A very brief description of the subsere and the kauri forest will suffice, as these various communities have been adequately described by Cockayne (1928). The succession of events is as follows. After the short-lived initial communities of grasses, and herbs there develops a tea-tree or manuka scrub. It is a typical shrub community with Leptospermum scoparium—the dominant—present in considerable numbers. In the early stages the ground is covered with a mass of Gleichenia circinata. Beneath this fern, which is often one metre and more deep, are to be found many of the forest tree seedlings, including Agathis. With the passage of time the fern dies out, Leptospermum scoparium grows larger, the allied Leptospermum ericoides enters the succession and becomes a co-dominant. Many individuals of the numerous species which go to make up the kauri forest are now seen as small saplings growing beneath the tea-tree canopy. L. ericoides is usually by now completely dominant, but it is eventually over-topped by individuals of Agathis which at this stage have a characteristic conical form. (Plate 43, fig. 2.) The Leptospermum ericoides persists for a considerable period, first as a co-dominant but later as a subdominant, and there comes into being what may be called “a forest of kauri rickers”. These rickers, which often form dense stands, increase in both diameter and height until eventually there is developed a community, little different from the mature forest. The main difference is the smaller size and very much larger number of individuals of kauri in the former type. The ground is by now covered with a heavy tussock-like growth of “kauri grass”, Gahnia xanthocarpa and Astelia trinervia, which frequently reaches a height of two to two and a-half metres. (Plate 44, fig. 2) By the time the mature kauri tree has developed, it has become the only dominant. Even in a dense stand, the trees are not close together but may be separated by distances of fifteen to twenty metres. There are also a number of smaller trees of other species but only one, Knightia excelsa, reaches into the sub-canopy layer. The canopy in the pure stand is quite irregular, being formed of the large rounded heads of the mature kauri tree. (Plate 43, fig. 1.)

The occurrence and distribution of natural regeneration was investigated by a series of quadrats laid down in the various successional stages, five types being recognised.

• (1) Young Leptospermum.

• (2) Older Leptospermum (usually lacks the ground layer of fern and rush found in the first type). (Plate 43, figs. 3, 4.)

• (3) Leptospermum–Agathis community.

• (4) Kauri ricker community. (Plate 44, fig. 1.)

• (5) Kauri forest. (Plate 44, fig. 2.)

A quadrat size of 36 sq. m. was decided upon from the species-area curve. Myers (1948) in a more comprehensive study of minimal area in Leptospermum shrubland also found that the minimal area for this community was 36 sq. m.

On every quadrat a list of all the species encountered was made. In addition, counts of all the individuals of Agathis were made. These individuals were classified into one of the six size-age groups listed below in order that a better picture of regeneration could be obtained. (Plate 45.)

Class I

Small seedlings with cotyledonary leaves only. The majority of these individuals are from the current year's seed fall, but retarded individuals are occasionally found which have come from a previous fall.

Class II

Seedlings with up to about twelve leaves and not more than ten centimetres in height; branching of the stem not apparent.

Class III

Seedlings between ten and fifty centimetres high, with a number of branches both lateral and terminal.

Class IV

Slender saplings with a stem not more than five centimetres in diameter. The height varies from fifty centimetres to four metres or more.

Class V

Saplings and small rickers five to fifteen centimetres in diameter.

Class VI

All individuals whose stem is more than fifteen centimetres in diameter.

A total of eighty-four quadrats were laid down, sixty of which were thirty-six square metres in area, the remaining twenty-four being one hundred square metres. The number of species occurring in the sixty smaller quadrats are eighty-six, with the addition of the twenty-four larger quadrats a further ten species were added, making a total of ninety-six.

Natural regeneration is considered first as the total number of individuals of Agathis in each of the five successional types, and secondly as the distribution of the size-age classes in these communities. In both cases, the data is taken from twelve 36 sq. m. quadrats in each of the five community types.

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

A second set of quadrats 10m. × 10m. provided the following results, which are very similar to those obtained from the smaller quadrats, apart from the larger numbers of individuals.

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

It can be seen from the above table that there is considerable variation in the total number of individuals of all size-age classes in the various communities examined. The smallest number of individuals is found in the young Leptospermum, the highest at the middle stage of the succession. The number of individuals in the kauri forest is also relatively low, much lower than would be expected even for the class I individuals. There is adequate seed available for the germination of many more seedlings.

In the young Leptospermum it is only to be expected that the density of Agathis will be low, since invasion by this and the other forest species has only just commenced. This fact was reflected in the absence of kauri from almost half of the quadrats and the small numbers on several of the remainder. The absence of individuals in classes V and VI is also due to the early age of the community at this stage. By the time an older Leptospermum community has developed, there has been a considerable increase in the number of individuals of all age classes, especially classes II and III. There has been a slight increase in the numbers in class IV, and class V is now represented by a solitary specimen. The increase in total numbers is probably due as much to the “accumulation” of seedlings as to the operation of any factor more favourable to their growth. In spite of the increase in numbers there is evidence to show that there is a considerable mortality of young kauri seedlings; these results only show the “end point” of their survival at the time of counting, they do not show what the trends in survival are.

The next stage of the sere is an ecotonal community in which Agathis becomes codominant with the giant tea-tree Leptospermum ericoides. The most noticeable feature of this stage is the presence for the first time of individuals of the oldest size-age class. It must not be thought that the individuals of this class in this community are at all comparable with those of the same class in the kauri forest. Many of them will not yet be producing seed and those that are, only small quantities. This added seed source much nearer to hand is probably partially responsible for the large increase in the number of class II seedlings. It is possible, too, that conditions in the community are more favourable for the germination and survival of small seedlings. On the whole, it can be said that natural regeneration is at its greatest in the Leptospermum-Agathis community.

There is a change from the sub- or co-dominance of Agathis by the time the ricker community develops. There is apparently a slight decline in the regenerative capacity of the sere as shown by the reduction in the number of class IV and V seedlings. Further evidence on this point is the absence of one or other of the various size-age groups on almost all of the quadrats. The ricker community gradually changes into a mature forest. In the kauri forest, there is a decrease in the number of individuals of Agathis and a steady increase in the size of those remaining. The diameter of any of the adult specimens examined was not less than sixty centimetres, and many were considerably larger than this. The first point to be noticed about this community is the small number of individuals of all age classes and more importantly the complete absence of individuals of class V in all probability showing that domination by Agathis is now on the downgrade because it is apparently not replacing itself. The largest individuals of class IV were not more than two or three centimetres in diameter and can scarcely be considered as potential replacements for some of the dominant individuals. If a community is to maintain itself as a stable climax, the dominant species must be able to reproduce themselves so that there is formed a more or less even aged stand. If the seedlings of tree species cannot meet the competition of lesser species, whether it be in the herb or shrub stratum, such trees must eventually disappear from the community. The small number of individuals of all age classes and the complete absence of class V indicates that, at least in some of the strata, Agathis is not competing successfully. Permanent or true dominance involves the ability to compete successfully in all strata of the