With the Development of statistical methods ecology is graduating to a more exact science. The days when the preparation of a species list or a vegetation map was the prime objective of an ecological study are now almost past. Closer attention is being given to the correlation of plants with their habitats, to the influence of the biotic factors on vegetation and to the microclimates in which the different communities develop. Consequently techniques by which data are accumulated are becoming more objective, with the result that different observers obtain similar records from a given ecological community, and the subjective or personal factor in making records is reduced to a minimum.

The following data were obtained by Botany students of Victoria and Auckland Universities in the course of their practical field work in Ecology during May, 1965. Data obtained from the same area by the 1963 (Victoria University) class have been included.

The scheme of work was designed to illustrate the Importance Value Index technique of Curtis and Macintosh (1951), and to study the vegetation above the timber-line on the Bruce Road, Mount Ruapehu. A preliminary examination indicated that, at any given altitude, the vegetation changed considerably with the shelter or exposure of the habitat, and was greatly influenced by cold air drainage and snow drifts of varying persistence. It was decided to study the vegetation of only one aspect, the north-west aspect in the neighbourhood of the Bruce Road. The graphs indicate that a reasonable uniformity of trend was achieved in the data obtained from each station, and the changes in the index for each species would seem to be related to its reaction to the habitat and the changing altitudes: but it should be stressed that the data as presented are applicable only to slopes facing north-west.

Methods

The class was divided into five groups of ten or more students, and a different group worked on the sub-alpine vegetation each day over a period of ten days. On the site each group was divided page 20 into parties of two or three persons. Each party recorded abundance, sociability and dominance on a scale of 1-5 for each species within two metres square quadrats at each of four predetermined altitudes. Thus, each day six or eight quadrats were recorded at each altitude. Frequency was determined as the percentage of quadrats at each altitude in which a species occurred.

The groups were instructed to ignore mosses (other than Rhacomitrium lanuginosum) and lichens; but in some cases these have been included though without specific determination. They have been grouped together in the tabulated data as ‘Sundry mosses and lichens’, since, having been included in the calculations, their omission would prevent checking of the total of the indices at each station. However no weight can be given to records of ‘mosses and lichens’, in Table I, since they were not recorded uniformly by all groups.

The practical details in the field, the checking of altitude and assistance in the identification of species were the responsibility of a member of staff or his deputy who supervised and co-ordinated the work of successive parties throughout the week.

The observations of abundance, dominance and sociability were carded and put on a computer as an experimental attempt at rapid analysis. The totals of each of these statistics for each species at each altitude were available the week after the cards had been completed, and a table for interpolating the ‘relative’ data was provided. With more efficient programming the Importance Value Index (I.V.I.) could have been obtained in the same time. Because absences were not carded, frequency and relative frequency were calculated manually from the summaries of the original data. The I.V.I. was obtained on a calculating machine by adding the figures for relative abundance, relative dominance and relative frequency for each species at each altitude. The sums of each of the contributing statistics and of the index were checked to confirm the accuracy of the calculations.

Importance Value Indices are presented both in tabular form (Table 1) and as graphs (Table 2) for each species at each different altitude. The species are arranged in the graphs according to three criteria: first, those species that are common throughout the range of altitudes sampled: secondly, according to the sequence in which they disappeared from the record as the altitude increased: and finally, in the approximate order in which they appeared in the record as the altitude increased above 4000 feet.

Discussion of Methods

Though these data were obtained as part of an undergraduate field course in ecology, I am confident that most of the work was done conscientiously, and this is confirmed by the uniformity of the page 21 data obtained by different parties and different groups from the same altitudes. It is probable that some errors of identification of plants occurred, and some of the identifications of grasses are suspect. However such errors as were made are unlikely to be of any great significance. In general they would in all probability be restricted to plants that occurred but rarely in the community and which would contribute little towards the I.V.I.

To increase the size of the quadrats, whether measured by individuals or by area, will cause more species to be included in each and so the number of species that are common to all quadrats will be more. To study a greater number of quadrats increases the number of species that occur in but a few quadrats. Thus both size and number of quadrats will affect frequency and especially relative frequency.

The effect of size can be eliminated if the quadrat size in different plant communities is related to the minimum homogeneous floristic area as determined from the species area curve. In the area under consideration the homogenous floristic area at 4.100ft. appears to be about 16 square metres; and at 4,700ft. about 12 square metres. Thus square quadrats on a side of two metres (four square metres) are between one-third and one-quarter of the homogeneous floristic area. This relationship permits the more common species to achieve 100 per cent. frequency and avoids confusing them with those less common plants which would nevertheless appear in most of the minimum homogenous floristic areas. A finer calibration of the distribution of a species in the community is thus possible.

The large number of quadrats at some altitudes has resulted in the sampling of a larger area, and has thus included more species in the data for that station, but since these contribute little volume to the vegetation and have low frequency the variation in the number of quadrats at any one site has not greatly affected the I.V.I. of the major species at that site.

Some slight allowance must be made for the assessment of alitude. Most of the plots recorded at any given altitude were placed within 12 or 15 feet of that altitude, as determined by a sensitive aneroid altimeter.

Table 1. Importance Indices of Species Recorded on North-West Aspects of Bruce Road Mount Ruapehu, Tongariro National Park.

Table 1 (continued)

Table 2. Graphical Presentation of Importance Indices for Prominent Species on North-West Slopes, Bruce Road, Mount Ruapehu.

Table 2 (continued)

Results

This method of analysis has permitted a clear-cut assessment of each of the 74 species that were encountered in some 270 quadrats studied at 18 altitudes between 4,000 and 5,600 feet on Bruce Road, Mount Ruapehu. No species were recorded at every one of the 18 altitudinal stations, and only six species (Dracophyllum recurvum, Anisotome aromatica, Senecio bidwillii, Gaultheria colensoi, Celmisia spectabilis, Rhacomitrium lanuginosum and another unidentified moss), were recorded over the whole of the altitudinal range. Three other species were recorded regularly between 4,100 and 5,600 feet, namely Wahlenbergia pygmaea, Helichrysum bellidioides and Drapetes dieffenbachii. These nine species contribute a total of 2283 I.V.I. units out of a possible 5,100 units for the whole analysis, and may be considered as the most important species in the community since they contribute jointly 44.8 per cent. of the total I.V.I. Only three other species, Notodanthonia setifolia, Gentiana bellidifolia and Coprosma pumila are recorded regularly between 4,100 and 5,500 feet. Of these the first would be about three times as important as either of the other two.

An examination of the graph indicates that a number of species seem to have an upper limit of 4,500 to 4,600 feet. Among these species may be listed Neopanax colensoi, Pimelea buxifolia, Olearia nummularifolia, Hypolaena lateriflora, Gleichenia circinata, Dracophyllum filifolium, D. subulatum, Hebe odora and H. venustula.

Within this range of altitudes (4,000-4,500 feet) the ten species that contribute most to the vegetation in the order of their importance are Gleichenia circinata, Chionochloa rubra, Dacrydium laxifolium, Dracophyllum recurvum, Celmisia spectabilis, Coprosma cheesemanii, Hebe odora, H. venustula, Rhacomitrium lanuginosum and Dracophyllum filifolium, which together contribute 944 of the possible 1,800 I.V.I. units attributed to this region.

Between 4,500 and 5,000 feet at which altitude such species as Podocarpus nivalis, Dacrydium laxifolium, Chionochloa rubra, Hebe tetragona, Myrsine nummularia and Gaultheria antipoda reach or approach their upper limit of growth, the plants that contribute most to the I.V.I. in the order of their importance are Dracophyllum recurvum, Rhacomitrium lanuginosum, Celmisia spectabilis, Anisotome aromatica, Notodanthonia setifolia, Gaultheria colensoi, Pentachondra pumila, Dacrydium laxifolium, Senecio bidwillii, Chionochloa rubra and Wahlenbergia pygmaea. These eleven species contribute together 1,000 I.V.I. units of a possible 1,800 distributed through the six stations in this region.

Above 5,000 feet the most obvious change is in the increased importance of Helichrysum bellidioides and Drapetes dieffenbachii, in an area where the vegetation is becoming very sparse and page 27 scattered. The most important species in such vegetation as does occur above 5,000 feet are Rhacomitrium lanuginosum, Dracophyllum recurvum and Gaultheria colensoi; each of which contributes more than 180 units to the total 1800 I.V.I. ratings in this area. These with Anisotome aromatica, Helichrysum bellidioides and Notodanthonia setifolia contribute 59 per cent. of the possible I.V.I. units of the area.

A few species extend from lower altitudes into this highest region and seem to find their maximum altitude at 5,200 feet. Included in these are Pentachondra pumila, Euphrasia cuneata, Ourisia vulcanica, Poa caespitosa, and Hebe tetragona. The sharpness of the exclusion of these species from the records at this altitude may be due to the fact that the next station is at 5,380 feet, an increase of 180 feet in elevation.

It is thus clear that there are two regions where changes in the plant communities can be discerned. That at 4,500-4,600 feet is not so readily obvious; but that between 4,900 and 5,000 feet was anticipated by the field parties, though it was defined by them as a change in the floristic lists, rather than as a change in physiognomy of the vegetation.

Physiognomically, the vegetation between 4,000 and 5,600 feet on the north-west aspects of the Bruce Road, Mt. Ruapehu, would appear to be one plant association in which Dracophyllum recurvum, Rhacomitrium lanuginosum, Celmisia spectabilis, Senecio bidwillii, Anisotome aromatica and Gaultheria colensoi are the most important species. This may be divided into perhaps three plant communities separated by the presence or absence of certain species. These communities may be described as follows.

The Gleichenia-Chionochloa-Dacrydium laxifolium community that exists between 4,000 and 4,600 feet would owe its physiognomic appearance to Gleichenia circinata, Chionochloa rubra, Hebe odora, H. venustula and Phyllocladus alpinus. The Dracophyllum recurvum-Rhacomitrium-Celmisia spectabilis community, differs physiognomically by the absence of Gleichenia, Hebe odora, H. venustula and Phyllocladus whose places are taken by Podocarpus nivalis, Pentachondra pumila and Hebe tetragona. This latter group is present but less prominent in the previous community. This community appears to lie between 4,500 and 5,100 feet on north-west slopes of the Bruce Road.

The third community, above 5,100 feet, would be Rhacomitrium-Dracophyllum recurvum-Gaultheria colensoi. It is notable because of the increased importance of Helichrysum bellidioides and Drapetes dieffenbachii; but in spite of the moss present the community is very open, and provides a very poor coverage of the ground.

Discussion of the Results

It will be noted in the graphs that the record of a species at its highest altitude is often graphed as a high peak out of character with the rest of the graph. This is probably because, in most cases, the working parties were taken to the highest altitude and worked their way downwards. As a new species was observed in the vegetation at the new station it was natural that each party wished to include it in their records, and so it was consciously included in the plots at that altitude. It was no longer a novelty at the next station down the mountain, and so fell into its natural place as a species in the random placing of quadrats. These peaks therefore show a bias when the species is first observed as a new species in the record — a human touch — but the bias is not continued in subsequent stations. When this interpretation was put to the students it was confirmed and admitted as a course that had been adopted by several parties.

One aspect that must be recognised in interpreting these results is that the Importance Index is just what it is called. It is in fact the sum of the relative abundance, the relative dominance and the relative frequency attributable to each species for which data were obtained at each station. These statistics represent the contribution that a species makes to the community in respect of (a) the number of plants within the quadrats (abundance), (b) its influence on the other species through its shading, competition or aggressiveness (dominance), and (c) its contribution to the community through its distribution (frequency). There is no estimate or measurement of the volume or productivity of a species in the quadrat, and no absolute assessment of the space occupied by the plant. Consequently the Index is not a measure of production. In fact, the plots at the highest altitudes were very sparsely covered by vegetation. Thus, in the absence of any data concerning the amount of bare ground, the Index is purely a measure of the contribution of a species to that vegetation which is present, regardless of whether the ground is completely covered (as it was at the lowest altitudes) or very sparsely covered, as at the highest.

In the three communities ten or eleven species contributed more than half the I.V.I. of the important vegetation. Dominance was shared by several species at each altitude, and, with the possible exception of Gleichenia circinata, no one species exerted complete dominance over any appreciable area. Importance, as indicated by I.V.I., was shared by several species, and, though these may change from station to station, they are not necessarily the physiognomic species of the community.

Summary

Some 270 quadrats were distributed over eighteen stations at different altitudes between 4,000 and 5,600 feet on north-western slopes adjoining the Bruce Road, Mt. Ruapehu, Tongariro National Park, and abundance, dominance and frequency ratings obtained for each species encountered.

Importance Value Indices were calculated for each species at each station. These have been graphed against increasing altitude.

Some twelve species were found to have their upper limit of growth about 4,500 feet, six others at 5,000 feet, while only seven species were recorded throughout the range from 4,000 feet to 5,600 feet.

On the basis of these changes in floristic composition, and changing importance of species as altitude increased, three subcommunities have been proposed within the one plant association that was studied.

Acknowledgements

I wish to acknowledge the contribution of the students of the Victoria University of Wellington Botany II and III classes of 1963 and 1965 who gathered the basic data that made this analysis possible. The 1965 class was joined for a week by the contemporary classes from Auckland University who contributed their share of data. The reliability of the data may be reasonably attributed to the leaders of the student groups who conscientiously endeavoured to assure that each party in their group followed instructions, and to the staff member, Mr. F. B. Sampson, and his deputy, Mr. John Braggins (a student), and to Mr. E. A. Alcock, Botany Department, Auckland University, who shared the supervision and co-ordination of the field work, and helped the various groups with the identification of plants.

Assistance has also been received from individual students, who wish to remain anonymous, in determining frequencies, carding the data for the computer, and drawing the initial graphs of Indices. Most of the 1965 Wellington class assisted at one time or another in working out various of the indices. Finally, in both 1963 and 1965 most of the staff of the V.U.W. Botany Department spent at least a week on the ecology field work, and if they did not actually take part in this project, they were in charge of other projects that kept those students not on Bruce Road busy.