Introduction

In a Country of such varied relief as New Zealand the distribution of plants is frequently controlled by the ecological factors associated with altitude. There is adequate geological evidence that this relief has persisted throughout the Quarternary. The existence of such relief introduces some doubt into the climatic interpretation of macrofossil floras, as the transport of plant remains from higher altitudes can result in the preservation at one place of fossil assemblages representing more than one climatic zone.

Plant remains are all potentially easily transportable by water to the lake or estuarine sediments which have preserved them in the past, and are preserving them today, but the possible distances of transport are less well known.

A second point of doubt concerns the proportion of vegetation preserved as fossils. The ultimate preservation of plant remains depends much on their relative resistance to damage, either by abrasion in transport or by microbiological decomposition during sedimentation. But little is known of the potential fossilisation of plants of the New Zealand flora.

In this paper, records of recent plant remains in mountain areas are compared with the distribution of source plants in such areas. Such comparisons can outline a direction to be followed in resolving the two different questions which must be answered before valid paleoclimatic interpretation can be made from fossil floras:

— Do the plant remains in a basin represent only the vegetation on the shores of that basin, or do they represent, as well, plants growing in a higher and colder climatic zone?

— What proportion of the total flora around the basin of deposition is likely to be preserved?

Description of Main Sampling Area

Sampling was done on and about Lake Pounui, which is at about 30 m altitude in the eastern foothills of the Rimutaka Range (N.Z.M.S. Sheet N.165). The lake is 1.7 km long and 0.8 km wide and not more than 10 m deep, and fed by streams entering it through aggraded page 14

FIG. 1: Sketch vegetation map, and sample localities, L. Pounui. Scale approx. 20 chs to one inch.

valleys. The immediately surrounding greywacke hills rise to about 250 m and supply pebbles to the lake shore beneath bluffs, and silts to the rest of the lake. The lake itself is in the Lower Warm Temperate zone of Zotov (1938). The Rimutaka mountains, reaching 1000 m are 6 km distinct; the nearest country over 800 m, within the Cold Temperate Zone (Zotov loc. cit.) is over 2 km distant.

Prevailing winds are northerly to westerly and are frequently strong. Southerly and easterly winds are less frequent and the lake is relatively sheltered from them by the surrounding hills. There is no record of rainfall at the lake but it is 1575 mm annually at Wairongomai, 8 km away, lying to the east of the axial range, as does Lake Pounui.

Methods

Vegetation: An assessment, based on an 8 point abundance scale, was made of the native vegetation around the lake. A map (Fig 1.) was compiled: the categories of vegetation being delimited by change in physiognomic species.

Sediment sampling: On the lake margin and in the shallow water of the swamps, samples were scooped from the bottom with a handheld canister of 400 ml capacity.

In open water samples were taken with a simple limnological dredge consisting of a canvas cone, with a steel reinforced mouth, and two of its three attached lines weighted. This apparatus, dragged from a dinghy, took samples each of about 10 m length of lake bottom and generally yielded about 100 cc of sediment. The location of sample stations is shown on Fig. 1.

Vegetation

The greater part of the lake shore is surrounded by extensive remnants of indigenous forest, by swamp vegetation in the aggraded valleys and by fire-induced scrubland dominated by indigenous species. The vegetation on the south-eastern shore has been converted to grassland farming but carries trees remnant from coastal forest. The vegetation (Fig. 1) consists of the following:

The first part of Table I shows the specific composition of these communities which are, apart from the coastal forest and swamp vegetation, a vegetation analogous to that described in detail by Druce (1958) in the western foothills of the Rimutaka Range.

Results of Sediment Sampling

There are two groups of samples: Firstly those from the swamps marginal to the lake, where plant remains deposited by direct fall and wind transport would be expected to dominate, with water transport being the case only in flood periods. The second group comprises those samples taken in open water, where there is a possibility of water transport before ultimate settling and preservation.

From dry ground adjacent have come:

Metrosideros robusta, Nothofagus solandri var. solandri, page 16 N. truncata and their hybrid, and Pseudopanax arboreum; from more than 2 km away a solitary leaf of Nothofagus menziesii, either blown from higher country, or, more likely, carried from a riparian tree by the free-flowing portion of the stream above the swamp.

A total of 57 species was recorded in the communities around and on the swamp but only 9 are represented as potential fossils. Eight of these have coriaceous leaves, slow to decompose on the forest floor; only Pseudopanax arboreum has a softer, rapidly decomposing leaf.

Station 1 was from shallow water equidistant from a rocky bank and from a marginal swamp. The matrix was entirely organic, of leaves and twigs, and yielding large quantities of the shrubs Leptospermum ericoides, L. scoparium and Cyathodes fasciculata, all abundant near the sample station. This matrix contained occasional leaves of other plants: Nothofagus solandri var. solandri, A. truncata and their hybrid, all absent from shrubland on the bluffs above the sample station, and whose leaves must have been blown and washed at least 60 m from source plants.

Deeper water samples (Stations 2-7) were from a matrix of grey clayey silt. Plant remains washed from this silt represent the more common species of the vegetation of the surrounding hills: Coprosma rhamnoides, Leptospermum (2 spp.) Cyathodes fasciculata, Nothofagus solandri var. solandri, N. truncata and their hybrid, and Pseudopanax arboreum. Only Typha muelleri and possibly the unidentified Cyperaceous nut represent the swamp or semi-aquatic

Underlining across the table indicates species found in swamp or lake sediments. Botanical names for the pteridophytes, gymnosperms and dicotyledons are those of Allan (1961) except where otherwise indicated.

vegetation on the lake shore. The leaves are all coriaceous. In all, the lake sediments yielded 8 of the possible total of 32 species available.

Notably absent from the lake sediments were remains of woody plants of the swamp vegetation. It can be concluded that fragments of such plants are effectively trapped in the dense foliage of the swamp floor plants. This filtering effect of the swamp may also account for the absence of any plant remains from higher altitudes.

Support for this contention comes from two localised records of plant remains from a free-flowing river and a lake fed by such rivers:

Conclusion

The results of sampling of potential basins of fossil deposition have shown that plant remains in a densely vegetated swamp are almost exclusively of plants growing in or close to that swamp. The exception, Nothofagus menziesii, of distant origin, both horizontally and vertically, could have been trapped by the swamp vegetation after stream transport over a distance exceeding 2 km.

In the lake sediments the plant remains represented only the dominant members of the vegetation surrounding the lake, and no remains were found that had been transported any distance to the lake. The absence of remains allochthonous to the lakeside vegetation supports the idea that the swamps in all the feeding streams act as filters, catching plant remains brought from higher country by these streams.

In the case of plant remains carried by unimpeded rivers, the transport of such remains from higher zones of vegetation is evidence that potential macrofossils can be carried considerable distances, and remain recognisable. In almost all cases the remains were of coriaceous leaves, a fact in agreement with general principles of fossilisation.

The nature of the sediments containing plant remains can also contribute to interpretation of fossil floras.

In the case of Lake Pounui, the sediments in the swamp and in the deep waters of the lake were all of fine texture, being those page 19 carried through the swamps by slow moving streams at or near base level (Cotton 1945). In the freeflowing Mead River and on the shores of Lake Taupo, the sediments are coarse, up to boulder size; this material having been carried by rivers further above base level and unimpeded by swamps.

Such relationships must be considered in the interpretation of macrofossil floras. Floras from deep series of uninterrupted silts and clays with highly organic strata are likely to have been associated with relatively slow flowing rivers and the presence of swamps, and are less likely to contain elements from higher zones of vegetation. Plant macrofossil beds intercalated with conglomerates should indicate the existence of more rapidly flowing rivers, and imply the possibility of inclusion in the fossil flora of plant remains from higher zones of vegetation than that around the basin of deposition.

Acknowledgements

I am grateful to those who helped in this study, both in the field and in the preparation of the manuscript: Mr. A. Rohde, formerly of ‘Wharekauhau’, Western Lake Road; Professor H. D. Gordon; Dr. G. Dimbleby; Dr. N. T. Moar. Especial thanks are due to my wife for her invaluable field aid, and to Mr. A. Reid for draughting.