative capacity of the seeds, and this may also have contributed to the result obtained. It was impossible to have any replicates in this series of germinations; the results obtained indicate their great desirability.

The conclusion to be drawn when the above factors are taken into account is that, as far as kauri seeds are concerned, moisture supply appears to play an important role in the number of seeds germinating and that an adequate supply is apparently essential at all stages.

An interesting observation was made in 1948 before these experiments were commenced. Seed sown in the field, fully protected from predators, during early April, following a very dry summer together with further dry weather after sowing, resulted in few seedlings. At the time it was thought that this was due to the low moisture content of the soil, and in view of the above results, this is probably the case, although doubtless other factors were also operative.

Animals and Kauri Seed

It had long been suspected by the writer that some animal was destroying large numbers of the sound seeds before they were able to germinate. In order to find out the nature of the animal, seed was germinated in the field both with and without the protection of wire gauze cages which excluded all animals except subterranean ones. Studies were also made with seeds which had been marked on the wing with aluminium paint.

Birds were early eliminated as the predators; the mesh of the wire covering on the cages being small enough to exclude them. A search had, therefore, to be made for some lower organism as numbers of the seeds were eaten beneath the cages.

Small quantities of these marked seeds were placed in many locations in all of the communities studied. The seeds showed on re-collection and examination that most of the sound seed was eaten during the first night, or at the latest during the second night. The blind seed did not appear to be touched in any way at all. The seed which had been eaten was scattered over an area little greater than that it originally occupied and was all eaten in a characteristic manner. It had the appearance of having been crushed by a pair of powerful jaws and the oily contents then eaten. A number of metal pipes three inches in diameter and approximately fifteen inches in length were sunk into the soil in one of the Leptospermum communities for the purpose of trapping insects. These traps proved efficient in use, and a number of wetas (Hemideina thoracica) were caught. The weta is a large orthopterous insect with heavy mandibles; a number were taken into the laboratory and left overnight with kauri seed. This seed was found to be eaten in the same manner as that recovered in the field.

The attacks of leaf-eating insects on seedlings have been mentioned by Millar (1925); damage caused by leaf-eaters has been observed on a number of kauri seedlings, in no case is it considered likely that they would succumb. One of the native parrots (Nestor occidentalis) has been reported by Oliver (1930) to feed on kauri seeds, using its strong beak to break open the cones. Sando (1936) thinks that the attacks of the parrots are a factor in causing the premature fall of kauri

cones. These parrots have been observed in the Waitakeres, but the present writer is not of the above opinion.

One can also mention here the effect of an unidentified fungus which causes the “damping off” of considerable numbers of young seedlings soon after germination. The usual symptoms are in the nature of collar rot. Sando (1936) mentions a fungal infestation of immature or unopened cones, the organism responsible being Pestalozzia funerea (Fungi imperfecti).

The effect of the weta eating the seed is considered to be important as a biological factor in natural regeneration, with next in order the collar rotting fungus. The actions of birds attacking the cones and insects the seedlings appear to be of much lesser importance.

The action of man as a biotic agent must also be mentioned. Vast areas of forest in New Zealand have been destroyed by man's agency, and it is in fact through the direct action of man that we find the numerous Leptospermum thickets in various parts of the country. Man by his very acts of destruction, is causing the devastated areas to regenerate and, as is shown in the present paper, the kauri forest is regenerating by the natural process of succession in many of these areas. Dansereau (1951) says “the mere breaking up of the continuity of certain communities can be genetically harmful, even in the case of climax associations. There is reason to suspect that such is the case of the kauri (Agathis australis) forest of northern New Zealand.”

The Survival of Small Kauri Seedlings

Observations were made on the survival of small seedlings in the young Leptospermum, Leptospermum-Agathis and in a ricker community. Counts of the individuals were made at more or less monthly intervals. (See Table V.)

Mortality is high soon after germination, at this stage the unidentified collarrotting fungus is the causal organism. A slight increase in numbers at this time is probably due to individuals that have come from seeds that were slower to germinate. During the winter months there is usually a steady fall in the number of individuals until the period of high temperatures and low soil moistures in late December, January and February, when there is a second, usually quite sudden, drop in numbers. The death of the seedlings in this case is apparently due to the high temperatures together with the low moisture supply causing them to die from wilting. The older seedlings are apparently just as liable to succumb to these summer temperatures as those which are younger.

A comparison between the survival of young seedlings and soil moisture content during the summer months shows the relationship of low soil moistures and the seedling deaths Sando (1936) mentions the leaf-litter as a factor preventing regeneration, also McKinnon (1945) who says “the depth of the litter layer and the fibrous humus layer of the soil leads to heavy mortality amongst seedlings during prolonged dry weather, roots not having penetrated to the A horizon of the soil.”

The following table has been constructed in an attempt to relate all the available field information. A comparison between the number of size-age class I seedlings on an area of one thousand square metres and the number of sound seeds which could be found on a similar area indicates that few seeds can germinate and the seedlings survive for one year. The major portion of this difference can possibly be attributed to the weta destroying a large number of the sound seeds, thus preventing their germination; low soil moistures at the time of seed-fall and high relative humidities causing loss of viability are further factors affecting the germination of the seed under field conditions. After germination, some of the seedlings die by the action of a collar-rotting fungus, and when somewhat older low soil moistures are a cause of death by wilting. Numerous other agencies and combinations of such agencies must also take their toll An exact comparison between the amount of seed estimated to be on the ground and the number of seedlings after, say, one year can only be an

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Table V.—Showing a Comparison of Various Field Data.

	1	2	3	4	5	6	7	8
Community.	Average No. of individuals of Class 1 per 100 sq. metres from quadrat data).	Average No. of sound seeds per annum per 1000 sq. metres (from seed traps).	Per cent. germination Columns 1 and 2.	Average No. of individuals of Class II per 1000 sq metres (from quadrat data)	Observed per cent. survival from first to second year. (Based on field counts.)	Estimated No. surviving from first to second year. (Columns 1 and 5.)	Calculated No. of individuals of Class II expected after six years (from Column 6).	Per cent. survival comparing Columns 4 an 7.
Kauri forest	236	58,000	0.407	296	—	—	—	—
Ricker	252	39,000	0.646	627	38.3	96.5	579	92.4
Leptospermum-Agathis Young	109	5.500	1.98	542	88.2	96	576	>100
Leptospermum	14	4,000	0.35	146	34.1	4.76	29	19.5

approximation; nevertheless, some interesting points emerge. From field counts on the survival of seedlings a percentage survival from size-age class I to size-age class II was worked out, and this has been applied to the actual number of class II seedlings recorded from the quadrats. The figure obtained was then multiplied by six, which is thought to be the maximum age of this class, no account being taken of earlier deaths apart from the fact that only a percentage of the actual number of class I seedlings is used as a basis for the calculation. It has already been said that natural regeneration is at its best in the old Leptospermum community, and this fact is supported by the above data. The percentage germination in the old Leptospermum is almost six times that in the preceding young Leptospermum and considerably more than that in the ricker and kauri forest communities. This greater germination must be brought about by the occurrence of environmental conditions more favourable to germination at this stage in the sere. The percentage of seeds germinating in the kauri forest is low, almost as low as in the youngest Leptospermum, where the seed source is by no means as close as it is in the kauri forest. In the transition of seedlings from less than to more than one year old the actual per cent. survival is highest in the old Leptospermum. The number of seedlings calculated to be present in the ricker community after six years is almost the same as the number actually present. Before any definite conclusions can be reached much more long term information is required.

Discussion

It has been shown that kauri seed is able to germinate under a wide range of laboratory conditions. Under natural conditions, however, before the seeds are able to germinate they are subjected to the attacks of an insect which destroys a large proportion of the sound seed before it has a chance to germinate, thus very considerably reducing the number of seedlings which may eventuate from a seed-fall.

Another factor of importance is the relative humidity of the asmosphere in which the seed is present. Kauri seed stored in the laboratory under ordinary atmospheric conditions loses its viability rapidly, and it has been shown that one of the factors concerned is a high humidity in the storage atmosphere. The best germinations were obtained from seed which was stored at constant relative humidities of 50–65%. These levels are much below those recorded in the forest communities, where there is almost full saturation for many hours each week. It is therefore evident that the seed must germinate soon after reaching the ground or else it will lose viability because of high humidities. There must also be a ready supply of moisture available when it reaches the ground or else germination is unable to proceed. Field observations of soil moisture content indicated that the surface soil horizons are driest at

Acknowledgments

The writer is pleased to acknowledge the assistance given to him by Professor V. J. Chapman. He would also like to thank Messrs. J. A. McPherson and A. D. Mead, of the Auckland City Council, for the assistance which they and their departments gave to him. Finally, he would like to thank his various friends, both staff members and students mainly from the Botany Department, Auckland University College, for their help in various ways.

Bibliography

Baldwin, H. I., 1942. “Forest Tree Seed.” Waltham, Mass.

Cheeseman, T. F., 1914a. “Illustrations of the N.Z. Flora,” Wellington.

——, 1914b. “Age and Growth of the Kauri,” Trans. N.Z. Inst., 40, 9–19.

——, 1925. “Manual of the N.Z. Flora” Second Edn. Wellington.

Cockayne, L., 1908. “Report on a Botanical Survey of the Waipoua Kauri Forest.” Wellington.

——, 1928. “Die Vegetation der Erde” XIV. The Vegetation of N.Z. Second Edition, Leipzig.

——, 1932. “Polymorphy in N.Z. Conifers and its relation to Horticulture,” reprinted from “Conifers in Cultivation,” Report of Conifer Conference 1931, 151–164.

Cranwell, L. M., 1938. “Fossil Pollens,” The Key to the Vegetation of the Past. N.Z.J. Sci. & Tech. 19, 628–45.

——, 1940. “Pollen Grains of the N. Z. Conifers.” Ibid 22, 1B–17B.

—— and Moore, L. B., 1936. “The Occurrence of Kauri in Montane Forest on Te Moehau.” Ibid 18, 531–43.

—— and Von Post, L., 1936. “Post Pleistocene Pollen Diagrams from New Zealand” Georgrafiska Annaler 18, 308–47.

Dansereau, Pierre, 1951. “Description and Recording of Vegetation Upon a Structural Basis.” Ecology 32, 172–229.

Eames, A. J., 1913. “The Morphology of Agathis australis,” Anns. Bot. 27, 1–38.

Evans, W. P., 1934. “Microstructure of N.Z. Lignites, Part III,” NZ.J. Sci. & Tech. 15, 365–85.

——, 1937. “Note on the Flora which yielded the Tertiary Lignites of Canterbury, Otago and Southland.” Ibid. 19, 188–93.

Foster, F. W., 1930. “A note on the Height Growth of Kauri Seedlings and Small Saplings.” Te Kura Ngahere 2, 34–5.

Foley, T. A., 1950. “Rate of Fall of Kauri (Agathis australis Salisb.) seeds in quiet air.” Forest Res. Notes 1, 32–42.

Cornier, B. J., 1950. “The Climates of N.Z.: according to Thornthwaites Classification,” in N.Z. Weather and Climate, N.Z. Geog. Soc. Misc. Publn. No. 1.

Harrison-Smith, J. L., 1938. “The Kauri as a Host Tree,” N.Z.J. Fort. 4, 173–77.

Hochstetter, F. Von, 1867. “New Zealand, its Physical Geography, Geology and Natural History with special reference to the results of Government Expeditions in the Provinces of Auckland and Nelson.” Stuggart.

Hocking, G. H., 1935. “A Note on the germination of some Native species.” N.Z.J. Fort. 3, 225–27.